CAMBRIDGE STUDIES IN

CHINESE HISTORY, LITERATURE AND INSTITUTIONS

General Editors
PATRICK HANAN & DENIS TWITCHETT

Taxation and Governmental Finance

in Sixteenth-Century Ming China
 Other books in the series

GLEN DUDBRIDGE: The HsUyu Chi: A Study of Antecedents

to the Sixteenth-Century Chinese Novel
STEPHEN FITZGERALD: China and the Overseas Chinese: A Study
of Peking's Changing Policy, 1949-1970
CHRISTOPHER HOWE: Wage Patterns and Wage Policy in
Modern China, 1919-1972
DIANA LARY: Region and Nation: The Kwangsi Clique
in Chinese Politics, 1925-1937
 Taxation and
Governmental Finance in
Sixteenth-Century
Ming China
by
RAY HUANG
Professor of History, State University College,
New Paltz, New York

CAMBRIDGE UNIVERSITY PRESS

 CAMBRIDGE UNIVERSITY PRESS
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, Sao Paulo, Delhi

Cambridge University Press

The Edinburgh Building, Cambridge CB2 8RU, UK

Published in the United States of America by Cambridge University Press, New York

www.cambridge.org
Information on this title: www.cambridge.org/9780521104876

© Cambridge University Press 1974

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 1974

This digitally printed version 2009

A catalogue record for this publication is available from the British Library

Library of Congress Catalogue Card Number: 73-79311

ISBN 978-0-521-20283-1 hardback

ISBN 978-0-521-10487-6 paperback
 Contents

List of figures page vii

List of tables viii
Preface by D. C. Twitchett x
Acknowledgements xii
A note on weights and measures xiv
The Ming emperors xv
Map of Ming provinces xvi

1 Fiscal organization and general practices 1

I Governmental organization 4
II Rural organization and basis of taxation 32

2 The heritage of the sixteenth century and major fiscal problems 44

I The level of state income and modifying factors 46
II Land and population data 60
III The maintenance of the army 63
IV The monetary problem 69

3 The land tax—(i) Tax structure 82

I The complexities of the tax structure 84
II Regional variations 98
III The service levy and its partial absorption by the
land taxes 109
IV Further modifications of the tax structure 134

4 The land tax—(ii) Tax administration 141

I Tax administration by the local government 142
II Factors affecting the general administration 154
[v]
vi Contents
III The level of collection page 163
IV Disbursement of tax income 175
V A final analysis of the land tax system 182
5 The salt monopoly 189
I The mechanism of the salt monopoly 189
II Governmental control and manipulation 195
III Administrative cycles in the sixteenth century 205
IV Revenues, salt prices, and their effects on consumers 212
V Responsibility for the failure 221
6 Miscellaneous incomes 225
I Revenues from commerce and industry 226
II Administrative income 244
III Cash income from commutation of services and supplies 252
IV Non-cash income 257
V Summary of miscellaneous incomes 262
7 Financial management 266
I The ministry of revenue in the sixteenth century 268
II Inter-provincial and inter-ministerial administration 277
III Appropriation of military supplies 284
IV Fiscal retrenchment under Chang Chii-cheng 294
8 Concluding observations 306
I The risks of over-simplification 307
II Ming financial administration and its place in Chinese
history 313
List of abbreviations 324
Appendixes
A. Landed properties not subject to regular taxation 325
B. Customary fees and extra services collected by the
magistrate of Shun-an county, Chekiang, 1561 327
C. Barter rates and excise tax on each yin of salt, 1535 328
D. Partial returns of the land survey of 1581, as recorded
in Ming Shih-lu 329
Notes to the text 331
Bibliography 361
Glossary index 377
General index 383
 Figures

1 Ming provinces page xvi

2 The Grand Canal (Ts'ao Ho), ca. 1610 54
3 Land tax structure in the late sixteenth century 83
4 Diagram showing distribution of tax revenue 182
5 Northern frontier army posts in the late sixteenth century 289

[vii]
 Tables

1 Financial burden to the taxpayer of tax grain delivered or

commuted, for four prefectures in South Chihli, ca. 1585 page 101
2 Service levy collection in Chang-chou prefecture, Fukien, 1572 126
3 Service levy distribution in six counties in the southeastern
provinces, 1572-1621 127
4 Service levy distribution in three counties in north
China, 1608-20 130
5 Quotas of fish duty and local business tax of Yung-chou
prefecture, Hukwang province, 1571 140
6 Provincial tax quotas in piculs of grain based on the
tabulation of 1578 164
7 Estimated land tax rates of Hang-chou prefecture in 1572 167
8 Estimated land tax rates of Fen-chou prefecture in 1608 169
9 Estimated land tax collection and disbursement in piculs of
grain, ca. 1578 177
10 Distribution of land tax revenue from Lin-fen county
1591: in grain 178
11 Distribution of land tax revenue from Lin-fen county
1591: in silver 178
12 Distribution of land tax revenue from Wu-hsien county
1575: in grain 180
13 Cash income from salt monopoly, as of 1578 214
14 Estimated annual income of the salt monopoly, ca. 1575-1600 216
15 Miscellaneous incomes, ca. 1570-90 227
16 Collection quotas of eight inland customs stations,
1599-1625 231
17 Collection offish duty, 1578 245
18 Estimated annual proceeds from miscellaneous incomes,
ca. 1570-90 263
19 Reported annual revenues from miscellaneous sources of
income of the T'ai-ts'ang Treasury, ca. 1570-90 265
20 Reported disbursement of the T'ai-ts'ang Treasury, 1551-7 270

[ viii ]
 List of tables ix
21 Reported revenues of the T'ai-ts'ang Treasury, 1567-92 page 270
22 Reported deficits of the T'ai-ts'ang Treasury, 1583-90 270
23 Estimated basic revenues of the T'ai-ts'ang Treasury,
ca. 1521-90 271
24 Cash disbursement in Peking by the T'ai-ts'ang Treasury,
ca. 1580 276
25 Major items of supply disbursed by the fourteen frontier
army posts in 1575 291
26 Revenues of the fourteen frontier army posts, 1578 291
 Preface

When some twenty years ago I finished the first draft of my own study of
T'ang financial administration, I began to undertake the preliminary
reading for a similar study of the Ming period, feeling that it would be
possible to pose and answer many questions which cannot even be
formulated in detail for earlier periods because of the lack of evidence.
I soon found myself frustrated by the complexity of the task. The sheer
bulk of source material which has to be covered is daunting in itself, and
since then far more of the Ming historical literature has been made
generally accessible. But in addition the subject matter proved to be
infinitely more complicated than in the case of the T'ang. The main problem
was that whereas the earlier dynasties had systematically attempted to
impose a uniform and universal set of comparatively simple institutions
throughout their empire, and incorporated these in a tightly drafted
system of centrally codified administrative law, the gradual abandonment
of this concept of uniform government from the late eighth century on-
wards, and the decentralization of control over detailed policy-making and
enforcement, had led by Ming times to a situation of bewildering local
diversity. It had reached the point where in some fields it was no longer
possible to make simple generalizations about the empire as a whole.
Eventually I was side-tracked into other things and abandoned my
plans.
Since Professor Huang and I first met in the early 1960s and began
discussing the subject matter of this book, it became clear that far from
exaggerating the difficulties of the subject, I had underestimated them.
The comparatively straight-forward and clear-cut picture given in the
'Shih-huo chih' sections of the Ming dynastic history, from which I had
gained my own first overall impressions of the subject was gradually
transformed into an infinitely complex and ill-defined mosaic of often
apparently unrelated detail.
Although many detailed articles and a few important large-scale studies
of individual aspects of Ming finance have appeared during the past
decades in China and Japan, Professor Huang's is the first attempt in any
[x]
 Preface xi
language to give a general account of financial policy, placing this mass of
newly discovered detail in a broader historical perspective.
The reader may sometimes find that his account of certain aspects of
policy contains apparent anomalies and even internal contradictions. This,
however, reflects the fact that in many fields government policy and local
practice did contain glaring inconsistencies and anomalies. Ming govern-
ment, notably effective as it was in many respects, was not a tidy and uni-
form system, particularly at the local level. The detail in this volume,
overwhelming as it may sometimes appear, is also far from complete. But
at the present primitive stage of our knowledge of the subject, and of the
detailed history of the period, it is important to present the evidence in full,
rather than rushing to erect facile generalizations. The study aims to
present a general framework to which further detail may be related, not to
provide yet another grand historical pattern.
It is to be hoped that not only will it stimulate further detailed studies
of financial history, but that it will also provide reliable guidelines on these
aspects of government policy for the growing numbers of young historians
working on late Imperial China. In particular it should help scholars
working in the important area of local history in Ming and Ch'ing times
to interpret the perplexing masses of statistical and administrative detail
included in local gazetteers and elsewhere.
DENIS TWITCHETT
1973
 Acknowledgements

It is my sincere hope that the publication of this book will bring some
measure of satisfaction to those who, during the past several years when it
was under preparation, generously helped me in many different ways:
provided inspiration and guidance, supplied bibliographical materials,
answered specific questions, read parts of the manuscript and offered
valuable criticisms. Any errors which remain are of course my own.
My thanks are due to Dr William Theodore de Bary, Vice-President and
Provost, Columbia University; Professor Wing-tsit Chan, Chatham
College; Professor Albert Feuerwerker, University of Michigan; Professor
Ping-ti Ho, University of Chicago; Professor James T. C. Liu of Princeton
University; Professor John Meskill, Barnard College; Professor F. W.
Mote, Princeton University; Dr Joseph Needham, Master of Gonville and
Caius College, Cambridge; Professor Morris Rossabi, Case Western
Reserve University; Reverend Henry Serruys; Mr Wei-ying Wan of the
East Asian Library at the University of Michigan; Dr Eugene Wu of the
Harvard-Yenching Library and his staff, among them in particular
Mr George C. Potter; Dr K. T. Wu, Division of Orientalia of the Library
of Congress; and Professors Lien-sheng Yang and Ying-shih Yu, both of
Harvard University.
Professor John K. Fairbank and his committee at the East Asian
Research Center, Harvard University, gave me a fellowship grant in 1970
when I was completing the first draft. In particular I wish to express my
gratitude to Professor Fairbank for his kindness in advising me how to
deal with the subject matter. In his opinion, a study in depth of a specific
topic may, by extension, be related to the study of other topics within the
same area, without the necessity of covering these in similar detail.
Originally I had intended to write on the fiscal administration of the entire
Ming period, but the material had a constant tendency to get out of
control, and on Professor Fairbank's advice I eventually arrived at the
present format. Having benefited from his insight, I record this here with
appreciation, in the hope that others may continue to benefit from it.
For a great many years Professor and Mrs Chao-ying Fang have been
[xii]
 Acknowledgements xiii
giving valuable guidance to scholars in the field of Chinese studies. Having
the good fortune to work closely with them in 1967 on the Ming Biographi-
cal Project, I took every advantage of their immense knowledge of the
Ming dynasty and their willingness to share it with others. Dr L. Carrington
Goodrich, Professor Emeritus at Columbia University and editor of the
Ming Biographical Project, has at my request read practically every word
I have written for several years. His criticism is always offered in a tone of
affection. Charles O. Hucker, Professor of History at the University of
Michigan, is a special friend. I made his acquaintance by correspondence
more than twelve years ago when I was a Ph.D. candidate, and have never
ceased to be grateful for his continuing help since then. Professor Denis
C. Twitchett, Cambridge University, has helped me to prepare the final
draft for publication and graced the present volume with a preface. My
indebtedness to them all is great.
The basic research for this book was done several years ago with free
time obtained through fellowship grants. An area study fellowship jointly
sponsored by the American Council of Learned Societies (ACLS) and the
Social Science Research Council (SSRC) covered one term in 1966. Summer
research grants were made available by the Center for Chinese Studies at
the University of Michigan and the Research Foundation of the State
University of New York. My grateful acknowledgement to them, however,
should not be construed to mean that my sponsors endorse the opinions
in this book.
I am grateful for the assistance rendered by Miss Hilary Beattie,
Research Fellow at Newnham College, Cambridge. She has made a signifi-
cant contribution to improving the style of my writing; I alone am respon-
sible for the inadequacies inherent in the original draft. Although I have
retained American spelling in this volume, I hope that the text will be
equally acceptable to readers on both sides of the Atlantic.
Closing this acknowledgement, I would like to extend my deepest
appreciation to my wife, Gayle, who for seven years has shared my hopes
and endured the hardships that I brought to the family through my
interest in historical research. Her enthusiasm is always a source of my
strength. Only after completing this book did I get in touch with
Dr Ch'iian Han-sheng. Several articles recently published by him in the
Journal of the Institute of Chinese Studies will be of invaluable interest
to readers of this volume. While I am most grateful to Dr Chiian for
sending me off-prints, unfortunately it has not been possible to include
the titles in the present bibliography.

Mulberry Close, Cambridge R. H.

12 July 1973
 Weights and measures

The Ming tried to standardize weights and measures. Though metal scales,
weights, and measuring receptacles were issued by the ministry of works
no examples have yet been discovered. So far the closest thing found is an
ivory scale made in the Chia-ching period, which, being an engineering
scale, differs from the fiscal standard.
The following equivalents are based on specialist studies of Ming paper
currency and copper coins and weights and measures, and are known to be
relatively accurate, although their absolute accuracy cannot be guaranteed.

1. Measurement of Length. The cKih, or 'foot', is approximately 12.3

inches.
2. Weights. The chin, or 'catty', is approximately 1.3 pounds. It is
divided into 16 Hang, or 'taels', each of which of which is about 1.3 ounces.
3. Capacity. The dry measuring unit is generally known as the shih,
though some scholars, such as Rieger, Sun and de Francis, prefer to
romanize the term as tan or dan. Here translated as 'picul', it equals about
107.4 liters.

Except for the division of the catty into 16 taels, the fiscal units always
followed the decimal system. Ming accounts do not use a decimal point
but enumerate the fractions of the basic fiscal units by name. Each fraction
of a basic unit has its own special term. For example, a millionth of a tael
is called a wei, and a trillionth a mo. All these cumbersome figures have
been converted into the basic units and whenever 'decimal points' and
'decimal digits' are mentioned they refer to these converted figures.
'Billion' is used to mean one thousand million, and 'trillion' (as in the
preceding paragraph) one million million.

Numbers in the text are, in general, spelled out up to 100, but figures
are used for percentages, units of currency and series of numbers.
[xiv]
 The Ming emperors

Temple name Reigned Era name

T'ai-tsu
Hui-ti
1368-98
1398-1402
Hung-wu
Chien-wen
mm
T'ai-tsung, Ch'eng-tsu 1402-24 Yung-lo mx
Jen-tsung 1425 Hung-hsi &m
Hsiian-tsung 1425-35 Hsiian-te mm
'Km
Ying-tsung 1435-49 Cheng-t'ung itm
Ching-ti 1449-57 Ching-t'ai
Ying-tsung (restored) 1457-64 T'ien-shun mm
Hsien-tsung 1464-87 Ch'eng-hua & Ik
Hsiao-tsung 1487-1505 Hung-chih
Wu-tsung 1505-21 Cheng-te
Shih-tsung 1521-66 Chia-ching
Mu-tsung 1566-72 Lung-ch'ing 81
Shen-tsung 1572-1620 Wan-li
Kuang-tsung 1620 (one month) T'ai-ch'ang nm
Hsi-tsung 1620-7 T'ien-ch'i m^
Chuang-lieh-ti 1627-44 Ch'ung-chen

[XV]
 MONGOLIA
LIAOTUNG
(MANCHURIA)j

/ NORTH
T'ai-yuan « CHIHLI

SOUTH
/ CHIHLI /"Nanking

/ KWEICHOW }V~*y~ \
Kun-ming / /-N ^-—^"'v/'

<~l~P KWANGTUNG \

Chinese frontier
provincial boundaries
-n-ruxn Great Wall of China

FIG. 1. Ming provinces.

[xvi ]
 1
Fiscal organization and
general practices

Most governmental institutions under the Ming preserved certain features

reminiscent of previous dynasties, the T'ang, Sung and Yuan, yet at the
same time developed distinctive characteristics of their own. The dynasty's
fiscal administration was no exception. Such Ming practices as the employ-
ment of supervising secretaries as fiscal auditors, the organization of six
ministries, the circulation of government notes, the utilization of the Grand
Canal as a trunkline connecting the north and the south, the trading of tea
for horses with nomadic groups, and bartering salt under governmental
monopoly for military supplies in the frontier region, were clearly copied
from earlier dynasties. On the other hand the Ming administration saw
more concentration of power in the emperor's hand, smaller administrative
overheads, a greater reliance on agrarian sources for state income and
a stricter control over maritime trade which virtually committed China to
a state of limited seclusion.
The imitation of the past was not totally unjustified. Infiscalmatters all
dynasties in post-T'ang China faced the same fundamental problems. To
secure the durability of the empire, imperial control over the nation's
financial resources had to be firm and thorough. Yet the empire's vastness
and its regional diversities worked against a policy of centralization, as
did pre-modern methods of transport and communications. The gap be-
tween aims and the limited available means of achieving them was not
easily bridged. In these circumstances it was not surprising that all dynasties
tended to draw on past experience.
The particular style of Ming administration, geared to the concept of
a village commodity economy, could be labelled conservative, and could
indeed in the light of economic development in China in the sixteenth and
seventeenth centuries, be regarded as anachronistic. This conservative
outlook was, however, the inevitable outcome of political centralization
in a vast country under pre-modern conditions.
It must be noted that economic development in China before the Ming
showed considerable regional imbalance. It was based on a few handicraft
industries and on foreign trade, which affected only certain geographical
I [1] HTF
2 Fiscal organization and general practices
areas. As the Ming administrators saw it, to promote those advanced
sectors of the economy would only widen the economic imbalance, which
in turn would threaten the empire's political unity. It was far more desirable
to keep all the provinces on the same footing, albeit at the level of the more
backward sectors of the economy. The Ming fiscal policies were largely
determined by this attitude.
Such an approach seems absurd to the modern historian, in that it
sacrificed China's long-term economic growth for the sake of a short-term
political aim. Ming administrators, however, did not have the foresight
to visualize the contributions that industry and commerce could ultimately
render to a modern state. For the small states of Western Europe govern-
ment promotion of industry and commerce could facilitate a swift capitalist
transformation of their economies. In China, which was vastly larger, it
could not have effected such rapid and far-reaching change. Moreover,
unlike the emerging national states in Europe or even the rival feudatories
in Japan, Ming China never placed itself on a competitive basis with its
neighbors; it could therefore afford the price of backwardness. Within
their own horizons Ming administrators saw no compelling reason to
revise their policies. On the contrary, there was good reason for them to
proceed on their traditional course, in perfect accord with the traditional
Confucian doctrine that agriculture alone provided the economic foun-
dation of the state.
The tragedy was that despite their striving for simplicity and uniformity,
and their willingness to base their policies on the minimum level of
economic activity, the Ming administrators never fully accomplished their
purpose. China's internal diversity simply defied any uniform control
from the center, especially in the area of fiscal administration. In agri-
culture, there were wide regional variations in climate, soil, topography,
and labor supply, not to mention the enormous variety of crops, differing
market conditions, multiple kinds of land tenure and leases, and the
diverse standards of units and measurements across the country. Legis-
lating from the capital, the Ming court could not take all these factors
into account. Without attempting to reconcile these differences it was to
find that to proclaim a uniform land tax law was one thing and to have it
applied in every corner of the empire was another.
It was generally assumed by many late Ming writers that in the early
years of the dynasty the empire's fiscal programs had been carried out in
full, and that only toward the end did the administration become corrupt.
The assumption is only partially true. There is sufficient evidence that even
at the very beginning government regulations were compromised, imperial
decrees discounted, and official data to some extent tampered with. This
was not necessarily the act of dishonest officials. In the early stages it was
due largely to the fact that fiscal schemes conceived from above failed to
 Fiscal organization and general practices 3
correspond to conditions at lower levels, or else the desired degree of
centralization exceeded the technical capacities of contemporary govern-
ment. As a result, imperial laws had to be adjusted. Local manipulations
and revisions were necessary. Indeed, in the dynasty's later years deviation
from prescribed procedures became a general practice; by then widespread
abuses were only to be expected.
This lack of exactitude in the fiscal machinery had numerous con-
sequences. An outstanding example was that Ming officials customarily
compensated for shortages in one item of state revenue with funds and
materials derived from another. What we regard as the salt gabelle in
Ming times actually contained portions of the land taxes. After the middle
period the land taxes also became inseparable, if not totally indistinguish-
able, from several other sources of income. Government income and
expenditures in Ming times can be compared to water channels in a swamp,
in that they had constant tendency to diverge and re-converge.
These complexities do not facilitate our task. Most of the Ming institu-
tions do not lend themselves to precise classification or definition. They
were in flux, their changes being determined in the main by external
environment rather than internal development. They owed more to the
manipulations and compromises of the administrators than to any logical
growth of their own. In this book, the greatest difficulty has been to confine
the material under topical headings without creating a misleading im-
pression that a jigsaw puzzle is a checkerboard. We have therefore, chosen
to rely more heavily on description rather than on tabulation of data. In
places the logical sequence of presentation is slightly modified to suit the
material. Cross references in parentheses mark topics treated in different
chapters or different sections of the same chapter. While this is not the
ideal way of writing a financial history, we hope that it will permit a more
realistic view of the subject.
Similarly a mixed approach is adopted in the present chapter. It may be
noted that throughout the Ming period, except for the reign of Hung-wu,
reorganization of officialdom was seldom attempted. In its span of 276
years the dynasty saw China's commodity economy transformed into
a money economy, taxes in kind and statutory labor largely replaced by
silver payments, and a conscripted army gradually replaced by a recruited
army; yet, surprisingly, few fiscal offices were created in the middle period
of the dynasty and still fewer offices abolished. This was possible because
the functions of government agencies were not always fixed by stature, but
were rather based on customary usage. Administrative procedures more-
over were rarely superseded by new laws. Old and new statutes alike were
simultaneously preserved. The obsolete sections were simply disregarded
and the surviving clauses selectively applied to individual cases according
to the circumstances. All offices in fact went through some kind of evolu-
4 Fiscal organization and general practices
tionary process, their functions being readjusted from time to time, some-
times so smoothly that the change went unnoticed even by contemporaries.
This introductory chapter on the fiscal structure must therefore deal with
its form as well as its dynamics.

I GOVERNMENTAL ORGANIZATION
The nature of the Ming monarchy and its role in public finance
Under the Ming system there was no central authority administering the
empire's finance other than the emperor himself. The premiership was
abolished in 1380 and never revived. The grand-secretaries limited their
functions to rescript-drafting. Although they were consulted by the
emperor and did participate in decision-making, they were never officially
accorded any power. The minister of revenue supervised routine fiscal
matters, yet he could never act without prior imperial approval. The
emperor, truly an autocrat, received memorials from a large number of
censors, supervising secretaries and ministerial officials, down to division
heads and section chiefs. Suggestions and criticisms on fiscal matters could
be initiated by practically anyone, regardless of his area of specialization
and current assignment. The emperor was also approached by scores of
imperial commissioners, governors and governors-general. Their proposals
and requests could be handed to the ministry of revenue for comment, or
else, when important issues were involved, referred to the Nine Ministers'
Conference for deliberation.1 But the final decision was always the
emperor's own.
In practice, in the course of the dynasty, there were a few persons who
managed to exercise the power reserved to the throne, notably Grand-
Secretary Chang Chii-cheng during the reign of Wan-li, and 'eunuch
dictators' Wang Chen, Liu Chin, and Wei Chung-hsien in the Cheng-t'ung,
Cheng-te, and T'ien-ch'i eras respectively. Yet, although Chang gained his
powerful position through informal arrangements with his fellow officials
behind the back of a youthful emperor, even then the formality of seeking
the sovereign's approval on major and minor matters was not disregarded
(7, m). The eunuchs, with the indulgence of the emperors, violated virtually
every fundamental law of the empire. All of them were later condemned,
or charged with conspiracy. It is noteworthy that none of the three eunuchs
died a natural death and Chang Chii-cheng was posthumously disgraced.
Outwardly the imperial power could not be delegated and never was.
In general, official and office titles in this work follow Charles O. Hucker,' Governmental
Organization of the Ming Dynasty', Harvard Journal of Asiatic Studies, 21 (Dec. 1958),
pp. 1-66. This article is reprinted in Studies of Governmental Institutions in Chinese
History, ed. J. L. Bishop (Cambridge, Mass. 1968). But the Ming emperors are referred
to throughout by their reign titles rather than temple names.
 1,1 Governmental organization 5
Several emperors ascended the throne when they were minors but no
formal regency was ever established for them.
In the realm of fiscal administration some of the matters brought to the
emperor's attention could be exceedingly trivial. Even minor issues - such
as the relocation of a particular business tax station, or from which pro-
ductive area a particular county was to draw its salt supply, or how many
rolls of silk were to be awarded to a foreign tributary mission, - were all
submitted to the emperor for final approval. The sovereign was expected to
attend to all administrative details, the precedent having been established
by the dynastic founder. In one eight-day period Hung-wu is said to have
received 1,660 memorials in which 3,391 issues were discussed.2
Apparently the founding emperor felt that this personal control over
what seemed to be insignificant matters was justified. In the early years of
the dynasty taxes were collected almost exclusively in kind. The govern-
ment had to avoid the accumulation of tax proceeds at the intermediate
level, in order to avoid over-burdening the service facilities, the weakest
section of the whole fiscal establishment. The solution was for each
revenue-gathering agency to make delivery directly to a dispensing agency;
items of income and expenditure were matched up so as to cancel each
other out. Under this system bulk deliveries were few, but there were
numerous modest-sized consignments of materials and goods moving from
one end of the empire to another. The disposal of each of those small con-
signments, therefore, could not be treated lightly, for each was a thread in
the fiscal fabric. If control over any one of them were relaxed it would
permit precedents to be established that might ruin the whole system. In
later years, owing to deviations from fiscal regulations at the local level,
imperial control over the nation's financial affairs slackened considerably,
but the custom of the emperor attending to minor matters persisted. The
assertion of the imperial prerogative was now limited, however, to those
selected cases which were actually brought to the court's attention.
It seems outwardly strange that, even though after Hung-wu few
emperors effectively exercised their legislative power, and no sweeping
fiscal reforms were ever proclaimed by them, yet toward the end of the
dynasty, taxes in kind and labor services were largely commuted to silver
payments, the fiscal responsibilities of the military bureaucracy were
reduced to a minimum, the ranks of the army were filled with recruited
soldiers and army posts were paid from the capital. It may be added that
even the use of silver in private business transactions, which came to be
universally accepted in the later period, had been forbidden in earlier
years. The answer to the puzzle is that even though ostensibly the succeed-
ing emperors had in general refrained from revising the dynastic founder's
fiscal legislation, they did authorize occasional exceptions to the established
procedures. Such changes followed a pattern. They usually started with
6 Fiscal organization and general practices
a petition from a lower-echelon bureaucrat, the requested exception
applying to his office alone. The emperor's approval, most likely granted
along with other routine matters, then established the required precedent.
Sooner or later similar petitions were submitted and approved until the
original exception had become a common practice. Thereafter no formal
petition was required; either an imperial decree directed the remaining
offices to follow suit or the latter proceeded to carry out the reform without
explicit approval. In this way a major reform might take several decades
to accomplish. The chiin-yao method (3, m), for instance, took almost
fifty years from the time of its earliest experimental use to become generally
accepted. The Single Whip Reform took even longer, and even at the
collapse of the dynasty it had not attained a final, definite form.
The removal of earlier restrictions, especially when the restrictions could
not be effectively enforced, took much less time. An example is the relaxa-
tion of the use of silver as a common medium of exchange. It started with
a petition from a prefect in Kwangsi whose original proposal applied only
to the legalization of copper cash in private trade (2, in). Apparently, the
precedent-establishing effect quickly caught up with the approval of this
petition and soon the earlier decree forbidding the use of silver bullion,
too, became defunct.3 This exercise of legislative power by the Ming
emperors sometimes had a very similar effect to the judiciary decisions
handed down by a supreme court in the Western world.
Another paradox in the administration was that in spite of their auto-
cratic power many later Ming emperors found themselves in a defensive
position vis-a-vis the bureaucrats who raised questions concerning their
spending habits. There were several reasons for this. One was that in the
eyes of the Confucian bureaucrats the private life of the emperor was
a matter of public concern and subject to discussion if not criticism.
Another was that since the early sixteenth century few Ming emperors,
with the notable exception of the last emperor, Ch'ung-chen, had shown
any genuine interest in state affairs. Some of the emperors were either too
indolent to care or too incompetent to make a decision. Thus disagree-
ments between the monarchs and their courtiers on strictly public matters
were few. Controversies were usually related to the emperor's private life.
Thirdly, according to the theory that all under the heaven belonged to the
emperor, the Ming system made no clear-cut distinction between state
income and the emperor's personal income, and between governmental
expenditure and the emperor's expenditure. The sovereign's personal
spending was therefore closely related to public finance. Frequently the
issues raised by the courtiers concerned exorbitant palace expenses,
excessive procurement programs, the misconduct of eunuch commissioners
in charge of those programs, and land grants to the emperor's favorites
and relatives. But since the final authority rested in the sovereign, bureau-
 1,1 Governmental organization 7
cratic action was limited to remonstrance, resignation, attempted impeach-
ment of those who carried out the emperor's orders, and exaggeration of
portents as heaven-sent warnings to the wayward emperor. When all these
failed there was no recourse left. Furthermore, in carrying out these steps
the bureaucrats always risked reprisals from the throne.
With the exception of the dynastic founder, who was self-taught, all
Ming emperors received a Confucian education.4 Thus indoctrinated, they
were expected to follow the frugal standards of their forefathers, to respect
public opinion, and to remain above malice and self-indulgence. The
actual effect of these moral precepts is difficult to gauge. In view of the
unattractive personalities of most Ming emperors it is easy for us to
conclude that the so-called Confucian morality was no more than a fiction.
Yet considering the unlimited power at the disposal of the crown, one
might wonder what would have happened had this power never been
subject to any moral restraint at all.
Grand-secretaries, outwardly powerless, could also exercise considerable
positive influence from behind the throne. It is true that newly-appointed
grand-secretaries serving mature sovereigns were in general more inclined
to be subservient. But by tradition the grand-secretaryship under the
Ming was a long-term appointment. In normal circumstances grand-
secretaries were expected to hold office for life. Many of them did so and
served under several emperors. Because of their prestige and the public
trust which they enjoyed, they could act as intermediaries between the
emperor and the ministerial officials, thus providing a stabilizing force in
the court. There were numerous examples of such senior statesmen in the
early Ming period. Even in the later years towering figures of this kind were
by no means lacking. Chang Chii-cheng's case may be somewhat too
complicated to accord exactly with this description. But Yang T'ing-ho
(in office, 1507-24), who introduced a drastic austerity program in the
court upon the accession of Chia-ching, and Yeh Hsiang-kao (in office,
1607-24), who masterminded the transfer of some 7 million taels of silver
from the emperor's personal treasury to replenish the empty state coffers
in the T'ai-ch'ang-T'ein-ch'i periods, show well how grand-secretaries
could be instrumental in carrying out popular policies and restoring public
confidence.5
Nevertheless, none of the deterrents to unlimited exercise of imperial
power - including Confucian morality, reverence for the standards set up
by imperial ancestors, public opinion, or the influence of senior statesmen
- had the effect of law. If an emperor chose to defy all these and was
determined to exercise his absolute power to the full, there was no way of
checking him. This situation actually prevailed during the reign of Wan-li.
In the 1590s the emperor dispatched eunuchs to collect business taxes in
the provinces, superseding the civil officials (7, m). Many courtiers, after
8 Fiscal organization and general practices
remonstrating to no avail, handed in their resignations. The emperor was
so annoyed by the opposition that he ignored their resignation requests.
Some officials subsequently left without authorization and the emperor, in
turn, kept the vacant offices unfilled. The sequence of events precipitated
what amounted to a 'constitutional crisis', and the deadlock remained
unbroken until the sovereign's death in 1620.

Palace expenditures and the eunuchs

The complexities of palace expenditures in Ming times can best be under-
stood by first visualizing the layout of the Ming palace complex.
The center core of the palace was the Forbidden City, a walled and well-
guarded compound of about a quarter of a square mile of elegant archi-
tecture, consisting of the residential quarters of the imperial household,
many ceremonial halls, studies, libraries and the office of the grand-
secretaries. Encircling the Forbidden City was the Imperial City, also
walled and restricted, which was in turn surrounded on all sides by the city
of Peking proper. Ministerial offices were located outside the walls of the
Imperial City. Inside the wall, encompassing an area of three square miles,
was a huge maintenance area.6
Installations inside the Imperial City included treasuries, receiving and
storage depots, material processing and manufacturing plants such as
a bakery, confectionery, dispensary, distillery, leatherworks, tailor's shop,
silversmith's shop, printing shop, dyeing shop, etc. There was even an
arsenal which manufactured firearms and gunpowder. Most of those
supply agencies were manned by eunuchs, some of them exclusively, others
with a handful of civil service officials nominally in charge. Next to
them were other eunuch agencies which, drawing supplies from the afore-
mentioned warehouses and shops, provided household services in the
Forbidden City. Offices of the civil service physically located inside the
Imperial City included the supervising secretariat, which had the function
of inspecting the supply depots, and the court of imperial entertainments,
which was nominally subordinate to the ministry of rites.
The total number of these offices and service and supply installations was
over fifty. Apart from the eunuchs, artisans performing statute labor and
their hired assistants constituted a major portion of the palace population,
which, even in the later fifteenth century, exceeded 100,000. Because the
salaries of Ming officials were only nominal (2, i), the amount of cash
paid to the functionaries by the ministry of revenue remained fairly small
and insignificant, but it appears from the accounts that prodigious quanti-
ties of grain were rationed to the palace personnel, including soldiers called
to perform constructional labor in the compound.
While the service agencies were primarily concerned with serving the
 1,1 Governmental organization 9
imperial household, their operation was not totally separated from
governmental functions. The enormous quantity of silk fabrics processed
in the complex was issued mainly to clothe the palace women, but was also
awarded to foreign tributary missions and the civil and military bureau-
cracy, sometimes in large quantities. The imperial smithy produced silver
utensils for the Forbidden City, but also cast metal frames for important
state papers. The court of imperial entertainments provided delicacies for
the emperor's table, but also prepared state banquets, served dinner to
officials on special duty, rationed meat and wine to them, and provided
food for sacrifical services. As for manufacturing military ordnance, there
is no need to stress that it was a matter of state function rather than palace
consumption. All these shops and depots must have used up considerable
quantities of materials for their own plant maintenance.7
In brief, the Ming organization followed the familial principle so closely
that the palace and government were virtually one and indivisible. The
underlying principle was that the emperor shared his material goods with
his bureaucrats. It also involved something akin to the inseparability of
church and state. One of the primary functions of the Ming monarchy was
to perform endless court rituals in an ostentatious setting. The construction
of palace buildings and the numerous ceremonies both in and outside the
palace, including imperial accessions, weddings, and many other such
activities, always incurred immense costs. It was difficult to ascertain
which was the emperor's personal spending and which was state expenditure.
No ministry of imperial household was ever established; no accounts of
the net expenses of the Forbidden City were ever prepared or published.
Figures referred to as palace expenses came from two sources. One was the
delivery reports and procurement orders of the several ministries which
submitted materials and goods to the palace compound; the other was the
papers of the supervising secretaries who audited the accounts and made
inventories of the warehouses. No complete set of data is currently avail-
able. The customary method of estimating palace expenses by Ming officials
was to sort out a few essential items for consideration, such as cotton cloth,
silk fabrics, tea, wax (used for candle manufacture), and dyes, all of these
in large quantities. The merging of palace maintenance with public finance
was undoubtedly a serious handicap to fiscal operations.
More must be said about the organizational weakness of the logistical
system. Even though the palace warehouses were designated as 'belonging'
to the ministries of revenue, works, and war, in practice the ministerial
officials were only responsible for maintaining the levels of their supplies.
They had little authority to distribute the materials stored in the palace
grounds, that being a prerogative of the sovereign. In managing the ware-
houses, the rule was that the civil officials kept the books and the eunuchs
kept the keys.
10 Fiscal organization and general practices
The ministries held their silver bullion outside the palace. The only
agency that received silver payments in the Imperial City was the Inner
Ch'eng-yun Treasury. Each year the ministry of revenue delivered to it
about one million taels of silver derived from land taxes, which was ear-
marked as the emperor's privy purse (see Gold Floral Silver, 2, i). Rents
from the imperial estates which constituted the cash incomes of the dowager
empresses were also collected by the ministry and handed over to the Inner
Ch'eng-yiin Treasury. The treasury had little to do with palace maintenance,
however. It was in reality an intermediate agency. The funds received by
it were usually spent on gifts and personal grants, contributions to religious
institutions, and purchases of gems and curios. Any surplus was transferred
to the Tung-yii Treasury inside the Forbidden City.8
The Kuang-huei Treasury, also located inside the Imperial City, was
concerned with petty cash only. All revenues arriving in Peking in copper
coins and paper currency had to be surrendered to this coffer. Its operation
caused considerable irregularity in the fiscal system. For instance, normally
when the inland customs duties were collected in silver, the proceeds were
under the jurisdiction of the ministry of revenue. But if in a particular year
the deposits of the Kuang-huei Treasury ran short, the emperor would
then order that the next year a percentage of the customs duties be collected
in coins and paper notes. In the end the palace treasury was replenished,
but at the same time the ministry of revenue lost a portion of its normal
income from this source. The expenditures of the Kuang-huei Treasury
were also irregular; but a major portion of the copper coins and paper
money was distributed by order of the emperor to officials at festival times,
and disbursed as petty cash for lesser purchase orders in the capital.9
After the early fifteenth century the eunuchs were organized into twenty-
four agencies, but no chief eunuch was appointed. The director of cere-
monies was generally regarded as the senior eunuch; but even then at no
time did he assume any responsibility as a fiscal officer.10 It may be said
that the emperor was his own treasurer inside the palace compound.
Eunuchs in Ming times became notorious. Customarily they were dis-
patched by the emperors to the provinces to supervise a variety of state
functions. Yet for much of the dynasty there is no evidence that they
seriously interfered with governmental operations, despite their occasional
abuses of power. Some of them were stationed at the sea ports to receive
tributary emissaries, one was in Nanking to take charge of imperial tombs,
and others attached to the army as personal agents of the crown.11 As long
as the normal operating procedures of the government were not altered
because of the presence of these eunuchs, they could not be considered to
have supplanted the bureaucracy. In the 1590s however the Wan-li
Emperor commissioned them to manage local business taxes and in the
last two decades of the dynasty the Ch'ung-chen Emperor put them in
 1,1 Governmental organization 11
charge of city defense. In these two instances at least the integrity of the
civil service was threatened.
Throughout the dynasty senior eunuchs were frequently dispatched on
procurement missions. Supplies thus procured included palace furniture,
porcelain, and silk fabrics. Palace furniture was made in Nanking; the
procurement usually caused little controversy except for the fact that on
their return trips the eunuchs often demanded extra government transporta-
tion on the Grand Canal to carry their own private cargo. Porcelain was
manufactured in Kiangsi; the items in demand included flower vases, bowls
more than ten feet in diameter, chess games, porcelain screens, sacrificial
vessels, and all the chinaware for the court of imperial entertainments.
Silk fabrics were woven in Chekiang and South Chihli, of varying type,
color and design, to the imperial specifications. The manufacture of these
two last items, which might number as many as a quarter of a million in
a single order, called for statutory labor and the requisition of materials
in the local districts, plus extra logistical support. It was a constant cause
of tension between the civil officials and the eunuchs, because the demand
for materials and labor cut into the incomes of the ministry of revenue and
the ministry of works; it affected the local administrations of the provincial
officials as well. One more area of tension between the eunuchs and the
bureaucrats was the purchase of palace supplies. All the warehouses in the
palace grounds received relatively fixed quotas of materials derived from
tax proceeds or requisitions from the provinces. But whenever shortages
occurred, the extra purchases were still the responsibility of the respective
ministries, using funds derived from their own accounts. A special item of
supply might have been commuted to silver payments, in which case the
ministry intercepted the money but was responsible for supplying the
article. The eunuchs always demanded larger and more extravagant orders,
which were consistently opposed by the bureaucrats. Only in a few instances
would the emperor's decision favor the latter. This kind of conflict, com-
bined with the other vested interests of the eunuchs in charge of the supply
depots, was one of the reasons why tax payments in the later period could
not be totally commuted to silver.12 The lack of such a sweeping reform, as
we shall see later, affected the empire's tax structure deeply.

The minister of revenue and the ministry of revenue

The hu-pu shang-shu in Ming times is now recognized by us as the minister
of revenue, but the translated title with its sense of authority derived from
Western usage may be somewhat too complimentary to the Ming office-
holders. Throughout the Ming period no hu-pu shang-shu ever rose to
become a Cochrane or a Colbert. The Ming office had never been intended
to be that of a policy-maker. Since its establishment in 1368 the ministry
12 Fiscal organization and general practices
of revenue had gone through major reorganizations several times in
Hung-wu's reign. In 1372 there was one shang-shu and under him four
bureaus. In 1373, however, the office was demoted and split up. Five
shang-shu were appointed to take charge of five sections, none of them to
assume an overall command. Only after 1380 was the position of a single
chief administrator revived and the shang-shu, with his rank elevated to
2a, gained some semblance to a minister.13
Even then the Ming emperors often treated their ministers of revenue
with outrageous brutality, paying scant regard to the dignity of their office.
In 1385 Minister Yii T'ai-su was chained to his desk by Hung-wu because
he did not conduct business as speedily as the sovereign wished. In 1421
Minister Hsia Yuan-chi was imprisoned by Yung-lo merely because he
tried to dissuade the monarch from launching another expedition into the
Mongolian desert. Only on the emperor's death in 1424 did he regain his
freedom. In 1441 Minister Liu Chung-fu was pilloried by Cheng-t'ung
simply because he suggested that palace horses be entrusted to civilian
stables for maintenance. And in 1547 Minister Wang Kao was first beaten
in public and later exiled by orders of Chia-ching on a false charge of
receiving bribes, the true offense being his lack of enthusiasm for pur-
chasing ambergris, a rare kind of incense that the emperor wanted for his
Taoist altar.14
Of the 89 ministers of revenue appointed after 1380, 25 left office through
retirement, 22 were transferred to other duties, 16 were dismissed, 7 died in
office, 7 resigned because of illness or mourning, 3 were executed, 2 were
banished with their names removed from the civil service register, 1 was
exiled, 1 left office without authorization, 1 was killed in battle, and the last
minister, Ni Yiian-lu, committed suicide at the fall of the dynasty.15 This
leaves only 3 unaccounted for; the reasons for their leaving office cannot
be ascertained from the materials currently available. Even among those
who later left through retirement or otherwise, 3 ministers, namely Chin
Lien in 1451, Han Wen in 1507, and Pi Tzu-yen in 1633, had been arrested
while still in office. At least 5 more relinquished their positions in order to
avoid surrender to the palace eunuchs. After fighting losing battles with
the eunuchs, Ch'in Chin retired in 1527, Ma Shen in 1569 and Wang
Ying-chiao in 1622; Pi Ch'iang resigned in 1586, and Wang Lin was
transferred in 1585. The list shows the insecurity of the ministers; their
tenure was very much dependent upon the whims of their autocratic masters
and the favor of the latters' trusted attendants.16
Perhaps the only minister of revenue during the dynasty who managed
to exercise some initiative and handle his office with a sense of authority
was Ko Tse, who served under Yung-lo, Hung-hsi, and Hsiian-te. He is
said to have countermanded the latter two emperors' decrees which made
abrupt tax reductions without his consent. As a senior statesman and an
 1,1 Governmental organization 13
imperial tutor, Ko was able to defy the throne without causing himself
serious trouble. Nevertheless, in the end his policy was overruled.17
Most of the time the minister of revenue acted as the fiscal adviser to
the crown; under normal circumstances his ministerial duties involved only
very limited planning. Since the early years of the dynasty when taxes were
paid in kind and labor services performed by the taxpayers in person, the
military installations, the palace grounds, and several other disbursing
agencies in Peking were assigned fixed quotas of materials and services to
be furnished by designated collecting agencies. In the sixteenth century
most but not all of the materials and services were commuted to silver
payments; but the cash incomes were never consolidated. The disbursing
agencies simply maintained separate expense accounts in silver derived
from the commutation, still paid by the same collecting agencies which
earlier had furnished them with supplies and services. The only exception
was the palace. Due to the lack of a central fiscal agency in the Imperial
City, the ministry of revenue handled part of the silver consignments paid
in lieu of palace supplies. In addition, the ministry of works, the court of
the imperial stud under the ministry of war, and the court of imperial
entertainments under the ministry of rites all maintained large separate
expense accounts, as did the army installations. The ministry of revenue
performed some bookkeeping functions connected with the delivery of
silver payments, but could not exercise budgetary control over the
appropriation of funds. The surpluses from those expense accounts were
held by the respective offices as their own treasury reserves and remained
beyond the ministry of revenue's reach. The same procedure governed the
several fiscal offices in Nanking. When the dual-capital system was put
into effect in 1421, government organs were duplicated in the auxiliary
capital. By customary arrangement the ministry of revenue in Nanking
performed certain regional functions in the south. But this office, sub-
ordinate only to the throne, was by no means a branch office of the ministry
of revenue in Peking. The minister of revenue in Nanking reported directly
to the emperor, maintained his own quota of receivables, and controlled
silver vaults, granaries, and warehouses.18 Even auditing the separate
expenditures and making inventories of the supply depots in Nanking
were functions of the censorial officials, not the ministry of revenue in
Peking.
All this means that the revenues and appropriations had already been
established on a semi-permanent basis. The utmost the minister of revenue
could do was to make minor readjustments, revising the commutation
rates, recommending further commutations, and sometimes switching the
deliveries of materials and funds from one point of origin to another, all
on an ad hoc basis. The ministry of revenue under the Ming was therefore
much less an executive agency than a gigantic accounting office. Had there
14 Fiscal organization and general practices
not been a rapid increase in military expenditures in the later part of the
dynasty which forced the ministry to take a more active role in govern-
mental finance, the minister of revenue could have continued to remain for
all practical purposes the emperor's chief accountant.
Yet, merely to supervise the flow of supplies across the empire was an
awesome task. To ensure that all tax payments both in commodities and in
silver were properly delivered, the ministry of revenue prepared the so-
called stub-books (k'an-ho). Like bus tickets carrying detachable coupons,
the sheets of the stub-book were made of two, three, or more parts. But
the most commonly used kind had only two parts, consisting of a detach-
able coupon and a stub for each sheet. The remaining book with stubs on
it was sent to the receiving agency. Over the perforation line separating the
coupon and the stub the vermilion seal of the ministry was applied, and
the serial number of the delivery was recorded. Each part thus carried half
of the seal imprint and the writing on its edge. During the transaction the
deliverer and the receiver pieced the coupon and the stub together to
verify the order. This eliminated subsequent dispute over the delivery and
quantity of the article. In the rare situations when the delivery involved an
intermediate agency, the extra coupon on each sheet of the stub-book
served the purpose for intermediate delivery. When the deliveries were
completed the receiving agency reported to the ministry with a 'fulfillment
of the order' (fung-kuari).19 Any shortage or irregularity was also reported
to the ministry of revenue.
This device reflected the fact that the fiscal system relied on centralized
directing with decentralized handling. Supplies, in commodities or in cash,
were rarely handled in large volume during transit. The aggregate figures
of the various articles appearing in official documents had in most cases
been compiled for statistical purposes only. In practice it was typical for
a receiving depot to receive supplies from scores of different delivery
agencies; at the same time a delivery agency also made articles available
to dozens of different service installations. The volume of the transaction
was always kept at the lowest possible level to avoid transportation and
storage difficulties. Furthermore, the commodities thus handled, including
not only grain, hay, and cotton wadding, but also such things as indigo,
hemp, and sesame seeds, were of so many kinds that there was no way for
the accounts to be combined. This system of control caused a fantastic
accumulation of bookkeeping. The ministry of revenue had to scrutinize
all the accounts, usually down to the county level; in 1385 it reviewed the
accounts of 2,437 fiscal offices.20 The situation was not greatly improved
in later centuries, when the use of silver had become more widespread. In
1572 the ministry is said to have consolidated twenty-two kinds of account-
ing books and discontinued the use of another twenty-eight kinds. We do
not know how many kinds were retained after the reform.21 As late as 1632
 1,1 Governmental organization 15
Minister of Revenue Pi Tzu-yen submitted a memorial to the Ch'ung-chen
Emperor in which tax arrears across the empire were listed. The list covers
four and a half pages of fine print in the present reproduced form. Among
many other items, the minister was able to report to the emperor that
Wu-hsien in South Chihli still owed the palace some 28 taels of silver,
which was supposed to cover the portion of honey requisitioned from that
county.22
Considering its assignment, the staff of the ministry of revenue was
surprisingly small. The number of officials assigned to it was listed as
51 in 1390, plus 160 lesser functionaries handling the clerical work. In the
late 1570s there were 74 officials plus 165 lesser functionaries. Except that
from time to time dozens of imperial university students were attached to
the ministry to gain administrative experience, the size of the ministry
virtually remained unchanged through the rest of the dynasty.23
The minister of revenue had no executive officer, no comptroller, and no
chief statistician. All this is understandable when one remembers that he
himself for most of the time was serving in those capacities under the
emperor. Nor did the minister have a planning staff. Even though the
ministerial organization provided a 'business assistant' and a 'document
checker', those officials carried out only housekeeping and clerical super-
vising duties within the ministry; they were by no means under-secretaries.
When Minister Ni Yiian-lu in 1643 promoted a brilliant but hitherto
unknown student as his business assistant and placed five staff members
under him to take charge of office routine, the delegation of responsibility
was considered a novelty.24
Under the minister of revenue were two vice-ministers, but neither was
active within the ministry. One of the vice-ministers was customarily
appointed superintendent of imperial granaries. He maintained his separate
office and reported directly to the emperor. The T'ai-ts'ang Treasury,
which after its establishment in 1442 handled all the silver bullion under the
ministry's jurisdiction, was placed under his supervision. To add to the
confusion, sometimes this vice-minister was also promoted to become an
extra hu-pu shang-shu, thus becoming equal in rank and title with the
minister, though in practice he limited his function to that of treasurer and
granary superintendent and did not interfere with other business of the
ministry.25 The other vice-minister was usually given a field assignment
such as commissioner in charge of the Grand Canal, or superintendent
of military supplies to Manchuria, and thus also was absent from the
ministry's office.
In carrying out his duties the minister of revenue dealt directly with
thirteen bureaus of the ministry, which were organized to correspond to
the thirteen provinces of the empire. After 1575 matters concerning North
and South Chihli, the salt monopoly, inland customs houses, the granaries
16 Fiscal organization and general practices
on the Grand Canal and imperial stables and pastures, were ordered to be
placed under the jurisdiction of the Fukien, Szechuan, Shantung, Yunnan,
Kwangsi, and Kweichow bureaus respectively. The assignment was
arbitrary, mainly because the aforementioned bureaus, which supervised
the fiscal administration of remote provinces normally had less volume
of work and could take on the extra duty.26
Within each bureau there were supposed to be three or four staff mem-
bers of civil service status, but the positions were not always filled. Even
when they were, some of the occupants were often detailed to the frontier
provinces to manage field installations. For about a century before 1570,
the staff members were not even required to report to the ministry daily;
they held official positions on an inactive basis in order to gain seniority.
It was Wang Kuo-kuang, minister of revenue from 1572 to 1576, who first
demanded that all personnel on the ministry's payroll actually report to the
office.27 When in the decade after 1610 the positions of bureau heads were
again left vacant, Minister of Revenue Li Nii-hua acted simultaneously as
the head of several bureaus under his own command.28 It seems surprising
that with its limited manpower the ministry could afford these further
reductions of its authorized strength. This can be explained by the nature
of the ministry's routine, which was basically clerical and technical, some-
thing in which the civil service bureaucrats were least interested and
for which they were least qualified. The mainstay of the office personnel
were the lesser functionaries. Self-trained, informally recruited, and often
regarded by the civil service regulars with disdain, they nevertheless became
indispensable due to their special knowledge of procedural details and
customary usage. Apparently they held their positions as a life-time
profession and passed their trade on to their friends and relatives. Ku
Yen-wu has pointed out that in the early seventeenth century practically all
the lesser functionaries on duty with the ministry were natives of Shao-
hsing prefecture, Chekiang province.29 The official papers and accounts of
the ministry were reviewed and audited by a specially appointed supervising
secretary. Sun Ch'eng-tse, supervising secretary in that capacity during the
reign of Ch'ung-chen, indicated in one of his memorials that while con-
ducting the auditing he dealt with the lesser functionaries directly, and no
civil service officials appeared on the scene.30
Service agencies directly subordinate to the ministry were few. The
superintendency of paper currency, the office of currency supply and office
of plate engraving were rarely mentioned after the fifteenth century, due
to the gradual discontinuation of the production of paper money. Even
though coining money was a function of the ministry of revenue, the mint
in Peking was controlled by the ministry of works. Only in 1622 for the
purpose of increasing coin production was an additional mint created
under the ministry of revenue, operating simultaneously with the existing
 1,1 Governmental organization 17
mint and other mints under the jurisdiction of the ministries in Nanking
and provincial offices.31 All those mints, however, were manufacturing
plants. They were not grouped together under an imperial mint. The ware-
houses, which were under the control of the ministry of revenue but
actually in the hands of the eunuchs in the palace grounds, have already
been noted.
The ministry of revenue had no regional offices in the provinces. But
the inland custom stations outside Peking were managed by ministerial
personnel, usually secretaries of the several bureaus on leave of absence
from the ministry (6, i). Their tour of duty was limited to a one-year term.
Habitually these officials referred to their posts as 'branch offices, ministry
of revenue'. The title is misleading. The offices were strictly customs duty
collecting stations and no other ministerial functions were performed by
them. Moreover, in the later sixteenth century the inland customs duty was
gradually merged with other business taxes under provincial jurisdiction.
The customs stations in reality became operated jointly by the ministerial
officials and local prefects. Towards the end of the dynasty, vice-ministers
of revenue were put in charge of the services and supplies of field armies.
They became imperial commissioners of a sort. In performing their field
assignments they reported directly to the emperor rather than to the
ministry. Other officials of lesser ranks, such as directors and vice-directors
of territorial bureaus, were likewise dispatched to the frontier to take
charge of army logistics.32 For all practical purposes they were on loan
from the ministry to the respective governors-general on the border. In no
significant way did those officials extend the ministry's authority to the
field; nevertheless they maintained a close liaison between the army
command and the office in Peking.

