彼得森经济研究所 美国是否正在经历一场将提振中产阶级的制造业复兴?(英) 2024.10 17页
彼得森经济研究所 美国是否正在经历一场将提振中产阶级的制造业复兴?(英) 2024.10 17页
    Note: The author is grateful for comments from Gary Hufbauer, Adam Posen, and
    participants at seminars at Harvard Kennedy School and the Peterson Institute. He
    is especially grateful to Steve Weisman and Madona Devasahayam for guidance
    and editorial assistance, Barbara Karni for copyediting, Julieta Contreras for data
    verification, and Christopher Ong for research assistance.
The historic trend of the declining share of jobs in manufacturing in the United                 Robert Z. Lawrence
                                                                                                 is nonresident senior
States has bedeviled politicians and policymakers over many years. Elected                       fellow at the Peterson
in 2020 in the wake of an economic downturn aggravated by the COVID-19                           Institute for International
pandemic, President Joseph R. Biden Jr. made the goal of his economic policies                   Economics and the Albert
                                                                                                 L. Williams Professor of
to “build back better” and restore the middle class by reviving industrial                       International Trade and
jobs, especially in the Midwest, which he labeled “growing the economy                           Investment at Harvard
                                                                                                 Kennedy School. This
from the bottom up and the middle out.”1 The emphasis on manufacturing
                                                                                                 Policy Brief is a revised and
was reinforced by an economic nationalist goal of returning jobs supposedly                      updated version of chapter
sent overseas back to US shores—“making and building it in America,” as the                      10 of his forthcoming
                                                                                                 book, Behind the Curve:
administration proclaimed.2                                                                      Can Manufacturing
    This emphasis is reflected in the special incentives for US manufacturing in                 Still Provide Inclusive
                                                                                                 Growth? (Peterson
President Biden’s programs. He raised the threshold local content requirement for
                                                                                                 Institute for International
procurement by the US government. The Infrastructure Investment and Jobs Act                     Economics, 2024).
(also referred to as the Bipartisan Infrastructure Law) requires that all iron, steel,
manufactured products, and construction materials used in its projects be made
in America.3 The CHIPS and Science Act appropriates $24 billion in tax credits
for manufacturing semiconductors in the United States and another $39 billion
to provide incentives for investment in chip facilities and equipment in the United
States.4 The Inflation Reduction Act (IRA) provides tax credits for clean energy
investments and production and the purchase of electric vehicles (EVs) that are
assembled in North America and have batteries that use minerals that are mined
and refined in the United States or a country with which the United States has a
free trade agreement.5
    The administration’s policies have expanded the base of manufacturing
employment on the margins in recent years. But despite their appeal, these
policies are unlikely to be the key to achieving middle class growth, because
manufacturing no longer plays the role it played in the past in providing
opportunities for workers without college degrees to join the middle class.6
Manufacturing can still help achieve other goals, such as providing hardware for
the digital revolution; weapons for national security; and the EVs, wind turbines,
and solar panels that are vital for decarbonization. But the sector is now too
small to play a major role in reviving America’s depressed regions and providing
significant opportunities for American workers.
    The appeal of creating manufacturing jobs is not likely to subside politically
in the United States. But the United States is hardly alone in experiencing a
long trend of declining manufacturing employment. As a result of increased
automation and technological change and the shift in demand toward services,
manufacturing accounts for a declining share of employment and output in all
high-income economies and many emerging-market economies. Remarkably,
declining trends in the share of manufacturing employment are evident even in
countries such as Germany, South Korea, Japan, Singapore, and China, despite
their substantial trade surpluses in manufactured goods (Lawrence 2020, 2024).7
4   Justin Badlam, Stephen Clark, Suhrid Gajendragadkar, Adi Kumar, Sara O’Rourke, and Dale
    Swartz, The CHIPS and Science Act: Here’s What’s in It, McKinsey & Company, October 4, 2022.
5   For a comprehensive analysis of the IRA, see Bistline, Mehrotra, and Wolfram (2023).
6   This mistaken view is shared by Robert Lighthizer, US trade representative in the Trump
    administration, who advocates increasing tariffs to rejuvenate American manufacturing. See
    “Donald Trump’s former trade chief makes the case for more tariffs,” By Invitation: American
    trade policy, The Economist, March 8, 2024. In Lawrence (2024), I elaborate on this argument.
