The Political Economy of the Abe

Government and Abenomics Reforms

This volume seeks to explain the political economy of the Abe government and
the so-called “Abenomics” economic policies. The Abe government represents
a major turning point in postwar Japanese political economy. In 2019, Abe
became the longest serving Prime Minister in Japanese history. Abe’s
government stood out not only for its longevity, but also for its policies. Abe
came to power promising to reinvigorate Japan’s economy under the banner of
Abenomics. He pursed a host of structural reforms and industrial promotions
to increase Japan’s potential growth rate. Abe also achieved important
legislative victories in security policy. However, the government also faced
significant controversies. The book will hold appeal to scholars and students
specializing in the study of Japanese politics, comparative political economy,
the politics of contemporary advanced democracies, macroeconomic policy,
labor market reforms, corporate governance, gender equality, agricultural
reforms, energy and climate change, and East Asian security.
takeo hoshi is Professor of Economics at the University of Tokyo. His
research interest includes corporate finance, banking, monetary policy and the
Japanese economy. He received the 2015 Japanese Bankers Academic
Research Promotion Foundation Award, 2006 Enjoji Jiro Memorial Prize of
Nihon Keizai Shimbun, and the 2005 Japan Economic Association-Nakahara
Prize. His book Corporate Financing and Governance in Japan: The Road to
the Future (2001) co-authored with Anil Kashyap received the Nikkei Award
for the Best Economics Books. He co-authored The Japanese Economy (2020)
with Takatoshi Ito. Hoshi received his PhD in economics from the
Massachusetts Institute of Technology.

phillip y. lipscy is Associate Professor of Political Science and Director

of the Centre for the Study of Global Japan at the Munk School of Global
Affairs and Public Policy, University of Toronto. He researches
international and comparative political economy with a focus on Japan.
He is author of Renegotiating the World Order: Institutional Change in
International Relations (2017) and co-editor of Japan under the DPJ: The
Politics of Transition and Governance (2013). Lipscy obtained his PhD in
political science at Harvard University, an MA in international policy
studies and BA in economics and political science at Stanford University.
The Political Economy of
the Abe Government and
Abenomics Reforms
Edited by
takeo hoshi
University of Tokyo
phillip y. lipscy
University of Toronto
University Printing House, Cambridge CB2 8BS, United Kingdom
One Liberty Plaza, 20th Floor, New York, NY 10006, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
314–321, 3rd Floor, Plot 3, Splendor Forum, Jasola District Centre,
New Delhi – 110025, India

79 Anson Road, #06–04/06, Singapore 079906

Cambridge University Press is part of the University of Cambridge.

It furthers the University’s mission by disseminating knowledge in the pursuit of
education, learning, and research at the highest international levels of excellence.

www.cambridge.org
Information on this title: www.cambridge.org/9781108843959
DOI: 10.1017/9781108921145
© Cambridge University Press 2021
This publication is in copyright. Subject to statutory exception
and to the provisions of relevant collective licensing agreements,
no reproduction of any part may take place without the written
permission of Cambridge University Press.
First published 2021
A catalogue record for this publication is available from the British Library.
ISBN 978-1-108-84395-9 Hardback
Cambridge University Press has no responsibility for the persistence or accuracy of
URLs for external or third-party internet websites referred to in this publication
and does not guarantee that any content on such websites is, or will remain,
accurate or appropriate.
In memory of Masahiko Aoki
Contents

List of Figures page x

List of Tables xiv
List of Contributors xvi
Acknowledgments xxv
List of Abbreviations xxvii

Part I Introduction 1
1 The Political Economy of the Abe Government 3
t a k e o h o s h i a n d ph i l l i p y. l i p s c y

Part II Political Context 41

2 Expansion of the Japanese Prime Minister’s Power and
Transformation of Japanese Politics 43
ha r u k at a t a k e n a k a
3 Why Does the Abe Government’s Approval Rating Always
Recover? 68
m as a ru k o h n o
4 The LDP under Abe 87
yu k i o m a e d a a n d s t e v e n r . r e e d
5 The Third Arrow of Abenomics: Est. in 2013 – or 2007? 109
t a k at o s h i i t o

Part III Macroeconomic Policy 137

6 Abenomics, Monetary Policy, and Consumption 139
j o s h u a k. h a u s m a n , t ak a s h i u n a y am a , a n d
j o h a nn e s f . w i e la n d

vii
viii Contents

7 The Great Disconnect: The Decoupling of Wage

and Price Inflation in Japan 170
t a k e o h o s h i an d a n i l k k a s h y ap
8 Abenomics, the Exchange Rate, and Markup
Dynamics in Japanese Industries 200
k yo j i fu k a o an d sh u i c h i r o n i s h i o ka
9 The Crisis That Wasn’t: How Japan Has Avoided
a Bond Market Panic 239
m a rk b a m ba a n d da v i d e . w e i ns t e i n

Part IV Third Arrow of Abenomics 269

10 Abe’s Slight Left Turn: How a Labor Shortage
Transformed Politics and Policy 271
steven k. vogel
11 Abe’s Womenomics Policy: Did It Reduce
the Gender Gap in Management? 310
n ob u k o na g a s e
12 Corporate Governance Reforms under Abenomics:
The Economic Consequences of Two Codes 357
h id ea k i mi ya j ima an d t ak u ji s a ito
13 Abenomics and Japan’s Entrepreneurship and Innovation:
Is the Third Arrow Pointed in the Right Direction
for Global Competition in the Silicon Valley Era? 394
k en j i e . k u s h i d a
14 Japanese Agricultural Reform under Abenomics 421
p a t ri c i a l . m a c l a c h l a n a nd k a y s h i m i z u
15 The Politics of Energy and Climate Change in Japan
under the Abe Government 445
t r e vo r i n c e r t i a n d p h i l l i p y . li p s c y

Part V Foreign Policy 477

16 Japan’s Defense Reforms under Abe: Assessing
Institutional and Policy Change 479
a da m p . l i f f
Contents ix

17 The Enduring Challenges of History Issues 511

m ar y m . m c c ar t h y

Index 534
Figures

1.1 Election results in the Japanese lower house, 1986–2017 page 6

1.2 Passage rate of Diet legislation (%), 1980–2019 7
2.1 The chain of delegation in the parliamentary system 46
2.2 The “1955 system” 50
2.3 After the 1994 reform 52
2.4 After the 2001 reform 54
2.5 Politicians with ministerial experience 56
2.6 Public officials in the Cabinet Secretariat 57
2.7 Expansion of special policy units in the Cabinet
Secretariat (under the director general of cabinet affairs
and the assistant deputy chief cabinet secretary) 57
2.8 Officials in the Cabinet Office 58
2.9 Councils set up under the cabinet 59
2.10 Bills submitted by the Cabinet Secretariat 59
2.11 After the 2014 reform 60
3.1 Public confidence in the Abe government 71
3.2 Confidence in the government and economic indicators 74
3.3 Distribution of government evaluation in comparison 78
3.4 (A) Views on new national security law and priming
effect; (B) Priming effect and views on constitutionality
among pro-legislation respondents; (C) Priming effect
and views on constitutionality among anti-legislation
respondents 79
4.1 Number of votes cast for the LDP and the DPJ,
2003–2014 89
4.2 Cabinet approval from December 2012 to October 2019 97
4.3 Approval and reason for approval from December 2012
to October 2019 99
4.4 Disapproval and reason for disapproval from
December 2012 to October 2019 100
5.1 Agricultural, forestry and fishery product: exports 121

x
List of Figures xi

5.2 Visitor arrivals, international 125

5.3 International passengers, Narita and Haneda airports 127
5.4 Changes in the GPIF portfolio allocation 130
6.1 Abenomics’ financial market effects 143
6.2 Japanese actual and expected inflation 147
6.3 (A) Annualized quarter-on-quarter GDP growth since
2007. (B) Comparison of annualized contributions to
GDP by component during Abenomics (2012–2018), the
lost decade excluding the Great Recession (1994–2007),
and the Great Recession (2007–2012) 150
6.4 Actual and forecast output and consumption 152
6.5 Checking FIES data consistency with the national accounts 155
6.6 FIES results 156
6.7 FIES results 157
7.1 Japanese price inflation, 1971–2018 176
7.2 Japanese wage inflation, 1971–2018 177
7.3 Wages and bonuses in Japan, 1993–2018 181
7.4 Composition of the Japanese workforce (establishments
with more than thirty employees) 182
7.5 Proportion of firms with excess full-time workers
(percentage) 185
7.6 Japanese price and wage level, 1981–2018 (2012=100) 188
7.7 Impulse response of core-core inflation to Cholesky One
Standard Deviation Innovations, 1981:Q1–1997:Q4 190
7.8 Impulse response of core-core inflation to Cholesky One
Standard Deviation Innovations, 1998:Q1–2018:Q4 192
7.9 Correlation between unobserved trends for core-core
CPI inflation rate and wage inflation rate for all workers 194
8.1 GDP gap, 1990–2015 201
8.2 Japan’s exchange rate (monthly) 202
8.3 Japan’s market and PPP rates 204
8.4 Markups in manufacturing 205
8.5 Markup dynamics in Japan 215
8.6 Markups in selected industries 219
8.7 Markups before/after Abenomics 220
8.8 Log prices in manufacturing 228
9.1 Japan’s interest rate gap 244
9.2 Per capita expenditure versus GDP 246
9.3 Sources of government revenue 248
xii List of Figures

9.4 Government revenues and expenditures 249

9.5 Japan’s debt levels 250
9.6 Log money supply 252
9.7 Japan 2020–2100: government expenses as a share of
GDP by population group (Case 1) 258
9.8 Japan 2020–2100: government expenses as a share of
GDP by population group (Case 2) 259
9.9 Japan 2020–2100: government expenses as a share of
GDP by population group (Case 3) 259
10.1 Japan’s population: actual and projected, 2000–2065 272
10.2 The unemployment rate in Japan, 1995–2018 273
10.3 Job offers to applicants ratio in Japan, 1995–2019 273
10.4 The rise of economic inequality in Japan, 1995–2015
(Gini coefficient, before taxes and transfers) 279
10.5 The increase in non-regular workers in Japan,
1988–2019 (percentage) 282
10.6 Labor productivity in Japan and the United States,
2000–2018 (GDP per hour worked, US dollars) 287
11.1 Female average wage as percent of male average wage,
by educational attainment, age 23–59 when not
otherwise specified 322
11.2 Wage distribution of males for permanent contract
regular workers 324
11.3 Wage distribution of females for permanent contract
standard workers 324
11.4 Wage distribution of males for all nonstandard workers 325
11.5 Wage distribution of females for all nonstandard workers 325
11.6 Ratio of managers in narrow definition among
employees aged 22–59 by gender and firm size 327
11.7 Gender ratio of managers among all workers and among
permanent contract workers including non-seishain 332
11.8 (A) Ratio of those in different managerial positions by
age group, birth cohort of 1965–1969, university
graduates at large firms; (B) Ratio of those in different
managerial positions by age group, birth cohort of
1980–1984, university graduates at large firms 333
11.9 Percentage of workers in any managerial position at
large firms by gender 335
List of Figures xiii

11.10 Ratio of section chief among permanent contract

employees by gender, comparison of those working at
firms with 100–299 employees to those with 300–499 336
11.11 Ratio of section heads to other subordinated managers
by gender, comparison of those working at firms with
100–299 employees to those with 300–499 337
12.1 Transition of ownership structure of Japanese firms 360
12.2 Transition of board structure of Japanese firms 362
15.1 Japanese energy mix before and after Fukushima
(2006–2018) 450
15.2 Cross-national renewable energy share excluding
hydropower, 1990–2018 451
15.3 Public opinion on nuclear power in Japan: “What do you
think should be done with the current number of nuclear
plants in Japan?” 457
15.4 Japanese electricity prices, 1970–2017 458
15.5 CO2 emissions intensity (Mt CO2/2010 USD billions),
Japan vs. European G7 countries, 1960–2015 463
15.6 Percentile rank of Japan in the climate policy component
of the Germanwatch Climate Change Performance Index 466
Tables

3.1 Evaluations of the Abe government and the national

security law page 76
3.2 Different sequence of questions 77
5.1 Prime ministers since 2001 112
5.2 Japan’s FTAs (signed and in effect), as of January 2020 116
5.3 GPIF portfolio allocation benchmark and tolerance band 131
6.1 Announcement effects 145
6.2 Macro summary statistics 149
7.1 ARMA models for price and wage inflation 178
7.2 ARMAX models for wage inflation with unemployment
rate (UNEMP) 183
7.3 ARMAX models for wage inflation with unemployment
rate: full-time and part-time workers 186
7.4 VAR system with core-core inflation, wage inflation, and
bonus inflation, 1981:Q1–1997:Q4 189
7.5 VAR system with core-core inflation, wage inflation, and
bonus inflation, 1998:Q1–2018:Q4 193
8.1 Summary statistics for output prices and marginal costs,
1994–2006 212
8.2 Summary statistics for markup growth, 1994–2006 216
8.3 Decomposition analysis for markup growth in
manufacturing, 2006–2015 221
8.4 Decomposition analysis for markup growth in services,
2006–2015 225
8.5 Import competition and markups (manufacturing) 230
8.6 Import competition and markups (services) 232
8.7 Prices and the exchange rate 233
9.1 Sustainable tax rates based on privately held net debt 255
9.2 Sustainable tax rates based on BOJ-inclusive net debt 257
10.1 Types of labor adjustment in Japan, 1993–2018
(percentage) 281

xiv
List of Tables xv

10.2 The foreign worker program by industry 295

11.1 Ratio of managers in broader definition (denominator,
all workers) 329
11.2 Ratio of managers in broader definition by gender
(denominator permanent full-time workers) 331
11.3 Probit analysis of being a manager, marginal effects 341
11.4 Estimate of Abe’s policy on female promotion by firm size 343
12.1 Major events related to corporate governance reforms
under Abenomics 364
12.2 Distribution of institutional investor ownership 369
12.3 Characteristics of boards of directors 372
12.4 The effect of outside directors on corporate performance 376
12.5 Firm characteristics classified by the level of relational
shareholdings 379
12.6 Average firm performance and behavior 383
12.7 The effect of institutional ownership on firm
performance and behavior 386
13.1 Abenomics third arrow strategies over time 397
13.2 Silicon Valley ecosystem characteristics compared to
Japan’s impediments in the 1990s, changes by 2016 398
13.3 Selected venture capital–related KPIs 401
13.4 Select KPIs for human capital 406
13.5 Select KPIs for university–industry–government ties 409
13.6 Select KPIs for industry–university–government
collaboration 411
Contributors

Mark Bamba is a PhD candidate in economics at Princeton University.

His research interests are in international trade, economic geog-
raphy, and the Japanese economy. He received his BS and MS in
applied physics from Columbia University.
Kyoji Fukao is president of the Institute of Developing Economies,
Japan External Trade organization (IDE-JETRO), and Specially
Appointed Professor at the Institute of Economic Research,
Hitotsubashi University. He has published widely on productivity,
international economics, economic history, and related topics in
journals such as the Journal of Political Economy, the Quarterly
Journal of Economics, and the Review of Income and Wealth. In
addition, he is the editor of many books on productivity and eco-
nomic history, including The World Economy: Growth or
Stagnation (Cambridge University Press, 2016) with Dale
W. Jorgenson and Marcel Timmer.
Joshua Hausman is an associate professor of public policy and eco-
nomics at the University of Michigan and a faculty research fellow at
the National Bureau of Economic Research. His research interests
are in economic history and macroeconomics with a focus on the US
economy in the 1930s and the Japanese economy today. Josh holds
a BA in economics from Swarthmore College and a PhD in econom-
ics from the University of California, Berkeley. From 2005 to 2007
he worked as a research assistant at the Federal Reserve Board, and
in 2010 he worked as a staff economist at the Council of Economic
Advisers. Josh won the 2013 Allan Nevins Prize for the best disser-
tation on US or Canadian economic history.
Takeo Hoshi is Professor of Economics at the University of Tokyo. His
research interests include corporate finance, banking, monetary pol-
icy, and Japanese economy. He received the 2015 Japanese Bankers

xvi
List of Contributors xvii

Academic Research Promotion Foundation Award, the 2006 Enjoji

Jiro Memorial Prize of the Nihon Keizai Shimbun, and the 2005
Japan Economic Association-Nakahara Prize. His book Corporate
Financing and Governance in Japan: The Road to the Future (MIT
Press, 2001), co-authored with Anil Kashyap, received the Nikkei
Award for Best Economics Book. He co-authored The Japanese
Economy (MIT Press, 2020) with Takatoshi Ito. Hoshi received his
PhD in economics from the Massachusetts Institute of Technology.
Trevor Incerti is a PhD candidate in political science at Yale University.
His research focuses primarily on the use of political office for private
or corporate gain, with an emphasis on regulation, corruption, and
energy policy. Prior to Yale, Trevor worked as a data scientist at
TrueCar, Inc., and as an economic consultant at Compass Lexecon.
He holds BAs in political economy and Asian studies from UC
Berkeley and an MA in political science from Yale University.
Takatoshi Ito is Professor of International and Public Affairs at the
School of International and Public Affairs, Columbia University. He
is an expert on international finance, macroeconomics, and the
Japanese economy who served from 2006 to 2008 as a member of
the Prime Minister’s Council on Economic and Fiscal Policy. He also
held senior positions in the Japanese Ministry of Finance and at the
International Monetary Fund. Ito served as dean of the University of
Tokyo’s Graduate School of Public Policy for the past two years and
as professor at Japan’s National Graduate Institute for Policy
Studies. He has served as a visiting professor at both Columbia and
Harvard and taught at other institutions. He earned his PhD in
economics at Harvard University.

Anil K Kashyap is the Stevens Distinguished Service Professor of

Economics and Finance at the University of Chicago Booth School
of Business. His research focuses on banking, business cycles, cor-
porate finance, price setting, and monetary policy. His research has
won him numerous awards, including a Sloan Research Fellowship,
the Nikkei Prize for Excellent Books in Economic Sciences, and
a Senior Houblon-Norman Fellowship from the Bank of England.
Since 2016 he has been an external member of the Bank of England’s
Financial Policy Committee. In 2017 he was awarded the “Order of
the Rising Sun, Gold Rays with Neck Ribbon” by the government of
xviii List of Contributors

Japan. His bachelor’s degree is from the University of California at

Davis and his PhD is from the Massachusetts Institute of
Technology.
Masaru Kohno is Professor of Political Science at the School of Political
Science and Economics at Waseda University in Tokyo, Japan.
Before joining Waseda, Professor Kohno taught at the University of
British Columbia (1994–1998) and at Aoyama Gakuin University
(1998–2003), and he was a national fellow at the Hoover Institution
at Stanford University (1996–1997). He has published a number of
books and articles on Japanese politics and Japan’s foreign relations,
including Japan’s Postwar Party Politics (Princeton University Press,
1997), Seido (Institutions) (University of Tokyo Press, 2002), and
Seiji o Kagaku Suru Koto wa Kano-ka (Is It Possible to Study Politics
Scientifically?) (Chuo Koron Shin-sha, 2018). Professor Kohno is
currently engaged in a variety of research activities, including survey
experiments in aid-receiving countries, inquiries into citizens’ atti-
tude toward hate-speech regulations, and text-analysis of popular
song lyrics regarding the concept of freedom in North America.

Kenji E. Kushida is the Japan Program Research Scholar at the Walter

H. Shorenstein Asia-Pacific Research Center at Stanford University.
He is an international research fellow at the Canon Institute for
Global Studies. Kushida’s research and publications are focused on
the following streams : how politics and regulations shape the devel-
opment and diffusion of Information Technology such as AI; the
institutional underpinnings of the Silicon Valley ecosystem; Japan’s
transforming political economy; Japan’s startup ecosystem; the role
of foreign multinational firms in Japan; and Japan’s Fukushima
nuclear disaster. He spearheaded the Silicon Valley–New Japan pro-
ject that brought together large Japanese firms and the Silicon Valley
ecosystem. Kushida holds a PhD in political science from the
University of California, Berkeley. He received his MA in East
Asian Studies and BAs in economics and East Asian Studies with
honors, all from Stanford University.

Adam P. Liff is Associate Professor of East Asian International

Relations at Indiana University’s Hamilton Lugar School of Global
and International Studies, where he also serves as founding director
of its 21st-Century Japan Politics & Society Initiative. His research
List of Contributors xix

focuses on international security affairs and the Asia-Pacific – espe-

cially Japanese and Chinese security policy; USA–Asia-Pacific strat-
egy; the USA–Japan alliance; and the rise of China. Aside from his
duties at Indiana University, Dr. Liff is a Non-resident Senior Fellow
at the Brookings Institution’s Center for East Asia Policy Studies. He
holds a PhD and an MA in politics from Princeton University, as well
as a BA from Stanford University.

Phillip Y. Lipscy is Associate Professor of Political Science at the

University of Toronto. He is also Chair in Japanese Politics and
Global Affairs and director of the Centre for the Study of Global
Japan at the Munk School of Global Affairs & Public Policy. His
research addresses substantive topics such as international cooper-
ation, international organizations, the politics of energy and climate
change, international relations of East Asia, and the politics of
financial crises. He has also published extensively on Japanese polit-
ics and foreign policy. Lipscy’s book Renegotiating the World
Order: Institutional Change in International Relations (Cambridge
University Press, 2017) examines how countries seek greater inter-
national influence by reforming or creating international organiza-
tions. His is also co-editor of Japan Under the DPJ: The Politics of
Transition and Governance (Brookings Institution, 2013).
Patricia L. Maclachlan is Professor of Government and the Mitsubishi
Heavy Industries Professor of Japanese Studies at the University of
Texas at Austin. A student of interest-group politics and political-
economic reform in Japan, her publications include Consumer
Politics in Postwar Japan: The Institutional Boundaries of Citizen
Activism (Columbia University Press, 2002) and The People’s Post
Office: The History and Politics of the Japanese Postal System,
1871–2010 (Harvard University East Asia Center, 2011). She is the
author, with Kay Shimizu, of a forthcoming book from Cornell
University Press on the reform of Japanese agricultural cooperatives.

Yukio Maeda is Professor of Political Science at the Inter-Faculty

Initiative in Information Studies, University of Tokyo. His research
interests include public opinion and voting behavior; approval rat-
ings for prime ministers; methodology for public opinion polling and
social survey; and data preservation and dissemination in social
sciences. He advises the Association for Promoting Fair Elections
xx List of Contributors

(aka Meisuikyo) for its post-election survey and the polling section of
Kyodo News for its election forecast. He was a member of the Social
Science Japan Data Archive from 2006 to 2017. He also served as
a Japanese local committee member for many international academic
surveys, including Comparative Study of Electoral Systems (CSES),
Asian Barometer, World Value Survey (WVS), and East Asian Social
Survey (EASS).

Mary M. McCarthy is an associate professor of politics and inter-

national relations at Drake University in Des Moines, Iowa. She
specializes in Japan’s domestic and foreign policies and is editor of
the Routledge Handbook of Japanese Foreign Policy (2018). Her
current research and most recent publications examine the historical
legacies of the Asia-Pacific War on Japan’s foreign relations, with
a particular focus on the comfort women issue. Dr McCarthy is
a Mansfield Foundation US–Japan Network for the Future Scholar
and was a 2014 Japan Studies Resident Fellow at the East-West
Center in Washington, DC. She received her BA in East Asian studies
and her PhD in political science from Columbia University.

Hideaki Miyajima is executive vice president for financial affairs at

Waseda University; Professor of Japanese Economy, Graduate
School of Commerce, Waseda University; and advisor, Waseda
Institute for Advanced Study (WIAS). He finished the PhD course-
work in economics at the University of Tokyo and received his PhD
at Waseda University. He first worked as a research associate at the
University of Tokyo Institute of Social Sciences and then moved to
Waseda University. He has been appointed to several positions:
Faculty Fellow at the Research Institute of Economy, Trade &
Industry; Special Research Fellow at the Policy Research Institute
(Ministry of Finance); Research Fellow at EHESS (École des hautes
études en sciences sociales, Paris); and adjunct professor at Chung-
Ang University (Seoul). He has published numerous papers in the
Journal of the Japanese and International Economies, the Review of
Financial Studies, the Journal of Banking and Finance, and
Corporate Governance. He also edited several books, including
Corporate Governance in Japan (Oxford University Press, 2007),
Changes and Continuity in Japan (Routledge Curzon Press, 2002),
and Policies for Competitiveness (Oxford University Press, 1999).
List of Contributors xxi

Nobuko Nagase is a professor of labor economics and social policy at

Ochanomizu University in Tokyo. Her research interests include
work and family issues in the Asian perspective, comparing Japan
with other East Asian and Western economies. She has pursued her
interests in the wage structure and work choice in Japan, given labor
market regulations, social security, and other social institutions as
child care, elderly long-term care, pensions, tax, and other institu-
tions on work. Her interests also extend to marital behavior and
childbirth timing. She obtained her PhD in 1995, one of the first
females granted such a degree in economics by the University of
Tokyo. Nagase has served at various Japanese governmental
appointments. Since 2012, she has been a member of the Science
Council of Japan. Results of her research have been published in
various scholarly peer-reviewed journals in Japan and overseas.
Shuichiro Nishioka is an associate professor at West Virginia
University’s Department of Economics. His research covers broad
issues in international trade, economic development, and East Asian
economies. His current research focuses on how competition, glo-
balization, and policy changes have created diverse economic, social,
and political outcomes in the United States, China, and Japan.
Nishioka received his PhD in economics from the University of
Colorado at Boulder.

Steven R. Reed is Emeritus Professor of Political Science at Chuo

University. He is a co-editor of and a contributor to the Japan
Decides series published by Palgrave. His most recent publication is
Political Corruption and Scandals in Japan, with Matthew
M. Carlson (Cornell University Press, 2018). He is currently working
on the development of patronage politics and religious groups and
politics in Japan.

Takuji Saito is Associate Professor of Finance at the Graduate School of

Business Administration, Keio University. His research interests
include corporate finance, corporate governance, board of directors,
family firms, and CEO succession. He holds a BA, an MA, and a PhD
in economics from Hitotsubashi University. Prior to his current
position, he was research fellow (postdoctoral) (2004–2006) of the
Japan Society for the Promotion of Science and lecturer (2007–2008)
xxii List of Contributors

and associate professor (2009–2011) at the Faculty of Economics,

Kyoto Sangyo University.

Kay Shimizu is a research assistant professor in the Department of

Political Science at the University of Pittsburgh. Her research
addresses institutional design and their effects on economic govern-
ance with a special interest in central local relations, small busi-
nesses, and property rights. Her publications include Political
Change in Japan: Electoral Behavior, Party Realignment, and the
Koizumi Reforms (Brookings Institution, 2009; co-edited with
Steven R. Reed and Kenneth McElwain) as well as articles in Socio-
Economic Review, the Journal of East Asian Studies, Current
History, and the Social Science Japan Journal. She is the author,
with Patricia L. Maclachlan, of a forthcoming book on agricultural
cooperative reform from Cornell University Press. Shimizu received
her PhD in political science from Stanford University.
Harukata Takenaka is Professor of Political Science at the National
Graduate Institute for Policy Studies in Japan. He studies the effect of
political institutions on the prime ministerial leadership and the role
of the Japanese House of Councillors in Japanese politics. His publi-
cations include Failed Democratization in Pre-war Japan: the
Breakdown of a Hybrid Regime (Stanford University Press, 2014),
Sangiin to ha Nanika (The role of the Japanese House of Councillors)
(Chuokoron Shinsha, 2010; winner of the Osaragi Jiro Rondan Prize
2011), and Shusho Shihai (Dominance of the Japanese prime minis-
ter) (Chuko Shinsho, 2006). Takenaka holds a BA from the Faculty
of Law of the University of Tokyo and a PhD in political science from
Stanford University.

Takashi Unayama is Professor of Economics at Kyoto University. His

research interests are household behavior, public policy, household
surveys, and issues in the Japanese economy. He especially focuses
on consumption-associated topics such as responses to tax cuts/
increases, public pension/child benefit payments, and retirement.
He is a member of the Statistics Commission and the Fiscal System
Council of Japan. He holds a BA and a PhD in economics from the
University of Tokyo.
List of Contributors xxiii

David E. Weinstein is Carl S. Shoup Professor of the Japanese Economy

at Columbia University. He is also director of the Center on Japanese
Economy and Business, director of the Japan Project at the National
Bureau of Economic Research, and a member of the Federal
Economic Statistics Advisory Committee. Previously, Professor
Weinstein was Chair of the Economics Department; senior econo-
mist as well as a consultant at the Federal Reserve Bank of
New York, the Federal Reserve Bank of San Francisco, and the
Federal Reserve Board of Governors; and a visiting professor at the
London School of Economics. Prior to joining the Columbia faculty,
Professor Weinstein held professorships at the University of
Michigan and Harvard University. He also served on the Council
of Economic Advisors from 1989 to 1990. Professor Weinstein
earned his PhD and MA in economics from the University of
Michigan and his BA at Yale University. He is the recipient of
many grants and awards, including five National Science
Foundation grants, an Institute for New Economic Thinking grant,
a Bank of International Settlements Fellowship, and a Google
Research Award.
Johannes Wieland is an associate professor of economics and
Distinguished Endowed Chair in Macroeconomics and Public
Finance at the University of California, San Diego. He is also
a research associate at the National Bureau of Economic Research.
His research interests are in macroeconomics and economic history,
with a focus on economies in deep recessions. Johannes holds a BA in
economics from the University of Cambridge, an MPhil in economics
from the University of Oxford, and a PhD in economics from the
University of California, Berkeley. From 2018 to 2019 he worked as
a visiting research economist at the Federal Reserve Bank of Chicago.

Steven K. Vogel is Chair of the Political Economy Program, the Il Han

New Professor of Asian Studies, and a professor of political science
at the University of California, Berkeley. He specializes in the polit-
ical economy of the advanced industrial countries, especially Japan.
He is the author of Marketcraft: How Governments Make Markets
Work (Oxford University Press, 2018) and Japan Remodeled: How
Government and Industry Are Reforming Japanese Capitalism
(Cornell University Press, 2006). His first book, Freer Markets,
xxiv List of Contributors

More Rules: Regulatory Reform in Advanced Industrial Countries

(Cornell University Press, 1996), won the Masayoshi Ohira
Memorial Prize. He has been a columnist for Newsweek Japan and
the Asahi Shimbun, and he has worked as a reporter for the Japan
Times in Tokyo and as a freelance journalist in France. He has a BA
from Princeton University and a PhD in political science from the
University of California, Berkeley.
Acknowledgments

This book is the result of two conferences that we organized when we

were colleagues at the Shorenstein Asia Pacific Research Center
(S-APARC) at Stanford University. The first conference was held in
February 2018; the second was in March 2019. We felt lucky to work
with so many top researchers in the field of Japanese political economy.
We thank all the authors, who provided excellent analyses of various
aspects of the political economy of Japan under the Abe administration.
Many authors attended both conferences and contributed to active
discussion in addition to writing up their own chapters.
The project benefited tremendously from participants who
contributed to the conferences as discussants. They include Seung
Cho, Saori Katada, Yong Suk Lee, Huiyu Li, Kenneth McElwain,
Curtis Milhaupt, Yuhei Miyauchi, Thuy Lan Nguyen, Katheryn Russ,
Dan Smith, Rob Weiner, Scott Wilbur, Yu Jin Woo, and John Zysman.
Our project also received excellent feedback from participants at the
2018 Yale Workshop on Japanese Politics and Diplomacy, organized
by Frances Rosenbluth, and two faculty seminars organized by our
colleagues at S-APARC in 2018 and 2019. We also thank the
anonymous reviewers for their helpful suggestions.
The two conferences that led to this book were part of the Stanford
Juku in Political Economy, which we inaugurated in 2013 to promote
scholarship and foster a sense of community among scholars who study
Japan. We have been delighted to see the Juku inspiring similar
initiatives by past participants. This book continues a long tradition
of collaborative work on contemporary Japan convened at S-APARC,
including most recently Political Change in Japan: Electoral Behavior,
Party Realignment, and the Koizumi Reforms (Brookings Institution,
2009), by Steven Reed, Kenneth McElwain, and Kay Shimizu, and
Japan under the DPJ: The Politics of Transition and Governance
(Brookings Institution, 2013), by Kenji Kushida and Phillip Lipscy.

xxv
xxvi Acknowledgements

Financial support for this project was provided by S-APARC and by

the Freeman and Spogli Institute for International Studies (FSI), both at
Stanford University; by the Japan Foundation Center for Global
Partnership; and by the Japan Society for the Promotion of Sciences
(JSPS). We thank them for making this project possible.
Indispensable in organizing the conferences and collecting chapters
from the authors was the superb support of staff members at S-APARC,
especially Meiko Kotani and Amanda Davidovich. We are grateful to
them for their support of this project as well as for their vital
contributions to the Japan Program during our time at Stanford.
Joe Ng at Cambridge University Press saw value in our project and
guided us through the process of publishing this book. We thank him,
as well as Gemma Smith, Gayathri Tamilselvan, and Wade Guyitt for
their help and patience. We also thank Koteki Cho at the University of
Tokyo for indexing the manuscript.
This book is dedicated to the memory of Masahiko Aoki, who was
our esteemed colleague at Stanford University. His work on economic
theory, comparative institutional analysis, and the Japanese political
economy is very well known. His class on the Japanese economy at
Stanford attracted and inspired many students, including a younger
version of one of us (Lipscy). When a group of us sought to revive Japan
studies at S-APARC by establishing the Japan Program in 2011, he
assumed early leadership as founder. He was a dear friend, mentor, and
colleague. Masa Aoki was one of the pioneers who created the study of
the Japanese political economy as we know it today. Although we
started this project after he had passed away, we hope our work sits
well with the intellectual trajectory that he originated.
Abbreviations

AI artificial intelligence
APCs agricultural production corporations
ASEAN Association of South East Asian Nations
AWF Asian Women’s Fund
BOJ Bank of Japan
CDPJ Constitutional Democratic Party of Japan
CEFP Council of Economic and Fiscal Policy
CEO chief executive officer
CGC Corporate Governance Code
CLB Cabinet Legislation Bureau
CNN Cable News Network
CO2 carbon dioxide
COVID coronavirus disease
CPI Consumer Price Index
CPTPP Comprehensive and Progressive Agreement for
Trans-Pacific Partnership
CRRA constant relative risk aversion
CSD collective self-defense
CWNC company with nominating committee
DARPA Defense Advanced Research Projects Agency
DPJ Democratic Party of Japan
EEOL Equal Employment Opportunity Law
EPA Economic Partnership Agreement
ETF exchange-traded fund
EU European Union
FDI foreign direct investment
FIES Family Income and Expenditure Survey
FIT feed-in tariff
FLCBs farmland consolidation banks
FOIP Free and Open Indo-Pacific
FTA Free Trade Agreement

xxvii
xxviii List of Abbreviations

G7 Group of Seven
GATT General Agreement on Tariffs and Trade
GDP gross domestic product
GPIF Government Pension Investment Fund
ICT Information and Communications Technology
ILO International Labour Organization
IMF International Monetary Fund
INDEC Intended Nationally Determined Contributions
IoT Internet of Things
IPOs initial public offerings
ISS Institutional Shareholder Services
IT information technology
J-NSC Jiminto (Liberal Democratic Party) Net Supporters
Club
JA Japan Agricultural Cooperatives
JANE Japan Association of New Economy
JCG Japan Coast Guard
JILPT Japan Institute for Labor Policy and Training
JIP Japan Industrial Productivity
JGBs Japanese Government Bonds
JMA Japan Medical Association
JSC Japanese Stewardship Code
JSDF Japan Self-Defense Forces
JSP Japan Socialist Party
KPIs key performance indicators
LCC low-cost carriers
LDP Liberal Democratic Party
MAFF Ministry of Agriculture, Forestry and Fisheries
M&A mergers and acquisitions
METI Ministry of Economy, Trade, and Industry
MHLW Ministry of Health, Labor and Welfare
MOE Ministry of the Environment
MOF Ministry of Finance
MOFA Ministry of Foreign Affairs
MP Member of Parliament
MTDPs Medium-Term Defense Programs
NATO North Atlantic Treaty Organization
NDPG National Defense Program Guidelines
List of Abbreviations xxix

NEDO New Energy and Industrial Development

Organization
NIH National Institute of Health
NIPSSR National Institute of Population and Social Security
Research
NISA Nuclear and Industrial Safety Agency
NRA Nuclear Regulatory Authority
NSC National Security Council; Nuclear Safety
Commission
NSF National Science Foundation
NSS National Security Secretariat
OECD Organisation for Economic Co-operation and
Development
PARC Policy Affairs Research Council
PAYGO pay-as-you-go
PFG Party for Future Generations
PM prime minister
PPP purchasing power parity
PR proportional representation
QQE quantitative and qualitative easing
R&D research and development
ROA return on assets
ROE return on equity
SMD single-member district
SMEs small and medium-sized enterprises
SNTV-MMD single nontransferable vote, multimember district
SRJE Secretariat for the Revitalization of the Japanese
Economy
STOVL short take-off and vertical landing
TEPCO Tokyo Electric Power Company
TFP total factor productivity
TLOs Technology Licensing Offices
TPP Trans-Pacific Partnership
TSE Tokyo Stock Exchange
UN PKO United Nations Peacekeeping Operations
UNESCO United Nations Educational, Scientific and Cultural
Organization
UNFCCC United Nations Framework Convention on Climate
Change
xxx List of Abbreviations

US/USA United States of America

VAT value-added tax
VAR vector auto regression
VC venture capital
part i

Introduction
1 The Political Economy of the Abe
Government
t a k e o ho s h i a n d p h i l l i p y . l i p s c y

1.1 Introduction
The government of Shinzo Abe, which ruled Japan from 2012 to 2020,
represents an important turning point in Japanese politics and political
economy. Abe became the longest-serving prime minister in Japanese
history, reversing a trend of short-lived leaders. But his government
stands out not only for its longevity but also for its policies: Abe
implemented a variety of significant changes, among the most import-
ant being a series of economic reforms to reinvigorate Japan’s economy
under the banner of “Abenomics.” In this introductory chapter, we
provide a broad review of the Abe government and its policies and
point out several apparent puzzles that motivate the volume.
We argue that the Abe government is the clearest manifestation of
a new Japanese political system that represents a full transition away
from the 1955 system. The new system is characterized by a strong
prime minister with centralized authority, careful management of pub-
lic opinion, the strategic use of elections for party discipline, and a focus
on policies with broad, popular appeal. Abe utilized and strengthened
Japan’s new political institutions to implement significant policy
changes across a wide range of issue areas, including monetary and
fiscal policy, the labor market, corporate governance, agricultural
reforms, and national security.
The first two “arrows” of Abenomics (“bold monetary policy” and
“flexible fiscal policy”) were expansionary macroeconomic policies
that sought to normalize Japan’s economy by bringing an end to
persistent deflation. To implement “bold” monetary policy, Abe
appointed Haruhiko Kuroda as governor of the Bank of Japan (BOJ),
and the BOJ engaged in an unprecedented expansion of its balance
sheet. “Flexible” fiscal policy meant that the government pursued fiscal
expansion to stimulate the economy while being mindful of fiscal

3
4 Takeo Hoshi and Phillip Y. Lipscy

prudence. Thus, fiscal policy also turned more expansionary, but the
government raised the value-added tax (VAT) rate in April 2014 as the
starting point for fiscal consolidation in the long run. The VAT rate was
scheduled to be raised again in October 2015, but less-than-stellar
economic growth prompted the government to postpone the tax hike
twice before it was finally enacted in October 2019. During Abe’s
tenure, deflation came to an end, and, until the 2020 coronavirus
disease (COVID-19) pandemic, the Japanese economy grew relatively
consistently. Nonetheless, the 2 percent inflation rate target, which was
supposed to be achieved in two years, remained unattained even after
seven years, and the goal of raising nominal GDP to 600 trillion yen by
2020 was out of reach even before the pandemic.
Many economic reforms were included in the third arrow of
Abenomics, also known as “growth strategy to promote private invest-
ment.” Here, Abe pursed various industrial promotions and a host of
structural reforms to increase Japan’s potential growth rate. Among
other things, Abe’s government sought reforms in unproductive sectors
such as agriculture; deregulation of the electricity distribution market;
promotion of innovation and entrepreneurship, especially in promising
industries such as artificial intelligence (AI), robotics, and healthcare;
“womenomics” policies to empower women in the labor market; other
labor market reforms to establish “work-life” balance; corporate gov-
ernance reform; and attempts to open up Japan to the rest of the world,
including participation in the Trans-Pacific Partnership (TPP) and the
promotion of inbound tourism and inward foreign direct investment.
The reforms were first packaged in the growth strategy that was
announced in June of 2013. The emphasis on various economic
reforms changed over time as the growth strategy was revised annually.
In foreign policy, Abe’s approach represented a mix of nationalism
and pragmatism. This was a surprise to many foreign observers who
initially reacted with consternation to Abe’s ascendance, fearing he
would pursue a revisionist agenda and exacerbate tensions with
regional neighbors (Oros 2017, pp.129–130). Abe faced international
criticism when he visited the controversial Yasukuni Shrine in 2013,
and conflicts over history issues with South Korea intensified after
2017, when President Moon Jae-in rose to power. However, Abe’s
foreign policymaking largely reflected a pragmatic approach, con-
strained by both geopolitical and domestic realities, such as the con-
tinuing rise of China and the electoral codependence of the Liberal
The Political Economy of the Abe Government 5

Democratic Party (LDP) on the more moderate Komei Party (Liff &
Maeda 2019). Among other things, Abe refrained from visiting
Yasukuni Shrine after 2013, negotiated an agreement with South
Korea on compensating comfort women in 2015, became the first
Japanese prime minister to visit Pearl Harbor in 2016, and pursed
diplomatic rapprochement with China (see Chapter 17). After the
victory of Donald Trump in the 2016 US presidential election, Abe
pursued a “bromance” foreign policy that established himself as one of
the president’s closest foreign counterparts.
Abe orchestrated important shifts that built on long-term evolution-
ary trends in Japanese security policy (see Chapter 16). These included
several major legislative victories: secrecy legislation in 2013 to facili-
tate greater intelligence sharing; security legislation in 2015 that allows
the Japan Self-Defense Forces (JSDF) to engage in collective self-
defense under limited circumstances; and anti-terrorism legislation in
2017 that increases the government’s domestic surveillance authority.
The Abe government reversed a decade-long decline in Japanese
defense spending, which reached postwar highs after 2015. The Free
and Open Indo-Pacific (FOIP) vision, developed by the Japanese
Ministry of Foreign Affairs, was embraced by the Trump administra-
tion during Abe’s tenure, though the substance of FOIP remained
somewhat vague and contested (Hosoya 2019).
Several of Abe’s personal priorities proved elusive. Revision of the
Japanese constitution was a long-standing goal for Abe, and his large
Diet majorities created an unprecedented opportunity. However, Abe
struggled throughout his tenure to galvanize sufficient political and
popular enthusiasm for the idea (McElwain 2020). Despite consider-
able efforts, Abe also ultimately could not make progress on the issue of
North Korea abductions of Japanese citizens or conclude a peace treaty
with Russia.
Abe’s government also invited significant controversy. Abenomics
reforms came under fire for a variety of shortcomings, such as stagnant
consumption growth (Chapter 6), lack of significant inflation despite
unprecedented monetary expansion (Chapter 7), and inadequate atten-
tion to climate change (Chapter 15). Journalists assailed Abe’s heavy-
handed approach toward the media, particularly liberal news outlets
such as Asahi (Fackler 2016). The Abe government adopted an assert-
ive approach toward international organizations despite widespread
criticism, withdrawing from the International Whaling Commission
6 Takeo Hoshi and Phillip Y. Lipscy

and threatening to pull funding from UNESCO over its handling of the
Nanking Massacre (Lipscy 2020). The government’s security legisla-
tion invited large-scale protests from civil society groups fearing
entanglements in military conflicts and potential violations of human
rights.

1.2 Puzzles
The track record of the Abe government raises some intriguing puzzles
for scholars of Japanese politics and political economy. First, how was
Abe able to remain in and exert power for so long? After Koizumi,
Japan saw a revolving door of six prime ministers come to power and
follow a predictable pattern of collapsing approval ratings and resig-
nation within about a year. Abe himself was among these short-lived
prime ministers, lasting only a year in 2006–2007. What was different
in 2012? Figure 1.1 shows lower house election results for the Japanese
Diet from 1986 to 2017. Abe won three consecutive landslide victories
in the lower house – in 2012, 2014, and 2017 – which saw his LDP and
coalition partner Komei capture about two-thirds of total seats. The

100% JCP JCP JCP JCP JCP

JCP JCP Other Other
Other Other Other
Other Other
Other Other Ishin
90% Other
Other Ishin
Ishin Hope
DPJ
80%
SDPJ DPJ DPJ
SDPJ
DPJ DPJ CDPJ
70%
Komei Komei Komei Komei Komei
DPJ
60% SDPJ
Komei DPJ SDPJ
Komei Komei
50% Komei
SDPJ

40%

30% LDP
LDP LDP LDP LDP
LDP Komei
LDP LDP
LDP
LDP
20%

LDP
10%

0%
1986 1990 1993 1996 2000 2003 2005 2009 2012 2014 2017

Figure 1.1 Election results in the Japanese lower house, 1986–2017

The Political Economy of the Abe Government 7

magnitude of these victories is comparable to the 2005 election under

Koizumi, which was considered a historical landslide at the time
(Kabashima & Steel 2007). How was Abe able to win decisive victories
with such reliability and frequency?
Second, how was Abe able to overcome intraparty divisions that
bedeviled governments of his predecessors under both the LDP and
Democratic Party of Japan (DPJ)? Prime ministers of both parties
struggled to reconcile the conflicting priorities of reformist politicians,
primarily drawn from urban districts, and conservatives representing
rural districts. In contrast, Abe successfully marginalized potential
opponents within the LDP, such as Shigeru Ishiba, and faced limited
internal challenges. One illustration of this is Figure 1.2, which shows
the passage rate of legislation submitted to the Diet from 1980 to 2019,
separated by cabinet-submitted legislation and legislation submitted by
lawmakers. The DPJ government was characterized by an unusually
low passage rate of cabinet-submitted legislation and a high passage
rate of non-cabinet legislation, reflecting internecine conflict within the
party (Kushida & Lipscy 2013). In contrast, the Abe government saw

Passage Rate of Diet Legislation (%), 1980–2019

100%

90%

80%
Cabinet-Submitted Legislation
70% Non-Cabinet Legislation

60%

50% DPJ LDP

(Abe)
40%

30%

20%

10%

0%
1980
1982
1984
1986
1988
1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018

Figure 1.2 Passage rate of Diet legislation (%), 1980–2019

8 Takeo Hoshi and Phillip Y. Lipscy

the passage rate of cabinet-submitted legislation recover, and non-

cabinet legislation fell to historically low levels. The Abe government
maintained meaningful discipline vis-à-vis backbenchers despite its
large majorities.
There are also interesting puzzles related to the economic policies of
Abenomics. First, despite extremely expansionary macroeconomic pol-
icy, especially monetary policy, why did economic growth and inflation
remain low? The targeted inflation rate of 2 percent, which was sup-
posed to be reached quickly by 2015, was not achieved even five years
later. Economic growth remained below the target growth rates of
3 percent in nominal terms and 2 percent in real terms. Although
Japan may have escaped deflationary conditions, the recovery was
less impressive than many expected given the unprecedented size of
the macroeconomic stimulus.
Second, despite continued concerns about mounting government
debt, why was the government able to roll over the existing debt and
issue new debt with higher prices (i.e., lower yields) without a credible
prospect for fiscal consolidation? The second arrow of Abenomics
(flexible fiscal policy) was supposed to be fiscal expansion in the short
run combined with a clearly specified path for fiscal consolidation in
the long run. Fiscal expansion was implemented, but fiscal consolida-
tion proved more challenging. The consumption tax hike is a good
example. After the tax rate was increased from 5 percent to 8 percent
in April 2014, another tax hike from 8 percent to 10 percent was
planned for October 2015. It was postponed twice and only went into
effect in 2019.
Third, how was Abe able to implement some structural reforms in
areas such as agriculture and the labor market, which were character-
ized by entrenched resistance in the past? For example, Abe systemat-
ically undermined Japan Agricultural Cooperatives (JA), part of the
LDP’s traditional support base, while pursuing the TPP – eventually
enacted as the Comprehensive and Progressive Agreement for Trans-
Pacific Partnership (CPTPP) after US withdrawal – which included
significant reductions in agricultural protection (see Chapter 14).
Many LDP politicians feared a repeat of the 2007 upper house and
2009 lower house elections, when Ichiro Ozawa took advantage of
rural alienation with the LDP to guide the DPJ to major victories in the
countryside (Lipscy 2017). Why was Abe willing to gamble rural
support to pursue agricultural reforms? Conversely, why were reforms
The Political Economy of the Abe Government 9

less ambitious in other areas, such as climate change policy, promotion

of inward direct investment, policies to increase corporate entries and
exits, and measures to encourage fluidity of the labor market?

1.3 A New Japanese Political System

Abe’s government represents the culmination of political reforms in
Japan since the mid-1990s, which fundamentally altered the Japanese
political system. It is now appropriate to characterize Japan as operat-
ing under a new political system that represents a full transformation
from the 1955 system, which ended in 1993. The following character-
istics are defining features of this new political system.
First, electoral reform in 1994 shifted the incentives of Japanese
lower house politicians from particularism – the narrow targeting of
pork to organized supporters – to broad appeal to the Japanese median
voter. The consequences of these reforms are well known.1 Japanese
political leaders now face considerable electoral incentives to pursue
economic reforms that enhance macroeconomic growth and increase
the prosperity of Japanese households at the expense of traditional,
organized interest groups.
Second, bureaucratic scandals and administrative reforms since the
late 1990s have shifted political authority from traditionally powerful
ministries – such as the Ministry of Finance (MOF) and the Ministry of
Economy, Trade, and Industry (METI) – to politicians, particularly the
Cabinet Office of the Prime Minister (Estevez-Abe 2006; Takenaka
2006; Mulgan 2017). Expansion of the administrative capacity of the
Cabinet Office also elevated the authority of the prime minister vis-à-
vis backbenchers, who previously exerted influence in specialized pol-
icy areas (Schoppa 1991). The personnel of the Cabinet Secretariat
expanded by more than three times since 2000. Abe’s Cabinet
Secretariat came to employ about 3,000 officials, a doubling since the
Koizumi years and comparable to the staff of the Ministry of Foreign
Affairs (see Chapter 2).
Third, the Japanese upper and lower houses are governed by
distinct electoral rules, which creates significant challenges for
Japanese political parties. The upper house is still largely governed
by the traditional SNTV-MMD (single nontransferable vote, multi-
member district) system, with a strong rural bias due to an emphasis
on prefectural districts and the tendency for sparsely populated rural
10 Takeo Hoshi and Phillip Y. Lipscy

districts to become decisive, single-member districts. Furthermore,

the upper house has nearly co-equal authority, requiring a two-
thirds lower house majority to override its decisions. The need to
appeal to very distinct constituencies in each house contributes to
two patterns in recent Japanese politics that stymie policymaking.
The first is divided government or a “twisted Diet,” in which differ-
ent parties control each house. The second is intraparty conflict, in
which a party successfully captures majorities in both houses but
becomes internally divided. This incongruity proved particularly
challenging for the DPJ, which drew from a primarily urban support
base while incorporating Ozawa’s traditionalist approach to appeal
to rural areas (Lipscy 2017).
Fourth, prime ministers are critically dependent on high public
approval ratings in order to maintain party cohesion. The primary
leverage Japanese prime ministers can exercise vis-à-vis backbenchers
within their party is the threat of withdrawing official support during
national elections. This threat is not credible when the prime minister is
unpopular and hence unable to implicitly threaten an early election.
Japanese prime ministers therefore have strong incentives to maintain
their popularity with the general public. Abe maintained relatively
robust public approval ratings during his tenure, averaging slightly
below 50 percent according to polling by NHK (Nippon Hoso
Kyokai), in the same range as Junichiro Koizumi (51.4 percent), the
only other prime minister to enjoy longevity in recent years.2 All other
recent prime ministers had lower average approval ratings and stepped
down when their approval ratings sank to around 30 percent. This
contrasts to prime ministers under the 1955 system, who routinely
remained in power despite average approval ratings in the 30 percent
range, such as Kishi, Sato, Tanaka, and Nakasone. This creates an
additional impetus to support policies that appeal broadly to
Japanese voters.

1.4 Taking Advantage of Japan’s New Political Institutions

Japan’s institutional transformation began in the 1990s, but Abe is the
first prime minister who established a stable governance structure that
has some chance of replication by future prime ministers. As Abe
himself noted, “it is true that prime ministers are in a strong position
under a single-member district electoral system. However, it is not
The Political Economy of the Abe Government 11

unconditional strength: if prime ministers were unconditionally power-

ful, my first administration would not have ended in one year.”3
Two previous governments that emerged after the 1990s appeared to
mark the beginning of a new era in Japanese politics: the 2001–2006
Koizumi government and the 2009–2012 DPJ government. Koizumi’s
tenure exhibits some similarities to Abe’s. His rule was buoyed by
strong public approval, and he implemented consequential changes,
including transportation sector reforms and the privatization of the
postal savings system (Kabashima & Steel 2007; Reed, McElwain &
Shimizu 2009). However, his leadership style, which relied on personal
charisma and sought to “destroy the LDP,” was defined by internal
conflict within his own party. Koizumi’s approach could not be repli-
cated by subsequent prime ministers. This included Abe himself, who
succeeded Koizumi in 2006 only to become the first in a string of six
prime ministers who lasted for only about one year each.
The DPJ came to power in a historic lower house victory in 2009,
marking the first decisive electoral defeat of the LDP by an opposition
party (Lipscy & Scheiner 2012; Kushida & Lipscy 2013; Funabashi &
Nakano 2017). However, the DPJ struggled to implement meaningful
reforms due to persistent internecine conflict and resistance from the
Japanese bureaucracy. The party also lost control of the upper house in
2010, making it difficult to pass legislation in a “twisted Diet.” In other
words, prior to Abe, Japan’s new political institutions were not associ-
ated with a clear and replicable governance model.
The Abe government broke this trend. Unlike Koizumi, Abe ruled as
a political insider, largely eschewing public attacks on his own party or
the bureaucracy. Abe also maintained party discipline and cohesion
despite the LDP’s large majorities in both houses of the Diet, a marked
contrast to the Koizumi and DPJ governments. A popular view of
the second Abe government is that it stood out for competent manage-
ment and discipline centered on a core group of officials in the Cabinet
Office, particularly Chief Cabinet Secretary Yoshihide Suga.4
However, many of the governance practices implemented under Abe
were practical measures that could be adopted by future governments.
This section briefly reviews what Abe government officials believe they
did differently compared to previous Japanese administrations.5
The second Abe government was supported by a relatively stable core
group of close advisors surrounding the prime minister. Abe brought
back many officials from his first tenure, most notably Yoshihide Suga,
12 Takeo Hoshi and Phillip Y. Lipscy

Hiroshige Seko, Katsunobu Kato, and Seiichi Eto among politicians and
Takaya Imai, Eiichi Hasegawa, Hiroshi Suzuki, and Motoya Nakae
among former bureaucrats.6 Suga and Imai played a particularly out-
sized role as respectively chief cabinet secretary and executive secretary
to the prime minister. These officials developed close relationships with
Abe during his first administration and remained loyal even after Abe
resigned (Hoshi 2014, p.176). Imai attributes his close relationship with
Abe in part to the fact that his uncle had worked directly under
Nobusuke Kishi, Abe’s grandfather, when the latter was minister of
commerce and industry under the Tojo government in 1941 (Oshita
2017, pp.52–53).
These close advisers to Abe note that the lessons from the first Abe
government in 2006–2007 were critical in shaping their management
style during Abe’s second tenure as prime minister. Abe himself has
publicly stated that learning from the mistakes of his first tenure was
key to his second prime ministership.7 Furthermore, many of the
reforms that would be implemented under the banner of Abenomics
after 2012 were developed within the prime minister’s Council of
Economic and Fiscal Policy (CEFP) during Abe’s first tenure as prime
minister (Chapter 5).

1.4.1 Managing Public Opinion

Officials close to Abe sought to manage public opinion to consistently
maintain the prime minister’s popularity at a high level. This included
initiatives to buoy Abe’s public approval (primarily by focusing on
economic reforms under the banner of Abenomics), a careful strategy
to time and selectively pursue policies unpopular with the general
public, and a media strategy that sought to minimize political fallouts
from negative news events. This focus on public approval reflected
perceived lessons from Abe I, during which Abe’s public approval
rating plummeted quickly, triggering internal discord within the LDP,
a major upper-house election loss, and Abe’s resignation.
Abenomics was designed by the Abe administration to enhance
macroeconomic performance and appeal to the reformist inclinations
of the general public. Abe saw Abenomics as a defining feature of his
government and credited his success in enacting controversial legisla-
tion, such as the secrecy and security laws, to strong public support
underpinned by his economic policies. Suga similarly characterized the
The Political Economy of the Abe Government 13

Abe government as using its high public approval rating to make

progress on diplomatic and security reforms that could cause tempor-
ary unpopularity (Oshita 2017, pp.13–14, p.182). Officials close to
Abe note that Abe’s personal priority always remained his long-
standing ambition to “escape from the postwar order” by enacting
constitutional reform, but this was seen as an impossibility without
successful economic measures to maintain strong public approval
(Tazaki 2014, p.135).
Abe’s team also explicitly sought to manage the prime minister’s
popularity by sequencing and timing legislation so that unpopular or
controversial laws were followed by more popular initiatives. Suga is
said to have predicted that passing the Secrecy Law in 2013 would lead
to a roughly 10 percent hit to the prime minister’s approval rating.
However, the legislation was deemed sufficiently important and pro-
posed after upper house elections to minimize electoral impact.
Similarly, after seeing the prime minister’s popularity decline after
controversies surrounding security legislation enabling collective self-
defense, Imai coordinated a pivot back to popular economic reforms
under the banner of the “second stage” of Abenomics and “dynamic
engagement of all 100 million citizens” (Oshita 2017, pp.180–181,
p.241). Furthermore, Abe’s nationalist credentials may have contrib-
uted to the resilience of his public approval ratings: Chapter 3 suggests
that Abe’s core conservative supporters were willing to return to the
fold even after temporary drops in popularity due to controversies and
scandals.
Another perceived lesson of Abe I was the failure to manage the
media and news cycle. Abe’s advisors concluded that the prime minis-
ter’s popularity suffered during his first tenure due to poor media
management. They inferred that attempting to manage scandals
through the resignation of officials is often counterproductive, as it
immediately leads to questions about the prime minister’s judgement in
making the initial appointment. During Abe I, the pattern was set early
when Masaaki Honma resigned as chairperson of the Research
Commission on the Tax System in December 2006 due to a personal
scandal. Abe’s government was then beset by successive scandals and
resignations, which dominated the news cycle and diminished the
prime minister’s popularity. In his second government, Abe reversed
this approach and resisted calls for resignation of his ministers and
lieutenants even during major scandals, except in exceptional cases
14 Takeo Hoshi and Phillip Y. Lipscy

where it was judged that the official would not be able to survive
(Tazaki 2014, pp.76–77).
The second Abe government also sought to exert direct control over
the media cycle. Responding to scandals through indecision and even-
tual resignation, as was the pattern under Abe I, meant the media
would spend many days covering new developments. Abe’s advisors
sought to limit media coverage to no more than one or two days by
acting decisively: either firmly resisting resignation and moving on or
opting for immediate resignation. In cases where multiple resignations
were necessary, Abe sought to announce them simultaneously to limit
the duration of negative media coverage (Tazaki 2014, pp.84–85). The
Abe government “proved much better at dealing with the media,
including how to limit scandal coverage” (Carlson & Reed 2018).
Another change was elimination of burasagari interviews. These
were brief, daily interviews that created opportunities for the media
to ask (often inconvenient) questions directly to the prime minister. The
burasagari interviews were canceled by the DPJ government following
the Great Eastern Japan Earthquake of 2011. The media hoped the Abe
administration would revive the practice, but Abe did not, and the TV
media lost the ability to attach the prime minister’s face to negative
stories. Instead, the soft-spoken and careful Suga became the public
face of the government to the media. Suga maintained careful message
discipline, claiming to follow Colin Powell’s dictum that “reporters
have a right to ask questions, and I have a right not to answer them”
(Hoshi 2014, p.177).
According to Tazaki (2014, p.91), Abe referred to the cancellation of
burasagari interviews as “the only meaningful achievement of the DPJ
government.” The elimination of burasagari interviews also improved
discipline within the cabinet by making it less feasible for bureaucrats
or politicians to work around the formal decision-making process by
going straight to the prime minister and getting him to publicly commit
to a policy direction in front of the media. This was an important
problem in the DPJ government, in which prime ministers often over-
turned policy in media interviews without careful consultation.
Abe also ended the previous practice of scheduling prime minister
interviews according to a set schedule that rotated between NHK and
the private media. Instead, he responded only to interview requests he
wanted to field, which led to a disproportionate share of interviews
granted to the relatively sympathetic and conservative Yomiuri and
The Political Economy of the Abe Government 15

Sankei (Tazaki 2014, pp.86–88). Furthermore, the LDP under Abe

sought to shape narratives on social media by supporting the
“Jiminto [LDP] Net Supporters Club (J-NSC),” a volunteer group of
over 19,000 members that promotes narratives supportive of the gov-
ernment and critical of opposition parties.8
The Abe government’s media strategy, though perhaps effective at
buttressing his popular approval, came under sharp criticism by jour-
nalists and advocates of press freedom. For example, the government
was criticized for threatening to invoke the Broadcast Law and Radio
Law to silence broadcasters that did not conform to the government’s
interpretation of political impartiality (Fackler 2016; Kingston 2016).
The Abe government and its conservative allies were particularly out-
spoken about media coverage of historical memory issues, such as
reporting on the comfort women issue by the Asahi Shimbun
(Chapter 17). Japan’s World Press Freedom Index ranking fell from
an all-time high of eleventh place in 2011 to sixty-seventh by 2019
(Reporters Without Borders 2019).

1.4.2 Relations with the LDP and the Bureaucracy

Close advisors to Abe claim that they were politically inexperienced
during Abe I and had limited ability to manage the internal politics of
the LDP. This meant Abe was blindsided by internal discord that
sought to elevate Fukuda as an alternative as the prime minister’s
popularity diminished. In contrast, during Abe II, his advisors sought
to keep a careful pulse on the party and nip any signs of anti-Abe
movements in the bud. One important aspect of Abe’s strategy in
managing LDP backbenchers was the use of lower house elections as
a disciplinary mechanism.
Abe strategically called lower house elections in 2014 and 2017,
enhancing his credibility vis-à-vis potential political opponents within
his own party and the bureaucracy. For this purpose, the 2014 election
was critical. An important rationale for Abe calling the 2014 election
was an internal challenge from LDP politicians supported by the
Ministry of Finance, which opposed Abe’s proposed postponement of
a consumption tax hike. Although they chose not to publicly frame the
election in these terms in order to maintain an impression of party
unity, there was serious concern that the government could be brought
down in the absence of an election. Abe and Suga feared that LDP
16 Takeo Hoshi and Phillip Y. Lipscy

opponents would withhold support from the enabling legislation for

consumption tax postponement, threatening the survival of the prime
minister (Tazaki 2014, pp.9–10, p.17). By calling the election and
running on postponing the consumption tax, Abe secured a clear public
mandate and strengthened party discipline.
Although the specific context was different, the strategy echoed that
of Koizumi in 2005. Koizumi faced opposition from “postal rebels” –
those opposed to privatization of the postal system – within his own
party. In response, Koizumi expelled the rebels and dissolved the lower
house, securing a strong mandate for reform through electoral victory.
Both Koizumi and Abe silenced their internal opponents by demon-
strating their willingness to call elections, and in turn, the credibility of
this threat hinged on the prime minister’s public support. This strategy
also reflected a perceived failure of Abe I, when the prime minister
readmitted the postal rebels who had been expelled from the LDP by
Koizumi. This led to criticism that Abe was reversing Koizumi’s pol-
icies despite inheriting his large lower-house majority from the 2005
postal election. This provided an additional rationale under Abe II to
call lower house elections when the prime minister announced signifi-
cant policy shifts.
Abe also sought to improve coordination at all levels – within the
Cabinet Office, within the LDP, and with bureaucrats. During Abe I,
Abe appointed relatively young and ambitious politicians, such as
Yuriko Koike and Hiroshige Seko, to special advisor roles. However,
this proved dysfunctional: the special advisors lacked formal legal
authority over bureaucrats and had limited sway over backbenchers
in the LDP, limiting their influence (Tazaki 2014, pp.52–53). There
was no coordination mechanism among the special advisors, which led
to communication problems. Under Abe II, all of Abe’s immediate
advisors convened daily in a meeting for twenty minutes to coordinate
strategy and perspectives: such a coordination mechanism did not exist
under Abe I or the DPJ government, leading to scattered decision-
making and inadequate message discipline. In addition, to enhance
coordination between the prime minister and the Diet, officials close
to Abe, such as Seko, actively participated in the Diet Affairs
Committee in order to exchange information about priorities and
coordinate legislative strategy with LDP Diet members.
The Abe government also sought to incorporate lessons from failures
of the preceding DPJ government. Under the DPJ, politicians close to
The Political Economy of the Abe Government 17

the prime ministers had become special advisors, playing a role trad-
itionally occupied by executive secretaries. Abe saw this as an ineffect-
ive strategy, as politicians had a strong incentive to take credit for their
accomplishments, rather than working behind the scenes in support of
the prime minister. Abe thus primarily relied on Imai, a METI bureau-
crat, to play this role.
The Abe government also exerted strong control over the bureau-
cracy. Abe strengthened personnel authority over the bureaucracy and
centralized authority under the prime minister, particularly through the
creation of the Cabinet Bureau of Personnel Affairs in 2014 (Chapter
2). Abe made a point of putting the bureaucracy on notice in
January 2013, within weeks of taking power, by overturning a long-
held precedent that the Japanese Supreme Court appoints its own
justices, exerting his authority under the constitution to reject the
court’s preferred appointee (Tazaki 2014, pp.63–64).
Abe personally credited Suga with making calculated personnel deci-
sions to maximize political control and minimize bureaucratic resist-
ance. Even before he became chief cabinet secretary, Suga had
established a reputation as an effective manager of Japan’s traditionally
powerful bureaucracy. Suga published a book while the LDP was still
out of power, subtitled “A Politician’s Resolution: Spur the
Bureaucrats to Action,” in which he criticized the DPJ’s handling of
bureaucrats and laid out his principles for bureaucratic management
(Suga 2012). Suga notes that information asymmetries are a crucial
problem for the cabinet in dealing with the bureaucracy. For example,
he notes that bureaucrats often act like their hands are tied by local
governments or constituents when they are in fact pressuring localities
to adopt their preferred positions (Hoshi 2014, p.171).
The Abe government sought to overcome these information asym-
metries by expanding the independent capacity of the Cabinet Office,
cultivating close allies of the prime minister among the bureaucracy – in
no small part through control of personnel decisions – and appointing
former bureaucrats like Imai and experienced politicians like Suga to
key positions. For example, the 2013 appointment of Yuji Sato as
commandant of the Japan Coast Guard from within the ranks of the
agency broke a long-standing norm of appointments of career bureau-
crats from the Ministry of Transportation (Yamaguchi 2016, pp.100–
103). This signaled to the entire bureaucracy that their bureaucratic
turf was not secure, but the specific appointment came with a strong
18 Takeo Hoshi and Phillip Y. Lipscy

normative justification that was difficult to resist – that Japan Coast

Guard officers defending Japanese territory at the front lines should be
led by one of their own.

1.4.3 The Abe Model and COVID-19

Abe and his close advisors governed with a deliberate strategy that
sought to take advantage of Japan’s new electoral and administrative
institutions. These practices did not depend on any unique attributes of
Abe or his advisors and could thus plausibly serve as a model for future
prime ministers. Abe sought to maintain a high degree of public
approval by promoting popular economic reforms, timing unpopular
policies after elections to allow time to pivot back to popular issues,
and employing a media management strategy to limit critical news
coverage. Abe’s robust public approval ratings, combined with the
credible threat of calling early lower house elections in an electoral
system that emphasizes party label (e.g., Krauss & Nyblade 2005;
Reed, Scheiner & Thies 2012), diminished resistance from backbench-
ers within the LDP. Furthermore, the Abe government built on prior
administrative reforms by further consolidating authority in the
Cabinet Office. Practical mechanisms to coordinate policymaking
within and among the cabinet, the Diet, and the bureaucracy were
implemented to avoid the chaotic decision-making of prior administra-
tions, including Abe’s own.
Although the Abe model appeared to function effectively for much of
the prime minister’s tenure, the 2020 COVID-19 pandemic exposed
important shortcomings. In some respects, Japan’s response to
COVID-19 reflected strengths. The country initiated its response rela-
tively early, though the stringency of its measures lagged behind other
developed countries (Hale et al. 2020). Unlike populist leaders in the
United States and Brazil, the government relied heavily on scientific
experts like Shigeru Omi and Hitoshi Oshitani to orchestrate its
response. These experts emphasized measures that appear prescient in
retrospect, such as a cluster-based strategy through contact tracing,
elimination of high-risk situations for contagion – the so-called three
“Cs” of closed spaces, close-contact settings, and large crowds – and an
emphasis on cloth masks during a period when the World Health
Organization (WHO) and Western countries were discouraging their
use.9 Japan fared relatively well in terms of reported health outcomes:
The Political Economy of the Abe Government 19

Japanese COVID-19 cases and fatalities remained low in international

comparison, and there was no significant spike in excess deaths that
would suggest undercounting.
However, Abe’s response to COVID-19 was also characterized by
seeming indecision and hesitance. Shutting down the economy to
prevent contagion would reverse the macroeconomic gains of
Abenomics, which the prime minister saw as critical to his popularity
and legacy. The government faced constitutional restrictions that
foreclosed legally enforceable lockdowns. However, even accounting
for this, the government shied away from measures that could poten-
tially reduce economic activity, such as providing financial incentives
for businesses to stay closed. As his approval rating declined, Abe
attempted to stick to his standard playbook by doubling down on
Abenomics growth themes. This strategy may have backfired, as
initiatives to stimulate the economy, like the “Go To” travel cam-
paign to encourage domestic travel, cut against efforts to limit
COVID-19 contagion.
The Abe government also struggled to reach consensus on major
decisions, such as the scope of government stimulus and declaration
of a national emergency.10 Abe’s diminishing popularity under pan-
demic conditions reduced his leeway to threaten a snap election,
loosening his grip on his party and emboldening other powerbrokers
like Secretary General Toshihiro Nikai. The pandemic also elevated
the role of local governments in critical areas like testing and business
closures. Although Japanese prime ministers exercise strong authority
over the central government, local governments retain considerable
autonomy (Horiuchi 2010). This provided a platform for governors
like Yuriko Koike and Hirofumi Yoshimura to propose more strin-
gent measures and highlight shortcomings of the central govern-
ment’s response.11
Although Abe’s COVID-19 response did not create a health crisis, it
did create a serious public opinion problem for Abe. Most international
leaders saw their approval ratings spike as their countries responded to
the pandemic, but Abe’s approval ratings slid to the lowest levels of his
tenure.12 A majority of Japanese citizens disapproved of the govern-
ment’s COVID-19 response.13 Abe announced his resignation on
August 28, 2020, citing deteriorating personal health and apologizing
for “leaving various other policies still on the way to being realized and
the coronavirus epidemic at hand.”14
20 Takeo Hoshi and Phillip Y. Lipscy

1.4.4 Idiosyncrasies and Good Fortune

Although the Abe government took advantage of Japan’s political
institutions and governed with a calculated strategy to remain in
power and achieve policymaking success, it also benefited from several
idiosyncratic factors. While Abe and his close advisors responded to the
perceived failures of prior governments, it is less clear that they would
have been able to stay in power as long as they did without other
supportive factors (see Chapter 4). This suggests that while Abe’s
model of governance may be replicable, the same degree of legislative
productivity and longevity may prove elusive for his successors.
Abe’s tenure was characterized by opposition collapse, most import-
antly the disintegration of the DPJ. This gave Abe room to selectively
pursue policies unpopular with the general public, such as nuclear
restarts and the Antiterrorism Special Measures Law, without great
fear of electoral consequence. In fact, across most substantive issues,
the Abe government’s policy positions were less popular with the
public than those of most opposition parties (Horiuchi, Smith &
Yamamoto 2018). Opposition collapse gave Abe strong credibility in
calling lower house elections, in which the failure of opposition con-
solidation in single-member districts produced a large structural advan-
tage for the LDP. Although opposition failure reflects some structural
factors, such as the cross-cutting pressures of the lower and upper
house electoral systems, it is unlikely to remain a permanent feature
of Japanese politics.
Another factor that worked in Abe’s favor was the absence of major
international or domestic shocks that undermined Japan’s economy
during most of his tenure. Abe came to power in 2012, after the worst
consequences of the 2008 subprime crisis and 2011 Euro crisis had
subsided. Japan’s two largest trading partners, the United States and
China, were characterized by mostly stable, if somewhat sluggish,
growth from 2012 to 2019. Abe became the longest-serving prime
minister in Japanese history without having to face a major upheaval
like a financial crisis or natural disaster that seriously threatened
Japanese economic growth. The 2020 COVID-19 pandemic was argu-
ably the first serious test of his leadership.
Abe also benefited from the relative absence of major crises in foreign
policy. While the rise of China remains an important long-term chal-
lenge for Japan, Abe escaped serious diplomatic crises, such as the 2010
The Political Economy of the Abe Government 21

boat collision incident near the Senkaku/Diaoyu islands during the DPJ
government. North Korea’s aggressive development of nuclear weap-
ons and missile technology remained persistent challenges, but US
threats to unleash “fire and fury” against the regime in 2017 did not
materialize. The ascendance of US President Donald Trump, which
caused havoc in many Western capitals, was less disruptive for Japan:
Abe and Trump quickly developed a personal chemistry, and Japan
was credited with successful “Trump management” (Hikotani 2017).
Of course, these outcomes also reflect conscious policy choices by
Abe and his allies (Chapters 16 and 17). The opposition remains weak
in part because the LDP successfully painted the DPJ as solely respon-
sible for the Fukushima nuclear disaster, despite the fact that regulatory
deficiencies long predated the DPJ’s rule (Kushida 2014). Positive
external economic conditions do not guarantee strong economic per-
formance in Japan, as was seen most clearly in the 1990s. Under
a different prime minister, relations with President Trump may have
proven more difficult. Nonetheless, Abe clearly benefited from some
good fortune.

1.5 Assessing Abenomics

One of the purposes of this volume is to evaluate and assess the track
record of Abenomics. As we discussed at the beginning of this intro-
ductory chapter, Abenomics combined expansionary macroeconomic
policies (first two arrows) with structural reforms and industrial pro-
motion (third arrow). Four chapters of the volume (Chapters 6 through
9) discuss the macroeconomic policies of Abenomics, and six chapters
(Chapters 10 through 15) examine some key aspects of the economic
reforms. To provide a background to these chapters, this section briefly
examines how Abenomics was implemented and evolved over time.

1.5.1 First Arrow

As soon as the Abe administration started in December 2012, the
government began pressuring the BOJ to enhance monetary easing
more aggressively. Before the Abe administration started, major central
banks – including the Federal Reserve, the Bank of England, and the
European Central Bank – cut interest rates aggressively and embarked
on asset purchases at an unprecedented scale after the global financial
22 Takeo Hoshi and Phillip Y. Lipscy

crisis of 2007–2009. However, the BOJ was more tentative. It finally

did cut the policy rate to zero and started an asset purchase program of
35 trillion yen in “Comprehensive Monetary Easing” in October 2010,
but BOJ officials often downplayed the importance of monetary easing,
arguing that monetary expansion cannot solve deflation. In late
January 2013, within one month of the start of the Abe government,
the BOJ introduced the price stability target of 2 percent inflation rate
and open-ended asset purchasing until the price stability target was
reached.
After Haruhiko Kuroda replaced Masaaki Shirakawa as the new
BOJ governor in late March 2013, the BOJ introduced “Quantitative
and Qualitative Easing” (QQE) on April 4 and announced it would
double the amounts of monetary base, long-term JGB holdings, and
ETF holdings in two years. These measures were implemented to
achieve the target inflation rate of 2 percent within two years. The
BOJ declared that the policy “will support the positive movements that
have started to appear in economic activity and financial markets,
contribute to a further pick-up in inflation expectations that appear
to have risen, and lead Japan’s economy to overcome deflation that has
lasted for nearly 15 years” (Bank of Japan 2013, p.4).
The first arrow of Abenomics looked promising, at least initially.
Inflation expectations rose, the yen depreciated, and stock prices rose.
Careful economic analysis by Hausman and Wieland (2014) estimated
that the bold monetary policy under Abenomics should increase the
real growth rate by about 1.0 percent. The targeted inflation of 2 per-
cent within two years, however, remained elusive. When Hausman and
Wieland (2015) updated their analysis, they found that the VAT rate
hike in 2014 reduced consumption, and increased imports (the reason
being unclear) also hurt.
The BOJ continued the initial plan of QQE, except for an expansion
of its size in October 2014. Then, in February 2016, the BOJ introduced
the negative interest rate policy and started to apply a negative interest
rate (−0.1 percent) on a part of bank reserves. Many interest rates fell,
and commercial banks objected (as their interest margins were further
compressed), but the inflation rate did not increase very much.
In September 2016, the BOJ started to target the ten-year JGB yield
(initially at 0 percent) in addition to the negative short-term interest rate.
The BOJ also clarified the commitment to “inflation-overshooting,”
saying, “[The] Bank will continue expanding the monetary base until
The Political Economy of the Abe Government 23

the year-on-year rate of increase in the observed CPI (all items less fresh
food) exceeds the price stability target of 2 percent and stays above the
target in a stable manner” (BOJ 2016, p.3). When the VAT rate increase
from 8 percent to 10 percent in October 2019 was in sight, the BOJ
renewed its commitment to QQE, announcing that the “Bank intends to
maintain the current extremely low levels of short- and long-term
interest rates for an extended period of time, taking into account uncer-
tainties regarding economic activity and prices including the effects of
the consumption tax hike scheduled to take place in October 2019”
(BOJ 2018, p.1).
Even with the gradually escalated monetary expansion of the QQE,
the targeted inflation of 2 percent was not reached. Faced with the
COVID-19 crisis in early 2020, the BOJ turned to even more aggressive
monetary easing, much like other central banks in advanced econ-
omies. On May 22, BOJ Governor Kuroda and Finance Minister Aso
jointly announced they would do “whatever it takes” to cushion the ill
effects of the pandemic on the Japanese economy.15 With the sharp
economic downturn, achieving the inflation target of 2 percent became
seemingly impossible.
Why the super-expansionary monetary policy under Abenomics for
seven years was not successful in achieving the inflation target is a great
puzzle. The question is taken up in Chapters 6 and 7 of this volume.
One success of Abenomics monetary policy was depreciation of the yen
(or correction of over-valuation of the yen during the previous years).
Chapter 8 examines the implications on exchange rate changes on the
competitiveness of Japanese industries.

1.5.2 Second Arrow

The second arrow of Abenomics was called “flexible” fiscal policy,
which envisioned fiscal expansion when there is a demand shortage
while being prudent in the long run. Since the demand shortage was
considered to be a serious issue when Abenomics started, one of the first
policies of the Abe government was to put together a supplemental
budget that exceeded 10 billion yen as part of the announced
20 billion yen stimulus package of January 2013 (Ito & Hoshi 2020,
pp.232–235). Unlike monetary policy, the fiscal expansion of
Abenomics was not a departure from existing policy. Japan’s fiscal
policy had already turned expansionary right after the global financial
24 Takeo Hoshi and Phillip Y. Lipscy

crisis. The Aso government put together three stimulus packages with
a stated total size of more than 90 billion yen between October 2008 and
April 2009. The DPJ government between 2009 and 2012 also came up
with six stimulus packages with stated total amount of 80 billion yen.
The stimulus package of January 2013 was followed by more in
December 2013 (18.2 billion yen), December 2014 (3.5 billion yen),
and August 2016 (28.1 billion yen).
Although fiscal policy under the Abe administration was expansion-
ary, the Abe government also maintained the goal of achieving the
primary balance by fiscal year 2020, which was originally set in 2006
under the Koizumi government. To achieve that, the Abe administra-
tion planned to follow through with the two-step VAT increase that
was legislated in 2012 by the DPJ’s Noda government: a first tax hike in
April 2014 followed by the second in October 2015. The Abe govern-
ment indeed raised the VAT tax rate from 5 percent to 8 percent on
April 1, 2014, but the economic slowdown that ensued convinced the
government to postpone the second VAT increase to 10 percent until
April 2017 and then again to October 2019, when it was finally
implemented.
By 2018, the government formally abandoned the goal of eliminat-
ing primary deficits by fiscal 2020 and set fiscal 2025 as the new
deadline. Even the new delayed goal is not likely to be achieved.
According to the Cabinet Office (2020a), the simulation by the gov-
ernment just before the COVID-19 shock showed the primary bal-
ance for fiscal 2020 is −2.7 percent of GDP, which would improve
only to −1.3 percent by fiscal 2025 under the baseline scenario. Even
under the optimistic scenario where the nominal GDP growth rates
quickly rises to 3.5 percent by fiscal 2025 (as opposed to 1.3 percent in
the baseline case), the primary balance for fiscal 2025 would still
be −0.5 percent of GDP, and the primary balance would become
positive only in fiscal 2027.
Thus, the second arrow may have succeeded in stimulating the
aggregate demand by fiscal expansion, but the effort to establish fiscal
balance did not progress as quickly as planned. Nonetheless, the ana-
lysis in Chapter 9 shows that the fiscal situation does not look very bad.
Even though the goal of primary balance by 2020 was postponed,
Japan’s fiscal situation is still sustainable in the long run with a small
amount of tax increases if the growth of government expenditures does
not exceed that of GDP per capita.
The Political Economy of the Abe Government 25

Japan’s fiscal situation deteriorated quickly as the government imple-

mented spending measures to fight the effects of COVID-19. As of
May 2020, the government put together two supplementary budgets
financed by increased issues of government bonds. The planned bond
issues increased from 32.5 trillion yen (about 6 percent of GDP) in the
original budget for fiscal year 2020 to 90.2 trillion yen (about 16 per-
cent of GDP) after the two supplementary budgets. Based on this
additional spending, achieving primary balance even in fiscal 2027
became very difficult. According to the Cabinet Office (2020b), the
latest simulation shows that the primary balance would not be achieved
until 2029.

1.5.3 Third Arrow

Finally, let us examine the third arrow. We do this by reviewing
each year’s growth strategy adopted by the Abe government. All these
documents are available at the website of the Cabinet Office.16 Many
documents also have English versions.
The first growth strategy of the Abe administration was published
in June 2013 with the title “Japan Revitalization Strategy: Japan Is
Back.” It collected various policies/ideas on economic reforms and
interventions under three “plans” (industry revitalization plan, stra-
tegic market creation plan, and strategy of global outreach). From the
start, the growth strategy included both structural reforms likely to
enhance Japan’s potential growth and old-fashioned industrial target-
ing policies that may not be effective anymore. There is a huge litera-
ture on Japan’s industrial policy and its effectiveness, and we cannot
review that here.17 Although there is no clear consensus on the overall
effectiveness of the industrial targeting policy, many argue that the
policy became less effective after Japan’s catch-up phase of economic
growth ended. When Japan was following more advanced economies,
it was relatively easy for the Japanese government to identify promis-
ing industries by looking at the past experience of advanced econ-
omies. Now that Japan has become an advanced economy itself, it is
hard for the government (or anyone else) to know which industry will
become important in the near future.
The 2013 growth strategy was criticized for lacking focus. For
example, Pilling (2013) argued that “Mr. Abe’s 94-page plan, replete
with a 48-page road map and 37 subcomponents of industrial
26 Takeo Hoshi and Phillip Y. Lipscy

revitalization, leaves no policy stone unturned. Sadly, neither does it

present a grand vision.”18 It included some sensible goals such as
encouraging economic restructuring and reducing the cost of doing
business in Japan, but it also included various industrial targeting
ideas (to promote healthcare-related industry, clean energy, agricul-
ture, and so on). The 2014 revision of the growth strategy responded to
the criticism that the original lacked focus and identified the following
ten key reforms.
1 Enhancing corporate governance
2 Reforming investment of public and quasi-public funds
3 Accelerating industrial restructuring and venture businesses, pro-
moting provision of funds for growth
4 Corporate tax reform
5 Promotion of innovation and a robot revolution
6 Enhancing women’s participation and advancement
7 Enable flexible working practices
8 Attracting talent from overseas
9 Aggressive agricultural policy
10 Vitalizing the healthcare industry and providing high-quality
healthcare services
The list was still a mixture of sensible structural reforms (such as
accelerating industrial restructuring) and industrial targeting (agricul-
ture, healthcare, and robotics), but it was much more streamlined and
focused compared with the original growth strategy of 2013.
The 2015 revision of the growth strategy argued:
Abenomics has shifted from the stage where the focus was primarily placed
on solving the lack of demand with the aim of overcoming deflation to a new
“second stage” where steadfast policies are required to overcome the yoke of
supply constraints due to the decreasing population

and listed “realization of revolution in productivity by investment in

the future” and “promotion of local Abenomics” as the two central
policies. The latter policy was aimed at “reviving a vigorous Japan
where human resources and funds as well as technologies and infor-
mation to support them are flowing freely and actively all over and in
every corner of Japan.” The “second stage of Abenomics” was sup-
posed to “put Japan back onto a growth path to become a leading
nation in the world by promoting these two policies as the two wheels
The Political Economy of the Abe Government 27

of a cart.” Looking through more details, we find that the 2015

revision includes the ten focus areas of the 2014 revision but add
more (such as the need to respond to the fourth industrial revolution,
countermeasures to the low birth rate, and the rebuilding of tourism
as the driver of regional economies), which ends up somewhat blur-
ring the focus again.
The central part of “local Abenomics” was Machi, Hito, Shigoto
Sōgō Senryaku (Comprehensive Strategy on City, People, Jobs), which
aimed at creating more jobs and opportunities in local areas. As Hatta
(2018) points out, the policy emphasis was mostly on subsidy policies
that encourage the relocation of corporations and people from the
Tokyo metropolitan area to rural areas rather than structural reforms
and deregulations in such areas as agriculture, fishery, and tourism that
would allow the regional economies to grow.
After Shinzo Abe was re-elected as LDP president (and hence prime
minister) in the fall of 2015, he announced “new three arrows” as
Abenomics entered the second stage. The new three arrows were (1)
a strong economy that creates hope; (2) support for child raising that
fosters dreams; and (3) social security that gives citizens a sense of
reassurance. Each new arrow had a numerical target attached: nominal
GDP of 600 trillion yen for the first arrow; a fertility rate of 1.8 for
the second arrow; and a reduction to zero of the number of people who
quit working to take care of elderly parents or other family members
for the third arrow.
The relative emphasis on industrial targeting seems to have started
rising in the 2015 growth strategy. Although it continued to include
economic deregulation and structural reforms, it started to emphasize
the role of government in preparing for the challenges of the fourth
industrial revolution, for example. The leaning on industrial policy
became clearer in the 2016 revision of the growth strategy. The 2016
revision starts out by arguing that
in the First Stage of the Growth Strategy, the Abe administration has cut into
areas where the so-called “bed-rock” regulations remain, such as electricity,
agriculture and healthcare, which have been considered taboo

and continues to point out that the government also carried out
“reforms that have been thought to be impossible,” such as a drastic
corporate tax cut, TPP, and corporate governance reform. Then it
states: “On the contrary, private entities are not fully in action.”
28 Takeo Hoshi and Phillip Y. Lipscy

The 2016 growth strategy identifies three new missions: (1) strategic
cultivation of new “promising markets”; (2) “revolution in productiv-
ity” to overcome supply constraints and labor shortage coming from
the population decline; and (3) “enhancing capabilities of individuals”
to facilitate coming industrial transformation. These new missions are
expected to be carried out mostly by the government leadership.
Central to the endeavor was “10 strategic public–private joint pro-
jects” that (1) respond to the fourth industrial revolution (IoT, Big
Data, artificial intelligence), (2) achieve a world leading healthcare
country, (3) overcome environment-energy constraints and expand
investments, (4) change sports to a growth industry, (5) revitalize
markets for transaction of existing houses, (6) improve productivity
of service industry, (7) reinvigorate mid-ranking companies, SMEs and
microenterprises, (8) promote exports of agriculture, forestry and fish-
ery products, (9) become a tourism oriented country, and (10) stimu-
late domestic consumer sentiment. Although the 2016 growth strategy
still mentions national strategic special zones and other measures to
advance regulatory reforms, the emphasis shifted to numerous “pro-
jects” conducted under the leadership of the government.
The growth strategy revisions in 2017, 2018, and 2019 maintained
the emphasis on industrial policy. The 2017 revision, which was titled
“Investments for the Future Strategy 2017,” listed “five basic paths” to
achieve “Society 5.0.” The list included structural reform ideas such as
corporate governance reform, administrative reform, and regulatory
reform, but more attention was paid to various policy supports to
targeted industries including healthcare, automotive products, distri-
bution and production networks, economic and social infrastructure,
and fintech. The 2018 revision (“Future Investment Strategy 2018”)
also stressed “flagship projects” to achieve “Society 5.0,” which aimed
at building the next-generation mobility system; creating the next-
generation healthcare system; promoting innovations for energy
conservation and decarbonization; promoting fintech and a cashless
society, shifting toward digital government; building next-generation
infrastructure; introducing smart technologies to agriculture, forestry,
and fishery; building smart cities; and increasing productivities of
SMEs and micro businesses. These projects for Society 5.0 remained
central in the 2019 revision (“Action Plan of the Growth Strategy
2019”), though there was also renewed emphasis on social security
reforms for all generations and regional promotions.
The Political Economy of the Abe Government 29

The 2020 revision (“Action Plan of the Growth Strategy 2020”)

continues the emphasis on the projects for Society 5.0. The plan argues
that the COVID-19 crisis revealed many shortcomings for achieving
Society 5.0 and that the various projects have to be accelerated. 19
The (original) third arrow of Abenomics covered policies in many
areas, so the brief overview in this section leaves out many important
details. Chapters 10 through 15 of this book examine several policy
areas of the Abenomics growth strategy more carefully. In some areas
including corporate governance reform (Chapter 12) and womenomics
(Chapter 11), authors find the government pushed for structural
reforms with some success. Chapter 10 on labor market reform finds
that the government shifted the emphasis away from flexibility of the
labor market to improving worker welfare. This may not have been
growth-oriented but ended up benefiting workers. In agriculture
(Chapter 14), the new industrial policy was actually more market
oriented than the old industrial policy. In that sense, Abenomics
marked progress over previous policy, although the plan to provide
funds for modernizing agriculture through a public–private fund
(Agriculture, Forestry and Fisheries Fund Corporation for
Innovation, Value-Chain and Expansion Japan) did not succeed.20 In
innovation policy (Chapter 13), the policy mostly targeted what the
private sector actors would have achieved even without government
help anyway, but the government policy helped in legitimizing growth
of an innovation ecosystem. In energy policy (Chapter 15), the Abe
government made some progress in areas like deregulation of the
electricity retail market, but weak efforts on decarbonization and
continued emphasis on fossil fuels, particularly coal, invited sharp
criticism from the international community.

1.6 Chapter Summaries

We will close by providing an overview of the contributions to this
volume. Each chapter examines a substantial puzzle or evaluates
a policy of the Abe administration in a specific issue area. The volume
is divided thematically into the following sections: political context of
the Abe government; Abenomics macroeconomic policy; structural
reforms under the third arrow of Abenomics; and foreign policy.
Part II of the book examines the politics and policy formulation of
the Abe government: how did Abe govern, and what does it tell us
30 Takeo Hoshi and Phillip Y. Lipscy

about contemporary Japanese politics? In Chapter 2, “Expansion of

the Japanese Prime Minister’s Power and Transformation of Japanese
Politics,” Harukata Takenaka examines the growth of prime minister-
ial power, which began in the mid-1990s and accelerated under Abe.
Using a wealth of data, he illustrates how Abe expanded his authority
over both party backbenchers and bureaucrats, creating the founda-
tions for stable rule and the implementation of reforms.
One puzzle of the Abe government is why the prime minister and
LDP remained popular and succeeded in elections despite weak voter
support for the party’s policies. Two papers examine this question. In
Chapter 3, “Why Does the Abe Government’s Approval Rating Always
Recover?,” Masaru Kohno argues that Abe’s popularity tended to
recover from temporary dips because the dips usually reflected
a softening of support among his core conservative supporters, who
were inclined to revert to support after key events passed. In Chapter 4,
“The LDP under Abe,” Yukio Maeda and Steve Reed examine Abe’s
strategy toward elections during his tenure and consider how Abe
secured large electoral victories despite often adopting unpopular pol-
icy positions. They offer a cautionary note that Abe’s successors will
have a tougher time replicating his success as the opposition consoli-
dates and restless backbenchers demand greater say.
One intriguing puzzle about Abe is the apparent divergence between
his first tenure as prime minister in 2006–2007 and his more recent
tenure after 2012. Abe’s first government saw a rapid collapse in
approval ratings, a major upper house election defeat, and resignation
amid high-profile scandals. However, in Chapter 5, “The Third Arrow
of Abenomics: Est. in 2013 – or 2007?,” Takatoshi Ito finds surprising
continuity between Abe’s two administrations: the seeds of Abenomics
reforms under Abe II were planted during Abe I. Abe’s tenure in
2006–2007 not only provided lessons about governance: it also set
the stage for structural reforms after 2012.
Part III of the book shifts to an evaluation of Abenomics economics
policy. We begin by analyzing the so-called first and second arrows of
Abenomics: aggressive monetary easing and flexible fiscal policy.
Chapter 6, “Abenomics, Monetary Policy, and Consumption” by
Joshua Hausman, Takashi Unayama, and Johannes Wieland, asks
why consumption growth remained sluggish even though the first
arrow of Abenomics successfully stimulated real economic activities
and financial markets. The chapter finds that even the consumption of
The Political Economy of the Abe Government 31

those households who would have benefited most under Abenomics

remained sluggish. The chapter also offers potential explanations for
this puzzle. Chapter 7, “The Great Disconnect: The Decoupling of
Wage and Price Inflation in Japan” by Takeo Hoshi and Anil
Kashyap, looks at another puzzle concerning the first arrow of
Abenomics. They ask why nominal wage growth (and hence price
inflation) stagnated despite tight labor markets. They argue that the
increase in the proportion of part-time (low-wage) jobs and the lack of
response of wages for full-time (high-wage) jobs to the labor market
condition can explain the stagnant wage growth. One clear impact of
the first arrow was depreciation of the yen. Chapter 8, “Abenomics, the
Exchange Rate, and Markup Dynamics in Japanese Industries” by
Kyoji Fukao and Shuichiro Nishioka, examines how the depreciation
of the yen helped the competitiveness of Japanese industries. They do
this by estimating the changes of product markups over time. The key
question is whether Abenomics successfully stopped (and reversed) the
trend of declining product markups.
Mark T. Bamba and David E. Weinstein examine the second arrow
of Abenomics in Chapter 9, “The Crisis That Wasn’t: How Japan Has
Avoided a Bond Market Panic.” They evaluate the track record of the
Japanese government, including the Abe government, on both expan-
sionary fiscal policy and fiscal consolidation. One of the enduring
puzzles of the Japanese economy is why Japan has managed to escape
a bond market crisis despite having the largest public debt in the world.
Bamba and Weinstein find that, contrary to early predictions, the
Japanese government has acted responsibly by cutting spending and
raising taxes, overcoming supposed political impediments to fiscal
consolidation. As a result, Japan’s government finances can be sustain-
able with modest additional tax increases unless future expenditures
start rising rapidly.
In Part IV we turn to an assessment of the third arrow of Abenomics,
which focuses on structural reforms in a variety of sectors of the
Japanese economy. In Chapter 10, “Abe’s Slight Left Turn: How
a Labor Shortage Transformed Labor Politics and Policy,” Steven
K. Vogel examines the Abe government’s labor market policy and
identifies an important shift away from policies to facilitate labor
adjustments by corporations toward policies to promote better work-
ing conditions, such as work–life balance. He attributes this to demo-
graphic shifts in Japan that transformed a labor surplus into a labor
32 Takeo Hoshi and Phillip Y. Lipscy

shortage. One area of the third arrow emphasized by Abe from early on
was policies to enhance the participation and to increase the presence of
women in the labor market. Chapter 11 by Nobuko Nagase titled
“Abe’s Womanomics Policy: Did It Reduce the Gender Gap in
Management?” studies this womenomics policy in Abenomics. The
chapter focuses on the question of whether the policy achieved one of
the stated goals of increasing the number of women in management
positions. Using the data up to 2017, the chapter finds that, while the
probability that women become managers increased slightly, it
remained much smaller compared to that for men.
One area of reform that appeared to progress relatively quickly
under Abenomics is corporate governance. This is the subject of
Chapter 12, “Corporate Governance Reforms under Abenomics: The
Economic Consequences of Two Codes,” by Hideaki Miyajima and
Takuji Saito. They review the progress of corporate governance
reforms in Japan in the last two decades and examine how the process
accelerated under the Abe administration. The chapter finds that many
corporations changed the structure of corporate governance as seen in
the number of outside directors or the board committee structure, for
example, but it fails to find much impact on corporate behavior such as
investment rate and cash holdings. Another important objective of the
third arrow was encouraging innovation in Japan. Kenji Kushida docu-
ments changes in Japan’s ecosystem for innovation in his Chapter 13,
“Abenomics and Japan’s Entrepreneurship and Innovation: Is the
Third Arrow Pointed in the Right Direction for Global Competition
in the Silicon Valley Era?” The chapter finds that Japan’s startup
ecosystem improved remarkably under Abe’s government, but much
of the improvement would have happened without Abenomics. Japan’s
environment for entrepreneurship and innovation had already started
to take off even before Abe. The main role of Abenomics, if any, was to
provide legitimacy for entrepreneurship and innovation in Japanese
society.
As we pointed out earlier, it is puzzling how Abe pursued agricultural
reforms and the TPP, risking alienation of agricultural interests, a critical
part of the LDP’s traditional support base. Chapter 14, “Japanese
Agricultural Reform under Abenomics” by Patricia Maclachlan and
Kay Shimizu, tackles this puzzle. They argue that agricultural reform
for the Abe administration was analogous to postal reforms under
Koizumi. Abe sought to weaken resistance to agricultural reform by
The Political Economy of the Abe Government 33

undercutting the institutional foundations of Japan Agricultural

Cooperatives (JA), the conservative farm lobby. Although reform
momentum slowed after 2017, Abe succeeded in permanently kneecap-
ping the Japanese farm lobby.
Trevor Incerti and Phillip Lipscy examine energy policy under Abe in
Chapter 15, “The Politics of Energy and Climate Change under the Abe
Government.” They argue that Abe’s reforms in the energy sector
prioritized low energy prices in support of macroeconomic growth
and industrial policy interventions to support Japanese business inter-
ests. Although Abe publicly expressed support for international efforts
to combat climate change, Japanese policy during his tenure came
under sharp international criticism for lack of ambition and continued
support for traditional energy sources, particularly coal. The authors
argue that this reflected both institutional constraints and the prime
minister’s preferences, which prioritized economic growth over
decarbonization.
In Part V, the last two chapters of the book evaluate Abe’s foreign
policy in two broad areas: security policy and history issues. Adam Liff
takes up security policy in Chapter 16, “Japan’s Defense Reforms
under Abe: Assessing Institutional and Policy Change.” He examines
the evolution of Japanese security policy under the Abe administration
and concludes that they represent a significant shift toward centraliza-
tion of national security policy decision-making in the executive. This is
broadly consistent with the patterns we observe in domestic policy-
making. However, Liff also cautions that Abe’s reforms are largely
evolutionary in nature, building on long-standing trends in Japanese
security policymaking. Furthermore, important constraints on radical
reforms remain in place, such as fiscal constraints on military expan-
sion and domestic resistance to amendment of Article 9 in Japan’s
constitution.
In Chapter 17, “The Enduring Challenges of History Issues,” Mary
McCarthy considers the Abe government’s approach to historical
memory and reconciliation. McCarthy notes that the Abe government
sought to reinstate a sense of national pride in the Japanese nation and
move beyond narratives of Japan as a perpetrator or victimizer. Abe
achieved some notable successes in historical reconciliation, such as the
2016 visit to Hiroshima by US President Barack Obama – the first in
history by a sitting US president – and a reciprocal visit to Pearl Harbor.
However, the comfort women issue proved intractable despite a 2015
34 Takeo Hoshi and Phillip Y. Lipscy

agreement with South Korea that declared the issue “resolved finally
and irreversibly.” McCarthy attributes this to the democratization and
globalization of historical memory by various domestic, transnational,
and international actors, which has diminished the ability of govern-
ments to shape narratives.
Together, the chapters of the book paint the picture of a dynamic
government that took full advantage of political institutions that central-
ize power around the prime minister. Abe exercised this power effectively
in no small part due to his high and stable levels of popularity, which he
managed deliberately by focusing on popular economic reforms, sequen-
cing and timing unpopular policies, and appealing to rising Japanese
nationalism. The chapters that evaluate Abenomics reforms find mean-
ingful and significant progress across a wide range of issue areas that
overturn the conventional wisdom of Japan as a stagnant country mired in
policy paralysis. However, progress was not uniform, and the momentum
of reform appeared to slow down over time, shifting increasingly toward
questionable industrial policy interventions.

Notes
1. The large literature on the consequences of the 1994 electoral reform
includes Rosenbluth (1989), Sakakibara (1991), Ramseyer and
Rosenbluth (1993), McCubbins and Rosenbluth (1995), Cox,
Rosenbluth, and Thies (1999), Hirano (2006), Scheiner (2006),
Rosenbluth and Thies (2010), McElwain (2012), and Catalinac (2016).
2. NHK Yoron Chosa, www.nhk.or.jp/senkyo/shijiritsu/.
3. Abe as quoted in Oshita (2017), p.15 (our own translation).
4. For example, see journalistic accounts by Mori (2016) and Matsuda
(2016).
5. The information in this section is based on author interviews as well as
published interviews of government officials in Abe’s inner circle by
journalists.
6. These are the names listed by Abe himself as his closest personal allies.
See Oshita (2017), pp.11–12.
7. “Abe Naikaku Sori Daijin Shunin Kisha Kaiken,” 2012, www
.kantei.go.jp/jp/96_abe/statement/2012/1226kaiken.html; Tazaki (2014),
p.52; Oshita (2017), p.11.
8. “19,000-strong LDP supporters’ group debates how to knock down
election rivals,” Mainichi, October 18, 2017, https://mainichi.jp/eng
lish/articles/20171018/p2a/00m/0na/010000c.
The Political Economy of the Abe Government 35

9. Evan Dyer, “Some health experts questioning advice against wider use
of masks to slow spread of COVID-19,” March 31, 2020, www.cbc.ca
/news/politics/covid-19-pandemic-coronavirus-masks-1.5515526.
10. Daiki Nonaka, “Abe naikaku wa naze ‘ichiritsu 10man en wo ukeireta
no ka?” Toyo Keizai Online, April 23, 2020, https://toyokeizai.net
/articles/-/346133; “Kinkyu jitai sengen wa naze 4gatsu 7ka?” FNN
Prime Online, April 14, 2020, www.fnn.jp/articles/-/31768.
11. Eric Johnston, “Tokyo and Osaka governors vie for national attention,”
Japan Times, June 21, 2020.
12. “Covid-19 has given most world leaders a temporary rise in
popularity,” The Economist, May 9, 2020.
13. NHK Yoron Chosa, August 2020, www.nhk.or.jp/senkyo/shijiritsu/.
14. Shinzo Abe, Press Conference by the Prime Minister, August 28, 2020,
https://japan.kantei.go.jp/98_abe/statement/202008/_00004.html.
15. “Japan’s finance minister, central bank governor vow to use all tools to
combat virus fallout,” Reuters, May 22, 2020, www.reuters.com/article/
us-health-coronavirus-japan-economy/japans-finance-minister-central-
bank-governor-vow-to-use-all-tools-to-combat-virus-fallout-idUSKBN22
Y170, accessed May 30, 2020.
16. See www.kantei.go.jp/jp/singi/keizaisaisei/kettei.html.
17. For a brief introductory discussion, see Ito and Hoshi (2020,
pp.316–328).
18. See also Hoshi (2014).
19. “Social security reforms for all generations” are not so much on national
pension or other social security systems. The policy calls for continued
employment to age seventy, more mid-career hiring, and supports for
those workers who entered the job market during the “job ice age” (mid-
1990s to mid-2000s).
20. The fund, aka A-FIVE, was established in 2013, but accumulated huge loss
(more than 10 billion yen as of the end of 2019). In December 2019, the
government decided to stop investment and plans to liquidate A-FIVE.

References
Bank of Japan. 2013. Introduction of the “Quantitative and Qualitative
Monetary Easing.” Available at: www.boj.or.jp/en/announcements/rel
ease_2013/k130404a.pdf, accessed March 21, 2020.
Bank of Japan. 2016. New Framework for Strengthening Monetary Easing:
“Quantitative and Qualitative Monetary Easing with Yield Curve
Control.” Available at: www.boj.or.jp/en/announcements/releas
e_2016/k160921a.pdf, accessed March 21, 2020.
36 Takeo Hoshi and Phillip Y. Lipscy

Bank of Japan. 2018. Strengthening the Framework for Continuous Powerful

Monetary Easing. Available at: https://www.boj.or.jp/en/announce
ments/release_2018/k180731a.pdf, accessed March 21, 2020.
Cabinet Office. 2020a. Chūchōki no Keizai Saisei ni kansuru Shisan
(Simulation on Economic and Fiscal Conditions in the Medium to
Long Run), January.
Cabinet Office. 2020b. Chūchōki no Keizai Saisei ni kansuru Shisan
(Simulation on Economic and Fiscal Conditions in the Medium to
Long Run), July.
Carlson, Matthew M., and Steven R. Reed. 2018. Political Corruption and
Scandals in Japan. Ithaca, NY: Cornell University Press.
Catalinac, Amy. 2016. From Pork to Policy: The Rise of Programmatic
Campaigning in Japanese Elections. Journal of Politics 78(1): 1–18.
Cox, Gary, Frances McCall Rosenbluth, and Michael Thies. 1999. Electoral
Reform and the Fate of Factions: The Case of Japan’s LDP. British
Journal of Political Science 29: 33–56.
Estevez-Abe, Margarita. 2006. Japan’s Shift toward a Westminster System:
A Structural Analysis of the 2005 Lower House Election and Its
Aftermath. Asian Survey 46(4): 632–651.
Fackler, Martin. 2016. Abe Seiken Ni Hirefusu Nihon No Media. Tokyo
Japan: Futabasha.
Funabashi, Yoichi, and Koichi Nakano. 2017. The Democratic Party of
Japan in Power. New York: Routledge.
Hale, Thomas, Sam Webster, Anna Petherick, Toby Phillips, and Beatriz Kira.
2020. Oxford COVID-19 Government Response Tracker, Blavatnik
School of Government.
Hatta, Tatsuo, ed. 2018. Economic Challenges Facing Japan’s Regional
Areas. Singapore: Palgrave Pivot.
Hausman, Joshua, and Johannes Wieland. 2014. “Abenomics: Preliminary
Analysis and Outlook.” Brookings Papers on Economic Activity: Spring
2014: 1–63.
Hausman, Joshua, and Johannes Wieland. 2015. “Overcoming the Lost
Decades? Abenomics after Three Years.” Brookings Papers on Economic
Activity: Fall 2015: 385–413.
Hikotani, Takako. 2017. “Trump’s Gift to Japan: Time for Tokyo to Invest
in the Liberal Order.” Foreign Affairs (September/October): 21–27.
Hirano, Shigeo. 2006. Electoral Systems, Hometowns, and Favored
Minorities: Evidence from Japan’s Electoral Reforms. World Politics
59(1): 51–82.
Horiuchi, Yusaku. 2010. Understanding Japanese Politics from a Local
Perspective. International Political Science Review 30(5): 565–573.
The Political Economy of the Abe Government 37

Horiuchi, Yusaku, Daniel M. Smith, and Teppei Yamamoto. 2018.

Measuring Voters’ Multidimensional Policy Preferences with Conjoint
Analysis: Application to Japan’s 2014 Election. Political Analysis 26(2):
190–209.
Hoshi, Hiroshi. 2014. Kanbo Chokan: Sokkin No Seijigaku. Tokyo: Asahi
Shimbun Shuppansha.
Hoshi, Takeo. 2014. “Can Abenomics Restore Japan’s Growth?” Macro
Economy Proceedings 9.
Hosoya, Yuichi. 2019. FOIP 2.0: The Evolution of Japan’s Free and Open
Indo-Pacific Strategy. Asia-Pacific Review 26(1): 18–28.
Ito, Takatoshi, and Takeo Hoshi. 2020. The Japanese Economy, Second
Edition. Cambridge, MA: MIT Press.
Kabashima Ikuo, and Gill Steel. 2007. “The Koizumi Revolution.” PS:
Political Science and Politics 40(1): 79–84.
Kingston, Jeff. 2016. Press Freedom in Contemporary Japan. New York:
Taylor & Francis.
Kushida, Kenji E. 2014. The Fukushima Nuclear Disaster and the
Democratic Party of Japan. Japanese Political Economy 40(1): 29–68.
Kushida, Kenji E., and Phillip Y. Lipscy. 2013. Japan under the DPJ: The
Politics of Transition and Governance. Stanford, CA: Brookings
Institution/Shorenstein APARC.
Krauss, Ellis S., and Benjamin Nyblade. 2005. “‘Presidentialization’ in
Japan? The Prime Minister, Media and Elections in Japan.” British
Journal of Political Science 35: 357–368.
Liff, Adam P., and Ko Maeda. 2019. “Electoral Incentives, Policy
Compromise, and Coalition Durability: Japan’s LDP–Komeito
Government in a Mixed Electoral System.” Japanese Journal of
Political Science 20(1): 53–73.
Lipscy, Phillip Y. 2017. “DPJ Election Strategy: The Dilemma of Landslide
Victory.” In The Democratic Party of Japan in Power, edited by
Yoichi Funabashi and Koichi Nakano, pp.138–157. New York:
Routledge.
Lipscy, Phillip Y. 2020. “Reformist Status Quo Power: Japan’s Approach
toward International Organizations.” In The Crisis of Liberal
Internationalism: Japan and the World Order, edited by G.
John Ikenberry and Yoichi Funabashi. pp.107–132. Washington, DC:
Brookings Institution Press.
Lipscy, Phillip Y., and Ethan Scheiner. 2012. Japan under the DPJ: The
Paradox of Political Change without Policy Change. Journal of East
Asian Studies 12(3): 311–322.
Matsuda, Kenya. 2016. Kage No Kenryokusha: Naikaku Kanbo Chokan
Suga Yoshihide. Tokyo: Kodansha.
38 Takeo Hoshi and Phillip Y. Lipscy

McCubbins, Mathew, and Frances McCall Rosenbluth. 1995. “Party Provision

for Personal Politics: Dividing the Vote in Japan.” In Structure and Policy
in Japan and the United States, edited by Peter F. Cowhey and
Mathew McCubbins, pp.35–55. Cambridge: Cambridge University Press.
McElwain, Kenneth Mori. 2012. The Nationalization of Japanese Elections.
Journal of East Asian Studies 12(3): 323–350.
McElwain, Kenneth Mori. 2020. “The Perils and Virtues of Constitutional
Flexibility: Japan’s Constitution and the Liberal International Order.”
In The Crisis of Liberal Internationalism: Japan and the World Order,
edited by Yoichi Funabashi and G. John Ikenberry, pp.303–324.
Washington, DC: Brookings Institution Press.
Mori, Isao. 2016. Souri No Kage: Suga Yoshihide No Shotai. Tokyo:
Shogakukan.
Mulgan, Aurelia George. 2017. The Abe Administration and the Rise of the
Prime Ministerial Executive. London: Routledge.
Oros, Andrew. 2017. Japan’s Security Renaissance: New Policies and
Politics for the Twenty-First Century. New York: Columbia University
Press
Oshita, Eiji. 2017. Abe Kantei “Kenryoku” No Shotai. Tokyo:
Kadokawa.
Pilling, David. 2013. “It is the First Arrow of Abenomics that Matters.”
Financial Times, November 14, 2013.
Ramseyer, Mark, and Frances McCall Rosenbluth. 1993.
Japan’s Political Marketplace. Cambridge, MA: Harvard University
Press.
Reed, Steven R., Kenneth Mori McElwain, and Kay Shimizu. 2009. Political
Change in Japan: Electoral Behavior, Party Realignment, and the
Koizumi Reforms. Stanford, CA: Shorenstein APARC.
Reed, Steven R., Ethan Scheiner, and Michael Thies. 2012. “The End of LDP
Dominance and the Rise of Party-Oriented Politics in Japan.” Journal of
Japanese Studies. 38(2): 353–376.
Reporters Without Borders. 2019. 2019 World Press Freedom Index.
Available at: https://rsf.org/en/ranking/2019, accessed March 16,
2020.
Rosenbluth, Frances McCall. 1989. Financial Politics in Contemporary
Japan. Ithaca, NY: Cornell University Press.
Rosenbluth, Frances McCall, and Michael F. Thies. 2010. Japan
Transformed: Political Change and Economic Restructuring. Princeton,
NJ: Princeton University Press.
Sakakibara, Eisuke. 1991. “The Japanese Politico-Economic System and the
Public Sector.” In Parallel Politics: Economic Policymaking in the
The Political Economy of the Abe Government 39

United States and Japan, edited by Samuel Kernell, pp.50–80.

Washington, DC: Brookings Institution Press.
Scheiner, Ethan. 2006. Democracy without Competition in Japan:
Opposition Failure in a One-Party Dominant State. Cambridge:
Cambridge University Press.
Schoppa, Leonard J. 1991. “Zoku Power and LDP Power: A Case Study of
the Zoku Role in Education Policy.” Journal of Japanese Studies 17(1):
79–106.
Suga, Yoshihide. 2012. Seijika No Kakugo: Kanryo Wo Ugokase
(A Politician’s Resolution: Spur the Bureaucrats to Action). Tokyo:
Bungei Shunju.
Takenaka, Harukata. 2006. Shusho Shihai: Nihon Seiji No Henbo. Tokyo:
Chuko Shinsho.
Tazaki, Shiro. 2014. Abe Kantei No Shotai. Tokyo: Kodansha.
Yamaguchi, Noriyuki. 2016. Sori. Tokyo: Gentosha.
part ii

Political Context
2 Expansion of the Japanese Prime
Minister’s Power and Transformation
of Japanese Politics
harukata takenaka

2.1 Introduction
How has the power of the Japanese prime minister changed since the
1990s? This question has been a subject of intensive academic contro-
versy in recent years.1 Many argue that the Japanese prime minister has
increased in strength in terms of disciplining his party, has come to have
more discretion in making appointments, and now has a greater ability
to formulate and coordinate policies.
Most researchers ascribe the increase in the prime minister’s leader-
ship to the two political reforms carried out since the 1990s.2 These
were the political reform of 1994 and the reorganization of the govern-
ment agencies of 2001. The 1994 reform changed the electoral system
from a single non-transferable vote system to a system combining first-
past-the-post and proportional representation. It also imposed stricter
regulations on the collection of political funds while introducing
a public subsidy system for the political parties. The administrative
reform of 2001 expanded the authority of the prime minister, strength-
ened the Cabinet Secretariat, and created the Cabinet Office to support
the prime minister’s policy formulation.
Taking into account the contributions of past research, this chap-
ter analyzes comprehensively how the power of the Japanese prime
minister as an office in the Japanese parliamentary system has grown
from the 1990s through the reshuffling of the fourth Abe cabinet in
2018. Expansion of the prime minister’s power is a factor that
enabled Prime Minister Shinzo Abe to stay in power for a long time
and to remain in power, the issues raised in the Introduction to this
book.
I undertake this study because the current literature raises three
challenges if we are to more deeply understand the nature of the
Japanese prime minister’s power. The first challenge is empirical.

43
44 Harukata Takenaka

Most of the work on the growth of the Japanese prime minister’s power
derives its findings from political developments up through the
Koizumi administration, which was in power between 2001 and
2006. But we have witnessed new political developments since then.
Namely, Prime Minister Shinzo Abe kept his power for over seven years
after his comeback in 2012, and he carried out several institutional
reforms. Yet the implications for the nature of the Japanese prime
minister’s power over the political process under the second, third,
and fourth Abe cabinets has not been thoroughly examined.3
Some studies find that under the second Abe administration the
power of the prime minister expanded even further, as demonstrated
by an increase in the number of policy councils and the introduction
of new organizations such as the Cabinet Bureau of Personnel Affairs
and the National Security Council (NSC).4 Yet, several questions can
be addressed more extensively. The first question is whether the
prime minister’s power has expanded under Abe relative to other
administrations. The second is how the additional institutional
reforms carried out under the second Abe administration, such as
the introduction of the NSC of 2013 and the civil service reform of
2014, affected the prime minister’s power vis-à-vis other political
actors.
The second challenge is to more systematically analyze the prime
minister’s power. This consists of two different elements: the political
clout he holds in the ruling parties; and the set of powers he has within
the government. Most previous studies address only one of these. Even
those that take both into account have not located the position of the
Japanese prime minister as an office in a parliamentary system and have
not systematically examined how these changes affected the prime
minister’s power over other political actors and organizations in the
system from a single analytical perspective. To fully understand the
evolution of the Japanese prime minister’s power, it is necessary to
locate the Japanese prime minister as an office in a parliamentary
system and analyze how the changes in the two elements of his power
have transformed the prime minister’s relationship with other actors
and organizations in the Japanese parliamentary system. To locate the
position of the prime minister as an office in a parliamentary system,
this chapter provides an analytical perspective, relying on a principal–
agent model.
Prime Minister’s Power in Japanese Politics 45

The third challenge is that there remains plenty of room for numer-
ical data to show the increase in the prime minister’s power. Most
researchers rely on episodes from the Koizumi administration as evi-
dence for the expansion of the prime minister’s power.5 Certainly,
some studies refer to items beyond the period of the Koizumi adminis-
tration as evidence, such as the figures on ministerial appointments by
the prime minister, the enhanced role of the politicians and bureaucrats
close to the prime minister, and the increase in the number of councils
in the government.6 Yet it is possible to provide more diverse evidence.
This chapter offers more material, paying particular attention to the
role of the Cabinet Secretariat, one major supporting institution of the
prime minister.
So far, this chapter has stated its objectives. It proceeds as follows.
The next section presents an analytical perspective to examine the
evolution of the prime minister’s power in the Japanese parliamentary
system. The third section reviews how various institutional reforms led
to the expansion of the prime minister’s power up through the Koizumi
administration.
The fourth section explores the changes in the prime minister’s
power in recent years. It focuses on developments under the second
Abe administration. It shows how the prime minister’s power in mak-
ing ministerial appointments and formulating policies has expanded. It
then demonstrates that the administration implemented two more
reforms that expanded the role of the Cabinet Secretariat. It emphasizes
how the civil service reform of 2014 changed the prime minister’s
relationship with public officials. The last section addresses a possible
counterargument against the thesis of this chapter and discusses the
implications of these changes for the nature of the Japanese parliamen-
tary system.

2.2 Analytical Perspective

In understanding the increase in the Japanese prime minister’s
power, a study by Kaare Strøm on the nature of the parliamentary
system provides important insights.7 He argues that the parliamen-
tary system is characterized by a chain of delegation among different
sets of political actors. Each relationship can be perceived as one
between a principal and an agent. The chain of delegation goes from
the voters to the members of the assembly, then from legislators
46 Harukata Takenaka

Voters MPs PM Minister Civil Service

Figure 2.1 The chain of delegation in the parliamentary system

Source: Strøm, “Parliamentary Democracy and Delegation,” p.65

(MPs) to the prime minister (PM), from the prime minister to

the ministers, and lastly from the ministers to public officials
(Figure 2.1).
Strøm refers to two characteristics inherent in this delegation under
parliamentary democracies. The first is indirectness. For example, the
voters cannot control political offices, except the representatives.
The second is singularity. Each principal employs only one agent or
several noncompeting agents, and each agent reports only to one
principal.
He also argues that each principal is subject to agency problems, such
as adverse selection and moral hazard.8 Adverse selection occurs when
the principal chooses inappropriate agents. Moral hazard is the risk
that the agent will not act in accordance with the interests of the
principal. There are five ways to mitigate such problems: screening,
selection, contract design, monitoring, and institutional checks.9 The
first three are solutions ex ante, which a principal can use before hiring
an agent. In screening, the principal examines the quality of possible
candidates and chooses the appropriate one; in selection, agents offer
their credentials to the principal to be chosen. And the principal can
design a contract with the agent in such a way that the interests of the
two actors match.
The last two are mitigations ex post facto, which the principal can
use after hiring the agent. The principal can monitor the behavior of the
agent to see whether the agent respects the principal’s interests. Or, the
principal can create procedures through which the actions of the agent
can be rejected or corrected.
When we analyze the prime minister’s power, it is important to exam-
ine it vis-à-vis other actors. Theoretically, a cabinet is collegial, and the
prime minister cannot make decisions unilaterally. Yet, today, the prime
minister is at the center of policy formulation. He receives public atten-
tion and is often treated in the media as if he formulates policies.10 If the
cabinet is successful in formulating some policies, normally he receives
the credit. If the cabinet fails, he usually receives the most blame.
Prime Minister’s Power in Japanese Politics 47

Given such a political environment, it is sensible to assume that the

prime minister wants as much control and discretion as possible in the
process of policy formulation – in other words, that he wants to retain
independence from his principals (the MPs), while maximizing control
over his agents (the ministers).
This chapter pays particular attention to the prime minister’s rela-
tionships with the MPs as well as the ministers. It starts with his
relationship with the politicians in the parliament.
It is possible to assume that the MPs rely on both ex ante and ex post
mechanisms to control their agent, the prime minister. Projecting ex
ante control, they can choose as prime minister the person who is likely
to respect their interests. As means of ex post control, they can reject
legislation proposed by the prime minister or even take a vote of no
confidence to remove him/her.
Yet what makes the prime minister distinct from other actors is that
he has prerogatives to enhance his independence from his principals.
MPs are of two kinds: the members of the majority parties; and the
members of the opposition. What is important for the prime minister in
managing the government is to retain independence from the members
of the ruling parties.
In other words, in a parliamentary system, the cabinet led by the
prime minister prepares most policy and legislation. To enact them, he
needs the support of more than half of the MPs. He expects to obtain
this support from the ruling party. If his position is weak within
the ruling parties, he has to formulate policies in congruence with the
demands of the members of the ruling parties. In the meantime, if he
retains significant power among them, he can have more discretion in
formulating policy.
Of course, diverse factors affect the prime minister’s power within
the ruling parties. For example, the nature of the cabinet is important.
In general, a prime minister in a single-party government has more
power than one in a coalition government. Even under a single-party
government, which institution has the power to endorse candidates is
crucial.11 When the right of endorsement belongs to local organiza-
tions, it weakens the prime minister’s position against backbenchers.
When the party leadership retains this power, it increases the prime
minister’s leverage.
In the Japanese parliamentary system, there are two important
sources of power over backbenchers for the prime minister: the
48 Harukata Takenaka

authority to endorse candidates in his party; and the privilege to

distribute financial resources to the backbenchers.
Now I turn to the relationship between the prime minister and his
agents, the ministers. It is possible to assume that the prime minister
wants control over the ministers in formulating and implementing
policies. Among the five means he can use to avoid adverse selection
and moral hazard, the most effective are probably screening and
monitoring.
Control over his agents really comes from two different sources. One
derives from his political clout as the head of a ruling party. If he has
sufficient influence in his party, he can have more discretion over the
appointment of ministers. The second source relates to powers the
prime minister holds in the government. They originate from legal
rights, which the prime minister can rely on to formulate policies, and
from the organizations and offices in the government, which he can use.
Strøm takes it as a given that the prime minister entrusts much of
policy formulation to the ministers. In his model, the prime minister
does not formulate policies on his own. Strøm also emphasizes that
delegation is indirect and singular and that “principals cannot bypass
their agents.”12
Yet, the degree of delegation to and control over other actors is
a matter of institutional arrangements. It is probable that the prime
minister is provided with a set of organizations which enable him to be
less dependent on other actors and to have more control over policy
formulation. For example, the prime minister can formulate policy on
his own, using organizations and public officials who directly report to
him. This will reduce the cost of monitoring his agents. Or the prime
minister can delegate formulation of important policies to a limited
number of ministers. When many ministers are involved, the cost of
monitoring rises. If the prime minister limits the number of ministers,
he can reduce this cost. Further, it is possible to provide the prime
minister with legal rights to increase his command over the bureau-
cracy under the ministers, although Strøm assumes that the bureau-
cracy is beyond the direct control of the prime minister.
The issue here is how to set the range of organizations and offices
contributing to the prime minister’s power. For example, he can ask
his ministers to formulate policies. This implies that the prime minis-
ter can indirectly make use of the organizations under the ministers.
Yet, expanding the relevant organizations to include ones the prime
Prime Minister’s Power in Japanese Politics 49

minister can turn to indirectly makes it difficult to determine the

range of organizations available to the prime minister. Thus, it is
necessary to set some criteria. This chapter relies on legal stipula-
tions – that is, it includes the organizations and offices which have
clear legal authority to assist the prime minister’s policy formulation
and coordination.
In the case of the Japanese prime minister, Article 26 of the Cabinet
Law puts the prime minister in charge of the Cabinet Secretariat.
Likewise, Article 6 of the Cabinet Office Law makes the prime
minister the head of the Cabinet Office, which was established in
2001. Article 2 of the former Prime Minister’s Office Law also made
the prime minister the head of the Prime Minister’s Office, which
existed until 2001.
In the light of these provisions, we consider that the Cabinet
Secretariat, the Prime Minister’s Office, and the Cabinet Office, as
well as political offices and civil servants within these organizations,
comprise the prime minister’s power. The important political offices in
these organizations are the chief cabinet secretary and the deputy chief
cabinet secretaries, as well as Cabinet Office ministers.

2.3 Evolution of the Prime Minister’s Power

2.3.1 The Prime Minister under the “1955 System”

With this analytical perspective, it becomes possible to understand how
the prime minister’s power expanded in three sequences: the 1994
political reform; the 2001 administrative reform; and the 2014 civil
service reform. Most changes contributed to enhancing the prime
minister’s power over his agents through screening, a means of ex
ante control, and monitoring, a mechanism of ex post control. The
1994 reform expanded his influence over the backbenchers of the
ruling parties, his principals, because it became more difficult for
them to rely on vetoes to rebel against legislations proposed by the
prime minister, restraining him as an ex post mean of control. It also
enhanced his influence over his ministers, his agents, as it had become
possible for him to exercise more effective screening. The 2001 reform
strengthened the Cabinet Secretariat, created the Cabinet Office, and
introduced new political offices, the Cabinet Office ministers. This
enabled the prime minister to delegate policy formulation to the
50 Harukata Takenaka

Cabinet Office ministers and these organizations. And this has lowered
the cost of monitoring his agents, as he can now delegate important
policies to a smaller number of ministers and organizations. The 2014
reform also expanded the power of the prime minister over the senior
officials of the ministries through screening.
To understand the evolution of the Japanese prime minister’s power,
it is necessary to demonstrate how his power was curtailed before the
reforms under the “1955 system” – a term used to characterize the
party system as well as how the political process evolved between 1955
and 1993 in Japan.
Under the 1955 system, the prime minister could not have much
independence from his principals, the backbenchers. The backbenchers
kept a firm grip on the prime minister, so that the prime minister would
not formulate policies violating their interests, as shown by the
thick arrow from the MPs to the PM in Figure 2.2. In the meantime,
the prime minister could not have strong control over his agents, the
ministers, as demonstrated by the thinner arrow from the PM to the
ministers. This was because the electoral system made the power of
endorsement held by the party leadership insignificant, making the
government controlled by the conservative Liberal Democratic Party
(LDP) closer to a coalition government.
There were two reasons why the power of endorsement did not
contribute to the prime minister’s power. Between 1947 and 1993,
Japan used a single nontransferable vote system for the lower house,
in which several politicians were elected from each district. For the LDP
to obtain majority in the lower house, it had to run several candidates
in each district. Therefore, the LDP candidates often competed among
themselves. Endorsement by the LDP did not contribute to making
distinctions among the LDP candidates. Further, since it was relatively
easy to be elected as an independent, the LDP politicians did not have to
be concerned about a possible withdrawal of their candidacy by the
party leadership. Thus, the prime minister could not use the threat of
withdrawing endorsement to discipline party politicians to secure sup-
port for his policies.

Voters MPs PM Minister Civil Service

Figure 2.2 The “1955 system”

Prime Minister’s Power in Japanese Politics 51

The electoral system also gave the factions in the LDP strong auton-
omy. To increase the size of their faction, faction leaders recruited new
politicians and had them run in different districts. Even if some factions
should rebel against the prime minister, expelling those who rebelled
from the LDP could not be a real threat because they had a good chance
of being elected as independents. Thus, it was difficult for the prime
minister to implement policies opposed by a faction.
As principals, the LDP backbenchers could rely on vetoes as valuable
ex post mechanisms to restrain the prime minister. After selecting their
agent, they could make their agent, the prime minister, respect their
interests because they could veto his policies without much fear of
penalty.
In the meantime, as a principal, the prime minister could not rely on
screening as an effective tool of ex ante control over the ministers.
Between 1955 and 1993 the LDP controlled both chambers of the Diet
and formed a single-party government. But in practice the LDP govern-
ment was closer to a coalition government because of the factions’ strong
autonomy. Each faction was virtually an independent political party.
The prime minister did not have much discretion in choosing minis-
ters. To gain support from the factions he usually assigned them
ministerial portfolios according to their relative size. In appointing
ministers, he also had to respect recommendations from the factions.
He therefore had to delegate policy formulation to ministers he had not
really chosen. And he did not have much power to discipline them,
because the ministers (except those from his own faction) remained
more loyal to the faction bosses than to the prime minister.
To make things worse, the prime minister’s legal authority was
limited. He had the right to convene cabinet meetings but not the
power to propose policies. It was the ministers who proposed policies.
The prime minister had two supporting organizations, the Cabinet
Secretariat and the Prime Minister’s Office. But these organizations
were weak. A major responsibility of the Cabinet Secretariat was the
“coordination” of policies, which involved various ministries. One
major mission of the Prime Minister’s Office was also the “coordin-
ation of policies and businesses” of different ministries. The difference
between the two offices with regard to coordination was not clear. In
practice, the sections responsible for coordination in the two organiza-
tions were integrated, and the same public officials were concurrently
assigned to the two sections.
52 Harukata Takenaka

Yet, since his control over his ministers was limited, the prime
minister could not effectively use these organizations to coordinate
policies. Thus, the ministries remained very autonomous.