Other ministries
All the other five ministries were in one way or another involved in fiscal
administration. Some of their interests in governmental finance were
derived from concerns shared with the ministry of revenue. These matters
usually caused no serious problem. For instance, the ministry of personnel
was concerned with the sale of ranks; the ministry of justice with the
commutation of punishments to fines; and the ministry of rites with awards
to foreign tributary missions and the licensing of Buddhist monks and
Taoist priests. As long as those ministries did not actually operate fiscal
installations to manage the funds, differences between the ministries could
be easily resolved. Such overlapping interests, after all, could have been
expected in any form of government.
Under the Ming system, however, the ministry of war to a lesser extent
and the ministry of works to a greater extent were virtually competing
18 Fiscal organization and general practices
with the ministry of revenue in fiscal management. The division of authority
went back to the early days of the dynasty when taxes were rarely levied in
cash. Since those days the land taxes in a large area extending over four
provinces had been reduced to one-half in exchange for the services
provided by the taxpayers in stabling of government horses. But when the
services were discontinued, the ministry of revenue was not authorized to
increase the taxes. Instead, the civilian households were ordered to deliver
silver payments to the court of the imperial stud to provide the army with
combat horses.33 The payments were still assessed according to the previous
quota of'studs' and designated as 'feeding fees', although there were now
no horses in their care. In reality this meant that half of the regular land
taxes in the affected areas was collected by the ministry of war (3, n).
The ministry of works not only collected the forest produce levy at many
stations and intercepted part of thefishduty, but also requisitioned materials
and funds from all provinces. The duties and functions of the ministry
included construction of palaces, mausoleums, public buildings and city
walls, undertaking water-control and reclamation projects, exploiting
water and forest resources, and manufacturing military equipment and
ships. In addition, right from the early days of the dynasty when no funds
were provided for those undertakings, the ministry of works called upon
corvee labor and collected materials all across the empire. While unskilled
labor was provided by the general population, skilled labor was levied
from those households which were registered as artisans. The weavers in
South Chihli and the porcelain workers in Kiangsi fell into this category,
as did the artisans of numerous trades in the maintenance area within the
Imperial City. The requisition of materials was even more extensive. Bows
and arrows were contributed from most prefectures. Lumber was derived
from the produce levy and glue from fish duty. Leather and fur were
furnished by those who were registered as hunters. Dyes and potash alum
were supplied by the districts specializing in those products. When corvee
labor was called up for water-control projects, not only did those on call
provide their own tools, but each man also brought along some supplies
from his home locale, including wooden poles, nails and ropes. The material
and labor requisitions could be both scheduled and unscheduled. The
former were assessed to every prefecture of the empire and the prefect in
turn divided the quotas among his subordinate counties. Some of the
assessments were collected annually, others biennially or triennially.34
Since all these obligations in Ming times gradually began to be commuted,
the ministry of works slowly became a recipient of silver payments in the
provinces. The income may be considered as a huge expense account,
directly derived from taxation, which enabled the ministry to remain self-
sufficient. It is in fact true to say that the ministry was gradually transformed
into a taxation agency that rivalled the ministry of revenue.
 1, I Governmental organization 19
The deterioration of the empire's population data further intensified
the conflict of interest. In the early years the land tax and the service levy
(see 1, II below) were separate impositions; those with artisan registration
were clearly distinguished from the general population. From the mid
fifteenth century onwards the population registration became a matter
of formality. The registered statutory laborers had already absconded and
many miscellaneous requisitions, such as those assessed on hunters and
fishermen, had become uncollectable. Most counties had adopted the
solution of transferring the burden of service and supply obligations to
landed property, converting these quotas into surcharges on land or even
adding them to the land tax proper. The commutation of the service levy
therefore in the end resulted in both the ministry of revenue and the ministry
of works chasing the same tax income.
For extraordinary quantities of supplies needed in Peking, the procedure
was 'local procurement' (tso-pan). For instance when a major construction
project was undertaken, the required lumber might exceed 1 million taels
of silver in value. Obviously, the order could not be filled from the regular
requisitions. By local procurement the ministries directed the provincial
officials in the lumber-producing areas to furnish what was needed, the
cost being compensated by tax proceeds derived from the same provinces.
In other words, either the lumber was paid for with tax proceeds due to the
central government or the taxpayers submitted the required raw materials
in lieu of regular tax payments. The ministry of works, deriving its cash
income from commutation of service obligations, which were scattered
over the empire, and restricted in amount, often found the resources at
its disposal on the spot insufficient to pay the costs. It would petition the
emperor and by order of the sovereign the ministry of revenue had to make
the local tax proceeds under its jurisdiction available for the occasion. The
only large item of tax proceeds was derived from the land taxes. The
procurement orders of the ministry of works subsequently invaded that
domain too. 35
It seems that the ministry of works in the early part of the dynasty had
been an operating agency. But clearly as time went by it became more and
more concerned with fiscal management. The spectacular water-control
projects under the direction of Liu Ta-hsia in the later fifteenth century and
P'an Chi-hsun in the sixteenth century (7, i) were executed with little active
participation from the ministry. It merely allocated materials, funds, and
labor to the project directors. Land reclamation, similarly, slipped into the
hands of provincial and local officials. For palace construction either
ministerial officials or eunuchs were put in charge. But inasmuch as the
service facilities and technicians were under the control of the eunuchs,
they alone could carry out the operation. Even when a ministry staff
member was commissioned as the project director, he could not do more
20 Fiscal organization and general practices
than assume the overall responsibility and carry out the financial planning.36
To supply material and labor for the manufacturing plants and warehouses
inside the palace compound was mainly, though not exclusively, the
responsibility of the ministry of works, and so was the budgetary allocation
for eunuch procurement missions. All this deprived the ministry of its
operating capacity and reduced it to a logistical branch or service support
agency of the court. The Ming system created numerous palace installations
without a ministry of imperial household to coordinate them. It was the
ministry of works, in its changing role, that came close to filling this gap.
The unexpected role sometimes confused the ministerial officials and
caused them to be inconsistent. In the event of an extraordinary order for
silk fabrics initiated by the eunuchs, the bureaucrats at the ministry of
works would, generally as a matter of principle, join their civil service
colleagues in objecting to its excessiveness. But when the cost of the order
was to be borne by the ministry of revenue and the latter ministry argued
against surrendering its own tax quotas for this purpose, the matter became
a dispute between the two ministries. At this point the ministry of works'
officials would change their stand and say that such a supply order was,
after all, only reasonable.37
The net result of the divided fiscal authority had serious consequences,
however. Throughout the Ming period no central treasury was ever
established. The T'ai-ts'ang Treasury under the ministry of revenue was
only one of the silver vaults in Peking. It exercised no control over the
Ch'ang-yin Treasury or the Chieh-sheng Treasury, which were subject to
the court of the imperial stud and the ministry of works respectively, nor
the silver vault inside the court of imperial entertainments, not to mention
the Inner Ch'eng-yiin, Kuang-huei, and Tung-yii Treasuries inside the
palace. The ministry of revenue in Nanking also had its own treasuries.
Except by a specific order from the emperor, silver deposits could not be
transferred from one unit to another. In the Wan-li period, even after the
emperor's order for transfer, no agency would yield its bullion readily,
knowing well that the order had been initiated by some rival office rather
than by the emperor himself. Usually it memorialized the throne for
reconsideration and applied delaying and haggling techniques to avoid the
loss of silver to the other agency, giving in grudgingly only after all means
of resistance had been exhausted.38

Fiscal agencies of the central government in the provinces

Throughout the Ming period no regional treasuries were ever established
in the provinces by the central government. Owing to the monolithic
structure of the state, the primary fiscal function of all provincial and local
governments was to serve the central government. There was no need for
 1,1 Governmental organization 21
the central government to add regional offices in the provinces. The few
fiscal installations considered in the present category were exceptional
cases, each specializing in an inter-provincial service or a special type of
revenue. They were the imperial commissionership in charge of the Grand
Canal (2, i), and the managerial offices of the salt monopoly (5, i), tea and
horse trade, maritime and inland customs duty, and the forest levy (6, i and
iv). It may be noted here that in theory most of these offices were under
provincial jurisdiction. But when the censors were dispatched to exercise
surveillance over the salt monopoly and tea and horse trade, they actually
became fiscal officers operating in the provinces under the emperor's
authority. They reduced the operating agencies, still nominally under
provincial jurisdiction, to the status of their direct subordinates. In order
to avoid duplication, detailed descriptions of these offices will be postponed
until the discussion of the services and revenues under their management.
However, it may be observed that in general the efficiency of these
offices was low, and the revenues under their management never reached
levels comparable to those of the T'ang and Sung. In part this was due to
the fact that at the beginning of the dynasty the Ming had neglected
commerce as a major source of state income. Although the material
requisitions under the Ming were comprehensive, more emphasis was
placed on goods that were actually needed for use by the state; consider-
ation of their intrinsic financial value was secondary. The aforementioned
managerial offices were moreover scattered over the empire without an
effective chain of command to back them. As will be seen in the forth-
coming chapters, this lack of organizational support induced some of the
agencies in the later part of the dynasty to maintain a de facto joint
administration with the provincial officials.

Provincial government and local government

The primary consideration behind the organization of the provincial and
local governments was that of fiscal management. An edict in 1373 assigned
upper, middle, or lower grades to the prefectures of the empire. The first
category included those prefectures which produced an annual land tax
income of 200,000 piculs of grain* or more, the prefect to be ranked 3b;
the second category embraced those whose annual income was less than
200,000 piculs but more than 100,000 piculs, the prefect being ranked 4a.
All other prefects were of the lowest category and ranked 4b. Apparently
in the early years of the dynasty it was expected that the tax quotas of all
territorial units would be revised from time to time to reflect changes in
population and land data; the status of each unit, together with the rank of
its administrator, would then be adjusted accordingly. The system was too
* For explanation see: A note on weights and measures, p. xiv.
22 Fiscal organization and general practices
cumbersome to be carried out in full; by 1371 the empire already had
12 provinces, 120 prefectures, 108 subprefectures and 887 counties.39 (In
later years their numbers increased to 13, 140, 193, and 1,138 respec-
tively.) Nor was any periodic revision of the tax quotas of the territorial
units seriously attempted, most districts retaining their tax quotas on
a permanent or semi-permanent basis. Even though in the sixteenth century
there were still scattered cases where counties were promoted to sub-
prefectures or vice versa, they were increasingly rare in occurrence. In the
meantime officials were ranked equally according to the levels of admini-
stration, with the exception of the two prefects in Nanking and Peking
who were one grade higher than their counterparts elsewhere.40
The local government could be arranged in either four or three tiers.
The four-tier pattern included the province, prefecture, subprefecture and
county in descending order. The three-tier type involved a subprefecture
directly subordinate to the provincial government (without any inter-
mediate prefecture), or else a county directly subordinate to a prefecture
(without any intermediate subprefecture).41 There were also subprefectures,
which, while subordinate to prefectures, had no counties under them. The
two metropolitan areas, namely North Chihli and South Chihli, each the
size of a province, had no provincial administration, their prefects and
subprefects reported directly to the central government. Consequently, in
the national accounts the 8 prefectures of North Chihli and 15 prefectures
and 3 independent subprefectures of South Chihli were never consolidated
under the respective metropolitan areas. The fiscal data of these 26 units
were placed on the same footing as those of the 13 provinces.
These different patterns again reflected a primary concern for fiscal
administration. The guiding principle behind it was that the county was
a basic tax-collecting unit, the prefecture a basic accounting unit, and the
province a revenue transit unit. Whenever possible, the seat of a county
government was located within one day's travel distance from its borders.
The size of the county, therefore, was more or less predetermined. The
offices of the higher echelons also had to be centrally located among their
subordinate units to facilitate operations. It was moreover necessary that
all those governmental offices be situated in walled cities and population
centres for convenience with regard to military operations, their own
maintenance, and land and water transportation. In special circumstances
the Ming court could actually relocate a sizeable city merely to suit
administrative purposes. Yet, territorial organization could not completely
ignore geographical and historical factors. In the selection of a prefectural
capital, in particular, they had to balance the demands of organizational
uniformity and the physical advantages of the sites. The creation of sub-
prefectures eliminated the difficulty when the divergent demands could
not all be reconciled. Most of them had the effect of breaking the large
 1,1 Governmental organization 23
prefectures into administratively manageable units; others accommodated
those districts which had an intermediate level of revenue but were, how-
ever, disadvantageously located geographically. Created to adjust im-
balances, the subprefectural government had no distinctive features of its
own. When the subprefecture was subordinate to a prefecture, it functioned
as a branch office of the prefect. If the subprefecture was immediately sub-
ordinate to a province, it functioned as a prefecture of a minor order. In the
latter case its fiscal accounts were put on the same footing as those of the
prefectures.42
When a level of government became unnecessary in the chain of com-
mand, it was eliminated. Since fiscally the provincial government served
mainly as a relayer of revenue, it was not required in the two metropolitan
areas whose prefectures and subprefectures were conveniently close to the
northern and southern capitals.
According to their organizational principles there was no great difference
between the central government and local government. All provincial and
local officials down to county magistrates were appointed by the throne.
All revenues, in a sense, were imperial income. There was no expenditure
of the central government that could not be paid by provincial and local
officials. This monolithic order was slightly compromised during the later
years, notably in the southern provinces after the mid sixteenth century.
Nevertheless even then the parcelling-out of funds was never clear, definite
and thorough. In most instances every item of income was shared by both
the central government and the provincial government. The relative fiscal
autonomy in specific areas could only be considered, at least in theory, as
growing out of standing authorizations which permitted their officials to
disburse some of the funds on the spot on behalf of the throne, without
frequent requests for prior approval. The funds thus disbursed were not
recognized as institutionalized local revenue.43
Most of the provincial and local officials delivered portions of their tax
proceeds to the various receiving depots according to established pro-
cedures. Some of those depots were located on the northern frontier, some
in Peking or Nanking, and some quite close to the point of origin of the
taxes. Some but not all of the deliveries were consolidated at prefectural
and provincial levels. A consignment delivered outside the fiscal officer's
district constituted an item of ch'i-yiin, literally a 'checked-out item',
which will however be referred to as 'transferred revenue'. Upon the
completion of delivery the item was removed from the local administrator's
jurisdiction. The balance retained within the territory was called ts'un-liu,
literally a 'staying-in item', but will be termed 'retained revenue'. Any tax
income could simultaneously comprise portions of transferred revenue and
portions of retained revenue for the collecting district. From the retained
revenue the local officials drew their salaries. The stipends due to govern-
24 Fiscal organization and general practices
ment students and imperial clansmen were also derived from this portion,
as were local relief payments authorized by the emperor. Any surplus was
kept by the provincial or local official for the crown, and could not be
disposed of without the emperor's approval. From time to time the central
government directed the provincial and local officials to make local
procurements (tso-pari), build ships, construct palaces for imperial princes,
or make unscheduled deliveries; all such expenses were also met from the
retained revenue. This practice meant that most of the provincial and local
officials, down to county magistrates, were simultaneously regional
treasurers of the imperial government. An item of retained revenue was
neither a surplus nor a local revenue, but represented the account managed
by a local official when he acted in the capacity of a minor regional treasurer
of the empire.
All territories were expected to be self-sufficient and only in exceptional
cases were grants-in-aid dispatched from adjacent districts by order of the
central government. In the 1580s and 1590s, when Yunnan was fighting
a border war with Burmese leader Nanda Bayin, the governor repeatedly
petitioned Peking for aid and for some time the emperor authorized
subsidies from Szechuan. In 1594, however, the governor was finally
instructed to manage the affair with the resources of the province and no
more subsidy was forthcoming.44 It reaffirmed the principle that inter-
provincial aid was not to be lightly granted.
This decentralized operation under centralized control meant that
among officials of all levels, the county magistrate always had the heaviest
fiscal responsibilities. For this reason, in discussing the particular functions
of the various offices, it is easier to start from the county upward than to
descend from the province downward.
Except for the maritime and inland customs duties, the forest produce
levy, the salt revenue, and some administrative incomes which were
collected and managed by special agencies and higher offices, practically
all state revenues passed through the hands of the county magistrate. The
collection of land taxes, including the surcharges, was completed at the
county level. Most of the magistrates also handled local business tax, stamp
tax, store franchise fees, license fees, excise on wine and vinegar, fines,
payments for rationed salt, and part of the fish duty. (For such miscella-
neous incomes, see chapter 6.) The magistrate managed public land in his
district, filled material requisitions, called statutory laborers to serve at
various state agencies and when the services were commuted, collected
payments for them. Wherever land reclamation programs were in effect, or
horse-stabling services by civilian households were required, or the pro-
ceeds from military farming had to be collected, all responsibility for them
fell on the magistrate. In addition he conducted the periodic population
registration, compiled the Yellow Book, organized the village communities
 1, I Governmental organization 25
in the so-called li-chia service system, appointed tax captains for tax
deliveries (1, n), reported natural disasters for tax remissions and admini-
stered famine relief. While there were few imperial agencies in the provinces,
there were still fewer provincial and prefectural fiscal agencies in the
counties. Normally the offices of the higher echelons were located in
the walled cities which also happened to be provincial and prefectural
capitals. In a few cases the higher offices operated fiscal installations such
as business tax stations immediately adjacent to those cities. Otherwise,
the territory of a county was under the undivided authority of its magis-
trate. 45
The costs of office maintenance, apart from salaries, were not borne by
the ordinary revenue, but derived from special materials and labor requisi-
tions from the village communities, the li-chia. These materials and services
were not furnished merely for the maintenance of the magistrate's own
office, however. Allocations from them were also made to all superior
offices, because aside from li-chia requisitions no other funds were provided
for operating expenses. The procedure was not changed even after the
commutation of material and labor requisitions into silver payments. Such
entires as '20 taels of silver, this county's contribution to the wages paid to
the prefect's sedan chair bearers', and '18 taels, our share of stationery
expenses for the provincial administration commission', appear in all late
Ming editions of local gazetteers. Some of the proceeds from local collec-
tions went all the way to the capital. In the later sixteenth century the two
county magistrates in Peking were still responsible for providing food,
money, and stationery for the triennial civil service examinations, which
were a function of the imperial government.
Regional offices of the central government, though few in number, also
derived their office maintenance funds from local collections. The messen-
gers, doormen, and guards for the Ch'ing-chiang-p'u forest-produce
collecting station, an agency of the ministry of works, were always supplied
by Shan-yang county in which it was located.46 This practice in itself
probably to an extent inhibited the central government from establishing
too many branch offices in the provinces, because their operating expenses
were inevitably an extra burden to the local population. A related practice
was that officials in charge of special functions were concurrently given
provincial assignments. The commissioner in charge of the Grand Canal,
for instance, was concurrently appointed the governor of Huai-an, so that
he could use local resources to establish his logistical base.
The upkeep of local police, militia, transport offices and postal stations,
and the normal maintenance of waterworks within the county, were also
among the magistrate's duties. Since most counties were expected to be
self-sufficient, the li-chia collection fell more heavily on those counties
along major land and water routes. Sometimes the magistrate appealed
26 Fiscal organization and general practices
and secured a grant-in-aid from an adjacent district. Such a grant, as has
been mentioned, was an exception rather than a rule.
The understating of the central government was duplicated at the local
level. Governing an area of some 500 to 1,000 square miles, with a popu-
lation ranging from 30,000 to 250,000, the magistrate's regular staff in-
cluded only three members of civil service status: the vice-magistrate, the
assistant magistrate, and the docket officer. Some of the officers in charge
of the county granary, police force, business tax, postal station, and fish
duty might also have civil service status, but such positions were not
established in all counties.47 Most of the workers in the county government
were lesser functionaries. Sometimes the county magistrates assigned
normal administrative duties to the district Confucian instructors. The
government had no functionaries in the rural areas; during the reign of
Hung-wu in fact the magistrates were not even permitted to leave their
cities.

At the intermediate level, the prefect's fiscal responsibility was largely

supervisory. The prefect saw to it that all the scheduled tax deliveries were
properly carried out and the reserves were kept in good order. He also
operated a number of revenue and service agencies including a prefectural
granary, police, postal, business-tax and fish-duty stations, though not all
these installations existed in every prefecture. In some areas there were
major watergates in the canals and rivers, and government mines, pastures,
dyeing, weaving, and miscellaneous manufacturing plants, which also
required the prefect's supervision.
At the beginning of the dynasty taxes were fixed for each prefectural
unit. Since the later fourteenth century prefectural tax quotas were relatively
settled (2, i), only a few internal readjustments being made from time to
time. The prefect had a certain unspecific authority to readjust the quotas
of the subordinate counties. Those readjustments could be formal or
informal. Approval from provincial authorities or the throne and the
concurrence of surveillance officials might or might not be sought, depend-
ing on the circumstances. Normally no prefect would increase or decrease
a county's tax quota outright, but he could suggest the switching of tax
deliveries, the modification of surcharges, or the revision of commutation
rates, thus in effect reapportioning to some extent the tax burden of the
subordinate counties. Some vigorous prefects and their county magistrates
actually revised internal tax procedures, directed the reapportionment of
payments among individual taxpayers, and even carried out local land
surveys.48 But all this depended upon the prefect's or the magistrate's
personal character, prestige, and resourcefulness. The local official
exercised his discretionary powers at a certain risk, and he expected no
explicit authorization from his superiors. He was protected only by the
 1,1 Governmental organization 27
general understanding that since he was obliged to produce his regional
tax quota he should have some freedom in devising the best possible
method. After his revised tax procedure had been in effect for some time it
would become established and gain the respectability of customary law.
On the other hand he had also to take into account the possibility of local
resistance and impeachment by censorial officials if he carried his program
too far.
Most prefectures supervised only a handful of subprefectures and
counties, usually less than ten. But there were exceptional cases. K'ai-feng
prefecture of Honan had four subprefectures and thirty counties. Tsi-nan
prefecture of Shantung had four subprefectures and twenty-six counties.
The size of prefectural government could also vary widely. A prefect
might be assisted by one vice-prefect or seven,49 and the number of his
other staff members varied similarly. The record office of the prefectural
government, or the county government, which employed a sizeable number
of lesser functionaries, was organized into six sections to correspond to six
ministries of the central government.
As might be expected, the duties of a subprefect resembled those of
a prefect administering a smaller district or of a magistrate administering
a large county.

Executive organs were less integrated at the provincial level. The provincial
administration office was the chief fiscal agency, but the surveillance
commissioner's office was also authorized to inspect water-control projects,
the tribute grain, land reclamation, salt administration, postal services,
and sometimes military defense. The investigatory function of the surveil-
lance commissioner in the course of time often came to exceed its original
limits.50 Practically every surveillance commissioner produced some income
of his own, partly through fines and confiscations and partly through
commuting the services and supplies required for the various projects and
programs under his supervision. One of the functions of the surveillance
commissioner was to check and rectify abuses in tax administration. As an
extension of this, some surveillance commissioners set up ground rules for
collecting silver payments as service levy. In so doing they were virtually
acting as tax legislators. Two outstanding surveillance commissioners who
did this in the sixteenth century were P'an Chi-hsiin (1521-95) in Kwang-
tung (3, i) and P'ang Shang-p'eng (Chin-shih, 1553) in Chekiang (3, m),
both of whom were instrumental in devising the Single Whip Reform.
The provincial administration office was headed by two administrative
commissioners, of the left and the right, the former being the senior. The
office kept vital statistical records, corresponded with the ministries on
budgetary, taxation and procurement matters, and was responsible for all
cash deposits, granary reserves, and warehouse stocks within the province.
28 Fiscal organization and general practices
In the early years of the dynasty the administration office had a much
smaller operating capacity. Few provincial administration offices had
revenue offices directly under them. After the supply depots were taken
over from the army in the early fifteenth century, all granaries in the interior
provinces were managed by prefects and magistrates; provincial granaries
existed only in the frontier areas.51 But the material-processing and manu-
facturing plants, warehouses, and armories were placed at the provincial
level so that the administration office could fulfill the function of a main-
tenance and transit agency. There were also provincial mints in most pro-
vinces.
The fiscal function of the provinces steadily increased during the middle
period, due to the increasing use of silver which permitted some degree of
concentration of financial resources at the provincial level. The decline of
the military colony system (wei-so), which forced the provincial officials
to organize their own defense, was also conducive to the extension of fiscal
authority in the provinces. This development was by no means balanced,
however. Provincial officials, especially in the southern region, had con-
siderable freedom in administering those revenues designated as 'military
supplies' (ping-hsiang); yet in many other matters they were still much
restricted by the standing operating procedures.
The appointment of provincial governors, which began in 1430, was the
cause of considerable organizational ambiguity. Originally the governor-
ship was not intended to be a permanent office, nor was the governor nrsant
to be a regional executive officer. A governor was an individual delegated
by the central government to tour a particular province or a part of
a metropolitan area. But in practice governors came to hold their positions
as regular appointments; they consolidated their offices at the provincial
level and treated the administrative commissioners as their staff members.
Yet the fiscal responsibility of the provincial administration office was
never completely shifted to the office of the governor. In general governors
submitted their memorials directly to the emperor and the administrative
commissioners maintained their regular channel of communication with
the ministries, the former reporting on specific matters and the latter on
routine business. Right to the end of the dynasty it was the provincial
administrative commissioner, not the governor, who was held responsible
by the ministries for fiscal delinquencies in a province, such as tax arrears.
Moreover the fiscal administration of more than 100 separate adminis-
trative units proved too burdensome a task for the provincial administration
office. Numerous difficulties arose if the office had to confine its activities
within the provincial capital. In the Yung-lo period, senior staff members
of the commissioners were authorized to operate branch offices within the
provinces. Originally an informal arrangement, this delegation of authority
later became formalized. As the system developed, both the administrative
 1,1 Governmental organization 29
commissioners and surveillance commissioners had 'circuit intendants'
strategically located in the provinces, some performing specific functions
in wide areas, others supervising general administration within limited
districts. There was a great variety. Furthermore even some of the pre-
fectures in the metropolitan areas were placed within the authority of
circuit intendants from the adjacent provinces.52 The earlier purpose of
establishing circuits was to expedite field operations and to exercise close
scrutiny over local administration. But the decentralization, once started,
tended to gain momentum. From the sixteenth century onwards many
circuit intendants made on-the-spot decisions on fiscal matters; they some-
times approved local tax legislation initiated by the prefects. Even the
court in Peking began to assign specific duties to the individual circuits. On
balance, the circuit intendants, performed a useful function: they intro-
duced some measure of decentralization within a much too centralized
governmental structure. A point to be observed is that in normal account-
ing a circuit was not a regular fiscal unit. Only on special occasions did
the circuit intendants actually handle revenue, and they had no formal
responsibility as fiscal officers.

Fiscal administration of the army

At the beginning of the dynasty military commanders took a much more
active role in public finance than they did in the later period. Most of the
granaries were placed under the army's control. The army itself had its
own administrative organs starting at the lowest level with the military
colony (wei) and independent battalion (so), and surmounted by the five
military commissions at the capital. The lowest grade in the military
bureaucracy was 6b and army officers always held higher ranks than their
civil service counterparts.
After the first quarter of the fifteenth century, nevertheless, the influence
of the whole military hierarchy rapidly dwindled. The first blow came in
1425, when the Hung-hsi Emperor assigned a number of civil officials to
assist the unlearned military commanders with their administrative duties.
From this precedent evolved the military defense circuits of the later days.
Thenceforth the military defense circuit intendants, being civil officials of
the surveillance branch of the provincial government, all but took over the
administrative functions of the district commanders.53 In 1435 by order of
the Cheng-t'ung Emperor, all granaries hitherto under military manage-
ment were turned over to the civil administration, the only exceptions
being those in Liao-tung (Manchuria), Kan-su,* Ning-hsia,* Wan-ch'iian,
and some coastal areas where the military colonies were located beyond
* Under the Ming those were frontier military districts, not to be confused with the
modern provinces bearing the same names.
30 Fiscal organization and general practices
the territories of counties and prefectures.54 This realignment significantly
reduced the fiscal authority of the army officers and further integrated land
revenues derived from military and civilian sources. From local gazetteers
it can be seen that in the later period provincial officials handled the pro-
duction of the military colonies, paid in kind or in cash, in just the same
way as they did the land taxes. The two kinds of revenue began to merge
(7, n). This in turn meant that the army was accountable to the civil govern-
ment for its grain production.
All these measures, nevertheless, did not diminish the nominal fiscal
responsibilities of the regional military commissioner, nor of the army
officer on administrative duty at the provincial level. An imperial edict
dated 1554 still required him to submit annual reports specifying the food
production and consumption of his command, and quarterly reports on
fort repairs and equipment maintenance. The reports were transmitted to
one of the five chief military commissions in Peking, and thence either to
the ministry of revenue or to the ministry of works according to the nature
of the business.55 It is doubtful, though, that such reports had much
practical value, aside from their use for general reference purposes, since
by then the regional commissioner could no longer be considered a fiscal
officer in the same way as he had been earlier.
For his own office maintenance the regional military commissioner made
a special requisition from the military households in his territory, re-
sembling the li-chia collection of the civil government. A local gazetteer
from north China dated 1609 indicates that the procedure remained in
effect up to that time.56 The collection was supervised by the civil govern-
ment, but the proceeds were submitted to the commissioner's office.
On the northern frontier the regional military commissioners remained
autonomous until the mid fifteenth century, but thereafter there was an
even more complete takeover of these districts by the civil bureaucracy.
Governors and governors-general were appointed to administer the frontier
posts and territories, with regional inspectors and vice-inspectors to assist
them. They built a new chain of command, reducing the military command-
ers to the status of their deputies and subordinates. In fact, few officers
were ever appointed as military commissioners thereafter. The senior army
officer in a frontier region was usually a vice military commissioner
{tu chih-hui fung-chih) or assistant commissioner {tu chih-hui ch'ien-shih).
Those titles signified the holders' ranks rather than their assignments. The
assignment was termed tsung-ping, or commander-in-chief, a position that
differed drastically from that of a regional administrator. In the later part
of the sixteenth century even the maintenance of the army at the lower
levels fell into the hands of civil officials. Swarms of them, holders of
nominal assignments as directors and secretaries of the ministry of revenue,
and as prefectural judges and vice-magistrates of the interior provinces,
 1,1 Governmental organization 31
moved into the border lands. Some of them took over army administrative
functions down to the battalion level.57

The aristocracy
The aristocracy in Ming times could never lawfully interfere with govern-
mental operation. Only a few distinguished military commanders were
granted aristocratic titles as dukes, marquises and earls, as were some of
the emperors' in-laws. They received no fiefs; none commanded a personal
army. Most of them were stipendiary members of the military hierarchy,
although they passed their titles to their descendants. Even the most power-
ful among them, the Mu family in Yunnan, whose members held the
position of senior army commander in that province throughout the
dynasty and occupied large tracts of land, did not come anywhere near
reducing that province to a private domain.
Imperial princes, including the emperor's uncles, younger brothers, and
sons with the exception of the heir apparent, were all removed from the
capital upon attaining maturity. They were given territorial titles, but in
reality the so-called principalities existed only on paper, reminiscent in
a way of the British aristocracy in recent days.58 On the other hand each
prince under the Ming received a magnificent residence provided by the
state, usually in a remote province but never near the coastal region. In
addition he was granted an annual stipend, normally 10,000 piculs of
grain. His small staff was appointed and salaried by the state, while
household attendants were furnished by the local districts. A detachment
of some 3,000 men, also on the payroll of the imperial army, was assigned
to each prince of the first degree as his personal guard. The only function of
a prince of the first degree was to provide moral leadership for all imperial
clansmen under his jurisdiction. The title of the prince and its accompanying
privileges were inherited by the eldest son, while other descendants were
granted lesser titles and smaller stipends. All imperial clansmen, that is,
descendants of the founding emperor, were supported by the state for life.
Their given names were assigned by the ministry of rites and entered in the
'jade genealogy'. None of them, including the princes, could leave his city
of residence without the emperor's approval. None of them was permitted
to enter the civil or military service. They were not eligible for the civil
service examination and not allowed to engage in trade. Towards the end
of the fifteenth century, the number of imperial clansmen grew to un-
manageable proportions. Only in the later sixteenth century was the last
rule revoked owing to the government's continued defaulting on their
stipends.59
The emergence of aristocratic estates, embracing large tracts of land, was
a development of the later fifteenth century. At about the same time the
32 Fiscal organization and general practices
emperor's in-laws, along with the eunuchs, began to take advantage of the
state monopoly of salt and thereby created considerable disturbance in the
empire's fiscal administration. But such abuses, even then connived at by
the emperor in question, were usually corrected during subsequent reigns.
The privileges of the aristocrats were never institutionalized.

II RURAL ORGANIZATION AND BASIS OF TAXATION

The Yellow Book
Ten years before the founding of the dynasty, Hung-wu had already decreed
that all the population and households under his control be properly
registered. In 1370 he personally directed that each household be issued
with a registration certificate.60 The first Yellow Book was compiled in 1381
and with it came the li-chia system. Thereafter census-taking was supposed
to be carried out once every ten years. The last census was conducted in
1641-2, only two years before the collapse of the dynasty.
The population records were made in four sets, so that copies could be
deposited with the county, prefectural, and provincial governments. The
fourth copy was submitted to the imperial government, which constructed
a special depository outside the city walls of Nanking for their storage
(2, n). This last copy had a yellow cover, from which the term 'Yellow
Book' (huang-ts'e) was derived.
Most of the households under registration were classified into one of
four categories, namely, the general population, hereditary military families,
artisans, and saltern households.61 The most complicated category was
probably that of artisans. These were classified according to trade as
masons, carpenters, weavers, printers, etc. Apparently in the early Ming
the members of a household were supposed to be permanently confined
to their district of registration. Private travel, for which passports were
required, was not forbidden outright but was nevertheless discouraged.62
Those who stayed outside their districts of registration for prolonged
periods had to report to local officials. Persons who were not bona fide
merchants and applied as such were punished. After the mid fifteenth
century these restrictions could no longer be enforced and were gradually
relaxed. There is some evidence nevertheless that local officials still
occasionally issued passports in the sixteenth century.
The vocational classification, applicable to households instead of
individuals, implied that a family trade was to be inherited in perpetuity.
Sons and nephews were expected to follow in the steps of their fathers and
uncles. Yet the state never insisted on rigid social stratification. No decree
was issued to effect class segregation. No law was ever proclaimed to
prohibit marriages between different groups. The vocational registration
 1, II Rural organization and basis of taxation 33
was designed entirely for maintaining the army and servicing the govern-
ment. The state merely demanded that each type of household provide it
with a specialized type of service. In practice as long as sufficient carpenters
could be summoned to provide unpaid labor in the state construction
projects, the government was unconcerned if the sons of a registered
carpenter took to professions other than carpentry. Even in the early years
substitutes were accepted without question. The hereditary military families
filled the vacancies in the army, but their remaining members chose their
occupations freely and were as eligible for the civil service examinations as
the general population. In fact a number of top civil officials under the
Ming originally came from hereditary military families.63
Another feature of the system was that merchants did not appear as
a major category in the register though some city-dwellers, while registered
as general population, were marked as 'wealthy households' or 'shop-
keeper households'. They were required to make good the special purchase
orders occasionally issued by the government, and to make forced con-
tributions now and then.
The coverage of the population registration was so broad that few
escaped the service obligation. The nobility, government officials, and
qualified students, along with their family members, were partially or
totally exempted. Care was taken that Buddhist monks and Taoist priests
be licensed for a fee. According to a policy declared during the Hung-wu
period, the licensed ecclesiastics ought to have performed general service,64
but in practice they were excused. With these few exceptions each subject
made some form of service contribution to the state. Hunters for instance
had to surrender to the government a certain number of animal skins each
year.65 Similarly, salt workers toiled for the state to fulfill their productive
quota in exchange for a small grain payment. Even musicians were liable
to be called to perform without compensation.66 Households of this sort
which did not fall into the major categories were registered under the
heading 'miscellaneous'.
As the demand for unpaid services continued to grow, there was a
multiplication of these miscellaneous households in the later part of the
dynasty. A family assigned to tend to the vegetable garden of an imperial
prince was called a 'vegetable household', one acting as caretaker of the
tomb of the nobility a 'tomb household'. The family of a palace woman
was designated a 'female household', a term that applied somewhat in-
congruously to her male relatives. But on account of their having placed
a daughter or a sister at the service of the inner court, they were exempted
from the services normally required from the general population.
34 Fiscal organization and general practices

The li-chia system and the service levy

The most essential services were furnished by the rural communities. In
the villages the population was organized into the li-chia. Every 110 house-
holds formed a //, or a village community. Each // was divided into ten chia,
or sections, consisting of 10 households each. The remaining 10 households
reckoned to be the largest and most affluent among the 110, took it in
annual turns to be the li chief. Likewise each year a particular chia was
called to service. Under the direction of the //chief of the year, it performed
the local tax collection and delivery, and met all material and labor requisi-
tions on behalf of the entire li. The other units paid their regular taxes, but
were not liable to service obligations that year. Thus in a decennial period
all households took a one-year turn at discharging their service obligations.
After the ten-year cycle a new census was taken and all li-chia were re-
organized in accordance with the changes that had occurred during the
decade. With certain variations, the wards and precincts in the cities were
organized along similar lines.
The service obligations fulfilled by the li-chia were referred to by con-
temporaries as /, sometimes translated as 'service'. It was more than labor
service however, for it also included the contribution and handling of
materials and always some small cash payments. As it was basically a form
of taxation it will be in this work referred to as 'service levy'.
When taxes were paid in kind the local collectors were responsible for
the measuring, sorting, packing, temporary storage and final delivery of the
commodities, sometimes but not always involving long-distance haulage.
Corvee labor in the Ming again far exceeded the prototype that was
customarily employed in local road and waterwork maintenance, the latter
kind could be and was in fact called for beyond the regular li-chia rotation.
The li-chia obligation included labor services such as providing office
attendants for all echelons of the administration, from the county all the
way up to the imperial government. Aside from the doormen, guards,
messengers, and sedan-chair bearers mentioned earlier, the cooks, buglers,
boatmen, patrolmen, jailers, grooms at government stables, receiving men
in the warehouses, operators of canal watergates, and clerical assistants
wherever required, were also drafted from the general population.
Material requisitions from the li-chia were quite extensive. First of all,
each community supplied the local government offices with its share of the
required stationery, oil, charcoal, and candles. Military equipment in-
cluding swords, bows, arrows, and winter uniforms were also contributed
by the population. With few exceptions, all communities had their contri-
bution quotas. Botanic and mineral drugs for the imperial academy of
medicine were selected from the best specimens from various locales,
and practically every community had to furnish its share of these. Local
 1, II Rural organization and basis of taxation 35
delicacies were submitted to the court of imperial entertainments, as was
the calendar paper for the imperial astronomy. Palace supplies, notably tea,
wax, pigments, and lacquer, were regularly delivered by the areas which
produced them. To take an extreme case, when the Ning-kuo prefecture of
South Chihli was called upon to provide sweeping brooms for the imperial
palace, the prefectural quota was then divided among all the li-chia in the
subordinate counties.67 All the articles cited above were scheduled and
their quantities fixed on an annual basis. The local gazetteers called them
sui-pan, or 'annual contributions'. There were other items for which
quantities were not fixed and deliveries were spread over a number of years,
for example colored paper requisitioned once every three years and sulphur
and nitrates once every ten years for the ministry of works.68 The gazetteers
referred to these as tsa-pan, or 'miscellaneous contributions'.
In a different category from requisitions were purchase orders of the
central government that were fulfilled by the local government. As already
explained, these were called tso-pan or local procurement, and their costs
were supposed to be paid with the district's retained revenue. As will be
seen later (3, m), when after the mid sixteenth century some of the tso-pan
orders were unpaid or only partially paid, they were transformed into
sui-pan and fulfilled by the village communities. The sui-pan, tsa-pan and
tso-pan thus became the three 'contributions' of the local community.
Although they might appear under slightly different headings in the
financial sections of the county gazetteers, their occurrence was uni-
versal.69
Theoretically at least, all the materials on requisition enumerated so far
could have been furnished by the villagers in the form of local produce.
The service levy however also embraced a number of items that made cash
payment inevitable. It must be noted that other than the li-chia requisition,
the government had no funds for providing banquets at the local level,
for entertaining visiting dignitaries, or even for escorting and executing
prisoners. The expenses of official travel, constructing and repairing office
buildings, delivering New Year and birthday greetings to the emperor,
erecting monuments, and sponsoring local candidates for civil service
examinations were likewise borne by the li-chia, which was the regular and
only source of supply.
The actual assessment of the service levy on the villagers varied from one
county to another. The basic fiscal unit was the ting, or able-bodied male,
but the call for material and labor services was directed not to individuals,
but to households. In principle, the apportionment of the burden was to
take into account the number of male adults in the household and its
property-holding. Unlike the land taxes, which applied general rates to all
taxpayers, the service levy thus maintained some spirit of progressive
taxation. At the beginning of the dynasty all households were classified
36 Fiscal organization and general practices
into upper, middle, and lower categories, so that that service obligations
could be distributed accordingly.70 The i was therefore neither a poll tax
nor a property tax, but a combination of both. The general trend in the
later part of the dynasty was to emphasize the latter more than the former.
The gradual commutation of service obligations into silver payments and the
partial absorption of the service levy by the land taxes caused numerous
problems everywhere (3, in). The fundamental difficulty stemmed from the
fact that the two types of taxation had been based on divergent principles.
The demand for adaptation to local needs also conflicted with the require-
ment that at least minimal imperial uniformity be maintained.
The li-chia system and the service levy were clearly designed to suit the
village commodity economy. The extensive labor requisitions provided
the outlet for under-employment in the villages and the material requisi-
tions permitted local products to be submitted as tax payments without the
need to market them. When there was a fixed demand for service by the
government, the system was not at all absurd. Even though it resembled
drawing water from a deep well, not merely bucket by bucket but also drop
by drop, the villages' delivery of materials and provision of labor was
regular and automatic. This saved the government many logistical problems
and reduced its administrative overhead. Yet during the middle of the
dynasty the gradual sophistication of governmental functions and the
increasing demand for services, against a background of fundamental
economic change, made both the li-chia and the collection of the service
levy outdated. Though the chun-yao method and the Single Whip Reform
were provided as remedies for this, throughout the Ming period these
village organizations were never abolished, nor was the fiscal scheme of
defraying the operating costs of the government by direct impositions on the
villages discarded. The result was that the land taxes were complicated.
Since the service levy was partially and indirectly assessed on landed
properties it made the tax burden on the latter not only difficult to readjust
but also difficult to calculate.

Other service obligations fulfilled by the general population

In the early years of the dynasty the additional obligations were in the area
of transportation. One such was the 'tax captaincy' (liang-chang) estab-
lished in 1371. In general it applied to the central and eastern provinces
where the population density and the presence of large landowners made
the system workable. Local officials were ordered to divide their territories
into tax districts of an annual tax assessment of 10,000 piculs of grain
each. The largest taxpayer of each district became the tax captain. He was
responsible for mobilizing the population in his district and conducting
the designated delivery. An order of 1373 further specified that each tax
 1, II Rural organization and basis of taxation 37
captain should be assisted by one bookkeeper, twenty measurers, and
a thousand porters, all drafted from the tax-paying population.71 The
li-chia system and the tax captaincy complemented each other. The /*,
composed of 110 households, was roughly the size of an average village.
Tax captains, who numbered thirty or forty in a medium-sized county,
resembled headmen of townships. A tax captain might have ten, twenty, or
thirty li under his jurisdiction. The li chief collected tax payments in the
villages and furnished the tax captain with the required manpower. The
latter consolidated the payments and planned and carried out the delivery.
All administrative details pertaining to packing, routing, and temporary
storage of the tax grain, selection and procurement of transportation,
organizing the convoy and maintaining its logistical support were decided
by him. The transportation costs had been collected from the taxpayers on
a pro rata basis, but any losses and spoilage of the commodities in his
custody had to be made good by the tax captain.
The tax captain who was unpaid, received his commission from the local
magistrate, but had to proceed to the Nanking ministry of revenue to
receive the stub-book in person, and was held accountable for tax arrears
in his district. On the other hand he wielded unspecified power in the rural
areas. In the days of Hung-wu he was given audience by the emperor: in
1381 the sovereign is said to have received 1,325 tax captains from Kiangsi
and Chekiang on one day.72 Tax captains and their family members could
also in those days utilize their positions as stepping stones to governmental
posts. When they committed a misdemeanor, the punishment was drastically
reduced. A death sentence could be commuted to beating or fines.73
The foremost study of tax captaincy to date is that by Liang Fang-chung.
Liang's investigation, based on local gazetteers, discloses that the system
was definitely institutionalized in South Chihli, Chekiang, Kiangsi,
Hukwang and Fukien. It was also applied, probably to a lesser extent, in
Shantung, Shensi and Honan. Even in some remote provinces, such as
Szechuan, where it was not generally introduced, similar delivery agents
were sometimes appointed.74
The organization of the army transportation corps on the Grand Canal
(2, i) in time reduced the usefulness of the system. From the mid fifteenth
century the district under each captaincy was reduced in size and at the
same time the position was filled by several households jointly, signifying
that the Ming court could no longer commandeer the top gentry for this
purpose. The lesser landowners who came to undertake the task could
never be as effective tax expediters as their predecessors.75 In a rural area it
was obviously much easier for a man who owned 10,000 mou of land (see
p. 40) to give orders to owners of 500 mou than vice versa. Presumably
the decline of the tax captaincy as an intermediary between the govern-
ment and the general population also affected the operation of the li-chia
38 Fiscal organization and general practices
system. Inadequate control over the rural areas subsequently became one
of the fundamental defects of the fiscal system.
The tax captaincy survived as long as the dynasty, however. Right up to
the Ming collapse the annual delivery of some 214,000 piculs of highly
polished grain for palace consumption in Peking was still handled by
civilian agents called chieh-hu, or 'tax transmitting households', a modified
form of the captaincy. This consignment, along with many other mis-
cellaneous deliveries, was never taken over by the army transportation
corps (4, i).76

The imperial postal system operated 1,030 relay stations in the pro-
vinces.77 It was nominally subordinate to the ministry of war, but its
logistical support was entirely decentralized. In the Hung-wu period the
duty of servicing the stations was assigned to individual taxpayers of
substantial means, outside the li-chia system, or to political prisoners whose
punishments were commuted to service obligations. During the middle
period the burden was gradually shifted to the general population. At the
same time the postal stations began to find that their primary function was
no longer sending on documents, but rather providing transportation and
hostel services for travelling officials and foreign tributary missions. The
demand for sedan chairs, horses, boats, food and drink, along with labor
requirements, rose sharply. Local li-chia, sometimes assisted by adjacent
districts, filled their requisition orders. The additional service for the postal
stations, the i-cKuan, was however separated from the regular li-chia
account, though the burden was borne by the same taxpayers. This was
partly because the services and supplies were delivered via a different route
but it was also due to the fact that the i-cKuan account was an unstable one,
with a constant tendency to expand.

In general, the service levy under the li-chia system increased steadily during
the dynasty. The chiin-yao (equalization of labor service) and the min-
chuang (militia service) were introduced in the late fifteenth century, and
for regional defense, the ping-hsiang (military supply) was added in the
sixteenth century. These will be dealt with in later chapters (3, m and iv).

Main features of land tax assessment

The land tax (fu) provided by far the largest item of the state income. Even
excluding the surtaxes and surcharges, it still produced on average some
27 million piculs of husked grain each year. The salt revenue, the second
largest item, generated approximately only 10 per cent of this amount in
comparable monetary value. But the collection of land taxes was a very
complicated matter. To explore these complexities is one of the main
 1, II Rural organization and basis of taxation 39
purposes of this work and will occupy a large part of subsequent chapters.
Here only the most outstanding features of the tax assessment will be
outlined.
The land tax followed the previous 'two tax system', which meant that
the assessment was based on productivity. The' summer tax', paid basically
in wheat, was originally set to be collected in the eighth lunar month of the
year. The 'autumn grain', consisting basically of husked rice, was to be
collected in the second month of the year after the harvest. Land which
yielded double crops paid taxes twice.78 Under previous dynasties the
summer tax also comprised a number of commodities such as cotton, silk,
and tea, and the Ming in general retained those items.
For the basic assessment the general fiscal unit was the picul of grain,
either in husked rice or wheat depending upon local produce. A picul of
wheat was considered as equal in value to a picul of husked rice, though the
former was worth considerably less. But the equalization was only for the
purpose of simplifying the accounting and no taxpayer could take advan-
tage of the price differential. When the commodities were commuted to
silver payments commutation rates for husked rice were generally higher
than those for wheat.
As early as the Hung-wu era substitutes were accepted. Land taxes in
Yunnan were customarily paid in precious metals and mercury and even
cowrie shells.79 In other regions sorghum, millet, and beans were accepted
at discount rates. The acceptance of substitutes prior to the settlement of
the regional tax quota should not be confused with commutation. The
order was rather the reverse, with substituted commodities calculated in
terms of quantities of staple grains so that the figures could be integrated
into the national accounts. Such substitutes, at all events, did not constitute
a high percentage of the total tax income.
Commutation took place occasionally in the early years but became
frequent in the sixteenth century. It could be ordered in two stages. For
instance, a picul of husked rice was first commuted to a bolt of cotton
cloth, and then the cloth was commuted to 0.3 taels of silver. The two stages
could be separated by a span of over a century. There were also cases where
one was permanent and the other temporary. The commutation rate might
or might not follow market prices. Occasionally commutation was offered
at deliberately lower rates as a form of tax relief. A particular rate therefore
applied only to one specific tax consignment. It is difficult to say which order
of commutation was permanent and which temporary. In general an order
which had been in effect over a score of years could be considered to be
permanently binding, yet there was no absolute guarantee that it would not
be reversed or modified. In the sixteenth century court orders became more
explicit, usually declaring whether the commutation was to be temporary or
permanent.80 The rates moreover corresponded more closely to the market
40 Fiscal organization and general practices
prices. Despite the numerous commutations husked rice and wheat re-
mained the basic tax standards throughout the dynasty. Even after a county
collected more than 90 per cent of its land taxes in silver, the silver was still
computed from the grain assessment.
Taxpayers were required to deliver the payments to distant granaries and
warehouses. In the beginning the state was not concerned with the collection
of transportation fees. Local officials fixed such rates simply to prevent
excessive collection by tax captains. But after some transportation was
taken over by the government, as in the case of the tribute grain transported
on the Grand Canal, it began to count the transportation fee as a part of
its regular income. Even after the basic assessment was commuted to
silver, the transportation fee was still assessed as if bulky cargo were in
transit and still computed by the distance of haulage. In a few extreme cases
the surcharges thus computed exceeded the value of the basic payment.
The actual burden of' 1 picul of grain' therefore varied widely, depend-
ing upon whether the payment was made in silver, grain, or another
commodity, what kind of transportation cost was involved, and the com-
mutation rate applicable to that picul of grain. The most 'expensive' picul
could cost the taxpayer seven times the amount paid for the least expensive
picul.
Surtaxes were distinguished from the surcharges. They usually involved
hay, hemp and raw silk, collected along with grain payments in the pro-
ductive areas. Nor should they be confused with li-chia requisitions, though
in some special cases the same commodity, such as silk, could be duplicated
in both surtaxes and li-chia annual contributions. Occasionally a county
could have submitted a certain number of rolls of silk fabric in lieu of raw
silk in its tax quota, only to receive another local procurement order from
the court also asking for silk fabrics, although the latter were often specified
to be of higher quality.