7   Singapore had a very large trade deficit as a share of GDP from 1962 until 1997, averaging
    25 percent of GDP. But between 1998 and 2018, it averaged a trade surplus of 15 percent of
    GDP.
8   See Amanda Chu, Oliver Roeder, and Alex Irwin-Hunt, “Inside the $220bn American Cleantech
    Project Boom,” Financial Times, August 16, 2023.
9   David Brooks, “The American Renaissance Is Already at Hand,” New York Times, September 7,
    2023.
PB 24-12 | OCTOBER 2024                                                                                                  3
Figure 1
US manufacturing employment, January 2019–July 2024
millions of jobs
13.0
12.8
12.6
12.4
12.2
12.0
                                                    January 2021
 11.8                                               Start of Biden administration
 11.6
 11.4
                          1.37 million
 11.2                 Manufacturing jobs lost
                        (Jan.–Apr. 2020)
11.0
         Jan  May     Sep    Jan  May      Sep    Jan    May   Sep     Jan  May     Sep    Jan  May   Sep     Jan  May
         2019                2020                 2021                 2022                2023               2024
Source: Federal Reserve Bank of St. Louis, Federal Reserve Economic Data (FRED), https://fred.stlouisfed.org/series/
MANEMP.
10      Paul Krugman, “Biden and America’s Green Push,” New York Times, August 17, 2023.
11      White House, The Biden-Harris Economic Blueprint, September 2022, p. 17.
12      The White House, Statement from National Economic Advisor Lael Brainard on US Steel,
        December 21, 2021.
PB 24-12 | OCTOBER 2024                                                                                                        4
Figure 2
US manufacturing and nonfarm employment, January 2019–July 2024
90
85
       Jan  May      Sep    Jan  May      Sep    Jan    May     Sep    Jan  May      Sep    Jan  May       Sep    Jan  May
       2019                 2020                 2021                  2022                 2023                  2024
Source: Federal Reserve Bank of St. Louis, Federal Reserve Economic Data (FRED), https://fred.stlouisfed.org/series/
MANEMP.
nonfarm employment increased by 5.8 percent. The result was a steady decline
in the manufacturing employment share of nonfarm employment (figure 3). As
of July 2024, the data showed no renaissance in manufacturing employment.
Employment simply recovered to 2019 levels and stagnated.
     Output growth in manufacturing has also been relatively slow. Between the
first quarter of 2019 and the first quarter of 2024, the ratio of real manufacturing
output to GDP (both measured in 2017 dollars) fell 5.2 percent (figure 4).
Manufacturing’s role in the expansion of US output has been minimal.
     The slow growth in manufacturing employment combined with the sector’s
declining output share resulted in slower growth in labor productivity growth in
manufacturing than in the economy as a whole. In the second quarter of 2024,
labor productivity in the business sector, for example, was 12.7 percent higher
than in the same period in 2017. In contrast, labor productivity in manufacturing
fell by 2.0 percent.13 Between 2017 and 2022 (the latest data available), the
increase in total factor productivity in manufacturing—the more comprehensive
measure of the rise in output relative to all inputs used in manufacturing—was
less than 0.1 percent.14 By contrast, for decades until 2010, faster productivity
growth in manufacturing than in the rest of the economy was the norm.15
13    Bureau of Labor Statistics, Labor Productivity and Cost Measures: Major Sectors: Nonfarm
      Business, Business, Nonfinancial Corporate, and Manufacturing, August 1, 2024.
14    Bureau of Labor Statistics, Total Factor Productivity and Related Measures, Major Industries,
      March 21, 2024.
15    For example, output per worker was 2.5 percent faster in manufacturing than in the economy
      as a whole between 1987 and 2000 and 3.69 percent faster between 2000 to 2010 (Lawrence
      2024).
PB 24-12 | OCTOBER 2024                                                                                                    5
Figure 3
Manufacturing employment as percent of nonfarm employment in the United States,
January 2019–July 2024
percent
8.8
8.5
8.4
8.3
8.2
8.1
8.0
7.9
7.8
      Jan      May   Sep   Jan  May    Sep     Jan    May    Sep    Jan  May      Sep    Jan  May      Sep    Jan  May
      2019                 2020                2021                 2022                 2023                 2024
Source: Federal Reserve Bank of St. Louis, Federal Reserve Economic Data (FRED), https://fred.stlouisfed.org/series/
MANEMP.