2.3.2 Growth of Prime Ministerial Power

The 1955 system started to change with the 1994 reform, which
replaced the single nontransferable vote system with a system combin-
ing plurality and proportional representation. This transformed the
relationship between the prime minister and the backbenchers as well
as the factions. As principals, the backbenchers’ control over the prime
minister became weaker, as shown by the thinner arrow from the MPs
to the PM in Figure 2.3. It became more difficult for the backbenchers
to rely on the veto of cabinet legislation as an ex post means of
restraining the prime minister. The reform also changed the relation-
ship between the prime minister and the ministers. The prime minister
came to have a tighter control over the ministers by using an ex ante
means of control – screening – more effectively, as shown by the thicker
arrow from the PM to the ministers.
The 1994 reform made it no longer possible for political parties to
run several candidates in a single district. As a result, the relationship
between the prime minister and his principals, the backbenchers,
changed, because it was more difficult to be elected as an independent.
For backbenchers, it became vital to obtain endorsement from their
party. And as the president of the ruling party, the prime minister came
to have significant influence over who should receive endorsement.
Likewise, it was now impossible for the factions to freely let their
members run for election. Now they had to ask the LDP president–
who, after 1994, was usually the prime minister – to endorse their
members as candidates.
The reform also introduced new rules on political funds, making it
much harder for the backbenchers and the factions to collect political
funds. Political funds came to be concentrated in the political parties.

Voters MPs PM Minister Civil Service

Figure 2.3 After the 1994 reform

Prime Minister’s Power in Japanese Politics 53

Before, it was possible for the backbenchers and the factions to collect
political contributions from firms. Since the reform (with a transitional
period), firms can now make contributions only to the political parties.
At the same time, it introduced a public subsidy system for the political
parties. The backbenchers have become more dependent on the party
leadership for financial resources. As the prime minister sits at the top
of the party leadership, it is fair to assume that he can influence the
distribution of political funds to the backbenchers.
Thus, the backbenchers’ ex post means of restraining the prime minister
has become less effective. The cost of rebelling against the prime minister
has increased enormously. As a result, the prime minister has more power
in the ruling party and more independence from his principals.
Now, let us turn to the prime minister’s power as principal over the
ministers. First, the prime minister’s power in making ministerial
appointments has expanded. It has become possible for the prime
minister to rely on screening as a more effective way of ex ante control
over his agents. This reflects the decline of political clout of the factions.
After the reform, there was a gradual erosion of the custom of
distributing ministerial portfolios among the factions in proportion to
their size. This became apparent when Keizo Obuchi, then the foreign
minister, became the prime minister in 1998. When Prime Minister
Obuchi formed his cabinet, he succeeded in appointing about half of
the ministers. The increase in the prime minister’s power over these
appointments became conspicuous when Koizumi Jun’ichiro, who had
served as minister of welfare as well as minister of postal services in past
administrations, became prime minister in 2001. He appointed all the
ministers at his discretion, except the positions saved for the LDP
politicians of the House of Councillors.
The administrative reform of 2001 expanded the prime minister’s
power in policy formulation in terms of legal responsibilities. In add-
ition, a major organization supporting the prime minister’s policy
formulation, the Cabinet Secretariat, was assigned new authority to
design policies and draft legislation. Further, a new organization and
offices, the Cabinet Office and its ministers, were introduced to assist
the prime minister.
With this reform, the prime minister could use monitoring as
a more efficient mechanism of ex post control because the cost of
monitoring was reduced. He could delegate formulation of important
policies to a smaller number of politicians and organizations.
54 Harukata Takenaka

Voters MPs PM Minister Civil Service

Cabinet Secretariat/
Cabinet Office

Figure 2.4 After the 2001 reform

Through this delegation, he, as a principal, could now more closely

monitor his agents: the ministers, politicians, and public officials
under his direct control. This new means of control is illustrated by
the heavy arrow from the PM to the Cabinet Secretariat and the
Cabinet Office in Figure 2.4.
The Cabinet Law was revised in 1999 to give the prime minister the
formal right to propose policies at cabinet meetings. This gave the prime
minister the authority to formulate policies in which he was interested.
The reform of 2001 also gave the Cabinet Secretariat the power to
initiate policies and prepare legislation. Before the reform, as it was only
permitted to coordinate among ministries, the preparation of a bill by the
Cabinet Secretariat was rare. It also introduced the Cabinet Office and its
ministers, to support the prime minister. Note that the Cabinet Office
ministers are under direct control of the prime minister. Under the
Cabinet Office, bodies such as the Council on Economic and Fiscal
Policy and the Council on Science and Technologies were established to
design policies.
The effect of this reform became clear under the Koizumi cabinet.13
Prime Minister Koizumi made full use of the Council on Economic and
Fiscal Policy to design key economic policies, such as the postal service
privatization. He delegated much of the management of this council to
the Cabinet Office minister for fiscal and economic policy. He also
relied on the Cabinet Secretariat to draft important bills, such as the
bills to deploy the Self-Defense Forces in 2001 to assist multinational
forces fighting the “war on terror” and to Iraq after the Iraq War.14 He
also used the Cabinet Secretariat to prepare legislation for the postal
service privatization, a policy he considered very important.
With the 2001 reform, it is fair to say that the Japanese political
system had become distinct from the 1955 system. The result I will refer
to as the “2001 system” to emphasize this difference. Its distinctive
Prime Minister’s Power in Japanese Politics 55

features are the cabinet with the nature of a single-party government,

a strong prime minister, and an electoral system combining first-past-
the-post and proportional representation.

2.4 The Power of the Prime Minister and the Second

Abe Administration

2.4.1 Further Growth of Prime Ministerial Power

After the Koizumi cabinet, the power of the Japanese prime minister
grew even more in several aspects. First, the prime minister expanded
the power of appointment. The power of appointment became even
stronger under the second Abe administration, which had started in
2012. As a principal, Prime Minister Abe has more discretion in
choosing ministers than former prime ministers did. The expansion
of the prime minister’s power over advancement is demonstrated by
the long tenure of the ministers. All the ministers who were appointed
when the second Abe cabinet was formed in December 2012 served
until September 2014. This is unprecedented in postwar Japanese
history. And until the reshuffling of the third Abe cabinet, in
August 2017, the three most important ministers – the chief cabinet
secretary, the finance minister, and the foreign minister – had not
changed from the inauguration of the cabinet. This is also epoch-
making.
The flip side of these phenomena is that it has become more difficult
for politicians to become ministers, as demonstrated by Nakakita.
Previously, most LDP politicians who had been elected at least six
times to the lower house were appointed ministers at least once.
Nakakita calculates the ratio of LDP politicians who did not become
ministers despite having been elected more than at least six times to the
total number of LDP politicians who were elected more than at least six
times. He computes this figure for every election year and finds that,
while it remained below 15 percent in the past, it had risen to 27.5 per-
cent before the 2014 election.15
Following Nakakita’s lead, I collected data on the appointments for
all cabinets after 2001. I calculated the ratio of politicians who had
been elected more than at least six times and who had become ministers
to the total number of LDP politicians who were elected more than at
least six times. The result indicates the expansion of power in making
56 Harukata Takenaka

ministerial appointments under the second Abe administration.

Initially, the number was more than 90 percent: almost all LDP politi-
cians could become ministers after having been elected six times. But it
gradually declined, and under the Abe administration it has fluctuated
between 75 percent and 60 percent (Figure 2.5).
In addition, the prime minister’s capacity to formulate policies in
the government kept growing after the 2001 reform. Since the reform
there has been a continuous expansion in the number of civil servants
in the Cabinet Secretariat (Figure 2.6). In 2000, before the reform,
there were 822. In 2006, under the Koizumi administration, the
number reached 1,438. Under the Abe administration it rapidly
expanded, from 2,331 in 2012 to 2,971 in 2018. This is more than
three times the figure before the reform and twice the figure under the
Koizumi administration.
Further, the number of sections under the cabinet affairs officer and
the assistant deputy chief cabinet secretaries, which are important
offices in the Cabinet Secretariat, has multiplied from ten in 2001 to
thirty-nine in 2017 (Figure 2.7). These sections formulate and coordin-
ate policies to which the cabinet attaches importance. This indicates the
expansion of the role of the Cabinet Secretariat in policy formulation.

More than Six With Ministerial Experience

Number of LDP politicans elected more than six times

120 100%
90%
100
80%
70%
80
60%
60 50%
40%
40
30%
20%
20
10%
0 0%
ku kud s
iz iz s
d d iR i

es

e dR s
d s
Ab th es
t K t K riR ri

iz iz s
i i
1s 1s Res
be be

da a

3r Ab rd es
3r 3r um um
oi oi es
d d iR i

d d o
d 3 eRe

3r Ab 1s be

1s e
es
um um

Fu Fu Re
Ko Ko e

Ab n e
3r e
2n 2n umzum

Ko Ko e
1s 1s Mo Mo

2n 2n As
A b Ab

e Ab
R

d e2 tR

R
tA tA

tR
d e A
d d
2n 2n

4
z

h
3r

4t

Figure 2.5 Politicians with ministerial experience

Prime Minister’s Power in Japanese Politics 57

Official quotas Concurrent officials Total

3500

3000

2500

2000

1500

1000

500

0
00
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
18
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
Figure 2.6 Public officials in the Cabinet Secretariat

45
40
35
30
25
20
15
10
5
0
oi oiz s
2n 2n um mi
Ko o es
3r um mi

iz s
1s mi
Fu be
da
oy so
a
od nd n

a tR s
d s
2n n Res

e e
3r b 3rd es
3r be st e
Ab nd s

d s
4t es

e
t K t K Re

Ko e

od 1s e
2n e

d 2 Re

3r e
am
N 2 Ka

Ab b

A 1 b

Ab
z u
d dK iR
iz izu
d iR
u

at A

N a R

e R
R
tA
ku

d dA

d e A
1s 1s ori

h
M

2
H
d

n
2n

Ka

A
d
3r

Figure 2.7 Expansion of special policy units in the Cabinet Secretariat (under
the director general of cabinet affairs and the assistant deputy chief cabinet
secretary)
58 Harukata Takenaka

Official quotas Concurrent officials Total

3500

3000

2500

2000

1500

1000

500

0
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

Figure 2.8 Officials in the Cabinet Office

The organization of the Cabinet Office enlarged as well. The number

of public servants working in the office increased from 2,412 in 2001 to
3,165 in 2017 (Figure 2.8). The businesses assigned to the Cabinet
Office grew from 76 in 2001 to 100 in 2016.
The number of committees chaired by the prime minister also grew
dramatically, from four in 2001 to twenty-three in 2018 (Figure 2.9).
Note that the Cabinet Secretariat serves as the secretariat for many of
these committees (thirteen of them, as of 2018).
Above all, the number of bills drafted by the Cabinet Secretariat
soared (Figure 2.10). Before the reform, the Cabinet Secretariat seldom
drafted bills – just eight of them between 1990 and 2000. Since the
reform the Cabinet Secretariat prepares bills every year; it drafted more
than ten bills in 2005 as well as in 2012.

2.4.2 More Reforms

The second Abe administration introduced a third reform. This was the
civil service reform of 2014, which created the Cabinet Bureau of
Personnel Affairs. The reform contributed to extending the prime
Prime Minister’s Power in Japanese Politics 59

Councils under the Cabinet

Councils which the Cabinet Secretariat serves as the secretariat

25

20

15

10

0
99
00

18
01
02
03
04
05
06
07
08
09
10
11
12
13
14
15
16
17
19
20

20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
20
Figure 2.9 Councils set up under the cabinet

16

14

12

10

0
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018

Figure 2.10 Bills submitted by the Cabinet Secretariat

60 Harukata Takenaka

minister’s influence over the bureaucracy, in particular over high-

ranking officials in the ministries. It enabled the prime minister to rely
more on screening as an ex ante means of control over the bureaucracy.
Before the reform, the ministries had substantial autonomy in
appointing high-ranking officials. Ministers could design personnel
policies on their own. What they needed to do was to seek approval
from the prime minister, chief cabinet secretary, and deputy chief
cabinet secretaries for these appointments. The number of positions
for which such approval was necessary was limited to 200 high-ranking
positions above the director-general level in all ministries.
Now, the Bureau of Personnel Affairs of the Cabinet Secretariat
makes a list of about 600 officials as candidates for the positions
above deputy director-general level in the ministries. And each minister
must consult with the prime minister and chief cabinet secretary before
making appointments for those positions. The reform has made public
officials more loyal to the prime minister. In other words, the prime
minister can now bypass the ministers, who are the direct principals of
the public officials, and can project more control over the public offi-
cials through screening. This is illustrated by the thick arrow from the
PM to the civil service in Figure 2.11.
In 2013, the second Abe administration further expanded the policy
formulation capacity of the prime minister by introducing the NSC.16
This enhanced the prime minister’s ability to integrate foreign and
security policy and increased his control over the foreign and defense
ministers. The 2001 reform gave the prime minister more effective
coordination over domestic policies. Yet, the coordination of security
policies between the Ministry of Foreign Affairs and the Ministry of
Defense was not as efficient. The NSC addressed this weakness of the
secretariat.

Voters MPs PM Minister Civil Service

Cabinet Secretariat/
Cabinet Office

Figure 2.11 After the 2014 reform

Prime Minister’s Power in Japanese Politics 61

The NSC is the committee which has become responsible for coord-
inating foreign policy and security policy under the leadership of the
prime minister. Before the NSC, there was a similar organization called
the Security Council. But it focused on defense policy.
The second Abe cabinet succeeded in passing the bill to establish the
NSC in 2013. The most important committee of the NSC is the Four
Ministers Meeting, in which the prime minister, the chief cabinet
secretary, the foreign minister, and the defense minister discuss
“Japanese security policy which involves foreign policy and defense
policy” several times a month.17 This meeting has become an effective
venue in which to formulate and coordinate security policy.
The second Abe cabinet also created the National Security Secretariat
(NSS) as the secretariat of the NSC within the Cabinet Secretariat. This
new organization has made more effective coordination of security
policy in the Cabinet Secretariat possible. Now the prime minister can
more closely monitor the formulation of security policy by the foreign
minister and the defense minister, as well as public officials in the NSS.
Prime Minister Abe succeeded in introducing the NSC on his second
attempt. He made the first try in his first administration. When he ran
for LDP president in September 2006, he proposed to strengthen the
coordination capabilities of the organizations supporting the prime
minister in foreign and security policy by establishing an organization
similar to the National Security Council of the United States. He
intended to make communication between the governments of Japan
and the United States on security policy more effective.18 The first Abe
cabinet drafted a bill to introduce the NSC in 2007. The responsibilities
and organization of the NSC of 2007 were very similar to those of the
NSC established in 2013.
Traditionally, some ministries, in particular the Ministry of Foreign
Affairs, opposed increases in the power of the Cabinet Secretariat in the
field of external policy. It was concerned that the Cabinet Secretariat
would take power from the ministry. By the time the first Abe cabinet
tried to introduce the NSC, in 2007, the power of the prime minister
had grown, and the Ministry of Foreign Affairs did not resist strongly.
The first Abe cabinet nonetheless met resistance from bureaucrats over
the design of the new organization. The issue was the eligibility of the
position of the head of the secretariat. The NSC of 2007 was to have
a secretariat independent of the Cabinet Secretariat, whereas the cur-
rent secretariat, the NSS, is a part of the Cabinet Secretariat.
62 Harukata Takenaka

In 2006, Prime Minister Abe set up a special council to discuss how

to design the NSC. The council submitted a report in February 2007,
which proposed to introduce the permanent position of national secur-
ity advisor to the prime minister and an independent secretariat for the
NSC. The national security advisor could concurrently serve as the
head of the secretariat and that the prime minister could appoint
a Diet member as his advisor. This meant that it would be possible
for a politician to lead the secretariat.
The officials in the Cabinet Secretariat were concerned that a security
advisor leading an independent secretariat of the NSC might compete
with the chief cabinet secretary.19 Such a situation might undermine the
power of the Cabinet Secretariat as well. In the process of drafting the
legislation for the NSC, the officials in the Cabinet Secretariat weak-
ened the position of the security advisor by leaving it to the prime
minister to decide whether to fill the position and restricted eligibility
for the position of head of the secretariat to bureaucrats. When Prime
Minister Abe introduced the NSC in 2013, the eligibility remained the
same as in the NSC proposed in 2007. More importantly, the NSS
became a part of the Cabinet Secretariat and was put under the com-
mand of the chief cabinet secretary. This design is consistent with the
interests of the officials in the Cabinet Secretariat.

2.4.3 Prime Minister Abe’s Reliance on the Cabinet Secretariat

The second Abe administration has made full use of the Cabinet
Secretariat to design important policies. It is possible to see a general
pattern in how the Abe administration formulates policies. When
Prime Minister Abe initiated a policy to which he attached political
significance, he appoints a minister responsible for the policy and set up
a special division in charge in the Cabinet Secretariat. In addition, he
often created a conference body to discuss the substance of the policy,
with that special division as the secretariat.
At the onset of his second administration, Prime Minister Abe made
a commitment to three economic policies, called the “three arrows”:
“bold monetary policy,” “flexible fiscal policy,” and “growth strategy
to promote private-sector investment.” To design the third arrow, he
relied on the Cabinet Secretariat. He appointed Amari Akira minister
for economic revitalization, responsible for the growth strategy. He
also set up the Secretariat for the Revitalization of the Japanese
Prime Minister’s Power in Japanese Politics 63

Economy (SRJE) in the Cabinet Secretariat as a special section to design

a set of policies for the growth strategy. For concrete policy deliber-
ation, he created the Committee on Industrial Competitiveness. The
SRJE served as its secretariat. The committee discussed a series of
microeconomic policies and proposed the Strategy to Revitalize Japan
2013, the Strategy to Revitalize Japan 2014, and the Strategy to
Revitalize Japan 2015, which included various important economic
policies, such as corporate tax reduction and electricity market
liberalization.
From the beginning of 2016, the second Abe administration put
more weight on the formulation of social welfare and labor policies.
One of those policies was the Human Resources Development
Revolution, which Prime Minister Abe launched in June 2017 to
boost spending on education and human resources development. In
August 2017 he appointed Mogi Toshimitsu minister in charge of the
Human Resources Development Revolution and set up a special div-
ision called the 100 Year Life Time Design Office as the section respon-
sible for formulating concrete policies. In September he also set up the
100 Year Life Time Design Council, with that office as its secretariat.
The office and the committee designed the details of policies such as the
complete liberalization of childcare and education for children three to
five years old.
The second Abe administration adopted the same style to formulate
other important policies, such as Trans-Pacific Partnership negoti-
ations and the policy to allow the government to exercise the right of
collective defense under certain conditions.

2.5 Toward a Westminster Model?

This chapter has made a systematic analysis of the changes in the
Japanese prime minister’s power and demonstrated how it has
expanded since the 1990s. It has offered an analytical perspective
with a principal–agent model to examine the evolution of the
Japanese prime minister’s power, paying attention to the nature of
the parliamentary system.
The Japanese prime minister’s power has grown in five ways. First,
due to the 1994 reform, the prime minister, as an agent of the members
of the Diet, and in particular as an agent of politicians of the ruling
parties, has become more and more independent.
64 Harukata Takenaka

Second, again thanks to the 1994 reform, the prime minister can
rely more on screening as an ex ante mechanism of control, as he can
now appoint ministers more according to his preferences. Third, as
a result of the 2001 reform, he can put specific policies under his
direct control by making the Cabinet Secretariat and some ministers
responsible for formulating such policies. In other words, the prime
minister can reduce the cost of ex post control, monitoring, by
delegating important policies to the Cabinet Secretariat and a few
ministers.
Fourth, with the 2014 civil service reform, the control of the prime
minister now extends to the public officials. Fifth, through the creation
of the NSC in 2013, he has more influence over the formulation of
security policy and two important agents, the foreign and defense
ministers.
We claim that with the first two of the three important reforms, the
political reform of 1994 and the administrative reform of 2001, the
Japanese political system became a “2001 system,” characterized by
a prime minister with more power. We also emphasize that the second
Abe administration is politically important in that the prime minister’s
power expanded even further under this administration, both in the
ruling party (the LDP) and in the government, when compared with
past administrations.
To the contention that the power of the Japanese prime minister has
expanded, we can expect an important challenge to be raised. This is to
refer to the recurrence of briefly tenured prime ministers between 2006
and 2012.20 This observation has important implications for the nature
of the Japanese political system. Thus, it is necessary to evaluate its
validity.
Two factors are behind this phenomenon. One is the issue of man-
agement. The purchase of a great set of pots does not make you a good
cook – you need skill as well. Likewise, effective institutions do not
enable all prime ministers to manage their administrations effectively.
Two administrations, the first Abe cabinet and the Hatoyama cabinet,
could stay in power only briefly because of poor management by the
prime minister. Abe did not effectively screen ministerial candidates,
and he formulated too many policies, which were often contradictory,
in a short period of time. Hatoyama undermined the relationship
between Japan and the United States over the issue of relocating the
Futenma Airfield.
Prime Minister’s Power in Japanese Politics 65

There is a second, more important factor. It is the Japanese version of

divided government, as I have explained elsewhere.21 For most of the
period from 2006 to 2012, the opposition dominated the House of
Councillors, which made it very difficult for the prime minister to
manage the administration. The Fukuda, Aso, Kan, and Noda cabinets
all suffered from divided governments and had a hard time legislating
bills. For example, Prime Minister Noda had to accept the LDP’s
demand to dissolve the lower house in return for passing a bill to
increase the consumption tax rate.
The repetition of briefly tenured prime ministers because of divided
governments shows that, although the prime minister’s power has
expanded, the House of Councillors is his Achilles’ heel. This has
important implications for the nature of the current Japanese political
system. As the result of a series of reforms, the Japanese parliamentary
system has shifted closer to a Westminster model.22 Yet the House of
Councillors makes the Japanese system still distinct from the
Westminster model, as it is much stronger than the second chamber
assumed under that model.
The Japanese political system also stands apart from the
Westminster model in one other aspect. That is the overall relationship
between the executive and the legislature. The power of the cabinet
over the legislative organization, as well as parliamentary business and
legislation, is much more limited in Japan than in the Westminster
model. How that constrains the Japanese prime minister remains to
be explored by future research.

Notes
* An earlier version of this chapter was published as “Expansion of the
Japanese prime minister’s power in the Japanese parliamentary system:
Transformation of Japanese politics and the institutional reforms,” Asian
Survey 59:5 (September/October 2019): 844–865, from University of
California Press. The article was revised for the publication of this book.
1. Margarita Estevez-Abe, “Japan’s shift toward a Westminster system:
A structural analysis of the 2005 lower house election and its
aftermath,” Asian Survey 46:4 (2006): 632–651; Tina Burrett, “Abe
Road: Comparing Japanese prime minister Shinzo Abe’s leadership of
his first and second governments,” Parliamentary Affairs 70:2 (2017):
400–429; Ellis S. Krauss and Robert J. Pekkanen, The Rise and Fall of
66 Harukata Takenaka

Japan’s LDP (Ithaca, NY: Cornell University Press, 2014);

Satoshi Machidori, Shushō Seiji no Seido Bunseki (Institutional analysis
of prime minister politics) (Tokyo: Chikura Shobō, 2012); Aurelia
George Mulgan, The Abe Administration and the Rise of the Prime
Ministerial Executive (Milton Park: Routledge, 2018); Koji Nakakita,
Jimintō (Liberal Democratic Party) (Tokyo: Chukō Shinsho, 2017);
Frances McCall Rosenbluth and Mike F. Thies, Japan Transformed
(Princeton, NJ: Princeton University Press, 2010); Tomohiko Shinoda,
Koizumi Diplomacy (Seattle: University of Washington Press, 2007);
Tomohiko Shinoda, Contemporary Japanese Politics (New York:
Columbia University Press, 2013); Harukata Takenaka, Shushō Shihai
(Dominance of the prime minister) (Tokyo: Chukō Shinsho, 2006).
2. For arguments which attach importance to the series of institutional
reforms since 1990s, see Estevez-Abe, “Japan’s shift toward a west
minster system”; Machidori, Shushō Seiji no Seido Bunseki; Nakakita,
Jimintō; Mulgan, Abe Administration; Shinoda, Koizumi Diplomacy;
Shinoda, Contemporary Japanese Politics; Takenaka, Shushō Shihai.
3. The second, third, and fourth Abe cabinets were formed in
December 2012, September 2014, and December 2017, respectively.
I call the three cabinets together the second Abe administration.
4. Mulgan, Abe Administration; Harukata Takenaka, “Kesshō”
(Conclusion), in Futatsu no Seiken Kōtai (The Two Changes of
Government), edited by Harukata Takenaka, pp.273–289 (Tokyo:
Keisō Shobō, 2017).
5. Estevez-Abe, “Japan’s shift toward a west minster system”; Shinoda,
Koizumi Diplomacy; Takenaka, Shushō Shihai.
6. Nakakita, Jimintō; Machidori, Shushō Seiji no Seido Bunseki;
Satoshi Machidori, “Minshutō Seiken ka ni oeru Kantei Shudō”
(Leadership of the prime minister’s office under the DPJ
administrations), in Seiken Kōtai to Seitō Seiji (Changes in the
Government and Party Politics), edited by Jun Iio (Tokyo: Chuō
Kōron, 2013); Mulgan, Abe Administration.
7. Kaare Strøm, “Parliamentary democracy and delegation,” in
Delegation and Accountability in Parliamentary Democracies, edited
by Kaare Strøm, Wolfgang C. Müller, and Torbjörn Bergman,
pp.54–106 (Oxford: Oxford University Press, 2003).
8. Strøm, “Parliamentary democracy and delegation,” p.62.
9. Arthur Lupia, “Delegation and Its Perils,” in Delegation and
Accountability, pp.33–54. Strøm also refers to these means, introducing
Lupia’s arguments (“Parliamentary Democracy and Delegation,” p.63).
10. Thomas Poguntke and Paul Webb, “Presidentialization of politics in
democratic societies,” in The Presidentialization of Politics, edited by
Prime Minister’s Power in Japanese Politics 67

Thomas Poguntke and Paul Webb, pp.14–15 (Oxford: Oxford

University Press, 2005). Krauss and Nyblade emphasize the increase in
media attention to the prime minister in Japanese politics. See also Ellis
S. Krauss and Benjamin Nyblade, “‘Presidentialization’ in Japan? The
prime minister, media and elections in Japan,” British Journal of
Political Science 35:2 (2005): 357–368.
11. On the factors that affect the relationship between the party leadership
and backbenchers, see Shaun Bowler, D. M. Farrell, and R. S. Katz,
“Party cohesion, party discipline, and parliaments,” in Party Discipline
and Parliamentary Government, edited by D. M. Farrell Bowler and
R. S. Katz, pp.3–22 (Columbus: Ohio State University Press, 1999);
Michael Gallagher, “Introduction,” in Candidate Selection in
Comparative Perspective, edited by Michael Gallagher and
Michael Marsh, pp.1–19 (Newbury Park, CA: Sage Publications, 1988).
12. Strøm, “Parliamentary Democracy and Delegation,” p.62.
13. Takenaka, Shushō Shihai.
14. For the role of the Cabinet Secretariat in legislating these bills, see
Shinoda, Koizumi Diplomacy.
15. Nakakita, Jimintō, pp.66–67.
16. On the introduction of the NSC, see Adam P. Liff, “Japan’s National
Security Council: Policy Coordination and Political Power,” Japanese
Studies 38:2 (2018): 253–279; Yasuaki Chiyoji, Kawari yuku Naikaku
Anzen Hoshō Kikō (Evolution of the security organization of the
cabinet) (Tokyo: Hara Shobō, 2015), pp.161–234. See also Mulgan,
Abe Administration, pp.35–36.
17. National Security Council Law, Article 2.
18. Adam P. Liff, “Japan’s National Security Council,” Sankei Shimbun,
August 23, 2006; Tsuyoshi Sunohara, Nihon ban NSC to ha nanika
(What is the Japanese version of the NSC?) (Tokyo: Bunshun Shinsho,
2014), pp.39–40; Chiyoji, Kawari yuku Naikaku Anzen Hoshō Kikō,
pp.162–163.
19. Nihon Keizai Shimbun, April 7, 2007; Asahi Shimbun, April 6, 2007.
20. On the ineffective governance under the DPJ governments, see Phillip
Y. Lipscy and Ethan Scheiner, “Japan under the DPJ: The paradox of
political change without policy change,” Journal of East Asian Studies
12:3 (2012): 311–322.
21. Harukata Takenaka, “The frequent turnover of Japanese prime
ministers: Still a long way to a Westminster model,” in Looking for
Leadership, edited by Ryo Sahashi and James Ganon, pp.46–82 (Tokyo:
JCIE, 2015).
22. Estevez-Abe, “Japan’s shift toward a Westminster system”; Machidori,
Shushō Seiji no Seido Bunseki.
3 Why Does the Abe Government’s
Approval Rating Always
Recover?
masaru kohno

3.1 Introduction
One of the most noteworthy facts about the government of Shinzo Abe
is that, although his cabinet approval rating often drops quite dramat-
ically, it always bounces back.1 As a result, his government has been
able to maintain a relatively high level of public support, which no
doubt has contributed to the unprecedented longevity of Abe’s tenure
as Japan’s prime minister.2
In the spring of 2017, for example, Abe’s approval rating started
dropping in various opinion polls, as a result of the so-called Mori-
Kake scandals,3 as well as a series of improper behavior and gaffes
made by several members of his own cabinet. At one time, many
observers even described the government as being on the verge of
collapse, using such expressions as “shini-tai,” which literally means
“dying body.” By the end of the summer, however, there were signs
that the approval rating was improving. It was precisely in reaction to
this turn-around of his rating that Abe swiftly dissolved the House of
Representatives and called for a general election in September.4
That the approval rating recovers after having fallen is something
characteristic of the Abe government. Looking back at the records of all
cabinets established in Japan after World War II, the approval rating
tends to be highest when the cabinet is first formed, and it begins to

* The original version of this chapter appeared as “Naze Abe Naikaku no Shijiritsu wa
Fukkatsu Suru no ka,” Chuo Koron 131 (November 2017): 84–97. The funding for
this research was provided by the Office of International Affairs at Stanford
University, and I thank Professor Michael Tomz at Stanford for inviting me to
participate in his project and for including, in one wave of the surveys he conducted in
Japan, the questions and experiment from which the main findings of this chapter are
drawn. I thank Munenori Kita and Susumu An’naka for their assistance in data
analysis and valuable comments. My gratitude also goes to Robert Nordstrom and
Victoria Edwards for their help in translating the original article.

68
Why Does the Abe Government’s Approval Rating Always Recover? 69

drop in subsequent months. In Abe’s case, too, the rating was highest in
the earliest period in office and, in that respect, there is nothing unusual
about the pattern of his tenure. However, what is unique about the Abe
government is that its approval rating does not fall into gradual decline
even after experiencing serious crises in which the negative rating (i.e.,
“disapproval”) becomes higher than the positive approval. That Abe has
survived not only one but several instances of significant drops in
approval rating – and has thus maintained relatively high public support
for a long period – is very unusual in Japan’s postwar political history.
Why does the Abe government’s approval rating always recover?
Answering this question, I believe, provides an important key to under-
standing contemporary Japanese politics, and the present chapter is
devoted to such a task. I begin, in Section 3.2, by analyzing the monthly
polls released by the Japanese mass media to trace the changes in public
attitudes toward the Abe government. In Section 3.3, I critically exam-
ine some of the conventional interpretations of public support for the
Abe government. In Sections 3.4 and 3.5, I attempt to solve the riddle of
Abe’s resurgent popularity based on a survey experiment which
I conducted in November/December 2015, two months after the pas-
sage of controversial national security legislation caused public support
for his government to sink to its lowest level. Section 3.6 concludes with
discussion on the broader implications drawn from the analysis and
findings presented.

3.2 Changes and Characteristics of the Abe Government’s

Approval Rating
I begin by analyzing the long-term trend as well as short-term fluctu-
ations in the public’s assessment of the Abe government based on the
series of monthly public opinion polls conducted by Japanese mass
media. The time-series graph, depicted in Figure 3.1, is taken from
the Yomiuri Shimbun, one of the most widely circulated newspapers,
for the period of December 2012 through August 2017. Note this line
graph does not show the approval rating per se but represents the
difference in the positive (approval) rating and the negative (disap-
proval) rating, which I call hereafter the “level of confidence” in the
cabinet, following Masumi Ishikawa’s usage of the term (Ishikawa
1984). I opt to use this confidence measure rather than a simple
approval score because approval and disapproval ratings may increase
70 Masaru Kohno

simultaneously when public opinion becomes polarized into two

camps with sharply opposing opinions in the midst of a salient political
event. This may occur because, when such a controversial event is
absent, respondents can choose to withhold their opinions in survey
by selecting “I don’t know” or “I don’t want to answer.”
Using a Bayesian technique, vertical lines are drawn in Figure 3.1 to
denote what can be inferred as significant decreases from one month to
the next in public support for the Abe government. (Using the same
technique, significant increases in the cabinet’s popularity can be iden-
tified, but they are not shown in the figure for the sake of clarity.) Figure
3.1 is illustrative of a number of important characteristics of public
attitudes toward the Abe government. Here I summarize four points
that are perhaps the most critical.
First, as noted earlier, Abe’s approval rating did indeed drop at many
points during this period. At the same time, Figure 3.1 also demon-
strates that these drops were followed by recoveries and, in some cases,
the post-recovery rating even surpassed the pre-drop rating.
Second, confidence in the Abe government often suffered severely
when public opinion was sharply divided on controversial legislations,
such as the “Act on the Protection of Specially Designated Secrets”
(which allows the government to designate sensitive information as
“special secrets” protected from public disclosure) and the new
national security law (which allows Japan to exercise collective self-
defense). As far as these controversial bills are concerned, important
political events – such as the cabinet’s endorsement, the Diet deliber-
ation, the passage through the lower house, and the final enactment –
seem to influence the level of public confidence at various points. Some
of the most obvious cases of such relevant events are annotated in
Figure 3.1.
Third, among the political events that led Abe’s popularity to drop,
the sharpest declines can be observed in two cases, one following the
controversy over the new national security law in 2015 and the other
following the surfacing of the Mori-Kake scandals in 2017. In particu-
lar, the case of the national security law is noteworthy, because this was
the legislation that Abe himself prioritized above anything else.5
Despite his commitment to the bill, the cabinet’s positive (approval)
rating dropped below its negative (disapproval) rating for the first time
since Abe came into office in late 2012. Yet the Abe government
recovered even from this episode. As displayed in Figure 3.1, public
Why Does the Abe Government’s Approval Rating Always Recover? 71

Yomiuri Shimbun Polls

2014.11
2013.06
2013.07

2013.12

2014.06
2014.07
2014.10

2015.06
2015.07

2016.10

2017.03

2017.06
2017.07
60
Confidence in Government
40

Diet Deliberation
on New National
Security Law
20

Bill on Protection of
Designated Secrets
Passed in LowerCabinet Decision
0

House to Grant Collective

Self-Defense
-20

0 20 Months in Office 40 60

Figure 3.1 Public confidence in the Abe government

confidence reached bottom when the law passed in September but

began to recover gradually thereafter.
Fourth, the passage of the 2015 national security law seems to have
additional implications for the fate of the Abe government. Seen from
a rather long-term perspective, Figure 3.1 suggests that the overall trend
of the cabinet approval rating can be divided into two parts: the first
from the government’s formation to the enactment of this law; and
the second part thereafter. The two parts are indeed in contrast, in that
the first shows a gradual decline, while the second trends upward at least
until the Mori-Kake scandals surfaced in 2017. Here thus lies the
importance of the national security law. Although the cabinet’s approval
rating dropped notably following the passage of this controversial legis-
lation, this seems to have become a watershed event that, in retrospect,
improved the public’s view of the government in the long run.
Thus far, I have identified trends and important traits of the changes
in the people’s attitudes toward the Abe government based on the
publicly available polling results. It is possible that, because these
results reference only one particular source (the Yomiuri Shimbun),
they may be biased. To respond to this concern, I have replicated the
analysis using polling data from three other media sources, including
the Asahi Shimbun, the Nihon Keizai Shimbun, and the Sankei
Shimbun. These replications (not reported) hardly change the basic
72 Masaru Kohno

observations already mentioned – namely, the tendency for the cabinet

approval rating to recover directly following declines and a long-term
improvement after Japan’s new security legislation was deliberated and
finally enacted in the summer of 2015. Thus, I submit that the patterns
depicted in Figure 3.1 are reliable characterizations of the public assess-
ment of the Abe government.

3.3 Questioning the Conventional Interpretations

How, then, can one account for the ups and downs of Abe’s approval
ratings? As noted earlier, the repeated combination of temporary drops
and subsequent recoveries defies much of the previous experiences of
Japanese governments established throughout the post–World War II
period. In fact, the resurgent popularity of the Abe government pre-
sents itself as a puzzle, one that is difficult to explain from ordinary
analytical perspectives.
Take, for example, political ideology as an explanation. It is almost
taken for granted, among political commentators and social critics in
Japan, that those who actively support Abe are staunch conservatives.
Abe himself has often described, in speeches and writings, the experi-
ence of Japan under the reformist government led by the Democratic
Party of Japan from 2009 to 2012 as that of a “nightmare.” His
supporters are also identified as being unsympathetic, even hostile, to
liberal and leftist parties in their political orientations. On the surface,
the Abe government does seem very effective in rallying the conserva-
tive base, even in comparison with some of the previous hardline prime
ministers like Nobusuke Kishi and Yasuhiro Nakasone.6 However,
a perspective focusing on political ideology alone stands insufficient
when accounting for the fact that public support for the Abe govern-
ment drops steeply with salient political issues rising on the agenda. If
support for Abe were ideologically motivated, why would these
instances necessarily result in such drastic drops?
As exemplified by other countries, it is possible that public support
for the national leader tends to increase with debates concerning
important issues, particularly on defense or foreign affairs,
a phenomenon referred to in the literature of political science as the
“rally ’round the flag” effect.7 The patterns observed in Figure 3.1 are
contrary to this effect. For example, in accordance with the thesis of the
“rally ’round the flag” effect, the debate and passage of the legislation
Why Does the Abe Government’s Approval Rating Always Recover? 73

concerning Japan’s national security in 2015 should have increased the

Abe government’s popularity because this was undoubtedly the matter
that the Abe government pursued vigilantly and in a manner most
aligned with the opinions of its core conservative supporters. Why,
then, do we see a drop in its popularity, or a “reversed rally”?
Of course, sensitive political issues spur debates not only on the
conservative side of the ideological spectrum. One could argue that,
particularly in these intense times, liberal parties and leftist groups in
the opposition camp critical of Abe become motivated and mobilized as
well. Does such a revitalization of the opposition offer a good explan-
ation? The answer is no. Surely some voters in Japan support the Abe
government only reluctantly, because they feel there isn’t a viable
alternative. It is thus possible that such issues concerning designated
state secrets and the right to exercise collective self-defense may have
intensified dissatisfaction among these voters. However, if dissatisfied
over these issues, why would they go back to supporting the Abe
government so soon? If the hypothesis concerning the activization of
liberal/leftist opposition were correct, these voters’ anger and criticisms
would accumulate in the long run; in reality, however, public support
rebounded each time after these contentious issues faded away from the
main political scene. The opposition intensification may account for
a temporary decline in Abe’s rating, but it cannot account for why the
rating bounces back.
Let us, then, explore a completely different line of explanation. The
discussion thus far has been based on the premise that the patterns of
public approval/disapproval of the Abe government result from polit-
ical events. However, another widely held view (especially among some
economists) is that the popularity of the Abe government is most
strongly tied to public perceptions of its economic policy,
“Abenomics.” Can this perspective offer a good explanation?
To test this economy-driven hypothesis, I have looked at a number of
economic performance indicators together with Abe’s approval rat-
ings. Figure 3.2 provides a set of time-series plots showing the active
opening-to-application ratio in the job market and the stock price
indicator Nikkei TOPIX (30-day average), side by side with the cabinet
confidence score.8
According to Figure 3.2, there is no clear correlation between the
observed economic indicators and public confidence in the Abe govern-
ment. The job-opening indicator, for instance, has improved
74 Masaru Kohno

Poll Average from Four Media Sources

.8 Active Opening-to-Application Ratio 1:6

1600

60
1200 1400

40
TOPIX

20
1000

0
800

-20

0 20 40 60
Months in Office
confidence job opening
TOPIX

Figure 3.2 Confidence in the government and economic indicators

consistently throughout Abe’s tenure and therefore cannot explain the

temporary ups and downs in Abe’s popularity. As for the stock price
indicator, one might detect a correlation between the Nikkei TOPIX
and the confidence level during the government’s first several months,
but no clear relationship is later observable. Further, in the midst of the
2015 debate on security legislation, when the confidence in the cabinet
sank to its very lowest, stock prices were rising rather than dropping. In
addition to these measures, I have also tried examining several other
economic indicators as well, including unemployment, price levels, and
the Bank of Japan’s forecast reports, but none displays correlation with
the trend of cabinet approval/disapproval ratings (not reported).
In the absence of empirical evidence, it is thus difficult to regard
“Abenomics” as the key determinant for the Abe government’s popular-
ity. I do not intend to imply that economic factors have no influence on
politics. I even admit, indeed, that how voters perceive the government is
affected by their perceptions of economic performance at some funda-
mental or unconscious level. However, in the case of the Abe govern-
ment, to claim that the cabinet approval rating is driven by the
underlying economic conditions would not be persuasive, particularly
in accounting for the characteristic combination of its drops and recov-
eries. Nor does such an economic hypothesis offer insight as to why the
passage of the new security legislation in 2015 was a watershed moment
resulting in a long-term increase in support for the government.
Why Does the Abe Government’s Approval Rating Always Recover? 75

In sum, the conventional perspectives that emphasize political ideol-

ogy or economic performance as the basis for Abe’s popularity are not
convincing. We are thus left with an empirical puzzle. To solve this
puzzle requires an alternative approach that will explore more directly
the mechanism of how public attitudes toward the Abe government are
determined in the course of political controversy.

3.4 National Security Legislation and Abe’s Popularity

The enactment of the new national security law in September 2015
under Abe represented a significant transformation of Japan’s defense
and alliance policy, departing markedly from the constitutional inter-
pretation his predecessors had adhered to throughout the postwar
years. Anticipating the historic importance of this event, I conducted
an online public opinion survey to gauge the political attitudes of
ordinary Japanese citizens, including their evaluations of the Abe gov-
ernment and of the controversial national security law. In this section,
I present findings from this survey. More specifically, this survey was
conducted in cooperation with Nikkei Research, one of the major
survey firms in Japan, on a sample of approximately 3,600 Japanese
men and women.9 The survey took place in late November and early
December 2015, two months after the national security law passed in
the Diet. As noted, public support for the Abe government had just
started recovering around that time.
At the very beginning of the survey, all respondents were asked to
answer questions regarding both whether they do or do not support the
current Abe government and how they evaluate the national security
law that had recently been enacted. Table 3.1 shows, as expected,
strong correlation between support for the new law and approval of
the Abe government. A closer look at the table also reveals nuanced
aspects of this correlation – that is, when the comparison is made
among only those respondents who answered with clearly stated opin-
ions on both questions, the correlation coefficient is estimated at
0.7836, while it drops to 0.3887 when the set of comparison includes
all respondents. This gap, of course, results from the standard format of
the survey, which allowed the respondents to choose either “I don’t
know” or “I don’t want to answer” from among other answer options.
Note also there is a discrepancy between the number of respondents
who withheld their opinions on the matter of cabinet approval and the
76 Masaru Kohno

Table 3.1 Evaluations of the Abe government and the national security
law

Are you in favor of or against the new law?

Opinions
withheld (don’t
In know/no
favor Against answer) Total

Do you or do Support 933 113 413 1,459

you not Not support 130 1,078 452 1,660
support the Opinions 58 98 312 468
Abe withheld
government? (don’t know/
no answer)
Total 1,121 1,289 1,177 3,587

number who withheld their opinions regarding national security legis-

lation. The number of those withholding their opinion on Abe’s cabinet
totaled 468 (13 percent of the sample), while the number of those
withholding their opinion on security law was much higher, at 1,177
(33 percent of the sample). Thus, on average, from the respondents’
point of view, whether to approve or disapprove of Abe was cognitively
a much simpler question compared to how to evaluate the appropriate-
ness of the newly adopted national security law.
Incidentally, Table 3.1 shows the aggregated results based on the
entire survey sample. This display may be misleading because, in
actuality, I embedded a random assignment procedure into this part
of the survey so that the respondents were divided into two groups
a priori and faced the questions in different ways. Specifically, although
both groups were given the exact same set of questions (with the exact
same set of response options), their order (i.e., whether the question(s)
regarding national security legislation or the question regarding Abe
government was asked first) was reversed between the two groups.
Table 3.2 clarifies the contrasting order of questions that each group
of respondents was actually assigned.
As can be seen in Table 3.2, those in Group A were first asked whether
or not they support the Abe government. This sequencing was thought
Why Does the Abe Government’s Approval Rating Always Recover? 77

Table 3.2 Different sequence of questions

Group A
Question 1: Do you or do you not support the current Abe government?
Question 2: Are you in favor of or against the law regarding national security
that was passed in the last Diet session?
Question 3: Do you think the law regarding national security is constitutional
or unconstitutional?
Group B
Question 1: Are you in favor of or against the law regarding national security
that was passed in the last Diet session?
Question 2: Do you think the law regarding national security is constitutional
or unconstitutional?
Question 3: Do you or do you not support the current Abe government?

to be better at eliciting attitudes toward the Abe government that the

respondents possess in their ordinary environments. In Group B, by
contrast, I asked the respondents for their opinions on the national
security law prior to asking about their approval or disapproval of
Abe. This sequencing was aimed at reminding the respondents – while
they were still formulating their opinions about the Abe government – of
the controversial legislative event that took place two months earlier.
Hence, by dividing the sample into these two groups, the survey took on
a very simple experimental design. The “treatment” in this experiment
was informational priming – that is, having the respondents in Group
B reflect on the controversial national security issue before they were
asked to express their approval/disapproval of the Abe government.
If this treatment had no effect, we would expect no difference in the
average response between Group A and Group B. In the current
research in political science, simply changing the order of questions in
a survey is perhaps too simple an experimental design, one that may
yield only a very limited effect on the outcome. This may be an advan-
tage, however, in that, if we observe a treatment effect using this least
complicated design, we can be more generally convinced of the possi-
bility that reflection on the important national security controversy
would influence assessment of the Abe government. Even prior to the
summer of 2015, as discussed in Section 3.2, Abe’s approval rating had
fallen on several occasions when confronting sensitive political issues.
It was therefore expected in this experiment as well that the
78 Masaru Kohno

0.5

0
Approve Disapprove Opinion withheld

Group A Group B

Figure 3.3 Distribution of government evaluation in comparison

respondents in Group B, compared to those in Group A, would be less

supportive of Abe’s government.
Figure 3.3 summarizes the results. There seems to be a group-based
difference in how the respondents expressed their assessment of the
Abe government. As can be seen, there was virtually no difference
between the two groups as far as those who claimed not to support
Abe. However, the number (ratio) of respondents who approved the
government is markedly lower in Group B, the group primed with
questions regarding the national security legislation in advance.
Instead, more respondents in Group B withheld their opinions concern-
ing their evaluation for the Abe government, by choosing to answer
either “I don’t know” or “I don’t want to answer.” Through this simple
experiment, we can thus confirm that salient political events, such as
the passage of the controversial national security legislation, do influ-
ence the approval rating of the Abe government.

3.5 Solving the Puzzle of Abe’s Resurgent Popularity

The results confirm, under an experimental setting, the predicted pat-
tern of change in the public assessment of the Abe government, but this
is only the very first step in solving the entire puzzle. Having now
established that major political events lead to decreased public support
for Abe, we can now explore who – with what attributes – are more
likely to change their attitudes and behavior.
In the aforementioned survey, I asked the respondents many stand-
ard questions regarding their social and economic backgrounds, such
as sex, age, income, education, occupation, etc. While analyzing the
varying influences of these socioeconomic factors might discern some
Why Does the Abe Government’s Approval Rating Always Recover? 79

Figure 3.4 (A) Views on new national security law and priming effect

valuable insights, nothing would be more intriguing than investigating

the interactions between the priming effect and the respondents’ opin-
ions on the national security law itself. By focusing on this aspect of the
results, it may be possible to identify what type of respondents are most
likely to stop supporting the Abe government in the event of heated
public controversy.
Figure 3.4 (A) shows the main results of ordered-logit analysis,
which compares the different marginal effects of informational prim-
ing, or the varying likelihood with which the assignment to Group B as
opposed to Group A induces changes in the respondents’ evaluations
about the Abe government. For the sake of clarity, the vertical line is
drawn at the effect being equal to 0 to represent the base line of Group
A. Each of the points in the figure denotes the estimated marginal
effects of informational priming, together with their respective 95 per-
cent confidence intervals. The three possible (mutually exclusive) cat-
egories of responses are: (1) supporting the government; (2)
withholding opinion; and (3) opposing the government. Thus, for
each category, if the estimated point falls to the right of 0, it implies
that the probability of Group B respondents providing that category of
answer is greater than that of Group A. When the point is estimated to
80 Masaru Kohno

the left of 0, it implies that the probability decreases. The further the
predicted value from the baseline, the stronger the priming effect.
According to Figure 3.4 (A), as far as those respondents opposed to
the new national security law, there is no big difference between Group
B and baseline Group A. In other words, for these anti-legislation
respondents, the informational priming – that is, reminding them of
the law – does not significantly influence their evaluations of the Abe
government after all. Presumably, these respondents represent leftists
or liberals, rather than conservatives.
As for those respondents who are in favor of the new national
security law, on the other hand, the informational priming seems to
have a marked effect. Assignment to Group B decreases the likelihood
that they approve the Abe government, while it increases the likelihood
that they express disapproval or withhold their opinions. Judging from
their preference for the national security law, these respondents are
presumed to be conservatives, not leftists or liberals. Hence, one can
conclude, for the first time with empirical evidence, that it is the
behavior and attitudinal changes of these rather conservative voters
that explain the short-term drops in Abe’s cabinet popularity in time of
political controversies.
Let us take a step further. Priming, as it was carried out in the experi-
ment, with the information about the legislative debate, elicits memories
of not only the content of the national security law but also the political
context in which the law was enacted. Looking back at the arguments
made for and against the legislation, those in favor usually emphasized the
point about increasing Japan’s deterrence capabilities and its greater
responsibilities in regional and global affairs, while those who opposed
were extremely critical of the procedure and tactics Abe used to push for
this major policy change. Because the new law forced a fundamental
revision in constitutional interpretations but not an actual constitutional
amendment, critics of Abe were outraged from the standpoints of consti-
tutionalism and the rule of law. Theoretically, then, the respondents’
opinions about the new national security law may have been influenced
not only by their policy preference per se but also by their judgment about
the importance of due process and governmental responsibility of uphold-
ing the provision and inherent value of the present constitution.
This raises the question: does the respondents’ judgment about the
constitutionality of the new national security law affect their evaluation
of the Abe government? And how does such a judgment interact with the
Why Does the Abe Government’s Approval Rating Always Recover? 81

Figure 3.4 (B) Priming effect and views on constitutionality among pro-
legislation respondents

priming effect of the experiment? To explore these issues, I have subdiv-

ided both pro-legislation and anti-legislation groups of respondents based
on their expressed opinions about the constitutionality of the law. The
results of these analyses are summarized in the following two figures,
although the subdivision causes an inevitable small sample size problem.
Figure 3.4 shows that, among respondents who are against the new
national security law, there is no tangible difference, attributable to the
informational priming, regardless of whether they think the law is
unconstitutional or constitutional (although the small sample size
makes it hard to be conclusive on the latter point). For these anti-
legislation respondents, the estimated effects for the primed groups
are more or less fixed around the same point as the baseline groups.
Presumably, those opposed to the new legislation have established
beliefs about the law regardless of whether or not they are explicitly
reminded about it and, based on these beliefs, they do not change their
views on the Abe government after all.
Figure 3.4 reveals, on the other hand, significant interactions
between the priming effects and the respondents’ views on the consti-
tutionality, among those who are in favor of the new national security
82 Masaru Kohno

Figure 3.4 (C) Priming effect and views on constitutionality among anti-
legislation respondents

law. That is, while no clear effect can be observed for those who think
the new law unconstitutional, it is among those respondents who think
that the law is constitutional that the assignment to Group B decreases
the likelihood that they approve the Abe government.
Who are those respondents in favor of the newly adopted national
security law with the view that the law is constitutional? Surprisingly,
and perhaps ironically, they are those respondents whose positions are
closest to Prime Minister Abe himself. Here lies the underlying mechan-
ism of Abe’s resurgent popularity. When important political issues
surface and become a matter of public controversy, these supporters
closest to Abe seem to withhold their enthusiastic support, as if they
want to “cool off” the situation in the midst of heated and polarized
debate. They are the ones causing some nontrivial drops in the level of
public confidence. Precisely because these drops are caused by those who
are closest to Abe’s own policy preference, they are always temporary
phenomena, and the popularity recovers eventually. Meanwhile, the
experiment flatly rejects the hypothesis that the mobilization of leftist
or liberal forces caused drops in Abe’s popularity. Had this been the
Why Does the Abe Government’s Approval Rating Always Recover? 83

case, the Abe government would not have been able to maintain the high
level of public approval in the long run.

3.6 Conclusions
Under a democratic system, voters consider and prioritize many factors
in evaluating the government. Even among experts on the subject, there
is no solid agreement as to what ultimately determines the approval
rating revealed in public opinion polls. Further, how voters evaluate
their leaders in government is unlikely to be uniform but varies across
different democracies, especially between a country like the Unites
States, where the three branches of government are clearly separated,
and a country like Japan, where the legislative and executive (adminis-
trative) branches overlap in their membership and functions.
With regard to the government of Shinzo Abe in Japan, a diverse set of
interpretations have been provided for explaining its relatively high
popularity, but this chapter has focused on a particular and repeated
pattern which the conventional perspectives cannot fully account for,
namely the combinations of short-term falls and subsequent recoveries
in public confidence occasioned by important political controversies.
The latter half of the chapter has been devoted, based on a simple survey
experiment, to empirical analysis of how the passage of the national
security law in 2015 impacted public support for the Abe government.
The findings, among other things, confirm that typical drops in the
cabinet’s approval rating occur and that they occur not because of the
leftist/liberal opposition strengthening their criticism of the government
but rather because of the conservative voters becoming temporarily
hesitant to endorse the Abe government. Here lies the mechanism of
why Abe’s popularity always bounces back eventually.
The data and analyses presented here yields many insights and
implications about contemporary Japanese politics. I now close this
chapter with discussion on three points in particular.
First, when the public debate on the national security legislation was
at its most intense both inside and outside the Diet, some observers
argued that the government’s approval rating was dropping because
the public still did not fully understand its key concept of collective self-
defense. The results and analysis presented in this chapter refute such an
elitist view that underestimates voters’ knowledge of policy matters. As
demonstrated, temporary drops in the cabinet approval rating were
84 Masaru Kohno

caused by those respondents who not only supported the newly adopted
national law but considered it constitutional. These respondents were
likely to be well aware of both the purpose of changing Japan’s security
policy and the potential problems associated such a change. Insufficient
explanation, on the part of government, and hence the absence of
accountability may have been at the root of the shift in these voters’
attitudes and behavior, but, certainly, it was not their ignorance that
explains the temporary decrease in Abe’s popularity.
Second, the findings from the experiment show that voters with policy
positions most closely aligned with the Abe government do not rally
behind the prime minister during times of highly contentious public
debate. This is clearly at odds with the image of so-called populists who
are described as being emotionally motivated and taking political actions
accordingly. Although the word “populism” seems to have an established
position in the writings of some political scientists, the meaning of the
concept is still vague, and there is no clear, shared academic understand-
ing of how the word should be interpreted. In any case, the analysis of this
chapter implies that Japanese voters are far more cool-headed than popu-
lists, balancing their positions and attitudes on political issues, at least in
the case where crucial national policies are at stake. A discourse that
attaches too much importance to the dangers of populism facing Japan
is therefore unwarranted, and such a thinking does not provide a good
prediction on the future course of Japanese politics.
Third, to expand on the previous point, one individual who under-
stands that Japanese voters are not easily persuaded by emotional argu-
ments is none other than Prime Minister Shinzo Abe himself. Since
returning to office in December 2012, Abe has, on several occasions,
sought the public’s affirmation of the need for constitutional amend-
ment. However, he has always avoided pressing the issue in national
elections. This, I believe, is an indication that Abe has correctly recog-
nized that voters are levelheaded with a sense of balance in forming
opinions and taking political actions. As it is recalled, the most recent
general election was held in the fall of 2017, amid North Korean nuclear
weapons tests and missile launches. At that time, whether or not to
revisit Article 9 of the constitution and clarify the role of the Japan Self-
Defense Forces became an issue on the political agenda. Nevertheless,
Abe avoided prioritizing the issue even in this electoral setting.
In conclusion, the Abe government has faced a dilemma since its
formation. On the one hand, it has maintained a stable, relatively high
Why Does the Abe Government’s Approval Rating Always Recover? 85

level of support concentrated around a base of conservative voters. On

the other hand, Abe has never been able to fully seize their hearts, failing
to maintain their united support at the time when the public is polarized
around contentious political issues. This situation has not changed ever
since he came into office and is unlikely to change in the near future.

Notes
1 In Japanese, the word commonly used to measure cabinet approval/
disapproval in opinion polls is “shiji/fu-shiji,” which can be translated
as “support/non-support.” In this chapter, I use “approval” and
“support” interchangeably.
2. Abe was prime minister of Japan briefly between September 2006 and
August 2007. This chapter, however, focuses on the so-called Second Abe
cabinet, which was formed in December 2012 after his Liberal Democratic
Party won a landslide in the House of Representatives election.
3. The term “Mori-Kake” refers to two scandals combined, one involving
a private educational enterprise called Moritomo Gakuen, which sought to
construct a new elementary school in Osaka, and the other involving Kake
Gakuen, a private school that bid successfully to open a veterinary medicine
department in Ehime. The Moritomo scandal began when the head of
Moritomo petitioned to expand its kindergarten into a full-fledged
elementary school and negotiated to lease a parcel of government-owned
land with a steep discount. It is unknown how the discounted cost was
calculated, but Prime Minister Abe’s wife was named the school’s honorary
principal, aggravating the scandal. The Kake scandal surfaced when
a document was found in the Ministry of Education, Culture, Sports,
Science and Technology that hinted at Abe’s direct involvement in pushing
the administrative process for launching that same department. Kake
Gakuen is headed by a longtime personal friend of Abe’s. See “The Kake
and Moritomo scandals and the bureaucracy,” Japan Times, October 20,
2017.
4. The lower house election was held on September 22, 2017. Abe’s Liberal
Democratic Party, in coalition with the Komei (Clean Government)
Party, won a clear majority of the parliamentary seats.
5. To be precise, the Diet passed, on September 19, 2015, a package of
eleven bills related to Japan’s national defense, its relations with
allies, and its role in regional security and international affairs. The
most controversial element was that the new law, though under
limited circumstances, enabled Japan’s Self-Defense Forces to
participate in military activities for “collective self-defense” with
86 Masaru Kohno

important allies. Previously, the government had upheld the

interpretation that, while Japan’s constitution guaranteed the right
to “self-defense,” any act of collective self-defense was not permitted
even when an ally and its forces were attacked during their operation
of protecting Japan.
6. Nobusuke Kishi, grandfather of Shinzo Abe, who served as a cabinet
member during World War II, returned to the postwar political scene and
became prime minister from 1957 to 1960. Kishi is known for his role in
consolidating Japan’s security ties with the United States. Yasuhiro
Nakasone served as prime minister from 1982 to 1987. He is known
for his leadership in pushing for the privatization of state-owned
enterprises as well as for expanding Japan’s contribution in its bilateral
defense arrangement with the United States.
7. See Mueller (1970) for a classic study on the subject.
8. I have used, for this figure, the average confidence score of the Yomiuri
Shimbun, the Asahi Shimbun, the Nihon Keizai Shimbun, and the Sankei
Shimbun.
9. Through a stratified random sampling procedure, respondents for this
survey were drawn from categories defined by gender and region in
proportion to demographic data reported in the most recent edition of
the Jūminkihondaichō (Basic Residence Register). Due to the limited
number of internet users among the elderly, the sample is restricted to
individuals between twenty and sixty-nine years of age. The median
respondent in the sample is thus slightly younger than the median
Japanese resident. In previous comparisons between Nikkei Research
online samples and the Japanese population, the samples are also
shown to skew slightly toward those with higher income and those who
are more educated.

References
Ishikawa, Masumi. 1984. Deeta Sengo Seijishi. Tokyo: Iwanami.
Mueller, John E. 1970. “Presidential popularity from Truman to Johnson.”
American Political Science Review 64: 18–33.
4 The LDP under Abe
yuki o m aeda and s teven r . r eed

4.1 The Most Successful Prime Minister in LDP History?

Prime Minister Shinzo Abe is arguably the most successful prime
minister in Liberal Democratic Party (LDP) history. First and foremost,
he has led the party to an unprecedented series of landslide victories. In
the elections of 2012, 2014, and 2017, the LDP defeated the leading
opposition party by an average of over 230 seats, winning over 60 per-
cent of the seats in each case. The LDP also won landslide victories in
the upper house in the 2013, 2016, and 2019 elections, a total of six
consecutive national elections. The party has never experienced such
a run of landslides. The only other LDP prime minister in the conversa-
tion is Junichiro Koizumi. Koizumi led the LDP to landslide victories in
the 2001 upper house election and the 2005 general election. However,
the 2003 general election and 2004 elections were losses. Koizumi was
thus only two for four, while Abe is six for six.
In order to win consecutive elections, a prime minister must remain
prime minister through several elections. Abe is the longest-serving
prime minister in Japanese history. Staying in power requires maintain-
ing the support of his party. The LDP is a factionalized party, and the
competition to become the next prime minister has generally prevented
any one leader from staying in power for long. Abe has also maintained
a relatively high level of cabinet support throughout his second term.
High approval ratings do not guarantee an LDP prime minister long
tenure, but low approval ratings now doom him to a short tenure. This
point was driven home by the six short-term prime ministers that
followed Koizumi’s landslide victory for the LDP in 2005, starting
with Abe’s own first term. Low support rates played a large role in
forcing each of the six out of office after only one year. In his second

* This work was supported by JSPS KAKENHI Grant Number JP17H00971.

87
88 Yukio Maeda and Steven R. Reed

term Abe has managed to keep his approval ratings as high as those of
Koizumi. Abe has also made several bold policy proposals and
delivered on some of them, a feat that has been rare among LDP
prime ministers (Hayao 1993). One may well doubt the wisdom and
feasibility of some of his proposals, but one cannot deny their boldness.
After all of this praise, it is wise to remember that Abe’s first term was
a disaster. Abe succeeded Prime Minister Koizumi and was greeted
with high hopes that were immediately dashed. He accepted the postal
rebels who won their seats in 2005 back into the LDP, destroying both
the fact and, perhaps more importantly, the image of LDP of a unified
party. If Abe’s legacy was based on his first term, he would now be
considered one of the least successful prime ministers in LDP history.
The fact that Abe was one of three consecutive failed LDP leaders
before 2009 but was one of, if not the, most successful LDP prime
minister between 2012 and 2017 suggests that the reasons for his
success are to be found not in Abe’s personal characteristics but in
the situation he faced and what he and the LDP learned during the three
years and three months of DPJ government. Abe certainly took full
advantage of the opportunities presented by the situation he faced in
the 2012 election, but we can never know if someone else could have
done as well.

4.2 How Does Abe Win Landslide Victories?

Abe’s most impressive feat has been his series of successive landslide
election victories starting with the 2012 election. Abe’s victories have
not been produced by increasing the LDP vote or by pursuing popular
policies. Instead, the secret of his success is, first and foremost,
a fragmented opposition. A fragmented opposition allows the LDP to
win more seats with fewer votes. But it also means that voters and
interest groups can be certain that the LDP will be in power after the
next election, allowing the party to mobilize the “organizational vote”
based not on policies but on providing access to governmental
policymaking.

4.2.1 Increasing the LDP Vote?

First, the LDP under Abe has not won by increasing the LDP vote.
Indeed, the LDP received fewer votes in the landslide victories of 2012,
The LDP under Abe 89

30,000,000

25,000,000

20,000,000

15,000,000

10,000,000
LDP
DPJ
5,000,000

0
2003 HR 2005 HR 2009 HR 2012 HR 2014 HR

Figure 4.1 Number of votes cast for the LDP and the DPJ, 2003–2014

2014, and 2017 than it did in the 2009 election it lost to the Democratic
Party of Japan (DPJ) in a landslide. The LDP won because the DPJ lost
even more votes than did the LDP, as illustrated in Figure 4.1. Thus,
“although there was a clear counter-wave reversing the swing for the
DPJ in 2009, there was no such swing in favor of the LDP” (Reed,
Scheiner & Smith 2013, p.44). DPJ voters in 2009 failed to vote in
2012: “For every ten extra abstainers the DPJ lost nine votes!” (Reed,
Scheiner & Smith 2013, p.45). Turnout dropped from 69.28 percent in
2009 to 59.32 percent in 2012. The LDP vote dropped by 1,658,836
votes between those two elections, but its percentage of the vote rose
from 26.1 percent to 27.0 percent because of falling turnout. These
findings are consistent with the hypothesis that those voters who were
drawn to the polls by the prospect of an alternation in power in 2009
were disappointed in the DPJ government and stayed home in 2012.
The shift in the percentage of votes was all about a rejection of the DPJ
and had little to do with a desire to return to an LDP government. The
LDP was virtually guaranteed to win the 2012 election no matter what
they did because they were the only alternative to the failed DPJ.
The LDP won more seats despite losing votes because the opposition
fragmented and turnout dropped. In the single-member districts, the
LDP won 64 seats in 2009 but 237 seats in 2012 because the opposition
fragmented and divided the opposition vote. Neither turnout nor the
90 Yukio Maeda and Steven R. Reed

LDP vote recovered to previous levels in 2014 or 2017, but the oppos-
ition remains fragmented. Abe’s landslide victories have been, there-
fore, quite different from Koizumi’s historic victory in 2005, which
increased turnout, the LDP vote, and the LDP percentage of the vote.
Koizumi also defeated an opposition that was led by a single party. His
victory was solid. There is good reason to believe that, if the LDP had
learned the lessons of the 2005 election, it would not have lost the 2009
election. Instead, the three leaders who followed Koizumi, starting with
Abe himself, watered down Koizumi’s policies and returned to trad-
itional LDP practices. Abe’s victories are more vulnerable. An oppos-
ition party attractive enough to draw nonvoters back to the polling
booths and raise turnout would threaten the LDP. If such a party
demonstrated a potential to defeat the LDP, it could become
a magnet for other opposition parties and candidates leading toward
a more unified opposition.

4.2.2 Popular Policies?

Abe has not won by pursuing popular policies. In fact, many of his
signature policies are vote losers. He has made several bold moves toward
the right. The clearest examples are the reinterpretation of the constitution
to allow collective defense in the face of nearly unanimous opposition of
constitutional experts. Indeed, “The right to CSD [Collective Self-
Defense] had previously been banned under the interpretation of the
Cabinet Legislation Bureau (CLB), but the intransigent former head of
the CLB had been shuffled to the Japanese Supreme Court and replaced
with a more pliant Abe appointee who approved of the reinterpretation.”
(Pekkanen, Reed & Scheiner 2016, p.15). The reinterpretation has been
criticized as “constitutional revision by fiat” (kaishaku kaiken) and
resulted in a temporary drop in Abe’s popular support. Abe has also
pressed forward on revising the constitution following the proper legal
procedures and made foreign policy moves to address two issues that have
festered since the end of the on the right: Korean “comfort women” and
the northern islands “lost” to Russia. Neither of these latter two moves
seem successful, but they have been bold.
There has long been talk of what might happen if someone played
“the nationalist card” in Japan. These discussions usually centered on
the popular author and governor of Tokyo Prefecture, Shintaro
Ishihara. Parties representing Ishihara’s positions have run under
The LDP under Abe 91

many different names and leaders over time but have never experienced
the slightest electoral success. The 2014 Party for Future Generations
(PFG) was probably the right wing’s best chance. Ishihara led the party,
and they had a high-profile candidate in Toshio Tamogami,
a decorated and outspoken retired chief of staff of the Air Defense
Forces. He was also a darling of the “internet right” (netto uyoku),
which engages in practices that have proven effective in Europe and the
United States, such as propagating fake news over the Internet (Schäfer,
Evert & Heinrich 2017). When the votes were counted, however,
Tamogami finished fourth out of four candidates, well behind the
Communist candidate. The far left won over six times the votes of the
far right. The only PFG candidates to win seats were two elderly
incumbent candidates who could have won without any party nomin-
ation. “In other words, the party label won exactly zero seats – zero in
PR and made no contribution to the two SMD victories” (Pekkanen &
Reed 2016, p.68).
Abe has not won elections by attracting right-wing voters. Neither
has he pulled the party very far to the right. A Yomiuri poll of LDP
members reports that 45 percent want the LDP to maintain a centrist
line, 28 percent prefer to move to the right, and 15 percent prefer
a move to the left (March 26, 2019). The UTokyo-Asahi Survey
(2019) reports that the LDP Diet delegation moved sharply to the
right and become more ideologically unified under Abe but that LDP
voters did not move at all. The LDP under Abe has moved to the right,
but not because their voters or their members support those policies.
The movement to the right is an elite-led phenomenon, and voters have
not followed.
The most convincing evidence that Abe’s policies are not popular
comes from two experiments, one on the 2014 election (Horiuchi,
Smith & Yamamoto 2018) and the other on the 2017 election
(Horiuchi, Kuriwaki & Smith 2019, p.21). Each of these articles
are based on an experimental design that allows them to estimate
the effect of including a given issue position on the probability of
supporting a manifesto. The authors use the actual positions of the
various parties on the major issues of each campaign but present
respondents with randomly generated combinations of those issue
positions allowing them to estimate the impact of any specific
policy position separately from other factors that affect voting
behavior.
92 Yukio Maeda and Steven R. Reed

In 2014 they find that “of the eleven policy positions where support
is estimated to be significantly different (at the 0.05 level) from support
for the LDP’s position, nine turn out to be more popular than the LDP’s
position on the corresponding policy dimension. . . . The only two
positions that are estimated to be significantly less popular than the
LDP’s . . . were proposed by opposition parties on the extreme left,
rather than the LDP’s main competitors in the election” (Horiuchi,
Smith & Yamamoto, 2018, pp.11–12). The findings for 2017 are
similar. In both elections the major opposition parties proposed more
popular polices than did the LDP.
Abe knows that his policies are unpopular and does whatever he can
to keep them out of the news. For example, it became clear in the 2017
election that Abe’s most problematic policy position was that of
restarting nuclear power plants (Horiuchi, Kuriwaki & Smith 2019,
p.24). His government has thus avoided announcing any plans for any
new plants and has been extremely careful about restarting the old ones
for fear of arousing negative public opinion (Kamikawa 2018, vol 2,
pp.111, 154, 163, 266, 303). Abe also keeps the Diet closed after
raising controversial issues. For example, the security related bills
passed in September 2015 and the regular session ended on
September 27, but Abe did not call a special session of the Diet and
waited for the next regular session, postponing debate until
January 2016 (Asahi, October 17, 2015). There is also some evidence
that Abe announces favorable news timed to distract from unfavorable
stories. For example, on December 9, 2014, the day before the contro-
versial Specially Designated Secrets Act would come into effect, Abe
announced that Apple would open a research institute in Yokohama
(Osaka 2015, pp.48–50).

4.2.3 Opposition Fragmentation

The most important single reason the LDP has won landslides under
Abe is that the opposition fragmented. Between 2003 and 2009 the DPJ
offered an increasingly viable challenger to the LDP, just as predicted
by Duverger’s Law (Reed 2007). After they took over the government,
however, they proved incapable of governing effectively (Kushida &
Lipscy 2013; Maeda & Tsutsumi 2015).
The DPJ presented a manifesto and campaigned as if their candidates
were all “on message.” However, once in office, the DPJ government
The LDP under Abe 93

largely ignored the manifesto, and individual officials pursued their

personal policy preferences. The signature policies of each of the three
DPJ prime ministers had not been part of the party’s election manifesto.
Hatoyama tried to solve the problem of US base concentration in
Okinawa. Kan, responding to the Fukushima nuclear disaster, pro-
posed ending dependence upon nuclear energy. Noda sought to raise
the consumption tax. Other policies changed every time a new cabinet
minister was appointed (Kamikawa 2018, vol 1, pp.185, 214). It was
impossible for voters to see a stable DPJ policy on any issue. The project
that attracted candidates to the DPJ between 2003 and 2009 was
defeating the LDP. The candidates thus spanned a broad policy
spectrum.
None of the characteristics described, however, make the DPJ sig-
nificantly different from the LDP from 1955 through 2009. The LDP
attracted a wide variety of candidates whose primary interest was in
participating in government policymaking. The first time the LDP had
run a campaign with all of their candidates technically “on message”
was in Koizumi’s postal privatization election of 2005. The difference
between the two parties is to be found not in their unity but in their
levels of institutionalization.
After long years of running the government with a diverse party, the
LDP had developed institutions and mechanisms for reaching consen-
sus. These mechanisms were cumbersome and prevented party leaders
from exercising strong leadership, but policies were enacted. Because
differences were hammered out behind closed doors, many of the
internal disagreements were invisible to the public eye.
Unsurprisingly, the DPJ had developed no such mechanisms. The DPJ
had a policymaking apparatus modeled on the LDP’s policy affairs
research committee but dismantled it after winning power, replacing
it with mechanisms designed to promote stronger political leadership.
One result, however, was that rank-and-file Diet members were
excluded from any involvement in policymaking, which led to confron-
tation and confusion within the party. The DPJ had criticized the LDP
decision-making process and could not simply imitate LDP institutions,
but they had no clear idea how to replace it. The decision-making rules
changed with every new prime minister, and they failed to develop any
stable procedure for reaching agreement (Maeda & Tsutsumi 2015).
Against a divided opposition, a party with less than half the vote
can win more than half of the seats if the electoral system is based on
94 Yukio Maeda and Steven R. Reed

single-member districts. This is the fundamental mechanism behind

Duverger’s Law, but it is wise to remember that Duverger’s Law
concerns elections, not the capacity of parties to govern. Italy offers
another clear example of Duverger’s Law working but resulting in
parties incapable of governing (D’Alimonte 2005). Constructing
a party capable of governing from the remains of a party system that
has disintegrated is not easy and takes time. Japanese opposition par-
ties are perfectly aware of the need to form a united front, but policy
and strategy differences continue to divide them. Until an opposition
party capable of presenting a credible challenge to the LDP emerges, the
LDP will continue to win “by default.”

4.2.4 The Organizational Vote

In their paper on the 2017 election, Horiuchi, Kuriwaki, and Smith
(2019, p.30) present preliminary evidence suggesting that the govern-
ing parties won many votes from those who did not support the party’s
policy positions. When voters are offered a choice between party or
coalitions, either of which has a realistic choice of winning control of
the government, the issue positions of the parties are among the most
important choice criteria. However, when only one party or coalition
has a realistic chance of winning, another criterion becomes important:
access to governmental policymaking.1
Interest groups have clear policy preferences but also a strong
interest in maintaining access to governmental policymaking. When
an interest group is confident that the same parties will be in power
after the next election, it may choose to support those parties
whose policies they oppose. Indeed, when that party or coalition
seems likely to govern for the foreseeable future, an interest group
may support it precisely because they oppose its policies: it is the
only effective way to oppose those policies. Supporting an oppos-
ition party may actually reduce a group’s voice in policymaking.
Organizations support the LDP not because they support LDP
policy but because they seek access to the relevant bureaucrats.
(Yomiuri Shinbun, March 18 and 19, 2019). In party systems with
alternation in power, interest groups find it wise to have a foot in
both camps. In party systems without alternation in power, how-
ever, interest groups find it wise to have both feet firmly planted in
the governing party camp.
The LDP under Abe 95

Between 2003 and 2009 the DPJ presented an increasingly credible

challenge and organizations were forced to consider having a foot in
both the LDP and DPJ camps. As a leader of the Japan Medical
Association (JMA) said, “In the Showa era the LDP was the only
choice. . . . In the Heisei era centrist parties emerged and some members
made overtures to the DPJ, . . . but I never stopped consulting the LDP.”
(Yomiuri Shinbun, March 27, 2019). However, after the consecutive
landslide victories and the fragmentation of the opposition since 2012,
the LDP under Abe has regained one of the primary advantages of
a predominant party: the virtual certainly that after the next election
the LDP–Komeito coalition will be in power. Because the opposition is
fragmented, there is no party threatening to take control of the govern-
ment and therefore no need to keep a foot in any of the several oppos-
ition party camps. After 2012, with no credible challenger in sight,
interest groups have again planted both feet firmly in the government
party camp. In fact, the LDP has made significant inroads into the labor
union vote, the traditional support base of the left. The LDP is now
responding to policy petitions (chinjō) offering unions some access to
governmental policymaking and, though data is sparse, it seems that
some union members seem to have been moving to the LDP. For
example, the Japan Electrical and Information Union (Denki Rengō),
a relatively conservative union, polled its members and found that, in
the 2017 election, more voted for the LDP than for the primary oppos-
ition candidate in their district (Yomiuri Shinbun, March 24, 2019).
Especially in elections that feature the LDP versus a single opposition
candidate, the opposition often campaigns on issues while the govern-
ment mobilizes organizations. Indeed, organizational campaigns seem
to be a way of avoiding issues. Organizational votes are thus a likely
place to look for those people who vote for the LDP despite preferring
the policies of an opposition party. Several DPJ candidates have also
been moving to the LDP for reasons similar to that of the organizations
already discussed: they want to participate in governmental policy-
making. Mitsuru Sakurai explained it clearly when he resigned from
the DPP: “Even if the party develops policies it cannot change society.
I want to work with the governing parties from an independent pos-
ition to make policies that change society” (NHK News Web,
September 30, 2019). Three DPJ SMD incumbents have joined the
LDP Nikai faction and are seeking to join the LDP. Three others who
have run for the DPJ in the past will be on the LDP PR list in the 2019
96 Yukio Maeda and Steven R. Reed

election, and one is running in the Iwate prefectural district (Yomiuri

Shinbun, March 25, 2019, June 11, 2019; Asahi Shinbun, June 24,
2019). None of the four who ran in the 2019 upper house election won
seats, which augurs poorly for those who will run in the next general
election.

4.3 How Does Abe Maintain High Cabinet Support Levels?

Before the electoral reform of 1994, prime ministers did not need to
worry about their approval ratings. Many successful prime ministers
had low support rates. For example, Kakuei Tanaka’s approval ratings
never went above 20 percent during the last twelve months of his
tenure, and Takeo Fukuda scored above 30 percent only four times.
On the other hand, Toshiki Kaifu was forced to step down even though
his approval rating at the time (September 1991) was 50 percent. In the
1970s, the golden age of factional struggles within the LDP, the average
approval rating for LDP prime ministers was 29.7 percent with
a standard deviation of 9.6, while that figure during the 2000s (includ-
ing the DPJ government) was 38.7 percent with a standard deviation of
14.4. Not only have average approval ratings risen 9 percentage points,
but rates have become more volatile and more important.
Popularity has become an important factor LDP MPs use when
selecting their president (who then becomes prime minister). Before
his first premiership, Abe stated that “recently poll results are published
every month. I cannot image I could be considered a candidate for LDP
president without support in the opinion polls” (Mainichi, May 25,
2006). And, as mentioned, low support rates can lead to short prime
ministerial terms. How did Abe manage to keep his support rates from
dropping like it had in his first administration?

4.3.1 Not Popular Policies

Popular policies did not explain Abe’s landslide victories. Neither do
they explain his cabinet support. Monthly Kyodo opinion polls ask if
people approve or disapprove of the cabinet and then ask respondents
to select from several options regarding the reasons why they either
approve or disapprove of.
Figure 4.2 presents the overall trajectory of Abe’s approval ratings.
Initially Abe enjoyed high approval ratings but gradually lost support,
The LDP under Abe 97

80
Approve
70 Disapprove
Don't know
60

50

40

30

20

10

0
Mar-13
Dec-12

Sep-13

Mar-14
Dec-13

Sep-14

Mar-15
Dec-14

Sep-15

Mar-16
Dec-15

Sep-16

Mar-17
Dec-16

Sep-17

Mar-18
Dec-17

Sep-18

Mar-19
Dec-18

Sep-19
Jun-13

Jun-14

Jun-15

Jun-16

Jun-17

Jun-18

Jun-19
Figure 4.2 Cabinet approval from December 2012 to October 2019
Source: Kyodo opinion poll (Tokyo Shinbun)

a typical pattern found for new prime ministers. The first serious drop
occurred in the summer and fall of 2015 when Abe pushed his contro-
versial national security–related legislation through the Diet amid the
strong opposition both inside and outside the Diet. However, his
support rebounded after three months. Abe was perfectly aware that
he would lose support for passing this legislation and thus refused to
accede to opposition party demands for an extraordinary session of the
Diet that would have kept the issue in the news. He simply waited for
the next regular session, which started in January of 2016, and the lack
of negative exposure seems to have allowed his approval ratings
to recover.
The next serious drop came during the summer of 2017. Diet deliber-
ations regarding licensing a new veterinary medicine department to the
Kake Gakuen, owned by PM Abe’s close friend, left the impression that
Abe’s friend was given a special treatment (Carlson & Reed 2018b,
pp.121–124). Then, just after the Diet session ended and during the
election campaign period for the Tokyo Metropolitan Assembly elec-
tion, Tomomi Inada, minister of defense and a close Abe ally, made
several gaffes which were absurd enough to call into question her
98 Yukio Maeda and Steven R. Reed

qualifications to serve as minister (Carlson & Reed 2018b, pp.113–115).

Abe himself made an unusually aggressive and inflammatory remark
against the anti-Abe demonstrators in a stump speech. The image of
“unfairness” and “arrogance” clearly contributed to the disastrous loss
for the LDP in Tokyo Metropolitan Assembly. The scandals appear to
have been one reason Abe chose to dissolve the Diet and call an election
in October 2017, but the most important reason was probably that the
opposition remained deeply divided, meaning that the LDP-Komeito
coalition was virtually certain to win another landslide under any cir-
cumstances (Pekkanen & Reed 2018, p.26). In any case, cabinet support
rebounded after the election.
The final dip in Abe’s approval occurred in the spring of 2018 was
also due to a scandal, the Moritomo Gakuen scandal, but it was more
the handling of these issues by the government than the scandal itself
that was questioned. In April 2018. Abe’s wife had developed ties to
a private school, and the owners, Mr. and Mrs. Kagoike, used their ties
to the first lady to strike a deal with the Ministry of Finance for the sale
of state-owned land for opening up a new elementary school (Carlson
& Reed 2018b, pp.117–121). The political impact of Moritomo scan-
dal dramatically increased once it was revealed that the Ministry of
Finance altered the public documents on a land deal with Moritomo
before submitting them to the Diet. The falsification of official docu-
ments by the Ministry of Finance to obscure Mrs. Abe’s involvement
and the subsequent testimony of Nobuhisa Sagawa, a high-ranking
bureaucrat of Ministry of Finance, pushed Abe’s approval ratings
down in March and April.
In the last two months shown in Figure 4.2, Abe’s approval rating
suddenly improved. This was the period in which Emperor Akihito
retired, ending the Heisei era, and was replaced by his son Naruhito,
starting the Reiwa era. The media was saturated with these events, and,
of course, Abe frequently showed up in the various ceremonies. The
media reports were overwhelmingly positive, and this may well explain
the rise in Abe’s popularity.
Like his electoral victories, Abe’s approval ratings rests on fragile
ground. Figure 4.3 shows the percentage of respondents that approve
of Abe for several specific reasons. Initially, more than one-quarter of
those surveyed endorsed Abe for his economic policy, “Abenomics.”
Abe spoke with confidence and did a good job of selling his policies,
and Abenomics probably helped the LDP win in 2012. However, the
The LDP under Abe 99

35
I trust Prime Minister
30 Prime Minister shows leadership
His economic policy is promising
25 His foreign policy is promising
There is no one suited to the job

20

15

10

0
Sep-13

Sep-14

Sep-15

Sep-16

Sep-17

Sep-18

Sep-19
Dec-12

Dec-13

Dec-14

Dec-15

Dec-16

Dec-17

Dec-18
Mar-13

Mar-14

Mar-15

Mar-16

Mar-17

Mar-18

Mar-19
Jun-13

Jun-14

Jun-15

Jun-16

Jun-17

Jun-18

Jun-19
Figure 4.3 Approval and reason for approval from December 2012 to
October 2019
Source: Kyodo opinion poll (Tokyo Shinbun)

approval based on his economic policy gradually eroded and never

climbed above 10 percent after May 2015. Abenomics looked good
in prospect, but, once the results began to come in, support waned. Abe
focused his 2014 campaign on Abenomics, but Horiuchi, Smith, and
Yamamoto (2018) show that the issue had no effect on support for
a manifesto. Similarly, approval for Abe based on his diplomacy has
been around 5 percent all through his second tenure. Since the second
half of 2015, Abe secured support from the public because people think
“there is no better alternative.” The wording may be vague enough to
allow different interpretations, but the choice “no alternative” is listed
last of six choices in a phone interview. Roughly half of Abe supporters
keep choosing this answer, which translates into one-quarter of all
respondents.
Turning now to the reasons for not supporting Abe, we find that,
whenever overall approval rating is high, there is no clear pattern for
the reasons given for disapproval. Once the approval rating falls below
the disapproval rating, however, the reason given for disapproval
becomes clear. Abe’s disapproval surged twice due to dissatisfaction
100 Yukio Maeda and Steven R. Reed

35
I can't trust Prime Minister
30 Prime Minister lacks leadership
His economic policy is not promising
His foreign policy is not promising
25
I don't like his personality (until March 2017)
He is not qualified for Prime Minister
20 (from April 2017)

15

10

0
Mar-13

Mar-14

Mar-15

Mar-16

Mar-17
Dec-12

Sep-13
Dec-13

Sep-14
Dec-14

Sep-15
Dec-15

Sep-16
Dec-16

Sep-17

Mar-18

Mar-19
Dec-17

Sep-18
Dec-18

Sep-19
Jun-13

Jun-14

Jun-15

Jun-16

Jun-17

Jun-18

Jun-19
Figure 4.4 Disapproval and reason for disapproval from December 2012 to
October 2019
Source: Kyodo opinion poll (Tokyo Shinbun)

with economic conditions, first in October 2014 and then in

April 2016. In each case, he was able to eliminate this concern by
postponing the scheduled consumption tax hike twice, before the
lower house election in November 2014 and then before the upper
house election in 2016.
When disapproval rises because of something other than the econ-
omy, the reason given for nonsupport is Abe himself. During the Kake
Gakuen scandal discussed already, a growing number of people disap-
proved of Abe because “I cannot trust the prime minister.” The same
dynamic appears again in April 2018 during the Moritomo Gakuen
scandal. At the height of the scandal, Abe’s disapproval rate went to
53 percent against approval of 37 percent (Tokyo Shimbun, April 16,
2018) and slightly less than one-third of those who disapproved did so
because they said they could not trust Abe. Abe’s support thus appears
vulnerable to scandals, even when there is no viable alternative on the
horizon.
The good news for Abe is that, between his first and second adminis-
trations, he learned how to minimize the damage caused by scandals. In
The LDP under Abe 101

his first administration, Abe stood by his scandal-tainted ministers even

after they had become clear liabilities. In the worst example, his minis-
ter of agriculture became the first sitting cabinet minister to commit
suicide. In his second administration, he forced ministers out as soon as
they became liabilities. Finally, the Akira Amari scandal was handled
so deftly that it did not affect his cabinet support at all (Carlson & Reed
2018a, p.155; Yomiuri, February 2, 2016). The bad news for Abe is
that, after the 2017 election, he returned to the standard LDP practice
of protecting his ministers (Carlson & Reed 2018b). He continued to
protect Defense Minister Inada long after she had become a liability,
and his handling of the two “Gakuen” scandals projected an image of
deception and abuse of power. We can be sure that more scandals will
occur in the future. When they do, the effect upon cabinet support will
depend as much on how Abe handles them as on the facts of the case.

4.3.2 Media Strategy

In his first administration, Abe started with high cabinet support that
quickly disappeared, and he was forced to resign after a year in office.
Like many politicians, Abe blamed negative media coverage for his
experience and came back to power with deliberate strategies for
dealing with the mass media (Osaka 2015). Abe pays close attention
to the content of news reports and has shown little tolerance toward
“biased” news stories that show him and his policies in a negative light.
The most important effect of his media strategy, however, has been to
reduce the impact of the media on elections. Whereas Koizumi won by
courting the media, Abe wins by suppressing it. The goal is issue-free,
low-information, low-interest, low-turnout elections that magnify the
effectiveness of a campaign based on organizational support.
Abe seeks appearances on television programs broadcast by stations
that are generally favorable toward him (AERA, March 7, 2016).
Japanese prime ministers generally do not appear on entertainment
programs, but in the spring of 2014 he appeared on the popular daily
TV show Waratte Iitomo (Yomiuri, March 22, 2014). He also showed
up as a special guest in a popular comedy performed in the theatre in
Osaka to promote his economic policy in April 2019 (Asahi, April 21,
2019).
Second, Abe and his close aides have tried to intimidate news anchors
and journalists who disseminate news reports thought to be “biased”
102 Yukio Maeda and Steven R. Reed

against him.2 Just before the general election of November 2014, Abe
appeared on a live TBS program and became indignant when the
station broadcast “man in the street” interviews suggesting that most
people did not feel like their lives had been improved by Abenomics
(jikkan ga nai). Even though most polls agreed with those “men in the
street,” Abe accused the station of selecting negative comments,
because his statistics showed that 60 percent of companies had raised
wages (Asahi Shinbun, November 29, 2014). On November 20, two
days after the incident, the LDP issued a document titled “Request for
fairness and neutrality of news reports during the election campaign
period” to the editors and directors of major TV stations. In the
document the LDP detailed specific procedures to make news reports
and programs “neutral.” The report had a clear chilling effect on TV
programming. The average daily airtime on election-related topics
during the 2014 campaign period was roughly half that of 2009 and
two-thirds that of 2012 (Osaka 2015, p.51). In addition, “man in the
street interviews” – a traditional repertoire for TV media reports –
disappeared from most news programs during the 2014 campaign
period (Mizushima 2015).
Abe also tries to set the agenda for each election and to shorten the
effective campaign period. In 2014 he held a news conference three
days before dissolving the Diet at which he announced that the election
would be “the Abenomics election.” The timing allowed him to appear
alone without sharing airtime with opposition leaders. In addition, he
minimized the time between dissolution and the beginning of the offi-
cial campaign period (Osaka 2015, p.49). The most interesting and
useful media reporting occurs during this period. Discussions before
dissolution are abstract and hypothetical. Discussion during the official
election period is tightly regulated. McElwain (2008) has shown that
both stricter regulations and shortening the official election period
work to the advantage of incumbents. Shortening the period between
dissolution and the official campaign period should do the same.
None of the main news outlets in Japan are as blatantly biased as
some of those found in the United States and Europe. Most notably,
even newspapers and television stations considered favorable to Abe
continue to report avidly on political and bureaucratic scandals. On the
other hand, none of the main news outlets possesses the kind of inde-
pendence from government interference enjoyed by such newspapers as
the New York Times and the Washington Post in the United States.
The LDP under Abe 103

4.4 How Has Abe Stayed in Power So Long?

For a prime minister to lead a party through consecutive victories as
Abe has done, he must keep his post as party leader for a long time. One
might think that winning elections and maintaining high support rates
would be sufficient to guarantee a long tenure. That would be the case
in most parliamentary documentaries, but it has not been the LDP
tradition. Winning elections and maintaining high support rates are,
of course, important, but they have not proven sufficient conditions for
maintaining a long tenure. The LDP has been a factionalized party
incapable of enforcing party discipline (Reed 2009). Between its found-
ing in 1955 until the 2000s, the sitting prime minister’s primary rival
was not a leader of the opposition but a member of his own party,
usually another faction leader. There were always rivals waiting in the
wings and looking for any opportunity to push the current prime
minister out and take their turn in the job.
Neither was the LDP capable of enforcing party discipline on votes in
the Diet. Diet members were able to campaign against and vote against
party policy with impunity. Policymaking was thus slow and the policy
enacted watered down by the multiple compromises that had to be
made to achieve consensus. Prime Minister Koizumi was the first LDP
prime minister to enforce party discipline on a Diet vote. In the 2005
election, many members of the LDP voted against Koizumi’s signature
policy of privatizing the post office. Koizumi managed to deny the LDP
nomination to these “postal rebels” and run LDP-nominated or -
supported candidates against these “assassins” (Reed & Shimizu
2009, pp.42–43). The result was a reversal of the recent losses and
a landslide victory for the LDP. One reason the next three prime
ministers were unable to hold onto Koizumi’s victory was that they
could not enact bold policies nor keep LDP candidates “on message”
during elections.
Abe was elected LDP president and prime minister in September of
2012. It was not an auspicious beginning. In a preliminary primary
polling, in which only party members vote, he finished second to
Shigeru Ishiba. Then, the final vote including members of the Diet
was indecisive so he was forced into a runoff. He did win, with 108
votes to 89, but he clearly did not start with a strong base in the party
nor with party members. Yet he led to the LDP to a landslide victory of
December 2012.
104 Yukio Maeda and Steven R. Reed

The key to Abe’s success in 2012 was simply timing. First, 2012 was
an election that the LDP was virtually certain to win, simply because
the DPJ government had been such a failure. Second, after “three years
and three months” out of power, a phrase oft repeated by Abe and his
supporters, he benefited from the determination to avoid another
“nightmare” of opposition parties in power. The three one-year LDP
prime ministers after Koizumi and the three one-year DPJ prime minis-
ters that followed, each voted out by a disastrous defeat, seemed to
confirm the importance of party unity. Abe thus benefited from an
unprecedented “honeymoon” period virtually free from internal
dissent.
The LDP was more unified and disciplined than it ever had been. Abe
used this period to begin a series of bold policy initiatives. These were
precisely the kind of initiatives that previously would have been bogged
down in the LDP’s internal consensus decision-making process, but
Abe also managed to construct a more top-down approach focused on
the prime minister’s office (the kantei). There was a great deal of
disagreement and dissatisfaction with this centralization of policy-
making, but backbenchers did not dare to voice them in public (Asahi
Shinbun, December 15, 2014). He also defied opposition from interest
groups, such as farmers, that had repeatedly blocked reforms in the
past (Machlachlan & Shimizu 2016; see also Chapter 14).
Abe also used his honeymoon period to undermine Ishiba, his pri-
mary rival within the party (Yomiuri Shinbun, March 29, 2019), and
acted to prevent new rivals from emerging. In 2015 Seiko Noda
planned to run against Abe simply to force an election, a tactic used
successfully in the past by Junichiro Koizumi. The claim was that an
election made the LDP look like a party open to debate, while running
unopposed made it look like decisions were made in the back room
behind closed doors. In the end, however, Noda was unable to obtain
the twenty endorsements from members of the Diet to qualify her to
run in the primary (Asahi Shinbun, 8 September 2015).
The fundamental secrets to Abe’s long tenure are, first, winning
elections; second, maintaining high levels of cabinet support; and,
third, preventing rivals from challenging him. He had an easier time
deterring rivals and maintaining party discipline than any prime minis-
ter before him, including himself in his own first term, as – after three
years and three months out of power – no one in the LDP wanted to
return to the nightmare of opposition parties in power.
The LDP under Abe 105

4.4.1 Prospects for the LDP after Abe

The LDP has been remarkably successful under Prime Minister Abe
since 2012. Will his successors be able to maintain this level of success?
First, there are several reasons to think that Abe’s successor will not be
able to maintain party discipline as well as Abe has. Second, there are
also good reasons to think that the opposition will not remain frag-
mented indefinitely, even though there is little hope that they can create
a viable alternative soon. Abe was blessed with an environment that
favored LDP success; his successors will not be so lucky. Instead of
starting with a guaranteed victory, they will follow one of the most
successful strings of victories in LDP history. Abe will be a hard act to
follow.
Abe’s successor will inherit a more centralized policymaking
apparatus and greater influence over cabinet selection and appoint-
ments to party posts. However, instead of facing backbenchers
fearful that any sign of disunity might result in another stint in
opposition, the successor will face rivals anxious for their chance
to lead the party and backbenchers frustrated by years of central-
ized leadership that left them with little to do, including many who
have never served as a governmental minister. In addition,
Secretary-General Nikai is actively recruiting candidates from the
opposition parties in a classic “if you win, you are LDP” strategy.
This strategy is guaranteed to frustrate any candidate who does not
receive the LDP nomination. Intraparty electoral completion is not
conducive to party unity.
If and when the opposition coalesces into a single viable alterna-
tive, the advantages of being a predominant party will fade. The
government will no longer be the only path to influencing govern-
mental policy, and all groups will feel the pressure to keep a foot
in both camps. Similarly, the media will be less vulnerable to
governmental intimidation. Finally, those candidates frustrated
with their role in the LDP – and especially those who failed to
get an LDP nomination – will find the option of running for an
opposition party more attractive.
Even if the opposition does not unify, scandals will inevitably
occur, hurting LDP popularity. If some event that the party cannot
keep off the front page focuses public attention on any one of the
LDP’s unpopular policies, that event could generate the voter
106 Yukio Maeda and Steven R. Reed

interest needed to raise turnout by drawing nonvoters out to vote

against the LDP. Currently the primary candidates for such an
issue appear to be constitutional revision, national security, and
restarting nuclear power plants. An event that makes Abe’s foreign
policy look like it might lead to war or another nuclear disaster
could also play this role.

Notes
1. Opposition parties have significant access to policymaking in many
democracies, but not in Japan.
2. For a list of incidents of the government or the LDP trying to influence the
content of news reports, see https://webronza.asahi.com/politics/articles/
2016100500005.html.

References
Asahi–Tōdai Kyōdō Chōsa (Asahi–University of Tokyo Joint Survey). 2019.
“Shuyō seitou ha dou hennka shita ka? (How have the main parties
changed?). Asahi Shinbun, June 13, 2019. www.asahi.com/senkyo/asa
hitodai/asahitodai15nen/?iref=pc_extlink.
Carlson, Matthew M., and Steven R. Reed. 2018a. Political Corruption and
Scandals in Japan. Ithaca, NY: Cornell University Press.
Carlson, Matthew M., and Steven R. Reed. 2018b. “Scandals during the Abe
administrations.” In Japan Decides 2017, edited by Robert J. Pekkanen,
Steven R. Reed, Ethan Scheiner, and Daniel M. Smith. New York:
Palgrave.
D’Alimonte, Roberto. 2005. “Italy: A case of fragmented bipolarism.” In
The Politics of Electoral Systems, Michael Gallagher and Paul Mitchell.
Oxford: Oxford University Press.
Hayao, Kenji. 1993. The Japanese Prime Minister and Public Policy.
Pittsburgh, PA: University of Pittsburgh Press.
Horiuchi, Yusaku, Daniel M. Smith, and Teppei Yamamoto. 2018.
“Measuring voters’ multidimensional policy preferences with conjoint
analysis: Application to Japan’s 2014 election.” Political Analysis 26(2):
190–209.
Horiuchi, Yusaku, Shiro Kuriwaki, and Daniel M. Smith. 2019. “Are
voters’ decisions consistent with their policy preferences?” Working
paper.
Kamikawa Ryunosuke. 2018. Denryoku to Seiji (Electric power and
politics). Tokyo: Keisho Shobo.
The LDP under Abe 107

Kushida, Kenji E., and Phillip Lipscy, eds. 2013. Japan under the DPJ.
Stanford, CA: The Walter H. Shorenstein Asia-Pacific Research Center.
Machlachlan, Patricia, and Kay Shimizu. 2016. “The Kantei vx the LDP:
Agricultural reform, the organized vote, and the 2014 election. In Japan
Decides 2014, edited by Robert J. Pekkanen and Steven R. Reed.
New York: Palgrave Macmillan.
Maeda, Yukio. 2018. “Public opinion and the Abe cabinet: Alternating
valence and position issues.” In Japan Decides 2017, edited by
Robert Pekkanen, Steven R. Reed, Ethan Scheiner, and Daniel
M. Smith. New York: Palgrave Macmillan.
Maeda, Yukio, and Hidenori Tsutsumi. 2015. Tōchi No Jōken: Minshutō Ni
Miru Seiken Un’ei to Tōnai Tōchi (Party government and party governance:
The case of the Democratic Party of Japan). Tokyo: Chikura Shobo.
McElwain, Kenneth Mori. 2008. “Manipulating electoral rules to
manufacture single-party dominance.” American Journal of Political
Science 52: 32–47.
Mizushima, Hiroaki. 2015. “Terebi Hodo no ‘Tsuyomi’ wo Fujita Abe
Jimin” (The LDP under Abe blocked the ‘strength’ of news reporting
on television) Janarizumu (Journalism) 305 (October 2015): 38–45.
Osaka, Iwao. 2015. “Terebi Hodo wo ‘Gakushu’ shita Abe Seiken no
Fukushugeki ga ‘Abenotahn’ no Honshitsu de Aru” (The essence of
“Abe’s turn” is revenge toward the mass media after learning the
pattern of reporting on television). Janarizumu (Journalism) 305
(October 2015): 46–54.
Pekkanen, Robert J., Steven R. Reed, and Ethan Scheiner. 2016.
“Introduction: Take a Second Look at the 2014 Election, It’s Worth
it.” In Japan Decides 2014, edited by Robert J. Pekkanen and Steven
R. Reed. New York: Palgrave Macmillan.
Pekkanen, Robert J., and Steven R. Reed. 2018. “Japanese politics between
2014 and 2017.” In Japan Decides 2017, edited by Robert Pekkanen,
Steven R. Reed, Ethan Scheiner, and Daniel M. Smith. New York:
Palgrave Macmillan.
Reed, Steven R. 2007. “Duverger’s Law is working in Japan.” Senkyo
Kenkyu (Electoral studies) 22: 96–106.
Reed, Steven R. 2009. “Party strategy or candidate strategy: How does the
LDP run the right number of candidates in Japan’s multi-member
districts?” Party Politics 15: 295–314.
Reed, Steven R., and Kay Shimizu. 2009. “Avoiding a two-party system: The
Liberal Democratic Party versus Duverger’s Law.” In Political Change
in Japan, edited by Steven R. Reed, Kenneth Mori McElwain, and
Kay Shimizu. Stanford, CA: The Walter H. Shorenstein Asia-Pacific
Research Center.
108 Yukio Maeda and Steven R. Reed

Reed, Steven R., Ethan Scheiner, Daniel M. Smith, and Michael F. Thies.
2013. “The 2012 election results: The LDP wins big by default.” In
Japan Decides 2012, edited by Robert Pekkanen, Steven R. Reed, and
Ethan Scheiner. New York: Palgrave Macmillan.
Schäfer, F., S. Evert, and P. Heinrich. 2017. “Japan’s 2014 general election:
Political bots, right-wing Internet activism, and Prime Minister Shinzō
Abe’s hidden nationalist agenda.” Big Data 5(4): 1–16. doi: 10.1089/
big.2017.0049.
5 The Third Arrow of Abenomics: Est.
in 2013 – or 2007?
t a k a t o s h i it o

5.1 Introduction
Abenomics, the package of economic policies adopted by Prime
Minister Shinzo Abe since 2013, is commonly credited with bringing
the Japanese economy out of deflation and two decades of stagnant
growth. Abenomics consists of three so-called arrows: (1) aggressive
monetary policy; (2) flexible fiscal policy; and (3) structural reforms for
growth. The first two arrows were quite visible in 2013–2014, lifting
the inflation rate from negative territory to positive territory and rais-
ing the growth rate above potential. The positive inflation rate and
above-potential growth rate have been maintained since then, lowering
the unemployment level to very low levels even with a higher participa-
tion rate for women and the elderly. The labor shortage has become
widespread by 2018.
This chapter is about the third arrow: structural reforms for growth.
Many people believe that, in contrast to the first two arrows, the
progress of the third arrow has been nonexistent or at best very slow.
A common excuse is that structural reforms take time. However, some
blame Shinzo Abe for lack of decisive leadership.
In this chapter I will argue that, first, contrary to the common belief,
several parts of the third arrow have achieved prominent successes; and
that, second, they originate in the economic policies of the first Abe
administration of 2006–2007, followed by the Fukuda administration
of 2007–2008. Important policy discussions and decisions were made
by the Council of Economic and Fiscal Policy (CEFP) of 2006–2008.
Structural reforms take time, indeed. For many achievements in the
third arrow after 2013, seeds were sown in 2007; frozen during the
opposition party’s rein in 2009–2012; then brought back to life in
the second Abe administration that started in December 2012.
The following lists success cases of Abenomics third arrow:

109
110 Takatoshi Ito

(1) EPA/FTA and agriculture promotion;

(2) re-internationalization of the Haneda airport;
(3) increasing foreign visitors to Japan;
(4) labor market reform: womenomics and regular workers vs. non-
regular workers;
(5) pension fund reform; and
(6) consolidation of stock/commodity exchanges.

5.2 Chronology and Institutions

After a decade of economic stagnation, the Liberal Democratic Party
(LDP) elected Junichiro Koizumi, an intraparty critic of the main-
stream, as a party leader. He became prime minister in April 2001,
with a slogan that he would “destroy” the LDP that voted for him.
What he meant was that he would fundamentally reform the base of
LDP supporters in order to promote deregulation and structural
reforms. He had prominent success in economic policies, including
ending a prolonged banking fragility; partially privatizing the post
office system including banking and insurance; and fundamentally
reforming the pension system. Koizumi was a popular prime minister,
winning both lower and upper house elections and serving for more
than five years.
In 2006, Koizumi handed down the LDP leader position to Shinzo
Abe, who served as a cabinet secretary under Koizumi. As the LDP had
a majority in the lower house that decides prime minister, Abe became
prime minister in September 2006, at the age of fifty-two. Abe inherited
a majority in both the lower and upper houses with a comfortable
margin. However, several political missteps made him unpopular,
and he lost the upper house election in July 2007. He abruptly resigned
as prime minister in September 2007, citing poor health. The period of
his first cabinet in 2006–2007 is considered a failure as he exited only
after one year. Yasuo Fukuda, senior LDP member, succeeded Abe as
prime minister but lasted only one year with any major achievement,
due to opposition parties’ majority in the upper house. Taro Aso
succeeded Fukuda in the middle of the global financial crisis. He
directed large fiscal stimulus, making half of the annual budget
financed by new issues of bonds. However, he had to go to a general
election as the four-year term was approaching the maturity.
The Third Arrow of Abenomics: Est. in 2013 – or 2007? 111

The Democratic Party of Japan (DPJ) became a largest party in

the lower house and formed the government. However, the three
years that the DPJ held the government, with three different prime
ministers, Yukio Hatoyama, Naoto Kan, and Yoshihiko Noda, was
marred by mismanagement in diplomacy, ineffective monetary and
fiscal policies, lack of growth policies, and mishandling of the East
Japan Great Earthquake and the Fukushima nuclear disaster. Both
the inflation rate and the growth rate sank into negative territory.
The DPJ turned out to be no better than the LDP, having three
prime ministers in three years. Thus, prime ministers in 2006 to
2012 were all short-lived, resulting in six prime ministers in six
years.
Voters gave the LDP a landslide victory in the December 2012
election. Shinzo Abe became prime minister, winning leadership of
the LDP in a September 2012 party election. The Abe administration,
since December 2012, turned out to be a political and economic suc-
cess. The growth rate and the inflation rate were lifted to positive
territory, and the unemployment rate continued to decline. Abe’s
approval rating remained above 40 percent. In November 2019, Abe
has become the longest-serving prime minister in the history of the
parliamentary cabinet system since 1885.1
Table 5.1 gives a brief history of prime ministers since April 2001 as
background to Shinzo Abe’s second term as prime minister.
The success of his long tenure is arguably based on the success of
Abenomics. So, the question is how Abenomics was formed and imple-
mented. I argue that most of the third arrow policies, announced in
June 2013, were discussed and brewed in the first Abe administration,
not in the just-born second Abe administration.
The driving force of the economic policies in the first Abe adminis-
tration was the Council of Economic and Fiscal Policy (CEFP). The
CEFP as an institution was created in January 2001, just three months
before Koizumi became prime minister, as a part of administrative
reform. The CEFP, chaired by the prime minister, includes the cabinet
secretary; ministers of economy, finance, METI, and internal affairs;
and the governor of the Bank of Japan, as well as two academics and
two business leaders. The two academics and two business leaders form
a group of private-sector committee members, which produces sugges-
tions for reform. By law, the CEFP meets at least once a month, and in
practice more than once.
112 Takatoshi Ito

Table 5.1 Prime ministers since 2001

Year/month/day Prime minister Party Days in power

April 26, 2001– Junichiro Koizumi LDP 1980

September 26, 2006
September 26, 2006– Shinzo Abe LDP 366
September 26, 2007
September 26, 2007– Yasuo Fukuda LDP 365
September 24, 2008
September 24, 2008– Taro Aso LDP 358
September 16, 2009
September 16, 2009–June 8, Yukio Hatoyama DPJ 266
2010
June 8, 2010–September 2, Naoto Kan DPJ 452
2011
September 2, 2011– Yoshihiko Noda DPJ 482
December 26, 2012
December 26, 2012– Shinzo Abe LDP 2822
September 16, 2020

Source: www.kantei.go.jp/jp/rekidainaikaku/.
1 As of November 20, 2019, Abe’s cumulative days, including the earlier 366 days,
became 2,887, surpassing a former record holder, Taro Katsura. On August 24, 2020,
Shinzo Abe became the longest uninterrupted prime minister with 2,799 days,
surpassing the former record holder Eisaku Sato.

Heizo Takenaka, an academic, became the minister in charge of

CEFP under Koizumi from April 2001 to October 2005. During his
leadership the CEFP achieved several major structural reforms.
Koizumi used the CEFP effectively to carry out what he thought
to be most important for revitalizing the Japanese economy.
One of CEFP’s tasks was to provide a framework for fiscal policy by
approving the “Basic Principles of Economic Policy” before the budget
process took place every year. In the early years of CEFP, this was
a contentious issue, because the budget process was solely owned by the
Ministry of Finance. However, soon it came to be understood that the
CEFP only sets the trend of the total budget, establishes a budget ceiling,
or targets the deficit level without discussing details of revenues or expend-
iture. This is dubbed in Japanese Honebuto no Hoshin, or literally a thick
skeleton guidance. The report is published in June every year.
The Third Arrow of Abenomics: Est. in 2013 – or 2007? 113

Abe in 2006 appointed Hiroko Ota as minister in charge of CEFP,

and the academic and business members were recruited anew.2 When
the 2006 CEFP was gathered for its first meeting, the members identi-
fied seven areas important for growth:3
(1) Reforms for globalization: promoting EPA; strengthening agricul-
ture; establishing open financial/capital markets;
(2) Labor market reforms: reform to pursue deregulation, equal oppor-
tunity, and corporate vitality at the same time; foster human capital;
increase mobility and diversity; assist second-chance employment;
(3) Productivity reforms: increasing productivity of the service sector
by deregulation; increasing productivity of manufacturing sector,
esp. SMEs by industry–academic alliance; increasing productivity
through innovation, intellectual property, and IT;
(4) Tax reforms: Increase competitiveness of the Japanese economy by
reforming corporate tax; fundamental reform of taxation system;
(5) Decentralization reforms: Enhance decentralization by shifting
mandates and responsibilities to prefectural and local governments;
grouping prefectures into regions; comprehensive reforms of local
allocation tax, subsidies, and local government bonds; assist pri-
vate-sector efforts to revitalize local communities;
(6) Social security reforms: redesign medical and long-term care mar-
kets by electronic record, deregulation, and reducing costs; redesign
nursery; make the social security system sustainable; increase trans-
parency and reliability by reforming social security agency;
(7) Government reforms: bureaucracy reform; administrative reform;
special account fiscal reform; public investment reform; reform of
independent administrative agencies; reform of government asset
and liability management.
These seven areas were studied and discussed in subsequent CEFP meet-
ings in 2006–2007 under the Abe administration and in 2007–2008
under the Fukuda administration.4 For a few important topics, the
CEFP created expert committees and working groups and issued detailed
reports.
On topic of globalization, the following committees and reports are
relevant to the discussions in the later sections of this chapter.
➢ The working group on “EPA and Agriculture” held meetings seven-
teen times between January 2007 and March 2008. It issued the
114 Takatoshi Ito

report “Accelerate EPA negotiation and strengthen agriculture

reforms” in May 2007.5
➢ The working group on “Financial and Capital Markets” held meet-
ings seventeen times between January 2007 and May 2008. It issued
two reports in April 2007 and May 2008.6
➢ The expert committee on the “Labor Market Reform” held twenty-
four meetings from December 2006 to September 2008 and pub-
lished four reports.
After the two years of activities in 2006–2008, the members appointed by
Abe were replaced under Prime Minister Taro Aso in September 2008,
but, when the DPJ took power by winning the general election of 2009,
the CEFP members were fired and its meetings suspended.7 The CEFP
homepage was shut down, although the law that established and
gave mandate to the CEFP was not abolished. During the DPJ years
(2009–2012), no CEFP members were appointed, its meetings were
suspended, and its website disappeared.
When Shinzo Abe returned as prime minister, he revived the CEFP.
In addition, he created another task force to discuss the “Japan
Revitalization Strategy.” The latter’s report became the platform to
announce the plan and to check progress of the third arrow. Broadly
speaking, policy measures and targets announced in the Japan
Revitalization Strategy are mostly supply-side policies: increasing
labor input, increasing the efficiency of capital, and promoting innov-
ation. The strategy can be seen as being aimed at raising the potential
growth rate. There are few demand-side policies. The report of Japan
Revitalization Strategy has been published annually in June each year
since 2013.8 The report of 2013 (Japan Economic Revitalization
Headquarters 2013) ran to ninety-four pages, with details of many
policy measures.
Soon after Abe became prime minister for the second time in
December 2012, he launched a package of economic policies widely
known as “Abenomics.”9 Abenomics consists of three arrows: aggres-
sive monetary policy; flexible fiscal policy; and growth strategies to
stimulate private-sector investment. In the second Abe administration,
“Basic Principles of Economic Policy” deals with the first and second
arrows while the “Japan Revitalization Strategy” deals with the third
arrow; that is to say, the former covers aggregate demand policies,
while the latter covers supply-side policies.
The Third Arrow of Abenomics: Est. in 2013 – or 2007? 115

5.3 FTA Promotion and Strengthening Agriculture

Prime Minister Abe, after becoming PM again in December 2012,
advocated for free trade. His contribution to Japan’s free trade
is four-fold. First, he took leadership and persuaded the skeptics
in the LDP to join in the negotiation toward agreement of the Trans-
Pacific Partnership (TPP).10 This was announced on March 15,
2013.11 Later, the TPP agreement was concluded and signed in
February 2016. Second, once President Trump withdrew from the
already signed TPP in January 2017, Abe took a leadership role in
regrouping the remaining eleven countries and concluding the TPP11
treaty (officially the Comprehensive and Progressive Agreement for
Trans-Pacific Partnership (CPTPP)) in March 2018 and implemented
in December 2018. Third, the Japan–EU Economic Partnership agree-
ment was signed in July 2018 and implemented in February 2019.
Fourth, the trade agreement between Japan and the United States of
America was negotiated successfully in October 2019 and imple-
mented in January 2020. The last three achievements are quite
remarkable in light of an increasing global wave of protectionism
led by the United States.12
I will argue that the pro-free trade stance of Prime Minister Abe is
based on the study and recommendation of the 2006–2008 CEFP,
which created a working group to promote free trade and strengthen
agriculture. It is well known that promoting more trade – both exports
and imports – is key for growth, but the agricultural sector has a strong
lobby against free trade. The working group used logic to promote both
free trade and agriculture.
When Abe formed the first cabinet in September 2006, Japan had
only three implemented Economic Partnership Agreements (EPAs),
with Singapore, Mexico, and Malaysia, although several others were
in the negotiation stage (see Table 5.2). In the two years when Abe and
Fukuda served as prime minister, five EPA/FTAs were signed, of which
four – with Chile, Thailand, Indonesia, and Brunei – were imple-
mented. In the same two years, negotiations started with four other
countries: Switzerland, Vietnam, India, and Australia. One with Peru
started negotiation in May 2009, when Taro Aso was prime minister.
In Table 5.2, the 2006–2008 shade shows the contribution of the CEPF
under Abe and Fukuda. The 2013–2019 shade shows the contribution
of the Abenomics third arrow after 2013.
Table 5.2 Japan’s FTAs (signed and in effect), as of January 2020

Start of negotiation/first meeting Share in Japan’s total

of negotiation Signed Effective trade (%)

1 The Japan–Singapore Economic Jan. 2001 Jan. 2002 Nov. 2002 Incl. in ASEAN
Partnership Agreement1
2 Japan–Mexico FTA2 Nov. 2002 Sep. 2004 Apr. 2005 1.21
3 Japan–Malaysia EPA Jan. 2004 Dec. 2005 Jul. 2006 Incl. in ASEAN
4 Japan–Chile EPA Feb. 2006 Mar. 2007 Sep. 2007 0.62
5 Japan–Thailand EPA Feb. 2006 Apr. 2007 Nov. 2007 Incl. in ASEAN
6 Japan–Indonesia EPA July 2005 Aug. 2007 Jul. 2008 Incl. in ASEAN
7 Japan–Brunei FTA June 2006 Jun. 2007 Jul. 2008 Incl. in ASEAN
8 Japan–ASEAN EPA April 2005 Apr. 2008 Dec. 2008 15.25
9 Japan–Philippines EPA Feb. 2004 Sep. 2006 Dec. 2008 Incl. in ASEAN
10 Japan–Switzerland EPA February 2009 September 2009 0.78
May 2007
11 Japan–Vietnam EPA January 2007 December 2008 October 2009 Incl. in ASEAN
12 Japan–India EPA January 2007 February 2011 August 2011 1.11
13 Japan–Peru EPA May 2009 May 2011 March 2012 0.21
14 Japan–Australia FTA April 2007 4.23
July 2014 January 2015
15 Japan–Mongolia EPA June 2012 February 2015 June 2016 0.04
16 Trans-Pacific Partnership3 March 2010 February 2016
4
July 2013
17 Comprehensive and Progressive November 2017 March 2018 December 2018 1.775
Agreement of Trans-Pacific
Partnership (TPP11)
18 Japan–EU Economic Partnership April 2013 July 2018 February 2019 11.53
6
19 Japan–US Trade Agreement on Goods April 2019 October 2019 January 2020 14.91

Total share of trade as of 19.18

December 2012: FTA 1–13
Total share of trade with FTA 1–18 36.75
Total share of trade including 51.66
Japan–US

1 The EPA with Singapore was revised in September 2007.

2 The FTA with Mexico was revised in April 2012.
3 The United States withdrew from the TPP in January 2017; the rest of the eleven countries agreed to implement the revised agreement under
the new name.
4 Japan participated in the ongoing TPP negotiation from the eighteenth negotiation meeting.
5 The trade ratio in TPP11 includes Canada (1.42 percent) and New Zealand (0.35 percent).
6 Technically speaking, the Japan–US Trade Agreement on Goods is not considered to be a free-trade agreement, as it does not cover service
trade. It is not in the list of FTAs and EPAs in the list maintained by the Ministry of Foreign Affairs.
Source: Ministry of Foreign Affairs, www.mofa.go.jp/policy/economy/fta/index.html; www.mofa.go.jp/mofaj/files/000490260.pdf; www
.mofa.go.jp/mofaj/ila/et/page23_002886_00001.html.
118 Takatoshi Ito

The first report of the Working Group on Economic Partnership

Agreements and Agriculture, published May 8, 2007 (CEFP 2007a),
was titled “Acceleration of Economic Partnership Agreements and
Strengthening Agricultural Reform.” The title set the innovative tone
for the 2006–2008 CEFP. First, the juxtaposition of EPA and agriculture
in the same working group was unusual, yet the logic to do so was very
clear. The agricultural lobby was well known to be a strong resistance to
movement toward free trade, so it was essential to get the agricultural
lobby on board – or at least not disrupt free trade policy. Traditionally,
free trade had been discussed and pushed by the Ministry of Economy,
Trade, and Industry (METI) and the Ministry of Foreign Affairs
(MOFA), while resistance was formed by the Ministry of Agriculture,
Forestry and Fisheries (MAFF). The CEFP working group was able to
discuss the issues that cut across these ministry boundaries.
Second, the traditional solution to the agricultural resistance had
been to provide an additional financial assistance. The typical
example was the lump-sum subsidies to the agricultural sector when
Japan agreed to concluding the Uruguay Round of GATT. Japan
resisted tariffication of rice imports, as it may result in large imports
of rice. A compromise between Japan and other countries was that
Japan agreed to increasing minimum access (guaranteed purchase of
certain amount of rice) as an alternative to tariffication. The minimum
access rice would be imported by the government agency but not to be
released to the high-quality rice market in Japan. Most of the min-
imum accessed rice would end up decaying in the warehouse before
being sold as animal feed or assistance to developing countries. Even
with this arrangement, the agricultural lobby demanded compensa-
tion. The total compensation exceeded 6 trillion yen distributed over
seven years starting in 1995, of which 500 billion yen was disbursed in
the two years following the decision – this latter amount made up
12.5 percent of the annual budget of the Ministry of Agriculture,
Forestry, and Fisheries. However, it was determined later the 6 trillion
yen was not efficiently used – if not wasted – on unproductive infra-
structure and building projects. When a discussion started in the
working group, it was made clear by several members that using
subsidies to buy off the agricultural lobby is not an option. Instead,
the committee considered deregulation and structural reforms to
raise agricultural productivities. Scale economies should be exploited
and branding should be important. With these reforms, several
The Third Arrow of Abenomics: Est. in 2013 – or 2007? 119

agricultural products may not only withstand tariff reduction but

export products abroad.
The CEFP report in May 2007 had two major sections, one on
promoting free trade by recommending to strengthen economic part-
nership agreements and another on strengthening agriculture:
The prompt implementation of a Japanese EPA strategy involves the matter
of adjustment in the agriculture sector, which has already become a major
issue. There is no question that immediate liberalization under present cir-
cumstances would be likely to incur large adjustment costs. However, if
liberalization is carried out according to a phased schedule at the same time
that structural reforms are made in the agriculture sector, then it would be
possible to strengthen the competitiveness of the agriculture sector while
keeping adjustment costs at a low level. If the implementation of EPAs
simultaneously with structural reforms can realize a competitive agriculture,
then using EPAs as leverage to open up the markets for agricultural products
in partner countries could lead to expand exports of agricultural products
from Japan. (CEFP 2007a)

The report argued for the urgency of promoting FTAs with Asian
countries, Australia, and the United States. If Japan lags behind other
countries, such as Korea, it suffers not only from the opportunity costs
of not reaping benefits of free trade but also from diversion of Japanese
exports to other countries that reach FTA agreements with the United
States and Europe. However, the report recognized that an FTA with
countries with strong agricultural sectors will have a large impact on
Japanese agricultural sectors. The report emphasized the importance of
having high-quality FTAs, which has high liberalization ratios.
Exemption to zero tariff should be limited to a small set of agricultural
goods. However, the report suggests a scheduled gradual reduction of
tariff rates in general, in order to give time to implement structural
reforms. What is emphasized is the importance of structural reforms,
not subsidies. Specifically, the report recommends the following actions
in negotiations for FTAs.
On the tariff side, the report recommends: lowering tariff rates to
zero for goods that already have very low tariff rates; prohibitively high
tariff rates can be lowered without influencing effective protection; the
“Gate Price System for pork imports” should be changed to regular ad
valorem system; and substituting the minimum access of rice to
a regular tariff system.
120 Takatoshi Ito

On the agriculture side, the report pointed out that the agricultural
sector faces very serious structural problems. Lots of cultivated land
has been abandoned because young people, potential heirs, are not
continuing family farms. Farmers are aging, and the abandoned land
problem is only getting worse. Thus, domestic production is declining.
The experts in the committee reckoned that policy must be geared
toward increasing productivity of agriculture by realizing scale
economies. If land plots that are abandoned or potentially abandoned
can be sold or leased to viable farmers, then scale economies may
be realized. As proposals, the report recommended: to establish
a workable farmland lease system and fixed-term tenant rights; to
require information disclosure and to promote of farmland intermedi-
ation by the private sector, instead of the decision by a local agricultural
committee that consists of nearby farmers; to create an environment for
fair competition with agricultural cooperatives; and to create a support
system for core farmers.
The committee believed that agricultural cooperatives enjoy
a regional monopoly power due to not competing with each other
and that they exercise dominant positions by requiring farmers to
have exclusive relationships ranging from purchasing fertilizers and
feeds to selling products to cooperatives to financial transactions. The
last proposal, a support system for farmers, was intended to help
stronger farmers, for example by introducing a reform to realize expan-
sion of the scale.
Did these recommendations of 2007 get realized? As of
January 2020, Japan has eighteen FTAs plus the Japan–US Trade
Agreement on Goods, which may not be regarded as a comprehensive
FTA, as shown in Table 5.2. Among the eighteen, six were signed
during the 2006–2008 CEFP or the Abe–Fukuda period; and five
more were signed after 2013.
Before Abe became prime minister in December 2012, there were
thirteen FTAs signed and implemented, with trade coverage of 19 per-
cent. With FTAs with Australia and Mongolia, trade coverage went
up to 23 percent. In 2018–2019, the Trans-Pacific Partnership
Agreement without the USA (TPP11) and the Japan–EU agreement
were implemented. This brought trade coverage to 37 percent. With
the Japan–US TAG, trade coverage is now 52 percent. Japan came
a long way in becoming a country with more than half of trade
covered by FTAs.
The Third Arrow of Abenomics: Est. in 2013 – or 2007? 121

(100 billion yen)

10,000
Target for 2019:1 trillion
9,000 Agricultural Products Forestry Products
8,000 Fishery Products

7,000
6,000
5,000
4,000
3,000
2,000
1,000
0
1990
1991
1992

1994
1995
1996
1997
1998
1999
2000
2001
2002
2003

2005
2006
2007
2008
2009
2010
2011
2012
2013
2014

2016
2017
2018
1993

2004

2015
Figure 5.1 Agricultural, forestry and fishery product: exports
Source: Ministry of Agriculture, Forestry anf Fisheries

The increase in FTAs under the Abe administration after

December 2012 has been accompanied by a conspicuous absence of mas-
sive subsidies to the agricultural sectors. Many of the recommendations on
tariffs from the 2007 report have been realized. For example, the Gate Price
System for pork has been significantly modified in TPP11 and the Japan–
EU EPA.
The promise of making agriculture stronger was actually supported
by efforts to export agricultural products (see Figure 5.1).
The 2006–2008 strategy to make agriculture stronger as well as push
FTAs without massive agricultural subsidies has been at least partly
realized.