Tax payments were determined by acreage under cultivation, the only

exception being the land taxes paid by the aboriginee tribes in the south-
west provinces where the chieftains handed over lump sums, the amount
having been arrived at through negotiation rather than investigation of
acreage. For other agricultural land the unit of measurement was the mou.
Five feet constituted a pace, and 240 square paces a mou. A standard mou,
therefore, covered approximately 6,000 square feet and was about the
acreage of two tennis courts combined. Normally such an area of agri-
cultural land in south China could be expected to produce 2 piculs of
husked grain annually.81
The standard mou was more a concept rather than an actual fiscal unit.
Contemporary sources indicate that on the richest soil in the Yangtze delta
a standard mou could yield 3 piculs of husked rice. There were also claims
 1, II Rural organization and basis of taxation 41
that the yield per mou could reach 4 piculs.82 In the arid northwest, on the
other hand, it was not uncommon that the yield per mou was below half
a picul. Furthermore, on the less productive soil dry cereals were planted
which had a lower market value. Sometimes wide variations in fertility
could prevail in the same locality due to a peculiar water supply situation.
The variations were further intensified by differences in labor supply. In
general the most fertile land needed the least labor. Conversely, the least
productive land demanded far more labor investment in irrigation and thus
often cut per capita production to the minimum, such as in the case
reported by Ho Liang-ch'iin (1506-73) in the mid sixteenth century in his
native Hua-t'ing county of South Chihli, that five mou of land required
the continuous toil of a husband-and-wife team.83 As universal application
of the standard mou would clearly be unfair, an alternative was found in the
so-called 'fiscal mou'.
Ping-ti Ho, in connection with his study of the Chinese population, has
amassed much data concerning fiscal mou conversion from an extensive
survey of local gazetteers. In general, each mou of standard measurement
which had a normal yield or better was still counted as a fiscal mou. The
less productive land was converted into fiscal mou by counting one and
a half, two, three, or even eight mou as a fiscal mou. The conversion was
not based on any uniform scheme decreed by the central government, but
rather every district was left to find its own formula. In some special cases
the 240-square pace standard was completely ignored and instead, the local
district introduced its own unit of measurement for the sake of con-
venience.84 The conversion method consequently showed a great variety;
some of them were undoubtedly based on local usage and had a historic
origin. Yet the conversion was by and large reasonable. After scrutinizing
the gazetteers, one is left with the impression that through all the various
devices an effort was made to guarantee that a fiscal mou of agricultural
land should have a minimum annual productivity of 1 picul of husked rice
or its equivalent. In many districts in the southern provinces, the level
could have been 2 piculs, though this observation cannot be sufficiently
documented as yet. The extant materials seem to suggest that even the fiscal
mou conversion was not formally authorized by the central government.
All this bears out the earlier observation that the attempt to impose a
uniform land tax throughout the vast empire extended centralization beyond
the technical capacity of the times. Though the Hung-wu Emperor pro-
claimed uniform tax rates for the individual prefectures, this goal was
nowhere attained. The uniform rates mentioned in the Ta-Ming Hui-tien,
namely 0.0335 piculs per mou and 0.0535 piculs per mou applicable to
private and public land respectively, were no more than a general guideline
in fixing the level of collection.85 Put into effect only in a few newly-settled
counties and prefectures in north China, even this guideline still required
42 Fiscal organization and general practices
local revisions and internal readjustments. In south China the taxable land
often included hills, ponds, and marshes, and frequently all these topo-
graphical features were present in the same piece of landed property. No
uniform rates were possible. Those properties which had above normal
yields also deserved to be taxed at higher rates. In addition, public land
inherited from earlier dynasties, the confiscated land added by the new
dynasty, and properties of which the tenure was in question, all presented
problems of readjustment, inasmuch as under the Ming no attempt was
made to separate rents due on government land from regular land taxes.
Consequently, after the fiscal mou conversion each county was still re-
quired to differentiate the tax rates into different grades. In the later
centuries, 5 or 6 grades could be considered typical for a county in north
China. In south China a county's tax grades were seldom fewer than 20.
In 1543 Hu-chou prefecture of Chekiang reported that it administered 599
grades of basic tax rates. Cheng Hsiao (1499-1566), writing at about the
same time, disclosed that in seven counties of this same province, the tax
rates ramified into 800 grades. With the inclusion of surtaxes and sur-
charges, those grades could swell to the thousands.86 Some of the com-
plexities undoubtedly accumulated during the dynasty's later days; but
their foundation had already been laid immediately after its birth.
Complicated and ever-changing topographical features also presented
a severe obstacle to the tax administration. There is no indication that under
the Ming this obstacle was overcome. A laconic entry in the Ming-shih
created an impression that during the Hung-wu period a national land
survey had been conducted and the returns had been edited into the 'fish-
scale book' (yii-lin fu-ts'e), so-called because the topographical charts
which outlined the contours and boundaries of every plot of land under
cultivation resembled fish skin in appearance.87 Recent studies, however,
point out that the statement is misleading. A land survey actually was
conducted in Chekiang and South Chihli in 1386 and completed early the
following year,88 but it was not a national land survey.89 While in other
districts the fish-scale book is occasionally mentioned, there is no evidence
that the field investigation followed a general standard or was under central
direction. The fish-scale book was in fact by no means a Ming invention;
its origin could be traced at least to the Sung. The Mongols had also
prepared it in several southern provinces.90 On the other hand in the case
of some districts in north China, such as Chi-hsien in Honan and Ta-ming
prefecture of North Chihli, it is known that no fish-scale book had ever
been compiled there until the sixteenth century.91 The multi-standard in
fiscal mou conversion is further evidence that in the early Ming period
there was little coordination in compiling land data. Doubtless it is too
much to expect the Ming administrators in the fourteenth or fifteenth
century to have overcome all the technical difficulties inherent in any
 1, II Rural organization and basis of taxation 43
attempt to establish a general standard for land classification, whereby all
the cultivated acreage in China could be compiled in a few categories
without disparity. Such a scheme might be difficult to effect even in modern
times. Yet it is clear that while this fundamental problem remained un-
solved the type of centralized administration attempted under the Ming
was over-ambitious.
 2
The heritage of the sixteenth century
and major fiscal problems
The origins of the problems treated in this chapter can for the most part
be traced to the founding of the dynasty. When Hung-wu proclaimed
himself emperor in 1368 he enjoyed freedom of action that few dynastic
founders in Chinese history could ever have dreamed of. He had over-
thrown a hated and much discredited alien rule and he was not obliged to
observe any existing statute or even customary practice unless it served
his purpose. The country, exhausted by civil strife, was ready to accept
deliverance by any form of law and order. In organizing his empire one of
Hung-wu's major concerns was the preservation of his own power. His
bureaucrats were expected to carry out the wishes of the crown, not to
exercise their own initiative. Local officials were not even permitted to
enter the rural areas.1 Villages were organized into self-governing units,
with 'virtuous elders' assuming the responsibility of disciplining the popu-
lace in each local community.2 In fiscal administration, priority was given
to accounting control rather than to field operations. The emperor's
frugality was such that both the government budget and administrative
overhead were reduced to a minimum.3 Since the supply procedure laid
stress on lateral transactions at the lower level, there was no need to build
up the logistical capacity of the middle echelon.
This approach eschewed the possibility of deriving administrative
efficiency from basically sound organization and depended rather on in-
cessant regulation and supervision by the sovereign himself. Though the
general level of taxation was low, tax legislation was promulgated by the
emperor in the capital, with little regard to local conditions. The emperor's
own writings indicate that his fiscal legislation was beset by numerous
technical difficulties.4 When tax arrears occurred, tax deadlines could not
be met, imperial decrees were misinterpreted, or regional fiscal accounts
could not be integrated at the intermediate level, the monarch refused to
retreat from his previous position. He resorted to a reign of terror, executing
those officials who had failed to carry out his orders, perhaps because he
thought that the technical difficulties could have been overcome by force.
The bloody persecutions in 1382 and 1385, both the result of discrepancies
[44]
 The heritage of the sixteenth century 45
in fiscal accounts, involved thousands of offenders.5 Leaving aside the
extreme cruelty which it entailed, the major result of this unrealistic
approach was thatfinancialorganization under the Ming was never placed
on a rational basis from the very beginning.
The lateral transfers of materials and goods at the lower level, as has
been pointed out in the previous chapter, were intended to save service
facilities. But Hung-wu, in his desire for economy, extended the use of this
type of transaction to extreme lengths. In 1388 the emperor invented
a method of supply which required each county to service an army unit in
its vicinity. The grain payments were not to go through the regular govern-
mental channels at all, but were to be handed by the taxpayers directly to
the soldiers. Ying-t'ien prefecture in South Chihli was ordered to experi-
ment with the new procedure for a year. The 5,000 soldiers on duty with
the Ching-wu Guard no longer received their rations and pay from their
paymasters. Instead, more than 5,000 civilian households fulfilled the
latter's function, delivering their tax payments to the homes of the army
personnel. The scheme was declared workable and was supposed to be
extended over the whole empire in 1390.6 Although this fantastic procedure
was later discontinued, the principle of disbursing tax revenues without
first consolidating the payments persisted and became a general practice
under the Ming. It was also followed by the Ch'ing until the late nineteenth
century (see chapter 8).
Hung-wu's fiscal scheme, while crude, served his purpose. The dynastic
founder had no ambition to expand his realm beyond the boundaries of
China.7 Within the country he was content to rule in a spirit of agrarian
simplicity.8 After Yung-lo took over in 1402, however, those restraints
were altogether abandoned. The new emperor, impatient with details, gave
but little thought to internal organization. As long as he could make the
populace deliver the needed supplies for his military campaigns and con-
struction projects (see 2, i below), he never gave serious thought to fiscal
procedures.
As the dynasty entered the second quarter of the fifteenth century, the
constitution of the empire imperceptibly altered. The power that had
founded the new dynasty was already a thing of the past. The army declined
and the navy all but disappeared. The emperor no longer ruled the nation
directly but merely controlled the bureaucracy. It was the absolute
obedience of bureaucrats to the throne that now supported the absolute
monarchy. Ideological cohesion among the civil officials was now of
greater importance than the military or financial strength of the state. The
general mood in Peking was now one of preserving the status quo. The
bureaucrats, who in effect represented the top gentry, were ready to main-
tain a consensus based on traditional Confucian political economy but
regarded any major innovation as unorthodox and therefore suspect.
46 The heritage of the sixteenth century
A general reconstruction of the financial establishment was now no
longer possible. Even when circumstances required institutional reforms
by the government, it merely effected makeshift revisions and compromises.
The administration's greatest handicap was that state revenues were
fragmented. Even though nominally under centralized control, in practice
they could not be integrated. The second half of the fifteenth century was
a particularly lackluster period in the dynasty's history. The emperors made
few major decisions on public affairs, yet palace expenditures increased
while all state institutions declined.
In the sixteenth century the crisis, developing on many fronts, could no
longer be ignored. The most pressing problems were the deterioration of
the armed forces and the monetary system, and most of the readjustments
made by fiscal administrators in this century were aimed at relieving these
two difficulties.

I THE LEVEL OF STATE INCOME AND MODIFYING FACTORS

The quota system
Sixteenth-century writers often create the impression that both state
expenditures and taxation had in their day reached precipitous heights.
The argument is misleading. On the contrary, the major problem in their
time was that the projected state income was too small and its level not
readjustable. In 1502 the ministry of revenue submitted a long report to the
Hung-chih Emperor listing all major state incomes. Of key importance
were the regular land taxes, estimated to have comprised close to 75 per
cent of all receivables. The annual amount, still calculated in grain, was
given as 26,799,341 piculs.9
Under the Sung, state income and expenditure had both been assessed in
terms of a standard fiscal unit, namely the string of copper cash, the value
of which has generally been considered comparable to that of the picul of
grain under the Ming. The Sung accounts show that by the mid eleventh
century the annual state budget had already reached a level of between
126 million and 150 million units.10 Although these accounts have yet to be
studied in detail to discount inflationary influences, there is little doubt that
the financial capacity of the Ming was well below that of its forerunner
some four centuries earlier. Moreover in Sung times, the state produced
nearly 3,500 short tons of copper and 5,000 short tons of iron annually in
the 1050s. In 1159, it generated an income of 2 million strings of copper
cash from the maritime tariff.11 The Ming records of the sixteenth century
show nothing comparable.
The narrow tax base in the sixteenth century was the direct result of the
fiscal policies of the Hung-wu Emperor, in whose reign a tax quota system
 2,1 The level of state income and modifying factors 47
was put into effect.12 In 1377 the monarch dispatched teams composed of
ministerial officials, imperial university students and eunuchs to tour the
178 local business tax stations and fix their tax quotas. 13 In 1385 he ordered
that stone tablets inscribed with the tax quotas of the provinces and pre-
fectures be erected inside the ministry of revenue's office.14 In 1393 the
income from the land taxes reached the figure of 32,789,900 piculs.15 The
sovereign, satisfied with this, subsequently declared that newly-reclaimed
land in the northern provinces was never to be taxed,16 when taxation by
regional quota became an unwritten law. Though minor readjustments
were made from time to time the basic settlement was never abandoned.
In 1412, in the reign of the Yung-lo Emperor, taxes from agricultural
land were said to have reached an all-time high of 34,612,692 piculs.17 The
nature of the increase has never been adequately explained. It is likely that
an additional quota for Annam, recently annexed as a new province, was
included in this total, 18 for when revenues in that territory proved to be
uncollectable the empire's land tax income was re-stated as being close to
30 million piculs.19
A further revision of the tax ceiling came in 1430. During the decade
when Peking was occupied with the Annam problem the landowners in the
Yangtze delta, resentful over the high surcharges added to their land taxes,
deliberately kept their payments in arrears until the uncollected amount
exceeded three years' total income. As a major concession to those districts
the Hsiian-te Emperor authorized a general tax reduction; some 3 million
piculs of the national total were written off, most of the benefit going to
the lower Yangtze districts.20 The lost tax income was not, however, com-
pensated for by additional collections elsewhere. Subsequently the annual
projected income remained at some 27 million piculs, regardless of increases
in population and land brought under cultivation. In fact in their reports
to the imperial government most provincial and local officials simply re-
submitted earlier acreage figures as current data (2, n). On the few occasions
when increases in cultivated acreage were reported, the general rule was
that they be utilized by the local administration for internal tax reapportion-
ment, and not serve as a basis for tax increases. Thus taxation became
virtually independent of cultivated acreage, a situation which the 1502
report of the ministry of revenue only confirmed. It should be noted that
fixed tax quotas were basically a Ming policy and that this approach was
never followed so rigidly by the T'ang or Sung.
In addition to the land taxes the Ming also relied heavily on requisitions
of labor and basic commodities, again in accordance with the Hung-wu
Emperor's fiscal policies. Here too the quota system was applied; its
unsatisfactory results will be discussed in later chapters.
48 The heritage of the sixteenth century

Inadequate budgeting
A misconception generally subscribed to by Confucian bureaucrats was
that a low level of taxation always automatically benefited the taxpayers.
Ming experience shows that this was not always the case. Insufficient tax
income meant that the government was unable to husband the empire's
resources to the best advantage, which in fact proved to be a disservice to
the taxpayer. In the sixteenth century inadequate financing resulted in the
failure of many governmental functions, notably coining money (2, iv) and
salt distribution (5, iv), thereby causing considerable distress to the people.
Furthermore, though formal taxation might fail to cover expenditures,
indispensable items still had to be paid for in one way or another. The
resulting unauthorized collections made effective auditing impossible and
the funds so obtained were usually misspent.
Rampant official corruption under the Ming was another result of in-
adequate budgeting. The operating expenses of many offices were clearly
below the necessary minimum level (4, v), and the rates of government
emoluments were absurd. The Ming civil service was very small. In 1371
provincial and local officials totalled only 5,488.21 In 1455 there were
1,520 civil officials in service at Peking.22 Even in the early sixteenth
century when the offices had been significantly enlarged, there were still
only 20,400 civil officials throughout the empire. The lesser functionaries
totalled about 51,000 persons but they served both the civil government
and the army.23 The state was unable to pay reasonable salaries to this
small administrative body.
The earliest salary schedule,fixedin 1392 and nominally in force through-
out the dynasty, was not generous. It provided annual grain payments of
1,044 piculs for the top rank, la, which were scaled down to 60 piculs
for the lowest rank, 9b.24 From the late fourteenth century, however, the
payment had been partially substituted by paper currency.25 In the fifteenth
century the items of substitution included silk fabrics, cotton cloth, pepper
and sapanwood. The conversion rates were so fixed that the substituted
portions, comprising 50 per cent and more for rank 4b and above and
smaller percentages for the lower ranks, were virtually unpaid. It is estimated
that the conversion rates adopted in 1434 reduced the payments to 4 per
cent of their original value.26 In 1432 some officials received confiscated
garments and salvaged materials as part of their salaries.27 In 1472 peas
were given out as partial emoluments. The following year in Nanking it
was found that the peas were suitable only for feeding horses.28 During the
three years from 1475 to 1477 even these substitutes were withdrawn.29
Officials supplemented their nominal salaries with the commutation
payments of the personal attendants (tsao-li) furnished them by the
government, a practice started in the early years of the fifteenth century; the
 2,1 The level of state income and modifying factors 49
officials sent the attendants home after demanding from them 12 taels of
silver per man. After 1429 the practice became generally recognized. The
ministry of war, which had called on such personal attendants as a part of
local corvee labor service, waived the service call altogether and instead
collected the payments and distributed them among the officials.30 The
minister of revenue, ranked 2a, in this way received an extra 144 taels of
silver annually, not from his own treasury, but from the war office. The
amount was probably just sufficient to pay his household staff.
In the late fifteenth century the portions of official salaries that had been
paid in grain were converted to silver, but the payments were so low that
the total of all "salaries paid in Peking remained an insignificant item in
the general account (7, i). The financial integrity of government officials
consequently became a relative matter.
Throughout the fifteenth century there were numerous cases in the
capital of officials heavily involved in corruption. The resulting scandals
exposed such senior dignitaries as the Chief Censor Liu Kuan (in office,
1415-28), Minister of Personnel Yin Min (in office, 1473-86), and Minister
of Rites Chou Hung-mo (in office, 1481-8).31 Minor infractions of the code
are too numerous to be cited. One private account indicates that the most
venerated Minister of Personnel Ni Yiieh (in office, 1500-1), who had a
name for righteousness and honesty, was in reality not free from the sus-
picion of irregular practices.32 In 1470 Minister of Personnel Yao Ku'ei
reported that his office was haunted by professional moneylenders seeking
information. Those persons offered selective loans to officials in the capi-
tal, and when they received provincial appointments accompanied them to
their new posts. Some of the appointees, the minister added, never again
freed themselves from the clutches of their creditors.33 The situation appar-
ently worsened in the sixteenth century. The recorded impeachment cases
reveal a general lowering of moral standards: offenses considered serious
in the previous century were now no longer recognized as such. Capital
officials, being virtually unpaid, received no travel allowances when they
were transferred to provincial posts. Ku Yen-wu in the early seventeenth
century confirmed that those officials had no way to reach their new posts
other than raising loans.34

The expandable and contractible fiscal unit

Though the budgetary ceiling of the proceeds from the land taxes remained
fixed at some 27 million piculs of grain, the government could not always
limit its expenditures to this prescribed level. One of the most convenient
ways of creating extra income was to make fiscal units expandable. The
population was ordered to provide the state with certain amount of goods
and labor which nominally were paid for with the tax grain collectable in
50 The heritage of the sixteenth century
the local districts, as local procurement orders (tso-pan, 1, n). In reality the
taxes thus written-off constituted only a token payment, a mere fraction of
the value of the conscripted labor and the commandeered goods. The Yung-
lo Emperor undoubtedly adopted this practice on a large scale. The records
of his military campaigns, palace construction, and six major maritime
expeditions indicate that their actual costs far exceeded the normal income
of the state; at a guess by two or three times.35 In 1422 Minister of Revenue
Ko Tse reported that of the land taxes due in the fiscal year from 1419 to
1421, less than 23 million piculs of grain had been actually turned in at
imperial granaries.36 The annual average over the three-year period was
less than 8 million piculs. There is reason to believe that the bulk of the tax
proceeds had been disbursed to pay for the goods and services of the
populace at inflated rates. No actual accounts are available but the records
relating to construction labor, the shipbuilding program, and procurement
orders for lumber and other commodities bear witness that no clear distinc-
tion was drawn between government purchases and outright requisitions.37
Official payments, even when these were made, were never intended to cover
the whole cost. In these circumstances allfinedistinctions relating to the state
budget, tax rates, regional quotas and accounting systems were irrelevant.
No Ming emperor after Yung-lo dared to try the taxpayer's tolerance so
far. Several of his successors in the early fifteenth century in fact ordered
tax concessions (including the tax reduction of 1430 by Hstian-te), that
were designed to mitigate the grievances of the general public after Yung-
lo's excessiveness.38 Nevertheless the restraint was only relative. Partly
owing to technical difficulties the basic fiscal unit, the picul of grain, still
had no absolute monetary value assigned to it. Occasionally it was expanded
by the government in order to make up for a deficit. For example, tax-
payers in Shansi were required to deliver portions of their tax payments to
the northern army posts. In 1443 the payments were converted at 0.25 taels
of silver per picul. In 1457, only fourteen years later, the rate was revised
to 1 tael per picul, thus increasing the actual payment four-fold.39 Such
practices not only distressed the taxpayers, but also greatly increased
accounting problems.
Because of the quota system and the insufficiency of tax income, the
higher offices tended to pass their fiscal responsibilities down to the lower
echelons, which then had to cover the deficits as best they could. It was not
uncommon for the drafted civilian tax agents to be called upon to bear the
burden. No saving resulted, merely individual injustice. These practices,
though of historic origins, were more widely applied under the Ming
because the projected governmental income was so low.
Closely related to the issue of the expandable fiscal unit was the practice
of collecting surcharges, which particularly affected the 'tribute grain'
(ts'ao-liang) - the 4 million or so piculs of husked rice collected as part of
 2,1 The level of state income and modifying factors 51
the land taxes in five provinces and South Chihli, and hauled by the army
transportation corps up the Grand Canal to Peking.40 The original practice,
devised under Hung-wu and Yung-lo, had been for the population to
deliver its tax grain at any depot designated by the state, paying all the
transportation costs incurred. After the army transportation corps took
over the long-distance delivery the costs were still paid by those districts
which furnished the grain and, even in the fifteenth century, by the particular
taxpayers in those districts. These surcharges were not credited toward their
tax payments but added as an additional service obligation. This might
not have created such a problem if the surcharges had been small, but they
were large. For the grain consignments from Hukwang, the farthest
province involved, the surcharges covering the handling losses alone added
80 per cent to the basic payments.41
The heavy surcharges were not entirely unjustified. Long-distance
delivery of grain involved many transfers before the destination was
reached. Porter and wagon services were required to clear the canal
sluices; lighters were called for when canals and rivers became too shallow.
These transfers resulted in losses. Husked rice, when dampened, ferments
easily. Quite frequently after each transfer the grain had to be sun-dried,
during which process its volume shrank drastically, by as much as 8.5 per
cent in five hours42 (as established by an official experiment).
Matters would have been simplified if the government had considered
such inevitable losses as part of the expected operating expenses. But
under the quota system and in accordance with Ming accounting methods
every item of receivables had to be counted down to the last peck, and
there was no way to cover extra expenses except by additional collections
from the taxpayers. In practice the surcharges were not consolidated as
a single item, but divided under six or more headings, including extras to
cover anticipated losses, mat-cover fees, lake-crossing fees, Yangtze river
ferry fees, etc.43 In the sixteenth century when portions of the tribute grain
were commuted to silver payments, the conversion had to be calculated as
including these miscellaneous items. The rationale behind it was that the
taxpayers and the tax-paying districts were already used to paying the sur-
charges, while the government needed the additional cash income. In fact,
even when the payments were made in kind the surcharges were calculated
to produce a surplus, which was already counted as a budgetary item. It
certainly could not be dispensed with.
To complicate the matter further, in 1471 the transportation corps also
took over a number of grain consignments which up to then had been
delivered by the taxpayers themselves at the auxiliary granaries on the
Grand Canal. These consignments, even though they originated from the
same local districts, were not combined with those which had been taken
over in the early fifteenth century. They were designated as 'added trans-
52 The heritage of the sixteenth century
mittance' (kai-tui), to be distinguished from 'regular transmittance'
(cheng-tui).^
The customary surcharges on added transmittance were much lower
than those on regular transmittance. In the mid sixteenth century when
a taxpayer in Yang-chou prefecture of South Chihli was assessed 1 picul of
grain in regular transmittance, he actually paid 1.73 piculs. For added
transmittance, however, he would only be required to pay 1.27 piculs.
A commuted picul of grain in regular transmittance cost the taxpayer about
1.2 taels of silver but in added transmittance entailed a payment which
varied from 0.5 taels to 0.7 taels depending when the commutation order
was issued.45 The resulting confusion is apparent: tax receivables in the
official accounts did not correspond to the actual financial burden of the
taxpayers. Sometimes there was no way of ascertaining the land tax rates
in a particular district, even when the basic assessment was quoted in
piculs. In other words, the fiscal unit was a relative, not an absolute index;
it was partly abstract and partly real.
The picul of grain could also be contracted to reduce its value. An
outstanding example is the so-called 'Gold Floral Silver' (chin-hua-yiri),
institutionalized in 1436. Outwardly, the move was to facilitate tax trans-
mittance. A total of 4.05 million piculs of tax grain from six southern
provinces and South Chihli, supposedly from those districts where no
adequate transportation was available, was commuted at the flat rate of
0.25 taels per picul, with virtually no surcharges added.46 In reality how-
ever, the quota was almost exclusively taken up by the richest prefectures
where transportation was not a problem. The circumstances under which
the commutation was ordered suggested that it was another major con-
cession by the court to the landowners in the districts concerned, who had
been showing dissatisfaction at paying higher taxes. The sponsors of the
proposal, Minister of Revenue Hu Ying (in office, 1431-6), Nanking
Minister of Revenue Huang Fu (in office, 1432-40), and Governor of
South Chihli Chou Sheng (in office, 1430-50), were all champions of tax
reduction. The Cheng-t'ung Emperor, who approved it, was then only
nine years old and could not have understood all the implications of the
move. Because the payment was set lower than the grain price and handling
costs, the measure thus alleviated the tax burden in these districts without
overtly revising their regional quotas in terms of piculs of grain. The
income, slightly over 1 million taels, was subsequently known as Gold
Floral Silver, and was never to be dispensed with throughout the rest of the
dynasty. In accordance with tradition, the bullion was turned in at the
Inner Ch'eng-yun Treasury, and was used by the emperor to pay army
officers in Peking. The rest became his privy purse and beyond the auditing
of the ministry of revenue.
The implications of the Gold Floral Silver were profound. It permanently
 2,1 The level of state income and modifying factors 53
cut off 15 per cent of the proceeds from the land taxes and created another
discrepancy between the nominal revenue and actual income.

Lack of service support

The system of lateral transactions accorded well with the establishment of
regional tax quotas; both emphasized operations at the lower level and
could be linked together on a semi-permanent basis. But this mode of
administration inevitably retarded the development of service facilities.
The financial establishment was, therefore, preoccupied with tax collection
and delivery and neglected to build up logistical capacity at the middle
level. The basic services rendered by the lower echelon to the higher offices
lacked service support in the form of maintenance and subsidies from the
latter. In the long run this clearly made operations more expensive. Much
effort was duplicated and the materials and labor were either wasted or
could not be accounted for.
This lack of service support can be seen in the administration of the
Grand Canal. Of the three dynasties which preceded the Ming, only the
Yuan concentrated on transportation by sea. Both the T'ang and Sung had
relied on inland waterways for the redistribution of the nation's financial
resources. Under the two previous dynasties the canal administration was
in the hands of one or more 'transportation commissioners' (chuan-yun-
shih). Those officials managed the regional surpluses turned over by the
provincial administrators and usually also controlled the salt monopoly.
Supplying tax grain was only a part of their function. Many of the com-
missioners were empowered to dispose of government properties at their
own discretion, and were expected to make a profit by buying merchandise
at one end of the waterway and selling at the other. Revolving funds were
provided to facilitate the operation. Their jurisdiction covered the construc-
tion of service ships, the organizing of the crews, and the maintenance of
the canal system. In other words, they combined the functions of regional
treasurer, transportation authority, and purchasing agent, and operated
on a commercial basis.47
The anti-commercialism of the Ming prevented government agencies
from participating in trade. Yet the entire tribute grain system and the
canal administration received no central financing. The waterway was
maintained by local corvee labor without any subsidies from the central
government.48 The transportation corps by the mid fifteenth century had
121,500 officers and men operating 11,775 grain boats. The personnel drew
their pay and rations from the 124 guards and independent battalions from
which they were detached.49 Even the construction of service craft, carried
out every ten years, was financed partially by deductions from the payroll
of the soldiers operating them, and partly by remittances from those
54 The heritage of the sixteenth century

Peking T'ung-chou
# ~ Chang-chia-wan
Ho-hsi-wu

•Tientsin"
Ching-hai

L'ing-hsien
Ho-chien •
Hsing-chi)
Ts'ang-chou "Xf—~N
Tung-kuang

—— Grand Canal
—— Chiang-nan Canal

0 Mis 100

FIG. 2. The Grand Canal (Ts'aoHo), ca. 1610.

 2,1 The level of state income and modifying factors 55
districts which furnished the tribute grain.50 The transportation costs, as
noted above, were collected from the taxpayers on a pro rata basis, any
surpluses on the surcharges were surrendered to the government treasury
and subsequently transferred to Peking (see 'speed-the-delivery money',
6, m). Thefinancingwas thus so decentralized that the imperial government
took no fiscal responsibility for the operation, but rather derived extra
administrative income from it. A censorial official was appointed canal
commissioner, and an army officer was made commander-in-chief of the
transportation corps. Neither built up logistical support for their sub-
ordinate units. Junior officers, down to captains and lieutenants, were made
financially responsible for the tax grain in their custody, which could only
be discharged at the T'ung-chou granary.
The decline of the service fleet was thus hardly surprising. In the late
fifteenth century it was frequently reported that soldiers' pay and rations
were in arrears and the service units short of personnel. Some detachments
of the transportation corps had collective debts of several thousand taels of
silver or more.51 In the early sixteenth century some army officers committed
suicide and another found sanctuary in a Buddhist monastery after failing
to discharge their delivery obligations.52 The transportation corps survived
merely because the court, realizing the inadequate maintenance of the
servicefleet,relaxed administrative discipline. Permission for army person-
nel to carry a limited amount of private cargo for profit had already been
authorized early in the fifteenth century.53 Toward the end of the century,
the soldiers, rarely-paid, had to rely on this private trade for support.
In 1480 it was reported that the southbound service boats were heavily
loaded with commercial cargo while civilian vessels returning to the south
had nothing to carry.54 The state lost through the unpaid customs duties.55
The operation of the Grand Canal is only one example of the inadequacy
of government service facilities under the Ming. That the logistical system
established by the Hung-wu Emperor was never reformed was partly due
to the low level of state income. In the sixteenth century every particle of
government revenue had already been counted as a budgetary item;
through standing delivery procedures it had already been disbursed before
it was collected. Even revenue agencies were not permitted to utilize their
internally-generated funds to improve their services. This lack of investment
prevented the expansion of government enterprises, a notable example
being the records of government mining (6, i).
The size of the industries under government control and the huge quanti-
ties of raw materials, goods and manpower mobilized by the Ming state
were apt to impress and mislead the observer. In reality these financial
resources were accumulated in an irregular and disorganized manner. The
higher offices which controlled them had neither the authority nor the
capacity to integrate the operations of their subordinate units. The
56 The heritage of the sixteenth century
budgetary income of each of those offices was actually only a projected
maximum. Only under the most ideal conditions could the funds arrive in
full, because the diffuse supply procedure inevitably led to delinquencies
by the delivery agencies responsible for handing them over; at the same
time, naturally no agency would deliver more than was required. In brief,
all such organizations had a 'built-in' tendency to degenerate. The govern-
ment shipyard near Huai-an may be cited as an illustration.
The Ch'ing-chiang-p'u dockyard was one of the largest public works
establishments in the empire, and constructed about half of the grain boats
needed by the army transportation corps. In its heyday it completed 746
craft within a year and in 1464 still had an annual capacity of 550 vessels,
an impressive figure by pre-modern standards.56 From the later fifteenth
century all the materials and labor needed by the dockyard were paid for in
silver. The earlier procedure whereby the soldiers operating the boats paid
30 per cent of the cost of new materials still persisted, however.57 These
funds were remitted by 82 guard units. The dockyard in turn was organized
into 82 small units, each with its own housing for which it was assigned
a narrow strip of land about thirty yards wide on a bank of the Huai river.
The total length of the adjacent units was over two and a half miles.58 The
commanding officers of the 82 units, generally captains on temporary duty
from their respective guards, were responsible for financing the ship con-
struction program. The central office merely allocated them raw materials
and labor from its own account, which was also derived from small
quantities of funds paid by scores of counties and prefectures. In the
sixteenth century many of the captains in charge of the dockyard units
were said to have incurred personal debts because of the managerial
responsibilities with which they were encumbered.59

The palace expenditures

Despite the smallness of state income palace expenditures steadily increased
during the fifteenth century. The size of the palace staff provides a yard-
stick for this. The Hung-wu Emperor, according to the traditional his-
torians, had in 1369 limited the number of palace eunuchs to sixty,60 a figure
which Hucker has shown to be grossly under-reported.61 Most likely it
included only those who had civil service ranks, and no household attend-
ants and auxiliaries.
It does seem, however, that at the founding of the dynasty the palace
staff was not too large and unwieldy. After 1420, the eunuchs took over
duties which previously had been assigned to palace women and their
offices soon proliferated into twenty-four bureaus and directories. A report
by the ministry of revenue in 1443 indicated that the eunuchs consumed
120,000 piculs of husked rice; if this figure, as seems likely, was an annual
 2,1 The level of state income and modifying factors 57
one, it suggests a palace staff of about 30,000 eunuchs.62 Even in the Chia-
ching period, when efforts had been made to reduce the palace personnel,
their number was still close to 10,000 in the mid sixteenth century.63
A modern scholar has estimated that by the end of the dynasty there were
70,000 eunuchs in service.64
The cooks serving in the court of imperial entertainments, all statutory
laborers, numbered 800 in the Hung-wu period. In the early years of the
Yung-lo period when kitchen services were provided in both capitals, their
number swelled to 3,000, and before the reign was over it had further
expanded to 9,000. Under Hsiian-te the court employed 9,462 cooks. An
austerity program in 1435 reduced their number to 5,000 but in 1487 it was
increased to 7,884, not far from the all-time high.65
Palace supplies increased in the same fashion. At the beginning of the
dynasty Ch'ang-chou prefecture in South Chihli each year delivered to the
palace 100 catties of tea. By 1431, the amount had increased to 290,000
catties, or about 200 tons.66 The supply of yellow wax used for candles was
expanded most rapidly in the late fifteenth century. In 1430 the annual
requirement was 30,000 catties, but by 1488 it had reached 85,000 catties,
and in 1503 was in excess of 200,000 catties.67
To provide charcoal supplies the Hsiian-te Emperor established a special
factory at I-chou, about 100 miles west of Peking. The charcoal burners
were statutory laborers from Shantung, Shansi, and three prefectures of
North Chihli; more than 30,000 of these served for three months each
year. The operations of the factory grew to such a scale that its supervisor
assumed the title of either extra transmission commissioner or minister of
works.68 In the early years of its operation the factory produced some
10,000 tons of charcoal and kindling wood annually. By 1442 its output had
increased six-fold. Transportation of the fuel to the capital became an
extra burden for the population along the route. The cost of hauling one
ton of charcoal over the short distance is estimated to have been 5 taels of
silver. Because of the uneconomical nature of this operation, toward the
end of thefifteenthcentury part of the required charcoal was purchased in
Peking and the statutory laborers at I-chou were gradually replaced by
soldiers. Even then each season 19,900 charcoal burners still had to be
enlisted from the general population. On the above estimate the annual con-
sumption of fuel and heating material in Peking in thefifteenthcentury would
have cost 500,000 taels of silver, had the court paid the expenses in cash.69
The exact cost of palace maintenance is beyond calculation. The supplies
were drawn from numerous warehouses and depots and the services in-
volved a sizeable number of the population. The rice consumed by the
palace personnel was derived from the regular land taxes. Fur and leather
were contributed by the local communities while silk fabrics were both
collected as surtaxes on the land taxes and obtained through local procure-
58 The heritage of the sixteenth century
ment orders. Some of the supplies, such as porcelain of the imperial design,
were in fact priceless because they were unique. Hsiian-te in 1433 ordered
443,000 pieces of chinaware from Kiangsi. When Hung-chih died in 1505,
his as yet unfilled order for porcelain still totalled 300,000 pieces.70 Some
of the articles seem to have been deliberately made to become collectors'
items. As far as is known palace supplies of all kinds submitted annually
must have been worth more than 5 million taels.
These items, largely covered by the service levy, cancelled much of the
effect of the basically low level of taxation. Since a major portion of those
accounts consisted of statutory labor and transportation costs, it was also
likely that the tax burden would fall more heavily on the poor and in-
articulate, who were often called upon to perform the unpaid services.
Closely related to the normal palace expenditures were the construction
costs of public buildings and mausoleums, and the maintenance of military
supernumeraries in the capital. The Peking garrison was converted into
a huge labor gang and construction materials were requisitioned from the
provinces. The building projects that went on incessantly from the early
fifteenth to the seventeenth century might have been a less significant factor
in governmentalfinancehad the latter been better organized. In view of the
dynasty's limited fiscal capacity it was clearly unsound to give priority to
such construction works. It was the army that suffered most from the
thoughtless appropriation of the financial resources. In addition to de-
priving the capital garrison of its combat readiness, several emperors in the
fifteenth and early sixteenth century also filled the officers' corps with
supernumerary unqualified personnel.
This was an abuse which the civil service, thanks to its rigid standards of
appointment, escaped. The unqualified personnel were absorbed exclusively
by the army. Emperors' in-laws, their favorites, and even the latter's
relatives were constantly admitted as supernumeraries of the officers' corps.
In the 1460s some of the capital garrison guards had close to 3,000 officers
each, whereas the total authorized strength of such a guards unit, including
officers and enlisted men, was 5,600 men.71 The sovereign's maternal uncles,
father-in-law and brothers-in-law usually became marquises and earls and
their sons and nephews then became generals and colonels. A favorite
eunuch's nephews, sometimes a large group of them, would be commis-
sioned as regimental commanders by the throne. It was not uncommon
for the father or brother of a woman attendant in the palace to receive the
honorific title of guard captain or even battalion commander,72 as a reward
for her diligent service, a method of acknowledging merit rather similar
to European royalty's granting of medals to cooks.
Army commissions under the Ming were traditionally hereditary. The
supernumeraries in the capital thus gradually proliferated beyond control,
in the same way as did the imperial clansmen residing in the provinces, who
 2,1 The level of state income and modifying factors 59
were also supported by the state for life. In 1371 the total number of army
officers was reported to be 12,980.73 Before the end of Hung-wu's reign it
had already increased to 28,000,74 and in 1455 it was reported that there
were 31,790 officers in Peking alone.75 In 1469 the total number of army
officers was said to have exceeded 80,000,76 and by about 1520 it had further
expanded to an unmanageable 100,000,77 most of whom undoubtedly
resided in the capital.
Thefinancialconsequences of this abuse of rank were somewhat strange.
Because the pay of the supernumeraries, being the same as that of the
regular army officers and civil officials, was low, it did not significantly
enlarge the state budget. The cash portion of their pay was actually derived
from the emperor's own purse, owing to the institutionalization of the
Gold Floral Silver. But there was only a limited quantity of materials
(including grain and cotton cloth) available to the ministry of revenue to
meet these salary obligations. The greater the number of persons on the
payroll, the smaller was the share of each. This in part explains the low
salaries received by the officials.
A major item of state income which was seriously affected by the super-
fluous personnel was the tribute grain. For a time the tribute grain to
Peking had fluctuated in volume, but from 1472 it had been settled at
4 million piculs annually. A small portion, usually about 300,000 piculs,
was forwarded to the frontier army posts.78 The amount available for
appropriation in Peking was therefore around 3.7 million piculs. The
distribution was effected through a rationing system. With few exceptions,
all civil officials, army officers, and lesser functionaries received 1 picul per
man per month regardless of rank, as part of their pay. Soldiers and arti-
sans on duty received half this amount.79 By the late fifteenth and early
sixteenth century there were no fewer than 300,000 men eligible for the
grain payment.80 The 1502 report by the ministry of revenue indicated
that the annual amount payable was 3.38 million piculs.81 Whenever there
was a shortage in grain delivery or an additional demand for payment, the
capital granary incurred a deficit. Thus the tribute grain, which entailed
such high delivery costs, contributed little to the empire's finances. While
it solved the food supply problem in the capital, most of its recipients, being
supernumeraries, performed no useful function to the state.
The situation was partially rectified by Grand-Secretary Yang T'ing-ho
(1, i). In 1522, at the accession of the Chia-ching Emperor, this austere
statesman summarily removed 148,700 superfluous persons from the pay-
roll, including numerous honorary officers who were stripped of their
ranks. The cutback saved the state 1.532 million piculs of grain annually.82
The commutation of this surplus into silver payments was one of the key
measures which enabled the empire to survive its financial crisis in the
mid sixteenth century (7, i).
60 The heritage of the sixteenth century

II LAND AND POPULATION DATA

Land data
In the 1587 edition of the Ta-Ming Hui-tien are three sets of land data, the
total figures of which are as follows:83 1393: 850,762,368 mou; 1502:
622,805,881 mou; 1578:701,397,628 mou. Until recently it was thought that
these figures represented cultivated acreage under taxation and that they
had corresponding statistical value. Only after a thorough investigation by
several Japanese scholars in the 1940s was the true nature of these figures
gradually revealed. It now appears that the 850 million mou figure of 1393
was not a fiscal record in the true sense but as Fujii Hiroshi has shown,
included waste land and land earmarked for reclamation.84 Nor was the
figure obtained through a national land survey as was earlier believed;
most of the provincial figures were simply random estimates. The taxable
acreage in the early Ming was less than 400 million mou, as shown by the
figures recorded in the Shih-lu.85
The 1502 figure seems to have been derived from the compilation of the
Yellow Book in that year, but the actual acreage was 200 million mou less,
being recorded in the 1510 edition of the Ta-Ming Hui-tien as 422,805,881
mou. Shimizu Taiji, who first discovered the discrepancy, points out that
the editors of the later edition of the Hui-tien, perhaps disturbed by the
small acreage reported by Hukwang province, arbitrarily added 200 million
mou to the provincial figure.86
According to these two scholars the 1578 figure similarly inflated the
Hukwang acreage by some 190 million mou. To be statistically consistent,
the total figure should be revised approximately 500 million mou.
In his contribution to Wada Sei's annotated translation of the economic
monographs in the Ming-shih, Fujii summarizes the revised acreages as
follows: 1381:366,771,549 mou; 1391:387,474,673 mou; 1502:422,805,892
mou; 1578: 510,612,728 mou. Though the findings are amply documented,
the documentary evidence is suggestive rather than conclusive, the de-
ductions are most logical, but no absolute proof can be found.87 Further-
more these revised figures are still no more than fiscal acreages which were
never intended to represent the actual acreages under cultivation (see fiscal
mou conversion, 1, n).
In point of fact it would have been extremely difficult to establish a uni-
form statistical standard in the fifteenth and sixteenth centuries. Once the
tax quota was put into effect cultivated acreages and population data
became relatively unimportant to the central offices, for they were no longer
the bases of taxation. The interpolation of vital statistics in official publi-
cations by senior bureaucrats is an indication that they were not seriously
concerned with quality control in data processing.
 2, II Land and population data 61
The expansion of agricultural land in the sixteenth century indicated
by Fujii's revised figures is of course related to our topic of study. The
numerous inconsistencies in the primary data submitted by the regional
offices make it difficult, however, to use thesefiguresas the basis of a com-
parative study. The limited quantitative surveys in the following chapters
are based on the regional data, whose internal consistency can at least be
ascertained.
Population data
Sixteenth-century population reports were even more subject to corruption
than were the official land data. Under-reporting was a universal tendency.
Unlike the cultivated acreage the population returns could not be checked
for even relative accuracy. The traditional Ming practice was merely to
spot the mathematical errors in the reports. When the breakdown figures
did not correspond to the aggregates, the whole set of books was rejected,
with a severe fine imposed on the district which had submitted them.88 In
order to avoid the fine, the counties and prefectures simply resubmitted
their earlier returns as current reports, sometimes with slight modifications
in the last two or three digits. Ping-ti Ho has shown, as an example of this,
that for the three decades from 1522 to 1552 during which the compilation
of the Yellow Book took place four times, Feng-hua county in Chekiang
persistently reported neither increase nor decrease in the district's 18,865
households, and Hsiang-shan county even recorded its population as
consistently comprising 17,812 persons throughout.89
The formality of submitting decennial population and land returns was
never discontinued, however. The Ming court designated an island on the
present Hsuan-wu lake outside Nanking as the site of an archive to store
these records. The whole region was declared a restricted area and before
the reports arrived thirty new bungalows were constructed for their
storage. Cooking and the lighting of candles were both forbidden on the
island, as measures of fire prevention. The format of the records was
meticulously prescribed and the stationery standardized. A number of
imperial university students were assigned to check the returns for con-
sistency.90 Much of the work was wasted, nevertheless. Wei Ch'ing-yiian
points out that Hsing-hua county in South Chihli in 1582 included 3,700
households as purporting to have residents 100 years old or older. In the
early Ch'ing an official reported that many late Ming records still retained
the personal names and property-holdings of the Hung-wu era.91
The inadequacy of the Yellow Book was fully recognized by contempor-
ary writers. Wang Shih-cheng (1526-90) called its compilation 'a children's
game'. The 1609 edition of the Wen-shang county gazetteer, Shantung
province, deplored its 'waste of paper and writing brushes'.92 The 1572
edition of the Ku'ai-chi county gazetteer, Chekiang, in accordance with the
62 The heritage of the sixteenth century
official report stated that the district's population was 62,004 persons, but
the editors also added that the actual number must have been more than
four times that figure.93
The official population data appear in the Ta-Ming Hui-tien as follows :94
1393: 60,545,812; 1491: 53,281,158; 1578: 60,692,856. These figures have
little intrinsic value. Nevertheless, they at least do not contradict the
assumption that population in the late sixteenth century must have signifi-
cantly increased from its level at the founding of the dynasty. Of the three
figures that of 1393 is undoubtedly the closest to reality, given the stern
administrative discipline and the severe punishment for falsification of
fiscal records during the Hung-wu period. Thefigureof 60 million or so is
also consistent with the reported salt production of the era.95 It is known
that in later centuries individual districts often reported population de-
creases that had occurred but not the increases.96 The 60 million level of the
later records can therefore be regarded as a sign of net growth.
Non-quantitative statements indicating population growth are abundant
in the contemporary sources. Ping-ti Ho, on the basis of a lengthy study
of the contemporary literature, suggests that the population level of 1600
was around 150 million.97 Suffice it to say that this estimate accords with
general descriptions of economic activities in the late sixteenth century,
with the suggested increase of salt production in the same period (5, n),
and with post-Ming population records.
As for the effects of population growth, Ming writers maintained almost
unanimously that a large population was a favorable factor in taxation.
This attitude seems to reflect ideological bias rather than objective analysis.
Little consideration was ever given to the lower per capita incomes resulting
from increased population. It is true that in the sixteenth century the Ming
state still possessed a considerable amount of empty space, notably in
Hukwang and Honan, and was therefore little concerned with population
pressure; the expanding handicraft industries moreover provided an
important outlet for additional manpower.98 This did not however alter the
fact that in certain districts the extra number of mouths to be fed might
already constitute a problem. The Ku'ai-chi county gazetteer, cited above,
raised this issue in suggesting that the county's natural resources could not
adequately support half its population.99 Food shortages in Fukien were
cited by Ming writers as well as modern scholars as one fundamental reason
why its residents pursued maritime trade in defiance of imperial law, which
eventually led to piracy on the coast after the mid sixteenth century.100
Many historians are currently showing interest in the urban culture of the
late Ming and there has been much discussion of the widespread production
of commercial crops and goods by landowners in the phenomenally
prosperous lower Yangtze valley.101 Inadequate data makes it difficult to
gauge the general standard of living under the Ming. However, P'eng
 2, III The maintenance of the army 63
Hsin-wei in his study of China's monetary history does suggest that there
was a lowering of the standard of living as a consequence of population
growth and that this affected taxation adversely.102
Another factor to be considered is that the Mingfiscalapparatus was so
immobile that it never developed the capacity to cope with the changing
variables of a dynamic society. The dogmatic Confucian belief that more
productive labor meant greater tax revenues was in reality only wishful
thinking. Even though it might be true economically, the Ming government
could not take advantage of it administratively. Because of the quota
system, the tax burden of a depopulated district or of a declining community
could not easily be lightened. The standard solution was to increase the
taxes of the remaining residents, which led to further abscondings and
uncollected revenues.103 Conversely, increased population seldom con-
tributed directly to higher tax proceeds. The surplus population tended to
be mobile, and therefore difficult to bring under tax control. Even when
more population could be registered and assessed, the local administrators
were unenthusiastic about reporting the actual returns, for fear that the
reported increases would prompt the court to readjust their regional tax
quotas upward. At most they would reapportion the tax burden among the
enlarged pool of taxpayers, easing their own collection problems and at the
same time earning for themselves the reputation of benevolent adminis-
trators.
Local deviations from standard procedures in compiling tax registers
started early. By the latefifteenthcentury, the so-called 'white-books' had
already made their appearance. As a formality the local officials still sub-
mitted the Yellow Book to the imperial archive but for actual tax adminis-
tration they compiled another set of data, conventionally known as white
books. It is said that when Wang Shu (1416-1508) was the governor of
South Chihli, in about 1479, he actually gave the local officials his personal
permission for the preparation of these records.104 In the sixteenth century
the data in these accounts appeared in some of the local gazetteers.105

Ill THE MAINTENANCE OF THE ARMY

The legend and legacy of a self-supporting army
The claim that under Hung-wu and Yung-lo the Ming army attained self-
sufficiency in grain supplies through military farming still remains an
obstacle to our understanding of the dynasty's financial history. This
legend was built up by late Ming writers who most likely had themselves
been misled by the earlier records. The primary sources contain plentiful
material for its construction. The total acreage allocated to military farming
is recorded in the Ta-Ming Hui-tien as having been in excess of eighty-nine
64 The heritage of the sixteenth century
million mou.106 Its total yield in 1403 is stated in the Shih-lu to have been
more than 23 million piculs, almost comparable to the entire income from
the land taxes.107 The Ming-shih further asserts that each soldier participat-
ing in the program was allocated fifty mou of land, from which he was
expected to produce a minimum of 18 piculs of grain. He kept 12 piculs as
his own pay and allowance, and turned in the surplus at military granaries.
On the frontier, 70 per cent of the soldiers were enrolled in the program,
and 80 per cent in the interior.108 These estimates imply at least one million
soldiers occupied in grain production. It might be conjectured that such
a prodigious program, if carried out in full, could have averted any prob-
lems over army supplies in Ming times.
The above assumption is, however, not only untrue but also impossible.
In the late sixteenth century even some Ming writers began to question
the validity of the earlier reports. The acreage allocations showed that
Szechuan alone had 65,954,526 mou. One author, working out the garrison
strength in that province around the year 1400, discovered that the
allocation would have granted every soldier 4,500 mou of land.109 Other
evidence contradicting this claim can readily be found in contemporary
writings. The Hung-wu Emperor himself disclosed that two military
commands in South Chihli, even after putting soldiers on farm duty for
over twenty years, had not come close to self-sufficiency in food supplies.110
Entries in the Shih-lu, indicating that army units in the late fourteenth
and early fifteenth century relied on supplies from civilian sources, are
too numerous to be cited in full.111 In 1404, even after Yung-lo's vigor-
ous promotion of military farming, its production in many areas could
not meet half the food requirements of the cultivators themselves.112 The
total figure for the yields of the empire's military farming program was
steadily revised downward until in 1423 it was quoted as being close to
5 million piculs.113 Actual production could well have been much lower.
In 1429 Minister of Revenue Ko Tun (in office, 1427-31) brought it to the
Hsiian-te Emperor's attention that many guard units assigned fewer than
a dozen soldiers to farm duty and produced less than 100 piculs of grain,
while each unit needed tens of thousands of piculs for its annual main-
tenance.114
This suggests that this vaunted self-sufficiency and the national statistics
associated with it represented a projected target only, and involved over-
estimates at all levels. It is impossible to ascertain the actual results of the
program in its early years, partly because of the systematic suppression by
Ming bureaucrats of information that tended to discredit the dynastic
founders, and partly owing to the absence of documentation from the
army officers, who as a group left no records. Wang Yu-ch'iian, who
directed a collective research project on military farming under the Ming,
calls its acclaimed success 'exaggerated' and 'without factual basis'. 115
 2, III The maintenance of the army 65
The level of attainment must be viewed in the light of the conditions
under which the farming program was organized. It was put into effect with
practically no prior planning or preparation. No field investigation was
ever conducted; no pilot project was carried out. Nor was any office of
control set up. The emperors, by issuing orders, merely forced the military
commanders to carry them out with whatever resources were at their
disposal. The extant information shows that the distribution of seeds and
draft animals, and the allocation of manpower and farm land was at no
time brought under centralized control.116 There was little possibility
therefore for the program to develop into a permanent system.

The criss-crossing supply lines

The Ming army was organized by the Hung-wu Emperor as a gigantic
military machine. By 1392 it was estimated to have some 1.2 million men
on active service,117 and to have enrolled 1.7 million to 2 million hereditary
military households.118 But it lacked internal cohesion. Not all the house-
holds appearing in the roster had chosen to be so registered. They included
large number of political prisoners, criminals, exiles, and minority groups
from the north-west.119 During the Hung-wu period large numbers of
civilian households in Hukwang, Kwangtung, Fukien and Shansi had
been pressed into military service and placed under military registration.120
Desertions seem to have been quite common even in the early years.121
Each hereditary military household, in addition to furnishing a soldier
for active duty, also answered the service calls of the guard unit to which
it belonged. One of these obligations was to provide 30 per cent of the
military equipment required, the other 70 per cent to be furnished by the
civilian population in the adjacent counties and prefectures. This procedure
was possibly established in the later years of the Hung-wu period.122
The soldiers did not receive their regular pay in cash, but from time to
time the emperor handed out general awards of paper currency to the
officers and men on active duty. Such grants were unscheduled. The regular
pay consisted of a monthly ration of 1 picul of grain per man plus a small
amount of salt for family maintenance.123 Winter uniforms were issued in
the form of either ready-made garments or cloth and cotton wadding,
depending on the conditions of supply.124 Since under the Hung-wu
Emperor a soldier often received as much as 25 kuan of paper currency
annually, the payments were fairly adequate. During the middle of the
Yung-lo period, around 1410-20, however, the general awards were
gradually cut off, and were thereafter totally suspended except on the
occasions of a new emperor's accession.125
The wei-so units were in reality of two different types. On the northern
frontier, in certain districts on the coast, and in Kweichow, the guard and
66 The heritage of the sixteenth century
battalion headquarters functioned virtually as local governments. Practi-
cally all the population of these regions during the early years of the
dynasty was under military registration and all the land was government-
owned. Here the standard land allocation for military farming, fifty mou
per soldier, was in general carried out.126 But productivity in these areas
was generally poor and the farming program was undoubtedly also handi-
capped by problems of security and other physical factors. In the interior
the wei-so had no territorial responsibility. The army units were stationed
among the general population, and only limited and dispersed amounts of
land were available for military farming.127 The fragmentary data indicate
that in general each soldier was allocated twenty mou or less.128 The
regulation of 1402 which required each soldier on farm duty to turn in
12 piculs of surplus grain in addition to providing 12 piculs for his own
maintenance was apparently difficult to carry out in full.129 Clearly the
standard had been set at the maximum possible level.
After the introduction of the farming program in 1371,130 the guard units
were expected to cut down on supplies requisitioned from civilian sources.
In spite of this the earlier supply procedure continued. As explained in the
previous chapter no overall logistical command had ever been established.
Following the directives of the ministry of revenue, local magistrates
supplied grain consignments to the adjacent army units. Deliveries at the
lowest level followed standing orders and developed into a network of
supply lines criss-crossing the empire.131 In our present state of knowledge
there is no way to discover how much tax revenue was thus disbursed. The
complexity of the accounting system remained a fundamental problem
that was never solved throughout the dynasty. It was possible for army
officers to exaggerate the achievements of the farming program when they
were in fact being subsidized by the civil government. In 1407, at a time
when Yung-lo was issuing stern orders that military farming be expanded,
a censorial official reported that its actual production could not be effectively
ascertained.132 Significantly, the same problem existed in later centuries.