Figure 4
Ratio of real manufacturing output to GDP in the United States, 2019Q1–2024Q1
98
96
 94
                                                                                                        5.2% decrease
                                                                                                      (2019Q1 to 2024Q1)
 92
90
88
          Q1    Q2 Q3      Q4   Q1   Q2 Q3    Q4      Q1    Q2 Q3    Q4    Q1    Q2 Q3     Q4   Q1    Q2 Q3      Q4     Q1
                 2019                 2020                   2021                 2022                 2023            2024
Note: Figures are measured in 2017 chain dollars. Data are seasonally adjusted.
Source: US Bureau of Economic Analysis, Chain-Type Quantity Indexes for Value-Added by Industry.
PB 24-12 | OCTOBER 2024                                                                                                   6
Figure 5
US spending on manufacturing construction, 2018–24
200
150
100
                                                                   $76.4 billion
 50
          Jan      Jul      Jan      Jul      Jan      Jul     Jan        Jul      Jan    Jul   Jan    Jul     Jan
          2018              2019              2020             2021                2022         2023           2024
   In sum, in the first three and a half years of the Biden administration, there
was no renaissance in manufacturing. Employment growth stagnated after the
pandemic recovery, the share of manufacturing in employment and output
declined, and manufacturing labor productivity did not rise.
16       Federal Reserve Bank of St. Louis, Federal Reserve Economic Data, Total Construction
         Spending: Manufacturing in the United States.
17       US Department of Treasury, Unpacking the Boom in US Construction of Manufacturing
         Facilities.
18       White House, President Joe Biden: Investing in America.
PB 24-12 | OCTOBER 2024                                                                                      7
19   The detailed state data by investment category and aggregate investments by state do not
     correspond exactly to the aggregates by type of investment and of all states. Summing the
     detailed underlying data yields an aggregate investment of $887.2 billion rather than the
     $897.8 billion reported by the White House. As the difference is very small (for reasons that are
     unclear), it does not make a significant difference to the calculations in table 1 and 2.
20 Bureau of Economic Analysis, Current Cost of Private Fixed Assets by Industry.
21   The White House announced a total of $730.1 billion for manufacturing, but the investments
     in the excluded states are small and mainly in clean power, which is not counted as
     manufacturing. Alaska ($8 million) and the District of Columbia ($5 billion) have investments
     only in clean power, and Hawaii has investments of $95 million in clean manufacturing and
     $135 million in clean power. The aggregate manufacturing number is therefore distributed over
     the 48 states in the continental United States.
22   Bureau of Economic Analysis, Total Full-Time and Part-Time Employment by NAICS Industry.
23   Utilities are very capital intensive. In 2022, the net capital stock in utilities was $3.531 trillion
     and employment was 556,000. An increase of $159.9 billion would increase the capital
     stock by 4.5 percent and employment by just 25,178 jobs. The estimated total employment
     associated with all the announced investments in manufacturing and clean energy would equal
     1.725 million.
24   See Federal Reserve Bank of St. Louis, Federal Reserve Economic Data (FRED), All Employees,
     Manufacturing.
PB 24-12 | OCTOBER 2024                                                                               8
    This approach obviously paints with a very broad brush. The estimates
are intended to provide a sense of the order of magnitudes of the potential
employment impacts of the investments rather than precise estimates. In fact,
capital-labor ratios differ—by industry, firm, and state—and as it will take several
years for the investments to be completed, the ratios are likely to change.
    The intention here is to capture the long-run additions to manufacturing
employment if these investment projects are carried out. These estimates do not
include the jobs created in planning, constructing, and equipping the facilities.
The estimates also exclude the long-run value-added and employment generated
in nonmanufacturing sectors.
    It should be emphasized that these estimates take account only of positive
prospective additions to employment associated with the investments and ignore
factors that could offset this growth, such as employment declines associated
with the shutting down of manufacturing production dependent on fossil fuels,
declines from auto parts production rendered obsolete by EVs, depreciation
of existing manufacturing assets, possible increases in productivity growth
from automation, declines in other manufacturing industries and price and
employment changes since 2022.
    Moreover, some of the investments might not actually be made—because
of changes in firm plans; implementation difficulties, such as failure to obtain
permits, the federal and state grants the announced investments were predicated
upon, and/or sufficiently skilled workers; and changing demand conditions. Since
grants are still being made, however, it is also possible that the tabulations could
underestimate the ultimate magnitudes of the investments eventually spurred
by the programs.