5.4 Haneda Airport Re-internationalization

Haneda airport used to be the main airport of Tokyo, handling both
domestic and international flights after World War II. As demands for
air services grew, the government decided to build a new airport. After
a few years of deliberation, a location – which was not on the short list –
60 kilometers away from central Tokyo was chosen in 1966. However,
opposition from farmers around the proposed airport site meshed with
122 Takatoshi Ito

radical students protesting in university campuses at the time. With

violent oppositions and protests, which claimed some casualties, the
airport’s construction was delayed by several years. It eventually
opened on May 20, 1978, with one runway and a curfew on operation
hours – a compromise in the face of complaints of noise during take-
offs and landings. The second runway was not built until 2002, due to
continuing opposition.
Since the opening of Narita airport, the government introduced
a strict demarcation: all international flights are handled at Narita,
with none allowed at Haneda.13 However, Narita was, and still
is, a very inconvenient airport. It takes a long time to reach the central
part of Tokyo; an airport curfew limits the flexibility of flight schedul-
ing; feeder traffic from other cities is small; and having only one
runway, and later with the second shorter runway, it is impossible for
the airport to grow into an Asian hub for international connections.
Haneda airport, as demands for domestic travels grew, expanded its
capacity in the 1980s and 1990s. The third runway was built over
landfill in the sea. By moving runways toward the water, noise from
take-offs and landings were minimized for the neighboring residential
area, and thus the airport became operational twenty-four hours a day.
The fourth runway was planned in early 2000s, and construction began
in 2007, later being completed in 2010.
With this background, CFEP members wanted to reform aviation
policy: to re-internationalize Haneda in order to make Tokyo an
attractive city for an international finance hub; and to adopt an
“open sky” for international flights (except for congested Narita and
Haneda airports) in order to increase competition and distribute
benefits to local airports.
The fourth runway would increase the capacity of Haneda airport
such that international flights can be introduced without sacrificing
domestic needs. As the proposition of Haneda re-internationalization
first surfaced, a strong opposition emerged. One part of opposition is
just bureaucratic inertia, as changing a policy may suggest that policy
was failing. Efforts to improve the Narita airport by buying oppos-
ition-owned land may also be hampered: if Haneda is taking busi-
nesses away from Narita, why continue buying land? Another point
of opposition came from those airlines that had disproportionately
large landing slots at Narita.14 The incumbent vested interests always
oppose reforms.
The Third Arrow of Abenomics: Est. in 2013 – or 2007? 123

The first goal of reformers was to have Haneda re-internationaliza-

tion listed as a topic in the Basic Policies of 2007, published on June 19,
2007. An important change in policy can happen if it is written in the
Basic Policy, a sort of annual report of CFEP activities. Re-
internationalization of the Haneda airport could trigger an important
change in aviation policy. Hence, in deliberations toward the Basic
Policy of 2007, vigorous debate took place both in CEFP meetings
and behind the scenes.
Those opposed to a drastic change in aviation policy wanted to
minimize the expansion of flights at Haneda airport, both in distance
to destination and regularity. They argued the principle of demarcation
of Haneda for domestic flights and Narita for international flights
should not be changed. They wanted to limit the international flights
at Haneda to short-distance destinations and “charter” flights.
Reformers countered the bureaucratic inertia argument by pointing
out that the Narita airport had a curfew (10pm–6am), while the
Haneda airport did not. Allowing Haneda to have international flights
to any destination during late-night and early-morning hours would
complement rather than displace Narita. Businesspeople value highly
a city with good and quick access. Only several international flights can
be added at Haneda, so value-added destinations should be chosen.
Opening Haneda would make businesspeople attracted to doing busi-
ness in Tokyo, it was felt – reformers had in mind destinations like
New York, Paris, and London. In addition, since Haneda is connected
to almost all domestic airports with high-frequency services, having
international flights would increase the convenience of non-Tokyo
residents who travel abroad and vice versa.
In the end, the Haneda re-internationalization section of the Basic
Report 2007 was formed only after heated debate in the 2006–2008
CEFP, with some discussions being recorded in the transcript and some
not.
In chapter 2 of the Basic Report 2007 (CEFP 2007b),
“Enhancing growth potential,” there is a subsection on “Reform
addressing globalization.” Three points were emphasized: pushing
EPAs with more countries; reforming financial and capital markets
to build an international financial center; and promoting the
“Asian open skies” and internationalizing Haneda airport. The
CFEP Basic Report 2007 (CFEP 2007b, p.17) was written as
a compromise between reformers and resistance:
124 Takatoshi Ito

For Haneda Airport, regular “charter” flights to destinations including

Europe and the United States will be allowed for the late-night hours
(23:00 to 06:00). . . . After 2010, 30,000 regular international flights a year
will be added. For destination, not only distance, but demand and import-
ance of routes will be judged. Considerations will be given to destinations
that fit as counterparts of Haneda, with a priority to closer distance ones,
depending on negotiations of air treaty. (p.17; emphasis added)

Although distance is mentioned, the phrase “destinations that fit as

counterparts of Haneda” clearly implies destinations such as
New York, Paris, and London. Reformers won out on this point.
However, a final battle between reformers and resistance had to be
fought in 2008. On May 20, 2008, Minister Fuyushiba of the Ministry of
Land, Infrastructure, Transport, and Tourism was invited to participate in
the CFEP meeting (CFEP, May 2008). CFEP members pressed the minis-
ter that destinations for Haneda’s daytime international flights, after its
capacity would be increased by 40 percent in 2010, should include
business centers around the world as well as capital cities in Asia. CFEP
members also proposed promotion of low-cost carrier (LCC) entry to the
Tokyo area. CFEP members also suggested the open sky negotiations
should be started between Japan and the EU and between Japan and the
United States. Lastly, in order to attract more Asian tourists to Japan,
CFEP members recommended that visa requirements be loosened.
Minister Fuyushiba emphasized that an understanding of Chiba
prefecture is necessary in pushing internationalization of Haneda air-
port, because flight paths for landing and take-off at Haneda go over
Chiba prefecture, while dirt and sand for landfill on the fourth runway
also came from Chiba prefecture. Chiba prefecture also suffered pains
to build Narita airport. This reasoning was expected. As for the destin-
ation, Minister Fuyushiba offered that, during daytime, Seoul,
Shanghai, Beijing, Taipei, and Hong Kong can be considered. For the
late-night hours, he mentioned that flights to Europe and the United
States would be allowed. The logic of reformers won the case.

5.5 Foreign Tourists to Japan

One of the big successes in the third arrow category is the increase in
incoming foreign visitors to Japan – “inbound tourism.” Spending by
foreign tourists in Japan, counted as goods and service exports, was
identified early on as an extra boost for increasing aggregate demand. In
The Third Arrow of Abenomics: Est. in 2013 – or 2007? 125

Million Persons
35
Abenomics
30 Visit Japan

25

20

15

10

0
2009
2003
2004
2005
2006
2007
2008

2010
2011
2012
2013
2014
2015
2016
2017
2018
2019
Figure 5.2 Visitor arrivals, international
Source: Japan National Tourism Organization

2003, Prime Minister Junichiro Koizumi announced the Visit Japan pro-
gram, designed to increase visitors to Japan.15 He aimed for 10 million
visitors by 2010, double the number of visitors in 2003. However, the
number of visitors remained below 10 million until 2013.16 Figure 5.2
shows the number of international visitors to Japan over time. There was
no sense of being successful in promoting foreign visitors to Japan until
2013.
Since then, however, the number of visitors has increased sharply,
from 10 million in 2013 to almost 20 million in 2015 to more than
30 million by 2018. There were several factors behind this success.
First, the growth strategy of Abenomics had a component of promot-
ing foreign tourism in Japan. However, there were very few concrete
policy measures associated with the announced plan. Second, the yen
depreciation from 80 yen per dollar in mid-2012 to 120 yen per dollar
in 2014 meant foreigners saw visiting Japan as a bargain. Third, visa
waver countries have been expanded gradually. Fourth, and probably
the most important, the airport capacity and allotment of landing
slots to international flights at the Haneda airport were expanded
dramatically in steps from 2010 to 2020. It made it possible to
accommodate more flights for tourists as well as business people
destined to Tokyo.
126 Takatoshi Ito

The third arrow of Abenomics announced in 2013 had a component

to promote foreign visitors to Japan, although details were not pro-
vided (Prime Minister’s Office 2013, p.25):
Promoting globalization by advancing Cool Japan and increasing the num-
ber of foreign visitors to Japan and encouraging foreign direct investment in
Japan Target:

• Triple the overseas sales of broadcast contents by 2018 from the

current level (6.3 billion yen).
• Double foreign companies’ direct investment in Japan to 35 trillion
yen by 2020.
• Increase the number of foreign visitors to Japan to 10 million in 2013
and over 30 million in 2030.
The 10 million foreign inbound target in 2013 was hit on time, and the
30 million visitor target in 2030 would be achieved in 2018, 12 years
earlier – a sign of huge success exceeding expectation.
With yen depreciation from 2012 to 2014, costs of staying and
moving around in Japan were lowered for foreign visitors. Hence
many visitors to Japan started to explore cities and regions other than
Tokyo and Kyoto. The benefits of inbound tourism were felt by many
local communities. Promotion of inbound tourism became one of the
tools to help businesses in rural communities.
Visa requirements for foreign visitors have been significantly relaxed
from 2013 to 2015. Among other changes, visa requirements were
removed for Thai passport holders and Malaysian passport holders in
July 2013, which had significant positive impacts on the numbers of
visitors from these countries. The number of visitors from Thailand
increased by 76 percent from the one-year period before the visa waiver
of July 2013 to the subsequent one-year period. For the comparison of
the same period, the number of Malaysian visitors increased by 57
percent.17 For Chinese visitors, requirements for obtaining a multiple-
entry visa became significantly relaxed in January 2015.
The policy measure that had most impact on increasing foreign visitors
to Japan was arguably the building of the fourth runway and re-
internationalization of Haneda airport.18 It is no surprise that the increase
in tourism coincided with expansion of capacities at Haneda and Narita
airports. In 2010, the fourth runway of the Haneda airport was com-
pleted, a dedicated international terminal was built, and long-range
The Third Arrow of Abenomics: Est. in 2013 – or 2007? 127

(million persons)

35 Arrival (Japanese) 110

Departure (Japanese)
30 Arrival (Foreigners) 100
Departure (Foreigners)
25 90

20 80

15 70

10 60

5 50

0 40
Haneda

Haneda

Haneda

Haneda
Haneda

Haneda

Haneda

Haneda

Haneda

Haneda

Haneda

Haneda

Haneda

Haneda
Narita

Narita

Narita

Narita

Narita
Narita

Narita

Narita

Narita

Narita

Narita

Narita

Narita

Narita
2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020

Figure 5.3 International passengers, Narita and Haneda airports

Source: Ministry of Justice

international flights were reintroduced at Haneda airport. The number of

international destinations increased from eleven in October 2010 to seven-
teen in March 2011, with more being added in the following years. The
number of international flights has steadily increased, and the flight con-
trol system used to manage the four runways has improved. As a result, the
number of international passengers (both Japanese and foreigners) has
increased steadily from 2011 to 2019 (Figure 5.3).
The number of international passengers at Haneda airport jumped
up in 2011, reflecting the introduction of regular international flights in
October 2010, in addition to international charter flights. The number
jumped again in 2014 as more slots were assigned to international
operations. Increases from 2015 onward resulted from negotiation of
aviation treaties with other countries. The final expansion of slots took
place in March 2020. The number of international passengers at
Haneda airport was expected to increase sharply again. But, inter-
national flights started to be cancelled in March due to COVID-19,
and many new routes were on hold.
The number of international passengers at Haneda airport has con-
tinued to climb since 2014. What is remarkable is that the number of
international passengers has been rising in Narita airport too. Both
airports saw more passengers, though with some differentiation in pas-
senger characteristics. Haneda airport, being closer to downtown
128 Takatoshi Ito

Tokyo, caters to business passengers who fly in and out of Tokyo. Narita

airport increasingly serves passengers connecting flights between North
America and Asia and passengers on low-cost carriers, with an LLC
terminal (terminal 3) having been built in April 2015. Haneda’s rise as
an entry point to Tokyo and to Japan in general was achieved without
hurting Narita airport.
As the number of international passengers rose in both Haneda and
Narita, it proved the assertion that Haneda would not erode Narita’s
importance. More flights can be accommodated for Tokyo. All the hard
work to convince skeptics and those resistant to the value of internation-
alizing Haneda in 2007–2008, as described in Section 5.4, paid off.

5.6 Pension Fund Reform

One of the third-arrow reforms that has had a strong impact on capital
markets was the reform of the Government Pension Investment Fund
(GPIF), which manage funds for corporate pension plans in Japan. The
Japanese pension system is in principle a pay-as-you-go (PAYGO) type:
social security contributions (taxes) are collected in connection with
personal income tax, and the revenues are used for pension benefits in
that fiscal year.19 Benefits to the pensioners can be more generous when
the population of pensioners is smaller compared to that of the workers
who pay social security contributions, under the pure form of the
PAYGO system. When the economy is growing – and the younger
generation earns a higher lifetime wage than the older generation –
the PAYGO system is sustainable, and all generations are better off
with the system.
When the current pension system was introduced in the early 1970s,
the Japanese economy satisfied these conditions, meaning that the
PAYGO system was sustainable and all generations were happier
with it. However, the growth rates of the Japanese economy precipi-
tously declined, and the old-age dependency ratio – the ratio of the
number of elderly people to the working-age population – rose higher
due to lengthening life expectancy and a fast-declining fertility rate.
The Japanese government had modified the PAYGO system slightly to
build up “reserves” by not paying out all the contributions to the
elderly in the same year. The reserves were built up to the level of
more than 120 trillion yen (close to one-quarter of GDP) by the early
2010s, and the GPIF was managing the portfolio.
The Third Arrow of Abenomics: Est. in 2013 – or 2007? 129

Traditionally the GPIF had invested close to two-thirds of its port-

folio in government bonds. However, investing mostly in government
bonds in an economic environment with low interest rates is far from
the norm among large public pension funds in other advanced coun-
tries. In order to realize returns for future generations, large public
pension funds had invested in equities. Also, shifting portfolios from
bonds to equities would stimulate the domestic equity markets.
Among the proposals in the Abe administration’s Japan
Revitalization Strategy of 2013, the reform of GPIF was briefly men-
tioned. In the section on “strengthening Japan’s international competi-
tiveness as a business hub,” the plan states: “The government will
consider ways to desirably manage public pensions and financial assets
held by incorporated administrative agencies, etc. (public and quasi-
public funds)” (Prime Minister’s Office 2013, p.72). The expert panel,
chaired by Takatoshi Ito, was formed in the summer of 2013 to discuss
how to reform the GPIF. The report was issued in November 2013
(Cabinet Secretariat 2013). Among its recommendations was: lowering
the share of domestic bonds, mostly Japanese government bonds
(JGBs); setting investment return targets; diversifying portfolios by
investing in new types of assets (so-called alternatives); increasing the
ratio of active investments; and improving governance by establishing
a “collegial decision-making” system, where full-time experts without
conflict of interest should play the role. The last point was emphasized
as the old system was that the president of the GPIF was solely respon-
sible for portfolio and administration decisions.
One sticking point, a very core issue, was why it makes sense to
invest in equities. Most experts argued that the public pension fund as
a long-term investor should invest in equities and reap the benefit of
equity premium, a well-known concept in finance. A few members
argued against investing in equities, not to mention increasing the
equity share. They argued that pensioners dislike risks and that they
want the fund to invest in safe assets, such as government bonds. The
majority argued that the beneficiaries from the public pension fund
portfolio investment are pensioners of future generation, and those
who receive pensions, supplemented by long-term returns from the
public pension fund, would prefer to raise returns while controlling
risk. Another objection was that having large shares in equities of
individual companies may mean that the government can intervene in
management through the voting power of the public pension fund. The
130 Takatoshi Ito

majority of the members argued that more diversification and more

equity investment have to come with enhanced governance, where
investment decisions have to be made solely on the ground of financial
investment.
The GPIF shifted its portfolio according to the expert group recom-
mendation. In December 2014, about a year after the experts’ report,
the GPIF announced it would lower the benchmark shares of domestic
bonds from 60 percent to 35 percent and raising shares of domestic
equity from 12 percent to 25 percent; those of foreign equity from
11 percent to 15 percent; and foreign equities from 12 percent to
25 percent (see Figure 5.4).
The shifts away from domestic bonds, from the benchmark of 67 per-
cent to 60 percent in June 2013 and from 60 percent to 35 percent in
December 2014, were drastic decisions. Actual portfolio shares of the
four asset classes were adjusted gradually from June 2013 to 2015 and
beyond.
Another change that experts emphasized in the report was that the
benchmark shares are not the mark that should be targeted all the time,
but the tolerance band around it should be wisely used. Within the

GPIF portfolio allocation

(%)
1.48 1.46 2.06 1.77 1.46 2.34 2.62
100 5.0 4.23 5.0
0.0
4.30 5.08 3.27 Short-term
12.35 12.90 assets
90 9.0 12.90 12.0
13.49 15.18 15.59 15.98 17.41 25.0 (others)
19.64 20.89 22.32
8.0
80 9.82
9.79
11.0
10.03 10.13
10.60 11.06
Foreign
11.06 Equities
11.0 12.14
70 14.57 15.73 15.0
13.14
12.63 13.08
12.92 12.0 16.29
17.22 16.47
17.26 Foreign
60 18.23
Bonds
19.80
50 25.0
22.00 23.39

Domestic
40 Equities
67.0
30 60.14 61.81 60.0 59.87 58.03 55.22 55.43 53.36 Domestic
49.61
20 43.13
39.39 Bonds
37.95
35.0

10
0
2012.2013.2013.2013.2013.2013.2014.2014.2014.2014.2014.2015.2015.
12 03 06 06 09 12 03 06 09 12 12 03 06
traditional old benchmark (2013/06) new benchmark
benchmark (2014/10)

Figure 5.4 Changes in the GPIF portfolio allocation

The Third Arrow of Abenomics: Est. in 2013 – or 2007? 131

Table 5.3 GPIF portfolio allocation benchmark and tolerance band

Old New
Traditional benchmark benchmark
benchmark (June 2013) (October 2014)

Domestic bonds 67% (+/− 8%) 60 (+/− 8%) 35 (+/− 10%)

Domestic equities 11% (+/− 6%) 12 (+/− 6%) 25 (+/− 9%)
Foreign bonds 8% (+/− 5%) 11 (+/− 5%) 15 (+/− 4%)
Foreign equities 9% (+/− 5%) 12 (+/− 5%) 25 (+/− 8%)
Short-term assets (others) 5% 5% NA

tolerance band, the actual share can deviate from the benchmark for
a sustained period time. This became useful in later years when the
bond yield precipitously declined. The change in the benchmark with
tolerance bands is shown in Table 5.3.
So, was the change in the GPIF portfolio and governance a triumph
of Abenomics that started in 2013? The short answer is no. In fact, the
2013 expert group recommendation had roots in the expert group on
globalization in the 2007–2008 CEFP.
One of the priority areas of the 2006–2008 CEFP was to improve
efficiency of the Japanese capital markets in order to provide savers and
pensioners with higher returns. The CEFP created a working group in
order to come up with concrete proposals. It identified that GPIF
should be a model for other public pension funds in investing more
aggressively into the equity market.
“The Second Report of the Expert Committee on Reforms
Addressing Globalization” recognized the governance problem of the
GPIF and recommended several directions of reform.20
1 Clarification of mission as pension fund management organization
2 Governance: Ensuring independence and professionalism, transpar-
ency and accountability
3 Split of management funds
4 Establishment of the New Fund (baby funds)
5 Qualification and treatment of directors, managers, and employees
6 Asset management from long-term perspective and performance
evaluation
132 Takatoshi Ito

I argue that the origin of the GPIF reform, which materialized after
2013, was the second report of the Expert Committee on Reforms
Addressing Globalization.21 In the report, a reform of the GPIF was
advocated: first, the governance has to be strengthened; and second,
returns to portfolio were very low compared to public funds in other
countries. The report was discussed in the CEFP meeting on May 23,
2008. MrYoichi Masuzoe, then minister of health, labor, and welfare,
gave rather negative views toward diversifying the portfolio of the
GPIF, insisting that the public does not want the pension fund to take
a risk by investing more into equities. The negative reaction from the
minister in charge basically killed the proposal at the time.
This report in 2008 was indeed a precursor of the expert report of
2013. The seed of the GPIF reform that became implemented in
2013–2016 was indeed sown in the first Abe administration, but it
came to full bloom only five years later.

5.7 Concluding Remarks

In this chapter, selected aspects of the third arrow of Abenomics
were highlighted. I argued that they have roots in the 2006–2008
CEFP formed under the first Abe cabinet. Several policy areas were
revived after a five-year hiatus when Prime Minister Abe came
back to power. Expanding the FTAs was set in motion back in
2006–2008, but it was not until 2013 and later that it got into full
swing, including FTAs with the EU and the TPP11. The crucial
decision to re-internationalize Haneda airport, without limitation
on the distance to destination of flights, was also decided in
2006–2008. It took many years, until March 2020, to realize the
full potential of the fourth runway, but the advances were put on
the schedule back in 2008. The GPIF reform was a bit more
unsuccessful in 2006–2008, but the idea was revived in the experts
panel that was formed in 2013. An unsuccessful reform plan in
2008 was successfully implemented in 2013.
There are several other topics that have links between the
2006–2008 CEFP and Abenomics after 2013, including labor reform
and stock exchange consolidation. The details of those topics will be
investigated in the future.
One lesson is that reforms take time, especially in reaching the end of
implementation. Another lesson is that the leader in charge matters.
The Third Arrow of Abenomics: Est. in 2013 – or 2007? 133

Strong leadership with capable and skillful advisors is indispensable for

structural reforms.

Notes
1. He served as prime minister from September 2006 to September 2007;
and then became prime minister again in December 2012 and he is still
prime minister as of this writing (end-December 2019). He became the
longest cumulatively serving prime minister on November 20, 2019;
and he will likely become the longest continuously serving prime
minister on August 24, 2020, as no general election is necessary or
foreseen during 2020.
2. The author of this paper was appointed as a member of the CEFP in
September 2006 and served until September 2008. Although the author
tries to be neutral, some tendency to glorify the work of the 2006–2008
CEFP may remain.
3. Private-sector member’s paper, October 24, 2006, www5.cao.go.jp/keizai-
shimon/minutes/2006/1024/item1.pdf.
4. Meeting records – including papers and documents submitted to the
meeting, minutes, and transcripts of the meeting – are available at the
CEFP homepage archives (in Japanese): www5.cao.go.jp/keizai-shimon/
minutes/index.html. Only a limited amount of information is disclosed in
English: www5.cao.go.jp/keizai-shimon/english/publication/index.html.
5. The “EPA and Agriculture” report is available in Japanese at www5
.cao.go.jp/keizai-shimon/special/global/epa/pdf/item1.pdf and in
English at www5.cao.go.jp/keizai-shimon/english/publication/pdf/070
703economic.pdf.
6. Two reports of the working group are available in Japanese at
www5.cao.go.jp/keizai-shimon/special/global/finance/pdf/item1.pdf
and www5.cao.go.jp/keizai-shimon/special/global/finance/pdf/item2
.pdf; and in English at www5.cao.go.jp/keizai-shimon/english/publica
tion/pdf/070723_financial_wg_02.pdf and www5.cao.go.jp/keizai-
shimon/english/publication/pdf/080523financial_wg.pdf.
7. It is not clear why the DPJ regarded the CEFP as an undesirable
institution. I can only guess that the DPJ regarded the CEFP as an
institution with an agenda for market-oriented liberalization that had
caused inequality and that the DPJ wished to show up individual
politicians as taking initiatives – for example, the task force of
government business sorting (Jigyo Shiwake Iinkai), which was led by
a DPJ politician, Renho.
134 Takatoshi Ito

8. Links to all past Japan Revitalization Strategy reports can be found on

the website of the Japan Economic Revitalization Headquarters: www
.kantei.go.jp/jp/singi/keizaisaisei/kettei.html.
9. LDP politician Hidenao Nakagawa coined the word “Abenomics.”
10. The DPJ also considered joining in the TPP negotiation, but it could not
make a decision before the regime ended.
11. Japan joined in the group of negotiating countries from the 18th
negotiation meeting of July 2013.
12. See Ito and Hoshi (2020, table 11) for the past FTAs and EPAs that
Japan negotiated before TPP.
13. Initially, China Airlines, an airline based in Taiwan, remained in
Haneda in order to separate it from airlines from Mainland China.
Later, China Airlines moved to Narita.
14. For historical reasons, Northwest Airlines had the largest share of slots
(rights to land and depart) at Narita. Japan Airlines as a national flag
carrier had the second-largest share. It was expected that their
dominance would be diluted if Haneda opened slots to presumably
new entries to international flights.
15. In January 2003, Prime Minister Koizumi announced that he would aim
to achieve 10 million foreign visitors to Japan by 2010. The Ministry of
Land, Infrastructure, Transport, and Tourism established the “Visit
Japan Campaign” secretariat and promoted Japan as a destination.
They focused on developing favorable destinations in Japan; making
infrastructure signs multilingual; relaxing visa conditions; expanding
airports; and inviting low-cost carriers.
16. In addition to their other tragic effects, the East Japan Great Earthquake
and the Fukushima Daiichi nuclear disaster resulted in reduced foreign
visitors to Japan in 2011 and 2012.
17. The list of relaxations of visa requirements from 2013 to 2015 and the
numbers of Thai and Malaysian visitors before and after the reform are
available (in Japanese) from Tourism Agency, MLITT, www.mlit.go.jp
/common/001080017.pdf.
18. Haneda airport was the only major airport in Tokyo until Narita airport
was opened in 1978. Since then, the two airports were strictly segmented,
with Narita for international flights and Haneda for domestic. Narita
airport had very limited domestic feeder flights, while Haneda was
opened for daytime “charter flights” to Seoul in 2003; to Shanghai in
2007; Hong Kong in 2008; and Beijing in 2009.
19. See Ito and Hoshi (2020, section 8.7) for details about Japan’s pension
system.
20. See www5.cao.go.jp/keizai-shimon/english/publication/pdf/080523fin
ancial_wg.pdf.
The Third Arrow of Abenomics: Est. in 2013 – or 2007? 135

21. See www5.cao.go.jp/keizai-shimon/english/publication/pdf/080523fin

ancial_wg.pdf.

References
Cabinet Secretariat. 2013. Final Report. Panel for Sophisticating the
Management of Public/Quasi-public Funds. November. www
.cas.go.jp/jp/seisaku/koutekisikin_unyourisk/index-e.html.
Council of Economic and Fiscal Policy (CEFP), Cabinet Office. 2007a.
“Acceleration of Economic Partnership Agreements and Strengthening
Agricultural Reform.” The first report of the working group on
Economic Partnership Agreements and Agriculture, May 8, 2007. www5
.cao.go.jp/keizai-shimon/english/publication/pdf/070703economic.pdf.
Council of Economic and Fiscal Policy (CEFP), Cabinet Office. 2007b. “The
Basic Policy of 2007.” www5.cao.go.jp/keizai-shimon/cabinet/2007/de
cision070620.pdf.
Council of Fiscal and Economic Policy (CFEP), Cabinet Office. 2008.
“Minutes of meeting.” www5.cao.go.jp/keizai-shimon/minutes/2008/0
520/minutes_s.pdf.
Ito, Takatoshi, and Takeo Hoshi. 2020. The Japanese Economy, 2nd ed.
Cambridge, MA: MIT Press.
Prime Minister’s Office, Japan Economic Revitalization Headquarters. 2013.
“Japan Revitalization Strategy: Japan Is BACK.” June 14. www
.kantei.go.jp/jp/singi/keizaisaisei/pdf/en_saikou_jpn_hon.pdf.
part iii

Macroeconomic Policy
6 Abenomics, Monetary Policy,
and Consumption
j o s h u a k . h a u s m a n, t a k a s h i u na y a m a ,
a n d jo h a n n e s f. w i e l a n d

6.1 Introduction
Prime Minister Shinzo Abe took office in late 2012 and began an
economic program dubbed “Abenomics.” Abenomics consists of
three parts, commonly called arrows: monetary expansion, fiscal
expansion followed by consolidation, and structural reforms. In
2015, two of us provided an interim update on Abenomics (Hausman
& Wieland 2015).1 We focused on the first arrow, monetary expan-
sion. While we found large, positive effects of the Bank of Japan’s
policies on financial markets, we saw little evidence of substantial
effects on the real economy. We noted the disappointing response of
net exports to a weaker yen and the disappointing response of con-
sumption to a lower real interest rate.
This (early 2019, at time of writing) is a good time to revisit the
progress of Abenomics, since longer time series now allow for more
definite conclusions. In particular, we shall take advantage of the
availability of household-level data on consumption during
Abenomics. In addition to more complete data, Abe’s recent reelection
in October 2017 means that Abenomics may continue for some time.
Understanding what Abenomics has achieved thus far may be useful for
predicting its future effects.
As in our prior work, we largely focus on the first arrow, the monet-
ary policy component of Abenomics. Still, the components of

* This manuscript was prepared for the conference on “The Political Economy of
Japan under the Abe Government” at Shorenstein Asia Pacific Research Center,
Stanford University. We thank the Statistical Bureau of Japan for allowing us to
use the microdata from the Family Income and Expenditure Survey. We are
grateful for comments and encouragement from Huiyu Li (discussant), three
anonymous referees, Takeo Hoshi, John Leahy, Phillip Lipscy, and Valerie
Ramey. Naoto Watanabe and Kazuma Yamamoto provided superb research
assistance.

139
140 Joshua K. Hausman, Takashi Unayama, Johannes F. Wieland

Abenomics cannot be entirely separated. To draw conclusions about

the effect of monetary policy on consumption, for instance, we cannot
avoid discussing the effect of the April 2014 consumption tax increase
on consumption.
More broadly, and as this volume demonstrates, the economics of
Abenomics cannot be completely isolated from the politics. Abe admin-
istration economic policies both were influenced by political consider-
ations and had political repercussions. Hoshi and Lipscy argue in the
Introduction to this volume that the Abe administration was strongly
focused on maintaining public support, with policies supporting eco-
nomic growth part of this effort. Of course, this is not to suggest that
higher GDP growth and lower unemployment were not themselves
goals. No doubt they were. But some observers believe that the larger
priority of the Abe administration was on obtaining sufficient political
support to achieve conservative political goals, with the ultimate
objective being to change Article 9 of the Japanese constitution.2
Our results suggest that Abenomics may have been less successful in
delivering political support for Abe over the medium-to-long run than
commonly perceived. We show that, while financial markets and to
a lesser extent GDP growth responded as hoped to Abenomics,
Abenomics failed to significantly increase household consumption,
likely because it failed to significantly increase household disposable
income. An evaluation of the political impacts of Abenomics, and in
particular those of the monetary policy arrow, is outside the scope of
our paper. But it would be worthwhile to investigate the subject:
perhaps Abe has maintained political support despite – and not because
of – his economic policies, or perhaps positive political repercussions of
financial and labor market performance during Abenomics have out-
weighed negative political repercussions from the lackluster growth of
disposable income and consumption.
We begin in Section 6.2 by exploring the effect of expansionary
monetary policy on Japanese financial markets. We find that, over the
past six years, expansionary monetary policy has lowered nominal
interest rates, weakened the yen, and raised stock prices. In
Section 6.3 we turn to an examination of actual and expected inflation,
since many economists (e.g., Romer 2014; Kuroda 2013) predicted
that the first arrow of Abenomics would work by raising actual and
expected inflation. Here results are more disappointing. While actual
inflation has risen since 2012, expected inflation has not. And both
Abenomics, Monetary Policy, and Consumption 141

actual and expected inflation remain (in early 2019) far below the Bank
of Japan’s 2 percent target.
As shown in Section 6.4, despite the lack of progress raising inflation,
Abenomics has been accompanied by strong output and employment
growth. GDP growth per working age person (ages 15–74) in Japan
averaged 1.6 percent per year between 2013 and 2018, roughly 10 per-
cent more than US growth in this period and 50 percent more than
Japanese growth in the 1993–2007 period. Along with strong output
growth, the unemployment rate fell from 4.4 percent in 2012 to
2.4 percent in December 2018.
Growth occurred despite, not because of, consumption spending, as
detailed in Section 6.5. Consumption grew at only a 0.4 percent annual
rate between 2012 and 2018, whereas GDP grew at a 1.1 percent annual
rate. We consider why consumption spending has grown so slowly
despite strong overall GDP growth, lower real interest rates, and rising
asset prices. We start by noting that slow aggregate consumption growth
must reflect either (1) a combination of positive effects of monetary
policy and negative effects of other shocks or (2) the lack of a positive
effect of monetary policy on consumption. To distinguish between these
possibilities, we look at household-level consumption data.
Household-level data from the Family Income and Expenditure
Survey allow us to compare the path of consumption across households
that are differentially exposed to monetary policy based on their house-
hold characteristics. These characteristics include measures of house-
hold indebtedness, asset holdings, income, and age. We see no obvious
differences in household consumption patterns across these groupings.
Thus we find little evidence in the cross section that the effect of
monetary policy during Abenomics was large.
This finding means that there is not necessarily a need to appeal to
negative aggregate shocks to explain slow consumption growth.
Nonetheless, we consider the plausibility of three such possible shocks:
(1) demographic trends; (2) the April 2014 consumption tax increase;
and (3) slow household income growth. Of these, we find that the last is
likely to have mattered most.

6.2 Financial Markets

Figure 6.1 in this chapter reproduces figure 1 in Hausman and Wieland
(2015). The figure shows the evolution of Japanese financial markets
142 Joshua K. Hausman, Takashi Unayama, Johannes F. Wieland

before and after Abenomics. In general, the initial financial market

effects of Abenomics persisted or continued in the same direction.
Nominal interest rates – see Figure 6.1 (A) – fell when Abenomics
began and continued to fall thereafter. The decline was most notable
at the long end of the yield curve; thirty-year bond yields fell more than
a percentage point between October 2012 and October 2018.
Market inflation forecasts – Figure 6.1 (B) – were an exception to the
general pattern of persistent financial market effects. After initially
rising, two-year and ten-year inflation swap rates are now (as of
January 2019) only slightly positive, below where they were when
Abenomics began. Figure 6.1 (C) shows the result of subtracting infla-
tion swap rates from nominal bond yields. The resulting ten-year real
bond yield fell 0.7 percentage points between October 2012 and
October 2018. As mentioned, and as is shown in Figure 6.1 (D),
Abenomics was also associated with a persistent weakening of the
yen. This weakness is apparent both in the nominal rate vis-à-vis
the dollar (the solid line) and in the real, trade-weighted exchange
rate (the dashed line).
Despite a sharp correction in fall 2018, stock prices remain much
above their pre-Abenomics levels – see Figure 6.1 (E). The Nikkei 225
index is now (in December 2018) 132 percent above its
November 2012 level. This is, however, still 45 percent below the
index’s peak in December 1989. In Hausman and Wieland (2014) we
argued based on a Campbell-Shiller decomposition (Campbell &
Shiller 1988) that the Japanese stock market would likely revert to
a level consistent with the level of dividends. Dividends had increased
by 9.1 percent at the time, so our best (long-run) forecast for future
stock prices was that they would fall until they were 9.1 percent above
their November 2012 level. Since then, dividends have increased sub-
stantially more, to 76.9 percent above their November 2012 level,
suggesting that a larger gain in stock prices is likely to be sustained.
The dividend–price ratio has by December 2018 closed 85.1 percent of
its gap relative to November 2012. This is a marked change from the
previous year. As of December 2017 only 14.9 percent of the gap was
closed. The rapid convergence reflects both a sharp drop in the stock
index as well as dividend growth of roughly equal proportion.
A skeptical observer might wonder how much, if any, credit
Abenomics deserves for the behavior of financial markets shown in
Figure 6.1. Stock price gains in Japan, for instance, are not independent
Abenomics, Monetary Policy, and Consumption 143

4
3
Nominal bond yields (%)

Inflation swap rates(%)

2

2
1

0
0

2-year

-2
10-year 2-year
30-year 10-year
-1

2007m1 2009m1 2011m1 2013m1 2015m1 2017m1 2019m1 2007m1 2009m1 2011m1 2013m1 2015m1 2017m1 2019m1
(a) (b)

Yen/$ exchange rate and real exchange

160
4

Nominal exchange rate

Real exchange rate

140
rate (1/2007 Yen/$)
Real bond yields (%)
2

120
0

100
–2

2-year
80

10-year

2007m1 2009m1 2011m1 2013m1 2015m1 2017m1 2019m1 2007m1 2009m1 2011m1 2013m1 2015m1 2017m1 2019m1
(c) (d)
140
120
Index (1/2007=100)
100
80
60

Nikkei
Topix
40

2007m1 2009m1 2011m1 2013m1 2015m1 2017m1 2019m1

(e)

Figure 6.1 Abenomics’ financial market effects

Abenomics begins at the vertical line, November 2012. Real bond yields are
calculated as the difference between nominal bond yields and inflation swap
rates.
Sources: See data appendix

of the general rise in developed country stock markets. To obtain some

evidence on causality, we expand our analysis of announcement effects
in Hausman and Wieland (2014). We identify dates on which there was
substantial, surprising news about the course of Japanese monetary
policy and then examine the one-day change in interest rates, the
exchange rate, and stock prices on these dates. The dates are:
144 Joshua K. Hausman, Takashi Unayama, Johannes F. Wieland

• November 15, 2012 – as a candidate, Shinzo Abe calls for monetary

easing.
• January 22, 2013 – the Bank of Japan announces its 2 percent
inflation target.
• February 5, 2013 – the governor of the Bank of Japan, Masaaki
Shirakawa announces that he would resign early.3 This is viewed as
the prelude to the appointment of a more dovish governor.
• April 4, 2013 – the Bank of Japan announces its “qualitative and
quantitative” monetary easing program.
• October 31, 2014 – the Bank of Japan expands its annual asset
purchases from 60–70 trillion yen per year to 80 trillion yen per year.
• December 18, 2015 – the Bank of Japan extends the maturity of its
bond purchases and commits to other, small expansions of its easing
program.
• January 29, 2016 – the Bank of Japan introduces a negative interest
rate on additions to reserves held by financial institutions at the Bank
of Japan.
• July 29, 2016 – the Bank of Japan announces a near doubling of its
exchange-traded fund (equity) purchases.
In studying these events, it is important to recognize that monetary
news could either be good or bad. Any monetary announcement could
be either stronger or weaker than expected and thus have opposite
effects on financial markets. Since the sign of the news shock is ambigu-
ous, we look for a more restrictive set of patterns. First, are the
announcement effects (both negative and positive) large? Second, is
the comovement of bond yields, the exchange rate, and the stock
market consistent with a monetary shock?
Table 6.1 shows that in most cases Bank of Japan easing was accom-
panied by falling nominal rates, a weakening yen (an increase in the
number of yen per dollar) and rising stock prices. This comovement is
consistent with monetary policy shocks. The exceptions are also
informative. As we pointed out in Hausman and Wieland (2014),
given prior news about the Abe administration’s monetary policy
objectives, there was little news content in the Bank of Japan’s formal
announcement of a 2 percent inflation target on January 22, 2013.
Thus the strengthening of the yen and decline in stock prices on this
date are unsurprising. Likewise, while the December 18, 2015,
announcement was on paper mildly expansionary, it did not signal
Abenomics, Monetary Policy, and Consumption 145

Table 6.1 Announcement effects

2-year 10-year 30-year %Δ

Date bond bond bond %Δ ¥/$ Topix

November 15, −0.5 −0.9 2.8 1.15 2.09

2012
January 22, 2013 −0.1 −0.8 −0.1 −0.99 −0.44
February 5, 2013 −1.2 −2.9 −2.2 1.36 1.37
April 4, 2013 0.3 −11.4 −21.7 3.55 2.70
October 31, 2014 0.9 −1.1 −5.0 2.85 4.28
December 18, −1.7 −3.0 −4.1 −1.14 −1.76
2015
January 29, 2016 −4.1 −12.9 −18.0 1.95 2.87
July 29, 2016 9.7 8.0 1.0 −3.05 1.20

* The table shows changes in basis points – bond yields; and percent – exchange rate
and stock prices.

expansion relative to prior expectations. The Japan Times headline

read: “BOJ adjustments signal limits to monetary easing” (Sakamoto
2015).
The July 29, 2016, announcement is an intriguing case: on this date,
the stock market rose, but other asset prices suggested contractionary
news – bond yields rose and the yen significantly strengthened. The
explanation appears to be that while the announcement of more
exchange-traded fund purchases was good news for the stock market,
it was less monetary easing than had been hoped for. Reuters quoted an
economist at Mizuho Securities, Norio Miyagawa, as saying: “The BOJ
did not live up to expectations” (Kihara 2016).
Importantly, the responses to monetary policy shown in Table 6.1
are not only qualitatively consistent with an effect of monetary policy
on financial markets: they are also quantitatively large. All daily
exchange rate movements in the table are either in the top or bottom
fourth percentile of all daily exchange rate movements between 2010
and 2018. The table includes the largest and fourth-largest daily depre-
ciation as well as the third-largest appreciation in this period. Daily
stock price movements tend to be more volatile than exchange rate
movements. Still, half of the announcement dates fall in the top or
bottom fifth percentile between 2010 and 2018. And October 31, 2014
146 Joshua K. Hausman, Takashi Unayama, Johannes F. Wieland

was the ninth-largest stock market appreciation over this period. These
large changes in asset prices suggest that monetary news had a large
influence on financial markets.

6.3 Inflation and Inflation Expectations

Figure 6.2 (A) shows three measures of prices in Japan: the overall CPI;
the CPI excluding food and energy (so-called core-core prices); and the
GDP deflator. For the two CPI measures, we show both the actual data
and the data excluding the effect of the April 2014 3 percentage point
increase in the consumption tax. All indices show some increase during
the Abenomics period. As of 2019, deflation appears to have ended. But
none of the measures suggest that inflation is near the Bank of Japan’s
2 percent target. Between December 2017 and December 2018, both
the overall CPI and the CPI excluding fresh food and energy rose
0.3 percent. At such low levels of aggregate inflation, many prices
continue to fall: over the December 2017 to December 2018 period,
price indices for the food, housing, and transportation and communi-
cation categories all fell.
Figure 6.2 (B) provides information on the inflation expectations and
perceptions of Japanese households and firms. The solid line shows the
inflation expectations of Japanese households. These expectations
come from the Bank of Japan “Opinion Survey,” in which households
are asked: “By what percent do you think prices will change one year
from now?” Household expected inflation is surprisingly high –
roughly 4 percent – both before and after Abenomics begins. The
Bank of Japan also asks households what they believe inflation was
over the past year. Perhaps unsurprisingly, household inflation percep-
tions generally track household inflation expectations. They are far
above actual inflation. But, unlike expected inflation, perceived infla-
tion rose during Abenomics.
The dashed line in Figure 6.2 (B) shows the inflation expectations
of firms from the Bank of Japan’s Tankan survey. Unfortunately,
before 2014 firms were not asked about their inflation expectations.
Thus the data do not allow for a pre- to post-Abenomics comparison.
They show, however, that the inflation expectations of firms have
declined during Abenomics from 1.5 percent in 2014 to 0.9 percent
in 2017.
Abenomics, Monetary Policy, and Consumption 147

106
CPI, 2/2007 = 100
CPI ex. cons. tax
CPI ex. fresh food and energy
CPI ex. food and energy ex. cons. tax
103

GDP deflator, 2007:Q1= 100

Price indices (SA)
100 97
94

2007m1 2009m1 2011m1 2013m1 2015m1 2017m1 2019m1

(a)

BoJ Opinion Survey: 1 year-ahead CPI forecast

BoJ Opinion Survey: Past year CPI estimate
Tankan Survey: 1 year-ahead CPI forecast
108
Inflation (%)
6 4
2
0

2007m1 2009m1 2011m1 2013m1 2015m1 2017m1 2019m1

Survey date
(b)

Figure 6.2 Japanese actual and expected inflation

Abenomics begins at the vertical line, November 2012. In (A), direct effects of
the consumption tax are excluded from the CPI by assuming that the
148 Joshua K. Hausman, Takashi Unayama, Johannes F. Wieland

Actual and expected inflation outcomes of Abenomics are thus disap-

pointing. The lack of an increase in inflation expectations means that,
contrary to the hopes of the Bank of Japan and many economists, higher
expected inflation did not lower real interest rates and thus did not spur
spending. Real interest rates – Figure 6.1 (C) – fell, but this is more
because nominal rates fell than because inflation expectations rose.

6.4 Output and Employment

Despite the lack of progress raising actual and expected inflation, real
economic performance has improved during Abenomics along most
dimensions. (The prominent exception to this good performance is
consumption, to which we will turn in Section 6.5.) Table 6.2 updates
table 2 in Hausman and Wieland (2015), while changing the defin-
ition of working age to include people aged sixty-five to seventy-
four.4 Across all dimensions shown in the table, performance during
the Abenomics period (2013–2018) is better than or equal to that
during the lost decades. While GDP growth (the first column of panel
A) was roughly the same in both periods, when adjusted for the recent
decline in the working age population (the first column of panel B) it
was roughly 50 percent more rapid during the Abenomics period.
During this time, growth per working-age person was also faster than
that in the United States (the second column of panel B). The remain-
ing columns of panel B show that higher growth during Abenomics
was driven both by an increase in productivity growth and by a large
increase in the employment-to-population ratio.
To better understand the timing and sources of growth during
Abenomics, Figure 6.3 (A) shows growth by quarter during
Abenomics, and Figure 6.3 (B) shows contributions to GDP growth
by component. Quarterly growth during Abenomics has been generally

Caption for Figure 6.2 (cont.)

consumption tax raised twelve-month headline (excluding food and energy)
inflation 1.9 percentage points (1.5 percentage points) in April 2014, and
2.1 percentage points (1.7 percentage points) from May 2014 through
March 2015. These are the figures suggested by the Bank of Japan, www
.stat.go.jp/english/data/cpi/report/2014np/pdf/fu8.pdf.
Source: See data appendix
Table 6.2 Macro summary statistics

Panel A: Raw data

Real GDP Money market

growth (% Unemployment interest rate
Year change) rate (%) CPI inflation (%) (%)

1974–1992 average 4.0 2.3 4.8 6.7

1993–2007 average 1.2 4.1 0.1 0.6
2008–2012 average −0.2 4.6 −0.2 0.2
2013–2018 average 1.1 3.2 0.5* 0.0

Panel B: Adjusted for working-age population

Real GDP per person age Multifactor productivity Employment/population

15–74 (% change) (% change) age 15–74 (%)

Year Japan USA Japan USA Japan USA

1974–1992 average - - - - 64.4 64.2

1993–2007 average 1.1 2.1 0.7 1.1 64.3 67.5
2008–2012 average 0.2 -0.2 0.2 0.7 64.0 63.6
2013–2018 average 1.6 1.5 0.9** 0.4*** 66.9 63.9

* This excludes the direct effect of the April 2014 increase in the consumption tax from 5 percent to 8 percent (see http://www.stat.go.jp/english/
data/cpi/report/2014np/pdf/fu8.pdf). Including the consumption tax, average CPI inflation was 0.9 percent. **2013–2016 ***2013–2017.
Note: Growth rates are year-over-year.
Source: See data appendix.
150 Joshua K. Hausman, Takashi Unayama, Johannes F. Wieland

10
Real GDP growth (%, annualized)
-10 0
-20

2007q1 2010q1 2013q1 2016q1 2019q1

Date
(a)

1995–2007
1.5

2008–2012
Avg. year-on-year growth rate (p.p.)

Abenomics (2013–2018)
0 .5 -.5 1

GDP Res. Inv. Gov. purchases Imports(-)

Consumption Nonres. Inv. Exports GDP
Contribution to GDP growth Per working age
(b)

Figure 6.3 (A) Annualized quarter-on-quarter GDP growth since 2007. Quarters
since Abenomics began are more lightly shaded; (B) Comparison of annualized
contributions to GDP by component during Abenomics (2012–2018), the lost
Abenomics, Monetary Policy, and Consumption 151

positive. Particularly notable are the six consecutive quarters of posi-

tive growth from 2016:Q3 through 2017:Q4.
Turning to Figure 6.3 (B), we see that while overall growth during
Abenomics (the leftmost columns) was similar to that during the 1990s
and early 2000s, its composition was different. Growth was driven less
by consumption and more by investment and government purchases.
Exports made substantial contributions to growth in both periods, but –
particularly during Abenomics – this positive contribution to growth
was offset by rising imports. The average annual contribution of net
exports to growth during Abenomics was only 0.2 percentage points.
When we examined contributions to growth during Abenomics three
years ago (Hausman & Wieland 2015), we identified two puzzles: the
slow growth of net exports; and the slow growth of consumption. The
former has to some extent resolved, with net exports now contributing to
growth, albeit less than might be hoped.5 Lackluster consumption growth
remains a puzzle, however, and is the focus of the rest of this chapter.

6.5 Consumption
Consumption grew at an average annual rate of just 0.4 percent
between 2012 and 2018; the slow growth of consumption relative to
GDP meant that over this six-year period its share of GDP fell from
58.6 percent to 55.7 percent. To put this is in perspective, between
1956 and 2018 the consumption share of US GDP has never fallen by
more than 1.7 percentage points in a five-year period. The slow growth
of consumption during Abenomics has not been driven by just one year
of poor performance; GDP growth exceeded consumption growth in
each of the five years from 2014 through 2018.
The slow growth of Japanese consumption was disappointing not
only in absolute terms but also relative to forecasts. Figure 6.4 shows

Caption for Figure 6.3 (cont.)

decade excluding the Great Recession (1994–2007), and the Great Recession
(2007–2012).
Contributions are calculated as in Japan’s national accounts (https://www
.esri.cao.go.jp/en/sna/data/sokuhou/files/2011/qe114_2/pdf/kiyoe2.pdf). The
figure also shows annualized working-age adjusted GDP growth.
Source: See data appendix
152 Joshua K. Hausman, Takashi Unayama, Johannes F. Wieland

120
Data (2007=100)
Forecast 2012m10
Forecast 2013m10
115

Forecast 2015m4
Forecast 2007m10
110
GDP
105 100
95

2005 2010 2015 2020 2025

(a)
120

Data (2007=100)
Forecast 2012m10
Forecast 2013m10
Forecast 2015m4
115

Forecast 2007m10
Consumption
110
105
100

2005 2010 2015 2020 2025

(b)

Figure 6.4 Actual and forecast output and consumption

Note: Forecasts are from Consensus Economics. Abenomics begins at the
vertical line (2012).
Abenomics, Monetary Policy, and Consumption 153

Consensus Economics forecasts for real GDP and consumption made

in 2007, 2012, 2013, and 2015. Actual GDP has significantly disap-
pointed relative to forecasts made in 2007; it is also lower than was
forecast in 2013, the initial year of Abenomics. But real GDP is above
the level forecast in 2012, before Abenomics began, and it is tracking
closely the forecast made in 2015. By contrast, actual consumption has
disappointed relative to all forecasts, including those made in 2015.
One possibility is that the expansionary effect of monetary policy on
consumption was offset by other aggregate shocks. Another is that
monetary policy only weakly affected consumption. We use household
expenditure microdata to distinguish between these two hypotheses.
These data allow us to distinguish between these hypotheses because if
monetary policy affects consumption, we would expect it to have
heterogenous effects on individual households; the effect of monetary
policy on an individual household’s consumption growth ought to
depend on characteristics of the household, including the household’s
debt position, age, and employment status. An increase in expected or
perceived inflation, for instance, ought to benefit a household more if it
has a large nominal debt burden (Fisher 1933; Auclert 2019). Wong
(2016) finds stronger consumption responses to monetary policy
shocks among the young, due to their greater propensity to refinance.
And Cloyne, Ferreira, and Surico (2020) estimate stronger consump-
tion responses for mortgagers than for outright owners or renters.
If we observe these sorts of heterogenous effects, it would imply that
monetary policy is affecting household consumption; the lack of aggre-
gate consumption growth would then suggest the presence of other
negative aggregate factors, such as demographic trends or the 2014
consumption tax increase. By contrast, if we do not observe heterogen-
ous consumption patterns consistent with monetary policy’s predicted
effect, it would suggest that monetary policy transmission to house-
holds is not occurring in the expected way.

6.5.1 Data
Our data come from the Family Income and Expenditure Survey
(FIES). This is a monthly rotating panel of 9,000 Japanese households.
It is designed to provide an accurate representation of income and
spending patterns of the entire Japanese population.6 Before using
these data to investigate the effect of monetary policy on consumption
154 Joshua K. Hausman, Takashi Unayama, Johannes F. Wieland

during Abenomics, we check that the FIES consumption data align with
the patterns we observe in the national accounts. We do this because
there have been concerns about the reliability of consumption statistics,
particularly given the divergence between slow consumption growth
and relatively strong GDP growth. Indeed, in 2016 the Bank of Japan
started publishing a new indicator for household consumption called
the Consumption Activity Index. Also in 2016, the government set up
two special committees to improve the FIES. Fortunately, however, in
spite of these efforts, no serious problems in the aggregate consumption
data have been found. That is, it turned out that the stagnation of
consumption was not a statistical problem but a real one.
We see this result in Figure 6.5 (A), which compares aggregate
consumption implied by the FIES with the national accounts and the
Bank of Japan Consumption Activity Index (travel balance adjusted).
To avoid differences resulting from different methods of deflating, the
comparison is in nominal terms. The series track each other closely, at
least until 2018. Figure 6.7 (C) compares yearly nominal income in the
FIES, our most reliable household income measure in the dataset, with
nominal wages and nominal gross national income from the national
accounts. Yearly income tracks wage growth fairly well, but it is
somewhat more volatile.

6.5.2 Heterogenous Monetary Policy Effects?

In Figures 6.6 and 6.7 we compare the consumption patterns since
Abenomics began among households with different characteristics.
Our graphical analysis is analogous to the specifications in Wong
(2016) for age and Cloyne, Ferreira, and Surico (2020) for home
ownership.7 Figure 6.6 (A) starts by comparing the path of consump-
tion across households with high, low, and medium liability to income
ratios. Insofar as Abenomics at least initially induced some expectation
of higher inflation, we would expect it to have particularly benefited
households with high debt-to-income ratios. We do not, however, see
any notable differences in the consumption paths of these three groups.
In Figure 6.6 (B) we plot the average propensity to consume across the
three groups to account for potentially different income exposure to
monetary policy. Again we see largely parallel trends, with the excep-
tion of an uptick in the propensity to consume in the high liability to
income group in 2017. Still, over the entire 2012–2017 period there is
Abenomics, Monetary Policy, and Consumption 155

(a)
Trend in Consumption Measures 2005-2018
110

FIES (Consumption Expenditure

National Accounts
Consumption Activity Index (BOJ)
105
Year 2012 = 100
100 95
2005q1

2006q1

2007q1

2008q1

2009q1

2010q1

2011q1

2012q1

2013q1

2014q1

2015q1

2016q1

2017q1

2018q1
(b)
Trend in Income Measures 2005-2018
115

FIES (Yearly Income)

Gross National Income
Wage (National Accounts)
110
Year 2012 = 100
105
100
95

2005q1

2006q1

2007q1

2008q1

2009q1

2010q1

2011q1

2012q1

2013q1

2014q1

2015q1

2016q1

2017q1

2018q1

Figure 6.5 Checking FIES data consistency with the national accounts
Source: See data appendix

not any obvious, large divergence between the change in the average
propensity to consume in the high versus low liability to income group.
156 Joshua K. Hausman, Takashi Unayama, Johannes F. Wieland

(a) (b)

(c) (d)

Figure 6.6 FIES results

Source: See data appendix

Figure 6.6 (C) performs a similar exercise with households divided

by their financial asset to income ratios. Since expansionary monetary
policy during Abenomics led to a large increase in asset prices, all else
equal we would expect to see more consumption among households
with higher asset to income ratios. Indeed, Unayama and Komura
(2014) show that there were significant wealth effects on consumption
caused by the stock price increase in the early stage of Abenomics. They
conclude that roughly 40 percent of household consumption growth
between 2012 and 2013 could be attributable to this wealth effect. But
Figure 6.6 (C) suggests that this effect was short-lived; we again see no
marked difference among the groups, at least between 2013 and 2016.
In Figure 6.7 (A) we turn to differences among households by home-
ownership status. One would expect households with mortgages
(proxied by households with new houses) to benefit from the lower
interest rates caused by monetary expansion. But these data provide no
Abenomics, Monetary Policy, and Consumption 157

(a) (b)

(c) (d)

(e) (f)

Figure 6.7 FIES results

Source: See data appendix

evidence for such an effect. Both the consumption paths and the aver-
age propensity to consume – see Figure 6.7 (B) – for recent mortgagers
and renters moved roughly in parallel; if anything, the propensity to
consume among recent mortgagers has declined.8
158 Joshua K. Hausman, Takashi Unayama, Johannes F. Wieland

Figure 6.7 (C) examines consumption trends among different income

groups. Selection and endogeneity issues complicate interpretation; still
it is striking how similar the consumption trends during Abenomics are
across the three income groups, a result preserved when we look at the
propensity to consume in the low, middle, and high income groups as
shown in Figure 6.7 (D).
Figure 6.7 (E) shows how consumption patterns vary with age.
This way of categorizing households has the benefit of avoiding
selection issues that complicate the interpretation of the other
graphs: for instance, by changing interest rates, Abenomics may
have changed which households have high levels of debt, assets, or
income. But Abenomics does not cause changes in the age of
household heads. The analysis of the United States in Wong
(2016) suggests that we should find larger consumption responses
among the young, as they tend to be more indebted and thus will
benefit more from lower interest rates. During Abenomics, by
contrast, the consumption paths again reveal no clear differences
among groups. In Figure 6.7 (F) we do observe a large relative
decline in the average propensity to consume among households
with heads aged forty or below. This relative decrease is driven by
an increase in relative income for the young. Thus, in stark con-
trast to Wong (2016), the relative savings rate among young
households increased during Abenomics.
In sum, both in the aggregate and in the cross section it is
difficult to detect an influence of monetary policy on consumer
expenditure. Contrary to basic predictions, the large monetary
expansion during Abenomics appears not to have increased house-
hold consumption even among those groups most likely to be
affected. Of course, this is a tentative conclusion: the cross-
sectional evidence in Figures 6.6 and 6.7 is not definitive, given
concerns about selection and omitted variables. Nonetheless, the
lack of cross-sectional monetary policy effects is striking and sug-
gests that aggregate effects were also weak.
That monetary policy has not had significant positive effects on
consumption does not rule out the possibility that other shocks
have contributed to slow consumption growth. Here we briefly
evaluate the plausibility of three commonly mentioned candidate
shocks.
Abenomics, Monetary Policy, and Consumption 159

6.5.3 Demographics
One of the defining features of the Japanese economy is its declin-
ing total and, in particular, working-age populations. But while
a declining population undoubtedly contributes to slow growth, it
is unclear how it explains a decline in the consumption share of
GDP. Standard life-cycle theory suggests that the savings rate falls
in retirement. Indeed, in the FIES dataset the average propensity to
consume out of disposable income is higher for older households; it
was 85.2 percent in 2016 for households in which the household
head was aged sixty or older, while it was 68.4 percent when the
household head was less than sixty. Thus, the aging of the
Japanese population would be expected to increase consumption
relative to income as retirees draw down their savings, thus lower-
ing the aggregate household savings rate.

6.5.4 The April 2014 Consumption Tax Increase

A natural candidate for slowing consumption growth is the April 2014
consumption tax. In that year, OECD figures show disposable income
falling 1.2 percent. Cashin and Unayama (2016a) show that consumers
reacted to the increase in the consumption tax as if it were a decline in
permanent income. From the perspective of consumer theory, this is
surprising, since the expressed purpose of the consumption tax was to
reduce the deficit rather than to finance additional government spend-
ing. In its “Highlights of the Budget for FY 2014,” the Ministry of
Finance noted (p. 8): “As decided in the recent Comprehensive Reform
of Social Security and Tax, all consumption tax revenue that includes
increased revenue due to the raised consumption tax rate (excluding
national/local consumption tax revenues; excluding current local con-
sumption tax revenue) shall be used for social security.” It goes on to
make clear that most of the money from the tax increase would be used
to maintain existing benefits rather than for benefit expansion (p. 8).
To illustrate why such a tax increase would not be expected to reduce
the consumption growth rate, consider the case of a consumer with
CRRA utility in consumption and a real income stream fyt g. The
discount factor is β and the real rate of interest is 1 + R. Suppose the
consumption tax in each period is τ ct and the income tax in each period
y
is τt . Then the intertemporal budget constraint is
160 Joshua K. Hausman, Takashi Unayama, Johannes F. Wieland

X
∞
ð1 þ τc Þct X
∞ y
ð1  τt Þyt
t
t ¼
t¼0 ð1 þ RÞ t¼0 ð1 þ RÞt

and the Euler equation is

1 1
n¼ cσ ¼ βð1 þ RÞ cσ :
ð1 þ τct Þ t ð1 þ τctþ1 Þ tþ1

Substituting back into the budget constraint yields the consumption

function,
" #1σ
1β X ∞ y
ð1  τt Þyt
ct ¼ :
1 þ τct t¼0 ð1 þ RÞt

With log utility the elasticity of substitution of consumption to the

gross tax rate is equal to the intertemporal elasticity of substitution,
 1σ . Thus, an increase in the consumption tax by 3 percentage points
would cut consumption 3σ percent. But this is a decline in the level of
consumption. The growth rate of consumption is likely to be higher
after the increase in the consumption tax.
To see this, note that the government budget constraint is
X
∞ c y
τ ct þ τ t yt X
∞
gt
t
t ¼ b0 þ ;
t ð1 þ RÞ t ð1 þ RÞt

where b0 is the initial level of debt, and gt is government spending. If

government spending does not change when the consumption tax
increases, then the higher revenue from the consumption tax must be
reflected in either lower future consumption taxes or lower future
income taxes. In either case, consumption in the future will be higher
because the consumption tax shifts a future tax liability to the present.
This logic implies that one ought to expect high consumption growth
after a consumption tax increase, and that consumption will eventually
exceed its pre-tax trajectory.
By contrast, the data show that both the level and the growth rate of
consumption were low after the consumption tax increase in
April 2014 – see Figure 6.4 (B). This is all the more mysterious since
near-term expected future consumption tax increases ought to have
increased consumption through standard intertemporal substitution
Abenomics, Monetary Policy, and Consumption 161

effects.9 These intertemporal substitution effects may, however, be

small (Cashin & Unayama 2016b). The model is only consistent with
slow consumption growth after April 2014 if consumers believe that
the higher consumption tax will precipitate higher government spend-
ing in the future. In that case, permanent income falls causing
a permanent decline in consumption and no increase in the consump-
tion growth rate. Of course, the model omits many frictions and
behavioral elements that affect consumption behavior. We still believe
the model to be useful, however, for two reasons. First, if there were
any context in which a model without borrowing constraints were
likely to be a good description of household consumption behavior, it
might be Japan, where there are relatively few hand-to-mouth con-
sumers (Cashin & Unayama 2016b; Hara, Unayama & Weidner 2016)
and where the consumption tax increase was explicitly advertised to be
for long-run deficit reduction rather than for higher government
spending.
Second, even outside of the strict optimizing framework, the basic
intuition that higher taxes used for deficit reduction will not perman-
ently reduce consumption goes through. Consider the case of a simple
Keynesian consumption function ð1  τct Þct ¼ a þ bð1  τyt Þyt . A higher
consumption tax would reduce consumption today, but then lower
taxes in the future (due to unchanged government liabilities) would
lead to higher consumption in the future. Thus, again, while the con-
sumption tax would reduce the level of consumption, the expected
growth rate of consumption would be high.

6.5.5 Slow Household Income Growth

The household savings rate has been roughly stable during Abenomics.
According to the OECD, it was 2.7 percent in 2012 and 2.6 percent in
2016.10 The OECD forecasts a rise in the savings rate to 3.0 percent in
2017 and 4.2 percent in 2018.11 In line with the stability of the savings
rate, our analysis of the micro-data suggest that monetary policy did
not increase consumption and thus reduce savings.
By definition, the combination of a stable savings rate and slow
consumption growth implies slow disposable income growth. This
is confirmed in the data, which show real net household disposable
income rising at a 0.5 percent annual rate between 2012 and
162 Joshua K. Hausman, Takashi Unayama, Johannes F. Wieland

2016, half the rate of overall GDP growth. Slow disposable

income growth during Abenomics was caused both by higher
taxes (e.g., the consumption tax) and by a declining labor share
in GDP. The national accounts data show that compensation of
employees as a share of nominal GDP fell from 51.3 percent in
fiscal year 2012 to 50.4 percent in fiscal year 2016 and 50.3 per-
cent in fiscal year 2017.12 An explanation of the declining labor
share is beyond the scope of this chapter but is likely important to
understanding the disappointing growth of consumption during
Abenomics. The work by Hoshi and Kashyap in Chapter 7 of
this volume investigates the behavior of wages during
Abenomics; in doing so, it contributes to an explanation for
slow labor income growth and thus, we would argue, slow con-
sumption growth during Abenomics.

6.6 Conclusion
We have reviewed the impact of Abenomics on the Japanese
economy, updating the analysis in Hausman and Wieland (2014)
and Hausman and Wieland (2015). Abenomics continues to have
large, positive effects on financial markets: nominal rates are
lower, stock prices higher, and the yen weaker. Evidence that the
first arrow of Abenomics, expansionary monetary policy, has
caused at least part of these developments comes from an analysis
of announcement effects. On days with significant, positive mon-
etary policy news, financial markets show large, positive move-
ments with comovement across asset classes consistent with
a monetary policy shock.
These financial market developments have not been accompanied by
progress in raising inflation expectations. Market and household infla-
tion expectations are now (in early 2019) roughly where they were in
late 2012, before Abenomics began. Despite this lack of progress on
inflation expectations, working-age adjusted output growth and
employment growth have been strong; output growth would have
been stronger had consumption growth not persistently disappointed
relative to expectations.
The coincidence of weak consumption growth and weak gains in
inflation expectations is puzzling given the very large monetary
Abenomics, Monetary Policy, and Consumption 163

expansion that is the first arrow of Abenomics. We use microdata from

the Family Income and Expenditure Survey to distinguish between two
explanations: either positive monetary policy effects on consumption
were offset by other aggregate shocks, or the aggregate effects of
monetary policy have been small. Only in the first case should we see
large differences in consumption across households differentially
exposed to monetary policy shocks. Across a variety of dimensions,
such as age and asset holdings, we find little evidence of differential
responses to monetary policy shocks, suggesting that monetary policy
also had relatively small effects in the aggregate.
This poses a puzzle for some monetary theories. New Keynesian
models in particular tend to predict large output expansions from
monetary expansions, even after being modified with incomplete
markets or lack of common knowledge (Del Negro, Giannoni &
Patterson 2015; McKay, Nakamura & Steinsson 2016; Angeletos
& Lian 2018). And it seems difficult to conceive of a more clear-
cut case of a central bank committed to monetary expansion than
the Bank of Japan under Abenomics. In other contexts, there have
been large effects on output and consumption of monetary regime
changes, most famously in the United States in spring 1933
(Eggertsson 2008; Hausman, Rhode & Wieland 2019a). We hope
that future work will lead to a better understanding of why mon-
etary policy had less of an impact during Abenomics than either
theory or historical experience predicts.

Notes
1. There is a growing literature on Abenomics and its economic effects.
See, for instance, Botman, Danninger, and Schiff (2015), De Michelis
and Iacoviello (2016), Haidar and Hoshi (2015), Hoshi (2013), Ito
(2013, 2014, 2016), Miyao and Okimoto (2017), and Patrick (2014).
2. See, for instance, Sieg, Yoshikawa, and Kajimoto (2013) and Holland
(2016).
3. For this date, we look at the two-day change in asset prices, since it is
unclear whether the resignation announcement came after financial
markets had closed.
4. This change does not affect our qualitative conclusions.
5. For more on net exports during Abenomics, see Fukuda and Doita
(2016), Kurachi, Ando, and Shoji (2015), Nguyen and Sato (2015),
164 Joshua K. Hausman, Takashi Unayama, Johannes F. Wieland

Rodnyansky (2018), Sasaki and Yoshida (2017), Shimizu and Sato

(2015), and Thorbecke (2017).
6. More precisely, the survey covers 96.6 percent of all households.
Certain small categories, such as one-person student households, are
excluded. See www.stat.go.jp/english/data/kakei/1564.html.
7. Our analysis builds on Hausman and Wieland (2015), in which we
used aggregate consumption data from the FIES to look for
heterogenous effects of monetary policy. The FIES micro-data
allow us to examine more household groupings and to examine
average propensities to consume, which provides a much more
powerful test for the existence of differential effects of monetary
policy across household groups.
8. In a separate paper (Hausman, Unayama & Wieland 2019b), we
discuss how specific housing market institutions in Japan may limit
the transmission of monetary policy to consumption via the housing
market. We show that Abenomics lowered new mortgage interest
rates but did relatively little to lower rates on existing mortgages:
limited refinancing, variable rate mortgages indexed to an
unchanging prime rate, and infrequent household moves all limit
pass-through from lower new mortgage rates to lower rates on
existing mortgages.
9. After raising the consumption tax from 5 percent to 8 percent in
April 2014, the government planned to increase it to 10 percent in
October 2015. This planned increase was then postponed until
April 2017 and then further postponed to October 2019.
10. The discrepancy with the much higher savings rate in FIES occurs
because imputed rent is counted as a part of consumption in the
national accounts but not in the FIES; the national accounts reports
the net saving rate, which includes housing depreciation; disposable
income in the FIES is calculated using only employees and no job
households; and expenditures on high-cost items (e.g., durable goods)
and asset income are underreported in the FIES. See Unayama (2010) (in
Japanese) for details.
11. Data on the savings rate is in the process of being revised due to an error in
the Monthly Labour Survey. The revisions are significant; the Japanese
Cabinet Office estimate of the savings rate for fiscal year 2016 was, for
instance, revised up from 2.5 to 2.8 percent (see https://www.esri.cao.go
.jp/en/sna/data/kakuhou/files/2017/pdf/announce_en_20190125.pdf).
Still, we believe it unlikely that these revisions will change the basic story of
a roughly stable savings rate between 2012 and 2016.
12. The fiscal year runs from April 1 to March 31. Data are from the
Cabinet Office “Annual Report on National Accounts for 2017
Abenomics, Monetary Policy, and Consumption 165

Summary (Flow Accounts)” available at https://www.esri.cao.go.jp/en/

sna/data/kakuhou/files/2017/2017annual_report_e.html. The data on
employee compensation may be slightly affected by ongoing revisions
resulting from a data error in the Monthly Labour Survey. But the
changes appear almost certainly to be so small as to not affect our
conclusions.

References
Angeletos, George-Marios, and Chen Lian. 2018. “Forward guidance
without common knowledge.” American Economic Review 108(9):
2477–2512.
Auclert, Adrien. 2019. “Monetary policy and the redistribution channel.”
American Economic Review 109(6): 2333–2367.
Botman, Dennis, Stephan Danninger, and Jerald Schiff, eds. 2015. Can
Abenomics succeed? Overcoming the legacy of Japan’s lost decades.
Washington, DC: International Monetary Fund.
Campbell, John Y., and Robert J. Shiller. 1988. “The dividend–price ratio
and expectations of future dividends and discount factors.” Review of
Financial Studies 1(3): 195–228.
Cashin, David, and Takashi Unayama. 2016a. “The impact of a permanent
income shock on consumption: Evidence from Japan’s 2014 VAT
increase.” RIETI Discussion Paper Series 16-E-052.
Cashin, David, and Takashi Unayama. 2016b. “Measuring intertemporal
substitution in consumption: Evidence from a VAT increase in Japan.”
Review of Economics and Statistics 98(2): 285–297.
Cloyne, James, Clodomiro Ferreira, and Paolo Surico. 2020. “Monetary
policy when households have debt: New evidence on the transmission
mechanism.” Review of Economic Studies 87(1): 102–129.
Eggertsson, Gauti B. 2008. “Great expectations and the end of the
Depression.” American Economic Review 98(4): 1476–1516.
Fisher, Irving. 1933. “The debt-deflation theory of Great Depressions.”
Econometrica 1(4): 337–357.
Fukuda, Shin-ichi, and Tsutomu Doita. 2016. “Unconventional monetary
policy and its external effects: Evidence from Japan’s exports.”
Developing Economies 54(1): 59–79.
Haidar, Jamal Ibrahim, and Takeo Hoshi. 2015. “Implementing structural
reforms in Abenomics: How to reduce the cost of doing business in
Japan.” NBER Working Paper No. 21507.
Hara, Ryota, Takashi Unayama, and Justin Weidner. 2016. “The wealthy
hand to mouth in Japan.” Economics Letters 141: 52–54.
166 Joshua K. Hausman, Takashi Unayama, Johannes F. Wieland

Hausman, Joshua K., and Johannes F. Wieland. 2014. “Abenomics:

Preliminary analysis and outlook.” Brookings Papers on Economic
Activity 2014(1): 1–63.
Hausman, Joshua K., and Johannes F. Wieland. 2015. “Overcoming the lost
decades? Abenomics after three years.” Brookings Papers on Economic
Activity 2015(2): 385–413.
Hausman, Joshua K., Paul W. Rhode, and Johannes F. Wieland. 2019.
“Recovery from the Great Depression: The farm channel in spring
1933.” American Economic Review 109(2): 427–472.
Hausman, Joshua K., Takashi Unayama, and Johannes F. Wieland. 2019.
“Abenomics, the housing market, and consumption.” Keizai Bunseki
200: 37–62.
Holland, Tom. 2016. “The hidden agenda behind Japan’s Abenomics.”
South China Morning Post, September 12.
Hoshi, Takeo. 2013. “Will Abenomics restore Japan’s growth?” Prepared
for Macro Economy Research Conference “Prospects for Growth in the
World’s Four Major Economies,” sponsored by Nomura Foundation
and held at Brookings Institution, Washington, DC.
Ito, Takatoshi. 2013. “‘Abenomics’: Early success and prospects.”
Economy, Culture, and History: Japan Spotlight, September/
October, 4–7.
Ito, Takatoshi. 2014. “We are all QE-sians now.” IMES Discussion Paper
Series 2014-E-5.
Ito, Takatoshi. 2016. “Japanization: Is it endemic or epidemic?” NBER
Working Paper 21954.
Kihara, Leika. 2016. “BOJ eases policy by doubling ETF buying,
underwhelms expectations.” Reuters, July 29.
Kurachi, Yoshiyuki, Masatoshi Ando, and Kanako Shoji. 2015. “Changes in
the environment surrounding Japan’s exports: An approach focusing on
global trade volume and export share.” Bank of Japan Review April:
1–9.
Kuroda, Haruhiko. 2013. “Overcoming deflation and after.” Speech at the
Meeting of Councillors of Nippon Keidanren (Japan Business
Federation), Tokyo, December 25.
McKay, Alisdair, Emi Nakamura, and Jón Steinsson. 2016. “The power of
forward guidance revisited.” American Economic Review 106(10):
3133–3158.
Michelis, Andrea De, and Matteo Iacoviello. 2016. “Raising an inflation
target: The Japanese experience with Abenomics.” European Economic
Review 88: 67–87.
Abenomics, Monetary Policy, and Consumption 167

Miyao, Ryuzo, and Tatsuyoshi Okimoto. 2017. “The macroeconomic

effects of Japan’s unconventional monetary policies.” RIETI
Discussion Paper Series 17-E-065.
Negro, Marco Del, Marc P. Giannoni, and Christina Patterson. 2015. “The
forward guidance puzzle.” Working Paper.
Nguyen, Thi-Ngoc Anh, and Kiyotaka Sato. 2015. “Asymmetric
exchange rate pass-through in Japanese exports: Application of the
threshold vector autoregressive model.” RIETI Discussion Paper 15-
E-098.
Patrick, Hugh. 2014. “Japan’s Abenomics: So far so good, but will it
prevail?” Columbia Business School Center on Japanese Economy and
Business Occasional Paper Series 67.
Rodnyansky, Alexander. 2018. “(Un)competitive devaluations and firm
dynamics.” Working Paper.
Romer, Christina D. 2014. “It takes a regime shift: Recent developments in
Japanese monetary policy through the lens of the Great Depression.” In
NBER Macroeconomics Annual 2013, edited by Jonathan A. Parker
and Michael Woodford, pp. 383–400. Chicago, IL: University of
Chicago Press.
Sakamoto, Mie. 2015. “BOJ adjustments signal limit to monetary easing.”
Japan Times, December 19.
Sasaki, Yuri, and Yushi Yoshida. 2017. “Factor decomposition of Japan’s
trade balance.” RIETI Discussion Paper 17-E-042.
Shimizu, Junko, and Kiyotaka Sato. 2015. “Abenomics, yen depreciation,
trade deficit, and export competitiveness.” RIETI Discussion Paper 15-
E-020.
Sieg, Linda, Yuko Yoshikawa, and Tetsushi Kajimoto. 2013.
“Special Report: The deeper agenda behind ‘Abenomics’.”
Reuters, May 23.
Thorbecke, Willem. 2017. “A yen for change: The strong yen and the
Japanese automobile industry.” RIETI Discussion Paper 17-E-005.
Unayama, Takashi. 2010. “Discrepancy between saving rates in SNA and
Family Income and Expenditure Survey and its implications
(Japanese).” RIETI Discussion Paper Series 10-J-003.
Unayama, Takashi, and Norihiro Komura. 2014. “Effects of stock prices on
consumption: Evidence from stock price hike during Abenomics.” PRI
Discussion Paper Series (No. 14A-09), Ministry of Finance (in
Japanese).
Wong, Arlene. 2016. “Population aging and the transmission of monetary
policy to consumption.”
168 Joshua K. Hausman, Takashi Unayama, Johannes F. Wieland

6. A Data appendix
Campbell-Shiller decomposition: Total returns (net and gross) and
price of MSCI Japan index. www.msci.com/end-of-day-data-
country.
Consumption Activity Index: Bank of Japan. https://www.boj.or.jp/
en/research/research_data/cai/index.htm/.
CPI: Ministry of Internal Affairs and Communications, Statistics
Bureau. www.stat.go.jp/english/data/cpi/.
Disposable income: From the OECD, see https://data.oecd.org/hha/
household-disposable-income.htm.
Earnings and Hours: Statistics Japan.
Employment rate: FRED series LREM74TTJPA156 N (Japan) and
LREM74TTUSA156S (USA).
Exchange rate – nominal: We use noon buying rates in New York, as
reported by the Federal Reserve (FRED series DEXJPUS). For the
announcement table we use the Bloomberg series USDJPY.
Exchange rate – real: Bank for International Settlements trade-
weighted index. We use the broad index including sixty-one coun-
tries. www.bis.org/statistics/eer/broad1901.xlsx.
Forecasts: Consensus Economics.
Government bond yields: Bloomberg series GJGBX, where X is the
years to maturity.
Household savings rate: National accounts data from the OECD; see
https://data.oecd.org/hha/household-savings-forecast.htm.
Family Income and Expenditure Survey from https://www.e-stat
.go.jp/en/stat-search/files?page=1&layout=datalist&tou
kei=00200561&tstat=000000330001&cycle=7&year=20160&
month=0&tclass1=000000330001&tclass2=000000330004&t
class3=000000330006&stat_infid=000031586344&result_
back=1&result_page=1.
Inflation swaps: Bloomberg series JYSWIT2, JYSWIT10.
Money market interest rate: IMF, International Financial Statistics.
Multifactor productivity: OECD multifactor productivity. https://d
ata.oecd.org/lprdty/multifactor-productivity.htm.
National Accounts: Cabinet Office. Pre-1994 data are from the IMF
GDP index, International Financial Statistics. We splice the two data
series together, assuming the level of GDP is the same in 1994:Q1.
Abenomics, Monetary Policy, and Consumption 169

National Accounts (USA): Real GDP: Federal Reserve Bank of

St. Louis, FRED database series GDPCA (annual) and GDPC1
(quarterly). Consumption share: NIPA table 1.1.10.
Stock prices: Nikkei 225 price data are from Yahoo finance and
FRED series NIKKEI225. TOPIX is Bloomberg series TPX.
Unemployment rate: FRED series LRUNTTTTJPM156N.
Working age population: FRED series LFWA74TTJPA647N
(Japan) and LFWA74TTUSA647N (USA).
7 The Great Disconnect
The Decoupling of Wage and Price Inflation
in Japan
takeo hoshi and anil k kashyap

7.1 Introduction
Japan’s persistently low inflation is puzzling in two respects. First, the
Bank of Japan (BOJ) since 2013 has engaged in unprecedented monetary
stimulus. This began with a massive asset purchase program (concentrated
on Japanese government bonds), which was followed by driving short-
term interest rates below zero and ultimately targeting the yield on the ten-
year bond rate. Despite these efforts, and the conventional view that
a determined central bank can raise the price level, inflation has remained
below the BOJ’s target of 2 percent per year.
Second, the labor market in Japan has been relatively tight. There is
no broadly accepted estimate of the natural rate of unemployment in
Japan, but speeches by BOJ officials often indicate that they believe
that, once the unemployment rate reaches about 3 percent, the only
people unemployed are in that situation because of search frictions.
With an unemployment rate below 2.5 percent, one would expect to see
rising wage and price inflation (even if the monetary policy stance were
neutral). Likewise, standard estimates of potential growth in Japan are
about 0.5 percent. The growth rate for the economy has regularly been

* The views expressed are those of the authors only and should not be attributed to
any of the institutions with which they are affiliated. We are grateful to
comments by Joshua Hausman, Yuzo Honda, Daiji Kawaguchi, Sachiko
Kuroda, Etsuro Shioji, Shigenori Shiratsuka, Takashi Unayama, Johannes
Wieland, and participants of the conferences on the Abe administration at
Stanford University (February 2018 and March 2019), the NBER Japan Project
Meeting (July 2019), the Japan Economic Association Meeting (October 2019),
and seminars at the Financial Services Agency of Japan, Hitotsubashi University,
Kobe University, and University of California, Santa Cruz. Kashyap thanks the
Initiative on Global Markets at the University of Chicago for financial support.
We thank Masahiro Abe, Masayoshi Amamiya, Yoichi Arai, and Jess Diamond
for helpful conversations at the early stages of this research. Dongchen Zou and
Jun Xu provided expert research assistance. All mistakes are ours alone.

170
The Great Disconnect 171

above this level since early 2013. Thus, adherents of the Phillip’s curve
might also find the subdued inflation puzzling.
In this chapter we explore the role of Japan’s dual labor market in
driving the low inflation outcomes in Japan. Economists have long studied
the differences in employment conditions between full-time regular work-
ers and part-time employees in Japan. As the Bank of Japan (2017) notes,
there has been a steady rise in part-time workers, and their wages are more
sensitive to market conditions than full-time workers. However, the level
of part-time workers’ wages is far below those for full-time workers. This
change in composition has depressed firms’ wage bills. We investigate the
effects of this factor on inflation outcomes in Japan.
We begin our analysis in 1981, just after the second global oil shock,
which marked the beginning of Japan’s relatively low inflation era.
Starting in 1981, inflation gradually drifted down before rising a bit and
then settling around zero by the middle of the 1990s. This pattern holds
for both headline measures of inflation and those that exclude food and
energy prices. We show that the behavior of wages and prices in this
period is very conventional. Wages are reliably connected to labor market
tightness and wage inflation has important explanatory power for prices.
Sometime in the late 1990s, these patterns begin to change. We date the
shift as starting in 1998. Precisely identifying the exact break point is hard,
but there are several good reasons to impose a break in 1998. One is that
the Bank of Japan gains its legal independence then. A second is that the
acute phase of the Japanese banking crisis begins in late 1997. A third,
most importantly, is that past research has found that a number of other
structural relations in the economy changed around this time (Hamada,
Kashyap & Weinstein 2011).1 Thus, our analysis focuses on a number of
differences between the pre- and post-1998 eras.
We organize our investigation into five parts. First, Section 7.2
briefly reviews the literature on inflation in Japan. Past work can be
separated into two broad approaches. One begins with the kind of
Phillips curve that has been estimated in many countries and then tries
to modify it to account for particular factors in Japan. The other is
closer to our approach and puts the structure of the labor market at the
center of the story.
Section 7.3 presents the evidence that supports our contention that it
is wise to analyze inflation separately before and after 1998. We include
both our own statistical analysis and review relevant related prior work
that argues for the important structural break around 1998.
172 Takeo Hoshi and Anil K Kashyap

Section 7.4 covers our summary of the labor market facts. One novel
aspect of our analysis is to study separately the determination of bonuses
from wages. We find that the bonuses and wages exhibit different dynam-
ics and that their responsiveness to labor market tightness also varies.
More importantly, we find that there was a significant shift in the connec-
tion between the unemployment rate and wage inflation after 1998.
Section 7.5 investigates the joint links between prices and wages. We
show how the trends in wages and prices changed in 1998 and how the
trend changes also correspond to a shift in the dynamic relationship
between the two series.
Section 7.6 presents our conclusions.

7.2 Literature Review

There is an enormous literature on the Japanese inflation experience over
the last two decades. Since the BOJ was given formal independence in
1998, many of these papers have focused on the operating tactics and
strategies of the BOJ. Many observers criticized the BOJ for being ambiva-
lent about whether it should fight the ongoing deflation and if so whether
it in fact had the necessary tools.
One influential analysis by Bernanke (2000) argued that the BOJ had
been insufficiently aggressive in using the tools at its disposal to combat
deflation. In a 2003 speech, Bernanke (2003) followed up this analysis and
gave a number of suggestions about steps that could be taken to raise the
price level in Japan. Svensson (2003) offered what has come to be known
as the “foolproof” way to overcome deflation by relying on monetary
induced changes in the exchange rate to eventually drive prices up.
These diagnoses were in part a reaction to the BOJ’s contention
that deflation was occurring for reasons outside of its control. For
example, in January 2002 Masaru Hayami, the first post-
independence BOJ governor, gave a speech describing recent BOJ
policy decisions (Hayami 2002). One key passage was (with emphasis
added by us):
We changed the main operating target from call rates to a quantitative
indicator of liquidity, that is, the outstanding balance of current accounts
at the Bank of Japan, and we have substantially increased the balance.
Because this was a very drastic easing measure and was unprecedented in
the history of central banking worldwide, we had to consider various issues
The Great Disconnect 173

very carefully before deciding it. Above all, it was uncertain whether and how
expansion of liquidity provision in a situation where the rate is virtually
zero percent would influence the economy. Needless to say, the mechanism
through which such an expansion affects the economy has not been demon-
strated either theoretically or empirically.

As late as 2011, the Bank of Japan was still making similar arguments.
For instance, Governor Shirakawa gave a speech (with a nearly identi-
cal title to Hayami’s) in which he claimed (Shirakawa 2011):
The prolonged, albeit moderate, deflation since the end of the 1990s cannot be
explained only by short-term and cyclical factors. A more fundamental cause is
the long-term downtrend in the growth potential of Japan’s economy. When the
growth rate continues to decline for a protracted period, people’s expectations
for future income growth are reduced and firms and households restrain their
spending. As a result, downward pressure on prices continues . . .
It is sometimes said that deflation will be solved only if the central bank
provides liquidity more aggressively. Provision of ample liquidity is import-
ant, of course, but this alone does not solve the problem of deflation.
These arguments were met with continued criticisms from external
observers.2
Yet, since 2013, when Governor Kuroda took the helm at the BOJ,
these critiques no longer apply. The BOJ has become very aggressive
and implemented many of the ideas advocated by its critics. It has
stopped questioning its responsibility for inflation outcomes and has
disavowed the idea that it is impossible to end deflation. The recent
literature, therefore, has shifted to try to explain why, even with
extremely accommodative monetary policy, inflation in Japan has
remained below 2 percent.
One strand of research takes a Phillips curve approach and seeks to
relate inflation to measures of slack in the economy and expectations.
Bernanke (2017) surveys this literature and adds his own conjectures
about the reasons for persistently low levels of Japanese inflation. The
conventional view is that the sensitivity of inflation to measures of slack
have declined in many advanced economies including Japan. The Bank
of Japan (2017) gives a specific example of this argument, contrasting
the estimated slope of the Phillips curve before and after 2013 and
concluding that the curve is much flatter after 2013.
This reduced sensitivity is supplemented with the view that expect-
ations for inflation in Japan are unanchored. Kuroda (2017) argues
174 Takeo Hoshi and Anil K Kashyap

that expectations in Japan are much more adaptive or backward-

looking than in other countries, so that factors such as oil price swings
that change realized inflation can be particularly important in Japan.
Bernanke (2017) argues the long period when inflation undershot the
BOJ’s target might still be influencing expectations. In this sense, the
BOJ may still be suffering from the past performance despite the change
in leadership.
Not everyone accepts the idea that the Phillips curve is broken in
Japan. Gagnon (2017) argues instead that potential output and full
employment have been chronically mismeasured in Japan. As he
notes, standard measures of potential output are constructed under
the assumption that, over the sample being considered, output is
both above and below potential. If instead the economy was con-
sistently depressed so that potential output was never reached (and
the economy was never at full employment), then standard statistical
approaches to deduce potential are doomed. Gagnon argues that
this was likely the case in Japan between 1995 and 2016, so the
more puzzling issue is why there was not more deflation. He conjec-
tures that downward wage rigidity prevented an acceleration of
deflation. He predicts that, as the economy continues to recover, it
will soon pass the level of full employment and inflation will appear
when it does. This explanation seems less persuasive after having
seen that steady growth that continued through the middle of 2019
did not lead to a pickup of inflation.
Some researchers have focused on the wage Phillips curve rather than
the price Phillips curve, and they have emphasized the role of the labor
market structure in influencing wage and price dynamics. An excellent
collection of the essays on this topic is Genda (2017). The sixteen
chapters in that volume explore a variety of reasons why wages have
stagnated in the presence of a tight labor market. Three common
themes emerge from its essays. One is that the dual labor market is
important and that full-time workers (or regular workers) and part-
time workers (or nonregular workers) are paid and treated very differ-
ently. The increasing share of part-time workers in the recent decades is
highlighted in many of the chapters.
A second theme is the prevalence of downward wage rigidity and
the extent to which firms hesitate to cut wages. One essay, by
Yamamoto and Kuroda, notes that there are many firms that never
cut wages over a ten-year period. These same firms also fail to raise
The Great Disconnect 175

wages when the economy improves. Some authors call this “upward
wage rigidity.”
A third consideration is the supply side of the labor market. One
aspect of labor supply is the steady demographic shift in Japan. As
society has aged, many elderly people continue to want to work past
mandatory retirement age and have stayed in the labor market, often by
taking part-time positions. This trend has perhaps been supplemented
by a collapse in training within firms. On-the-job-training was once an
important factor in raising skill levels, and the decline in this practice
has reduced the supply of certain types of high-skilled workers. In cases
where there is a mismatch between the hiring needs of firms and the
supply of suitable workers, wage increases may not be very effective in
clearing the market, and the wage can stagnate even with a tight labor
market.
The hypothesis that we pursue in this chapter is very much in line
with the arguments found in Genda (2017). None of the essays specif-
ically matches our story, though it is rightly viewed as a convex com-
bination of the different pieces of evidence in the book. Our main
contribution is to explicitly connect inflation developments to the
wage developments.

7.3 Structural Breaks in Japanese Inflation

This chapter examines two price inflation indicators and two wage
inflation indicators. Figure 7.1 shows the two price series from the
first quarter of 1971 to the last quarter of 2018: the so-called “core-
core” inflation rate that is the consumer price index excluding food and
energy (the solid red line in the figure) and the headline inflation rate as
computed from the total consumer price index (the dashed blue line in
the figure). For both of the CPI measures we adjusted the series to
remove the mechanical effects of the three increases in the consumption
tax on April 1 in 1989, 1997, and 2014. To eliminate seasonal fluctu-
ations, the figure shows the four-quarter change in the natural loga-
rithm of the series. Throughout our analysis, we will measure wage and
price inflation this way.
We construct the wage series by dividing a measure of earnings by
a measure of hours worked. We distinguish between two components
of earnings for a worker: wages (including overtime), and bonuses.3
The source of the data is the Monthly Labour Survey.4 The wage series
176 Takeo Hoshi and Anil K Kashyap

20%

16%

12%

8%

4%

0%

−4%
1971
1973
1975
1977
1980
1982
1984
1986
1989
1991
1993
1995
1998
2000
2002
2004
2007
2009
2011
2013
2016
2018
Headline Inflation Core-Core Inflation

Figure 7.1 Japanese price inflation, 1971–2018

Source: Authors’ calculation based on Ministry of Internal Affairs and
Communications, Statistics Bureau, Consumer Price Index

is calculated by dividing the contracted salary by the total number of

hours worked (including overtime). The bonus series is calculated by
dividing the bonus payments received in the last twelve months by the
total number of hours worked for the last twelve months. We look at
the last twelve months because the bonus is typically paid only twice
a year: once in June or July and then in December. We use series that
relate to establishments with thirty or more employees. Figure 7.2
shows the two wage inflation indicators: wage inflation in the solid
blue line and bonus inflation in the dashed red line.
Simple inspection of Figures 7.1 and 7.2 reveals the three main
points that are important for our subsequent analysis. First, the
period from the start of the sample for price inflation series and
wage inflation series through 1980 looks qualitatively different
from the rest of the data. This period contains the end of the “high-
growth era,” when Japan’s economy averaged more than 9 percent
growth, and the two spikes in price inflation that accompanied the
disruptions in global oil markets in 1974–1975 and 1980, although
the spike for 1980 is smaller than that for 1974–1975 and is nearly
absent for wage and bonus inflation. Price inflation, even excluding
food and energy, is much higher and more volatile than the rest of the
sample. Our measures both of inflation and of wage and bonus
growth rise sharply and then fall during this period.
The Great Disconnect 177

35%
30%
25%
20%
15%
10%
5%
0%
−5%
−10%
−15%
1971
1973
1975
1977
1979
1981
1983
1985
1987
1989
1991
1993
1995
1997
1999
2001
2003
2005
2007
2009
2011
2013
2015
2017
Wage Inflation Bonus Inflation

Figure 7.2 Japanese wage inflation, 1971–2018

Source: Authors’ calculation based on Ministry of Health, Labour and Welfare,
Monthly Labour Survey

The episode of the 1970s is so different from what followed after

the 1980s that any econometric exercises to examine the connec-
tions between the variables would be driven by the events of the
1970s (even though they are not representative of the rest of the
sample). We do not believe the experience from the 1970s is
relevant for our question about why inflation has been so low for
the last two decades. So for the remainder of the paper we begin
the analysis in 1981.
The second important observation is that, starting in 1981, inflation
begins falling and declines gradually through 1988 before rising a bit.
For the entire time through 1997 inflation sits below 4 percent.
Headline inflation moves more than the core-core measure, but the
two variables track each other closely. Wages also move along with the
price measures.
Finally, note that all these measures fall to about zero by 1998.
Starting in 1998 all price and wage inflation rates hover around zero.
Price inflation measures also become more stable after 1998 than in the
prior sample.
To confirm these visual impressions that the periods before and after
1998 are different, we estimated Box-Jenkins-style time series models
for each of the four series in Figures 7.1 and 7.2 and tested for the
stability of the models. Table 7.1 consists of four panels showing the
178 Takeo Hoshi and Anil K Kashyap

best-fitting models for each of the four series. The models were selected
to fit the 1981 to 1997 sample as parsimoniously as possible. The
model was then reestimated over the post-1998 sample to investigate
whether the dynamics of the series were stable.

Table 7.1 ARMA models for price and wage inflation

Headline CPI inflation

Sample: 1981–1997 Sample: 1998–2018

Variable Coefficient Standard error T statistic Coefficient Standard error T statistic

Constant 0.0240 0.0124 1.93 0.0003 0.0023 0.13

AR(1) 0.9473 0.0451 20.99 0.7986 0.0757 10.55
MA(3) 0.3772 0.1190 3.17 0.1406 0.1217 1.15
MA(5) 0.4360 0.1488 2.93 −0.1945 0.1096 −1.77
R2 0.88 0.67
Approximate 0.78 0.01
P-value Q (at lag 17) (at lag 19)
statistic

Core-core inflation

Sample: 1981–1997 Sample: 1998–2018

Variable Coefficient Standard error T statistic Coefficient Standard error T statistic

Constant 0.0291 0.0135 2.15 −0.0014 0.0023 −0.60

AR(1) 0.9802 0.0333 29.43 0.8968 0.0483 18.55
MA(2) 0.3842 0.1019 3.77 0.2015 0.1070 1.88
R2 0.93 0.84
Approximate 0.95 0.58
P-value (at lag 17) (at lag 19)
Q statistic

Wage inflation for all workers

Sample: 1981–1997 Sample: 1998–2018

Variable Coefficient Standard error T statistic Coefficient Standard error T statistic

Constant 0.0370 0.0068 5.41 0.0022 0.0021 1.05

AR(1) 0.5827 0.1418 4.11 0.2352 0.1111 2.12
AR(2) 0.3089 0.1242 2.49 0.3026 0.1242 2.44
R2 0.88 0.67
Approximate 0.78 0.01
P-value (at lag 17) (at lag 19)
Q statistic
The Great Disconnect 179

Table 7.1 (cont.)

Bonus inflation for all workers

Sample: 1981–1997 Sample: 1998–2018

Variable Coefficient Standard error T statistic Coefficient Standard error T statistic

Constant 0.0391 0.0130 3.01 −0.0065 0.0170 −0.38

AR(1) 0.8997 0.0482 18.67 0.8517 0.0556 15.33
R2 0.80 0.72
Approximate 0.53 0.00
P-value (at lag 17) (at lag 19)
Q statistic

For the headline CPI inflation rate, the best-fitting model had one
autoregressive term and lagged moving average terms at quarters three
and five. For the core-core inflation rate, the best fitting model also had
one autoregressive term and a lagged moving average term at quarter
two. The wage series was modeled as having two lagged autoregressive
terms, at quarters one and two. The best-fitting model for the bonus
series had a single autoregressive term at quarter one.
We draw four conclusions from these estimates. First, for the two price
inflation series, the coefficient on the autoregressive term is close to one,
indicating a high degree of persistence. This is consistent with (unreported)
unit root test results that suggested that inflation series (especially headline
CPI) is I(1). As Cecchetti et al. (2017) observe in many countries,
a relatively good model for inflation assumes that the first difference of
inflation follows a first-order moving average process. Our result suggests
that this characterization applies for Japan. Second, for both of the price
inflation series, the levels of persistence fall after 1998. Both the autore-
gressive coefficients and the moving average coefficients are smaller in
magnitude and have a lower degree of statistical significance. For the
headline series, one of the MA coefficients flips sign.
Third, the two wage series are also very persistent (but not as much
as are prices). The models have large coefficients on the autoregressive
terms. Fourth and finally, as with the price inflation measures, the
persistence declines after 1998, although for the bonus series the drop
is inconsequential.
Formal tests for coefficient stability show that, for all four series, the
coefficients across the two samples are significantly different. For
180 Takeo Hoshi and Anil K Kashyap

example, a Chow test that supposes a break date of the first quarter of
1998 rejects the equality of coefficients at levels of significance below
1 percent for three of the four series. The exception is the core-core
inflation rate, for which the standard Chow test using F-statistics fails
to reject the null hypothesis that the coefficients are equal across the
two sample periods. In that case, the test using the likelihood ratio
rejects the null at 5 percent level, and the test using the Wald statistic
rejects the null at below 1 percent.
Alternatively, a Quandt-Andrews test allows for a potential break at
any date (other than those that are too close to the beginning or the end of
the sample period). We consider a potential break in the middle 85 percent
of the sample (1986:Q4 to 2013:Q2). This test also generally rejects the
null hypothesis that there were no break points. Again the exception is the
core-core inflation rate. According to the likelihood ratio test, the most
likely breakpoint of 1995:Q2 is not statistically significant, while the Wald
statistic suggests the breakpoint of 1993:Q4 and is statistically significant.
For headline inflation, the break date that is mostly likely is 1999:Q4
according to the likelihood ratio test and 2008:Q3 for Wald test. For wage
inflation, the likelihood ratio test selects 2000:Q1 and Wald test selects
1998:Q1 as the most likely breakpoint. Suggested breakpoints are a little
bit earlier for bonus inflation and are 1995:Q2 and 1992:Q2.
Combining all of this direct evidence, with the aforementioned ana-
lysis in Hamada, Kashyap, and Weinstein (2011) suggesting that
a number of other macro relations changed around 1998, leads us to
examine separately the pre- and post-1998 data for the remainder of
the paper. Our focus now shifts to trying to understand what might be
responsible for the shift.

7.4 Wages and the Labor Market

The structure of the labor market began to shift following the Japanese
banking crisis. Figure 7.3 shows annual data for the levels of average
hourly wages (as opposed to rates of changes in Figure 7.2) in Japan from
1993 to 2018. The starting point for the graph is dictated by data
availability because the full-time and part-time distinction only becomes
available in 1993. Each panel of the figure shows the wage rate for full-
time and part-time workers along with the average for the economy that
accounts for both types (the green solid line) for wages (panel A) and
bonuses (panel B). To calculate these wage series, we start with an estimate
The Great Disconnect 181

2,200 1300

2,100 1200

2,000 1100

1,900 1000

1,800 900

1,700 800
2001
1993
1994
1995
1996
1997
1998
1999
2000
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
All (left axis) Full-time (left axis) Part-time (right axis)

Figure 7.3 Wages and bonuses in Japan, 1993–2018

Source: Authors’ calculation based on Ministry of Health, Labour, and
Welfare, Monthly Labour Survey

for the level of wages in yen for full-time, part-time, and average workers
in 2018, using earnings in yen (and number of hours worked) reported in
the Monthly Labour Survey. We then project the wage series back in time
using indices of wages and bonuses by different types of workers at
establishments with thirty or more employees.
Panel A of Figure 7.3 shows the movements of the level of wages. Note
that the left axis is the scale for wage series for full-time workers and all
workers and that it differs from the scale on the right axis that measures
wage for part-time workers. The average level of wages peaked in 1999
and did not exceed that level again until 2017. The patterns for full-time
and part-time wages are very different. The wage for full-time workers has
been remarkably steady, sitting around 2,100 yen per hour, and have only
slightly increased during the last twenty years. The part-time wage has
risen by about 20 percent, from below 1,000 yen to just under 1,200 yen
per hour, between 1999 and 2018. The part-time wage also has been more
sensitive to improved business conditions since 2013.
Bonus earnings divided by total hours worked is very different than
the wage levels. The average bonus peaked in 1997 and then trended
down, as Panel B of Figure 7.3 shows. The level has been recovering
since 2014, but the 2018 level is less than 80 percent of the 1997 level.
The movements of the bonus for full-time workers are essentially the
same as those of the average. This is because few part-time workers
182 Takeo Hoshi and Anil K Kashyap

Number of Employees (million)

23 8

22 7

21 6

20 5

19 4

18 3
1990

2008
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
Full-time (left axis) Part-time (right axis)

Figure 7.4 Composition of the Japanese workforce (establishments with more

than thirty employees)
Source: Authors’ calculation based on Ministry of Health, Labour, and
Welfare, Monthly Labour Survey

receive bonuses. As we can see from Panel B, the level of bonus divided
by total hours worked has been very low (and trending down) for
part-time workers.
Figure 7.4 shows that the main reason for the average regular and
overtime wages staying flat despite rising regular and overtime
wages for part-time and (to a lesser degree) full-time jobs is the
changing composition of the workforce. We use the employment
index for each type of job at establishments with thirty or more
employees in the Monthly Labour Survey and convert it into actual
number using the ratio of the index value to the actual value
in year 2018. Here we show the annual average of each series.
The number of full-time workers is measured on the left axis (in
millions) while the level of part-time workers is measured on the
right axis (in millions). The number of full-time workers peaked in
1994, dipped a bit afterward, then began a precipitous decline in
1998 that continued through 2003. The number of part-time work-
ers was relatively stable until 1996 and then began a climb that
continued over the next twenty years. The replacement of high-cost
full-time employees with much less expensive part-time workers is
a major reason why average regular and overtime wages in 2018
are almost the same as in 1997.
The Great Disconnect 183

Section 7.3 documented changes of dynamics for both prices and

wages. For wages, the reason for the changing dynamics is specifically
attributable to the interaction of the sharp recession that began in 1997
and the dual labor market. Unfortunately, data on wages for part-time
and regular workers in the Monthly Labour Survey is only available
starting in 1993. This prevents us from being able to directly show that
there was a break in the link between labor market conditions and
wages for full- and part-time workers. Consequently, we proceed indir-
ectly, building up evidence in three steps.
First, using data on average wages for all workers, we show that the
connection between average wages and labor market tightness changed
around 1997. We do this by augmenting the ARMA models for the two
measures of wage inflation that we considered Section 7.3 with the
unemployment rate, which is the most widely used measure of labor
market slack (or tightness) and checking whether the response of wage
to labor market conditions changed before and after 1997.
This analysis is presented in Table 7.2. We use the same autoregressive
specifications that we used in Section 7.3 for each wage measure. Panel

Table 7.2 ARMAX models for wage inflation with unemployment rate
(UNEMP)

A. Wage inflation for all workers

Sample: 1981–1997 Sample: 1998–2018

Variable Coefficient Standard error T statistic Coefficient Standard error T statistic

Constant 0.1074 0.0120 8.94 0.0352 0.0116 3.04

AR(1) 0.2659 0.1141 2.33 0.0945 0.1040 0.91
AR(2) 0.2305 0.1384 1.67 0.2016 0.1209 1.67
Lagged UNEMP −0.0756 0.0121 −6.25 −0.0234 0.0078 −2.99
R2 0.78 0.28

B. Bonus inflation for all workers

Sample: 1981–1997 Sample: 1998–2018

Variable Coefficient Standard error T statistic Coefficient Standard error T statistic

Constant 0.0888 0.0473 1.88 0.1334 0.0394 3.39

AR(1) 0.8635 0.0822 10.50 0.7708 0.0716 10.76
Lagged UNEMP −0.0195 0.0189 −1.03 −0.0342 0.0084 −4.07
R2 0.80 0.76
184 Takeo Hoshi and Anil K Kashyap

A of Table 7.2 examines the regular wage inflation. The panel shows the
estimation results for two different sample periods: 1981–1997 and
1998–2018. For both samples, one period lag of the seasonally adjusted
unemployment rate is statistically significant, but the size of the coeffi-
cient in the later period is less than one-third that of the earlier period.
Thus, the wage does respond to labor market conditions in the expected
direction in both periods, but the magnitude of the effect drops substan-
tially after 1998. These estimates imply that, all else equal, between
1981 and 1997, when unemployment rose by 0.2 percentage points in
one quarter, inflation would fall by about 0.15 percentage points in the
next quarter. A similar increase after 1998 would predict less than
a 0.05 percentage point reduction in the next quarter.
The results for bonus wage inflation in Panel B are very different. For
the bonus, the responsiveness to the unemployment rate increased after
1998. The estimated coefficient on the unemployment rate for the pre-
1998 period is not statistically significant. The point estimate suggests
that a 0.2 percentage point reduction in the unemployment rate would
raise bonuses by about 0.4 percentage points before 1998, while the
reduction in the same magnitude would lead to a 0.7 percentage point
increase in bonuses in the post-1998 period.
Given the fraction of total compensation that comes from bonuses,
the overall sensitivity of earnings to labor market conditions is much
lower after 1998. For instance, in 2018 the fraction of overall earnings
from regular and overtime wages over the course of a year was 80 per-
cent. So the fact that these wages have decoupled from labor market
conditions is much more relevant than the increase in sensitivity for
bonus wages.
The flattening of the wage Phillips curve in Japan has been noted in
various studies, including Muto and Shintani (2020). So the estimates in
Table 7.2 can be read as confirming the observation that average wages
now seem only weakly connected to excesses in the labor market.
To investigate what might be behind this change, we turn to data on
market conditions for full-time workers. We do this by looking at how
the proportion of enterprises that report excess full-time workers
changed over time using data from Rōdō Keizai Dōkō Chōsa (Survey
of Labour Economy Trend) conducted by the Ministry of Health,
Labour, and Welfare in February, May, August, and November of
each year.
The Great Disconnect 185

Proportion of Firms with Excess Full Time Workers

(%)
35

30

25

20

15

10

0
1990

1992

1994
1995
1996
1997

1999

2001

2003

2005

2007
2008
2009
2010

2012

2014

2016

2018
1991

1993

1998

2000

2002

2004

2006

2011

2013

2015

2017

2019
Figure 7.5 Proportion of firms with excess full-time workers (percentage)
Source: Ministry of Health, Labour, and Welfare, Survey of Labour Economy
Trend

Figure 7.5 shows the proportion of enterprises that report having

excessive full-time workers. This variable is clearly cyclical, and the level
of excess workers jumps in 1998 and stays elevated till 2004, even though
there is a partial economic recovery in 2000. By 2005, the excess employ-
ment indicator is back to the 1997 level, and it remains there for two more
years before spiking again during the global financial crisis. Looking back
at Figure 7.4, it is clear that the periods of excess correspond to times
when full-time employment falls. Our contention is that this adjustment
was incomplete, and, if firms had not hoarded workers, the number of
full-time employees would have fallen much more sharply – and presum-
ably would have recovered more quickly.
This hunch leads to our third piece of evidence that explores whether
the labor market condition has been equally important for full-time
and part-time wages. To do this we estimate ARMAX models for
wages separately for full-time and part-time workers. Because the
wage data for full-time and part-time jobs separately are available
only from 1993, we conduct the analysis for the post-1998 period only.
These results are shown in Table 7.3. The first panel compares the
sensitivity of full-time wages to that of part-time wages. For both full-time
and part-time wages, the specification with a second-order autoregression
186 Takeo Hoshi and Anil K Kashyap

Table 7.3 ARMAX models for wage inflation with unemployment rate:
full-time and part-time workers
Sample period is 1998:Q1–2018:Q4.

A. Wage inflation for both types of workers

Full-time workers Part-time workers

Variable Coefficient Standard error T statistic Coefficient Standard error T statistic

Constant 0.0265 0.0116 2.28 0.0595 0.0121 4.92

AR(1) 0.0473 0.1091 0.43 0.3603 0.0871 4.14
AR(2) 0.2430 0.1221 1.99 0.1692 0.1533 1.10
Lagged UNEMP −0.0168 0.0079 −2.12 −0.0344 0.0086 −3.99

R2 0.18 0.58

B. Bonus inflation for both types of workers

Full-time workers Part-time workers

Variable Coefficient Standard error T statistic Coefficient Standard error T statistic

Constant 0.1252 0.0376 3.32 0.1170 0.0633 1.85

AR(1) 0.7734 0.0714 10.84 0.7005 0.0913 7.68
Lagged UNEMP −0.0313 0.0080 −3.90 −0.0354 0.0139 −2.55
R2 0.75 0.62

is augmented by adding the (economy-wide) unemployment rate. For

both full-time and part-time workers, the wage growth is negatively
correlated with lagged unemployment rate, and the coefficient estimates
are statistically different from zero. The magnitude of the unemployment
rate coefficient, however, is different. One-fifth of a percentage point
increase in the unemployment rate would predict a 0.03 percentage
point decrease in full-time worker wages in the next quarter, while the
same increase would lead to twice as large a decline in part-time wages.
Notice also that the R2 is substantially lower for the model for full-time
wages.
Panel B examines the bonus wage inflation. In this case, we use the
specification with a single autoregressive term (as in Table 7.2). Here
the full-time workers’ bonuses show a response to the unemployment
rate that is as large as the part-time workers’ (though we need to note
that the bonuses for part-time workers are much smaller than bonuses
The Great Disconnect 187

for full-time workers). One-fifth of a percentage point increase would

reduce the bonuses for full-time workers by 0.06 percentage points and
that for part-time workers by 0.07 percentage points.
In sum, Table 7.3 suggests that wages for full-time workers do not
move as much as the wages for part-time workers in response to
changes in labor market conditions. For bonuses, the responses for
both types of workers are similar.
Given that the unemployment rate is not predictive of what happens to
full-time wages, and the evidence in Section 7.3, we are lead to the
following account of the data. The market for full-time workers was
often out of balance after 1998. This meant that labor market tightness
did not influence full-time wages. Instead, the marginal equilibrating factor
in the labor market occurs via part-time workers. When the market is
relatively tight for full-time workers, wages respond for part-time workers.
This is not so surprising since many people would prefer to work full-time,
so to keep them in part-time positions, their wages must rise.
On the other hand, the bonuses for both full-time workers and part-
time workers do respond to labor market condition. This is consistent
with the finding in Section 7.3 that the response of bonuses to changes
in the unemployment rate did not decline in the post-1998 period (and
actually increased).

7.5 Wages and Prices

We now turn to the implications of these facts for prices. In brief, the
story we propose is that the break in the labor market helps us under-
stand some, but not all, of the puzzling features for inflation. In par-
ticular, prior to 1998 the labor market operated in a fairly conventional
way: slack helped determine wage dynamics and wage changes were an
important determinant of price changes. After 1998, there was
a persistent excess supply of full-time workers, and the marginal adjust-
ments in labor came through the adjustment of part-time workers. Full-
time wages were not an important factor in determining price changes,
and part-time wage changes were a statistical significant factor in
affecting prices, but their quantitative impact was small. We proceed
by first documenting the important pre-1998 facts before showing the
post-1998 patterns.
Figure 7.6 plots the levels of headline CPI and core-core CPI with the
indices for wages and bonuses for all workers. The structural break
188 Takeo Hoshi and Anil K Kashyap

Headline CPI Core-Core CPI Wages Bonuses

140

130

120

110

100

90

80

70

60

50
1981
1982
1983
1984
1985
1986
1987
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
Figure 7.6 Japanese price and wage level, 1981–2018 (2012=100)
Source: Ministry of Internal Affairs and Communications, Statistics Bureau,
Consumer Price Index, and Ministry of Health, Labour, and Welfare, Monthly
Labour Survey

around 1998 already discussed is apparent in this figure as well. Before

1998, all the series were trending up, but they became stagnant or
started to fall after 1998.
The bonus wage series stands out as a particularly extreme ver-
sion of this pattern, as the level at the end of the sample is about
25 percent below the 1998 level. Recall that we had computed
bonus wages by taking bonus payments over the last four quarters
divided by total hours over that period. That smoothing is import-
ant because of the erratic timing of when bonus payments occur. If
we measure bonus inflation as the simple four-quarter change in
bonuses, the resulting inflation is much, much more volatile.5
We next compute a vector auto regression (VAR) of a three-
variable system that includes the core-core inflation rate, wage
inflation, and bonus inflation for the sample between 1981 and
1997. Given the ambiguity over degree of integration of the vari-
ables we also include a linear time trend in the system. We focus
on the standard diagnostics for describing VARs: Granger causality
The Great Disconnect 189

tests, impulse response functions, and variance decompositions.6

For the latter two diagnostics, we have ordered the series in the
system so that the price inflation variable is first, wage inflation
is second, and bonuses are last. We then use a Cholesky decom-
position to impute the explanatory power of the different series.
The standard deviations of the respective equations, which repre-
sent the size of the shocks for the impulse responses, are 0.0021,
0.0070, and 0.0113 for core-core inflation, wage inflation, and
bonus inflation, respectively.
Panel A of Table 7.4 shows the Granger causality tests for core-core
inflation. Both the wage inflation and bonus inflation rates are signifi-
cant predictors of core-core inflation prior to 1997.

Table 7.4 VAR system with core-core inflation, wage inflation, and
bonus inflation, 1981:Q1–1997:Q4

Panel A. Granger causality from wages inflation to core-core inflation:1981:

Q1–1997:Q4
Included observations: 67

Dependent variable: Core-core inflation

Excluded Chi-sq df Prob.

Wage inflation 9.433 4 0.051
Bonus inflation 20.343 4 0.000
All 18.363 8 0.019

Panel B. Variance decomposition of core-core inflation rate (%): 1981:

Q1–1997:Q4

Quarter Core-core inflation Wage inflation Bonus inflation

1 100.00% 0.00% 0.00%

2 98.37% 0.07% 1.56%
3 87.30% 6.18% 6.52%
4 73.64% 7.40% 18.97%
5 58.83% 16.67% 24.50%
6 50.45% 23.29% 26.26%
7 45.29% 27.35% 27.35%
8 41.47% 31.42% 27.11%
9 39.50% 33.02% 27.47%
10 38.20% 33.52% 28.28%
190 Takeo Hoshi and Anil K Kashyap

The impulse response functions are shown in Figure 7.7 for the
response of the core-core inflation series to the three types of
shocks. The solid blue shows the mean response at each quarter,
and the dashed lines show 95 percent confidence intervals. By

To Core-Core Inflation
0.0030
0.0025
0.0020
0.0015
0.0010
0.0005
0.0000
-0.0005
-0.0010
1 2 3 4 5 6 7 8 9 10
To Wage Inflation
0.0030
0.0025
0.0020
0.0015
0.0010
0.0005
0.0000
-0.0005

1 2 3 4 5 6 7 8 9 10
To Bonus Inflation
0.0030
0.0025
0.0020
0.0015
0.0010
0.0005
0.0000
-0.0005
-0.0010
1 2 3 4 5 6 7 8 9 10

Figure 7.7 Impulse response of core-core inflation to Cholesky One Standard

Deviation Innovations, 1981:Q1–1997:Q4
The Great Disconnect 191

using the Cholesky decomposition, the wage series are presumed

to have no impact on price inflation in the first period. The
impact of the wage shock on core-core inflation grows and
reaches a maximum between five and eight quarters. The pass-
through is meaningful, a one-standard-deviation shock to wages
moves prices by about two-thirds of a standard deviation. For
bonus wage inflation there is also a notable link to core-core price
inflation. The impulse response shows that the maximum effect
occurs at around four quarters and the quantitative effect is about
the same as for wages.
Panel B of Table 7.4 shows the decomposition of the variance of
core-core price inflation. Both wage and bonus inflation explain
important amounts of the variance of prices. For instance, at the eight-
quarter horizon, together they explain over half the variation in price
inflation.
The overall results from Tables 7.4 and Figure 7.7 paint a very
conventional view of inflation determination and the role of the labor
market. When wages or bonuses rise, these increases are passed on to
prices, and quantitatively this mechanism explains a lot of the move-
ments in prices.
The linkage between wages and prices changed dramatically after
1998. Table 7.5 and Figure 7.8 repeat the analysis of the same VAR
system for the period 1998–2018.7 One interesting observation is
that the R2 values for wage inflation equation drop significantly in
the later period; the core-core fit is also a bit lower (though still
high).
More importantly, the Granger causality tests in Panel A of Table 7.5
suggest neither wage nor bonus inflation has any additional explana-
tory power for the core-core inflation process beyond what can be
explained by the lagged core-core inflation rate. Even both series com-
bined have no incremental ability to predict core-core inflation during
this period.
The impulse response functions in Figure 7.8 point to the same conclu-
sion. An increase in wages or bonuses does not have any visible impact on
core-core inflation. The decomposition of the variance of core-core infla-
tion (Panel B of Table 7.5) suggests that almost all the variation in core-
192 Takeo Hoshi and Anil K Kashyap

To Core-Core Inflation
0.0030
0.0025
0.0020
0.0015
0.0010
0.0005
0.0000
-0.0005
-0.0010
1 2 3 4 5 6 7 8 9 10

To Wage Inflation
0.0030
0.0025
0.0020
0.0015
0.0010
0.0005
0.0000
-0.0005
-0.0010
1 2 3 4 5 6 7 8 9 10

To Bonus Inflation
0.0030
0.0025
0.0020
0.0015
0.0010
0.0005
0.0000
-0.0005
-0.0010
1 2 3 4 5 6 7 8 9 10

Figure 7.8 Impulse response of core-core inflation to Cholesky One Standard

Deviation Innovations: 1998:Q1–2018:Q4

core inflation is explained by its own shocks. For example, at the eight-
quarter horizon, wage and bonus inflation explain less than 5 percent of
the core-core inflation variation.
The Great Disconnect 193

Table 7.5 VAR system with core-core inflation, wage inflation, and
bonus inflation: 1998:Q1–2018:Q4

Panel A. Granger causality from wages inflation to core-core inflation: 1998:

Q1–2018:Q4
Included observations: 84

Dependent variable: Core-core inflation

Excluded Chi-sq df Prob.

Wage inflation 3.2335 4 0.52
Bonus inflation 2.9511 4 0.566
All 9.1963 8 0.326

Panel B. Variance decomposition of core-core inflation rate (%): 1998:

Q1–2018:Q4

Quarter Core-core inflation Wage inflation Bonus inflation

1 100.00% 0.00% 0.00%
2 99.18% 0.79% 0.03%
3 98.51% 0.88% 0.61%
4 98.53% 0.90% 0.57%
5 98.22% 0.81% 0.98%
6 97.52% 0.84% 1.64%
7 96.59% 0.95% 2.45%
8 95.69% 1.06% 3.25%
9 95.07% 1.07% 3.86%
10 94.69% 1.07% 4.24%

The contrast between the 1981–1997 period and the 1998–2018

period is clear. Before 1998, both wages and bonuses had significant
and sizable impacts on prices. After 1998, those linkages between
wages and prices were lost. Combined with the results from Section
7.4, the VAR estimation suggests that, around 1998, not only did the
response of wages to labor market conditions get weaker but prices also
became disconnected from the wages.
We also tried another approach to examine how the connection
between price inflation and wage inflation changed over time. We
start by estimating an unobserved components model of the following
194 Takeo Hoshi and Anil K Kashyap

form for each of the four series: core-core inflation, wage inflation for
all workers, and bonus wage inflation for all workers:
yt ¼ μt þ σ y εt

μt ¼ μt1 þ σy vt
where y is the series, μ is the unobserved trend, and ε and ν are identical
and independently distributed normal random variables with mean
zero and variance one.
We then compute the correlations between the estimated μ for the
price series and each of the wage series for rolling windows of twenty-
five quarters. Figure 7.9 shows the evolution of the correlation between
the core-core inflation and wage inflation for all workers. Here each
point shows the correlations calculated for twenty-five quarters cen-
tered on the date. For example, the observation for 1995:Q1 is the
correlation calculated over [1992:Q1, 1999:Q1].
The correlation is positive and quite stable between 1989 and 2000,
which covers rolling windows moving from [1986:Q1, 1993:Q1] to
[1997:Q1, 2004:Q1]. As the rolling windows ceases to include the pre-
1998 period (which first occurs in 2001:Q1), the correlation drops
substantially. The correlation seems to move up again in the late
2000s, but it never reaches the level of the 1990s. The result is again
consistent with the idea that prices and wages disconnected around
1998.

1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0
1984

1987
1989
1991
1993
1995
1996
1998
2000

2004
2006
2007
2009
2011
2013
2015
1985

2002

Figure 7.9 Correlation between unobserved trends for core-core CPI inflation
rate and wage inflation rate for all workers
Source: Authors’ estimation
The Great Disconnect 195

7.6 Conclusion
This chapter looks at why Japan’s inflation has not changed much
despite the extremely aggressive monetary easing and the expansion of
the real economic activity under Abenomics. Our explanation focuses on
the role of the dual labor market that made wages less sensitive to the
labor market conditions. Japan’s labor market has become increasingly
segmented into full-time workers, many of whom are on lifetime
employment and hence rarely fired, and part-time workers, whose
employment can be adjusted more flexibly over business cycles.
Following the recession in the late 1990s, many firms ended up hoarding
an excessive number of full-time workers. Attempts by many firms to
increase the proportion of part-time workers to gain the ability to adjust
their employment more effectively in globally competitive markets,
which preceded the recession, intensified the surplus of regular full-
time workers.8 The surplus of regular workers meant that there was
no pressure on their wages. The wages for part-time workers do depend
on business conditions, but, since the level of their wages is much lower
than full-time workers, the substitution of part-time for full-time
employees has meant that the average wage for the economy has stag-
nated even as economic conditions have improved.
While these broad trends are known, especially to labor econo-
mists, their role in understanding inflation has been less studied. We
provide some evidence on aspects of the story that bear on inflation.
We find that the link between labor market conditions and regular
and overtime wages that existed before 1998 substantially weakened
after 1998. Looking at wages for full-time workers and part-time
workers separately in the period after 1998, we find that the response
of the wages for full-time workers to tightness of the labor market is
much weaker than that of the wages for part-time workers. The
increasing proportion of part-time workers, who are paid much less
than full-time workers, during the period also dampened the wage
increases that might have occurred as unemployment fell. Since the
response of the wages for part-time workers after 1998 itself is
weaker than the response of the average wage for all workers before
1998, the excess of full-time workers (and reduced sensitivity of their
wages to the labor market condition as a result) is not likely to
explain all of the weakening of wage responses to labor market
conditions.
196 Takeo Hoshi and Anil K Kashyap

For bonuses, we do not find any weakening in their link to labor

market conditions. The average bonuses for all workers respond more
strongly to the unemployment rate after 1998. Moreover, the bonuses
for both full-time workers and part-time workers (although they
receive much less of this kind of compensation than full-time workers)
respond to the labor market conditions.
Thus, one reason why Japanese inflation does not seem to have
responded to economic recovery is the decoupling of average wages
and labor market conditions. Looking at the relationship between
wages and prices, however, shows that this is not the whole story. We
find that the link between wages (including bonuses) and prices
changed also around 1998. Before 1998, wage inflation and bonus
inflation were important determinants of price inflation. The link,
however, seems to have disappeared after 1998. Further understanding
the causes of the disconnect between wages, bonuses, and prices after
1998 is an important topic for future research.

Notes
1. For instance, one widely studied question is the connection between
monetary policy and the economy. Many studies have found a large shift
at some point during the 1990s, although they differ in the variables they
examine and the exact timing of the break. See, for example, Arai and
Hoshi (2006), Kimura, Kobayahi, Muranaga, and Ugai (2003), Kuttner
and Posen (2001), and Miyao (2000).
2. See, for example, Ito and Mishkin (2006) for a detailed assessment of the
first few years of the BOJ independence and IMF (2012) and Kuttner,
Iwaisako, and Posen (2015) for subsequent critiques.
3. In earlier versions of this paper we separated overtime and regular wages,
but the differences were mostly small, so we have combined them here.
4. See www.mhlw.go.jp/toukei/list/30–1.html, English page www
.mhlw.go.jp/english/database/db-l/monthly-labour.html. The working
paper version of this chapter (Hoshi & Kashyap 2020) includes an
appendix that describes more details about how we constructed the
wage data. The appendix also goes over the errors in calculating some
series in the Monthly Labour Survey that were revealed in January 2019
and how we have corrected the series.
5. For those interested, the working paper version of this chapter (Hoshi &
Kashyap 2020) has a graph documenting this claim.
The Great Disconnect 197

6. The full VAR coefficients are shown in the working paper version of this
chapter (Hoshi & Kashyap 2020).
7. Again, the VAR coefficient estimates are reported in the working paper
version of the paper.
8. For example, see Abe and Hoshi (2007) for some evidence on increased
adjustment of employment and reduced emphasis on lifetime
employment.

References
Abe, Masahiro, and Takeo Hoshi. 2007. “Corporate Finance and Human
Resource Management in Japan.” In Corporate Governance in Japan:
Institutional Change and Organizational Diversity, edited by
Masahiko Aoki, Gregory Jackson, and Hideaki Miyajima,
pp.257–281. Oxford: Oxford University Press.
Arai, Yoichi, and Takeo Hoshi. 2006. “Monetary Policy in the Great
Stagnation.” In Japan’s Great Stagnation: Financial and Monetary Policy
Lessons for Advanced Economies, edited by Michael Hutchison and
Frank Westermann, pp.157–181. Cambridge, MA: MIT Press.
Bank of Japan. 2016. “Comprehensive Assessment: Developments in Economic
Activity and Prices as well as Policy Effects since the Introduction of
Quantitative and Qualitative Monetary Easing (QQE).” Tokyo: Bank of
Japan. www.boj.or.jp/en/announcements/release_2016/k160921b.pdf.
Bank of Japan, 2017, “Outlook for Economic Activity and Prices
(July 2017).” Tokyo: Bank of Japan.
Bernanke, Ben S. 2003. “Some Thoughts on Monetary Policy in Japan.”
Remarks before the Japan Society of Monetary Economics, Tokyo,
Japan, May 31, 2003. Available at: www.federalreserve.gov/board
docs/speeches/2003/20030531/.
Bernanke, Ben S. 2000. “Japanese Monetary Policy: A Case of Self-Induced
Paralysis?” In Japan’s Financial Crisis and Its Parallels to U.S.
Experience, edited by Ryoichi Mikitani and Adam Posen, pp.149–166.
Washington, DC: Institute for International Economics.
Bernanke, Ben S. 2017. “Some Reflections on Japanese Monetary Policy.”
Ben Bernanke’s Blog, Brookings Institution, May 23, 2017.
Cecchetti, Stephen G., Michael Feroli, Peter Hooper, Anil K Kashyap, and
Kermit Schoenholtz. 2017. “Deflating Inflation Expectations: The
Implications of Inflation’s Simple Dynamics.” Working paper.
University of Chicago.
Gagnon, Joseph E. 2017. “There Is No Inflation Puzzle – Japan Edition.”
Realtime Economics Watch, Peterson Institute for International
198 Takeo Hoshi and Anil K Kashyap

Economics. https://piie.com/blogs/realtime-economic-issues-watch/there-
no-inflation-puzzle-japan-edition.
Genda, Yuji. 2017. Hitode-busoku na no ni naze chingin ga agaranai no ka
(Why are wages not rising despite the shortage of labor?). Tokyo: Keio
University Press.
Hamada, Koichi, Anil K Kashyap, and David E. Weinstein. 2011.
“Introduction.” In Japan’s Bubble, Deflation and Long-Term
Stagnation, edited by Koichi Hamada, Anil K Kashyap, and David
E. Weinstein, pp.1–16. Cambridge, MA: MIT Press.
Hayami, Masaru. 2002. “Revitalizing the Japanese Economy.” Speech at the
Keizai (Economic) Club, January 29, 2002. www.boj.or.jp/en/announce
ments/press/koen_2002/ko0201a.htm/.
Hoshi, Takeo, and Anil K Kashyap. 2020. “The Great Disconnect: The
Decoupling of Wage and Price Inflation in Japan.” NBER Working
Paper 27332.
International Monetary Fund. 2012. “Japan: 2012 Article IV Consultation.”
IMF Country Report No. 12/208. Washington, DC.
Ito, Takatoshi, and Frederic S. Mishkin. 2006. “Two Decades of Japanese
Monetary Policy and the Deflation Problem.” In Monetary Policy with
Very Low Inflation in the Pacific Rim, edited by Takatoshi Ito and
Andrew Rose, pp.131–193. Chicago, IL: NBER-University of Chicago
Press.
Kimura, Takeshi, Hiroshi Kobayashi, Jun Muranaga, and Hiroshi Ugai.
2003. “The Effect of the Increase in the Monetary Base on Japan’s
Economy at Zero Interest Rates: An Empirical Analysis.” In BIS
Papers No. 19 (Monetary Policy in a Changing Environment),
pp.276–312.
Kuroda, Haruhiko. 2017. “The Role of Expectations in Monetary Policy:
Evolution of Theories and the Bank of Japan’s Experience.” Speech at
the University of Oxford, June 8, 2017. www.boj.or.jp/en/announce
ments/press/koen_2017/ko170609a.htm/.
Kuttner, Kenneth, and Adam Posen, 2001, “The Great Recession: Lessons
for Macroeconomic Policy from Japan.” Brookings Papers on
Economic Activity 2: 93–185.
Kuttner, Kenneth, Tokuo Iwaisako, and Adam Posen. 2015. “Monetary and
Fiscal Policies During the Lost Decades.” In Examining Lost Decades,
edited by Yoichi Funabashi and Barak Kushner, pp.17–36. London:
Routledge.
Miyao, Ryuzo. 2000. “The Role of Monetary Policy in Japan: A Break in the
1990s?” Journal of the Japanese and International Economies 14:
366–384.
The Great Disconnect 199

Muto, Ichiro, and Kohei Shintani. 2020. “An Empirical Study on the New
Keynesian Wage Phillips Curve: Japan and the US.” B.E. Journal of
Macroeconomics 20(1).
Shirakawa, Masaaki. 2011. “Toward a Revitalization of Japan’s Economy.”
Speech at the Foreign Correspondents’ Club of Japan, February 7, 2011.
www.boj.or.jp/en/announcements/press/koen_2011/ko110207a.htm/.
Svensson, Lars E. O. 2003. “Escaping from a Liquidity Trap and Deflation:
The Foolproof Way and Others.” Journal of Economic Perspectives 17
(4): 145–166.
8 Abenomics, the Exchange Rate, and
Markup Dynamics in Japanese
Industries
k y o j i fu k a o a n d s h u i c h i r o ni s h i o k a

8.1 Introduction
Following the global financial crisis in late 2007, Japan’s economy –
even though it did not suffer from a collapse in house prices or toxic
assets – was hit harder by the crisis than the United States or European
countries, as several analysts have noted (e.g., Buiter 2009).1
Japan’s contraction of real GDP in the fourth quarter of 2008 was
nearly twice as sharp as that of the United States (a 12.1 percent decline
for Japan versus a 6.3 percent decline for the United States), and, as
Figure 8.1 shows, the decline in the GDP gap from 2007 to 2009 was
the sharpest among major advanced countries.
Japan’s contraction over the crisis was almost entirely due to the
steep decline in external demand, with the drop in net exports account-
ing for 11.8 percentage points of the 12.1 percent real GDP decline in
the fourth quarter of 2008. In contrast, the decrease in net exports was
responsible only for 0.15 percentage points of the 6.3 percent US GDP
decline, meaning that the fall in the US GDP was almost entirely due to
the decline in domestic demand.
Why did Japan experience such a steep decline in external demand?
There are several potential factors. The first factor was a contraction in
so-called triangular trade (e.g., Fukao & Yuan 2009). Until the crisis,
there was vigorous triangular trade, in which Asian advanced econ-
omies (such as Japan, South Korea, and Taiwan) exported to Asian
developing countries (such as China, Thailand, and Vietnam) key
components that were then assembled and exported as final goods to
American and European markets in return for US Treasury securities.