The decline of the military establishment

While the real achievements of the farming program in its early stages are
unclear, its decline after the first quarter of the fifteenth century is no
mystery. A decree of 1425 permanently reduced the annual surplus grain
quota to be provided by each soldier on farm duty to 6 piculs,133 a move
that was both thoughtful and realistic. No new source of supply was
provided to compensate for the reduced quota, however. The granting of
paper currency to soldiers as general awards had already been suspended.
Subsequent land tax reductions and occasional tax remissions further
undercut the capacity of local government to pay the army units.
 2, III The maintenance of the army 67
From 1435 the 12 piculs of grain that each soldier was supposed to
provide as his own pay was no longer checked in and handled by his unit
commander but was simply withheld by himself during the harvest.134 In
theory, farm land was never granted to individual soldiers; it was rather
the army personnel who were assigned to plots of public land. But in
reality no rotation was ever put into effect. The system merely turned
soldiers back into farmers and demanded from them grain payments that
were five to ten times the regular land taxes paid by the general population.
By the mid fifteenth century, if not earlier, soldiers had already started
selling and mortgaging the land in their custody. Such cases are frequently
mentioned in the late Ming records.135
The scale of the deterioration of the Ming armed forces, especially those
in the interior provinces in the 100 years or so following the mid fifteenth
century, seems to have been unprecedented in Chinese history. When
supplies were short, the army simply reduced the soldiers' rations. Part of
their pay and rations could also be converted to cloth, pepper and sapan-
wood. The conversion rates varied from one command to another and
there was a differentiation between married and unmarried soldiers.136 At
1 picul per month, the ration was not abundant to begin with, and the
conversion virtually constituted cancellation of payment. When in 1468
some soldiers were told that the converted portions of their pay were
henceforth to be reduced to 4 taels of pepper and sapanwood per man and
that the commodities were to be issued at distant warehouses, they simply
refused to accept the payment.137 A battalion headquarters in Shansi
reported in 1489 that its soldiers had received no grain ration for two
years, and cloth and cotton wadding for six years.138 Only in 1511 was the
Ming government able to settle the arrears, some dating back for nine
years or more, due to soldiers in Honan. The overdue rations were then
converted into copper coins at the rate of 20 cash per picul, a payment
which represented about 5 per cent of their original value.139 In 1528 the
emperor in a general order acknowledged that the soldiers in many wei-so
had not been paid for years.140
The result was that the army's ranks were thinned by desertion. The
military colony system could not have survived at all, had it not been for
the vigorous 'troop purifying' policy of the court. When a soldier deserted
the 'troop purifying censors' held his relatives and neighbors responsible.141
The case was usually settled when the scores of persons thus encumbered
would agree to purchase a substitute, finance his marriage and relocation,
and send him off to fill the vacancy.142 But all this was a slow process.
Before one case was closed another desertion occurred, until in the end the
guards and battalions were reduced to skeleton organizations.
During the Hung-wu era the military units in Kwangsi had com-
prised 120,000 men; by 1492 only 18,000, or 15 per cent, remained.143 The
3-2
68 The heritage of the sixteenth century
Nan-ch'ang Left Guard in Kiangsi province had an authorized strength
of 4,753 men; in 1502 only 141 of them were on barracks duty, less than
3 per cent of the original quota.144 The Ching-hua Battalion was authorized
to have 1,225 men; in the sixteenth century only 34 soldiers were in the
camp, plus some 300 men serving in the transportation corps.145
By the early sixteenth century it was typical for a military colony in the
interior to operate at 10 per cent of its prescribed strength. The seventy-
eight guard units around Peking, which in the early fifteenth century had
been backed up by 380,000 hereditary military households, provided the
capital garrison with only 50,000 to 60,000 soldiers. Many of them were
employed as construction laborers by the palace, while others took posi-
tions as grooms in army stables, office attendants, and household servants.
Not more than 10,000 of them still bore arms, and some of those were
actually paupers hired as substitutes.146
The army posts on the northern frontier fared somewhat better. At an
official count in 1487 some 300,000 men were serving on the frontier.147
This came about because it had been necessary gradually to replace the
weiso regulars in these posts by recruits.148 By about 1500 it was estimated
that recruited soldiers constituted half the combat strength in these units149
which in turn meant that the imperial government had to step up its
subsidies to the northern army posts. Until then the annuities delivered by
Peking had been limited to 480,000 taels of silver,150 afigurewhich had to
be increased significantly at the turn of the century. In the three years
from 1500 to 1502, in response to military pressure from the Mongolian
tribes, the ministry of revenue rushed another 4,150,200 taels of emergency
funds to the several commands.151 This trend, once established, was never
to be reversed. Supplying the army in the north became a crucial issue and
one of the most difficult tasks of the ministry of revenue (7, n).
Though the abovefigureswere by no means prodigious in terms of the
empire's total resources, the government had no income already earmarked
to meet this expenditure. The unrealistic legend of a self-sufficient army
still persisted. Thus not only its financing, but the wei-so system itself
became a burden to sixteenth-century administrators. Though its efficiency
was gradually being reduced to the minimum, it could neither be abolished
nor reorganized. It had been such a fundamental institution of the state
that latter-day administrators, with so few financial resources at their
disposal, did not even dare to think of introducing innovations into it.
Only in 1643, less than a year before the end of the Ming, did Minister of
Revenue Ni Yiian-lu boldly suggest that hereditary military households be
altogether eliminated, and his proposal was.rejected by the Ch'ung-chen
Emperor.152
 2, IV The monetary problem 69

IV THE MONETARY PROBLEM

Paper currency
The adverse effects of the inflation of paper currency under the two early
Ming emperors have been generally recognized by economic historians.153
The exact extent of the damage to governmental finance caused by their
inflationary policies has not yet been fully assessed however. Cursory
observation might lead one to the conclusion that the blunder had only
temporary effects; but in reality the inflation had long-term consequences
in that one mistake led to another. The failure of the paper currency
started a chain reaction, the unfortunate effects of which spread to the
minting of copper coins and subsequently to the use of silver in tax
administration. Though there were many reasons for the Ming dynasty's
failure to integrate itsfinancialaccounts and discard its tax quota system,
one of them was undoubtedly the lack of a workable monetary system, and
this in turn originated in the unrestricted issue of paper currency in the
early years of the dynasty.
The Hung-wu Emperor's attitude toward money is difficult to explain.
Over-confident in his own powers of regulation, he must have decided that
the paper currency would hold its arbitrarily assigned value. Yet there is no
indication that he ever understood that his inflationary policy in fact con-
stituted a form of taxation. It could have been anticipated that by defraying
expenditures with unlimited quantities of government notes, the circulation
of those notes would soon reach saturation point. There is no evidence that
he realized these implications.
Hung-wu is known to have circulated paper currency in the form of
general awards, and to have distributed it as special grants to imperial
princes and high-ranking officials, and as return gifts to foreign tributary
missions. He also used them to pay for grain purchases and handed them
out as relief for natural disasters. Though it used to be regarded as im-
possible to ascertain the amount of paper currency in circulation, its
general range can in fact be established. Its high level is surprising.
In the Shih-lu for the year 1390 are sixty-nine entries indicating various
occasions on which the dynastic founder made payments in paper cur-
rency.154 Of these sixty-nine casesfifty-threeeither specify the exact amounts
distributed or define the nature of the grants clearly enough to permit
calculation of the sums involved. Together they total 88,607,315 kuan.
While in the remaining sixteen entries the quantity is not reported, its
total value can be estimated with some accuracy as close to 7 million kuan.
Thus in that year the emperor handed out no less than 95 million kuan of
paper currency. The state income in government notes is recorded that
year as totalling 20,382,990 kuan.155 Deducting the latter figure from the
70 The heritage of the sixteenth century
former, one may conclude that in the year 1390 alone the Ming founder
inflated the market with some 75 million kuan of new notes. At the official
rate of exchange, 1 kuan per picul of husked grain, the amount of new
money in circulation was thus equivalent to two and a half years' income
from the land taxes. Even at the market value, which was by then 4 kuan
per picul of grain,156 the amount was still worth more than half a year's
land tax income.
This estimate does not conflict with the known productive capacity of
the superintendency of paper currency. The emperor's own statements
disclose that in the Chinese year of 1385, the workers in the superintendency
actually produced somewhere between 27 million and 34 million kuan of
new notes.157 This amount is considerably less than the 75 million kuan
circulated in 1390. Nevertheless in 1385 the market value of the paper
currency was 2.5 kuan per picul of grain.158 The total value of the new
currency put into circulation in that year was still not much below half
a year's income from the land taxes.
In the early years of the fifteenth century, currency inflation had already
got out of control. In 1404 an imperial censor suggested that salt be sold
to the entire population on a uniform rationing system, with payment made
exclusively in paper money. His original plan required the raising of a huge
annual income of 1 billion kuan. The ministry of revenue's counter-proposal
reduced the payment by half.159 Even though the rationing was never
carried out according to either the original plan or the counter-proposal,
the scale of the projected income gives some idea of the total amount of
currency then in circulation. It is evident that Yung-lo's policies also
contributed to this advanced state of inflation. Though he reduced the
official exchange rate to 30 kuan of currency per picul of grain,160 the actual
purchasing power of the paper money was below even this level.
When Hsiian-te ascended the throne in 1425 the paper notes had only
from one-fortieth to one-seventieth of their original value.161 Such rates,
nevertheless, were not price indices comparable to those in a modern
society where they fluctuate in accordance with supply and demand condi-
tions of the market. In the early fifteenth century private business was
limited in scope. Commercial transactions using large amounts of paper
currency were unheard of. The so-called market value of the paper currency,
while in part determined by the users, to a large extent still relied on the
regulating power of the state. Whenever this power was lacking the popu-
lace avoided the currency altogether; the result was therefore not further
devaluation but total disuse.
The Ming court naturally did not neglect its policing of private trans-
actions. In 1426 earlier laws proscribing the circulation of gold and silver
were reaffirmed.162 In 1429 a stiff fine was imposed on violators, each tael
of silver used in a transaction being assessed at 10,000 kuan of paper
 2, IV The monetary problem 71
currency.163 Such regulations could not be effectively enforced because
private business dealings were too scattered and silver was relatively
abundant. Hsiian-te, contrary to his own monetary policy, had himself
distributed 963,829 taels of the precious metal in general awards at his
accession.164
The most effective method of maintaining the value of paper currency
was to find outlets for the notes that were circulating among the populace.
In the early fifteenth century the government permitted the use of paper
currency to pay taxes and fines, but were careful to restrict the amounts
involved. Because of the widespread lack of confidence in the paper
currency, administrators did not dare accept such payments on a large
scale; this would entail financial loss, which could not be supported since
the government had no budgetary surplus.
The next step was to create new tax revenues and make them payable
exclusively in paper notes. At the suggestion of Minister of Revenue Hsia
Yiian-chi, Hung-hsi in 1425 introduced store franchise fees levied on shop-
keepers. The scope of the levies was widened and the rates increased by
Hsiian-te who in 1429 ordered that all the franchise fees in thirty-three
cities be increased five-fold. The highest fee paid by the shopkeepers was
500 kuan per month per store. Donkey carts entering and leaving the city
gates of Peking and Nanking also paid tolls, at 300 kuan a, time. In addition
a levy, assessed by tonnage was made on shipping using the Grand Canal,
generally bringing in about 500 kuan for a boat going through the whole
length of the waterway.165 The year-end report for 1431 indicated that
in that year 200 million kuan of paper currency had been produced
by these 'miscellaneous collections'.166 In 1433 the collection further
expanded to 288 million kuan.167 The deflationary effort was declared
a success. The emperor announced: 'Now both within and outside the
capital paper money is circulating well.' He ordered a uniform reduction
of the tax rates, in general to about one-third of the previous level. Further
reductions were ordered by Cheng-t'ung and Ching-t'ai, until by 1442 the
rates were no more than 10 per cent of what they had been at their highest
point. In 1452 the tolls on carts at the city gates, differentiated into four
categories, were 8, 4, 2 and 1 kuan respectively. The 4 kuan rate was less
than 2 per cent of the previous payment.168
The 1429 schedule was clearly excessive and must have caused serious
hardship to merchants and the general public alike. But once this drastic
step had been taken and its effects realized, it was unwise to abandon it
without devising follow-up measures to strengthen the paper currency as
a permanent institution. Clearly a golden opportunity for currency reform
presented itself at this point. To say that the Ming government's failure to
seize it was due to its lack of the technical knowledge is unconvincing,
because as early as 1425 a soldier had already submitted a plan for a
72 The heritage of the sixteenth century
currency reform, urging the replacement of the Hung-wu issue with new
notes.* The sovereign not only read it, but circulated it among his mini-
sters, indicating that the suggestion was a wise one which coincided with his
own opinions.169 In view of this, it can only be concluded that any currency
reform on such a scale was already beyond the capacity of the Ming court.
It should be noted that the plan submitted by the soldier was not limited
to the simple matter of issuing a different kind of paper money but also
pointed to the need for a more integrated fiscal administration, entailing the
gathering of accurate fiscal data and balancing the budget. Such a program,
if carried out at all, would have required a complete departure from Hung-
wu's concept of governmental finance. Hsiian-te, who was at that time
harassed by the problems of tax arrears and the Annam situation, could
not be expected to venture so far. Yet without the suggested reorganization
the introduction of new currency would be pointless and would only shake
public confidence still further. As it turned out, all the paper notes issued
afterwards still bore the regnal title of the Hung-wu Emperor.170
The empire's major problem was in the area of land taxes and military
farming. In the 1420s and 1430s the court, under pressure, only reduced
some of the payments without compensating for the tax losses thus in-
curred. Artificial outlets for paper money were restricted to heavy taxation
on the urban sector of the population, since in the cities the police power of
the state was much more effective. In a predominantly agrarian society,
however, such a basis of circulation was much too narrow; it was merely
a closed circuit. The 200 million kuan or more of paper money derived
from the annual miscellaneous collection could not have been worth more
than 500,000 taels of silver. At the same time the tolls and franchise fees
were already felt to be too heavy and could not be continued permanently.
The failure of the policy was a foregone conclusion.
When Hsiian-te died in 1435, a child emperor came to the throne. The
prefect of Wu-chou, Kwangsi province, petitioned for the legalization of
copper cash in private trade (at that time the unenforceable imperial law
forbidding the use of copper coins had not yet been rescinded). The petition
was approved, thus providing the population with an excuse to regard the
decrees proscribing the circulation of precious metals as no longer in force
either.171 When in the following year the court itself authorized the collec-
tion of land taxes in silver (see Gold Floral Silver, 2, i), the fate of paper
currency as legal tender was sealed. Its further devaluation was only to be
expected.
In 1436 it was reported that more than 1,000 kuan of paper were required
to purchase 1 tael of silver.172 Nonetheless, in the 1440s the value of the
* In the early Ming many learned men were drafted to become soldiers. This particular
man had been a political prisoner; Hsiian-te later discharged him from hereditary
military registration on reading his paper.
 2, IV The monetary problem 73
notes increased to a rate of less than 500 kuan per tael of silver,173 as
a consequence of the Ming court's final effort to re-enact the law forbidding
the use of copper coins. This step was in theory logical. The largest
denomination of the paper currency was 1 kuan. Though they were unable
to compete with silver, it was decided that the notes at their depreciated
value should replace copper cash for use as small change. In 1448 a regu-
lation to this effect was proclaimed and in Peking the police began to
arrest those who used the coins in transactions.174 But the government
notes could not easily be revitalized in view of their previous history, and
petty transactions were extremely difficult to regulate. When in the follow-
ing year the Ming armies suffered a crucial defeat at T'o-mo-pao and the
youthful emperor was captured by the Oirat chieftain Esen, during the
ensuing state of emergency in the capital the unenforceable law was simply
ignored. No effort was ever made thereafter to promote the general accept-
ance of government notes.
Paper currency was never entirely abandoned, however. Even though it
fell into disuse, it survived as a fiscal unit. Certain revenues which had been
assessed during the Hung-wu period in paper notes and later converted into
silver at the sixteenth-century exchange rate of 0.003 taels of silver per
kuan, thus dwindled preposterously. Nor was there ever any thorough
reorganization of the revenues created in the early fifteenth century for the
purpose of promoting the use of the paper currency. As a matter of principle
some of these items still had to be paid partially in paper notes. In 1466 it
was already reported that the notes 'could not attract the attention of
passers-by when they were piled up in the stores', yet even as late as the
1580s the magistrate in Peking was still required to purchase paper currency
in order to fulfill his tax-delivery obligations.175
The currency had no commercial value, except in so far as some dealers
specialized in the business of buying it for resale to those taxpayers who
had to include the notes in their tax payments. From 1488 the government
in general recognized 1 kuan as equivalent to 0.003 taels of silver,176 but
in the assessment of taxes one type of currency could not be substituted by
another. The total amount of taxes actually paid in the currency was
negligible.177 In 1527 the notes were no longer officially counted by the
kuan. The assigned value of the blocks varied from place to place.178
Until the mid fifteenth century the Ming court had used paper currency
to pay part of official salaries and sometimes the soldiers' pay as well.
Thereafter such practices were generally discontinued and the paper
currency became in effect a kind of 'ceremonial money'. It was distributed
to officials as nominal travel allowances, and occasionally meritorious
ministers and governors-general also received parcels of paper money,
never exceeding 1,000 kuan at a time, as a form of special honor from the
throne. On festival occasions the emperor handed it out as awards to his
74 The heritage of the sixteenth century
court.179 In 1618, during the Manchurian campaign the government still
paid some of the soldiers with the paper currency.180 The last time paper
notes were distributed as a part of the general awards seems to have been in
1620 on the accession of the T'ien-ch'i Emperor, more than one and a half
centuries after they had been discontinued in general use.181

Copper coinage
The coining of money was considered throughout the dynasty more as an
administrative function than a public service. The mint had no operating
budget and its statutory laborers received rations but no pay. Ordinarily
the necessary materials were supplied either by the artisans or by designated
merchants at prices determined by the government. Furnishing such supplies
constituted part of their service obligations. After the mid fifteenth century
most of the laborers paid silver rather than perform the service obligations
in person; the actual workers were hired substitutes. Many earlier practices,
such as holding the workers and suppliers financially responsible for
minting operations, continued however. Some of the foremen in the mint,
known as lu-fou, or 'furnace chiefs', supplied the needed charcoal.
A seventeenth-century handbook indicates that the melting of the metals
had to be carefully timed. Otherwise over-melting resulted in shortages of
raw materials, which the suppliers had to make up, and under-melting
yielded fewer coins than required, which became the financial liability of
the furnace chiefs.182
Most of the coins circulated under the Ming were of one denomination
only, known as Mien or wen, and sometimes referred to by Western
writers as the copper 'cash'. But cKien (mace) was also a unit of weight,
namely one-tenth of a tael, the Chinese ounce. In theory at least each coin,
a ch'ien, should also weigh 1 cWien. In other words 10 coins were supposed
to weigh 1 tael; and 160 coins 1 catty.183 Coins of larger denominations
were rarely minted. Hung-wu minted some coins in denominations of up
to 10 cKien. Similar coinage produced during the T'ien-ch'i period in the
seventeenth century was a total failure.184 The bullion theory of money
was a time-honored tradition and was probably even more reinforced
during the Ming than in earlier times. The dynasty's monetary policy must
have further dissuaded the people from putting their faith in abstract
money.
Copper coins, in principle, were to be made of pure copper. An alloy
containing an admixture of tin was acceptable, but lead tended to debase
the value.185 A memorial of 1505 requested the emperor's permission for
coinage to be allowed to contain anything from one-sixteenth to one-eighth
of tin, but this rule was at that time applied only to the coinage in Peking.186
In the fourteenth century, prior to the proscription of private trans-
 2, IV The monetary problem 75
actions in precious metals, it was decreed that 1,000 standard coins were
to be worth 1 tael of silver. After 1500, partly due to the rise in the price
of copper, the rate was revised to 700 and sometimes 800 coins per tael of
silver. The actual rates in private transactions varied considerably from
this norm, depending on local copper prices and the quality of the coins.187
The coins were produced not by stamping but by die-casting. An illus-
tration in the Tien-kung K'ai-wu shows that the two halves of the hollow
molds resembled an attache-case standing on one of its ends and contain-
ing dozens of impressions of coin. The molten metal was poured in from
the top. 188 After the coins were cast, a great amount of work was involved
in trimming and polishing the edges. One source indicates that up to one-
third of the metal had to be filed off in order to produce top-quality coins.189
The description of the trimming process is not entirely clear, but it seems
that a lathe-like device was used. Coins were fixed in position for trimming
by stringing them on a square rod, of which the two ends were clamped
tight. In the early sixteenth century, it is claimed, this process was elimin-
ated in favor of hand-trimming in order 'to save manufacturing costs'. 190
The workers, therefore, mixed more tin and lead in the alloy to facilitate
the trimming. This is said to have been a major cause of the deterioration
in the quality of government-minted coins, which in its turn encouraged
counterfeiting. These details are of considerable significance; for they show
that it was inadequate financing rather than lack of technological skill that
forced down manufacturing standards.
The minting of money was from the beginning retarded by the policy
of promoting paper currency. The government was unwilling to create
coinage that would compete with its own legal tender, and even more so
after the devaluation of the paper notes which put them roughly on a par
with the copper coins.191 It has already been observed that up to the mid
fifteenth century the Ming court from time to time proscribed the circula-
tion of coins. Such orders were ineffective as the populace simply resorted
to using the coins minted by previous dynasties.
Copper coins were minted, albeit infrequently, by Hung-wu, Yung-lo,
and Hsiian-te, but seventy years after 1433 no coins were produced at all.192
The records of the early production are incomplete but the fragmentary
figures show that the quantities were not enormous. The largest numbers
of coins minted by Hung-wu, for example, were the 222,401,956 cash in
1372 and 199,849,832 cash in 1374, both amounts being close to 200,000
taels of silver in value.193 Official records indicate that even at the high
point of production in the fourteenth century the state's normal annual
productive capacity was only 190,667,800 coins.194 The experience of the
Northern Sung indicates that to provide an adequate money supply the
state would have had to produce 2 to 3 billion coins annually, roughly
50 new coins per capita per year.195 Ming production, because of its con-
76 The heritage of the sixteenth century
stant fluctuation, never approached this level. Moreover a large proportion
of the coins minted in the early Ming circulated overseas. Cheng Ho's
expeditions had exported unspecified sums of copper cash,196 and in other
cases coins were also granted to foreign emissaries. In 1453 the Japanese
tributary mission alone carted off 50.118 million Ming coins.197 Up till then
the Ming court had no legitimate way to circulate its own coins in the
domestic market.
By about 1450, when the prohibition on the use of silver and copper
coins was rescinded, the acute demand for copper cash led to a 'copper
coin famine'. Yet throughout the rest of the century the government did
practically nothing to relieve the shortage. At the same time large quantities
of privately-minted coins appeared in the markets. The makers of these
counterfeits, which contained lead, iron, and sand, could imitate any of the
ancient issues which were in circulation.198
When the court finally took action in 1503 there were no funds available
to finance the production of coinage and the capacity of the mint in Peking
was very limited.199 The solution was to follow the fourteenth-century
precedent of assigning productive quotas to the provinces. The provincial
officials were then required to turn out a certain number of coins to the
specification of the central government. Ordinarily, coining money was
a profitable undertaking. Even in the late sixteenth century the minting of
copper coins could still yield a 40 per cent return (6, n). In this case how-
ever the central government was not authorizing the governors to mint their
own money. Its requisitioning of specified quantities of coins from them
was, as pointed out by a supervising secretary, in effect a new form of
taxation.200 Nanking, the southern capital, was required to produce
25.6 million coins. Many officials in the south jointly signed a petition to
the emperor stating that the local districts had bad years and the assigned
quota was much too heavy a burden on the populace. The emperor sub-
sequently reduced the quota to two-thirds.201 Yet this quota, even if
realized in full, would have been worth only 36,657 taels of silver, an
amount too insignificant to alleviate the coin shortage.
In 1505 it was acknowledged even in official circles that government
artisans could not compete with the workmanship of the counterfeiters.
A group of officials, after touring the Peking mint, reported that the
government workers 'clumsily hammered and chiseled the coins', which
were 'awkward, unbecoming, and shapeless'.202 When the order to coin
money was issued in 1503 the total production envisaged was about
200 million coins.203 By the summer of 1505 the ministry of revenue
reported that less than 20 per cent of the assigned quota had actually been
produced.204 Apparently the original plan had still not been fully carried
out even by 1509, in which year the whole project was suspended.205
The Chia-ching Emperor was the last Ming monarch to make a serious
 2, IV The monetary problem 77
attempt to maintain copper coinage as a state institution, ordering pro-
duction to be resumed in 1527.206 Despite efforts to improve the workman-
ship and increase the weight of the coins, monetary disorder worsened
during his reign. By the 1550s the market was so inundated with low-
quality counterfeits that the rate of exchange was 6,000 coins to 1 tael of
silver. Unscrupulous persons put shopkeepers out of operation by forcing
them to accept coins at the official rate of 700 per tael. A memorial to the
emperor in 1554 claimed that people were dying in the streets of Peking
because of the prevailing confusion in the monetary system.207 The logical
connection was not clearly stated but it is reasonable to deduce that there
was a sharp rise in food prices and in unemployment.
The economic historian, Li Chien-nung, sees in this the operation of
Gresham's Law that bad money drives out good. He claims that because
the Chia-ching coins were too heavy for their face value they were melted
down by counterfeiters for the metal which they contained; some of the
high-quality coins, moreover, were hoarded by the people at large.208 There
is a grain of truth in this theory but as a complete explanation it is un-
acceptable. Li ignores the factors of the quantity of government coinage in
the mid sixteenth century and its quality control.
The Chia-ching coinage has unfortunately been the subject of consider-
able misunderstanding. The Ming-shih states that in 1553 government
mints produced a total of 95 billion coins, the information being cited from
the Ta-Ming Hui-tien.209 P'eng Hsin-wei, a major authority on Chinese
monetary history, calls this fantastic figure 'impossible'. Himself a collec-
tor, P'eng declares that there is no trace of some of these coins said to have
been produced in such vast quantities, and suspects that while there might
have been a proposal for their mass production it could not have been put
into effect.210 One might add that the value of 95 billion coins in the mid
sixteenth century would have been roughly equivalent to the total cash
income of Peking over a period of twenty years. To attain such an output
most of the government mints would have had to be expanded a hundred
times over.
All the verifiable information indicates that the total amount of coinage
was very small. In 1527 the mints of Nanking and Peking turned out only
41,491,200 coins. In 1540 it was decided that the operation was unprofitable
and it was again suspended.211 The boldest proposal for large-scale pro-
duction during the entire period came from Supervising Secretary Yin
Cheng-mou (later minister of revenue, 1576-8) who in 1555 called for an
annual output of 650 million coins utilizing Yunnan copper supplies. This
would have required an annual investment of 390,000 taels of silver,212
but when the plan was put into effect, it was drastically scaled down and
only 20,000 taels of silver invested. The annual output was only slightly
more than 33 million coins, worth less than 50,000 taels of silver. Though
78 The heritage of the sixteenth century
the project realized a nominal profit of 150 per cent, this was only made
possible by demanding that the local population transport the money to
the central provinces. The Yunnan provincial administration repeatedly
pleaded that the burden was too great for the province to bear and in 1565
production was finally ended.213
The production of coinage at Nanking and Peking, suspended in 1540,
was later resumed, though the exact date is uncertain. But in a report
to the emperor in 1564, Grand-Secretary Hsu Chieh (in office 1552-68)
revealed that there was only 28,000 taels of silver invested in the Peking
mint.214 In short, there is no evidence that coin production could ever have
significantly exceeded the 100 million level in any one year in the Chia-
ching period, and in some years it was suspended completely. Even though
the Chia-ching Emperor attempted to establish a bi-metallic monetary
system and the copper coin supply did not necessarily have to accord
with the per capita ratio of the Sung, the new production, at a rate of less
than one coin per person, could not possibly have met the general demand,
especially in view of the higher level of economic activity in the Ming.
The lack of quality control of the Chia-ching coinage is confirmed by
Hsu Chieh's statements. The government mints produced many types of
coins, of which the grand-secretary attached samples for the emperor's
inspection. One of them, which was trimmed by means of the lathe-like
device, was known by the users as 'wheeled edges'. Another kind, which
had a painted surface, was termed 'golden backs' while the third type,
the surface of which was black and rough, was given the name 'paraffin
money'. There were still worse coins which did not fall into even these
classifications. Hsii convincingly argued that these shortcomings were the
fault of the government agencies rather than the general public. At his
suggestion the mint in Peking suspended its operations, as did the Yunnan
mints in the following year.215
Part of the responsibility for this failure rests with earlier emperors. The
coinage was not gradually built up to begin with. Handicapped by this
late start the government in the sixteenth century further underestimated
the magnitude of the problem. At a time when it should have put all its
resources into the production of a new currency, it looked merely to the
immediate profit, to the detriment of long-term planning. The new coins,
small in quantity and unstandardized, had no chance of success. Nor did
Ming coinage ever depart from the traditional design. Unable to replace
the issues of previous dynasties, the court merely proclaimed laws to
demonetize them, which only confused the users.216
After the failure in the 1560s no further attempts were ever made to
regulate exchange rates in private transactions. The people were advised by
successive imperial orders to handle the different kinds of coins 'as they
wished'.217 Up to the end of the dynasty the taxes paid in copper coins
 2, IV The monetary problem 79
were limited to certain business taxes in the cities and percentages of the
inland customs duties. Preference for one kind of coin over another con-
tinued in official collections,218 but this still does not mean that the Ming
coins were the legal tender. After 1571 government mints were occasionally
put back in operation,219 and a general order for coin production issued in
1576 even anticipated that the population would be able to pay the land
taxes in copper cash, a goal that was never realized.220 In general, the
government's attitude was cautious.221 It accepted the market value of
its own coinage, and occasionally managed to realize a small profit on
minting money by taking advantage of the fluctuations in its value (6, n).
In the late sixteenth century the use of silver in tax payments as well as in
private transactions spontaneously increased. The precious metal was
always handled in ingots, chunks, and bits, and no contemporary evidence
has ever come to light that suggests the slightest intention of minting silver
dollars. Modern China produced its own silver dollars for the first time in
the nineteenth century.222 It is difficult to dimiss the possibility that this
prolonged inertia has been somehow related to the painful experience with
copper cash under the Ming.

Unminted silver in tax administration and as a common medium

of exchange
The use of uncoined silver in the fiscal administration of a large country at
the dawn of modern times was a unique situation. When the Ming govern-
ment failed to develop an adequate income from silver mining (6, i), its
loss of control over money and credit was complete. This meant that fiscal
administrators had to perform their task without one of the essential tools.
The resulting hindrance to tax collection and delivery was evident, but
there were other implications besides.
The money supply situation in the late sixteenth century is still open to
speculation. Liang Fang-chung has estimated that between 1390 and 1486,
domestic mining produced a total of 30 million taels of silver, and that in
the last seventy-two years of the Ming, more than 100 million silver coins
were imported from overseas.223 P'eng Hsin-wei shows that under the
Yuan Chinese silver had steadily flowed into central Asia. By the time the
Ming came to use silver in official transactions it was already in short
supply. He interprets one seventeenth-century source to mean that towards
the end of the Ming there was only 250 million taels of silver in the hands
of the populace. This figure included silver utensils and jewelry that could
readily be converted into cash.224 Such hypotheses naturally require a great
deal of verification. Yet it seems to be a fact that the volume of silver in
circulation in the late sixteenth century was small. There is evidence that
farm prices were apt to drop sharply immediately after the harvest, when
80 The heritage of the sixteenth century
taxes had to be paid in silver, a subject that will be taken up later in con-
nection with the land tax administration (4, m). Price fluctuations would
have occurred under any circumstances but their severity suggests that
inadequate money supply could have been a key factor.
The use of silver bullion in tax administration was fundamentally un-
planned and not even adopted by choice. Its drawbacks could not have
been foreseen and modified. The commutation of taxes and service obliga-
tions in the sixteenth century was a long-drawn-out and quite unsystema-
tic process. The insufficiency of silver in circulation was doubtless one
reason why some of the obligations were never commuted. Ku Yen-wu,
describing conditions in the mid seventeenth century, still felt that collec-
tion of taxes in silver was a mistake and argued for a reversion to taxes in
kind.225
Tax delivery in the late sixteenth century involved basically a northward
movement of the bullion. The central and southern provinces submitted
their payments to Peking while the northern provinces, in addition to
making deliveries to Peking, also forwarded tax payments to the army posts
further north. After the mid sixteenth century annuities to those posts from
the capital were stepped up. The revenues from the salt monopoly moved
along the same path. The essential routine deliveries can be estimated to
have involved the movement of at least 5 million taels in a more or less
southeast to northwest direction.226 Undoubtedly a large proportion of the
bullion returned eventually to its points of origin. The northern frontier
was an absolute barrier as far as silver was concerned. Without this return
its northward movement could not have continued uninterruptedly for
over a century. Though the data are still inadequate, the work of several
modern scholars on the cotton and cotton-goods trade and the porcelain
industry, and contemporary records of the frontier, suggest that the silver
returned by roughly the same route as it came, destined for purchase of
basic commodities in the south and the southeast.227
This circuit placed diagonally like a huge paper clip on the map of Ming
China, while promoting the circulation of money, could also conceivably
have intensified the regional economic imbalance within the empire. It
forced the regions at one end of the circuit to earn their silver through
industrial production and those at the other end through governmental
services. Ku Yen-wu, who had travelled widely, pointed out that in his
time silver shortages were most acutely felt in Teng-chou and Lai-chou in
Shantung and west of Hu-hsien in Shensi, two major areas that were left
out of the bullion circulation.228 At the same time the slow process of tax
collection and delivery must have kept the bullion, which constituted
a significant proportion of the money supply, out of normal market
circulation for months.
A more detailed study of this subject will be the task of general economic
 2, IV The monetary problem 81
historians; but from the standpoint of financial history it can be shown that
the sixteenth-century use of silver in tax administration, especially without
effective backing by copper coins, had several disadvantages. Promoting
the use of unminted silver could actually discourage investment. Clearly
a man who was worth 1 billion copper cash could not possibly have kept
that number of coins in his possession. On the other hand a millionaire in
the late Ming could physically hoard a million taels of silver within his
household. A memorial dated about 1580 declared that there were many
households south of the Yangtze that possessed literally several hundred
thousand taels of silver each.229 The usual method of safe deposit was to
bury the bullion underground.230 Understandably, silver became widely
used for jewelry, ornaments, and utensils all of which, in theory a form of
ready cash, actually reduced the amount of money in circulation.
One advantage of the system, nevertheless, was the absence of inflationary
pressure. The long-term price structure in silver in the sixteenth century
was remarkably stable. Barring regional differences, seasonal fluctuations,
and the effects of natural disasters, basic commodities throughout the
century showed little price variation.231 The only notable exception was the
widespread fall in farm prices reported in the decade between 1570 and 1580,
which was possibly a consequence of the financial retrenchment carried out
by Chang Chii-cheng (7, in). State treasuries are known to have accumu-
lated large quantities of silver during his period in office. Prices began to
rise towards the end of the century, but still only gradually. It was only in
the early seventeenth century that military expenditures disturbed the
economy to such an extent that price increases became drastic. By this time
moreover one must also take into account the effects of increased importa-
tion of silver, a trend that was to continue for another 200 years.
 3
The land tax—(i) Tax structure

Land taxes in the late Ming period were undeniably as complex as personal
income tax in the twentieth-century United States, if not more so. Present-
day income tax is complex because the schedule has to deal with numerous
special cases, of which the upper income group is the most complicated of
all. The complexities of the Ming system, however, applied to all taxpayers,
to owners of 5 mou and 5,000 mou of land alike. This makes the tax
structure most difficult to describe. Furthermore even though tax regu-
lations at the local level under the Ming followed a general pattern, there
were numerous internal variations between one prefecture and another and
sometimes even among counties in the same prefecture.
The conventional way to examine a specific financial institution is to
start with a schematic presentation, explaining the terminology and pre-
senting diagrams. This approach has not been completely abandoned. (The
main features of the land tax system have already been enumerated (1, n)
and in Fig. 3, they are presented diagrammatically.) But it has its limitations
in dealing with the Ming taxation system. A more detailed investigation of
the problem would soon show that there is no way of citing all the varieties
and exceptions which occurred. In these circumstances generalizations could
easily be misleading.
This section therefore proceeds in the following order: first the tax
regulations of a selected county are described and the causes of their com-
plexities traced. It should provide valuable insight into other cases in the
same system. Next the major regional differences are outlined. The merger
of the regular land taxes with the service levy, a fundamental cause of all
these complexities and one of the most important developments in Ming
financial history, is the topic of the third section, and is followed by an
explanation of the remaining common tax denominations. While figures
are cited as required, discussion of the general level of tax collection is
reserved for the following chapter on tax administration.
In other words partial evaluation of specific cases precedes overall
generalization, and some conclusions are drawn prior to reviewing numeri-
cal data. An analysis of the dynamics of the system will precede the ex-
haustive description of its terminology.
[821
 oo

partial absorption
into the land tax KEY
(A) = always paid in silver. (S) = some commuted, applicable to
(DAS) =directly assessed in silver. surcharges also.
(FCSP) = further commutation to silver possible. (Tp) = could be used to pay iso-pan orders.
(M) = mostly commuted. (VR) = very rarely commuted.
(R) = rarely commuted.
(P) = permanently commuted.
regular trans-
mittance (S)

added transmit-
tance (S)
paid in grain tribute grain to Peking
white grain to Peking (VR)
+ surcharges covering southern grain to Nanking (S)
spoilages and transpor- frontier post delivery (S)
basic assess- tation (S) local granary delivery (M)(Tp)
regular land taxes
ment in piculs
on the mou of grain permanently commuted J [ G o l d Floral Silver (P)
V

THE commuted to other com-

basic assess- principally in cotton wadding,
LAND ment in other modities
accessory cotton cloth, sesame seeds,
TAX surtaxes commodities + surcharges covering dates (FSCP)
transportation
hay-I-transport (S)
farmland silk (S) principally in silk wadding (S), silk fabrics (S), and cotton
poll tax silk (M) wadding (R)
hemp (S)

horse payment (A), revenue from grassland reserved for salterns

variations of land taxes (M), rents on government pastures, palace estates, etc. (DAS)

fish duty (M), local business taxes (A), stamp tax (A), rents due
absorptions of uncollectibles to lost public land (M)

FIG. 3. Land tax structure in the late sixteenth century.

84 The land tax—(i) Tax structure
This revised order is dictated by the subject matter. Land taxes under
the Ming were at no time completely static and were subject to both
controlled and uncontrolled growth. The rationale of the changes can often
be discovered beneath the inconsistencies. If it can be ascertained where
control was exercised in the course of this growth and where it was lacking,
it should not be difficult to understand the structure and functions of the
land taxes.

I THE COMPLEXITIES OF THE TAX STRUCTURE

The extent of the complexity: the case of Shun-te
The financial section of the 1585 edition of the Shun-te county gazetteer
commences with the observation that in general the tenants in the district
paid their landlords a rent of 2 piculs of unhusked rice per mou, amounting
to 0.9 piculs when husked. Since the basic tax rates assessed in husked
grain averaged 0.03 piculs per mou, the collection represented only one-
thirtieth of the landlord's income. Because the tenant shared the crop with
the landlord on a fifty-fifty basis, the taxation level actually stood at one-
sixtieth of the total grain output. Why then, the gazetteer asked, were there
complaints of over-taxation?1 The modern historian might well ask the
same question. Before attempting to answer it however it is essential to
describe the local background of the tax assessment.
Shun-te was situated in Kuang-chou prefecture near modern Canton
on the southeast bank of the Hsi river. Both climate and soil were favor-
able to rice production. Hilly land comprised only 8 per cent of the district's
taxable acreage.2 In 1581 a land survey was conducted with the purpose of
tax reapportionment (Chang Chii-cheng's land survey, 7, m). As was the
case elsewhere, tax assessment in the county up till then showed consider-
able illogicality. Even though official accounts showed that 48,313 mou of
land were government property, such public estates remained unidentifiable
in practice. Taxpayers were assessed at higher and lower rates, yet it was
impossible to determine how many of the high rate payers held land which
had originally been government property. It was equally uncertain that all
the low rate payers had clear titles to their land. In other words, land tenure
had become confused and the rents on public estates had become hard to
separate from the regular land taxes on private property. Nor did the rate
differentiation bear much relation to soil productivity. When the survey
was completed the county government finally acknowledged, apparently
with Peking's approval, that the public land had been lost. Thereafter all
registered acreage was recognized as private property. The rents on the lost
public estates were collectively assumed by all land-holders under the new
registration. Full advantage was taken of the survey to reapportion the
 3,1 The complexities of the tax structure 85
regular land taxes among the landowners. The tax quota that the imperial
government had assigned to the county was thus more evenly distributed
among the taxpayers.
The reapportionment was beyond any doubt a great improvement. The
decision to hold the new landowners collectively responsible for the dues
on lost government properties was inevitably criticized, but nonetheless the
amount apportioned to each individual taxpayer was infinitesimally small.
The new formula for tax assessment was both fair and simple. It em-
bodied the following three major items :3
(a) All taxable land in the county was classified into upper, middle, and
lower grades according to soil productivity. Hilly land was included in the
lower category. The basic assessment on upper-grade land was 0.0404
piculs of husked rice per mou; that on middle-grade was 0.0273 piculs per
mou and on the lower grade 0.0172 piculs per mou.
(b) All taxable land, regardless of classification, was assessed at a uni-
form 0.0094 piculs per mou to compensate for the rents lost to the govern-
ment by the write-off of public estates.
(c) A 1 per cent surcharge was added to the grain payment to cover
losses due to handling. This applied to both the regular taxes and the
residual rents. In the gazetteer the regular land tax was referred to as
min-mi, literally 'civilian rice' and the shares of rent kuan-mi, literally
'official rice'. These two terms will be employed in the following para-
graphs to avoid confusion.
After fixing these three major provisions, the gazetteer editors refrained
from stating the total combined amount of kuan-mi, min-mi, and surcharges
collectible from each class of land. It seems rather strange that they should
have avoided such a simple mathematical computation since all they needed
to do was to add the particular min-mi rate to the general kuan-mi rate and
add 7 per cent. It can thus be worked out that the amount assessed on
upper-grade land was 0.053286 piculs per mou, on middle-grade 0.039269
piculs per mou, and on lower-grade 0.028462 piculs per mou. The ratio is
close to 5:4:3.
In point of fact such a computation was of little practical value because
by 1585 all the taxes in the county had been made payable in silver. There
was a difference between the conversion rates for min-mi and kuan-mi. For
the former it was 0.6028 taels of silver per picul, close to the market price in
the late sixteenth century; and for the latter it was 0.2653 taels of silver
per picul, a rate greatly to the advantage of the taxpayer.4 This makes the
procedure more cumbersome but does not prevent the calculation of the
total tax payments. The only difference being that this time the 7 per cent
surcharge has to be added separately. The separately converted items of
kuan-mi and min-mi, when added together, constitute the total payment.
The calculation shows that each mou of upper-grade land was assessed
86 The land tax—(i) Tax structure
at 0.0287262258 taels, the middle-grade mou at 0.0202767782 taels, and the
lower-grade at 0.0137623186 taels. The ratio is now close to 4:3:2.
The question might be asked why such vital figures were omitted from
the gazetteer. Equally puzzling is why, after the public estates were written-
off, the two tax denominations of min-mi and kuan-mi should still have
been preserved and different conversion rates applied to them. The first
question can be answered in full here; the second will be discussed under
the next heading Causes of the complexities, below.
It must be recalled that at no time in the Ming period was silver ever
announced as the official standard. The national accounts still used the picul
of grain as the basic fiscal unit and the local districts had to do likewise in
order to maintain the fiscal uniformity of the empire. Even though com-
mutation of tax payments from grain to silver became widespread, in theory
there was no guarantee that the central government in an emergency would
not instruct the county magistrate to deliver a certain number of piculs of
grain to another district. When such a situation occurred, the conversion
rates had to be discarded and the surcharge percentage revised. The tax
provisions in the gazetteer that have been outlined so far in reality involved
both permanent regulations and temporary procedures; they followed
both the central accounting system and local usage. The total tax quota
assigned in piculs of grain to the county could be regarded as permanent, in
accordance with the general practice of the time. The basic apportion-
ment of min-mi and kuan-mi was in principle settled in conformity with
the national standard. The commutation of grain payments to silver,
the adding of the 7 per cent surcharge and the particular conversion rates
were, however, subject to later readjustment.
Moreover the tax rates in silver computed above, each containing ten
digits after the decimal point, were not necessarily the amounts paid by the
individual taxpayer, but were only a general standard. The tax structure
had to be adjusted to include the partial collection of the service levy. By
1585 Shun-te county had seven major items of service levy (each could
also be broken into two or more components) that were totally or partially
levied on agricultural land. This was simple compared to many districts
where the surtaxes often exceeded twenty in number. This simplification
was promoted partly through the efforts of P'an Chi-hsiin, provincial
surveillance commissioner of Kwangtung (1, i). In 1559 P'an demanded that
all the districts under his jurisdiction collect a silver payment called the
chiin-p'ing-yin, or 'equalization silver payment', a concept not funda-
mentally different from that of the Single Whip Reform.5
The gazetteer shows that the equalization silver payment in Shun-te was
collected from three sources. The first was the pool of taxpayers consisting
of all able-bodied male adults, the ting. Secondly, 'abstract ting' were
formed, by assessing every fifty mou of taxable land as equivalent to a real
 3, I The complexities of the tax structure 87
ting for tax purposes. Their total was then added to that of the male adults.
Thirdly an extra 'piggy-back' assessment was added to each picul of the
regular min-mi collection.6
The most notable feature of this formula is the way in which it combined
poll and property taxes, the former being represented by the first source
and the latter by the second and third sources. The second and third sources
could not be combined, because they represented different principles of
taxation. The units of fifty mou of taxable land involved acreage alone,
regardless of soil productivity. Their purpose was to make the taxation
broadly based, ensuring that each landowner paid his fair share. Those
who owned less than fifty mou also paid fractions of the abstract ting
assessment. The amount added to the min-mi, on the other hand, reflected
both acreage and grain yields, greater emphasis probably being given to the
latter. A man who was assessed a large amount of min-mi as his basic tax
payment was likely to own land of good quality and enjoy a substantial
grain income; it was unlikely that such a man would hold huge tracts of
lower-grade land. Since the equal silver payment meant that such a man
was taxed twice, both on his entire acreage and on his total grain income,
the system also implied some measure of progressive taxation, which
should in theory have been effective even when the affluent owner divided
his property into small lots and registered them under different names.
The equalization silver payment was designed to pay all the annual
office expenses of the magistrate's own administration to replace part of
the requisitions from the li-chia. A similar collection called chiln-yao
(equalization of corvee labor, (3, n)) was designed to replace all the labor
service requisitions which were also managed by the county magistrate. At
the time the gazetteer was published there was as yet no indication that
the two kinds of service levy might be combined, because the chiln-yao
collection had just been changed from quinquennial to annual collection
and the procedure was not all settled.7 The chiin-yao, which also combined
the features of a poll and property tax, was however derived from two
segments instead of three. It imposed a collection on the ting or able-bodied
adults and added another surtax to the min-mi, but omitted the conversion
of every fifty mou of taxable land into an abstract ting. Its rates were, of
course, different from those of the equalization silver payment.
The county in addition contributed its share of palace supplies and
furnished raw materials under the jurisdiction of the ministry of works.
The materials were partially delivered in kind by the magistrate and par-
tially submitted in silver payments. Owing to the accounting system, which
was based on allocation of funds prior to collection rather than after it, the
assessments remained as three separate items (supplies to the palace, in
kind, materials to the ministry of works, in kind, and materials to the
ministry of works, in cash), each involving a surtax on both the min-mi and
88 The land tax—(i) Tax structure
the kuan-mi. In every case the kuan-mi rate was considerably less than that
on the min-mi, in accordance with the earlier principle that tenants on
government land, who paid a higher percentage of their grain income to the
state, should be assessed with lighter service obligations than private land-
owners. But because each taxpayer's basic assessment now included both
kuan-mi and min-mi these three service assessments resulted in six different
kinds of surtax on his property.8 The county magistrate was also required
to deliver silver payments to subsidize postal stations in the adjacent dis-
tricts, the funds being derived from a single surtax on min-mi? Like other
counties, Shun-te was obliged to maintain its quota of militia, which was
usually supervised by the military defense circuit intendant. About two-
thirds of the necessary funds were derived from a surtax on min-mi and the
remaining third was paid by the abled-bodied male adults, the ting.10
The above items represented three surtaxes on kuan-mi, seven surtaxes
on min-mi, one combined tax on the total acreage of taxable land, plus three
different kinds of assessment on the able-bodied male adults in the house-
hold. In addition to these there were other minor kinds of service levy,
such as funds collected to pay for military equipment in a nearby army
unit.11 The accounts being small, they are here regarded as nuisance taxes
and excluded from the discussion.
At this point it begins to become apparent that what initially seemed to
be a single tax formula could develop into a great maze of surcharges and
surtaxes. We have made several mock case-studies of the tax schedule for
different taxpayers holding varying amounts of land of upper, middle, and
lower grade and with different numbers of male adults in their households.
Using longhand it takes at least one and a half hours to settle the most
simple case, for the decimal digits in the fiscal units can in theory be ex-
tended indefinitely. Yet on the other hand the actual level of taxation,
despite the fourteen kinds of surtax, was low. For small landowners with
less than 30 mou of mainly lower-grade land each, and no more than two
male adults in their household, the tax burden was in general less than
5 per cent of their estimated grain income. The progressive feature of the
tax formula can also be verified. For a medium landowner who had 300
mou of mostly upper-grade land, and five to six adults in his household,
the total tax burden could be close to 10 per cent of his estimated grain
income.
The low level of taxation can be further confined from the aggregate
figures listed in the gazetteer. In 1585 the min-mi and kuan-mi plus the
7 per cent surcharge together totalled 34,689 piculs. The amount was
actually paid with 17,952 taels of silver.12 Following the detailed description
of the tax procedure in the gazetteer, we computed the additional tax
burden borne by the agricultural land as a result of the apportionment of
the service levy, seven items in all, but not including the portions assessed
 3, I The complexities of the tax structure 89
as poll taxes. The aggregate was 11,324 taels. The total burden on the
taxable acreage was therefore 29,276 taels of silver. Dividing this figure
by the 883,706 mou of land under registration13 shows the average tax
burden to have been 0.0332 taels per mou. Since hilly land in Shun-te
comprised only 8 per cent of the total taxable acreage, its lower yields
should not have affected the average grain output in the district significantly.
The opening paragraph of the gazetteer mentions that an average mou of
land produced 1.8 piculs of husked grain. The local grain price being
somewhere between 0.5 and 0.6 taels of silver per picul (4, n), the average
mou must have produced an annual income of around 1 tael of silver, not
counting miscellaneous crops planted out of the regular season. The general
level of taxation was therefore close to 3.5 per cent of farm income. An
unofficial source indicates that in the late sixteenth century the grain price
in Kuang-chou prefecture could fall as low as 0.3 taels per picul,14 perhaps
due to the effects of deflation in the decade between 1570 and 1580 (2, iv
and 7, in). That would reduce the estimated annual income to 0.54 taels of
silver per mou but even then the level of collection would still have been
no more than 6.12 per cent of farm output.
It is not difficult, however, to imagine the actual situation when this
complicated tax schedule was applied to sixteenth-century agrarian society
by an understaffed local government. Ming writers left volumes of com-
plaints against the corruption and abuses of rural collectors, office clerks,
and influential landowners. Such malpractices took various forms and
employed different methods, and were only to be expected.
The fundamental fallacy of the taxation system was neither a high level
of collection nor a lack of egalitarian provisions in the tax legislation. On
the contrary, those objectionable features were notably absent. The weak-
ness of the system was the complexity of the tax schedule, which even the
gazetteer could not spell out in full. Tax rates stated to twelve or fourteen
decimal digits were grotesque, and had never been seen before Ming times.
Faced with the same absurdity at about the same time, the prefectural
gazetteer of Sung-chiang in South Chihli declared: 'Silver should not be
counted beyond one-thousandth of a tael; grain should not be counted
beyond one-thousandth of a picul. Anything beyond that must be omitted,
otherwise we shall all fall into the mists and fogs of the wicked. ' 15 No such
basic reform was ever carried out in the Ming however. The first imperial
edict ordering the elimination of figures after the fifth place after the
decimal point in all fiscal accounts was issued in 1685 by the K'ang-hsi
Emperor of the Ch'ing dynasty.16 Even so the cumbersome digits still
appeared in Ch'ing ledgers for another fifty years. As has already been
observed, Shun-te county also started with tax rates containing no more
than four digits behind the decimal point. The proliferation of digits could
be stopped neither by the gazetteer writers' appeals, nor the emperor's
90 The land tax—(i) Tax structure
edicts, for it was a direct consequence of the manifold complexities of the
basic tax structure.