    Given that the economy is close to sustainable full employment (the
unemployment rate was about 4 percent in August 2024), most of the workers
who take jobs in these operations are most likely currently working in other
jobs rather than unemployed or out of the labor force. This means that the
estimates should be regarded as depicting potential changes in the industrial
and geographic composition of employment rather than aggregate employment
additions at the national level.
Impacts by State
Some states could experience very substantial percentage increases in demand
for manufacturing employment (table 1).25 The states with manufacturing
employment gains from the announcements of 30 percent or more are Arizona
(139.2 percent increase), Idaho (79.2 percent), West Virginia (57.5 percent), New
York (47.8 percent), Oregon (47.3 percent), and New Mexico (46.3 percent).
A second group of states are above 20 percent. They include Nevada
(28.3 percent), Louisiana (26.4 percent), Texas (25.7 percent), and Georgia
(21.3 percent).
    The location of these investments represents a redistribution of US
manufacturing geographically, as the employment effects are concentrated in
25   Because they use more recent data and a slightly different methodology, the estimates in table
     1 here differ from the announcement employment numbers in table 10.1 in my book (Lawrence
     2024).
PB 24-12 | OCTOBER 2024                                                                                                   9
Table 1
Estimated impact of announced investments (as of July 2024) on wage and salary employment in
manufacturing in the United States, by state
Note: Data are based on 48 states. Alaska, Hawaii, and the District of Columbia are excluded. The estimates in this table differ from the
announcement employment numbers in table 10.1 in my book (Lawrence 2024) because of different methodologies and data used.
Sources: US Bureau of Economic Analysis, https://www.bea.gov/data/employment/employment-by-state; White House, Investing in America,
https://www.whitehouse.gov/invest (accessed August 5, 2024).
PB 24-12 | OCTOBER 2024                                                                                                            11
Table 2
Effect of announced investments on employment, by group of states (percent above 2022 levels)
States Biden won in 2020 53.6 57.7 63.5 50.2 12.5 0.9
26   Bureau of Economic Analysis, SAEMP27N Full-Time and Part-Time Wage and Salary
     Employment by NAICS Industry.
PB 24-12 | OCTOBER 2024                                                                                 12
     Consider, for example, the effect on the Rust Belt (defined here as Illinois,
Indiana, Michigan, Ohio, Pennsylvania, and Wisconsin). These states are
disproportionately dependent on manufacturing jobs. In 2022, these states had
27.1 percent of US employment in manufacturing but just 18.4 percent of US
employment overall. The Rust Belt’s employment share of the announcements is
just 14.8 percent, and demand for manufacturing employment in these states due
to the announcements is projected to increase by just 7.3 percent, far less than
the 13.3 percent increase in overall manufacturing demand in the United States
(table 2).27 The announced investments are projected to create 251,448 jobs, equal
to 17.3 percent of the 1.45 million manufacturing jobs these states lost between
2000 and 2022. For the Rust Belt, these investments will not represent a rebirth.
Electoral Considerations
Most Republicans opposed the IRA, which was responsible for the clean-tech
dimension of these investments, and only 24 House Republicans supported the
CHIPS Act. Despite Republicans’ lack of support, the announced investments
have not been biased toward states that voted for either presidential candidate
in 2020. The states that President Biden won in 2020 accounted for 53.6 percent
of manufacturing employment in 2022 and 63.5 percent of the manufacturing job
decline between 2000 and 2022. The employment share of these states from the
announced investments would equal 50.3 percent (see table 2).
    Investments are more concentrated in swing states (defined as states in
which the margin of victory in the 2020 election was less than 3 percent).
Three of these states (Arizona, Nevada, and Georgia) are among the top 10
in terms of percentage increases in manufacturing employment tied to the
announcements; North Carolina, another swing state, with a 19.4 percent
increase in manufacturing employment tied to the announcements, ranked
11th (table 1). Together these four states, which are sometimes the focus of
an electoral strategy emphasizing swing Sun Belt states, have 49 votes in the
electoral college.
    The manufacturing employment increases associated with the
announcements are smaller in Rust Belt swing states, such as Michigan
(10.2 percent), Pennsylvania (1.1 percent), and Wisconsin (1.0 percent), which
together account for 44 electoral college votes.
    Whether these investments will have a political impact remains to be seen.
The investments could increase support for politicians who voted for the
programs, in which case Democrats would benefit. But they could also increase
support for incumbents, in which case the effects could be more balanced. If
these investments are to help the administration politically, they point to focusing
on their relatively large effects in the marginal western and southern states rather
than Rust Belt states, especially Pennsylvania and Wisconsin.