* This research was supported by the Joint Usage and Research Center, Institute of
Economic Research, Hitotsubashi University. We thank Katheryn Russ, Yuhei
Miyauchi, and seminar participants at various universities.

200
Markup Dynamics in Japanese Industries 201

Figure 8.1 GDP gap, 1990–2015

Note: (1) We use output gap in percent of potential GDP from International
Monetary Fund, World Economic Outlook Database, April 2019. (2) The
shaded area is the period of the financial crisis.

Asian developing countries specialized in relatively low-value-added

assembly and manufacturing processes, and Asian advanced economies
specialized in high-value-added production of key components.
However, the sudden and sharp decline in US external demand led to
a rapid contraction in this triangular trade. American imports fell
during the last quarter of 2008 at an annual rate of 19.6 percent,
and, as developing countries’ exports to the United States decreased,
so did their imports from Asian advanced economies. As a result,
Japan, South Korea, and Taiwan all experienced a significant decline
in exports.
The second factor was financial issues related to international trade.
The health of Japanese banks is also an important factor for Japanese
exporters. Using data that cover financial crises in Japan from 1990 to
2010, Amiti and Weinstein (2011) show an empirical link between the
health of banks (that provide trade finance) and the change in the ratios
of exports to domestic sales. A large share of Japan’s exports consists of
capital and durable consumer goods such as cars, electric machinery,
202 Kyoji Fukao and Shuichiro Nishioka

Figure 8.2 Japan’s exchange rate (monthly)

Note: (1) Market rates are derived from various sources including the Bank of
Japan. (2) Real exchange rates are calculated from Corporate Price Index (CPI)
or GDP deflators. (3) Real effective exchange rates are based on GDP deflators.

machine tools, and their components. As such, Japanese exports are

sensitive to financial shocks due to the higher default risk and higher
capital requirements associated with global commerce.
The third factor was the appreciation of the Japanese yen over the
crisis period. As shown in Figure 8.2, the average real effective exchange
rate of the yen from 2008 to 2012 was 18 percent higher than that in
2007. Like Japan, South Korea was deeply involved in the triangular
trade; however, the decline in South Korea’s GDP gap was much smaller
than in Japan’s GDP gap (see Figure 8.1), partly because the Korean won
substantially depreciated (rather than appreciated) during the same
period (Shin 2013). China fixed its exchange rate in 1995 to the US
dollar and maintained that fixed rate until China liberalized its currency
in 2005. However, China has maintained strict rules for individuals and
banks to hold foreign currency, and the Chinese yuan did not appreciate
substantially against the US dollar during the crisis period.
Markup Dynamics in Japanese Industries 203

The main cause of the yen appreciation was differences in monetary

policy across countries. Following the crisis, most countries adopted an
easy-money policy. As a result, both short-term and long-term interest
rates declined substantially in the wake of the crisis. In contrast,
because of long-term economic stagnation, interest rates in Japan
were already low before the global financial crisis, and, because of the
zero lower bound on nominal interest rates, the Bank of Japan did not
have much leeway for further monetary easing.
Since the Liberal Democratic Party of Japan (LDP) won the forty-
sixth general election of the House of Representatives on December 16,
2012, Prime Minister Shinzo Abe carried out his comprehensive eco-
nomic policy – Abenomics – to try to boost the economic growth of
Japan. Abenomics consists of three main strategies – the so-called three
arrows – of aggressive monetary easing, bold fiscal stimulus, and
structural reforms of the Japanese economy. Two of the three arrows
were implemented as soon as Abe replaced the previous cabinet led by
Yoshihiko Noda, a leader of the Democratic Party of Japan. First, on
January 10, 2013, Prime Minister Abe unveiled a $117 billion fiscal
stimulus bill to spend more on public works and disaster recovery and
to provide support for small businesses. Second, on March 20, 2013,
Prime Minister Abe appointed Haruhiko Kuroda, a well-known sup-
porter of aggressive monetary easing, to lead the Bank of Japan to boost
the inflation rate.
By the time Prime Minister Abe returned to the office in late 2012,
most advanced economies had begun to recover from the financial
crisis, and their interest rates were gradually increasing. This change
in the global economic landscape enabled the Abe cabinet and the Bank
of Japan to use drastic monetary easing – the first arrow of Abenomics.
As a result, the Japanese yen began to depreciate from early 2013. As
shown in Figure 8.2, the average real effective exchange rate from 2013
to 2017 was 22 percent lower than that in the period 2008–2012. In
2014, for the first time since the Plaza Accord in 1985, the yen/dollar
market exchange rate became cheaper than the purchasing-power-
parity–adjusted yen/dollar exchange rate (see Figure 8.3).
This chapter examines the impact of the first arrow of Abenomics –
aggressive monetary policy – on markup dynamics in Japan. The
evolution of markups has important implications for macroeconomic
dynamics including profitability of firms, the prevalence of zombies
(i.e., Caballero, Hoshi & Kashyap 2008),2 and income allocation
204 Kyoji Fukao and Shuichiro Nishioka

Figure 8.3 Japan’s market and PPP rates

Note: (1) Market and PPP exchange rates are from the OECD. (2) The shaded
area is the period of Abenomics (2013–2015).

between capitalists and workers (i.e., Autor et al. 2020). While the
decline in markups increases the risk of bankruptcy and the prevalence
of zombies, the increase in markups tends to create income inequality
within a country. Figure 8.4 illustrates the trends of aggregate markups
in Japanese, American, French, and German manufacturing sectors.3
Markup is roughly one when an industry is perfectly competitive and
producers operate at the break-even point; and markup is higher than
one as market competition weakens. As shown in Figure 8.4, the US
manufacturing markups have increased over the last two decades – see
Basu (2019) for the review of the US estimates – and the aggregate
markup did not fall over the crisis period. German and Japanese
markups declined substantially over the crisis period and bounced
back strongly after the crisis.4 The French manufacturing markups
were relatively stable over the period.
As we will show, the underlying forces behind the fall and rise of
markups differ substantially over the crisis versus the Abenomics
period. The fall and rise in markups over the financial crisis period
(2006–2012) depend largely on productivities. Because of the sudden
Markup Dynamics in Japanese Industries 205

Figure 8.4 Markups in manufacturing

Note: (1) The trends for the US, French, and German manufacturing sector are
computed from the KLEMS database (SNA 2018). (2) The trend for the
Japanese manufacturing sector is computed from the JIP 2018 database (SNA
2008).

decline in external demand, Japanese producers were unable to adjust

inputs in a short period of time. As a result, productivities fell, and
markups declined over the period of 2006–2009. The recovery period
(2009–2012) was a mirror image of the crisis period (2006–2009):
productivities improved, and markups increased. Over the
Abenomics period (2012–2015), however, the rise in output prices
was responsible for the rise in markups, and markups increased more
in the manufacturing sector than in the service sector.
Consistent with the aim of the first arrow of Abenomics, aggressive
monetary policy created inflation over the 2012–2015 period. However,
its impact on output prices was heterogeneous across industries: output
prices increased more in the manufacturing sector than in the service
sector. The key factor behind our results, we believe, is how Japanese
firms compete in global markets. Although some large Japanese export-
ers invoice their products in the Japanese yen, most Japanese exporters
use US dollars for global commerce (see Ito et al. 2016). As a result,
during the period of strong yen, most Japanese exporters are forced
206 Kyoji Fukao and Shuichiro Nishioka

to reduce markups because competition in foreign markets prevents

Japanese exporters from increasing their local prices (e.g., Atkeson &
Burstein 2008; Bergin & Feenstra 2009). Our findings are consistent
with the two fundamental changes in global commerce over the 1990s
and 2000s: a decline in exchange pass-through over the past several
decades in the United States (e.g., Gust, Leduc & Vigfusson 2010), and
the surge of China – which controls foreign exchange rates – in global
commerce (i.e., Bergin & Feenstra 2009). These findings in the United
States suggest that, during the period of strong yen, most Japanese firms
face competitive pressures from foreign firms. For example, consider that
a Japanese manufacturer competes with a Chinese counterpart in the
Japanese market. When the US dollar depreciates against the yen,
because the Chinese yuan is nearly fixed, a Chinese manufacturer is
able to reduce its price in Japan, and the Japanese manufacturer needs
to cut its price to compete. Indeed, with the productivity growth of East
Asian competitors and the resurgence of American information technol-
ogy (IT) industries, Japanese manufacturers faced global competition,
were unable to change their local prices in foreign markets, were forced
to cut their prices in Japan, and had to reduce their profits in terms of the
yen during the period of the yen appreciation.
In the next section, we discuss the theoretical consideration of meas-
uring markups. Section 8.3 describes the data and Section 8.4 illus-
trates markup dynamics across industries in Japan. Section 8.5
examines how import competitions influence output prices. Section
8.6 concludes this chapter.

8.2 Measuring Markups at the Industry Level

There are the two broad strands of theoretical approaches to estimate
product markups: one relies on demand-side information on product
prices and quantities and computing markups from price elasticities of
demand (e.g., Atkeson & Burstein 2008; Feenstra & Weinstein 2017);
and the other relies on supply-side information on outputs and inputs
and computing markups from the output elasticities of an input divided
by revenue shares of that input (e.g., Hall 1988; Diewert & Fox 2008;
De Loecker & Warzynski 2012; De Loecker et al. 2016). Our approach
is a variant of the supply-side approach. In particular, we use the two
measures of markups without estimating production functions: first,
we follow Diewert and Fox (2008) and, under the two assumptions of
Markup Dynamics in Japanese Industries 207

constant returns to scale and no adjustment costs in inputs, we compute

markups from the revenue divided by total costs; and second, we follow
Roeger (1995) and compute markups from the output price divided by
imputed marginal cost. Basu (2019) explains pros and cons about the
alternative measures of markups. Although we impose the two restrict-
ive assumptions, we believe our method is appropriate for the industry-
level analysis.
We assume that industry i at time t uses a production function that
converts labor (Lit), capital (Kit), and materials (Mit) into real output
(Yit). We use the following constant returns to scale time-specific Cobb-
Douglas revenue production function:
β γ
Yit ¼ Ωit Lαitit Kitit Mitit ð1Þ

where αit+βit+γit =1, and Ωit is the industry- and time-specific total
factor productivity (TFP).
In Equation 1, the output elasticities, αit, βit, and γit, are industry- and
time-specific. Thus, we can allow structural transformation in produc-
tion techniques over the period. We first use a profit maximization
problem and derive markups from the revenue divided by total costs.
Our approach is a constant returns to scale version of Diewert and Fox
(2008). Here, we assume that industries can hire their composites of
labor, capital, and intermediate inputs at wit, rit, and pit where input
prices are industry-specific and exogenous for producers. In particular,
by observing output price of industry i at year t (Pit), we solve the profit
maximization problem.
By imposing all the first-order conditions from this optimization
problem, we can derive the first measure of markups (Markup 1 in
Tables 8.2, 8.3, and 8.4) from the revenues divided by total costs:
Pit Yit
μ1it ≡ ð2Þ
wit Lit þ rit Kit þ pit Mit
The second approach, which will be used for a decomposition analysis
of markup dynamics, is similar to Roeger (1995) who derives markups
from a dual problem of a profit maximization problem. We use the cost
minimization problem (to produce one unit of output). By imposing all
the first-order conditions from this optimization problem, we can
derive marginal cost as a function of input prices and a productivity:5
208 Kyoji Fukao and Shuichiro Nishioka

1 β γ
cit ¼ ½wαit r it p it  ð3Þ
Ωit it it it
And, our second measure of markup (Markup 2 in Tables 8.2, 8.3, and
8.4) is obtained from the observed output price divided by imputed
marginal cost from Equation 3:
Pit
μ2it ≡ ð4Þ
cit
In Equation 4, αit, βit, and γit are output elasticities of labor, capital, and
real spending on materials, respectively. Moreover, the first-order con-
ditions imply that output elasticities are cost shares of inputs. As
Syverson (2011) notes, if cost shares can be measured and the scale
elasticity is either estimated or assumed, then the output elasticities can
be directly derived. Therefore, by using cost shares as proxies for
output elasticities, we can derive the two measures of markups without
confronting various data and econometric problems (see, for example,
Ackerberg et al. 2015; De Loecker et al. 2016). Our strategy is useful
for us because we are unable to estimate production function param-
eters from firm-level data. Note that the two measures of markups from
Equations 2 and 4 give us almost identical dynamics over the period we
consider. However, we are unable to obtain the levels of markups from
Equation 4 because all the price and quantity variables are normalized
to a base year. Finally, because we compute industry-level markups
aggregated across producers, we cannot distinguish the changes in
industry-level markups arising from the changes in market shares
(i.e., Autor et al. 2020) from the changes within each firm in the
industry.
Markup derived from Equation 4 reflects the observed changes in
output prices, input prices, and productivities. It is straightforward to
decompose the markup growth into the following five factors:

Δlnðμ2it Þ ¼ ΔlnðPit Þ þ ΔlnðΩit Þ  Δait lnðwit Þ  Δβit lnðrit Þ  Δγit lnðpit Þ ð5Þ
where Δln(xit)=ln(xit)−ln(xi,t−1) for xit, and ln(Ωit)=ln(Yit)−[αitln(Lit)+βit
ln(Kit)+γitln(Mit)].
In Equation 5, Δln(Pit) is the contribution of the change in output
price, Δln(Ωit) captures the contribution of productivities including
scale effects,6 and the last three terms are the contributions of the
Markup Dynamics in Japanese Industries 209

changes in wage rates, capital services prices, and intermediate prices,

respectively.

8.3 Data

8.3.1 JIP Database

This chapter uses the Japan Industrial Productivity (JIP) Database
2018, which covers the period of 1994–2015, as well as the previous
version of the same database (JIP Database 2015), which covers the
period of 1970–2012.7 We mainly use the former data because it covers
the financial crisis period (2006–2012) as well as the Abenomics period
(2012–2015). The two versions of the JIP Database, however, differ in
how to develop the data. First, the JIP Database 2018 uses the
Laspeyres real input index, whereas the JIP Database 2015 uses the
Törnqvist real input index. Second, the JIP Database 2018 is based on
the 2008 SNA, whereas the JIP Database 2015 is based on the 1993
SNA. Following the 2008 revision of the SNA, in the JIP Database
2018, R&D expenditure is counted as output, and R&D investment is
capitalized and included in the capital stock. As a result, these two
measures give us slightly different levels and changes of markups.
The JIP databases comprise various types of annual macroeconomics
variables necessary for estimating markups. It is important to note that
the database reports the variables including capital services costs8 and
quality- and working hour-adjusted labor service input indices.9
Because we observe all the cost and revenue information, we do not
need to estimate production functions to obtain output elasticities. The
JIP Database 2018 reports the data for 100 industries; however, we
exclude the nine nonmarket industries10 and seventeen resource-
intensive or highly regulated industries.11 Therefore, we focus on sev-
enty-five industries in the JIP Database 2018: fifty-one manufacturing
and twenty-four service industries. As such, our aggregate variables of
the manufacturing and service sectors do not include these nonmarket
and resource-intensive industries.

8.3.2 Capital Services

To use Equations 2 and 4 to compute markups, we need to have the
price of capital services (rit). Without this variable, we cannot obtain
210 Kyoji Fukao and Shuichiro Nishioka

total costs, cost shares, output elasticities, and markups from

Equations 2 and 4. There are two approaches to derive capital services
prices (Timmer et al. 2007): the first is the ex ante approach proposed
by Hall and Jorgenson (1967), which estimates capital services prices
from the opportunity costs of holding capital assets; and the second is
the ex post approach, which is based on the internal rates of returns
treated as residuals, or the value added minus labor compensation
divided by capital stock. While the ex post approach is popular in the
literature, this approach can be justified only if final product markets
are perfectly competitive. Moreover, markups from Equation 2 are
unity by construction. Because we are estimating markups – which
should reflect the degrees of imperfect competition – we should use
capital service price based on the ex ante approach. The JIP Database
follows Hall and Jorgenson (1967) and computes prices of capital
services as follows.12
Consider that capital stock (Kit) consists of several capital services. In
particular, Kitk is the quantity of the kth capital service, and rtk is its
corresponding price. We also introduce the following notation: Iitk is
the quantity of the kth investment good acquired to produce in industry
i in time t, and ptk is its corresponding market price of this investment
good. As is standard in the literature, the cumulated stock of past
investments in the kth capital good has the following property:

Kki;tþ1 ¼ ð1  δk Þ Kkit þ Iitk ð6Þ

where δk is the depreciation rate of the kth investment good.

Capital costs in the JIP databases are developed from Hall and
Jorgenson’s approach (1967), which takes account of debt and equity
financing and business income tax. It uses the following equation to
obtain capital costs:

1  zt
rkt Kkit ¼ pkt Kkit ðit þ δk  Δpkt =pkt Þ ð7Þ
1  τt
where the corporate tax rate is τt, the weighted average cost of capital is
it = btitL+(1−τt)(1−bt)itS where bt is the debt to asset ratio, itS is the risk
free interest rate, and itL is the long-term loan rate, and the present
value of depreciation deductions on investment is zt = δkτt/(it+δk).
Equation 7 captures opportunity cost of holding capital stock. If
a producer does not use capital stock Kitk, the producer would sell
Markup Dynamics in Japanese Industries 211

Kitk at ptk, invest ptkKitk into assets, and gain itptkKitk over a year.
Moreover, if the producer would invest in the risk-free asset, the
producer would avoid deducting the depreciation of capital stock (δk
ptkKitk). Finally, if the producer invests in the risk-free asset, the produ-
cer would lose the potential gain of the appreciation of capital stock as
an asset (ΔptkKitk). Therefore, Equation 7 represents the opportunity
cost of holding capital stock of an asset type k.
The industry-level stock of capital in this database is Kit = ∑kKitk, and
the industry-level price of capital services is rit = ∑krtkKitk/Kit. As noted
before, we mainly use the 2018 version of the JIP database. Thus, R&D
stock and other intangible assets (including, for example, in-house
software) are included in capital stocks and capital costs.

8.4 Markups in Japan

8.4.1 Markup Dynamics over the Lost Decade (1994–2006)

Table 8.1 reports the changes over the period of 1994–2006 for the key
production variables before the global financial crisis. Throughout this
chapter, because all the price and quantity indices are normalized
to year 2011 in the JIP Database 2018, we focus on the growth rates
and examine the changes over the period. As shown in the first column
of Table 8.1, the aggregate value of real manufacturing output did not
increase much over the period of 1994–2006 (a 7 percent increase in
real output). From the second through seventh columns in the same
table, we report the summary statistics across the fifty-one manufac-
turing industries. While the mean value of real output growth over the
period was a 3 percent increase, the median value was a 1 percent
decline: slightly more than half of Japanese manufacturing industries
(i.e., twenty-six of fifty-one industries) experienced a decline in real
output over the period of the lost decade. The maximum value (a
107.6 percent increase) comes from semiconductor devices and inte-
grated circuits (JIP code 40), whereas the minimum value (a 99.6 per-
cent decline) comes from textile products (JIP code 13), which appears
to reflect the evolution of comparative advantages and disadvantages
of Japan over the period. In the second panel of Table 8.1, we report the
summary statistics for the service sector. The mean value of real output
growth across the twenty-four service industries increased by 21.4 per-
cent, and the median value increased by 15.3 percent: the majority of
Table 8.1 Summary statistics for output prices and marginal costs, 1994–2006

1. Manufacturing

Summary statistics across 51 manufacturing industries

Aggregate % of > 0 Mean Median Min Max sd

(1) (2) (3) (4) (5) (6) (7)

Real output 0.070 49.0% 0.030 −0.010 −0.996 1.076 0.410

Employment −0.147 25.5% −0.165 −0.134 −0.933 0.835 0.297
Real capital stock 0.155 62.7% 0.148 0.130 −0.753 1.296 0.414
Real spending on materials 0.030 41.2% −0.050 −0.041 −0.985 0.673 0.369
Cost share of labor −0.021 49.0% −0.008 −0.001 −0.101 0.061 0.037
Cost share of capital −0.015 47.1% −0.016 −0.009 −0.184 0.118 0.057
Cost share of materials 0.036 62.7% 0.023 0.020 −0.128 0.174 0.065
2. Services

Summary statistics across 24 service industries

Aggregate % of > 0 Mean Median Min Max sd

(1) (2) (3) (4) (5) (6) (7)

Real output 0.201 70.8% 0.214 0.153 −0.176 1.077 0.323

Employment 0.100 66.7% 0.060 0.060 −0.346 0.584 0.243
Real capital stock 0.277 87.5% 0.521 0.439 −0.791 2.389 0.662
Real spending on materials 0.257 79.2% 0.253 0.226 −0.258 1.385 0.369
Cost share of labor −0.033 25.0% −0.029 −0.038 −0.135 0.083 0.055
Cost share of capital −0.026 41.7% −0.029 −0.025 −0.241 0.119 0.078
Cost share of materials 0.059 75.0% 0.058 0.045 −0.088 0.256 0.089

Note: We report percentage change from 1994 to 2006 for all the variables but cost shares. Percentage point changes between 1994 and 2006 are
reported for cost shares.
214 Kyoji Fukao and Shuichiro Nishioka

the Japanese service industries (i.e., seventeen of twenty-four indus-

tries) increased their real outputs over the period. Overall, Japan has
deindustrialized over the period.
Similar to the US experience (i.e., Pierce & Schott 2016) in the
2000s, manufacturing employment (hour- and quality-adjusted) in
Japan’s manufacturing sector declined by 14.7 percent. The real
capital stock, on the other hand, increased by 15.5 percent. In the
table, we also report the changes in cost shares of labor, capital,
and intermediate inputs. If capital and labor are gross substitutes
in the production process (e.g., Karabarbounis & Neiman 2014),
cost shares of labor should decline and those of capital should
increase. However, we cannot find clear evidence that cost shares
of labor declined in Japanese manufacturing industries. We also
find that most of the manufacturing industries spent more on
intermediate inputs (a 3.6 percentage point increase at the aggre-
gate level) and less on capital services (a 1.5 percentage point
decline at the aggregate level). As reported in the second panel,
the results are much clearer in the service sector. Similar to the
findings in Boehm, Flaaen, and Pandalai-Nayar (2019), most of the
service industries spent more on intermediate inputs (a 5.9 percent-
age point increase at the aggregate level) and less on labor services
(a 3.3 percentage point decline at the aggregate level) and capital
services (a 2.6 percentage point decline at the aggregate level),
suggesting that, facing domestic as well as foreign outsourcing
opportunities, more expensive factor inputs were replaced by
cheaper intermediate inputs.
Figure 8.5 illustrates the evolution of markups computed from
Equation 2 over the period of 1994–2015 for the manufacturing
and service sectors. The figure shows that markup had declined
over the financial crisis period and bounced back before
Abenomics began, and the dynamics are similar in the manufactur-
ing and service sectors. This finding – the cyclical relationship
between the two sectors – is interesting: the financial crisis
impacted not only the tradable manufacturing sector through
trade linkage but also the nontradable sector probably through
its linkage with intermediate input transactions. Prior to the crisis,
however, the trends over the lost decade period (1994–2006) were
very different: markups had constantly increased for the service
sector, whereas they did not increase in the manufacturing sector.
Markup Dynamics in Japanese Industries 215

Figure 8.5 Markup dynamics in Japan

Note: (1) We use Equation 2 to compute markups. (2) The shaded areas are
the period of the financial crisis (2007–2009) and that of Abenomics
(2013–2015).

Table 8.2 summarizes the decomposition exercise of markups over

the period of 1994–2006. The first row, “Markup 1,” reports markups
from Equation 2, and the second row, “Markup 2,” reports markups
from Equation 4. We use Equation 5 mainly and decompose the
markup growth into the contributions from output price, productivity,
wage rate, price for capital services, and price for intermediate inputs.
In the first column of the table, a reduction in marginal costs (i.e.,
a 4.4 percent increase of a 1.9 percent increase in the aggregate
markup) contributed positively, and a decline in output price (i.e.,
a 2.5 percent decline) contributed negatively to markups in the aggre-
gate manufacturing sector. Among the decline in marginal cost, prod-
uctivity growth (i.e., a 6.0 percent increase) and decline in capital
services prices (i.e., a 3.4 percent increase) contributed to the improve-
ment of markups in this sector.
In the first column in the second panel of the table, we report the
decomposition analysis for the aggregate service sector. Although the
Table 8.2 Summary statistics for markup growth, 1994–2006

1. Manufacturing

Aggregate Summary statistics across 51 manufacturing industries

Sign Relation % of > 0 Mean Median Min Max sd

(1) (2) (3) (4) (5) (6) (7)

Markup 1 + 0.015 58.8% 0.005 0.015 −0.180 0.218 0.088

Markup 2 + a=b+c 0.019 60.8% 0.028 0.022 −0.135 0.438 0.111
Output price + b −0.025 35.3% −0.082 −0.036 −1.158 0.784 0.341
Marginal cost − c=d+e 0.044 64.7% 0.111 0.065 −0.581 1.038 0.324
TFP + d 0.060 70.6% 0.095 0.046 −0.180 0.790 0.178
Input prices − e=f+g+h −0.016 58.8% 0.015 0.034 −0.650 0.536 0.196
Wage rate − f −0.016 29.4% −0.009 −0.009 −0.044 0.038 0.019
Price for capital services − g 0.034 100.0% 0.043 0.037 0.001 0.113 0.027
Price for intermediates − h -0.035 41.2% −0.019 −0.015 −0.725 0.522 0.205
2. Services

Aggregate Summary statistics across 24 service industries

Sign Relation % of > 0 Mean Median Min Max sd

(1) (2) (3) (4) (5) (6) (7)

Markup 1 + 0.079 66.7% 0.069 0.048 −0.153 0.351 0.146

Markup 2 + a=b+c 0.093 70.8% 0.115 0.087 −0.165 0.749 0.199
Output price + b −0.027 54.2% −0.034 0.005 −0.515 0.151 0.154
Marginal cost − c=d+e 0.120 75.0% 0.149 0.094 −0.200 0.714 0.226
TFP + d 0.007 41.7% 0.001 −0.031 −0.367 0.424 0.196
Input prices − e=f+g+h 0.113 91.7% 0.148 0.124 −0.024 0.625 0.136
Wage rate − f 0.038 70.8% 0.031 0.014 −0.066 0.241 0.066
Price for capital services − g 0.065 95.8% 0.108 0.065 −0.005 0.575 0.130
Price for intermediates − h 0.010 58.3% 0.008 0.011 −0.155 0.086 0.047

Note: (1) We report percentage change over the 1994–2006 period. (2) See Equations 2 and 4 for markups and Equation 5 for the decomposition
analysis.
218 Kyoji Fukao and Shuichiro Nishioka

service sector’s productivity growth was very slow (see Fukao 2012)
and its output price declined, markups increased substantially by
replacing labor and capital with intermediate inputs. Overall, despite
competitive pressure on output prices, our findings suggest that
Japanese manufacturing industries sustained markups by improving
productivities, and Japanese service industries improved markups by
changing the composition of inputs toward cheaper ones.
Columns 2-7 in Table 8.2 summarize the changes in markups and
their contributions at the industry level. The table shows that most
Japanese manufacturing industries (i.e., thirty-one of fifty-one
industries) experienced increases in markups. While less than
a half of the industries experienced declining output prices, more
than half of the industries reduced marginal costs. The aggregate
results (as reported in the first column) appear to suggest that
Japanese manufacturing improved productivities: at the industry
level, 70.6 percent of the industries improved productivities over
the relatively long period of the lost decade. The trend is very
different for service industries. Only 41.7 percent of the industries
improved productivities. Moreover, marginal costs declined and
contributed positively to markups in service industries mainly
because prices of capital services declined.
The first panel of Figure 8.6 illustrates the trends in markups for
three selected manufacturing industries: textile products, pharma-
ceutical products, and automobile industries. The second panel of
the figure illustrates the trends for three selected services: whole-
sale, hotels, and financial services. The figure shows that there are
clear winners and losers over the two lost decades in Japan. While
the pharmaceutical products and wholesale industries survived
strongly over the period with high and increasing markups, most
of the declining – or comparative disadvantaged – industries
including the textile industry reduced markups substantially. The
financial crisis negatively impacted financial services, and, consist-
ent with Amiti and Weinstein (2011), it impacted exporting indus-
tries, including the automobile industry. The decline of the textile
industry is not surprising because most Japanese retailers outsource
textile and apparel products from Asian countries such as China,
Indonesia, and Vietnam.
Markup Dynamics in Japanese Industries 219

Figure 8.6 Markups in selected industries

8.4.2 Markups Dynamics over Abenomics (2006–2015)

To present the results over the Abenomics period, we normalize the
variables in year 2012 – the eve of Abenomics – so that we can examine
the evolution of markups before and after the Abenomics period. As
briefly discussed in Section 8.1, our estimates suggest that markups in
Japan’s manufacturing sector recovered after Abenomics. Figure 8.7
illustrates the box plots of markups over the period of Abenomics for
the manufacturing and service sectors. While markups had been con-
stant until 2012 in the manufacturing sector, markups increased across
broad industries after 2013. The evolution in markups precisely sug-
gests how important the dollar/yen exchange rate is for Japanese
manufacturers’ profitability. For the service sector, the rise of markups
began much earlier, but the growth in markups was not as strong as the
manufacturing sector over the Abenomics period.
Panels 1 through 3 in Table 8.3 summarize the decomposition exer-
cises of markups for the manufacturing sector over the three periods of
2006–2009 (the financial crisis period), 2009–2012 (the recovery
period), and 2012–2015 (the Abenomics period), respectively.
Although each period is very short, we can find some stylized facts.
220 Kyoji Fukao and Shuichiro Nishioka

Figure 8.7 Markups before/after Abenomics

Note: The bottom and top values of the boxes are 25th percentile and 75th
percentile values, respectively.

As before, we use Equation 5 and decompose markup growth into the

contributions of the five factors. As reported in the first column of the
table, at the aggregate level in the manufacturing sector, markups
declined by 4 percent over the financial crisis period, increased by
3.6 percent over the recovery period, and improved by 3.9 percent
over the Abenomics period. The fall and rise in markups over the
financial crisis and its recovery periods (2006–2009 and 2009–2012)
depend largely on productivities. Because of the sudden decline in
external demand, Japanese producers were unable to adjust inputs
(labor and capital, in particular) in a short period of time. As a result,
productivities fell, and markups declined over the period of
2006–2009. The recovery period (2009–2012) was a mirror image of
the crisis period (2006–2009): productivities improved, and markups
increased. Over the Abenomics period (2012–2015), the rise in output
prices was responsible for the rise in markups, and markups increased.
Although marginal costs were stable over the period, the decompos-
ition of marginal costs suggests that productivity growth (a 2.4 percent
increase) offset the rise in material prices (a 2.7 percent decline). Our
Table 8.3 Decomposition analysis for markup growth in manufacturing, 2006–2015

1. Financial crisis period (2006–2009)

Aggregate Summary statistics across 51 manufacturing industries

Sign Relation % of > 0 Mean Median Min Max sd

(1) (2) (3) (4) (5) (6) (7)

Markup 1 + -0.036 25.5% -0.035 -0.041 -0.181 0.124 0.063

Markup 2 + a=b+c −0.040 27.5% −0.035 −0.043 −0.188 0.127 0.069
Output price + b −0.007 51.0% −0.014 0.002 −0.326 0.184 0.117
Marginal cost − c=d+e −0.033 31.4% −0.021 −0.036 −0.229 0.364 0.131
TFP + d −0.025 27.5% −0.025 −0.032 −0.194 0.239 0.092
Input prices − e=f+g+h −0.007 45.1% 0.004 −0.010 −0.110 0.200 0.061
Wage rate − f 0.000 49.0% −0.001 0.000 −0.047 0.066 0.016
Price for capital services − g −0.004 23.5% −0.004 −0.004 −0.016 0.008 0.005
Price for intermediates − h −0.003 47.1% 0.009 −0.003 −0.106 0.214 0.058
Table 8.3 (cont.)

2. Recovery period (2009–2012)

Aggregate Summary statistics across 51 manufacturing industries

Sign Relation % of > 0 Mean Median Min Max sd

(1) (2) (3) (4) (5) (6) (7)

Markup 1 + 0.032 72.5% 0.024 0.030 −0.060 0.107 0.044

Markup 2 + a=b+c 0.036 74.5% 0.032 0.035 −0.059 0.120 0.047
Output price + b −0.029 31.4% −0.027 −0.017 −0.279 0.289 0.098
Marginal cost − c=d+e 0.065 76.5% 0.059 0.048 −0.210 0.319 0.104
TFP + d 0.051 78.4% 0.051 0.044 −0.107 0.241 0.070
Input prices − e=f+g+h 0.013 62.7% 0.008 0.018 −0.268 0.081 0.060
Wage rate − f 0.008 72.5% 0.008 0.003 −0.007 0.047 0.011
Price for capital services − g 0.018 100.0% 0.020 0.019 0.003 0.041 0.009
Price for intermediates − h −0.013 37.3% −0.020 −0.008 −0.286 0.054 0.055
3. Abenomics period (2012–2015)

Aggregate Summary statistics across 51 manufacturing industries

Sign Relation % of > 0 Mean Median Min Max sd

(1) (2) (3) (4) (5) (6) (7)

Markup 1 + 0.040 92.2% 0.046 0.048 −0.029 0.204 0.039

Markup 2 + a=b+c 0.039 92.2% 0.044 0.046 −0.030 0.184 0.037
Output price + b 0.044 88.2% 0.048 0.041 −0.083 0.179 0.056
Marginal cost − c=d+e −0.005 47.1% −0.005 −0.012 −0.192 0.154 0.072
TFP + d 0.024 72.5% 0.025 0.028 −0.134 0.169 0.058
Input prices − e=f+g+h −0.029 15.7% −0.029 −0.028 −0.155 0.110 0.039
Wage rate − f −0.004 23.5% −0.003 −0.001 −0.026 0.007 0.006
Price for capital services − g 0.003 100.0% 0.003 0.003 0.000 0.018 0.003
Price for intermediates − h −0.027 11.8% −0.030 −0.028 −0.156 0.107 0.038

Note: See Table 8.2.

224 Kyoji Fukao and Shuichiro Nishioka

results from the manufacturing sector are consistent with the view that
exchange rate dynamics are crucial for the markup dynamics in Japan.
The depreciation of the yen during Abenomics improved profitability
of Japanese manufacturing industries because the competitive pres-
sures on output prices weakened.
Columns 2–7 in Table 8.3 also summarize the changes in product
markups and their contributions at the industry level. The results in the
table suggest that many Japanese manufacturing industries (i.e., forty-
seven of fifty-one industries) increased markups over Abenomics.
While nearly 90 percent of the industries (i.e., forty-five industries)
increased output prices, slightly less than half of the industries (i.e.,
twenty-four industries) increased marginal costs. Similar to the aggre-
gate results in column 1, the increase in marginal costs depends mainly
on the rise of input costs due to the depreciation of the yen.
Table 8.4 reports the decomposition analysis for the service sector.
Although the aggregate trends over the three periods were very similar
to those in the manufacturing sector, the factors behind these trends are
very different. Over the financial crisis period, markups declined by
2.5 percent mainly due to a decline in output prices (a 1.9 percent
decline). Although output prices continued to decline over the recovery
period (a 2.3 percent decline), markups increased due to the declines in
payments for wages and prices for capital services. This occurred
because capital services prices declined substantially after the financial
crisis. Over the Abenomics period, markups did not increase as much as
those in the manufacturing sector because increases in costs of inter-
mediate inputs due the depreciation of the yen exceeded its slow
productivity growth. Industry-level data reported in columns 2–7 are
consistent with the aggregate results. Markup dynamics in the Japanese
service sector are susceptible to price changes because the productivity
growth was much smaller than that in the manufacturing sector.

8.5 Global Competition, the Exchange Rate, and Markups

The JIP Database 2018 does not include trade variables. Therefore, in
this section, we use the JIP Database 2015 (Fukao et al. 2004), which
includes various trade variables. The JIP Database 2015 reports the
data, based on the 1993 SNA, for 108 industries; however, we exclude
the sixteen non-market industries,13 fourteen energy-intensive or
highly regulated industries, and four agriculture industries.14
Table 8.4 Decomposition analysis for markup growth in services, 2006–2015

1. Financial crisis period (2006–2009)

Aggregate Summary statistics across 24 service industries

Sign Relation % of > 0 Mean Median Min Max sd

(1) (2) (3) (4) (5) (6) (7)

Markup 1 + −0.024 16.7% −0.045 −0.054 −0.194 0.079 0.066

Markup 2 + a=b+c −0.025 16.7% −0.042 −0.045 −0.176 0.090 0.068
Output price + b −0.019 37.5% −0.023 −0.013 −0.160 0.044 0.045
Marginal cost − c=d+e −0.006 29.2% −0.018 −0.034 −0.210 0.144 0.083
TFP + d −0.016 20.8% −0.028 −0.026 −0.248 0.129 0.079
Input prices − e=f+g+h 0.010 58.3% 0.010 0.020 −0.088 0.138 0.049
Wage rate − f 0.004 54.2% 0.002 0.003 −0.066 0.054 0.029
Price for capital services − g −0.001 50.0% −0.002 0.001 −0.080 0.140 0.038
Price for intermediates − h 0.007 83.3% 0.009 0.003 −0.008 0.065 0.016
Table 8.4 (cont.)

2. Recovery period (2009–2012)

Aggregate Summary statistics across 24 service industries

Sign Relation % of > 0 Mean Median Min Max sd

(1) (2) (3) (4) (5) (6) (7)

Markup 1 + 0.039 87.5% 0.049 0.043 −0.095 0.181 0.064

Markup 2 + a=b+c 0.042 87.5% 0.054 0.052 −0.093 0.192 0.066
Output price + b −0.023 29.2% −0.015 −0.006 −0.139 0.058 0.043
Marginal cost − c=d+e 0.065 91.7% 0.068 0.055 −0.086 0.331 0.079
TFP + d 0.017 58.3% 0.007 0.015 −0.141 0.194 0.062
Input prices − e=f+g+h 0.048 87.5% 0.061 0.054 −0.008 0.156 0.044
Wage rate − f 0.015 87.5% 0.023 0.020 −0.022 0.098 0.025
Price for capital services − g 0.032 100.0% 0.038 0.023 0.006 0.125 0.035
Price for intermediates − h 0.001 54.2% 0.000 0.002 −0.058 0.065 0.021
3. Abenomics period (2012–2015)

Aggregate Summary statistics across 24 service industries

Sign Relation % of > 0 Mean Median Min Max sd

(1) (2) (3) (4) (5) (6) (7)

Markup 1 + 0.015 70.8% 0.033 0.030 −0.052 0.137 0.052

Markup 2 + a=b+c 0.015 66.7% 0.030 0.030 −0.054 0.134 0.052
Output price + b 0.029 87.5% 0.037 0.037 −0.080 0.148 0.045
Marginal cost − c=d+e −0.014 33.3% −0.007 −0.014 −0.097 0.107 0.059
TFP + d 0.011 62.5% 0.022 0.006 −0.067 0.160 0.052
Input prices − e=f+g+h −0.025 12.5% −0.029 −0.033 −0.082 0.030 0.029
Wage rate − f −0.006 33.3% −0.010 −0.006 −0.043 0.017 0.018
Price for capital services − g 0.003 66.7% 0.006 0.002 −0.004 0.049 0.013
Price for intermediates − h −0.022 4.2% −0.026 −0.020 −0.058 0.001 0.015

Note: See Table 8.2.

228 Kyoji Fukao and Shuichiro Nishioka

Figure 8.8 Log prices in manufacturing

Therefore, we focus on seventy-four industries: forty-nine manufactur-

ing and twenty-five service industries from years 1990–2012.
Figure 8.8 illustrates the trends in various output, export, and import
price indices in the manufacturing sector over the period before
Abenomics. Output prices in the manufacturing sector declined by
20.4 percent over the 1990–2012 period. At the industry level, export
prices declined across broad industries: forty-five of forty-nine indus-
tries in the manufacturing sector that faced pro-competitive pressure
and reduced their export prices. In the service sector, however, the
declines in output prices were not as steep as those in manufacturing
industries.
As suggested by the results of the survey on exchange rate risk
management, Ito et al. (2016) find that Japanese exporters use invoi-
cing in terms of the US dollar, which creates exchange rate risk when
the yen strongly appreciates. In the broad literature of exchange rate
pass-through, the literature has found evidence indicating a decline in
exchange rate pass-through to import prices, particularly in the United
States. While Goldberg and Knetter’s (1997) survey suggests that the
pass-through was around 0.5 in the 1980s, Gust, Leduc, and Vigfusson
(2010) show that the pass-through rate had gradually declined over the
Markup Dynamics in Japanese Industries 229

1990s and 2000s. As in Bergin and Feenstra (2009), the emergence of

Chinese exporters in global commerce could explain a decline in the
exchange rate pass-through particularly for Japanese exporters. Since
China pegged or controls its exchange rate relative to the US dollar, the
depreciation of the US dollar during the great financial crisis did not
change the prices of Chinese goods in the US markets. If Japanese
exporters competed in prices not only with American firms but also
with Chinese exporters, Japanese exporters were unable to increase
their local prices in the United States. As a result, Japanese exporters’
markups in terms of the yen could decline. Similarly, these findings in
the United States suggest that, during the period of strong yen, most of
Japanese firms face competitive pressures from foreign firms in
Japanese markets.
While we estimate markups from producer-side information, mark-
ups derived from demand-side information are also well studied – see,
for example, Atkeson and Burstein (2008); Edmond, Midrigan, and Xu
(2015); De Blas and Russ (2015); and Feenstra and Weinstein (2017).
Most of these studies predict that, conditional on marginal costs,
a producer will raise its markup as its market share increases.
Atkeson and Burstein (2008) show that this property holds in
a model with monopolistic competition and product differentiation,
whereas Feenstra and Weinstein (2017) show this property holds in
a model with monopolistic competition, differentiated products, and
general translog preferences. To better understand the impact of global
competition on two types of markups (Markup 1 and Markup 2), we
examine the following reduced-form equation:
Δμit ¼ γ0 þ γ1 ΔIit þ εit ð8Þ
where Iit=IMit/(PitYit−EXit+IMit) is the import share similar to Bernard,
Jensen, and Schott’s (2006) import competition index.
Columns 1–4 in Table 8.5 report the OLS results of estimating
Equation 8 from the sample of manufacturing industries. Column 1
uses markups from Equation 2, Column 2 uses markups from Equation
4, and columns 3 and 4 use log output price as a dependent variable and
use log marginal cost and its components as control variables. The
results in the table suggest that the changes in markups and output
prices are negatively and statistically significantly associated with the
changes in import shares. As import shares in Japan’s domestic markets
increase, output prices as well as markups decline. To check the validity
Table 8.5 Import competition and markups (manufacturing)

OLS TSLS

Δln(MU1) Δln(MU2) Δln(P) Δln(P) Δln(MU1) Δln(MU2)

Dependent variable: (1) (2) (3) (4) (5) (6)

Δimport share −0.362*** −0.364*** −0.222** −0.237*** −0.548*** −0.496***

(0.083) (0.084) (0.104) (0.073) (0.177) (0.130)
Δln (marginal cost) 0.653***
(0.063)
Δln (input cost) 0.915***
(0.048)
Δln (TFP) −0.411***
(0.041)
Lagged Δln (dependent variable) 1.026*** 0.704***
(0.280) (0.135)

Observations 1078 1078 1078 1078 980 980

R-squared 0.037 0.034 0.658 0.771
Hansen J test (p-value) 0.244 0.061
Kleibergen-Papp LM test (p-value) 0.001 0.000

Notes: (1) Dependent variable is the one-year first-difference in log dependent variable. All explanatory variables are first differenced. (2) Standard
errors that are clustered at the county-industry level are reported in parentheses. (3) ***, **, and * indicate significance at the 1%, 5%, and 10%
confidence levels, respectively. (4) We use the yearly data from years 1990–2012. (5) Columns 5 and 6 report the results from the two stage least
squares (TSLS). We treat the lagged first difference of log dependent variable as an endogenous variable, and instruments it with the two- and three-
period lagged levels. See Bazzi and Clemens (2013). (6) Hansen’s (1982) J test examines the joint null hypothesis of the validity of the model and
instruments. Finally, we report the Kleibergen and Paap (2006) LM test for weak or underidentification of instrumental variables.
Markup Dynamics in Japanese Industries 231

of the results, the results from the two-stage least squares (TSLS) are
reported in the fifth and sixth columns of Table 8.8. In these models, we
include the lagged first differences of the dependent variables and treat
them as endogenous variables, which are instrumented by two- and
three-period lagged levels of the dependent variables.15 The results
from the TSLS not only support our OLS results but also suggest that
elasticities of import competition on markups are much higher in the
manufacturing sector. As reported in Table 8.6, the results from the
service sector are very different. The service industries do not face
similar competitive pressures from foreign competitors. Throughout
the models in columns 1–6, we do not find any negative associations
between markups and import shares.
Throughout the chapter, we argue that the first arrow of Abenomics
improved markups in manufacturing industries because of the depreci-
ation of the yen. Thus, markups in Japanese manufacturing are sensi-
tive to the exchange rate. In Table 8.7, we report the results when we
regress the three types of the prices with the yen rate (the yen divided by
the US dollar). Here, an increase in the exchange rate corresponds to
depreciation of the yen. The first through sixth columns report the
results from the manufacturing sector, and the last two columns report
the results from the service sector. Not surprisingly, depreciation of the
exchange rate mitigates pro-competitive pressures in the manufactur-
ing sector. We do not find a similar tendency in the service sector.

8.6 Conclusion
This chapter examined the evolution of markups over the period of
Abenomics. Prior to Abenomics during the global financial crisis, the
aggregate markup in the manufacturing sector declined by 3.6 percent
from 2006 to 2009 and bounced back by 3.2 percent from 2009 to
2012. The fall and rise of productivities contributed the most to the
U-shape recovery of manufacturing markups. Over the Abenomics
period (2012–2015), markups increased more in the manufacturing
sector than in the service sector, and the rise in output prices was
responsible for this trend. We discussed some suggestive implications
as to why the first arrow of Abenomics – aggressive monetary policy –
has effectively increased manufacturing industries’ markups by easing
pro-competitive effects from foreign competitors. Our results suggest
some important implications for policymakers. Particularly when
Table 8.6 Import competition and markups (services)

OLS TSLS

Δln(MU1) Δln(MU2) Δln(P) Δln(P) Δln(MU1) Δln(MU2)

Dependent variable: (1) (2) (3) (4) (5) (6)

Δ import share 0.348 0.253 0.147 0.151 0.610 0.462*

(0.289) (0.166) (0.105) (0.104) (0.383) (0.259)
Δln (marginal cost) 0.176**
(0.082)
Δln (input cost) 0.150*
(0.073)
Δln (TFP) −0.210*
(0.110)
Lagged Δln (dependent variable) −0.462*** −0.578***
(0.163) (0.162)

Observations 550 550 550 550 500 500

R-squared 0.003 0.002 0.204 0.208
Hansen J test (p-value) 0.00962 0.143
Kleibergen-Papp LM test (p-value) 0.0263 0.272

Note: See Table 8.5.

Table 8.7 Prices and the exchange rate

Manufacturing Services

OLS TSLS OLS TSLS

Δln(P) Δln(Pex) Δln(Pim) Δln(P) Δln(Pex) Δln(Pim) Δln(P) Δln(P)

Dependent variable: (1) (2) (3) (4) (5) (6) (7) (8)

Δln (exchange rate) 0.051*** 0.312*** 0.382*** 0.027* 0.238*** 0.285*** −0.014 0.010
(0.019) (0.058) (0.056) (0.016) (0.061) (0.063) (0.017) (0.011)
Lagged Δln (price) 0.724*** 0.504*** 0.515*** 0.836***
(0.111) (0.088) (0.092) (0.054)

Observations 1,078 1,078 1,078 980 980 980 550 980

R-squared 0.006 0.027 0.048 0.002
Hansen J test 0.244 0.0613 0.0613 0.005
(p-value)
Kleibergen-Papp LM 0.001 0.000 0.000 0.022
test (p-value)

Note: See Table 8.5.

234 Kyoji Fukao and Shuichiro Nishioka

Japanese manufacturing firms face global competition with exporters

from the countries with pegged or controlled exchange rate regimes, it
is important to pay attention to markup dynamics to avoid bankruptcy
risk and the prevalence of the zombies who merely survived because of
financial assistance. Finally, the rise or fall of markups carries an
important implication for increasing inequalities between workers
and capitalists. Abenomics may have created inequality by shifting
profits more to capitalists. This could be one crucial side effect that
policymakers should also consider.

Notes
1. Japan’s GDP growth rate (a 7.1 percent decline) in 2009 was the slowest
among the major advanced countries (International Monetary Fund,
World Economic Outlook Database, April 2019).
2. Zombies are firms that cannot cover debt servicing costs from current
profits over an extended period.
3. We will explain our strategy of computing markups at the industry level
in Section 8.2. We use the revenues (nominal output) divided by total
costs for Figure 8.4. To precisely compare the aggregate trends in the
manufacturing sector from the four countries, we report the markup
trends computed from the same GDP accounting method of the 2008
System of National Accounts (SNA).
4. Our results are particularly interesting because previous studies (e.g.,
Klitgaard 1999; Obstfeld 2011; Kiyota, Nakajima & Nishimura 2009)
do not include this period.
5. See Dekle, Fukao & Ungor (2011) who compare average costs between
Japan and the United States by decomposing changes in marginal costs
in the manufacturing sector into changes in factor prices and total factor
productivity (TFP). By adding changes in output prices, we can analyze
markup dynamics.
6. Even if we allow flexible returns to scale, Equation 5 holds as an identity.
7. See Fukao et al. (2004).
8. The database uses the ex ante approach that was initiated by Hall and
Jorgenson (1967) to impute prices of capital services. See Caballero and
Lyons (1992), Barkai (2020), and Karabarbounis and Neiman (2018)
for similar applications.
9. Following the methodology in Gollop and Jorgenson (1980), quality-
adjusted labor input in the JIP Database 2018 is constructed by
combining data from several labor force surveys. Therefore, labor
input reflects the number of workers, hours worked, and differences in
Markup Dynamics in Japanese Industries 235

hourly wages by sex, age, and educational attainment (see Fukao et al.
2004).
10. The nine nonmarket industries are housing (84), public administration
(91), education (92), medical (93), senior care (95), other public services
(99), and unspecified industries (100).
11. The seventeen additionally excluded industries are rice and wheat
production (1), agricultural services (2), forestry (3), fisheries (4),
mining (5), tobacco (12), petroleum products (23), coal products (24),
electricity (60), gas and heat supply (61), water supply (62), water
supply for industrial use (63), sewer (64), waste disposal (65),
construction (66), civil engineering (67), and real estate (85).
12. One notable difference of the JIP database over other industry-level
GDP databases (e.g., EU KLEMS database) is that the JIP database
reports the nominal rate of capital services computed from the ex ante
approach.
13. The sixteen nonmarket industries are housing (72), private education
(80), private medical (82), private hygiene (83), other public services
(84), public education (98), research (99), public medical (100), public
hygiene (101), social insurance and social welfare (102), public
administration (103), nonprofit medical (104), nonprofit social
insurance and social welfare (105), nonprofit research (106), other
(107), and activities not elsewhere classified (108).
14. The fourteen additionally excluded industries are forestry (5), fisheries
(6), mining (7), tobacco (14), petroleum products (30), coal products
(31), construction (60), civil engineering (61), electricity (62), gas and
heat supply (63), waterworks (64), water supply for industrial use (65),
waste disposal (66), and real estate (71).
15. See Arellano and Bond (1991) and Bazzi and Clemens (2013) for
discussions.

References
Ackerberg, D., K. Caves, and G. Frazer. 2015. “Identification Properties of
Recent Production Function Estimators.” Econometrica 83:
2411–2451.
Amiti, M., and D. E. Weinstein. 2011. “Exports and Financial Shocks.”
Quarterly Journal of Economics 126(4): 1841–1877.
Arellano, M., and S. Bond. 1991. “Some Tests of Specification for Panel
Data: Monte Carlo Evidence and an Application to Employment
Equations.” Review of Economic Studies 58(2): 277–297.
236 Kyoji Fukao and Shuichiro Nishioka

Atkeson, A., and A. Burstein. 2008. “Pricing-to-Market, Trade Costs, and

International Relative Prices.” American Economic Review 98(5):
1998–2031.
Autor, D. H., D. Dorn, L. F. Katz, C. Patterson, and J. Van Reenan. 2020.
“The Fall of the Labor Share and the Rise of Superstar Firms.” NBER
Working Paper 23396.
Barkai, S. 2020. “Declining Labor and Capital Shares.” Quarterly Journal of
Economics 135(2): 645–709.
Basu, S. 2019. “Are Price-Cost Markups Rising in the United States?
A Discussion of the Evidence.” Journal of Economic Perspectives 33
(3): 3–22.
Bazzi, S, and M. A. Clemens. 2013. “Blunt Instruments: Avoiding Common
Pitfalls in Identifying the Causes of Economic Growth.” American
Economic Journal: Macroeconomics 5(2): 152–186.
Bergin, P. R., and R. C. Feenstra. 2009. “Pass-Through of Exchange Rates
and Competition between Floaters and Fixers.” Journal of Money,
Credit and Banking 41(1): 35–70.
Bernard, A. B., J. B. Jensen, and P. K. Schott. 2006. “Survival of the Best Fit:
Exposure to Low-Income Countries and the (Uneven) Growth of U.S.
Manufacturing Plants.” Journal of International Economics 68:
219–237.
Boehm, C. E., A., Flaaen, and N. Pandalai-Nayar. 2019. “Input Linkages and
the Transmission of Shocks: Firm-Level Evidence from the 2011
Tohoku Earthquake.” Review of Economics and Statistics 101(1):
60–75.
Buiter, W. 2009. “Green Shoots: Grounds for Cautious Pessimism.” VoxEU.
https://voxeu.org/article/global-recession-has-not-turned-corner.
Caballero, R. J., T. Hoshi, and A. K. Kashyap. 2008. “Zombie Lending and
Depressed Restructuring in Japan.” American Economic Review 98(5):
1943–1977.
Caballero, R. J., and R. K. Lyons. 1992. “External Effects in US Procyclical
Productivity.” Journal of Monetary Economics 29: 209–225.
Dekle, R., K. Fukao, and M. Ungor. 2011. “The Japan-US Exchange Rate,
Productivity, and the Competitiveness of Japanese Industries.” In
Japan’s Bubble, Deflation, and Long-Term Stagnation, edited by
K. Hamada, A. K. Kashyap, and D. E. Weinstein, pp.105–128.
Cambridge, MA: MIT Press.
De Blas, B., and K. N. Russ. (2015). “Understanding Markups in the Open
Economy.” American Economic Journal: Macroeconomics 7(2):
157–180.
De Loecker, J., P. K. Goldberg, A. K. Khandelwal, and N. Pavcnik. 2016.
“Prices, Markups, and Trade Reform.” Econometrica 84(2): 445–510.
Markup Dynamics in Japanese Industries 237

De Loecker, J., and F. Warzynski. 2012. “Markups and Firm-Level Export

Status.” American Economic Review 102(6): 2437–2471.
Diewert, W. E., and K. J. Fox. 2008. “On the Estimation of Returns to Scale,
Technical Progress and Monopolistic Markups.” Journal of
Econometrics 145: 174–193.
Edmond, C., V. Midrigan, and D. Y. Xu. 2015. “Competition, Markups,
and the Gains from International Trade.” American Economic Review
105(10): 3183–3221.
Feenstra, R. C., and D. E. Weinstein. 2017. “Globalization, Markups, and
US Welfare.” Journal of Political Economy 125(4): 1040–1074.
Fukao, K. (2012). The Lost Two Decades and Japanese Economy. Tokyo:
Nikkei Simbun Publishing.
Fukao, K., T. Inui, H. Kawai, and T. Miyagawa. 2004. “Sectoral Productivity
and Economic Growth in Japan, 1970–98: An Empirical Analysis Based
on the JIP Database.” In Growth and Productivity in East Asia: East
Asian Seminar on Economics Volume 13, edited by Takatoshi Ito and
Andrew Rose, pp.177–228. Chicago, IL: University of Chicago Press.
Fukao, K., and T. Yuan. 2009. “Why Is Japan so Heavily Affected by the
Global Economic Crisis? An Analysis Based on the Asian International
Input-Output Tables.” VoxEU. https://voxeu.org/article/why-has-japan
-been-so-hard-hit-global-crisis.
Gollop, F., and D. Jorgenson. 1980. “U.S. Productivity Growth by Industry,
1947–73.” In New Developments in Productivity Measurement and
Analysis, edited by J.W. Kendrick and B.N. Vaccara, pp.15–136.
Chicago, IL: University of Chicago Press.
Gust. C., S. Leduc, and R. Vigfusson. (2010). “Trade Integration,
Competition, and the Decline in Exchange-rate Pass-through.” Journal
of Monetary Economics 57: 309–324.
Hall, R. E. 1988. “The Relation between Price and Marginal Cost in U.S.
Industry.” Journal of Political Economy 96(5): 921–947.
Hall, R. E., and D. W. Jorgenson. 1967. “Tax Policy and Investment
Behavior.” American Economic Review 57(3): 391–414.
Ito, T., S. Koibuchi, K. Sato, and J. Shimizu. (2016). “Exchange Rate
Exposure and Risk Management: The Case of Japanese Exporting
Firms.” Journal of the Japanese and International Economies 41:
17–29.
Karabarbounis, L., and B. Neiman. 2014. “The Global Decline of the Labor
Share.” Quarterly Journal of Economics 129(1): 206–223.
Karabarbounis, L., and B. Neiman. 2018. “Accounting for Factorless
Income.” NBER Working Paper #24404.
238 Kyoji Fukao and Shuichiro Nishioka

Klitgaard, T. 1999. “Exchange Rates and Profit Margins: The Case of

Japanese Exporters.” FRBNY Economic Policy Review April:
41–54.
Kiyota, K., T. Nakajima, and K.G. Nishimura. 2009. “Measurement of the
Market Power of Firms: the Japanese Case in the 1990s.” Industrial and
Corporate Change 18(3): 381–414.
Kleibergen, F., and R. Paap. 2006. “Generalized Reduced Rank Tests Using
the Singular Value Decomposition.” Journal of Econometrics 133(1):
97–126.
Obstfeld, M. 2011. “The Yen and Japan’s Economy, 1985–2007.” In
Japan’s Bubble, Deflation, and Long-Term Stagnation, edited by
K. Hamada, A. K. Kashyap, and D. E. Weinstein, pp.51–104.
Cambridge, MA: MIT Press.
Pierce, J.R., and P.K. Schott. 2016. “The Surprisingly Swift Decline of US
Manufacturing Employment.” American Economic Review 106(7):
1632–1662.
Roeger, W. 1995. “Can Imperfect Competition Explain the Difference
between Primal and Dual Productivity Measures? Estimates for U.S.
Manufacturing.” Journal of Political Economy 103(2): 316–330.
Shin, J. S. 2013. The Global Financial Crisis and the Korean Economy.
London: Routledge.
Syverson, C. 2011. “What Determines Productivity?” Journal of Economic
Literature 49(2): 326–365.
Timmer, M.P., T. van Moergastel, E. Stuivenwold, G. Ypma, M. O’Mahony,
and M. Kangasniemi. 2007. “EU KLEMS Growth and Productivity
Accounts Version 1.0 Part I Methodology.” Working Paper.
9 The Crisis That Wasn’t
How Japan Has Avoided a Bond
Market Panic
m a r k b a m b a a n d d a v i d e. w e i ns t e i n

9.1 Introduction
In the two decades since Takayama et al. (1999) published their work
indicating substantial generational imbalances in Japan arising from
Japan’s aging population, a large number of papers have been pub-
lished arguing that the Japanese government’s fiscal situation is not
sustainable.1 Many have interpreted these results as implying that
a Japanese default or a hyperinflation was imminent.2 The fact that
yields on Japanese government bonds (JGBs) have remained low for
almost twenty years suggests markets never feared a JGB crisis. Some
have suggested that this may reflect a bubble in the JGB market, but if
so, it is probably the longest-lasting bubble in the history of the world.
This paper explores the history of Japanese fiscal policy over the past
two decades with the aim of better understanding where previous
forecasts have erred and how policies adopted in the second Abe
administration have affected Japan’s fiscal sustainability. Japan pro-
vides an important case study of how a country facing intense fiscal
pressures can avoid a hyperinflation or financial panic. We find that
there were three key forces that likely improved Japan’s fiscal situation
relative to more pessimistic predictions. First, the Japanese government
has shown remarkable ability to hold down per capita expenditures on
social pensions and healthcare. Second, the Japanese government has
been able to raise taxes substantially. Third, the remarkable monetary
expansion pursued by the Bank of Japan (BOJ) has resulted in
a dramatic decline in the amount of government bonds held by the
private sector.

* An earlier draft of this paper was circulated with coauthors’ names Mark
T. Greenan and David E. Weinstein.

239
240 Mark Bamba and David E. Weinstein

Some of these policies can be directly attributable to “Abenomics,”

while for others the link is less clear. We argue that Abe’s “first arrow,”
expansive monetary policy, was a clear break from past trends and
served to lower the net amount of public debt held by the private sector
by 70 percent between 2012 and 2017. As a result, Japan’s government
debt has been largely monetized. This fact means that while inflation
remains a possibility, there is not much scope for a major private sector
selloff of Japanese government bonds (JGBs) because so many of them
are held by the BOJ and government institutions.
On the fiscal side, we see little evidence of Abe’s second arrow
(expansionary fiscal policy). The share of government expenditures in
gross domestic product (GDP) has stayed remarkably stable at
24–25 percent for every year between 1998 and 2018.3 Total govern-
ment expenditures fell as a share of GDP and tax revenues rose. While
this fiscal austerity was not in line with Abe’s initial positions, it did
serve to substantially close the Japanese government budget deficit and
reduce the gross and net debt levels (as a share of GDP). Thus, the Abe
administration was more fiscally conservative than one might have
expected at the outset.
In order to understand the disconnect between studies predicting
unsustainable policies and market reaction, it is worth bearing in
mind a few points. The first is that one must always be wary of any
such prediction because of the impossibility theorem that governs
forecasts of asset prices. If markets knew with certainty that the
Japanese government would default next year, interest rates would
skyrocket today, forcing a default today. But if the default is today,
then the original “certain” forecast of a default must have been wrong –
the default occurred today rather than next year. Given the self-
fulfilling nature of defaults and the resulting difficulty in predicting
them, we should avoid definitive claims or counterclaims about
whether a crisis is inevitable. Nothing is certain; at best, we can talk
about default probabilities.
The second point to realize is that discussions on fiscal sustainability
can be a bit confusing because showing that a government’s fiscal situ-
ation is not sustainable does not necessarily mean that a crisis is immi-
nent. Typically, fiscal situations are deemed unsustainable if there is some
difference between expected future expenditures by the government and
expected future government receipts. For example, both economists and
Japanese young people know that the amount of social pension payments
The Crisis That Wasn’t 241

relative to social pension contributions is likely to decline. While this

means that the current Japanese system is unsustainable because future
generations won’t get the same benefits as past generations, it does not
mean that a crisis will be forthcoming. Default can be averted by revisions
in pension indexing formulas, raising the retirement age, raising co-pays
on medical insurance, increasing taxes, more inflation, etc. In other
words, fiscal forecasts reveal a future but not necessarily the future. As
Dickens’s Scrooge avers, we “may change these shadows . . . by an altered
life.” Put in the more prosaic language familiar to economists, economic
forecasts are built on assumptions about future events and the behavior
of government and private agents that may or may not come to pass. But
the fact that we have had two decades of these forecasts provides us with
an interesting opportunity to compare what actually happened in Japan
with what economists predicted would happen.
In order to do this, we are going to update Broda and Weinstein
(2005), which used data up to the year 2000 to forecast what would
happen to Japan’s fiscal situation over the next 100 years. This paper,
optimistically titled, “Happy News from the Dismal Science,” is an
outlier in the literature, as it was, along with Morgan (2004), among
a small set of papers arguing that only modest changes were required to
avert a crisis. Broda and Weinstein (2005) implemented Blanchard’s
(1990) fiscal accounting system. In the Blanchard framework, fiscal
sustainability depends on three factors: the current level of Japanese
government liabilities and the forecast of future government expend-
itures and revenues. Broda and Weinstein extended this framework by
including the assets and liabilities of the Bank of Japan.
The simplicity of the Blanchard framework has advantages and disad-
vantages. On the positive side, fiscal sustainability comes down to just
a few variables: the level of net debt, the difference between interest and
growth rates, monetary policy, and the path of expenditures and taxes.
The principle disadvantage is that the calculation does not involve an
economic model. The Blanchard approach is essentially an intertemporal
budget accounting exercise, but as such it does lay bear the basic factors
that need to come into play in order to make a sustainability calculation.

9.2 A Simple Framework for Fiscal Sustainability

To help fix ideas, we first summarize the Blanchard (1990) framework as
used in the Broda and Weinstein (2005) paper. In this setup, government
242 Mark Bamba and David E. Weinstein

expenditures can be divided into three categories: public pension pay-

ments and medical benefits for the elderly, Ht; all other expenditures,
except interest, Gt; and interest on the debt, where Bt is the level of
government net debt and it is the interest rate on that debt. The govern-
ment’s intertemporal budget constraint can then be written as:
ðGt þ Ht þ it Bt1 Þ  Tt ¼ ðBt  Bt1 Þ þ ðMt  Mt1 Þ; ð1Þ
where Gt is other government non-interest expenditures; Ht is transfers
to the old; Bt and it are the level of government net debt and the interest
rate on debt in period t, respectively; Tt denotes government revenues;
and Mt is the money supply in period t. In this setup, any fiscal deficit
(given on the left-hand side) must be matched by either increasing net
debt or increasing the money supply.
If we divide both sides of Equation 1 by nominal GDP and rearrange
terms, we obtain
1 þ it λt
b t ¼ gt þ h t  τ t þ bt1  mt ; ð2Þ
1 þ ηt 1 þ λt
where τt is Tt/GDPt, ηt is the growth rate of nominal GDP, and λt is the
growth rate of the nominal money supply. Starting with an initial level
of debt-to-GDP is given by b0, we can express the level of debt-to-GDP
in period n as:
Xn  nt    
1 þ it λt 1 þ it n
bn ¼ gt þ ht  τt  mt þ b0 : ð3Þ
t¼1
1 þ ηt 1 þ λt 1 þ ηt

Equation 3 is simply the result of accounting and involves no theories of

economic behavior. In this setup the “sustainable” tax rate is defined
as the value of τt that sets level of debt in the year 2100 equal to that in
2000.
Broda and Weinstein (2005) did not formally model the evolution of
each term in Equation 3; they just used rules of thumb to produce
“plausible” paths under various assumptions about fiscal and monet-
ary policy. More sophisticated papers model the evolution of each of
the variables with time subscripts. However our interest is not in testing
or calibrating models but rather in seeing what actually happened to
each of these series empirically with an eye to assessing whether or not
taxes in Japan will have to rise if the future behavior of the Japanese
government looks like the past behavior.
The Crisis That Wasn’t 243

9.2.1 Assessing Assumptions

The first key assumption in Broda and Weinstein (2005) was the
interest rate gap or the difference between the nominal interest rate
and the nominal growth rate. An increase in interest rates relative to
growth rates tends to raise the sustainable tax rate because existing
debt becomes more costly to repay. Since forecasting this variable is
difficult, Broda and Weinstein (2005) used a number of interest rate
gaps ranging from 1 percent to 4 percent, with a preferred level of
2 percent.
Figure 9.1 shows the evolution of this rate gap from 2000 to 2017.
There are several notable features of the plot. First, while there have
been some large swings in Japanese nominal GDP over the past seven-
teen years, the long-run growth rate of Japanese nominal GDP has been
almost constant: only rising by an average of 0.3 percent per year. By
contrast, real GDP grew by 0.9 percent per year, with the difference due
to the fact that the GDP deflator fell by about 0.6 percent per annum.
Second, interest paid on Japanese debt fell consistently over this time
period to levels below 1 percent. These two facts mean that the average
interest rate gap was only 1.0 percent. This low interest rate gap –
which fell below zero during Kuroda’s term (2013–present) – meant
that it was easier for the Japanese government to finance the debt.
Two other important variables are the expenditures on the “elderly”
(ht) and the “young” (gt). In Broda and Weinstein (2005), elderly
expenditures were composed of government expenditures on social
pensions and elderly medical care, which they assumed to be paid
from age sixty-five onwards. Expenditures on the young constituted
all other expenses. Given Japan’s rapidly aging population, a crucial
feature of any sustainability calculation is getting the path of these
variables right. Indeed, the principle reason that Broda and Weinstein
(2005) was, in Hubbard and Ito’s opinion, an “optimistic” assessment
of Japan’s future can be traced back to forecasts of the development of
these variables.
Broda and Weinstein (2005) showed (in their “Case 3”) that – if one
assumed that per capita pension benefits rose with GDP per worker,
their proxy for wages – Japan needed to raise taxes dramatically in
order to stabilize the debt level. However, a major reason for the
optimism of Broda and Weinstein arose because, in their baseline
cases, they assumed that Japan would be able to hold down the path
 Japan’s Interest Rate Gap
4%
Average Interest Rate =1.3%
3%
Nominal Interest Rate
2%

1%

0%
1998 2000 2002 2004 2006 2008 2010 2012 2014 2016
-1%
Nominal Growth Rate
-2%
Average Nominal Growth Rate = 0.3%
-3%

-4%

-5%
Nominal interest rate is the total interest paid divided by gross debt. Averages are calculated using geometric means from 2000 to 2017. Interest-payment and GDP data is from the
2017 National Accounts. Gross debt data is from the BOJ’s Flow of Funds.

Figure 9.1 Japan’s interest rate gap

The Crisis That Wasn’t 245

of per capita expenses on the elderly to rise only as fast as GDP. Given
that GDP was essentially flat over the past seventeen years, this implied
we should have seen no growth in per capita payments to the elders.
It is important to remember that, in terms of the generational
accounting approach of Takayama et al. (1999), this sort of fiscal
discipline produces large generational imbalances because it implies
that current and future Japanese workers will pay taxes that will rise
with their earnings but will receive pensions that are indexed to a lower
rate. In other words, the current generation of Japanese workers is
making a large net transfer to retirees without a prospect of getting
a similar transfer in the future from their children. In this sense, Japan’s
policy is not sustainable because future generations will get much less in
benefits per dollar of taxes paid than past generations. However, it is
important to also realize that this does not mean that crisis is imminent.
Whether one assumes that pensions rise with wages – or with GDP
per worker, as in Broda and Weinstein (2005) – is a crucial assumption.
On one level this assumption seems non-controversial because current
pension formulas tie the pensions to wages. However, governments are
free to change the rules governing what they will pay retirees, and the
Japanese have been particularly apt at (quietly) doing it in a variety of
ways. One important step in this direction was implemented in 2004 as
the Japanese government reformed the social security law. In this
reform, the government changed the social pension indexing formula
in such a way that benefits would be reduced if the number of contribu-
tors to the system declined or life expectancy rose. Takayama (2004)
estimated that this would slow the growth rate of benefits to be 0.9 per-
cent per year less than wages. Moreover, the Japanese government has
been aggressive at holding down healthcare fee and price increases
while at the same time increasing co-payments (thereby shifting health-
care costs from the government to private citizens).
We can see the impact of these policy changes in Figure 9.2, which
displays nominal GDP, nominal GDP per worker, per capita expend-
itures on the elderly, and per capita expenditures on the young all
normalized to equal one in the year 2000. As we noted earlier, nominal
GDP rose by a paltry 3.6 percent, and GDP per worker rose by 18 per-
cent. What is most striking is that per capita expenditures on the elderly
(i.e., government expenditures on social pensions and elderly medical
care divided by the elderly population) declined by 11 percent between
2000 and 2017. This decline was not matched by a decline in wages, as
 Per Capita Expenditure Versus GDP
1.4

1.3

1.2

1.1

0.9

0.8

Value (Normalized to One in 2000)

0.7

0.6
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
Per Capita Expenditure on the Elderly (Normalized) Per Capita Expenditure on the Young (Normalized)
Nominal GDP (Normalized) GDP per Worker (Normalized)
Per Capita Expenditure on the Elderly refers to expenditure on social pensions, elderly medical care, and long-term care divided by the elderly population
(ages 65+). Per Capita Expenditure on the Young refers to all other expenditures (excluding interest payments) divided by the non-elderly population
(age 0-64). Expenditure and GDP data is from the 2019 National Accounts. Historical population data is from the Japan Statistics Bureau.

Figure 9.2 Per capita expenditure versus GDP

The Crisis That Wasn’t 247

Japanese hourly earnings in manufacturing rose slightly over this

period. Thus, while the number of elderly rose by over 50 percent
between 2000 and 2017, total expenditures on them grew by less
because of a reduction in per capita benefits. Interestingly, per capita
expenditures on the young did not exhibit this trend, rising faster than
even per worker GDP. Here, however, the opposite logic applies. While
the per capita expenditures rose, total expenditures rose much slower
because the number of young people fell over this time period.
Bringing this discussion back to what Japan’s experience tells us
about fiscal sustainability, we see that, while expenditures on the
elderly have been rising, they have done so at a much slower rate
than was forecast even in the most fiscally restrained version of Broda
and Weinstein (2005) and substantially slower than forecasts of pen-
sion benefits that predicted them to rise with wages, which most
economists assumed would rise roughly proportionally with GDP per
worker. In fact, wages in Japan rose much slower; average nominal
hourly earnings fell 6 percent between 2000 and 2017, and average
wages fell 7 percent.4 Both of these forces served to hold down the
growth of pension obligations, and this slow growth, along with
restraint on healthcare spending, worked toward preventing a crisis
from emerging.
A second important factor in terms of understanding Japan’s
response to its aging society is the impact of the 2004 pension reforms
on government revenues. As Takayama (2004) discusses, the reforms
phased in a 4.72 percentage point increase in the social pension tax
between 2004 and 2017. Moreover, the Abe government implemented
a 3 percentage point increase in the consumption tax in 2014.
We can see the impact of these policies in Figure 9.3, which shows
Japanese government revenue as a share of GDP and its sources.
Between 2000 and 2017, government revenues rose from being just
over 30 percent of GDP to 36 percent of GDP. Much of this increase
came from social contributions (which rose by 3 percentage points) and
sales taxes (which rose by 2 percentage points). To give some sense of
how important these tax increases are, we can return to the Broda and
Weinstein (2005) forecasts, which suggested that, if Japan could hold
down the per capita rate of pension rate increases to the GDP growth
rate, a 6-percentage-point-of-GDP tax increase would render the debt-
to-GDP ratio stable over a 100-year horizon in all but the most gener-
ous public expenditure scenarios.
 Sources of Government Revenue
35.0
4.2 4.1 4.1
3.7
30.0 2.4 2.4 2.6 2.7
2.3 2.4 2.5 2.5 2.6
2.4 2.5 2.5 Sales Tax
2.4 2.3
25.0 12.5 12.6 12.9 13.0
10.5 10.7 12.4 Social Contribution
9.6 10.0 10.0 10.2 11.3 12.0 12.4
10.1 10.0 11.3 11.5
20.0 Property Tax
2.3 2.3 2.2 2.1 2.2 2.2 2.3 2.3
2.2 2.3 2.3 2.3
15.0 2.3 2.2 2.4 2.3 Corporate Tax
3.7 2.3 2.3
3.8 3.7 4.3 4.9 4.9 3.9 4.1 4.5 4.3 4.2 4.5
3.2 3.7

Percent of GDP
3.2 2.6 3.2 3.4
10.0 Individual Tax
5.0 4.9 4.1 4.2 4.4 4.6 5.0 5.1 5.1 5.1 5.1
4.3 4.8 4.6 4.7 4.9 5.2 5.3
Other
5.0
7.3 7.3 6.8 7.2 7.4 7.3 6.9 6.8 6.6 6.3 6.1 6.1 6.1 6.3 6.5 7.1 6.8 6.5
0.0
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
Corporate Tax refers to the line item "Payable by corporations and other enterprises." Sales Tax refers to the line item
"Value-Added Taxes." Data is from the 2017 National Accounts

Figure 9.3 Sources of government revenue

The Crisis That Wasn’t 249

Government Revenues and Expenditures

45
Tohoku Earthquake

40
Percent of GDP

Financial Crisis Financial Crisis

35

30
2003 Results
25
1990 1995 2000 2005 2010 2015 2020
Government Revenues Government Expenditure
Data is from the 2017 National Accounts

Figure 9.4 Government revenues and expenditures

Another way of seeing the impact of these reforms is by looking at

the evolution of government revenues and expenditures, expressed as
a share of GDP, which is shown in Figure 9.4. Japanese government
expenditures show a clear upward trend driven by the aging society.
Most of the big increases followed the economic stimulus packages
after the 1997 and 2007 financial crises and the reconstruction pack-
age following the Tohoku earthquake. With the exception of these
extraordinary events, expenditures have been relatively stable as
a share of GDP. More striking is the time path of government rev-
enues, which since 2000 (and especially since 2009) have been rising
faster than expenditures. These measures to suppress costs and raise
revenues have helped to significantly improve Japan’s long-term fiscal
outlook.
A final component of the sustainability calculation concerns the
evolution of Japan’s government debt position. The dotted line in
Figure 9.5 shows the steady rise of Japan’s gross debt that has been
the source of so much consternation. However, what is less well publi-
cized is the financial assets held by the government of Japan. The
government owns a large amount of financial assets (principally foreign
and domestic government bonds, currency, and equities), and these
assets reduce the net liabilities of the Japanese government and there-
fore reduce the fiscal pressure.
One can see this formally by modifying Equation 1 so that we
introduce gross assets and liabilities. Since the definition of net debt is
 Japan’s Debt Levels
250

200

150

100

Percent of GDP
50

0
1996 1998 2000 2002 2004 2006 2008 2010 2012 2014 2016
BOJ-Inclusive Net Debt Net Debt Held Privately Gross Debt
Gross debt includes all line items in general government liabilities except for "Equity and investment fund shares" and "Financial
derivatives and employee stock options." BOJ-Inclusive Net Debt is gross debt less general government asset less BOJ net worth.
Net Debt Held Privately is gross debt less general government asset less the BOJ's JGB holdings. Data is from the BOJ's Flow of Funds.

Figure 9.5 Japan’s debt levels

The Crisis That Wasn’t 251

government financial assets less liabilities (i.e., Bt = Lt – At), we can

rewrite Equation 1 in terms of gross debt as
½Gt þ Ht þ it ðLt1  At1 Þ  Tt ¼ ðLt  Lt1 Þ
ðAt  At1 Þ þ ðMt  Mt1 Þ:

It is immediately apparent that any combination of gross assets and

liabilities that satisfies Equation 4 will also satisfy Equation 1. In other
words, given a particular path of government expenses and taxes, all
that matters for the government budget constraint is the net asset
position of the government. Thus, if one does the accounting correctly,
it does not matter how one wants to do fiscal accounting using net or
gross debt. Directly below the gross debt line in Figure 9.5 lies the BOJ-
inclusive net debt share of GDP. Here, we combine the net worth of the
BOJ with the net debt of the government, so our numbers differ slightly
from official net debt numbers, which exclude the BOJ net worth, but
the differences would hardly be visible in the plot. Interestingly, both
net and gross debt have been falling as a share of GDP in recent years,
which reflects the fiscal consolidation that we saw in Figure 9.4.
Finally, we also plot the net debt held by the private sector, which
equals the net debt of the Japanese government less the JGB holdings of
the BOJ. This number is useful for understanding the extent of bond
holdings by the private sector. As one can see, private net holdings of
JGBs fell sharply after Abe became prime minister, driven largely by the
BOJ’s open market operations. One implication of this plot is that it
would likely be quite difficult for the private sector to precipitate a major
selloff of JGBs for the simple reason that so many of them are held by the
BOJ or the government. Thus, while it is possible that a rise in inflation
might cause losses in the BOJ’s bond portfolio and cause difficulties if the
BOJ were to try to tighten monetary policy by selling JGBs, this problem
would only be an issue to the extent that the BOJ were not comfortable
letting the price level rise. In other words, there is relatively little scope
for a loss in confidence in JGBs per se to drive the crisis.
One final important element of the calculation is the path of monet-
ary policy. Auerbach and Obstfeld (2005) and Eggertsson and
Woodford (2003) made strong cases for large-scale open market oper-
ations in order to help stimulate the Japanese economy. The idea that
Japan should be following some price-level target along with a positive
inflation target has been widely accepted – see, for example, Ito and
252 Mark Bamba and David E. Weinstein

Mishkin (2005) – but the idea seems to invoke deep ambivalence on the
part of economists and policymakers. On the one hand, the theory of
a price-level target dictates that the central bank should allow future
inflation to be substantial in order to restore the price level to the level
that would have obtained if there had been no deflation. On the other
hand, the reality of the implied inflation rates required to achieve
a price-level target makes many economists uneasy even as they advo-
cate it.
For example, if one believes that Japan should have targeted the price
level in 1997 (as in Ito and Mishkin 2005) and that consumer price
inflation should have averaged 2 percent per year since then (due to
biases and other factors), then one should also believe that the price
level today should be about 50 percent higher than it is. This creates
a bit of a conundrum for policymakers: is the large increase in the
money supply and its implications for future inflation a fundamental
part of the solution or a fundamental part of the problem?
This issue also raises the question of what monetary policy is “nor-
mal” for Japan. As one can see from Figure 9.6, the growth rate of
money supply has been remarkably constant over long horizons in

Log Money Supply

34.5

34.0

33.5
Kuroda is
Governor of BOJ
33.0

32.5

32.0

31.5
1990

1992

1994

1996

1998

2000

2002

2004

2006

2008

2010

2012

2014

2016

2018

Figure 9.6 Log money supply

The Crisis That Wasn’t 253

Japan. We see a substantial increase between 1995 and 2002, but, aside
from that increase (and until the policies of Governor Kuroda), we see
little change in the long-run growth rate of the money supply. It is not
obvious from this picture how far back in Japanese history we should
go to see normal money-supply growth. If we go back to the 1990s, it
looks like the level of money in the economy is about 50 percent higher
than what we might have expected by extrapolating the trend. Of
course, one might argue that the five years following the collapse of
the Japanese bubble was an exceptional period, but this raises ques-
tions. Was monetary policy normal during the bubble years? Following
the first oil shock and the collapse of Bretton Woods? During the high-
growth period? In other words, without a theoretical model (e.g.,
a price-level target), it is difficult to look at long-run trends for guidance
about what monetary policy makes sense for Japan.
One fact that is indisputable is that the recent open market policy of
the Bank of Japan has had a dramatic impact on the amount of govern-
ment bonds held by the private sector. As one can see in the dashed line
in Figure 9.5, the net debt held privately, which equals the net amount of
bonds held by agents other than the Japanese government and the Bank
of Japan, has fallen precipitously as a percentage of GDP.
This figure provides another important reason why a JGB crisis hasn’t
materialized – there simply are not that many private entities holding
Japanese government debt. Obviously, if inflation started to rise, the Bank
of Japan might feel conflicted about its desire to reduce money growth and
not precipitate a crisis in the bond market, but it seems like the BOJ could
easily prevent any spike in JGB interest rates by buying up the small
amount of remaining JGBs in the market. This might involve some higher
inflation, but it is hard to see how this would turn into a default or even
hyperinflation. For example, if Japan ended up with an inflation rate of
5 percent per year for ten years, Japan’s price level would approximately
hit the 1997 price level (adjusted by a 2 percent per year inflation rate),
more or less what economists had advocated when writing the initial
price-level target papers. Avoiding a future like this hardly seems
a reason for a central bank to create a crisis in the bond market.

9.3 A New Sustainability Calculation

With this data preview as a background, we are now able to update the
results from Broda and Weinstein (2005). These forecasts used data up
254 Mark Bamba and David E. Weinstein

through 2000 to forecast expenditure and debt levels for the next 100
years. We are now seventeen years into that forecast, so it is reasonable to
wonder how well they did. In order to do the update, we now replace
forecast values for the years 2000–2017 with actual data and then com-
pare both how the various forecasts performed over this time period and
what the revised forecast looks like over the next eighty-three.
Broda and Weinstein (2005) considered three scenarios in their
forecasts that allowed for different paths of government benefits. In
Case 1, per capita expenditures on the elderly (which we define as ages
sixty-five and older) were assumed to rise at the same rate as GDP
growth while total expenditures on the remaining population were
assumed to be a constant share of GDP. In Case 2, per capita expend-
itures (both on the young and on the elderly) were assumed to always
be proportional to GDP. Finally, in Case 3, per capita expenditures (on
both the young and the elderly) were assumed to rise with GDP per
worker. These three cases differ dramatically in the implied growth
rates of Japanese government expenditures for the elderly and young.
In other words, Case 3 is the most generous in that all per capita
expenditures rise with GDP per worker; Case 1 maintains a generous
policy of expenditures for the young but implies more fiscal discipline
for expenditures on the elderly; and Case 2 imposes fiscal discipline on
both the young and the elderly.
Our sustainability numbers differ slightly from those of Broda and
Weinstein (2005) because there have been a number of data revisions in
the intervening years principally to population forecasts and GDP. In
particular, we use the most recent National Institute of Population and
Social Security Research (NIPSSR) forecasts for population (from the
2019 Japan Statistical Yearbook), which differ both from the Faruqee
and Muhleisen (2001) forecasts as well as the original NIPSSR fore-
casts used by Broda and Weinstein (2005). In addition, we update our
historical average values of expenditures using data from the seventeen
years since 2000, when Broda and Weinstein (2005) took their data.
Table 9.1 shows the results of this exercise. These numbers can be
interpreted as the hypothetical tax rate, if implemented today, that
would leave the level of privately held net debt in 2100 the same as it is
in 2018, assuming that future expenditures follow three possible scenarios
(as outlined earlier) and if the money supply were to grow proportionally
to nominal GDP. In each case, the first column reports the sustainable tax
rate computed using current data, and the second column replicates the
The Crisis That Wasn’t 255

Table 9.1 Sustainable tax rates based on privately held net debt

Sustainable tax rates

Case 1 Case 2 Case 3

Forecasts 2018 2003 2018 2003 2018 2003

Rate gap
0 33.1 33.6 26.4 27.0 42.7 42.5
1 33.9 34.5 28.1 28.9 42.4 42.1
2 34.7 35.3 29.6 30.7 42.1 41.5
3 35.3 36.0 31.0 32.1 41.9 41.1
4 35.9 36.6 32.2 33.4 41.7 40.7

2003 tax rate 29.1

Current tax rate 35.6

Sustainable Tax Rate refers to the hypothetical tax rate (measured as general
government revenues divided by nominal GDP) that, if implemented today, would
leave the level of privately held net debt in 2100 the same as it is in 2018. Entries are in
percentages of GDP. In Case 1, per capita expenditures on the elderly are proportional
to GDP, while expenditures on the young are proportional to GDP. In Case 2, per
capita expenditures (both on the young and on the elderly) are proportional to GDP. In
Case 3, per capita expenditures (both on the young and on the elderly) are
proportional to GDP per worker. Both 2003 and 2018 results use population
forecasts from NIPSSR.

Broda and Weinstein calculation using the data they had available (but
adjusted for data and accounting revisions that have occurred since then
to make the numbers comparable with column 1).5
There are two interesting features of this table. The first is that there’s
been remarkably little change in the sustainable tax rate between 2003
and 2018. In Cases 1 and 2 the sustainability tax rates fell slightly, and
in Case 3 (the most generous case) they went up slightly.
Taken at face value, the fact that sustainable tax rates have barely
changed since 2003 makes it difficult for us to determine which of the
three cases best reflects Japan’s fiscal outlook. The drop in the sustain-
able tax rate in Case 1 occurred largely because of the cost controls on
per capita benefits for the elderly and NIPSSR’s new population fore-
casts that expect a faster population decline. However, the same popu-
lation forecasts are also responsible for the rise in the rate for Case 3, as
256 Mark Bamba and David E. Weinstein

we now expect a faster working-population decline (relative to the

whole population) than was predicted in 2003. For all three cases, the
lower-than-predicted expenditures on the elderly have pushed down
the updated sustainable tax rate, while the higher-than-predicted
expenditure on the non-elderly (often triggered by recession-induced
fiscal stimulus packages) have pushed up the updated sustainable
tax rate. These offsetting forces have left the sustainable tax rates more-
or-less unchanged. However, it is important to note that, while past
under-predictions in non-elderly expenditures were mostly driven by
short-term expenditures (often in response to recessions) that will likely
stabilize in the future, past over-predictions in elderly expenditures
were due to long-term cost-cutting measures that will likely continue
in the future. This observation leads us to favor Case 1 or 2 over Case 3,
as the higher tax rate in Case 3 depends wholly on the assumption that
the future path of per-capita benefits for the elderly will rise at a much
faster rate than we have observed over the past twenty years.
The second notable feature is that the rise in Japanese taxes has pushed
Japan closer to the sustainable level. For example, in the preferred specifi-
cation of the Broda and Weinstein (2005) paper (a 2 percentage point rate
gap), government revenues needed to rise by 1.6 percentage points of GDP
in Case 2 (our baseline case), 6.2 percentage points in Case 1 and 12.4 per-
centage points in Case 3. Thanks in large part to the increases in govern-
ment revenue (from 29.1 percent to 35.6 percent), government revenues as
a percent of GDP are now 6 percentage points above sustainability for
Case 2 (versus 1.6 percent below in the previous forecasts), 0.9 percentage
points above sustainability for Case 1 (versus 6.2 percent below in the
previous forecasts), and 6.5 percentage points below sustainability for
Case 3 (versus 12.4 percent below in the previous forecasts). The rise in
government revenues is driven primarily by the large tax increases that
Japan has already enacted (in social contributions and sales tax).
Our updated numbers therefore indicate that, at this level, fiscal
policy is sustainable in most scenarios (Cases 1 and 2) and is much
closer to sustainability than it was previously for Case 3. Again, if one
believes that fiscal discipline will deteriorate in the future (i.e., in
Case 3), then Japan will have to raise taxes, but maintaining the current
level of aggressiveness at holding down cost increases may be sufficient
to avoid tax increases.
Table 9.2 repeats the same exercise using net debt (including the
BOJ’s net worth) as the relevant measure of debt rather than privately
The Crisis That Wasn’t 257

Table 9.2 Sustainable tax rates based on BOJ-inclusive net debt

Sustainable tax rates

Case 1 Case 2 Case 3

Forecasts 2018 2003 2018 2003 2018 2003

Rate gap
0 33.1 33.6 26.4 27.0 42.7 42.5
1 34.8 34.5 29.0 28.9 43.3 42.1
2 36.4 35.3 31.3 30.7 43.9 41.5
3 37.9 36.0 33.6 32.1 44.4 41.1
4 39.3 36.6 35.6 33.4 45.1 40.7

2003 tax Rate 29.1

Current tax rate 35.6

Sustainable Tax Rate refers to the hypothetical tax rate (measured as general
government revenues divided by nominal GDP) that, if implemented today, would
leave the level of BOJ-inclusive net debt in 2100 the same as it is in 2018. Entries are in
percentages of GDP. In Case 1, per capita expenditures on the elderly are proportional
to GDP, while expenditures on the young are proportional to GDP. In Case 2, per
capita expenditures (both on the young and on the elderly) are proportional to GDP. In
Case 3, per capita expenditures (both on the young and on the elderly) are
proportional to GDP per worker. Both 2003 and 2018 results use population
forecasts from NIPSSR.

held net debt, which was used in Table 9.1. Using net debt would be
appropriate if we expect the BOJ to unwind its JGB holdings in the
coming years. These more conservative estimates raise sustainability
tax rates across the board as a higher debt level today implies higher
interest payments. Yet we find only modest increases in sustainability
tax rates – rising by less than 2 percentage points for a 2 percentage
point rate gap (our preferred specification) and less than one percentage
point for a 1 percentage point rate gap (which better reflects the twenty-
year historical average). Thus, even if we exclude the BOJ’s recent
easing from our analysis, our estimates suggest we can remain optimis-
tic about Japan’s fiscal outlook, especially given the increase in the
consumption tax that was implemented in 2019.
Figures 9.7, 9.8, and 9.9 further explore the reasons for the different
forecasts in Cases 1–3. Each figure portrays historical and forecast
258 Mark Bamba and David E. Weinstein

Japan 2020–2100: Government Expenses as a Share of GDP

by Population Group (Case 1)
60
2018
50
Previous
Financial Crisis Predictions
40
Percent of GDP

30 Updated
Prediction
Historical Levels
20
Previous
Prediction
10 Updated
Prediction
0
1980 2000 2020 2040 2060 2080 2100 2120

Total Expenditure Elderly (65+) Young (<65)

Total (old predictions) Elderly (old predictions) Young (old predictions)

In Case 1, per capita expenditures on the elderly are proportional to GDP while expenditures
on the young are proportional to GDP

Figure 9.7 Japan 2020–2100: government expenses as a share of GDP by

population group (Case 1)

expenditure levels for each of our three cases. The lighter (blue) curves
show the forecasts for the expenditure levels in total and for the young
and elderly based on where Japan was in the year 2000. The black lines
to the left of the vertical bar show what actually happened to the series
between 2000 and 2017. The black lines to the right of these series
show updated forecasts of the series.
Comparing the historical expenditure levels with what was predicted
in 2003 (the section of the plots to the left of the vertical bar), we can
make a few observations. The first is that most of the discrepancies
between the actual numbers and the 2003 predictions were a result of
the increases in government spending stemming from the financial crisis
and the Tohoku earthquake. We can see in all three plots that the actual
expenditures on the young (which includes all expenditures that are not
for the elderly and would therefore include expenditures from a fiscal
stimulus) overshot the predicted expenditure levels by several percent-
age points starting in 2008 and have been slow to fall back to pre-crisis
levels. While these expenditure levels may not return to levels seen
before the crisis, our forecasts assume that they will continue to fall to
The Crisis That Wasn’t 259

Japan 2020–2100: Government Expenses as a Share of GDP

by Population Group (Case 2)
60

2018
50
Updated
Percent of GDP

40 Prediction

30 Updated
Prediction Previous
Historical Levels Prediction
20

10 Updated
Previous
Prediction
Prediction
0
1980 2000 2020 2040 2060 2080 2100 2120

Total Expenditure Elderly (65+) Young (<65)

Total (old predictions) Elderly (old prediction) Young (old predictions)

In Case 2, per capita expenditures (both on the young and the elderly) are proportional to GDP

Figure 9.8 Japan 2020–2100: government expenses as a share of GDP by

population group (Case 2)

Japan 2020–2100: Government Expenses as a Share of GDP

By Population Group (Case 3)
60
2018
50 Previous
Prediction
Financial Crisis
Percent of GDP

40
Updated
Updated
30 Prediction
Prediction
Historical Levels

20

10 Previous
Updated Prediction
Prediction
0
1980 2000 2020 2040 2060 2080 2100 2120

Total Expenditure Elderly (65+) Young (<65)

Total (old predictions) Elderly (old prediction) Young (old predictions)

In Case 3, per capita expenditures (both on the young and the elderly) are proportional to
GDP per worker

Figure 9.9 Japan 2020–2100: government expenses as a share of GDP by

population group (Case 3)
260 Mark Bamba and David E. Weinstein

some extent as the remaining effects from these temporary shocks taper
off and Abe’s government aims to balance its budget in the coming
years. The second observation is that expenditures on the elderly – the
part of government expenditures that is much less sensitive to economic
fluctuations – has undershot the 2003 predictions in all three cases. This
is particularly prominent in Case 3, where the actual expenditure
numbers were almost 3 percentage points of GDP lower than what
was predicted (a difference of 20 percent from the actual numbers),
offsetting the recession-induced higher expenditures on the young. This
slower-than-predicted rise in elderly expenditures can in part be attrib-
uted to the cost controls on elderly expenditures that were discussed
earlier. However, the pronounced discrepancy for Case 3 also highlights
a point already made – that the assumption of per capita expenditures
on the elderly rising with per worker GDP also contributed to the
significant overestimate in elderly expenditures.
Our updated expenditure forecasts (to the right of the vertical bar)
reflect these two observations. First, we predict a higher per capita
expenditure on the young (as a share of GDP) than we did in our
2003 results, but below what we have seen following the financial
crises.6 This assumption is why, for all three plots, the dotted dark
line (the updated predictions on young expenditures) immediately
after the vertical bar is above the dotted light line (the 2003 predic-
tions on young expenditures) but still below the historical expend-
iture levels immediately before the red bar. In other words, we
predict that, all else equal (i.e., if the population stayed the same),
non-elderly government expenditures should fall to a level between
what we have today and what we saw before the recession. Second,
we lower our predictions for per capita expenditures on the elderly
(as a share of GDP) to reflect the lower-than-expected expenditure
levels we have seen over the past few years.7 This is the primary
reason our updated forecasts for elderly expenditure lie slightly
below the 2003 predictions for the whole forecasting duration in
all three plots. It is worth noting that, while these modifications shift
the expenditure curves upward or downward (by altering their
respective proportionality constants), this does not change the fun-
damental assumptions. In Cases 1 and 2, we still predict per capita
expenditure on the elderly to rise proportionally with GDP, an
assumption that has matched the actual data reasonably well in the
past; and in Case 3, we expect per capita expenditures to rise
The Crisis That Wasn’t 261

proportionally to GDP per worker, an assumption that has signifi-

cantly overestimated elderly expenditures.
This would suggest there is reason to be optimistic about Japan’s
sustainability. In both Cases 1 and 2, we find that Japan’s tax rate
is already at a sustainable level. And for the one scenario that
indicates Japan’s current tax rate is not sustainable (Case 3), we
find that the underlying assumption led to significant overestimates
of elderly expenditures, suggesting that it may not accurately reflect
the ability of the Japanese government to rein in expenses.

9.4 Conclusion
This paper reexamines Japan’s fiscal and monetary policy over the past
two decades and finds that the Japanese government has been remark-
ably adept at restraining benefit growth, raising taxes, and monetizing
debt without generating inflation. Japan still needs to continue to
balance the generosity of its social pension program with future tax
increases, but the simplest interpretation of why markets have not
responded to studies showing Japanese finances are troubled is that
they have confidence in Japanese policymakers’ ability to work out
good compromises. If this approach continues, Japan may very well
avoid either a financial crisis or a major inflationary episode.

Notes
1. See, for example, Asher and Dugger (2000), Ihori, Doi, and Kondo (2001),
Dekle (2002), Kotlikoff (2004), Doi, Hoshi, and Okimoto (2011),
Imrohoroglu and Sudo (2011), Hoshi and Ito (2014), and Braun and
Joines (2015).
2. For example, Asher and Dugger (2000) conclude that “Japan’s
government debt and deficit are on an explosive and unsustainable
trajectory.” Similarly, Kotlikoff (2004) writes, “There seems to be
ample potential for inflation to take off at any moment in Japan and
with it, interest rates.” Likewise, Fukao (2003) writes, “In my view, the
current deflation in Japan can be regarded as a negative bubble; people
are shifting assets from stock and real estates to cash, deposits, and
government bonds. They are blindly buying government-backed
financial assets even though the credit worthiness of the government is
rapidly deteriorating. This negative bubble is clearly unsustainable.”
262 Mark Bamba and David E. Weinstein

3. See Economic and Statistical Research Institute, www.esri.cao.go.jp/en/

sna/data/sokuhou/files/2019/qe193_2/gdemenuea.html.
4. OECD (2019), “Labour: Hourly earnings (Edition 2018),” Main
Economic Indicators (database), https://doi-org.ezproxy.cul.columbia.e
du/10.1787/1a63d074-en (accessed December 24, 2019); OECD
(2019), “Average annual wages,” OECD Employment and Labour
Market Statistics (database), https://doi-org.ezproxy.cul.columbia.edu/
10.1787/data-00571-en (accessed December 24, 2019).
5. In addition to accounting revisions, the most recent System of National
Accounts (SNA) time series also spans a shorter time period. While the
2003 results used data from 1980 through 2000, the most recent SNA
does not have data before 1994. The time period over which we take
averages of expenditures has therefore been modified, so that for
expenditures on the young we take the average from 1994 to 2000
(the maximum length, given the data that we have), while for
expenditures on the elderly we take the average from 1998 to 2000
(we use a three-year time span for elderly expenditures so that we are
consistent with our updated results, which will be discussed later).
6. We make this prediction based on the historical average of government
expenditures on the young between 2000 and 2015. We use a sixteen-year
time range (including both the pre- and post-crisis years) so that we can
average out the volatility in spending as a result of economic fluctuations.
7. More specifically, we estimate this value by taking the average per capita
expenditure on the elderly over the previous three years (i.e., 2015 to 2017).
8. For the 2003 replications, we take the average from 1994 to 2000.
9. Again, the average is taken for 1994–2000 for the 2003 replications.
10. For the 2003 replications, we take the average over the years
1998–2000.
11. Again, the average is taken for 1998–2000 for the 2003 replications.

References
Asher, David Louis, and Robert H Dugger. 2000. Could Japan’s Financial
Mount Fuji Blow Its Top? Cambridge, MA: MIT International Science
and Technology Initiatives, Center for International Studies,
Massachusetts Institute of Technology.
Auerbach, Alan, J. , and Maurice Obstfeld. 2005. “The Case for
Open-Market Purchases in a Liquidity Trap.” American Economic
Review 95(1): 110–137.
Blanchard, Olivier. 1990. “Suggestions for a New Set of Fiscal Indicators.”
OECD Economics Department Working Papers No. 79. doi: http://dx
.doi.org/10.1787/435618162862.
The Crisis That Wasn’t 263

Braun, R. Anton, and Douglas H. Joines. 2015. “The Implications of

a Graying Japan for Government Policy.” Journal of Economic
Dynamics and Control 57: 1–23.
Broda, Christian, and David E Weinstein. 2005. “Happy News from the Dismal
Science: Reassessing the Japanese Fiscal Policy and Sustainability.” In
Reviving Japan’s Economy, edited by Ito Takatoshi, David Weinstein,
and Hugh Patrick, pp.40–78. Cambridge, MA: MIT Press.
Dekle, Robert. 2002. “Population Aging in Japan: Its Impact on Future
Saving, Investment, and Budget Deficits.” Department of Economics,
University of Southern California.
Doi, Takero, Takeo Hoshi, and Tatsuyoshi Okimoto. 2011. “Japanese
Government Debt and Sustainability of Fiscal Policy.” Journal of the
Japanese and International Economies 25(4): 414–433.
Eggertsson, Gauti B., et al. 2003. “Zero Bound on Interest Rates and
Optimal Monetary Policy.” Brookings Papers on Economic Activity 1:
139–233.
Faruqee, Hamid, and Martin Mühleisen. 2001. “Population Aging in Japan:
Demographic Shock and Fiscal Sustainability.” IMF Working Paper
WP/01/40. Research and Asia and Pacific Departments.
Fukao, Mitsuhiro. 2003. “Financial Strains and the Zero Lower Bound: The
Japanese Experience.” BIS Working Papers No 141.
Hoshi, Takeo, and Takatoshi Ito. 2014. “Defying Gravity: Can Japanese
Sovereign Debt Continue to Increase without a Crisis?” Economic
Policy 29(77): 5–44.
Ihori, Toshihiro, Takero Doi, and Hiroki Kondo. 2001. “Japanese Fiscal
Reform: Fiscal Reconstruction and Fiscal Policy.” Japan and the World
Economy 13(4): 351–370.
Imrohoroglu, Selahattin, and Nao Sudo. 2011. “Will a Growth Miracle
Reduce Debt in Japan?” IMES Discussion Paper Series 2011-E-1.
Ito, Takatoshi, and Frederic S. Mishkin. 2005. “Monetary Policy in Japan:
Problems and Solutions.” In Reviving Japan’s Economy, edited by
Ito Takatoshi, David Weinstein, and Patrick Hugh, pp.131–201.
Cambridge, MA: MIT Press.
Kotlikoff, Laurence J. 2004. “Assessing Japan’s Long-Term Fiscal Health.”
US Department of State Bureau of Intelligence and Research.
Morgan, Peter. 2004. “Japan’s public pension reforms.” HSBC Securities
(Japan) Limited Analyst Report. February 18.
Takayama, Noriyuki. 2004. “The Balance Sheet of Social Security Pensions
in Japan.” Hitotsubashi University Repository.
Takayama, Noriyuki, Yukinobu Kitamura, and Hiroshi Yoshida. 1999.
“Generational Accounting in Japan.” In Generational Accounting
around the World, pp.447–470. Chicago, IL: University of Chicago Press.
264 Mark Bamba and David E. Weinstein

Technical Appendix

In this section, we detail our data sources and methodology.

Data
1. The Cabinet Office’s Annual Report on National Accounts for 2017 is
the source of our data for historical expenditures, government revenues,
interest paid on debt, transfers made to the elderly, and GDP. This is
available (at the time of this writing) here: www.esri.cao.go.jp/en/sna/
data/kakuhou/files/2017/2017annual_report_e.html.
The Excel file with GDP can be obtained under the specification of
“Fiscal Year” at “Current Prices” in section “IV. Main Time Series,”
under “1. Gross Domestic Product (Expenditure Approach).” In the
spreadsheet, row 48 (called “5. Gross domestic product (expenditure
approach) (1+2+3+4)”) is used.
The Excel file containing expenditure, revenue, and interest data can
be obtained under the “V. Supporting Tables” section under a link
called “(6–2) Account classified by the Sub-sectors of General
Government (GFS)”. In the spreadsheet, we use row 9 (called “1
Revenue”) for government revenues, row 85 (called “2 Expense”) for
expenditures, and row 95 (called “24 Interest”) for the government’s
interest expenses (the interest expense is used to create Figure 9.1 and
also to calculate the non-elderly expenditures, which is defined as total
expenditures minus elderly expenditures and interest expenses. This
spreadsheet is also used to create Figure 9.3 using row 12 (“1111
Payable by individuals”), row 13 (“1112 Payable by corporations
and other enterprises”), row 16 (“113 Taxes on property”), row 24
(“11411 Value-added taxes”), and row 45 (“12 Social contributions”).
The Excel file containing data on elderly expenditures can be
obtained from the link “(9) Transfers from General Government to
Households (Social Security Transfers)” from the Supporting Tables.
From the spreadsheet, we use row 11 (“a. Public pension (excluding
children allowances)(*2)”), row 23 (“(3) New medical care system for
The Crisis That Wasn’t 265

the Elderly”), row 38 (“(9) Long-term care insurance”), row 20 (“(b)

Pensions” – these are pensions under Seamen’s insurance) and rows 27,
30, 33 (all are called “(b) Long-term”), which correspond to each of the
long-term expenditures from the three mutual benefit association line
items (“a. Federation of national public personnel mutual aid
associations,” “b. Pension fund association for local government
officials,” and “c. Others”).
2. Future population estimates are taken from Japan’s NIPSSR.
Future population predictions are available (at the time of this
writing) here: www.stat.go.jp/english/data/nenkan/68nenkan/1431–
02.html using the link “Future Population.” We use the “65 and
older” columns for the elderly population, the “15–64” column for
the working-age population, and the sum of the “15–64” column and
the “0–14” column for the young population.
3. Historical population numbers are taken from Japan’s Statistics
bureau using their E-Stat portal: www.e-stat.go.jp/SG1/estat/ListE.do?
bid=000001039703&cycode=0 (in case this link becomes unavailable,
the site is also accessible through the link called “Time Series
2000–2015” from the Statistics Bureau website: www.stat.go.jp/eng
lish/data/jinsui/2.htm).
The population time-series called “Population by Age (Five-Year
Groups and 3 Groups) and Sex (as of October 1 of Each Year) – Total
population, Japanese population (from 2000 to 2015)” is used. Historical
population numbers for the young (below 65 years old), elderly (65 and
older), and working (15–64) are available in this workbook for the years
2000–2015. For 2016 and 2017 (the years that follow the most recent
population census), we use data from the 2017 annual report under the
table title: “Population and Percentage distribution by Age (5-Year Age
Group) and Sex – Total population, October 1, Each Year.”
4. Historical debt level data comes from BOJ’s Flow of Funds.
Historical levels for each fiscal year can be found on the BOJ’s time-
series data search tool found here: www.stat-search.boj.or.jp/index_en
.html.
Historical balance sheet data is available on this website using the
“Flow of Funds” link (available here, at the time of this writing: www
.stat-search.boj.or.jp/ssi/cgi-bin/famecgi2?cgi=$nme_a000_en&lstSel
ection=FF). For gross debt, we sum together all categories in the
General Government’s liabilities except for “Equity and investment
fund shares” and “Financial derivatives and employee stock
266 Mark Bamba and David E. Weinstein

options.” For the government’s assets (for calculating BOJ-inclusive

net debt), we sum all categories in the General Government’s assets. For
the BOJ’s net worth (for calculating BOJ-inclusive net debt), we sum all
of the central bank’s asset categories except for “Others” and subtract
the liabilities the BOJ has in the “Loans” and “Currency and deposits”
categories. For the BOJ’s net assets (for calculating the privately held
net debt), we sum all of the central bank’s asset categories except for
“Currency and deposits” and “Others” and subtract the liability the
BOJ has in the “Loans” category.
These Flow of Funds statistics and categories can also be seen in more
reader-friendly Excel tables in BoJ’s annual Flow of Fund releases
that can be found here: www.boj.or.jp/en/statistics/sj/index.htm/.
However, these only provide the statistics for one year, so we still use
the time-series data tool to obtain the full balance-sheet time series.

Method
As discussed briefly in the main text, an important step in projecting
future expenditures is the determination of the proportionality
constants for each of the assumptions. For expenditures on the
young, we take three approaches. For Case 3, we compute the young
expenditure (as a share of GDP) multiplied by the working population
divided by the population of the young for each year, and take the
average of this value over the period from 2000 to 2017.8 To project
the future expenditure on the young, we would multiply this average by
the projected young population and divide by the projected working
population. For Cases 1 and 2, we first compute the average young
expenditure (as a share of GDP) divided by the young population and
take the average of this value for the period from 2000 to 2017 (this
would be the average per capita expenditure on the young as a share of
GDP).9 For Case 2, we then multiply this average by the projected
population of the young. For Case 1 (where young expenditure is
a constant share of GDP) we simply take the first projection of the
young expenditure level as a share of GDP (i.e., for FY2018) for Case 2
and then assume that this expenditure level (as a share of GDP) remains
constant. For expenditures on the elderly, we compute similar
averages. For Case 3, we compute the elderly expenditure (as a share
of GDP) multiplied by the working population divided by the elderly
population for each year and take the average of this value over the
The Crisis That Wasn’t 267

three-year period from 2015 to 2017.10 To project the future

expenditure on the elderly, we multiply this average by the projected
elderly population and divide by the projected working population. For
Cases 1 and 2, we compute the elderly expenditure (as a share of GDP)
divided by the elderly population and take the average of this value for
the years 2015 through 2017.11 For both Cases 1 and 2, we then
multiply this average by the projected elderly population for
a particular year.
part iv

Third Arrow of Abenomics

10 Abe’s Slight Left Turn
How a Labor Shortage Transformed
Politics and Policy
st e ven k. vo g el

10.1 Introduction
As Japan shifted from labor surplus to labor shortage, the government
and industry became less concerned about how to shed excess workers
and more concerned about how to increase the workforce and to
recruit and retain workers.1 The second administration of Prime
Minister Shinzo Abe that took power in 2012 changed course from
reducing labor costs by maximizing employer flexibility to boosting
labor participation and productivity by improving employee welfare.
The earlier reforms had helped firms to restructure in the 1990s and
2000s, but they took a toll over time: undermining employment secur-
ity, increasing inequality, suppressing wages, and slowing economic
growth. The new turn in labor reform delivered results: an increase in
female and elderly workforce participation plus some improvement in
working conditions and greater flexibility in work styles. If the govern-
ment and industry follow through with these initiatives, they could
deliver “win-win” solutions that promote both economic growth and
worker welfare. This chapter explains how demographic trends, polit-
ical dynamics, and corporate strategies combined to reshape Japan’s
labor market regime, and concludes with an assessment of the pro-
spects for positive change.
Japan is at the forefront of a tectonic demographic shift common
among advanced industrial countries, as its population growth has
declined and now reversed. Japan’s population growth gradually slowed
over the postwar period, reaching a peak of 128 million in 2010 and
declining since. The decline is expected to accelerate in the years to come,
even if fertility increases (Figure 10.1).2 Government and industry lead-
ers have been aware for several decades that demographic trends were
likely to generate labor shortages, but economic stagnation prolonged
the labor surpluses of the 1990s and delayed the tightening of the labor

271
272 Steven K. Vogel

(Millions)

140
High Fertility
130 Medium Fertility
120 Low Fertility

110

100

90

80
Actual Projected
70

60
00 10 20 30 40 50 60
20 20 20 20 20 20 20

Figure 10.1 Japan’s population: actual and projected, 2000–2065

Source: National Institute of Population and Social Security Research (IPSS),
based on data from Population Census of Japan and the Vital Statistics of
Japan; projections calculated by IPSS

market. By the 2010s, however, the trend was unmistakable, as

unemployment dipped to 2.4 percent and the job offers to applicants
ratio rose to 1.6:1 by 2018 (see Figures 10.2 and 10.3).
Nonetheless, Japan’s labor policy did not transition overnight.
Rather, the government continuously combined employer-friendly pol-
icies with more worker-friendly ones, gradually shifting the emphasis
from the former to the latter.3 The first Abe administration (2006–
2007) and the Democratic Party of Japan (Minshuto) government
(2009–2012) began the turn toward more worker-friendly policies,
but this new orientation only really took hold with the second Abe
administration’s “work style reform” (hatarakikata kaikaku)
announced in 2015.
The second Abe administration represented a peculiar hybrid, neither
neoliberal nor social-democratic, conservative in outlook yet progressive
in the primary direction of its labor reform. The political parties of the
left have long called for raising the status of women and non-regular
workers, improving childcare and eldercare policies, curbing excessive
overtime hours, and lifting wages. Yet the conservatives under Abe
Abe’s Slight Left Turn 273

6.0
5.0
4.0
3.0
2.0
1.0
0.0
1995

1997

1999

2001

2003

2005

2007

2009

2011

2013

2015

2017
Figure 10.2 The unemployment rate in Japan, 1995–2018
Source: OECD.Stat

1.8

1.6

1.4

1.2

1.0

0.8

0.6

0.4

0.2

0.0
1995 2000 2005 2010 2015 2019

Figure 10.3 Job offers to applicants ratio in Japan, 1995–2019

Sources: Ministry of Internal Affairs and Communications, Japan Macro
Advisors

embraced a similar mix of policies, albeit for different reasons. They

sought to bolster the birth rate by improving work-life balance, to
increase the labor force and raise productivity by improving working
conditions, and to boost GDP by increasing disposable income. They
274 Steven K. Vogel

also hoped to broaden their political support and to take away

a potential wedge issue from the opposition parties.
This chapter traces the evolution of policy reforms and business
practices from the 1990s with special attention to the “Abenomics”
era since 2012. The shift from labor surplus to shortage was the key
driver of change, but industry policy preferences and the policy process
help to explain the specific substance of reforms. Corporate strategies
reflect Japan’s distinctive market institutions, and these strategies in
turn shape corporate preferences for policy change. And those prefer-
ences are filtered through a policy process that has gradually transi-
tioned from corporatist accommodation toward a more top-down
model. Yet this transition has been far from complete, leaving many
elements of corporatist bargaining in labor policy formation.
Japanese company strategies reflect their specific “comparative insti-
tutional advantages,” including strong government-industry ties and
long-term collaborative relationships with their business partners
(interfirm networks, or keiretsu), their banks (the main bank system),
and their workers (“lifetime” employment).4 In Albert Hirschman’s
terms, these long-term relationships generate a propensity toward
“voice” (negotiation) over “exit” (severing ties) with these partners
(Hirschman 1970). This perspective then generates a fairly specific set
of predictions about how Japanese firms will respond to hard times.
They will squeeze their suppliers but not abandon them; press their
main banks for better terms but not pull away from them; and negotiate
cost-saving measures with their workers but not lay them off. And they
will leverage these relationships to survive the downturn. So when the
economy encountered a financial crisis in the 1990s, firms responded
with this distinctive pattern of adjustment.5
Japanese companies have favored long-term employment because
they benefit from employee loyalty, collaborative labor-management
relations, and the ability to invest in training workers in firm-specific
skills. Yet they have also counted on the flexibility of a dual labor
market, with a large tier of “non-regular” workers (hiseiki rodosha),
that permits them to control wage costs and to adjust more quickly to
shifts in demand. The non-regular workers often work full time but
receive lower wages, fewer benefits, and less job security. In the 1990s,
firms reduced new recruits, declined to renew contracts of non-regular
workers, negotiated wage restraint, and deployed a wide range of
measures to lower labor costs without laying off regular workers
Abe’s Slight Left Turn 275

(seiki rodosha or shain). And over time they increased the non-regular
share of the workforce, meaning that the long-term employment sys-
tem survived but covered a smaller share of the workforce. In short,
firms adapted to new conditions with incremental adjustments while
striving to preserve valued institutions.
Firms’ business strategies in turn shaped their preferences for policy
reforms. In the 1990s, they asked the government to give them more
flexibility in deploying labor. They wanted more leeway with working
hours and conditions, and they wanted to extend their non-regular
worker safety valve with a new tier of even more temporary workers:
agency temps. But they did not press for greater freedom to hire and fire
core workers. Japan Business Federation (keizai dantai rengokai, or
Keidanren for short) Vice Chairman Yoshio Nakamura puts it this
way: “We support voluntary reductions but not forced layoffs.
Companies would not resort to mass layoffs even if they had the
freedom to do so, because Japanese managers have a strong commit-
ment to employment stability.”6 Now that firms face a tighter labor
market, their corporate strategies have shifted, and this has in turn
altered their policy preferences. They want policies that will help them
to bring more people into the workforce and to raise productivity.
Industry preferences for labor market reform were mediated
through a distinctive political process. Scholars have appropriately
characterized Japan’s postwar political economy as one of “corpor-
atism without labor” (Pempel & Tsunekawa 1979; Wilensky 2002).
By this they mean that the business sector enjoyed systematic repre-
sentation in policy deliberations, via peak associations such as
Keidanren and/or sectoral industry associations, but that labor did
not share the same privileged status. Industry associations benefited
from direct representation in many government policy organs, such
as policy councils (shingikai), and a special relationship with the
ruling Liberal Democratic Party (LDP). This contrasts with Western
European corporatism, under which business and labor shared more
equal status.
But Japanese labor policy is a partial exception to the corporatism
without labor pattern, reflecting something closer to corporatism with
labor. That is, labor policy formulation runs through the Ministry of
Health, Labor, and Welfare’s (MHLW) Labor Policy Council (rodo
seisaku shingikai). And the council adheres to International Labor
Organization (ILO) guidelines for tripartite deliberations. The council
276 Steven K. Vogel

and its subcommittees have equal numbers of business, labor, and

public interest members (mostly academics). Prior to the 2000s, the
ministry would typically ask an academic study group (kenkyukai) to
examine a particular topic and issue a report. The Labor Policy Council
would deliberate on a proposal, which the ministry would develop into
a draft law. The neutral members on the council would serve as inter-
mediaries between industry and labor. Then the council would pass the
compromise proposal to the ministry officials, who would consult the
key LDP labor policy experts (rodo-zoku). “The ministry officials are
involved in the whole process, but they have to operate behind the
curtain,” explains Kazuo Sugeno, a long-standing member of the coun-
cil and its chair from 2005 to 2009. “They cannot tell the politicians
what to do. So they have the neutral council members tell the politicians
what is theoretically reasonable and politically feasible.”7 Labor policy
formation reflects corporatism with labor in the sense that the MHLW
officials tend to be more sympathetic with the unions than officials
elsewhere in the government. This means that labor has better repre-
sentation on labor-specific issues than on other policy items, and that
political maneuvering sometimes takes the form of shifting the agenda
into or out of the council, toward or away from the ministry.
In the 1990s and into the early 2000s, political leaders were focused
primarily on economic recovery, and they viewed reforms that would
facilitate corporate restructuring – including the labor reforms detailed
in the following section – as critical to that strategy. So industry
preferences for policy reforms were the primary driver of the substance
of reforms. But the second Abe administration officials had a different
understanding of the relationship between labor market reform and
economic performance, which was strongly influenced by the demo-
graphic and labor market trends already noted. They wanted to
improve working conditions more than to help firms cut costs, because
this would expand the workforce and raise productivity. They wanted
to raise the status of women and non-regular workers more than to
make it easier for employers to hire non-regular workers, because this
would make it easier for women to combine work and childrearing.
And they wanted to raise wages because this could lift consumer
demand. The policy process differed in this period as well, with Prime
Minister Abe and the Cabinet Office playing a stronger role in leading
policy change.
Abe’s Slight Left Turn 277

The first phase of policy reforms and corporate adjustments, roughly

from 1995 through 2012, succeeded in the short run yet failed in the long
run. At the micro level, companies made reasonable judgments to reduce
costs without threatening the core logic of an employment system that had
served them well, including job security for their core workers. Over the
longer term, however, these adjustments cumulated to take a toll at the
macro level. They suppressed wage growth to the point where this under-
mined demand and compromised overall economic performance. And
they increased inequality between regular and non-regular workers,
between men and women, and between older and younger workers.
The new direction in policy, beginning in 2006 but accelerating since
2012, has the potential to leverage the new market environment to
deliver reforms that will strengthen the economy as well as improve
work life. Yet there is no guarantee that the government and industry
will pull off this win-win solution.

10.2 Government Reforms and Corporate Practices,

1995–2012
As Japan succumbed to a financial crisis in the early 1990s, government
and industry leaders began to criticize the very institutions that had
been heralded for the postwar economic miracle: a powerful central
bureaucracy, collaborative interfirm networks, and the “lifetime”
employment system. And some of them called for a drastic shift toward
the US liberal market model. In practice, however, as stressed in the
previous section, most Japanese firms favored incremental reforms to
allow them more flexibility in reducing labor costs rather than more
fundamental reforms that would jeopardize valued institutions, includ-
ing close management–labor collaboration based on stable employ-
ment. Japan Federation of Employers’ Organizations (nihon keieisha
dantai renmei, or Nikkeiren for short) leaders were particularly cau-
tious with their proposals, stressing the need to protect employment
and to preserve worker participation in management. But they did
propose reforms in a widely publicized report in 1995, calling for
increased freedom in hiring non-regular workers, greater flexibility in
setting work hours, and more performance-based wages.8
The Japanese government reformed labor regulation accordingly. It
enhanced firms’ flexibility to employ non-regular workers and pro-
vided them more options for labor adjustment short of layoffs, yet it
278 Steven K. Vogel

preserved long-term employment for core workers. Specifically, it eased

restrictions on private employment agencies; opened up a new category
of non-regular workers (agency temps); and loosened restrictions on
working hours and other working conditions. But it did not make it any
easier for firms to dismiss regular workers.
Yet, even to proceed this far, the government had to bypass the stand-
ard policy process centered on the Labor Policy Council. The government
designed the Deregulation Subcommittee of the Administrative Reform
Committee (gyosei kaikaku iinkai kisei kanwa shoiinkai) to promote
liberal market reforms by transcending the usual consensual policy pro-
cess. The government stacked the committee with advocates of reform
and excluded opponents, structuring it to report directly to the prime
minister rather than to a particular ministry. The committee turned to
labor issues in 1995, and its recommendations prompted a series of
reforms. Some committee members felt that regulatory reform in other
sectors was likely to cause labor dislocation, so they wanted to reform
labor markets to facilitate the reallocation of workers. Some wanted to
open up new business opportunities in the employment matching and
temp agency sectors. The opposition parties and the unions viewed the
committee’s intrusion into labor issues as a blatant violation of the
postwar norms of labor policy formation. They pressed for implementa-
tion plans to run through the Labor Policy Council, even if the initial
recommendations came from outside bodies. And they adopted more
outsider political strategies, including public protests, direct challenges
in the Diet, and vocal opposition in election campaigns.
In 1997, the government amended the Equal Employment
Opportunity Law (danjo koyo kikai kintoho) to remove some special
protections for female workers, such as those governing overtime and
nighttime work. In 1998, it revised the Labor Standards Law (rodo
kijunho) to give employers more flexibility with work hour rules. The
Japan Trade Union Confederation (nihon rodokumiai sorengokai, or
Rengo for short) succeeded in imposing conditions, however, such as
requiring worker consent, that made it difficult for employers to exer-
cise this flexibility.9 In 1999, the government permitted private com-
panies to provide employment placement services and allowed
employers greater freedom in hiring agency temps. Prior to this, the
government had provided placement services but had not allowed
private companies to do so, fearing that these companies would exploit
workers. The agency temps gave employers a category of workers
Abe’s Slight Left Turn 279

outside the existing tier of non-regular employees whom they could

deploy for shorter periods of time with lower levels of benefits and job
security. The government made some modest adjustments in welfare
provisions for unemployed workers, but it did not substantially bolster
the social safety net. It continued to rely more on government policies
and private-sector practices that maintain employment (job protection)
than on policies to support those who lose their jobs (worker
protection).
The first Abe administration began the shift toward more worker-
friendly reforms, partly due to the backlash from Prime Minister
Junichiro Koizumi’s economic reforms. By 2005, the media were
sounding the alarm over increasing economic inequality in Japan, and
some critics were blaming Koizumi’s reforms, especially labor market
reforms such as the liberalization of agency temps (see Figure 10.4).
The labor market was not the only factor driving inequality, but it did
contribute by expanding the gaps between those with jobs and those
without and between those with stable jobs and those in precarious
employment.
The government also revised the Part-Time Labor Law to establish the
principle of balanced treatment for regular and non-regular workers, but
it left some room for “legitimate” discrimination in a distinctively
Japanese way. In the Japanese context, the two categories of workers
are distinguished not only by job duties but also by their levels of
“constraint” (kosokusei). That is, the (mostly male) regular workers
are expected to comply with employer demands for overtime work or

0.55

0.5

0.45

0.4

0.35

0.3
1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015

Figure 10.4 The rise of economic inequality in Japan, 1995–2015 (Gini

coefficient, before taxes and transfers)
Source: OECD.Stat
280 Steven K. Vogel

transfers, whereas the (mostly female) non-regular workers are not.10

The bill viewed non-regular workers as equivalent to regular workers if
they have the same responsibilities, indefinite work contracts, and terms
for changes in assignments. In practice, this meant that very few non-
regular workers were judged to be equivalent to regular workers – even
if they were performing similar duties – and thus they would not be
entitled to receive equal pay and benefits (Miura 2012, pp.76–78).
The DPJ took power in August 2009 on a platform that centered on
changing politics by throwing out the long-ruling LDP and wielding
power over the bureaucracy, not on an alternative economic paradigm.
The party issued a campaign manifesto with some specific policy pro-
posals, such as parental allowances, but without an overall framework
for economic policy. The new administration was slow to come up with
a growth strategy once it took power, issuing an outline in
December 2009 and then a more detailed plan in June 2010. The DPJ
was reliant on Rengo for political support, yet it was careful not to
appear as a pawn of labor interests. The administration promised some
measures to improve working life and promote stable employment, but
it focused particularly on the agency temp business. The DPJ had
cosponsored a proposal in May 2009, before taking power, that
would have prohibited agency temps in manufacturing sectors and
eliminated one type of temp work (toroku-gata), in which workers
register for individual contracts that only pay when a job is available.
The government submitted the bill to the Diet in March 2010, but it
dropped both of these provisions as a concession to the LDP and the
Komeito Party in November 2011. The bill that eventually passed in
March 2012 prohibited the hiring of agency temps on contracts of less
than thirty days and set guidelines for improving working conditions
for agency temps.
Japanese firms took advantage of the policy reforms enacted after
1995 to reduce costs and enhance flexibility while avoiding layoffs.
Their first impulse to cope with labor costs was in fact not labor
adjustment at all but rather to increase sales or reduce administrative
costs.11 And when they turned to labor adjustment, they favored
moderate measures such as limiting overtime, reducing bonuses, and
not renewing the contracts of non-regular workers (Table 10.1). Their
core strategy was to negotiate wage restraint and reduce the workforce
by attrition. They also deployed a distinctively Japanese tactic of
Abe’s Slight Left Turn 281

Table 10.1 Types of labor adjustment in Japan, 1993–2018 (percentage)

1993 2000 2005 2010 2015 2018 2019

Limit overtime 26 11 5 17 10 13 18
Increase holidays and vacation 6 2 2 8 10 13 16
Non-renewal of contracts 7 2 1 3 1 1 1
Reduce mid-career recruiting 18 4 1 6 1 1 2
Permanent transfers 11 7 4 12 13 15 17
Temporary transfers 8 5 2 5 7 7 8
Temporary layoffs 4 1 0 4 1 0 1
Voluntary retirement 2 2 1 1 1 1 1
No labor adjustment implemented 62 79 89 66 72 70 64

Source: Ministry of Health, Labor, and Welfare, Survey on Labor Economy Trends

transferring workers to affiliated firms, essentially deploying their cor-

porate groups as employment networks.
Companies were extremely reluctant to lay off workers. A Japan
Institute of Labor survey found that respondents were concerned
that workforce reductions would undermine morale (51.5 percent),
lead more employees to quit (17.2 percent), reduce productivity
(14.3 percent), or make it more difficult to recruit new employees
(11.3 percent) (Japan Institute of Labor 2002). Gregory Jackson
finds that Japanese firms were far less likely than US firms to have
workforce reductions greater than 10 percent in 1991 (5.9 percent
of Japanese firms versus 20.6 percent of US firms) and 2001
(2.0 percent versus 9.2 percent) (Jackson 2005). Japanese job
tenure actually increased during hard times, from 10.9 years in
1990 to 12.0 years in 2005 and 13.5 years in 2017, while US job
tenure decreased in good times.12
Over time, companies increased the non-regular share of the work-
force by hiring more non-regular workers, including agency temps,
rather than regular workers. The share of non-regular workers in the
workforce jumped from 20.3 percent in 1994 to 24.9 percent in 1999,
and 31.4 percent in 2004, and then continued to rise more slowly to
37.9 percent in 2018 (Figure 10.5).13 Corporations also diverged
more from each other in strategy after the 1990s: those firms that
benefited the most from long-term employment, such as
282 Steven K. Vogel

45%
40%
35%
30%
25%
20%
15%
10%
5%
0%
1988

1991

1994

1997

2000

2003

2006

2009

2012

2015

2018
Figure 10.5 The increase in non-regular workers in Japan, 1988–2019
(percentage)
Source: Ministry of Internal Affairs and Communications, Labor Force Survey

manufacturers, adhered to it; while those that benefited the least, such
as retailers, abandoned it.14
Japanese industry avoided major labor dislocation by favoring mod-
est policy reforms and gradual labor adjustment strategies, yet these
measures took a toll over time.15 Many firms reduced employment via
attrition rather than layoffs, but this created a generation gap between
older workers with permanent positions and younger ones with more
precarious employment. Likewise, firms increased the share of tempor-
ary workers rather than abandoning core workers, but this undermined
employment security and productivity growth overall.16 Over the long
term, corporate cost-cutting reduced employee welfare, strained labor–
management relations, and weakened macroeconomic performance.