Causes of the complexities

The tax procedure of Shun-te county was by no means the most compli-
cated among those in operation in the late sixteenth century, though it was
not the simplest either. Because every district had its own particular prob-
lems of taxation, it is impossible to cite a typical case. Yet most of the
complexities of procedure had common origins. The effects might be
different but the causes were similar. The tax administration of Shun-te
county has been described in considerable detail because its gazetteer
enumerated step by step in a more illustrative way the method of arriving
at the rates of collection.
Among the many factors contributing to the complexities of the tax
structure, agricultural methods cannot be ignored. Rice cultivation,
especially when it involved terrace farming, had a predetermined effect on
land-holding in that it was conducive to the parcelization of agrarian land.
The irrigation of rice paddies on different levels, complicated by the topo-
graphical contours, made large-scale operation virtually impossible. Ho
Liang-chiin's statement that a husband-and-wife team could properly
cultivate only between five and twenty-five mou of land in his native
Hua-t'ing county, South Chihli, must also hold true of many southern
provinces.17 This resulted in the division of agricultural land into small units,
sometimes regardless of ownership. Ho further attested that one household
might own scores of plots of land scattered in different locales, the total
tax payments on the smaller plots ranging from 0.1 to 0.2 taels of silver.18
In Chia-shan county, Chekiang province, in about 1525 the average
taxable acreage in each li was recorded a s ' slightly over 3,000 mou \ 1 9 When
divided among the 110 tax-paying households, this means that the average
land-holding of each household, including hills and ponds, must have been
close to 30 mou, or less than 5 acres. It was quite common in those days
for such small parcels of land to be further divided into even smaller plots,
sometimes even of fractions of a mou, for sale or mortgage. Fu I-ling's most
revealing study presents many actual cases of such fragmented landowner-
ship, drawn from sixteenth-century deeds of sale, mortgage agreements, and
family compacts, recently discovered in a small village in Yung-an county,
Fukien province.20 Fu's sample is small, for the type of documents in his
possession, never having been regarded as respectable source material by
traditional historians, must now be extremely rare. Wei Ch'ing-yiian's
photographic reproduction of a household's tax record of 1644 from Ch'i-
men county, South Chihli, shows a similar case. The taxpayer in question
owned altogether less than 32 mou of land, consisting of eight small plots
 3,1 The complexities of the tax structure 91
scattered over four different villages.21 The fact that such documents can
still be found here and there shows that small-scale land-holding by marginal
landowners was by no means unusual in the late Ming. The parcelization of
agricultural land can be further verified from the descriptions in the 1566
edition of Hui-chou prefectural gazetteer, South Chihli, and the 1572
edition of KWai-chi county gazetteer, Chekiang province.22
This situation created a double problem in tax administration. Tax
legislation had simultaneously to deal with many small landowners and
a few medium and large landowners. To suit the small taxpayers, the rates
had to be fixed on a minute scale; the range for maneuver was limited and
the surtaxes had to be added very carefully. In principle, heavier service
levy obligations should have been assessed on the larger landowners, but
the fragmentation of agricultural land-holdings made it easy for affluent
households to obtain separate registrations under different names, thus
evading heavier payments. The tax formula of Shun-te county shows an
attempt to break through this dilemma. It abandoned the graded ting and
made all able-bodied males pay at uniform rates. By the device of abstract
ting and a dozen or so surtaxes on the basic payment it managed to
equalize the tax burden somewhat. Many of these complications could
undoubtedly have been eliminated, if the medium and large landowners had
all had their landed properties in consolidated and compact estates. In that
case the heavier tax rates could have been assessed directly on such affluent
proprietors.
While the unit of land tax assessment, the mou, was rather small, the
unit of payment was much too large. Unlike the T'ang and Sung, which
used the copper cash as the fiscal unit, the Ming never developed an
effective monetary system (2, iv). When unminted silver was used in
taxation, the problem was that the smallest weights of the precious metal
were too large for ordinary payment. In the sixteenth century 0.6 taels of
silver could be considered as the normal price of a picul of husked rice in
south and central China. Even if the basic assessment were fixed in grain
at the simplified rate of 0.03 piculs per mou, the rate in silver would still be
counted in thousandths of a tael. The surtaxes on the basic payments could
not be fixed on an overall percentage basis. Instead, there was a separate
rate for each item; in Shun-te, for instance, the chiin-yao rate was settled
at 0.1403 taels of silver on each picul of min-mi in basic payment?* In
practice, no taxpayer was assessed a min-mi payment even in whole piculs;
it always involved hundreths and thousandths of a picul. The result was
endless computation of smaller and smaller fractions. Technically speaking,
it would have been much easier to combine the several surtaxes into a larger
percentage, if these surtaxes had also been fixed on a percentage basis to
begin with. The merging of the surtaxes would have followed naturally.
As illustrated by the case of Shun-te, the several items of service levy
92 The land tax—(i) Tax structure
represented different expenditures. The persistence of lateral transactions
in tax payments made it extremely difficult to merge these revenues. Even
when silver became widely used in tax collection, no regional treasuries
were ever established by the imperial government nor did the provincial
governors set up central treasuries. As banking techniques were not
developed to handle public funds, cash flows had to follow earlier pro-
cedures and the numerous consignments of commodities that had moved
from one end of a province to another and from one end of the empire to
another were merely replaced by equally numerous silver consignments.
There is very little substance to the theory that the use of silver constituted
a significant improvement in fiscal administration; it was rather a case of
the same tune played on a different instrument.
As long as the central government did not revise general fiscal pro-
cedures, it would not relax its control over local administration. Tax
collection by the local districts continued to provide the expenses of
offices on different levels. Not only did every account have its own quota
which had to be neither over- nor under-collected, but in addition it was
often delivered by a separate agent to a separate deadline. According to
Ho Liang-chun, up to the mid sixteenth century even the rural collection of
these separate items was handled by different persons. The lack of inte-
gration was so complete that the villagers were visited by squads of tax
agents once every several days.24 At this point local officials felt compelled
to initiate some basic reform, regardless of explicit approval from higher
levels of government.
The various kinds of service levy were in reality not merely separate
surcharges on the land taxes. They were different types of taxes partially or
wholly imposed according to the same tax formula but deriving from
different principles of taxation. The equalization silver collected in Shun-te
county was designed to replace certain li-chia material requisitions and in
principle the handling of the supplies had involved labor service. The
chiin-yao payments substituted for corvee labor; the militia service had
originally been organized within local communities. These three kinds of
impositions therefore had to be imposed on the ting. In view of the low
level of land taxes in the county one might think that no great harm would
have been done and the benefit of immense simplicity would have been
achieved if the said service obligations had been altogether waived and the
entire financial burden transferred to landed properties. This was im-
possible, however. The principle that all male adults were liable to state
service had to be upheld, regardless of whether they owned properties or
not. To abandon the principle would have impaired imperial uniformity.
The advantage of adhering to the ting as a taxable unit was that should the
taxpayer be unable to pay he could be conscripted to perform the service in
person. This is why only the three kinds of collection mentioned above
 3,1 The complexities of the tax structure 93
retained some features of a poll tax. Yet none of these items could remain
as a simple poll tax; much of the burden still had to be borne by agricultural
land. The tax structure failed to achieve the desired simplicity because it
served multiple principles and purposes.
The retaining of kuan-mi as a fiscal item is another instance of the
strong influence of tradition. In the sixteenth century public land in many
provinces had been completely lost (3, n). This was acknowledged by
a number of districts besides Shun-te.25 But there were also counties and
prefectures which claimed that they had preserved such estates in their
districts.26 The imperial government, unable to make any effective inquiry,
simply had to take the position that as long as local districts continued to
deliver the rents on public land in accordance with the established quotas it
would not question the details. Yet in spite of this it never decided, as
a general policy, to write off such public domains. The result was that most
counties retained kuan-mi as a separate item in the accounts. In Kiangsi
the government land had probably been lost before the mid sixteenth
century.27 Yet the provincial accounts in 1610 still maintained that of its
revenue derived from agricultural land, 657,274 piculs of grain was
collected as kuan-mi and 1,871,519 piculs as min-mi.28 Thus it was not only
Shun-te that observed the formality of preserving fictitious rents on non-
existent public land.
It is interesting to observe, however, that through the ingenuity of local
administrators the defunct fiscal term was actually utilized to serve certain
practical ends. The intention was to connect it with the collection of the
Gold Floral Silver. Ever since 1436 some 4 million piculs of grain of the
land taxes had been permanently commuted to silver at the discount rate
of 0.25 taels per picul, the proceeds becoming the emperor's privy purse
(2, i). Shun-te was apportioned some 4,000 taels of this item, known as
Gold Floral Silver. From 1585 the revenue collected as kuan-mi was de-
livered to Peking as payment towards this separate account.29 When the
proceeds from this proved insufficient, a small portion of the income from
min-mi was used to make up the balance. The commutation of kuan-mi,
settled by local administrators at 0.2653 taels of silver per picul, in reality
represented the basic rate for Gold Floral Silver plus a small surcharge of
close to 6 per cent. The demand that the rents on the lost government
properties should be made up by all taxpayers in the district might be un-
reasonable, but through this arrangement the arbitrariness of the order
was mitigated by the favorable conversion rate. The magistrate could
claim that a certain number of piculs of grain were still collected as kuan-
mi; at the same time by classifying this as a separate item of tax collection
he assured himself of sufficient income to meet the demand for the Gold
Floral Silver, which was handled separately when it reached the capital.
For each taxpayer to be simultaneously assigned a separate kuan-mi and
94 The land tax—(i) Tax structure
min-mi payment was also convenient for the apportionment of additional
surcharges in the future, in that, as already pointed out, the kuan-rni
assessment was based on acreage and the min-mi assessment on grain
income.
The reason why the tax structure of Shun-te county seems more com-
plicated than that of other districts is because its gazetteer spells out the
elaborate mathematical formulas behind the tax rates. Such details were in
general omitted by official publications elsewhere.
The omission of these formulas by no means implies a simpler tax
structure, however. Just as a modern accountant in computing a client's
income tax return appears to have simplified the tax procedure by keeping
the technical details to himself, in the same way, the unpublished tax
formulas of other counties and prefectures can be detected from the
decimal digits in their tax rates.
The criss-crossing supply lines that formed a monstrous web over the
empire presented another substantial obstacle to tax reform. They were not
only difficult to disentangle but also difficult to comprehend. The require-
ment for the submission of native goods to the capital alone significantly
complicated the tax schedules in numerous districts, because in principle
the costs had to be compensated for by collecting the exact amount from
the tax-paying public. In 1590 Shanghai county in South Chihli was re-
quired to submit 12 taels of silver to the Imperial Academy of Medicine in
lieu of locally-produced drugs. In theory this item, along with several
others, was still apportioned to every taxpayer and to each mou of taxable
land in the district.30
The situation of Shun-te ^was by comparison more favorable, since it
had fewer items of supplies to be submitted to Peking than did the counties
in Hukwang, Chekiang, and South Chihli. It was not called upon to make
distant deliveries of its tax grain. Nor was it required to provide stabling
services for government horses, as were the northern counties and pre-
fectures (3, n). Apart from the matter of the lost public estates land tenure
in Shun-te was rather simple, and never caused such problems as did land
leases in Fukien and Kiangsi (4, n).
While these additional complexities will still form the subject of some of
the following sections and chapters, a general observation can be made
here that most of them had long-established historic origins. The stabling
service was ordered by the Hung-wu Emperor. The lack of centralized
financing for the operations of the Grand Canal, which resulted in the
differentiated and excessive surcharges on tax grain in transit, originated
in the Yung-lo period. The first permanent commutation of tax grain to
silver in 1436 was intended only to solve a temporary problem, and the
rate was fixed without any thought to its effect on future administration.
The lateral transference of tax revenues was the general policy of the
 3,1 The complexities of the tax structure 95
dynastic founder. The improbable-seeming fact is that by the late sixteenth
century the basic structure of land taxes under the Ming could no longer be
radically simplified because it had already become too complicated. Any
meaningful reform would undoubtedly have entailed a complete admini-
strative upheaval.
The taxation system as a whole could still, however, have been some-
what rationalized if silver had been formally declared the national fiscal
standard and the remaining taxes in kind been altogether commuted. In
1572 in Ku'ai-chi county, Chekiang, the land taxes collected in silver com-
prised 82 per cent of the total tax revenue.31 By 1591, in Lin-feng county,
Shansi, 95 per cent of the land taxes were paid in silver.32 In Shun-te
county, Kwangtung, the commutation of land taxes and their associated
surcharges was a hundred per cent complete by 1585. Even the tribute
grain delivered to Peking in the late sixteenth century remained at the level
of only approximately 2.5 million piculs, less than 10 per cent of the empire's
land tax quota (7, i). Yet no step was taken to eliminate the picul of grain
as the basic fiscal unit.
There were many reasons for this failure to adopt a silver standard and
effect a total commutation. The reform would have affected the vested
interests of the eunuchs in charge of the palace warehouses. The court itself
would moreover have been tightly restricted by a silver budget, whereas
under the existing system it could still count on some of the grain quotas
in the local districts as units contractible and expandable at will. It could
assign to them lower or higher conversion rates as it liked and impose local
procurement orders on them with little reference to current prices. Apart
from all these there were problems resulting from disparities in local grain
prices and silver supplies in the various regions of the country.
Even if most of these problems could in time have been solved, any
meaningful reconstruction of the system would have required a different
concept of statecraft, for it would inevitably have affected the entire im-
perial organization. To declare the use of silver as a national standard
would have been easy; but the establishment of regional treasuries in the
provinces by the ministry of revenue would have called for the separation of
imperial income from local income. The strengthening of logistical capacity
at the provincial level would have required the augmentation of office
personnel and an increase in the administrative budget. The total elimi-
nation of taxes in kind and in corvee labor would have demanded a higher
level of formal taxation, an adjustable budget to replace the tax quota
system, and more effective control of the rural areas. Needless to say,
government in the late Ming had neither the capacity nor the ambition to
initiate such a fundamental reform, which Ch'ing reformers were still
unable to accomplish even in the late nineteenth century.
Tax structure in the sixteenth century thus included many elements
96 The land tax—(i) Tax structure
whose only rationale was earlier precedent. The ad hoc decisions of earlier
emperors were honored as customary laws. As time went on it became
increasingly difficult to make even minor readjustments to them. In the
beginning the imperial government had arrogated to itself too wide an
area of jurisdiction while developing too little capacity to maintain effective
control. Now it was neither able to expand the latter nor willing to relin-
quish the former.
Because the central government could not implement the necessary re-
forms, effort to simplify the tax procedures relied exclusively on the
local officials, usually the county magistrates and prefects. Inevitably
the scope of their reforms was narrow and their attempts uncoordinated.
In the late sixteenth century, most of these officials served a three-year
term in their posts. Owing to the 'law of avoidance' 33 they were often
sent to districts where, in Ku Yen-wu's words, 'the native dialect was
unintelligible and local customs foreign to them'. 34 By the time an official
became acquainted with the exact situation in the district under his juris-
diction, all too often his term of office had already expired.
Officials had varying degrees of success in carrying out local tax reforms
but many of them endured lengthy discouragement and frustration. In 1547
the magistrate of Ku'ai-chi county, Chekiang, attempted to consolidate
the district's sixty-four different grades of tax assessment into three major
categories. His suggestions were submitted to his immediate superior, the
prefect of Shao-hsing. The prefect then had to consult with his assistant
prefect in charge of taxation (kuan-liang fung-p'ari) as to the feasibility
of the new formula and its likely effects on prefectural administration
and public opinion. At this point the prefect specifically pointed out
that any reallocation of taxes should be limited to the surtaxes and sur-
charges, as the basic assessment was a matter of 'fundamental law' and
thus beyond discussion. The scheme was then referred to the resident
circuit intendant, who represented the provincial administration. On his
direction a bulletin was posted in front of the offices of the prefect and the
magistrate, inviting 'concerned elders and experienced persons' to express
their opinions. The more influential persons in the district were canvassed
separately. Since Ku'ai-chi was also a salt-producing district, portions of
the land taxes collected from the registered saltern households had been
designated as part of the salt revenue, and it was therefore necessary to
obtain the consent of the imperial censor on salt administration to the sug-
gested reform. Only after all the interested parties had been consulted and
the proposal amended several times was the new tax legislation submitted,
through the provincial administrative office to the governor for final
approval. The district's tax rates were indeed subsequently reclassified into
three major categories, but within each category there were so many
exceptions, and exceptions to the exceptions, that it is difficult to say how
 3,1 The complexities of the tax structure 97
35
many grades of assessment were actually retained. As a result the essence
of the original proposal was greatly diluted.
Most obstruction of district reform was likely to come from the local
gentry. They, as retired officials or holders of academic degrees, enjoyed
limited tax reductions and exemptions on corvee labor, in theory applic-
able only to themselves and their families, and all based on complicated
schedules. In general they took advantage of loopholes in the existing tax
procedures to extend their privileges. Consequently they were liable to be
affected by any tax reforms. When dissatisfied they could deliberately keep
their payments in arrears, or else apply pressure on the local administrator
via influential persons. Sometimes they approached a higher office to seek
arbitration. Such approaches could be formal or informal. The formal
legal proceedings which they initiated at higher offices were termed
'appeals', but in reality constituted civil suits against the magistrate. Yang
Chieh, magistrate of Ku'ai-chi county, in 1572 recorded five cases in the
district (one of them dating from the previous century), in which local
influential landowners appealed to higher offices in order in overturn the
magistrate's tax ruling. Yang simply referred to these cases as 'lawsuits'.
In at least one case the magistrate's decision was overruled and on another
occasion the case was brought to the attention of Peking.36
The most celebrated county magistrate whose tax legislation met strong
local opposition was Pai Tung, who served in Tung-ao county of Shantung
in 1574. Pai, in applying the Single Whip Reform in his district, collected
the regular land tax at 0.011 taels per mou, plus a service levy charge of
0.0092 taels per mou. Every ting in the county was also assessed with an
annual payment of 0.13 taels. These rates were not exceedingly low for
north China, but the formula was simple. 'After the first year of appli-
cation, 11,000 households which had previously absconded returned to
their home areas': Pai's achievement was even recognized by Chang
Chii-cheng. But on account of 'local dissatisfaction', he was impeached by
a supervising secretary and only on Chang Chti-cheng's intervention was
the case suspended.37
It is safe to assume that in the late Ming most local officials would
secure some consensus from the elite group in their districts before pro-
ceeding to procedural revisions. In about 1609 the magistrate of Wen-
shang county, Shantung, explained the irrationality of the district's tax
schedule by the following observations in the local gazetteer: 'But our
silk-robed gentlemen all insisted on their own views; none was willing to
compromise. Their quarrelsome arguments almost ended in lawsuits.'38
There were, of course, conscientious local officials who did not yield to
gentry power and indeed fought it tenaciously. Yet such heroism was
rarely rewarded and all too often demanded considerable self-sacrifice
from such conscientious magistrates. The public image of the ideal official
98 The land tax—(i) Tax structure
was one who could carry on his administration without a fight. It is
understandable therefore that many local officials refrained from making
innovations. Laissez-faire was the best policy to protect themselves and
their career advancement.39
The complexities of the tax structure were thus the product of many
divergent causes. They depended on methods of rice cultivation, the nature
of the monetary system, differing principles of taxation, the efforts of the
local government to make enactments conform to central legislation, the
failure to apply banking techniques to the management of public funds,
former ad hoc modifications which could not subsequently be done away
with, the inability of the central government to introduce general reforms,
and the limited authority of local officials and their restricted freedom of
action. The land tax under the Ming, after undergoing so many conflicting
compromises, was no longer simply afiscalsystem; it must simultaneously
be regarded as a political and social institution.

II REGIONAL VARIATIONS
The Yangtze delta
Traditionally the Yangtze delta, comprising Soochow, Sung-chiang,
Ch'ang-chou, and Cheng-chiang prefectures in South Chihli, and Chia-
hsing and Hu-chou prefectures in Chekiang, was a source of both profit and
problems to the imperial administrators. Benefiting from the rich soil and
high water-table around the T'ai lake, the region boasted an unrivalled
level of agricultural production. Because of convenient transportation by
water its food surpluses could be exported. Handicraft industries in the
delta area were also extremely advanced. Successful exploitation of this
rich region would have enabled the dynasty to solve most of its financial
problems. Extracting more revenue from this territory was not a simple
matter, however.
Natural forces undoubtedly contributed to the difficulties. Contemporary
writers reveal that in numerous cases taxable land in the Yangtze delta was
washed away byflood,and new alluvial land was created after inundations.40
The situation was worsened by human errors. The traditional method of
irrigation, as Shimizu has pointed out, was constantly working against the
forces of nature. Dikes were constructed with no regard for the underlying
water-table and streams were channeled against their natural direction.
While profitable in the short term, waterworks in themselves increased both
the frequency and intensity of flood disasters, which in turn periodically
altered the local landscape.41 It has been generally assumed that once a
fish-scale book (1, n) was compiled, all landed property was controlled by
a permanently binding record. This was not the case because land surveys
 3, II Regional variations 99
could not be frequently repeated and the features of the terrain were never
in a fixed state.
With local topography constantly changing, the administrators were
compelled to adopt occasional piecemeal remedies, without an accurate
set of land data to base them on. When taxable land was washed away, the
lost revenue had to be made up from the property of the remaining land-
owners. Any new alluvial land was hard to discover and easily evaded
taxation. Even the most alert and conscientious administrators could do
no better than patch up the tax structure. No adequate solution to this
perennial problem was ever found.
Land-holding in this region was further complicated by legal ambiguities.
These can be traced back to the Southern Sung which had, in order to
relieve pressing financial problems, forcibly purchased cultivated land
from owners in the delta area. From Ku Yen-wu's description we know
that the policy was extremely unpopular and possibly created more prob-
lems than it solved; even at the Sung dynasty's fall the promised install-
ments of the sale prices had not been paid off. The Yuan continued to hold
the properties in question as government land and extended its acreage.42
When the Hung-wu Emperor took over the territory by military force he
too, circumvented the legal problem. He simply declared that the populace
in this area, especially in Soochow prefecture, had supported his political
enemy and that their properties were thereby confiscated. No agency was
ever established to administer the confiscated estates and the rents on them
were merged with regular land taxes.43 This ambiguity, compounded by
the problem of a changing topography, made the registration of deeds com-
pletely ineffective. The same problem also existed in Shun-te county; but
the government properties in that district prior to the 1581 land survey
comprised only 5 per cent of the cultivated acreage. In the delta area the
confiscated land and other government estates had at the beginning of the
dynasty been built up to such an extent as to make private property look
insignificant by comparison. For example, Ch'ang-shu county registered
933,763 mou of government land in 1391, as opposed to 308,737 mou
under private ownership.44 Land in the former category exceeded 70 per
cent of the total acreage.
There was no indication that at the outset the government had connived
at the private sale of government properties. This practice nevertheless
started in the earlyfifteenthcentury, as ownership could neither be identi-
fied on the ground nor checked against the survey records. The time-
honored tradition of dividing up land for sale or mortgage intensified the
confusion. Ku Yen-wu reports that sometimes the buyers had no idea what
kind of property they had purchased.45 For more than a century the Ming
government did nothing to eliminate the ambiguity. Only when the more
substantial taxpayers constantly kept their taxes in arrears did provincial
4-2
100 The land tax—(i) Tax structure
officials manipulate the tax schedule to give them covert reductions (see
below).
By the mid sixteenth century the cumulative effects of the confused
situation in land tenure had become too troublesome to be ignored; at the
same time the public estates in the registers brought no real financial
benefit to the state. In 1547 the prefect of Chia-hsing, Chao Ying, suggested
that all taxable land be recognized as privately owned. How the proposal
wasfinallyapproved is unknown, as no official authorization can be traced.
Local gazetteers indicate that the residual rents due on the lost government
properties were apportioned among all landowners in the same way as
was done in Shun-te county many years later.46 Several prefectures in the
delta area immediately seized the opportunity to follow suit.47 This huge
write-off of government properties was undoubtedly one of the most
important milestones in Mingfinancialhistory. The circumstances suggest
that the creation of public estates in the delta area, having commenced
without formal legislation, was rescinded almost two centuries later, again
without any general announcement.
The merging of public land with private property created a special
situation, however. It increased the average assessment on all taxpayers
significantly, owing to the substantial acreage involved in the write-off. In
Chia-ting county, Soochow prefecture, the 1547 reapportionment resulted
in an overall average of 0.3 piculs of husked rice per mou in basic payment.48
The average in neighboring Ch'ang-chou county was 0.37 piculs per mou,
and in T'ai-ts'ang subprefecture 0.29 piculs.49 These overall averages were
about ten times the prevailing rates in most districts outside the delta area.
Considering the higher productivity obtaining in the prefectures of this
region the rates were by no means entirely unreasonable. The actual pay-
ment was further reduced by the partial conversion into Gold Floral Silver
and other hidden advantages. But at a time when imperial impartiality was
held to be thefirstprinciple in taxation these prejudicial rates presented the
local landowners with a cause of grievance.
Service charges on long-distance tax deliveries and ad hoc commutations
were by no means unique to the delta area, but there their effects were more
disturbing. The four prefectures in South Chihli were allocated a total
payment of 1,206,950 piculs of tribute grain (ts'ao-liang) to be sent to
Peking: that is, about one-third of the empire's annual quota. In general,
surcharges of 55 per cent had to be added.50 With the creation of 'regular
transmittance' and 'added transmittance' in 1471, the tribute grain was
further split into two kinds, the added transmittance paying only half the
surcharge rate (2, i). In addition there was a levy of 214,000 piculs of'white
grain' (pei-liang), the highly-polished ordinary and glutinous rice destined
for palace consumption, part of which went to the emperor's table and
part to sacrificial services. It was collected exclusively from the delta area
 3, II Regional variations 101
TABLE 1. Financial burden to the taxpayer of tax grain delivered or com-
muted, for four prefectures in South Chihli, ca. 1585

Form of tax payment Estimated cost

for 1 picul of grain in taels of silver

White grain 1.91

Tribute grain, regular transmittance 1.30
Tribute grain, added transmittance 0.90
Southern grain 0.75
Grain for local delivery 0.63
Commuted to cotton cloth 0.44
Commuted to Gold Floral Silver 0.26

and stringent standards were imposed; delivery was to be carried out by the
taxpayers, not the army transportation corps. During its transit on the
Grand Canal the delivery agents had to pay for all services required at
rapids and sluices, not to mention the extra demands made by the eunuchs
at the receiving depots.51 Part of the tax payments from the delta area,
64,391 piculs in all, was designated as 'southern grain' {nan-Hang), to be
delivered by the taxpayers at Nanking, the southern capital. The delivery
distance was short, however, and the surcharges much less.52 Other
deliveries at local granaries involved even lower surcharges, as it was
possible for most of the consignments to be transported by the taxpayers
in person. Thus there were in all five kinds of grain delivery, each subject
to a different rate of service charge. Since the central government recognized
a picul of tax grain only when it was actually checked in at its designated
granary, thefivekinds of delivery in reality created five classes of tax pay-
ment for the local district.
Apart from this, the four prefectures were also required to pay 365,135
taels of Gold Floral Silver, 4 piculs of grain payment being converted to
a tael. The payment was far below the local grain prices. Part of the grain
payment was also commuted to cotton cloth, 322,774 rolls in all, at a rate
of 1 or 2 piculs per roll depending upon the quality of the textiles. This
commutation also represented a substantial saving to the taxpayers.53
Using the local gazetteer records of prices, surcharges and delivery costs
in the late sixteenth century, the actual financial burden in silver to the
taxpayers for each picul of grain in basic assessment is calculated in
Table I.54
Obviously such a schedule was neither sufficiently logical nor suitable to
be put into effect as it stood; it was necessary to contrive revisions without
overtly contradicting imperial orders. A scheme for revision was indeed
devised by Chou Sheng (governor of South Chihli from 1430 to 1450). He
intended to use the inequalities in surcharges and commutation rates to
102 The land tax—(i) Tax structure
counterbalance the inequalities in the basic assessments. It must be recalled
that in the fifteenth century the institution of government land had not yet
been abolished and most landed properties, after changing hands several
times, ended up paying either extremely high or extremely low tax rates.
Some of those rates were so astonishingly high, reaching as much as
2 piculs of grain per mou, that even the local officials confessed that they
did not know their origin.55 Not wishing to disturb all these, Chou instead
introduced a new element into the tax structure which he called t h e ' levelling
grain' (p9 ing-mi).
First he ordered that all basic assessments, regardless of whether the land
tenure could be verified and whether the rates were high or low, should
remain as they were. But instead of adding the various kinds of surcharges
to these assessments, or converting them, he imposed another universal
service charge to be paid by all taxpayers. In the first year of application he
actually boosted the rates by 90 per cent. It was the basic assessment plus
the 90 per cent horizontal service charge that was termed the levelling
grain. Despite this, it was not paid entirely in grain. Payment in kind was
made only by those taxpayers who already had low rates of basic assess-
ment, 0.4 piculs per mou or less. Taxpayers who had assessments higher
than this paid their taxes in cotton cloth or Gold Floral Silver, and actually
benefited from the commutation. The ordinary surcharges and delivery
obligations were subsequently waived entirely. The 90 per cent horizontal
charge provided the local government with a sinking fund which enabled
it to meet the cost of the various kinds of delivery. In other words, the tax-
payers discharged their tax responsibility in their home districts by paying
the levelling grain, and the local officials carried out the allocation after
the collection. This created more centralized management at the county
level and above. The 90 per cent universal service change was calculated to
produce a surplus in the first year, and the surplus was carried forward to
the next year so as to reduce the rate. The collection was therefore not
totally limited by the quota system but exhibited some features of an
annual budget. Throughout his tenure as governor, Chou's annually
readjustable service charge was never less than 50 per cent, however.56
Chou Sheng was made to retire in 1450 following widespread criticism
and threats of impeachment proceedings, firstly for abuse of his powers
of taxation and later for his mismanagement of the sinking fund.57 The
sinking fund system was then abolished in perpetuity; no local official
was henceforth permitted to collect a horizontal service charge which
was aimed at creating a surplus. But the levelling grain as an in-
stitution survived, being applied to all prefectures in the delta area and
remaining in effect until the end of the dynasty. Unlike Chou Sheng's
system of levelling grain the combined service charges in the sixteenth
century were calculated and fixed at the minimum level and in reality
 3, II Regional variations 103
merely confirmed the quota system. After the tax reallocation that followed
the elimination of government land in 1547, the levelling grain no longer
had to be calculated from the basic assessment but could now be assessed
directly on taxable land in place of the basic assessment. After making
a land survey in 1581 Shanghai county in Sung-chiang prefecture reported
that it had registered 1,494,775 mou of taxable land. The district's tax
liability was then assessed as 391,307 piculs of levelling grain.58 Likewise
Wu-hsien in Soochow prefecture registered 714,129 mou of taxable land in
1589, on which 157,193 piculs of levelling grain was assessed.59 The levelling
grain still differed from the tax quota in the national accounts in that
it incorporated the several kinds of surcharges and commutations on a
district-wide basis and distributed them evenly among all taxpayers. The
national accounts were still based on the items actually received at the
designated granaries, and commuted items were still counted in terms of
grain rather than their converted values.
Outwardly the collection of the levelling grain, especially after it was
substituted for the basic assessment, represented the greatest possible
simplification of the tax structure. Yet closer examination shows that the
reform did not go far enough. The levelling grain was still an abstract tax
unit, the greater part of which still remained to be converted to silver in the
late sixteenth century. From now on every taxpayer had to submit his
payment partially in grain and partially in silver, a portion of the latter
being used to cover the cost of providing cotton cloth. Furthermore, since
the levelling grain was permanently fixed at the minimum level possible,
there was no room for further adjustments. This inhibited the local govern-
ment from assuming fiscal responsibility in executing the transaction. It
merely allocated the delivery of the materials and funds to civilian agents
drafted from the local communities, thus reviving the tax captaincy. When
the inevitable shortages occurred, the delivery agents had to make up the
deficiencies personally. There is much evidence in local gazetteers that the
surcharges and transportation costs authorized by the government scarcely
covered the actual expenses, and that many delivery agents went bankrupt.60
The delivery was moreover a community service. There was no way to
prevent a delivery agent, as an official appointee, from requisitioning
materials, labor, and funds from the populace to subsidize his function.
Such details will be discussed further under the heading of tax adminis-
tration (4, i), but it should be noted here that the tax structure had not
been made sufficiently broad to cover all the costs incurred.
The collection of the levelling grain, started by Chou Sheng in the
fifteenth century, embodied the same principle as the Single Whip Reform
(3, in). The levelling grain consolidated the land tax proper, while the
Single Whip Reform extended the consolidation to the collection of
the service levy. Both combined the various kinds of tax denominations
104 The land tax—(i) Tax structure
at the pre-collection stage and allocated the revenue afterwards, and both
were subject to the same limitations. Even when local consolidation was
accomplished, lack of integration on the upper levels remained. The various
items of income, including the Gold Floral Silver, cotton cloth, glutinous
rice, the southern grain and tribute grain, had already been permanently
allocated as expended items in the national account. The lateral trans-
actions and criss-crossing supply lines continued. Local officials were
never authorized to obtain sufficient operating expenses to carry out
expanded administrative functions, as is shown by the abolition of the
sinking fund. The separate delivery system meant that these numerous
items remained irreducible components of the tax structure.

North China
In North Chihli and parts of Shantung and Honan, the tax structure was
complicated by the provision of stabling services for government horses in
lieu of land taxes, by the fifteenth-century creation of palace estates and
aristocratic estates, and by the existence of the government land and
pastures that remained thereafter.
The 'horse administration' (ma-cheng) started in the early years of the
dynasty. Both Hung-wu and Yung-lo made persistent efforts to procure
horses outside China proper. In the later fourteenth and early fifteenth
century horses were purchased from or bartered with Korea, Manchuria,
the regions along the northern frontier, Kokonor, and even Samarkand.
Official records indicate that by 1424 the Ming government was already in
possession of 1,736,618 horses and that the number of horses was still
expanding at a compound rate of 10-15 per cent annually.61 Though both
these figures seem to have been grossly exaggerated, it is clear that the
number of mounts was too large to be maintained by government agencies
and some of them had to be farmed out to civilian households for upkeep.
Hung-wu started this by farming out government horses to several pre-
fectures north of the Yangtze river. After Peking was chosen as the capital,
north China became the major area for horse-raising. Stabling services were
provided by the population in seven prefectures of North Chihli, three
prefectures of Shantung, and three prefectures and one county of Honan. 62
Up to 1568 all the horses distributed to civilian households were used
for breeding. Their total number was variously reported in the fifteenth
century as 100,000 or 120,000.63 Every five studs were assigned as far as
possible to a group of five households. These households had their land
taxes and other service obligations either exempted or reduced, and instead
fed the horses and provided veterinary attention at their own expense.
Every year each group delivered a horse fit for military service to Peking.
In addition the households were expected to produce a certain number of
 3, II Regional variations 105
yearlings, which they had to maintain until they were required. From the
later fifteenth century as a general rule each stud was expected to produce
two yearlings every three years. The households were held responsible for
making good any losses of horses or short-fall in the number of yearlings.64
From 1466 the provision of horses for military service was gradually
commuted to silver payments.65 In 1468 the court of the imperial stud con-
structed a silver vault known as the Ch'ang-ying Treasury to handle these
funds.66 By the early sixteenth century few horses were actually delivered.
In 1528 for example a total of 25,000 horses should have been submitted,
but only 3,000 were in fact provided, silver payments being substituted for
the rest.67 The commutation rates varied widely, ranging from 12 to 30 taels
per horse, depending on the date of commutation and the administrative
district.68 At the same time horse-breeding on the local level continued.
Imperial censors were dispatched to the prefectures to brand the new horses,
to dispose of the surplus yearlings, and to supervise the collection of pay-
ments compensating for any losses.69 To simplify tax administration most
counties separated their taxable land into two categories, termed na-liang-
ti, or 'tax-paying land', and yang-ma-ti, or 'horse-maintenance land'. The
division was necessary because the horse administration operated in three-
year cycles.70 In addition to their annual payment, the horse custodians
had to guarantee the prescribed rate of fecundity (two yearlings every three
years). The accounts could not be settled until the censors handed out
their decisions. The amount of land assigned to the rearing of horses
was generally large. For instance in around 1500, of the 355,999 mou of
registered taxable land in Wan-p'ing county, North Chihli, 142,143 mou,
or about 42 per cent of the total acreage, was set aside for this purpose. 71
This situation remained unchanged for the greater part of the sixteenth
century.
In 1568 as a major measure of tax reform the court ordered that all the
studs, by then totalling 100,000, should be sold for cash. The sale was com-
pleted by 1581.72 The horse administration could have been terminated
thereafter, had it not been for the special fiscal procedure obtaining in
Peking. The court of the imperial stud, which was responsible for procuring
combat horses for the army, required a budgetary income of its own and
continued therefore to maintain the account. The annual payments from
the local districts, still based on the previous quotas for horses, continued
to be delivered to the Ch'ang-ying Treasury.73 The populace, now freed of
the burden of raising horses, no longer risked having to pay for shortages
in the triennial cycle. The horse payments, stabilized by the fixing of an
annual quota for every district, could now be re-admitted as part of the
land taxes. In these districts of north China the tax-paying land was thus
merged again with the horse-maintenance land. The annual payment to
the court of the imperial stud was apportioned among all taxpayers in the
106 The land tax—(i) Tax structure
same way as was the Gold Floral Silver in south China. All this undoubtedly
constituted a great improvement. In the tax ledgers the horse payments
were still kept separate from the regular land taxes, which meant that basic
land tax assessments in these particular districts always appeared unusually
low when compared with the rates prevailing elsewhere.
In some counties in North Chihli the stabling service continued after
1581. The horses assigned to them for upkeep were no longer studs how-
ever, but mounts purchased by the court of the imperial stud on the northern
frontier. The upkeep was short term and the magistrates appointed custo-
dians on an ad hoc basis. In 1590, both Wan-p'ing county and Hsiang-ho
county were obliged to maintain one horse for every 650 mou of taxable
land. The magistrate issued a certificate to the appointed custodian who
made monthly collections directly from neighbouring households in a local
community containing 650 mou of taxable land. Ku-an county in 1585 had
a similar arrangement, but the average acreage assigned for the main-
tenance of each horse was 430 mou.74

The emergence of palace estates (huang-chuang) and aristocratic estates

(chuang-fieri) was basically a development of the fifteenth century. The
spread of these estates reached its high point by about 1500. Most of them
were concentrated in four prefectures in North Chihli, namely Shun-t'ien,
Ho-chien, Cheng-ting and Pao-ting. Princely estates also appeared in
Shantung and Honan.75 Throughout the sixteenth century efforts were
made to check their expansion and to bring them under control. Not until
the early seventeenth century did the princely estates start to increase once
more, by which time the most affected area was Hukwang province in the
south.
These estates also originated partly from the complexities and ambi-
guities of land tenure. In order to encourage reclamation in north China,
Hung-wu in 1390, 1393, and 1395 repeatedly declared that all uncultivated
land in the three northern provinces was open to occupation by the first
comer. All landed properties developed after these dates were to be per-
petually tax-exempt.76 By thefirstquarter of thefifteenthcentury these early
decrees had already given rise to unmanageable confusion. Within the
same prefecture or county, and sometimes in the same village, some land-
owners paid taxes and others did not. Disputes and lawsuits among the
populace were frequent. Moreover in some regions there were still tracts
of land which had at one time or another been utilized by the government
as pastures. It was difficult to say whether the latter-day occupiers of these
properties were rightful owners or merely squatters.77
The creation of palace estates started in 142578 when Hung-hsi appro-
priated some of the acreage in North Chihli as crown properties. Hsiian-te
distributed some waste land, theoretically ownerless, to his senior army
 3, II Regional variations 107
officers in 1431, thus creating the earliest aristocratic estates on record.79
With the support of these two precedents there was a great scramble for
untaxed fields in the later fifteenth century. Princes, princesses, favorite
eunuchs, and emperors' relatives by marriage went so far as to 'discover'
untaxed land and then petitioned to the throne to have it granted to them
as their private estates, usually specifying the acreage and location.80 In
many cases they infringed the rights of the original occupiers and reduced
them to the status of tenants.81 In 1489 Minister of Revenue Li Ming (in
office, 1487-91) reported that in the vicinity of Peking there were 5 palace
estates totalling 1,200,000 mou, plus 332 eunuch and aristocratic estates
totalling 3,310,000 mou. 82
After Chia-ching's accession in 1521 vigorous efforts were made to
register such properties. The aristocratic estates were drastically reduced
and certain properties which had been seized from private owners were
returned. The most notorious chuang-fien holder, Chang Yen-ling, Mar-
quis of Shou-ning, was executed on the order of the emperor.83 In principle
thereafter no chuang-fien holder could keep the same estates in his family
for more than five generations. No holder was allowed to dispatch private
managers to his estates; the rent was collected and remitted to him by the
local magistrate.84
In the late sixteenth century these estates, together with the former
pastures, could be classified into the following six categories (their acreages,
locations, and incomes are included in appendix A): (a) palace estates,
(b) princely estates, (c) other aristocratic estates, (d) pastures belonging to
the capital garrison, (e) pastures belonging to the court of the imperial stud,
and (/) land assigned to the imperial stable and imperial zoo.
All these estates were excluded from the taxable acreage and their income
separated from the regular land revenue. In reality, however, the rents from
these properties bore a strong resemblance to land taxes. With few
exceptions they were administered by local magistrates. Even the renting
rates were on a par with regular land taxes. In Hsiang-ho county, North
Chihli, the land tax rate after absorbing a part of the service levy came to
0.027 taels per mou, quite close to the renting rates which stood in general
at 0.03 taels per mou.85 By the late sixteenth century most of the pastures
and zoo land had already been converted to cultivation, at similar levels of
rent. Since the aristocratic recipients of land grants, unless cultivating the
land themselves, were not permitted to live on their estates, their permanent
tenants differed little in fact from landowners. These estates imposed on
the local administrators the further problem of rent collection, and
another squad of agents had to be appointed for this purpose.
The chuang-fien assigned to Prince Te in Tung-ch'ang prefecture, Shan-
tung, covered 451,495 mou, scattered over six counties and two sub-
prefectures. Consisting of waste land reclaimed after the flooding of the
108 The land tax—(i) Tax structure
Yellow river, its rental grossed only 6,552 taels. According to the local
gazetteer, in 1600 the rents were still managed by the local magistrate.86
The problem of the ambiguity of land ownership did not end here. In
1529 the magistrate of Chi-hsien, Honan province, was said to have 'un-
covered' 1,174,046 mou of untaxed land in his district. In order 'not to
make further investigations', he simply reapportioned the district's tax
quota so as to include the new acreage.87 A similar reapportionment in
Hsin-hua county increased the taxable acreage to ten times that of the
Hung-wu period, and in Yung-ch'eng county by over sixteen times. The
reapportionment, while permitting general tax-rate reductions in these
counties, widened regional disparities. Ku Yen-wu, who was not usually
an advocate of tax increases, declared that these were cases 'where tax
quotas should have been increased but were not'. 88 In 1556 Honan pro-
vince reported that it had 14,080,975 mou of such land under its juris-
diction. The provincial gazetteer referred to it as 'land previously untaxed
but now under government control'. 89 In the early seventeenth century
Szu-shui county, Shantung, still registered some land of which the title
was in question. It was called pei-ti, or 'untitled fields'.90 Such properties,
along with the alluvial land in Hukwang, were coveted by the princes as
potential chuang-fien.

Other irregularities
There were many other irregularities in the tax structure. The diverse
standards of land measurement and classification have already been cited.
Owing to regional peculiarities, these variations sometimes led to most un-
usual situations. One such example is found in Shun-an county of Chekiang,
where forested land which yielded timber could be as profitable as rice
paddies or even more so. Because of the technical difficulties involved,
however, no adequate way of surveying the mountainous region was
ever devised. In 1558 the conventional definition of one mou of such
land was stated to be the area over which a man's shouting voice could be
heard.91 The ting was a fiscal unit for the purposes of service levy and in
most cases had nothing to do with the land tax proper. But in Yung-chou
prefecture, Hukwang, from the Southern Sung onwards the pool of ting
had paid a third of the district's land taxes, each ting being assessed 0.3
piculs of grain in basic payment. This practice, devised in order to accommo-
date local ethnic minorities in the prefecture's tax ledger, was not totally
eliminated even as late as 1571.92 In the city of Hang-chou, Chekiang, the
urban population paid a small tax based on the estimated number of rooms
in their residences; the total acreage was included in the district's land
data and the proceeds were counted towards the land taxes.93 This practice,
the origin of which can be traced to the T'ang, was still in effect in that
 3, III The service levy 109
city in 1579, but nowhere else. In Fukien province most monastic properties
had been tax-exempt from the early years of the dynasty. In the mid six-
teenth century when the state decided to tax these estates it was discovered
that the Buddhist abbots never controlled the land but only collected a
nominal rent from it. Such properties had in fact been ostensibly donated
to the temples by their owners in order to evade tax and service obligations;
in practice the original owners still sold, mortgaged, and rented the land in
question at will. When T'an Lun (1520-77) was governor of Fukien in 1564,
he suggested confiscating 60 per cent of all the monastic estates, a move
which could not in the event be carried out because every piece of land
in this category involved several interested parties and even its original
ownership was not clear. There is no evidence that the problem was ever
solved. Though the principle of making such properties liable to taxation
might be upheld in theory, its actual execution depended on the local
magistrates who had to find the most feasible formula that they could. The
extant sources thus reveal only numerous proposals and counter-proposals
towards a settlement.94

Ill THE SERVICE LEVY AND ITS PARTIAL ABSORPTION

BY THE LAND TAXES
The divisions of the service levy up to 1500
At the beginning of the dynasty, the service levy (/) was clearly separated
from the regular land taxes (fu). They were two different types of taxation,
the former imposed on the able-bodied male (ting), the latter on each mou
of land. Such a division could not be absolute, however, as it would have
been unrealistic to impose service obligations on individual households,
while ignoring completely their ability to perform them. In a predominantly
agrarian society the major criterion for determining this ability was land-
holding. It was thus quite natural that the two types of taxation in the
sixteenth century should gradually be merged.
The service levy was imposed on the local communities, the li-chia.
While it embodied the principle of progressive taxation, there was no fixed
standard for its actual allotment. From the entries of the Shih-lu, the
descriptions of the local gazetteers and many articles and monographs on
the subject, one gains the impression that there was considerable variety in
its actual administration at the local level. For instance, in most prefectures
the government requisitions for raw materials were fulfilled by the chia whose
turn it was to perform that year. Yet one entry in the Shih-lu reveals that
in Chungking prefecture, Szechuan, the // chief of the year fulfilled 30 per
cent of the order and the remaining 70 per cent was contributed by all the
households in the /f.95 In Hu-chou prefecture, Chekiang, certain labor
110 The land tax—(i) Tax structure
services were assigned to those well-to-do landowners who had sufficient
acreage to require 2.5 piculs of seed grain or more.96 On the other hand
there is plenty of evidence that in many districts the issue was settled by the
ad hoc decision of the // chief. Another entry in the Shih-lu declares that in
practice 'the poor furnished the labor while the rich supplied the funds.
All depended on what a person possessed.'97
The sources seem to agree, however, that during the early years of the
dynasty both material and labor requisitions were light. It is equally clear
that most of the requisitions from the general population were very narrow
in scope. Heavy service obligations, such as servicing the postal stations,
were as a rule required directly from the affluent landowners in the com-
munities, outside the operation of the li-chia system. Under these circum-
stances it was possible for the system to function in an informal way.
In the fifteenth century the demand for raw materials and labor services
expanded significantly and since the land tax quota was not raised, extra
government expenses at all levels had to be defrayed by li-chia contributions.
It has already been mentioned that when the government could not pay
official salaries even at the subsistence level, it permitted its officials to
conscript personal attendants from the general population, and that the
service of these attendants was then at once converted to silver payments.
Understandably it was then necessary to conscript more personal attendants
for actual service from the same li-chia communities but under a different
name. In the fifteenth century tax increases were generally accomplished by
these means.
The li-chia system had been designed to augment a somewhat crude
system of local government that was adapted to a simple agrarian society
(1, n). The idea of allowing ten peasant families in a village under the
direction of a // chief to decide among themselves who should pay what
share of the operating expenses of the state was in itself somewhat fan-
tastic. Clearly, the system could not survive unchanged at a time when the
government was becoming more sophisticated and wealthy families were
increasingly able to evade their fair share of the financial burden. A revision
was obviously needed.
The first major revision of nation-wide scope was the introduction of the
chun-yao method in 1443, a date which has been fixed by both Yamane
Yukio and Heinz Friese independently.98 The method did not replace the
li-chia system: in fact it relied on the li-chia for its operation. Its major
feature was the splitting of the previous ten-year service cycle into two
five-year cycles. Until then individual households had been called on to
fulfill the demand for raw materials and labor one year in every ten. After
the introduction of the chiln-yao method they answered service calls twice
in the same decennial period, in one year performing labor service which
was also called chun-yao, and in the other year contributing materials and
 3, III The service levy 111
delivering tax payments, this also being known as li-chia. Here li-chia
became a fiscal term, and is not to be confused with the li-chia organization
which continued to administer both kinds of service levy. For the tax-
payers the previous nine-year interval between service cycles was now
reduced to a shorter respite of four years."
In principle when the chiin-yao method was applied to a county the
magistrate published a list of all the regular labor services required from
the district and graded each item according to the weight of the burden it
imposed. The individual households in the li-chia communities were also
graded in three main categories (upper, middle and lower), each of which
was in turn subdivided in three. The classification naturally placed con-
siderable emphasis on property-holding. Those households which were
incapable of providing regular services were listed together in a separate
book called shu-wei-ts'e, literally' rats' tails'. They were expected to provide
miscellaneous supplementary services. When it came to assigning indivi-
duals to specific services, the grade of the taxpayer was supposed to match
the grade of the service assignment. In this way it was possible to check
some of the abuses perpetuated by the li chiefs and also to eliminate the
problem whereby some households were repeatedly called on to perform
services while others were completely left out.
But all this was only a general concept, and some of the practices also
followed precedents set by the Sung. In general it can only be said that the
chiin-yao method increased the frequency of service cycles, broadened the
participation of the general population and effected an official listing of
the labor services and prior classification of the service burden. It also laid
more emphasis on property-holding, increased governmental supervision
and minimized village autonomy, and as a result facilitated the commuta-
tion of services to silver payments. All these features were inter-related.
The chiin-yao method was not entirely successful, in that it was at
different times both accepted and repudiated by the Ming court. But the
local districts nevertheless out of sheer necessity carried on with the reform.
In 1488 the court finally ordered that all districts should compile the
chiin-yao records,100 by which time many prefectures had in fact already
done so.
Another service levy, the min-chuang or militia service, started early in
the 1430s,101 but was formally established as a national institution only in
1494. In that year the court directed that each // be required to provide two
or three militiamen depending the size of the district.102 Because this was
a new service and constituted only a part-time assignment, the min-chuang
became a separate item. Similarly the i-chuan, the provision of logistical
support for the postal stations (1, n), owing to its distinctive nature, was
not merged with the other labor services either.
By 1500 all districts thus had the following four kinds of service levy,
112 The land tax—(i) Tax structure
listed here in the conventional order in which they appear in the local
gazetteers:
(a) li-chia: tax collection and supplying of raw materials, including the
short-term labor service involved in making the delivery.
(b) chiin-yao: full-time and year-round labor service.
(c) i-chuan: equipping and servicing the postal stations.
(d) min-chuang: militia service.

From the chiin-yao method to Single Whip Reform

After the turn of the century further reforms were needed. The demand for
materials and labor services was still expanding, government expenses were
increasing and officialdom steadily proliferating. The postal stations were
busily engaged in providing hostel services in addition to sending on
documents. The palace personnel had also increased. The court therefore
continued the practice of adding hidden extras to the tax collection. To
take only one example: Hui-chou prefecture in South Chihli had the
reputation of being a rich district and had always been required to provide
the ministry of works with fixed quantities of varnish and f ung oil. Up to
1493 these materials had been acquired through local procurement orders
(tso-pari), that is, their costs were deductible from the district's land tax
proceeds. The next year however the prefect received orders from Peking
that henceforth the same supplies were to be treated as annual requisitions
(sui-pari), which meant that the items would have to be contributed by the
local population at no charge to the government. The total cost, 3,777 taels
of silver, was allotted by the prefect to the five counties under his juris-
diction and the county magistrates in turn passed on the quotas to the
li-chia communities.103 The grain quota which had previously been ear-
marked to pay for the order was therefore made available to pay for
further new procurement orders. The increase in material requisitions
as happened elsewhere inevitably brought about a concomitant rise in
labor requirements because the extra supplies had to be delivered, and
more receiving men had to be posted at the warehouses. Even though
after the adoption of the chiin-yao method the villagers performed two
one-year terms of government service, their tax-paying resources were
limited. Rural administration once again fell into difficulties owing to the
increased pressures on the li-chia system.
Meanwhile the use of silver in tax payment, which started in the early
fifteenth century, was slowly becoming more popular. In the sixteenth
century, owing to the influx of silver from overseas, the amount of the
previous metal in circulation greatly increased.104 It might be assumed that
in these circumstances it would be simple to abolish the service levy and
transfer the financial burden to agricultural land in a relatively short time.
 3, III The service levy 113
The measures actually adopted were, however, much more complicated
than this. The four kinds of service levy represented very different types of
income and even more ways of disbursing it. No Ming official would ever
think of amalgamating them completely, and the process of merging them
with the land tax therefore followed a definite sequence. In the following
discussion they will be treated separately, item by item.
Of the four the i-chuan was the quickest to undergo the transfer. In part
this was because the obligation of equipping the postal stations always
involved certain property qualifications. In addition each county usually
serviced only one station and the i-chuan account was therefore simpler
than the other items. Though the network of postal stations was a nation-
wide institution, the system of providing logistical support for each one was
completely decentralized, so that it took scores of years for the effects of
the reform to be felt. It should be noted that the service obligations of these
stations were virtually unlimited. As long as an official presented a travel
pass from the ministry of war he was entitled to free transportation,
lodging, food and drink, and in the sixteenth century there were few restric-
tions on the issue of such passes. Local government, with its fixed and
limited budget, was understandably reluctant to assume these financial
responsibilities even after the payments were commuted to silver. A sug-
gestion to this effect had already been made to, and approved by, the
emperor in 1490 though it is uncertain how the order was transmitted.105 It
took another recommendation from the minister of war in 1507 before
a general order for the commutation was actually issued.106 In Chang-chou
prefecture, Fukien, the transformation of i-chuan from a service levy to
a surtax on the land taxes was recorded as having been accomplished in
1520. An additional assessment of 0.12 taels of silver on each picul of grain
in basic payment provided the needed funds. The rate was high, being
approximately 25 per cent of the basic payment. The tax burden was thus
entirely transferred to agricultural land.107 Similar transfers with regard to
i-chuan took place in North Chihli in 1524, and in Chao-chou prefecture,
Kwangtung, in 1528. The expenses were likewise reapportioned to all land-
owners.108 By the second quarter of the sixteenth century the reform, while
not complete, was quite widespread. Hang-chou prefecture continued to
draft individual households to serve the postal stations until the 1550s,
causing many of the draftees to go bankrupt. When finally the district
decided to reapportion the service obligations to all taxpayers, the necessary
funds were not entirely derived from the land but came in part from the
ting collection.109
It should be noted here that in many districts the reform covered only
the i-chuan account but did not completely finance all the operations of the
postal stations. The actual expenses of many such stations could not be
restricted by a fixed budget. Their deficits were usually borne by some
114 The land tax—(i) Tax structure
of the attendants on duty, notably the warehouse receiving men, whose
service obligations were included in the chiin-yao account.110
The transformation of the min-chuang, also a relatively simple account,
took place somewhat later. It was undoubtedly speeded up by the cam-
paign against the wo-k'ou pirates in the mid sixteenth century. In this type
of fighting the military governors found that volunteers were more effective
than the drafted militiamen and most districts were subsequently instructed
to provide silver payments to pay recruits, rather than to continue supply-
ing draftees. After the campaign ended these arrangements still continued.111
Unlike the i-chuan, however, which in most cases completely lost the
characteristics of a service levy, the min-chuang, even after being commuted,
still retained some features of a poll tax. Such a case, in Shun-te county,
Kwangtung, was discussed in the opening section of this chapter. There
were some exceptional cases, such as in Ning-kuo prefecture, South Chihli,
and Hsin-hua county, Hukwang, where the whole financial burden was
transferred to land.112 But most districts in south China apportioned the
maintenance costs of the militia between agricultural land and able-bodied
males, the ting.
In many northern provinces which had not been affected by the wo-k'ou,
the min-chuang did not remain a separate account. Some districts began to
merge it with the chiin-yao after the middle of the century. Wen-shang
county, Shantung, reported that the militia in this and neighboring districts
existed only on paper.113 Chang-te prefecture in Honan paid its militiamen,
who were only occasionally called to serve, with funds derived from other
sources; there was no permanent appropriation for their maintenance.114
It was transformation of the two remaining types of service levy into land
taxes that proved to be the most difficult. The li-chia and chiin-yao had been
assessed on individual households for a great many years and were un-
doubtedly bona fide poll taxes. The li-chia service included not only
material supplies but also their delivery, which, though seemingly a labor
service, was under the Ming always inseparable from the material con-
tributions. Since the government never attempted to develop its logistical
capacity at the intermediate level, the county, prefectural and provincial
offices had neither the manpower nor the funds to take over this delivery
function. The only solution was to demand that the civilian agents handle
the supplies and be accountable for the articles at the designated depots.
Even though in the sixteenth century the court at Peking repeatedly author-
ized the commutation of the regular land taxes to silver payments, palace
supplies were rarely included in this. Two separate estimates made by Ming
ministerial officials show that around the year 1600 the annual deliveries
of such miscellaneous supplies at the palace warehouses were worth be-
tween 4 and 5 million taels of silver. Part of the supplies came from the
land taxes and part from li-chia requisitions.115 The eunuchs at the ware-
 3, III The service levy 115
houses, like the civil officials, were virtually unpaid by the government, nor
did the warehouses have any budget to cover normal operating expenses.
Customarily the eunuchs extracted from the delivery agents fees called
'cushion money' and 'courtesy money'. 116 If the supplies had been com-
muted to silver payments, the accounts would then have been controlled
by the ministries and the extra income of the palace personnel cut off.
Resistance by the eunuchs played a fundamental role in retarding the
reform of the li-chia service.
The chiin-yao service comprised a number of job assignments which were
most resistant to commutation. The measurers in the government depots
had to make the inventories of tax items and were personally responsible
for making up any shortages, sometimes even accountable deficits.117
Jailers were financially responsible for their assignments because most of
the prisoners in their custody could be freed upon paying a fine. Patrolmen
were in the same category. As late as 1590 the patrolmen on the coast at
Shanghai, South Chihli, were responsible for seizing a fixed quota of
contraband salt. In Chinese leap years, which had one extra month, these
quotas were increased by one-twelfth.118 The so-called doormen in the
various offices were actually building-superintendents of a sort, and had to
keep the buildings in good repair, paying minor maintenance costs them-
selves. In other words these assignments involved fiscal responsibilities for
which certain individual taxpayers were held accountable.
Such provisions reflect the true nature of Ming administration. Not
merely corrupt practices, they were a direct result of the overtly low level
of taxation, the small number of officials, tight budgetary control, and the
relative self-sufficiency of the various offices, all of which were of long
historic standing.
Throughout the greater part of the sixteenth century many provincial
and local officials sought ways of commuting the remaining service levy
into silver payments. A more drastic and fundamental reform proved
impossible; changes had to be introduced on a small scale and at the lower
levels of administration.
As a first step in reforming the supply procedure the operating budget
of the local government had to be widened. The equalization silver
payment introduced by P'an Chi-hsiin in Kwangtung (3, i) for instance,
authorized the county magistrate to collect twice as much as the previous
li-chia account had provided for, partly from taxable land and partly from
the ting. As the account was small, even after this 100 per cent increase it
produced less than 1,000 taels in a rich county.119 Once freed from li-chia
service, taxpayers actually benefited by the procedural change, though
nominally they paid altogether twice as much as before. Similar collections
in other provinces were termed kang-yin (summarized payment) or li-chia-
yin (community payment).120
116 The land tax—(i) Tax structure
All these payments had the following features in common: The decennial
contributions by the villagers were replaced by an annual payment. Part
of the financial burden was transferred to the land, the remainder being
borne by the ting. A uniform rate of payment was declared, eliminating the
former differentiated grading and allocation by the // chief. The collection
was expected to provide the local government with a cash income sufficient
to defray most of its operating expenses, and to permit it to purchase some
of the supplies for Peking. The delivery agents drafted from local com-
munities, while not totally eliminated, were more adequately subsidized.
Certain small consignments of supplies were either delivered by the local
government or consolidated at the prefectural level, with some official
representative being appointed to accompany the civilian agents.121 All
this was a slow process. The men who introduced these progressive changes
were as a rule officials of considerable prestige122 and their revisions of the
system were carefully timed to coincide with other contemporary develop-
ments. The suppression of the wo-k'ou provided a good opportunity for
such reforms. During the state of emergency the provincial and local
officials in the southern provinces gained some degree of fiscal autonomy,
with the result that the reform definitely gained momentum during and
after the mid sixteenth century. Nevertheless the reform was never a
thorough one. In the Yangtze delta, for instance, the delivery of large
quantities of supplies remained a communal obligation until the end of the
dynasty.
An equally patient and persistent effort was made to reform the chun-yao
which until the wo-k'ou campaign was the largest service item in many
districts. It represented the most difficult task of the entire reform move-
ment. Commencing earlier than the other procedural changes, it took
longer and in the end achieved less satisfactory results. By the beginning
of the sixteenth century most districts had already divided these labor
services into two categories. Those job assignments which entailed no
fiscal responsibilities were called yin-cKai, literally 'silver service', and
included such duties as those of sedan-chair bearers to the county magis-
trate, cooks in government schools, buglers in the prefectural government,
and so on. The commutation of such services was a simple matter. As soon
as payments were made the local government could always hire the needed
substitutes. The other category included jailers, doormen, patrolmen, and
some others, and was called li-ch'ai, literally 'labor service'. (Note the
spelling of this term, which is not to be confused with li-chia.) The term in
fact implied that thefiscalresponsibilities of the draftees were indissoluble.
Substitutes, though acceptable, had to be hired by the draftees themselves
so that they could continue to be held accountable.123 In general the
approach of the reformers was to commute the yin-cKai right away and to
reduce the U-cKai assignments, or sometimes to limit theirfiscalresponsi-
 3, III The service levy 117
bilities. Again, such reforms could not be thoroughly implemented. Even
in the districts where the commutation of labor service was quite advanced,
a number of U-cKai assignments usually still remained (6, i).
Though they converted the yin-cWai into monetary payments, the pro-
vincial officials still did not call for abolition of the li-chia system. By im-
posing the financial burden partly on the land and partly on the ting, the
local administrators still adhered to the system whereby the villagers were
liable to answer service calls once every ten years. But in addition to the ting
payment, an additional assessment was imposed on the taxable land belong-
ing to the households which took their turn of duty that year. From this
practice developed the famous shih-tuan-chin or 'ten-sectioned tapestry
method', first put into operation in certain districts of South Chihli and
later extended to Chekiang and Fukien.124
In its early stages the ten-sectioned tapestry method differed little
from the operation of the li-chia system. Every year one section (chid) of
households from each village community (//) answered the service call,
but paid the obligation in cash. Soon the magistrates discovered that while
each year the number of households on call was the same, some years the
households included more ting and thus, understandably, often owned
more land. In other years the situation was reversed, and there were
fewer ting and less land. Since the service requirements were relatively
stable, the process meant dividing a constant by a variable. When there
were more taxable units the financial burden of each was comparatively
light, but conversely, the burden on each unit might be considerably in-
creased when there were too few of them on call. To avoid these annual
fluctuations the officials then proceeded to allocate the total ten years'
labor requirements uniformly, in advance, on all the ting and the entire
taxable acreage in the district, although the ten-year service cycle was still
retained. By collecting at uniform rates, the yin-cKai account would show
a surplus in the years when more ting and more land were involved. This
surplus was then carried forward to subsidize the service account of sub-
sequent 'lean' years when the households on duty happened to be smaller
and to own less land.
It should be observed here that from this point on the li-chia organi-
zation, the ten-year service cycle, and the rotation system actually handi-
capped rather than facilitated the collection. The next logical step was to
free the chiln-yao account from all of them. It was much simpler to divide
the ten years' payment at uniform rates into ten equal parts, and assess
them on the entire ting pool and the total taxable acreage on an annual
basis. When the tax administration entered this final stage the district
could be said to be well on the way to the Single Whip Reform. The
various kinds of payment for the service levy, having all been settled at
annual, uniform rates, could now be consolidated into a single payment
118 The land tax—(i) Tax structure
for each individual taxpayer. The consolidated payment could further be
combined with the commuted portions of the regular land taxes. The tax
grain not so commuted and any remaining U-cKai assignments had of
course to be omitted from this scheme.