    It is also possible that the political impact of these policies could differ
from their economic impact. Autor et. al (2024) report that while the economic
impact of President Trump’s tariffs on China had few positive effects, they did
strengthen Republican support politically. Thus similarly, the expressive role of
these programs in demonstrating concern for workers could ultimately be more
powerful than the purely economic impact.
28   It uses a comprehensive measure of the US labor force that includes not only nonfarm wage
     and salary workers but also workers in agriculture and the self-employed.
29   Bureau of Labor Statistics, Employment Projections 2022–2032, September 6, 2023.
PB 24-12 | OCTOBER 2024                                                                                      14
Table 3
Projected changes in US employment between 2022 and 2032, by sector
                                                                                       Change in number of
Sector (NAICS code)                                                                    jobs (thousands)
Total 4,665.5
motor vehicle parts, which constituted the bulk of the 87,700 jobs projected
to be lost in transportation equipment manufacturing, presumably because of
EVs. Employment was also expected to fall in printing (63,400 jobs lost), paper
(44,100), furniture (29,800), primary metals (28,500), textiles (22,000), and
apparel (7,700). Thus, although, with the notable exception of transportation,
the sectors targeted by the Biden program with the CHIPS Act were expected to
show some employment growth, these gains were offset by more pervasive jobs
losses in the rest of manufacturing.
    These estimates of manufacturing employment growth could be too low, but
the analysis points to the perils in proclaiming a manufacturing “renaissance”
based on selective anecdotes from a few industries and projects rather than a
comprehensive consideration of the entire sector.
CONCLUSION
Although it will spur rapid manufacturing employment growth in nontraditional
locations, the Biden programs have not created a broad renaissance in US
manufacturing, and they are unlikely to do so in the future. The effects of
the programs on the Rust Belt states that experienced large manufacturing
employment losses since 2000 are likely to be modest, and the impact will not
significantly change the sectoral composition of the US labor market.
    Manufacturing still has an important role to play in providing the goods
necessary to rebuild US infrastructure, promote the digital revolution, and ease
the transition to a decarbonized US economy. But because of its relatively
small overall employment share—and the growing bias toward hiring more
educated workers—the sector no longer provides noncollege workers with the
opportunities it used to. Although the Biden programs may achieve important
social objectives, they are therefore unlikely to improve the opportunities
for most workers without college degrees or help most of the country’s
disadvantaged places.
    Additional announcements are likely as grants from the Biden programs are
awarded. The results from updated numbers are unlikely to change the central
conclusion of this analysis, however. Even if additional announcements were to
increase the manufacturing employment estimates used here by an additional
50 percent, manufacturing’s role in overall US employment would remain very
modest. Additional policies with a focus far wider than manufacturing are
needed to improve the prospects of non-college-educated workers and left-
behind communities.30
REFERENCES
Autor, David, Beck Anne, Dorn David, and Hanson Gordon. 2024. Help for the Heartland?
   The Employment and Electoral Effects of the Trump Tariffs in the United States. NBER
   Working Paper 32082. Washington: National Bureau of Economic Research.
Bistline, John, Neil Mehrotra, and Catherine Wolfram. 2023. Economic Implications of the
     Climate Provisions of the Inflation Reduction Act. Brookings Papers on Economic
     Activity (Spring). Washington: Brookings Institution.
30 For proposals for some these policies, see chapter 11 of Lawrence (2024).
PB 24-12 | OCTOBER 2024                                                                 16
Langdon, David, and Rebecca Lehrman. 2012. The Benefits of Manufacturing Jobs.
   Washington: US Department of Commerce.
Lawrence, Robert Z. 2020. Trade Surplus or Deficit? Neither Matters for Changes in
   Manufacturing Employment Shares. PIIE Working Paper 20-15. Washington: Peterson
   Institute for International Economics.
Lawrence, Robert Z. 2024. Behind the Curve: Can Manufacturing Still Provide Inclusive
   Growth? Washington: Peterson Institute for International Economics.
© 2024 Peterson Institute for International Economics. All rights reserved.
This publication has been subjected to a prepublication peer review intended to ensure
analytical quality. The views expressed are those of the author. This publication is part
of the overall program of the Peterson Institute for International Economics, as endorsed
by its Board of Directors, but it does not necessarily reflect the views of individual
members of the Board or of the Institute’s staff or management.