10.3 Government Policy 2012–2015: Abenomics

Prime Minister Abe and the LDP returned to power in December 2012.
Abe had resigned after only one year in September 2007. His first
tenure was plagued by recurring scandals, gaffes, and low public sup-
port ratings. He was also physically exhausted, suffering from ulcera-
tive colitis. Most observers assumed that was the end of the line for Abe
as a force in Japanese politics. But he retained some loyal supporters
within the party, who eventually helped him to mount a comeback. He
was considered an underdog in the September 2012 election for party
president, with former defense minister Shigeru Ishiba enjoying greater
Abe’s Slight Left Turn 283

popularity. Ishiba won more votes from party members at large, but
Abe prevailed in a runoff on the strength of his support among LDP
Diet members.17
The DPJ was vulnerable in the December 2012 lower house election
due to the widespread impression that it had failed to govern effect-
ively. It rattled USA–Japan relations by threatening to pull out of an
agreement to relocate a US base in Okinawa, and it floundered in
advancing its domestic policy agenda, partly because it waged war on
the bureaucracy. It was blamed for missteps after the Fukushima
earthquake, tsunami, and nuclear meltdown triple disaster of
March 2011. Most important, however, it failed to effectively address
the issue of greatest concern to the Japanese public: the economy. The
LDP won a resounding victory, making Abe prime minister, and the
new administration quickly focused on economic policy under the
slogan of “Abenomics.” The program had three “arrows”: monetary
stimulus, fiscal stimulus, and structural reform. The first two arrows
paid off in the short term, with a weaker yen, higher exports, a rising
stock market, and stronger economic growth.
Abe’s team was also concerned that wages were not rising and that
this was hampering consumer demand. Abe himself prodded industry
to increase wages. At the same time, the administration felt that the
government had been spending too much money on job preservation
subsidies and unemployment insurance. It sought to shift funds toward
policies that would support labor mobility, such as vocational training,
professional accreditation, and career counseling.
The administration referred to the third arrow as its growth strategy,
including agricultural reform, pharmaceutical market reform, corpor-
ate governance reform, deregulation zones in major cities – and labor
market reform. It unveiled its growth strategy in June 2013 (Prime
Minister’s Office 2016). Administration officials argued that corporate
performance would have to improve to generate more employment.
They would support the reallocation of employment into new areas,
especially the service sector. But they made it clear that they would not
liberalize immigration, except for highly skilled workers. They began
with a mix of neoliberal reforms designed to give employers more
flexibility, on the one hand, and measures to strengthen worker protec-
tion, on the other. They centered deliberations on the growth strategy
in the Industrial Competitiveness Council (sangyo kyosoryoku kaigi).
Yasuchika Hasegawa, chairman and CEO of Takeda Pharmaceuticals
284 Steven K. Vogel

and chairman of the Japan Association of Corporate Executives (keizai

doyukai), was assigned the role of promoting labor market reforms
within the Council. “We foresaw that the labor market was changing,
and we would have to prepare reforms for an era in which labor would
be more scarce,” recalls Hasegawa. “Japan might serve as an example
for the world.”18
The Council revisited two proposals that had been considered but
not adopted during the first Abe administration: a white-collar exemp-
tion for working hour restrictions; and a proposal to allow monetary
compensation to settle disputes regarding unfair dismissals. The stand-
ard workweek had been set at forty hours in 1987 (effective in 1994),
and employers were required to pay an overtime premium beyond that.
The government created a discretionary work hour system for specific
classes of workers, less than 2 percent of total workers, in the
same year. This meant that most white-collar workers were still eligible
for overtime pay on an hourly basis (Koshiro 2017). The administra-
tion submitted the white-collar exemption revision to the Diet in
April 2015 but then abandoned the effort due to strong opposition.
The government also announced that it would create more flexible
labor rules in six special economic zones in large cities, but MHLW
officials successfully fought off the proposal, arguing that the country
could not have different labor laws for different regions. Instead, they
proposed to set up special employment assistance offices for foreign
companies and to enact more modest amendments to the Temp Agency
Law that would apply nationwide.
The clearest sign of the shift in priorities – from promoting employer
flexibility to enhancing working life – came in the form of
“Womenomics,” yet another component of the Abenomics
program.19 This initiative included policies to make it easier for
mothers to work outside the home and to encourage companies to
appoint more female executives. As with other initiatives, it was insti-
tutionalized with a bureau in the Cabinet Office and a designated
cabinet member to lead the effort. The Diet passed the Promotion of
Women’s Participation and Advancement in the Workplace Law (josei
katsuyaku suishinho) in 2015, requiring companies with 300 or more
employees to publish data on the number and status of female employ-
ees and to develop action plans for promoting women in the
workforce.20
Abe’s Slight Left Turn 285

Government and industry leaders discussed modifications to the

basic logic of the Japanese employment system, including via
a ministry policy council in 2014 (Ministry of Health, Labor, and
Welfare 2014a). In the arcane lexicon of Japanese labor relations,
these discussions often centered on a debate over skill-based wages
(shokunokyu) versus occupation-based wages (shokumukyu). In prac-
tice, Japan’s “skill-based” system gave considerable weight to senior-
ity, since skill was substantially defined by seniority. So shifting to
occupation-based wages would mean less reward for seniority and
more for specialized expertise. Japanese labor unions resisted this
change because they felt that it would undermine the core values of
the Japanese employment system, including job security. Reform advo-
cates sometimes employed the Japanese-English terms for
a “membership-style” (menbashippu-gata) employment system,
based on the employee’s membership in a corporate community, versus
a “job-style” (jobu-gata) employment system, based on the employee’s
vocation. METI officials were particularly eager to pursue this transi-
tion because they believed that greater labor mobility and diversity in
work styles would help Japan to adapt to the information economy (Ito
2017). Meanwhile, government and industry considered narrowing the
gap between regular and non-regular workers by promoting hybrid
categories, such as “limited regular workers” (gentei shain) who would
be less obligated to accept different job assignments, to relocate, or to
work extra hours (Ministry of Health, Labor, and Welfare 2014b). In
one vision, employees would have a choice between higher wages with
more obligations or lower wages with fewer obligations.
Abenomics took a further turn with the 2015 announcement of
three new arrows. Politically, the new arrows were meant to assure
voters that Abenomics had not lost momentum and to demonstrate
that the government was occupied with issues of public concern.
They also indicated a shift in economic strategy toward policies
that would support a higher birthrate; increase the workforce, espe-
cially among women and the elderly; make workers feel more secure;
and raise wages to increase disposable income. The three new arrows
would be: a strong economy that gives hope; a childcare system that
fosters dreams; and a social security regime that provides reassurance
(Prime Minister’s Office 2015). Unlike the first three arrows, these
came with specific targets, albeit without deadlines: the largest nom-
inal GDP in postwar history of 600 trillion yen, a birthrate of 1.8
286 Steven K. Vogel

children per woman, and no one having to quit a job to care for
family members.
The Abe administration’s slight left turn in labor policy reflected
a classic LDP strategy that had served it well for decades: preempting
the opposition. The LDP embraced environmental protection and social
welfare policies in the early 1970s, when the opposition was threatening
the party’s dominance by advocating those policies. Prime Minister
Koizumi (2001–2006) stole the thunder from the opposition by advo-
cating structural economic reforms, and he then went one step further by
embracing the opposition’s critique of the LDP as clientelistic and cor-
rupt, proclaiming that he would clean it up. The opposition parties and
the labor unions had been advocating more generous family policies and
equality in the workplace for decades – and here was Abe suddenly
saying the same thing. “The LDP leaders are clever,” laments Rengo’s
Yoko Murakami. “They want to increase their support. So they take the
best ideas from the Democratic Party and make them their own.”21

10.4 Government Policy 2015–Present: “Work Style Reform”

The government announced its “Work Style Reform” (hatarakikata
kaikaku) as one element of the three new arrows, including measures to
address some of the core weaknesses of the labor system, such as the
gap between regular and non-regular employees and excessive work
hours. Government officials were concerned that the earlier round of
reforms had increased inequality and provoked a backlash against
reforms. “Given the polarization of Japanese employment into two
groups, regular and non-regular workers,” explains one MHLW offi-
cial, “labor deregulation only exacerbated that gap, dampened
demand, and undermined competitiveness. So we need to moderate
this polarization before proceeding further with market reform.”22
LDP leaders also wanted to shift labor policy in a direction that
would promote economic growth. They were especially concerned
about the size of the labor force, and labor productivity that lagged
behind that of the United States (Figure 10.6). Given the decline in
population, Japan could only grow the economy by increasing the
labor participation rate and/or by raising productivity. Government
officials estimated that, to achieve 2 percent annual GDP growth from
2016 to 2040, Japan would have to raise productivity growth from
0.9 percent in the 1990–2015 period to 2.9 percent or to reverse the
Abe’s Slight Left Turn 287

70

60
50
40
US
30 Japan
20
10
0
2000 2002 2004 2006 2008 2010 2012 2014 2016 2018

Figure 10.6 Labor productivity in Japan and the United States, 2000–2018
(GDP per hour worked, US dollars)
Source: OECD.Stat

projected 0.9 percent annual decrease in the workforce with

a 1.1 percent increase. Since they deemed that they would be unlikely
to reach either of those targets, they concluded that they would need to
make some headway on both (Shiozaki 2017). And higher productivity
and higher wages could support a virtuous cycle by boosting consumer
demand. Government officials were also concerned that precarious
work was reducing fertility. The ministry calculated that a couple’s
average number of children depends heavily on its employment status:
1.90 children for couples with both male and female as regular work-
ers; 1.79 children for a regular worker male and a non-regular worker
female; 1.09 children for a non-regular worker male and a regular
worker female; and 1.36 children for both non-regular workers.23
The Abe administration promoted the Work Style Reform in
a similar manner to other major initiatives, by creating a new promo-
tion office (suishinshitsu) within the Cabinet Office, assigning a cabinet
member with responsibility for the initiative, and launching an advis-
ory council to promote reforms. These were characteristic tactics
within the administration’s broader strategy to transfer power from
the ministries to the Prime Minister’s Office (shusho kantei) and the
Cabinet Office (naikakufu).24 Yet the ministries were hardly powerless
in this new environment. After all, many of Abe’s top advisors in the
Prime Minister’s Office and the Cabinet Office came from the minis-
tries. So the ministry officials exercised their influence through repre-
sentation in these agencies, interaction with other officials in the
288 Steven K. Vogel

various offices and headquarters, and coordination of the advisory

councils. Furthermore, METI had a particularly advantageous pos-
ition, with officials in important spots and strong representation
throughout. In fact, METI official Takaya Imai served continuously
in the key position of executive secretary (hishokan) to the prime
minister after 2012 (Shimizu 2016). “METI has mastered this new
environment,” concedes one MHLW official. “Now they take their
ideas straight to the Cabinet Office, and they coordinate there.”25 In
the case of the work style reform, ironically, the Cabinet Office simply
did not have the staff to handle the workload. So the MHLW took on
much of the detailed work of crafting reform proposals, with a dozen or
so officials moving to the Cabinet Office on a temporary basis.26
The Cabinet Office reform committees under the Abe administration
operated differently from the old ministerial policy councils, which
often balanced the various vested interests in a particular policy
domain. They also differed from the Labor Policy Council subcommit-
tees characterized by open debate between management and labor.
They included the prime minister and top cabinet leaders plus the
very top representatives from industry associations and other organ-
izations. The meetings were shorter and more orchestrated by the staff,
with a predetermined agenda, more concise presentations, and less
open discussion. The Council for the Realization of Work Style
Reform (hatarakikata jitsugen kaigi) met for one year and issued its
report in March 2017. As in the case of the Deregulation Committee,
the opposition parties and the labor unions viewed this as a violation of
the established policy process. But they were less incensed this time
because at least they had one representative on the council, Rengo
Chairman Rikio Kozu. And they hoped that the council would defer
to the Labor Policy Council on the details of implementation.
Meanwhile, the MHLW bureaucrats, who tended to identify with
workers and support labor protections, still exercised considerable
influence over the specifics. Yasuhisa Shiozaki, who was MHLW min-
ister from 2014 through 2017, reports that he met considerable resist-
ance from the career civil servants within his own ministry. Shiozaki is
a former Bank of Japan official known for his expertise on financial
affairs and his strong advocacy of market reforms, especially corporate
governance reforms. Shiozaki recalls that ministry officials insisted that
subsidies to promote labor mobility should go only to companies that
were increasing employment. The officials argued that increasing
Abe’s Slight Left Turn 289

employment was their core mission, written right into Japan’s basic
labor law. In Shiozaki’s view, in contrast, subsidies should go to firms
increasing turnover instead. “I really wanted to rewrite the law to state
that the government should promote job mobility, not job expansion,”
he laments, “but I ran out of time.”27
Shiozaki managed to make one administrative change, following
a recommendation from the Regulatory Reform Promotion Council
(kisei kaikaku suishin kaigi). He created a new Basic Subcommittee
(kihon bukai) within the Labor Policy Council that would take
a broader view of labor policy and would not adhere to standards of
tripartite representation. The subcommittee would have fewer repre-
sentatives of both labor and industry and more public interest repre-
sentatives such as academics. “The ministry officials are probably
plotting how to unravel this right now,” Shiozaki speculates, only
months after leaving the post.28
The work style reform agenda was extremely broad: equalizing
pay for equal work; reducing working hours; enhancing flexibility
in work styles (including telework and multiple jobs); improving
training for women and younger workers; enhancing medical leave;
extending childcare and elder-care provisions; expanding oppor-
tunities for the elderly; bringing in more foreign workers; and
more. In practice, deliberations focused on working hours, equal
pay for equal work, and the white collar exemption for overtime
rules. The working hours issue rose in prominence after a heavily
publicized incident in 2015 in which a young woman at Dentsu,
Japan’s largest advertising agency, committed suicide due to exces-
sive work hours – a particularly jarring example of the phenom-
enon of death by overwork (karoshi). The council reframed the
white-collar exemption as a “high-level professional” (kodo puro-
feshonaru) exemption. Advocates insisted that it would only cover
high-wage professionals, but the unions worried that it would be
expanded over time to cover a broader range of employees.
Meanwhile, council members deliberated over upper limits on
overtime. They were unable to agree, so they asked Keidanren
and Rengo to negotiate directly, and the two sides came up with
a monthly limit of 100 hours of overtime (Council for the
Realization of Work Style Reform 2017). The Rengo negotiators
were caught by surprise, however, when protesters targeted Rengo
headquarters, decrying them for selling out.
290 Steven K. Vogel

Prime Minister Abe made a speech on equal pay for equal work in
January 2016 and issued a statement in February. “Prior to that, the
government had been cautious on this issue,” notes the ministry’s
Noriko Kawamura. “So these statements were a surprise. They indi-
cated a change in direction.”29 The equal pay for equal work proposals
did not encounter open opposition, but there was plenty of less visible
resistance. Many employers liked the option to hire a second tier of
workers with lower pay, benefits, and job security. This allowed them
to restrain wages and to achieve flexibility while preserving security for
their core workers. Moreover, the unions primarily represented regular
workers, so they were ambivalent about moves that could mean lower
pay, benefits, or security for their core members. Hence the Democratic
Party (Minshinto) remained cautious as well.30 “We support equal pay
for equal work,” notes Diet member Takanori Kawai, “but we do not
want this to come at the expense of the core workers. We want to raise
the non-regular workers, not lower the core workers.”31 Former
MHLW minister Shiozaki reported that the Prime Minister’s Office
made it clear that he should not pursue reforms that would threaten the
interests of Japan’s core workers, since this was a vital support group
for the party.32
The council report issued recommendations on the full range of
agenda items but included more detailed plans for the higher-priority
items such as work hours and equal pay for equal work. It also outlined
ten-year implementation programs for each item (Council for the
Realization of Work Style Reform 2017). The Labor Policy Council
subcommittees reviewed the recommendations in mid-2017 and
offered their general approval, allowing the proposal to move toward
submission to the Diet. One labor representative issued a dissenting
opinion on the high-level professional exemption to overtime rules
(Ministry of Health, Labor, and Welfare 2017b).
The LDP had been reluctant to propose controversial bills, despite its
overwhelming Diet majority, because it did not want to erode its
popularity. “We have been putting off bills where the opposition
disagrees since we passed the security bills [in 2015],” notes the
LDP’s Kozo Yamamoto. “We have not wanted to push through too
much too quickly.”33 Party leaders knew that the high-level profes-
sional exemption would provoke opposition, so they craftily combined
it with the more popular overtime limits and equal work for equal pay
provisions into a combined bill. This put the unions and the opposition
Abe’s Slight Left Turn 291

parties in an awkward position because they supported many of the

measures in the bill but not all. Rengo issued detailed comments on its
concerns but was reluctant to press too hard. “We support the main
direction of the work style reform council report,” explains Rengo’s
Murakami. “But we want the details to be worked out through the
Labor Policy Council” (Murakami interview, 2017; Japan Trade
Union Confederation 2017). The Democratic Party was divided
between those, like Kawai, a former private-sector union leader, who
were eager to compromise to get results and those, like Hiroshi Ogushi,
the party Policy Research Council chair, who wanted the party to
differentiate itself from the LDP. “I feel that something is better than
nothing,” explains Kawai, “so we should not be too rigid.”34 “We need
to point out the problems with these proposals,” Ogushi insists.35
The government did not include the dismissal compensation pro-
posal in the bill, judging that it would be too controversial. In fact,
support for the measure was tepid and variable even within the business
community. “Most companies do not want to dismiss workers because
they have a policy of cultivating human resources over the long term,”
explains Keidanren’s Satoshi Mukuta. “So many of our member com-
panies do not feel that they need this reform. We want to proceed
extremely cautiously because this could undermine the tradition of
stable employment.”36 Prime Minister Abe insisted on at least two
occasions (in 2013 and 2016) in Diet hearings that the government
had no intention of using the compensation system to make it easier for
companies to dismiss workers. Rather, it sought to give workers who
won unfair dismissal suits the option of seeking monetary compensa-
tion if they did not want to return to their company (Japan Business
Federation 2017, p.36). A policy council deliberated this proposal from
2015 through 2017 and referred it to the Labor Policy Council in
May 2017 (Ministry of Health, Labor, and Welfare 2017c). By 2017,
the government proposal stipulated that employees who had won an
unfair dismissal judgment could request monetary compensation, but
that the employer could not mandate this form of resolution. The
unions nonetheless continued to oppose the measure because they
feared that this system could make some companies more likely to
dismiss workers, and it could lead to more employer-friendly reforms
down the line.
The fall 2017 extraordinary session of the Diet was slated to be the
Work Style Reform Diet, but Abe called a snap election for October 22.
292 Steven K. Vogel

Abe’s popularity had recovered somewhat after a dip over the summer,
and he hoped to garner a victory before the opposition – including
a new party headed by Tokyo Governor Yuriko Koike, the Party of
Hope (kibo no to) – could mobilize. The risk paid off, as the LDP won
a resounding victory. In fact, the election provoked a split in the
Democratic Party, with some members running on the Party of Hope
platform and others forming a new Constitutional Democratic Party of
Japan (rikken minshuto) led by former chief cabinet secretary Yukio
Edano.37
The work style reform legislation was expected to pass smoothly
in the following Diet session, but the government ran into trouble in
February 2018 when Prime Minister Abe cited faulty data in Diet
deliberations to support the expansion of the discretionary work
system, whereby certain categories of office workers may be paid
based on presumed work hours rather than actual hours tallied. In
March, things got worse for Abe when the Ministry of Finance
disclosed that it had altered documents submitted for an investiga-
tion into allegations of favoritism toward Moritomo Gakuen, an
ultra-conservative school operator associated with Abe’s wife Akie.
Abe’s support ratings plunged, and deliberations on the labor reform
bill further bogged down. The administration decided to remove the
discretionary labor measure from the bill and to postpone it for
a year. Meanwhile, it modified the bill to delay implementation of
the working hours and equal pay for equal work provisions by
one year for small businesses. And it upgraded the proposed working
hours restrictions from ministerial ordinance to law, thereby allow-
ing for more severe penalties for companies that do not comply.
The Work Style Reform Law passed the Diet in June 2018. It
limited overtime to 100 hours a month and 720 hours a year, effect-
ive April 2019 for large companies and April 2020 for small com-
panies. But it exempted construction workers; taxi, bus, and truck
drivers; and physicians for five years. It prohibited unjustifiable dif-
ferences in wages and allowances for regular and non-regular work-
ers performing similar duties, effective April 2020 for large
companies and April 2021 for small companies. As already noted,
employers often justified pay gaps on the basis that the two categories
of workers were subject to different levels of constraint because the
core workers were required to work overtime, switch job assign-
ments, or relocate on demand. So the government would differentiate
Abe’s Slight Left Turn 293

between compensation that could be justified rationally on the basis

of these different levels of obligation to the firm, such as base salary,
and those that could not, such as commuter stipends. And the bill
exempted high-level professionals in certain specialized fields with
annual income above 10.75 million yen from overtime rules, pro-
vided that labor and management reach an accord on implementation
and the employees consent, effective April 2019 (Nihon Keizai
Shinbun 2018a, 2018b). The government would monitor compliance
with these rules rather strictly, which could deter many companies
from using the exemption.38
Meanwhile, the government began work on the next stage of labor
policy reform, a “human resources revolution” (hitozukuri kakumei)
that would focus on education and training issues. Prime Minister Abe
vowed to use revenues from the October 2019 increase in the consump-
tion tax for free preschool and higher education subsidies and to press
companies to allocate a greater share of revenues to wages and invest-
ment. “The prime minister’s office is taking the lead with the human
resources revolution too,” notes Rengo’s Murakami. “They have
already decided what they want to do.”39 The government launched
a reform council, the “100-Year Life Era Vision Council” (jinsei hya-
kunen jidai koso kaigi) in September 2017 to deliberate reforms. The
Council reported in June 2018, proposing free preschool starting in
October 2019 and free university for low-income students starting in
April 2020. It also discussed proposals for promoting workforce par-
ticipation among the elderly and strengthening adult education, espe-
cially in the information technology sector (Nihon Keizai Shinbun
2018c).
In June 2018, the Abe administration abruptly announced a program
to allow more foreign “guest workers” in selected industries with
severe labor shortages, reversing its stance of not liberalizing immigra-
tion while taking care not to label the program as such. The program
differed from the government’s previous policy for foreign workers
because it would issue working visas rather than “technical internship
trainee” visas, and it would allow workers in a broader range of
sectors. The LDP encountered internal resistance from its own right
wing and protests from the opposition parties in Diet debates. “We
opposed the bill because the government was moving too fast. It had
not done the proper preparation,” asserts Naoko Sakaue, general
manager of the Executive Office at the Democratic Party for the
294 Steven K. Vogel

People.40 “We are concerned that the burden of supporting these

workers will fall to local governments. And they are going to need
more support for that.” Nonetheless, the LDP’s Shiozaki felt that the
opposition parties were half-hearted in their dissent. “They did not
fight us that hard because everyone knows that we need these work-
ers,” he explains. “And we limited the program to workers in specific
impacted sectors, and we promised to change the policy if the labor
shortage ends.”41 Shiozaki’s LDP colleague Shintaro Ito, who repre-
sents a district in the Tohoku region, agrees: “We do not have enough
people to work in jobs like restaurants and hotels in my district, so we
are really going to need these foreign workers to help us.”42
The legislation that passed in December 2018 (effective April 2019)
created two new categories for work visas. Type 1 visa workers would
be required to have some knowledge of Japanese language and
a minimal level of skill in one of 14 fields (see Table 10.2). The
government would allow up to 345,000 foreign workers over five
years in fields with labor shortages estimated at 1.45 million during
that period (Nihon Keizai Shinbun 2018d). Government officials did
not expect the program to depress wages for Japanese workers, since
the number of foreign workers would be considerably smaller than the
expected shortage, and employers would be required to pay the foreign
workers wages equivalent to those of Japanese workers. The foreign
workers could stay for up to five years, but they could not bring family
members. They would be able to change jobs within their designated
field, unlike the trainees in the past, leading some critics to worry that
they might cluster in urban areas over time rather than remaining in the
rural areas most desperate for workers. Type 2 visa workers would be
required to demonstrate higher levels of skill. They would receive one-
to-three-year renewable visas, and they would be permitted to bring
family. The Ministry of Justice Immigration Bureau would be reorgan-
ized as a full-fledged agency with about 500 more officials, and the
agency would be expected to monitor the workers and organize sup-
port systems for them as well as to oversee their entry into Japan. The
hiring companies would also bear some responsibility for supporting
foreign workers, helping to locate housing and briefing workers on
Japanese customs (Ministry of Health, Labor, and Welfare 2019b;
Nihon Keizai Shinbun 2018e).
Abe’s Slight Left Turn 295

Table 10.2 The foreign worker program by industry

Number of
Field expected workers Ministry

Elder care 60,000 Ministry of Health, Labor,

Building maintenance 37,000 and Welfare
Synthetic materials 21,500 Ministry of Economy, Trade,
Industrial machinery 5,250 and Industry
Electronics 4,700
Construction 40,000 Ministry of Land,
Shipbuilding 13,000 Infrastructure, and
Auto maintenance 7,000 Transportation
Aviation 2,200
Hospitality 22,000
Agriculture 36,500
Fishing 9,000 Ministry of Agriculture,
Beverages 34,000 Forestry, and Fisheries
Food services 53,000

Source: Ministry of Health, Labor, and Welfare

10.5 Corporate Practices 2012–Present: Can Policy Reforms

Transform Working Life?
So with all of this reform activity, have company employment practices
really changed? Government officials understand that simply declaring
a new policy is not sufficient to alter long-standing practices and norms.
They have moved beyond setting guidelines to actively reshape corpor-
ate practices by monitoring behavior through regional branches, publi-
cizing model behavior, and sanctioning violations. Yet ministry officials
do not have the staff to monitor all of Japan’s companies. They have
local bureaus throughout the country, but their staff members are only
able to inspect a small percentage of workplaces.43 So they launched
public relations campaigns to promote core goals – such as gender equity
and work-life balance – and devised programs to publicize model com-
panies. Many of their guidelines are not requirements, and many of the
requirements do not stipulate penalties for noncompliance.
With respect to gender equity, the ministry devised an “eruboshi”
point system to recognize companies for their effort, with three levels of
296 Steven K. Vogel

achievement based on five criteria: female hiring, job tenure, work style
(including work hours), share of executives, and multiplicity of career
courses (Ministry of Health, Labor, and Welfare 2016). As of
December 2018, the ministry had recognized 507 companies with the
third and highest level (pink), 264 with the second (red), and 4 with the
first (orange) (Ministry of Health, Labor, and Welfare 2017d).
The ministry set up special headquarters in its regional branches
to monitor progress on work style reforms in 2016, well before the
legislation passed. On equal pay for equal work, the ministry
issued initial guidelines in 2016 and more definitive guidelines in
2018. Ministry officials focused primarily on fair treatment of
different categories of workers within firms, but they also moni-
tored imbalances across firms. The ministry could not impose
penalties for noncompliance, but it could issue administrative guid-
ance (nonbinding directives). And companies would have an incen-
tive to comply because judges might refer to the guidelines in the
case of disputes.44 On work hours, the ministry began publicizing
model companies on its website in 2015, even before the primary
reform council had begun its deliberations. It issued administrative
guidance to 29,097 companies with high rates of overtime from
April 2018 to March 2019: 69.6 percent admitted to violations,
including illegal overtime (40.4 percent of the companies) and
unpaid overtime (6.4 percent) (Ministry of Health, Labor, and
Welfare 2019c).
Meanwhile, Keidanren joined the effort to promote work style
reform. It does not normally get into the business of telling its members
what to do, but it made an exception in this case. It promoted more
vacation days in 2016, and its follow-up survey found that 70 percent
of respondents were aware of the campaign and 55 percent had taken
action in response (Japan Business Federation 2017; see also Japan
Business Federation 2019). It then launched a more ambitious “Work
Style Reform Challenge 2017” comprised of four elements: a specific
action plan; a joint declaration (along with other business associations)
to rethink corporate practices; a campaign to promote taking holidays;
and related public relations activities. The action plan called on mem-
ber companies to specify key performance indicators on three core
goals: reducing long working hours, taking more vacation time, and
deploying more flexible work styles. The companies would set specific
progress targets with defined deadlines.
Abe’s Slight Left Turn 297

The evidence to date suggests that companies are implementing work

style reforms. The outstanding question is how profound and how
lasting these changes will be.45 A Japan Institute for Labor Policy and
Training (JILPT) Survey reports that 82.8 percent of companies with
more than 300 employees were implementing work style reform policies,
while 51.0 percent of companies with fewer than 50 employees were
doing so (Kikanshi rengo tsushin 2019). A government survey finds that
companies were promoting vacation time (60 percent of companies),
monitoring long work hours (62 percent), introducing flex time (23 per-
cent), allowing telework (8 percent), introducing new job categories such
as limited regular workers (10 percent), improving conditions for those
with child care or elder care duties (52 percent), and transmitting
a message on work style reform from top management (28 percent)
(Ministry of Health, Labor, and Welfare 2018a). A Keidanren survey
shows that 87 percent of member companies had transitioned non-
regular workers to regular, limited regular, or unlimited contract status
(Japan Business Federation 2017). The share of employees working
more than 60 hours a week declined from 12.2 percent in 2004 to
8.5 percent in 2014 and 7.7 percent in 2017. But the share of employees
working more than 48 hours per week remained high by international
standards, at 26.4 percent versus 16.6 percent in the United States and
8.3 percent in Germany in 2017 (Ministry of Health, Labor, and Welfare
2019d, pp.14–15). Despite all of this reform activity, the JOBS Research
Center reveals that most workers do not feel that their work style has
changed much (25.1 percent) or at all (28.3 percent) due to work style
reform (JOBS Research Center 2019).
The work style reform became something of a national campaign,
with a plethora of books that reference the term, mostly “how to”
books on implementing change at the company level. The National
Diet Library search engine provided the following results for titles that
contain “hatarakikata kaikaku”: 17 books in 2013 or before, 50 in
2015, 177 in 2016, and 909 in 2017. And a Nikkei Shinbun article
search found 17 articles with “hatarakikata kaikaku” in the title in
2013, 58 in 2014, 114 in 2015, 698 in 2016, and 2,057 in 2017.
Meanwhile, female workforce participation increased from 63.4 per-
cent in 2012 to 69.4 percent in 2017 (OECD.Stat 2019a). Nobuko
Nagase finds that the Womenomics program, from the initial
announcement in 2013 and continuing through the law of 2015, had
a modest but significant impact on the promotion of women to
298 Steven K. Vogel

managerial positions in large firms.46 The female share of corporate

executives increased steadily from a low base: from 14.4 percent in
2012 to 19.4 percent in 2017 for subsection chief (kakaricho); from
7.9 percent to 19.4 percent for section chief (kacho); and from 4.9 per-
cent to 10.6 percent for department chief (bucho) (Ministry of Health,
Labor, and Welfare 2018b). According to the World Economic Forum
Global Gender Gap report, which is frequently cited in Japan, Japan’s
ranking for economic opportunity and participation rose modestly
from 118th of 144 countries in 2016 to 115th in 2019, but its overall
gender gap ranking declined from 111th to 121st due to a drop in its
political empowerment score (World Economic Forum 2019).
Meanwhile, elderly (over sixty-five) labor participation increased
from 19.9 percent in 2010 to 24.7 percent in 2018 (Ministry of
Internal Affairs and Communications 2019b).
Despite all this reform, Japanese companies have not delivered what
workers want most: higher wages. “I thought managers would have
increased wages more than they have, given the high demand for
workers,” concedes the LDP’s Yamamoto. “The managers say they
can’t raise wages because they have fixed costs. They cannot do this
unless they are confident that they have a bright future.”47 Japanese
real annual wages have been roughly stagnant since 1990, declining
from 4.388 million yen in 2011 to 4.334 million yen in 2018 (in 2018
constant prices).48

10.6 Future Prospects

Japan’s tight labor market presents an opportunity to remedy the
weaknesses of the postwar model, such as gender discrimination and
long work hours, while preserving or bolstering strengths, such as
employment security and collaborative labor–management relations.
And it offers the possibility of retooling the system to better fit an era
characterized by more service sector work, more variable work modes,
and more options for telework and collaboration across distances.
In the best-case scenario, work style reforms would deliver a win-win
solution: an increased workforce and higher productivity for the econ-
omy, as well as a healthier working life, less inequality, and greater
diversity for society. In essence, companies would expand the work-
force and enhance productivity by improving worker welfare. They
could attract more workers by upgrading the status of non-regular
Abe’s Slight Left Turn 299

workers, creating job categories that offer security without unreason-

able demands, and reducing work hours. They might even improve
productivity by reducing work hours if they could shift the primary
goal from clocking hours to delivering results. After all, workers who
are happier with their workplace and better matched for their jobs
should perform better.
Companies could recruit more women by providing a better work
environment, including daycare facilities and childcare and elder-care
leave, more limited working hours, and more opportunities for advance-
ment. And they could crack down on sexual harassment; “maternity”
harassment (matahara), including social pressure on pregnant employ-
ees to quit; and “power” harassment, abuses of seniority. Measures to
support female and foreign workers should increase productivity by
expanding the talent pool and bringing in novel practices and ideas.49
Work style reform could give workers more choices about work
hours and work style and more flexibility to change course over
a lifetime. The postwar system delivered real benefits for Japan’s core
workers in the form of stable employment and steadily rising wages,
but this came at the cost of inflexibility and discrimination. Workers
who desired a different style of work or who wanted to change direc-
tion mid-career were at a severe disadvantage. In pursuing win-win
solutions, government and industry would also address some of
Japan’s most pressing social problems: childcare and elder care, gender
discrimination, sexual harassment, and a lack of flexibility in career
choices.
But will this really happen? The evidence presented here suggests
several tentative conclusions. First, the “deregulatory” reforms of the
1995–2005 period had detrimental effects at the macro level, exacer-
bating inequality and suppressing aggregate demand. Second, reforms
since that time have progressed in a direction better matched to the
needs of Japanese workers, companies, and society at large. Third, the
government is serious about moving in this direction and is going
beyond announcing reforms to monitoring and directing firm behavior.
Fourth, companies have made some real changes in human resource
practices. And fifth, the Womenomics and work style reforms have
delivered some initial results at the aggregate level, including increased
female and elderly workforce participation.
Just because the political dynamics have changed, however, does not
mean that Japan will realize the best-case scenarios sketched here. The
300 Steven K. Vogel

reforms are being propelled by a conservative party and industry

leaders focused more on increasing the birth rate and spurring eco-
nomic growth than on promoting equal rights, diversity, and worker
welfare as goals in themselves. Meanwhile, companies have been slow
to revise entrenched practices or to challenge traditional social norms.
They have been reluctant to raise wages or to improve working life in
ways that incur substantial costs. Many companies have embraced
work style reform in form more than substance, continuing to pursue
cost-cutting while touting employee welfare. Moreover, Japanese com-
panies have been pursuing corporate governance reforms that make
them more sensitive to shareholder concerns. This opens up the very
real possibility that the Abe administration’s corporate governance
reforms could undermine its labor market reforms. That is, Japanese
companies have been enjoying rising profits while sharing a smaller
percentage of those profits with workers. Japanese firms’ return on
equity rose steadily from 6 percent in 2010 to 9.8 percent in
2018 (Nihon Keizai Shinbun 2019). Yet the labor share of income
declined from 74.7 percent in 2008 to 66.3 percent in 2019 (Ministry
of Finance 2019).
Japan is unusual in a comparative context in that it has the political
will to shift in the direction of policies that support both economic
growth and worker welfare. The United States, France, and other
European countries continue to move in the other direction, weakening
labor protections. Japan’s right–left political cleavage on labor and
other welfare issues is narrowing, not expanding. The Japanese case
thus raises the intriguing question of whether changing demographics
and labor market developments could realign politics in other indus-
trial countries as well.
Oddly enough, the Abe administration’s slight left turn left out the
Left. Japan will make more headway toward “win-win” solutions if the
progressive parties and the labor unions participate actively in shaping
policy reforms and negotiating firm-level responses. The center-left
parties undermined their cause through missteps when they were in
power 2009–2012 and through infighting before and after. The unions
lack resources, but they have also focused too much on defensive
efforts – such as blocking the proposed white-collar exemption and
the compensation system for dismissals – and too little on a more
proactive agenda of empowering workers and transforming labor rela-
tions from the ground up. Japanese workers are left in the awkward
Abe’s Slight Left Turn 301

position of relying on a center-right ruling party and industry associ-

ations to promote their interests. The LDP and Keidanren, to their
credit, have made some real changes in a positive direction. Over the
longer term, however, Japan’s workers cannot rely on the LDP and
Keidanren to advocate for them.
Ultimately, the fate of Japan’s work style reform hinges on the ability
of ministry officials, industry associations, labor unions, and individual
managers and workers to persist in monitoring progress and to insist on
results. The ministry publicity campaigns and administrative guidance
certainly help, but they will require backup via more forceful means of
enforcement. And if managers really want their workers to work
shorter hours, they may have to turn off the lights, lock the doors,
and head home themselves.

Notes
1. This chapter is a revised and updated version of Vogel (2018a). The
author thanks Maurice Ang, Makoto Fukumoto, Marina Hamada,
Chihiro Iwatsuki, Sae Kobayashi, Toma-Jin Morikawa, Ivo Plsek,
Reina Sasaki, and Amanda Zhao for research assistance; Kazutoshi
Koshiro, James Lincoln, Ivo Plsek, several anonymous reviewers, and
the editors and chapter authors of this volume for comments; and the Il
Han New Chair and the Center for Japanese Studies at the University of
California, Berkeley for research funding.
2. Japanese life expectancy rose from 72.0 years in 1970 to 82.0 in 2005
and 84.2 in 2017 (Organization for Economic Co-operation and
Development Data 2019); while fertility (births per woman) declined
from 2.14 in 1970 to a low of 1.26 in 2005 and rose to 1.43 in 2017
(World Bank Data Bank 2019).
3. Watanabe (2014) describes a more unidirectional neoliberal thrust to
Japan’s labor reforms. Miura (2012, p.68), however, identifies both of
these elements in policy reforms.
4. Hall and Soskice (2001, p.37) use the term “comparative institutional
advantage” to depict how firms produce some kinds of goods more
efficiently than others because the institutions of that particular
national political economy support those activities.
5. Some segments of this chapter build on Vogel (2006), especially
pp.79–83 and pp.115–126, and Vogel (2018b), pp.96–100. Vogel
(2006) reviews the theoretical underpinnings of this analysis of firm
strategy and policy preferences in more detail, tapping insights from
302 Steven K. Vogel

institutional economics and economic sociology and comparing Japan

to the United States and Germany.
6. Interview with Yoshio Nakamura, vice chairman, director general, and
representative director, Keidanren, Tokyo, August 6, 2013.
7. Interview with Kazuo Sugeno, Professor Emeritus, University of Tokyo,
Tokyo, September 11, 2017.
8. Japan Federation of Employers’ Associations (1995). Nikkeiren merged
with Keidanren in 2002.
9. Miura (2012, p.131). The Japanese Confederation of Labor (Domei)
and the Federation of Independent Labor Unions (Churitsu Roren)
merged to form Rengo in 1989, and the General Council of Trade
Unions of Japan (Sohyo) joined them in 1990. Union leaders hoped to
consolidate their power by joining forces, but Rengo has not emerged
as a dominant force in setting the policy agenda, even in the DPJ
years.
10. As of 2019, 76.8 percent of regular workers were men, whereas
55.7 percent of non-regular workers were women (Ministry of
Internal Affairs and Communications 2019a).
11. Ministry of Health, Labor, and Welfare (2003) reports, for example,
that firms dealt with high labor costs in the following ways: 36.9 percent
saw sales growth and product development; 15.6 percent cut general
costs; 12.7 percent introduced more performance-based wages;
11.6 percent added non-regular workers; and 8.1 percent cut wage
costs or reduced workers by attrition.
12. Ministry of Health, Labor, and Welfare (2019a) for Japan; Lazonick
and O’Sullivan (2000) for the United States. Kambayashi and Kato
(2017) find that Japan’s economic stagnation since the 1990s had little
discernible adverse effect on the job stability of prime-age male workers
with at least five years of tenure, but it did undermine job stability for
youth and mid-career hires.
13. The non-regular category breaks down into the following subcategories:
temporary (paato) and part-time (arubaito) 25.6 percent; contract
workers (keiyaku shain) 5.4 percent; agency temps (haken shain)
2.5 percent; independent contractors (itaku) 2.2 percent; and other
1.4 percent (Ministry of Internal Affairs and Communications 2017).
A Japan Institute of Labor Policy and Training (JILPT) study (2017,
pp.302–303) finds that Japan has preserved the long-term employment
system for core workers. Employers, workers, and the general public
continue to support this system; but the long-term employment system
covers a smaller share of total workers, the seniority wage system has
declined, and workplace solidarity has atrophied.
Abe’s Slight Left Turn 303

14. These patterns of variation across sectors are documented in more detail
in Vogel (2006, pp.115–204).
15. Schaede (2008, pp.174–199) depicts the policy reforms and corporate
adjustments during this period as bolder and more strategic than
characterized here, stressing that job mobility increased.
16. Fukao (2017, p.7) estimates that, if firms had not increased the share of
non-regular workers after 1988, the average quality of labor would be
8 percent higher today.
17. The LDP party president is elected based on a formula that combines
votes from party members and Diet members. The party president
normally serves as prime minister when the party is in power.
18. Interview with Yasuchika Hasegawa, director, chairman of the board,
Takeda Pharmaceuticals, Tokyo, September 13, 2017.
19. Dalton (2017) characterizes Womenomics as neoliberal in thrust
because the program does not address the status gap between regular
and non-regular workers or the problem of long work hours. Yet the
Abe administration went on to focus on precisely these issues, as
described later in this chapter.
20. The ministry reports that 99.5 percent of the large companies required
to report complied (Ministry of Health, Labor, and Welfare 2017a).
21. Interview with Yoko Murakami, executive director, Department of
Working Conditions, Japanese Trade Union Confederation (Rengo),
Tokyo, September 20, 2017.
22. Interview with Hiromitsu Ohtsuka, senior deputy director, Labor
Relations Law Division, Labor Standards Bureau, Ministry of Health,
Labor, and Welfare, Tokyo, August 12, 2016. Non-regular workers earn
65.8 percent of what regular workers earn (not including overtime and
bonuses, which increase this gap further), and women regular workers
earn 72.9 percent of what male regular workers earn (Ministry of
Health, Labor, and Welfare 2019a). Fukao (2017) argues that the
productivity gap between regular and non-regular workers is actually
larger than the wage gap. If he is right, then the policy implication would
be that Japan needs to increase the share of regular workers, because
companies invest more in training these workers and get higher
productivity out of them, more than justifying the higher wages.
23. Kawamura (2017, p.4), calculation based on Ministry of Health, Labor,
and Welfare (2012).
24. On this broader strategy, see George Mulgan (2018), especially pp.32–86.
25. Ohtsuka interview, 2016.
26. Interview with Hiromitsu Ohtsuka, senior deputy director, Labor
Relations Law Division, Labor Standards Bureau, Ministry of Health,
Labor, and Welfare, Tokyo, September 21, 2017.
304 Steven K. Vogel

27. Interview with Yasuhisa Shiozaki, member, House of Representatives,

Liberal Democratic Party, Tokyo, September 19, 2017.
28. Shiozaki interview, 2017.
29. Interview with Noriko Kawamura, senior deputy director, Dispatch and
Contract Labor Policy Planning Division, Employment Security Bureau,
Ministry of Health, Labor, and Welfare, Tokyo, August 12, 2016;
Kawamura (2017, p.11).
30. The DPJ merged with the Japan Innovation Party (ishin no to) and
Vision of Reform (kaikaku kesshu no kai) in March 2016; the
combined party was renamed the Democratic Party.
31. Interview with Takanori Kawai, member, House of Councillors,
Democratic Party, Tokyo, September 12, 2017.
32. Shiozaki interview, 2017.
33. Interview with Kozo Yamamoto, member, House of Representatives,
Liberal Democratic Party, Tokyo, September 13, 2017.
34. Kawai interview, 2017.
35. Interview with Hiroshi Ogushi, member, House of Representatives,
Democratic Party, Tokyo, September 13, 2017.
36. Interview with Satoshi Mukuta, senior managing director, Japan
Business Federation, Tokyo, September 14, 2017.
37. Ogushi was reelected as a member of the Party of Hope. Kawai was not
up for reelection because he is an upper house Diet member.
38. Interview with Hayato Kume, deputy director, Working Conditions
Policy Division, Labor Standards Bureau, Ministry of Health, Labor,
and Welfare, January 9, 2019.
39. Murakami interview, 2017.
40. Interview, January 16, 2019.
41. Interview, January 7, 2019.
42. Interview, January 7, 2019.
43. The Regulatory Reform Promotion Council proposed that the ministry
outsource a portion of this work to private agencies to expand its reach
(Yomiuri Shinbun 2017).
44. Kawamura interview, 2016.
45. Kojima, North, and Weathers (2017) take a more skeptical view.
I concur that the reforms are not likely to achieve all of their lofty
goals but maintain that the direction of reform has shifted from
employer-friendly flexibility to more worker-friendly measures.
46. See Chapter 11 in this volume.
47. Yamamoto interview, 2017.
48. OECD.Stat (2019b). Genda (2018) and collaborators investigate this
puzzle at length. They offer a wide range of explanations, but many
stress the pathologies of the dual labor market. Employers may not
Abe’s Slight Left Turn 305

raise wages for core workers, for example, because they can increase
the workforce in the short term by hiring non-regular workers
instead. Or they may be reluctant to raise wages for core workers
because they know that they will not be able to lower wages even if
demand softens.
49. Fukao (2017, pp.3–5) notes that Japan ranks at the top among rich
countries in literary proficiency, yet it stands out in its low usage of the
literary proficiency of females at work.

References
Council for the Realization of Work Style Reform. 2017. “Hatarakikata
kaikaku jikko keikaku” (Work Style Reform Realization Plan),
March 28.
Dalton, Emma. 2017. “Womenomics, ‘Equality’ and Abe’s Neo-liberal
Strategy to Make Japanese Women Shine.” Social Science Japan
Journal 20: 95–105.
Fukao, Kyoji. 2017. “Secular Stagnation and the Labor Market in Japan.”
Paper for the Conference on Prospects for Sustained Growth in Asia,
Seoul, September 7–8.
Genda, Yuji, ed. 2018. Hitode busoku nanoni naze chingin ga agaranai no ka
(Why Don’t Wages Go Up with the Labor Shortage?) Tokyo: Keio
University Press.
George Mulgan, Aurelia. 2018. The Abe Administration and the Rise of the
Prime Ministerial Executive. Abingdon: Routledge.
Hall, Peter, and David Soskice, eds. 2001. The Varieties of Capitalism: The
Institutional Foundations of Comparative Advantage. Oxford: Oxford
University Press.
Hirschman, Albert. 1970. Exit, Voice, and Loyalty: Responses to Decline in
Firms, Organizations, and States. Cambridge, MA: Harvard University
Press.
Ito, Sadanori (Ministry of Economy, Trade, and Industry). 2017. “Daiyonji
sangyo kakumei to hatarakikata kaikaku hitozukuri kakumei ni tsuite”
(The Fourth Industrial Revolution: Work Style Reform and the Human
Resources Revolution). Briefing materials.
Jackson, Gregory. 2005. “Stakeholders Under Pressure: Corporate Governance
and Labor Management in Germany and Japan.” Corporate Governance:
An International Review 13: 419–428.
Japan Business Federation. 2017. “Hatarakikata kaikaku no genjo to kongo
no kadai,” (Work Style Reform: Current Status and Future Issues).
September.
306 Steven K. Vogel

Japan Business Federation. 2019. “Keiei rodo seisaku tokubetsu iinkai

hokoku” (Report of the Committee on Management and Labor Policy).
Japan Federation of Employers’ Organizations. 1995. “Shin jidai no
nihonteki keiei” (Japanese Management for a New Era).
Japan Institute of Labor. 2002. “Jigyo saikochiku to koyo ni kansuru chosa”
(Survey of Corporate Restructuring and Employment).
Japan Institute of Labor Policy and Training (JILPT). 2017. Nihonteki koyo
shisutemu no yukue (The Evolution of the Japanese Employment
System). Tokyo: JILPT.
Japan Trade Union Confederation. 2017. “Hatarakikata kaikaku kankei
shiryo” (Documents on Work Style Reform). July 4.
JOBS Research Center. 2019. “Hatarakikata kaikaku ni kansuru ishiki
chosa” (Opinion Survey on Work Style Reform). Tokyo: JOBS
Research Center.
Kambayashi, Ryo, and Takao Kato. 2017. “Long-term Employment and Job
Security Over the Past 25 Years: A Comparative Study of Japan and the
United States.” Industrial and Labor Relations Review 70: 359–394.
Kawamura, Noriko. 2017. “Doitsu rodo doitsu chingin ni tsuite no saikin no
kento jokyo ni tsuite” [On Recent Deliberations on Equal Pay for Equal
Work]. Dispatch and Contract Labor Policy Planning Division,
Employment Security Bureau, Ministry of Health, Labor, and Welfare.
Briefing materials. July 28.
Kikanshi rengo tsushin. 2019. “Hatarakikata kaikaku okonatteiru 6-wari”
(60 Percent Have Implemented Work Style Reform). September 20.
Kojima, Shinji, Scott North, and Charles Weathers. 2017. “Abe Shinzo’s
Campaign to Reform the Japanese Way of Work.” Asia Pacific Journal
15(3): 1–17.
Koshiro, Kazutoshi. 2017. “Hatarakikata kaikaku to ‘datsu jikan chingin’
ronsoko” (Work Style Reform and the Debate Over Non-Hourly
Wages). Unpublished manuscript.
Lazonick, William, and Mary O’Sullivan. 2000. “Maximizing Shareholder
Value: A New Ideology for Corporate Governance.” Economy and
Society 29: 14–27.
Ministry of Finance. 2019. “Hojin kigyo tokei chosa” (Statistics of
Corporations). www.mof.go.jp/pri/reference/ssc/results/h30.pdf, accessed
December 27, 2019.
Ministry of Health, Labor, and Welfare. 2003. Chingin hikiageto no jittai ni
kansuru chosa (Survey of Wage Increases).
Ministry of Health, Labor, and Welfare. 2012. “21 seiki seinensha judan
chosa” (Longitudinal Survey of Adults in the Twenty-First Century).
Ministry of Health, Labor, and Welfare. 2014a. “Tayo na seishain no fukyu
kakudai no tame no yushikisha kondankai hokokusho” (Council of
Abe’s Slight Left Turn 307

Experts Report on Promoting the Diversity of Permanent Employees).

July.
Ministry of Health, Labor, and Welfare. 2014b. “Tayo na seishain” (The
Diversity of Permanent Employees). Briefing materials.
Ministry of Health, Labor, and Welfare. 2016. “Josei katsuyaku suishinho ni
motozuku nintei seido” (The Certification System under the Promotion
of Women’s Participation and Advancement in the Workplace Law).
Briefing materials.
Ministry of Health, Labor, and Welfare. 2017a. “Dainiji abe naikaku hossokuji
kara no josei katsuyaku no shinpo” (Progress on Women’s Empowerment
Under the Second Abe Administration). Briefing materials.
Ministry of Health, Labor, and Welfare. 2017b. “Hatarakikata kaikaku
o suishin suru tame no kankei horitsu no seibi ni kansuru horitsuan
yoko no toshin” (Report on the Draft Outline for the Preparation of the
Laws to Promote Work Style Reform). Press release with attached
documents. September 15.
Ministry of Health, Labor, and Welfare. 2017c. “Tomei katsu koseina
rodofunso kaiketsu shisutemuto no arikata ni kansuru kentokai
hokokusho” (Advisory Council Report on a Transparent and Fair
Labor Dispute Settlement System). May.
Ministry of Health, Labor, and Welfare. 2017d. “Josei katsuyaku shinpoho
ni kakaru nintei jokyo.” (The Current Status of Certification under the
Promotion of Women’s Participation and Advancement in the
Workplace Law). Briefing materials. September 30.
Ministry of Health, Labor, and Welfare. 2018a. Rodo keizai doko chosa
(Survey on Labor Economy Trends). November.
Ministry of Health, Labor, and Welfare. 2018b. “Koyo kinto kihon chosa”
(Basic Survey on Employment Equity).
Ministry of Health, Labor, and Welfare. 2019a. Chingin kozo kihon tokei
chosa (Basic Survey on Wage Structure). www.mhlw.go.jp/toukei/list/c
hinginkouzou.html, accessed November 19, 2019.
Ministry of Health, Labor, and Welfare. 2019b. “Aratana gaikoku jinzai
ukeire seido ni tsuite” (On a New System for Admitting Foreign
Workers). Briefing materials. February.
Ministry of Health, Labor, and Welfare. 2019c. “Chokikan rodo ga
utagawareru jigyojo ni taisuru heisei 30-nendo no kantoku shido
kekka o kohyo shimasu” (2018 Inspection Results for Companies
Suspected of Excess Working Hours). Press release. September 24.
Ministry of Health, Labor, and Welfare. 2019d. “Hatarakikata kaikaku ga
mezasu mono” (The Goals of Work Style Reform). Briefing materials.
January 8.
308 Steven K. Vogel

Ministry of Internal Affairs and Communications. 2017. Rodoryoku chosa

(Labor Force Survey).
Ministry of Internal Affairs and Communications. 2019a. Rodoryoku chosa
(Labor Force Survey).
Ministry of Internal Affairs and Communications. 2019b. Statistics Bureau
of Japan Statistics. www.stat.go.jp/english/data/roudou/results/month/,
accessed February 19.
Miura, Mari. 2012. Welfare Through Work: Conservative Ideas, Partisan
Dynamics, and Social Protection in Japan. Ithaca, NY: Cornell University
Press.
Nihon Keizai Shinbun. 2018a. “Hatarakikata kaikakuho ga seiritsu” (The
Work Style Reform Law Passes). June 29.
Nihon Keizai Shinbun. 2018b. “Kuwashiku wakaru hatarakikata
kaikakuho” (Understanding the Work Style Reform Law). July 5.
Nihon Keizai Shinbun. 2018c. “Hitozukuri kakumei koso kettei” (The
Human Resources Revolution Plan Is Set). June 14.
Nihon Keizai Shinbun. 2018d. “‘Erabareru kuni’ e seido sekkei” (System for
“a Chosen Nation” Is Set). December 8.
Nihon Keizai Shinbun. 2018e. “Gaikokujin ukeire kakudai kaisei nyukanho
ga seiritsu” (Law Passes to Expand the Admission of Foreign Workers).
December 8.
Nihon Keizai Shinbun. 2019. “ROE, 3-nen buri teika 18-nendo wa 10%
ware” (ROE Down for the First Time in Three Years, Below 10% in
2018). January 7.
OECD.Stat. 2019a. https://stats.oecd.org/Index.aspx?DataSetCode=LFS_S
EXAGE_I_R, accessed April 5, 2019.
OECD.Stat. 2019b. https://stats.oecd.org/, accessed December 8, 2019.
Organization for Economic Co-operation and Development Data. 2019.
https://data.oecd.org/healthstat/life-expectancy-at-birth.htm, accessed
November 13, 2019.
Pempel, T. J., and Keiichi Tsunekawa. 1979. “Corporatism Without Labor?
The Japanese Anomaly.” In Trends Toward Corporatist Intermediation,
edited by Philippe C. Schmitter and Gerhard Lehmbruch, pp.231–270.
London: Sage.
Prime Minister’s Office. 2016. “Nihon saiko senryaku 2016” (Japan Revival
Strategy 2016). June 2.
Prime Minister’s Office. 2015. “Ichioku sokatsu kokumin kaigi no kaisai ni
tsuite” (On the Creation of the 100 Million People Participation Plan).
October 21.
Schaede, Ulrike. 2008. Choose and Focus: Japanese Business Strategies for
the 21st Century. Ithaca, NY: Cornell University Press.
Abe’s Slight Left Turn 309

Shimizu, Masato. 2016. “Ikioizuku Abe keisansho naikaku no ikisaki” (The

Abe-METI Cabinet is Gaining Strength). Nihon Keizai Shimbun,
June 21.
Shiozaki, Yasuhisa. 2017. Briefing materials. August 18.
Vogel, Steven K. 2006. Japan Remodeled: How Government and Industry Are
Transforming Japanese Capitalism. Ithaca, NY: Cornell University Press.
Vogel, Steven K. 2018a. “Japan’s Labor Regime in Transition: Rethinking
Work for a Shrinking Nation.” Journal of Japanese Studies 44: 257–292.
Vogel, Steven K. 2018b. Marketcraft: How Governments Make Markets
Work. New York: Oxford University Press.
Watanabe, Hiroaki Richard. 2014. Labor Market Deregulation in Japan and
Italy: Worker Protection Under Neoliberal Globalisation. London:
Routledge.
Wilensky, Harold L. 2002. Rich Democracies: Political Economy, Public
Policy, and Performance. Berkeley and Los Angeles: University of
California Press.
World Bank Data Bank. 2019. https://data.worldbank.org/indicator/SP
.DYN.TFRT.IN?name_desc=false, accessed November 13, 2019.
World Economic Forum. 2019. Global Gender Gap Report 2020. Geneva:
World Economic Forum.
Yomiuri Shinbun. 2017. “Rokisho gyomu o minkan itaku” (Labor
Standards Inspections to be Outsourced to the Private Sector). May 22
evening edition, 1.
11 Abe’s Womenomics Policy
Did It Reduce the Gender Gap
in Management?
nobuko nagase

11.1 Introduction
The underutilization of female human resources is an urgent problem
in East Asian countries that face declining births, aging populations,
and expected future declines in their labor forces. Many East Asian
countries, such as Japan and Korea, also have a large gender wage
gap. Prime Minister Abe announced that promoting female labor
force participation would be one of his core growth policies when
he came into office in late 2012. However, promoting higher labor
force participation is not enough if family responsibilities keep
women in the low earnings sector. Enabling females to build their
careers is also an important goal. Nagase (2018a) found a significant
effect of “Womenomics” on women’s long-term employment pos-
ition after their first childbirth. Long-term employment, or seishain1
in Japan, connects to one’s future higher wage growth; as such, the
increase in work continuation of mothers in long-term employment is
a welcome change. This chapter will review developments concerning

* The author would like to thank Japan’s Statistics Office for providing the Labour
Force Survey microdata. The data was applied for and used based on funding
provided by the Japan Society for the Promotion of Science (JSPS) Grant in Aid
for Scientific Research (C) No. 15K03503 for a project entitled “East Asia’s
Declining Birthrates, Institutions to Match Spouses, and Labor Markets”
(Project Leader: Nobuko Nagase). I would also like to thank the Ministry of
Health, Labor, and Welfare for providing the Basic Survey on Wage Structure
microdata. The data was applied for and used based on funding provided by the
Joint Usage and Research Center Programs of the Institute of Economic
Research, Hitotsubashi University, “Student Loan Burden on Youth
Development and Family Formation” (Project Leader: Nobuko Nagase). I thank
Dr Takeo Hoshi, Dr Lorrain Dearden, Dr Hiroshi Ono, and participants at the
Seminar at Stanford University in 2018 and 2019 and the Japan Economy
Network at the Bank of Japan in 2019 for their helpful comments. I thank
Dr Elaine McCrate for her helpful comments and her warm support.

310
Abe’s Womenomics Policy 311

Japanese female labor, placing Abe’s Womanomics policy as part of

larger policy developments. It will then closely examine the gender
promotion gap, by employing two governmental micro data sources,
the Labor Force Survey and the Basic Survey on Wage Structure, to
see whether Abe’s policy had any significant effect in increasing
women in management.
For the readers who are not familiar with Japanese labor market, this
chapter will first provide an overview of traditional Japanese employ-
ment practice and its effect on the gender gap in the workplace.
Secondly, it will review various measures taken from 1980s up to
today by the government to close the gap and their effects. Thirdly,
the paper will examine developments in the earnings distribution by
gender from the late 1990s to the 2010s: though the gap is closing, data
still show a large gap. Finally, this chapter will discuss in more detail
Abe’s new policy to increase the number of women in management in
Japan. In particular I focus on a major law, Josei Katsuyaku Suisin Hō
(or The Act on Promotion of Women’s Participation and Advancement
in the Workplace), proposed in 2014 and passed in the summer of
2015, that mandates enterprises with more than 300 employees to take
action to actively use female employees in management, although the
mandate is rather weak. The chapter will then investigate whether the
period of the Abe administration has had a significant impact on the
number of women being promoted to management levels.

11.2 Background: Traditional Japanese Employment Practice

and the Equal Employment Opportunity Law
The gender gap is large in Japan. The World Economic Forum
announced in 2017 November that Japan ranked 114th among 144
countries in the gender gap index. With respect to economics, a large
gender wage gap and a low percentage of females in professional,
engineering, and management professions contributed to the low
ranking.
In this section, I will review the development of female labor in
Japan. In 1985, the Equal Employment Opportunity Law (EEOL)
passed. However, despite the law, wage regressions have shown that
the coefficient gap between the genders increased between 1984 to
1994 at large firms, even though the endowment difference narrowed
(Hori 1998; Abe 2010, 2011).
312 Nobuko Nagase

Before the passage of the EEOL, large firms selectively employed

university graduate males to be on a male-only fast promotion track,
trained them as future managers, and promoted them to chief. Often
then they became section heads and then department managers
through tournament competition. In contrast, females were only
hired in the slow track and given clerical assistant jobs, with the
expectation that women would retire at marriage or childbirth. After
the passage of the EEOL in 1985, firms could no longer explicitly
discriminate in hiring by sex, but in fact they maintained similar hiring
and training practices. They formally renamed the employment track
that was previously only for university graduate males as the “fast track
course” while also allowing females to apply, and they renamed the
female hiring track as the “slow track course.” Only a small number of
females were employed along with males in the newly created “fast
track course” after the passage of the EEOL, and the slow track course
remained dominated by females. Studies found that many of those
women in the fast track course quit work after several years – for
example, Ohuchi (1999).
Still about half of firms with more than 5,000 employees today have
a course-based hiring practice. This ratio has remained stable accord-
ing to Ministry of Health, Labor, and Welfare’s Basic Survey of Gender
Equality in Employment Management. The percentage of firms that
have such a practice is increasing among firms with 300 to 1000
employees, although such practice is most often found in larger
companies.
In June 2008, the Ministry of Health, Labor, and Welfare established
a committee on “The Gender Wage Gap under Changing Wage and
Labor Practice.” After the Asian financial crisis of 1998, Japanese
employment practice changed substantially. Hiring of long-term
employees lessened and hiring of fixed-term employees increased.
Females in the slow track course were often replaced by fixed-term
employees and dispatched workers, especially in larger firms. The
Japan Institute of Labor Policy and Training (JILPT) in cooperation
with the committee conducted a wage regression of full-time workers
using the Basic Survey on Wage Structure of 2000 and 2006. It found
that the gender wage gap is larger when firms have “course based
hiring” (JILPT 2009; Fujii 2009) and that the wage gap is closing
mainly due to the shrinking gender age gap among long-term employ-
ees. This in turn happened because more women stayed single and
Abe’s Womenomics Policy 313

worked as long-term employees longer, even though many still quit

when they had children. The analysis also showed that females are
much less likely to become managers, and this contributed to the
gender wage gap (Fujii 2009).
In recent years, the distribution of the gender wage gap has gained
renewed attention internationally. Whether the gap is glass ceiling or
sticky floor has been newly investigated by using quantile regression
on Swedish and American data (Albrecht et al. 2003). Hara (2018)
conducted a similar distributional gender wage gap analysis in Japan
using selected data on full-time employees from 1980 to 2015 using
the Basic Survey on Wage Structure. Using the Firpo-Fortin-Lemieux
decomposition (Firpo et al. 2009), Hara found that the raw gender
wage gap generally decreased from 1980 to 2015 at most points in the
wage distribution, especially in recent years, except at the top where
the gap grew substantially after 2010. She also showed that the gap
was larger in the lowest part of the wage distribution among large firms
presumably due to course-based hiring practices, while the gap was
larger in the highest part of the distribution at smaller firms in 2015.

11.3 Recent Measures to Close the Gender Wage Gap in Japan

The government has sought to close this large gender gap by three
methods. One is to change the incentives and constraints faced by
women when deciding whether to stay in long-term labor contracts
after marriage and childbirth. In Japan, the wage prospects of long-
term employees compared to hourly part-time, fixed-term contract and
other non-seishain employees are very different, as will be shown in
Section 11.6. The prospect of higher wages in the latter sector is poor.
Long-term employees are given more training and have a higher prob-
ability of promotion. Therefore, promoting work continuation in
seishain positions at marriage and at childbearing is a meaningful
measure for closing the gender wage gap. Measures of this sort included
increasing the income replacement rate of the childcare leave allow-
ance, allowing seishain mothers to return to their jobs on a part-time
basis after childbirth upon their request, reducing work hours overall,
introducing five days a year of sick leave for the care of pre-school
children, and reforming the childcare system. Some of these measures
were taken before the accession of the Abe administration in late 2012
and some during his administration.
314 Nobuko Nagase

The second government strategy is to change the incentives of firms

to promote women to management, through a series of nudges to
change corporate culture. Several measures of this kind have been
adopted during the Abe administration. The government has aimed
to nudge firms to hire and promote more females through statutory
requirements to disclose their gender gap data, such as the gender ratio
in management and the gender gap in hiring. This may negatively
impact the recruitment of new female hires at companies with poor
records. In addition, the government has argued that firms in general do
better with diversified management boards.
Third, the government also sought to increase the wage of nonstand-
ard workers, who are poorly paid and highly disproportionately
female. There are two policies of this sort. The first aims to reduce
the incentive of a spouse to keep their income low, by attempting to
change spousal tax deduction rules and to change firms’ dependent
spouse allowance linked to tax deduction rules, as will be described
later. The second introduces the new equal-work/equal-wage principle
irrespective of job contract status.
Early measures taken to encourage women to continue work after
childbirth included the introduction of childcare leave in 1992, the
introduction of a childcare leave allowance in 1995, and an increase in
the leave allowance from 25 percent to 40 percent of base pay in 2001.
However, the introduction of childcare leave increased women’s work
continuation after marriage but had no effect after childbirth (Nagase
1999). Nor did raising the childcare leave allowance of 2001 produce
any change in behavior (Asai 2015). According to my interviews, many
women felt that it was difficult to take the leave unless their bosses,
colleagues, and husbands were supportive. However, corporate culture
was not supportive enough, and husbands came home late.2
This effort to retain women in the long-term labor market did not
succeed for some time. Other problems remained in corporate culture.
Under Japanese employment practice, large firms internally train new
hires in different sections of a firm by relocation, evaluation, and gradual
promotion to chief, then to section head. Since firms invest in workers’
firm-specific training, hoarding occurs during economic downturns, and
overtime occurs during economic booms, compared to more frequent
layoffs and hiring among US firms (Houseman & Abraham 1993;
Abraham & Houseman 1989). Because of the higher firm-specific train-
ing, returns to tenure are higher compared to those in the USA
Abe’s Womenomics Policy 315

(Hashimoto & Raisian 1985, 1992). However, overtime work and

sudden orders by the company to relocate are hard to reconcile with
family needs. Consequently, many female workers resigned at marriage
in the 1980s and then at pregnancy in the 1990s and 2000s. Some wished
to take care of their own children, but others felt corporate culture made
it difficult to take childcare leave (Teramura 2014; Enmi & Brinton
2015). Once women quit their “good jobs” – that is, the seishain jobs
that they had begun upon graduation – they often reentered the labor
force as part-time nonstandard employees at a wage near the minimum
wage. However, the National Fertility Survey of 2002, 2005, and 2010
revealed that there was almost no increase in the average employment
rate of mothers across five-year birth cohorts of their infants, observed
when the first child was one year old. It was 26.8 percent in 2005 to
2009, not much higher than the 24.4 percent in 1990 to 1994.
Eventually, employment of seishain mothers did start to go up around
2009 and then accelerated after 2013. The short hour option mandate
of 2009 – which allowed workers with a child under three to work six
hours a day, applied to firms with more than 100 employees from 2010
and to all firms from 2012 – was key to the rising employment of
seishain mothers. By using time differences in enactment of the law as
a natural experiment, Nagase (2014, 2017) found that the law had
a significant and powerful effect on both childbirth and labor supply.
The National Fertility Survey of 2015 also showed that new mothers’
employment rate when their first child was age one was 29.1 percent for
those who had a child in 2005–2009 but rose to 38.3 percent among
women with first births in 2010 to 2014. However, Takeishi (2010) and
Matsubara (2012) argue that the short hour option may negatively
affect female careers. In part because of this, increasing women’s pro-
motion is another important feature of Womenomics.
Formal infant care provision was also helpful for maternal work con-
tinuation in general (Nagase 2003a; Unayama 2011; Nagase 2018a) – or at
least helpful for work continuation among mothers in nuclear families
(Asai et al. 2015) – but the provision was low in metropolitan areas and
higher in rural areas. The Noda administration in 2012 prepared and
enforced a significant reform concerning both day care and kindergarten,
better coordinating their regulation. Before this reform, the development of
childcare slots in urban areas was slow until 2008,3 despite the series of
“Angel Plans” in 1990s (Nagase 2007) and remained slow till 2012
(Nagase 2018a).
316 Nobuko Nagase

In the Japanese type of internal training system, employees are often

trained to become managers. However, according to interviews, even if
women were employed in the fast track course right after the EEOL reform,
they were more likely to be placed in clerical jobs than males (Ohuchi
1999). Such different allocation of jobs was reduced after some years, but
bosses still often gave women routine tasks, especially after childbirth,
while males were given more challenging assignments (Yamaya &
Nagase 2012; Nagase & Yamaya 2011). In order to be promoted to
manager, internal valuation accumulation points should exceed those of
coworkers. Many firms had grading systems that gave low evaluations to
childcare leave takers and short hour workers. Some of our interviewees
reported that at their firm, when a woman had a child, her promotion
would come much later than that of comparable males who have no need
to take childcare leave. A survey of seishain mothers by 21seiki Shokugyo
Zaidan (2014) found that, on average, the respondents felt that their
bosses’ provision of training and feedback declined after they took child-
care leave.

11.4 Abe’s Policy Concerning Female Promotion and Closing

the Wage Gap
Starting in 2013, the Abe administration continued to expand policies to
encourage women to continue work as seishain. With strong leadership,
it rapidly increased the number of childcare facilities, and the growth of
childcare slots accelerated after a reform of childcare in 2012. The
administration again increased the childcare leave allowance in the first
six months after birth from 50 percent to 67 percent of base pay.
In Section 11.4.1, I will discuss laws that preceded Abe administration
and then the new legislation that passed during the Abe administration to
encourage firms to promote women to management. This is primarily the
Josei Katsuyaku Suishin Hō of 2015. I will then discuss another import-
ant policy concerning women’s wage – that is, policy to close the wage
gap between standard and nonstandard workers – in Section 11.4.2.

11.4.1 Laws That Nudge Firms to Promote Women

to Management
Starting around 2003, the government began to focus attention on the
practices of firms that may discourage women’s work continuation.
Abe’s Womenomics Policy 317

The Law to Support the Next Generation of 2003 mandated that firms
with more than 300 employees assess their firm practices for compati-
bility with family caregiving responsibilities, starting in 2004 as meas-
urement for low birth rate. In addition, the goal of increasing the share
of women leaders to at least 30 percent by 2020 was adopted in 2003 as
the Gender Equality Head Office Decision. It was reaffirmed in
December 2010 as the Cabinet Decision, when the third Basic Plan
for Gender Equality was adopted.4 The “2020–30” target – achieve
a 30 percent target for females in authoritative positions by 2020 – is an
old policy that has been hard to achieve. The interim report showed
that actual progress was far from the target.5 In the fourth Gender
Equal Participation Basic Plan of 2015, the target was modified to, for
example, 25 percent women for the lowest managerial category,
kakarichō or section chief, at private enterprises, an increase from
their actual representation of 16.2 percent in 2014. Similarly, the
Plan established a target of 15 percent women for kachō, or section
heads, by 2020, up from 9.2 percent in 2014. Finally, it set a target
around 10 percent for buchō, or department manager, the highest rank
targeted, which had female representation of 6.0 percent in 2014. This
Basic Plan also set a target of about 5 percent, to be rapidly increasing to
10 percent, for women on boards of directors, while the actual percent-
age was 2.8 percent in 2014. These numbers are calculated from the
Basic Survey on Wage Structure, as the percentage of employees in
managerial positions among those who have employment contracts
without termination, working at firms with more than 100 employees.
It seems that even the lower target is still hard to achieve. The White
Paper on Gender Equality (2019) shows that, in 2018, women were
18.3 percent of kakarichō, 11.2 percent of kachō, and 6.6 percent of
buchō. This is progress, but it is still short of the revised target.
When Prime Minister Abe came to office in 2013, he announced that
he hoped firms listed on the stock exchange would each appoint at least
one woman on their boards.6 In 2014 June, the Cabinet Decision was
made to mandate all listed firms to disclose the number of female
members on their boards, effective in the new fiscal year starting in
April 2015. In August 2015, the Abe cabinet also passed the Act
Concerning the Promotion of Women’s Career Activities, Josei
Katsuyaku Suisin Hō, to be implemented in April 2016.
Josei Katsuyaku Suisin Hō nudges enterprises to take actions to
promote women. The companies must gather data on the ratio of
318 Nobuko Nagase

women among new hires, the gender gap in average years of tenure, the
overtime work hours of average workers, and the ratio of women
among managers. Other types of data, such as the number of women
and men taking childcare leave, must be analyzed, although firms are
not mandated to disclose more than one of them.7 Large firms are also
required to make action plans. Specifically, the law requires firms with
more than 300 employees to analyze the status of women within their
firms and then to make action plans to increase their active
participation.8 The action plans should have specific timelines, meas-
ures, and the actual targets to be reached. The plan should be
announced to workers and made public. It must also be registered at
the local MHLW ministry office.
Those firms with good performance are given the Eruboshi recogni-
tion, which consists of three grades. These firms can put the Eruboshi
logo on their products.9 In March 2016, the government announced
that the recognized firms would also be given preferred points in public
procurement.10 In May 2016, the government announced that 46 firms
got Eruboshi recognition. The number increased to 291 by
March 2017, 579 by March 2018, 837 by March 2019, and 1028 by
March 2020.

11.4.2 Other Laws That Aim to Change Japanese Labor

Practice and Close the Gender Gap
The more recent Womenomics policies of the Abe cabinet concern
three things. The first policy is putting a cap on the long overtime
hours among seishain workers. The second is closing the wage gap
between standard and nonstandard workers, such as so-called part-
time workers (arubaito), “dispatched workers,” “fixed term contract
workers” and “shokutaku.” These workers are disproportionately
female. The third is reform of the tax code. Regarding the first of
these, the overtime reform was made into law. From the beginning of
2016, the buzzword of the Abe cabinet was “Hatarakikata Kaikaku” –
“Change the Way We Work.” When Abe reorganized the cabinet in
2016 August, he said, “The greatest challenge we meet is the
Hatarakikata Kaikaku.” The Hatarakikata Kaikaku Hō was proposed
in 2016 and passed the Diet in 2018. Among its contents, it more
strictly regulates overtime work for large firms starting in April 2019
and one year later for smaller firms.
Abe’s Womenomics Policy 319

The Hatarakikata Kaikaku Hō also included guidelines for the prin-
ciple of “equal work, equal pay,” regardless of whether one has
a standard or nonstandard contract. The guidelines, announced in
late 2018, say that, in principle, if workers engage in the same work
and have equal capability, experience, and performance, the pay
should be equal, irrespective of contract types. It is not easy to define
what equal work is. Therefore, companies are to explain to workers the
rational reasons for differences in pay and benefits, as well as for
differences in training between standard and nonstandard employees
and to remedy any irrational differences. The prefectural labor offices
are to negotiate disputes between employer and employees. This prin-
ciple is to be implemented starting in April 2020 for larger firms and
one year later for smaller firms. The effect of this policy is yet to be seen.
Finally, Japanese tax policy has for a long time worked to support
nonstandard work arrangements that often were for married females,
who hoped to supplement the household budget. Scholars argued that
they were accepting low wage and short work hours because their work
motive was only to bring in some addition to the household budget
However, in the late 1990s to early 2000s, nonstandard work arrange-
ments increased among young workers who needed to support their own
living. This happened because of a decline in demand by employers for
long-term employment. Moreover, the number of single mothers
increased, who also needed to support their living but many of them
also worked in nonstandard employment. Among these nonstandard
workers, part-time and arubaito are hourly paid workers, whose work
hours can be either short or full-time and are paid near minimum wage.
Higuchi (1995), Nagase (2002, 2003b, 2018a, 2018b), Abe (2002),
Akabayashi (2006), and Abe and Ohishi (2009) point out that the protec-
tion for dependent housewives through the spousal tax deduction and
social security tax exemption is a core mechanism that limits the income
of married females. Married women increase their labor supply when
children grow older, but, when their annual income reaches 1.03 million
yen,11 the threshold for the tax deduction for dependent spouses, or reaches
1.30 million yen, the threshold for the social security tax exemption for
dependent spouses, many of them adjust their work hours to hold their
income below these thresholds, because otherwise the greater tax burden is
large compared to the increased income achieved by working longer hours.
Takahashi (2010) and Bessho et al. (2014) say that the effect of the
tax structure on labor supply is not large. However, the resulting
320 Nobuko Nagase

incentives for good workers to limit their hours reduces pressure for
higher wages in the part-time labor market. Through wage arbitration
between married part-time workers and unmarried workers, the con-
duct of the former, who benefit from such tax provisions, not only
negatively affects their own wage level but also that of unmarried part-
time and arubaito workers who do not benefit from keeping income
below the thresholds (Nagase 2018b).
The Abe cabinet repeatedly announced, in 2014 and 2016, that it
would change the rule concerning the spousal income tax deduction
threshold of married females. Abe also urged large companies to change
their own spousal allowance rules so that they would not affect married
females’ work hour choice.12 However, the Abe cabinet in the end did
not abolish the spousal tax deduction threshold. Instead, it surprisingly
increased the spousal tax deduction threshold to 1.5 million yen.
Theoretically this raises the total working hours at which unmarried
workers have to compete with married women who benefit more from
the national tax spousal deduction and social security tax exemption. It
will increase the work hours of married females who are willing to work
with low earnings, unless those earnings exceed the sum of their social
security tax exemption and their husbands’ spousal allowance.13 The
effect of this last policy is yet to be determined, as it was announced only
in late 2016 and implemented in 2018.
In subsequent sections, I will use Japanese micro data to illustrate
changes and the present situation of gender wage gap in Japan; I will
also show how female promotion differed by cohort and determine
whether Abe’s policy made any difference.

11.5 About the Data Used in this Analysis

The Labor Force Survey, collected by the Statistics Office, is
a nationally representative monthly survey covering about 40,000
households and about 100,000 individuals over the age of 15.14 The
earlier Labor Force Special Survey (Tokutei Chosa) collected detailed
information on educational attainment, income, and employment
before it was terminated in 2001. From 2002, the Labor Force Survey
incorporated questions previously included in the Labor Force Special
Survey. From January 2002, the survey began to use a longer question-
naire in the fourth visit that asked not only about labor force status but
also about educational attainment, job tenure, and annual income in
Abe’s Womenomics Policy 321

the previous year for all relevant individuals in the household. It

provides data on about 10,000 households, or 18,000 to 22,000 indi-
viduals every month. To look at long-term trends, the analysis of this
chapter uses microdata of the Labor Force Special Survey conducted
each February from 1988 to 2001 and data from the January to March
Labor Force Survey from 2002 to 2018 to make a connected time series
from 1988 to 2017.
Since the LFS series only has categorical data on the previous year’s
annual income, in twelve categories, the hourly wage rate is computed
as the midpoint of the annual income category divided by the product
of weekly work hours times fifty weeks.15 One can identify managers in
the LFS by using the variable on occupations.
I will also use another microdata set, the Ministry of Health, Labor
and Wealth’s Basic Survey on Wage Structure, 2002–2017. It is
a survey of firms with more than five employees conducted every July.
It collects data about salary, bonus, overtime pay, and work hours. The
sample includes private enterprises and workplaces with more than five
standard employees and public enterprises with more than ten standard
employees; the sample size is as large as 1 million or more per year. In
2005, the questionnaire was reformed to be more sensitive to the
increased number of nonstandard workers. Before 2005, it did not
ask full-time workers if they were seishain. From 2005, it revised its
questionnaire sheet and surveyed seishain workers and non-seishain
workers, including those with fixed- term contracts. It also inquires as
to whether one is working part-time. It has questions about managerial
status, but only for those who work at firms larger than 100 employees.

11.6 Change in the Gender Wage Ratio and Wage Distribution

Figure 11.1 shows the time series average gender wage ratio for work-
ers aged twenty-three to fifty-nine, displayed by educational category,
using the LFS. Until 2012, the standard retirement age was sixty, so
that, for the longer time series, excluding those sixty and over will more
closely standardize the comparison.16
“Gender wage ratio, total” represents the average ratio of women’s
to men’s wages among all workers, including both permanent contract
standard workers and nonstandard employees. This ratio is increasing
at the aggregate level. The average wage of females was around 50 per-
cent of that of males in 1988, but the figure increased to around
322 Nobuko Nagase

90

85

80

75

70

65

60

55

50
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
gender wage ratio, total
seishain, or permanent standard employees, total
seishain, high school
seishain, university
seishain, university, age 23–39

Figure 11.1 Female average wage as percent of male average wage, by

educational attainment, age 23–59 when not otherwise specified
Source: Author’s calculations using the Labor Force Special Survey 1988–2001
and Labor Force Survey 2002–2017 January–March

65 percent in 2017. When the figure is limited to those who have

permanent seishain contracts, the gender wage ratio is higher, and the
female wage, which was over 60 percent of the male wage in 1988, was
approximately 75 percent in 2017. Because more women than men
work as nonstandard workers, and because the average wage of non-
standard workers is low, the overall gender wage ratio is smaller when
considering all workers. Restricting the estimate to permanent stand-
ard contract workers thus produces a higher estimate of the gender
wage ratio.
Now consider the trends in the gender wage ratio by educa-
tional attainment for permanent contract standard workers. The
gender wage ratio was smaller in the late 1980s, especially among
workers with less education. For those with high school education
or less, the wage ratio was a little over 60 percent in 1988 and
around 70 percent in 2017. In contrast, the gender wage ratio
among university graduates who were seishain was higher
Abe’s Womenomics Policy 323

compared to the ratio among less educated workers in 1988. The

average wage of university-educated females was about 70 percent
of the corresponding male wage in 1988. However, the subsequent
improvement was small, with the ratio remaining a little over
70 percent in 2017. This may be because female university educa-
tional attainment is increasing faster than that of males, so that
the average age and average experience of female university gradu-
ates are lower overall. Therefore, I also calculated the ratio for
university graduate workers aged twenty-three to thirty-nine,
expecting an increase in gender wage ratio. Now the ratio is
even larger, around 80 percent of the male wage in 1988, but it
also stayed the same during this period, around 80 percent of the
male wage in 2017. Therefore, the stubborn gap among university
graduates cannot only be due to the average experience gap by
gender.
The figures show that the overall gender wage ratio among seishain
workers is increasing over time. And while it is larger among those with
university education, the gap is closing more slowly for this group than
for the less educated, among permanent-contract standard workers.
The gap may be caused by difference in promotion.
I will now show the kernel estimation of the wage distributions
of male and female workers in this section using selected years of
the Labor Force Survey, first for permanent contract standard
workers and then for nonstandard workers. Comparing Figures
11.2 and 11.3 – that is, the distribution of wages for permanent
contract standard workers, by sex – women’s wage distribution is
much narrower than men’s: men are more likely than women to
have an hourly wage rate over 3,000 yen. Figures 11.4 and 11.5
show that, for nonstandard workers, the wage distribution is con-
centrated at the lower end for both males and females, though
more so for females.
All the distributions are right-skewed. For both permanent con-
tract standard workers and nonstandard workers, the female wage
is moving to the right over time, and their wages are on average
increasing. On the other hand, male workers’ wage distribution is
moving to the left and deteriorating. Despite these gap closing
trends, the wage distribution difference between genders continues
to be very large. Difference in promotion may be one of the
324 Nobuko Nagase

.0001 .0002 .0003 .0004 .0005

Density of full-time permanent workers
0

0 2,000 4,000 6,000 8,000

Hourly wage rate in yen

Male 1996 Male 2003

Male 2012 Male 2017

Figure 11.2 Wage distribution of males for permanent contract standard

workers
.0006
Density of full-time permanent worker
.0002 0 .0004

0 2,000 4,000 6,000 8,000

Hourly wage rate in yen

Female 1996 Female 2003

Female 2012 Female 2017

Figure 11.3 Wage distribution of females for permanent contract standard

workers
Abe’s Womenomics Policy 325

.0006
Density of non-standard employees
.0002 .0004
0

0 2,000 4,000 6,000 8,000

Hourly wage rate in yen

Male 1996 Male 2003

Male 2012 Male 2017

Figure 11.4 Wage distribution of males for all nonstandard workers

Density of nonstandard employees
.0002 .0004 .0006 .0008
0 .001

0 2,000 4,000 6,000 8,000

Hourly wage rate in yen

Female 1996 Female 2003

Female 2012 Female 2017

Figure 11.5 Wage distribution of females for all nonstandard workers

326 Nobuko Nagase

reasons for the large gender wage gap, which I will examine in the
next section.

11.7 Change in the Gender Composition of Managers

11.7.1 Definition of Managers

In this section, I will analyze the promotion difference by gender in the
Japanese labor market. As mentioned in Section 11.4.1, the Japanese
Cabinet Office had explicit targets for women in managerial positions,
for example, to increase the female proportion of kachō, section heads,
to 15 percent by 2020.
We need, however, to be careful about the definition of managers
and thus about the choice of data to employ. Ohi (2004) extensively
reviewed the Japanese governmental statistics concerning workers in
managerial positions using publicly available data. She noted two
different definitions of managers in the Japanese governmental statis-
tics. According to one definition, workers’ occupations are classified as
“managers” if they manage subordinates. According to the other def-
inition, workers have a managerial title in the firm’s salary table but
they may not have subordinates. Ohi calls the former “managers in
narrow definition” and the latter “managers in broader definition.”
The Census and the Basic Employment Survey are household surveys
that use the narrow definition of managers. Ohi observes that the Basic
Employment Survey classifies about 500,000 persons in total as man-
agers; for firms with more than 100 employees, the number is 300,000
in 2002, or 2.0 percent of workers. In contrast, the Basic Survey on
Wage Structure collects data from firms, and, using the broad defin-
ition, it enumerates those who are titled as buchō (department head),
kachō (section head), kakarichō (section chief), shokuchō (foremen),
and sonota shokkai (other managers), among workers at firms with
more than 100 employees. Ohi shows from the public data that, in this
definition, managers altogether comprised 23.4 percent of all workers
in firms with more than 100 employees in 2004, or 3.40 million
employees.
The government uses both definitions when it employs targets to
increase the proportion of female managers. The Cabinet Office uses
the narrow definition in the Labor Force Survey to compare the gender
gap in management internationally. The Cabinet Office also uses the
Abe’s Womenomics Policy 327

Basic Survey on Wage Structure to examine the increase in the propor-

tion of all managers who are women. However, their target is just the
overall proportion of managers who are women. In the following
section, I will look at the percentage of workers who are managers,
separately by gender.

11.7.2 Change in the Ratio of Managers by Gender Using

Different Definitions
Figure 11.6 was made by the author using the Labor Force Survey.
While the governmental target only specifies the percentage of all
managers who are female, I will examine different indicators – that is,
the ratio of female managers among female workers and the ratio of
male managers among male workers. In my view, it is important to

8.0%

7.0%

6.0%

5.0%

4.0%

3.0%

2.0%

1.0%

0.0%
1988
1989
1990
1991
1992
1993
1994
1995
1996
1997
1998
1999
2000
2001
2002
2003
2004
2005
2006
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017

All workers, male All workers, female

Firms over 100 employees, male Firms over 100 employees, female
Firms over 500 employees, male Firms over 500 employees, female

Figure 11.6 Ratio of managers in narrow definition among employees aged

22–59 by gender and firm size
Source: Made by the author from Labor Force Survey Series
328 Nobuko Nagase

compare the proportions of managers by gender, not just the proportion of

all managers who are women, because the number of male workers is
larger than the number of females. Ohi shows that the overall proportion
of managers in private enterprises with more than 100 employees using the
Basic Employment Survey (using the narrow definition) was 2.0 percent in
2002. The proportion for the same population using the Labor Force
Survey (also using the narrow definition) was a comparable 2.5 percent
in 2002 by my calculation. The two statistics differ, but the ratio of
managers is similar because the definition of managers are both based on
the Statistics Office’s Japanese occupational standard classification. I show
a longer time series after Ohi’s study. Figure 11.6 shows the same trend as
found in Ohi: the overall ratio of managers decreased from early 1990s to
1998. Then, as shown in triangle marks, the ratio of male managers
among male employees went up after 2000, plateaued in the mid-2000s,
then declined continuously after then. The decline was more notable for
larger firms with more than 500 employees. The ratio of male managers to
male workers was once over 7 percent, but it was below 4 percent by 2017.
On the other hand, female managers among female workers, presented in
circle marks, show that the ratio is just very low and unchanged, regardless
of the firm size. Figure 11.6 also shows the ratio of managers of each
gender by controlling firm size to over 100 employees to compare with
broader definition shown in next two tables, which show a similar trend.
Table 11.1 was made by the author using the Basic Survey on Wage
Structure, showing ratio of managers in the “broader category,” as Ohi
puts it. The definition of managers in this data is whether one is given
managerial title in salary payment. It includes those who does not have
any subordinates. The Basic Survey on Wage Structure only asks about
managers for firms with more than 100 employees. Therefore, I divided
the number of male and female managers by the number of male and
female workers, respectively, at firms with more than 100 employees.
We see that the ratio of managers – that is, including all department
heads, kachō, kakarichō, foremen, and other managers – declined for
males from 27.9 percent in 2002 to a low of 23.5 percent in 2017, while
for females, the bottom was 3.2 percent in 200 and went up to 4.9 per-
cent in 2017. The decline in ratio of males with managerial titles must
have contributed to the sluggish male income after 2000s. Though ratio
of male managers increased somewhat after 2014, it was still lower in
2017 than during the Lehman shock of 2008. In contrast, for females,
an increase is evident from around 2013. We see that the increase in
Table 11.1 Ratio of managers in broader definition (denominator, all workers)

Department heads, Section heads, or Section chiefs, or

All managers or bucho kacho kakaricho Other managers

male female male Female male female male female male female

2002 27.9% 4.0% 3.4% 0.1% 7.4% 0.6% 6.6% 1.3% 8.1% 1.7%
2003 27.7% 4.2% 3.4% 0.1% 7.5% 0.6% 6.6% 1.3% 7.8% 1.8%
2004 25.5% 3.2% 3.2% 0.1% 7.4% 0.6% 5.7% 1.1% 7.2% 1.2%
2005 26.1% 3.3% 3.6% 0.2% 8.0% 0.6% 6.7% 1.2% 6.0% 1.1%
2006 27.0% 3.8% 4.0% 0.2% 7.8% 0.8% 6.7% 1.3% 6.4% 1.4%
2007 26.1% 3.9% 3.4% 0.2% 7.8% 0.8% 6.3% 1.4% 6.3% 1.3%
2008 25.2% 3.8% 3.3% 0.2% 7.5% 0.8% 6.2% 1.3% 6.2% 1.4%
2009 24.1% 4.0% 3.4% 0.2% 7.2% 0.8% 5.8% 1.3% 5.8% 1.5%
2010 24.6% 3.9% 3.5% 0.2% 7.9% 0.8% 6.3% 1.4% 5.1% 1.3%
2011 23.7% 4.1% 3.4% 0.3% 7.1% 0.9% 5.9% 1.5% 5.3% 1.3%
2012 24.1% 4.0% 3.8% 0.3% 7.1% 0.8% 5.5% 1.3% 5.7% 1.5%
2013 23.7% 4.3% 3.3% 0.2% 7.1% 0.9% 5.8% 1.5% 5.8% 1.5%
2014 23.8% 4.6% 3.1% 0.3% 7.3% 1.0% 5.8% 1.5% 5.8% 1.6%
2015 24.5% 4.6% 3.4% 0.3% 7.5% 1.1% 6.2% 1.7% 5.7% 1.5%
2016 24.3% 4.9% 3.2% 0.3% 7.6% 1.1% 6.2% 1.9% 5.7% 1.5%
2017 23.5% 4.9% 3.1% 0.3% 7.2% 1.1% 5.9% 1.7% 5.7% 1.7%

Source: The Basic Survey on Wage Structure, author’s calculation

330 Nobuko Nagase

female managers, especially for the lowest category of kakarichō,

started around 2010 and for section heads, or kachō, around 2014.
However, the ratio shows a different picture if we limit the denom-
inator to workers with permanent employment contracts as in Table
11.2.17 Now when we look at the ratio of all managers for males, the
first left column in Table 11.2, we do not see any decline in the ratio of
male managers among permanent employment contract workers: it
was 28.4 percent in 2002 and 29.4 percent in 2017.
The difference comes from the increase of fixed-term contract
workers over years. The percentage of permanent employment con-
tract workers among all workers at firms with more than 100
employees was 86.3 percent for males and 69.6 percent for females
in 2005. For males, that ratio went down to 77.6 percent in 2017 and
for females to 52.3 percent. When we compare the result of ratio of
female managers among all workers with that among permanent
employment contract workers, the ratio nearly doubles for females
in 2017, 4.9 percent vis-à-vis 9.1 percent. For females, the increase in
the ratio of managers is more evident among permanent contract
workers but not as much when fixed-term contract workers are
included. Table 11.2 shows that the ratio was 4.4 percent in 2002,
while it rose to 9.1 percent in 2017. The proportion of female
managers for the lowest category of kakarichō or section chief started
around 2010 and for kachō or section heads around 2014, the same
trend shown in Table 11.1.
Due to the continued rise over years of nonstandard employment
among males, the ratio of all males who are managers between Table
11.1 and Table 11.2 differs only by 0.5 percentage points in 2002 but
by 6 percentage points in 2017. For the ratio of male kachō or section
heads, for example, the difference was only 0.1 percentage points in
2002 but 2 percentage point in 2017, meaning that in 2002 section
heads managed mostly males who had permanent contracts, but in
2017 section heads had more male fixed-term contract workers to
manage (see Figure 11.7).
Takeuchi (1995) investigated the labor practice at a large financial
firm concerning the promotion of male entrants of the firm in the
university graduation year of 1966. He showed that 99 percent of
male university graduates of 1966 were promoted to section chief
exactly at the fifth year of tenure. Then the tournament promotion to
section heads was made from the age of thirty-four among section
Table 11.2 Ratio of managers in broader definition by gender (denominator permanent full-time workers)

Department heads, Section heads, or Section chiefs, or

All managers or bucho kacho kakaricho Other managers

male female male Female male female male female male female

2002 28.4% 4.4% 3.5% 0.1% 7.5% 0.6% 6.7% 1.5% 8.2% 1.8%
2003 28.3% 4.6% 3.4% 0.2% 7.6% 0.7% 6.8% 1.5% 8.0% 1.9%
2004 26.4% 3.6% 3.3% 0.2% 7.6% 0.7% 5.9% 1.3% 7.5% 1.3%
2005 29.7% 5.3% 4.0% 0.3% 9.1% 1.1% 7.7% 2.0% 6.7% 1.7%
2006 30.8% 6.3% 4.1% 0.4% 9.0% 1.2% 7.8% 2.2% 7.4% 2.2%
2007 30.3% 6.6% 4.0% 0.4% 9.1% 1.4% 7.3% 2.3% 7.3% 2.2%
2008 29.6% 6.7% 3.8% 0.4% 8.9% 1.4% 7.4% 2.3% 7.2% 2.4%
2009 28.6% 6.9% 4.0% 0.4% 8.5% 1.4% 7.0% 2.3% 6.8% 2.5%
2010 29.3% 7.0% 4.1% 0.4% 9.4% 1.5% 7.6% 2.6% 6.1% 2.2%
2011 28.4% 7.2% 3.9% 0.4% 8.6% 1.6% 7.2% 2.7% 6.2% 2.2%
2012 29.7% 7.5% 4.7% 0.5% 8.8% 1.6% 6.9% 2.5% 7.0% 2.7%
2013 29.1% 8.2% 4.0% 0.5% 8.8% 1.7% 7.3% 2.8% 7.0% 2.9%
2014 29.7% 8.6% 3.8% 0.5% 9.1% 1.9% 7.3% 2.9% 7.2% 3.1%
2015 30.4% 8.7% 4.1% 0.5% 9.3% 2.0% 7.8% 3.2% 7.0% 2.7%
2016 30.5% 9.0% 4.0% 0.5% 9.6% 2.1% 7.8% 3.5% 7.0% 2.7%
2017 29.4% 9.1% 3.7% 0.5% 9.1% 2.1% 7.5% 3.2% 7.0% 3.0%

Source: The Basic Survey on Wage Structure, author’s calculation

332 Nobuko Nagase

35.0%

30.0%

25.0%

20.0%

15.0%

10.0%

5.0%

0.0%
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017

All managers among permanent contract workers, male

All managers among permanent contract workers, female
All managers among all workers, male
All managers among all workers, female

Figure 11.7 Gender ratio of managers among all workers and among
permanent contract workers including non-seishain
Source: The Basic Survey on Wage Structure, author’s calculation

chiefs, but with more variation in promotion age. Among section

heads, promotion to vice department heads started in their forties.
This is very old and anecdotal evidence for the entrants of the class of
1966 at a large financial firm. Labor practice may differ by industry.
This type of age-based promotion, however, was also depicted by
Imada & Hirata (1995) in their studies of the 1980s.
To see how such age-based promotion has changed, I looked at
promotion using the Basic Survey on Wage Structure 2002–2017 for
the birth cohort of 1965–1969 and for 1980–1984, for whom I can see
the longest period of promotion by age groups. I added up the ratio of
those in different managerial position for both birth cohorts. For the
former group, they graduated from university around 1987–1994,
about twenty years later than those in Takeuchi’s study.
Figure 11.8 (A) left side shows for the 1965–1969 birth cohort males
and the right side for females. The denominator is the data for those
who are employed at firms with more than 1,000 employees and for
university graduates who are permanent contract employees following
Abe’s Womenomics Policy 333

Takeuchi (1995). Figure 11.8 (B) presents the same information for the
1980–1984 birth cohort, who would have graduated from university
around 2002–2008.
Figure 11.8 (A)’s left-side figure of male promotion shows that the
ratio of higher-rank managers such as section heads and depart-
ment heads increases with age as found in Takeuchi. We see that
the ratio of section chiefs decreases for males with age as more
males are promoted to higher positions. The right-side figure of
female promotion finds that only 25 percent were any kind of
manager by age forty to forty-four, and only 35 percent by age
fifty to fifty-four. Nearly 70 percent of female university graduates
in permanent employment were in nonmanagerial positions in their
early fifties. On the other hand, the comparable figures were
60 percent and 63 percent for male university graduates. Less

Birth cohort 1965–1969, university graduates working at large firms

Males Females

70% 70%

60% 60%

50% 50%

40% 40%

30% 30%

20% 20%

10%
10%
0%
0%
30–34 35–39 40–44 45–49 50–54
30–34 35–39 40–44 45–49 50–54

Department heads (buchō) Department heads (buchō)

Section heads (kachō) Section heads (kachō)
Other managers Other managers
Section chiefs (kakarichō) Section chiefs (kakarichō)

Figure 11.8 (A) Ratio of those in different managerial positions by age group,
birth cohort of 1965–1969, university graduates at large firms
Note: Among workers without fixed term of contract, including non-seishain
employees, by age and birth cohort, in firms with more than 1,000 employees.
Source: The Basic Survey on Wage Structure, author’s calculation
334 Nobuko Nagase

than 40 percent of males in their early fifties were in nonmanager-

ial positions. Unlike men, the ratio of section chief does not
decrease with age for women, presumably since promotion to
section chief occurs at varied timing and at later ages for women
compared with men. Note that the birth cohort of 1965–1969 is
the cohort that entered firms after the enactment of the EEOL;
however, there are still very large differences in promotion by
gender.
Figure 11.8 (B) shows the promotion history of the younger birth
cohort of 1980–1984. By age thirty-five to thirty-nine, more than
30 percent of university graduate males were managers of some kind,
but the ratio was only 16 percent for university graduate females at age
thirty-five to thirty-nine. The promotion speed was twice as fast for

Birth cohort 1980–1984, university graduates working at large firms

Males Females

35% 35%

30% 30%

25% 25%

20% 20%

15% 15%

10% 10%

5% 5%

0% 0%
25–29 30–34 35–39 25–29 30–34 35–39

Department heads (buchō) Department heads (buchō)

Directors (kachō) Directors (kachō)
Other managers Other managers
Section chiefs (kakaricho) Section chiefs (kakarichō)

Figure 11.8 (B) Ratio of those in different managerial positions by age group,
birth cohort of 1980–1984, university graduates at large firms
Note: Among workers without fixed term of contract, including non-seishain
employees, by age and birth cohort, in firms with more than 1,000 employees.
Source: The Basic Survey on Wage Structure, author’s calculation
Abe’s Womenomics Policy 335

males for the birth cohort of 1980–1984, compared with females. The
promotion gap in gender did narrow for the birth cohort of 1980–1984
compared with that of 1965–1969, not because women’s promotion
sped up but because men’s promotion became slower.
However, changes in policies and labor practice over time also
increased female promotion among younger birth cohorts. Figure 11.9
shows percentage of those in any managerial positions by age and by
birth cohort among large firms but including all educational groups, not
limited to university graduates. We clearly see from Figure 11.9 that,
while the managerial ratio of males at age thirty-five to thirty-nine is
lower for the 1980–1984 birth cohort than the 1965–1969 cohort,
falling from 41 percent to 32 percent, the female ratio at age thirty-five
to thirty-nine increased from 7 percent to 11 percent. The 1980–1984

Male Female
0.6 0.6

0.5 0.5

0.4 0.4

0.3 0.3
0.2 0.2
0.1 0.1
0 0
30 9
35 4

40 9
45 4

60 9
4
50 9

55 4

30 9
4

4
9

45 4
9

9
–2
–3
–3

–4

–5
–6
–4
–5

–2
–3

–6
–3

–4

–4

–5

–5
25

25

60
35

40

50

55

Age group Age group

birth cohorts birth cohorts
∼1944 1945–1949 ∼1944 1945–1949
1950–1954 1955–1959 1950–1954 1955–1959
1960–1964 1965–1969 1960–1964 1965–1969
1970–1974 1975–1979 1970–1974 1975–1979
1980–1984 1985–1989 1980–1984 1985–1989

Figure 11.9 Percentage of workers in any managerial position at large firms by

gender and by birth cohort
Note: Among workers without fixed term of contract, including non-seishain
employees, by age and birth cohort, in firms with more than 1,000 employees.
Source: The Basic Survey on Wage Structure, author’s calculation
336 Nobuko Nagase

birth cohort was in the twenty-nine to thirty-three age group in 2013,

and thirty-two to thirty-six in 2016 when Josei Katsuyaku Suisin Hō was
enacted, so they may be the target generation for promotion during the
Abe administration. We also see that promotion of women was very
slow up to birth cohort of 1955–59, who were the generation hired
before EEOL, but sped up in the age group of 45–49 after then.

11.7.3 Differences in the Ratio of Managers at

Policy-Affected Firms
Figures 11.10 and 11.11 let us explore whether there is any observ-
able change in the proportion of females in managerial positions at
times when the government nudged firms to change their practices
concerning women employees. The first of these statutory nudges,
the Law to Support the Next Generation of 2003, mandated that
firms with more than 300 employees assess their firm practices
for compatibility with family caregiving responsibilities starting
in 2004. Abe initiated the discussion of the second such nudge in
2013. It was adopted as Josei Katsuyaku Suishin Hō, the law in
2015 and implemented in 2016, also mandated for firms with more
than 300 employees. Figure 11.10 depicts the proportion of men
and women who attained section chief or kakarichō, the lowest

16.0%
14.0%
12.0%
10.0%
8.0%
6.0%
4.0%
2.0%
0.0%
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
Male, firmsize Female, firmsize Male, firmsize Female, firmsize
300–499 300–499 100–299 100–299

Figure 11.10 Ratio of section chief among permanent contract employees by

gender, comparison of those working at firms with 100–299 employees to
those with 300–499
Note: Age twenty-seven to forty-three, permanent contract employees only.
Source: The Basic Survey on Wage Structure, author’s calculation
Abe’s Womenomics Policy 337

45%

40%

35%

30%

25%

20%

15%

10%

5%

0%
2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017
Male, firmsize Female, firmsize Male, firmsize Female, firmsize
300–499 300–499 100–299 100–299

Figure 11.11 Ratio of section heads among low ranking managers by

gender, comparison of those working at firms with 100–299 employees
to those with 300–499
Note: Age twenty-seven to fifty-four, permanent contract employees only.
Source: The Basic Survey on Wage Structure, author’s calculation

managerial position, by firm size. The denominator is the number

of permanent employees, since I assumed the change would be
most evident among long-term employees as most human resource
departments target their policy toward such workers. For longest
time series, permanent employee status is only available, which is
close to the seishain definition. The change should also be more
evident at the lowest managerial position. However, when differ-
entiating between firms with 300–499 employees and those with
100–299 employees, the changes are not much different by firm
size. We do see some increase in promotion for females in 2004
and 2013, but we also see a similar trend for males. Moreover, the
change in Figure 11.10 after 2003 only raised the percentages one
time, after which the trend flattened. Casual review of Figure 11.10
does not seem to identify a significant effect of the law.
Figure 11.11 illustrates the corresponding data for section heads,
among those who are either section chief or section heads. Here, the impact
for year 2015 to 2016 among females is more notable, and the law could
have impacted female promotion to section heads among section chiefs.
338 Nobuko Nagase

11.8 Method of Estimation

My central question for this chapter is whether women had a greater
probability of attaining managerial employment during the Abe adminis-
tration. I will also ask whether females are treated differently compared
with males in promotion. I will use the broad definition of managers in the
Basic Survey on Wage Structure. The broad definition of managers is
more connected to a firm’s wage table than in managing subordinates, so
that the promotion difference may give an explanation to the large gender
wage gap. My dependent variable, Manager, equals 1 if given any type of
managerial position, and Manager equals 0 if not. I will estimate two
probit equations for this dependent variable, Equations 1 and 2.
Manageri ¼ a1 þ b1 Educationi þ b2 female  Educationi
þ c1 Agei þ c2 female  Agei þ d1 Tenurei þ d2 female  Tenurei
þ e1 female þ f1 ðyears of policy 2013 to 2017Þ
þ f2 ðyears of policy 2016 to 2017Þ
þ f3 female  ðyears of policy 2013 to 2017Þ
þ f4 female  ðyears of policy 2016 to 2017Þ
þ g2 firmsizei þ h1 seishaini þ h2 female  seishaini
þ i1 year dummy þ ui
ð1Þ
In Equation (1), I have a policy period dummy for the years 2013 to
2017, interacted with gender, to see if women had a greater probability of
becoming managers during the Abe administration. I also include a 2016–
2017 year dummy interacted with gender, as this is when Josei Katsuyaku
Suisin Hō came into effect. The Josei Katsuyaku Suisin Hō was passed in
2015 and implemented from April 2016. For this reason, I added the year
dummy 2016–2017 and crossed it with the female dummy. If their
coefficients, f3 and f4, are significantly positive, it means that, during the
Abe administration, the probability of a female being a manager
increased, especially after implementation of the law. Gender, and the
two periods of the Abe administration, are also tested for direct effects.
Other coefficients are also important. We know from previous case
studies of promotion policies in various large firms (Imada & Hirata
1995; Takeuchi 1995; Hanada 1987) that being male, accumulating
tenure, reaching a certain age, and being a university graduate are
important for attaining managerial positions. Being seishain is
a prerequisite. Then lower-level managers compete for the next highest
Abe’s Womenomics Policy 339

rank. Thus I will also see whether educational degrees, age, a seishain
contract, and tenure affect males and females differently. Finally, I will
control for individual years in this period, firm size, and year.
Manageri ¼ a1 þ b1 Educationi þ b2 female  Educationi þ
c1 Agei þ c2 female  Agei þ d1 Tenurei þ d2 female
Tenurei þ e1 female þ f1 ðyears of policy 2013 to 2017Þ
þ f2 female  ðyears of policy 2013 to 2017Þ
þ g1 ðfirmsize100  299Þ þ g2 ðfirmsize100  299Þ  femalei
þ g3 ðfirmsize100  299Þ  ðyears of policy 2013 to 2017Þ
þ g4 ðfirmsize100  299Þ  ðyears of policy 2013 to 2017Þ  femalei
þ h1 seishaini þ h2 female  seishaini þ i1 year dumy þ ui
ð2Þ
In Equation (2), I use variation in firm size to see if Abe’s policy Josei
Katsuyaku Suisin Hō had a statistically significant effect on increasing the
probability of female managers at targeted firms. The term that tests for
this effect is a cross term with firm size 100–299, the policy period, and
gender. (The dataset does not include information on firms with fewer
than 100 employees.) If Abe’s policy was effective for increasing female
promotion, then g4 should be negative and significant, since the law did
not require firms of size 100–299 to make the gender statistics public or to
make scheduled plans for active participation of female employees. The
policy was actually enacted in 2015 and implemented in 2016, but there
were talks about this policy from 2013 and a proposal to the Diet in 2014;
therefore, I made a policy dummy spanning the years 2013 to 2017.
Finally, I estimated Equations 1 and 2 for seishain workers only,
while dropping the terms seishain and seishain*female.

11.9 Results of the Estimation

11.9.1 Promotion of Females, Long-Term Employment

Practice, and the Effect of Abe’s Policy
The first set of results in Table 11.3, on the left-hand side, shows the
estimated marginal effects of Equation 1 for all workers. We can say that
the promotion of females went up significantly by 1 percent during the
2013–2017 Abe administration period and additionally by 0.5 percent
during the 2016–2017 period when Josei Katsuyaku Suisin Hō was
implemented.
340 Nobuko Nagase

However, this progress must be considered in the context of trad-

itional Japanese employment practice. While higher education and
higher age both increase the probability of promotion to managerial
positions, the effects of education and age for females are significantly
less than those for males. The probability of becoming managers
strongly increases with age and peaks at ages forty-five to fifty-four,
but the effect is also around 10 percent lower for females.
As shown in on the left side of Table 11.3, being in a seishain position
greatly increases the probability of promotion. Comparing the estimates
for all workers and for seishain only, the gender difference in advance-
ment to manager by age and education is even more pronounced among
seishain, especially in their forties and fifties. Males in their fifties are
nearly 60 percent more likely to be a manager than males in their teens or
in early 20s, but the effect is 13 percent less for females in their fifties.
Tenure effect shows that if women stay in the firm for a long period,
tenure square term shows that females tenure will be compensated in
promotion a little higher than males.
The long-term employment practice of seishain evidently negatively
affects the allocation of firm specific human capital investment to
females. Firms may be concerned that females will quit their work
when they form families, while such conduct is less likely for males.
This in turn generates statistical discrimination against females in
hiring and promotion. According to the results, females were more
rewarded only after staying with the firm for a long period, only after
many males surpassed them in promotion.
Interesting results include the marginal effects of firm size, which
show that promotion is less likely at larger firms compared to smaller
firms when other factors are controlled for. The year dummies reveal
that promotion was slow during the recession of 2002 to 2006 for both
sexes relative to 2012. These were years of bad macroeconomic per-
formance, with the economy in deep recession in 2003. Additionally,
from 2002 to around 2010, females were clearly less likely to be
promoted to managers than males after considering age, education,
tenure and other variables. However, the sign on the coefficient of
females shifted to positive in 2011 and became significantly positive
during Abe’s cabinet, 2013–2017, as compared with 2012.
In summary, the first column of Table 11.3 shows that education,
tenure, and age are significantly less valued for females for promotion
in long-term employment. However, there are more forces to push
Abe’s Womenomics Policy 341

Table 11.3 Probit analysis of being a manager, marginal effects

All employees, firm Seishain employees,

size over 100 firm size over 100

Coefficient t values Coefficient t values

Female -0.0096 −1.2 −0.0099 −1.0

Ed 14 yrs 0.0751 148.1 *** 0.0912 148.0 ***
Ed 16 yrs or more 0.1476 501.9 *** 0.1763 497.8 ***
Ed 14 yrs* female −0.0105 −13.5 *** −0.0135 −13.8 ***
Ed 16 yrs or more* female −0.0207 −28.0 *** −0.0252 −26.9 ***
Age 21–24 0.0412 7.5 *** 0.0465 7.0 ***
Age 25–29 0.1629 26.6 *** 0.1888 26.2 ***
Age 30–34 0.3430 49.4 *** 0.3824 49.0 ***
Age 35–39 0.4771 65.4 *** 0.5160 64.8 ***
Age 40–44 0.5610 75.5 *** 0.5945 74.7 ***
Age 45–49 0.5966 79.6 *** 0.6247 78.4 ***
Age 50–54 0.5984 79.6 *** 0.6231 78.0 ***
Age 55–59 0.5522 73.0 *** 0.5743 70.8 ***
Age 60–64 0.5104 65.4 *** 0.4839 57.3 ***
Age 21–24* female −0.0122 −1.5 −0.0153 −1.5
Age 25–29* female −0.0386 −5.3 *** −0.0500 −5.5 ***
Age 30–34* female −0.0681 −10.5*** −0.0868 −10.4***
Age 35–39* female −0.0893 −15.3*** −0.1138 −15.0***
Age 40–44* female −0.0997 −18.2*** −0.1268 −17.7***
Age 45–49* female −0.1028 −19.3*** −0.1295 −18.4***
Age 50–54* female −0.1051 −20.2*** −0.1315 −18.9***
Age 55–59* female −0.1047 −20.2*** −0.1301 −18.7***
Age 60–64* female −0.1073 −21.4*** −0.1326 −19.2***
tenure 0.0130 266.2*** 0.0155 248.4***
tenure^2 −0.0002 −144.2*** −0.0002 −124.8***
tenure* female −0.0007 −6.3*** −0.0002 −1.3
tenure^2* female 0.0001 17.9*** 0.0000 10.1***
year2002 −0.0085 −13.0*** −0.0110 −13.7***
year2003 −0.0105 −16.3*** −0.0135 −16.8***
year2004 −0.0205 −32.4*** −0.0256 −32.5***
year2005 −0.0037 −5.3*** −0.0067 −7.8***
year2006 0.0055 7.9*** 0.0050 5.8***
year2007 0.0136 18.9*** 0.0155 17.4***
year2008 0.0089 12.4*** 0.0096 10.8***
year2009 0.0030 4.2*** 0.0022 2.5***
year2010 0.0030 4.2*** 0.0023 2.6***
year2011 −0.0052 −7.3*** −0.0070 −8.0***
342 Nobuko Nagase

Table 11.3 (cont.)

All employees, firm Seishain employees,

size over 100 firm size over 100

Coefficient t values Coefficient t values

year2002* female −0.0230 −14.3*** −0.0291 −14.5***

year2003* female −0.0160 −9.8*** −0.0200 −9.9***
year2004* female −0.0237 −14.7*** −0.0298 −14.8***
year2005* female −0.0156 −9.5*** −0.0193 −9.3***
year2006* female −0.0117 −7.1*** −0.0161 −7.7***
year2007* female −0.0130 −8.0*** −0.0166 −8.0***
year2008* female −0.0017 −1.0 −0.0031 −1.4
year2009* female −0.0041 −2.4** −0.0060 −2.8**
year2010* female −0.0030 −1.8* −0.0035 −1.6
year2011* female 0.0017 1.0 0.0012 0.5
year2013−2017 −0.0043 −7.6*** −0.0051 −7.2***
year 2013–2017* female 0.0119 8.7*** 0.0131 7.6***
year2016−2017 −0.0033 −7.6*** −0.0035 −6.4***
year 2016–2017* female 0.0056 5.6*** 0.0072 5.7***
firm5000− −0.0429 −114.6*** −0.0538 −114.2***
firm1000−4999 −0.0290 −76.2*** −0.0362 −76.0***
firm500−999 −0.0149 −34.4*** −0.0185 −34.3 ***
firm 100–299 0.0089 22.4 *** 0.0114 23.1 ***
seishain 0.1253 249.9 ***
seishain * female 0.0034 2.5 ***
Pseudo r2 9,893,284 8,528,542
observation 0.2643 0.2325

Note: statistically significant *** 1 percent, ** 5 percent, * 10 percent.

women into managerial positions after 2013 and even more so in

2016–2017 compared to 2013–2014, although the marginal effect
was small.

11.9.2 Using Variation in Firm Size to Assess Abe’s Policy

Table 11.4 shows the effect of Abe’s policy by using Equation 2, which
examines differences in firms’ response to the policy change by firm
size. Josei Katsuyaku Suisin Hō did not mandate firms with 100–299
employees to make action plans. However, firms with more than 300
Abe’s Womenomics Policy 343

Table 11.4 Estimate of Abe’s policy on female promotion by firm size

All employees, firm Seishain employees,

size over 100 firm size over 100

Coefficient t values Coefficient t values

Female −0.0165 −6.0 *** −0.0199 −6.5 ***

Ed 14 yrs 0.0788 154.4 *** 0.0955 154.4 ***
Ed 16 yrs or more 0.1478 502.4 *** 0.1763 498.1 ***
Ed 14 yrs*female −0.0136 −17.4 *** −0.0173 −17.8 ***
Ed 16 yrs or more* female −0.0221 −29.7 *** −0.0268 −28.5 ***
Age 25–29 0.1137 68.4 *** 0.1364 68.7 ***
Age 30–34 0.2861 153.9 *** 0.3267 153.7 ***
Age 35–39 0.4205 211.6 *** 0.4638 210.5 ***
Age 40–44 0.5073 246.8 *** 0.5468 244.5 ***
Age 45–49 0.5450 260.4 *** 0.5799 256.8 ***
Age 50–54 0.5487 260.8 *** 0.5802 255.7 ***
Age 55–59 0.5030 237.7 *** 0.5315 230.4 ***
Age 60–64 0.4608 195.5 *** 0.4378 163.0 ***
Age 25–29* female −0.0294 −13.2 *** −0.0384 −13.8 ***
Age 30–34* female −0.0615 −31.5 *** −0.0781 −31.4 ***
Age 35–39* female −0.0845 −47.7 *** −0.1072 −46.7 ***
Age 40–44* female −0.0959 −57.4 *** −0.1213 −55.3 ***
Age 45–49* female −0.0994 −60.8 *** −0.1245 −57.5 ***
Age 50–54* female −0.1021 −63.4 *** −0.1270 −59.1 ***
Age 55–59* female −0.1021 −62.6 *** −0.1261 −57.5 ***
Age 60–64* female −0.1060 −57.5 *** −0.1307 −48.6 ***
tenure 0.0130 263.9 *** 0.0154 246.1 ***
tenure^2 −0.0002 −146.1 *** −0.0002 −127.1 ***
tenure* female −0.0007 −6.0 *** −0.0001 −1.0
tenure^2* female 0.00005 18.0 *** 0.00004 10.4 ***
year2002 −0.0084 −12.8 *** −0.0108 −13.4 ***
year2003 −0.0102 −15.7 *** −0.0131 −16.3 ***
year2004 −0.0200 −31.4 *** −0.0250 −31.6 ***
year2005 −0.0029 −4.2 *** −0.0058 −6.7 ***
year2006 0.0064 9.1 *** 0.0059 6.9 ***
year2007 0.0142 19.6 *** 0.0162 18.1 ***
year2008 0.0094 13.1 *** 0.0102 11.4 ***
year2009 0.0033 4.6 *** 0.0025 2.8 ***
year2010 0.0036 5.1 *** 0.0030 3.4 ***
year2011 −0.0047 −6.6 *** −0.0065 −7.4 ***
year2002* female −0.0216 −13.3 *** −0.0273 −13.5 ***
year2003* female −0.0147 −9.0 *** −0.0184 −9.0 ***
344 Nobuko Nagase

Table 11.4 (cont.)

All employees, firm Seishain employees,

size over 100 firm size over 100

Coefficient t values Coefficient t values

year2004* female −0.0226 −14.0 *** −0.0285 −14.1 ***

year2005* female −0.0143 −8.6 *** −0.0177 −8.4 ***
year2006* female −0.0114 −6.9 *** −0.0157 −7.6 ***
year2007* female −0.0133 −8.1 *** −0.0169 −8.1 ***
year2008* female −0.0012 −0.7 −0.0025 −1.2
year2009* female −0.0041 −2.4 ** −0.0060 −2.8 ***
year2010* female −0.0030 −1.8 * −0.0035 −1.6 *
year2011* female 0.0016 0.9 0.0011 0.5
year2013–2017 −0.0031 −5.5 *** −0.0031 −4.4 ***
firm 100–299 0.0426 119.1 *** 0.0531 120.7 ***
year 2013–2017* female 0.0140 10.2 *** 0.0157 9.2 ***
firm100−299* female −0.0084 −10.3 *** −0.0106 −10.4 ***
firm100–299* year2013–2016 −0.0094 −16.0 *** −0.0126 −17.1 ***
firm100–299* female* 0.0023 1.7 * 0.0033 1.9 *
year2013–2017
seishain 0.1262 251.3 ***
seishain*female 0.0022 1.6
Pseudo r2 9,893,284 8,528,542
observation 0.2644 0.231

Note: statistically significant *** 1 percent, ** 5 percent, * 10 percent.

employees were mandated to do this. The law was implemented in

2016, but it was discussed in 2014 and passed in 2015, so I chose to
examine the effect of the law throughout the 2013–2017 period.
If the cross term of the firm size 100–299, the Abe policy period, and
being female is positive and barely significant, that means that women
working at firms with 100–299 employees were slightly more likely to
be promoted to manager during 2013–2016 relative to larger firms –
or, to put it differently, that firms with more than 300 employees did
not have more females promoted to managers during the Abe cabinet.
The results do not show that females had a significantly higher
probability of promotion at mandated firms during the Abe adminis-
tration period; that is, overall promotion to managers did not acceler-
ate specifically at firms with more than 300 employees during this
Abe’s Womenomics Policy 345

period. This is understandable considering the weak type of the man-

date. However, from Table 11.3 we did see that female promotion
accelerated more generally during the Abe administration, though the
marginal effects were small. The descriptive statistics is shown in the
Appendix.

11.10 Discussion and Conclusion

This paper has reviewed changes in policies and law in Japan after
1980s to today that aimed to increase the number of women in the
labor force, promote females in management, and close the gender
wage gap. We constructed a long-term microdata series using the
Labor Force Special Survey (Tokutei Chosa) and Labor Force Survey
to look at long-term trends from 1988 to 2017. Since the LFS only
gives data on managers in narrow definition, we also used the Basic
Survey on Wage Structure of the Ministry of Health, Labor, and
Welfare to look at those in managerial positions in the broad defin-
ition, by gender.
Previous research of myself has shown that, after around 2009, work
continuation of new mothers as seishain started to increase with the
introduction of a short hour option that firms were mandated to make
available for workers with infants. This trend of greater work continu-
ation as seishain accelerated during Abe’s administration and his
Womenomics policy of 2013. New mothers’ continuing labor partici-
pation as seishain will help to reduce the gender pay gap in the long run,
since, once females quit their permanent standard employment at
marriage or at childbirth, most only re-enter the labor market as poorly
paid hourly nonstandard employees.
However, previous research did not reveal whether those females,
who continued their work as seishain, would be as likely to be
promoted as males. Therefore, this paper has investigated the man-
agerial and career prospects of females. The paper showed that the
closing of the gender gap in managerial positions was still quite
limited, although some improvement was shown during Abe’s policy
period.
We found that the closing of the promotion gap has been slow even
among the more highly educated seishain workers. This paper showed
that there is a profound gender difference in the evaluation of age,
education, and tenure for promotion to manager in Japan. We did find
346 Nobuko Nagase

some acceleration of women’s accession to managerial positions during

the 2013–2017 period.
Even though a menu of policies was rolled out, most were aimed at
reducing the opportunity cost of having children. They were not tar-
geted at making fundamental changes to the Japanese long-term
employment system.
The results found in this chapter indicates that further drastic change
is required in labor law and labor practice to close this large gender
wage gap and the gap in promotion prospects.
Why are the gaps still so large? A substantial proportion of women may
be hired in the slow track but not the fast track, compared to male
university graduates. Another reason may be lack of training, even though
one might be in the same track as males. One can only become manager
through adequate experience and training. If such opportunity is offered
to females only in a limited way, even if the law nudges firms to promote
women, firms will have a harder time finding adequate candidates. In
addition, often women choose the slow track, as long work hours and
company expectations of relocation may harm their family life. In add-
ition, more seishain women than men are taking one year of leave at
childbirth and then returning to work as short work-week employees after
the leave. This may also hinder their promotion.
We found that gender-based evaluation is greatly evident in the
Japanese long-term employment system, which was once highly praised
as the cause of strong economic performance in Japan in the 1980s. For
more women to make the best of their capacity, the Japanese type of long-
term employment system needs to undergo further change.
To conclude, Abe’s policy had a statistically significant effect on
increasing women’s representation in management for companies in
general. This being said, the gap in gender is still very large. A more
fundamental change in labor practice is needed to close the large gender
wage gap and the gender gap in managerial positions. This should
include a change in hiring practice, a change in training, and a change
in relocation practices, as well as reforming the EEOL.

Notes
1. “Seishain” in Japan is sometimes translated as “regular full-time
employment.”
Abe’s Womenomics Policy 347

However, “seishain” is often associated with a work contract without

termination. Moreover, wages are often expected to increase with
tenure, and higher work commitment is also expected. The same
workers are also often referred to as “standard” employees. Therefore,
I will use “permanent standard employee,” “standard worker,” or
“long-term employee” as translations of “seishain,” as well as using the
term itself.
2. Most of this research is based on interviews that have been published
only in Japanese, including my own research, as cited in the sections that
follow.
3. For the first time in December 2007, the governmental document gave
a specific target and timeline – 38 percent in ten years – for the
percentage of registered childcare places for infants, in the Report of
the National Council for Pathways to Supporting Children and Families
(Kodomo to Kazoku wo Oensuru Nihon Juten Senryaku Kaigi). The
national average of registered childcare places for children below three
years of age was then 20 percent, but lower in most major urban areas.
This report was adopted as a reference material of the National Social
Security Council Report (Shakai Hosho Kokumin Kaigi) of
November 2008 and then recognized 38 percent in ten years as
a national target in the Cabinet Office’s Medium Term Plan of
Economy and Finance (Keizai Zaisei no Chuchoki Hoshin to 10 nen
Tembou) of December 2008.
4. The 2020–30 target was initially decided in June of 2003 by the Gender
Equality Promoting Head Meeting, Danjo Kyodo Sanaku Suisin
Honbu. The Council on Gender Equality in 2007 explained that
“women leaders” meant representatives at the national and local
congress, managers over the rank of kachō at enterprises, and
professionals who need high professional knowledge. The percent of
women among private corporate managers was described as still
6.5 percent as of 2009 but was targeted to be 10 percent by 2015 in
the third Basic Plan for Gender Equality.
5. For example, in 2013 the proportion of women was 7.9 percent in the
House of Representatives and 18.2 percent in the House of Councilors.
Women were 6.9 percent of department managers in private enterprises
in 2011. In 2010, women were 18.9 percent of medical doctors,
5.6 percent of Ministers, and 25.7 percent of new hires of national
bureaucrats in the fast promotion track.
6. The Gender Equality Bureau Cabinet Office showed in their White
Paper of 2011 that among the 41,973 board members of 3,608 listed
firms, only 515 were women, comprising only 1.2 percent. In
manufacturing, construction, finance, and transportation, women
348 Nobuko Nagase

board members comprised less than 1 percent. Toyokeiai Yakuin

Shikiho shows the number of women board members among listed
firms as of July each year. The number was 538 or 1.2 percent in 2006
and remained at 1.2 percent until 2009. In 2011, it was 585 or
1.4 percent and in 2012, 630 or 1.6 percent. Many large firms did
seem to respond to Abe’s request. In 2013, it was 691 or 1.8 percent;
in 2014, it rose to 816 and 2.1 percent; and in 2015, 1,142 and
2.8 percent. For example, among 92 listed banks, there were no
women on the board among 772 executive board members in 2011.
Even when non-executive directors and audit and supervisory board
members are included, among 1,370 members, only 6, or 0.4 percent,
were female in 2011. The figure of female officers in listed companies
rose to 4.1 percent as of 2018 according to the Gender Equality Bureau.
7. Firms must collect and publish one type of data among fourteen options.
The options include the female ratio of employees, the gender gap in
average tenure, the gender gap in application to hiring ratio, monthly
overtime work hours, and the gender gap in management. Firms are
asked to analyze the reason for the gender gap. Action plans should have
targets and planned periods, which must be made available to the public
and registered at the local MHLW office. The action plan can be, for
example, increasing male childcare leavers and providing a model for
diverse career plans. As many as 12,236 firms registered their gender
statistics on the MHLW’s database as of May 2020 May. Also 15,288
firms made open their action plan on the same database: http://positive-
ryouritsu.mhlw.go.jp/positivedb/. From April 2020, the law was
renewed, and these firms must disclose at least two, but yet just two.
8. From April 2020, the law was renewed, and firms with more than 100
employees will be required to publicize their action plans.
9. To be recognized, the hiring ratio and tenure should be about the same
for men and women, average work hours should be less than fewer
hours, the ratio of female managers should be above the industry
average, and the ratio of promotion to kachō or middle managerial
positions from section chief should be about the same by gender. In
addition, firms should demonstrate the development of diverse career
paths by allowing changes from the slow track to the fast track within
the company or by recruiting seishain from nonstandard employees.
The Eruboshi recognition has three stages, and the recognition may be
obtained based on their present status.
10. There will be a budget for preferred procurement of around 5 trillion
yen according to the Cabinet Office, even though Eruboshi comprise
one among lists for preferred points.
Abe’s Womenomics Policy 349

11. The net decline in after-tax household income due to the dependent
spousal tax deduction when spouses work longer hours became
a political issue, so that, in 1987, a new spousal special deduction was
introduced to prevent the decline in net household income that results
due to the increase of income tax of husbands. Yet many firms gave
spousal allowances as a supplement to the main income earners, and
since this allowance was often linked to one’s spouse not having taxable
income, that spouses still adjusted their work hours so that income tax
should not be levied.
12. Toyota, for example, announced in 2016 that it would abolish its
spousal allowance step by step in five years and would instead offset
the decline in salary by increasing the children’s allowance.
13. As to the issue of social security tax for a dependent spouse, the
government is not overtly making any changes. However, the
government introduced new rules in October 2016 that following part-
time workers should be taxed the social security insurance for
employees and also be given rights to the employee pension. This is for
those who earn more than 1.06 million a year, worked twenty hours per
week, and are not students, but it is limited to those who work at firms
with more than 500 employees. The government, in this way, is hoping
to decrease the fraction of dependent wives who are in the social security
tax category of “dependent spouses” despite that fact that they are short
hour workers. In 2019, Pension Law reform further mandated that the
rule for part-time workers will be expanded to firms with more than 100
employees in 2022, and further to those with more than 50 employees in
2024.
14. Details of the sampling methods, estimation methods and the sampling
errors of the estimates used for the Labour Force Survey are available in
English: www.stat.go.jp/english/data/roudou/pdf/samplerr.pdf, accessed
September 13, 2017.
15. The question about annual income in the previous year offers respondents
twelve possible answers: 0 yen; >0 yen and less than 0.5 million yen; ≥0.5
yen and <0.99 million yen; ≥1 million yen and <1.49 million yen;
≥1.50 million yen and <1.99 million yen; ≥2.00 million and <2.99 million
yen; ≥3.00 million and <3.99 million yen; ≥4.00 million yen and
<4.99 million yen; ≥5.00 million yen and <6.99 million yen; ≥7.00 million
yen and <9.99 million yen; ≥10.00 million yen and <14.99 million yen;
15.00 million yen or over. Because only upper and lower bounds are
available, the midpoint in each category is used to represent income. For
the top income group, 17.00 million is used. Estimates of the wage rate are
obtained by dividing the annual income by the number of work hours in the
350 Nobuko Nagase

last week of the month multiplied by 50. The bottom and top 1 percentiles
of the wage data were not used in the analysis.
16. Many firms still have a mandatory retirement age of sixty. However,
a new law in 2012 mandated that, from April 2013, firms must continue
to offer employment opportunity to their workers until age sixty-four.
This, however, can be in the form of nonstandard employment. The law
passed because of the rise of the standard age of receiving the employee
public pension.
17. A time series of seishain is available from 2005. To include the time series
of 2002 to 2005, the definition of workers with permanent contract, that
those of seishain and nonstandard employees are recommended for use
by the ministry. In 2005, seishain was 81.4 percent of all male workers,
and 44.1 percent of all female workers. Permanent contract nonstandard
workers were 4.7 percent for males and 15.5 percent for females in the
same year. In 2017, the respective seishain figures by gender were
74.2 percent 43.5 percent, and permanent contract nonstandard
workers by gender, 3.6 percent and 8.8 percent.