The Single Whip Reform and its limitations

In the past half century Chinese and Japanese scholars have produced
a score of articles and monographs on the Single Whip Reform or i-fiao-
pien-fa, devoting particular attention to the question of its origins. Most
scholars now seem to agree that as a national institution it can be said to
have been initiated in 1531, the year in which the term appeared in an
official memorandum brought to the attention of the Chia-ching Emperor.125
This investigation, while satisfying our curiosity, does not substantially
add to our knowledge, because the i-fiao-pien-fa itself has not been pre-
cisely defined.
In the 1570s the statesman Yii Sheng-hsing (1545-1607), in attempting
to show that the Single Whip was a general and not a specific term, argued
that it could refer to any one or a combination of the following: (a) elim-
inating the household with property classification as a unit of taxation,
(b) paying taxes in silver, (c) reapportionment of tax surcharges and trans-
portation costs uniformly among all taxpayers, (d) combination of the
service levy with the land taxes. It was not necessary for tax reform measures
to fulfill all these conditions in order to qualify for the description i-fiao-
pien-fa™ Yii's opinion has recently been echoed by modern scholars.
Liang Fang-chung observes that the i-fiao-pien-fa 'varied in refinement
and depth', 127 while John K. Fairbank calls the Single Whip Reform 'no
more than a combination of tendencies'.128
It is in fact easier to define the Single Whip Reform in a negative rather
than a positive way. While it is difficult to prescribe a minimum level for
the operations of the Single Whip, its upper limitations can be defined. We
therefore venture to offer the following definition, based on the statements
of the above scholars.
The Single Whip represents the various efforts of Ming administrators in the
sixteenth century to attain an ideal situation in which the service levy would be
totally eliminated, and the li-chia system, in form or in spirit, cease to exist;
where any remaining poll taxes would be combined with the land taxes and the
taxpayer would discharge all his financial obligations to the state in the form of
one single, consolidated silver payment which could however be spread over the
year into several installments.
Obviously no district ever carried out the Single Whip Reform to such
an ideal level and even in its most advanced forms it only approximated
to this goal. It is outwardly perplexing that such a reform movement, after
 3, III The service levy 119
so much painstaking preparation, was never brought to final fulfillment.
The explanation of the problem however is simply that the Ming financial
structure could not accommodate such a thorough reorganization.
In the 1570s and 1590s when the Single Whip Reform reached its height,
the central government established neither a regional treasury nor a general
purchasing agency. The logistical capacity of provincial and local govern-
ment, while slightly improved, was still quite inadequate to enable it to
dispense with all the unpaid services of the populace. The budget was not
expanded in the slightest. Delivery procedures, which required a specific
revenue agency to correspond to each specific disbursing agency, remained
virtually unchanged. In 1592, for instance, the magistrate of Wan-p'ing
county, Peking, reported that he was required to make cash deliveries to
twenty-seven different depots and agencies designated by the central
government, yet the total amount involved in the transaction was less than
2,000 taels of silver.129 Under these conditions it was still necessary to rely
on the services of the taxpayer. To absolve the populace of all service
obligations would have created many logistical gaps which the existing
fiscal machinery could never have bridged.
All this means that the Single Whip Reform, though it was an import-
ant milestone in Chinese economic history and possessed many positive
features, was destined to remain limited in scope. Its abortive outcome
could have been foreseen. From this point of view, the descriptions of the
i-fiao-pien-fa by a group of mainland Chinese scholars in the 1950s are
clearly exaggerated and misleading.130 Even the pioneering work of Liang
Fang-chung, which paved the way for all subsequent studies including the
present one, is too optimistic in tone. Without removing the miscon-
ceptions derived from these earlier writings, further meaningful discussion
of the land tax structure after its absorption of the service levy would be
impossible.
In the first place the palace in Peking persistently thwarted the move
toward a total tax commutation. Aside from the remaining items of tax
grain, cotton wadding, cotton cloth, and animal fodder that were still paid
in kind, practically all districts up to the end of the dynasty still contributed
lacquer, tea, wax, metals, bow and arrows, etc. Ni Yiian-lu, who was to
become the last minister of revenue under the Ming, in 1635 argued un-
availingly for the commutation of these supplies. In 1643, only a year
before the end of the dynasty, he brought up the question again, this time
in his capacity as minister of revenue. He listed fifty-eight selected items for
commutation, but owing to eunuch opposition only eight of them were
subsequently commuted.131
Secondly the li-cttai, which entailed service by the taxpayer in person,
was reduced but not totally eliminated. Even two of the most vigorous
promoters of the Single Whip Reform, Hai Jui (1514-87) and P'ang
120 The land tax—(i) Tax structure
Shang-p'eng (fl. 1550-75) failed to dispense with the measurers drafted to
serve in the government granaries that were directly under their super-
vision.132 In the 1570s when the reform was carried out to its furthest extent
in the coastal prefectures of Chekiang, the regular conscripted patrolmen
were actually eliminated. Their quota of contraband was compensated for
by adding the amount to the chiin-yao account, which was paid by all the
taxpayers. Yet the prefect of Shao-hsing doubted whether the same pro-
cedure could be applied in the case of the salt patrols and in the end an
'exception' was made of the salt patrolmen.133
It is confusing when local gazetteers sometimes mention wages for the
conscripted H-cHai assignments. The wages listed for doormen, measurers,
etc. known as pang-chii, or 'published standard', were simply the official
rate, which provided a guideline for the draftees in hiring substitutes and
also for the local officials in dividing the financial burden of one job assign-
ment among several draftees.134 The 1566 edition of Hui-chou prefectural
gazetteer indicates that when a taxpayer was assessed 1 tael of silver in
U-cKai assignment he might end up paying 5 or 6 taels.135 In Chao-ch'ing
prefecture, Kwangtung, the actual cost to the taxpayer could be as much
as a hundred times the published standard.136 Some districts collected
a U-cKai wage from the general population, paid it to the draftees and
required them to undertake the fiscal responsibility of the job assignment.137
The collection of a H-cKai payment from all taxpayers cannot therefore
be regarded as evidence that the service obligation had actually been
commuted. In most cases it merely meant that the draftee was subsidized
by the public according to the published standard. The basic problem was
that the operating budget of the local government was so unrealistically
small, and so rigid, that it was difficult for it to function without making
some taxpayers responsible for its deficits.
Chung-mou county in Honan adopted the Single Whip Reform in 1584
but up to 1626 the district still had 127 U-cKai assignments. Some of these
labor services were performed at higher offices and were thus outside the
magistrate's control.138 Hsiang-ho county in North Chihli in 1620 still
retained 419 li-cKai assignments and its gazetteer deplored that the financial
liabilities of the draftees and their substitutes were such as to lead to the
'dissolution of their properties and their own collapse'.139
The third point to be remembered is that in north China the reform was
introduced later than in the southern provinces. When officials in the south
applied the procedures already in effect in their home districts to the north,
it caused uproar. The major issue was that of hu-yin, or 'household pay-
ments'. In the sixteenth century most northern districts, especially Shan-
tung, had imposed this in addition to the ting. The household payments
were also divided into three major grades and nine sub-grades, the highest
possible rate being 40 taels of silver per year.140 It is possible that the
 3, III The service levy 121
higher the rate of household payment the less revenue would have to be
raised from the ting; and the lower-grade ting would be assessed only
a token payment. The application of the Single Whip Reform would have
eliminated the hu-yin, which many local leaders defended as an instrument
of progressive taxation. The advocates of the Single Whip Reform in their
turn asserted that progressive taxation was never really achieved because
most of the households which should have fallen into the upper tax-bracket
were in any case tax-exempt.141 In the 1570s when the controversy was
brought to Peking the court was unable to arbitrate (7, in). Later, when the
Single Whip Reform was carried out in many of these districts, the house-
hold payment was generally dropped, but the differentiated ting payment
was retained, a compromise formula which in 1590 received the emperor's
approval.142 By about 1600 Lu-ch'eng county in Shansi, Huai-jou and
Hsiang-ho counties in North Chihli, and Ts'ao-hsien and Tsou-hsien in
Shantung had all declared their adoption of the Single Whip Reform, yet
all of them still retained their graded ting payments.143 A few districts even
stubbornly retained the household payments. The implication here was
that the household grading had to be settled in the villages; by thus assert-
ing the principle of tax apportionment by the community they omitted one
of the main features of the reform, which was the standardization and
integration of fiscal procedures.
A fourth limitation was that the Single Whip Reform only revised the
procedure of tax collection but by no means simplified the basic tax
structure. On the contrary, it complicated it. The tax structure, as illus-
trated by the case of Shun-te, still comprised a great variety of components.
The utmost the local government could do was to create a fiao-pien-yin or
combined silver payment which consolidated the various kinds of payments
into two categories, one applicable to land, the other to the ting. Because
the numerous tax denominations were still retained, this measure did not
simplify bookkeeping and may well conceivably have added to the volume
of office work.
In some districts, owing to technical difficulties, it was impossible to
arrive at even these consolidated rates.144 The local government then merely
totalled up the separate payments for each individual taxpayer; in effect it
provided no more than an accounting service, though it did prevent dis-
torted interpretations of tax regulations by village leaders. An example is
to be found in the tax bill issued by the magistrate of Ku'ai-chi county,
Chekiang, in the early Lung-ch'ing period (1567-72) which is cited by
Liang Fang-chung as an instance of the Single Whip Reform. The bill,
probably dated 1568, and preserved in a hand-copied rare book in the
Harvard-Yenching Library, indicates that a typical taxpayer in this county
was assessed fourteen different kinds of tax payment.145
Though it is difficult to say just when the reform came to a halt, it is
122 The land tax—(i) Tax structure
probable that in the south, especially in the seaboard provinces, the con-
solidation had already reached its limits by 1590. The provincial officials
had now done all that was in their power to do and could go no further.
In north China the movement was probably carried on for another fifteen
years. Shansi province, for instance, declared its adoption of the Single
Whip Reform only in 1588.146 Tax reform virtually stopped in the early
seventeenth century because of increased military expenditures and other
governmental needs. This does not mean however that the portions of the
service levy which had been absorbed into the land taxes reverted to the
previous system. Rather, the need for additional taxation unavoidably led
to further forms of communal contribution. This meant that the system
evaded budgetary control, revived village apportionment and once more
gave priority to services performed by the taxpayers in person. The Ming-
shih summarizes the situation thus: 'Less than a score of years after the
general adoption of the Single Whip method, its guiding rules and principles
were thrown into confusion. The li-chia services and the tax captaincy,
though nominally abolished, in fact still existed.'147

Methods of absorbing the service levy into the land taxes

To merge a part of the service levy with the land taxes the officials generally
adopted one of three methods. One much used in Chekiang, Fukien and
Kwangtung was to take as the taxable unit the number of piculs of grain in
the regular land tax assessment and add a surcharge to it, similar to the
surtax on personal income tax currently in effect in the United States. The
second method was to assess the service levy payment directly on each mou
of taxable land, creating in effect a general rate increase. This practice was
widely adopted in north China. The third method was a combination of
the above two. As practiced in Soochow and Sung-chiang prefectures in the
Yangtze delta, it meant that the districts collected the li-chia and chiin-yao
portions of the service levy from each mou of taxable land, while the
i-chuan and min-chuang were derived from a surtax on each picul of level-
ling grain. In section I of this chapter it was observed that Shun-te county,
Kwangtung, also adopted a slightly modified form of this mixed approach.
The 'piggy back' or surtax approach was the most direct. Its major
drawback was that the computation had to start on an infinitesimally small
scale, as has already been explained in the case of Shun-te. When in the late
sixteenth century Chia-ting county, South Chihli, adopted this method its
chiin-yao account showed that 0.0147445814487 taels of silver had to be
added to every picul of levelling grain in basic assessment.148 The thirteen
digits after the decimal point were only a beginning; they had yet to be
multiplied by the fractional piculs in the levelling grain assessments of the
individual taxpayers.
 3, III The service levy 123
The assessment of the service levy directly on acreage was more popular
in north China because the land grading within each district was more
uniform. Otherwise, this type of assessment would have been just as
complex as that of the basic payment, when every category of land, in-
cluding rice paddies, dry fields, marshy land and hilly land, had to be
assessed separately. Those districts which chose direct apportionment of
the service levy by applying a flat rate to the mou usually ignored the pro-
ductivity of the land. This was only justifiable when the service levy account
was small and the economic injustice thus created was negligible. Even
then, this tax formula was still criticized.149
The advantages of the mixed formula are hard to understand. Possibly
the inconsistency was the result of the uncoordinated tax legislation in the
past, so that when the different kinds of service levy were gradually added
to the land, it resulted in one surtax on the basic land tax payment and
another assessed directly on the acreage. Tax exemption may also have
been a factor, in that in some districts one kind of service levy could be
exempted while another could not.
Tax exemption was always a matter of great complexity. The Ming
system allowed service levy exemptions for officials both on active service
and in retirement. It was also granted to persons who had acquired the
status of imperial university students, by examination or purchase. The
exemption table was complex in itself. The exemption rate was basically
determined by rank; but differed for civil and military personnel, and for
active and retired status. Capital officials were entitled to full exemption,
but provincial officials only half.150 Exemption was applicable to both the
ting and the taxable land. The exemption for land was not assessed by
acreage, because a mou of land meant different things in different districts.
It was based rather on the number of piculs of grain in the basic tax pay-
ment. For instance the 1545 schedule allowed an official of 3a or 3b rank
serving in Peking an exemption of twenty ting payments plus 20 piculs of
grain. This did not mean that the official could deduct 20 piculs of grain
from his basic tax payment; the deductible amount was the portion of
service levy that the local government imposed on 20 piculs of grain of
the basic land tax payment. Grand-Secretary Sheng Shih-hsing (in office,
1578-91) estimated that such exemption was generally worth 6 taels of
silver in all.151 The exemption table was published in 1531, revised in 1545
and modified in 1567. In principle, the ting exemption and the land exemp-
tion were not interchangeable. In other words no-one could claim that
because he had fewer ting in his household than his exemption allowed for,
he was therefore entitled to more land exemption. But as provincial
officials kept revising their local rates of exemption the court likewise
contradicted its own legislation by instructing them to be 'liberal' with
the exemptions. The result was that each district interpreted the imperial
124 The land tax—(i) Tax structure
regulations in its own way, and most permitted the interchanging of
exemptions.
In making the two kinds of exemption interchangeable, the local per unit
rates on the ting and on the land had to be fixed in a simple mathematical
ratio in order to facilitate their computation. This will be seen in the cases
cited below. Since the imperial regulations on tax exemption took the picul
of grain in the basic assessment as the taxable unit, this caused no problem
to those districts which took the surtax approach to the service levy. But
for those districts which had apportioned the service levy to each mou of
land, readjustments had to be made. For instance Wu-chin county in
South Chihli gave rank 3a an exemption of the service levy on 670 mou of
land instead of on the 20 piculs of grain in basic payment.152 The number
of exemptions could vary greatly in each district. Chi-chou prefecture in
Shantung exempted only 910 ting out of the district's 62,832.153 Since the
exemptions comprised no more than 1.5 per cent of the taxable total, it
made no great difference how the matter was handled. But in Shanghai
county, South Chihli, 12,789 ting were eligible for service levy exemption,
that is 17.8 per cent of the 73,623 ting registered in the district, and great
care was therefore called for.154 In addition to the officials and degree-
holders, the county's exemptions also covered a large number of salt
producing ting.155 The county's tax formula of 1586 granted exemptions
from individual payments for the li-chia and chun-yao service levies, but the
assessment of the i-chuan and min-chuang levies was totally on land, and
no exemption was allowed.156

The extent of absorption

It is at present impossible to publish complete and consistent statistics
showing the extent of tax consolidation between 1570 and 1590. Using
information from secondary sources, and from rare editions, reprints and
microfilms of Ming gazetteers, the present investigation has covered 175
counties, that is about 16 per cent of the Ming total of 1,138. The sources
seldom provide the necessary information, however. In the sixteenth cen-
tury detailed tax accounts were compiled in the form of Shih-cheng-ts9e
(actual collection records)157 which, as handbooks for local officials, are
not known ever to have been published. The gazetteers only summarize
essential information thought to be of permanent value, and thus the tax
formulas which they present are either incomplete or else subject to dis-
tortions which make any accurate estimate impossible. Yet the search is
not completely in vain, for there survive a handful of cases which permit
us a glimpse of the way in which tax consolidation was actually carried out.
In the Yangtze delta, the case of Shanghai county may be used for
illustration. In 1590 that district, after making its tax exemptions, had
 3, III The service levy 125
1,210,007 mou of taxable land (including inferior land of which the acreage
had been converted to fiscal mou) and 60,834 ting liable for service levy
assessment. For the li-chia collection the rate was 0.0042 taels of silver
per ting and 0.0021 taels per mou (note the 2:1 ratio). For the chim-yao
account, the rate was 0.0148 taels both per mou and per ting (rate 1:1).
This formula in effect allocated 94.2 per cent of the financial burden of
these two items to the land, leaving only 5.8 per cent to the ting. The county
gazetteer also disclosed that a surtax was assessed on all the district's
agricultural land without any exemptions. Called fieh-i-yin, or 'money for
subsidizing services', it produced 3,128 piculs of husked rice plus 10,010
taels of silver. This paid for the county's contributions to the militia
service, postal stations, and many other miscellaneous services.158 Taking
this factor into account, the share of the burden borne by the ting was
further reduced to less than 4 per cent and very likely close to 3 per cent,
depending upon the actual price of the husked grain.
It seems to have been common in the Yangtze delta for a high per-
centage of service levy to be absorbed by the land taxes. Ping-ti Ho pre-
sents, for instance, a set of data for the Soochow prefecture's seven counties
in 1617, showing that amounts of service levy absorbed by the land tax
ranged from 65.5 per cent in Wu-hsien to 92.6 per cent in T'ai-ts'ang sub-
prefecture.159 His sources furnished no information about the 'money for
subsidizing services', which is known to have been collected in Soochow
prefecture from 1538.160 If this payment were included, the percentage of
the levy borne by the land would have been slightly higher.
Nevertheless the high rate of absorption in the Yangtze delta was a
special case. Its larger cultivated acreage, higher productivity and resulting
higher land tax rates created larger tax accounts which made absorption to
this extent a rather simple matter, as was pointed out by contemporaries.161
The total service levy obligation (including that paid in husked rice) borne
by the taxable land in Shanghai county must have exceeded 30,000 taels
in value. But when apportioned to the district's regular land taxes, which
totalled 155,573 piculs of husked grain and 111,433 taels of silver, the
extra payment probably amounted to only 15 per cent of the basic payment
(assuming the localriceprice to be 0.55 taels per picul). No district outside
the delta area can have enjoyed a situation so favorable to tax consolidation.
Furthermore the size of the service levy accounts in the delta area can be
misleading, as its prefectures had far more service obligations than actually
appeared in the official accounts. An example is the cost of transporting
white grain and cotton cloth to Peking, part of which was still borne by the
delivery agents and not subsidized by the service levy (3, n).

In the southeast provinces outside the delta area, there were a number of
districts in which service levy accounts were significantly large in proportion
126 The land tax—(i) Tax structure
TABLE 2. Service levy collection in Chang-chou prefecture, Fukien, 1572

Collected from the ting Collected from land Total

Service levy collection
account tls. tls. 0/
/o (tls.)
/o
Li-chia 3,004 43.4 3,910 56.6 6,914
Chiin-yao 11,786 53.5 10,255 46.5 22,041
I-chuan 0 15,955 100.0 15,955
Min-chuang 14,335 53.5 12,384 46.5 26,719
Total 29,125 40.6 42,504 59.4 71,629
service levy

to their land taxes. Chang-chou prefecture in Fukien was one of these. In

1572 the total proceeds from the land taxes of the prefecture's ten counties
had an estimated value of only 56,262 taels of silver.* On average each
county's share was less than 3 per cent of that of Shanghai. But the pre-
fecture's service levy accounts totalled 71,599 taels of silver.
To transfer the service burden to taxable land, the prefecture adopted the
surtax method, in other words taking the picul of grain in the basic land
tax assessment as the fiscal unit. For the li-chia collection, the rates on the
ting and on the picul were imposed in a 2:3 ratio whereas for both the
chiin-yao and min-chuang accounts the rates on the two kinds of taxable
unit were the same. The i-chuan was assessed only on the land, not on the
ting. The distribution according to this tax formula can be calculated as
shown in Table 2.162
The tax formula was generally described as ting-ssu-fien-liu, meaning
that 40 per cent of the burden of service levy was borne by the ting and
60 per cent by the land and the actual apportionment approximated to
this. It should be noted that the 42,504 taels of silver imposed as a surtax on
the 56,262 taels of the regular land taxes meant a tax increase of 75.5 per
cent.
It was not unusual for a county's service levy to come close to or to
exceed its regular land taxes. Though it was rare in north China there were
many such cases in the southeastern provinces. One contributory factor
was the expansion of the min-chuang account (which included military
supplies, 3, iv) as a result of the wo-k'ou campaigns. In addition, for one
reason or another the regular land taxes in certain districts had been fixed
at very low levels in the early years of the dynasty, but the operating ex-
penses of the local government could not be kept at proportionally low
rates. The expenses of higher offices which were to be provided by the local
districts were moreover requisitioned from them according to household
* Part of the regular land taxes in the prefecture was still paid in kind. It has here been
converted to silver at 0.5 taels per picul, which was the general conversion rate in the
prefecture.
 3, III The service levy 127
TABLE 3. Service levy distribution in six counties in the southeastern provinces,
1572-1621™
Land tax
increase
Service levy Estimated after
Total
1 vJlcLl VdlUv Ul absorption
(a) borne by (b) borne by service regular of service
District the, ting the land levy land levy
(county and (c = a + b ) taxes (d) (e - b/d)
Year province) tls. % tls. % (tls.) (tls.) (%)
1572 Chang-p'ing, Fukien 2,189 32.4 4,558 67.6 6,747 3,185 143.1
1582 K'ai-hua, Chekiang 3,457 27.3 9,191 72.7 12,648 9,808 93.7
1585 Shun-te, Kwantung 5,304 31.9 11,324 68.1 16,628 17,952 92.1
1612 Sui-an, Chekiang 2,218 29.0 5,427 71.0 7,645 9,528 58.6
1621 Hsi-hsien, S. Chihli 7,657 32.7 16,012 67.3 23,669 24,940 64.2
1621 Hsiu-ning, S. Chihli 7,559 26.6 13,785 73.9 21,344 16,010 86.1
All six counties 28,384 32.0 60,297 68.0 88,681 81,423 74.1

numbers rather than their land tax quotas. In Table 3 are listed six such
counties in which the service levy accounts appeared disproportionately
large in comparison with the regular land taxes.
Of these six, Shun-te has been discussed in section i of this chapter.
Chang-p'ing county was in Chang-chou prefecture, described above.
K'ai-hua county was situated in a mountainous region and had more
timber-bearing land than arable acreage, while Sui-an county specialized
in producing raw silk. Hsi-hsien and Hsiu-ning counties were located in
Hui-chou prefecture, which, as the home of many wealthy merchants, was
constantly required to send contributions of goods to Peking and thus
accumulated large service levy accounts (6, i).
It is clear that in those districts where the service levy accounts were dis-
proportionately large, absorption of about 70 per cent of these accounts
into the land taxes resulted in a significant increase in the latter, in some
cases by as much as or more than the basic payment. This situation occurred
here and there in all the southeast coastal provinces, though the data are
too fragmentary to permit any complete assessment. For example in
Fu-ning subprefecture, Fukien, after the adoption of the Single Whip
Reform in 1578, each picul of grain of the basic assessment was in general
converted to a payment of 1.37 taels of silver.164 In the early 1570s in
Lung-hsi, Nan-ching, and P'ing-ho counties, also in Fukien after the
absorption of the service levy, approximately every 0.963 piculs of grain
of the basic assessment was converted to 1.2 taels of silver.165 Since local
grain prices are unlikely to have exceeded 0.6 taels per picul, the actual
payment in all these districts seems to have been double the basic assess-
ment or even slightly more.
128 The land tax—(i) Tax structure
With tax rates at this level, the absorption of the service levy into the
land taxes could be said to have reached saturation point. This does not
mean that when in any district the actual combined tax payment came to
twice the basic assessment, the tax-paying power of the agricultural land
was automatically exhausted. Ming land tax rates after all were never fixed
at very high levels (4, in), nor with great uniformity and precision. Because
the level of taxation varied from one prefecture to another, there should
have been considerable capacity for readjustment.166 Yet the regional rates,
having been in effect in these districts for centuries, had become local
institutions, to which the population had adapted themselves along with the
terms of land tenure, leases, and rents, all wrapped together in one package.
To effect significant tax increases was therefore extremely difficult. Ad-
mittedly in tax collection '1 picul of grain' might in fact be more than
a picul, and often was, but when it involved 2 piculs popular opinion
considered it excessive. The absorption of the service levy into land taxes
was designed merely to convert one form of taxation to another; yet
to some landowners it represented an outright tax increase. It is worth
noting that at the high point of the tax reform, though efforts were
made to restrict the number of ting and to minimize their significance
(see below), no proposal was ever made to abolish the ting altogether.
The principle that the service levy was basically a poll tax was always
upheld.
In districts where the land tax quotas were not usually low, the absorp-
tion of the service levy was not such a critical matter. In the southern
provinces outside the delta area, 60 per cent of the service burden was
usually apportioned to the land. A good case study can be made of the
eight counties of the Hang-chou prefecture, Chekiang, in 1572.167 Here the
payment consolidation resulted in increasing the regular land taxes by
anything from 50 per cent to 60 per cent in four of the counties, and by
34 per cent to 47 per cent in the other four. The percentages were in general
larger in the smaller counties and vice versa. In the prefecture as a whole
the overall percentage increase was 38.1 per cent. In other words the pre-
fecture's land tax, after absorbing 60 per cent of the service levy, was
increased by nearly 40 per cent. The data are presented in Table 7 (p. 167).

It is impossible to make this kind of survey for any districts in the interior
southern provinces as the available information is either incomplete or
thoroughly confused. Even though these provinces declared their adoption
of the Single Whip Reform at an early stage, they did not carry it out as
thoroughly as in the coastal provinces. When in 1566 the governor of
Hukwang ordered all local units under his jurisdiction to proceed with the
Single Whip Reform, the emphasis was in fact only on the chiin-yao
account.168 The accounts of Yueh-chou prefecture in 1570 still showed
 3, III The service levy 129
many items of taxes in kind contributed by the li-chia communities.169 Yung-
chou prefecture in 1571 still adhered to the procedure whereby taxpayers
took their turn at performing labor service every five years, despite the fact
that the service had been largely replaced by silver payments.170
The assessment of the service levy on taxable land should not have
created serious problems in these provinces however. In general the service
levy accounts in their counties were noticeably small as compared to the
land tax quotas. Kiangsi, a prosperous province, in 1610 had a regular
land tax quota of 2,616,889 piculs of grain, whereas its service levy accounts
are said to have totalled 687,660 taels of silver.171 The latter could not have
been worth more than half of the former.* The 1582 accounts of P'eng-ts'e
county, Kiangsi, show that its service levy was 6,502 taels of silver as
opposed to the 13,277 piculs of grain of the district's regular land tax
quota. 172 The service account was quite high in the interior provinces, but
the ratio of service levy to basic land tax never reached the levels shown in
Table 3.

In north China the service levy accounts were usually noticeably smaller
than in the southern provinces because the northern counties as a whole
delivered far less raw materials to Peking. The demand for militia service
was never so urgent as in the regions affected by the wo-k'ou campaigns.
Probably the most significant reason was that these provinces, in particular
North Chihli, retained many unlisted labor services that were still per-
formed by the populace.173
It is rarely possible to reconstruct in detail the allocation of the service
levy to the land taxes in the northern provinces. Some counties published
the total numbers of ting but omitted the payments on individual ting.
Others specified the various grades of ting payment but omitted the total
number of ting in each category. Each county for convenience adopted one
set of figures as the basis of its permanent tax legislation. Any other figures
were regarded as variables to be periodically adjusted when necessary.
Most local gazetteers therefore only printed the permanent data and
omitted the variables. In the following three cases however, all dating from
the early seventeenth century, both the ting numbers and the per unit rates
are clearly enumerated in the sources, and the regular land taxes, including
the surcharges, have been converted to silver. The distribution of the service
levy is computed in Table 4.174
In general in the northern provinces it can be estimated that the land
taxes absorbed 45 per cent to 50 per cent of the service levy. This sample
is obviously extremely small but the service levy in the northern provinces
* This is only a general estimate. The regular land tax in the process of collection was
broken down into scores of different consignments, some commuted, some paid in
kind. The surcharges and conversion rates differed from one consignment to another.
5 HTF
 130 The land tax—(i) Tax structure
TABLE 4. Service levy distribution in three counties in north China, 1608-1620

Land tax
increase
Service levy Estimated after
Total value of absorption
(a) borne by (b) borne by service regular of service
District the ting the land levy land levy
(county and (c = a+b) taxes (d) (e = b/d)
Year province) tls. tls. o/ (tls.) (tls.) (%)
/o
1608 Fei-hsien, Shantung 3,296 55.3 2,657 44.7 5,953 16,074 16.5
1608 T'an-clreng Shantung 3,046 56.6 2,336 43.4 5,382 14,162 16.5
1620 Hsiang-ho, N. Chihli 1,387 58.7 975 41.3 2,362 8,990 10.8
All three counties 7,729 56.4 5,968 43.6 13,697 39,226 15.2

was likewise relatively small, so as to reduce the risk of error in estimating

the total land taxes after the absorption. The tax accounts of 33 counties
and subprefectures in Shansi, 21 in Shantung, 3 in Honan, and 3 in North
Chihli, reveal clearly the relative smallness of their service levy in proportion
to the regular land taxes; in most cases the former can be valued at be-
tween 25 per cent and 40 per cent of the latter. Only 5 counties had service
levy accounts slightly in excess of 50 per cent of their regular land taxes.
As far as can be ascertained only Wen-shang county in Shantung and
Ho-ch'iu county and Pao-te subprefecture in Shansi actually derived more
income from the service levy than from the regular land taxes. In Table 8
it is assumed that the 7 counties and 1 subprefecture of Fen-chou, Shansi,
allowed 50 per cent of the service levy to be collected as a surtax on the land
taxes; the tax increases resulting from the absorption fall into a rather
narrow range, between 13 per cent (in P'ing-yao county) and 22 per cent
(in Lin-shih county).175 Even if the assumed extent of the absorption were
to err by 20 per cent (meaning that anything between 30 per cent and
70 per cent of the service obligation could have been transferred to the
land), the estimated final tax payment would still involve a margin of error
of only about 7 per cent, which may be regarded as permissible. Table 8 is
presented in the following chapter because it also deals with the final rates
per mou (p. 169).
The consequences of tax consolidation
It is difficult to generalize about the Single Whip Reform, consisting as it
did of a great amount of complex and uncoordinated tax legislation. In the
1950s a group of mainland Chinese historians maintained unreservedly
that it was beneficial to the poor and detrimental to the interests of the
rich, and that anyone who criticised the reform must be a ' corrupt reaction-
ary who spoke from the position of the landlord class'. 176 Such dogmatic
arguments are more emotional than rational. It cannot be denied though
 3, III The service levy 131
that the Single Whip Reform had many positive features. The subscription
of informal tax allocation within the villages by uniform published rates
was a great improvement. The elimination of taxes in kind and of services
personally performed by the taxpayers was indeed a step, albeit a restricted
one, toward modernization of the tax structure. The reform at least clarified
local tax accounts, but even so it is hard to see how it could unquestionably
qualify as an act of economic justice.
The above historians based their arguments on the oversimplified notion
that prior to the reform the service levy had been assessed entirely on the
ting and that the Single Whip transferred the major portion of this tax
burden to landowners. They did not realize that the ting assessment was
never a simple matter of counting heads but always involved some degree
of property qualification. This principle of progressive taxation can be
traced back to the middle of the T'ang dynasty and it was in fact the
i-fiao-pien-fa which terminated it as D. C. Twitchett has pointed out.177
By the mid sixteenth century progressive taxation could no longer be
maintained. Large landowners either divided their properties into smaller
plots for which they secured separate registrations, or else by claiming
exemption avoided the service obligation altogether. Heavy service obli-
gations usually fell on men of lesser means. This term however in con-
temporary writings refers to medium-grade landowners, that is, house-
holds who owned roughly 100 mou of land or more, and not to the lowest
income group. Heavy service obligations usually entailed considerable
fiscal responsibilities and it would have been unwise of the officials to
impose them on draftees who fell below this level, and would therefore
have no hope of discharging them at all.
In that it relieved the difficulties and uncertainties of these lesser land-
owners, the reform was a good one. From the administrative point of view
it was also a practical solution to the problem. Yet it should be noted that
the reform did not transfer most of thefinancialburden of the service levy
to the large landowners but rather spread it over the majority of the tax-
payers, including small-holders who might possess only five mou of land.
The intention was that the allocation of the service levy to the entire
population should result in a set of uniform rates so low that the wealthy
households would find it hardly worthwhile to evade them, and the really
poor would not be greatly harmed by them. The editor of a Shantung local
gazetteer, in advocating the reform, declared that the annual payment
could be met 'by taxpayers of meager means with one day's wages'.178
Even the reformers did not cite general economic justice as one of the
major virtues of the Single Whip.
In practice the impact of the reform varied from one region to another,
depending on local conditions and the detailed legislation of each county.
Some counties, unwilling to abandon the principle of progressive taxation
5-2
132 The land tax—(i) Tax structure
entirely, proceeded to modify the uniform rates. The creation of abstract
ting in Shun-te county (3, i) was one example and the retention of graded
ting in many northern districts was another. In 1607 Fei-hsien, in Shantung,
collected 0.13 taels of silver each from 23,775 of the county's ting, but the
remaining 4,320 ting who owned no land whatsoever were instructed to
pay at one-third of the uniform rate.179 In 1606 the magistrate of Ts'ao-
hsien, also in Shantung, insisted on maintaining a uniform payment of
1 ting on every forty mou of taxable land.180 The freezing of a district's ting
quota at a permanent level apparently derived from the practice of pro-
portional assessment, by which it was attempted to retain some property
qualification with regard to the ting. Thefiscalaccounts of Huai-jou county,
North Chihli, by 1604 already showed fractions of a ting, which were
certainly not arrived at through counting heads.181 All these compromise
devices were exceptional, and contrary to the spirit of the Single Whip
Reform, which was on the whole aimed at universality and uniformity
rather than simple economic justice.
The motives of the reform's opponents seem to have been mixed. Defense
of the gentry's economic privileges was very likely one of them; Ho T'ang
(1474-1543) for instance criticized the reform on the grounds that it harmed
the rich.182 A very different argument was put forward by Ko Shou-li
(1505-78, minister of revenue for six months in 1567) who idealistically
defended the principle of progressive taxation in the interests of the
marginal landowners. It must be emphasized that when the Single Whip
Reform was introduced in north China, it was not limited to changes in the
service levy but also involved reapportionment of the regular land taxes.
The levelling of the differential mou conversion rates and surcharge
collections had been achieved in the southern provinces as much as a cen-
tury earlier, but was still new in the north. Ko's evidence shows that the
reform was carried out there with little preparation or prior investigation.183
It is therefore not difficult to imagine how much distress it caused to poor
taxpayers who possessed inferior land and had considerable numbers of
male adults in their households. The modifications introduced to try and
remedy this situation constitute additional evidence that the reform was
not always equitable.
The entire reform could in fact be described as a modification designed
to remedy an unsatisfactory situation. It was based on a concept of taxa-
tion alien to thefiscalstructure on which it was superimposed. When these
underlying discrepancies were allowed to persist it meant that the declared
uniform rates created equality only in the mathematical sense. Though
ostensibly the basis of taxation was widened in reality its scope was narrowed
because the level of taxation was thereafter increasingly restricted by the
capacity of those taxpayers who were least able to pay. In other words the
strength of the chain was determined by that of its weakest link.
 3, III The service levy 133
The reform's lack of thoroughness has already been mentioned. By the
late sixteenth century, the service levy had grown to sizeable proportions.
Even from the incomplete information available it can be shown that the
listed items were rarely worth less than 3,000 taels of silver in any one
county. In the southeastern provinces, with few exceptions, the service levy
account was valued at a minimum of 7,000 taels, and not infrequently ran to
over 20,000 taels. Even so these collections did not cover all the expenses they
were actually supposed to, particularly those of local administration and
tax collection. This will be discussed in connection with tax administration
in the following chapter.
The tax consolidation furthermore resulted in legislation in which the
conflicting needs of imperial uniformity and regional autonomy were
curiously combined. Despite the fact that the diverse methods adopted and
the numerous tax formulas resulting from them did officially become law,
all counties still meticulously preserved the standard tax denominations.
The general scheme applicable to all regions had been created by the central
government on the basis of the proclamations of the dynastic founder and
of later decrees. No item of taxable income could be eliminated from it. In
principle each had itsfixedquota, established either by explicit orders from
Peking or through customary usage. The local officials merely had to work
out the detailed procedures for collection. Though the various items could
be consolidated in actual collection, in bookkeeping they had to be listed
separately. Tax regulations, outwardly uniform, were thus extremely
diversified in practice, a finding that is confirmed by the way in which the
northern counties treated the ting payments.
The restricting effects of centralized control on local officials were tre-
mendous, in view of their struggles with decimal digits. Needless to say the
tax structure all too often sacrificed rationality for the sake of outward
conformity. The accounts of Shanghai county in 1590 (above, p. 125) showed
that the district's ting payment was collected from 60,834 ting.18i A more
rational approach would have been to waive the payment entirely and
transfer the financial burden to the district's 1.2 million mou of taxable
land, which would have entailed an increase of only 0.001 tael, or less than
1 copper cash per mou. Yet in order to maintain the principle of a poll tax,
the collection of the ting payment continued in this county as elsewhere,
apparently at the cost of a large volume of paper work and considerable
administrative expense.
The fact that the Single Whip Reform required all miscellaneous tax
items to be clearly specified and committed to writing, while it undeniably
curbed abuses by rural tax agents, nonetheless at the same time created
many opportunities for lesser yamen functionaries to manipulate the
accounts in the office. The tax system had simply accumulated too many
complexities which might have been avoided but were not.
134 The land tax—(i) Tax structure

IV FURTHER MODIFICATIONS OF THE TAX STRUCTURE

Military supplies
All the tax revenues discussed in this chapter, including those in the present
section, were permanent items. Once they appeared in the tax ledgers, they
were expected to remain there. While minor readjustments were made
from time to time, the general level of collection was considered to be fixed.
The only exception to this was the item known asping-hsiang, or 'military
supplies'.
The 'military supplies' owed their inception to the wo-k'ou campaign.
The pirates' marauding raids in the south-eastern provinces exposed the
total inadequacy of the wei-so system (2, in), and volunteers had to be
recruited to supplement the guard units. It was necessary to organize new
fighting units and to procure equipment for them, and to hire or hastily
construct ships, all of which needed money. In the 1550s Peking was itself
threatened by Altan's invasion and the treasury in the capital had no funds
to spare. The court merely authorized the military governors in the south
to solve the financial problem as best they could.
The emergency left little time for deliberation. The solution was the
ping-hsiang, which took the form of a universal surcharge on all existing
tax revenues from which extra collection was feasible, and could be made
in a hurry. In some special cases portions of existing quota revenues were
also diverted to meet the crisis. In addition dozens of new sources of
revenue, usually minor nuisance taxes, were created for the same purpose.
All were indiscriminately labelled 'military supplies'. Strictly speaking, the
ping-hsiang was not an item of revenue but rather a general designation for
a special expenditure. Its fuller implications will be discussed later (7, II).
Here only its disruptive effects on the land tax structure will be described.
There were many methods of imposing the ping-hsiang on the land tax.
For instance hilly land in Ku'ai-chi county, Chekiang, was very lightly
taxed; during the military campaigns its rates were therefore increased.185
Fukien province took to taxing monastic properties which had previously
been exempt.186 In most districts the militia service was commuted and the
resulting payments diverted to the ping-hsiang account. Additional im-
positions were also made directly on the regular land taxes, usually in the
form of a surtax on each picul of grain in the basic assessment. The most
ingenious device for increasing taxes however was the so-called fi-pien.
There is no adequate English equivalent for this term. Of its two com-
ponents, fi means to lift up, and pien to organize. In the 1550s taxpayers in
most counties still answered calls for li-chia and chiln-yao service in rotation
at five to ten year intervals, even though the service obligations had been
largely commuted to silver payments. By the device of fi-pien the military
 3, IV Further modifications of the tax structure 135
governors called those households who were scheduled to perform service
the following year to active duty in the current year. In reality no raw
materials or labor services were demanded of them, but their silver pay-
ments were surrendered to the war chest. The services required in the second
year were then fulfilled by those households who should have been on call
in the third year.187 The complication was that in many districts the ting
payment had been partially reapportioned on the land, and the fi-pien
payment thus represented a surtax on a surtax.
The whole procedure was extremely cumbersome, and in addition necessi-
tated several kinds of tax agents making collections in the rural areas, some
for the regular service levy account for the local government, some for the
fi-pien account for the military authorities, others for the li-chia and the
chun-yao levies, quite apart from those who collected the regular land taxes.
Ho Liang-chiin protested, with humorous exaggeration: 'When everyone
in the village turned tax collector, who was left to cultivate the land?' 188
This absurd situation was one of the reasons which prompted many local
administrators to proceed with the Single Whip Reform.
The ping-hsiang was not intended to be permanent, but was expected to
be removed as soon as the campaigns were over. Military action actually
dragged on for years and even in the late sixteenth century when the threat
from the pirates had diminished, demobilization was only partial. Because
the services of the recruited army could not be dispensed with, the ping-
hsiang therefore had to remain. In the course of time its earlier ad hoc
features were gradually eliminated and in most districts it was consolidated
into a single surtax, with about 40 per cent being paid by the ting and
60 per cent by taxable land. In this way it differed little from the min-
chuang account. Yet in theory at least the ping-hsiang was collected for
regional defense while the min-chuang continued to provide an auxiliary
force. The former was adjustable, depending on the actual expenditures of
the regional armed forces but the latter was part of the district's fixed tax
quota. In 1572 the ping-hsiang account in Chang-p'ing county, Fukien, was
reduced by half after coastal defences were lightened, while the min-chuang
account remained fixed as before.189 In accounting the two items were
almost always listed separately. Thus each taxpayer continued to con-
tribute his share of militia service as well as his share of military supplies.
In assessing the general level of collection in the previous section, how-
ever, the ping-hsiang figures have been merged with the min-chuang account
in the interests of simplicity; it would have created unnecessary compli-
cations to introduce the question of the military supplies in the midst of
describing the evolution of the service levy. In effect the item was a two-
layer surtax on the regular land tax, as illustrated in Figure 3 on p. 83.
136 The land tax—(i) Tax structure

Accessory surtaxes
Accessory surtaxes were really 'built-in' surtaxes, levied on taxable
acreage and paid in agricultural products, which dated from the early
years of the dynasty. Though separate from the basic assessments, they
still differed from the other types of surtax which were created subsequently
for administrative reasons. They were surprisingly few in number.
There is a general misconception that the Ming government collected
large quantities of cotton, cotton cloth, silk wadding and silk fabrics as
surtaxes on the land taxes. The misunderstanding arises from entries in the
Ming-shih and Shih-lu which indicate that the Hung-wu Emperor ordered
all the landowners in the empire to set aside part of their land for pro-
ducing hemp, cotton and silk; all land taxes were to include these products
and landowners who failed to comply were to submit cotton cloth and silk
fabrics at punitive rates. Such a decree was actually issued in 1365 (before
the founding of the Ming empire) and again in 1368.190 By 1385 however the
government realized that it could not enforce the compulsory production
of these commodities and therefore rescinded the previous orders. 191 In
the case of silk production, landowners who had already received the
specified acreage for planting mulberry trees were to be taxed in accordance
with the earlier decrees. The silk quotas of 1385 for each county sub-
sequently became permanent. This collection, known as nung-sang ssu-
chuan, or 'farmland silk', was almost ubiquitous; but each county's quota
was never too large, and might be infinitesimally small, owing to the
government's inability to enforce mulberry cultivation. Apart from this
small quota, landowners were free to plant whatever crops they wished.
Cotton was not even mentioned.
The silk and silk fabrics which appeared in large quantities in the tax
ledgers came from areas which specialized in their production, and their
regional quotas were clearly specified in the local gazetteers as basic assess-
ments. These areas submitted silk as tax payments instead of grain, just as
they had done in previous dynasties. This had nothing to do with the
farmland silk.
One such county, Sui-an, in Chekiang, had a total of 367,332 mou of
taxable land but a grain tax quota of only 1,685 piculs, while its quota of
silk wadding was 111,663 taels, or close to 10,000 pounds.192 The value of
the latter item was approximately four times that of the former. Lin-an
county in the same province had a grain quota of 11,362 piculs, while its
silk wadding quota was 110,416 taels. Lin-an in addition had to provide
its quota of farmland silk, which in 1572 was only 2,517 taels of silk
wadding. Since each tael of silk wadding was worth approximately 0.04
taels of silver, the latter item was worth only about 100 taels in cash.193
In other regions, the revenues from farmland silk were either small or
 3, IV Further modifications of the tax structure 137
negligible. In 1608 Wen-shang county, Shantung, raised 419 taels of silver
from this source.194 In 1620 Chung-mo county, Honan, thus collected 1,100
taels of silk wadding, its 33,320 mulberry trees apparently having survived
from the time of the early Hung-wu decrees.195 The market value of the
county's farmland silk was less than 50 taels of silver. The total proceeds
in 1562 from farmland silk in Hui-chou prefecture, South Chihli, were
only slightly in excess of 10 taels of silver, each of its counties providing
1 or 2 taels.196
Another surtax, also in silk, was called jen-ting ssu-chuan, or 'poll tax
silk'. It was collected only in North Chihli, Honan, and Hui-chou pre-
fecture in South Chihli, mentioned above. Its origins are unclear. Most
likely it was inherited from the previous dynasty, which had permitted
some districts to submit silk in partial discharge of their service obligations.
The gazetteer editor of Hui-chou prefecture argued at some length that the
collection was in fact the result of a misinterpretation of a tax-ruling dating
from before the founding of the dynasty. He asserted that it was an unjust
and mistaken enactment by the ministry of revenue that cost the prefecture
6,146 taels of silver a year.197
Cotton wadding and cotton cloth were never collected as surtaxes. Raw
cotton, known as ti-mou mien-hua-jung, or 'cotton wadding assessed on
acreage', was produced in certain areas of Szechuan, Shensi, Shantung and
North Chihli that specialized in its cultivation. For the purposes of basic
tax assessment, every 10 catties of cotton were equivalent to 1 picul of
grain.198 To confuse matters, however, grain quotas in other regions could
also at times be paid partially in cotton and cotton cloth. Shansi province,
for instance, had no basic assessment in these two commodities, yet it
constantly delivered them in lieu of grain payment. This constituted a kind
of commutation, and eased problems of tax transportation.
Several counties in the south did pay a surtax in hemp, but the quotas
were unevenly distributed and their effect was negligible. More hemp was
in fact derived from the proceeds of the fish duty (6, i).
The only item in the list of accessory surtaxes which significantly affected
the taxpayers was the ma-ts9ao, or 'horse fodder'. The rate was sixteen
bundles of hay per 100 mou of taxable land, the collection being limited to
Shantung, Shansi, Shensi, Honan, and North and South Chihli.199 At the
official rate sixteen bundles of hay were worth approximately 1 picul of
husked rice in south China during the late sixteenth century.200 The burden
of this surtax was heavier than it appears however, because of the difficulties
of transporting such a bulky commodity. As a percentage of the total it was
fairly appreciable because the basic assessment was usually low; in certain
counties this surtax might in fact add an extra 20 per cent to the basic
payment.
These accessory surtaxes were of importance to the ministry of revenue
138 The land tax—(i) Tax structure
because even though unevenly distributed and raised in tiny amounts, they
constituted increasingly indispensable items of ready cash (7, i). Commuted
ma-ts'ao in fact yielded more silver to the T'ai-ts'ang Treasury than did the
total proceeds from the inland customs duties.