Reference
Abraham, K., and S. Houseman. 1989. “Job Security and Work Adjustment:
How Different Are US and Japanese Practices?” Journal of the Japanese
and International Economies 3(4): 500–521.
Abe, Y. 2002. “Does the \1.03 million tax exemption limit depress the hourly
wage of part-time workers?“ Japan Labor Bulletin 41(9): 6–7.
Abe, Y. 2010. “Equal employment opportunity law and the gender wage gap
in Japan: A cohort analysis.” Journal of Asian Economics 21(2):
142–155.
Abe, Y. 2011. “Equal employment opportunity law and labor force behavior
of women in Japan.” Journal of the Japanese and International
Economies 25(1): 39–55.
Abe, Y., and A. Oishi. 2009. “The 1.03 million yen ceiling and earnings
inequality among married women in Japan.” Economics Bulletin 29(2):
1510–1519.
Akabayashi, H. 2006. “The labor supply of married women and spousal tax
deductions in Japan – a structural estimation.” Review of Economics of
the Household 4(4): 349–378.
Albrecht, J., A. Börklund, and S. Vroman. 2003. “Is there a glass ceiling in
Sweden?” Journal of Labor Economics 21(1): 145–177.
Abe’s Womenomics Policy 351

Asai, Y. 2015. “Parental leave reforms and the employment of new mothers:
Quasi-experimental evidence from Japan.” Labour Economics 36:
72–83.
Asai, Y., R. Kambayashi, and S. Yamaguchi. 2015. “Childcare availability,
household structure, and maternal employment.” Journal of the
Japanese and International Economies 28: 172–192.
Bessho, S. I., and M. Hayashi. 2014. “Intensive margins, extensive margins,
and spousal allowances in the Japanese system of personal income taxes:
A discrete choice analysis.” Journal of the Japanese and International
Economies 34: 162–178.
Firpo, S., N. M. Fortin, and T. Lemieux. 2009. “Unconditional Quantile
Regressions.” Econometrica 77(3): 953–973.
Eunmi, M., and M. Brinton. 2015. “Workplace matters: The use of parental
leave policy in Japan.” Work and Occupations 42(3): 335–369.
Fujii, K. 2009. “Research on Gender Wage Gap.” Business Labor Trend,
September 11–16 (in Japanese).
Hanada, M. 1987. “Competition within Human Resource System : The
Human Resource Management of Japanese Firms seen from the
Promotion System (Jinji Seido kara mita Nihon Kigho no JInji
Senryaku). Soshiki Kagaku (Organization Science) 21(2): 44–53
(in Japanese).
Hara, H. 2018. “The Gender Wage Gap across the Wage Distribution in
Japan: Within and Between Establishment Effects.” Labour Economics
53: 213–229.
Hashimoto, M., and J. Raisian. 1985. “Employment Tenure and Earnings
Profile in Japan and the United States.” American Economic Review 75
(4): 721–735.
Hashimoto, M., and J. Raisian. 1992. “Employment Tenure and Earnings
Profile in Japan and the United States: Reply.” American Economic
Review 82(1): 346–354.
Higuchi, Y. 1995. “Sengyoshufu Hogo Seisaku no Keiziteki Kiketsu
(Economic Consequence of Social Protection Policy for Housewives).
In “Jakusha” Hogo Seisaku no Keizai Bunseki (Consequence of Social
Protection for the “Vulnerable”), edited by T. Hotta and N. Yashiro.
Tokyo: Nihon Keizai Shinbunsha (in Japanese).
Hori, H. 1998. “Danjo Chingin Kakusa no Shukusho Keiko to sono Yoin
(Decline in Gender Wage Gap and the Factors Contibuting to the
Change).” Nihon Rodo Kenkyu Zasshi No. 456 (in Japanese).
Houseman, S., and K. Abraham. 1993. “Female Workers as Buffer in the
Japanese Economy.” American Economic Review 83(2): 45–51.
352 Nobuko Nagase

Imada, S., and S. Hirata. 1995. Promotion Structure of White Collar

Workers (Howaito Kara no Shoshin Kozo). Tokyo: Japan Institute of
Labor (in Japanese).
JILPT. 2009. Report on Research for Gender Wage Gap under Changing
Wage and Labor Practices, Surveys on Firm Wage and Employment
Practices, and Survey on Employer and Labor Union Leaders. JILPT
Research Series No. 52 (in Japanese).
Matsubara, M. 2012. “The Impact of Long-term Use of the Short Working
Hour System on Careers” (Tanjikan seishain seido no chokiriyo ga
kyaria ni oyobosu eikyo). Nihon Rodo Kenkyu Zasshi (Japanese
Journal of Labour Studies) 54(10): 22–33 (in Japanese).
Nagase, N. 1999. “Work and childbearing choice of married women in
Japan: The effect of labor practices” (Shoshika no yoin: shugyo
kankyo ka kachikan no enka ka – Kikonsha no shugyo keitai sentaku
to shussan jiki no sentaku). Journal of Population Problems (Jinko
Mondai Kenkyu) 55(2): 1–18 (in Japanese).
Nagase, N. 2002. “Wife allowance and tax exemption behind the low wage
of part-time workers.” Japan Labor Bulletin 41(9): 8–9.
Nagase, N. 2003a. “Childcare policy for economic revival of
urbanities” (Toshi saisei to hoiku seisaku).” In Economic Analysis
of Urban Revival (Toshi Saisei no Keizai Bunseki), edited by
F. Yamazaki and Y. Asada, pp.248–278. Tokyo: Toyo keizai
shinposha (in Japanese).
Nagase, N. 2003b. “Standard and non-standard work arrangements, pay
difference, and choice of work by Japanese mothers.” In Nonstandard
Work in Developed Economies: Causes and Consequences, edited by
S. Houseman and M. Osawa, pp.267–300. Kalamazoo, MI:
W. E. Upjohn Institute for Employment Research.
Nagase, N. 2007. “Shoshika ni kakawaru Seisaku ha Doredake
Jikkosaretanoka? Hoiku to Jido Ikusei ni kansuru Seisakuno Kadai
(How much of the Policy for Childcare and Child Development Has
Been Put into Action?).” Financial Review 87: 3–22 (in Japanese).
Nagase, N. 2010. “Danjo Chingin Kakusa no Kaisho ni Mukete Naniga Sei
CHurituteki na Seido ka (What Is Meant by Gender Neutral Practices, in
Search for Closing Gender Wage Gap).” Rodo Chosa (Labor Research),
February 14–22 (in Japanese).
Nagase, N. 2014. “The effect of a short-hour option mandate on first
childbirth, birth intention and work continuation after childbirth:
Evidence from Japan” (Ikujitanjikan no gimukaga daiisshishussan to
shugyokeizoku shussan iyoku ni ataeru eikyou: hokaisei wo
shizenjikken tosita jisshou bunseki).” Journal of Population Studies
(Jinkogaku Kenkyu) 50(1): 29–53 (in Japanese).
Abe’s Womenomics Policy 353

Nagase, N. 2017. “The Effect of Family Friendly Policies on Fertility and

Maternal Labor Supply.” Stanford Asia Health Policy Program
Working Paper No. 42.
Nagase, N. 2018a. “Has Womenomics Worked?” Asian Economic Policy
Review 13(1) 68–101.
Nagase, N. 2018b. “Inequality between Standard Employment and
Non-Standard Employment: Policies to Promote Human Resource
Investment among Females and Youth.” Japan Labor Review (Nihon
Rodo Kenkyu Zasshi) Special Issue No. 691: 19–38 (In Japanese).
Nagase N., and M. Brinton. 2017. “Gender division of labor and second
births: Labor market institutions and fertility in Japan.” Demographic
Research 36(11): 339–370.
Nagase, N., and Mana Yamaya. 2011. “Career Prospects of University
Graduate Females in Large Japanese Firms: The Course Based Hiring
Practices, their Anxieties and Possibilities (Daikigyo Kinmu no Daisotu
Seishain Josei no Shugyo Keizoku Fuan – kosu betu jinji ni chumoku
shite).” Kyaria Dezain Kenkyu (Career Design Research) 7: 185–197 (in
Japanese).
Nagase, N., and Mana Yamaya. 2012. “Women in Managerial Position
at Private Firms: Labor Practice and Division of Labor within
Households (Minkan Daikigyo no Josei Kanrishoku no Kyaria
Keisei).” Kyaria Dezain Kenkyu (Career Design Research) 8:
95–105 (in Japanese).
Ohi, M. 2004. “Changes in Managers Seen from Statistics, the Numbers, the
Speed of Promotion, and Relative Wage versus Non-Managers (Suji de
miru Kanrishokuzo no Henka: Ninzu, Shoshinsokudo, Ippanshoku
tono Sotaichingin).” Nihon Rodo Kenkyu Zasshi (Japanese Journal of
Labour Studies) 545: 4–17 (in Japanese).
Ohuchi, S. 1999. “Daisotsu Howaito Kara no Kigyonai Kyaria Keisei:
Sogoshoku, Ippanshoku no Jittai Chosa yori (Career Development of
White Collar University Graduates within Firm: Sogoshoku and
Ippanshoku).” Nihon Rodo Kenkyu Zasshi (Japanese Journal of
Labour Studies) 471 (in Japanese).
Takahashi, S. 2010. “An estimation of the work disincentive effects of the
spousal deduction and the social security systems on the labor supply of
Japanese married women (Haigusha kojo oyobi shakai hosho seido ga
nihon no kikon josei ni oyobosu roudou yokusei koka no sokutei).”
Nihon Rodo Kenkyu Zasshi (Japanese Journal of Labour Studies) 52
(12): 28–43 (in Japanese).
Takeishi, E. 2013. “Measures to shorten working hours: Their current state
and areas of concern (Tanjikan kinmu seido no genjo to kadai).” Shogai
Gakushu to Kyaria Desain 10: 67–84 (in Japanese).
354 Nobuko Nagase

Takeuchi, H. 1995. Japanese Meritocracy (Nihon no Meritokurashi).

Tokyo: University of Tokyo Press (in Japanese).
Teramura E. 2014. “The effect of the workplace norm on female
employment and work quit (Josei no ahugyo, rishoku kodo no
jisshou bunseki: Shokuba no funiki ga ataeru eikyo no kensho).”
Kyaria Dezain Kenkyu (Research on Career Design) 10: 175–185 (in
Japanese).
21seiki Shokugyo Zaidan. 2014. Ikuji wo shinagara Hataraku Josei no
Shoshin Iyoku ya Mochibeshon ni kansuru Chosa (Study on the
Factors affecting Promotion Motives of Women with Children in Long-
term Standard Employement). www.jiwe.or.jp/research-report/2013m
otivation_women_to_work (in Japanese).
Unayama, T. 2011. “Marriage and female labor participation: Evidence
from Japanese prefectural cross-section data” (Kekkon, shussan to
shugyo no ryoritsu kanousei to hoikujo no seibi).” Nihon Keizai
Kenkyu 65: 1–22 (in Japanese).

Appendix Descriptive Statistics for Table

11.3 and Table 11.4

All employees, firm Seishain employees firm

size over 100 size over 100

standard standard
means dev min max means dev min max

managers 0.222 0.416 0 1 0.253 0.435 0 1

female 0.299 0.458 0 1 0.263 0.440 0 1
Ed 14 yrs 0.148 0.355 0 1 0.146 0.353 0 1
Ed 16 yrs or more 0.362 0.481 0 1 0.388 0.487 0 1
Ed 14 yrs*female 0.083 0.275 0 1 0.077 0.266 0 1
Ed 16 yrs or more* 0.073 0.260 0 1 0.072 0.259 0 1
female
Age 21–24 0.076 0.265 0 1 0.076 0.265 0 1
Age 25–29 0.131 0.338 0 1 0.133 0.340 0 1
Age 30–34 0.134 0.340 0 1 0.138 0.345 0 1
Age 35–39 0.134 0.341 0 1 0.140 0.347 0 1
Age 40–44 0.131 0.337 0 1 0.136 0.343 0 1
Abe’s Womenomics Policy 355

(cont.)

All employees, firm Seishain employees firm

size over 100 size over 100

standard standard
means dev min max means dev min max

Age 45–49 0.120 0.325 0 1 0.124 0.330 0 1

Age 50–54 0.115 0.319 0 1 0.118 0.322 0 1
Age 55–59 0.100 0.300 0 1 0.099 0.298 0 1
Age 60–64 0.041 0.197 0 1 0.018 0.133 0 1
Age 21–24* female 0.035 0.183 0 1 0.034 0.181 0 1
Age 25–29* female 0.050 0.218 0 1 0.048 0.213 0 1
Age 30–34* female 0.041 0.199 0 1 0.038 0.191 0 1
Age 35–39 * female 0.036 0.187 0 1 0.033 0.178 0 1
Age 40–44* female 0.035 0.183 0 1 0.030 0.170 0 1
Age 45–49* female 0.032 0.175 0 1 0.026 0.160 0 1
Age 50–54* female 0.029 0.168 0 1 0.024 0.153 0 1
Age 55–59* female 0.024 0.153 0 1 0.019 0.137 0 1
Age 60–64* female 0.010 0.099 0 1 0.005 0.069 0 1
tenure 13.262 11.159 0 49 14.281 11.042 0 49
tenure^2 300.39 403.31 0 2401 325.88 402.11 0 2401
tenure* female 2.825 6.640 0 49 2.785 6.777 0 49
tenure^2* female 52.06 178.74 0 2401 53.69 183.33 0 2401
year2002 0.070 0.254 0 1 0.080 0.271 0 1
year2003 0.068 0.252 0 1 0.078 0.268 0 1
year2004 0.069 0.253 0 1 0.078 0.269 0 1
year2005 0.061 0.239 0 1 0.060 0.238 0 1
year2006 0.062 0.242 0 1 0.062 0.240 0 1
year2007 0.059 0.236 0 1 0.058 0.234 0 1
year2008 0.058 0.235 0 1 0.057 0.231 0 1
year2009 0.057 0.231 0 1 0.055 0.228 0 1
year2010 0.058 0.233 0 1 0.056 0.230 0 1
year2011 0.056 0.230 0 1 0.054 0.227 0 1
year2002* female 0.018 0.132 0 1 0.020 0.140 0 1
year2003 * female 0.017 0.130 0 1 0.019 0.138 0 1
year2004* female 0.018 0.133 0 1 0.020 0.140 0 1
year200 * female 0.019 0.136 0 1 0.016 0.125 0 1
year2006* female 0.019 0.135 0 1 0.016 0.124 0 1
year2007* female 0.019 0.136 0 1 0.016 0.124 0 1
year2008* female 0.017 0.131 0 1 0.014 0.118 0 1
year2009* female 0.017 0.131 0 1 0.014 0.119 0 1
year2010* female 0.018 0.132 0 1 0.015 0.121 0 1
356 Nobuko Nagase

(cont.)

All employees, firm Seishain employees firm

size over 100 size over 100

standard standard
means dev min max means dev min max

year2011* female 0.017 0.130 0 1 0.015 0.120 0 1

year2013–2017 0.321 0.467 0 1 0.304 0.460 0 1
year 2013–2017* female 0.101 0.302 0 1 0.083 0.277 0 1
year2016–2017 0.133 0.340 0 1 0.127 0.333 0 1
year 2016–2017* female 0.044 0.204 0 1 0.037 0.188 0 1
firm5000– 0.243 0.429 0 1 0.248 0.432 0 1
firm1000– 0.245 0.430 0 1 0.246 0.431 0 1
firm500– 0.133 0.339 0 1 0.131 0.337 0 1
firm100– 0.264 0.441 0 1 0.261 0.439 0 1
firm100–* female 0.083 0.275 0 1 0.073 0.259 0 1
firm100–* female* 0.082 0.275 0 1 0.021 0.145 0 1
year2013–2017
firm100–* year 0.082 0.275 0 1 0.078 0.268 0 1
2013–2017
seishain 0.862 0.345 0 1
seishain* female 0.227 0.419 0 1
sample size 9,893,284 8,528,542
12 Corporate Governance Reforms
under Abenomics
The Economic Consequences of Two
Codes
h i d e a k i m i y a j i m a an d t a k uj i s a i t o

12.1 Introduction
Abenomics advanced a policy mix of bold monetary policy, flexible
fiscal policy, and a growth strategy for stimulating private investment,
which were called the “three arrows.” Corporate governance reforms
were positioned as one of the most important issues for the growth
strategy, the third arrow. Policy measures to achieve growth through
corporate governance reforms are not unique to Abenomics. The pur-
suit of corporate governance reforms from the standpoint of promoting
growth became a global trend following the global financial crisis (Kay
2012; OECD 2012). However, in United States, the United Kingdom,
and continental Europe, the focus of attention was the problem that
shareholder pressure was so strong that firms engaged in excessive risk-
taking or myopic business practices.1 In contrast, the central issue in
corporate governance reforms in Japan is that the existing system has
made firms take risk-averse or conservative approaches to business
management because the influence of shareholders (employees)
remains too weak (strong) on corporate decisions.

* An earlier version of this paper was presented at seminars in Stanford University,

Yonsei University, Hitotsubashi University, RIETI, John Hopkins University,
Seoul National University, Waseda University and the SASE annual meeting, the
Japan Association for Applied Economics annual meeting, and the Waseda-
RIETI-ECGI conference. We would like to thank Curtis Milhaupt for his
insightful comments. Johan Jidinger provided outstanding assistance with this
project. In compiling the data for this research, we were supported by
a KAKENHI research grant (I5H01958, 19H00603, 17K03810) from the
Ministry of Education, Culture, Sports, Science, and Technology. We were also
supported by the JSPS Core-to-Core Program (A. Advanced Research
Networks). A longer working paper version of this chapter is available at SSRN:
https://ssrn.com/abstract=3684372.

357
358 Hideaki Miyajima and Takuji Saito

The basic idea of Abenomics was to induce greater efficiency in

corporate management in the following ways. On the one hand, the
Japanese Stewardship Code (JSC) encouraged institutional investors to
engage in corporate management and actively exercise their voting
rights. On the other, the Corporate Governance Code (CGC) promoted
the appointment of independent outside directors, adoption of
a performance-related compensation system, and reduction of rela-
tional shareholdings. Thus, reforms of corporate governance systems
encouraged greater risk-taking by firms.
To what extent did the corporate governance reforms under
Abenomics, which centered on the implementation of the two codes,
affect corporate governance systems in Japan? Former governance
systems had already been changing since the 1997 banking crisis in
Japan. Japanese firms were once commonly characterized by the main
bank system, cross-shareholding, and insider boards consisting of
internally appointed directors. However, corporate reorganization
efforts after the banking crisis separated hybrid firms, which had
significantly changed their external governance systems, from trad-
itional Japanese firms that retained their old institutional characteris-
tics (Jackson & Miyajima 2007).2 What impact did the reforms under
Abenomics have on this evolution of corporate governance?
If the corporate governance reforms promoted changes in corporate
governance systems, to what extent then did they change Japanese cor-
porate behavior? According to “Japan Revitalization Strategy 2014,”
which extensively summarizes the plans under Abenomics, the govern-
ment intended to change the mindset of corporate management by
strengthening corporate governance and to enhance mechanisms for sup-
porting aggressive business decisions for winning in global competition.
Furthermore, the government expected firms to actively utilize their earn-
ing high profit for capital investment, bold business reorganizations, and
mergers and acquisitions (M&A), instead of accumulating retained earn-
ings. Did the corporate governance reforms, which was supported by two
pillars – the JSC and the CGC – contribute as expected to enhance capital
efficiency and corporate risk-taking?
This chapter aims to provide answers to these questions. To this end, we
first follow the evolution of corporate governance system before
Abenomics and highlight the characteristics of corporate governance
reforms under Abenomics. Then we examine the effects of introducing
the JSC and the CGC in Sections 12.3 and 12.4 respectively. In Section
Corporate Governance Reforms under Abenomics 359

12.5, we also examine the effects of the reforms on relational sharehold-

ing. In Section 12.6 we repost the extent to which the reforms affected
corporate behavior. Section 12.7 concludes.

12.2 Toward Governance Reforms under Abenomics: Twenty

Years of Experience

12.2.1 The First Wave of Governance Reform

The upsurge of governance reform under Abenomics is not the first
phenomenon for the past decades. Its first wave coincided with the period
beginning with the banking crisis of 1997 and ending around 2004. After
the banking crisis, strong doubts were expressed about the effectiveness of
the Japanese-type corporate governance system, and a serious effort to
take on its reform of this governance system was initiated.3 During this
period, a series of reforms modeled after a US-type governance system
were proposed, but partly implemented, notwithstanding the heated
debate over the pros and cons. In parallel with debate and reforms, the
governance system of Japanese firms evolved substantially.
First, the role of main banks in corporate governance retreated
substantially following the banking crisis of 1997. The worsening of
the non-performing loan problem meant that the banking sector, which
had been expected to get involved in firms that confronted financial
distress, was itself facing a financial crisis, and severe doubts arose
regarding the function of the main bank system (Hoshi & Kashyap
2004). The prolonged disposal process of nonperforming loans finally
got underway with the implementation of the Financial Reconstruction
Program in October 2002 at the Koizumi cabinet, which clearly out-
lined a procedure for disposing of nonperforming loans. During this
time, a new framework for private and public arrangements including
the realignment of the banking sector took shape. An important conse-
quence of this realignment was that it became difficult, due to rigorous
classification of borrowers, for banks to continue to hold on to client
firms with deteriorating performance (at-risk borrowers), so the impli-
cit rule that the main bank would shoulder more of the bad loans of
client firms was no more effective.
Second, the regulations with regard to stock ownership by banks and
business enterprises were revamped. After the banking crisis, the legal
requirements pertaining to shareholding between banks and business
360 Hideaki Miyajima and Takuji Saito

enterprises were clearly changed in the direction of restricting bank-

business enterprise holdings. The Law of Restriction on Banks
Shareholding dramatically pushed banks to reduce their
shareholding.4 The BIS regulations and introduction of the current
accounting standards promoted a cutback in bank shareholding
(Miyajima & Kuroki 2007).
Furthermore, the reform of the pension system in 2000 brought insti-
tutional change that had a major impact, fundamentally changing the
management of employee pension and national pension system reserve
funds. These funds, which had not been consigned to risk assets, were
invested in the stock. In 2001, the Government Pension Investment Fund
(GPIF) was established, and substantial reserves were redirected into
domestic stocks. As a result, the ownership structure changed dramatic-
ally. Figure 12.1 shows the long-term trend of insider and outsider own-
ership ratio followed by the classification of Franks, Mayer & Mayajima
(2014). The insider ratio is the aggregated ratio of share held by banks
(excluding trust accounts of trust banks), insurance companies, and cor-
porations. The outsider ratio is the aggregated ratio of shares held by
foreign investors, individuals, mutual funds, and pension trusts. It is
impressive that the insider ownership that exceeded over 60 percent and
was very stable for almost three decades dramatically declined since the

Insider Outsider Foreigners

70%

60%

50%

40%

30%

20%

10%

0%
55

60

65

70

75

80

85

90

95

00

05

10
19

19

19

19

19

19

19

19

19

20

20

20

Figure 12.1 Transition of ownership structure of Japanese firms

Note: Data is from Shareholder Survey by Tokyo Stock Exchange.
Corporate Governance Reforms under Abenomics 361

middle of the 1990s, and instead the outsider ownership was majority in
aggregated level.
Third, a debate was initiated over the redesign of the internal govern-
ance system – and more specifically, the compensation system and the
board system –around the time of the banking crisis. The 1997 revision of
the Commercial law made it possible to introduce stock options for the
first time; subsequently there occurred an apparent boom in the adoption
of stock options. Furthermore, there was also a spike in media coverage
of the debate over the pros and cons of making a transition to US-type
boards, whose primary function is monitoring. A fierce debate unfolded
between those who asserted that there was no other way to preserve
Japan’s competitiveness and those who asserted that the introduction of
US-type arrangements would not be suited to Japan’s other economic
systems centering, among which employment practices play a central
role, and that simple convergence on the US model would in fact reduce
efficiency.5 Against the backdrop of this conflict, the debate over the
revision of the Company Law proceeded, and, as a result, the law was
revised in 2002 to make it possible to choose a board modeled on a US-
type monitoring board, which was called the committee system.6
However, the fierce debate and consequent law amendment did not
lead to a substantial outcome. Figure 12.2 traces the changes of board
structure in Japanese firms. Although the board of directors has been
decreasing with the introduction of executives officer system, the com-
mittee system was rarely chosen among listed firms.7 The introduction
of outside directors was very slow.
In the early 2000s, the corporate governance system of Japanese firms
diversified. The main bank system, insider ownership, and insider
boards are no longer characteristics shared among Japanese firms. The
landscape of the diversification is summarized in Jackson and Miyajima
(2007, figures 1–3), where the hybrid forms of governance system that
combined the relation-based internal governance with market-based
external governance were emerged, while firms keeping the trajectories
of traditional Japanese companies still coexisted substantially.

12.2.2 Interlude: The Reverse Course and Gradual Change

Since around 2005, against this backdrop of increasing outside owner-
ship, the pressure of outsiders became influential, and even a smaller
boom of hostile takeovers has occurred.
362 Hideaki Miyajima and Takuji Saito

35 Mean number of directors (Left: People) 100%

Mean ratio of outside directors (Left: %) 90%

30
Ratio of firms with at least one outside 80%
25 director (Right: %) 70%

20 60%
50%
15 40%
10 30%
20%
5
10%
0 0%
1998
1999

2008
2009
2010
2011

2019
2000
2001
2002
2003
2004
2005
2006
2007

2012
2013
2014
2015
2016
2017
2018
Figure 12.2 Transition of board structure of Japanese firms
Note: Data is from Yakuin Shikiho (Directory of Executives) by ToyoKeizai
Inc.

First, the exercise of voting rights by institutional investors is

beginning to take on real meaning. In the first half of the 2000s, in
response to the release of guidelines by the Welfare Pension Fund
Association (in 2005 renamed the Pension Fund Association) and
others, investment advisory firms and trust banks have established
their detailed standards for exercising their voting rights. With the
establishment of standards for exercising voting rights, from the mid-
2000s the likelihood that dissenting votes would be cast increased,
and, as a result, the possibility that resolutions proposed by the firm
would be vetoed became a reality.
Second, paralleling this development, there was a modest scale of
activism wave, centering on cash-rich medium-sized firms. Drawing
particular attention were the activities of the Murakami Fund from
2001 to 2004 and Livedoor’s bid to take over Nippon Broadcasting
System in 2004. And in 2006, Oji Paper’s public takeover bid of
Hokuetsu Paper attracted attention for being the first hostile takeover
attempt by a business enterprise. In 2007, the Bulldog Sauce incident
saw the first application of takeover defenses.8
The eruption of the subprime loan crisis in the summer of 2007 and
the Lehman Brothers collapse in September 2008 were turning points
Corporate Governance Reforms under Abenomics 363

that vastly increased skepticism toward the U.S.-type market economiza-

tion trend that had begun in 1997. As selloffs by foreign institutional
investors proceeded, critics emphasized that the swing toward a market
economy had gone too far. There was a rapid decline in interventions by
activists, and use of stock options flagged. The Democratic Party of
Japan’s cabinets paid particular attention to the growing inequality and
were passive toward efforts to impose a stronger US-type governance
system. Furthermore, 2011 saw the eruption of the Olympus scandal,
which was based on an attempt to take care of massive losses through
dodgy accounting. The occurrence of scandals at firms that were con-
sidered advanced in terms of corporate governance gave to rise to skepti-
cism about the role of external directors. Further revisions of the
Company Law were considered from 2011, but, as expectations of the
US model diminished, the focus of corporate governance shifted primarily
to efforts to prevent misconduct.

12.2.3 Corporate Governance Reforms as a Growth Strategy

in Abenomics
Following the first boom of corporate governance reforms in the early
2000s, progress toward reforms began again, when the second Abe
cabinet was formed at the end of 2012, replacing the Democratic Party
of Japan cabinet. Since then, corporate governance reforms have been
pursued as a growth strategy – the third arrow of Abenomics – and
a series of related policies have been implemented.
As part of the Japan Revitalization Strategy of June 2013, the cabinet
announced its aim to promote corporate governance reforms by strength-
ening the fiduciary duties of institutional investors through the JSC, which
was implemented in February 2014. The revised Japan Revitalization
Strategy of June 2014 proposed strengthening Japanese corporate govern-
ance with a view to “revive the earning power of Japanese firms.”
Consequently, reforms were rapidly implemented in 2015, with the draft-
ing and enactment of the CGC. That same year, the revised Companies
Act came into force. In the revision of the Japan Revitalization Strategy of
June 2016, top priority was given to the implementation of effective
governance premised on structural reforms. Important events on govern-
ance reforms are summarized in Table 12.1.
The aim of the corporate governance reform package is clearly
described in Japan Revitalization Strategy 2014:
364 Hideaki Miyajima and Takuji Saito

Table 12.1 Major events related to corporate governance reforms under

Abenomics

December 2012 Formation of the second Abe cabinet.

Establishment of the Headquarters for Japan’s Economic
Revitalization.
February 2013 ISS recommends voting against the appointment of a top
executive “if the board after the shareholder meeting does
not include at least one outsider.”
June 2013 The Japan Revitalization Strategy puts forth the principles
to be followed by institutional investors in fulfilling their
fiduciary duty (e.g., encouraging firms through dialogue
to pursue medium- to long-term growth).
June 2013 Toyota’s general shareholder meeting; three outside
directors appointed. Kyocera, Sumitomo Corporation,
and FANUC follow suit.
January 2014 Start of the JPX400 index.
Canon announces the appointment of outside directors.
February 2014 Creation of the Japanese Stewardship Code (JSC).
June 2014 Revision of Japan Revitalization Strategy. It puts forth
enhanced corporate governance as part of the Emergency
Structural Reform Program.
August 2014 Ito Review, “Competitiveness and Incentives for
Sustainable Growth: Building Favorable Relationships
between Companies and Investors.”
Dai-ichi Life decides to disclose voting results.
October 2014 Revision of ISS proxy voting policies (ROE criterion, board
composition, creation of an audit committee).
March 2015 The Financial Services Agency and the Tokyo Stock
Exchange announce the first draft of Corporate
Governance Code (CGC).
May 2015 Amendment of the Companies Act (audit and supervisory
committee; explanation for not appointing outside
directors; stricter requirements for outside directors).
June 2015 Mizuho Bank announces further dissolution of cross-
shareholding.
2015 Revision of the Japan Revitalization Strategy
Implementation of the CGC.
June 2016 Japan Revitalization Strategy 2016; shift of reforms from
formality to substance.
December 2016 JSC adopted by 214 financial institutions.
Corporate Governance Reforms under Abenomics 365

Table 12.1 (cont.)

March 2017 CGS Study Group’s first-term report.

May 2017 Revised JSC (disclosure of detailed voting results).
June 2017 The GPIF publishes its voting principles and requests insti-
tutions entrusted with asset management to disclose
detailed voting results.
Future Investment Strategy 2017 (cabinet decision).
Aug. 2017 Corporate governance reports; description on the treatment
of retired top executives.
December 2017 CGS Study Group, second term; follow-up conference;
group governance.
March 2018 Revision of the CGC.
June 2018 Future Investment Strategy 2018; operational guidelines for
group governance.
November 2018 Creation of the Study Group on Fair M&A.
June 2019 Operational guidelines on group governance.

What should be done to increase Japanese companies’ earning power, in

other words, medium to long-term profitability and productivity and to pass
the fruits of such increase on to the people (households) evenly? First, it is
important to strengthen the mechanism to enhance corporate governance
and reform corporate managers’ mindset so that they will make proactive
business decisions to win in global competition for the purpose of attaining
targets including globally compatible level in return on equity. Particularly,
companies that have achieved the highest earnings in several years should be
encouraged to proactively use their earnings for new capital investment, bold
business realignment, mergers and acquisitions, and other deals, instead of
accumulating internal reserves.

In this strategy, it was assumed that the conservative behavior of manage-

ment could be changed by setting ROE as a performance criterion, enhan-
cing shareholder engagement, appointing independent directors, and
refining compensation systems. The characteristics of the corporate gov-
ernance reforms under Abenomics can be summarized as follows.9
First, the corporate governance reforms in Japan clearly attempted to
rebalance the influence among various corporate stakeholders by
strengthening shareholder rights, implicitly acknowledging that the
balance had been biased in favor of employees (insiders). Linking
corporate governance reforms to corporate growth is a worldwide
trend after the global financial crisis (Kay 2012; OECD 2012), but
366 Hideaki Miyajima and Takuji Saito

the problems differ by country. In the United States and the United
Kingdom, excessive risk-taking induced by performance-related com-
pensation and myopic management due to strong pressure from share-
holders are the targets of reform efforts.
Second, in general, strengthening shareholders’ rights can associated
with various cost. For instance, it can lead to a higher likelihood of
increasing dividends and shareholder returns at the cost of long-term
investment, while closer monitoring of corporate management by share-
holders can make managers likely to reduce their discretion and initiatives,
and enhancing performance-related compensation system can induce cor-
porate managers to manipulate stock prices for their personal gain.
Therefore, the adoption of the reforms implies that the benefits of strength-
ening corporate governance were considered to exceed their such costs.
Last, it is important to note that strengthening corporate governance
was expected to lead to corporate growth and increased productivity, as
part of a mechanism for equally distributing the benefits of increased
earnings to the public. Since corporate governance reforms directly
increased the proportion of profits allocated to shareholders and corpor-
ate managers, wage of labor would not rise without an increase in the
added value. The key to “equitably distributing the benefits of the reforms
to the public” is that corporate growth spurred by increased investment
must lead to increased productivity and higher value-added levels of
labor, which must then lead to higher wages and income. A kind of
“trickle-down” mechanism was therefore assumed for the distribution
of gains to employees.

12.3 Effects of Stewardship Code

12.3.1 Changes in the Behavior of Institutional Investors

The JSC was published in February 2014 and was modeled on the UK
Stewardship Code of 2010. The JSC has seven principles, including that
institutional investors: have a clear policy on voting rights; have a clear
policy on managing conflicts of interest; and publicly disclose their
voting activity. Immediately after the release of the JSC, 127 institu-
tions agreed to comply with it, including the GPIF, the world’s largest
public pension fund, with the total reaching 214 institutions in 2017.
The implementation of the JSC has led to the significant changes in
the behavior of institutional investors. First, asset owners such as
Corporate Governance Reforms under Abenomics 367

pension funds are involved in the engagement process for the first time.
In Japan, there were no active pension funds equivalent to CalPERS in
the United States (Jacoby 2010). The JSC required Japanese asset
owners to actively monitor asset management firm. The acceptance of
the JSC by the GPIF had significant impact on the behavior of institu-
tional investors. The GPIF asks asset management firms to comply with
the JSC and strictly implement stewardship activities. As a result, to get
business from the GPIF, it is basically imperative for asset management
firms to strictly follow the JSC. Furthermore, in 2017, the GPIF
announced its “Voting Principles,” which require asset management
firms to disclose their voting results for individual proposals in line with
the new disclosure policy from the Financial Services Agencies.
The second change was that implementation of the JSC led to institu-
tional investors setting up independent voting divisions as well as strength-
ening voting standards. For example, Mitsubishi UFJ Trust Bank
established a committee for exercising voting rights within the asset
management department. This committee discussed the basic policy on
the exercise of voting rights and the policy for voting on individual
resolutions, and these were then finalized by the managing director. In
addition, to verify the monitoring results, a stewardship committee was
organized, with the majority being outside third-party members.
In addition, the JSC encouraged financial institutions to revise the
standard for the exercise of voting. The standards which many institutions
revised include (1) profitability standards for appointment of directors; (2)
governance standards; and (3) standards for payout policies. For example,
related to (1), Institutional Shareholder Services (ISS) announced a new
recommendation, which explicitly set a return on equity (ROE) as
a proposed standard for the appointment of directors in October 2014
and suggested voting against the top management at companies with
average ROE below 5 percent over the past five consecutive years.
Regarding governance systems, most institutions set the standard for
appointing multiple outside directors. The ISS recommended, for
example, voting against appointment of a top executive “if the board
does not include at least one outsider, regardless of independence.”
Regarding payout policy, Daiwa SB Investment Ltd. set the following
standard: “for companies with excessive cash, over 50 percent of the
total profit will be expected to be paid as dividend.”
Third, an important outcome of the JSC is to transform insurance
companies from “silent” partners to “long-term shareholders with
368 Hideaki Miyajima and Takuji Saito

a voice.” The average shareholding of insurance companies declined

from 14 percent in 1995 to 11 percent in 2004 and to 5.7 percent in
2014. Also, insurance companies are investing in fewer firms, with
a decrease for Nippon Life from 1,412 in 2000 to 1014 in 2014, and
a decrease for Dai-ichi Life Insurance from 1,054 to 501.10 In addition,
they first disclosed the results of their voting activity after the agreeing
to the JSC. In 2017, Dai-ichi Life Insurance decided to disclose their
voting results for individual firms. While the ratio of votes against
companies’ proposals was low, life insurance companies emphasized
long-term dialogue with the companies in which they invested, rather
than the exercise of their voting right.11

12.3.2 Improving the Attractiveness of the Japanese Stock

Market
Corporate governance reforms under Abenomics were expected to
improve the attractiveness of the Japanese stock market through the JSC
coupled with reorganization of corporate boards following the implemen-
tation of the CGC, which is discussed in Section 12.4. The reforms were
expected to mitigate the information asymmetry between institutional
investors and publicly traded firms in Japan. In particular, the JSC was
expected to increase the institutional investor ownership ratio and
strengthen their influence on corporate decision-making. Let us next
examine the extent to which firms’ ownership structure changed during
the corporate governance reforms. Table 12.2 focuses on firms listed on
the First Section of the Tokyo Stock Exchange (hereinafter, TSE-1 firms)
and shows the institutional investor ownership ratio.
First, it should be noted that the average institutional investor own-
ership ratio significantly increased from 19.8 percent in 2012 to
27.8 percent in 2017. The latter percentage exceeds the highest per-
centage observed before the global financial crisis (approximately
23 percent in 2005 and 2006) by almost 5 percentage points. Of the
increase of about 8 percentage points between 2012 and 2017,
increased shareholding by foreign institutional investors accounts for
4.7 percentage points, and increased shareholding by domestic institu-
tional investors accounts for 3.3 percentage points. The increased
shareholding by domestic institutional investors was due to the
increase in stockholding of the GPIF that started in 2015 and due to
ETF purchases by the Bank of Japan.
Corporate Governance Reforms under Abenomics 369

Table 12.2 Distribution of institutional investor ownership

Change from
2010 2012 2014 2017 2012 to 2017

Mean (%) 19.5 19.8 23.3 27.8 8.0

Median (%) 16.0 16.4 20.4 25.6 9.3
Standard deviation 15.4 15.6 16.2 16.5
(%)
Coefficient of 0.790 0.788 0.694 0.592
variation
Quintiles of market value of equity at the end of each year
First quintile
Mean 6.0 6.5 8.9 13.1 6.6
Median 2.7 4.9 7.4 11.3 6.4
Standard deviation 6.4 6.6 7.6 8.5
Second quintile
Mean 12.1 11.7 15.1 20.4 8.6
Median 9.9 9.5 14.0 18.6 9.2
Standard deviation 9.8 9.1 10.3 11.7
Third quintile
Mean 17.4 18.2 22.3 27.9 9.7
Median 15.6 17.1 21.2 26.6 9.5
Standard deviation 10.3 11.0 12.1 13.0
Fourth quintile
Mean 26.5 26.9 30.0 34.0 7.1
Median 25.3 26.7 29.7 33.6 6.9
Standard deviation 12.5 13.3 13.1 14.5
Fifth quintile
Mean 36.9 37.2 40.7 44.6 7.4
Median 36.2 37.8 41.1 45.6 7.8
Standard deviation 13.8 13.8 13.4 13.5

Note: This table shows the distribution of institutional investor ownership of the firms
listed on the First Section of the Tokyo Stock Exchange.

A second point to note is a decrease in the variation among firms in

the institutional investor ownership ratio. This stands in contrast to the
fact that the variation across different firm sizes increased from 1997 to
2005 (Miyajima & Kuroki 2007). The behavior of foreign institutional
investors is strongly biased as they tend to prefer larger firms with
370 Hideaki Miyajima and Takuji Saito

higher liquidity, and familiarity. As Table 12.2 shows, the average

institutional investor ownership ratio was 37.2 percent for the firms
in the fifth quintile of market capitalization as of the end of FY 2012. In
contrast, the averages are 18.2 percent for the firms in the third quintile
and 6.5 percent for the firms in the first quintile. A similar trend was
widely observed in the United States and other countries in the 1990s
(Gompers & Metrick 2001). However, as progress was made in cor-
porate governance reforms in Japan, the investment activity of foreign
institutional investors expanded to include small- and mid-cap stocks.
Table 12.2 shows that increases in the institutional investor ownership
ratio are large among the firms in the second and third quintiles of
market capitalization, at 8.6 percent and 9.7 percent, respectively, both
of which are above the overall average of 8.0 percent.
As for differences in institutional ownership level across large firms,
it is clear that the firms included in the MSCI Japan Index or the JPX-
Nikkei 400 (JPX400) index equally face pressure from institutional
investors, while small and medium-sized firms are subject to less pres-
sure from institutional investors. Corporate governance reforms, how-
ever, began to lessen this difference, and small and medium-sized firms
that had received little pressure from institutional investors have been
increasingly subject to the engagement of these shareholders.

12.4 Corporate Governance Code: Board Reforms

12.4.1 Major Aims of the Corporate Governance Code

The second pillar of corporate governance reforms under Abenomics was
the Japanese CGC, which was modeled on the OECD Principles of
Corporate Governance. Intensive discussion and drafting on the CGC
was started in late 2014, and it was soon adopted in June 2015, reflecting
the serious efforts of the Abe cabinet to move forward on this issue. The
CGC comprises five general principles, thirty principles, and thirty-eight
supplementary principles (a total of seventy-three items). The five general
principles are: (1) securing the rights and equal treatment of shareholders;
(2) appropriate cooperation with stakeholders other than shareholders;
(3) ensuring appropriate information disclosure and transparency; (4)
responsibilities of the board; and (5) dialogue with shareholders.
We focus on the effects of the CGC on board structure and cross-
shareholding, considering that they would have the most significant
Corporate Governance Reforms under Abenomics 371

impact on the essential characteristics of corporate governance of

Japanese firms.

12.4.2 Changes in the Board Structure

Board structure of Japanese firms have changed dramatically under the
corporate governance reforms, because the rules related to board struc-
ture have been revised. First, Principle 4.8 of CGC requires firms to
appoint at least two independent outside directors. If this requirement
is not satisfied, they must provide a sufficient reason for not adopting
this practice. Second, the revised Companies Act made it possible for
listed firms to choose the new option of becoming a company with
audit and supervisory committee (kansa tou iinkai secchi gaisha), in
addition to the conventional system of a company with company
auditors (kansa-yaku kai secchi gaisha) and previously introduced
new system of a company with nominating committee (shimei iinkai
tou secchi gaisha). The audit and supervisory committee must have
more than three directors, and the majority of them must be outside
directors, indicating that companies with audit and supervisory com-
mittees must have at least two outside directors. Third, Principle 4.10
of the CGC states that firms should employ optional approaches (for
example, optional nomination committee and optional remuneration
committee, to which independent outside directors make significant
contributions), as necessary, to further enhance governance functions.
Contrasting to the fact that outside directors had previously been
appointed at a slow pace, the revision of the rules promoted appointment
of outside directors. As seen in Table 12.3, the average number of outside
directors drastically increased from 1.2 in 2012 (before introduction of the
CGC) to 2.7 in 2018 (after introduction of the CGC). The proportion of
outside directors on corporate boards also jumped from 14.0 percent to
29.3 percent. As Figure 12.2 shows, the average number of outside
directors gradually increased since 2004, but it took ten years to exceed
one. By 2018, however, the percentage of firms with at least two outside
directors reached 94.6 percent, which shows that most firms complied
with the CGC instead of providing explanations for noncompliance.
The JPX400 is an index of major firms that have been primary
investment targets of foreign institutional investors. The proportion
of JPX400 firms with no outside director had already declined to
24 percent by the end of FY 2013. Many of the firms in this group
372 Hideaki Miyajima and Takuji Saito

Table 12.3 Characteristics of boards of directors

JPX400 Non-JPX400
All firms firms firms

2013 2018 2013 2018 2013 2018

Mean number of directors 8.6 9.2 10.4 10.6 8.1 8.8

Mean number of outside 1.2 2.7 1.8 3.3 1.0 2.5
directors
Number of outside directors 37.6 0.3 24.3 0.0 41.3 0.4
= 0 (%)
Number = 1 (%) 31.7 5.1 26.1 1.3 33.2 6.0
Number = 2 (%) 17.1 49.4 22.5 31.1 15.6 53.7
Number = 3 or more (%) 13.7 45.2 27.2 67.7 9.9 39.9
Outsider majority board (%) 2.6 4.9 6.3 7.8 1.5 4.2
Company with nominating 2.5 2.9 6.0 8.8 1.5 1.5
committee (%)
Company with audit and - 24.4 - 15.0 - 26.6
supervisory committee (%)
Company with voluntary 5.9 31.4 14.6 52.4 3.5 26.5
nominating committee (%)
Company with voluntary 8.6 34.9 19.1 54.9 5.7 30.2
compensation committee
(%)

Note: This table shows the board structure of the firms listed on the First Section of the
Tokyo Stock Exchange. The information is from the Board Survey by Tokyo Stock
Exchange and Yakuin Shikiho (directory of directors) published by Toyo Keizai
Shinposha.

had voluntarily appointed outside directors before the introduction of

the CGC because of the strong demand from institutional investors.
From a different angle, however, this means that more than 20 percent
of these firms, including Toyota, Canon, and FANUC, maintained an
all-insider board of directors, despite pressure from institutional invest-
ors. Therefore, another result of the CGC is that firms that had avoided
voluntarily appointing outside directors implemented major reforms
around the time of the CGC’s introduction.
As of 2018, the proportion of firms that became companies with an
audit and supervisory committee in response to the 2015 amendment
to the Companies Act was 24.4 percent, and about a quarter of TSE-1
Corporate Governance Reforms under Abenomics 373

firms became this type of company. In particular, a number of relatively

small firms made the transition. Of non-JPX400 firms, 26.6 percent
made this change, whereas 15.0 percent of JPX400 firms did so. Even
though the introduction of the CGC brought significant changes to
Japanese firms’ boards, few firms applied a US-type system of company
with nominating committee. The proportion of companies with nom-
ination committee rose by only 0.4 percentage points between 2013
and 2018. A reason might be that firms are very wary of the possibility
that outside directors in the nomination committee gain control over
important personnel-related decisions.
It should be also noted that a number of firms set up a nomination
committee and a compensation committee, which were voluntary, when
the CGC was introduced. By 2018, about 30 percent of the TSE-1 firms
had a voluntary committee. More than half of JPX400 firms have
a voluntary committee. If the company with nominating committee is
also taken into account, about 60 percent of firms have a nomination
committee.
Japanese firms’ boards of directors significantly changed when the
CGC was introduced; however, these boards remain less independent
than boards of the firms in the United States where the average propor-
tion of independent outside directors on a board of directors is around
70 percent (Gordon 2007). Therefore, even if firms set up a voluntary
committee when the Code was introduced, they did not fully transition
from a management board system to a monitoring board system in
which the board specializes in supervising management activities.

12.4.3 Effects of Board Reforms

We examine whether the increase in the number of outside directors
following the implementation of the CGC and other policies lead to
better performance. The reforms were implemented based on the
assumption that there were firms with fewer outside directors than
optimal, that the introduction of the CGC would induce such firms to
increase outside directors, and that management policies would subse-
quently change to improve corporate performance. Fauver et al. (2017)
analyzed the effect of board-related reforms in forty-one countries and
reported that the reforms led to an increase in corporate value, espe-
cially in countries that adopted a comply-or-explain rule.
374 Hideaki Miyajima and Takuji Saito

On the other hand, many Japanese firms voluntarily appointed out-

side directors since the early 2000s, meaning that many firms had
already had optimal number of outside directors. It is therefore possible
that the introduction of the CGC has given firms with no need for
outside directors no choice but to appoint them. It is also possible that
such firms appoint outside directors just to comply with the CGC and
do not take advantage of their presence on the board. In both cases,
appointing outside directors would not affect management policies or
corporate performance. Duchin, Matsusaka & Ozbas (2010) show
that on average the increase in the number of outside directors intended
to comply with the Sarbanes–Oxley Act and listing regulations did not
lead to an increase in corporate value or profit.
To examine the validity of the two scenarios, we estimate the causal
effect of an increased number of outside directors on corporate per-
formance and corporate behavior using the introduction of the CGC as
exogenous shock. It is difficult to accurately ascertain the effects of
a corporate governance system on corporate performance due to the
reverse causality (i.e., corporate governance affects corporate behavior
and performance, while corporate performance affects governance).
Furthermore, there is the problem of endogeneity as it is impossible to
control for all factors that influence both performance and governance.
We use the difference-in-difference method and instrumental variables
approach for overcoming this problem.
We assume that the appointment of outside directors following the
introduction of the CGC was intended to comply with Principle 4–8,
which stipulates that a firm should appoint at least two independent
outside directors. In other words, the increase in the number of outside
directors was caused not by changes in corporate attributes or other
factors but by an exogenous measure (i.e., a policy). Therefore, estima-
tion of the causal effect of the appointment of outside directors is
considered possible by comparing changes in the corporate perform-
ance and behavior of firms that did not fulfill the CGC’s principle prior
to its implementation but subsequently appointed outside directors,
with changes in the corporate performance of firms that already ful-
filled the principle even before the CGC’s implementation.
We empirically analyze a sample of TSE-1 firms in 2013. The
dependent variable is change in corporate performance before and
after the introduction of the CGC, and the independent variables
include change in the number of outside directors. The instrumental
Corporate Governance Reforms under Abenomics 375

variable is a dummy variable that takes a value of 1 if the number of

outside directors was less than two before the introduction of the Code
and zero otherwise. Approximately 60 percent of firms had fewer than
two outside directors before the introduction of the Code, meaning that
they did not fulfill Principle 4–8. About 80 percent of these firms had
appointed two or more outside directors as of one period after the
introduction of the Code and fulfilled this principle. In fact, at the first
stage of analysis, in which the dependent variable is change in the
number of outside directors, the coefficient of the dummy variable is
approximately 1 and is statistically significant at the 1 percent level.
Table 12.4 shows the results of estimating the effects of change in the
number of outside directors on firm performance. The results do not
show that an increase in the number of outside directors led to
a statistically significant improvement in performance as measured by
ROA, ROE, and Tobin’s q. These results suggest that the second
scenario is valid. It is possible, however, that the effectiveness of outside
directors varies depending on corporate attributes. Therefore, analysis
is performed including interaction terms (Change in number of outside
directors × Firm size (sales); Change in number of outside directors ×
Foreign ownership ratio; Change in number of outside directors ×
Managerial ownership ratio). The coefficient of the interaction term
(Change in number of outside directors × Foreign ownership ratio) is
negative and significant in some estimation models.
This result implies that a greater performance improvement follows
an increase in the number of outside directors for a firm that has a lower
foreign ownership ratio and faces less discipline from capital markets.
For example, a 10 percentage point decrease in the foreign ownership
ratio is associated with a 0.8 percentage point increase in ROA if the
number of outside directors increases by two. The coefficient of the
interaction term (Change in number of outside directors × Managerial
ownership ratio) is positive and significant in some estimation models.
More specifically, a 10 percentage point increase in the management
ownership ratio is associated with a 1.7 percentage point increase in
ROE if the number of outside directors increases by two. This result
implies that greater performance improvement follows an increase in
the number of outside directors for a family-run firm where the found-
ing family has strong managerial control.
As for the increased risk-taking anticipated under Abenomics, on
average there is no evidence of an increase in the number of outside
Table 12.4 The effect of outside directors on corporate performance

Time = −1 to 1 −1 to 2 −1 to 3 −1 to 1 −1 to 2 −1 to 3

Dependent variable: ROA

Δnumber of outside directors −0.179 −0.059 −0.110 −2.686 ** −2.120 −2.809 *
(0.303) (0.215) (0.150) (1.181) (1.366) (1.491)
x LN(Sales） 0.233 *** 0.219 ** 0.239 *
(0.080) (0.107) (0.133)
x Foreign ownership −0.026 −0.041 *** −0.023
(0.019) (0.015) (0.021)
x Managerial ownership 0.028 0.011 0.013
(0.019) (0.038) (0.033)
Dependent variable: ROE
Δnumber of outside directors 0.461 0.165 −0.350 −5.816 −2.589 −7.049 ***
(0.793) (0.350) (0.253) (4.896) (3.237) (1.992)
x LN(Sales） 0.660 * 0.338 0.584 ***
(0.390) (0.282) (0.166)
x Foreign ownership −0.057 −0.077 *** −0.029
(0.037) (0.025) (0.023)
x Managerial ownership 0.080 0.036 0.085 **
(0.062) (0.045) (0.037)
Dependent variable: Tobin’s q
Δnumber of outside directors −0.010 −0.012 0.007 −0.435 ** −0.290 −0.310
(0.027) (0.022) (0.028) (0.199) (0.186) (0.219)
x LN(Sales） 0.032 * 0.024 0.022
(0.017) (0.016) (0.020)
x Foreign ownership −0.001 −0.003 0.000
(0.002) (0.003) (0.003)
x Managerial ownership 0.006 0.008 * 0.016 **
(0.006) (0.004) (0.006)

Note: This table shows the second stage results of instrumental variable regressions. In parentheses are robust standard errors corrected for
clustering at the firm level. *** denotes 1 percent level of significance, ** 5 percent level of significance, and * 10 percent level of significance.
378 Hideaki Miyajima and Takuji Saito

directors leading to an increase in capital expenditure or R&D expend-

iture even when corporate attributes are taken into account.

12.5 Corporate Governance Reforms and Reduction

of Relational Shareholdings

12.5.1 Promoting Sales of Relational Shareholdings

Principle 1–4 of the CGC demands that firms explain the reason for
relational shareholding:
When companies hold shares of other listed companies as cross-
shareholdings, they should disclose their policy with respect to doing so. In
addition, the board should examine the mid- to long-term economic ration-
ale and future outlook of major cross-shareholdings on an annual basis,
taking into consideration both associated risks and returns. The annual
examination should result in the board’s detailed explanation of the objective
and rationale behind cross-shareholdings.
The amount of stocks held by banks fell to about 3 percent of stocks
held by TSE-1 firms as a result of the rapid decline of cross-
shareholding with industrial corporations that started in 2000. Thus,
relational shareholding consequently started to be mainly conducted
between nonfinancial companies. The introduction of the CGC is
considered to have provided an opportunity for those firms to reexa-
mine whether they should continue their relational shareholding or sell
stocks held for relational purposes.
The market positively reacted to firms’ reductions of relational share-
holding. For example, Yamaha announced after the end of the trading
session on November 28, 2017, that it would sell 8 million cross-held
shares of Yamaha Motor and, at the same time, decided to use the
proceeds from the sale to buy back 3.7 percent (7 million shares) of
Yamaha’s outstanding shares, with a spending limit of 25 billion yen.
The stock price of Yamaha rose by 7 percent next day as the market
positively viewed the firm’s improved ROE and shareholder returns.12

12.5.2 Relational Shareholding and Corporate Governance

Reform
Table 12.5 shows firm characteristics classified by the level of relational
shareholdings. Firms with high relational holdings have lower
Corporate Governance Reforms under Abenomics 379

Table 12.5 Firm characteristics classified by the level of relational

shareholdings

Ratio of the value of relational

shareholdings to assets

All 1 (low) 2 3 4 (high)

Value of relational 4.13 0.26 1.76 4.23 10.27

shareholdings to
assets (%)
Number of relational 48.6 10.0 44.0 66.3 74.2
shareholdings
Cross-shareholdings to 8.1 1.4 5.5 9.7 16.0
total share (%)
ROA (three-year 6.3 8.7 6.1 5.5 5.1
average) (%)
ROE (three-year 5.3 7.5 4.9 5.4 3.4
average) (%)
Tobins’s q (three-year 1.00 1.15 0.99 0.96 0.92
average)
Institutional ownership 21.4 21.4 22.5 21.4 20.3
(%)
Foreign ownership (%) 13.3 14.1 14.3 12.9 12.0
Debt to assets (%) 18.2 18.3 19.0 19.1 16.3
Total assets (million 420,564 300,185 562,480 510,099 308,559
YEN)
Sales (million YEN) 344,468 246,934 461,497 415,399 253,281
Market share value 188,633 151,455 231,680 228,999 142,082
(million YEN)
R&D to sales (three- 1.70 1.20 1.81 1.72 2.08
year average) (%)
CAPEX to PPE (three- 13.5 16.3 13.4 12.5 12.1
year average) (%)
M&A to assets (three- 0.12 0.12 0.15 0.11 0.10
year average) (%)

Note: This table shows firm characteristics classified by the ratio of the value of
relational shareholdings to assets. Sample is the firms listed on the First Section of the
Tokyo Stock Exchange.
380 Hideaki Miyajima and Takuji Saito

profitability, lower CAPEX and M&A expenditures, and a lower per-

centage of foreign ownership. This result is consistent with Ikeda,
Inoue & Watanabe (2018) that shows firms with high cross-
shareholding were likely to have had low performance due to enjoying
the so-called quiet life. The Abe cabinet and other policymakers have
paid considerable attention to this inverse correlation between high
relational/cross-shareholding and corporate performance.
The JSC states that institutional investors should become actively
engaged in managerial decisions of firms in which they hold stocks.
Promoting efficient use of capital is one of the main concerns in such
investor engagement. The CGC states that firms should provide
reasons for relational shareholding and explain their validity. After
the introduction of the CGC, all publicly traded firms explained their
reasons for relational shareholding in their corporate governance
reports, following Principle 1–4. These movements might make man-
agers decrease the relational shareholdings.
For capturing a firm’s decisions on relational shareholding, Jidinger
and Miyajima (2019) analyze the actual number of shares of relational
shareholding, which is available from the end of FY 2010, when
amendment of information disclosure rules first required firms to dis-
close details of their relational shareholdings.13
Focusing on firms that have a high relational shareholding ratio and
are therefore often called “bedrock firms” of cross shareholding, they
show that the percentage of firms that decided to sell increased consid-
erably, from 36 percent in 2012 before the CGC to 51 percent in 2015
and to 89 percent in 2017. A substantial increase in the number of
stocks sold was also observed following the enactment of the CGC in
2015. From 2015 to 2017, the total number of stocks to sell increased
from 330 in 2012, to 482 in 2015, and to 1073 in 2017.
Thus, the corporate governance reforms have been effective not only
for all listed companies but also implicitly for firms with high relational
shareholdings. To better understand the effects, we first estimate the
determinants of the decision to sell relational shareholding.

12.5.3 Determinants of Selling Relational Shareholdings

Which stocks did “bedrock firms” sell? Jidinger and Miyajima (2019)
analyzed their annual decisions on selling relational shareholdings by
estimating, first, a fixed-effect model in which the dependent variable is
Corporate Governance Reforms under Abenomics 381

the number of firms issuing stocks sold by the bedrock firm each year;
and second, a logit model focusing on each individual shareholding
relationship, in which the dependent variable indicates sale or con-
tinued shareholding for each year.
First, the results of their analysis revealed the following tendencies:
a bedrock firm with a higher ratio of relational shareholdings to total
assets was more likely to sell; a bedrock firm having smaller latent gains
from relational shareholdings was more likely to sell; stocks with larger
latent gains were sold before those with smaller latent gains; and
a bedrock firm with a greater need to secure liquidity by selling share-
holdings (i.e., a firm with a higher debt ratio or a lower interest
coverage ratio) was more likely to sell. Put simply, the decision on
whether to sell relational shareholdings was basically made in
a financially rational manner.
Second, the introduction of the CGC has a significant positive rela-
tionship with sale of shares in both (1) the analysis where the number of
firms issuing the stocks sold by a bedrock firm is the dependent variable
and (2) the analysis of the decision on whether to sell a shareholding in
a firm. For example, the estimation results for the first analysis shows
that, all things being equal, the introduction of the CGC increased the
number of firms issuing stocks sold by a bedrock firm from 2.5 to 3.4.
Third, a bedrock firm with a high foreign ownership ratio was less
likely to sell relational shareholdings. This result contrasts with the
tendency for TSE-1 firms with a higher institutional investor ownership
ratio to have, on average, a lower relational shareholding ratio and
a lower cross-shareholding ratio. The characteristic of bedrock firms
that their relational shareholding ratio is high when the proportion of
their shares held by institutional investors is high implies that one of the
motivations for relational shareholding is to block pressure from cap-
ital markets. The corporate governance reforms have not been effective
enough in mitigating this entrenchment effect.
Lastly, the second analysis made it possible to examine the cross-
shareholding relationship between two firms and revealed that the
existence of a cross-shareholding relationship formed the foundation
of relational shareholding. The corporate governance reforms, cen-
tered on the JSC and the CGC, induced firms to sell relational share-
holdings in other firms despite the existence of cross-shareholding
relationships with them.
382 Hideaki Miyajima and Takuji Saito

In sum, corporate governance reforms based on a comply-or-explain

rule was effective in reducing bedrock firms’ relational shareholdings,
which had hardly changed previously.

12.6 Consequences of Corporate Governance Reforms under

Abenomics

12.6.1 Corporate Performance after Corporate Governance

Reforms
Since the start of the second Abe cabinet in 2012, cross-shareholding
has declined, the institutional investor ownership ratio has increased,
and dialogue of institutional investors with firms on long-term corpor-
ate management and financial policy has become increasingly active
due to the introduction of the JSC and the CGC. A question now arises
as to whether these changes influenced firms’ decision-making and gave
rise to the virtuous economic circle envisioned by Abenomics.
Table 12.6 shows the performance of TSE-1 firms. From 2012, when
the Abe cabinet was formed, to 2017, mean ROE increased by 2.8 percent-
age points, from 5.8 to 8.6 percent, and median ROE increased by
2.4 percentage points. Also, mean and median ROE in 2017 exceeded
values before the global financial crisis by 1.6 and 1.3 percentage points,
respectively. To examine what caused this increase, ROE was decomposed
into net profit margin (net profit divided by sales), total asset turnover
(sales divided by total assets), and financial leverage (total assets divided by
equity). Mean and median net profit margin increased by 1.9 and 1.4 per-
centage points, respectively, between 2012 and 2017, which contributed
to the increase in ROE. In contrast, both mean and median total asset
turnover fell by 0.04 between 2012 and 2017, which lessened the increase
in ROE. Similarly, mean and median financial leverage decreased by 0.29
and 0.12, respectively, also lessening the increase in ROE. These results
show that the increase in ROE between 2012 and 2017 is attributable to
improved profitability but not to increased asset efficiency or use of debt.
Contrary to what was anticipated under Abenomics, investment did
not markedly increase. From 2012 to 2017, the mean of the ratio of
capital expenditure to tangible fixed assets increased by only 1.2 per-
centage points from 15.9 to 17.2 percent. Over the same period, the
mean R&D intensity (R&D expenditure divided by sales) decreased by
0.2 percentage points, from 2.7 to 2.5 percent. While the mean of
Table 12.6 Average firm performance and behavior

ROE ROA Operating income Dividend Payout to CAPEX to Acquisition R&D Debt to Cash to
(%) (%) to sales (%) to equity equity fixed assets to assets to sales assets assets

2007 6.1 7.3 6.7 0.021 0.030 0.170 0.002 0.024 0.188 0.187
2008 −1.6 5.2 4.4 0.024 0.036 0.157 0.002 0.027 0.215 0.199
2009 1.5 4.8 4.3 0.018 0.023 0.120 0.002 0.028 0.209 0.229
2010 5.4 6.5 6.1 0.018 0.023 0.126 0.002 0.026 0.200 0.248
2011 5.0 6.3 5.9 0.019 0.024 0.140 0.003 0.027 0.193 0.254
2012 5.8 6.4 5.9 0.019 0.024 0.159 0.003 0.027 0.189 0.272
2013 7.5 7.0 6.6 0.019 0.024 0.168 0.003 0.025 0.179 0.275
2014 7.3 7.1 6.8 0.019 0.026 0.169 0.003 0.025 0.175 0.280
2015 7.4 7.3 7.2 0.022 0.031 0.174 0.003 0.025 0.177 0.297
2016 7.9 7.3 7.3 0.023 0.032 0.172 0.004 0.026 0.176 0.307
2017 8.6 7.5 7.7 0.024 0.031 0.172 0.004 0.025 0.170 0.311
2018 8.1 7.5 7.8 0.026 0.035 0.191 0.004 0.025 0.167 0.327

Note: This table shows average performance and behavior of the firms listed on the First Section of the Tokyo Stock Exchange.
384 Hideaki Miyajima and Takuji Saito

a measure of cash acquisitions of firms (total acquisition divided by

total assets) increased by 0.1 percentage points, from 0.3 to 0.4 percent,
its median was zero. This shows that most firms do not engage in
corporate acquisition, and some large acquisitions (cross-border
M&A) seems to have driven mean values.
Payout for shareholders showed a statistically significant increase
during the Abenomics period. From 2012 to 2017, the mean ratio of
annual dividends to equity rose 0.5 percentage points, from 1.9 to
2.4 percent. Also, the mean ratio of total annual dividends plus stock
buybacks to equity increased 0.7 percentage points from 2.4 percent to
3.1 percent. The increase in this ratio accounts for about a quarter of
the increase in ROE, which shows that a major part of the profit
increase was used for dividends and stock buybacks.
Not all the profit increase was used for payout. Some of it was used to
reduce debt, and the rest was held by firms as retained earnings. The
mean cash and equivalents ratio ([cash and equivalents + securities] ÷
[total assets − cash and equivalents − securities]) increased by 3.8 per-
centage points, from 27.2 to 31.1 percent.
Taken together, these results show that, after the start of the Abe
administration in 2012, firms’ earning creation increased due to depre-
ciation of the yen, corporate tax cuts, and other factors, and ROE rose
mainly because of firms’ improved profitability. The increased profit
led to increased payout but not to increased investment. Therefore, the
amount of cash held by firms increased, contrary to what was antici-
pated under Abenomics.

12.6.2 Effects of Corporate Governance Reforms

We empirically examine whether the corporate governance reforms
contribute to the changes in firm performance and behavior. If the
reforms have improved the environment for institutional investors to
play a role in corporate governance, the effects of institutional owner-
ship on firm performance and behavior is expected to become stronger.
To test this hypothesis, the following simple model is estimated.
Pi;t ¼ F ðINSTi;t ; CGR; INST  CGR; Xi;t ; Yt Þ
Pi,t is a variable that measures firm performance or behavior. ROE,
ROA, and the operating profit margin are used as measures of
Corporate Governance Reforms under Abenomics 385

corporate performance. The ratio of debt to total assets, the ratio of

total annual dividends to equity, the ratio of total annual dividends plus
stock buybacks to equity, the ratio of capital expenditure to physical
fixed assets, R&D intensity, the ratio of acquisitions to total assets, and
the cash and equivalents ratio are used as measures of corporate
behavior. Xi,t denotes control variables that measure various corporate
attributes such as firm size. Yt is a year dummy. INSTi,t denotes the
institutional investor ownership ratio or the foreign institutional
investor ownership ratio at the beginning of each year. CGR is
a dummy variable that takes a value of 1 for 2014, when the JSC was
implemented, and subsequent years. INST*CGR is an interaction term
between the institutional investor ownership ratio and the CGR
dummy. The main focus of this analysis is on this interaction term. If
the CGR changed institutional investors’ attitudes, a higher institu-
tional investor ownership ratio would lead to a greater change in
corporate behavior or performance.
The analysis considers TSE-1 firms and the period from 2010 to
2018. A fixed-effects model is used for the estimation to control for
firm-specific effects. It must be noted that causal relationships are not
fully considered in the following analysis and that some results may
reflect reverse causality.14
Table 12.7 summarizes the estimation results. For 2010 and subse-
quent years, the institutional investor ownership ratio has a significant
positive association with ROA and operating profit margin. These
results are consistent with the results of previous studies, including
those by Aggarwal et al. (2011) and Ferreira and Matos (2008).
However, the coefficient of the interaction term is not statistically
significant: the result does not show that the positive association
between institutional investor ownership ratio and ROA was enhanced
by the implementation of the JSC. Therefore, the analysis does not find
evidence that firms with a higher institutional investor ownership ratio
gained extra profitability due to the implementation of the JSC.
The effects of the corporate governance reforms on investment are not
clear. There is a positive and significant association between capital
expenditure and the institutional investor ownership ratio. The coefficient
of the interaction term, however, is negative and significant, indicating
that, while capital expenditure tends to increase as the institutional
investor ownership ratio increases, this tendency weakened after the
implementation of the JSC. In addition, the institutional investor
Table 12.7 The effect of institutional ownership on firm performance and behavior

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10)

Operating CAPEX to
income to Payout to Dividend fixed Acquisition R&D Debt to Cash to
Variables ROE ROA sales equity to equity assets to assets to sales assets assets

Institutional 0.024 0.043*** 0.039*** 0.019*** 0.012*** 0.092*** 0.001 0.004* -0.116*** 0.025**
ownership (0.021) (0.007) (0.008) (0.006) (0.002) (0.029) (0.003) (0.002) (0.016) (0.012)
Institutional 0.001 −0.004 0.000 0.015*** 0.003** −0.058*** 0.001 −0.000 0.001 −0.016*
ownership (0.011) (0.005) (0.005) (0.003) (0.001) (0.015) (0.001) (0.002) (0.010) (0.009)
× CGR
dummy
Observations 15,456 15,456 15,448 15,322 15,322 15,299 15,299 9,689 15,322 15,322
R-squared 0.053 0.111 0.117 0.058 0.129 0.027 0.031 0.049 0.231 0.068
Number of 1,982 1,982 1,982 1,964 1,964 1,964 1,964 1,252 1,964 1,964
firms

Note: Firm fixed effects are used to control for time-invariant firm characteristics. Control variables are total assets, foreign sales to total sales, debt
to assets, cash to assets, cash flow to assets, Tobin’s q and year dummies. In parentheses are robust standard errors corrected for clustering at the firm
level. *** denotes 1 percent level of significance, ** 5 percent level of significance, and * 10 percent level of significance.
Corporate Governance Reforms under Abenomics 387

ownership ratio is not positively associated with M&A expenditure and

R&D expenditure throughout the sample period, even during the reform
period. The above estimation results do not provide evidence that capital
investment, M&A expenditure, and R&D expenditure increased due to
reforms, despite what was anticipated under Abenomics.
On the other hand, the relationship between the institutional
investor ownership ratio and payout is clearly strong. The institutional
investor ownership ratio is significantly positively associated with the
ratio of total annual dividends to equity and with the ratio of total
annual dividends plus stock buybacks to equity. Previous studies have
found a similar tendency in the United States (Grinstein & Michaely
2005; Crane, Michenaud & Weston 2016). The coefficient of the
interaction term between the institutional investor ownership ratio
and the CGR dummy is positive. In particular, it is statistically signifi-
cant at the 1 percent level in the model whose dependent variable is the
ratio of total annual dividends plus stock buybacks to equity. The
coefficient of the interaction term is 0.015 whereas the coefficient of
the institutional investor ownership ratio is 0.019, which suggests that
the effect of the implementation of the CGR on profit allocation was
quite substantial. It is therefore likely that, after implementation of the
Code, the tendency for shareholder returns to increase with an increas-
ing institutional investor ownership ratio became stronger.
The institutional investor ownership has a significantly negative
effect on the ratio of debt to total assets and a positive effect on the
ratio of cash to total assets. The interaction terms have opposite signs.
These results show that, contrary to what would normally be assumed,
firms with a high institutional investor ownership ratio had less debt
and more cash. After the implementation of the CGR, increased insti-
tutional investor ownership reduced this tendency. However, the esti-
mated coefficients of interaction term are small, and the overall effect
(the coefficient of institutional ownership + that of interaction term) is
still negative for debt dependence and positive for cash holdings, sug-
gesting that the corporate governance reforms did not mitigate the
tendency for debt reduction and cash accumulation.
In sum, under the favourable macro condition of Abenomics, profit
increased due to an increase in profit margins. But, the increased profits
were not necessarily directed to investment, and some of the increased
profits were used for payout and debt reduction, with firms holding the
rest as cash.
388 Hideaki Miyajima and Takuji Saito

12.7 Conclusion
The year 2015 is often referred to as the first year of corporate govern-
ance in Japan. Yet corporate governance reforms under Abenomics did
not bring discontinuous changes to corporate governance systems. It
had already been evolving since the 1997 banking crisis. Japanese firms
were once commonly characterized by the main bank system, cross-
shareholding, and boards consisting of internally appointed directors.
However, their reorganization efforts after the banking crisis separated
hybrid firms – whose external governance systems had significantly
shifted to market-based systems and had been combined with trad-
itional relationship-based internal governance mechanisms – from
traditional Japanese firms that continued to lag behind in terms of
external and internal governance (Jackson & Miyajima 2007). What
kinds of changes, then, did the reforms bring to Japanese corporate
governance systems?
The series of reform measures implemented were significant because
they encouraged institutional investors to exercise their voting rights
and engage in corporate governance; strongly promoted board
reforms; promoted efficient capital use and reduction of cross-
shareholdings by inducing firms to reexamine their “bedrock” rela-
tional shareholdings; and increased the attractiveness of the Japanese
market, which led to increased investments from overseas.
Comparison of the changes resulting from corporate governance
reforms under Abenomics with results from the first peak of reforms
following the banking crisis can be summarized as follows. Whereas
earlier reforms introduced a hybrid structure into Japanese corporate
governance systems and made these systems more diverse, the subsequent
corporate governance reforms led to a further transition of the corporate
governance system into a hybrid structure and promoted fine-tuning of
hybrid governance systems. The reforms advanced the structural transi-
tion of governance systems at traditional Japanese firms that lagged
behind in implementing reforms by promoting an increase in the institu-
tional investors, reorganization of the board of directors, and elimination
of relational shareholding. The reforms also induced firms whose govern-
ance systems already had hybrid structures to fine-tune their systems,
through stock buybacks conducted in response to an increase in institu-
tional investors and through an increase in independent outside directors
and creation of optional committees.
Corporate Governance Reforms under Abenomics 389

Our analysis finds no evidence that the corporate governance reforms

led to improved corporate performance through institutional investors’
increased influence or appointment of independent directors. Our analysis
shows that the positive relationship between institutional investor owner-
ship and real investment rather weakened during the reform period and
finds no clear evidence of institutional investor ownership influencing
M&A and R&D expenditures. Those results imply that the corporate
governance reforms did not have an impact strong enough to change
corporate managers’ attitudes toward risk and investment.
On the other hand, we can clearly find the effects of corporate
governance reforms on profit distribution. After the reforms, the
responsiveness of stock buybacks to the institutional investor owner-
ship ratio increased by more than 60 percent and would further increase
with consideration given to the institutional investor ownership ratio
itself. The direct result of the reforms was increased distribution of
profits to shareholders. Increased profits gained during the Abenomics
period were also directed to debt reduction, and therefore the corporate
governance reforms did not sufficiently curb the deleveraging. The
increased shareholder returns and debt reduction did not deplete
firms’ profits, which resulted in accumulation of cash and equivalents.
In sum, the reforms under Abenomics were a powerful driving force
behind making governance systems more shareholder-centric but have
not succeeded in changing corporate managers’ attitudes toward risk in
order to increase investments. Therefore, if a trickle-down effect was
expected as a way to equally distribute the benefits of the reforms to the
public, this has not been sufficiently realized yet.
It is too early to say that Abenomics has failed, or that pressure from
myopic shareholders has intensified, based solely on the findings that
corporate behavior has not sufficiently changed and shareholder returns
have increased. It takes time for any reforms to change corporate
behavior. Also, there has been no evidence of stock buybacks being
used for increased compensation, as discussed by Almeida, Fos &
Kronlund (2016) in the context of the US market, and increased share-
holder returns have not substantially reduced real investment. It is true,
however, that the vision of Abenomics has not been realized as antici-
pated. In order for the Japanese economy to grow in the future, we need
to prudently examine what should be expected of corporate governance
reforms and how the reforms should be complemented by other policies.
390 Hideaki Miyajima and Takuji Saito

Notes
1. Researchers’ interest in the problem that pressure from the stock market
could shorten the time horizon among corporate managers peaked
around 1990 (Stein 1988; Shleifer & Vishny 1990; Porter 1992,
1994). Research on such myopic behavior subsequently became less
active but has regained attention in recent years (Almeida, Fos &
Kronlund 2016; Edman, Fang & Huang 2018). This resurgence
reflects shifting interest in this issue in the United States and the United
Kingdom.
2. For details, see Aoki, Jackson & Miyajima (2007) and Miyajima,
Ogawa & Saito (2018).
3. For the reform process, see Jackson & Miyajima (2007); for details on
the ground-breaking legal reforms around banking crisis, see Shishido
(2007).
4. The law set the target to reduce the equity holding to be the same level as
its equity capital, from 27 trillion yen in 2001 to 9 trillion yen in 2004.
5. For further detail, see Jackson & Miyajima (2007) and Aoki (2010).
6. In 2002, institutional reform was implemented to bolster the status of
the auditor.
7. Only forty-three firms chose the committee system in 2003 in TSE 1st
section, right after the Company Law was amended.
8. For further detail, see Buchanan, Chai & Deacon (2012); Hamao and
Matos (2018), and Becht et al. (2017).
9. For further detail, see Milhaupt (2017) and Tanaka (2017).
10. Based on Miyajima and Suzuki (2018), which counts the number of
firms for which the insurance company was listed in the top thirty
shareholders.
11. The average rate of voting against proposals is 5 percent for insurance
companies compared with 20 percent for trust banks and asset
management firms (based on materials from the Financial Services
Agency).
12. “Yamaha up 7%: Positive Response to Buyback and End of Cross-
Shareholding Relationship,” Nihon Keizai Shimbun, November 29,
2017.
13. The amendment required all listed firms to disclose the following: the
name of firms; the number of shares held; and the book and current
values of those shares.
14. The estimation method used here does not completely eliminate the effects
of investors’ preference for high-performing firms, firms that actively make
high-risk investments, and firms with high shareholder returns.
Corporate Governance Reforms under Abenomics 391

References
Aggarwal, R., I. Erel, M. Ferreira, and P. Matos. 2011. “Does governance
travel around the world? Evidence from institutional investors.” Journal
of Financial Economics 100: 154–181.
Almeida, H., V. Fos, and M. Kronlund. 2016. “The real effects of share
repurchases.” Journal of Financial Economics 119: 168–185.
Aoki, M. 2010. Corporations in Evolving Diversity: Cognition, Governance,
and Institutions. Oxford: Oxford University Press.
Aoki, M., G., Jackson, and H. Miyajima. 2007. Corporate Governance in
Japan: Institutional Change and Organizational Diversity. Oxford:
Oxford University Press.
Crane, A., S. Michenaud, and J. Weston. 2016. “The effect of institutional
ownership on payout policy: Evidence from index thresholds.” Review
of Financial Studies 29: 1377–1408.
Becht, M., J. Franks, J. Grant, and H. Wagner, 2017. “The returns to hedge
fund activism: An international study.” Review of Financial Studies 30:
2933–2971.
Buchanan, J., D. Chai, and S. Deakin. 2012. Hedge Fund Activism in Japan:
The Limits of Shareholder Primacy. Cambridge: Cambridge University
Press.
Duchin, R., J. Matsusaka, and O. Ozbas. 2010. “When are outside directors
effective?” Journal of Financial Economics 96: 195–214.
Edmans, A., V. Fang, and A. H. Huang. 2018. “The long-term consequences
of short-term incentives.” ECGI Finance Working Paper, 527.
Fauver, L., M. Hung, X. Li, and A. Taboada. 2017. “Board reforms and firm
value: Worldwide evidence.” Journal of Financial Economics 125:
120–142.
Ferreira, M., and P. Matos. 2008. “The colors of investors’ money: The role
of institutional investors around the world.” Journal of Financial
Economics 88: 499–533.
Franks, J., C. Mayer, and H. Miyajima. 2014. “The ownership of Japanese
corporations in the 20th century.” Review of Financial Studies 27:
2580–2625.
Gombers, P., and A., Metrick. 2001. “Institutional investors and equity
prices.” Quarterly Journal of Economics 116: 229–259.
Gordon, J. 2007. “The rise of independent directors in the United States,
1950-2005: Of shareholder value and stock market prices.” Stanford
Law Review 59: 1465–1568.
Grinstein, Y., and R. Michaely. 2005. Institutional holdings and payout
policy. Journal of Finance 60: 1389–1426.
392 Hideaki Miyajima and Takuji Saito

Hamao, Y., and P. Matos. 2018. “US-Style Investor Activism in Japan: The
First Ten Years.” Journal of the Japanese and International Economies
48: 29–54.
Hoshi, T., and A. Kashyap. 2004. “Japan’s Economic and Financial Crisis:
An Overview.” Journal of Economic Perspectives 18: 3–26.
Ikeda, N., K. Inoue, and S. Watanabe. 2018. “Enjoying the quiet life:
Corporate decision-making by entrenched managers.” Journal of the
Japanese and International Economies 47: 55–69.
Jacoby, S. 2010. “Foreign Investors and Corporate Governance in Japan:
The Case of CalPERS.” In Corporate Governance and Managerial
Reforms in Japan, edited by H. Whittaker and S. Deakin, pp.93–133.
Oxford: Oxford University Press.
Jackson, G., and H. Miyajima. 2007. “Introduction: The diversity and
change of corporate governance in Japan.” In Corporate Governance
in Japan: Institutional Change and Organizational Diversity, edited by
M. Aoki, G. Jackson, and H. Miyajima, pp.1–47. Oxford: Oxford
University Press.
Jidinger J., and H. Miyajima. 2019. “Does regulation matter? Effects of the
corporate governance reforms on the relational shareholdings in
Japan.” RIETI Discussion Paper Series 20-E-003.
Kay, J. 2012. “The Kay Review of UK Equity Market and Long-term
Decision Making.” Investor Management Association.
Milhaupt C. 2017. “Evaluating Abe’s Third Arrow: How Significant are
Japan’s Recent Corporate Governance Reforms?” Columbia Law and
Economics Working Paper, no. 561.
Miyajima, H., and F. Kuroki. 2007. “The unwinding of cross-shareholding
in Japan: Causes, effects, and implications.” In Corporate Governance
in Japan: Institutional Change and Organizational Diversity, edited by
M. Aoki, G. Jackson, and H. Miyajima, pp.79–124. Oxford: Oxford
University Press.
Miyajima, H., R. Ogawa, and T. Saito. 2018. “Changes in corporate
governance and top executive turnover: The evidence from Japan.”
Journal of the Japanese and International Economies 47: 17–31.
Miyajima, H., and K. Suzuki. 2018. “Investment Behaviors of Japanese
Insurance Firms from 1990–2015.” Mimeo.
OECD. 2012. “Corporate Governance, Value Creation and Growth: The
Bridge Between Finance and Enterprise.” OECD Publishing.
Porter, M. E. 1992. “Capital disadvantage: America’s failing capital
investment system.” Harvard Business Review 70(5): 65–82.
Porter, M. E. 1994. Capital Choices. Cambridge, MA: Harvard University
Press.
Corporate Governance Reforms under Abenomics 393

Shishido, Z. 2007. “The Turnaround of 1997: Changes in Japanese

Corporate Law and Governance.” In Corporate Governance in Japan:
Institutional Change and Organizational Diversity, edited by M. Aoki,
G. Jackson, and H. Miyajima, pp.310–329. Oxford: Oxford University
Press.
Shleifer, A., and R. Vishny. 1990. “Equilibrium short horizons of investors
and firms.” American Economic Review 80: 148–153.
Stein, J. 1988. “Takeover threats and managerial myopia.” Journal of
Political Economy 96: 61–80.
Tanaka, W. 2017. “Current state and outlook of corporate governance
reform – With a focus on the board of directors’ system.” Working
paper at Waseda University: WCG WP #2017–002.
13 Abenomics and Japan’s
Entrepreneurship and
Innovation
Is the Third Arrow Pointed in the Right
Direction for Global Competition in the
Silicon Valley Era?
k e n j i e. ku s h i d a

13.1 Introduction
As Japan’s political economy evolves over time, a critical question is
whether it can adjust effectively to the rapidly transforming dynamics
of global competition. This chapter asks whether the Abe administra-
tion’s “third arrow” of structural reforms is useful in fostering innov-
ation and strengthening the global competitiveness of Japan’s
economy. To answer this, we must understand the dynamics of global
competition and how the locus of value-added activity has transformed
over time.
Over the past thirty years or so, the rise of Silicon Valley and venture
capital-backed firms has been dramatic as they created new industries,
transformed business models, and became the world’s most market-
valued and cash-rich companies. These firms – including Apple,
Google, Amazon, Salesforce, Oracle, Facebook, and others – origin-
ated from a coherent set of institutions: the “Silicon Valley model”
(Aoki 2001), which grew out of a specific political economic context,
developing partly as a competitive response to Japanese industrial
prowess of the 1970s and 1980s (Zysman & Newman 2006).
The Silicon Valley model’s core strength is to create and/or use new
promising technologies, fostering large numbers of companies that
scale up rapidly. To grow at breakneck speed, the successful firms
often spearhead new principles of competition and create new indus-
tries, often by disrupting and replacing incumbent firms (Kenney
2017). In contrast, a key weakness of Japanese firms over the past

394
Abenomics and Japan’s Entrepreneurship and Innovation 395

two decades has been their slow adjustment to rapidly changing

dynamics of value creation – an inability to generate or ride waves of
new technological paradigms and dynamics of competition that dis-
rupted entire industries (Kushida 2015).
Japan’s challenge is therefore how to introduce greater flexibility in
the constellation of institutions comprising the traditional Japanese
“model” (Aoki 2001) in order to better allow firms to capture value
in new ways. Japan’s interlocking and complementary institutions have
long hindered Japanese corporations’ ability to adjust rapidly to new
conditions (Aoki, Jackson & Miyajima 2007). A robust startup eco-
system can increase the flexibility and speed of adjustment, even if the
Japanese economy remains led by large firms. For example, Japan’s
Information and Communications Technology (ICT) sector has exhib-
ited a pattern of adjustment entailing new entrants forcing incumbents
to adjust dramatically, with this reorienting by incumbents transform-
ing the industry – despite incumbents not being displaced (Kushida
2012).
It is naive to assume that simply attempting to duplicate the institu-
tions of Silicon Valley will fuel Japan’s economic success in and of itself.
The world is littered with failed attempts to create local versions of
Silicon Valley (Lerner 2009). Piecemeal adoption of institutions is
unlikely to succeed anyway, since most existing institutions are com-
plementary: they depend on each other. Silicon Valley evolved through
historically specific conditions that were very different from those in
Japan.
Thus, the goal for Japan should not be piecemeal adoption but
instead strengthening the institutional underpinnings that enable
Japan’s startup ecosystem to flourish and coexist with the large-firm-
dominated model as part of Japan’s political economic evolution. The
aim is to foster flexibility rather than replace the core.
This chapter takes the Silicon Valley model and evaluates how much
the “third arrow” of Abenomics adds flexibility to some of the core
institutions of Japan’s traditional model by pushing toward or adding
institutional elements of the Silicon Valley model.
It concludes the following: Abenomics’ third arrow legitimizes the
institutions supporting Japan’s startup ecosystem by amplifying trajec-
tories of change already underway – not by forcing change in areas that
were headed in a different direction. By introducing concrete numerical
Key Performance Indicators (KPIs), the third arrow reforms enable the
396 Kenji E. Kushida

Abe administration to take credit for KPIs achieved. Many seem

achievable without policy action, with industries and economic funda-
mentals already heading in that direction. Some KPIs disappear, mostly
those achieved or not on track, and new ones appear every year. Since
institutions supporting the startup ecosystem are emerging in parallel
to the existing institutional configuration, the legitimacy provided by
Abenomics third arrow reforms do represent positive developments
and policy for helping innovation in Japan’s economy.

13.2 The Third Arrow of Abenomics

Abenomics third arrow reforms were articulated in a June 2013 docu-
ment, the “Japan Revitalization Strategy,” including the alluring and
hitherto almost unthinkably optimistic subtitle “Japan Is BACK.” It
began by recognizing Japan’s two decades of slow growth, outlining
a roadmap to regain GDP growth and increase productivity.
The third arrow reforms are nearly unprecedented in their breadth,
scope, and the inclusion of specific KPIs. They are updated annually,
and over the years trendy conceptions articulated by the World
Economic Forum at Davos, such as “Society 5.0” and “fourth indus-
trial revolution,” have been added. The strategy name also evolved
from “Japan Revitalization Strategy” to “Future Investment Strategy”
in 2017, becoming simply “Growth Strategy” in 2019. The number of
KPIs almost tripled from the initial 2013 strategy, with many dis-
appearing along the way – and not only those that were achieved (see
Table 13.1).

13.3 The Silicon Valley Model and Japan’s Transformation

This section first provides a brief overview of each Silicon Valley
institution supporting a startup ecosystem, noting how Japan has
been evolving over the past two decades in this area, then introducing
and evaluating relevant Abenomics KPIs.
Key characteristics underpinning Silicon Valley as an economic eco-
system, sorted into underlying institutions, include: (1) finance; (2)
human capital; (3) industry–university–government interactions; (4)
industrial organization; (5) entrepreneurship culture; and (6) business
infrastructure (Dasher et al. 2015).
Abenomics and Japan’s Entrepreneurship and Innovation 397

Table 13.1 Abenomics third arrow strategies over time

# KPIs (#
Year Title new)

2013 Japan Revitalization Strategy: Japan Is BACK 24

2014 Japan Revitalization Strategy 2014 Revision: Challenge 48 (25)
Towards the Future
2015 Japan Revitalization Strategy 2015 Revision: Investment 51 (9)
into the Future, Productivity Revolution
2016 Japan Revitalization Strategy 2016: Towards the Fourth 60 (14)
Industrial Revolution
2017 Future Investment Strategy 2017: Reforms Towards 60 (13)
Society 5.0
2018 Future Investment Strategy 2018: Transformation 70 (19)
Towards “Society 5.0,” “Data-Driven Society”
2019 Growth Strategy (2019) 63 (5)

Source: Prime Minister’s Office

Japan’s political economy has been evolving significantly to embrace

Silicon Valley–model institutional characteristics to support a startup
ecosystem. Observers of Japan’s nascent startup ecosystem in the late
1990s noted the numerous hindrances, but by the mid-2010s the
changes were significant (Kushida 2017) (see Table 13.2).

13.3.1 Finance
13.3.1.1 Silicon Valley: Venture Capital
Venture capital (VC) is central to the Silicon Valley model. As
financial intermediaries that invest in private companies, VCs take
equity stakes and commonly assume active roles in monitoring and
helping portfolio companies. The goals of VCs are to maximize
financial return by exiting investments, which are accomplished
through either Initial Public Offerings (IPOs) or M&A (Metrick &
Yasuda 2010).
Venture capital in Silicon Valley evolved as a distinct, prominent
investment mechanism in the 1970s and grew rapidly after American
regulatory shifts, in particular the relaxation of US Labor Department
398 Kenji E. Kushida

Table 13.2 Silicon Valley ecosystem characteristics compared to Japan’s

impediments in the 1990s, changes by 2016

Silicon Valley Japan in 2018: changes

startup ecosystem Japan in the mid-1990s: that facilitate startup
characteristic impediments ecosystem

Financial system: Bank-centered, traditional New small cap financial

Venture capital financial markets markets, growing VC
industry, rise of
independent VCs
Labor market: fluid, Long-term employment Increasing labor mobility,
diverse, highly with seniority ties especially in IT sector
skilled creating illiquid labor and with foreign firms.
markets. Best and Lower prestige and
brightest locked into opportunity with large
large firms for entire firms
career
Industry–university– Numerous formal Active efforts by
government ties regulatory constraints universities, private
on universities, lack of venture capital, and
brain circulation government to spin out
successful startups with
university technology
“Open” innovation Closed innovation with Firms more interested in
with large firms large firms in-house open innovation,
and small firm R&D and uninterested participation in VC
symbiosis in business with funds, business with
startups startups
Social system Entrepreneurship seen as Rising attractiveness of
encouraging low prestige vis-à-vis entrepreneurship as
entrepreneurship large firms and large firms enter
government, financially competitive crises,
risky increases cases of
successful startups
Professional services Small size of professional Law firms and accounting
ecosystem ecosystem firms setting up startup-
focused practice areas
to foster and benefit
from growing startup
ecosystem

Source: Kushida (2017)

Abenomics and Japan’s Entrepreneurship and Innovation 399

restrictions, that allowed pension funds to invest into VC funds.

Extremely high returns by VCs after the 1980s from major IPOs such
as Intel, Apple, and others provided venture capital with social legitim-
acy as an industry, establishing it as the driver of American industrial
innovation (Kenney 2011).

13.3.1.2 Japan: From Bank-Centered to the Rise of Independent VCs

Japan’s postwar financial model centered around banks. Banks were
the primary financial intermediaries, and the “main bank” system was
a mechanism of corporate governance entailing banks monitoring firms
and stepping to take over management for poorly performing firms
(Aoki & Patrick 1994). Japan’s venture capital industry, originating in
the 1970s, was peripheral to Japan’s overall financial system until the
late 1990s. Japan’s VC industry was long dominated by financial
institutions and corporations, and loans were the primary mechanism
of investment.
Japan’s VC industry transformed in the 2000s. Investments levels
grew, but – more importantly – loans disappeared, replaced by equity
options. Independent and professional VCs appeared and took center
stage, displacing traditional financial institutions and industrial cor-
porations whose employees were long criticized for being large-firm
employees with little experience in investing or growing startups.
A new generation of independent VCs emerged around 2008, and by
2014 and 2015 independent funds were the largest VC investors
(Center 2018). Prominent examples of independent VCs include
World Innovation Lab, Globis Capital Partners, and B Dash
Ventures, with fund sizes ranging from around $400 million to
$10 million – small by Silicon Valley standards but historically large
for Japan. While Japan’s VC industry remains tiny compared to that of
Silicon Valley – less than one twenty-eighth in 2015 – so does any other
country’s venture capital industry in comparison.
Maturation of Japan’s VC industry accelerated in the late 1990s when
small capitalization markets were introduced – a complementary institu-
tion to VC by enabling young, fast-growing companies to IPO quickly.
Until two competing small cap markets were created in 1999, the average
length for Japanese companies to IPO was just over twenty years. The long
time horizon and difficulty for VCs to exit their investment through IPOs
partly explains the prominence of loans by Japanese VCs until around
2000 (Milhaupt 1996; Kenney, Han & Tanaka 2004).
400 Kenji E. Kushida

The small cap markets were so successful at producing fast IPOs that
recent observers are concerned about too many small-scale IPOs hin-
dering the development of “unicorns” – private companies valued over
$1 billion, often used by the popular press to evaluate startup
ecosystems.

13.3.1.3 Abenomics: Venture Investment Targets and Unicorns

The Abenomics third arrow includes specific KPIs for strengthening the
startup ecosystem. A 2016 KPI aimed to double investments into
startups as proportion of nominal GDP by 2022. This KPI was reached
sooner, with the period of 2016–2018 reaching 0.058 percent, up from
0.025 percent in 2014–2016. It is highly plausible that this target
would have been achieved naturally even if nothing were done –
a pattern identified in other Abenomics reforms such as inward
Foreign Direct Investment (FDI).
While this KPI does not address the quality of VC, it does provide
explicit normative support for VC and the startup ecosystem.
Legitimacy can be significant in an industry subsegment that was long
a small and far less prestigious area of finance.
Another KPI, added in 2018, aims for twenty Japanese “unicorns”
by 2023. In 2017, Japan had two “unicorns”: Preferred Networks,
which provided artificial intelligence for factory robotics autonoma-
tion; and Mercari, an online consumer-to-consumer flea market ser-
vice, joined by Smartnews, a news curation service, in 2019. Mercari’s
2018 IPO was the largest in Japan’s startup history, raising around
$550 million, decreasing the number of unicorns to two (compared to
117 in the US, 73 in China, and 18 in the United Kingdom as of mid-
2018, according to CB Insights).
While the target of increasing unicorns seems justifiable, the prolif-
eration of unicorns may not signify success of the ecosystem (Kenney &
Zysman 2019). Moreover, given the lower cost and ease of listing small
firms on the Tokyo Stock Exchange Mothers market compared to the
US NASDAQ, it might even be a desirable strategy for VCs to list larger
numbers of smaller startups on the Mothers market (Kushida 2018).
The successful large IPO of Mercari, however, suggests, that larger
IPOs are indeed possible in Japan’s institutional environment; multiple
viable strategies exist.
Given that large IPOs and windfall returns historically legitimized
the US VC industry (Kenney 2011), this Abenomics KPI of
Abenomics and Japan’s Entrepreneurship and Innovation 401

Table 13.3 Selected venture capital–related KPIs

Year On
[Goal date] KPI added KPI value track?

[2022] Double VC investments 2016 2016–2018: Yes

as % of nominal GDP 0.058%
2014–2016:
0.025%
2012–2014:
0.028%
[2023] Aim for twenty Japanese 2018 2019: 2 -
unicorn companies (private or public)

increasing the number of unicorns may be useful to the ecosystem in

providing a focal point for Japanese VCs to aim for large IPOs (see
Table 13.3).

13.3.2 Human Capital

13.3.2.1 Silicon Valley: Global Top Talent, Fluid Employment
Silicon Valley draws global top talent, with high employment fluidity.
Top firms attract employees who settle locally and move among vari-
ous companies, bringing valuable technological expertise and know-
how with them (Saxenian 1994). Top universities such as Stanford,
University of California Berkeley, and UC San Francisco Medical
School are fed by global talent pools of students and professionals.
The Silicon Valley labor force includes people who have deep
expertise in every stage of a startup, from initial startup to rapid
growth to increasing maturity. The region also creates and attracts
people with cutting-edge skills in high-demand areas for numerous
historical waves of technologies, from semiconductors to various
types of software.1
The positive role of immigrants, particularly those with high-end
skills, has been a dramatic feature of Silicon Valley (Saxenian 2006). To
take a recent snapshot, the percentage of foreign-born population in
Silicon Valley was 36.4 percent in 2012, exceeding that of California
overall (27 percent), and is almost three times that of the US average
402 Kenji E. Kushida

(13 percent). Silicon Valley has continually benefited from flows of

immigrants from various areas of the world that create bridges with
the economies of their home countries. Cross-national production
networks with places like Taiwan and software outsourcing to India
also evolved through human interpersonal networks. A recent study
shows that just over 50 percent of “unicorn startups” in the US were
founded by immigrants (Anderson 2016).

13.3.2.2 Japan: Rigid Employment and Closed, Evolving Recently

Japan’s postwar employment model centered around long-term
employment arrangements by large firms for highly educated white
collar workers and was largely closed to foreign talent. Low employ-
ment fluidity constrained the development of Japan’s startup ecosystem
for much of the postwar period, choking it of potential entrepreneurs
and mid-career hires to provide startups with high quality employees.
From the late 1990s, regulatory shifts and large firms engaging in
downsizing and greater reliance on temporary and contract workers
began altering the system of long-term employment; it shrank to cover
a smaller core of workers despite remaining central to large firms
(Vogel 2006). At the same time, regulatory shifts such as allowing
stock options directly supported alternative employment schemes
benefiting the startup ecosystem. The influx of foreign firms into
Japan also increased employment opportunities for elite college gradu-
ates, creating a pathway for many of them to pursue entrepreneurship
after some years of working at foreign firms (Kushida 2017). By 2018,
large firms such as NEC, Sony, Hitachi, and Toshiba had shed tens of
thousands of workers, and other major firms such as Sharp and Nissan
were no longer under Japanese management. Top Japanese talent was
no longer “locked up” in large firms, and by 2018 media articles were
noting that wages for mid-career hires at startups often surpass those at
large firms (Nikkei 2019) – a change from the past when cash-strapped
startups could offer neither stock options nor enticing wages.
The Japanese startup ecosystem recently evolved to the point that
founders of the top funded startups are overwhelmingly graduates of
Japan’s top universities (Kushida 2018). Moreover, most of the previ-
ous employers of the entrepreneurs were foreign firms, notably con-
sulting and financial firms (Matsui, Kotosaka & Fukuda 2017).
Notably, however, they are almost all domestic universities and the
founders are Japanese (if not citizenship, then by name).
Abenomics and Japan’s Entrepreneurship and Innovation 403

Although Japan’s core lifetime employment system remains intact,

and coexists with a parallel set of arrangements for the startup ecosys-
tem, Japan’s white collar employment systems remain overwhelmingly
optimized for Japanese citizens. While severe labor shortages for
unskilled labor in factory work and construction historically led to
surges of foreign workers in the late 1980s, they were tightly limited
to particular sectors. Recently, with labor shortages in areas such as
healthcare, eldercare, and various service industries looming, regula-
tions are shifting to accommodate temporary, relatively unskilled or
narrowly defined skilled professions.

13.3.2.3 Abenomics: Increasing Labor Mobility, Addressing Skill

Gaps, and Limiting Foreign Labor
Abenomics’ third arrow reforms include specific KPIs to promote labor
mobility, although, given the low unemployment rate, they seemed
achievable regardless. The 2013 strategy aimed to double “job changes
without unemployment” from 1 million by 2017. By 2016, the number
was 3.3 million, achieving the KPI. Another KPI aimed to reduce long-
term unemployment (greater than six months) by 20 percent by 2018 –
also achieved. Another KPI aimed to increase the employment rate of
young people ages twenty to thirty-four from 74 percent in 2012 to
78 percent by 2020: a rate of 78 percent was achieved by 2016.
Another measure aimed at employment flexibility was to increase the
number of companies offering remote work (telecommute) options by
30 percent by 2020, from a base of 11.5 percent in 2012 and 13.9 per-
cent in 2017.
Specific KPIs aim to upgrade the skills base of existing employees and
citizens to the needs of IT-based skillsets. Included in a section titled
“strengthening human capital development and utilization needed in
the AI era,” these KPIs focus on certification and skills training in
various areas related to computer and software skills – for example,
aiming to double the number of certified professional practical skills
training programs in AI-related fields and enrolling 500,000 candidates
for a new “IT passport” exam. Also included were the goals of creating
100 certified “Fourth Industrial Revolution Skill Acquisition” profes-
sional development courses and aiming for 1 million people enrolled in
adult or professional school programs at universities and vocational
skill schools. Upgrading skills is important for any economy, but these
404 Kenji E. Kushida

KPIs were not targeting high-value skills at the frontier of value cre-
ation to compete with Silicon Valley leaders.
Women have been systematically underrepresented in long-term
employment arrangements, given the tradeoff for flexibly deploying
the workforce in exchange for a commitment to long-term employ-
ment. Specific KPIs aim for greater executive and higher-level oppor-
tunities for women. An initial 2013 KPI of increasing the
employment rate for women aged twenty-five to forty-four to 73 per-
cent was achieved. However, the goal of increasing the proportion of
women in leadership positions to 30 percent by 2020 was not on
track to being achieved (with 6.9 percent in 2012 and 9.8 percent in
2015), and this KPI disappeared in 2018. Instead, new KPIs in 2018
included increasing the proportion of women in executive positions
at publicly traded companies to 10 percent and managerial positions
in private sector companies to 15 percent by 2020. These were at
3.7 percent and 10.9 percent, respectively, in 2018 and were unlikely
to be achieved.
Abenomics has embraced vastly expanded temporary worker per-
mits to address acute labor shortages in numerous service-oriented
jobs, both skilled and relatively unskilled. Notably, there is no path
toward citizenship, and the numbers are far less than projected labor
shortfalls.
A 2013 KPI stipulated 5,000 highly skilled (certified) foreign profes-
sionals by 2017, targeting 10,000 by 2020. In 2016, the KPI was
updated to aim for 20,000 by 2022. Between May 2012 and
December 2017, 10,572 people were certified, so this KPI was on
track. A foreign worker bill enacted in April 2019 was designed to
attract 345,000 foreign workers over five years: mostly semiskilled
blue-collar labor in fourteen sectors facing severe labor shortages
including construction, the hotel industry, cleaning, and eldercare.2
The severe labor shortage continues, however, since the Ministry of
Health, Labor, and Welfare in 2015 projected a shortage of 377,000
people in caregiving alone by 2025.3
There were some efforts to internationalize Japan’s education sys-
tem. One goal was to have 1,500 full-time university faculty positions
for young and foreign applicants by 2016. This was achieved by mid-
2016 as measured by the number of new, fully funded faculty positions,
although most were filled by young Japanese rather than foreign fac-
ulty. Another KPI was to double the number of foreign students in
Abenomics and Japan’s Entrepreneurship and Innovation 405

Japan from around 140,000 in 2013 to 300,000 in 2020, which was on

track to being achieved if including Japanese language training insti-
tutes – some of which became notorious for serving as a back door to
foreign labor. These KPIs are not goals for Japan’s top universities that
could be responsible for churning out the highest level of human capital
(see Table 13.4).

13.3.3 Industry–University–Government Interactions

13.3.3.1 Silicon Valley: Multifaceted and Multidirectional Ties
In the Silicon Valley model, industry–university–government relation-
ships are multifaceted and multidirectional. University–industry ties
are anchored in intellectual property development and transfers,
often occurring through Technology Licensing Offices (TLOs). The
Bayh–Dole Act of 1980 allowed universities to own intellectual prop-
erty from government-funded research, incentivizing universities to
commercialize research.
However, university–industry ties reach far beyond TLOs to include
licensing, academic spin-offs, collaborative research, contract research,
consulting, ad hoc advice and networking for practitioners, teaching,
personnel exchanges, and student supervision (Grimaldi et al. 2011).
Industry personnel often spend time in universities, while university
faculty and researchers commonly take sabbaticals or spend other time
in company labs. University researchers often move into industry – and
sometimes back again.
The government, historically, was critical in establishing Silicon
Valley through defense contracting and developing semiconductors
for the Cold War effort (Leslie 2000). Massive research budgets for
universities led to “spinout” startups from university-led research,
which led the semiconductor industry and other related industries.
More recently, the government has set major research trajectories
through funding organizations such as the National Institute of
Health (NIH) and the National Science Foundation (NSF). The
Defense Advanced Research Projects Agency (DARPA), part of the
Department of Defense, also significantly shaped Silicon Valley by
setting deeper underlying technological trajectories – such as the
Internet in its original form and, more recently, self-driving vehicles –
through contests in which universities and industry research labs com-
pete (Markoff 2016).
406 Kenji E. Kushida

Table 13.4 Select KPIs for human capital

Year
[goal date] KPI added KPI value On track?

[FY 2020] 73% of 2013 2016: 73% Achieved

women age 25–44
employed (68% in FY
2010)
[2020] 80% of male 2015 2015: 55.9
workers take paternity
leave
[2020] 30% of women in 2013 No
leadership positions 2015: 9.8%
2012: 6.9%
[2020] 10% of women as 2018 No
corporate officers at 2018: 4.1%
listed companies 2017: 3.7%
[2020] 15% of women as 2018 No
section chiefs or 2018: 11.2%
equivalent in private 2017: 10.9%
sector companies
[2020] 30% of companies 2018 Achieved
offer remote work 2018: 19.1% (with
options 2017: 13.9% pandemic)
2012: 11.5%
[2020] 78% of young 2013 Achieved
people (ages 20–34) 2016: 78%
employed 2012: 74%
[2018] Reduce long-term 2013 Achieved
unemployed (>6 months) 2015: 1.09 million
by 20% 2013: 1.42 million
[2017] 2 million job 2014 Achieved
changes without 2016: 3.3 million4
unemployment 2013: 1 million
[2020] Recognize 10,000 2013 5/2012–12/2017: Yes
highly skilled foreign 10,572 highly skilled
professionals, 20,000 by foreign professionals
2022 recognized
Abenomics and Japan’s Entrepreneurship and Innovation 407

Table 13.4 (cont.)

Year
[goal date] KPI added KPI value On track?

[FY2020] Double the 2013 5/2017 (May):188,384 Yes

number of foreign in universities,
students studying in 267,042 in Japanese
Japan (140,000 to language education
300,000) institutions
[2023] Double certified 2018 2017: 7 certified **New KPI
professional practical programs
skills training programs
in AI-related fields
[2023] Enroll 500,000 2018 Exam starting FY 2019 -
candidates for new “IT
passport” exam
[2020] Certify 100 2018 -
“Fourth Industrial May 2019: 56
Revolution Skill 2017: 23
Acquisition”
professional
development courses
[2022] 1 million enrolled 2017 No
in adult school 2016: Approx 500,000
(professional school) 2015: Approx 490,000
programs at universities
or vocational skill
schools
= KPIs removed in 2018

13.3.3.2 Japan: Evolving University–Industry Collaborations

Until the late 1990s, Japan’s university system faced numerous
challenges in collaborating with industry. For example, national
universities professors were government employees, prohibited
from taking outside positions such as startup company cofounders
or board directors. More fundamentally, powerful academic insti-
tutions such as the Science Council of Japan, organized under the
408 Kenji E. Kushida

Prime Minister’s Office but independent, with 2,000 members span-

ning humanities, social sciences, life and natural sciences, and engin-
eering, were deeply divided over industry–university ties.
A significant group of senior, vocal academics from elite universities
steadfastly opposed closer ties with industry – a backlash against the
prewar and wartime era when universities and scientists were sub-
sumed by the totalitarian state and mobilized toward wartime
efforts.
Regulatory reforms in late 1990s, however, enabled far closer
industry–university ties. Technology Licensing Offices were
allowed and encouraged, aided by government funding.
A “Japanese Bayh–Dole Act” transferred intellectual property
from government-funded university research to the university
rather than government. National universities were reorganized
into “independent university corporations,” providing professors
and researchers far more flexibility to engage with industry
(Tabata 2005). Led by the University of Tokyo, which set up an
affiliated venture capital fund, created on-campus startup incuba-
tor facilities, and began offering a variety of courses promoting
industry–university collaborations, Japanese universities shifted
considerably toward partnering with corporations.

13.3.3.3 Abenomics: Industry–University Collaborations

The Abenomics third arrow had several KPIs to strengthen university–
industry ties: to increase industry investments into university research;
increase commercializable intellectual property from universities; and
enhance collaboration with Small and Medium Enterprises (SMEs).
However, interestingly, many were not on a trajectory toward achieve-
ment and were removed by 2018.
For example, the KPI to “increase the number of patents granted to
universities by 2020” appeared in 2016 but disappeared in 2018. The
number of patents filed by university TLOs actually decreased, from
4,831 in 2012 to 3,685 in 2016. (Importantly, since patents may not
have inherent value, numerical targets for patents may not be product-
ive to begin with.) Another KPI to double joint research projects
between corporations and universities or research institutes by
2020 – revised up from the 2015 KPI of a 20 percent increase by
2018 – was removed: it was not on target if projected linearly. A KPI
Abenomics and Japan’s Entrepreneurship and Innovation 409

to “create industry–academia–government collaboration to support

200 pioneering technology projects utilizing locally developed tech-
nologies by 2021, with approximately 1,000 projects over the sub-
sequent five years,” also disappeared. This KPI was in a section to
increase competitiveness of SMEs; a cynical view would classify it as
a new form of subsidizing less competitive SMEs as a social welfare
measure, especially for regional areas lacking large corporate
employers. A KPI that was not removed, originating in 2016, is to
triple corporate investments into universities and national research
institutes by 2025. With investment levels increasing from
120.9 billion yen in 2015 to 124.4 billion in 2016, a linear projec-
tion will not triple the amount – yet, the KPI is far enough in the
future that the Abe administration will not be responsible (see Table
13.5 for an overview of KPIs).

Table 13.5 Select KPIs for university–industry–government ties

Year
[FY goal date] KPI added KPI value On track?

[2020] 50% increase in patents 2016 2016: 3,685 No

granted to universities 2012: 4,831
[2025] Triple investments from 2016 2016: Y124.4 billion No
corporations into universities 2015: Y120.9 billion
and national research
institutes
[2020] Double large-scale joint 2015 2015: 1,004 Unverified
research projects between 2013: 390
corporations and universities/
research institutes
[2021] Create consortia of 2016 Unverified
industry–academia–
government–financial sector
collaboration to support 200
pioneering technological
development projects utilizing
locally developed
technologies annually, aiming
for 1,000 projects over 5 years
410 Kenji E. Kushida

13.3.3.4 Japan: Strategic Industrial Policy

Japan’s government–industry relations have long been characterized by
strategic industrial policy: infant industry promotion, technological
selection, and government orchestration of industrial consortia, as
well as promoting dense information exchanges enabling informal
“administrative guidance” to shape competition. Japan’s traditional
style of industrial policy, though its effectiveness is much debated,
differed considerably from US government policies that shaped
Silicon Valley. Japan’s policies were far more strategic and specific.
An upside was that specific technologies could be pursued, but
a downside was that technological choices could lead to domestic
isolation if global markets shifted in a different direction (Kushida
2011). Since the 1990s, many of Japan’s industry-level regulatory
structures moved away from “managed” competition in which the
government micromanaged competition by segmenting markets and
requiring approval for various products and services to ex post regula-
tion of markets (Vogel 2006; Kushida 2014).

13.3.3.5 Abenomics Strategic Policy

Abenomics third arrow KPIs include numerous specific technological
targets, ranging from developing new technologies surrounding electri-
city generation to deploying robotics and advanced sensors in health-
care, infrastructure maintenance, and agriculture. Some seem
reminiscent of previous eras of industrial policy when ministries were
at their strongest. On the one hand, Japan’s rapidly aging and shrinking
population can provide opportunities for technology deployment.
Goals include increasing large hospitals’ electronic medical record
utilization to 90 percent by 2020, growing the market for robotics in
nursing care to 50 billion yen by 2020, and deploying 8,000 robotic
devices for nursing care by 2030. These measures can potentially
transform Japan’s demographic challenges into internationally com-
petitive value-added products and services (see Table 13.6).
Abenomics KPIs selecting particular technologies do run the inherent
risk of selecting what turns out to be the wrong technology – examples
of which Japan’s industrial policy history is littered with. Specific
numerical targets for fuel cell vehicles, fuel cell charging stations, and
“next generation gas turbines” seem ambitious but run the risk of
following a familiar pattern of “leading without followers” experi-
enced in Japan’s ICT industries (Kushida 2015). Strategic policy can
Abenomics and Japan’s Entrepreneurship and Innovation 411

Table 13.6 Select KPIs for industry–university–government

collaboration

Year On
[Goal date] KPI added KPI value track?

[2017] Attain number one ranking in 2013 2017–2018: 8th No

the innovation ranking (technology 2013–2014: 5th
section, World Economic Forum)
[FY 2020] increase the ratio of R&D 2013 2016: 3.42% Yes
investment, including both public
and private sectors, to 4% of GDP

morph into protecting troubled industries, as has occurred historically

as well (Callon 1995). KPIs such as “commercializing the world’s first
floating wind turbine by 2018,” extending a demonstration project off
the coast of Fukushima, might be seen as such. The ten-member con-
sortium included Japan’s nuclear plant manufacturers Mitsubishi
Heavy Industries and Hitachi, whose nuclear facilities were halted
after the 2011 Fukushima nuclear disaster.
There may be more industrial opportunity in broader goals, such as
the promotion of robotics and sensors in maintaining and inspecting
infrastructure (4), and targets for upgrading commercial office build-
ings and housing to become more energy efficient (6).
Measures such as contests, such as the DoD contests for automated
driving, which led to an explosion of commercialization interest and
research, are not part of Abenomics.
KPI policy targets were malleable. Several KPIs that were not on
track were removed in 2018, such as halving the cost of storage batter-
ies by 2020, Japanese firms gaining approximately 20 percent of world
market share of advanced electricity storage batteries, and aiming for
a 100 percent diffusion of LEDs for lighting.
While Abenomics third arrow strategic policies do provide oppor-
tunities in particular promising areas that are broadly thought to lead
value-added IT competition, which can increase overall corporate
flexibility, many KPIs are extremely specific to the point that they are
reminiscent of traditional industrial policy.
412 Kenji E. Kushida

13.3.4 Industrial Organization

13.3.4.1 Silicon Valley: Large Firm and Startup Symbiosis
In Silicon Valley, the ecosystem supporting large companies coexists
symbiotically with the ecosystem supporting startups. Large firms
engage in “open” innovation by buying startups, funding them, and
drawing from and contributing to the same labor force and technology
environment (Kenney 2000). This industrial structure grew out of the
history of Silicon Valley, in which waves of technological advancement
provided opportunities for new firms to grow quickly, displacing
incumbent firms. As the Silicon Valley model developed, venture cap-
ital and the startups it financed moved to the center of American
innovation dynamics (Kenney 2011). Since VCs can only exit through
IPOs and M&A, with a far higher number acquired by M&A, this
direct relationship is also a key underlying feature of Silicon Valley’s
industrial organization. Tracking waves of M&A over time reveal the
areas where large firms are rushing to capture new value – for example,
recently, AI-related startups.

13.3.4.2 Japan: New Levels of Symbiosis

A historical challenge for Japan’s startup ecosystem has been the lack of
collaboration between large firms and startups. The dominant postwar
model was large, prestigious firms populated by elite university gradu-
ates, with less prestigious SMEs as suppliers in vertical “keiretsu”
arrangements. The latter were squeezed during economic downturns,
and the large firms took leadership in technological and strategic
directions. As a result, Japanese large firms were inexperienced in
dealing with startups that had high potential technological or strategic
value. Layers of management tended to treat startups as simply as
suppliers and did not considering various strategic collaboration possi-
bilities, including M&A – especially integration after purchase.
This situation has shifted markedly over the past decade, as waves of
new startups are collaborating with large firms in new ways. Even in the
financial industry, startups offering accounting services to SMEs, for
example, are partnering with large and local banks, some of the most
traditional and conservative firms (Kushida 2018).
Part of the new ease of collaboration may be that many of the new
startups are created by people with the same elite backgrounds as large
company leaders and employees, which is a contrast from the past.
Abenomics and Japan’s Entrepreneurship and Innovation 413

There are still few cases of startups purchased by large Japanese

companies. Soracom, founded in late 2014, offering services embed-
ding small, low-cost chips to connect various devices to the Internet,
used intellectual property from NTT DoCoMo’s research labs that
were deemed valueless. Led by an outside management team,
Soracom was purchased by telecom company KDDI in 2017 for
approximately $400 million.

13.3.4.3 Abenomics: New Opportunities through Industry-Specific

Policy
There are no concrete third arrow KPIs directly aimed at facilitating
collaborations between large firms and startups. However, many of
the IT adoption and technology-oriented goals do create opportun-
ities for startups to partner with large corporations. For example, for
infrastructure, the goal to “use sensors, robots, and other noninva-
sive technologies to examine and analyzing critical and aging domes-
tic infrastructure, with a 20 percent utilization by 2020, and
100 percent by 2030” opens opportunities for startups to provide
technologies and services that large firms may not have or be able to
develop quickly enough. Conversely, startups may be unable to scale
their offerings effectively on their own, leading to potential partner-
ship opportunities.
Another Abenomics KPI, added in 2017 and targeted directly at
Fintech, is to have eighty banks adopt open APIs (Application
Protocol Interfaces), which allows third parties to connect directly to
banks’ IT systems, by 2020. This directly benefits startups with IT
expertise, and newly listed startups Money Forward and Free – provid-
ing accounting services for SMEs and personal finance management
tools – were quickly tied into incumbent banks’ systems. Other areas,
such as the adoption of robots in nursing care, may be fruitful areas of
collaboration for startups and large firms. Another third arrow KPI
aims to double the amount of cashless electronic payments to 40 per-
cent by 2020 (24 percent in 2018). QR code–based payment transac-
tions was another opportunity-laden area, allowing startups such as
Origami, Line, and Coiney to offer services in partnership with incum-
bent banks, thereby supporting the startup ecosystem.
Government support of this entire industry segment increases legit-
imacy for startups, who must partner with incumbent, and often con-
servative, financial institutions.
414 Kenji E. Kushida

13.3.5 Entrepreneurship Culture and Business Infrastructure

13.3.5.1 Silicon Valley: Entrepreneurs Celebrated, Monitored,
Evaluated, and Supported
In Silicon Valley, entrepreneurs are celebrated, with deep shared know-
how for monitoring and evaluation, so that failures, although unfortu-
nate, are considered positive experiences if they were “good failures”
(Dasher et al. 2015). The business ecosystem supporting startups in
Silicon Valley comprises of law firms, accounting firms, headhunting
firms, and interpersonal networks of angel investors and other mentors
who underpin the ecosystem. Accountants and lawyers, for example,
often play the roles of dealmaker, advocate, advisor, and other func-
tions that go well beyond the professional services they offer (Kenney
2000).

13.3.5.2 Japan: Startup Ecosystem’s Increasing Legitimacy after 2010

Japan’s early postwar period saw a flourishing of entrepreneurs who
built large, globally competitive companies: Honda Soichiro of Honda,
Morita Akio and Ibuka Masaru of Sony, and Inamori Kazuo of
Kyocera, for example. However, by the 1980s, large firms dominated
Japanese industry, in technological and market leadership, as well as
prestige and social position. Elite university graduates flocked to long-
term employment positions in large firms while entrepreneurs were
seen as those lacked the option of stable employment at large firms.
The prestige and social acceptance of entrepreneurship rose signifi-
cantly since the 1990s. A wave of startups based on Internet-related
technologies such as Rakuten emerged in the late 1990s, entering
mainstream large corporate Japan by joining major business associ-
ations and purchasing professional sports teams to become household
names. A social setback to the startup ecosystem in the mid-2000s
occurred when a prominent IT entrepreneur and outspoken critic of
the establishment, Takafumi Horie, CEO of Livedoor, was arrested
and jailed for securities fraud. Numerous entrepreneurs report the
“Horie shock” cast a chill over the startup ecosystem, driving entrepre-
neurs away from the public spotlight and refraining from showing off
their wealth in public, making it far more difficult to celebrate daring
and openly ambitious entrepreneurs in the manner of Silicon Valley.
However, after 2010, a normative shift occurred in Japan, coinciding
with severe performance crises at many large Japanese companies
Abenomics and Japan’s Entrepreneurship and Innovation 415

following the global financial crisis of 2007–2008. Large firms were no

longer the obvious answer to Japan’s economic challenges, and massive
accounting scandals at venerable firms such as Toshiba and Olympus
and the near collapse of Sharp led to the startup ecosystem gaining
legitimacy simply through performance crises of incumbents. The
most-read economic newspaper, the Nihon Keizai Shimbun, began
carrying front-page news about the startup ecosystem by the late
2010s.
Startup pitch contests and major events celebrating high growth
startups multiplied, attracting thousands of audience members and
making news. For example, the annual New Economy Summit,
launched in 2013, organized by the Japan Association of New
Economy (JANE), spearheaded by Rakuten, brought prominent
Silicon Valley entrepreneurs such as Larry Ellison, founder of Oracle,
and the founders of startups such as Dropbox, Lyft, and Box.com to
Japan. JANE also engages in policy discussions and forms policy
recommendations in areas such as accepting digital signatures for
contracts and documents. The Infinity Ventures summit, which began
in 2009, brings hundreds of companies to Kyoto annually and is a hub
for investors, entrepreneurs, and large firms to meet. From the govern-
ment, the New Energy and Industrial Development Organization
(NEDO), which subsidized the R&D of science and technology–
based startups under a program called the NEDO Technology
Commercialization Program, held annual pitch contests from 2014.
Until recently, the support ecosystem for startups was weak, with
accountants and lawyers having little if any experience with high
growth startups and little ability to play the multiple roles of Silicon
Valley professional firms. Over the past decade, however, Japan’s
startup business infrastructure developed significantly, with law firms
and accounting firms, along with human resources development and
placement firms, increasingly focused on startups. Tohmatsu Venture
Support, for example, a subsidiary of accounting firm Deloitte, estab-
lished itself as a startup ecosystem player by providing startup-oriented
services and holding weekly morning pitches and annual major events
for startups.

13.3.5.3 Abenomics: KPIs Strengthening Norms and the Ecosystem

Third arrow KPIs include an explicit target for social norms: a goal to
double Japan’s position in an international survey of entrepreneurship
416 Kenji E. Kushida

mentality, measured as the number of people who answered that they

are “entrepreneurs or prospective entrepreneurs” by 2025. The figure
was 3.8 percent in 2014, rising to 4.7 percent by 2017.
Cultivating “a positive feedback loop to foster innovative startups”
is the title of an entire category of KPIs, which also included an ambi-
tious goal to become number one in the World Economic Forum’s
innovation ranking (technology section) by 2017. This KPI was not
achieved, as Japan’s ranking slid from fifth in 2013–2014 to eighth in
2017–2018. It also benchmarked business entry and exit rates to the
United Kingdom, aiming for comparable numbers.

13.4 Conclusion
The sheer volume and specificity of Abenomics third arrow reforms has
been impressive. Many of the key performance indicators are set beyond
2020, likely beyond the Abe administration’s tenure, therefore insulating
him from blame if they are not attained. Others are achievable without the
government doing anything extra and are likely an easy mechanism of
taking political credit for phenomena already underway. Yet, the contrast
with previous administrations – which all suffered short tenure and were
criticized for the lack of policymaking continuity, let alone the ability to
think longer-term or strategically – is stark.
The third arrow KPIs are by and large pointed in the right direction to
increase the flexibility of Japan’s economy. Japan’s startup ecosystem has
been developed significantly over the past twenty years, enjoying greater
legitimacy in Japan’s political economy and society. While the startup
ecosystem is unlikely to replace Japan’s large firms as the economic core, it
is quite likely that the ecosystem can drive adjustment, especially by
accelerating and shifting the course of large incumbent companies. To
this end, Abenomics KPIs aim to strengthen the startup ecosystem, and
there are few that one could consider misguided. In the past, strategies of
attempting to make localized versions of Silicon Valley appeared and
disappeared in a variety of localities, and it is good news that the
Abenomics third arrow does not contain these types of measures.
Arguably the most important contribution to Japan’s transforming
economic model in facilitating a Silicon Valley model of entrepreneurship
and innovation is the explicit normative support for many of the
underlying institutions of the Silicon Valley model in finance, human
capital, university–industry–government ties, and social support of
Abenomics and Japan’s Entrepreneurship and Innovation 417

entrepreneurship. They represent the recognition that various aspects of

the Japanese model need to change, providing legitimacy to an area that
long suffered a lack of it. This sustained push toward legitimacy is one that
observers would not have expected during the years when prime ministers
rotated on an almost annual basis, each putting forth various piecemeal
reforms that did not encompass an underlying vision or strategy. It is, on
balance, a step forward to have specific KPIs, even if many were on track
to being achieved to begin with. They could have pushed further in
numerous areas, and some parts are reminiscent of historical industrial
policy. Moreover, it is far from obvious that Japanese firms will compete
effectively in a world of new technological advances in areas such as
artificial intelligence and potential disruptions of the automotive industry.
Firms from the Silicon Valley model will likely continue to redefine and
challenge industrial structures, threatening incumbent firms worldwide in
various industries. Yet strategic policies that do not actively active hinder
corporate adjustment and a vibrant startup ecosystem are not to be taken
for granted.