The absorption of miscellaneous and uncollectible items

All the items mentioned here will be discussed again in chapter 6, to which
they really belong. It is necessary to refer to them at this point however
because in practice many counties incorporated them into the land taxes.
One reason for this was that they were very similar in nature to those items
which had already been incorporated into the land taxes. Also they were
so small that it was impractical to collect them separately.
Their absorption was neither complete nor uniform, but those districts
which carried it out simply appended these items to the land tax accounts
as if the merging of the two had been properly authorized.201
Among these items were payments to the ministry of works, of two kinds.
One was known as ssu-ssu liao-chia, or 'material supplies to the four
bureaus', that is, the four branches of the ministry. Up to the mid sixteenth
century the four bureaus of the ministry occasionally requisitioned funds
from the counties to help cover their operating expenses. In 1566 it was
decided to impose an annual payment of 500,000 taels of silver on the
counties in perpetuity, to provide these offices with a fixed budget to
finance their administration.202 Though this item might be considered
a variant of the li-chia collection, its imposition did not relieve the counties
of the responsibility of supplying raw materials to the ministry. The other
type of payment was termed chiang-yin, or 'artisan payment'. Local
registered artisans were initially required to perform unpaid services at the
capital. In 1562 this statute labor was totally commuted and the ministry
instead ordered the magistrates to deliver the payment in a lump sum.203
It might thus be considered as a variant of the chiin-yao levy. Some
counties and prefectures were still able to collect the payments from house-
holds in the artisan category, but most chose to add them to the service
levy accounts. With the adoption of the Single Whip Reform, these pay-
ments also became a kind of two-stage surtax in that while they were im-
posed on the service levy, the latter was in its turn partially imposed on the
land tax.
One rather curious item was the payment for rationed salt, or hu-k'ou
shih-yen-chiao,firstlevied in 1404. Its original purpose was not to mono-
polize salt-retailing but to keep the government's paper currency in
circulation. Each adult was allowed a ration of 1 catty of salt per month,
the compulsory payment for which was 1 kuan of paper notes. For children
the ration and the payment were half that.204 From the very beginning
 3, IV Further modifications of the tax structure 139
the plan was never carried out in full. Up to the late fifteenth century the
actual distribution of salt under government auspices was effective only in
certain parts of Kwangtung and Shansi, and thereafter even this came to
a complete halt. In the early sixteenth century some counties still made the
payment in paper currency, which was by then virtually worthless. In the
late sixteenth century, however, it was everywhere converted into silver,
and became a poll tax at very minimal rates. Some counties collected it at
0.018 taels per capita per annum, others at 0.0028 taels.205 It cannot be
considered a regular item of the service levy because it was imposed not on
the ting, but on the entire population, including women and children.
Shun-te county, Kwangtung, for instance, converted the collection to
a poll tax in 1584. While the county registered 26,011 ting, the tax was
imposed on 41,656 persons, yet its total annual proceeds were only 717
taels of silver. Another strange feature of the payment was that relatives of
government officials were not only unexempt but were required to pay
at double the rate.206 Many other counties found the amounts involved
too small to warrant this cumbersome procedure. For example Liao-ch'eng
county, Shantung, was required to produce only 66 taels of silver from
this source, and therefore covered the item from the land tax rather than
by making a special collection.207
Other revenues similarly absorbed by the land taxes included store
franchise fees, the excise on wine and vinegar, the stamp tax on real estate
transfers, the fish duty, and even some of the local business taxes (see
chapter 6). Most of those items had been created under the Southern
Sung, and were sometimes referred to as ching-chih-ch'ien.208 The Ming re-
established them but never took them seriously. The quotas for these items
werefixedin the early years of the dynasty in kuan of paper notes. When
the paper currency failed, the logical thing would have been either to
readjust the quotas or to abolish the items altogether. As neither of these
alternatives was adopted, the only solution was to absorb them into the
land taxes. Table 5 shows how their quotas were converted to silver.209
A reverse procedure was sometimes applied, in that several items of
revenue that should have been integrated with the land taxes were instead
accounted for under other headings. The fang-chia, literally 'marsh-land
payment', was one of these. During the early years of the dynasty the
registered saltern households were collectively allocated tracts of land in
the coastal districts. Designated as 'grass land', these reservations were to
be used to produce fuel for salt production. By the earlyfifteenthcentury
some of the saltern households had already moved inland, and turned the
grass land and the adjacent areas into rice paddies, on which they still
enjoyed tax exemption. They thus gradually grew so affluent that the
government finally decided to tax them on their cultivated land instead of
demanding salt from them. But because both the land tax quota of each
140 The land tax—(i) Tax structure
TABLE 5. Quotas offish duty and local business tax, Yung-chou prefecture,
Hukwang province, 1571

Fish duty Local business tax

County (tls.) (tls.)

Ling-ling 32.87 Not listed

Ch'i-yang 12.83 22.00
Tung-an 5.54 3.09
Tao-chou 8.10 2.77
Ning-yuan 4.98 8.00
Yung-ming 1.78 19.54
Chiang-hua 3.57 4.97
Total 69.67 60.37

county and the salt revenue of each productive area were already fixed,
this extra tax income was not integrated with the regular land taxes but
turned over to the offices of the salt administration to compensate for the
remitted salt quota.210 In 1566 Shanghai county, South Chihli, for instance,
delivered 4,647 taels of silver to the salt administration office on this
account.211 Assessing the service levy on this group of landowners was
a perennial administrative problem.212 As salt producers they were supposed
to be exempted from service obligations but as bona fide landowners they
had to be taxed. These complications also affected the registration of
property transfers and the adjustment of tax rates. The fang-chia, found
notably in Chekiang and South Chihli, was really a variant of the land
taxes. Other such variants were the horse payment (3, n), rents on govern-
ment pastures, and the reed tax (6, in).
 4
The land tax—(ii) Tax administration

The operations of the tax system were considerably hindered by its rigidly
predetermined nature and numerous inconsistencies. This was illustrated
in the previous chapter. Despite the fact that in the late sixteenth century
its collection was put on a much broader basis, the earlier practice of
minutely-specified tax assessment continued. The general rates were com-
plicated by the addition of fractional sums; in the assessment of each
taxpayer's burden every item of state expenditure had to be considered
separately. Such a scheme would be difficult to carry out even in a modern
society, with all the benefits of computerized accounting. Local govern-
ment in the Ming was hardly equipped for the task.
Moreover, to maintain taxation on its broad and general basis, it was
necessary to protect individual taxpayers from unfair practice and extor-
tion, if their equality of treatment were not to remain purely theoretical.
The Ming government made no such attempt and consequently the tax
administration, while appearing fairly solid and well constructed at the
top, remained in a rather ramshackle state on the lower levels. As in the
rest of the power structure, responsibility in fiscal matters followed an
upward direction, from the lower offices to the higher, and from the
populace to the government. The higher offices were in no way accountable
to their subordinates, nor were they even compelled to solve their technical
problems. The impracticalities of the system thus accumulated at the
lowest level, creating a wide gap between theory and practice.
All this made it impossible to use the tax system to carry out coherent
economic policies. Its operation was instead conditioned by prevailing
social values, local gentry power, and customary usage. The administration
of the land tax reflected many conflicting characteristics of traditional
Chinese society. It was neither completely regular and consistent, nor yet
totally irrational. Tax laws might be enforced with extreme rigidity and
severity or with remarkable leniency, depending on circumstances.
Modern critics of administrative corruption in the Ming have tended to
create the impression that it resulted entirely from a gigantic conspiracy
of officialdom against the common people. This is neither true nor fair. In
[141]
142 The land tax—(ii) Tax administration
the following sections it will be seen that many Ming officials did do their
best to uphold justice in their administration, and it is in fact through
their writings that most of the abuses have come to light. Nonetheless their
efforts could hardly make up for basic organizational inadequacy.

I TAX ADMINISTRATION BY THE LOCAL GOVERNMENT

Preparation for tax collection
No county ever collected an entire year's land taxes all at once. This would
have imposed not only an intolerablefinancialburden on the taxpayers but
also a severe strain on the local government's capacity to handle the
materials and funds involved. The payment was always spread over several
installments, usually not less than four.1
In the sixteenth century tax deadlines differed from one region to another,
determined largely by local usage. But magistrates could also make minor
readjustments for the sake of administrative convenience, mainly to enable
them to meet their own deadlines in tax deliveries. A substantial part of the
payment was usually made immediately after the harvest of the major
crop, when the landowners were in the best position to raise the necessary
cash. In most counties in the south the collection started at the end of the
tenth lunar month, that is in November or early December. In certain
areas of Shansi the collection started early in the lunar year, but the first
two installments only partially covered the service levy payment, and the
substantial portions of the land taxes were collected in the spring and
autumn.2 In Shantung, payment for the service levy was collected in the
fifth lunar month, sometime in June, when wheat and barley were harvested.3
In the Yangtze delta, where the collection schedule was extremely com-
plex, each picul of levelling grain (3, n) simultaneously included silver
payment, cotton cloth, and grain for both distant and local delivery. In
Chia-ting county for instance, the collection started with an initial payment
of 0.33 taels of silver for each picul of levelling grain, to be made by the end
of autumn. Early in the lunar year the grain designated for distant delivery,
including tribute grain to Peking, was collected. This was followed by
collection of the remaining grain for local delivery, and in turn by that of
the remainder of the silver payment, a very small amount. All this had to
be completed between the third and fourth lunar month, some time in
May, so that tax collection would not extend into the busiest season of
farm work. Cotton cloth, or payment substituted for it, was collected after
the harvest.4
To prevent the boats carrying tribute grain to Peking from being delayed
by the springfloods,the court in 1564 and 1570 ordered that the loading of
grain at the ports of embarkation be completed before the end of the lunar
 4,1 Tax administration by the local government 143
year.5 This caused considerable difficulty to local officials who had to revise
their tax collection schedules accordingly.
The payments of individual taxpayers were computed from the 'actual
collection record' (Shih-cheng-ts'e) which, in theory at least, calculated the
rates of payment with a precision that enabled the collection to meet the
county's tax quota exactly.6 From 1583 all provinces and independent
prefectures compiled a 'comprehensive book of taxation and service levy'
(fu-i-cKuan-shu),1 which listed all service requirements, regular tax quotas,
surcharges, and surtaxes down to the county level, and was intended to
provide fixed annual budget over a period of ten years until its next
revision. In practice, however, the fixed budget was still no more than
a guideline and the actual collection had to be adjusted annually.
Even though the quotas of tribute grain, southern grain, and grain for
frontier and local deliveries were more or lessfixedfor each county, any of
the items could be paid partially in silver. The amount to be commuted and
the commutation rates were determined by the higher offices, and were
sometimes subject to revision. Changes in delivery procedures and rates
inevitably affected the collection at the local level, so that the account of
every individual taxpayer had to be recalculated.
In certain counties of Chekiang the magistrates classified the taxable
acreage in some areas as being of inferior quality and commuted the pay-
ments on it permanently to silver. Only the taxable land in the centers of
these counties still made partial payment in kind and was therefore still
affected by changes in commutation rates.8
Local procurement orders and extra li-chia requisitions could affect tax
allocation in the same way, because in principle every item of taxation had
to be evenly apportioned to every taxpayer and to every mou of taxable
land. Hui-chou prefecture in South Chihli was constantly burdened with
such procurement orders which, while frequent, sometimes involved only
very small payments. To save paper work the prefect paid some of the
orders in advance, using the funds derived from the commutation of
punishments to fines. Only after a dozen or so of these small orders had
accumulated did he apportion the consolidated payments to the subordinate
counties, at the same time collecting an extra amount to reimburse the
prefectural treasury for the funds advanced. This was an irregular practice
which in 1557 was discontinued on the orders of the governor. Thereafter
each procurement order had to befilledas soon as it was received and the
actual levying of the payment by the office of the prefect on the individual
taxpayers usually took two months.9
Preparation for tax collection thus necessitated an immense amount of
paper work, mainly at the county level. A regulation of 1558 stipulated that
each taxpayer receive in advance from his county magistrate a tax bill
(known as yu-?iehy or sometimes ch'ing-yu)910 the form of which was
144 The land tax—(ii) Tax administration
standardized by the ministry of revenue. Because internal tax regulations
varied from one area to another, strict observation of this standard was
impossible for many counties and it became necessary for the local officials
to design their own tax bills.11 In addition it was often found that an extra
assessment had to be made after the tax bills had already been distributed,
in which case the magistrate handed out a second notice called hsiao-p'iao
('petty bill'), explaining the nature of the payment, its authorization and
its rates.12 While an average county might have 20,000 tax-paying house-
holds, in a heavily populated county the number could exceed 100,000, yet
the revenue section in a county administration was rarely staffed by more
than half a dozen lesser functionaries. (In practice this clerical force was
augmented by irregular help; see 4, v.)
The county magistrate, occupied with such matters as settling the general
rates of payment, corresponding with higher offices about delivery pro-
cedures, drafting civilian tax agents, allocating tax consignments, co-
ordinating deadlines, and conducting tax hearings, was too busy to attend
to the technical details. The preparation of the tax bills was usually left to
the clerical workers, as a rule with little supervision. Many of these lesser
functionaries had been handling the office routine for many years; their
authority over tax apportionment was an outstanding feature of the
administration and a major source of abuse.13 Chao Yung-hsien (1535-96)
observed in about 1582 that in his native Ch'ang-shu county, South
Chihli, no taxpayer could hope to have his payment reduced, even after
part of his land was washed away byflood,until he had made a handsome
payment to a particular lesser functionary in the county government who
controlled the district's tax records, as had his forefathers for generations
before him.14
Basically local government was under-staffed, tax procedures had become
too complicated, and there was no effective way of checking the tax records
against actual conditions in the rural areas. In the late sixteenth century
many local administrators, while committed to reform, at the same time
also strove to maintain the tax accounts in their districts at a stable level.
They knew that any tax reapportionment, unless specifically designed as
a part of a reform program and carried out under close supervision, always
tended to open up an avenue for the clerical workers and rural collectors
to profit themselves at the expense of the taxpayers. One aim of the Single
Whip Reform was in fact to freeze the accounts permanently. After the
tax payments were settled in the form of the consolidated Single Whip
payment (i-fiao-pien-yiri), most local officials were averse to any further
revision of the tax records lest it disturb the settlement. While some officials
admitted that the reform did comprise many questionable features, they
still felt that it was a great improvement on conditions before the reform.15
One district declared that harm would be caused not only by any future
 4,1 Tax administration by the local government 145
rate increases but even by rate reductions.16 In the last quarter of the cen-
tury, the attempts of many counties to fix the number of ting (3, in),
reflected in part this desire for stability, as did the publication of the fu-i
ch'uan-shu and the inscribing of tax rules on stone tablets.17
As has already been observed, however, the desired stability could only
be maintained for very short periods. Freezing the tax accounts might have
been successful in areas where administration was simple and unsophisti-
cated; but in the more developed regions of central and south China where
taxation was complicated by numerous extraneous factors, the concept was
completely impractical.
Tax agents
Several early works on the Single Whip Reform have created the impression
that in the late sixteenth century the people paid their land taxes in an
orderly fashion at the county magistrate's office, sealing up their silver
payments themselves and depositing them in a wooden chest without any
interference from official supervisors.18 This account of affairs is by no
means completely untrue, and is described in several local gazetteers in-
cluding those of Ku'ai-chi county, Chekiang, and Huai-an prefecture,
South Chihli.19 The use of wooden chests also later spread to Shantung,
Shansi, Honan, and North Chihli, besides the southern provinces. The
orderly scene thus described was not the whole story however; these reports
omit all reference to the coercion, brutality and craftiness invariably
associated with tax collection in the Ming.
Coercion was in fact essential to the attainment of even a moderate level
of tax collection at this period. Some taxpayers suffered genuine hardship
in meeting their payments, while many wealthy landowners resorted to
procrastination in order to evade their obligations. It is quite likely that
100 per cent collection was never attained, for in the late sixteenth century
even 80 per cent collection was considered an achievement.20 In the year
1570 alone the uncollected amount within the commuted payments was
over 2 million taels of silver.21 Possibly this problem could have been
mitigated by allowing tax exemptions for marginal landowners, but such
a measure was quite impractical in the sixteenth century.
If the tax delinquent were a commoner he could be arrested and flogged.
Kuei Yu-kuang (1506-71) declared that many of them were flogged to
death,22 and Ku Yen-wu reported cases in Shantung and Shensi of people
committing suicide when they failed to meet the tax deadline.23 Most
officials refrained from such extreme measures, however. When Chang
Chii-cheng was in power in the 1570s tax delinquents were prosecuted with
vigor (7, m), but Chang was severely criticized by his contemporaries,
notably Wang Shih-cheng.24 One serious problem was that many chronic
tax delinquents were wealthy landowners who had purchased degrees in
146 The land tax—(ii) Tax administration
order to obtain immunity to corporal punishment and arrest by the county
magistrate. The magistrate could report them to the higher offices, but
when such cases were numerous this had no effect except to expose the
magistrate's own incompetence. A more effective remedy, often adopted in
the Western world, would have been to confiscate the property of the tax
delinquents, but this measure was eschewed as being contrary to the
traditional concept of good government. These delinquents were moreover
shrewd enough to avoid complete non-payment, but often made partial
payments accompanied by promises that the remainder would be forth-
coming.
The local officials had to rely on their own resources to handle the
problem. The best approach was usually felt to be moral exhortation. The
magistrate personally persuaded the county's top gentry to pay their taxes
on time, in the hope that the less prestigious would be induced to follow
their example. One governor in South Chihli in the 1530s is said to have
achieved his purpose by threatening to force any taxpayer who had an
overdue payment of more than 50 piculs to handle its long distance delivery
himself.25 In 1568 Sung-chiang prefecture experimented with the method
of grouping together all taxpayers of official status and demanding that
they settle their tax accounts as a group. The results of this are not recorded.
The gazetteer of a subprefecture in Shansi contains a discussion of various
ways of persuading the defaulters to pay, one of which was to deliberately
leak the information that warrants of arrest had been prepared for them.
The warrants were not actually used, in the hope that the offenders would
be frightened into paying up. 26
When tax payments were in arrears, efforts to collect them were con-
tinued for a time but after two or three years there was clearly no hope of
success. Some of the defaulters were of course actually unable to pay and,
humanitarian considerations apart, flogging or imprisoning them was of
little use. The accumulated arrears were a considerable hindrance to the
current collection and had therefore to be written off, something that be-
came quite frequent in the late Ming.27 The emperor might authorize such
a remission of uncollected taxes up to a specific date, either for a particular
district in response to a petition from its local officials, or for the entire
empire. In addition customary remissions were provided for on such
occasions as the birth of an imperial prince, the investiture of the heir
apparent, and the accession of a new emperor.
All this in fact encouraged tax delinquency. A law-abiding subject who
rushed to pay his taxes and later discovered that the payment had been
cancelled by a delayed order of remission got no refund, nor any credit
towards his next payment.28 On the other hand a taxpayer who held out
long enough could hope that imperial magnanimity would cancel his
delinquent record. Tax default in these circumstances could be contagious.
 4,1 Tax administration by the local government 147
In one district cases were reported where the taxpayers in certain local
communities collectively hired persons to befloggedby the magistrate on
their behalf so that they could keep their own payments in arrears.29
Evidently in the early days of the Single Whip Reform many of its pro-
moters had hopes of putting the entire tax-paying population under the
direct control of the civil government, eliminating the village intermediaries
entirely. Had this ambitious goal been accomplished, its effects might have
been profound. In practically every case, however, the intermediaries were
either retained or revived, for they were one source from which to make up
the county's fiscal deficit and were also indispensable as agents of tax
collection. The principle applied here was group responsibility. Though the
magistrate was helpless to deal personally with 20,000 potential tax de-
faulters, he could always make 200 drafted agents responsible for rural tax
administration and let them pass on the pressure to the villages. Wu-chin
county in South Chihli, one of the most advanced districts in experimenting
with tax reforms, started using wooden chests for personal tax delivery as
early as 1542, and was felt to have completed its Single Whip Reform by
combining the chiin-yao account with the regular land taxes in 1573. Even
so it never abolished all its tax agents and by 1603 some of those eliminated
earlier had already been reinstated.30

Distribution and function of tax agents

Taxation in the late Ming thus involved a kind of dyarchy, with the local
government handling the paper work and civilian agents carrying out the
actual operations. In Shanghai county in 1584 it was organized on the
following lines.
The county was divided intofifty-sixtax zones (ch'u), each on average
comprising about a dozen li-chia communities. Every tax zone had a general
tax expediter (tsung-ts'ui) whose responsibility it was to see that all tax-
payers under his supervision paid their taxes on time.31 Once every ten
days he reported to the magistrate, to account for the taxes in his zone and
was flogged if the payment was behind schedule. On the other days he
visited the local communities and individual tax-paying households. He
rarely handled the tax payments in person, however.
Forty-four of thefifty-sixzones had to provide tribute grain for Peking.
After the preliminary collection was completed by the local communities
the grain checker (shou-tui), conscripted from each zone, assembled the
consignments of rice, transported them to the port of embarkation and
delivered them to the representatives of the army transportation corps.
Granaries for interim storage and boats for short-distance haulage were
made available by the local government, but their repair and maintenance
rested with the grain checker who also hired and paid his own bookkeeper.
148 The land tax—(ii) Tax administration
On average each grain checker handled 1,300 to 1,400 piculs of tribute
grain and had personally to make up any arrears in payment from the local
communities. Only after he secured written receipts from the army trans-
portation corps were his responsibilities finally discharged and it was not
unusual for the army personnel to withhold the receipts to ensure satis-
faction of their own demands.32
In addition thirty-three tax deliverers (fing-chieh) were also conscripted
from the population to handle more distant deliveries and the collection of
silver payments. Eight of them were stationed with the county government
to take charge of the silver chests, and were known as chest heads (kuei-
fou) or silver heads (yin-fou).33 It must be emphasized that the personal
submission of silver payments by individual taxpayers at the county yamen
cannot be construed as 'direct collection by the government from the
people'. 34 The procedure only enabled the county administration to
exercise close control over it. The chunks and bits of silver submitted by
the taxpayers had to be re-melted and cast into the oval-shaped ingots
required for delivery, the standard differing from one kind of payment to
another. At the higher-level offices some payments had to be supplemented
by additional silver, termed 'extra drops' (ti-pu). The magistrate also
usually extracted a payment termed the 'melting charges' (huo-hao) which
will be described later.35 When the individual taxpayer delivered his silver
to the chest, in almost every case his own responsibility to the government
was then discharged. The bulk silver was subsequently transferred to the
custody of the chest head who was responsible for re-melting it, casting
ingots, and making up shortages. In many districts the chest heads were not
released even after the silver was safely stored in the county treasury.36 As
' tax deliverers', their functions were completely fulfilled only after the bulk
silver was dispatched, all the requirements of collection having been met. In
Shanghai each chest head collected about 1,000 taels of silver and sus-
tained a personal loss of rarely less than 50 taels in the course of his duty.37
Other tax deliverers besides the chest heads made personal trips to
Nanking and Peking, to transport white grain and cotton cloth (as de-
scribed in 3, n). The cloth was as a rule purchased by the tax-paying com-
munities from officially designated weavers at regulated prices.38
All the positions mentioned above, 133 in all, were filled on the orders
of the county magistrate; but the list of nominees was submitted by 'those
who owned more land, and were both prudent and just'. Each position was
if possible filled by one taxpayer, but often jointly by four or five households.
The term of service was one year but appointments were made in advance
for five years at a time and thus involved most of the medium-grade land-
owners in the district. The position of chest head, for instance, was usually
filled by a taxpayer who owned 300 to 400 mou of land or by several
landowners whose combined holding was at that level.39
 4,1 Tax administration by the local government 149
Apart from the abovementioned agents, the magistrate appointed a chief
scribe for the county (tsung-shu, or hsien-tsung). Shanghai, along with
two other counties, took it in turns to provide its superior prefecture with
a prefectural chief scribe (fu-tsung). The chief scribe, unlike all other tax
agents, had no fiscal responsibilities. His function was rather that of
general coordination, he directed the tax expediters and grain checkers,
and established the schedule for dispatching silver from the tax chests. The
authority of this position enabled its holder to gain considerable irregular
income, and it was accordingly much sought after. One passage in the
county gazetteer indicates that some persons were willing 'to spend several
hundred taels of silver' to secure the post of chief scribe.40
The comprehensive network of tax agents in Shanghai county described
above, which consisted of civilian auxiliaries operating at a level above that
of the li-chia communities, was by no means exceptional, and similar net-
works were found in the adjacent counties of Hua-t'ing and Ch'ing-p'u.41
The organization of these agents was perhaps most advanced in Wu-chin
county, Ch'ang-chou prefecture, where the grain checkers were allowed
to elect from amongst themselves tax deliverers to take the white grain to
Peking.42 Nor was the extensive utilization of civilian agents limited to
South Chihli. In the late sixteenth century Hai-yen county, Chekiang, had
161 rural tax agents.43 Shantung province, on its adoption of the Single
Whip Reform, made an attempt to dispense with the drafted agents by
recruiting merchants as substitutes, but the plan failed ignominiously, as
can be seen from the experience of Ts'ao-hsien, Tung-ao, and Wen-shang
counties, and Ch'ing-chou subprefecture.44 Tung-ch'ang prefecture adopted
the Single Whip method in 1587, and the description of its civilian tax
agents in the 1600 edition of the prefectural gazetteer shows that the
situation there was not much different from that in Shanghai.45 About the
turn of the century Wang Ying-chiao (died 1628, minister of revenue,
1621-22) reported that in North Chihli, a subprefecture could appoint as
many as 300 to 400 persons as silver chest heads to carry out tax collection.46
Hsiang-ho county had about 50 rural tax expediters above the li-chia
level.47 Although the chief scribe does not seem to have been universally
appointed, such a position is known to have existed in Wu-chin county and
Ning-kuo and Sung-chiang prefectures in South Chihli, Hang-chou pre-
fecture in Chekiang, Hsiang-ho county in North Chihli, and in Shantung
province during the period of general land survey in 1581.48
In the late sixteenth century there were remarkably few complaints
against the various tax agents with the exception of the chief scribe. The
sources seem to agree that before about 1570, tax abuses had been much
more widespread, but that thereafter considerable improvement was made.
The new tax agents are portrayed as representing predominantly the
middle-class landowners, who were the victims rather than the villains of
150 The land tax—(ii) Tax administration
the system. This situation is partially explained by the local tax reforms
carried out at this time (3, in). The elimination of many U-cKai service
assignments must have released a considerable number of landowners in
this category, who were now available to be drafted as auxiliaries in tax
collection.
Contemporary writers in fact manifested strong sympathy for the tax
agents and deplored the fact that the tax expediters were repeatedly beaten
with 'bloody bamboo sticks' through no fault of their own.49 A county
magistrate wrote in 1609:' The more the flogging, the more the tax arrears.' 50
Influential and arrogant landowners deliberately made the tax expediters
wait at their doorways. Some of them were still unable to clear their
responsibilities four or five years after taking the assignment. Similar
sympathy was expressed for the grain checkers, chest heads, and tax
deliverers, many of whom were apparently bankrupted as a result of the
unpaid assignments.51
The decrease in tax abuses was thus largely brought about by forcing
the medium-grade landowners to cover the shortages. It is difficult to
believe, however, that the general population was not adversely affected by
this. In one document, clearly dated between 1570 and 1590, tax expediters
are already called 'crafty and deceitful' and are accused of pocketing the
payments of certain taxpayers.52 Despite the use of the wooden chests, some
taxpayers continued to rely on the expediters or other intermediaries to
hand over payments for them, showing that personal influence was still
an important factor in tax collection.53 In Shanghai county as well as in
many other districts tax expediters were also authorized to collect mis-
cellaneous dues for the local government, such as contributions to local
irrigation projects.54 In the last analysis the pressure exerted from above
was absorbed by those on the lowest level, the least articulate and the least
able to resist it. Chang Chii-cheng himself said in 1576: 'The powerful and
influential households and the crafty and habitual swindlers are passed
with no question asked. [Their taxes] are instead exacted from lesser
households and poor people.'55 Clearly it would have been naive to expect
the use of the wooden chests alone to afford legal protection to the
taxpayers.

The tax payment

Whenever part of the land tax was ordered to be commuted to silver, it was
in fact paid in silver. There is no indication that at the local level the pay-
ment could be made instead in copper coins as frequently happened under
the Ch'ing. Ku Yen-wu's report of cases of suicide occurring before tax
deadlines, was actually designed to back up his argument that in the under-
developed provinces demanding tax payment in silver was unreasonable
even in the seventeenth century.56 The only place he knew of where tax
 4,1 Tax administration by the local government 151
payments could be made partially in copper coins was Te-chou prefecture
in Shantung. In the late sixteenth century silver must have been in even
shorter supply. Some administrators manifested concern over the fact that
paying taxes in silver could depress local grain prices. Ch'ung-te county in
Chekiang, for example, produced an insufficient supply of grain to meet
its needs. When the land taxes were collected after the harvest some tax-
payers, in their anxiety not to sell their rice, preferred to raise the silver for
the tax payment by depositing the grain with a pawnshop; the loan could
be paid back with interest after the household had sent its production of
silk to the market.57
It is not known for certain how accurately decimal fractions of silver
payments were settled in the actual collection. When the Ch'ing government
in 1661 prosecuted tax delinquents in the Yangtze delta area, however,
taxpayers were officially listed as being in arrears by such trifling amounts
as 0.001 or 0.0007 taels of silver.58 This would seem to indicate that the
measurement was actually carried out to four places of decimals-a
practice that had in all likelihood been started by the Ming.
The origin of the huo-hao, or melting charges, its likewise unclear. Ku
Yen-wu declared in the mid seventeenth century that he did not know
when the collection started.59 It was undoubtedly necessitated by the fact
that every higher office demanded full weight in silver payments, and the
melting down and re-casting of the metal always involved a loss of around
2 per cent. Kuei Yu-kuang mentioned that in the 1550s Soochow prefecture
collected a 3 per cent melting charge on all taxes paid in silver, still a reason-
able rate.60 When Hai Jui was magistrate of Shun-an county, Chekiang, in
1562, local custom entitled him to impose a surcharge of 5 per cent but he
reduced the rate to 2 per cent, on the grounds that the melting charges
would result in a surplus on some payments but a deficit on others.61
Throughout the sixteenth century the irregular collection of surcharges
was subject to criticism. In 1594 an imperial censor asserted that on a small
payment of 0.1 tael of silver the melting charge could be as much as 0.05
taels, an increase of 50 per cent.62 But this was probably an exceptional
case.
When chest heads were appointed it was their responsibility to meet the
losses resulting from melting and casting, and the local officials had no
legitimate reason to demand melting charges in addition to the tax revenue.
All local governments, however, derived insufficient funds from the li-chia
service levy to provide them with an adequate operating budget. Previously
they had had considerable freedom to use funds derived from the com-
mutation of punishments to fines but by the mid sixteenth century this
income was also being gradually surrendered to the central government
(6, n). This was why officials now often resorted to imposing melting
charges. In many counties though, the magistrates simply ordered the chest
152 The land tax—(ii) Tax administration
heads to pay for any extraordinary office purchases, the reception of visiting
dignitaries, and their own tours of inspection.63
Thus during the last quarter of the sixteenth century the 'melting charge'
was no longer an extra item of expenditure but became a source of income,
and officials began to supplement their salaries with funds derived from
this source. Those charges that were retained in the collecting offices of the
county government were called cKang-li (customary payment), while those
forwarded to superiors in higher offices, including the prefectural govern-
ment, were called hsien-yii (declared surplus). Neither legal nor strictly
illegal, they were accepted by officialdom as well as the general populace.64
Though the chest heads still assumed general responsibility for the funds,
this extra silver had to be extracted from the taxpayers. It was not officially
entered in their accounts, but added to the collection as a customary
practice.
There are some scattered indications that even in the early seventeenth
century the funds thus raised by local officials were not enormously large.
One provincial official in Chekiang described the prevailing practice as
follows: When silver was cleared from the tax chests, it was deposited in the
county treasury. It was not melted down and re-cast into ingots by the chest
heads until it was time for it to be dispatched. When it was released the
clerk in charge of the treasury automatically deducted 1 tael in every 50
for the magistrate.65 Ku Yen-wu writing in the mid seventeenth century
claimed that melting charges had grown to 20-30 per cent on major tax
payments and 70-80 per cent on minor ones.66

Tax delivery
Up to the mid sixteenth century the delivery of most taxes, with the notable
exception of the tribute grain, was carried out by the people. In the 1560s
and 1570s, however, more and more tax consignments were taken over by
the local government as part of the reform. On its adoption of the Single
Whip Reform in 1568 Shao-hsing prefecture announced that henceforth
tax consignments in excess of 500 taels of silver would be delivered by an
official, and consignments of between 300 and 500 taels by a lesser adminis-
trative functionary.67 In the Yangtze delta the delivery of cotton cloth was
also for a time taken over by the government. Wu-hsien in Soochow pre-
fecture for this reason developed some public land, the income from which
was earmarked to pay for the deliveries.68
Not all such reforms were successful as local government staffs were
usually too small to carry out all the deliveries. A decree of 1573 further
forbade the magistrates to dispatch auxiliary personnel such as medical
officers and army officers as delivery agents.69 The cost of the delivery was
also a problem. Even in the early seventeenth century the counties of
 4,1 Tax administration by the local government 153
Ch'ii-chou prefecture, Chekiang, still had no funds whatsoever to finance
the delivery of tax revenues to the provincial capital, a distance of over
250 miles.70 Another substantial obstacle to the reform was the widespread
corruption that prevailed in Peking.
It was an open secret that all the personnel at the receiving depots in the
capital exacted fees from the delivery agents.71 There were fewer compli-
cations with silver payments, the purity of which could be guaranteed by
the silversmith accompanying the delivery agent. Taxes in kind, on the
other hand, had to be checked in by the palace eunuchs, who were quite
capable of rejecting the entire consignment on the grounds of inferior
quality if any of them were dissatisfied with the size of their personal share
of the proceeds. At the warehouses the transaction between deliverers and
receivers was handled by professional intermediaries. Any delivery agent
who dispensed with their services ran the risk of seeing his cargo declared
sub-standard, but its acceptance was virtually guaranteed on the payment
of a suitable fee. In the late sixteenth century the most notorious inter-
mediary in tax delivery was Li Wei, Earl of Wu-ch'ing, the maternal grand-
father of the reigning Wan-li Emperor.72
In his last years in office Chang Chli-cheng managed to eliminate this
last obstacle, including the influence of Li Wei (7, in). The effect was only
temporary, and the fate of the delivery agents was always uncertain. In
1600 the docket officer of Chia-ting county was imprisoned in Peking
because the textiles brought by him were said to fall below official specifi-
cation.73 In 1621 when the governor of Shantung ordered civilian agents
to be replaced by official deliverers, one county expressed doubts as to
whether the official deliverers could survive any better than the civilian
ones had.74 Some officials actually felt that they showed themselves more
considerate of the general population by drafting civilian agents for tax
delivery. When a consignment of cotton goods was rejected in Peking, not
only was the official deliverer arrested but at the same time the thousands
of rolls of cotton cloth were returned to their district of origin, whose tax-
payers then had to send a new consignment. If the deliverer was a draftee,
he alone took the responsibility.
The reform was relatively more successful in the northern provinces such
as Shantung and Honan where the delivery distances were shorter.75
Official delivery was also more successful in handling the silver payments
than the taxes in kind. In areas like Sung-chiang prefecture in South Chihli
which had to dispatch bulky cargoes as palace supplies, the practice of
delivery by the people was virtually unchanged. Many such districts, after
taking over the handling of cotton goods from the drafted deliverers, soon
found that the delivery was too troublesome and handed the responsibility
back to the civilian draftees.76
In some coastal areas the draftees had yet another duty to perform. These
154 The land tax—(ii) Tax administration
counties were responsible for supplying the army units stationed on offshore
islands, where cliffs made access difficult, and no grain supplies were avail-
able locally. Usually the magistrates turned over a certain amount of silver
in lieu of so many piculs of grain to the drafted agents, and expected them
to solve the supply problem.77

II FACTORS AFFECTING THE GENERAL ADMINISTRATION

The grain quota and associate factors
In the late sixteenth century the imperial government could do little to
reshape the tax administration. The major problem faced by the ministry
of revenue was that it did not have a set of adequate land data, for the
cultivated acreage was never systematically surveyed throughout the entire
dynasty. The piecemeal land survey ordered by the Hung-wu Emperor
(2, n) and the national land survey ordered by Chang Chii-cheng in 1581,
which because of its complexities will be fully discussed later (7, m), were
conducted without any uniform standards and had mixed results. The data
thus obtained seem never to have been compiled at the national level and
the ministry of revenue certainly did not use them for taxation purposes.
The only land data available to the ministry in the late sixteenth century
were those dated 1578, entered in the 1587 edition of Ta-Ming Hui-tien,
and these were inaccurate.*
The absence of adequate acreage records certainly facilitated tax evasion,
yet this was not the most serious problem. The only places where it was
reported that large tracts of land remained completely untaxed were
Hukwang and certain coastal districts of Fukien,78 which could be con-
sidered as peripheral areas of the administration. In other regions un-
registered landed property was found only in isolated cases. Tax evaders
were continually dunned by conscientious local officials and though a few
violators, because of their official status, managed to defeat these efforts,
the magistrates might when frustrated publish their names along with their
land-holdings in the local gazetteers, to show that laxity in law enforcement
was by no means universal.79
A far more pressing problem was that of tax reapportionment. While the
local officials always had records of one sort or another, the central govern-

* The data in the Ta-Ming Hui-tien were reproduced from the preliminary draft of the
Wan-li Kifai-chi-lu. The latter work was compiled in 1582 and published immediately.
It is now available in microfilm, but with chapters 3 and 6 missing, at the University
of Chicago and Columbia University. As Chang Hsiieh-yen's preface indicates, the
results of the land survey of 1581 are not included. Its sources were mainly official files
and earlier publications collected from various quarters, including some personal
papers. While the work contains some useful information, its land data were acknow-
ledged even by its own editors to be inadequate. (See for instance 4/99-100 and 9/87.)
 4, II Factors affecting the general administration 155
ment lacked the overall statistics essential to any large-scale tax readjust-
ment. The regional quotas, little revised since the late fourteenth century,
therefore had to remain as the basis of current taxation.
The basic land tax assessment, in piculs of grain, varied widely from one
region to another. The three counties of Sung-chiang prefecture in South
Chihli had a combined tax quota approximately equivalent to that of the
whole of Kwangtung province, which had seventy-five counties plus one
subprefecture. In the previous chapter it was also observed that Shanghai
county alone had a tax quota three times larger than that of Chang-chou
prefecture, Fukien, which comprised ten counties. An extreme example is
provided by Tzu-yang county, in Shensi, one of the poorest regions, the
annual tax return of which was 341 piculs of grain, less than one-thousandth
of Shanghai's.80
By the sixteenth century there were of course increasing regional im-
balances in population density, agrarian productivity, and general economic
development. On the other hand regional quotas, that had been established
some 200 years earlier, obviously could neither reflect these imbalances
nor be used to readjust them. The taxation system had long since lost its
dynamic capacity to regulate the state of the economy or even to correspond
to it. Each county's land taxes became fixed and compartmentalized, and
since the local officials never stayed in their posts very long, local landlords
always had the upper hand over the tax administration.
The unsatisfactory monetary system certainly made central planning
difficult, if not impossible. Under the Sung dynasty the copper cash was
established as the universalfiscalstandard and even in the early years of the
Yuan, land taxes were at first assessed in copper coins. When taxes in kind
were needed, the required commodities were computed on the basis of the
cash assessment, as some local gazetteers indicate.81 This way the tax
accounts could easily be integrated. The Ming system, which calculated
most of the basic assessments in grain, represented a step backwards and
caused a great deal of confusion in accounting in the sixteenth century when
land taxes were largely paid in silver. Regional grain prices in silver varied
widely and were also subject to seasonalfluctuations.Some of the confusion
resulted from tax commutations in the fifteenth century which were made
for the sake of temporary convenience with no regard to grain prices. In the
sixteenth century when the multiplicity of surcharges, surtaxes, and the
service levy were partially absorbed into land taxes, it was difficult to say
which was thefiscalstandard, the tael of silver or the picul of grain. In 1578
Hukwang province was supposed to deliver a subsidy of 102,400 piculs of
grain to Kweichow, but the actual payment was made with 30,720 taels of
silver.82 In 1591 when Lin-fen county collected 48,449 piculs of autumn
grain, the collection target was actually 49,769 taels.83
In the late sixteenth century offices of various levels ordered further
156 The land tax—(ii) Tax administration
commutation of grain quotas, fixing rates themselves. In general, any
office that had authority over both the collecting agency and the disbursing
agency also had the power to issue commutation orders. For instance, only
the imperial government could order that a consignment of tribute grain be
commuted, since the recipients of the tribute grain were always under its
direct jurisdiction. But a provincial governor, who controlled both the
army units in his territory and the tax-collecting counties, could commute
the tax payments passing from the latter to the former.84 Even though
commutation rates in the late sixteenth century were generally close to local
grain prices, the proliferation of these rates, some permanent, some tem-
porary, some ordered by the central government, and others by provincial
officials, perplexed even the administrators themselves. Sometimes the
same tax consignment was divided into two parts, each with a separate
rate of commutation. In some cases the local administration offered the
taxpayers the option of paying the taxes either in grain or in silver at speci-
fied rates.85 Understandably, the central government could not maintain
a complete record of all these confusing details for some 1,200fiscalunits.
It is impossible to find any national accounts consolidating the total
amount of silver paid as land taxes.
Thus as long as the grain quota remained the only fiscal standard by
which to integrate the national accounts, the picul of grain had no absolute
value. The tael of silver did have a general value but no general accounts
were kept in it. At the same time there was still a lack of accurate information
on cultivated acreage. The court at Peking therefore had no knowledge of
the real tax-paying power of the individual districts, nor even an adequate
measurement of the current level of tax collection. This made it extremely
difficult to increase the regular tax rates. Often the increases had to be made
in the form of additional li-chia requisitions or surcharge readjustments.
These measures complicated the tax structure still further, and produced
only small and scattered incomes.

Complications caused by land tenure, land leases, and farm prices

There is virtually no documentation of land tenure in the late sixteenth
century. The Yangtze delta, regarded as the area with the greatest con-
centration of land ownership in the late Ming, has been studied by many
modern scholars, as well as contemporary writers, yet even so the exact
nature of conditions there in the sixteenth century remains to this day
a highly debatable issue. As even less is known about other regions, the
present discussion will focus mainly on this area.
Mainland Chinese scholars, reconstructing the situation from the random
notes left by contemporaries, tend to argue that in the late sixteenth and
early seventeenth century, one or two individuals in the lower Yangtze
 4, II Factors affecting the general administration 157
could own an amount of land comparable to the entire acreage of a sizeable
county. Such a situation would indeed have been appalling.* Nonetheless
thesefindingsare inconsistent with what we know about the general social
and economic conditions of the times and also with the tax regulations
published before the turn of the century. The following facts should also be
taken into consideration:
(a) In the exchange of letters between Chang Chii-cheng and the
governor of South Chihli in the 1570s it was indicated that the largest land-
owner in the delta region probably was in possession of 70,000 mou of land
(see note * below). Persons with such immense fortunes were already under
the governor's surveillance.86
(b) In Ch'ang-chou prefecture, the wealthiest landowner was said to own
20,000 or more mou of land.87
(c) In 1610, the most affluent and powerful households in Hua-t'ing
county were mentioned as being those who owned 2,000 mou of land or
more.88
* Most of these assertions are based on statements in contemporary writings which
repeat such cliches as kao-yii wan-cKing (1 million mou of rich soil) and Vien-lien
chiin-hsien (land-holding extending over more than one county and prefecture), both
are classical expressions and do not have to be taken literally.
In these studies Hsu Chieh (1503-83) is said to have owned either 240,000 mou
(Wang Fang-chung in Ming CWing She-hui hsing-Vai ti Yen-chiu, p. 132), or 400,000
mou of land (Wu Han, Hai Jui chi, p. 7). Tung Ch'i-ch'ang (1556-1637) is said to have
owned 1 million mou of land (Wang Fang-chung, p. 132). Another often-quoted
statement is taken from Chang Chii-cheng's letter to the governor of South Chihli. In
this letter Chang mentioned that a certain large landowner (or several of them) in the
Soochow-Sung-chiang region were in possession of 7 million mou of land, which
carried a basic tax assessment of 20,000 piculs of grain {Huang-Ming Ching-shih Wen-
pien, 327/12).
Hsu Chieh was indeed a large landowner. A great part of his holdings came to him
through some form of 'contract of commendation'. The actual landowners handed
their titles over to him, paid him a nominal rent, and relied on his influence in Peking
to help them evade service obligations to the government. The abovementioned
acreages, corresponding to the amount of cultivated land in a medium and a small
county, were taken from the statements of his accusers whose figures could easily
have been exaggerated. Chao-ying Fang, who has written Hsii's biography for the
Ming Biographical History Project, chooses to put the figures in parentheses. In
discussing the matter with the present writer, Professor Fang has given it as his opinion
that since the act was illegal it is impossible to ascertain the actual conditions.
Tung Ch'i-ch'ang's family was notorious in Sung-chiang prefecture. His son became
involved in unfair financial deals with persons of lower status and extorted money
from them. This precipitated a public riot during which Tung's house was burned
down. The incident became folklore and the acreage in question comes from such
tales. The story is retold by Hsueh Kuo-chen, in Ming-Ctiing chih-chi Tang-she
Yun-tung-k'ao (Shanghai, 1935), pp. 270-3.
Chang chii-cheng's statement clearly involved a technical error. In Soochow and
Sung-chiang prefectures, an acreage that was assessed with 20,000 piculs of grain in
tax payment would have been 70,000 mou, not 7 million mou. It seems that Chang
misplaced the character cKing (100 mou) for the character mou. The error has been
noted by the editors of Huang-Ming Ching-shih Wen-pien, and by Wang Fang-chung,
p. 132. Fu I-ling, however, did not notice the discrepancy: see Ming-Ch'ing Nung-
ts"un She-huei Ching-chU p. 82.
158 The land tax—(ii) Tax administration
(d) In 1611 when the magistrate of Ch'ing-p'u county made a rigorous
investigation into land ownership in his district he discovered that practi-
cally all the large estates had previously been split up into smaller units
and registered separately, involving altogether 160,088 mou of land. On
publication of the results of the inquiry, he classified these holdings into
groups ranging from 250 mou to 2,500 mou in size.89
(e) In 1636, Grand Secretary Ch'ien Shih-sheng (in office, 1633-6)
testified that in the entire delta region many affluent landowners possessed
holdings which could be counted in hundreds of mou only. Households
with holdings counted in thousands of mou comprised no more than
40 per cent of the group (which presumably included all owners of 200 mou
or more). Those who owned more than 10,000 mou were exceedingly rare.90
(/) Yeh Meng-chu, writing in the 1660s, stated that until the mid seven-
teenth century no person in Hua-t'ing, Shanghai, and Ch'ing-p'u counties
owned more than 10,000 mou of land. Thereafter larger holdings than this
did emerge.91
(g) In 1661 when the Ch'ing prosecuted gentry landowners in the delta
area for violation of tax laws, the list of offenders contained 13,517 names.92
The above facts, and many other descriptive statements in local gazetteers,
do seem to show that the degree of land concentration in the sixteenth
century, while phenomenal, has been exaggerated in recent years. There is
no evidence that any single household in the four prefectures of South
Chihli owned more than 70,000 mou of land. Owners possessing landed
properties larger than 10,000 mou were numbered merely in dozens or
scores throughout the whole region. Most large owners, however, had
holdings of between 500 and 2,000 mou. The combined holdings of those
who owned 500 mou or more could very well have exceeded 25 per cent of
the arable land in the counties. They represented a rather small segment of
the population, unlikely to be more than 1,000 households in each county,
and perhaps closer to 500 households.
Aside from those large landowners, there was also a substantial number
of medium-grade landowners whose holdings were less than 500 but more
than 100 mou. In Shanghai county alone it seems that no fewer than 1,000
of them were drafted as rural tax agents in a five-year period (4, i).
Marginal landowners were numerous. Soochow prefecture registered
597,019 tax-paying households and Ch'ang-chou prefecture 234,355 house-
holds, though it is unknown how many of these households owned no land
whatsoever.93 Circumstances nevertheless suggest that many tenant farmers
also had small pieces of land of their own, as was the case in more recent
times. Even in the 1660s when the concentration of land ownership was
further advanced, there were still many households owning a meager 3 to
5 mou.94
From an administrative point of view, the presence of a few extremely
 4, II Factors affecting the general administration 159
wealthy landowners was a fairly simple matter for they could easily be
identified and isolated. When Hai Jui was governor of South Chihli in
1570, he could demand forcibly that the largest landowner in his territory,
Hsu Chieh, reduce his land-holdings by a half. The exchange of letters
between the two, though no figures are listed, shows that the governor's
goal was at least partially realized.95 When Wang Ssu-jen, a fearless
official, was magistrate of Ch'ing-p'u county in 1610, he threatened to
confiscate the properties of those large landowners who had falsely re-
gistered their land under the names of others to evade the obligation of
serving as tax agents. The threat brought results.96
Officials found that the most difficult problem was to handle the group
immediately below the extremely wealthy, namely the men whose land-
holdings were counted in hundreds and thousands of mou. Powerful
enough to obstruct justice, they were also too numerous to be dealt with
easily. Not only did they keep their own payments in arrears, they also
often provided a protective umbrella for their friends and relatives. It was
not the few powerful landlord-officials in the locality that were the most
serious obstacle to tax administration, according to Feng Ch'i (1558-1603).
But he went on to say that 'when the common people sought the inter-
vention of the gentry, the governor supervising the Soochow district would
be in trouble'. 47
The presence of marginal landowners made rate readjustments very
difficult, especially in the case of an upward revision. Though tax rates
were generally low (4, HI), a slight increase or decrease could significantly
affect the marginal landowners who relied on the after-tax income for their
livelihood. In 1583 the tribute grain of Chia-ting county, Soochow pre-
fecture, was permanently commuted to silver payments. The tax benefit to
the landowners can be estimated to have been worth, on paper, less than
0.05 taels of silver per mou. Since local handling of the grain consignment
was also eliminated, the actual benefit was slightly more. Such a minor
difference could cause land prices in the district to rise and as a result land-
owners who had previously sold their land rushed to 'redeem' their pro-
perties. Disputes ensued and there were numerous lawsuits. Some marginal
landowners were undoubtedly involved, since the sources indicate that
some of the buyers were illiterates.98 It is not difficult to imagine the social
consequences of a tax-rate increase of a similar magnitude.
Another difficulty connected with marginal landowners was that they
had little capacity to endure fluctuations in farm prices. T'ang Shun-chih
(1507-60), writing in the mid century, disclosed that the price of each picul
of husked rice had by then climbed from 0.7 taels of silver to 0.9 taels,
apparently because of the disturbance caused by the wo-k'ou campaigns.
He also said that in the region south of the Yangtze the 'normal price'was
0.5 taels per picul; this is quite consistent with other sources and applies to
160 The land tax—(ii) Tax administration
most of the sixteenth century. T'ang was in fact arguing that the conversion
rates should be restricted to a range of between 0.5 taels and 0.7 taels."
For most items of commutation in the late sixteenth century this standard
was maintained. The so-called normal price did not, however, exclude
regional, seasonal, and annualfluctuations.It has already been mentioned
that immediately after the harvest grain prices usually dropped. About
1580, it was reported that in South Chihli each picul of husked rice was
worth only 0.3 taels.100 Grand-secretary Sheng Shih-hsing confirmed the
report with the words: 'Recently prices of husked rice have fallen con-
siderably; and 1 tael of silver can now buy 3 piculs.'101 When a landowner
had to sell his crop at depressed prices in order to pay his taxes, he could
end up paying twice as much as he expected. When a tax increase coincided
with a serious fall in farm prices, the price of farmland also declined
significantly. Marginal landowners were then reduced to such desperation
that they sold their properties for less than an average year's income. All
this sounds impossible, but Ko Shou-li reported that it happened in
Shantung in the 1570s.102 About a century later Yeh Meng-chu also reported
that a similar situation occurred in the lower Yangtze region. In 1663
landed properties which previously had been sold for 5 taels per mou were
said to be worth only 0.5 taels per mou. By then the minimum price of
each picul of husked rice was about 0.6 taels.103 Statements such as these
have to be accepted with caution, as they could involve an element of
exaggeration. Yet it is clear that farm credit was difficult to secure and
marginal landowners could not cope with the double squeeze of falling
farm prices and rising tax rates.
Tenancy rates in the delta area have generally been regarded as high, but
again there are no adequate data. The often quoted statement of Ku Yen-
wu, referring to the situation in Soochow prefecture in the seventeenth
century, is fairly vague, to the effect that' only one person out often owned
his land, while the other nine were tenant farmers'.104 Yet there is no doubt
that land tax collection always took the landlord's return into consider-
ation, as is explicitly pointed out in the Shun-te county gazetteer (3, i). The
1643 edition of the county gazetteer of Wu-hsien, Soochow prefecture,
contains the following statement: 'The annual income from top-quality
rice paddies is no more than 1.2 piculs per mou. After paying the regular
land taxes and surcharges, in silver and in kind, the remaining portion is
slightly over 0.7 piculs.'105 The picul in the first figure must have been in
husked rice, otherwise it would not have corresponded to the prevailing
rate of collection and the usual tax procedure. The statement also indicates
that the actual yield per mou was some 2.4 piculs, which is consistent with
other observations from the same area.106 After the usualfifty-fiftysharing
of the crop with the tenant, the landlord had a pre-tax return of 1.2 piculs.
In sum, the land taxes, somewhere between 0.4 piculs and 0.5 piculs per
 4, II Factors affecting the general administration 161
mou, took only about 20 per cent of the yield, but from the landlord's
point of view represented 40 per cent of his income.
For those marginal landowners who also happened to be landlords, tax
rates at this level were quite high and any further increase would have been
unbearable. Tax rates in Soochow were universally regarded as among the
highest in the empire. The year 1643, the eve of the dynasty's collapse, also
marked the high point of its taxation. The above case nevertheless shows
that land taxes under the Ming could not have been increased by much.
Taxpayers sometimes devised extraordinary methods to defeat the
administration. In some districts of Fukien and Kiangsi, a strange type of
land lease became popular in the sixteenth century, generally referred to as
i-fien-san-chu, or 'one field with three owners'. It originated from the
practice whereby the absentee landlord, in order to relieve himself of the
obligation of paying taxes and performing the accompanying service
obligations, nominally 'sold' his land to a second person at a reduced price
but still retained for himself afixedamount of grain income from the same
property. The so-called buyer therefore had to carry out all the obligations
to the government as the effective landowner yet at the same time provide
the original proprietor with a tax-free payment. This second party did not
cultivate thefieldshimself, but permanently leased the property to a tenant.
When the contract had remained in force for several generations, none of
the parties could be released from the linked arrangement. The first party
was called yeh-chu, or 'property owner' and the second ta-tsu-chu, or
'major tax-paying owner'. The tenant was occasionally referred to as
fen-chu or 'fertilizer owner'. Because of the part he played in developing
the land, for example by applying 'nightsoiP to the fields, or because he
paid an extra deposit to the owner, he also had a claim to the property in
perpetuity. The grain income over and above a certain specific amount was
his share and the land could not be transferred to another owner without
his consent.107
The practice of contracting for someone else to pay one's taxes, once
started, could be extended even beyond this tripartite relationship. In
Chang-chou prefecture, Fukien, in the 1560s and 1570s even the major
tax-paying owner, though a tax agent himself, secured yet another tax
agent. The latter received afixedamount of grain payment, barely sufficient
to cover the taxes plus a minimal managerial fee. He became a constant
problem to the local authorities, as his payment was rarely made in time
or in full, and was amusingly termed pei-tui, literally 'the empty-handed
taxpayer'.108
About 1572 the prefect of Chang-chou worked out a formula to re-
constitute the diffused land ownership. The principle was that there should
be only one owner on each piece of taxable land and that this strange
partnership or tax management system must be dissolved, either by one
162 The land tax—(ii) Tax administration
party buying out the rights of the others or through the partitioning of the
property. The plan was not realized, however. After the land survey of
1581, the provincial officials merely attempted to register all the parties
involved so that the tax payment could be apportioned to each of them. In
1612 it was reported that even the second plan was not carried out any-
where.109 In Fukien province this type of land tenure, with certain variations,
is known to have existed in Cheng-ho, Nan-p'ing, Sha-hsien, Yung-an,
Shao-wu, Lung-hsi, Chang-p'u, Ch'ang-t'ai, Nan-ch'ing, P'ing-ho and
Ch'eng-hai counties. Fu I-ling has published a number of old contracts
found in Yung-an county, Yen-p'ing prefecture, which confirm that this
tenure system not only survived the Ming dynasty, but actually lasted
until the late nineteenth century at least.110
The i-fien-san-chu system fully exposed the ineffectiveness of the tax
administration. The three-year term of office of provincial and local officials
was not long enough to permit them to build up sufficient experience of the
district to cope with native customs which could survive for centuries.
In reality the taxpayers were playing a game of hide and seek with the
collectors and there was very little hope that the level of taxation could be
readjusted to reflect the land-users' actual ability to pay. This depended
not on the productivity of land, nor in this case the landlord's grain income,
but on his assistant manager's stipend.
One further practice that seriously obstructed tax administration was that
whereby taxpayers detached their tax liability in the course of property
transactions. A wealthy landowner might select a small acreage from his
holding for sale. The price was set unusually low on the understanding that
the buyer would take over the greater portion of the seller's tax liability.
Conversely, this affluent person might also offer to purchase a large part
of his neighbor's land at premium prices but accept only a lesser portion of
the tax assessment. Through carrying out a series of such transactions
a large landowner could make only a token payment on his huge estates,
while dozens of lesser owners were paying excessive rates. When the tax-
paying power of the general public was undercut by such shrewd maneu-
vers, tax arrears steadily accumulated and the level of taxation was re-
stricted. This evil was recognized and often discussed by bureaucrats
throughout the sixteenth century.111 The practice seems to have spread all
over the empire but was clearly more common in south than in north
China, owing to the greater complexities of tenure and taxation in the rice-
growing region. Some of the abuses may have been corrected by the land
survey of 1581, but its effects could only have been temporary. The author
can testify from personal experience that the practice of separating tax
assessment from acreage in property transactions was still practiced in
certain parts of China right up to World War II.
 4, III The level of collection 163