Notes
1. For a recent empirical study of global talent pools of Hadoop, a big data
framework, revealing the prominence of Silicon Valley, see Tambe
(2014).
2. See https://asia.nikkei.com/Spotlight/Japan-Immigration/Japan-enacts-
divisive-foreign-worker-bill-to-ease-labor-shortage.
3. See www.mhlw.go.jp/stf/houdou/0000088998.html.
4. See www.mhlw.go.jp/file/05-Shingikai-12602000-Seisakutoukatsukan-
Sanjikanshitsu_Roudouseisakutantou/0000190839.pdf.

References
Anderson, S. 2016. “Immigrants and Billion Dollar Startups.”
Aoki, M. 2001. Toward a Comparative Institutional Analysis. Cambridge,
MA: MIT Press.
Aoki, M., G. Jackson, and H. Miyajima, eds. 2007. Corporate Governance
in Japan. Oxford: Oxford University Press.
Aoki, M., and H. T. Patrick. 1994. The Japanese Main Bank System: Its
Relevance for Developing and Transforming Economies. Oxford:
Oxford University Press.
418 Kenji E. Kushida

Callon, S. 1995. Divided Sun: MITI and the Breakdown of Japanese High-
Tech Industrial Policy, 1975–1993. Stanford, CA: Stanford University
Press.
Center, V. E. 2018. Benchaa Hakusho (Venture White Paper). Venture
Enterprise Center.
Dasher, R., N. Harada, T. Hoshi, K. E. Kushida, and T. Okazaki. 2015.
“Institutional Foundations for Innovation-Based Economic Growth.”
National Institute for Research Advancement.
Grimaldi, R., M. Kenney, D. S. Siegel, and M. Wright. 2011. “30 years after
Bayh–Dole: Reassessing academic entrepreneurship.” Research Policy
40(8): 1045–1057.
Kenney, M. 2000. Understanding Silicon Valley: The Anatomy of an
Entrepreneurial Region. Stanford, CA: Stanford University Press.
Kenney, M. 2011. “How venture capital became a component of the US
National System of Innovation.” Industrial and Corporate Change 20
(6): 1677–1723.
Kenney, M. 2017. “Explaining the Growth and Globalization of Silicon
Valley: The Past and Today.” BRIE Working Paper 2017–1.
Kenney, M., K. Han, and S. Tanaka. 2004. “The globalization of venture
capital: the cases of Taiwan and Japan.” In Financial Systems,
Corporate Investment in Innovation, and Venture Capital, edited by
A. Bartzokas and S. Mani, pp.52–84. Northampton, MA: Edward Elgar
Publishing.
Kenney, M., and J. Zysman. 2019. “Unicorns, Cheshire cats, and the new
dilemmas of entrepreneurial finance.” Venture Capital 21(1): 35–50.
Kushida, K. 2018. “Is the lack of “Unicorns” in Japan good news or bad
news? Injecting a historical institutional perspective into debates about
Japan’s startup ecosystem.” NIRA Opinion Paper 39.
Kushida, K. E. 2011. “Leading Without Followers: How Politics and Market
Dynamics Trapped Innovations in Japan’s Domestic “Galapagos”
Telecommunications Sector.” Journal of Industry, Competition and
Trade 11(3): 279–307.
Kushida, K. E. 2012. “Entrepreneurship in Japan’s ICT Sector:
Opportunities and Protection from Japan’s Telecommunications
Regulatory Regime Shift.” Social Science Japan Journal 15(1): 3–30.
Kushida, K. E. 2014. “Foreign Multinational Corporations and Systemic
Change in Japan.” In Syncretization: Corporate Restructuring and
Political Reform in Japan, edited by K. E. Kushida, K. Shimizu, and
J. Oi, pp.199–246. Stanford, CA: Shorenstein APARC.
Kushida, K. E. 2015. “The Politics of Commoditization in Global ICT
Industries: A Political Economy Explanation of the Rise of Apple,
Abenomics and Japan’s Entrepreneurship and Innovation 419

Google, and Industry Disruptors.” Journal of Industry, Competition

and Trade.
Kushida, K. E. 2015. “The politics of commoditization in global ICT
industries: A political economy explanation of the rise of Apple,
Google, and industry disruptors.” Journal of Industry, Competition
and Trade 15(1): 49–67.
Kushida, K. E. 2017. “Innovation and Entrepreneurship in Japan: Why
Japan (Still) Matters for Global Competition.” SVNJ Working Paper
Series (2017–2).
Kushida, K. E. 2018. “Departing from Silicon Valley: Japan’s New Startup
Ecosystem.” In Reinventing Japan: New Directions in Global
Leadership, edited by M. Fackler and Y. Funabashi, pp.81–102.
Westport, CT: Praeger.
Lerner, J. 2009. Boulevard of Broken Dreams: Why Public Efforts to Boost
Entrepreneurship and Venture Capital Have Failed – and What to Do
About It. Princeton, NJ: Princeton University Press.
Leslie, S. 2000. “The biggest ‘angel’ of them all: The military and the
making.” In Understanding Silicon Valley: The Anatomy of an
Entrepreneurial Region, edited by M. Kenney, pp.48–70. Stanford,
CA: Stanford University Press.
Markoff, J. 2016. Machines of Loving Grace: The Quest for Common
Ground between Humans and Robots. New York: HarperCollins.
Matsui, R., M. Kotosaka, and R. Fukuda. 2017. “Data about Mothers
CEOs’ Careers at time of listing.” https://iber.sfc.keio.ac.jp/?p=9564,
accessed October 1, 2019.
Metrick, A., and A. Yasuda. 2010. “Venture capital and the finance of
innovation.”
Milhaupt, C. J. 1996. “Market for Innovation in the United States and Japan:
Venture Capital and the Comparative Corporate Governance Debate.”
Northwestern University Law Review 91: 865.
Nikkei. 2019. “Sutatoapputenshoku, nenshuu 720 mangoe, joujoukigyougoe
(Startup job hires, annual salary exceeds 7.2 million yen, listed firm
salaries).” Nihon Keizai Shimbun.
Saxenian, A. 1994. Regional Advantage: Culture and Competition in Silicon
Valley and Route 128. Cambridge, MA: Harvard University Press.
Saxenian, A. 2006. The New Argonauts: Regional Advantage in a Global
Economy. Cambridge, MA: Harvard University Press.
Tabata, H. 2005. “The incorporation and economic structural reform of
Japan’s national universities.” Social Science Japan Journal 8(1):
91–102.
Tambe, P. 2014. “Big data investment, skills, and firm value.” Management
Science 60(6): 1452–1469.
420 Kenji E. Kushida

Vogel, S. K. 2006. Japan Remodeled: How Government and Industry Are

Reforming Japanese Capitalism. Ithaca, NY: Cornell University Press.
Zysman, J., and A. Newman, eds. 2006. How Revolutionary Was the
Digital Revolution? National Responses, Market Transitions, and
Global Technology in a Digital Era. Stanford, CA: Stanford Business
Press.
14 Japanese Agricultural Reform under
Abenomics
p a t r i c i a l . m a c l a c h l a n a n d ka y
s h i mi zu

14.1 Introduction
Japanese agriculture is in deep crisis. For more than a generation, the
sector has struggled with a raft of urgent problems, from decreasing
farm household incomes and a severe shortage of farm successors to the
spread of abandoned farmland. Until recently, however, the govern-
ment’s response was largely limited to small-scale reforms that tinkered
with these problems at the margins. To a significant degree, this policy
inertia reflects the influence of anti-reformist voices in the Ministry of
Agriculture, Forestry, and Fisheries (MAFF), the Liberal Democratic
Party (LDP), and Japan Agricultural Cooperatives (JA) on the agricul-
tural policymaking process. Indeed, it can be said that the survival of
this “conservative farm lobby” depends on the very political institu-
tions and economic policies that catapulted farming into such dire
economic straits in the first place.
But things changed under Prime Minister Shinzo Abe (2012–2020).
After assuming the prime ministership in 2012, Abe prioritized agri-
cultural reform as a signature objective of the structural reform, or
“third arrow,” component of Abenomics, his comprehensive plan for
the reinvigoration of the Japanese economy. As of this writing, it is
clear that the Abe government scored several modest but significant
policy victories that have rendered the farm sector more responsive to
market signals. These accomplishments were in turn enabled by the
government’s centralization of the agricultural policymaking process in
the Kantei, where reformist voices hold sway. These changes – substan-
tive and procedural – suggest that the Abe government was intent not
only on enhancing the competitiveness of Japanese agriculture but also
on diminishing the farm lobby’s already weakening influence on
Japanese policymaking.

421
422 Patricia L. Maclachlan and Kay Shimizu

These developments do not mean that the battle between the Abe
government and defenders of the agricultural status quo was a decisive
one. Many policy reforms chronicled in this chapter included signifi-
cant concessions to the farm lobby, which in turn experienced a partial
resurgence from 2017. Nevertheless, the accomplishments of agricul-
tural politics under Abenomics (Abenomikusu nōsei) will likely prevent
the farm lobby from ever recouping its former influence.
Nor did the Abe government operate in a vacuum. As we shall see,
long-term demographic and economic trends, along with the cumula-
tive effects of many small-scale reforms introduced by past govern-
ments, created a fertile context for proactive leadership on behalf of
market-oriented change – including reforms that were long considered
politically taboo.
To support these claims, we begin by summarizing what the Abe
government inherited in 2012. In Section 14.2, after outlining the
contours of Japan’s agricultural crisis, we explain the economic policies
and political-economic institutions that have functioned as “pillars” of
both the crisis and the conservative farm lobby’s political power: high
rice prices; protected rice markets; draconian regulatory controls over
farmland ownership and transfers; JA’s economic powers; and
a political exchange relationship between the organized farm vote
and the LDP. In Section 14.3, we assess the Abe government’s struc-
tural reform agenda for the farm sector and the impact of changes to
the agricultural policymaking process on policy outputs. We conclude
in Section 14.4 with a brief look at agricultural reform as a “legacy
issue” for Prime Minister Abe and speculate about the future of reform.

14.2 What Abe Inherited

14.2.1 Japan’s Agricultural Crisis

In 2012, Japanese farming was in the throes of a long-term, multi-
faceted crisis. For starters, the number of commercial farm households
was shrinking at a rapid rate, from 2.97 million in 1990 to 1.63 million
in 2010 (MAFF 2010). More problematically, farmers were graying at
a rate much faster than that of the general population; between 2005
and 2010, those age sixty-five or older increased from 58.2 percent of
the total to 61.6 percent (MAFF 2010). Even more troubling was
a growing shortage of farm successors; conservatively, as many as
Japanese Agricultural Reform under Abenomics 423

50 percent of active farmers lacked – and continue to lack – successors.

The successor shortage was in turn a driving force behind the rapid
spread of abandoned farmland, which totaled approximately 396,000
hectares by 2010 (MAFF 2010).
The successor shortage reflected a deepening economic crisis in rural
Japan. Thanks to decreasing rice prices, the high cost of farm inputs,
and a long-term decline in the (shrinking) general population’s demand
for rice and other foods, Japan’s past successes at achieving rough
income parity between rural and urban households had been largely
undone. And for as long as farm household incomes remained on
a downward slope, the sector struggled to attract young blood to the
soil. These demographic and economic challenges were reasons to fear
for the very future of Japanese agriculture.
Japan’s failure to end the crisis is rooted in political factors. As others
have explained (see, for example, Mulgan 2000 and 2005), postwar
agricultural policymaking was tightly controlled by deeply conserva-
tive forces within the MAFF, JA, and the LDP farm zoku (Nōrin zoku),
one of the party’s more powerful policy tribes (Saitō 2016, p.24).
Against this backdrop, agricultural policymaking followed
a predictable pattern. First, relevant subcommittees of the Policy
Affairs Research Council’s (PARC) Agriculture Committee (Nōrin
bukai), which served as the primary venue for zoku politicians to
channel JA’s wishes into the legislative process, would hammer out
the details of a legislative proposal. The proposal would then wind its
way up the party hierarchy, clearing first PARC and then the General
Council (Sōmuchō), which would adopt it as official party policy. Only
then would the MAFF complete the legislative drafting process, even-
tually forwarding a draft bill to the cabinet for approval and then
submission to the Diet. Although innovative governments or MAFF
officials would occasionally advocate on behalf of reformist measures,
pro–status quo farm zoku politicians would see to it that those meas-
ures were either defeated or watered down.
Buttressing the political balance of power within the agricultural
policy process was an exchange relationship between the LDP and JA
in which JA would deliver the farm vote to individual LDP politicians in
lower house districts under single nontransferable vote, multimember
district (SNTV-MMD) rules, and to a small number of “sponsored
candidates” (rieki daihyō giin, literally “interest representative Diet
members”) in the national PR tier in upper house elections. In return
424 Patricia L. Maclachlan and Kay Shimizu

for their electoral backing, which contributed to long-term LDP dom-

inance by compensating the party for its weak organizational roots at
the local level (Maclachlan 2014), JA and its farmer-members
extracted a slew of particularistic favors from a succession of LDP
governments.
Some of those particularistic favors comprised the pillars of the
conservative farm lobby’s longevity and regime-like qualities.
Foremost among them were high, government-guaranteed rice prices.
Each year, JA participated in negotiations with the MAFF and zoku
politicians to keep prices high under the purview of the 1942 Food
Staple Control Law (see Davis 2003), which was enacted to regulate the
rice supply under conditions of wartime scarcity but maintained for
several decades thereafter as legal cover for the perpetuation of the
LDP’s farm-friendly redistributive policies. Predictably, artificially high
prices produced a massive oversupply of rice in the Japanese market,
but, instead of lowering prices, the government implemented
a program in 1971 to curb output – the so-called rice paddy reduction
policy, or gentan. By the mid-1980s, as farm household incomes
reached their postwar peak, the government’s rice policies had pro-
duced gross economic inefficiencies: despite rapidly shrinking con-
sumer demand, the proportion of farmers engaged in rice production
was holding steady at roughly 70 percent, and most of those farmers
were tilling tiny plots of land only part time.
The high rice prices that proved so instrumental to the achievement
of income parity between rural and urban households by 1970 could
not have been sustained without closed rice markets, another import-
ant pillar of the old agricultural political order. Japan was no stranger
to food imports: it had been gradually opening up its markets from
1960 (Nōgyō to keizai henshū iinkai 2011, p.68). But rice imports
remained taboo, thanks largely to pressure from JA, which was first
and foremost an organization of rice producers.
The strict regulation of farmland ownership and transfers also
helped prop up the farm lobby. In 1952, Japan passed the
Agricultural Land Act, which entrenched the main provisions of post-
war land reform. Among other things, the act prohibited absentee
farmland ownership, placed draconian limits on farm size, drastically
curbed farmland leasing, and restricted farmland ownership to farm
households (nōka). An effective curb on the postwar reemergence of
the parasitic landlord class and a welcome boon to Japan’s once
Japanese Agricultural Reform under Abenomics 425

beleaguered family farms, the act spawned long-term economic ineffi-

ciencies by preventing both nonfarm household corporate participa-
tion in agricultural production and the attainment of larger economies
of scale through land consolidation. Most notably, the statute proved
a boon to small-scale, part-time (mostly rice) farmers, who by the early
1990s outnumbered their more efficient full-time counterparts by
a ratio of more than five to one.
Finally, Japan’s deepening farm crisis and the power of the conser-
vative farm lobby were supported by JA’s predominant position within
the postwar food supply chain, most notably for rice. For decades, JA
functioned as the MAFF’s designated collector and distributor of rice
under the 1942 Food Staple Control Law. Under that system, most
farm households automatically transferred their raw product to local
JA co-ops, rather than navigating markets on their own or through
alternative intermediaries. Meanwhile, the Agricultural Land Act and
other regulatory controls effectively barred food processors, retailers,
and other nonfarm household enterprises from participating in the
primary sector. This near absence of competition in turn positioned
JA to become the main provider of fertilizers, farm machinery, and
other farm inputs, as well as farm credit and insurance. All the while,
JA’s freedom to jack up its fees and commissions went virtually
unchecked, which meant higher production costs for farmers.
High rice prices; closed markets for rice; the regulatory regime for
farmland; JA’s inordinate power within the primary sector; and a long-
term electoral exchange relationship between JA and the LDP – these
five pillars enabled a virtuous (or vicious, depending on your perspec-
tive) cycle of mutual backscratching among agriculture’s vested inter-
ests. High prices, protected markets, and curbs on farm size meant
more rice farmers and regulations to keep MAFF officials occupied.
More rice farmers meant more fee-paying members for JA and more
votes. And more votes helped the LDP win elections. And so the cycle
would begin again – and again, and again, and again.
But by the end of the 1980s, the writing on the wall was clear:
Japanese agriculture was operating in a dangerous bubble. High, gov-
ernment-guaranteed rice prices were draining government coffers, flee-
cing consumers, and keeping inefficient farmers in business, while
closed rice markets were tweaking the ire of Japan’s trade partners.
JA had mushroomed into an economic juggernaut that dominated the
supply of (overpriced) farm inputs and financial services to producers,
426 Patricia L. Maclachlan and Kay Shimizu

and in ways that benefited its own organizational longevity as much

as – if not more than – the livelihoods of its farmer-members. All the
while, market signals had been dampened to the point where no one
really knew what a “market price” was. The situation was, in a word,
unsustainable.

14.2.2 Chipping Away at the Pillars of Protection

Japanese agriculture was undoubtedly in trouble in 2012, but the long-
term erosion of those five pillars was cause for optimism about the
future of structural reform.
For starters, market signals had slowly begun to take root in rice
pricing. In 1995, Japan abolished the 1942 Food Staple Control Law
and partially liberalized the retail sale of rice. (In the past, rice for
household use could only be sold through government-licensed vend-
ors.) Although the survival of gentan and other measures ensured that
rice prices would remain well above free market levels (whatever those
were), prices began to drop. Against this backdrop, farmers who
wished to preserve their incomes had no choice but to reform.
The trigger behind the Food Staple Control Law’s demise was
Japan’s inclusion in the 1994 Agricultural Agreement of the GATT
Uruguay Round of trade negotiations (Davis 2003). Although Japan
had secured permission to avoid the wholesale opening of its rice
market by subjecting rice imports to tariffication, for the first time in
decades, foreign rice – very expensive foreign rice – was now for sale in
Japan. What is more, Japan was now obligated to reduce its rice price
subsidies.
Japan had also loosened its restrictions on farmland transfers. As
before, ownership was confined to traditional farm households, but it
was now significantly easier for those households to buy and sell
farmland. Measures had also been introduced to facilitate farmland
leasing and to permit nonfarm household enterprises to participate in
agricultural production via investment in “agricultural production
corporations” (APCs); by 2012, up to 25 percent of APC shares
could be held by nonfarm household firms. In addition, the
Agricultural Land Act was amended in 2009 to allow nonfarm house-
hold enterprises to lease farmland directly.
Finally, the deregulation of farmland transfers and of the food pro-
cessing and retail sectors combined with a host of other developments
Japanese Agricultural Reform under Abenomics 427

to weaken JA dominance in the primary sector. By the early twenty-first

century, the Japanese food supply chain was becoming more integrated
as processors and retailers moved upstream and downstream. Lawson,
AEON, and other food retail chains were venturing into agricultural
production by contracting directly with local farmers and, since 2009,
leasing farmland directly. Many large-scale farmers, for their own part,
were processing their own products and selling them directly to con-
sumers or via their own retail outlets. For years, governments referred
to these value-added production chains involving the primary, second-
ary, and tertiary industries (1 + 2 + 3) as the “6th industrialization” (6ji
sangyōka) of agriculture, and were taking modest steps to promote the
process: for example, in 2008 the Diet passed the Law to Promote the
Coordination of Food, Commerce and Industry (Nōshōkō renkei
sokushin hō); MAFF and METI were cooperating to make resources
available to interested parties; and the government established private
and semiprivate organizations at the national and local levels to help
farm enterprises expand their operations.
Although far more could have been done, these changes boosted
agricultural competitiveness. They also helped loosen the glue that
held the conservative farm lobby together. First, the modest introduc-
tion of market signals in rice pricing, downward pressures on farm
household incomes, and diversification of farm enterprises encouraged
a growing number of producers to abandon JA in their search for
cheaper farm inputs – now more readily available outside of the co-
op network in the wake of retail-sector deregulation – and more
profitable branding and distribution opportunities. For decades, JA’s
political clout within the policy process and the electoral sphere had
been based on the nearly universal participation of farmers in the co-op
network; by 2012, those days were gone.
Second, demographic decline in the countryside combined with
decreasing prices and farmland deregulation were accelerating the
exit of inefficient part-time farmers. Between 2005 and 2010, the
number of full-time farm households increased by 1.8 percent, from
443,000 to 451,000, while that of part-time farm households dropped
by a whopping 22.4 percent, from 1.52 million to 1.18 million (MAFF
2010). In 2010, the ratio of part-time to full-time farm households
shrunk from more than five to one to roughly 2.6 to one. Meanwhile,
the average farm size was on an upward trajectory. Between 2005 and
2010, the average paddy farm increased by 56.6 percent, to 5.9
428 Patricia L. Maclachlan and Kay Shimizu

hectares (MAFF 2010). These trends were promising indicators of the

(very gradual) concentration of farmland in the hands of larger-scale,
full-time farm households. This was good for the future of agriculture
but bad for JA, for which part-time farmers had long been its most
reliable members and consumers of co-op services.
Chinks in the exchange relationship between JA and the LDP offered
further cause for optimism about the potential for reform. The cohe-
sion of the organized farm vote had slipped for several reasons, not
least the declining number of farmers in absolute terms and the rapid
diversification of the economic and political interests of those who
remained in the profession. Also important were the effects of electoral
reform. JA’s electoral struggles date back to 1982, when the upper
house’s national multimember district was transformed into a closed-
list, PR district; the change abolished individual candidacies, thus
weakening JA’s overall vote yield. The effects of the reforms were
partially alleviated first by the introduction of opportunities for interest
groups to influence the placement of candidates on the party list
(Koellner 2002) and then by the 2000 transition to an open-list system
that allowed voters to vote for either a candidate or a party
(Maclachlan 2014). But these changes were not nearly enough for JA
to match its postwar record of over a million votes, set during the early
1980s. Meanwhile, at the lower house level, the 1994 transition from
the SNTV system to a combination of SMD and PR districts weakened
the organized vote relative to the independent vote by encouraging
interparty competition along programmatic, rather than particularis-
tic, lines.
Electoral reform also weakened the exchange relationship between
farmers and LDP politicians by changing the structure of the LDP and
the career incentives of individual politicians. Since candidates can no
longer secure a seat in first-past-the-post SMDs by depending solely on
the support of one or a few interest groups like JA, they must cast their
campaign nets more broadly to include a range of different interests
and independent voters. And as rural candidates think beyond their
traditional base of organized support, the LDP’s farm zoku is losing its
clout as spokesperson for farmers in the policy process (see Krauss &
Pekkanen 2011). Individual politicians, for their own part, recognize
that they can no longer climb the LDP career ladder by defining them-
selves solely as representatives of farmers (Interview, Noda), a lesson
that Toshio Yamada – JA’s main “sponsored candidate” in the upper
Japanese Agricultural Reform under Abenomics 429

house, and by all accounts a minor player in the LDP hierarchy – has
learned only too well. Even the MAFF has seen changes to its behind-
the-scenes electoral functions following electoral reform. In the past,
the MAFF would select one of its own to run as an LDP-backed, JA-
sponsored candidate in upper house elections. But by the early 2000s,
the MAFF had abandoned such schemes, thus compelling JA to recruit
candidates – like Toshio Yamada in 2007 – from among its own
leadership corps (Yomiuri Shimbun, March 15, 2015). As a result of
these developments, JA’s access to the policy process has narrowed.
But as Abe himself knows, while the power of the organized farm
vote at the national level is now a shadow of its former self, it remains
important. It may be folly for candidates in rural, winner-takes-all
SMD races to rely exclusively on the farm vote, but alienating it
could very well spell defeat. Furthermore, in the so-called best loser
system, rural LDP candidates who run simultaneously in both their
SMD and PR districts have an added incentive to court the organized
vote in the SMD even if they have no hope of winning, since success in
the PR tier is based on the proportion of the popular vote captured in
the SMD.
Moreover, the decline of the anti-reformist, organized farm vote has
yet to produce a wholesale rift between farmers and the LDP. To the
contrary, given the absence of strong opposition parties, many farmers
continue to view the LDP as the only game in town. Older farmers see it
as their best bet for access to good pensions and other social-welfare
benefits. And many larger-scale, more competitive farmers – or farmers
who want to improve their competitiveness – are pinning their hopes on
the capacity of the LDP’s more progressive politicians to introduce
more market-friendly reforms. In sum, the LDP still has an electoral
constituency in the countryside; the difference is that this constituency
has diversified to the point where it is now far less wedded to JA’s
political preferences and somewhat more likely to support pro-market
reforms.

14.3 Abe’s Agricultural Agenda

If ever there were a symbol of the tug-of-war between neoliberal
reformers and advocates of the status quo in the LDP, it would be
twenty-first-century agricultural policy. Prime Minister Junichiro
Koizumi (2001–2006) kicked things off by vowing to subject
430 Patricia L. Maclachlan and Kay Shimizu

agriculture to deep structural reforms and, in the leadup to the 2005

lower house election, expelling from the LDP farm zoku members who
had voted against his postal privatization bills. But agricultural reform
proved an uphill battle for Koizumi: his pledge to reform JA, for
instance, met with howls of protest from the farm lobby, which threat-
ened to decisively defeat postal privatization if he dared to proceed
(Interview, Honma). Koizumi then watched as his successor, Shinzo
Abe (2006–2007), readmitted those rebels into the party, propped up
rice prices in the weeks before the much-dreaded 2007 upper house
election, and put agricultural reform on the backburner, where it by
and large remained during the next two LDP governments. The future
of agricultural reform looked brighter under Prime Minister Yukio
Hatoyama (2009–2010) of the Democratic Party of Japan (DPJ), who
seemed eager to reform the farm sector and refused to meet with JA
leaders. But in due course his government made peace with JA and
adopted farmer-friendly direct household income subsidies as the DPJ’s
signature agricultural policy (Ishii 2015). Most prime ministers recog-
nized that the future of Japanese agriculture depended on market-
oriented reform, but all were severely handicapped by resistance from
the LDP’s farm zoku, conservative farm bureaucrats, and JA, as exer-
cised through their control over the agricultural policy process.
All, that is, except Shinzo Abe. Upon assuming office for the second
time in 2012, Abe upheld structural reform as a priority of the third
arrow of Abenomics. Abenomics sought to invigorate the economy by
promoting private investment, improving the use of human resources,
and further integrating domestic markets into the international econ-
omy. To fulfill these goals in the agricultural sector, the government
introduced a range of policies, several of which have further weakened
the pillars of what remains of the farm lobby.
Of course, not all of Abe’s policies were successful, as his efforts to
integrate agriculture into the global economy attest. On January 20,
2017, the day US President Donald Trump was sworn into office, the
Japanese Diet ratified the Trans-Pacific Partnership (TPP); within
a week, President Trump rendered the Abe government’s hard-won
victory a hollow one by officially withdrawing the United States – the
most important player – from the pact. For Abe, the twelve-nation
comprehensive trade agreement had been Japan’s ticket to further
globalizing the domestic economy and expanding external sources of
demand for Japanese farm (and other) products as domestic demand
Japanese Agricultural Reform under Abenomics 431

continues to shrink. Although the impact of TPP on the agricultural

sector would have been cushioned by major concessions to vulnerable
farmers, the controversial pact was expected to empower more com-
petitive farmers and open new opportunities for farm exports. The
government went on to play a leading role in the conclusion of an
eleven-nation version of the pact, which included, among other things,
a small but significant relaxation of Japanese tariffs on beef imports;
comparable benefits were extended to the United States in a bilateral
trade agreement concluded in October 2019 (USTR 2019). All told,
these agreements will do little to significantly alter Japanese agricul-
ture. Fears of heightened competition within the beleaguered sector
did, however, trigger a new round of pledges by the Abe government in
late 2019 to boost Japanese food exports.

14.3.1 The Reforms

On December 10, 2013, the Abe cabinet approved the Plan for the
Creation of Agricultural and Regional Vitality (Nōrinsuisangyō · chiiki
no katsuryoku sōzō puran; see Kantei 2014). Revised on June 24, 2014,
the document served as the government’s “grand design” for agricul-
tural and rural revitalization, with the ultimate aim of doubling rural
incomes within ten years. The plan was rolled out amid great fanfare as
the harbinger of a new era of “aggressive agriculture” (seme no nōgyō),
and for good reason: it proposed an ambitious array of market-
oriented reforms, including several that addressed – directly or indir-
ectly – the economic causes of the deepening agricultural crisis and the
lingering power of anti-reformist vested interests.
The plan’s objective to promote farmland consolidation had already
seen some progress: the December 5, 2013, enactment of legislation to
establish prefectural-level “farmland consolidation banks” (nōchi chū-
kan kanri kikō; FLCBs). Designed to help “core” (i.e., full-time and
other economically viable) farmers lower costs by achieving greater
economies of scale, the FLCBs are a more powerful version of organ-
izations introduced in 1970 to authorize local governments to expand
farmland leasing opportunities (Yamashita 2008). By late 2014, all
forty-seven prefectures had FLCBs in place, and the Abe government
was crediting them for increasing the concentration of farmland in the
hands of business farmers. Over the previous three years, concentration
levels had stagnated after increasing from 27.8 percent to 48.1 percent
432 Patricia L. Maclachlan and Kay Shimizu

between 2006 and 2011, but they then increased from 48.7 percent in
2013 to 52.3 percent two years later (MAFF 2017). As of 2014, the
government’s ultimate goal was to increase the concentration of agri-
cultural lands in the hands of core farmers to 80 percent by 2023.
The effectiveness of the FLCBs has varied by prefecture. Among
those that have done fairly well, the leadership of prefectural governors
appears to be key (Interview, Honma). In Kumamoto Prefecture,
Governor Kabashima Ikuo, a strong advocate of agricultural reform,
has effectively pushed bureaucrats in the prefecture’s Bureau of
Agriculture (Nōsuibu) to promote farmland consolidation (Interview,
Kabashima); officials in the Bureau confirmed that it is now much
easier for core farmers to obtain leased land in the prefecture
(Interview, Kumamoto officials). Tottori Prefecture is another success
case; by the end of FY 2014, it had increased the concentration of
prefectural farmland among core farmers from 9 percent a year earlier
to 56 percent (MAFF 2016a, p.20). Other prefectures have been less
successful, including those with high concentrations of farmland in
near-urban areas, where farmers face strong incentives to sell their
holdings for lucrative commercial or residential purposes (Interviews,
Kohno and JA representatives).
The FLCBs have met with passive resistance from farm lobby stal-
warts. Local agricultural committees (nōgyō iinkai) – municipal-level
elected organizations that help govern the transfer of farmland usage
rights – have been a drag on leasing in areas where proactive prefectural
leadership is lacking. And so has JA, which views the FLCBs as a threat
to its local authority and the size of its membership. For the most part,
it is part-time farmers, the mainstay of JA membership, who lease out
their land. Those who lease it are normally either core farmers, who are
growing less dependent on JA services, or nonfarm household agricul-
tural enterprises, which have been gaining a foothold in agriculture
since the partial deregulation of farmland leasing in 2009. The FLCBs
also empower prefectural and local bureaucrats in agricultural affairs,
a development that JA leaders find troubling. Some co-op leaders doubt
the ability of bureaucrats to manage the complex negotiations between
landowners and farmers over the transfer of land rights and would like
to see those tasks delegated to JA authorities. For now, at least, gov-
ernment officials have ultimate say over these crucial transactions.
Another major government objective was the reform of gentan. In
late 2013, the government announced that official production targets
Japanese Agricultural Reform under Abenomics 433

for rice and the payment of subsidies to farmers who comply with those
targets would be abolished within five years. Farmers, in other words,
would now decide for themselves how much rice to produce. The
announcement was widely heralded as a long-overdue attack on rice
price supports that simultaneously protect small-scale, part-time farm-
ers while impeding the growth of more internationally competitive,
larger-scale farmers.
Although the Abe government made significant headway in gentan
reform, the program has not completely disappeared. The government
will continue to distribute information about the rice market to farmers
and subsidize the producers of feed rice – a growing subsector (Kobari
2015, p.91). Not surprisingly, JA’s national leaders have opposed
gentan reform altogether on the grounds that it will destabilize the
demand and supply – and hence price – of rice (Nihon keizai shimbun,
December 23, 2017). More to the point, gentan reform threatens to
drive many of JA’s part-time farmer-members out of business.
Also significant have been initiatives at various levels of government
to encourage and fund the “6th industrialization” (6ji sangyōka) of
agriculture – those value-added chains of production involving the
primary, secondary, and tertiary industries. Abe took a personal inter-
est in farms that venture down the value-added chain, targeting them as
models for the future by visiting them with local dignitaries and the
media in tow. In so doing, Abe not only championed the development
of a more competitive food system outside of JA’s grip: he also
appeared to be turning his back on inefficient, JA-dependent, part-
time farmers.
The reforms chronicled thus far have impacted JA indirectly, by
strengthening market forces and/or JA’s competitors. But as we have
seen, some ignited stiff resistance from JA leaders. In fact, it was JA’s
intense opposition to the TPP and nonfarm household corporate par-
ticipation in farming that sparked the June 2014 revisions to the
government’s Plan for the Creation of Agricultural and Regional
Vitality (Taniguchi 2015, p.10). But, rather than offer concessions to
JA, the revised plan targeted JA itself for reform.
The purpose of the Abe government’s JA reforms was to expand
opportunities for JA’s local multipurpose cooperatives and their
farmer-members to engage consumer demand and other market signals
more freely and effectively (see Maclachlan & Shimizu 2016; Mulgan
2016). To that end, the government laid out an audacious plan: the
434 Patricia L. Maclachlan and Kay Shimizu

abolition of the national organizations that stood atop the pyramid of

the JA Group and that exerted major economic, financial, and even
political controls over local cultivators and cooperatives. The “JA
Group” is a three-tiered network of organizations that spans the
national, prefectural, and local levels. Three of its four main national
organizations were targeted for reform: Zennō, which channels agri-
cultural inputs to farmer-members and distributes their raw product to
market; Zenchū, the all-important “control tower” (shireitō) of the
federation and primary JA spokesperson in government circles; and
Nōrinchūkin (or JA Bank), the group’s massive banking arm.
(Zenkyōren, which oversees the network’s insurance services, was not
targeted for reform.) The government proposed to abolish Zenchū and
its prefectural chapters (chūōkai); transform Zennō into a joint-stock
company, thereby stripping it of its sweeping anti-monopoly exemp-
tions; and reduce the responsibilities of the group’s nearly 700 local
cooperatives to carry out Nōrinchūkin’s financial functions. It also
pledged to reduce the scope of services offered to so-called associate
cooperative members (jun kumiaiin) – nonfarmer members who now
outnumber “regular” farmer members (sei kumiain) by a significant
margin – on the grounds that these services undermine the core prin-
ciples of agricultural cooperatives. In effect, the reforms pushed local
co-ops to focus on their core agricultural responsibilities by reducing
their “multifunctionality” and expanding their capacity to respond
more creatively to the rapidly changing agricultural economy.
After much political wrangling, the amendments to the 1947
Agricultural Cooperative Law that finally cleared the Diet in
August 2015 highlighted the lingering influence of JA and its defenders
in the LDP – a subject we explore in greater detail below. Nōrinchūkin
and the cooperatives’ associate members were spared the reformers’ ax.
The transformation of Zennō into a joint-stock corporation was articu-
lated as an option rather than a requirement. And Zenchū and its
prefectural chapters were preserved. Among the more significant
amendments was the conversion of Zenchū from a cooperative organ-
ization into a general incorporated association (ippan shadan hōjin),
comparable to that of any other trade or industry association in Japan.
Zenchū was also stripped of its authority to impose mandatory audits
on local cooperatives – which it had used to influence the management
and financial decisions of those co-ops – and to collect levies (fukakin)
from the co-ops to help fund, among other things, JA’s lobbying
Japanese Agricultural Reform under Abenomics 435

campaigns. Zennō, for its own part, was forbidden from withholding
various benefits from local co-ops that refused to sign on to a full menu
of Zennō services.
Two other laws were amended in August 2015. The members of
local agricultural committees (nōgyō iinkai) – long a drag on the
loosening of farmland regulations at the local level – were now to be
appointed by municipal mayors instead of elected by local farmers.
A majority of appointed members must be core farmers with special
certification from their local governments, thus increasing the
likelihood that the committees will support future government-led
efforts to rationalize local farming. Second, the Agricultural Land
Act was amended to raise the ceiling on nonfarm corporate owner-
ship of farmland via agricultural production corporations from
25 percent to just under 50 percent (Maclachlan & Shimizu 2015,
pp.463–464).
In an interesting postscript, the Abe government passed another law
in spring 2017 to further weaken the economic foundations of JA
power: the Law to Support the Strengthening of Agricultural
Competitiveness (Nōgyō kyōsōryoku kyōka shien hō). Included in
the provisions were stipulations requiring providers (read: JA) to pub-
licize the prices of fertilizers, pesticides, farm machinery, and other
agricultural inputs. The law also promotes new entrants into the farm
input manufacturing industry and the consolidation of existing players,
as well as subsidies for farmers seeking to upgrade their farm equip-
ment – all with the express purpose of lowering production costs and
hence increasing farm household incomes. Although the law’s passage
received relatively little media attention, it imposes a degree of trans-
parency on JA, which had long refused to disclose its pricing practices
(Interview, Yamashita 2017).
When stacked up against the Abe government’s original blueprint for
reform, these changes to the organizational and economic foundations
of traditional Japanese agriculture appear modest. But when assessed
in relation to what had once been politically feasible, they border on the
radical. JA’s stranglehold over local producers and co-ops has been
loosened, as has the grip of local agricultural committees on local
farmland transfers. And opportunities for nonfarm household enter-
prises like retail chains to participate in agricultural production have
expanded. What enabled the Abe government to accomplish this much
is the focus of the next section.
436 Patricia L. Maclachlan and Kay Shimizu

14.3.2 The Reform Process

The politics of agricultural policy reform underwent a sea change in the
wake of Shinzo Abe’s 2012 electoral victory. As we have seen, agricul-
tural policy for many decades had been hammered out by a troika of
interests – the MAFF, zoku politicians, and JA – that by and large
sidelined the prime minister. Building on fissures within that sub-
government that had been widening for years, Abe introduced major
changes to the policy process, the gist of which was captured in two of
the reigning political slogans of the day: “Kantei agricultural politics”
(Kantei nōsei, or “agricultural politics under Abenomics”) and
“Elevate the government, humble the party” (Seikō tōtei). As these
slogans imply, the point of these procedural reforms was to strengthen
the hand of Abe and his allies vis-à-vis both the bureaucracy and zoku
politicians. (For more on these trends, see Chapter 2 in this volume.)
The MAFF was downgraded to second fiddle right from the start. In
2013, the Kantei established the “Headquarters for the Creation of
Agricultural and Regional Vitality” that consisted of the entire cabinet,
as well as private sector representatives; it was this group, rather than
MAFF bureaucrats, that ultimately produced the Plan for the Creation
of Agricultural and Regional Vitality that was adopted by the cabinet in
December 2013. Established to coordinate the efforts of relevant min-
istries on behalf of broad policy goals, the headquarters drew heavily
on the recommendations of both the Kantei’s Council on Regulatory
Reform (Kisei kaikaku iinkai) and the Industrial Competitiveness
Council (Sangyō kyōsōryoku kaigi), which was established in the
Cabinet Office in January 2013; both councils consisted primarily of
corporate representatives with strong neoliberal proclivities. Against
this backdrop, the MAFF’s responsibilities were made crystal clear:
through its corresponding Headquarters for the Promotion of
Aggressive Agricultural Production (Seme no nōrinsuisangyō honbu),
the MAFF was to implement policies formulated by the Kantei
(Taniguchi 2015, pp.6–9; Murata 2016, p.16).
One revealing example of the MAFF’s downgraded status in the
agricultural policymaking system was the conspicuous absence of min-
isterial representation at the February 5, 2016, meeting of the State
Strategic Special Zones Advisory Committee (Kokka senryaku tokku
shimon kaigi) headed by the prime minister and chaired by Shigeru
Ishiba, the minister responsible for special economic zones. Two zones
Japanese Agricultural Reform under Abenomics 437

were established under the Abe government for agriculture, one in

Niigata Prefecture and the other in Yabu City, Hyogo Prefecture. At
issue during that February 2016 meeting was whether to lift the ceiling
on nonfarm household corporate investment in agricultural production
corporations. One of the participants, Mayor Sakae Hirose of Yabu
City, made an impassioned speech in favor of the measure (Murata
2016, p.11), which was in keeping with his long-standing advocacy of
agricultural policies that respond more effectively to the specific needs
of local communities (Interview, Hirose). The measure met with the
strong support of other committee members as well, including econo-
mist Heizo Takenaka, Koizumi’s former postal privatization czar; all
looked to the Yabu zone as a laboratory for experiments in farmland
deregulation that could later be introduced to the country at large
(Murata 2016, p.14).
The MAFF was not the only nonplayer in this milestone effort to
push the boundaries of farmland liberalization via APCs beyond what
had been achieved during the 2015 reforms: representatives of the farm
zoku were also absent. The zoku’s influence as keepers of the agricul-
tural status quo had already deteriorated by the time Ishiba convened
the February 2016 meeting. Following the rapid diversification of
political interests among the farm population and the changing elect-
oral incentives chronicled earlier, some zoku members had even
acquired something of a neoliberal veneer. As PARC chairperson,
Tomomi Inada – lower house (SMD) member from Fukui
Prefecture’s deeply agricultural 2nd district – defied opposition from
JA and her more conservative zoku colleagues as she pressed for Kantei
leadership on behalf of JA and farmland reform (Inada 2016). Hiroshi
Moriyama – lower house representative from Kagoshima Prefecture,
long-time JA favorite, and former LDP “rebel” – betrayed JA Zenchū
by publicly criticizing its lobbying activities and cooperating with the
Kantei on JA reform (Ishii 2015, p.7); Moriyama would later serve as
one of Abe’s ministers of agriculture. Koya Nishikawa – long-term
lower house member from Tochigi Prefecture and the “don” of the
farm zoku – may have advocated for special protections for Japan’s
“five sacred items” (rice, wheat and barley, beef and pork, dairy, and
sugar) during TPP negotiations (Interview, Yamashita 2015), but he
also pressed for ambitious market liberalizations while serving as min-
ister of agriculture and, from 2015, as head of a special LDP research
committee on agricultural and food strategies. In a telling indication of
438 Patricia L. Maclachlan and Kay Shimizu

his split personality as a rural representative, Nishikawa lost two

elections in his single-member district before 2017 and retained his
seat only by running as a double candidate in the PR tier. Meanwhile,
as these and other farm zoku politicians were being pulled in conflicting
policy directions, posts that had once been monopolized by zoku
leaders were falling to their enemies; in a brazen snub of JA and zoku
conservatives, for example, in October 2015 Abe filled the chairman-
ship of PARC’s all-important Agricultural Committee with Shinjiro
Koizumi, a young anti-zoku politician and neoliberal advocate who
took up the cause of agricultural reform with a zeal reminiscent of his
father’s postal privatization crusade. During his tenure, Koizumi
focused political attention on JA’s price-gouging practices, eventually
spearheading the 2017 enactment of the Law to Support the
Strengthening of Agricultural Competitiveness (Interview, Yamashita
2017).
Conservatives among farm zoku members were outraged by their
exclusion from the State Strategic Special Zones Advisory Committee’s
deliberations on the liberalization of farmland regulation in the special
agricultural zones. But their exclusion undoubtedly facilitated the
introduction of a significant reform: at the March 2, 2016, meeting,
the committee approved the agricultural zones’ exemption from post-
2015 farmland regulations with the backing of both Nishikawa and
then-minister of agriculture Moriyama (Murata 2016, p.13). Both
substantively and procedurally, the decision marked a major departure
from politics as usual within the agricultural policymaking system.
The decision did not, however, mark the ultimate fulfillment of Seikō
tōtei – the supremacy of the government and humbling of the party (i.e.,
anti-reformers within the LDP) – in contemporary agricultural politics.
In a nod to JA’s long-standing opposition to nonfarm household cor-
porate acquisition of farmland, recipients of the exemptions within the
special agricultural zones must submit to strict conditions and moni-
toring by government officials. Moreover, the authority of those agri-
cultural zones to grant such exemptions will expire in five years, thus
slapping a significant curb on the zones’ capacity to serve as laborator-
ies for change at the national level.
Both the processes and outcomes of the State Strategic Special Zones
Advisory Committee deliberations on exemptions from farmland regu-
lations within the special agricultural zones symbolized the state of
agricultural policymaking more generally since December 2012.
Japanese Agricultural Reform under Abenomics 439

Although the Kantei broke new ground in terms of weakening the farm
lobby’s drag on the reform process, in the end, concessions were
granted to agricultural conservatives within the party and in ways
that ensured the party’s cooperation during the policy implementation
stage. Yes, the Kantei took the lead to an extent never before seen in
postwar agricultural politics. But the farm lobby had yet to be perman-
ently muted.

14.4 Conclusion
It is too soon to fully discern the effects of the Abe government’s
reforms on the state of Japanese agriculture. We can, however, point
to a few trends, some – but not all – of which are promising. On the one
hand, the farm population continues to shrink. The number of com-
mercial farm households, for instance, decreased from 1.63 million in
2010 to 1.33 million in 2015 (MAFF 2016a). The average age of
farmers continues to inch upward – it now stands at 66.8 years – and
the farm successor shortage continues unabated. On the other hand,
the number of larger farms – 100 hectares or more in Hokkaido and 20
hectares or more elsewhere in Japan – is growing at a respectable clip
(MAFF 2016a, 2016b), and the expansion of abandoned farmland is
slowing.
The Abe administration was well aware that much more remained to
be done. For example, new sources of consumer demand for Japanese
foodstuffs must be cultivated at home and abroad, the rate of land
consolidation must speed up, and farming must become more profit-
able and hence attractive to younger generations of Japanese. Whether
these and related objectives can be fulfilled will require government
leadership and/or the initiatives of farmers themselves.
In more ways than one, agricultural reform has been to Shinzo Abe
what postal privatization was to Junichiro Koizumi. Substantively,
both prime ministers strived to invigorate free market forces within
their targeted sector with an eye to strengthening the country’s capacity
to cope with the long-term economic effects of a rapidly aging society.
Strategically, both set out to strengthen the Kantei’s leadership and
weaken the policy influence of conservatives in both the bureaucracy
and the LDP. Where Abe parted company from his mentor was in his
determination to weaken the institutional foundations of interest
group opposition to reform. Koizumi indirectly diminished the policy
440 Patricia L. Maclachlan and Kay Shimizu

influence of the postmasters1 that were opposed to postal privatization by

expelling their party allies – those “rebels” – from the LDP before the 2005
general election. But he did nothing to directly weaken the organizational
sources of interest group power, thereby enabling a resurgence of those
“forces of resistance” and a DPJ-orchestrated retreat from postal privat-
ization in 2012. Abe, by contrast, was intent on weakening the institu-
tional foundations of JA as a precondition for permanently expanding the
creative capacity of local co-ops and farmers and reducing JA Zenchū’s
influence over the policy process. Although the 2015 JA reforms fell far
short of the Abe government’s objectives, when assessed against the
backdrop of a steadily shrinking farm population, changing electoral
incentives, gradually liberalizing rice prices/markets and farmland regula-
tions, and the diversification of the food supply chain, they are likely to
further diminish JA’s economic and political power.
Abe also differed from his neoliberal predecessor in his reluctance to
stake his political career on the success of agricultural reform. Koizumi
made several unprecedented political threats on the road to postal privat-
ization, including a promise to resign should his privatization plans be
rejected by the electorate. Abe, by contrast, had other political fish to fry
besides agricultural reform and took pains to downplay agricultural
reform in his election campaigns (see Maclachlan & Shimizu 2015).
At first glance, the Abe government’s electoral victory in
October 2017 – following an electoral campaign in which the LDP
faced virtually no opposition from other parties on agricultural issues –
would seem to suggest that Abe had carte blanche to forge ahead with
his reformist agenda, as he did after the December 2014 general elec-
tion (Maclachlan & Shimizu 2015). But the situation was not quite so
simple. Despite its diminishing vote yields, JA continued to support
sympathetic LDP politicians by mobilizing both campaign volunteers
and financial donations (Asahi Shimbun, December 16, 2017). And JA
was instrumental in pressuring the government to generously subsidize
feed-rice production in the wake of gentan reform.
The last months of 2017 brought some other unwelcome develop-
ments for the proponents of reform, the most dramatic of which was
Nishikawa Koya’s electoral defeat in the October election. At age
seventy-four, Nishikawa was forced to run only at the SMD level,
since he was now a year beyond the cut-off age for simultaneous
candidacy in the PR tier. For all of Nishikawa’s advocacy on behalf
of “protecting that which should be protected,” his defeat proved
Japanese Agricultural Reform under Abenomics 441

a loss for agricultural reform (Hokkaido Shimbun, October 24,

2017). Meanwhile, a conservative backlash gathered momentum
within the party’s farm zoku. In August 2017, for example,
Shinjiro Koizumi was replaced as head of PARC’s Agricultural
Committee by Tetsuro Nomura, a former JA official from
Kagoshima Prefecture. The fallout from these shifts in the political
winds became clear in early December, when the Abe government
sought to liberalize the wholesale distribution system for agricultural
and marine products. But aside from the introduction of measures to
unify transportation and wholesale pricing by requiring providers to
disclose their fee schedules, the initiative failed. Ironically, after
speaking up on behalf of the reforms during the early stages of
party deliberations on the proposals, the young Koizumi remained
largely silent for the duration of the discussions (Daiyamondo,
December 23, 2017, p.12).
In sum, Abe’s reformist efforts lost momentum in the last years of
his prime ministership. But there is still hope for further change
within the farm sector. The gradual liberalization of rice markets
and farmland, increasing competition within the food supply chain,
the erosion of JA and farm zoku influence, and a host of other
changes introduced both before and during the Abe government are
unleashing new competitive energies among Japanese farmers.
Reform within the Japanese farm sector, in other words, has taken
on a life of its own and will likely continue with or without the
government’s involvement.

Note
1. Like JA, the postmasters associations in their heyday gathered more than
a million votes in upper house elections.

References
Davis, Christina L. 2003. Food Fights Over Free Trade: How International
Institutions Promote Agricultural Trade Liberalization. Princeton:
Princeton University Press.
Inada Tomomi. 2016. “Abe seiken no nōsei kaikaku” (The Abe
administration’s agricultural policy reforms). Nihon nōgyō no ugoki
191: 72–88.
442 Patricia L. Maclachlan and Kay Shimizu

Ishii, Hayato. 2015. “Nōsei undō no kakutai wa dono yō ni henkashitekita

no ka” (How has the object of agricultural lobbying changed?). Nihon
nōgyō no ugoki 189: 6–13.
Kantei. 2014. Nōrinsuisangyō · chiiki no katsuryoku sōzō puran (Plan for
the Creation of Agricultural and Regional Vitality). www.kantei.go.jp
/jp/singi/nousui/, accessed December 18, 2017.
Kantei. 2015. “Abenomics is Progressing! Towards the Reinvigoration of the
Japanese Economy.” March. www.kantei.go.jp/singi/keizaisaisei/pdf/k
aikakudankouEN.pdf, accessed April 7, 2015.
Kobari, Miwa. 2015. “[Yotsu no kaikaku] ni okeru kome seisaku no
minaoshi” (Revising rice policy with regards to the ‘four reforms’). In
Abenomics nōsei no ikikata: nōsei no kihon hōshin to minaoshi no
ronten (The direction of Abenomics agricultural policy: basic policies
and issues for reconsideration in agricultural policy), edited by
Taniguchi Nobukazu and Ishii Keiichi, pp.91–107. Tokyo:
Nōrintōkeikyōkai.
Koellner, Patrick. 2002. “Upper House Election in Japan and the Power of
the Organized Vote.” Japanese Journal of Political Science 3(1):
113–137.
Krauss, Ellis S., and Robert J. Pekkanen. 2011. The Rise and Fall of Japan’s
LDP: Political Party Organizations as Historical Institutions. Ithaca,
NY: Cornell University Press.
Maclachlan, Patricia L. 2014. “The Electoral Power of Japanese Interest
Groups: An Organizational Perspective.” Journal of East Asian
Studies 14(3): 429–458.
Maclachlan, Patricia L., and Kay Shimizu. 2015. “The Kantei vs. the LDP:
Agricultural Reform, the Organized Vote, and the 2014 Election.” In
Japan Decides 2014: The Japanese General Election, edited by
Robert Pekkanen, Steven Reed, and Ethan Scheiner, pp.170–182.
London: Palgrave.
Maclachlan, Patricia L., and Kay Shimizu. 2016. “Japanese Farmers in Flux:
The Domestic Sources of Agricultural Reform.” Asian Survey, 56(3):
442–65.
Ministry of Agriculture, Forestry and Fisheries (MAFF). 2010. Report on
Results of 2010 World Census of Agriculture and Forestry in Japan.
Tokyo: MAFF.
Ministry of Agriculture, Forestry and Fisheries (MAFF). 2016a. Report on
Results of 2015 World Census of Agriculture and Forestry in Japan.
Tokyo: MAFF.
Ministry of Agriculture, Forestry and Fisheries (MAFF). 2016b. FY2015
Annual Report on Food, Agricultural and Rural Areas in Japan:
Summary. Tokyo: MAFF.
Japanese Agricultural Reform under Abenomics 443

Ministry of Agriculture, Forestry and Fisheries (MAFF). 2017. FY2016

Annual Report on Food, Agricultural and Rural Areas in Japan:
Summary. Tokyo: MAFF.
Mulgan, Aurelia George. 2000. The Politics of Agriculture in Japan. London:
Routledge.
Mulgan, Aurelia George. 2005. “Where Tradition Meets Change: Japan’s
Agricultural Politics in Transition.” Journal of Japanese Studies 31(2):
261–298.
Mulgan, Aurelia George. 2016. “Loosening the Ties that Bind: Japan’s
Agricultural Policy Triangle and Reform of Cooperatives (JA).”
Journal of Japanese Studies 42(2): 221–246.
Murata, Yasuo. 2016. “[Kantei nōsei] de, shinjiyūshugi nōsei wo tenkai”
[The evolution of neoliberal agricultural policy via kantei-led
agricultural policy]. Nihon nōgyō no ugoki 191: 8–19.
Nōgyō to keizai henshū iinkai. 2011. Kiiwaado de yomitoku gendai nōgyō to
shokuryō · kankyō [Reading contemporary agriculture, food and
environment through key words]. Tokyo: Shōwadō.
Office of the United States Trade Representative (USTR). 2019. “Fact
Sheet on Agriculture-Related Provisions of the U.S.–Japan Trade
Agreement.”
Saitō, Ken. 2016. “Jimintō ga kangaeru Abenomikusu nōsei no nerai” [The
aims of Abenomics agricultural policy according to the LDP]. Nihon
nōgyō no ugoki 191: 20–44.
Taniguchi Nobukazu. 2015. “Abenomics nōsei no [zentaizō]:
zaikaishudōgata nōsei e no tenkan” (An overview of Abenomics
agricultural policy: the conversion to zaikai-led agricultural policy). In
Abenomics nōsei no ikikata: nōsei no kihon hōshin to minaoshi no
ronten (The direction of Abenomics agricultural policy: Basic policies
and issues for reconsideration in agricultural policy), edited by
Taniguchi Nobukazu and Ishii Keiichi, pp.1–22. Tokyo:
Nōrintōkeikyōkai.
Yamashita, Kazuhito. 2008. “The Issues in the Farmland System.” Tokyo
Foundation, December 16. www.tokyofoundation.org/en/articles/200
8/the-issues-inthe-farmland-system, accessed January 15, 2014.

Interviews

Ikuo Kabashima, governor of Kumamoto Prefecture, Kumamoto City,

June 4, 2014.
Masayoshi Honma, University of Tokyo, Tokyo, May 29, 2014.
JA representatives, Kumamoto City, June–July 2014.
444 Patricia L. Maclachlan and Kay Shimizu

Taro Kohno, LDP lower house member from Kanagawa Prefecture,

Tokyo, August 2014.
Sakae Hirose, mayor of Yabu City, Tokyo, June 5, 2014.
Takeshi Noda, lower house member from Kumamoto City, Tokyo, June 6,
2014.
Officials, Kumamoto Prefecture Bureau of Agriculture, Kumamoto City,
June 5, 2014.
Kazuhito Yamashita, Canon Institute for Global Studies, Tokyo,
January 5, 2015.
Kazuhito Yamashita. October 7, 2017.
15 The Politics of Energy and Climate
Change in Japan under the Abe
Government
t r e v o r i n c e r t i a n d p h i l l i p y . l i p s cy

15.1 Introduction
Prime Minister Shinzo Abe came to power in December 2012 prom-
ising to implement an ambitious agenda of structural reforms under
the banner of Abenomics and “Japan is back.”1 The Abe govern-
ment built on pre-existing institutional reforms and centralized pol-
icymaking authority, enabling significant shifts in diverse areas such
as monetary policy, agriculture, and security policy.2 However,
energy and climate change policy stand out for their surprising
stasis. Under Abe, Japan came under widespread international criti-
cism for unambitious climate change goals and its promotion of
coal-fired power plants. Japan’s position in major international cli-
mate change rankings fell to all-time lows (Aldrich, Lipscy &
McCarthy 2019). This is doubly puzzling because Abe came to
power in the aftermath of the 2011 Fukushima nuclear disaster,
which provided a clear opportunity to pursue major reforms in the
Japanese energy sector.
In this chapter, we will argue that Abe’s energy and climate
change policy primarily focused on supporting the growth object-
ives of Abenomics in a policy mix that we call “Abenergynomics”
(Incerti & Lipscy 2018). The emphasis was on maintaining or
lowering energy prices for consumers, increasing competition in
electricity distribution, and industrial policy interventions to target
sectors – such as hydrogen – where Japanese firms possess com-
petitive advantages. There was comparatively less emphasis on
policies to facilitate decarbonization, though the government gen-
erally adopted rhetoric supportive of international climate change
efforts.3

445
446 Trevor Incerti and Phillip Y. Lipscy

15.2 The Puzzling Stasis of Japanese Energy and Climate

Change Policy
The government of Shinzo Abe came to power in December 2012,
when Japan was still reeling from the aftermath of the March 11,
2011, Great Tohoku Earthquake and Tsunami, which led to
a nuclear meltdown at the Fukushima Daiichi nuclear plant (Lipscy,
Kushida & Incerti 2013, 2016). Prime Minister Abe’s Liberal
Democratic Party (LDP) benefited from the perception that the
Democratic Party of Japan (DPJ) mishandled the response to the
Fukushima disaster (Kushida 2014). Furthermore, the DPJ government
had stunned the international community in 2010 by announcing
Japan’s withdrawal from the second commitment period of the Kyoto
Protocol, the principal international framework to address climate
change. Japan, once a leader in energy efficiency and greenhouse gas
mitigation, was criticized as a “fossil” and a “villain” of global climate
change efforts.4
Hence, the Abe government came to power at a pivotal moment
when energy policy – both domestic and international – loomed large
as an urgent priority. Would Japan transition away from traditional
energy sources toward greener alternatives? Would nuclear power be
abandoned or resuscitated? Would large utilities, such as Tokyo
Electric Power Company (TEPCO), see their political influence dimin-
ish as a result of the Fukushima disaster? Could Japan reemerge as
a leader in energy conservation and international climate change
negotiations?
The track record of the Abe government on energy policy raises some
intriguing questions. First, despite a nuclear disaster that displaced
174,000 people and turned a 337 square kilometer area into
a radioactive wasteland,5 Japanese energy policy did not undergo
widely predicted transformations.6 Abe’s government reemphasized
nuclear and coal-fired power plants and weakened the DPJ-
implemented feed-in tariff, which guarantees a set price for renewable
energy generators. Japan shied away from ambitious policy measures
to encourage energy conservation, such as an emissions trading scheme
or a meaningful carbon tax.7 Japan’s long-term targets for decarbon-
ization of the energy supply remained well behind Western counter-
parts. Why did the Fukushima disaster not become the policy turning
point many foresaw?
Energy and Climate Change under the Abe Government 447

Second, why did the LDP maintain its support for nuclear energy,
and how did it manage to win election after election despite this
unpopular and seemingly anachronistic stance? Among major
Japanese political parties, the LDP adopted the most conspicuously
pro-nuclear platform after 2011. Nonetheless, the LDP won landslide
victories in all major elections under Abe’s tenure, and the opposition’s
efforts to galvanize public support around the nuclear issue failed. This
was not what most international observers expected after the Great
Tohoku Earthquake. The catastrophe led distant countries, such as
Germany, to reconsider and abandon their nuclear programs
(Wittneben 2012). Many observers expected Japan to follow
a similar path, shifting its attention to renewable energy and
a reinvigorated emphasis on energy efficiency.
Third, Japan’s status as a laggard on global climate change continued
under the Abe government. Japan is a party to the Paris Agreement of
the United Nations Framework Convention on Climate Change
(UNFCCC), which entered into force in 2016. However, the Abe
government was widely criticized for submitting unambitious emis-
sions mitigation targets and engaging in accounting shenanigans to
inflate its headline contributions (Robiou et al. 2016). This runs coun-
ter to the general pattern of Abe’s foreign policy, which emphasized
international engagement and leadership, particularly after the election
of US President Donald Trump (Solis 2020; Lipscy 2020). Japanese
public support for climate change mitigation has been consistently
high.8 What explains Japan’s continuing backsliding in this critical
area of international cooperation?

15.3 Energy Policy under Abe: Abenergynomics

Abe’s energy policy is best described as what we call “Abenergynomics”:
a set of policies designed to support the economic objectives of
Abenomics with relatively less regard for other factors, such as popular
opinion, opposition from utility companies, or environmental conse-
quences. Abe first and foremost prioritized policies to facilitate economic
growth. Consistent with this objective, the bureaucratic tussling between
the Ministry of Environment (MOE) and the Ministry of Economy,
Trade, and Industry (METI), which had tilted in MOE’s favor during
the DPJ government, was resolved in METI’s favor. Low electricity
prices were prioritized, even if this meant pursuing unpopular policies,
448 Trevor Incerti and Phillip Y. Lipscy

exacerbating CO2 emissions, or undermining electric utilities, part of the

LDP’s traditional support base. Consistent with the increasing industrial
policy orientation of Abenomics (Chapter 1), government support for
new technologies was targeted selectively to areas that clearly offer
a competitive advantage to Japanese firms.
Japanese policymakers have faced common constraints that limit the
scope for aggressive measures to reduce nuclear power or combat
climate change (Incerti & Lipscy 2020). At the most basic level,
Japan is a resource-poor country dependent on foreign energy sources:
there is no attractive near-term alternative to nuclear energy (Vivoda
2012). Japan’s slow economic growth and large public debt make it
difficult to invest in ambitious new solutions or generously compensate
losers from adverse policy changes.9 Institutional constraints have also
been important. Electoral reform in 1994 shifted Japan from a single
nontransferable vote/multimember district system (SNTV/MMD) to
a mixed system emphasizing plurality voting in single-member districts.
The new electoral system makes it difficult to sustain what Lipscy calls
“efficiency clientelism,” policy arrangements that encourage energy
conservation by imposing high energy prices on consumers while redis-
tributing the consequent revenues or rents to organized supporters of
the LDP (Lipscy 2012).
Despite these general constraints, Abe faced a clear opportunity to
significantly alter the course of Japanese energy policy. In several
respects, Abe assumed power during a moment when constraints
were less binding. First, the Fukushima disaster sharply diminished
the prestige and political standing of large utility companies, which
had resisted major reforms in the past.
Second, the Fukushima disaster also provided an obvious rationale for
Japanese officials to pursue major policy shifts in the energy sector.
Prime Minister Naoto Kan of the DPJ sought to use the disaster as an
opportunity to shift Japan toward renewable energy. However, the
disaster also created a potential opportunity to move policy in the
opposite direction, by shifting public attention away from climate
change toward more urgent priorities like securing a stable energy supply
and managing nuclear safety. Internationally, Japanese policymakers
could portray their country’s disappointing performance on climate
change as an unfortunate consequence of the nuclear disaster.
Third, Abe and the LDP benefited from fortuitous circumstances.
The Fukushima disaster reflected deep-seated problems with Japanese
Energy and Climate Change under the Abe Government 449

nuclear regulation, which had developed over many years of LDP rule
(Kingston 2012). The DPJ, particularly under Kan, was hostile toward
to the “nuclear village” – a collection of pro-nuclear allies from util-
ities, the bureaucracy, politicians, the private sector, and academia –
and a natural standard-bearer for dismantling cozy ties between politi-
cians, bureaucrats, and large utility companies. However, the
Fukushima disaster took place under the DPJ’s watch, severely dam-
aging the party’s public image. The LDP could therefore frame its
policy measures – even those that reverted to the status quo ante – as
a necessary return to stability.
Fourth, the disintegration and deep unpopularity of the DPJ after
2012 meant the anti-LDP vote was split, allowing the LDP to secure
landslide victories despite lukewarm public support under an electoral
system that typically rewards consolidation toward two parties
(McElwain 2012). Following the DPJ’s resounding defeat in the
December 2012 lower house election, Japan was left without a viable
opposition party. As such, the Abe government was able to stay in
power despite adopting several strategically chosen policies unpopular
with the general public.10
Under what we call Abenergynomics, Abe pursued an energy policy
mix that prioritized economic growth by lowering electricity prices and
supporting Japanese producers. This explains seemingly contradictory
policy measures, such as increasing competition in the electricity distri-
bution market while reducing incentives for competition from renew-
able power generation: in both cases, the policies were designed to
facilitate lower electricity prices. Abe was willing to pursue policies
unpopular with the general public, but only in cases – such as nuclear
restarts and the construction of new coal-fired plants11 – where the
policies would lower energy costs. Policy measures to mitigate green-
house gas emissions were generally shunned, except in exceptional
cases where support would clearly benefit Japanese firms.12

15.4 The Energy Policy of the Abe Government

In this section, we will provide a brief overview of the context of
Japanese energy policy and measures undertaken by the Abe govern-
ment. We will first discuss general trends in Japanese energy usage and
then discuss major policy measures implemented by Abe. We will
postpone our discussion of nuclear energy and climate change
450 Trevor Incerti and Phillip Y. Lipscy

mitigation to the following sections, where they will receive more

focused attention.
Figure 15.1 depicts Japan’s energy mix during 2006–2018, a period
that encompasses the 2011 Fukushima disaster. The figure includes
energy sources for both power generation and other uses, primarily
industry and automobile transportation. As the figure shows, prior to
the disaster Japan relied on a diversified energy mix predominantly
based on fossil fuels and nuclear power, with limited reliance on
renewables (a category traditionally dominated by hydropower). The
Fukushima disaster altered this energy mix by triggering a total shut-
down of nuclear power generation. Japan responded to the crisis by
increasing its reliance on imported fossil fuels, particularly natural gas,
and reducing overall energy consumption.
Figure 15.2 depicts the share of energy supply attributable to renew-
able energy for several political units of interest from 1990 to 2018.
The figure omits hydropower in order to focus on nontraditional

Figure 15.1 Japanese energy mix before and after Fukushima (2006–2018)
Note: Yearly data. Dashed vertical line at 2010 marks last annual data prior to
March 11, 2011, the date of the Great Tohoku Earthquake.
Source: International Energy Agency, “World Energy Statistics and Balances,”
2019
Energy and Climate Change under the Abe Government 451

Figure 15.2 Cross-national renewable energy share excluding hydropower,

1990–2018
Note: Renewables includes geothermal, solar, tide, wind, and biofuels and
waste.
Source: International Energy Agency, “World Energy Statistics and Balances,”
2019

energy sources. As the figure shows, Japan has lagged behind most of its
international peers in renewable energy for the entire period depicted,
with the exception of South Korea. This has often been attributed to the
political influence of incumbent utilities in Japan, which made large
investments in traditional power plants and prefer not to face competi-
tion from new entrants (Kameyama 2016, p.131). Renewable share
started to increase noticeably in the EU and the United States in the
early 2000s, but Japan’s share increased only slightly. Japan’s renew-
able share increased more rapidly after the Fukushima disaster, reflect-
ing the shutdown of nuclear energy as well as the feed-in tariff
implemented by the DPJ. However, even with the trend increase,
Japan’s renewable energy share remains well below that of Western
countries, and the gap is not closing.
During its tenure, the Abe government was active in passing energy-
related legislation covering issues such as energy efficiency, climate
452 Trevor Incerti and Phillip Y. Lipscy

change, electricity distribution, and nuclear power. Rather than going

through each specific bill, we will focus on five policy areas with high
political and substantive salience in which Abe made significant policy
changes: electricity deregulation; reform of the feed in tariff scheme;
hydrogen fuel cells; nuclear regulation; and climate change policy.

15.4.1 Electricity Deregulation

The Abe government promoted deregulation of the electricity retail
market as a core objective, listing it as a priority under the “third
arrow” of Abenomics.13 Abe sought deregulation to increase competi-
tion in the retail power sector and reduce prices for consumers.
Electricity market deregulation began in Japan in 1995 with an amend-
ment to the Electricity Utility Industry Law, which for the first time
allowed for independent producers of electric power (Wang & Mogi
2017). Deregulation of the electricity retail markets expanded through-
out the next decade, with the establishment of a wholesale market for
electricity in 2003 and a steady increase in users eligible to access the
deregulated markets (large factories in 2003, mid-size factories in
2004, and small factories in 2005) (Nakajima 2016). The expansion
faltered after the DPJ’s electoral victory in 2009 and the subsequent
Fukushima disaster, but it was revived after Abe assumed power. In
April 2016, all residential users were granted access to deregulated
markets, giving consumers a choice beyond Japan’s ten regional
power monopolies for the first time.
The LDP also decoupled electricity transmission and distribution in
order to allow all power retailers access to both transmission and
distribution (Wang & Mogi 2017). Traditionally, electricity gener-
ation, retail, and transmission were controlled by vertically integrated
regional monopolies. In April 2020, however, generation and trans-
mission and retail were required to operate separately: a producer of
electricity cannot also own transmission infrastructure or sell electricity
to consumers. By decoupling generation from transmission, regional
monopolies are no longer be able to discriminate against electricity
produced by rival generators. By decoupling generation from retail,
a producer with market power is no longer able to discriminate against
other retailers.
The goal of these reforms was to introduce competition and lower
electricity prices for Japanese producers and consumers. According to
Energy and Climate Change under the Abe Government 453

METI, electricity supplied by new entrants increased from 5.2 percent

in April 2016 to 14.7 percent by the end of 2018, and 18 percent of
residential users had applied to switch their electricity providers.14
However, as of our writing in 2020, the impact of electricity deregula-
tion on prices was not yet clear, especially as the decoupling of trans-
mission and distribution coincided with the onset of the COVID-19
pandemic, leading to unusual patterns of electricity demand.

15.4.2 Reform of Feed-in Tariffs

Although the DPJ government was generally characterized by “polit-
ical change without policy change” (Lipscy & Scheiner 2012; Kushida
& Lipscy 2013), one exception was the feed-in tariff (FIT) scheme
adopted by Prime Minister Kan in the aftermath of the 2011 disaster.
Kan insisted that passing this legislation would be one of three precon-
ditions for stepping down from his office, and the related legislation
was adopted in August 2011 with unanimous support even in the upper
house, which the DPJ no longer controlled.15
A feed-in tariff encourages the adoption of renewable energy by allow-
ing electricity generated from renewables to be sold back into the grid at an
above-market rate. The Japanese FIT proved remarkably successful at
promoting renewable – particularly solar – power generation. Between
2011 and 2015, solar power generation increased by over fivefold, leading
some to declare the formation of a “solar power bubble.”16
However, this rapid adoption of renewable energy came with several
costs. The increasing size of the renewables market meant Japanese utilities
were being forced to buy ever larger quantities of electricity at premium
rates, and these costs were largely passed onto consumers through higher
electricity prices. By 2015, the premium was estimated to be about 700 yen
(USD $6) per month for an average Japanese household.17 The FIT also
came under criticism from Japanese manufacturers, who saw higher elec-
tricity prices as a competitive disadvantage.
In June 2016, the Abe government reformed the FIT scheme to make
renewable power generation more difficult and less lucrative, and the
measures were implemented in April 2017. The new legislation intro-
duced cumbersome regulations that made new installations more diffi-
cult, and existing energy producers that failed to follow new guidelines
lost their rights to sell power into the grid at a price premium. In
addition, the legislation sharply reduced the feed-in tariff rate, with
454 Trevor Incerti and Phillip Y. Lipscy

the stated goal of reducing the rate to the residential electricity rate by
2019 and the spot market rate as soon as possible after 2020.18 This
policy change was devastating for Japan’s nascent solar power sector,
leading to an abandonment of planned solar installations of 28 million
kilowatts, equivalent to about 10 percent of household electricity
consumption in Japan.19

15.4.3 Hydrogen Society

Compared to other environmentally friendly technologies, the Abe gov-
ernment took a proactive stance on fuel cells. Fuel cells, which run on
hydrogen combined with oxygen from the air, have long been seen as
a cleaner alternative to fossil fuels. Unlike internal combustion engines, fuel
cells do not directly emit greenhouse gasses, and energy efficiency tends to
be higher due to the absence of moving parts.20 The basic technology has
been available for nearly two centuries, but it was limited to niche uses due
to high production costs and the volatile nature of hydrogen.
The Abe government actively promoted fuel cells. The promotion of
hydrogen was prioritized early on in the first Abenomics growth strat-
egy published in 2013.21 The same year, METI established the Council
for a Strategy for Hydrogen and Fuel Cells, and this was followed by
the 2014 publication of a Strategic Road Map for Hydrogen and Fuel
Cells.22 Honda had become the first major automotive company to
bring a fuel cell vehicle to market in 2008, with the Honda FCX
Clarity, and Toyota followed with the 2014 Mirai. In 2015, Abe
declared that Japan was seeing “the dawn of a true hydrogen
society.”23 METI developed an ambitious plan to support putting
40,000 fuel cell vehicles on the road in time for the later-postponed
2020 Tokyo Olympics, and Abe declared that hydrogen automobiles
would be the transportation method for participants in the games.24
Many outside observers are skeptical about fuel cells. Electric vehicles
enjoy many of the advantages of fuel cell cars without the need for expen-
sive new infrastructure and technological development.25 Hydrogen fuel
must be derived from other sources and transported to fueling stations,
which diminishes its environmental advantages. This problem is particu-
larly acute for Japan, which has few domestic natural resources and hence
must inefficiently transport hydrogen on container vessels from distant
locations such as Australia and Brunei.26 Calculations cited by METI
indicate that fuel cell automobiles would emit only slightly less CO2 than
Energy and Climate Change under the Abe Government 455

gasoline-hybrid vehicles on a “Well to Wheel” (emissions over the entire

lifespan of the vehicle, including production, manufacture, distribution,
transportation, etc.) basis, with large improvements only possible if the
hydrogen is derived from renewable sources.27
The Abe government’s support for fuel cells conforms to the pattern
of policymaking seen in some other areas of Abenomics: an industry
policy orientation that sought to boost growth by benefiting domestic
industry (see Chapter 1). Japanese automobile makers, particularly
Toyota and Honda, were at the forefront of fuel cell technology,
while they lagged behind in the electric vehicle market, ceding leader-
ship to foreign competitors such as BYD, General Motors, and Tesla.28
As of 2019, the share of electric vehicles in the Japanese market
(1.1 percent in 2018) lagged behind both China (4.5 percent) and the
United States (2.5 percent), not to mention aggressive promoters like
Norway (46.4 percent).29 As the main impediment to fuel cell automo-
biles is cost and the absence of fueling infrastructure, government
support on these two fronts would ultimately benefit Japanese auto-
makers. Providing support to the nascent market for fuel cell vehicles
was essentially industrial policy under the guise of environmentalism.

15.4.4 Discussion
The policies discussed thus far illustrate the logic of Abenergynomics:
the use of energy policy to stimulate Japanese economic growth. On the
face of it, deregulation of electricity distribution and scaling back of
FIT reflect seemingly contradictory priorities: the former challenged
incumbent utilities dependent on traditional power sources and sup-
ported new entrants, while the latter did precisely the opposite.
However, both policies sought to lower electricity prices for end users
and hence theoretically put more money in the pockets of consumers
and industry. Abe’s enthusiastic endorsement of hydrogen fuel cells
was essentially a form of industrial policy, subsidizing Japanese auto-
makers and related industries against foreign competitors.

15.5 The Paradox of Pro-nuclear Policy after Fukushima

In this section, we will examine the Abe government’s nuclear policy.
Despite overwhelmingly negative public opinion toward nuclear power
in post-Fukushima Japan, the LDP embraced a strikingly pro-nuclear
456 Trevor Incerti and Phillip Y. Lipscy

policy platform. This fits into the general theme of Abenergynomics:

nuclear restarts provided a stable, cheap source of electricity in the
short and medium terms, even while longer-term concerns about
nuclear safety remained contested. Abe’s pro-nuclear policy was viable
despite lingering public skepticism in large measure due to the collapse
of the DPJ and subsequent failure of opposition consolidation. The
nature of the “nuclear village” evolved, as the Fukushima disaster
undermined the credibility of large utility companies and led to more
stringent government oversight. However, compared to the DPJ, the
Abe government moved nuclear policy significantly back toward the
status quo ante.

15.5.1 Abe’s Nuclear Policy

The LDP adopted a conspicuously pro-nuclear policy after the
Fukushima disaster. As Figure 15.3 shows, the Japanese public grew
deeply skeptical toward nuclear power after 2011, with around 40 per-
cent supporting a reduction in nuclear power plants and 30 percent
supporting the closure of all plants. Seeking to capitalize on this anti-
nuclear sentiment, all major opposition parties proposed a complete
phase-out of nuclear energy, and many local courts and politicians
attempted to block nuclear restarts in their own backyards. The LDP
stood apart, proposing that Japan receive roughly 20–22 percent of its
electricity supply from nuclear power by 2030. This 20–22 percent
nuclear share was proposed by METI and approved by the Abe Cabinet
in 2014 and reiterated in 2018. Although this represented a reduction
from pre-Fukushima planning, which sought to push nuclear power
above 30 percent,30 it would require restarting most of Japan’s shut-
tered nuclear plants, a tall order given political, regulatory, and tech-
nical constraints.

15.5.2 Prioritization of Economic Growth over Appeals to

Anti-nuclear Sentiment
Japan’s 1994 electoral reform increased the incentives of politicians
to make broad appeals to consumers, reducing the viability of pol-
icies that raise consumer energy prices (Rosenbluth & Thies 2010;
Lipscy 2019). As Figure 15.4 shows, electricity prices in Japan have
been in secular decline, but they rose sharply after the Fukushima
Energy and Climate Change under the Abe Government 457

Figure 15.3 Public opinion on nuclear power in Japan: “What do you think
should be done with the current number of nuclear plants in Japan?”
Source: NHK Research, “Genpatsu to enerugi ni kansuru iken chousa” (Public
opinion poll on nuclear and energy), 2011–2015

disaster as Japan’s nuclear power plants shut down, increasing by

approximately 29 percent between 2010 and 2014. Abenergynomics
saw these price increases largely reversed, helped by a decline in
international fossil fuel prices as well as a permissive policy toward
coal-fired plants and nuclear restarts. There is an inherent tension
between reducing consumer energy costs and responding to anti-
nuclear sentiment. The Abe administration also faced pressure by
Japan’s businesses to lower electricity prices: Keidanren quickly
emerged as a strong supporter of nuclear restarts, stating that “the
process of restarting nuclear power plants must be accelerated to
a maximum extent” and arguing that restarts were necessary to
maintain economic competitiveness.31
Abe also promoted the export of Japanese commercial nuclear tech-
nology. The government signed a $22 billion export deal for the
construction of Turkey’s second nuclear reactor in 2013, reached
a civil nuclear agreement with India, and negotiated with Saudi
Arabia over a similar arrangement.32 Restarting Japan’s nuclear plants
458 Trevor Incerti and Phillip Y. Lipscy

Figure 15.4 Japanese electricity prices, 1970–2017

Note: Yearly data. Dashed vertical line is placed at 2010, as this was the
last year of data unaffected by the March 11, 2011, disaster.
Source: Ministry of Economy, Trade and Industry, 2019 Energy White Paper

was seen by Abe as a necessary advertisement for the facilitation of

these deals. Indeed, Vietnam pulled out of an $11 billion deal in 2016
due to safety fears.

15.5.3 Post-Fukushima Regulatory Reform and LDP

Pushback
In the wake of the Fukushima nuclear disaster, many observers pointed
to a lack of regulatory oversight as a key cause of the accident. Some
blamed “regulatory capture” by the nuclear power industry, which
resulted in subpar safety regulations.33 Japan’s nuclear regulatory
apparatus was overhauled by the DPJ to increase the independence of
regulatory authorities. Numerous regulatory deficiencies were identi-
fied and addressed.
First, in September 2012, the former Nuclear and Industrial Safety
Agency (NISA) was abolished and replaced with the Nuclear
Regulatory Authority (NRA) within the Ministry of Environment.
Energy and Climate Change under the Abe Government 459

The NRA was tasked with creating new nuclear regulations and deter-
mining whether existing Japanese plants could resume operations.
NISA was criticized for conflicts of interest due to its location within
METI, which is responsible for the promotion of nuclear power. Only
one year prior to the Fukushima disaster, METI released a plan to
supply 70 percent of Japan’s energy via nuclear power.34 Prime
Minister Kan’s initial “Basic Policy on the Reform of an
Organization in Charge of Nuclear Safety Regulation” included “the
separation of nuclear power regulation and promotion” as its first
suggested reform.35 A second advisory body under the administration
of the Cabinet Office, the NSC (Nuclear Safety Commission), was also
abolished and consolidated under the NRA (Ferguson & Jansson 2013;
Kingston 2013). The NSC ostensibly had the role of creating new
nuclear safety rules, but, as an advisory body, it could not actually
force NISA to adopt new standards.
Amakudari – descent from heaven, or the practice of retired bureau-
crats taking positions in the private sector – was also blamed for
regulatory capture. Amakudari is not necessarily pernicious: firms
may hire former regulators because regulators possess valuable skills,
particularly in an industry like nuclear power with high skill barriers
and few opportunities for outside employment. However, lucrative
positions offered to retired METI bureaucrats by the power sector
created an obvious potential conflict of interest. The DPJ attempted
to increase transparency by requiring the creation of a list of former
regulators reemployed in the nuclear industry (Ferguson & Jansson
2013; Kingston 2013). In addition, the DPJ passed a bill banning
former ministry employees from returning to employment within
their respective ministries after working for the NRA.36
Each of these regulatory reforms occurred under DPJ rule. Given that
the DPJ sought to eventually eliminate nuclear power in Japan, and the
Abe administration sought to restore Japan’s commercial nuclear produc-
tion, it is no surprise that the LDP reversed some of the DPJ’s reforms. Abe
and other senior LDP lawmakers publically called for swift decision-
making on nuclear restarts, a proposition at odds with the NRA’s goal of
ensuring safety at all costs. The Abe-led LDP therefore pushed to reverse or
limit some of the regulatory reforms instituted after the Fukushima disaster
as they sought to revive nuclear power in Japan’s energy mix.
Some of these efforts took place while the DPJ was still in power.
While the DPJ proposal to place the NRA within the Environment
460 Trevor Incerti and Phillip Y. Lipscy

Ministry ultimately passed, the LDP proposed a separate NRA struc-

ture that would have created an independent body outside of the
umbrella of any ministries. Critics feared this more “autonomous”
body would have limited authority and fall prey to pro-nuclear
appointments. The Environment Ministry has consistently promoted
renewable energy and clashed with METI over the implementation of
post-Fukushima energy policy. For example, after the Environment
Ministry revealed an estimate of a potential Japanese energy future
including 35 percent renewable energy in 2015, METI Minister Yoichi
Miyazawa stated that the plan was not “feasible” and “could not be
used as the basis for the energy mix.”37 The LDP’s proposed body may
therefore have been more vulnerable to METI and industry influence.
Three months after the Fukushima nuclear disaster, the DPJ established
an Energy and Environment Council in order to debate future energy
policy changes and formulate a national energy plan (Takao 2016). The
Council was chaired by the minister of state for national policy, effectively
eliminating METI’s former control over national energy policy.
Nevertheless, METI created its own Advisory Committee for National
Resources and Energy, even if it could no longer formulate national energy
policies. The Energy and Environment Council ultimately recommended
a phase-out of nuclear power by 2040, a policy Keidanren, METI, and the
LDP all opposed. Upon returning to power in December 2012, the LDP
promptly abolished the Energy and Environment Council and placed
METI back in charge of the creation of future energy plans.38 National
energy policy was thereafter debated and formulated within METI’s
Advisory Committee for National Resources and Energy, and most anti-
nuclear members of the Committee were removed from their positions
(Takao 2016).
Some LDP NRA appointments also came under scrutiny from anti-
nuclear advocates and proponents of regulatory reform. For example,
in 2014 the LDP nominated and appointed Satoru Tanaka, a nuclear
engineering university professor and former Japan Atomic Industrial
Forum official, who received payments from TEPCO, Hitachi-GE, and
engineering firm Taiheiyo Consultant in the three years prior to his
nomination. By contrast, none of the five original NRA commissioners
nominated in 2012 received payments of any kind (including for aca-
demic research) from a utility in the three years prior to their appoint-
ments. Tanaka’s appointment was symbolic to anti-nuclear advocates,
as he was assigned to replace seismologist Kunihiko Shimazaki, whose
Energy and Climate Change under the Abe Government 461

assessment of fault lines under Japan Atomic Power Company’s

Tsuruga and Kansai Electric’s Ōi plants effectively foreclosed early
restarts. Utility officials reportedly described Shimazaki’s retirement
as a “small victory.”39 Shimazaki continued to argue that Kansai
Electric’s calculation methods may underestimate the maximum size
of a temblor at the Ōi plant, a claim his successor Tanaka called
“groundless.”40
Finally, despite more stringent NRA safety criteria, the traditional
“nuclear village” of the LDP, METI, utilities, and Keidanren was
partially revived by the LDP. Upon returning to power, the LDP
immediately abolished the DPJ’s independent Energy and
Environment Council and reverted control of national energy policy
to METI. The LDP, Keidanren, and utilities pushed for quick nuclear
restarts in order to reduce electricity prices, lower production costs,
and increase the profitability of beleaguered utilities.

15.5.4 Discussion
Despite public skepticism toward nuclear power in post-Fukushima
Japan, the Abe government focused on restarting nuclear plants and
retaining nuclear power as a permanent component of Japan’s energy
mix. Nuclear power was an essential part of Abenergynomics due to its
ability to supply cheap electricity in the short term and relieve Japan
from excessive dependence on foreign energy sources. For sure, the Abe
government’s nuclear policy did not represent a complete return to the
status quo ante. Nuclear regulation under the NRA remained more
independent than under NIRA, and local governments and courts
sought to block nuclear restarts. It is therefore doubtful that nuclear
energy will ever return to comprising 30 percent of Japan’s energy mix.
However, it is also clear that the Abe government actively sought to
reverse elements of stringent nuclear regulatory policies put in place by
the DPJ.

15.6 From Climate Change Leader to Climate Change Villain

In this section, we will consider the Abe government’s approach toward
greenhouse gas emissions and international climate change negoti-
ations. Japan’s disappointing track record on climate change predates
Abe, and the Fukushima disaster sharply increased Japan’s reliance on
462 Trevor Incerti and Phillip Y. Lipscy

fossil fuels. However, Abe generally pursued policies that lowered

energy prices in order to boost economic growth, even as it encouraged
higher greenhouse gas emissions and invited international criticism.

15.6.1 Japanese Climate Change Policy Prior to Abe

Japan was an early leader in global energy conservation efforts. Japan’s
aggressive response to the 1970s oil shocks established the country as
a frontrunner in energy efficiency and rationalization. Japan’s response
to the oil shocks reflected “efficiency clientelism,” increasing the cost of
energy consumption on consumers while using the revenues or rents to
reward concentrated political supporters of the LDP (Lipscy 2019). As
concerns over greenhouse gases and climate change intensified in the
1980s and 1990s, Japanese policymakers saw an opportunity to
assume global leadership in the new post–Cold War international
order. Measures to mitigate greenhouse gas emissions would also
benefit Japanese firms, which had developed a competitive advantage
in energy conservation technologies, such as hybrid automobiles and
energy-efficient household appliances.
It was therefore natural for Japan to play a major role in the
UNFCCC negotiations that culminated in the 1997 Kyoto Protocol.
Japanese negotiators accepted an aggressive target of a 6 percent reduc-
tion in greenhouse gas emissions to build support for the agreement.
Japanese firms took advantage of the landmark agreement to showcase
their leading technologies: Toyota accelerated production of its Prius
hybrid to bring the automobile to market in time for the Kyoto
meeting.41
However, Japan’s track record under the Kyoto Protocol was deeply
disappointing. Figure 15.5 depicts carbon intensity for Japan against
the average of European G7 countries. Carbon intensity is a measure of
how much CO2 a country emits per unit of GDP. For many years,
Japan’s economy stood apart for its low carbon intensity, and signifi-
cant reductions in intensity were achieved after the 1970s oil shocks.
However, improvements essentially stopped after the 1990s, a period
during which global climate change concerns intensified. As the figure
shows, European countries did not experience comparable stagnation
and traded places with Japan. This reflected institutional changes in
Japan, such as electoral and administrative reform, which rendered
Energy and Climate Change under the Abe Government 463

Figure 15.5 CO2 emissions intensity (Mt CO2/2010 USD billions), Japan vs.
European G7 countries, 1960–2015
Note: CO2 emissions intensity is the amount of CO2 emitted by a country per
unit of economic output. Japan achieved a sharp reduction in CO2 emissions in
the aftermath of the 1970s oil shocks, but progress stagnated just as concerns
over global climate change intensified. The European G7 countries are France,
Germany, Italy, and the United Kingdom.
Source: International Energy Agency, “World Energy Statistics and Balances,”
2017

Japan’s earlier model of energy efficiency unsustainable (Lipscy &

Schipper 2013).
Japanese emissions actually rose during the Kyoto commitment
period rather than declining 6 percent as envisioned. In absolute
terms, Japan was the worst performer under the Kyoto Protocol: even
accounting for land-use change and forestry, Japan exceeded its target
by 44 metric tons of carbon dioxide equivalent (MtCO2e) (Shishlov
et al. 2016). Japan only met its international treaty obligations by
making heavy use of flexibility mechanisms, effectively transferring
financial resources to countries such as the Czech Republic, Ukraine,
and Estonia to take credit for their emissions reductions.
When it assumed power in September 2009, the DPJ government
was sharply critical of the LDP’s performance on climate change. Prime
464 Trevor Incerti and Phillip Y. Lipscy

Minister Yukio Hatoyama surprised the international community and

many Japanese policymakers by announcing that Japan would reduce
CO2 emissions by 25 percent from 1990 levels by 2020. This goal was
widely ridiculed as unrealistic, and the DPJ ultimately could not
develop concrete policies to follow through on Hatoyama’s ambitious
pledge. In fact, many signature DPJ policies (e.g., elimination of high-
way tolls, reduction in gasoline taxes) cut against the emissions goal by
encouraging greater consumption of fossil fuels. Ultimately, the DPJ
made essentially no meaningful domestic progress in reducing
emissions.
Furthermore, the DPJ government stunned the international com-
munity in 2010 by announcing that it would categorically oppose
the second commitment period of the Kyoto Protocol. Once at the
forefront of international climate change cooperation, Japan was dis-
paraged for being a “fossil” and a climate change “villain.”42 The 2011
Fukushima disaster exacerbated Japan’s predicament by sharply
increasing the country’s reliance on imported fossil fuels.

15.6.2 The Abe Government’s Climate Change Policy

During his first, short-lived stint as prime minister in 2006–2007, Abe
took some steps to promote Japanese leadership on climate change,
declaring to the Diet that he would set up a “21st-Century
Environmental Nation-Building Strategy,” and the Environment
Ministry announced that Japan would play a central leadership role
in the post-Kyoto climate negotiations.43 By the time Abe reassumed
power in 2012, the situation had changed considerably. Far from
playing a leadership role, Japan had exited the Kyoto Protocol, and
the aftermath of the Fukushima Daiichi disaster shifted attention away
from climate change toward more immediate priorities.
Upon assuming power in 2012, Abe declared that he would “recon-
sider from zero start” (i.e., from scratch) the ambitious CO2 reductions
target espoused by the DPJ government: a 25 percent reduction in CO2
from 1990s levels by 2020.44 In 2013, the Cabinet formally abandoned
the emissions reduction goal and announced a new target equivalent to
a 3 percent increase in CO2 emissions from 1990s levels.45 Japan came
under international criticism for what was widely perceived as an
unambitious emissions target. As discussed, the Abe government also
Energy and Climate Change under the Abe Government 465

slowed the adoption of solar power generation in Japan by altering the

feed-in tariff scheme.
Under the Paris Agreement of 2015, Japan was obligated to publish
Intended Nationally Determined Contributions (INDC) to the
UNFCCC outlining its proposed emissions reduction plan. Abe was
not personally enthusiastic about climate change mitigation, but he
was unwilling to give up Japan’s status as an important contributor in
global climate change efforts. To this end, he directed government
ministries to produce an INDC that would lower electricity costs in
Japan while making Japan’s target appear comparable to those of the
EU and the United States.46
To reconcile these contradictory goals, Japanese bureaucrats pro-
duced an energy mix that relied heavily on coal, which reduced antici-
pated electricity costs, while choosing a base year – 2013 – that would
paint Japan’s headline emissions reduction number in the most favor-
able light possible: a 26 percent reduction by 2030, compared to
18–21 percent for the United States and 24 percent by the EU. Using
the 1990 base year in the Kyoto Protocol, the number was far less
impressive, putting Japan (18 percent) far behind the EU (40 percent)
and in a range comparable to the USA (14–16 percent).47 Furthermore,
Japan’s INDC was criticized for relying on accounting gimmicks, such
as excluding land use, land use change, and forestry from the base year
but including it in the target year to inflate the headline emissions
reduction level.48
Abe’s government also came under criticism for encouraging the
construction of new coal-fired power plants. The Abenomics growth
strategy accelerated the construction of fossil fuel power plants: for
example, the Abe government reduced the environmental assessment
period necessary for new construction of such plants from three years
to one.49 Under Abe, Japan became the only G7 country pursuing the
construction of new coal-fired plants.50 The Abe government also
pushed back against international efforts to limit foreign aid funding
of coal-fired plants in developing countries, arguing that coal repre-
sented the cheapest source of electricity.51 This policy also benefited
Japanese firms, which were at the forefront of energy efficiency tech-
nology for coal-fired plants.
One well-known indicator of climate change policy is provided by
Germanwatch, a nongovernmental organization that ranks countries
according to their performance and policies on climate change
466 Trevor Incerti and Phillip Y. Lipscy

mitigation (Burck et al. 2017). Figure 15.6 depicts the percentile ranking
of Japan in the Climate Policy category of the Germanwatch Climate
Change Performance Index. The Climate Policy category reflects the
assessments of surveyed experts about a country’s national and inter-
national climate change policies. As the figure shows, expert perceptions
of Japan’s national policies were rising consistently from a low base
until 2012, the year Abe came to power. Thereafter, Japan’s ranking fell
sharply to among the lowest of all countries ranked. The same basic
pattern holds for the international component, although there is more
volatility before 2012, likely attributable to the inconsistent policy
pronouncements under the DPJ government. The most recent version
of the Germanwatch report described Japan’s long-term CO2 targets as
“too unambitious” and lacking “a concrete roadmap for implementa-
tion”; it noted that Japan’s ranking suffered from “continued large
provision of public finance for coal overseas” (Burck et al. 2019).

100

90 Abe Government
80
Percentile Rank of Japan

70

60
National
50 International
40
30
20
10

0
2005 2010 2015 2020
Year

Figure 15.6 Percentile rank of Japan in the climate policy component of the
Germanwatch Climate Change Performance Index
Note: The Climate Policy component of the index reflects expert assessments of
national and international climate policy. Japan’s percentile rank in both the
national and international measures have declined under Abe’s tenure to
among the lowest in the world. The slight bounce in 2018 may reflect
a change in wording of the survey sent to experts and should not necessarily
be interpreted as an improvement in Japan’s performance.
Energy and Climate Change under the Abe Government 467

15.6.3 Discussion
Japan’s climate change policy has failed to live up to its leadership
ambitions in the 1990s. In part, this reflected long-term factors that
preceded Abe. Japan also achieved high levels of energy efficiency
relatively early on, making incremental mitigation relatively costly in
international comparison (Victor et al. 2017). However, even account-
ing for these factors, it is clear that Abe placed a low priority on climate
change mitigation, consistently favoring policies that reduce domestic
energy prices at the expense of higher CO2 emissions. Abe was not
a climate change denier like some leaders in Australia and the United
States, who openly dismissed the scientific consensus behind global
warming. Abenomics growth strategies generally emphasized Japan’s
commitment to climate change, such as the “realization of a carbon-
free society.”52 The appointment of the young and charismatic Shinjiro
Koizumi as minister of environment in 2019 put the spotlight on
international criticism toward Japan on issues like coal-fired power
plants. However, as Abe announced his resignation in August 2020, it
was clear that Japanese policy had not lived up to the country’s stated
ambitions to assume a global leadership role on climate change.

15.7 Conclusion
As our exposition illustrates, the Abe government’s energy policy
reflected both continuity and disjuncture from previous periods. In
several respects, Abe’s assumption of power represented a return to
traditional LDP energy policymaking only briefly perturbed by the DPJ
in 2009–2012: the prioritization of economic growth over environ-
mental issues, and hence greater bureaucratic influence of METI over
the MOE; rejection of the DPJ’s radical post-Fukushima environmental
policies, such as the 25 percent emissions reduction target and anti-
nuclear policy; and a relative shift back toward Japan’s traditional
energy mix in favor of nuclear and coal at the expense of renewables.
Abe’s government also faced the same constraints that rendered the
DPJ’s energy policy inchoate. Japan’s low economic growth rate and
dire public finances limit the scope for ambitious policy initiatives.
Japanese industry is struggling under international competitive pres-
sures and maintains a credible threat of international relocation, mak-
ing policymakers hesitant to impose stringent energy conservation
468 Trevor Incerti and Phillip Y. Lipscy

measures. The Fukushima disaster sharply increased Japan’s reliance

on fossil fuels and made new nuclear power plant construction
a nonstarter for the foreseeable future. Japan’s current electoral system
makes it difficult to raise the price of energy consumption on the
general public, limiting policy tools to encourage energy conservation
and mitigate greenhouse gas emissions.
Nonetheless, we should not exaggerate the political constraints faced
by Abe. Abe’s government came to power with greater leeway to
pursue an aggressive climate change agenda compared to predecessors.
The collapse of the DPJ and subsequent splintering of the opposition
meant the LDP could reliably return to power despite adopting at least
some policy positions out of favor with Japanese voters.53 Abe chose to
use this leeway primarily to reinvigorate nuclear power, which had
become deeply unpopular after the Fukushima disaster. Under
a different prime minister, it is conceivable that the political opportun-
ity afforded by opposition failure could have been used for other
priorities, such as accepting higher electricity prices for the more vigor-
ous promotion of renewables, a national cap-and-trade system, or an
ambitious supply-side decarbonization strategy.
Abe’s policies did not represent a complete return to traditional
Japanese energy politics. The Fukushima disaster undermined the
political influence and credibility of large utilities, which now face
more stringent regulatory oversight and meaningful competition
compared to the status quo ante. Although Abe watered down the
DPJ’s feed-in tariff scheme, the basic principles remain in place, and
the government’s long-term planning still foresees a significant
increase in renewable energy generation by 2030 to levels exceeding
nuclear power. Hughes sees substantial cumulative changes in
Japanese energy policy as a form of “radical incrementalism” despite
significant institutional constraints (Hughes 2018). Abe’s promotion
of hydrogen fuel cells and associated infrastructure represent
a gamble that few of his international counterparts have been willing
to take.
Abe promoted himself as a defender of the international liberal order in
the wake of President Trump’s assumption of the power in the United
States, symbolized by Japan’s resuscitation of the Trans-Pacific
Partnership (Solis 2020). Similar leadership was not forthcoming in cli-
mate change. Japan’s performance on CO2 emissions reductions was no
more impressive than that of the United States, despite the latter country’s
Energy and Climate Change under the Abe Government 469

withdrawal from both the Kyoto and Paris agreements. Japan’s INDC
associated with the Paris Agreement relied heavily on accounting gim-
micks designed to exaggerate emissions reductions. Japan’s deteriorating
profile in international climate change cooperation preceded Abe.
However, Abe’s government failed to reverse the tide.
What are the future prospects for energy policy in Japan? The basic
preconditions for Abenergynomics are likely to remain in place under
LDP leadership, but the longer term is less certain. During Abe’s tenure,
Japanese opposition parties, including the most credible Constitutional
Democratic Party of Japan (CDPJ) – agreed on nuclear power elimin-
ation but remained split on a host of other issues, such as constitutional
revision. Certain aspects of Japan’s electoral institutions may make
opposition consolidation challenging (Carey & Shugart 1995;
Scheiner 2006). Nonetheless, as the DPJ illustrated, two-party consoli-
dation is not an impossibility in Japan.
It is important to note, however, that two-party consolidation could
actually exacerbate Japan’s energy predicament by constraining viable
policy options. For example, credible competition from a consolidated
opposition could make it more difficult for the LDP to continue pub-
licly unpopular measures like nuclear restarts and coal-fired power
plant construction, increasing Japan’s reliance on energy from unstable
sources like the Middle East and Russia. A credible opposition could
also increase pressure to keep a lid on energy prices, further limiting
demand-side measures on climate change.
Nonetheless, it may be difficult for Japanese policymakers to sidestep
international and domestic criticism over its climate change policies
indefinitely. The Fukushima disaster and the election of Donald Trump
shifted attention away from Japan’s disappointing track record on
greenhouse gas mitigation. However, the Japanese public still sees
climate change as an important priority. An eventual return to nor-
malcy could refocus international attention on Japan and build greater
pressure for more ambitious measures.

Notes
1. See Chapter 1 of this volume.
2. See Chapters 2, 6, 14, and 16.
3. Shinzo Abe, “Join Japan and act now to save our planet,” Financial
Times, September 24, 2018.
470 Trevor Incerti and Phillip Y. Lipscy

4. “Japan said “cast as villain” at Cancun climate talks,” BBC,

December 12, 2010; “‘Fossil’ Japan seen as obstacle in Cancun,”
Reuters, December 2, 2010.
5. This is the area of the area designated by the government as “Kitaku
Konnan Kuiiki [Difficult to Return to Zones],” in which there is no
realistic prospect for residents to be able to return for the foreseeable
future. See “Kitaku Konnan Kuiiki ni Tsuite [About Difficult to Return
to Zones],” Naikaku Genshiryoku Hisaisha Seikatsu Shien (Cabinet
Office Disaster Victim Livelihood Support Team), October 1, 2013.
6. See, e.g., Martin Fackler and Andrew Pollack, “Japan scraps plan to
build new reactors;
Prime minister shelves nuclear project and vows to focus on
renewables,” International Herald Tribune, May 5, 2011; Eric
Johnston, “Current nuclear debate to set nation’s course for decades,”
Japan Times, September 23, 2011.
7. Japan did implement a carbon tax in 2012 under the DPJ government,
but the level of the tax is trivial compared to other countries. See Carl
and Fedor (2016).
8. See, for example, Cabinet Office of Japan, Chikyu Ondanka Taisaku ni
Kansuru Yoron Chosa (Public Opinion Poll on Global Warming
Counter-Measures), 2016.
9. See discussion of the politics of compensation in, among others, Calder
(1991, 2017); Estevez-Abe (2008); and Naoi (2015).
10. Other examples include the 2015 Legislation for Peace and Security,
which reinterpreted Japan’s constitution to allow collective self-defense
under limited circumstances, and the 2013 Act on the Protection of
Specially Designated Secrets, which created criminal penalties for
leaking state secrets.
11. A public opinion poll by the Japan Consumers’ Co-operative Union
found that only 27 percent of respondents preferred nuclear energy
to be a 15 percent or greater share of Japan’s energy mix in 2030,
compared to the Abe government’s official target of 20–22 percent.
Similarly, respondents were asked whether they preferred fossil fuels
with “high CO2 emissions and low cost (coal)” or “lower CO2
emissions even if this means higher cost (natural gas),” and
72 percent of respondents expressed a preference for natural gas
over coal. See Japan Consumers’ Co-operative Union, “Kore kara
no denryoku no arikata ni tsuite no shohisha ishiki chosa” (Survey
of consumers regarding the future of electricity), May 8, 2015.
12. Experts who personally briefed Abe on environmental measures, such as
cap-and-trade, note that he expressed no interest in these initiatives.
Energy and Climate Change under the Abe Government 471

13. See, e.g., Cabinet Office, Yawaraka Seicho Senryaku: Abenomics wo

Motto Mijikani (Slow growth strategy: bringing Abenomics closer),
May 2015.
14. METI, Energy White Paper 2019, p.158.
15. House of Councillors, The National Diet of Japan, Honkaigi Tohyo
Kekka, www.sangiin.go.jp/japanese/joho1/kousei/vote/177/177–0826-
v004.html.
16. “Taiyoko Hatsuden, Utage no Ato” (Solar Power Generation, After the
Banquet), Nikkei Shinbun, May 13, 2017.
17. METI, Kaisei FITho ni Kansuru Chokuzen Setsumeikai (Explanation of
the revision of the FIT law), February 3, 2017, www.enecho.meti.go.jp
/category/saving_and_new/saiene/kaitori/dl/kaisei/fit_2017setsumei.pdf.
18. Ibid.
19. “Taiyoko Hatsuden, Utage no Ato” (Solar Power Generation, After the
Banquet), Nikkei Shinbun, February 13, 2017.
20. U.S. Department of Energy, 5 Fast Facts about Hydrogen and Fuel Cells,
https://energy.gov/eere/articles/5-fast-facts-about-hydrogen-and-fuel-cells.
21. Government of Japan, Nihon Saiko Senryaku: Japan Is Back, June 14,
2013.
22. Agency for Natural Resources and Energy, METI, METI has compiled
a Strategic Road Map for Hydrogen and Fuel Cells, www.meti.go.jp/
english/press/2014/0624_04.html.
23. “Betting on Hot Air,” The Economist, May 9, 2015.
24. Umair Irfan, “Japan thinks hydrogen will rule. Does anyone else?”
Sasakawa Blog, May 22, 2017, https://spfusa.org/sasakawa-blog/japan-
thinks-hydrogen-will-rule-anyone-else/; Shinzo Abe, “Dai Nihyaku
Ikkai Kokkai ni Okeru Abe Naikaku Sori Daijin Shisei Hoshin
Enzetsu,” January 20, 2020.
25. Chisaki Watanabe, “Japan Is at Odds with Elon Musk Over Hydrogen
Fuel Cells,” Bloomberg, February 9, 2017.
26. METI, Suiso Kihon Senryaku (Basic Hydrogen Strategy), December 16,
2017, www.cas.go.jp/jp/seisaku/saisei_energy/kaigi_dai2/siryou1-1.pdf.
27. METI, Suiso Shakai no Jitsugen ni Muketa Torikumi ni Tsuite (Looking
toward the implementation of a hydrogen society), February 24, 2015.
28. “Electric cars not ready for mass production yet: Toyota chairman to
Spiegel,” Reuters, November 17, 2017; Michael Barnard, “6 of 10 Big
Electric Car Companies Are in China,” CleanTechnica, November 23,
2017.
29. International Energy Agency, Global Ev Outlook (Paris: International
Energy Agency, 2019): table A7.
472 Trevor Incerti and Phillip Y. Lipscy

30. World Nuclear Association, Nuclear Power in Japan, 2017, www

.world-nuclear.org/information-library/country-profiles/countries-g-n/j
apan-nuclear-power.aspx.
31. Keidanren, A Proposal for Future Energy Policy, October 2013, www
.keidanren.or.jp/en/policy/2013/089.html.
32. Tom Corben, “Japan’s Nuclear Exports: Risky Business,” The
Diplomat, December 2017, https://thediplomat.com/2017/12/japans-
nuclear-exports-risky-business/.
33. See for example, Ferguson and Jansson (2013); Kingston (2013).
34. “Lessons from Fukushima,” East Asia Forum, 2016, www
.eastasiaforum.org/2016/03/07/lessons-from-fukushima/.
35. METI, Basic Policy on the Reform of an Organization in charge of
Nuclear Safety Regulation, 2015, www.meti.go.jp/english/earthquake/
nuclear/iaea/pdf/20110911/annex4.pdf.
36. Daniel Aldrich, “Post-Fukushima Nuclear Politics in Japan, Part I,” The
Monkey Cage, 2013, http://themonkeycage.org/2013/04/post-
fukushima-nuclear-politics-in-japan-part-i/.
37. “LDP proposes future energy policy heavy on nuclear power,” Japan
Times, April 7, 2015.
38. World Nuclear Association, Nuclear Power in Japan.
39. “Abe picks for NRA ‘undermine’ nuclear watchdog’s independence,”
Japan Times, June 11, 2014.
40. “Local consent needed despite OK to restart Oi nuclear plant,” Asahi
Shimbun, February 23, 2017; “Quake scale may be underestimated in
calculations for nuclear plants: Ex-NRA official,” Mainichi, June 17,
2016.
41. “Toyota Steps on the Gas: A Leaner, Tougher Company Gambles on
Global Leadership with New ‘Eco-Car,’” Washington Post, December
14, 1997.
42. “Japan said “cast as villain” at Cancun climate talks” BBC,
December 12, 2010; “‘Fossil’ Japan seen as obstacle in Cancun,”
Reuters, December 2, 2010.
43. Ministry of the Environment, 21 Seiki Kankyou Rikkoku Senrayku
(Twenty-First-Century Environmental Strategy), www.env.go.jp/guid
e/info/21c_ens/index.html.
44. See, for example, Cabinet Office, Nihon Saiko Senryaku: Japan Is
BACK (Japan Revival Strategy: Japan Is BACK), June 14, 2013.
45. Kameyama (2016), pp.139–140. The Japanese government modified
the base year to 2005 in order to portray this target as an emissions
reduction.
46. Ibid., p.150.
47. Ibid., p.149.
Energy and Climate Change under the Abe Government 473

48. Climate Action Tracker, Japan, http://climateactiontracker.org/coun

tries/japan.html.
49. Cabinet Office, Yawaraka Seicho Senryaku: Abenomics wo Motto
Mijikani (Slow growth strategy: bringing Abenomics closer), May 2015.
50. Chris Littlecot, “Snapshot of Japan Coal Phase Out Progress,” E3G,
www.e3g.org/library/snapshot-of-japan-coal-phase-out-progress.
51. Kameyama (2016), pp.155, 159–160.
52. Government of Japan, “Action Plan of the Growth Strategy,” June 21,
2019, www.kantei.go.jp/jp/singi/keizaisaisei/pdf/ap2019en.pdf.
53. See Chapters 1 and 4 in this volume.

References
Aldrich, Daniel, Phillip Y. Lipscy, and Mary M. McCarthy. 2019. “Japan’s
opportunity to lead.” Nature Climate Change 9: 492.
Burck, Jan, Franziska Marten, Christoph Bals, Niklas Höhne,
Carolin Frisch, Niklas Clement, and Kao Szu-Chi. 2017. The Climate
Change Performance Index 2018. Germany: GermanWatch.
Burck, Jan, Ursula Hagen, Niklas Höhne, Leonardo Nascimento, and
Christoph Bals. 2019. The Climate Change Performance Index 2020.
Germany: GermanWatch.
Calder, Kent E. 2017. Circles of Compensation: Economic Growth and the
Globalization of Japan. Stanford, CA: Stanford University Press.
Calder, Kent E. 1991. Crisis and Compensation. Princeton, NJ: Princeton
University Press.
Carey, John M., and Matthew Soberg Shugart. 1995. “Incentives to
Cultivate a Personal Vote: A Rank Ordering of Electoral Formulas.”
Electoral Studies 14(4): 417–439.
Carl, Jeremy, and David Fedor. 2016. “Tracking Global Carbon Revenues:
A Survey of Carbon Taxes Versus Cap-and-Trade in the Real World.”
Energy Policy 96: 50–77.
Estevez-Abe, Margarita. 2008. Welfare Capitalism in Postwar Japan: Party,
Bureaucracy, and Business. Cambridge: Cambridge University Press.
Ferguson, Charles D., and Mark Jansson. 2013. “Regulating Japanese
Nuclear Power in the Wake of the Fukushima Daiichi Accident.”
Federation of American Scientists Issues Briefs.
Hughes, Llewelyn. 2018. “Japan’s Radical Incrementalism in Power Market
Regulation and Renewable Energy.” In Japan’s Energy Conundrum:
A Discussion of Japan’s Energy Circumstances and U.S.–Japan Energy
Relations, edited by Phyllis Yoshida, pp.59–67. Washington, DC:
Sasakawa Peace Foundation.
474 Trevor Incerti and Phillip Y. Lipscy

Incerti, Trevor, and Phillip Y. Lipscy. 2018. “The Politics of Energy and
Climate Change in Japan under Abe: Abenergynomics.” Asian Survey
58(4): 607–634.
Incerti, Trevor, and Phillip Y. Lipscy. 2020. “The Politics of Energy in
Japan.” In The Oxford Handbook of Energy Politics, edited by
Kathleen J. Hancock and Juliann Emmons Allison. Oxford: Oxford
University Press.
International Energy Agency. 2017. Global Ev Outlook. Paris: International
Energy Agency.
Kameyama, Yasuko. 2016. Climate Change Policy in Japan: From the 1980s
to 2015. Abingdon: Taylor & Francis.
Kingston, Jeff. 2012. “Japan’s Nuclear Village.” Asia-Pacific Journal 10(37).
Kingston, Jeff. 2013. “Japan’s Nuclear Village.” In Critical Issues in
Contemporary Japan, edited by Jeff Kingston. New York: Routledge.
Kushida, Kenji E. 2014. “The Fukushima Nuclear Disaster and the
Democratic Party of Japan.” Japanese Political Economy 40(1):
29–68.
Kushida, Kenji E., and Phillip Y. Lipscy. 2013. Japan under the DPJ: The
Politics of Transition and Governance. Stanford, CA: Brookings
Institution/Shorenstein APARC.
Lipscy, Phillip Y. 2012. “A Casualty of Political Transformation? The
Politics of Japanese Energy Efficiency in the Transportation Sector.”
Journal of East Asian Studies 12(3): 409–439.
Lipscy, Phillip Y. 2019. “The Institutional Politics of Energy and Climate
Change.” Book Manuscript. Stanford University.
Lipscy, Phillip Y. 2020. “Reformist Status Quo Power: Japan’s Approach
toward International Organizations.” In The Crisis of Liberal
Internationalism: Japan and the World Order, edited by
Yoichi Funabashi and G. John Ikenberry. Washington, DC: Brookings
Institution Press.
Lipscy, Phillip Y., Kenji E. Kushida, and Trevor Incerti. 2013. “The
Fukushima Disaster and Japan’s Nuclear Plant Vulnerability in
Comparative Perspective.” Environmental Science & Technology 47
(12): 6082–6088.
Lipscy, Phillip Y., Kenji E. Kushida, and Trevor Incerti. 2016. “Were Japan’s
Nuclear Plants Uniquely Vulnerable?” In Learning from a Disaster:
Improving Nuclear Safety and Security After Fukushima, edited by
Edward D. Blandford and Scott D. Sagan, pp.157–182. Stanford, CA:
Stanford University Press.
Lipscy, Phillip Y., and Ethan Scheiner. 2012. “Japan under the DPJ: The
Paradox of Political Change without Policy Change.” Journal of East
Asian Studies 12(3): 311–322.
Energy and Climate Change under the Abe Government 475

Lipscy, Phillip Y., and Lee Schipper. 2013. “Energy Efficiency in the Japanese
Transport Sector.” Energy Policy 56: 248–258.
McElwain, Kenneth Mori. 2012. “The Nationalization of Japanese
Elections.” Journal of East Asian Studies 12(3): 323–350.
Nakajima, Daiki. 2016. “Japan’s Energy Market Reform: Full Retail Choice
in Electricity Market.” Presentation slides. New York: JETRO
New York. www.jetro.go.jp/ext_images/canada/pdf/powermktsemi
nar070416keynote.pdf.
Naoi, Megumi. 2015. Building Legislative Coalitions for Globalization in
Asia: Globalization as Legislation. New York: Cambridge University
Press.
Robiou du Pont, Yann, M. Louise Jeffery, Johannes Gütschow,
Joeri Rogelj, Peter Christoff, and Malte Meinshausen. 2016.
“Equitable Mitigation to Achieve the Paris Agreement Goals.”
Nature Climate Change 7: 38.
Rosenbluth, Frances McCall, and Michael F. Thies. 2010. Japan
Transformed: Political Change and Economic Restructuring.
Princeton, NJ: Princeton University Press.
Scheiner, Ethan. 2006. Democracy without Competition in Japan:
Opposition Failure in a One-Party Dominant State. Cambridge:
Cambridge University Press.
Shishlov, Igor, Romain Morel, and Valentin Bellassen. 2016. “Compliance
of the Parties to the Kyoto Protocol in the First Commitment Period.”
Climate Policy 16(6): 768–782.
Solis, Mireya. 2020. “Follower No More? Japan’s Leadership Role as
a Champion of the Liberal Trading Order.” In The Crisis of Liberal
Internationalism: Japan and the World Order, edited by
Yoichi Funabashi and G. John Ikenberry. Washington, DC: Brookings
Institution Press.
Takao, Yasuo. 2016. Japan’s Environmental Politics and Governance : From
Trading Nation to Eco Nation.
Victor, David G., Keigo Akimoto, Yoichi Kaya, Mitsutsune Yamaguchi,
Danny Cullenward, and Cameron Hepburn. 2017. “Prove Paris Was
More Than Paper Promises.” Nature 548: 25–27.
Vivoda, Vlado. 2012. “Japan’s Energy Security Predicament Post-
Fukushima.” Energy Policy 46: 135–143.
Wang, Nan, and Gento Mogi. 2017. “Deregulation, Market Competition,
and Innovation of Utilities: Evidence from Japanese Electric Sector.”
Energy Policy 111: 403–413.
Wittneben, Bettina B. F. 2012. “The Impact of the Fukushima Nuclear
Accident on European Energy Policy.” Environmental Science &
Policy 15(1): 1–3.
part v

Foreign Policy
16 Japan’s Defense Reforms under
Abe
Assessing Institutional and Policy Change
adam p. li ff

16.1 Introduction
Immediately after the 2012 election that returned him to power, Prime
Minister Shinzo Abe identified national security as one of his top
priorities.1 Over the past seven years, and with Japan confronting
what its 2019 defense white paper identifies as a “security situation
surrounding Japan [that] has become increasingly severe at extremely
high speeds,”2 Abe has exercised active and pragmatic leadership in
pushing through significant institutional and policy reforms. Japan’s
longest-ever serving prime minister, Abe has established himself in the
area of defense as one of its most consequential postwar leaders.
With Prime Minister Abe having just announced his planned resig-
nation as this chapter went to press (August 28, 2020), the Abe admin-
istration’s place in the history books assured for its longevity, and his
administration having identified as major concerns for Japan’s national
security “changes in the balance of power,” “uncertainty over the
existing order,” and the existing paradigm of [Japan’s] national secur-
ity being fundamentally changed” by emerging threats in new
domains,3 it is a particularly appropriate time to take stock and assess
the significance of the Abe government’s defense reforms heretofore. In
keeping with this volume’s unifying theme – an inquiry into whether
the Abe government represents a “major turning point” in the trajec-
tory of postwar Japan – this chapter asks the following specific ques-
tions. How significant have the Abe government’s defense reforms
been? How much, and what, has changed since December 2012; and
what continuities exist? Where major reforms have been achieved, how
has Abe’s government been able to pursue its ambitious security agenda
while avoiding the domestic political backlash that has frustrated his
predecessors? In other words, what are the enabling and constraining

479
480 Adam P. Liff

factors shaping the context in which Abe and his LDP allies pursue their
ambitious national security reform agenda today?
This chapter argues that defense reforms under Abe thus far already
constitute, in aggregate, a practically significant shift for Japan – albeit
one whose size and allegedly transformative nature is easily exagger-
ated. Institutionally, the Abe era has witnessed significant further cen-
tralization of national security policy decision-making in the
executive – especially through the creation of Japan’s first-ever
National Security Council and Secretariat. In terms of defense policy,
in response to Japan’s evolving threat environment the Abe govern-
ment has accelerated major changes to the force structure and posture
of Japan’s Self-Defense Forces (JSDF); doubled down on Japan’s long-
standing emphasis on its alliance with the United States as the “corner-
stone” of both Japan’s territorial defense and regional peace and sta-
bility, while also expanding Japan’s contributions to it; and pursued
more extensive defense links with third countries within and beyond
East Asia.
On the question of whether Abe’s time in office marks a “major
turning point” for Japan’s defense posture, much remains up for debate.
Of particular importance is a key caveat emerging from this chapter’s
analysis: contrary to widespread and often uncritical assertions of Abe as
a “nationalist” ideologue, when (1) baselined in Japan’s defense reform
trajectory already underway before his return to the Kantei and (2)
assessed in the context of Japan’s rapidly changing regional security
environment, the Abe government’s reform agenda is best understood
as a pragmatic and evolutionary response that to significant degree
constitutes a continuation (and acceleration) of longer-term trends.
Also important, yet often overlooked: key long-standing pillars of
Japan’s remarkably self-restrained defense policy remain in place.
Finally, regardless of Abe and his LDP’s particular ambitions in the
national security domain, at key moments they have significantly dialed
back in response to various domestic political signals from within and
outside the ruling coalition – including on signature policy issues that
constitute long-standing party priorities, such as collective self-defense,
major defense budget increases, and Article 9 reinterpretation/revision.
Though the defense reforms the Abe government has achieved thus
far are undoubtedly significant, this study also suggests at least two
major implications for thinking about a post-Abe era. First, because
many major reforms achieved under Abe build on longer-term trends
Japan’s Defense Reforms under Abe 481

and have attracted support from within and outside the conservative
wing of Abe’s “right-of-center” LDP – including more liberal LDP
members, Komeito (the LDP’s “pacifistic” junior coalition partner),
and moderates from the (now defunct) “left-of-center” Democratic
Party of Japan (DPJ) – much of the contemporary discourse appears
to exaggerate both the particular significance of Abe as an individual
and his ideology as a driving force in his approach to national security.
This, in turn, suggests that evolutionary defense reforms in response to
Japan’s changing security environment are likely to continue, even after
Abe is no longer in office. Second, several long-standing domestic
constraints on Japan’s defense policy, which have frustrated gener-
ations of conservative LDP leaders (Abe included) coveting a more
radical transformation of Japan’s defense posture, appear likely to
persist. Especially salient examples are Japan’s dire fiscal climate
(exacerbated by an aging and shrinking population), which severely
limits defense spending increases, and continued domestic political
resistance to formal revision of the existing clauses of Japan’s never-
amended 1947 Constitution’s “pacifist” Article 9.4

16.2 Japan’s Changing Defense Posture under Abe

Beginning with Abe’s inaugural press conference immediately follow-
ing the December 2012 election that returned him to power, it was clear
he came into office with an ambitious national security agenda – both in
terms of institutional reforms and actual defense policy.5

16.2.1 Defense Policy-Relevant Institutions

Although it is defense policy that typically attracts newspaper head-
lines, the institutional context in which security threats are analyzed,
new policy initiatives are formulated and implemented (or frustrated),
and crises are managed also has a very important role to play in shaping
outcomes. It therefore deserves significant attention in any analysis of
Japan’s defense reforms under Abe.
Consistent with a longer-term trend predating 2012, the Abe gov-
ernment’s institutional reforms have focused on accelerating central-
ization of foreign policy decision-making in the executive branch
(cabinet) and the Prime Minister’s Office (Kantei). They have been
remarkable successful. Indeed, Abe and the Kantei have arguably had
482 Adam P. Liff

more direct influence over foreign policy decision-making than any

previous administration.6 Key objectives of this push for greater con-
trol have been: (1) to improve strategic planning, interagency coordin-
ation, and crisis management; and (2) to enhance the speed and
flexibility with which leaders can react to rapidly changing (worsening)
and increasingly complicated strategic environment surrounding
Japan. Two institution-related developments since 2012 are particu-
larly noteworthy.
First, and most importantly, in December 2013 the Abe government
stood up Japan’s first-ever National Security Council (NSC) and
National Security Secretariat (NSS).7 As Abe himself stated earlier
that year, the purpose of the NSC was to serve as a “control tower”
that was “centered on the prime minister” and responsible for “flexible
and regular discussions of diplomatic and security affairs from
a strategic perspective.” Acknowledging the need for Japan to not
only engage in long-term strategic planning but also be able to react
and ensure interagency coordination in a potentially fast-moving
national security crisis, Abe also called for “an environment for rapid
responses based on strong political leadership.”8 Accordingly, the
NSC’s core feature is a regular meeting convening key national security
principals – i.e., the prime minister, chief cabinet secretary, minister of
foreign affairs, and minister of defense – for discussions of long- and
short-term security concerns. Since 2014, the NSC has met far more
frequently than any predecessor institution and has been supported by
a robust, new NSS. The NSS is headed by a secretary-general (Japan’s
de facto “national security adviser”) and staffed primarily by roughly
seventy to eighty civil servants seconded from various ministries, the
JSDF, the National Police Agency, etc.9 Inter alia, the NSS helps to
consolidate the policies of Japan’s various agencies into
a comprehensive national strategy and to draft major national security-
related documents.10 Originally composed of six core teams (coordin-
ation, strategic planning, intelligence, and three regional affairs teams),
in spring 2020 the administration added a new “economic security”
unit tasked with reviewing economics-related national security topics
(e.g., investments, telecommunications, cybersecurity) and staffed with
roughly twenty experts from relevant agencies.11
Though undoubtedly a breakthrough development, the NSC’s estab-
lishment during the Abe government’s first year was also the natural
culmination of a decades-long – and supra-partisan – reform effort to
Japan’s Defense Reforms under Abe 483

bolster political leadership over Japan’s bureaucracy, to transcend

stove-piping across the bureaucracy through enhanced interagency
coordination, and to more directly involve the JSDF in defense-
related consultations. Importantly, its establishment received support
from opposition parties and was driven by various factors, in particular
a recognition across Japan’s government and political class of (1) the
shortcomings of past institutions; (2) the likelihood that twenty-first-
century crises would be extremely fast-moving and thus require rapid
response and internal coordination; and (3) that the very meaning of
“national security” had expanded in such a way – e.g., to include not
only traditional defense and emerging domains of cyberspace, space,
and the electromagnetic spectrum but also terrorism, so-called gray
zone challenges, and economic and financial issues – as to require far
more effective and rapid sharing of expertise, interagency coordin-
ation, and whole-of-government responses.12
A second, less-heralded development under Abe, which also contrib-
uted to centralization of national security decision-making in the
executive, is more active political involvement in bureaucratic person-
nel decisions.13 Perhaps most famously, the Cabinet Bureau of
Personnel Affairs, established in 2014, reviews appointments to high-
level administrative posts.14 In the area of defense, even prior to the
Bureau’s establishment the Abe government had already begun to
directly elevate favored personnel. Examples include the first-ever
appointments of an active-duty officer, rather than a career bureaucrat,
as Japan Coast Guard (JCG) commandant and a new director-general
of the Cabinet Legislation Bureau (CLB) originating from the Ministry
of Foreign Affairs (rather than simply elevating the CLB’s deputy
director-general) – both in 2013.15 The timing of each appeared to be
motivated by the Abe government’s concerns related to defense policy.
The new JCG commandant had operational experience in the East
China Sea – judged to be important in light of the JCG’s role on the
front lines of Japan’s territorial dispute with China (which had taken
a major turn for the worse just before Abe returned to power in late
2012).16 With regard to the CLB, the outside appointment came in the
run-up to the Abe cabinet’s effective assertion of a favored “reinter-
pretation” of Article 9 to allow Japan to exercise its (UN Charter-
sanctioned) right to collective self-defense – something the CLB had
for decades deemed unconstitutional – albeit with important caveats
(see next section).17
484 Adam P. Liff

16.2.2 Defense Policy Shifts

New institutions, especially the NSC, NSS, and 2013 national security
strategy, transformed the context in which the Abe government subse-
quently pursued its ambitious slate of defense policy reforms. Of par-
ticular salience is the 2015 passage of “peace and security legislation,”
which consisted of revisions to ten existing laws as well as a new
International Peace Support bill.18 Inter alia, this package of legislation
gave the legal foundation for a major revision of the Guidelines for
Japan–US Defense Cooperation in 2015 and the Abe government’s
aforementioned constitutional reinterpretation to enable Japan to exer-
cise the right of collective self-defense “under limited conditions.”
Major components of this package of security legislation had been
under debate for years. Their net effect was to broaden the scope of
JSDF roles and missions to bolster deterrence and to expand the cir-
cumstances under which it could play a role in regional and global
security – within and beyond the US–Japan alliance – and outside
a strict “armed attack on Japan” scenario. As with many developments
during the Abe era, it was framed as enabling Japan to expand its
“proactive contributions to peace”19
Though a full review of the massive legislation package is beyond the
scope of this chapter, it basically affected three kinds of JSDF
operations:20 First, and related to collective self-defense, it expanded
the circumstances in which the JSDF can act in response to an armed
attack against a third country “that is in a close relationship with
Japan.” This expansion, however, came with three fairly strict condi-
tions unique to Japan and absent from the UN Charter itself, such as
that the armed attack against a third party to which Japan could
(theoretically) respond must itself pose a “threat to [Japan’s] survival”
(kuni no sonritsu).”21 Whether or not this newly asserted, if limited,
right to exercise collective self-defense is ever acted upon, it has created
new opportunities for defense planning, training, and exercises with
the militaries of the United States and other friendly countries. Second,
the security legislation enables JSDF overseas deployments in inter-
national peace support activities, though, importantly, mainly in non-
combat roles – e.g., logistical support for US armed forces, search-and-
rescue operations, and ship inspections.22 This, too, comes with condi-
tions unique to Japan, such as that support must be suspended if the
area becomes a combat zone. Similarly, though JSDF personnel are
Japan’s Defense Reforms under Abe 485

newly empowered to use weapons in certain situations to defend

themselves or others under their supervision in the event of attack,
they are expected to evacuate if combat breaks out.23 Finally, the
legislation enables JSDF to use weapons to protect foreign military
forces if those forces are involved in peacetime activities contributing
to Japan’s defense – e.g., military exercises or ISR operations. For
example, under these new authorities the JSDF conducted its first-ever
maritime escort (of a US Navy supply ship) and aerial escort (of a US
B-1 bomber) of US military forces in 2017.24 In certain circumstances,
it can also use weapons in UN peacekeeping operations or to rescue
Japanese nationals overseas.25 In short, though the legislation has
enabled practically significant policy changes, important restrictions
unique to Japan still exist on many of the new authorities. Both should
be acknowledged to ensure a nuanced understanding of the legisla-
tion’s impact. Furthermore, something being made legal does not
necessarily mean it will ever actually be exercised. The latter will be
a political decision.
Beyond the 2015 security legislation itself, the major pillars of the
Abe government’s defense reforms are reflected in Japan’s first-ever
National Security Strategy (2013), subsequent annual defense white
papers, and its two paired national defense program guidelines
(NDPG) and Medium-Term Defense Programs (MTDPs) – released in
2013 and 2018.26 The remainder of this section summarizes notable
aspects of three of these mutually reinforcing core pillars: Japan’s
efforts to bolster its indigenous defense capabilities; to strengthen its
alliance with the United States; and to deepen defense links with third
countries (most of whom, not coincidentally, are also close allies or
security partners of Washington).

16.2.2.1 Pillar 1: Bolstering Japan’s Defense Capabilities

Since long before Abe’s return to power, Japan’s evolving defense
posture had refocused toward procuring more advanced capabilities
and rationalizing JSDF force structure and posture to confront
a rapidly changing regional security environment. Key goals have
been to enhance situational awareness, strengthen missile defense,
and develop more expeditionary capabilities. Confronted with
novel challenges, successive administrations have also attempted
to improve “joint-ness” (e.g., coordination and interoperability
across the JSDF’s ground, maritime, and air services), to enhance
486 Adam P. Liff

Japan’s ability to respond not only to an armed attack but also to an

array of contingencies that are neither a pure peacetime nor a “war”
scenario (i.e., contingencies occurring in the so-called “gray zone”),
and to strengthen capabilities in both traditional and emerging
domains (especially cyberspace, space, and the electromagnetic spec-
trum). That said, the Abe government’s impact has been significant,
especially due to its exceptional stability and longevity – in contrast,
the average tenure of the six previous prime ministers was only
one year – and Abe’s personal commitment to championing reforms.
A particularly notable focus of the Abe government’s efforts to
bolster territorial defense has been a relative prioritization of
a possible contingency with China over or near Japan’s remote
southwestern islands, especially the contested Senkaku Islands
(also claimed by China as the Diaoyu Islands). This focus is
a response to China’s rapid military (especially naval) buildup
and its expanded military and paramilitary activities in the East
China Sea and western Pacific Ocean – especially since fall 2012
(see Section 16.4). Building off the DPJ-era (2010) NDPG,27 the
Abe government’s 2013 NDPG called for a “Dynamic Joint
Defense Force” focused on improving the JSDF’s ability to respond
to “an attack on remote islands.”28 Japan has since sought to
strengthen deterrence through improving intelligence, surveillance,
and reconnaissance as well as implementing shifts to force struc-
ture and posture not only to the JSDF but also to the law enforce-
ment-focused JCG, which serves on the front lines near the
Senkakus. Regarding JCG, a priority has been strengthening its
ability to be present in the waters nearby 24/7 and to bolster its
capabilities through major budget and personnel increases, expand-
ing and upgrading facilities from Okinawa on down, and procuring
new, purpose-built vessels (e.g., a new twelve-vessel Senkakus
Territorial Waters Guard).29 In defending remote islands, the
most recent (2018) NDPG emphasizes developing stand-off,
amphibious, and transport capabilities; sustained presence; and
missile and hyper-velocity gliding projectile units, inter alia.30
Additional, measures already implemented include deploying new
radar and missiles; developing more expeditionary capabilities
(most notably a postwar first: an Amphibious Rapid Deployment
Brigade); significantly bolstering intelligence, surveillance, and
reconnaissance; basing a new fleet of F-15s in Okinawa; and
Japan’s Defense Reforms under Abe 487

creating a Ground Component Command tasked with controlling

ground forces across Japan and bolstering their flexibility to deploy
rapidly in various contingencies.31
More generally, recent and noteworthy developments in the latest
(2018) NDPG include a new emphasis on “cross-domain operations”
(ryoki odan sakusen) and associated calls for new “cyberspace units
and electromagnetic operation units” and a new “space domain mis-
sion unit,” inter alia. the NDPG also notes decisions to bolster ballistic
missile defense, including purchasing Aegis Ashore land-based systems,
to procure much longer-range “stand-off” missiles, and to research
hyper-velocity gliding projectile units.32 (In June 2020, Japan’s defense
minister announced the apparent cancellation of Aegis Ashore, but as
this chapter went to press there remains considerable confusion about
what will ultimately happen.33) Investments in the most advanced
platforms include a significantly increased order of F-35 joint strike
fighters that, if carried out, will make Japan the largest non-US buyer
and will see Japan acquire its first-ever short–take off and vertical
landing (STOVL) fighters (F-35Bs).34 Japan already stood up its first
F-35A squadron in March 2019.35
The apparent decision in 2018 to modify Izumo-class helicopter
destroyer decks to support STOVL fighters, coupled with the
announced F-35B purchase, could potentially lead to Japan fielding
its first (de facto) very light carriers since 1945. But the ships’ small size
(even fully loaded, Izumo-class are roughly one-quarter the size of US
Nimitz- or Ford-class aircraft carriers and can only conceivably
embark a handful of jets) and number (Japan has only two), coupled
with various aspects of existing plans and government statements,
suggest Japan is not on the verge of developing carrier strike groups.
F-35s may primarily operate from land (e.g., short runways on remote
islands) and not be permanently embarked on MSDF ships.
Nevertheless, depending on whether and how things play out in the
coming years, some of these capabilities (long-range strike missiles,
F-35Bs) could potentially constitute what some argue is a de facto
acquisition of “offensive” (kogekigata) platforms seemingly prohibited
under a decades-old official interpretation of Article 9. (Likely antici-
pating related concerns, in the section on introducing STOVL aircraft
the MTDP explicitly notes, “There will be no change in the existing
Government opinion concerning equipment that cannot be possessed
under the Constitution.”36) As with manifold effective and literal
488 Adam P. Liff

reinterpretations of Japan’s never-revised Constitution since 1947,

however, the government may end up presenting these effective policy
changes as consistent with the long-standing official interpretation
allowing Japan to possess capabilities “minimally necessary for self-
defense” based on a judgment that Japan’s rapidly changing threat
environment necessitates more flexible deterrence and response
options.37 In short, this is something to keep an eye on.

16.2.2.2 Pillar 2: Alliance with the United States

Ongoing efforts to strengthen Japan’s alliance with the United States
are a constant of the past several decades, yet this movement has
significantly accelerated since 2012. As a testament to the alliance
being a key priority of Abe-era recent defense reforms, Japan’s 2017
defense white paper devoted an entire chapter (fifty-one pages) to
“strengthening of the Japan–US alliance.”38 Together with the security
legislation, the Guidelines for Japan–US Defense Cooperation, which
provides a general outline of the scope of and respective responsibilities
for operational coordination between the allies and was also revised in
2015, is most significant.39 Beyond such bilateral pronouncements,
direct consultations between US and Japanese counterparts in design-
ing their respective defense strategies has also increased, and their
substance is closely aligned.40
Key recent alliance-related developments include new and/or newly
explicit commitments of support for each other in particular security
contingencies; a major expansion of bilateral (and multilateral) train-
ing, exercises, and planning across traditional and nontraditional scen-
arios (including the space and cyber domains) in peacetime, during
a gray-zone contingency, or in the event of an armed attack; and new
institutional links and dialogues. Examples of the first category include
the first-ever US presidential statement on the mutual security treaty’s
applicability to an armed attack against Japan over the contested
Senkakus (2014), Japan’s aforementioned new commitments (reflected
in the 2015 security legislation and 2015 Guidelines) to defend US
forces under an expanded set of circumstances, and a 2019 joint
statement affirming “that a cyber attack could, in certain circum-
stances, constitute an armed attack for the purposes of Article V of
the US–Japan Security Treaty.”41 Examples of the second category
include expanded bilateral training and exercises, both in number
and scope. For example, bilateral exercises more than tripled between
Japan’s Defense Reforms under Abe 489

2015 and 2017.42 New focus has been placed on collaboration in cross-
domain operations (including cyber, space, and the electromagnetic
spectrum) and new operational contexts (e.g., bilateral anti-ship com-
bat drills).43 Meanwhile, in 2017 Japan conducted its first-ever peace-
time maritime and air escorts of US military platforms, based on new
authorities under the 2015 security legislation. Finally, examples of the
third category include the upgraded Bilateral Planning Mechanism and
the new standing Alliance Coordination Mechanism, which facilitates
frequent, real-time communication among alliance managers and,
importantly, is “always on,” and plans for the first-ever inter-
governmental dialogue on economic security.44 All these developments
are backed up by several successive US administrations’ commitments
to the Asia/Indo-Pacific region as a priority theatre. This commitment is
reflected in official rhetoric and the deployment and sale of America’s
most advanced military assets to Japan (e.g., F-35s, Aegis Ashore).

16.2.2.3 Pillar 3: Deepening Defense Links with Third Parties

A third pillar of the Abe government’s national security strategy has
been expanded defense links to countries beyond the United States,
especially close US regional allies such as Australia and the Philippines,
regional partners such as India and Vietnam, and even US NATO
allies.45 Japans’ efforts to diversify security ties are largely done
together with and/or in consultation with Washington. Indeed, the
2015 Guidelines emphasize “cooperation with regional and other
partners, as well as international organizations,” and “the global
nature of the US–Japan alliance.”46 For example, during a May 2019
summit, “President Trump and Prime Minister Abe highlighted the
need for an increasingly networked structure of alliances and partner-
ships, anchored by the US–Japan Alliance, to counter challenges to the
United States’ and Japan’s shared vision of a free and open Indo-Pacific
region” and “noted that the US–Japan Alliance is both a model and
a platform for binding those allies and partners closer together to
uphold the rules-based order in the region.”47
Japan’s security cooperation with third parties can be divided into
two general categories: (1) multilateral initiatives centered on the US–
Japan alliance; and (2) initiatives with no formal US involvement. For
a nonexhaustive list of examples of the former, the 2019 US–Japan
joint statement identifies the US–Japan Australia Trilateral Strategic
Dialogue, trilateral summits and joint exercises with India, and greater
490 Adam P. Liff

cooperation with the United Kingdom and France.48 Regarding the

latter, Japan’s bilateral security cooperation with Australia is arguably
most robust and has received significant attention.49 Yet recent years
have also seen a major proliferation of new bilateral and multilateral
initiatives with additional partners. In 2016 Japan tabled its first-ever
(2016) proposal for an ASEAN-wide defense framework, which was
updated in 2019.50 The JSDF now trains, or has concrete plans to train,
bilaterally with its counterparts in the UK, Canada, Australia, and
India.51 Most famously, perhaps, the Abe government has revived his
2007 call for countries to cooperate in ensuring a “free and open Indo-
Pacific” – an initiative inspired at least in part by concerns about China,
and to which the United States later signed on.52
Finally, the Abe government’s further loosening of a 1970s-era ban
on arms exports has made possible more advanced defense technology
cooperation with, and exports to, third parties – most likely US allies
and partners. The government has highlighted defense technology
cooperation as a way not only to improve Japan’s defense but also to
assist its defense industry and strengthen ties with other countries.53
Though it has not finalized any large-scale development contracts yet,
since 2013 the Abe government has signed new defense equipment and
technology transfer agreements with the United Kingdom, Australia,
France, India, Germany, the Philippines, Italy, and Malaysia.54

16.3 Institutional and Policy Continuities

Although institutional and policy shifts since 2012 are significant in
any practical sense, to appropriately assess the big-picture significance
of “the Abe era” it is equally important to recognize that (1) the Abe
government has (at least so far) failed to achieve major goals coveted by
Abe and generations of other LDP leaders (e.g., constitutional revision;
truly transformative defense spending increases); (2) key foundational
pillars of Japan’s postwar defense policy remain in place; (3) important
“successes” of the past seven years have in key instances ended up
significantly watered down from what was originally intended; and/
or (4) they generally build on a reform agenda that predates Abe’s
return to power. In other words, a balanced understanding of Japan’s
defense reforms under Abe – as well as the prospects for future institu-
tional or policy shifts after Abe’s tenure – necessitates an examination
not only of what has changed but also what has not. Failure to examine
Japan’s Defense Reforms under Abe 491

the latter risks exaggerating the pace and scale of reforms achieved
heretofore. And specific to this volume’s unifying theme, it risks signifi-
cantly overstating the extent to which the Abe era constitutes a “major
turning point” in the history of postwar Japan.

16.3.1 Institutional Continuities

The major theme of institutional reforms to national security–relevant
institutions under Abe has been the consolidation of policy decision-
making in the cabinet and the Prime Minister’s Office in particular. Yet,
it is important to stress that centralization of decision-making in the
executive branch is a decades-old trend that not only significantly
predates Abe but is also not an ambition unique to the LDP.55
Specific to the institutions of greatest interest to this study, it is worth
noting that Japan’s NSC was the culmination of a reform effort dating
back two generations (to the 1970s) and which had accelerated signifi-
cantly in the wake of major crises after 2001 (e.g., 9/11; 3/11).
Additionally, similar institutional reforms had been pursued – to vari-
ous degrees – by Abe’s LDP and DPJ predecessors. And in 2013, Abe,
the ruling coalition, and the DPJ joined forces to establish the NSC.56 In
fact, the bills to establish both the NSC and the Bureau of Personnel
Affairs received significant support from the DPJ.57
Abe’s intervention in key bureaucratic appointments after 2012
(e.g., directly appointing a CLB director-general) suggests
a significant, and very public, break from past practice. But even in
this (arguably) most extreme case of political intervention, it is worth
noting that the sentiments that led to his decision are hardly unique to
Abe. In fact, given the unchanging constant that is Japan’s never-
revised Article 9, behind some of the most radical defense policy shifts
of the postwar period (e.g., the JSDF’s establishment in 1954;
enabling arms exports to the United States in the 1980s) was
a forward-leaning, conservative prime minister pressuring the CLB
to adjust its effective interpretation. More recently, three former
DPJ presidents are on record criticizing what they saw as overreach
by the CLB during earlier stages of their careers. For example, while
running for the DPJ party presidency in 2002 Abe’s immediate prede-
cessor as prime minister (Yoshiko Noda, 2011–2012) reportedly not
only called for collective self-defense to be judged constitutional but
also stated he would appoint a sympathetic CLB director-general to
492 Adam P. Liff

get it done.58 It is also worth noting that most of Abe’s national

security-relevant appointments have been rather conventional,
including his national security advisors and cabinet ministers – who
are generally considered to be more moderate than he is when it comes
to national security affairs (e.g., Taro Kono, who has served as both
foreign and defense minister).

16.3.2 Policy Continuities

As noted, there is no question that the Abe government has achieved some
practically significant defense policy reforms. Yet when, in pursuit of
a more complete picture, one examines the empirical record since 2012
for evidence of continuity in Japan’s defense policy, the persistence of core
pillars is striking. Continuity is even more remarkable when one factors in
the objectively measurable transformation (read: worsening) of Japan’s
security environment that has occurred in the interim – a fact about which
there is overwhelming consensus among elites and the general public.
Four major pillars – fundamental change to any one of which would
potentially indicate a more radical shift in Japan’s defense posture – are of
particular note: Article 9; defense spending under 1 percent of GDP; the
USA–Japan alliance at the center of Japan’s national security strategy; and
Japan’s eschewal of an indigenous nuclear deterrent (despite being sur-
rounded by nuclear-armed China, Russia, and North Korea).
First, despite Abe’s numerous assertions since 2012 that constitutional
revision is a top priority of his government, his various revision efforts the
past seven years have failed. As of this writing, Article 9 remains
untouched. Even if the 2017 amendment proposal Abe put forward as
prime minister, which would simply add a new clause asserting that the
JSDF’s “existence” is constitutional, ultimately passes the Diet and a public
referendum – hardly a foregone conclusion – it is not clear what the
practical significance for Japan’s defense policy would be. Importantly,
this proposal represents a significant “walk back” from far more ambitious
revision proposals put forward by generations of LDP politicians, includ-
ing an official party-wide proposal in 2012, all of which had called for
revision to Article 9’s existing two clauses.59 Furthermore, both the gov-
ernment and the vast majority of the public already agree the JSDF is
constitutional. Even so, public opinion remains, at best, ambivalent
about even adding this new third clause.60 Regardless, barring a more
Japan’s Defense Reforms under Abe 493

ambitious revision of Article 9’s first and/or second clauses, truly radical
changes to Japan’s defense policy seem unlikely.
Indeed, in various scenarios ranging from “collective self-defense” to the
use of small arms in UN peacekeeping operations, despite post-2012
reforms major conditions unique to Japan on the use of force/weapons
in a combat zone and/or outside an unambiguous armed attack on Japan
remain significant. For example, even a deployment to support a UN PKO
in South Sudan in which JSDF personnel were, under strict conditions,
newly authorized to use weapons proved controversial. Furthermore, after
the violence in the area worsened, the Abe government ended the mission
without the JSDF ever using those authorities. Anticipating controversy if
JSDF personnel were actually placed in a situation where they needed to
use their weapons, to reassure the public Abe had gone so far as to pledge
to resign if any JSDF personnel were killed.61 To date, no JSDF personnel
member has ever died in combat. An “exclusively defense-oriented policy”
remains central to Japan’s “basic defense policy.”62
Second, though the official ban on spending more than 1 percent of
GDP on defense was formally lifted in 1987, a de facto 1 percent ceiling
is still basically in effect. In any practical sense, without more practically
significant increases to Japan’s defense budget, a fundamental trans-
formation of JSDF capabilities and/or its effective mission set – especially
any significant power projection – that some allege is already underway
seems unlikely. To be sure, Abe has pushed through budget increases
since 2012 that are meaningful and which global media annually hype as
“record-breaking” (which is technically true). However, these increases
are small. Spending remains below 1 percent of GDP. In fact, despite
recent “record-breaking” budget increases, in nominal yen terms Japan’s
2019 defense budget marked only a 4 percent increase relative to its
1997 budget (but an 8 percent increase over 2012).63 As an indication of
its cautious approach, the Abe government effectively ignored calls from
within Abe’s own LDP for the government to consider a truly trans-
formative budget increase (a ~100 percent increase to NATO members’
current pledge level of 2 percent of GDP) in the 2018 MTDP.64 In the
end, the actual MTDP announced plans for average defense budgets
between FY 2019 and FY 2023 of roughly 5.1 trillion yen per year (for
comparison, Japan’s 2019 defense budget is 5.01 trillion yen) – hardly
a radical departure from past spending levels.65
A third pillar of Japan’s postwar defense policy is the centrality of the
US–Japan alliance and the basic “division of labor” within it. Both
494 Adam P. Liff

remain in place and, despite some practically significant adjustments

that enable Japan’s prime minister to significantly expand JSDF sup-
port for its ally (e.g., limited collective self-defense; asset protection,
etc.), the core bargain has not yet fundamentally changed. Indeed, the
2015 Guidelines reassert that “an armed attack against either Party in
the territories under the administration of Japan” remains the security
treaty’s primary focus.66 Relatedly Japan continues to rely on the US
nuclear umbrella, a position that has not changed under Abe and that is
reflected rhetorically in the Abe government’s reassertion of Japan’s
long-standing “three non-nuclear principles” as the country’s “funda-
mental policy.”67 That all said, discussion of Japan acquiring more
robust “strike” capabilities is increasingly mainstream, and the appar-
ent cancellation of Aegis Ashore in June 2020 has prompted a surge in
discussion of this possibility. If some of the proposals under discussion
turn into concrete procurement decisions and policy changes,
a significant shift could be in store for the alliance. Nevertheless, as of
summer 2020 it is not clear what Japan will actually do.68 This is
another important space to watch.
In sum, despite recent reforms under Abe, it is also important to
recognize that several fundamental pillars of Japan’s defense policy
remain firmly in place. Looking forward, there is no clear evidence
that the public – to say nothing of Komeito, the LDP’s junior coalition
partner (see Section 16.4.2) – would sanction truly radical changes to
or elimination of these pillars, such as eliminating or rewriting Article 9
to enable a clear break from “exclusive defense” (senshu boei), pursu-
ing autonomous defense capabilities outside an alliance framework,
doubling defense spending, or acquiring an indigenous nuclear deter-
rent. The future is of course unknowable. The point to emphasize here
is only that such transformative changes to these pillars have not
happened, despite Abe and his allies having a strong grip on the
Kantei and clear majorities in the Diet for over seven years. This reality
is often overlooked as the discourse focuses disproportionately on
rhetoric and ambition over actual policy, and change over continuity.

16.4 Key Factors Shaping the Abe Government’s Defense

Reform Agenda
The analysis heretofore has argued that (1) the Abe government has
achieved major reforms to institutions and defense policy since
Japan’s Defense Reforms under Abe 495

December 2012 but also that (2) any discussion of whether it constitutes
a “major turning point” in Japan’s postwar defense policy must also
acknowledge what has not changed, as well as baseline the current reform
trajectory in trends already observable before Abe’s return to power –
both outside and within Japan. Whereas the previous two sections focused
primarily on measuring outcomes, this section examines the international
and domestic political context that has shaped them. First, it highlights
key external drivers of defense reforms and several reasons why the Abe
government has been able to pursue a more ambitious security agenda
than its predecessors’ without confronting a major sustained domestic
political backlash. Second, it identifies several factors that have caused the
Abe government to either abandon or significantly moderate key coveted
reforms. These variables are likely to continue to shape Japan’s defense
reform trajectory for the foreseeable future, including in a post-Abe era.

16.4.1 External Factors

Seen from Tokyo, the past decade has witnessed a profound transform-
ation of Japan’s regional security environment. Changing threat per-
ceptions manifest in elite discourse and public opinion, as well as
authoritative national security documents. For example, Japan’s 2018
NDPG notes about the “dramatically changing security environment”:
“At present, [the] security environment surrounding Japan is changing at
extremely high speeds. Changes in the balance of power in the international
arena are accelerating and becoming more complex, and uncertainty over the
existing order is increasing. In addition, rapid expansion in the use of new
domains, which are space, cyberspace and electromagnetic spectrum[, are]
poised to fundamentally change the existing paradigm of national security,
which has prioritized responses in traditional, physical domains, which are
land, sea and air.”69

Variables of particular relevance since 2012 have been China’s rapidly

expanding military capabilities and assertive and coercive policies vis-
à-vis its neighbors – including Japan – in support of its vast and
controversial sovereignty claims in the South and East China Seas;
North Korea’s rapidly advancing nuclear and missile programs; the
rapid emergence of transformative military technologies and qualita-
tively new security challenges, including in the “gray zone” (the
ambiguous space between peacetime and war); and alliance politics.70
496 Adam P. Liff

Over the past decade, concerns in Tokyo about a rapidly changing

balance of power and specific security challenges posed by China have
gone mainstream. Most conspicuously, in the time between Abe’s first
administration (2006–2007) and today, Beijing’s official defense
budget – widely considered to underreport actual military spending –
has increased from being roughly commensurate with Japan’s to nearly
four times as large (in 2019, $178 billion vs. $48 billion).71 Even setting
aside long-standing concerns (e.g., Beijing’s sizable nuclear arsenal),
recent years have witnessed the People’s Liberation Army’s rapid mod-
ernization across the board. Its Rocket Force now fields the world’s
largest arsenal of conventionally tipped ballistic missiles – many of
which range Japan’s territory and US military bases throughout the
greater region. China also now possesses the world’s largest navy in
terms of the number of ships, and its growth and modernization are
extremely fast. For example, China launched more submarines, war-
ships, amphibious vessels, and auxiliaries between 2014 and 2018 than
are currently serving in India’s or Britain’s entire navy.72 Furthermore,
and especially since 2012, not only does Japan’s government and public
generally see China’s military and paramilitary operations to assert its
claim to the Senkakus as a concrete threat to Japan’s territorial sover-
eignty and challenge to the Japanese central government’s administra-
tion of the islands, but China’s reliance primarily on paramilitary Coast
Guard – rather than military – forces in the East China Sea presents
a novel and complex deterrence challenge (and potential fait accompli
scenario) in the “gray zone.”73 In short, the nature and severity of
Japan’s perceived security challenge vis-à-vis China – a country with
a population ten times and an economy already three times the size of
Japan’s – has transforming in highly visible ways the past decade.
As it concerns North Korea, seen from Tokyo the combination of the
despotic Kim dynasty and its rapidly advancing nuclear and missile
capabilities present a clear and present danger to Japan’s security.74
Over the past decade, North Korea has conducted four nuclear tests,
and since the Abe government’s formation in December 2012 it has
rapidly accelerated its missile testing – six tests in 2013, nineteen in
2014, fifteen in 2015, twenty-four in 2016, twenty-one in 2017, twenty-
two in 2019, and nine so far (as of August) in 2020.75 Meanwhile, these
missiles are becoming more accurate, longer-ranged, and both faster and
easier to hide (and thus more difficult to destroy). In 2017, North Korea
launched missiles over Japan’s territory and into its exclusive economic
Japan’s Defense Reforms under Abe 497

zone, tested an alleged thermonuclear weapon, and test-fired two inter-

continental ballistic missiles it credibly claimed could range Washington,
DC. In January 2018, Abe summarized his take-away: “the security
environment surrounding Japan is its most severe since World War
II.”76 (Some former Trump administration officials agree.77)
Meanwhile, beyond China’s rapidly advancing military capabilities
and North Korea’s expanding arsenal of advanced missiles and nuclear
weapons, rapid changes in military technology represent a third, more
general driver of Japan’s defense reforms, including institutionally.
Changing technologies shape judgments about the need for rapid crisis
response capabilities, interagency cooperation, and centralized deci-
sion-making, inter alia. They also clarify the exigency of developing
new defense capabilities in emerging domains, a point captured in
Japan’s 2018 NDPG and its call for Japan to develop a “multi-
domain defense force” (tajigen sogo boeiryoku).78 Of particular con-
cern are the proliferation of extremely fast Chinese and North Korean
ballistic and cruise missiles, emerging hypersonic technologies, and the
growing prominence of warfare in space, cyberspace, and the electro-
magnetic spectrum.
Finally, Japan’s ally the United States has also been an important
driver of defense reforms, both directly and indirectly. The volume of
US calls for Japan to adopt a more proactive security posture and to
expand its contributions within the alliance and to regional and
global security has increased significantly over the past two decades,
and especially as the regional security environment has worsened
since 2012. On the other hand, the rapidly changing balance of
power and nature of Japan’s threat environment has deepened inse-
curity about the US ability and/or willingness to come to Japan’s
defense in a major national security crisis. (In both cases, the basic
concerns are by no means new but have arguably been exacerbated by
Trump administration rhetoric and policies toward allies.)
Nevertheless, and despite occasional claims that these insecurities
are causing Japan to rethink its strategic alignment toward
Washington relative to Beijing, neither official rhetoric or policy nor
public opinion suggest widespread ambivalence concerning the fun-
damental importance and centrality of Japan’s alliance with the
United States to Japan’s national security.79
498 Adam P. Liff

16.4.2 Internal Factors

While Japan’s changing external security environment has been
a major driver of the Abe government’s defense reforms, it is not
a sufficient condition. Furthermore, any “ideological” factor – such
as the alleged “nationalist” agenda so often attributed, often uncritic-
ally, to Abe and the LDP in public discourse – cannot by itself come
close to explaining changes the past seven years. After all, both Abe’s
and the LDP’s transparent desire for Japan to have a more robust
national defense capability are well documented and decades-old con-
stants in Japan’s political life (e.g., even the LDP’s 1955 founding
charter called for constitutional revision and collective security80). So
what else has changed to allow the current Abe government to push
reforms significantly further than its LDP predecessors – including
during Abe’s first stint as prime minister (2006–2007)? Beyond the
external factors identified in the previous section, several domestic
political developments, coupled with Abe and his allies’ pragmatic
willingness to moderate their policy ambitions in the face of political
pushback, have played a major role. Though none of these factors are
inherently unique to the current Abe-era period, in key instances their
relevance has become more pronounced since 2012.
Two enabling domestic political factors in particular have created
a far more permissive environment for Abe’s ambitious reform
agenda than that faced by his predecessors. First, public aversion to
military affairs in nominally “pacifist” Japan has moderated signifi-
cantly in response to generational change and deepening recognition
of Japan’s changing security environment. In a manner that implies at
least basic overlap with the Abe government, the public generally
views North Korea and China as threats to Japan’s “vital interests”;
feels both little affinity and a sense of threat vis-à-vis Beijing (in stark
contrast to widespread affinity toward the United States); and identi-
fies a robust JSDF and a close alliance with Washington as essential
for Japan’s national security.81 Accordingly, moderate defense
reforms in pursuit of those basic goals have not generated sustained
domestic political backlash. For example, the most controversial
defense reform carried out since Abe’s return to power was the
passage of the “peace and security” legislation in summer 2015.
Unsurprisingly, it caused massive protests, and the cabinet’s disap-
proval rating increased significantly. Yet what happened next is also
Japan’s Defense Reforms under Abe 499

quite telling. Even in this extreme case of security legislation that

provoked the largest anti-government protests since the 1960s the
popular backlash proved fleeting. Within only a few months, the Abe
cabinet’s support rating was back above water – where it remained
for many months after, later falling only in response to unrelated
political scandals.82 More importantly, whatever popular discontent
existed about the legislation or other recent defense reforms, it did
not translate into a major decline in support for the LDP or ruling
coalition at the next election. In the first lower house election after
the legislation passed – October 2017 – the ruling coalition lost a few
seats, but it still won a two-thirds super-majority.
A second major enabling factor for Abe’s defense reform agenda is the
replacement of the ideological, (literally) pacifist left that defined the left–
right political divide throughout much the Cold War with a moderate,
pragmatic center-left. This also reflects a longer-term trend.
Furthermore, many of the individuals who constitute the left-of-center
since 2012 have actual experience governing during the three-year DPJ-
era immediately preceding Abe’s return to power (2009–2012). This
matters because, though often differing on the policy details, elite sup-
port for bolstering Japan’s defense capabilities, strengthening the US–
Japan alliance, and expanding security ties with third parties basically
exists across most of the political spectrum. Accordingly, the volume of
elite opposition to key Abe-era defense reforms has moderated signifi-
cantly. In fact, in important instances Abe government reforms have
effectively built on reform initiatives pursued by the DPJ while it was in
power and/or which subsequently received support from the DPJ even
after it was back in the opposition. Examples of the former include the
major revision of the Guidelines for Japan–US Defense Cooperation, the
loosening of the 1976 “arms export ban,” and the shift in force posture
southwest and new emphasis on “gray zone” challenges in the East
China Sea. Each was initiated under the DPJ. The most important
example of the latter is the NSC’s establishment, which reflected in
significant part a supra-partisan reform movement aimed at expanding
politicians’ influence over Japan’s powerful bureaucracies and a desire to
bolster crisis management capabilities following a series of major
national crises (e.g., 3/11). The DPJ reportedly shared a draft proposal
with the new Abe government, helped create the 2013 NSC establish-
ment bill, and voted in support of it.83
500 Adam P. Liff

Because of Japan’s rapidly changing strategic environment,

a resulting shift in public opinion, generational change, and basic
agreement among political elites of various party stripes on the neces-
sity of at least some significant reforms, the Abe government has had
more domestic political room to maneuver than its predecessors.
Nevertheless, and though repeated national election victories and an
opposition in considerable disarray since the DPJ’s landslide defeat in
2012 may suggest there are few checks on Abe and his LDP, several less
conspicuous factors have also frustrated the Abe government’s ambi-
tions. In key instances, it has significantly watered down, or in some
cases effectively abandoned, more ambitious defense reform objectives.
On high-salience policy issues where there is clear opposition, or at
least ambivalence, to Abe/LDP positions – e.g., rewriting Article 9’s
existing text or rendering constitutional “use of force” (buryoku koshi)
in an armed attack scenario even if Japan itself does not face a clear
threat – Abe appears to have significantly dialed back his ambitions.
For example, had the actual 2014 Cabinet Decision “reinterpreting”
Article 9 or the 2015 security legislation that provided its legal founda-
tion contained language regarding collective self-defense or collective
security reflective of earlier LDP proposals, it is reasonable to expect
that the domestic political and popular backlash would have been
much more forceful. To avoid this, what the Abe government actually
produced was significantly watered down.84
On the one hand, the Abe government’s pragmatic moderation suggests
a responsiveness to public opinion and ability to read the political winds.
But also likely at play is a recognition of the LDP’s remarkable vulnerabil-
ity and weak mandate, contrary to what its Diet seat totals and cabinet
support rates would suggest. Especially important are (1) the LDP’s poten-
tially fleeting support among the general public and (2) the fact that it rules
in coalition with a political party (Komeito) whose core supporters gener-
ally oppose major components of the LDP’s defense reform agenda.
One underappreciated paradox of post-2012 Japanese politics is that
the ruling coalition’s string of national election victories and the LDP’s
single-party majorities do not actually indicate widespread popularity,
much less a stable base of political support. This is especially true as it
concerns popular support for more domestically controversial elements of
Abe and the LDP’s national security agenda (e.g., Article 9 revision). Three
trends in particular suggest public support for the LDP after 2012 may be
soft. First, lower house elections during the current “Abe era” have
Japan’s Defense Reforms under Abe 501

attracted the lowest-ever voter turnout in history and mark declines of 10

(2012), 17 (2014), and 16 (2017) percentage points, respectively, relative
to the last “pre-Abe era” election (2009).85 Second, voters with no party
affiliation make up the majority of the electorate.86 Third, public opinion
surveys demonstrate that, among those who support the Abe cabinet,
a major reason is simply the lack of alternatives.87 These circumstances
suggest that if the government steps on a high-salience political landmine
(e.g., gutting Article 9) that mobilizes voters who have been staying home
recently and/or causes a critical mass of floating voters to support another
party, the LDP might find itself in the wilderness again.
Also powerfully influencing the Abe government’s moderation of its
defense reform agenda is the fact that (1) the LDP rules in coalition with
a political party whose primary support base is a lay-Buddhist movement
(Soka Gakkai); and (2) the LDP’s performance in national elections
depends significantly on close cooperation with Komeito (especially
mutual stand-down agreements in single-member districts).88 The
LDP’s relative dependence on the supporters of its junior coalition
partner gives Komeito leverage within the coalition disproportionate to
its number of Diet seats, especially on issues highly salient to its pacifistic
support base. Why does this matter for defense policy? For starters,
without Komeito electoral support, the LDP would probably not have
had a single-party majority when the Diet passed the controversial
security legislation in 2015. Moreover, as it concerns Article 9, LDP
concessions to Komeito in intra-coalition negotiations deboned both the
Abe cabinet’s 2014 “reinterpretation” to enable exercise of collective
self-defense and the Abe government’s official 2017 proposal for revising
it. The former resulted in the aforementioned three strict conditions
unique to Japan, while the latter caused Abe to propose a revision that
originally came from Komeito itself and which represents a fundamental
departure from the LDP’s long-standing position that Article 9’s two
existing clauses must be revised.89 Since 2012, as well as before,90
Komeito has repeatedly frustrated forward-leaning LDP prime minis-
ters’ ambitious defense reform agendas. Unless an unexpected political
realignment occurs, it will likely continue to do so.
In sum, external factors and Abe and his allies’ long-held ambitions –
whether ideological and/or pragmatic responses to a changing security
environment – are undoubtedly major drivers of Japan’s defense
reforms since 2012. Yet domestic political variables matter greatly in
shaping the actual outcomes these past seven years. On the one hand,
502 Adam P. Liff

shifting public opinion and elite preferences have reduced the vehe-
mence and sustainability of any domestic political backlash against
moderate – as opposed to radical – reforms linked directly to specific
perceived external threats. On the other hand, several domestic fac-
tors – especially LDP vulnerability in national elections and Komeito’s
influence – have caused Abe and his government to pragmatically, yet
significantly, dial back their ambitions.

16.5 Conclusion
At his inaugural press conference in December 2012, Abe stated his
determination “to defend fully people’s lives, our territory, and our beau-
tiful ocean” and asserted that Japan’s national security “is not someone
else’s problem.”91 Over the past seven-plus years Japan’s regional security
environment has arguably grown more foreboding. In the interim, Abe
and his government have pursued an ambitious and practically significant
defense reform agenda intended to enable Japan to more effectively deter
and, if necessary, defend against perceived security threats. Key accom-
plishments include: bolstering executive control over national security
decision-making; strengthening deterrence through a rationalization and
expansion of the JSDF’s roles, missions, and capabilities; boosting cooper-
ation with Japan’s US ally; and expanding defense links with like-minded
third parties within and beyond the Asia-Pacific.
Though much commentary during the “Abe era” has asserted that
Abe’s allegedly “nationalist” agenda is a driving force, or even that he
has taken Japan on a radical trajectory away from its postwar “paci-
fism,” a systematic analysis of both change and continuity over the past
seven years suggests that the defining feature of his government’s defense
reform accomplishments is a kind of evolutionary pragmatism.
Narrowly focusing on perceived (or imputed) policy shifts while over-
looking the persistence of several core pillars of Japan’s remarkably self-
restrained defense posture, the extent to which the Abe government’s
achievements build on efforts initiated by his predecessors, or the rapid
changes to Japan’s security environment, easily leads to hyperbolic
claims about the pace and scale of change or the impact of idiosyncratic
factors such as ideology or Abe himself. As further evidence of this
evolutionary pragmatism, the empirical record reveals significant
instances of Abe moderating his reform ambitions in response to, or in
anticipation of, domestic political pushback – including on foundational
Japan’s Defense Reforms under Abe 503

LDP priorities dating back to its founding sixty-five years ago (e.g.,
Article 9; collective self-defense).
What comes next? Just as this chapter is going to press (August 29,
2020), Abe has abruptly announced his planned resignation over health
concerns.92 Regardless, barring major and unexpected external or domes-
tic political structural changes, this chapter’s analysis suggests that Japan’s
evolutionary defense reform trajectory is likely to continue even in a post-
Abe era. Reforms put in place since 2012 are unlikely to be reversed.
Rather, a further incremental loosening of long-standing constraints
seems likely, especially given inauspicious geopolitical trend lines.
Though Japan’s defense reforms the past decade have been practic-
ally significant and are likely to continue, another important takeaway
emerges from this chapter’s analysis. That despite nearly eight years in
power Abe’s government has failed to achieve more radical changes
long coveted by him and his party (e.g., elimination/revision of Article
9’s original clauses; truly transformative increases to defense spending;
rendering “full” collective self-defense constitutional) suggests that –
barring a major crisis in the region (or within the USA–Japan alliance) –
future reform-minded prime ministers will continue to face powerful
domestic political headwinds. After all, Abe’s ambitions were no secret
and, superficially at least, throughout most of his tenure the stars
appeared aligned for a more significant transformation: the LDP and
its ruling coalition cruised to significant victories in every national
election (including single-party majorities for the LDP for several
years); cabinets were exceptionally stable and enjoyed relatively robust
public support; and public opinion showed widespread recognition of
Japan’s worsening security environment. Nevertheless, as Abe noted
mournfully during the August 2020 press conference at which he
announced his planned resignation, popular resistance to perhaps his
most coveted objective – revision of Japan’s never-revised constitution –
proved insurmountable.93

Notes
1. Shusho Kantei, Abe naikaku sori daijin shunin kisha kaiken (Prime
Minister Abe’s inaugural press conference), December 26, 2012. www
.kantei.go.jp/jp/96_abe/statement/2012/1226kaiken.html.
2. Japan Ministry of Defense, Defense of Japan 2019, www.mod.go.jp/e/
publ/w_paper/2019.html, 210, 212.
504 Adam P. Liff

3. Japan Ministry of Defense, Defense of Japan 2019, 210, 212.

4. Article 9 states that Japan “forever renounce[s] war as a sovereign right
of the nation and the threat or use of force as means of settling
international disputes” and pledges that “land, sea, and air forces, as
well as other war potential, will never be maintained.” The Constitution
of Japan, November 3, 1946, https://japan.kantei.go.jp/constitutio
n_and_government_of_japan/constitution_e.html.
5. Shusho Kantei, Abe naikaku sori daijin shunin kisha kaiken; Shusho
Kantei, Kokka anzen hosho kaigi no sosetsu ni kansuru yushikisha kaigi
(Meeting of Experts Concerning NSC Establishment), February 15,
2013, www.kantei.go.jp/jp/96_abe/actions/201302/15ka_yusiki.html.
6. Aurelia George Mulgan, The Abe Administration and the Rise of the
Prime Ministerial Executive (New York: Routledge, 2017), ch. 3.
7. For a focused English-language analysis of the NSC, NSS, and
associated issues, see Adam P. Liff, “Japan’s National Security
Council: Policy Coordination and Political Power,” Japanese Studies
38(2) (2018): 253–279. Key Japanese-language studies include
Yasuhiro Matsuda, ed., NSC kokka anzen hoshō kaigi
(NSC National Security Council) (Tokyo: Sairyūsha, 2009); Tsuyoshi
Sunohara, Nihonban NSC to wa nani ka? (What is the Japan-style
NSC?) (Shinjuku: Shinchōsha, 2014); Ken Kotani, “Nihon-ban Kokka
Anzen Hosho Kaigi (NSC) no kinoteki tokucho” (Japan-style National
Security Council (NSC) and its Functional Features), Kokusai Anzen
Hosho, March 2015, 61–75; Ryoichi Oriki and Kaneko Masafumi,
Kokka anzen hoshō kaigi: Hyōka to teigen (National Security Council:
Evaluation and Recommendations), (Tokyo: PHP Kenkyūjō, 2015).
8. Shusho Kantei, Kokka anzen hosho kaigi no sosetsu ni kansuru
yushikisha kaigi (Meeting of Experts Concerning NSC Establishment),
February 15, 2013, www.kantei.go.jp/jp/96_abe/actions/201302/15k
a_yusiki.html.
9. Liff, “Japan’s National Security Council,” 265–270.
10. Kotani, “Japan-style National Security Council (NSC),” 61, 70–72.
11. “Japan aims to bridge security, economic realms with new NSS
division,” The Japan News, May 19, 2020.
12. Liff, “Japan’s National Security Council.”
13. For a focused study on related issues, see Giulio Pugliese, “Kantei
Diplomacy? Japan’s Hybrid Leadership in Foreign and Security
Policy,” Pacific Review 30(2) (2017): 152–168.
14. “Naikaku jinjikyoku, 5gatsu ni secchi” (Cabinet Personnel Bureau to be
established in May), Nikkei Shimbun, April 11, 2014, www.nikkei.com
/article/DGXNASFS11002_R10C14A4MM0000/.
Japan’s Defense Reforms under Abe 505

15. “Kaijo Hoanchokan ni hatsu no genba shusshin” (First-ever JCG

commandant from the front lines),” Nikkei Shimbun, July 18, 2013;
“Abe’s Legal Aide on Defense Reform Steps down Due to Ill Health,”
Kyodo, May 16, 2014.
16. Yuji Sato, “The Japan Coast Guard protects the Senkaku Islands to the
last,” Discuss Japan 35, October 18, 2016, www.japanpolicyforum.jp
/archives/diplomacy/pt20161018235004.html.
17. Adam P. Liff, “Policy by Other Means: ‘Collective Self-Defense’ and the
Politics of Japan’s Postwar Constitutional (Re-)Interpretations,” Asia
Policy 24 (2017): 139–172.
18. For an official and extensive summary of the legislation, see Japan
Ministry of Defense, Defense of Japan 2019, www.mod.go.jp/e/publ/
w_paper/2019.html, 245–267.
19. Atsuhiko Fujishige, “New Japan Self-Defense Force Missions under the
‘Proactive Contribution to Peace’ Policy: Significance of the 2015
Legislation for Peace and Security,” Japan Chair Platform, Center for
International and Strategic Studies, July 21, 2016, www.csis.org/ana
lysis/new-japan-self-defense-force-missions-under-%E2%80%9Cproa
ctive-contribution-peace%E2%80%9D-policy.
20. The following breakdown is adapted from Satoru Mori, “The New
Security Legislation and Japanese Public Reaction,” Tokyo
Foundation, December 2, 2015, www.tokyofoundation.org/en/art
icles/2015/security-legislation-and-public-reaction.
21. Japan Ministry of Defense, “Outline of the Legislation for Peace and
Security,” Defense of Japan 2017, www.mod.go.jp/e/publ/w_paper/pd
f/2017/DOJ2017_2–3-2_web.pdf. The practical significance of this
reinterpretation is contested, and even Abe’s own statements at times
appear contradictory. For a sample of debates in English, see Michael
Green and Jeffrey W. Hornung, “Ten Myths About Japan’s Collective
Self-Defense Change,” Diplomat, July 10, 2014, https://thediplomat
.com/2014/07/ten-myths-about-japans-collective-self-defense-change/;
Christopher W. Hughes, “Japan’s Strategic Trajectory and Collective
Self-Defense: Essential Continuity or Radical Shift?” Journal of
Japanese Studies 43(1) (2017): 93–126; Liff, “Policy by Other Means.”
22. “In new role, MSDF patrolling waters around Koreas to foil oil
smuggling,” Japan Times, January 13, 2018, www.japantimes.co.jp/n
ews/2018/01/13/national/previously-undisclosed-role-self-defense-
force-ships-watching-north-korea-smuggling-sea/.
23. Mori, “The New Security Legislation and Japanese Public Reaction.”
24. “Analysis: Low-risk mission aimed at inuring public to SDF’s new role,”
Asahi Shimbun, May 2, 2017, https://article.wn.com/view-lemonde/20
17/05/02/ANALYSIS_Lowrisk_mission_aimed_at_inuring_public_to_
506 Adam P. Liff

SDF_s_ne/#/related_news; “Japan–US joint operations increase amid

regional uncertainty,” Yomiuri Shimbun, March 30, 2018, www
.standard.net/World/2018/03/30/Japan-US-joint-defense-operations-
increase-amid-regional-uncertainty.
25. Mori, “The New Security Legislation and Japanese Public Reaction.”
26. Naikaku Kanbo, Kokka anzen hoshō senryaku ni tsuite (About the
National Security Strategy), December 17, 2013, www.cas.go.jp/jp/sir
you/131217anzenhoshou/nss-j.pdf. Provisional translations of white
papers and the NDPGs are available on the Ministry of Defense’s
website: see www.mod.go.jp/e/publ/w_paper/ and www.mod.go.jp/e/d
_act/d_policy/national.html.
27. Japan Ministry of Defense, National Defense Program Guidelines for
FY 2011 and beyond, December 17, 2010, www.mod.go.jp/e/d_act/d_
policy/pdf/guidelinesFY2011.pdf.
28. Japan Ministry of Defense, National Defense Program Guidelines for
FY 2014 and beyond, December 17, 2013, www.mod.go.jp/j/approach/
agenda/guideline/2014/pdf/20131217_e2.pdf.
29. Adam P. Liff, “China’s Maritime Gray Zone Operations in the East
China Sea and Japan’s Response,” in China’s Maritime Gray Zone
Operations, ed. Ryan D. Martinson and Andrew S. Erickson
(Annapolis, MD: Naval Institute Press, 2019): 207–231.
30. Japan Ministry of Defense, National Defense Program Guidelines for
FY 2019 and beyond, December 18, 2018, www.mod.go.jp/j/approach/
agenda/guideline/2019/pdf/20181218_e.pdf.
31. “GSDF to undergo biggest realignment since founding,” Yomiuri
Shimbun, March 23, 2018, www.standard.net/World/2018/03/23/Japan-
s-ground-defense-force-to-undergo-biggest-realignment-since-founding.
32. Japan Ministry of Defense, Medium Term Defense Program (FY 2019–
FY 2023), December 18, 2018, www.mod.go.jp/j/approach/agenda/gu
ideline/2019/pdf/chuki_seibi31-35_e.pdf.
33. “Exclusive: Japan may still build Aegis Ashore despite reports of
cancellation, source says,” Reuters, July 10, 2020.
34. “With massive F-35 increase, Japan is now biggest international buyer,”
Defense News, December 18, 2018, www.defensenews.com/global/asi
a-pacific/2018/12/18/with-massive-f-35-increase-japan-is-now-biggest-
international-buyer/.
35. Franz-Stefan Gady, “Australia, Japan to Hold First Joint Air Combat
Exercise,” September 11, 2019, https://thediplomat.com/2019/09/aus
tralia-japan-to-hold-first-joint-air-combat-exercise/.
36. Japan Ministry of Defense, Medium Term Defense Program (FY 2019 –
FY 2023), 11.
37. Liff, “Policy by Other Means.”
Japan’s Defense Reforms under Abe 507

38. Japan Ministry of Defense, Defense of Japan 2017, www.mod.go.jp/e/

publ/w_paper/2017.html, 261–311.
39. Japan Ministry of Defense, Guidelines for Japan–US Defense
Cooperation, April 27, 2015, www.mod.go.jp/e/d_act/anpo/pdf/shish
in_20150427e.pdf.
40. Adm. Dennis Blair, “Chairman’s Message: the State of the US–Japan
Alliance,” Sasakawa USA, May 7, 2019. https://spfusa.org/chair
mans-message/chairmans-statement-the-state-of-the-u-s-japan-
alliance/.
41. Japan Ministry of Foreign Affairs, “Joint Statement of the Security
Consultative Committee,” April 19, 2019, www.mofa.go.jp/files/0004
70738.pdf.
42. “Japan–US joint operations increase amid regional uncertainty,”
Yomiuri Shimbun, March 30, 2018, www.standard.net/World/2018/0
3/30/Japan-US-joint-defense-operations-increase-amid-regional-
uncertainty.
43. Japan Ministry of Foreign Affairs, “Joint Statement of the Security
Consultative Committee”; “Japan and US to conduct first joint anti-
ship drills,” Jiji, June 16, 2018.
44. Jeffrey W. Hornung, Managing the US–Japan Alliance: An
Examination of Structural Linkages in the Security Relationship
(Washington, DC: SPF USA, 2017), https://spfusa.org/wp-content/up
loads/2017/04/Managing-the-U.S.-Japan-Alliance.pdf; “Japan, U.S. to
establish new dialogue on economic security,” The Japan News,
May 16, 2020.
45. For a list of several of Japan’s recent defense agreements beyond the
United States, see Reference 37, “Situations Concerning the Conclusion
of Agreements,” in Japan Ministry of Defense, Defense of Japan 2019,
515.
46. Japan Ministry of Defense, Guidelines for Japan–US Defense
Cooperation.
47. US Department of State, “Fact Sheet,” May 27, 2019.
48. Japan Ministry of Foreign Affairs, “Joint Statement of the Security
Consultative Committee.”
49. Thomas S. Wilkins, “After a decade of strategic partnership: Japan and
Australia ‘decentering’ from the US alliance?” Pacific Review 31(4)
(2018): 498–514.
50. Japan Ministry of Defense, “Vientiane Vision: Japan’s Defense
Cooperation Initiative with ASEAN,” November 2016, www.mod.go.
jp/e/d_act/exc/vientianevision/; Japan Ministry of Defense, “Updating
the ‘Vientiane Vision,’” November 2019, www.mod.go.jp/e/d_act/exc/
admm/06/vv2_en.pdf.
508 Adam P. Liff

51. “Behind the Scenes/SDF trains with countries other than US,” Yomiuri
Shimbun, October 11, 2018.
52. Japan Ministry of Foreign Affairs, “Toward Free and Open Indo-
Pacific,” November 2019, www.mofa.go.jp/files/000407643.pdf.
53. Taisuke Hirose, “Japan’s New Arms Export Principles: Strengthening
U.S.-Japan Relations,” Japan Chair Platform (Washington, DC: Center
for International and Strategic Studies, October 14, 2014), https://csis-
prod.s3.amazonaws.com/s3fs-public/legacy_files/files/publication/141
014_Hirose_NewArmsExportPrinciples_JapanPlatform.pdf.
54. Japan Ministry of Defense, Defense of Japan 2019, 515.
55. For a recent English-language analysis incorporating a review of
associated literatures, see Mulgan, “The Abe Administration.”
56. Liff, “Japan’s National Security Council.”
57. “Cabinet Personnel Bureau to be established in May,” Nikkei Shimbun;
“Nihon-ban NSC Raishu Hassoku” (Japan-style NSC to Launch Next
Week), Yomiuri Shimbun, November 27, 2013.
58. Richard J. Samuels, “Politics, Security Policy, and Japan’s Cabinet
Legislation Bureau: Who Elected These Guys, Anyway?” Japan Policy
Research Institute, Working Paper, March 2004, www.jpri.org/publica
tions/workingpapers/wp99.html.
59. “Nippon koku kenpo kaisei soan,” Liberal Democratic Constitutional
Reform Promotion Headquarters, http://constitution.jimin.jp/docu
ment/draft/.
60. “Kenpo kaisei, sansei 51% . . . Jieitai ‘goken’ 76%” (51 percent agree
with constitutional revision . . . 76 percent believe JSDF constitutional),
Yomiuri Online, April 30, 2018, http://sp.yomiuri.co.jp/politics/20180
429-OYT1T50099.html.
61. “Last Japanese troops leave U.N. peacekeeping mission in South
Sudan,” Reuters, May 25, 2017; Jeffrey W. Hornung, “Japan’s
Mistaken South Sudan Withdrawal,” The Diplomat, June 7, 2017,
https://thediplomat.com/2017/06/japans-mistaken-south-sudan-with
drawal/.
62. Japan Ministry of Defense, Defense of Japan 2019, 25.
63. Japan Ministry of Defense, Defense of Japan 2019, 240.
64. “Boeihi ‘GDP 1% teppai’ Jimin teigen, jijitsujo no kuboka mo”
(Abolish GDP 1 percent defense budget, LDP proposal, de facto
aircraft carrier, too), Asahi Shimbun, May 26, 2018.
65. Japan Ministry of Defense, Medium Term Defense Program (FY 2019–
FY 2023), 34.
66. Japan Ministry of Foreign Affairs, “Japan–US Security Treaty,”
January 19, 1960, www.mofa.go.jp/region/n-america/us/q&a/ref/1
.html.
Japan’s Defense Reforms under Abe 509

67. Japan Ministry of Defense, Guidelines for Japan–US Defense

Cooperation.
68. Jeffrey W. Hornung, “Is Japan’s Interest in Strike Capabilities
a Good Idea,” WarOnTheRocks, July 17, 2020, https://waronther
ocks.com/2020/07/is-japans-interest-in-strike-capabilities-a-good-
idea/.
69. Japan Ministry of Defense, National Defense Program Guidelines for
FY 2019 and beyond.
70. For a concise summary of the official government perspective on Japan’s
security environment, see Japan Ministry of Defense, National Defense
Program Guidelines for FY 2019 and beyond, esp. 3–7.
71. “China’s Defense Spending Set to Rise 7.5%,” Bloomberg, March 4,
2019; “Japan’s government approves record defense spending,”
Reuters, December 20, 2018.
72. China Power Team, “How is China modernizing its navy?” China
Power, October 16, 2019, https://chinapower.csis.org/china-naval-
modernization/.
73. Liff, “China’s Maritime Gray Zone Operations.”
74. Shusho Kantei, “Abe Naikaku Sori Daijin nento kisha kaiken” (Prime
Minister Abe’s New Year Address), January 4, 2018, www.kantei.go.jp
/jp/98_abe/statement/2018/0104kaiken.html.
75. “The CNS North Korea Missile Test Database,” Nuclear Threat
Initiative, March 31, 2020, www.nti.org/analysis/articles/cns-north-
korea-missile-test-database/.
76. Shusho Kantei, “Abe Naikaku Sori Daijin nento kisha kaiken.”
77. Uri Friedman, “The World According to H. R. McMaster,” The
Atlantic, January 9, 2018, www.theatlantic.com/international/archive/
2018/01/hr-mcmaster-trump-north-korea/549341/.
78. Japan Ministry of Defense, Defense of Japan 2019, 211.
79. Adam P. Liff, “Unambivalent Alignment: Japan’s China Strategy, the
US Alliance, and the ‘Hedging’ Fallacy,” International Relations of the
Asia-Pacific 19(3) (2019): 453–491.
80. “To no seiko” (Party Principles), LDP, November 15, 1955.
81. Various poll data cited in Liff, “Unambivalent Alignment,” 478–480.
82. See aggregated poll data at “Japan Political Pulse,” Sasakawa USA,
https://spfusa.org/category/japan-political-pulse/.
83. Sunohara, Nihon-ban NSC, 124–130; “Nihon-ban NSC Raishu
Hassoku,” Yomiuri Shimbun.
84. Liff, “Policy by Other Means.”
85. Ministry of Internal Affairs and Communications, “Kokusei senkyo no
tohyoritsu no suii ni tsuite (Heisei 28nen 9gatsu)” (Changes in national
510 Adam P. Liff

election voter turnout), September 2016, www.soumu.go.jp/main_con

tent/000255919.pdf.
86. Aiji Tanaka, “Japan’s Independent Voters, Yesterday and Today,”
Nippon, August 16, 2012, www.nippon.com/en/in-depth/a01104/.
87. “NHK poll: Cabinet support rate at 46%,” NHK, January 9, 2017,
www3.nhk.or.jp/nhkworld/en/news/20180109_32/.
88. For detailed analysis, see Adam P. Liff and Ko Maeda, “Electoral
Incentives, Policy Compromise, and Coalition Durability: Japan’s
LDP–Komeito Government in a Mixed Electoral System,” Japanese
Journal of Political Science 20(1) (2019): 53–73.
89. Liff and Maeda, “Electoral Incentives, Policy Compromise, and
Coalition Durability,” 64–69.
90. J. Patrick Boyd and Richard J. Samuels, Nine Lives? The Politics of
Constitutional Reform in Japan (Washington, DC: East–West Center,
2005), esp. 27–61.
91. Shusho Kantei, Abe naikaku sori daijin shunin kisha kaiken (Prime
Minister Abe’s inaugural press conference), December 26, 2012, www
.kantei.go.jp/jp/96_abe/statement/2012/1226kaiken.html.
92. “Abe Shusho, Jinin Hyomei,” Asahi Shimbun, August 29, 2020.
93. Ibid.
17 The Enduring Challenges of History
Issues
mary m. mccarthy

17.1 Introduction
History issues have taken center stage in myriad ways during Shinzo
Abe’s tenure as prime minister. Although he or his administration was
not always the cause of these issues coming to the fore, the way in
which these issues were managed and addressed correspond to a major
twofold goal of Abe, both related to creating a legacy for himself and
his agenda. First is to reinstate a sense of pride in the Japanese nation
through historical memory. This is based on the sentiment that this was
lost through the way in which previous administrations chose to deal
with the legacies of twentieth-century history, particularly during the
1990s. Second is to move beyond narratives of Japan as perpetrator or
victimizer in the global narrative.
Perhaps not surprisingly, these two objectives have come into conflict
with each other, as they require control of collective memories both at
home and abroad, as well as the sending of distinct messages to different
target audiences. In today’s globalized society, governments no longer
have a monopoly on the creation of collective memory. Public memory
has become democratized and globalized, with domestic, transnational,
foreign, and international actors all contributing narratives and counter-
narratives that may or may not serve the interests of the state.
Despite this, the Abe administration has seen some successes with
regard to historical reconciliation. During his tenure, he hosted US
President Barack Obama in Hiroshima and visited Pearl Harbor. He
also dampened some of the controversy swirling around Yasukuni
Shrine. His most enduring challenge, however, has been the comfort
women issue. This is one that he has not only allowed to fester but has
enflamed. This is in part due to poor policy choices and an incomplete
understanding of the relevant issues, particularly those related to trans-
national civil society.

511
512 Mary M. McCarthy

In this chapter I will explore three cases: the comfort women,

Yasukuni Shrine, and the Pearl Harbor visit. I will assess the degree
to which Abe achieved his objectives in each of these areas, by first
establishing what his objectives were and then delving into the factors
that influenced both objectives and outcomes.
I argue that Abe is pushing a nationalist conservative agenda on
history issues but places practicality before sentiment. Thus, he will
officially uphold the Kono Statement and will send an offering in lieu of
a visit to Yasukuni Shrine on August 15. Still, he will do so while
making it publicly known what his true views and intentions are
through clear signals. Geopolitically he is savvy, but he may be over-
estimating the power of the state and underestimating the power of civil
society.

17.2 Abe’s Goals and Perspective

The Introduction to this volume discusses the Abe government since
2012 as one that seeks to make a full break with the 1955 system. The
phrases “transition,” “major turning point,” and “reinvigorate” are all
used to describe the Abe government’s approach, if not achievements.
Shedding the 1955 system is part of this transformation, but so is
breaking free of the lost decade of the 1990s and the L-shaped recession
that long plagued Japan – a decade and its aftermath that saw financial
woes but also witnessed uncertainty after natural disaster (the Great
Hanshin Earthquake) and man-made attack (Aum Shinrikyo terror-
ism), coupled with a complacency rooted in a new generation born in
relative prosperity and free of the struggles of its parents’ generation.
Abe ushered in his second tenure as prime minister with the bold
statement of “Japan is back.” His government has sought to move past
the postwar era (with the exception of what it identifies as the norma-
tive goods it produced, such as the US–Japan Alliance1) and what Abe
views as some of the missteps of his predecessors. He wants to continue
what he began during his first tenure as prime minister but to finish
with greater success at achieving his goals. We see this pattern across
a variety of issue areas, as this volume illuminates.
Abe’s objectives with regard to history issues are no different. His
perspective on these issues is driven by a belief that national confidence,
national strength, and national identity are rooted in a sense of pride
about one’s national history. A history in which a nation can feel pride
The Enduring Challenges of History Issues 513

is interpreted as one where the positives of one’s national history are

highlighted and the negatives are obscured. This should be instilled in
children at a young age through textbooks and other educational
materials. It should be promoted through the media. National pride
is also enhanced when one is not hampered by the past, such as through
a constitution that eliminates a state’s sovereign right to maintain
a military and wage war.
Shinzo Abe’s political career has always been embedded in the
nationalist conservative agenda that promotes such postures. He has
been an active member of the LDP Committee for Historical
Investigation (Jimintō rekishi kentō iinkai), the Diet Members’
League for the Fiftieth Anniversary of the End of the War (Shūsen
50nen kokkai giin renmei)2, and the Diet Members’ League to
Consider Japan’s Future and History Education (Nihon no zento to
rekishi kyōiku wo kangaeru giin no kai) (Hughes 2015; Penney 2012a).
All of these groups have advocated for the promotion of a historical
narrative such as described.
What the Introduction to this volume calls Abe’s “heavy-handed
approach toward the media, particularly liberal news outlets such as
Asahi” is not new. Although one goal is to manage his popularity and
the news cycle, another is to ensure a narrative that is friendly to the
nationalist conservative perspective. As deputy chief cabinet secretary
in 2001, Abe successfully pressured NHK to revise a documentary it
was planning to air on the comfort women (Hughes 2015; Penney
2012b).3 As prime minister over the last eight years, Abe’s government
has altered the way in which media outlets discuss history, avoiding
terminology or even entire topics that may be deemed inappropriate by
the current administration (Fackler 2016).
Through nationalist conservative groups and activities, throughout his
career Abe has vigorously sought to influence government policy as well
as textbook and media content, with regard to issues of history and
historical revisionism. Mochizuki and Porter (2013) attribute such polit-
ical goals as stemming from Abe’s personal relationship with his maternal
grandfather, Nobusuke Kishi, whose roles during the war led to his being
imprisoned as a suspected Class-A war criminal by the Allied forces but
who then went on to serve as prime minister 1957–1960 (p.35). Yet such
a perspective on history issues clearly is not unique to Abe and fits him
squarely within the nationalist conservative camp, of whom there are
many among the political elites of the Liberal Democratic Party (LDP).
514 Mary M. McCarthy

This cohort has been particularly active since the late 1990s, shortly
after Abe himself gained national office in 1993, the same year that the
1955 system officially “ended.” The 1990s were a catalyst for renewed
efforts by nationalist conservatives to mold the national and inter-
national narrative of Japan’s role in the war. The decade was a unique
period in Japan’s interaction with history issues, as a series of apologies
and reflections on the war followed in quick succession. These included
the 1993 Kono Statement, the 1995 Murayama Statement, the 1998
Japan–Republic of Korea Joint Declaration, and the 1998 Japan–China
Joint Declaration. Nationalist conservatives, like Abe, pushed back
against this trend.
In the early 1990s, after the Japanese government initially denied any
culpability in the 1930s/1940s-era comfort station system of sexual
servitude, Japanese researcher Yoshiaki Yoshimi emerged from the
Japanese government archives with proof of a government role in
these stations. What followed was an official government investigation
that culminated in the 1993 Kono Statement, which acknowledged the
role of the government and committed to learning from history. As part
of that commitment, the government supported the creation of the
Asian Women’s Fund (AWF), a joint public–private initiative to accept
moral responsibility and express atonement toward the comfort
women. The AWF provided a letter of apology from the sitting prime
minister and a certain amount of money to victims of the comfort
women system.
In 1994, on the heels of domestic political turmoil, Tomiichi
Murayama of the Japan Socialist Party (JSP) came into power. He
sought to create opportunities to make the upcoming fiftieth anniver-
sary of the end of the war a time for the government and the people to
reflect and express “apology and remorse.” The most concrete result
was the 1995 Murayama Statement, which was supported by a cabinet
decision. This is the most comprehensive Japanese government apology
to date, explicitly stating that “through its colonial rule and aggression,
[Japan] caused tremendous damage and suffering to the people of many
countries” and that it “must eliminate self-righteous nationalism.”
Two years later, under Keizo Obuchi of the LDP, two bilateral apolo-
gies for Japan’s wartime policies were forthcoming, first with South
Korea and then with China.
Some members of parliament were interested in furthering these
efforts. In collaboration with societal actors, the Democratic Party of
The Enduring Challenges of History Issues 515

Japan, the Social Democratic Party (formerly JSP), and the Japanese
Communist Party prepared a variety of bills toward that end. In 2001
several were consolidated into a single legislative proposal related to
the comfort women titled “Promotion of Resolution for Issues
Concerning Victims of Wartime Sexual Coercion Act” (Senji seiteki
kyōsei higaisha mondai kaiketsu sokushin ni kansuru hōritsuan)
(Yoshikawa 2002). The bill introduced a notion of national pride
that differed from the nationalist conservative one mentioned earlier,
in that it suggested that only in coming to terms with history and
resolving these issues with its neighbors could Japan “occupy an hon-
ored place in international society” (AWF 2001).
The nationalist conservative camp was incited by these develop-
ments. As Shibata (2016) argues, “The conservative political elite of
Japan were not prepared to incorporate Japanese war crimes into the
national narrative and collective memory” (91). They responded with
a three-pronged attack in the areas of textbooks, the media, and
government policy. Abe was part of these efforts as his political coming
of age was in the midst of this nationalist conservative resurgence. His
overall objective is to restore Japanese national pride by moving
beyond the vestiges of the wartime and postwar periods, as well as
what he views as the mistaken policies of the 1990s. Specific goals
toward this end are to rescind the Kono Statement, revise Article 9,
and worship at Yasukuni Shrine as prime minister.
Although Abe was not elected because of his views on history,4 he is
willing to expend some political capital to further his agenda on these
issues. He does so in calculated and strategic ways, with an acceptance
of practicality over sentiment. But he is continually endeavoring to
move the needle so that his overall objective is achieved, even as his
short-term goals have needed to be revised at times. In terms of history
issues, his focus has been on that of the comfort women.

17.3 Comfort Women

As mentioned, one of Abe’s specific goals is to rescind the Kono
Statement, which admitted Japanese government culpability in creating
a system of sexual slavery during the 1930s and 1940s. Within a month
of becoming prime minister in 2006, Abe called on his cabinet to
conduct a review of the Kono Statement. In March 2007, the cabinet
revealed its findings that the documents used in the 1990s for the
516 Mary M. McCarthy

government investigation that resulted in the Kono Statement con-

tained no evidence directly proving that Japan’s military authorities
or government officials were involved in the forced recruitment of
women (Niksch 2007). This finding was in keeping with Abe’s prior
statements about the Kono Statement and contradicts what is actually
declared in the Kono Statement.
The Abe cabinet’s declaration led to considerable pushback, both
domestically and internationally. Yoshiaki Yoshimi, a historian whose
academic research in the 1990s was one of the factors that spurred the
government of Japan to take action on the comfort women, criticized
the Abe government as “completely off the mark”5 and went on to
point out that “it should be made clear that it is the army that caused
this problem, not private agents (working for the army).”6
Perhaps even more significantly, the Abe government’s revisionism
fomented a cascade of national and regional legislative resolutions
calling on Japan to apologize for the comfort stations, starting with
US House Resolution 121. In an effort to forestall the passage of this
resolution, Prime Minister Abe apologized for the suffering of the
comfort women during his April 2007 visit to the United States, at
a press conference with President George W. Bush. Of course, Abe did
not succeed in this and the resolution passed on July 30. In the end,
Prime Minister Abe had to publicly declare his government’s intention
to continue to uphold the Kono Statement. And suffered the indignity
of a critical resolution passed by the national legislature of one of
Japan’s staunchest allies. The Netherlands, Canada, and the
European Parliament are among those who enacted similar resolutions
in the years following. Thus, while trying to shore up his own nation-
alist narrative within Japan, Abe unleashed the promotion of a global
narrative that directly opposed what he was trying to achieve.
Many of the chapters in this volume discuss the lessons learned by
Abe during his first tenure as prime minister, which allowed him to
better achieve his goals the second time around. Similarly, when it
comes to the comfort women issue, he has been committed and he
has been strategic. Although other initiatives took precedence, revision
of the Kono Statement (which was the next stage planned after the
2006/2007 review) was always on the agenda. And, during his second
term as prime minister, Abe used the media to set the stage.
In early 2014, the Sankei Shimbun and the broadcaster FNN
conducted a public opinion poll asking, “On the issue of the Kono
The Enduring Challenges of History Issues 517

statement, construed as an admission of forced recruitment of com-

fort women: there have been no public records to support claims of
military or government involvement, and the sloppiness of research
into ex-comfort women has been pointed out. Should the Kono
statement be revised?” The wording of the question was clearly
based on the 2007 Abe cabinet resolution. Of respondents, 58.6 per-
cent supported revision, while 23.8 percent were against. Prime
Minister Abe used these findings shortly thereafter to declare in
parliament that close to 60 percent of the Japanese public supported
revision of the Kono Statement (Sekiguchi 2014). His government
then proceeded to announce a plan to set up a team of experts to
examine the process of how the Kono Statement was created, with
the goal of highlighting its failings.
The expert findings, released in June 2014, stated that the main
research for the Kono Statement was the surviving women’s testimonies,
without supportive fact-checking research done by the government. This
was a reiteration of the long-held nationalist conservative stance that
there is no documentary substantiation of eyewitness accounts; it also
implied that oral testimonies are insufficient as evidence. The expert
findings further revealed that the Japanese government had worked
closely with the South Korean government in writing the Kono
Statement. This implied that the Kono Statement was a political docu-
ment rather than a culmination of research based on facts.
In response to the findings, the Abe government decided that it would
continue to uphold the Kono Statement, but it did so in a way that
suggested such a move was necessary for geopolitical reasons rather
than for reasons of historical justice or as part of a quest for the truth.
During a June 30 press conference, Chief Cabinet Secretary Yoshihide
Suga described how “the Japan–ROK relationship improved in the
context of giving consideration to the wishes of the ROK at the time
[of the Kono Statement], [but] the relationship is experiencing setbacks
again at this present time.”7 In other words, there were practical,
political reasons for the Kono Statement, and those reasons still exist
today, in terms of needing to ensure good Japan–Korea relations;
therefore, the Abe government will not revise or rescind the Kono
Statement at this time.
Many saw the outcome of the review as a hollowing out of the Kono
Statement, even as it was formally adhered to. The South Korean
Foreign Ministry “accused the panel report of ‘containing misleading
518 Mary M. McCarthy

content that undermines the credibility of the Kono Statement’” (Tiezzi

2014). Abe wanted to revise or even rescind the Kono Statement, but he
could not do so for practical reasons. Still, in order to achieve at least
a portion of his goals, he made it clear publicly that the decision was for
practical, geopolitical reasons. This allowed him to move closer to the
spirit of his goals (by calling into question the validity of the Kono
Statement), even without achieving the letter of that specific goal
(rescinding the Kono Statement).
The timing of these developments and geopolitical considerations
(and concerns) gave Abe another opportunity to pursue his goals. Since
2007, the tide of international public, government, and international
organization opinion had turned fully against Japan on the issue of the
comfort women (McCarthy 2018). This made Abe’s decision to con-
tinue to pursue it a bold one. Yet, at the same time, his calculations may
have been based on geopolitics. He may have foreseen that the same
practicality that led him to uphold the Kono Statement eventually
would lead South Korea and the United States to him on this issue.
The comfort women issue had become a significant strain on bilateral
relations between South Korea and Japan, which was hurting US stra-
tegic interests in the region. There was pressure on both sides to resolve
the issue, and South Korean President Geun-hye Park and Prime Minister
Abe stated that they intended to resolve the comfort women issue by the
end of 2015, in commemoration of the seventieth anniversary of the end
of World War II and the fiftieth anniversary of the normalization of
relations between the two countries. On December 28, the two foreign
ministers declared that a deal had been reached.
Although it does not move Japan closer to rescinding the Kono
Statement, the 2015 deal was a short-term win for Abe. The deal stated
that, basically, with an apology and a one-time contribution of funds,
“this issue is resolved finally and irreversibly.” This was less than what
was promised during the 1990s, and there is no educational component
of learning from history (McCarthy 2015). At the same time, most of
the Japanese public supported the deal (Kennedy & Nagakawa 2016;
Snyder & Glosserman 2016), and it was highly praised by the foreign
press, foreign pundits, and the US government. The pressure shifted
from being on Japan to make a deal to being on South Korea to uphold
the deal.
The reception in South Korea was such that the incoming Jae-in
Moon administration was unable to support its full implementation,
The Enduring Challenges of History Issues 519

despite a declaration in 2018 of acquiescence in upholding it

(Arrington 2018). The reality is that it was flawed from the outset, as
it did not take into account that collective memory is no longer
a monopoly of the state (McCarthy 2020). Any government apology,
even one highly influenced by diplomatic and geopolitical concerns,
needs to consider civil society. Without this, any agreement is doomed
to eventual failure (McCarthy 2016). Furthermore, short-term wins
will evaporate the next time the issue resurfaces, as it is bound to do
since it has not actually been resolved.
Still, we cannot assert that Abe’s goals were completely unmet. His
administration is generally seen as having made an overture at recon-
ciliation on this issue. It was the South Korean government that agreed
and then chose to renege. Therefore, even if reconciliation was not
achieved, and was not likely to have been achieved with such an
agreement, there was some success in shifting the global narrative in
terms of Japan being seen as offering restitution.

17.4 Yasukuni Shrine

In 1985, Yasuhiro Nakasone became the first postwar prime minister
to worship at Yasukuni Shrine in his official capacity. (A number of
prime ministers had visited previously, but they were never declared as
official visits (Rose 2008).) There was significant domestic criticism of
Nakasone’s action, and, when the government of China added its voice
to the criticism, Nakasone stopped worshipping there (Cheung 2017b,
p.55).
When Junichiro Koizumi was running for LDP president in 2001, he
promised conservatives of his party that, if he won the election among
the lower house legislators, he would visit Yasukuni Shrine annually
(Masshardt 2007; Tamamoto 2001). So he did. “In his first visit in
2001 . . . PM Koizumi yielded to domestic as well as external voices,
and his visit took place on August 13 instead of the symbolic date of
August 15 – the annual memorial day for the war’s end in 1945”
(Fukuoka 2013, p.33). However, before leaving office, he fulfilled his
ultimate goal, and August 15, 2006, “was the first visit by a prime
minister on that day since Prime Minister Yasuhiro Nakasone’s visit
21 years earlier” (Sakamoto 2017, p.2).8 Like Nakasone, Koizumi was
criticized by both domestic and foreign actors, particularly China.
However, unlike Nakasone, he did not bow to that criticism by stopping
520 Mary M. McCarthy

his visits to the shrine. He visited every year he was prime minister
(2001–2006). By the end of his term in office, more than half the
Japanese public opposed Koizumi’s visits (Shibuichi 2005; Fukuoka
2013). Members of the US Congress also criticized the visits, and the
Bush administration failed to invite Koizumi to speak before Congress,
as had been anticipated, arguably due to what it thought would be
criticism by Congress given Koizumi’s Yasukuni visits (author’s
interviews).
In the first prime ministerial visit since Koizumi’s in 2006, Abe visited
the Yasukuni Shrine in December 2013, after becoming prime minister
a year earlier.9 Like Koizumi had done, Abe explained it as a visit to pray
for peace, declaring, “Some people criticize a visit to Yasukuni as some-
thing to pay homage to war criminals, an idea based on
misunderstanding . . . I paid a visit to show (to the war dead) my deter-
mination to create an age where no one will ever suffer from tragedies of
wars.”10 However, CNN described the visit as having “ignited
a predictable firestorm of criticism and condemnation.”11 Along with
statements from China and South Korea, the US Embassy in Tokyo stated
that “the United States is disappointed that Japan’s leadership has taken
an action that will exacerbate tension with Japan’s neighbors.”12
This was in the aftermath of the United States trying to play a role in
discouraging visits to Yasukuni Shrine. In October 2013, then
Secretary of State John Kerry and Defense Secretary Chuck Hagel laid
a wreath at Chidorigafuchi Cemetery, which some have raised as an
alternative place to honor Japan’s war dead and one which does not
have the controversy of Yasukuni. Japanese officials have often laid
wreaths at Arlington National Cemetery, and this action by the US
government was intended to illustrate the parallel actions. It has not
resulted in Chidorigafuchi taking the place of Yasukuni in the national
story, however. And, certainly, it did not dissuade Abe from visiting the
shrine two months later.
Like Koizumi, Abe has had to negotiate his conservative support
base, which wants him to visit, along with his own conscience, which
I would argue also supports such visits, with the domestic and inter-
national criticism for such visits. In so doing, Abe has made different
choices than Koizumi. Since 2013, Abe has not visited the shrine in
person but has only sent offerings – on August 15, in memorial of the
end of World War II, and during the semiannual festivals at
Yasukuni.
The Enduring Challenges of History Issues 521

This difference may be explained by domestic rather than inter-

national considerations. Deans (2007) explores Koizumi’s visits to
Yasukuni as a domestic necessity in the face of unpopular economic
reforms. Although Abe also relies on a nationalist conservative base
and has presided over a time of economic reform, those reforms have
not seen the same type of pushback that Deans describes occurred
under Koizumi.13 Abe visited Yasukuni in 2013, in support of his
own conscience and those “hawkish LDP conservatives” who “consti-
tuted the core group that supported the return of Abe in 2012 during
the LDP presidential election” (Cheung 2017a, p.68), when “the
domestic political cost of his shrine visit was relatively low” (p.75).
However, as we headed into 2014, Abe viewed it to be more in his
domestic political interests to refrain from Yasukuni Shrine visits.
Cheung (2013) states: “Compromising on the Yasukuni visit could
benefit Abe on two domestic objectives in the short term: (1) to use the
foreign policy fulfillment [of progress in Japan–China relations] to
improve his declining approval rating and (2) to strategically use the
improved domestic support for his snap election planned in
December 2014” (p.70). In keeping with this argument, the
December 2013 visit could be viewed as having provided useful leverage
to secure a Japan–China Summit Meeting in November 2014, as specula-
tion swirled over whether Abe would visit Yasukuni again in December of
2014. (It was implicit that China would not agree to a summit if Abe was
planning a Yasukuni visit.) Thus, the 2013 visit established an opportun-
ity for Abe to seemingly compromise with the Chinese by not visiting the
shrine in 2014, thus securing the summit. This was also a strategic domes-
tic political move as the summit meeting occurred immediately prior to
dissolving the lower house and calling a snap election, the results of which
the Japan Times described as tightening Abe’s “grip on power.”14 And he
did not damage the level of support from his nationalist conservative base
because he visited the shrine in 2013 and he never declared that he would
not visit in the future.
In addition, Abe has been able to support Yasukuni Shrine in
other ways, in accordance with his general strategy toward history
issues, of seeking to alter the global narrative. In this respect, by
looking at Yasukuni Shrine and the comfort women issue in com-
parative perspective, we see similarities in strategy, despite the fact
that Abe has been perceived as being more assertive on the latter
than the former. Abe maintains that visiting Yasukuni Shrine is
522 Mary M. McCarthy

appropriate and that generations of Japanese born after the war

should not continue to apologize for the past. His repudiation of
the Kono Statement has similar sentiments behind it. He further
regards the Kono Statement as a document of political expediency
only and not representing historical facts. This is also in keeping
with his decision not to visit Yasukuni Shrine since 2013 (putting
practicality before sentiment).15 His administration has chosen to
uphold the Kono Statement, as he has chosen not to personally visit
Yasukuni since 2013. In the cases of both Yasukuni and the comfort
women, Abe is officially supporting a status quo that takes into
account both domestic and international perspectives. Yet, he is
simultaneously working to change the global narrative surrounding
these issues.
Despite official stances that may appear to say otherwise, Abe is
emboldening those around him to take actions that can undermine
a narrative of repentance, a narrative that he and his political agenda
do not support. For example, while Abe has not visited Yasukuni
Shrine since 2013, cross-party delegations of lawmakers have visited
in strong numbers.16 On August 15, 2019, Tomomi Inada, former
defense minister and current Abe aide, made an offering to the shrine
in Abe’s name, asserting Abe’s regret that he could not be there himself.
This annual visit and statement by an Abe ally has become regular
practice. Meanwhile, throughout this period, Abe has refused to defini-
tively declare that he will not visit.
Just like officially upholding the Kono Statement, while using the
2007 cabinet resolution and the 2014 review to hollow out the
Statement, this is a clear signal to his domestic supporters. At the
same time, it is an attempt to normalize such views on the global
stage. In the cases of both the comfort women and Yasukuni Shrine,
Abe has asserted that there is a “misunderstanding” globally, be it
a misunderstanding of what Yasukuni Shrine is or who the comfort
women were. So, while officially upholding (or not visiting), Abe is also
seeking to explain and educate. This is an effort to alter the global
narrative.
Initiatives to bring about such changes in the narrative are challenged
by the existence of counternarratives. The Japanese nationalist conser-
vative narrative is on one side and, on the other, Korean nationalist
narratives, Chinese nationalist narratives, and Japanese liberal pacifist
narratives, overlapping in support of each other, as it comes to comfort
The Enduring Challenges of History Issues 523

women and Yasukuni Shrine. Substate actors and transnational move-

ments have developed to charge Japan to accept war responsibility, as is
seen in actions such as the South Korean Supreme Court’s decision in
favor of Japanese company compensation for forced labor victims. Yet
those in support of the nationalist conservative stance also have an
international voice. One example that Masshardt (2007) shares is the
2003 publication Please Protect Yasukuni Shrine: Letters from
Brazilian Junior High and High School Students, which “provides an
apt example of the group’s [Nippon Kaigi, in this case] international-
ization strategy of seeking acceptance in order to extend the shrine’s
boundary” (p.324). It also seeks to change the global narrative.

17.5 Pearl Harbor

Changing the global narrative and instilling pride in Japanese youth
through historical education are not just about preserving Yasukuni
Shrine or rescinding the Kono Statement: they can also be about
showcasing a strong relationship with a former enemy. Post–World
War II reconciliation between the United States and Japan has been
a tremendous feat. Ninkovich (2007) argues that “postwar forgetful-
ness was quite deliberate, a byproduct of cold war imperatives in which
the recently defeated enemy had become a valuable asset” (p.87). From
1945 to 1951, US public opinion dramatically shifted from viewing
Japan as the country’s number one enemy to seeing Japan as “friendly.”
This was part of a strategic policy pursued by the US government to
shape US policy (both foreign and domestic) to align with geopolitical
considerations that arose with the advent of the Cold War.
In recent years, the US government has been troubled by Japan’s
tensions with its neighbors over lingering historical issues, as is exhib-
ited by its responses to Abe’s 2013 Yasukuni Shrine visit and the US
role in helping to facilitate the 2015 comfort women agreement behind
the scenes. As in the earlier age, geopolitics was at the fore, as concerns
over the ability for joint military exercises and other cooperative initia-
tives were at stake in a region where the United States has significant
interests in undisrupted sea lanes and military collaboration in the face
of a sometimes belligerent North Korea and a rising China.
With this in mind, and on the basis of President Barack Obama’s
personal commitment to a world free of nuclear weapons, Obama
visited Hiroshima in 2016. He laid a wreath in front of the cenotaph
524 Mary M. McCarthy

and gave a speech calling for the end of nuclear weapons. He also met
with survivors – all with Abe by his side. According to Daryl G. Kimbal,
executive director of the Washington-based Arms Control Association,
“It certainly was a powerful and important gesture of reconciliation
and remembrance.”17
Later that same year, Abe visited Pearl Harbor, with Obama at his
side. Abe declared, “We must never repeat the horrors of war again.
This is the solemn vow the people of Japan have taken,” and he called
US–Japan relations an “alliance of hope” as it illustrates what postwar
reconciliation can achieve.18
In these two visits, Abe’s goals were achieved. Not only were these
visits for the purposes of reconciliation and moving beyond the past,
but Obama’s visit preceded Abe’s. This was on the heels of a series of
steps taken by the US side. In 2010, Ambassador John Roos became the
first US ambassador to Japan to attend the annual August commemor-
ation of the bombing in Hiroshima. Ambassador Caroline Kennedy
followed suit, as well as visiting Nagasaki in 2013. And in April of
2016 John Kerry became the first sitting secretary of state to visit
Hiroshima. But in none of these visits, nor in Obama’s visit, was
there an apology. In fact, the Japanese government has never asked
the United States for an apology for the atomic bombings of Hiroshima
and Nagasaki. Instead, it was framed as a visit of peace and hope.
Abe’s visit to Pearl Harbor also allowed him to focus on “peace,”
a goal he has asserted in his reasoning for visiting Yasukuni Shrine as
well. Here Abe’s and the Obama administration’s goals coincided.
When Kerry visited Hiroshima he declared, “While we will revisit the
past and honor those who perished, this trip is not about the past. It’s
about the present and the future particularly, and the strength of the
relationship that we have built, the friendship that we share, the
strength of our alliance and the strong reminder of the imperative we
all have to work for peace for peoples everywhere.”19 Thus, there was
an ability to illustrate acknowledgement of the past (even if the events
hold a different meaning for each society) and reconciliation, without
apology. It could even be said to have served as a model for Japan and
its neighbors for how to move beyond the past.
The comfort women and Yasukuni are more difficult to resolve
because there is no reciprocal action that China or Korea can take, as
they were not the perpetrators of an action against Japan similar to the
atomic bombings of Hiroshima and Nagasaki. Both have also
The Enduring Challenges of History Issues 525

repeatedly called for official apology from Japan. In fact, what we have
often seen is an inability to move beyond the past. Frequently this is due
to domestic political considerations. Civil society plays a powerful role.
The Obama administration also had to consider domestic politics
and sentiments in its decision to have the president visit Hiroshima. The
National Air and Space Museum’s controversial Enola Gay exhibit
plan in the 1990s, which would have included what the organizers
hoped would be a balanced narrative incorporating artifacts from
Hiroshima and Nagasaki, sparked ferocious criticism from veteran’s
groups and media outlets (Heinrichs 2007). In the end, the exhibit was
considerably scaled down to a basic display of the fuselage.
Furthermore, it was World War II veterans in the US legislature who
were behind President George H. W. Bush’s inability to invite Prime
Minister Koizumi to speak before a joint session of Congress due to the
latter’s visits to Yasukuni Shrine. This criticism included a personal
letter from Representative Henry Hyde to Speaker of the House Dennis
Hastert.
By 2016, there were no longer any WWII veterans left in Congress.
Public views in the United States had also shifted. According to Pew
Research Center, the percentage of those Americans who thought the
atomic bombings were justifiable declined from 63 percent in 1991 to
56 percent in 2015, as the percentage of those who said it was unjusti-
fied rose from 29 to 34 percent. If looking at only eighteen- to twenty-
nine-year-olds, the percentage of those who said the bombings were
justified fell further to 47 percent.20
In many ways, Abe’s visit to Pearl Harbor was parallel to Obama’s to
Hiroshima. As Chief Cabinet Secretary Yoshihide Suga emphasized,
“This visit to Pearl Harbor was to console the souls of the war dead, not
to apologize.” Obama called the visit to the USS Arizona Memorial of
a Japanese prime minister “a reminder that even the deepest wounds of
war can give way to friendship and a lasting peace.”21 Abe’s visit to
Pearl Harbor might also be seen as a personal legacy issue. His grand-
father, whose experiences and perspectives have been so important in
shaping Abe’s, had actually visited Pearl Harbor six decades earlier;
although Abe, reportedly, was unaware of that when the trip was first
planned.22
On the other hand, some have been dismissive of this visit as not
engaging with the real difficulties in Japan’s history issues. Jeff
Kingston, of Temple University’s Japan campus, stated, “A symbolic
526 Mary M. McCarthy

gesture of contrition to your closest ally is easy . . . If he (Abe) really is

sincere about reconciliation diplomacy and overcoming lingering enmi-
ties he needs to visit similar symbolic sights (in China and Korea) . . .
and make similar remarks of remorse that are more specific about
Japan’s responsibility.”23
The important thing to note is that this visit must be examined in
terms of Abe’s twofold goal in historical legacy issues – as the origins of
his policies and with consideration of the degree to which they allow
him to achieve his goals. The visit to Pearl Harbor did not take away
from his goal of historical revisionism to engender national pride, as
there was no apology and the visit was parallel to a prior visit to
Hiroshima by President Obama. It also helped to provide talking points
for those engaged in trying to shift the narrative, as it is an example of
reconciliation that was, once again, free of apology. Still, it was viewed
by each outside party through the lens of their existing viewpoint of
Abe. In this way it did not actively change the global narrative about
Japan and war responsibility or historical justice. Meetings between
leaders are symbolically important and may have some effect on global
narratives, but – as the slowly evolving US public opinion about the
atomic bombings reveals – a population’s perspectives can endure and
are not easily malleable, even in the case of two close allies like the
United States and Japan.
This is revealed nowhere more starkly than in Japan–South Korea
relations. These are two countries that have a plethora of commonal-
ities, as liberal democracies and capitalist societies, and shared chal-
lenges. Yet historical legacy issues are impeding their ability to work
together to achieve common goals. Tensions have escalated over the
past two years and led to a tit-for-tat game that is close to spiraling out
of control. The Abe government has allegedly made a potentially dan-
gerous link between history issues and trade.24
The most recent trigger was the fall 2018 decision by the South
Korean Supreme Court in favor of Korean plaintiffs seeking compen-
sation from Japanese companies that used them as forced labor during
the 1930s and 1940s. The Japanese government maintains that issues
of compensation were resolved through the 1965 normalization of
relations and that South Korea must abide by that agreement under
international law. Frustrated with what it saw as a lack of diplomatic
progress, in summer 2019 Japan imposed restrictions on the export to
South Korea of three chemicals indispensable in chip and consumer
The Enduring Challenges of History Issues 527

electronic display production. South Koreans responded with

a consumer boycott of Japanese goods, and the government brought
a complaint against Japan to the World Trade Organization. In
September 2019, Japan removed South Korea from its “white list”
(the fast track for strategic material exports to “trusted” countries),
which means an increase in a broad range of trade restrictions, and then
South Korea followed suit. South Korea further threatened to pull out
of the General Security of Military Information Agreement (GSOMIA),
a 2016 intelligence-sharing pact with Japan.
With heightened tensions between the United States and both North
Korea and China, US Defense Secretary Mark Esper arrived in Seoul in
November 2019 declaring “that the only ones benefiting from friction
between Japan and South Korea were ‘Pyongyang and Beijing.’”25 The
United States has been widely criticized as remaining on the sidelines of
the Japan-South Korea conflict (Arrington & Yeo 2019; Glaser &
Mastro 2019). Yet, in the wake of Esper’s visit, South Korea agreed
to remain in the GSOMIA as talks with Japan on the trade disputes
continued. Still, without addressing the underlying history issues, it is
unclear what the long-term results will be. In the meantime we see not
only a negative spiral but an engineered unraveling of the ties that bind
these two countries, which is leading to a hardening of positions and
not a path to reconciliation.
The Japanese government’s stance that all issues of legal responsi-
bility and compensation were concluded with the San Francisco Peace
Treaty and the 1965 Japan–South Korea Agreement are not specific to
the Abe administration but have spanned all Japanese governments.
Reinforcing this in a global setting helps to promote this view as part
of the global narrative, one of Abe’s goals. This is what has led the Abe
government to assert its interest in third-country arbitration or the
International Court of Justice with regard to wartime labor and
Korea. Such a validation would aid in the task of achieving Abe’s
goal of shifting the global narrative in favor of Japan.26
However, at the same time, if the result is to harden the position
within South Korea and across the Korean diaspora, as is likely, the
2015 comfort women agreement is a lesson of how even a short-term
win may not help you to achieve a long-term goal. Global narratives are
more likely to be formed around personal stories of trial and triumph
than (even solid) legal arguments that remain at the state level.
528 Mary M. McCarthy

17.6 Conclusion
In conclusion, although the Abe administration has seen what some have
perceived as great diplomatic successes in the area of historical reconcili-
ation, with Obama’s visit to Hiroshima and Abe’s visit to Pearl Harbor, it
has been less successful in achieving the overall dual goal of enriching
national pride through historical revisionism and changing the global
narrative of Japan as perpetrator. This is due in part to the incompatibility
between these two goals, as well as the way in which the Abe administra-
tion has failed to take into account that the state no longer has a monopoly
on historical memory creation. Domestic, foreign, and transnational civil
society need to be considered and their interests addressed. State-to-state
reconciliation may be made through international agreements and meet-
ings between leaders, but, if the people are not part of the bargain, history
issues will continue to reemerge. This is exemplified by the failure of the
2015 comfort women agreement to provide a lasting resolution.

Notes
1. See Chapter 16 in this volume.
2. Later called the Diet Members’ League for a Bright Japan (Akarui Nihon
kokkai giin renmei).
3. Abe retracted his admission of pressuring NHK after a public outcry
against media censorship (Laurence 2005).
4. See Chapter 4 in this volume.
5. Reiji Yoshida, “Sex slave history erased from texts; ’93 apology next?,”
Japan Times, March 11, 2007, www.japantimes.co.jp/news/2007/03/1
1/national/sex-slave-history-erased-from-texts-93-apology-next/#
.V7dwwGWVchY, accessed July 24, 2016.
6. Quoted in Reiji Yoshida, “As Hashimoto self-destructs, party also
reels,” Japan Times, May 23, 2013, www.japantimes.co.jp/news/2013
/05/23/national/politics-diplomacy/as-hashimoto-self-destructs-party-
also-reels/#.V5V5DWWVchY, accessed July 24, 2016.
7. See https://japan.kantei.go.jp/tyoukanpress/201406/30_a.html, accessed
December 28, 2019.
8. After Nakasone, Prime Minister Kiichi Miyazawa reportedly visited the
Yasukuni Shrine in 1992, and Prime Minister Ryutaro Hashimoto
visited in 1996. Miyazawa’s visit was clearly private (or “secret” even)
(Sakamoto 2017, p.4), but Hashimoto’s visit is described as official by
some (Cheung 2017b, p.51) and private by others (Seaton 2008, p.165).
The Enduring Challenges of History Issues 529

9. This is in contrast to Abe’s decision during his first tenure as prime

minister (2006–2007) not to visit Yasukuni Shrine (Cheung 2013).
10. Quoted in Reiji Yoshida and Mizuho Aoki, “Abe visits Yasukuni,
angering Beijing and Seoul,” Japan Times, December 26, 2013, www
.japantimes.co.jp/news/2013/12/26/national/abes-surprise-visit-to-
yasukuni-sparks-criticism/, accessed July 24, 2016.
11. Ed Payne and Yoko Wakatsuki, “Japanese Prime Minister Abe visits
controversial Yasukuni war shrine,” CNN, December 28, 2013, www
.cnn.com/2013/12/25/world/asia/japan-pm-war-shrine/index.html,
accessed February 10, 2019.
12. Quoted in Ed Payne and Yoko Wakatsuki, “Japanese Prime Minister
Abe visits controversial Yasukuni war shrine,” CNN, December 28,
2013, www.cnn.com/2013/12/25/world/asia/japan-pm-war-shrine/ind
ex.html, accessed February 10, 2019.
13. In Chapter 4 in this volume, Maeda and Reed further assert that any
movement to the right by Abe is elite-driven and is not a reflection of the
voters.
14. “Abe tightens grip on power as ruling coalition wins 325 seats in Lower
House election,” Japan Times, December 15, 2014, www
.japantimes.co.jp/news/2014/12/15/national/politics-diplomacy/abes-
snap-election-pays-big-win/#.XtCxbC2ZOL8, accessed May 26, 2020.
15. Cheung (2013) describes this practicality when detailing how “Seko
Hiroshige, one of [Abe’s] senior political advisors, also openly
admitted in August 2006 to believing that Abe regards the Yasukuni
issue as a useful diplomatic card to China and thusly approaches the
issue strategically” (p.34).
16. As an example, the Japan Times has reported that August 15 at
Yasukuni Shrine saw sixty-six lawmakers in 2015, sixty-seven in
2016, and sixty-three in 2017.
17. Quoted in David Nakamura. “In Hiroshima 71 Years After First
Atomic Strike, Obama Calls for End of Nuclear Weapons.”
Washington Post. May 27, 2016.
18. “Remarks by President Obama and Prime Minister Abe of Japan at
Pearl Harbor,” December 27, 2016, https://obamawhitehouse
.archives.gov/the-press-office/2016/12/28/remarks-president-obama-
and-prime-minister-abe-japan-pearl-harbor, accessed January 4, 2020.
19. Quoted in Bradley Klapper, “Kerry makes gut-wrenching visit to
Hiroshima site of atomic bomb,” PBS. April 11, 2016, www.pbs.org/n
ewshour/politics/kerry-makes-gut-wrenching-visit-to-hiroshima-site-of
-atomic-bomb, accessed February 10, 2019.
20. Bruce Stokes, ”70 years after Hiroshima, opinions have shifted on use of
atomic bombs,” FactTank, August 4, 2015, www.pewresearch.org/fac
530 Mary M. McCarthy

t-tank/2015/08/04/70-years-after-hiroshima-opinions-have-shifted-on-
use-of-atomic-bomb/, accessed February 10, 2019.
21. Quoted in Jeff Mason, “On Pearl Harbor visit, Abe pledges Japan will
never wage war again,” Reuters, December 27, 2016, www.reuters.com
/article/us-usa-japan-pearlharbor/on-pearl-harbor-visit-abe-pledges-
japan-will-never-wage-war-again-idUSKBN14G1LH, accessed
February 10, 2019.
22. The Japanese government originally announced that this was to be the first
visit to Pearl Harbor by a sitting Japanese prime minister, but the US media
uncovered that it was the fourth such visit, including the one by Kishi in
1957; although Abe’s was the first with the accompaniment of a US
president. See Motoko Rich, “Japanese Leader’s Pearl Harbor Visit,
Called a First, Looks More Like a Fourth,” New York Times,
December 26, 2016, www.nytimes.com/2016/12/26/world/asia/pearl-
harbor-japan-shinzo-abe-visit.html, accessed May 25, 2020; “Hawaii
paper says two more Japanese leaders visited Pearl Harbor in the 1950s,”
Japan Times, December 24, 2016, www.japantimes.co.jp/news/2016/12/
24/national/politics-diplomacy/hawaii-paper-says-two-japanese-leaders-
visited-pearl-harbor-1950s/#.XtBrey2ZOu4, accessed May 25, 2020.
23. Quoted in Mason.
24. Neither side is formally admitting that its actions are in retaliation to the
other.
25. “Japan–South Korea friction flares again after GSOMIA intel pact
rescue,” Japan Times, November 25, 2019, www.japantimes.co.jp/ne
ws/2019/11/25/national/politics-diplomacy/japan-south-korea-
bickering-gsomia/#.XhGoxC2ZM0o, accessed December 23, 2019.
26. See “Failure of the Republic of Korea to comply with obligations
regarding arbitration under the Agreement on the Settlement of
Problems concerning Property and Claims and on Economic Co-
operation between Japan and the Republic of Korea,” July 19, 2019,
www.mofa.go.jp/press/release/press4e_002553.html, accessed May 25,
2020. The South Korean government’s position is available at “Facts
regarding the Judgments of the Supreme Court of the Republic of Korea
on the Victims of Forced Labor during the Japanese Colonial Rule over
the Korean Peninsula,” December 28, 2018, http://overseas.mofa.go.kr
/rw-en/brd/m_9983/view.do?seq=758681, accessed May 25, 2020.

References
Arrington, C. 2018. “South Korea ended its review of its ‘comfort
women’ deal with Japan. Here’s what you need to know.”
The Enduring Challenges of History Issues 531

Washington Post, January 11. www.washingtonpost.com/news/mon

key-cage/wp/2018/01/11/south-korea-ended-its-review-of-its-comfort
-women-deal-with-japan-heres-what-you-need-to-know/, accessed
September 8, 2019.
Arrington, C., and A. Yeo. 2019. “Japan and South Korea Can’t Get Along.”
Foreign Affairs, July 31. www.foreignaffairs.com/articles/japan/2019–07–
31/japan-and-south-korea-cant-get-along, accessed December 23, 2019.
Asian Women’s Fund (AWF). 2001. “Promotion of Resolution for Issues
Concerning Victims of Wartime Sexual Coercion Act.” www.awf.or.jp
/pdf/0204.pdf, accessed July 28, 2016.
Cheung, M. 2013. “Japan’s Domestic Politics and China Policy: The
Yasukuni Controversy (2006–07).” Journal of Contemporary East
Asia Studies 2(2): 25–46.
Cheung, M. 2017a. “Japan’s China policy on Yasukuni under Abe
(2012–2015): A political survival interpretation.” Journal of
Contemporary East Asia Studies 6(1): 62–78.
Cheung, M. 2017b. Political Survival and Yasukuni in Japan’s Relations
with China. Abingdon: Routledge.
Deans, P. 2007. “The Yasukuni Shrine: Contested Politics.” East Asia 24:
265–267.
Fackler, M. 2016. “The Silencing of Japan’s Free Press.” Foreign Policy,
May 27. https://foreignpolicy.com/2016/05/27/the-silencing-of-japans-
free-press-shinzo-abe-media/, accessed January 4, 2020.
Fukuoka, K. 2013. “Memory, Nation, and National Commemoration of
War Dead: A Study of Japanese Public Opinion on the Yasukuni
Controversy.” Asian Politics & Policy 5(1): 27–49.
Glaser, B. S., and O. S. Mastro. 2019. “How an Alliance System Withers.”
Foreign Affairs, September 9. www.foreignaffairs.com/articles/chin
a/2019–09–09/how-alliance-system-withers, accessed December 23,
2019.
Heinrichs, W. 2007. “The Enola Gay and Contested Public Memory.” In The
Unpredictability of the Past, edited by M. Gallicchio, pp.15–48.
Durham, NC: Duke University Press.
Hughes, C. W. 2015. Japan’s Foreign and Security Policy Under the “Abe
Doctrine.” London: Palgrave Macmillan.
Kennedy, T., and M. Nagakawa. 2016. “Public divided over ‘comfort
women’ agreement.” East Asia Forum, January 22. www
.eastasiaforum.org/2016/01/22/public-divided-over-comfort-women-
agreement/, accessed July 28, 2016.
Laurence, H. 2005. “Censorship at NHK and PBS.” JPRI Critique 12(3).
Masshardt, B. 2007. “Mobilizing from the Margins: Domestic Citizen
Politics and Yasukuni Shrine.” East Asia 24: 319–335.
532 Mary M. McCarthy

McCarthy, M. M. 2015. “Competing Voices and Compelling Visions: The

“Comfort Women” Issue in Japanese Domestic Politics.” Presented at
the Association of Asian Studies annual meeting, Chicago, IL,
March 2015.
McCarthy, M. M. 2016. “Comfort Women Agreement Must Engage Civil
Society.” East Asia Forum, January 28. www.eastasiaforum.org/2016/
01/28/comfort-women-agreement-must-engage-civil-society/.
McCarthy, M. M. 2018. “The Power and Limits of the Transnational
‘Comfort Women’ Movement.” In Handbook of Japanese Foreign
Policy, edited by M. M. McCarthy, pp.366–380. London: Routledge.
McCarthy, M. M. 2020. “The Creation and Utilization of Opportunity
Structures for Transnational Activism on WWII Sexual Slavery in
Asia.” In Agency in Transnational Memory Politics, edited by
J. Wustenberg and A. Sierp, pp.113–134. Oxford: Berghahn Books.
Mochizuki, M. M., and S. P. Porter. 2013. “Japan Under Abe: Toward
Moderation or Pragmatism?” Washington Quarterly 36(4): 25–41.
Niksch, L. 2007. “Japanese Military ‘Comfort Women’s’ System.”
Congressional Research Service Memorandum, April 3.
Ninkovich, F. 2007. “History and Memory in Postwar US–Japanese
Relations.” In The Unpredictability of the Past: Memories of the Asia-
Pacific War in US–East Asian Relations, edited by M. Gallicchio,
pp.85–120. Durham, NC: Duke University Press.
Penney, M. 2012a. “The Abe Cabinet – An Ideological Breakdown.” Asia-
Pacific Journal. http://apjjf.org/-Matthew-Penney/4747/article.html,
accessed August 18, 2016.
Penney, M. 2012b. “The NHK Comfort Women Documentary – 10 Years
Later.” Asia-Pacific Journal. http://apjjf.org/-Matthew-Penney/4617/ar
ticle.html, accessed July 28, 2016.
Rose, C. 2008. “Stalemate: The Yasukuni Shrine Problem in Sino-Japanese
Relations.” In Yasukuni, the War Dead, and the Struggle for Japan’s Past,
edited by J. Breen, pp.23–46. New York: Columbia University Press.
Sakamoto, K. 2017. “Visits to Yasukuni Shrine by the Prime Minister and
Japan-China Relations: What is Confusing the Debate?” Japan Institute
of International Affairs. www.jiia-jic.jp/en/resourcelibrary/pdf/Sakamoto_
Visits_to_Yasukuni_Shrine_by_the_Prime_Minister_and_Japan-China_R
elations_What_is_Confusing_the_Debate.pdf, accessed May 26, 2020.
Seaton, P. 2008. “Pledge Fulfilled: Prime Minister Koizumi, Yasukuni and
the Japanese Media.” In Yasukuni, the War Dead, and the Struggle for
Japan’s Past, edited by J. Breen, pp.163–188. New York: Columbia
University Press.
Sekiguchi, T. 2014. “Japan’s 1993 Comfort Woman Apology Returns to
Center Stage.” Wall Street Journal, February 26. http://blogs.wsj.com/j
The Enduring Challenges of History Issues 533

apanrealtime/2014/02/26/japan-comfort-women-deniers-claim-
government-on-their-side/, accessed July 28, 2016.
Shibata, R. 2016. “Japan’s Identity Crisis and Sino-Japanese Relations.” In
Disasters and Social Crisis in Contemporary Japan: Political, Religious,
and Sociocultural Responses, edited by M. R. Mullins and K. Nakano,
pp.81–103. London: Palgrave Macmillan.
Shibuichi, D. 2005. “The Yasukuni Shrine Dispute and the Politics of
Identity in Japan: Why All the Fuss?” Asian Survey 45(2): 197–215.
Snyder, S., and B. Glosserman. 2016. “The Japan-Korea Comfort Women
Deal: This is Only the Beginning.” Forbes, February 1. www.forbes.com
/sites/scottasnyder/2016/02/01/the-japan-korea-comfort-women-deal-
this-is-only-the-beginning/#6c4b9cc42e0b, accessed July 13, 2016.
Tamamoto, M. 2001. “A Land without Patriots: The Yasukuni Controversy
and Japanese Nationalism.” World Policy Journal 18(3): 33–40.
Tiezzi, S. 2014. “Japanese Report on the Kono Statement Draws Ire From
Seoul, Beijing.” The Diplomat, June 24. http://thediplomat.com/2014/0
6/japanese-report-on-the-kono-statement-draws-ire-from-seoul-beijing/,
accessed August 19, 2016.
Yoshikawa, H. 2002. “’Senji seiteki kyousei higaisha mondai kaiketsu
sokushin houan’ teishutsu to keika.” Kikan Sensou Sekinin Genkyuu
38: 26–33.
Index

100 Year Life Time Design Office, 63 Article 9, 492, 500

100-Year Life Era Vision Council, 293 ASEAN-wide defense framework, 490
1955 system, 50–52, 512 Asian open skies, 123
Asian Women’s Fund, AWF, 514
2001 system, 54 Aso, Taro, 110
asset purchase program, 22
6ji sangyōka, 427, 433 aviation policy, 122–124

abandoned land problem, 120 Bank of Japan

Abe, Shinzo, 3 Opinion Survey, Bank of Japan, 146
close advisors, Abe, 11, 16 banking crisis, 359
first regime, Abe, 110, 430, 516 Obama, Barack, 523
Abenergynomics, 447–449 Basic Plan for Gender Equality, 317
Abenomics, criticism, 5, 461–467 Basic Subcommittee, 289
Act Concerning the Promotion of Basic Survey on Wage Structure, 321
Women’s Career Activities, 317 bedrock companies, 380
Act on the Protection of Specially best loser system, 429
Designated Secrets, 70 Bilateral Planning Mechanism, 489
activism, 362 bilateral training, 488
administrative reform, 9, 43, 49, 53–55 board structure, 361, 371–378
adverse selection, 46 estimation, 373–378
Advisory Committee for National Broadcast Law, 15
Resources and Energy, 460 burasagari interviews, 14
agency problem, 46–48 bureaucratic control, 16–18
agency temps, 278, buryoku koshi, 500
Agricultural Cooperative Law, 434
agricultural crisis, 422 Cabinet Bureau of Personnel Affairs,
Agricultural Land Act, 424, 425, 435 17, 58, 483
agricultural production corporations, Cabinet Office, 53, 58, 287
APC, 426 Cabinet Secretariat, 9, 51, 53, 56
agricultural reform, 8, 32, 115–121, capital share, 214
421–441 carbon intensity, 462
agricultural subsidies, 118 chain of delegation, 45
Alliance Coordination Mechanism, 489 Chidorigafuchi Cemetery, 520
Amakudari, 459 childcare, 316
Amari, Akira, 62, 101 childcare leave, 314
Angel Plans, 315 childcare leave allowance, 314, 316
approval rating, 10, 12, 68–85, civil service reform, 49, 58–60
96–101, 521 Climate Change Performance
arms exports, 490 Index, 466

534
Index 535

climate change policy, 461–467 defense policy, 479–503

coal-fired power, 465 defense white paper, 479, 488
collective self-defense, 90, 484–485, demographic shift, 175, 271
493 Denki Rengō, 95
comfort women, 511, 514, 515–519 Deregulation Subcommittee of the
committee system, 361 Administrative Reform
company with audit and supervisory Committee, 278
committee, 371 dispatched workers, 312
compensation committee, 371 disposable income, household, 161
competition, foreign market, 206, 229 divided government, 65
Comprehensive and Progressive downward wage rigidity, 174
Agreement for Trans-Pacific DPJ
Partnership, CPTPP, 115 climate change policy, 446
Comprehensive Monetary Easing, 22 defense policy, 499
Comprehensive Strategy on City, disintegration, DPJ, 20, 92–94
People, Jobs, 27 election, DPJ, 89
constitutional revision, 90 faction, DPJ, 52
consumer price index, 175, 187 government, DPJ, 11, 92, 111
core core, consumer price index, 175, institution, DPJ, 93
187 labor policy, DPJ, 280
consumer theory, 159–161 lower house election, DPJ, 283
consumption growth, 151–162 nuclear policy, 459
consumption share of GDP, 151 dual labor market, 171, 183–187, 195,
consumption tax hike, 8, 159 274
corporate culture, 314, Duverger’s Law, 94
corporate governance, 32, 357–389
Corporate Governance Code, CGC, Economic Partnership Agreements,
358, 370–378, 381 EPA, 115–121
corporate performance, 384–387 economic reforms, 25–29
investment, 382, 385 economies of scale, 425
M&A expenditure, 387 ecosystem, 412, 414
payout, 384, 387 education system, 404
R&D expenditure, 387 efficiency clientelism, 462
ROA, 384 electoral reform, 9, 52–53, 428–429,
ROE, 382 437, 448
corporate strategies, 274 electoral system, 50
corporatism, 275 electric vehicles, 454
Council of Economic and Fiscal Policy, electricity deregulation, 452–453
CEFP, 111–114, 131 electricity prices, 447, 452, 455, 456
Basic Policy, CEFP, 123 electricity retail market, 452
Council on Economic and Fiscal Policy regional monopolies, 452
and the Council on Science and Electricity Utility Industry Law, 452
Technologies, 54 elitist, 83
Covid-19, 23, 25 employment flexibility, 403
CPI, 146 employment fluidity, 401, 402, 403
cross-domain operations, 487, 489 employment to population ratio, 148
cross-shareholding, 378–382 Energy and Environment Council, 460
energy mix, 450, 465
danjo koyo kikai kintoho, 278 energy policy, 33, 446, 449–455
defense budget, 493 entrepreneurship, 414–416
536 Index

Equal Employment Opportunity Law, Four Ministers Meeting, 61

278, 311 free and open Indo-Pacific, 490
equal pay for equal work, 290, 319 free trade, 115–121
eruboshi point system, 295, 318 free trade agreement, FTA, 115–121
Eto, Seiichi, 12 fuel cells, 454–455
exclusive defense, 494 Fukuda, Yasuo, 110
Expert Committee on Reforms Fukushima Daiichi nuclear plant, 446,
Addressing Globalization, 131 448
export prices, 228 full-time farmers, 427
external demand, 200–202 full-time worker
bonus, full-time worker, 181
Family Income and Expenditure Survey, excess, full-time worker, 184
FIES, 153 wage, full-time worker, 181
farm lobby, conservative, 425, 427,
436–439 GDP growth, 148, 243
farm vote, 423, 428 Gender Equal Participation Basic Plan,
farm zoku, 423, 428, 437 317
farmland consolidation, 431 Gender Equality Head Office Decision,
prefectural governors, leadership, 317
432 gender wage gap, 310–346
farmland consolidation banks, FLSB, average employment rate of mothers,
431 315
farmland ownership, 424, 426 distribution, 313
farmland transfer, 424, 426 estimation, 338–345
feed-in tariff, FIT, 453–454 managerial position, women,
fertility rate, 287 326–346
financial market, 141 promotion track, 312, 346
Financial Reconstruction Program, 359 ratio of managers, by gender,
first arrow, 3, 21–23, 139–163, 203, 327–336
240 gender wage ratio, 321–326
fiscal consolidation, 31 General Security of Military
fiscal policy, 3, 8, 23–25, 30 Information Agreement,
fiscal sustainability, 239–261 GSOMIA, 527
Blanchard framework, 241–242 generation gap, 282
forecast, 253–261 generational imbalances, 245
gross assets and liabilities, 249 gentan, 424, 432
interest rate gap, 243 gentei shain, 285
sustainable tax rate, 254–257 global financial crisis, 200
three scenarios, Broda and Weinstein monetary policy, global financial
(2005), 254 crisis, 203
five basic paths, 28 government expenditure, 249
fixed-term employees, 312 Government Pension Investment Fund,
flexibility mechanism, 463 GPIF, 128–132, 360, 367
Food Staple Control Law, 424, 425 portfolio, GPIF, 129
food supply chain, 427 tolerance bound, GPIF, 130
foolproof, 172 government revenue, 247–249
forced labor, 526 government-industry relationships,
foreign ownership ratio, 381 410–411
foreign policy, 4, 33, 60 gray zone challenges, 483, 486, 488
foreign workers, 293 Great Tohoku Earthquake, 446
Index 537

growth strategy, 4, 25–29 time series model, inflation, 177–180,

guest workers, 293 183–187, 188–194
Guidelines for Japan-U.S. Defense wage, inflation, 176
Cooperation, 484, 488 inflation expectations, 146, 173
gyosei kaikaku iinkai kisei kanwa inflation perceptions, household, 146
shoiinkai, 278 inflation target, 2%, 22
inflation-overshooting, 22
Haneda airport, re-internationalization, Initial Public Offerings, IPO, 397
121–124, 126 innovations, 32
Hasegawa, Eiichi, 12 insider ownership ratio, 360
Hatarakikata Kaikaku, 272, 286, 318 institutional investors, 366, 380
Hatoyama, Yukio, 111, 430, 464 ownership ratio, 369
Hayami, Masaru, 172 insurance companies, 367
Headquarters for the Creation of Intended Nationally Determined
Agricultural and Regional Vitality, Contributions, INDC, 465
436 interest groups, 94–96
health of banks, 201 intermediate inputs, 214
healthcare, 245 internal governance, 361
highly skilled foreign internet right, 91
professionals, 404 intertemporal substitution, 161
Hiroshima, 523 intra-party conflict, 7, 9
hiseiki rodosha, 274 investor engagement, 380
history issues, 33, 511–528 Ishiba, Shigeru, 7, 103, 282, 436
global narrative, 521, 526 Ishihara, Shintaro, 90
hitozukuri kakumei, 293 IT passport, 403
Honebuto no Hoshin, 112 Ito, Takatoshi, 129
honeymoon, 104
hostile takeover, 362 JA reform, 433–435
Human Resources Development Japan Agricultural Cooperatives, JA,
Revolution, 63, 293 423, 432, 433, 438
hydrogen fuel, 454 food supply chain, 425
hydrogen society, 454 JA Group, 434
LDP, relationships, 423–426
ideology, political, 72–73 Japan Business Federation, 275
Imai, Takaya, 12, 288 Japan Electrical and Information
import shares, 229 Union, 95
impossibility theorem, 240 Japan Federation of Employers’
Inada, Tomomi, 97, 437, 522 Organizations, 277
inbound tourism, 124–128 Japan Industrial Productivity Database,
Industrial Competitiveness 209
Council, 283 Japan Medical Association, 95
industrial policy, 25, 410, 455 Japan Revitalization Strategy, 114, 129,
industry-university-government 358, 363, 396
relationships, 405–411 Japan Trade Union Confederation, 278
inequality, economic, 279 Japanese Agricultural Cooperatives, JA,
infant care, 315 120
inflation Japanese Government Bond, JGB, 240,
bonus, inflation, 176 251
literature, inflation, 172–175 Japanese Stewardship Code, JSC, 358,
low inflation, 170 366–370
538 Index

Japan-EU Economic Partnership tightness, labor market, 183

agreement, 115 wage, labor market, 184
Japan-U.S. alliance, 488–489, 493, 497 labor mobility, 283
JCG, 486 labor participation rate, 286
JGB yield, 10 year, 22 Labor Policy Council, 275
Jiminto [LDP] Net Supporters Club labor productivity, 286
(J-NSC), 15 labor regulation, 277
jinsei hyakunen jidai koso kaigi, 293 labor share, 162, 214
job-style, 285 labor shortage, 271–301, 404
jobu-gata, 285 Labor Standards Law, 278
Josei Katsuyaku Suisin Hō, 284, 317 labor supply, 175
JPX400, 371 landslide victory, 87,
JSDF, 484–485 Law to Promote the Coordination
joint-ness, 485 of Food, Commerce and Industry,
overseas deployments, 493 427
overseas deployments, JSDF, 484 Law to Support the Strengthening
of Agricultural
kaishaku kaiken, 90 Competitiveness, 435
Kake Gakuen, 97 LDP
Kan, Naoto, 111, 448, 453 faction, LDP, 51, 87, 103
kansa tou iinkai secchi gaisha, 371 institution, LDP, 93
karoshi, 289 legislation, 5, 12–13
Kato, Katsunobu, 12 legislation, timing, 13
Keidanren, 275, 296, 457 level of confidence, 69
keiretsu, 412 liberalization, agriculture, 119
keizai dantai rengokai, 275 life-cycle theory, 159
Key Performance Indicators, KPI, 395, limited regular workers, 285
400, 403–405, 408, 410, 415 local Abenomics, 26–27
kibo no to, 292 local agricultural committees, 432, 435
kihon bukai, 289 long-term employment, 274, 313, 402
kisei kaikaku suishin kaigi, 289 Low Cost Carriers, LCC, 128
Kishi, Nobusuke, 513 lower house election, 6, 15, 20
kogekigata, 487
Koike, Yuriko, 16, 292 Machi, Hito, Shigoto Sōgō Senryaku,
Koizumi, Junichiro, 11, 16, 53, 54, 87, 27
90, 103, 110, 286, 359, 429, 439, main bank system, 359, 399
519 man in the street interviews, 102
Koizumi, Shinjiro, 438, 441, 467 Esper, Mark, 527
Komeito, 501 market inflation forecasts, 142
Kono Statement, 514, 515–518 markup, 200–234
resolution, 516 decomposition, 1994–2006,
Kuroda, Haruhiko, 3, 22, 173 215–218
Kyoto Protocol, 446, 462 decomposition, 2006–2015,
219–224
labor adjustment, 280 decomposition, markup, 208
flexibility, 275, 277, 290 estimation, markup, 206–209
labor dislocation, 278 markup, 1994–2006, 214
Labor Force Survey, 320 markup, 2006–2015, 219
labor market, 170, 174, 180–187 Masuzoe, Yoichi, 132
inflation, labor market, 187–194 media cycle, 14
Index 539

media strategy, 13–15, 101–102, 513 negative interest rate policy, 22

membership-style, 285 net exports, 200
menbashippu-gata, 285 netto uyoku, 91
METI, 288 new three arrows, 27, 285
military capabilities, China, 496 nihon keieisha dantai renmei, 277
military capabilities, North Korea, 496 nihon rodokumiai sorengokai, 278
military technologies, 497 Nikkeiren, 277
minimum access, rice, 118 Nishikawa, Koya, 437, 440
ministerial appointment, 53, 55 nōchi chūkan kanri, 431
Ministry of Agriculture, Forestry and Noda, Seiko, 104
Fisheries, MAFF, 436 Noda, Yoshihiko, 111
Mogi, Toshimitsu, 63 nōgyō iinkai, 432, 435
monetary announcement, 144 Nōgyō kyōsōryoku kyōka shien hō, 435
monetary easing, 21–23, 30, 203 nominal interest rates, 142
monetary expansion, 139–163, 170 nomination committee, 371, 373
monetary policy, 3, 8 non-performing loan, 359
age, monetary policy, 158 non-regular worker, 274, 281
announcement effects, monetary non-standard worker, 314
policy, 143 Nōrin bukai, 423
debt-to-income ratios, monetary Nōrin zoku, 423
policy, 154 Nōrinchūkin, 434
financial asset to income ratios, Nōrinsuisangyō · chiiki no katsuryoku
monetary policy, 156 sōzō puran, 431
heterogeneous effects, monetary North Korea, 21
policy, 153 Nōshōkō renkei sokushin hō, 427
homeownership status, monetary Nuclear and Industrial Safety Agency,
policy, 156 NISA, 458
income groups, monetary policy, 158 nuclear energy, 447, 448, 455–461
money supply, 252 regulatory capture, 458
monitoring, 46, 53 nuclear policy, 455–461
Moon Jae-in, 4, 518 Nuclear Regulatory Authority,
moral hazard, 46 NRA, 458
Mori-Kake scandals, 68 Nuclear Safety Commission, NSC, 459
Moritomo Gakuen, 98, 292 nuclear technology, export, 457
Moriyama, Hiroshi, 437 nuclear umbrella, 494
Murayama Statement, 514 nuclear village, 449, 461
Murayama, Tomiichi, 514 nudge, 314, 317

Nakae, Motoya, 12 Obuchi, Keizo, 53, 514

Nakasone, Yasuhiro, 519 occupation-based wages, 285
Narita airport, 121, 127 offensive platform, 487
National Security Council, 60–61 official campaign period, 102
national security council and oil shock, 462
secretariat, 480 on-the-job-training, 175
National Security Council, NSC, open market policy, 253
482–483, 491 open sky, 122
national security law, 70–71, 75–77, 79 organizational vote, 94–96,
National Security Secretariat, NSS, 423, 428
60–61, 482–483 Ota, Hiroko, 113
nationalist, 513 output prices, 224, 228
540 Index

outside director, 371 manufacturing, 211

Ozawa, Ichiro, 8 service, 211
Promotion of Resolution for Issues
Paris Agreement, 447, 465 Concerning Victims of Wartime
Park, Geun-hye, 518 Sexual Coercion Act, 515
parliamentary system, 45–49 Promotion of Women’s Participation
particularism, 9 and Advancement in the
part-time farmers, 427 Workplace Law, 284
Part-Time Labor Law, 279 public opinion, 12–15, 92, 498, 516
part-time worker public subsidy system, 53
bonus, part-time worker, 182 puzzle, monetary theory, 163
wage, part-time worker, 180
Party for Future Generations, 91 Quantitative and Qualitative Easing,
party leadership, 50 QQE, 22
Party of Hope, 292
pass through, exchange rate, 228 Radio Law, 15
pay-as-you-go, PAYGO, 128 rally-’round the flag effect, 72
peace and security legislation, 484–485, regular worker, 274
498 regulatory reform, 408
Pearl Harbor, 523–525 Regulatory Reform Promotion
personnel, bureaucratic, 16–17, 60, Council, 289
483, 491 regulatory reform, nuclear energy,
Phillips curve, 173–174 458–461
Plan for the Creation of Agricultural relational shareholding,
and Regional Vitality, 431, 436 378–382
Policy Affairs Research Council’s relational shareholding ratio, 380
Agriculture Committee, 423 renewable energy, 450, 453
policy formulation, 46, 60, 62 Rengo, 278
political funds, 52 revisionism, 516, 526
political reform, 9–10, 29–30, 43, 49 rice import, 426
populism, 84 rice markets, closed, 424
postal rebels, 103 rice paddy reduction policy, 424
potential output, 174 rice prices
press freedom, 15 government guaranteed,
price-level target, 251–253 rice prices, 424
primary balance, 24 liberalization, rice prices, 426
primary deficits, 24 right wing, 91
prime minister rodo kijunho, 278
legal authority, prime minister, 51 rodo seisaku shingikai, 275
policy formulation, prime minister, rodo-zoku, 276
53 ryoki odan sakusen, 487
power range, prime minister, 48–49
prime minister and ministers, sangyo kyosoryoku kaigi, 283
relationship, 48–49, 51–52, 53 Sato, Yuji, 17
prime minister and parliament, savings rate, 159
relationship, 47–48, 50–51, 52–53 scale economies, agriculture, 120
Prime Minister’s Office, 51, 287, scandals, 13, 100
481, 491 screening, 46, 53, 60
priming effect, 76–77, 78–83 second arrow, 3,
production variables, 1994–2006, 211 23–25, 240
Index 541

Secretariat for the Revitalization of the Suga, Yoshihide, 11, 17

Japanese Economy, SRJE, 63 Suzuki, Hiroshi, 12
security environment, 495–497
security policy, 5, 33, 60 Takenaka, Heizo, 112
seiki rodosha, 275 Tamogami, Toshio, 91
seishain, 313 tariff rates, 119
Seko, Hiroshige, 12, 16 tax policy, 319
seniority, 285 Technology Licensing Offices, TLO,
Senji seiteki kyōsei higaisha mondai 405
kaiketsu sokushin ni kansuru temporary worker, 404
hōritsuan, 515 territorial defense, 486
Senkaku Islands, 486 third arrow, 4, 25–29, 62, 109–133
senshu boei, 494 labor market reform, 283–286
shain, 275 TPP11 treaty, 115
shareholder rights, 365 Trade Agreement between Japan
Shiozaki, Yasuhisa, 288 and the United States of America,
Shirakawa, Masaaki, 173 115
shokumukyu, 285 trade coverage, 120
shokunokyu, 285 trade variables, 224
short hour option, 315 Trans Pacific Partnership,
Silicon Valley Model, 394, 396–416 TPP, 115, 430
immigrant, 401 triangular trade, 200
single non-transferable vote system, 50
sixth industrialization, 427, 433 U.S.-Japan Australia Trilateral Strategic
skill-based wages, 285 Dialogue, 489
small capitalization market, 399 unicorn, 400
SNTV-MMD system, 9 university-industry relationships,
social pension, 128–132, 405–409
243–247 unobserved components model, 193
social pension tax, 247 Uruguay Round, GATT, 426
social security law, 245 use of force, 500
social security tax exemption, 319
Society 5.0, 28 value added tax, VAT, 4, 24
solar power, 453 vector auto regression, VAR, 188
spousal tax deduction, 319 venture capital, VC, 397
startup, 412–413, 414–416 independent VC, 399
startup pitch contest, 415 veto, 51, 52
startup ecosystem, 395, 400, 402, 412, visa requirements, 126
414, 416 Visit Japan program, 125
stimulus package, 23 voter turnout, 89, 501
stock market, 368–370 voting right, 367
stock option, 361, 402 voting right, institutional
stock ownership, 359 investors, 362
stock prices, 142
Strøm, Kaare, 45–49 wage distribution, 323
indirectness, Strøm, 46 wealth effects, consumption, 156
singularity, Strøm, 46 Westminster model, 65
structural break, 1998, 175–180 white list, 527
structural reforms, 31, 394–417 Womenomics, 32, 284, 297,
subprime loan crisis, 362 310–346, 404
542 Index

Work Style Reform, 272, 286–294 yen depreciation, 125,

monitoring, 296 142, 231
Work Style Reform Law, 292 Yen/Dollar exchange rate, 203
working hour restrictions, 284 Yoshimi, Yoshiaki, 514, 516
working hours, reduction, 289
working-age population, 159 Zenchū, 434
Zennō, 434
Yasukuni Shrine, 519–523 zero interest rate, 22
yen appreciation, 202 zero lower bound, 203