III THE LEVEL OF COLLECTION

Provincial tax quotas in piculs of grain
The provincial tax quotas can be found in the Ta-Ming Hui-tien, along
with the registered taxable acreages. The rates per mou for each province
are thus easily calculated, as shown in Table 6.
The tabulation of course has its limitations. The acreage figures are
known in some cases to be fallacious and the basic assessment in piculs of
grain was not an accurate index of actual payment, as has already been
explained. Even so the listing is not completely meaningless. In the late
sixteenth and early seventeenth centuries the administrators in Peking
actually relied on these data in making fiscal readjustments and so from
the quality of the statistics one can visualize the quality of the administra-
tion.
The ministry of revenue was fully aware of the nature of these statistics.
Its policy was to utilize the tabulation as a general reference but not to be
absolutely bound by it. This attitude is illustrated by the additional tax
apportionment carried out in 1619. In order to finance the Manchurian
campaign, Minister of Revenue Li Nu-hua, with the Wan-li Emperor's
approval, added a surtax to the land taxes of the whole empire with the
single exception of Kweichow province. Since the value of the picul of
grain varied considerably in the national accounts the grain quota was not
taken as the basis for the apportionment. The financial burden was dis-
tributed in accordance with each province's registered acreage. The land
data, however, besides containing numerous inaccuracies and inconsis-
tencies were fifty years out of date. The ministry subsequently imposed
a universal increase at the rate of 0.0035 taels of silver per mou, a rather
small amount. This was not apportioned directly on each landowner, but
on the provinces and metropolitan areas, whose governors were authorized
to make whatever internal readjustments they wished.112
Since the acreages of Hukwang province and of Huai-an prefecture in
South Chihli were known to have been significantly inflated in the tabu-
lation, the general apportionment formula made an exception in the case
of these two areas. As a compromise, Hukwang was assessed 333,420 taels
of silver, corresponding to some 95 million mou of land instead of the
221 million given in the Hui-tien tabulation. Huai-an prefecture was like-
wise assigned a quota of 35,977 taels, corresponding proportionately to
slightly more than 10 million mou of taxable land, instead of the 13,082,636
mou listed in the national accounts.113 No exceptions were allowed for
other districts as it was understood that the general rate of increase was
small and the minor inequalities resulting from it were not a matter for
serious concern. Because the extra tax was apportioned on acreage rather
6-2
164 The land tax—(ii) Tax administration
TABLE 6. Provincial tax quotas in piculs of grain based on the tabulation
of 1578

Total basic
Registered acreage assessment Indicated average
Province (mou) (piculs of grain) (piculs per mou)

Chekiang 46,696,982 2,522,627 0.054

Kiangsi 40,115,127 2,616,341 0.065
Hukwang 221,619,940 2,162,183 0.009
Fukien 13,422,500 851,153 0.063
Shantung 61,749,899 2,850,936 0.046
Shansi 36,803,927 2,314,802 0.032
Honan 74,157,951 2,380,759 0.032
Shensi 29,292,385 1,735,690 0.059
Szechuan 13,482,767 1,028,544 0.076
Kwangtung 25,686,513 999,946 0.039
Kwangsi 9,402,074 371,696 0.039
Yunnan 1,799,358 142,690 0.079
Kweichow 516,686 50,807 0.096
South Chihli 77,394,662 6,011,846 0.078
North Chihli 49,256,836 598,622 0.012
Total 701,397,607 26,638,642 0.038

than on the grain quota, it also avoided over-burdening those regions which
already had high tax rates, such as South Chihli.
From Table 6 it can be observed further that the average grain assess-
ment, at 0.038 piculs per mou, was not significantly different from Hung-
wu's ideal of 0.0335 piculs per mou (1, n). Regional distribution varied
widely, nevertheless. Rather strangely, the tabulation suggests that the
highest assessments were on Kweichow and Yunnan, two underdeveloped
areas, while the lowest were on Honan, North Chihli, and Hukwang.
These surprising facts are explained by inconsistencies in the original
data. Land tax collection in Kweichow and Yunnan, and in certain dis-
tricts of Hukwang and Kwangsi, had never followed normal procedures.
From the days of the Hung-wu Emperor land taxes in Yunnan had been
collected in a variety of commodities including cinnabar and vermilion.114
Portions of the payments were made by the aborigine chiefs in lump sums,
and bore little resemblance to ordinary taxation. Up to the very end of the
Ming dynasty certain tax payments in Teng-ch'uan subprefecture were
still made in cowrie shells.115 In Kweichow no Yellow Book was ever
compiled.116 A seventeenth-century memorial to the last emperor, Ch'ung-
chen, summarized the situation as follows:
Most of the territory of Kweichow originally belonged to the native tribes. Our
taxation enables us to exercise some measure of control over them. By paying
taxes the tribesmen are also able to learn something about Han customs. But for
all practical purposes the cultivated land is in the hands of the aboriginal officials.
 4, III The level of collection 165
Acreage which pays regular taxes to the magistrates and prefects comprises no
more than 10 per cent of the total.117
Obviously neither the reported acreage nor the quoted assessments could
be taken seriously. Because of the small volume of revenue derived from
these two provinces, taxation in Yunnan and Kweichow seldom became
an important issue. Both territories in fact regularly received subsidies from
other provinces.
The low level of taxation apparently prevailing in North Chihli was un-
real, in that the local horse payment (3, n), having been directly assessed
in silver, was excluded from the grain quota tabulation. There was also
a deliberate policy of keeping the level of formal taxation low in the counties
and prefectures adjacent to the capital, so that the populace could be called
upon to perform the numerous kinds of services required in Peking.118
In 1623, when substantial land tax increases were imposed to meet the
Manchurian crisis, all the prefectures of North Chihli except one were
exempted from the collection. The exempted prefecture, Shun-t'ien, con-
tributed only 1,227 taels of silver. But for a period the population of North
Chihli was intensely involved in servicing the military units in transit and
making contributions to army supplies.119 This policy of waiving formal
taxation in exchange for unpaid services was consistent with the dynasty's
regular practice. It affected North Chihli more because the requisitions
came directly from the imperial government.
The lower per mou rate of Honan was part real and part illusory. In the
late sixteenth century the province had absorbed a substantial amount of
hitherto untaxed acreage (3, n), enabling many districts to reduce the per
mou collection by means of tax reapportionment. On the other hand the
counties and prefectures along the Yellow river were assessed a heavy
additional payment known as ho-kung (river-control expenses).120 Un-
doubtedly a great portion of this financial burden was carried by the
landowners.
Conditions in Hukwang, made more confusing by its unreliable acreage
figure, cannot be ascertained. The fundamental difficulty was caused by the
peculiar topographical features of the great lake region. In 1582, Minister
of Revenue Chang Hsiieh-yen (in office, 1578-83) observed that while
most of the agricultural production in the province was concentrated on the
lake shores, the area was often inundated.121 The lack of accurate land data
for the area around the Tung-t'ing lake has remained a problem even in the
present century.
Without going into further details, this brief evaluation gives one an
idea of the picture that the minister of revenue saw from his desk in the late
sixteenth and early seventeenth centuries. Any additional information
available to the ministry only confirmed and enlarged this basically rather
diffuse picture. The fiscal administration had to be carried out in an
166 The land tax—(ii) Tax administration
impressionistic and artistic manner. The modern scholar who wants to
ascertain the real level of land tax collection has to gather his information
elsewhere. Local accounts do provide one major source, but a great deal of
reconstruction is still necessary.

The level of collection indicated in local accounts

Most of the Mingfiscalaccounts cannot be used statistically. It is possible
to make a rough estimate of the level of formal taxation by scanning the
data, connecting seemingly unrelated factors and guessing the missing
links, but as a rule thefigurescannot easily be handled mathematically. An
accurate estimate of the level of land tax collection must take into account
the nature of the land data, conversion rates, farm prices, and accounting
methods, and there is never a case in which all these factors are absolutely
clear.
In Table 7 the 1572 accounts of Hang-chou prefecture, Chekiang, are
selected for a case study.122 The prefecture comprised nine counties grouped
around the prefectural capital, a city which still bears the name of the
district today. This prefecture was one of the few in whichfiguresfor land
data approximated to actual measurements.123 Even though several of its
counties were situated in moutainous regions, such as Yii-ch'ien and
Ch'ang-hua, their land taxes were remarkably small. On the other hand the
taxable acreage of those counties which yielded larger revenue, such as
Jen-ho and Hai-ning, comprised only negligible amounts of hilly and
marshy land.
Under the direct supervision of P'ang Shang-p'eng, one of the most
enthusiastic promoters of the Single Whip Reform, Hang-chou prefecture
compiled quite a detailed service levy account. It is assumed that 60 per
cent of the service levy was borne by taxable acreage, in accordance with
the general practice in Chekiang (3, m). In 1572 portions of the land taxes
in Hang-chou were still paid in kind, and these have here been converted
to silver, at a rate of 0.6 taels per picul of grain and 0.04375 taels of silver
per tael of silk wadding. These were the standard rates of tax commutation
in Hang-chou prefecture at the time.
The level of farm income here is not all clear. I-wu county in neighbouring
Ching-hua prefecture reported in 1592, however, that one mou of medium-
quality land yielded 4 piculs of unhusked rice.124 Around the same time
one mou of top-quality land in Chia-hsing and Hu-chou prefectures, also
in Chekiang, yielded 3 piculs of husked rice as well as other crops.125
Shang-yii county in Shao-hsing prefecture reported in the early seventeenth
century that one mou of top-quality land yielded 5 piculs of unhusked
rice.126 Hang-chou prefecture was surrounded by these districts and the
productivity of its soil was not considered to be inferior to theirs. It is safe
 4, III The level of collection 167
TABLE 7. Estimated land tax rates of Hang-chou prefecture in 1572

(e = d/c) (f = c + d )
(d) Pro- Total (g = f/a)
(b) 50% of portion land Per mou
Land (c) service of service taxes payment
(a) taxes Value levy levy to after after
Taxable assessed of land borne land absorp- absorp-
acreage in grain t taxes* by land taxes tion tion
County (mou) (piculs) (tls.) (tls.) (%) (tls.) (tls.)

Jen-ho 805,238 104,980 64,948 21,942 34 86,890 0.108

Ch'ien-t'ang 592,745 44,770 27,753 12,187 43 39,940 0.067
Hai-ning 932,706 109,243 57,061 19,461 34 76,522 0.082
Fu-yang 686,565 17,231 14,904 7,512 50 22,416 0.033
Yu-hang 614,898 28,244 19,771 6,430 36 28,201 0.043
Lin-an 184,806 11,362 7,953 4,015 50 11,968 0.065
Hsing-ch'en 127,998 6,709 4,696 2,217 47 6,913 0.054
Yu-ch'ien 147,259 5,099 3,569 2,028 57 5,597 0.038
Ch'ang-hua 90,199 3,680 3,031 1,820 60 4,851 0.054
Total 4,182,414 331,318 203,686 77,612 38.1 281,298 0.067
prefecture

t The grain assessment includes the basic assessment and surcharges.

* The value is arrived at by adding the commuted amount and the estimated value of taxes
still paid in grain.
NOTE: One item from each of the service levy accounts of Hai-ning and Fu-yang counties is
missing in the original. Estimates of these two items have been added to the respective accounts.

therefore to assume that medium-quality land in Hang-chou yielded

2 piculs of husked rice per mou. When husked rice was sold at 0.6 taels of
silver per picul, such medium-quality land would thus produce an annual
income of 1.2 taels per mou.
Hang-chou was also a silk-producing district. According to Mou K'un
(1512-1601), before the turn of the century the income from silk-producing
land around T'ai lake was as high as 5 taels of silver or more per mou. Even
inferior land planted with mulberry trees could produce an annual income
of between 1 and 2 taels per mou.127 Though these income estimates seem
rather high, when applied to Hang-chou prefecture they suggest that 1 tael
of silver per mou would be a conservative estimate for the average farm
income of the district's taxable acreage.
Thefinaltax rates thus represent 10.8 per cent of estimated farm income
in Jen-ho county, where the collection was the highest, and 3.3 per cent in
Fu-yang county, where it was lowest. The overall tax rate in the prefecture
is estimated to have been 6.7 per cent of the district's total grain and silk
production.
One more factor yet to be considered is that in the 1570s grain prices in
Chekiang and South Chihli were below their normal level. Hang-chou
168 The land tax—(ii) Tax administration
prefecture in 1579 permitted some taxpayers to submit 0.4 taels of silver in
lieu of one picul of grain in actual payment, this rate being applicable to
several small items still collected in kind. The offer was optional,128 and
even though its application was limited, the rate of cash payment must
have been close to local grain prices. If the price depression was general the
above estimate of farm income should be reduced by 50 per cent, or else
the tax percentage increased by 50 per cent. This would put the overall
collection at 10.05 per cent of agricultural output.
It is difficult to make any similar estimate for the northwest of the
country, mainly because the nature of the land data is unclear. Following
the lead of Ping-ti Ho who has pointed out that Fen-chou prefecture in
Shansi provided substantial information about its fiscal mou conversion
system, the level of tax collection in this prefecture has been worked out
as shown in Table 8.129
Fen-chou prefecture was situated to the south of the provincial capital,
T'ai-yiian. Its territory covered a large part of the Fen river valley, the most
productive area of the province. Both rice and wheat were staple crops. The
accounts of Fen-yang county indicate that, from the early years of the
dynasty up to 1577, its land taxes were assessed in rice and wheat, in
general maintaining a 2:1 ratio.130 The irrigation dams on the Fen river
tributaries, described at some length in the prefectural gazetteer, must have
been a considerable benefit to local agriculture.
The prefecture also comprised some territory on which there was very
little possibility of improvement. Yung-ning subprefecture and Ning-hsiang
county, which was subordinate to it, both belonged to this category. The
former, in the northwest corner of the prefecture and remote from the
central alluvial basin, encompassed miles of barren mountains stretching
all the way to the Yellow river.131 In Ning-hsiang county, agricultural land
was traditionally classified intofivegrades for tax purposes. After the land
survey of 1581 these grades were eliminated, but acreage which originally
had fallen into the two lowest grades was converted by counting 4 actual mou
as 1fiscalmou. The total cultivated land in the county was actually 748,137
mou, but after fiscal mou conversion it was re-stated as 236,982 mou.132
This information is important because it suggests that in order to arrive at
thisfiscalacreage, 681,540 mou of land, more than 90 per cent of the county's
total, had been classified as being of inferior quality; only 66,597 mou,
or less than 10 per cent, fell into the upper class and was not converted.
For the territory directly under the administration of Yung-ning sub-
prefecture, a different conversion formula was adopted. For dry fields
every standard mou on level ground, every 4 mou on slopes, and every
7 mou in the mountains, were counted equally as 1fiscalmou. Rice paddies
in the valleys had a reverse conversion rate, with every 0.8 mou of actual
measurement being counted as a fiscal mou; in other words each actual
 4, III The level of collection 169
TABLE 8. Estimated land tax rates of Fen-chouprefecture in 1608

(e = d/c) (f=c+d)
(d) Pro- Total (g = f/a)
(b) 50% of portion land Per mou
Land (c) service of service taxes payment
(a) taxes Value levy levy to after after
Taxable assessed of land borne land absorp- absorp-
acreage in grainf taxes* by land taxes tion tion
County (mou) (piculs) (tls.) (tls.) (%) (tls.) (tls.)

Fen-yang 1,007,172 48,617 42,090 8,661 21 50,751 0.050

P'ing-yao 1,014,313 52,493 49,340 6,160 13 55,500 0.055
Chieh-hsiu 602,032 27,491 26,289 4,719 18 31,008 0.051
Hsiao-i 905,417 26,406 25,795 3,768 15 29,563 0.033
Lin-hsien 316,044 14,491 13,397 2,088 16 15,485 0.049
Ling-shih 223,111 12,920 13,139 2,918 22 16,057 0.072
Yung-ning (S]) 469,821 27,075 23,491 4,682 20 28,173 0.060
Ning-hsiang 236,982 10,615 9,072 1,935 21 11,007 0.046
Total 4,774,892 220,108 202,613 34,931 17 237,544 0.050
prefecture

t The grain assessment includes the basic assessment and surcharges.

* The value is arrived at by adding the commuted amount and the estimated value of taxes
still paid in grain.
(S) = subprefecture.
NOTE: Of the prefecture's total grain quota including surcharges, all but 7,210 piculs or less than
3.3 per cent, had already been commuted to silver. The remaining amount has been converted here
at a rate of 1 tael per picul, to cover the price of the grain and the service charges.

mou counted as 1.25 fiscal mou.133 These conversion rates permit the con-
clusion that the productivity of a fiscal mou in the prefecture was supposed
to equal approximately 1 picul of wheat or its equivalent, such as 1.2 piculs
of barley. It could not have been much smaller as it represented the return
from 7 mou of standard measurement on the mountains as well as 0.8 mou
in the valleys. If 7 mou of land could not yield a return at this level it could
not have been economically cultivated, or at least such land would not
have been so widely cultivated as to constitute a general category for
purposes of taxation. Conversely, the estimated return could not have been
much larger, since it also represented the harvest from less than 1 standard
mou of irrigated land in the northwest of the country.
Though other counties in the prefecture reported their acreage without
disclosing the nature of the land data, it seems that they too used fiscal
mou conversion. Ling-shih county, for instance, acknowledged that its
acreage had been arrived at through 'flexible readjustments'.134
Grain prices in Shansi were about 30 per cent higher than in the southern
provinces, according to Hsu Cheng-ming, writing in the late sixteenth cen-
tury. The conversion rates of 0.8 taels of silver per picul for husked rice and
0.6 taels per picul for wheat probably reflect, fairly accurately, the market
170 The land tax—(ii) Tax administration
value of those commodities.135 In these estimates it is therefore assumed
that every fiscal mou in Fen-chou prefecture produced 1 picul of wheat, and
thus an average income of 0.6 taels. This level of return is perhaps rather
low for Fen-yang and P'ing-yao counties in the alluvial basin, but could
likewise be slightly too high for the mountainous districts where trans-
portation costs were higher and grain prices lower.
On the assumption that farm income was 0.6 taels per mou, these rates
show that taxation was highest in Ling-shih county, at 12 per cent of total
output and lowest in Hsiao-i county at 5.5 per cent. The average over the
whole prefecture was 8.3 per cent. This was a period when inflationary
tendencies were becoming evident (2, iv); the rise in grain prices would
lower rather than raise these percentages.

The level of collection estimated from miscellaneous sources

Casual observations by contemporary writers and summary statements in
local gazetteers sometimes offer leads to estimating the level of taxation in
the sixteenth century, but again there is never any direct attempt to assess
tax payments as a proportion of farm income. Even statements that specify
total payment in taels per mou are hard to find. In the main the late six-
teenth century was a period of readjustment when the situation was so
fluid and the tax schedule so complicated that even some local officials
were unaware of the exact level of taxation. It usually came to light only
when it became an issue of debate.
The following ten cases represent reports from many regions over a period
of a hundred years or so. The normal scholarly approach would be to
tabulate the findings, but they are so meager and scattered that it would be
unwise to treat them as statistics which could conceivably inhibit further
research. A more methodical arrangement of the data would of course be
possible, but it would reduce or obscure the background information vital
to understanding them. Though taxation in the late Ming undoubtedly
promises to be a useful key to the further understanding of traditional
China, the topic is at present extremely under-researched and likely to
remain so for a long time. It is therefore essential at this early stage that all
the factors relevant to the inquiry be presented in full. The cases are there-
fore discussed individually, as follows:
(a) In 1543 An-hua county, Hukwang, imposed a basic payment of
0.02675 piculs of grain on each mou of privately owned land. This assess-
ment was in turn converted to 0.4432658 taels of silver per picul, which was
sufficient to cover all the surcharges and surtaxes.136
Neither the productivity nor the nature of the mou was mentioned. In
1623, however, a Hukwang surveillance commissioner estimated that all
grades of land in the province produced on average a yield of 3 piculs of
 4, III The level of collection 171
unhusked rice.137 Recently a Japanese scholar has estimated 2 piculs of
husked rice as the average yield per mou in Hukwang in the late Ming.138
A conservative estimate for the yield in mid sixteenth century An-hua, not
far from the major area of rice cultivation, would therefore be at least
1 picul of husked rice per mou. Grain prices in Changsha prefecture, in
which the county was located, were traditionally low.139 If the price were
0.3 taels of silver per picul, the above tax formula, which actually amounted
to 0.01184 taels per mou, would mean that tax collection took 3.9 per cent
of farm income. The price of husked rice subsequently declined further to
0.2 taels per picul at the turn of the century,140 in which case the figure
would have been 5.9 per cent.
(b) In 1569 Li-yang county in South Chihli divided its land taxes into
several different categories. The converted silver payments, which ranged
from 0.04 taels to 0.0065 taels per mou, included all surcharges, surtaxes,
and the portions of the service levy collectible from the land.141
Li-yang was on the edge of the Yangtze delta area, and contained many
lakes and rivers, so that it would be reasonable to expect an average yield
of 1.5 piculs of husked rice per mou.142 At the normal grain price of 0.5
taels per picul, the highest rate represented less than 5.4 per cent of the
value of the crop, and the lowest less than 1 per cent. These unusually low
rates were due not to poor productivity, but to historical circumstances.
Nor were the differences in rates closely related to relative productivity
within the district, all of which was pointed out by the county gazetteer.
The payment of less than 1 per cent is the lowest ever found so far.
(c) In about 1570 Ko Shou-li wrote to the governor of Shantung to say
that in order to promote the Single Whip Reform the provincial adminis-
tration office had decided to convert all the province's basic grain payments
to consolidated Single Whip payments. Three grades of payment, at
0.9 taels, 0.8 taels and 0.6 taels, were to replace the picul of grain.143
It was the general practice in Shantung to convert three or more standard
mou of land to one fiscal mou,144 from which the minimum yield was con-
sidered to be 1 picul of wheat. On such a fiscal mou the basic payment was
0.05 piculs of grain. The price of wheat was quoted in Wen-shang county
in 1576 as 0.52 taels of silver per picul, but immediately after the harvest it
fell to 0.37 taels.145
These prices would mean that the highest payment expected by the
provincial authorities fluctuated between 8.6 per cent and 12.2 per cent of
crop yield, and the lowest between 5.8 per cent and 8.1 per cent.
(d) In about 1573, the prefect of Chang-chou, Fukien, disclosed that the
actual payment on every ten mou of taxable land in his district, including
all the extra assessments, amounted to around 1.2 taels of silver.146
The rent rates in Nan-ch'ing and P'ing-ho counties show that the average
productivity of each mou of taxable land here was approximately 2 piculs
172 The land tax—(ii) Tax administration
of husked rice.147 At the prevailing price of 0.5 taels per picul, the con-
solidated tax payment took 12 per cent of the grain income. The high level
of collection was in part due to the wo-k'ou campaigns, during which sur-
taxes were increased.
(e) Around 1580, Ts'ao-chou in Shantung devised the following tax
formula: in the territory directly administered by the subprefecture, every
2.7 standard mou was converted to a fiscal mou, on which a Single Whip
payment of 0.042 taels of silver was levied. In Ts'ao-hsien county, the fiscal
mou comprised 4.8 standard mou, and the consolidated payment was
0.071 taels. In Ting-t'ao county, thefiscalmou contained 3.6 standard mou
and the payment wasfixedat 0.052 taels.148
Assuming the same grain prices and estimated per mou yield as above,
it can be estimated that taxpayers in these districts spent 9.1 per cent,
8.7 per cent and 8.5 per cent of their income respectively in paying taxes.
When the price of wheat was 0.37 taels, those percentages would have to be
revised to 12.6 per cent, 12 per cent and 11.7 per cent.
(/) In 1584 land taxes in Shun-te county, Kwangtung, took on average
3.5 per cent of agrarian output. When the price of rice fell to 0.3 taels per
picul the figure would be 6.1 per cent (3, i).
(g) In K'ai-hua county, Chekiang, in 1620, every 34.428 mou of top-
quality rice paddy was consolidated for tax purposes into a single acreage
unit on which a basic payment of one picul of grain was assessed. This
payment, which included all the assorted taxes apart from the Manchurian
supplies, was converted to 1.614673 taels of silver, a significantly high
level.149
K'ai-hua was basically a timber-producing district, and its productivity
and grain prices cannot be ascertained. It can be assumed, however, that
the crop yield on one mou of top-quality land was unlikely to have been
less than 0.8 taels, since there are indications that the return from other
such land in the same province could be as high as 1.5 taels per mou. The
tax payment, simplified to 0.047 taels per mou, is therefore estimated not
to have exceeded 6 per cent of farm income, and could have been much
lower.
(h) As was noted earlier (4, n), in Wu-hsien, Soochow prefecture, South
Chihli, taxes were so heavy that its gazetteer in 1643 complained that about
0.4 piculs of grain had to be set aside as tax payment on each mou of land.
Even though this statement was made in the mid seventeenth century,
conditions in Soochow were not then significantly different from what they
had been in the late sixteenth century. The tax increases apportioned to
acreage in the seventeenth century represented only a very small increase on
the existing assessment, which was already considered high.
Basic assessments in the delta area always seem to have been about five
or six times as high as the prevailing rates elsewhere, but the tax burden on
 4, III The level of collection 173
the general population was much reduced by the commutations to Gold
Floral Silver and cotton cloth. The collection of service levy payments in
this area was, moreover, not proportionate to its regular land taxes. Pro-
ductivity in the delta area was generally regarded as the highest in the
empire. Of the four prefectures of South Chihli and two prefectures of
Chekiang, only the coastal area of Sung-chiang prefecture was considerably
disadvantaged by reason of its akaline soil, but there the landowners were
amply recompensed by generous fiscal mou conversion.150
The above statement in the Wu-hsien gazetteer appears to mean that
while taxation took 40 per cent of the landlord's pre-tax income, it actually
constituted no more than 20 per cent of agrarian production. The accounts
of Shanghai county in 1584 show that conditions were quite similar there.
On average the rate on each mou of taxable land was about 0.1 piculs of
grain in kind plus 0.08 taels of silver.151 The total payment is estimated to
have been close to 15 per cent of the yield. In 1621 the governor of South
Chihli pointed out that the four prefectures under his jurisdiction paid
higher taxes than any other district. After enumerating all the payments in
the tax register, he described a set of complicated rates, including payments
both in kind and in silver.152 The rates apparently did not exceed 20 per
cent of farm output, however.
Tax payment in the delta area was thus still two or three times that in
other districts, but definitely notfiveor six times higher. So far no evidence
has been found to indicate that tax collection in any district outside the
delta area was anywhere near this level.
(i) In 1643 Minister of Revenue Ni Yiian-lu, reported to the Ch'ung-
chen Emperor that land tax collection rates varied widely over the empire,
from 0.13 taels per mou to slightly over 0.2 taels, and warned that these
figures must conceal 'unauthorized extras'.153
Ni's statement is not altogether clear. The crucial question is how widely
these rates were applied. If they were only isolated cases there would have
been little cause for alarm. Seventy years earlier Chang-chou prefecture in
Fukien was already collecting at the rate of 0.12 taels per mou and at the
same time the level of collection in the Yangtze delta was close to 0.2 taels
per mou. From 1619 onwards the imperial government ordered further
tax increases on seven occasions until the additional surtaxes reached the
level of 0.0268 taels per mou. In addition there was a 10 per cent increment
on all taxes that exceeded 1 tael of silver in total amount.154 It is probable
therefore that the tax rates in some areas did reach the levels described by
by Ni Yiian-lu, even without the unauthorized extras to which he referred.
It is therefore likely that Ni was warning the emperor that most districts
had raised their taxes to this level during the period of confusion, which
means in turn that normal, authorized tax collection before the seventeenth
century was probably appreciably below the range of between 0.1 and 0.17
174 The land tax—(ii) Tax administration
taels per mou (the rates quoted by Ni, minus the new surtaxes). This is
consistent with most of the late sixteenth-century accounts and reports.155
With the single exception of the Yangtze delta, no district had a collection
close to 0.17 taels per mou and even payments in excess of 0.1 tael per mou
were unusual.
(j) The prefectural gazetteer of P'ing-liang, Shensi, dating from about
1600, contains a nostalgic description of taxation in the past. The writer
pictured a husband and wife team working 200 mou of land and producing
300 piculs of assorted grain, less than 30 of which were paid as taxes in
kind. One-tenth of the grain output, he continued, should be 'amply
sufficient' to pay all taxes.156 It is apparent that the rate of collection had
since exceeded this level, but unfortunately the author did not specify the
current proportion of the tax outlay.
Increase in new local government offices, the proliferation of imperial
clansmen, heavier demands for services and a decrease in the amount of
taxable land were among the factors cited here as the causes of tax in-
creases. Significantly the gazetteer writer also held that commutation of
taxes to silver payments was responsible for much distress to the general
population. The taxpayer had to sell his crop at disadvantageous prices
and in addition the collection of melting charges made the actual rates
higher than they seemed.

An overall estimate of the collection

Any attempt to establish the general level of collection at present involves
considerable degree of uncertainty, and it will take many years to accumu-
late enough new types of source material to add significantly to the in-
formation currently available.
On the other hand it is still possible to make certain general observations
in the light of the current state of knowledge. Formal taxation, that is, the
total payment including the regular land taxes, surcharges, surtaxes,
portions of the service levy collected on the land, and uncollectible items
absorbed into it, was in general until the turn of the century less than 10
per cent of the estimated farm output at the normal local grain prices. In
many districts the collection was far below this level. Though in a few
places it was exceeded, the only significant case was that of the Yangtze
delta, where the collection approached 20 per cent of farm income. These
estimates do not take into account unpaid services, unauthorized collections
and melting charges, nor supplementary income from other than staple
crops.
The average level of collection over the whole empire does not seem to
have exceeded 10 per cent of agrarian output either, an estimate which is
unaffected by the unusual case of the Yangtze delta. Though nominally the
 4, IV Disbursement of tax income 175
four prefectures' total basic assessment comprised about 10 per cent of the
empire's land tax quota, it was greatly reduced by the commutation of tax
payments and the absorption of the service levy.
Only a general conjecture can be made as to the monetary value of the
taxation. Of the 26.6 million piculs of grain in basic assessment (Table 6),
more than 80 per cent seems to have been commuted to silver payments
before 1600. Only some 4 million piculs are definitely known to have been
commuted at the rate of 0.25 taels per picul (Gold Floral Silver). The rest
of the commutation rates varied widely. In South China however, most
commutations fell within the range of between 0.5 and 0.7 taels per picul.
In north China 0.8 taels to 1 tael per picul could be accepted as the normal
range. The surcharges, surtaxes, and the absorption of other revenues could
raise the average value of the * picul' but the Gold Floral Silver, commu-
tation to cotton cloth, and the low commutation rate prevailing in Hukwang
of 0.3 taels per picul, would all tend to lower the average. If one then
assumes that the average value of all 'piculs', in kind and in silver, was
0.8 taels, the total value of the regular land taxes would be slightly more
than 21 million taels.157 An assessment of the service levy accounts of
thirty-five counties in seven different provinces shows that on average they
each collected 9,724.26 taels on this item. (The counties appear in Tables 2,
3,4,7 and 8.) The total collection of service levy throughout the empire was
therefore probably in the vicinity of 10 million taels. Even if it was only
partially absorbed by the land taxes, the service levy should at least have
raised the total revenue from agricultural land to 25 million taels, or even
close to 30 million taels.
It is difficult to arrive at a more precise estimate because of variations in
local prices, readjustment of surcharges, etc. Not only are the extant data
incomplete, but the accounting methods cannot be fully ascertained. One
highly significant fact that must also be taken into account is that the
amount actually collected rarely exceeded 80 per cent of the projected
revenue, and sometimes fell much below this.

IV DISBURSEMENT OF TAX INCOME

Principles governing the distribution
Since most administrative expenses were met by the service levy, income
from the regular land taxes was allocated for the following uses only:
rations and pay for the army, the support of imperial princes and imperial
clansmen, official salaries and the stipends of government students. From
the beginning of the dynasty the concept was established that the regular
land taxes meant food, that is their proceeds were expected to be consumed
by individuals or issued to them. From time to time tax revenue was also
176 The land tax—(ii) Tax administration
diverted to finance large-scale public works but the appropriation was
basically to feed the laborers. Local reserves were maintained to meet
famine needs, another food requirement. Tribute grain and southern grain
were simply large tax consignments geographically redistributed. The only
occasions when the proceeds of the land taxes were treated as disposable
cash was when they were used to pay for local procurement orders of the
central government.
As late as 1578 the imperial university in Nanking still received 3,500
piculs of husked rice from Ch'ang-chou prefecture, 100 piculs of wheat
from Ning-kuo prefecture and 100 piculs of green lentils from Ying-t'ien
prefecture, all in South Chihli, and also over 20,000 catties of dried fish
from Hukwang province, all these items being charged to the tax quotas
of the respective districts.158 Such arrangements as these were based largely
on past usage rather than careful planning. The commutation of taxes to
silver did eliminate some cumbersome aspects of the supply procedure but
it was still impossible to make a fundamental break with the old system.
For one thing, the distribution of tax proceeds could no longer be re-
adjusted ; few of the commuted items could be either discontinued or reduced.
Worse still, the accounts were hopelessly complicated by the diverse nature
of fiscal units, regional prices, surcharges, and accounting methods. The
ministry of revenue allowed the lesser functionaries to keep detailed
records but none of this general information reached the higher levels of
government. Though ministerial officials did occasionally cite national
figures in their reports, in most cases these were general estimates rather
than valid statistics. It is worth noting that the ministry of revenue con-
tinued to be organized in territorial bureaus, a system that was in turn
adopted by the Ch'ing.159
In the late sixteenth century the ministry of revenue could barely be
considered as an operating agency. The funds under its direct control were
very small, consisting only of a few miscellaneous items from the land
taxes. This income was spent largely on the northern frontier, which will
be considered in the course of the general discussion of the empire's fiscal
administration (7, i). The ministry supervised the administration of the
land taxes but rarely made innovations, preferring always to maintain the
status quo. Though the categories of 'transferred revenue' and 'retained
revenue' remained in effect (l,i), the government published detailed
accounts relating to the former only, paying little attention to the latter.
The preoccupation with transferred revenue was a practical one in that it
was principally used for the maintenance of Nanking and Peking and for
supplying the army posts on the northern frontier, for which items of
expenditure the ministry of revenue was directly responsible. As for re-
tained revenue, in many provinces the income in this category was no
longer adequate to finance the expenditures for which it was earmarked
 4, IV Disbursement of tax income 177
(see below). Furthermore, since the imperial government exerted constant
pressure on provincial and local officials to provide it with transferred
revenue, all shortages resulting from uncollectibles, tax arrears and tax
remissions had to be met from the retained revenue. When funds were in-
sufficient, the shortages in the latter were simply never made up.
The grain quota was allocated on a more or less permanent basis, as
shown in Table 9.160
Although this table is reconstructed from the detailed accounts in the
Ta-Ming Hui-tien of 1578, its findings are almost identical to those in
a summary prepared by the ministry of revenue in 1502, and also to an
account left by Han Wen (minister of revenue 1506-8).161 It appears there-
fore that the basic quotas, once established, were never revised. The
ministry could of course transfer certain tax funds if necessary, but such
readjustments were temporary and did not necessitate any permanent
alterations to the system. In any case most of the readjustments made in the
sixteenth century were achieved by modifying the commutation rates and
surcharges, and thus did not affect the grain quota.

TABLE 9. Estimated land tax collection and disbursement in piculs of grain,

ca. 1578

Collection Disbursement

Summer tax 4,600,000 Retained funds and supplies 11,700,000

Autumn grain 22,000,000 Delivered to frontier posts by taxpayers 3,300,000
Delivered to Nanking 1,500,000
Delivered to Peking 9,534,000
In grain at imperial granaries (4,000,000)
As * white grain' (214,000)
In cotton cloth and other supplies (900,000)
In Gold Floral Silver (4,050,000)
Otherwise permanently commuted (370,000)
Miscellaneous and unaccounted for 566,000
Total 26,600,000 26,600,000

In order to discuss the distribution of tax revenue in more detail it is

necessary to examine some local accounts.

Distribution of tax revenue: the case of Lin-fen

The 1591 accounts of Lin-fen county, Shansi, have been selected to illustrate
the distribution of tax revenue in the northwest. Even in 1591 the county
still retained such categories as summer tax and autumn grain. In the case
of the summer tax, the transferred revenue was delivered to five different
receiving agencies and the retained revenue to seven. In the case of the
autumn grain the transferred revenue was delivered to seven receiving
178 The land tax—(ii) Tax administration
TABLE 10. Distribution of land tax revenue from Lin-fen county 1591:
in grain

Expenditure Piculs %
Transferred; subsidies to army frontier posts 21,044 37.1
Retained; subsidies to local army units 1,102 1.9
Retained; payments to princes and imperial clansmen 27,551 48.5
Retained; stipends of government students 1,044 1.8
Retained; salaries, forwarded to prefecture, etc. 6,060 10.7
Total 56,801 100.0

TABLE 11. Distribution of land tax revenue from Lin-fen county 1591.
in silver

Expenditure Tls. %
Transferred; subsidies to army frontier posts 20,013 36.2
Retained; subsidies to local army units 889 1.6
Retained; payments to princes and imperial clansmen 21,678 39.2
Retained; stipends of government students 835 1.6
Retained; salaries, forwarded to prefecture, etc. 4,646 8.3
Miscellaneous and unaccounted for 7,282 13.1
Total 55,343 100.0

agencies and the retained revenue to sixteen. 95 per cent of the county's
land taxes were, however, paid and delivered in silver, which eliminated
most of the technical difficulties in handling diverse fiscal units. In the
above two tables those taxes that were still paid in kind have been converted
at 1 tael of silver per picul, in accordance with the prevailing rate in the
county. The fact that the gazetteer designates precisely the name and
location of each receiving agency enables one to identify the nature of each
item of expenditure in the delivery schedule. The gazetteer also lists the
original distribution in terms of piculs of grain so that it could be checked
against the national accounts and the initial delivery orders. The accounts
are summarized in Table 10.162 The same payments, when converted to
silver, result in slightly different percentages, as shown in Table 11.
Since the original was more a delivery schedule than a real disbursement
ledger, it is impossible to ascertain the distribution in any more detail. Even
so, certain features are clear. The northern army posts, unsurprisingly,
received a large share of the tax revenue. The smallness of the amount left
to cover administrative costs can be explained by the nominal nature of
Ming official salaries and by the collection of service levy, the proceeds
from which were kept separate from the regular land taxes. The most un-
expected finding is perhaps that the largest share of the revenue went to the
princes and imperial clansmen resident in the area.
 4, IV Disbursement of tax income 179
This state of affairs was not unusual, however. While the percentage
varied from one prefecture to another, there is no doubt that in many
districts of north China, especially in Honan and Shansi, grants to the
princes and imperial clansmen remained the largest single item to be met
from the retained revenue. At about the same time as the Lin-fen county
gazetteer was published, K'ai-feng prefecture in Honan and the counties
adjacent to it were faced with the same problem of meeting this enormous
item of expenditure. One writer estimated that these forty-three counties
and subprefectures had a total tax quota of close to 800,000 piculs. Some
300,000 piculs of this was delivered as transferred revenue, leaving 500,000
piculs or so as retained revenue. Grants to princes and imperial clansmen
took 300,000 piculs of the latter. In other words, in these forty-three
districts such grants accounted for approximately 60 per cent of the retained
revenue and close to 40 per cent of all income from the regular land taxes.
By contrast the salaries paid to officials and lesser functionaries and the
stipends of government students in the same districts when added together
did not exceed 50,000 piculs, about 6 per cent of the revenue.163
It must be pointed out, however, that these grants constituted the
'fictitious' portion of the expenditure budget and that in the late sixteenth
century no district made the appropriation in full. Though the grants to
high-ranking princes were usually delivered according to schedule, those
to lower-grade imperial clansmen were reduced, left in arrears, or else
simply embezzled by government officials. As early as 1502 payments to the
imperial clansmen in Honan and Shansi, if made strictly in accordance with
the official schedule, would have exceeded the total retained revenue of
these two provinces.164 Ever since 1498 the government had in fact been
officially reducing the payments.165 Even at reduced rates the grants were
still a serious drain on government income, and at the same time the
imperial clansmen continued to proliferate at a great rate. In 1529 it was
reported that there were 8,203 names in the 'jade genealogy';166 exactly
forty years later the number had swelled to 28,492.167 It has been estimated
that by the end of the Ming period the direct male descendants of the
dynastic founder numbered close to 100,000.168
There is no adequate estimate of how much revenue was actually delivered
to meet these requirements in any one year. A censorial official indicated
in 1562 that the total payment on paper exceeded 8 million piculs.169 The
problem would have been insoluble even if the local officials had delivered
the specified payments in full, yet no magistrate could even accomplish
a 100 per cent tax collection.170 The grants to the lower class of imperial
clansmen therefore came to depend on negotiations between local officials
and the resident princes, who had jurisdiction over the imperial clansmen
in the local districts. In the late sixteenth century there were frequent
reports of lower-ranking imperial clansmen refusing to have their children
180 The land tax—(ii) Tax administration
registered in the jade genealogy, being reduced to paupers, resorting to
armed robbery, or even conspiring with Mongolian invaders on the
frontier; some of them had not been paid for a great many years.171
Even though the enormous grants listed in the gazetteer were in part
fictitious, the administrative problems they created were all too real. They
still cut away a significant portion of the total tax revenue, wiped out any
surplus revenue in the county and prevented it being disbursed in more
constructive ways. They were also an undoubted disincentive to tax in-
creases, for no righteous official could think of flogging taxpayers to death
to extort extra revenue that would in the end benefit only the non-productive
imperial clansmen.

Distribution of tax revenue: the case of Wu-hsien

The accounts of Wu-hsien in Soochow prefecture, South Chihli, for the
year 1575 are presented here to illustrate the distribution of revenue in the
southern counties where no princes or imperial clansmen were in residence.
Tax revenues in this county were delivered to forty different receiving
agencies, the fiscal unit being the picul of grain. Some but not all of the
delivered items were commuted to silver payments. The value of some
items was distorted by conversion rates and surcharges but since each
large item listed below was subject to both higher and lower rates, the dis-
torting effects probably cancelled each other out. Table 12172 thus probably
gives a fairly accurate picture of the revenue distribution.
The high percentage of transferred revenue, 78.3 per cent of the total was
peculiar to the Yangtze delta, but the breakdown of the retained revenue

TABLE 12. Distribution of land tax revenue from Wu-hsien county, 1575:
in grain

Transferred revenue: piculs Retained revenue: piculs

to Peking, in tribute grain and to guards units in South Chihli 47,624

silver (inc. Gold Floral Silver) 64,238 to prefectural granary 4,544
to princes not in residence 1,404 stipends of government
to office of the imperial govern- students in Soochow prefecture 735
ment in Peking 2,594 stipends of government
to office of the imperial govern- students in the county 367
ment in Nanking 523 salaries of county government
to governor's offices, etc. 1,280 personnel, local charities 1,334

Total transferred 122,980 Total retained 54,604

Grand total 177,584 piculs
 4, IV Disbursement of tax income 181
was quite typical of most southern counties. The percentages of the re-
tained items were as follows:
piculs %
Maintenance of the army 47,624 87.2
Prefectural granary 4,544 8.4
Stipends of government students 1,102 2.0
Salaries and local charities 1,334 2.4
Total retained revenue 54,604 100.0
The percentage distribution of the retained revenue shows that the
contribution to the maintenance of army units dwarfed all other items of
expenditure. Though in other districts the amount was not always as high
as 87.2 per cent, it was rarely less than 50 per cent. The situation was
evidently of long standing for even in the late fifteenth century, of Chekiang's
total retained revenue of 1.3 million piculs, about 840,000 piculs were ear-
marked as rations and pay for the armed forces.173 That this continued in
the sixteenth century is shown by a personal letter written by Huo Yii-hsia
who indicated that, in 1561 when he was magistrate of Tz'u-hsi county,
30,000 piculs of its total tax quota of 37,000 were consigned to the army.174
In 1572 in Chang-chou prefecture, Fukien, there were complaints that
while only 7,000 piculs of grain were delivered to the prefectural granary,
subsidies to the Cheng-hai Guard alone cost the district 9,000 piculs.175
In around 1582, Chao Yung-hsien, a member of the Hanlin Academy, in
a memorial to the emperor declared that even so many years after the
wo-k'ou campaign Soochow prefecture was still required to deliver 55,000
piculs of grain plus 18,000 taels of silver to three local army posts, which
left 10,040 piculs of grain or the equivalent to pay the stipends of 'all the
officials, lesser functionaries, instructors, and students' of the six counties
and one subprefecture. 'If there is a famine', he added, 'the prefect and
the magistrates will be helpless.'176
Nonetheless it is doubtful whether the grain or other funds assigned
to the army were always delivered in full as specified. As early as 1480
the retained revenue in Chekiang was already insufficient to pay the
army units in that province.177 By the time the wo-k'ou campaigns began,
the guards units in the southern provinces had been reduced to skeleton
organizations (2, m) and could not have consumed the large quantities of
supplies and funds listed in the gazetteers. It was common knowledge that
tax delinquency was a much more serious problem in south than in north
China, and the repeated tax remission orders must have created local
budgetary deficits. The enormous local procurement orders that Peking
sent to the southern provinces, over which their governors haggled with
ministerial officials, must also have eaten into the retained revenue. In
short, the maintenance of the army was another appropriation in which it
might not be necessary to make up shortages.
The collection of the military supplies surtax (ping-hsiang, 3, iv) from
182 The land tax—(ii) Tax administration

A typical northern district:

Most of the retained revenue
paid to the princes and imperial
clansmen in residence

iy post

A typical southern district:

Most of the retained revenue
paid to the army posts in the
territory

FIG. 4. Diagram showing distribution of tax revenue.

the mid sixteenth century onwards was intended to close this gap. In
theory, the local districts made their regular, scheduled deliveries to the
guards units from the revenue from the ordinary land taxes, and the ping-
hsiang was then forwarded to the provincial military authorities.178 But in
some districts, including Wu-hsien, as was mentioned above, the two kinds
of appropriation were soon merged into one.179

V A FINAL ANALYSIS OF THE LAND TAX SYSTEM

Were the tax rates too high ?
In the sixteenth century the general consensus was that tax rates were too
high. Contemporary writers, memorialists, and local gazetteers all shared
the same opinion, even though they might be speaking of conditions in
different regions where the rates varied widely.
One fundamental difficulty is of course the lack of a rigid yardstick by
which to measure the level of taxation, which depended to a considerable
extent on the overall state of economic development, the aims of the ad-
 4, V A final analysis of the land tax system 183
ministration, and the extent of service that it was expected to provide, not to
mention the prevailing social values, such as concern for the protection of
the private landowner. When all these factors are taken into account it is
hard to agree with contemporary opinion.
In fact, for a huge empire which relied mainly on revenues from land,
an overall tax level of 10 per cent of agrarian output seems rather low. The
initial responsibility for this state of affairs rested with the dynastic founder.
A self-sufficient army and a system of village self-government meant that
Hung-wu had no need of a high level of taxation. He was also able to defray
his expenses by printing huge quantities of government notes. Several
gazetteer writers commented on the low level of taxation on landed pro-
perty in the early years of the dynasty.180 Once the quota system was put
into effect no fundamental revisions were ever made to it, though extra
taxes and service levies were added in a totally unplanned and insidious
manner. The central government knew nothing of regional conditions and
provincial officials had no power to act on their own authority. The demand
for uniformity in financial administration therefore continued. Even if the
Single Whip Reform did not abolish progressive taxation, it was an
acknowledgement that the principle of progressive taxation could not be
sustained. In this situation there was no possibility of imposing a high level
of taxation. Contemporary opinion viewed taxation purely as a revenue-
producing instrument, rather than as a regulating device. When it was not
used as such, the prevailing local conditions of land tenure, land leases,
tenancy, and interest rates combined to restrict taxation within its current
range, until any tax increase became unbearable to taxpayers at the lower
end of the scale. The level of taxation was therefore always regarded as high.
Ming fiscal administration was also severely handicapped by the fact
that there was no clear division between the emperor's privy purse and
public funds. In 1590 Minister of Revenue Sung Hsiin (in office from 1586)
before handing over to his successor Shih Hsing (in office until 1591)
advised the latter never to transfer any tax surpluses from the provinces to
the capital for fear that the mere knowledge of these surpluses might
stimulate the Wan-li Emperor to even greater personal extravagance.181
In Shansi and Honan, as was shown above, any tax increase was likely to
end up in the princely coffers. Many bureaucrats were so convinced that
any extra tax burden they might impose on the populace would soon be
transformed into imperial extravagance that they took a completely
negative attitude towards taxation. The Confucian tenet that the nation's
wealth should be 'preserved within the people' was taken to its literal
extreme, interpreting it to mean that any financial gain to the government
was bound to be a loss to the governed. In 1537 Minister of Works Lin
T'ing-ang was forced to retire merely because he had suggested a general
tax increase.182 Clearly, the supervising secretaries who censured him failed
184 The land tax—(ii) Tax administration
to realize that an inadequate level of taxation could be just as injurious to
people as over-collection.

The operating expenses of the local government: informal taxation

One direct result of the low level of taxation was that all local government
offices were under-staffed, their personnel under-paid, and their adminis-
trative functions inadequately carried out.
An unofficial count estimated that in the sixteenth century the Ming
government employed a total of 20,400 officials and 51,000 lesser function-
aries.183 But since the latter also served with the army, it is assumed that
only about 30 per cent of them were employed by county and prefectural
governments. This number of personnel, spread over 1,138 county offices,
meant that even the largest county would have no more than thirty salaried
positions, and a smaller one far fewer than that.
This limited manpower, which was responsible for all local adminis-
trative functions, including taxation, justice, local security, transportation,
education, public works and public welfare, was unable to carry out any
of them in other than a superficial way. It should also be remembered that
Ming officials were in addition required to perform numerous ceremonial
functions, and that by the late sixteenth century the volume of paper work
they handled was heavy even by modern standards.
No county, as has been noted, set aside more than 300 taels of silver to
cover all its annual salaries. The salary scale can also be checked from the
1578 schedule.184 The annual allowance of a prefect, who might exercise
civil jurisdiction over a million people, was 62.05 taels of silver, on which it
would have been difficult to support a small family. A county magistrate
received annually 27.49 taels which was considerably less than the sum of
36 taels collected to provide one day's rations for the emperor. In contrast,
by the end of the sixteenth century paupers hired as day laborers were
earning 0.03 taels a day185 and some mercenary soldiers were paid 18 taels
a year.186
The operating expenses of the local government were met from the
service levy which amounted to about 3,000 taels in an average county and
7,000 taels in a southern county. A few counties collected as much as
30,000 taels under this heading. Though it might be thought that these
amounts would have been ample for the purpose, in fact not all the funds
collected from the county were actually spent within the county. The
i-chuan for instance was largely used to meet the expenses of travelling
officials. Su T'ung-ping's study of the accounts of seventy-two counties
shows that on average each county spent close to 2,000 taels on this item,187
although the more impoverished districts were usually subsidized by others.
Nor was the min-chuang, or militia account, used to finance the normal
 4, V A final analysis of the land tax system 185
functions of civil administration. All that could be utilized for this purpose
were the li-chia and chiin-yao accounts, yet part of even these collections
had to be sent to the imperial and provincial governments. In fact after
paying the wages of its official attendants an average county might have
only 100 to 200 taels left over to meet office expenses. Ch'i-men county in
South Chihli, admittedly an extreme case, budgeted only 27.74 taels of
silver for this purpose.188 Ch'ii-chou prefecture in Chekiang did not have
sufficient funds even to provide for the upkeep of the prefect's official
residence.189 These circumstances were simply an encouragement to un-
authorized taxation, and in the late sixteenth century such informal
measures of taxation actually became institutionalized in many districts.
Hai Jui was able to publish a list of the 'customary payments' (ch'ang-li,
4, i) collectible in Shun-an county, Chekiang, until he himself abolished
them in 1561190 (see Appendix B). The magistrate added on extra to every
item of revenue that passed through his office; a small transit tax was even
imposed on the official salt that merchants shipped in and out of the
county. Gratuities were offered to the prefect, and the magistrate's assist-
ants and all the lesser functionaries in the office also took a regular share
of the spoils. In addition, the li chiefs in the county who served the magis-
trate in rotation could be made to pay extra office expenses. It was the
prevailing custom at the time for every // chief to pay the magistrate a fee
on appointment.191 Likewise persons hired as auxiliaries to the lesser
functionaries paid a fee for the privilege. Both Ni Yuan-lu and Ku Yen-wu
reported in the early seventeenth century that there were in each province
several thousand such auxiliaries who relied for a living on their quasi-
official status with the local government.192 The collection of melting charges
and the commissioning of tax agents have been described earlier. The low
level of taxation was, therefore, only illusory. Even European travellers
had observed that though the Chinese paid less in taxes than all others,
they were required to give much 'extraordinary and personal service'.193
Undoubtedly most of this financial burden, plus the squeeze exacted by
officials and their henchmen at all levels, had to be passed on to the general
populace. There were objections to increasing formal taxation on the
grounds that small and marginal landowners would be unable to pay, yet
the informal type of taxation, inconsistently administered, imposed the
greatest burden on precisely these lesser landowners since they had least
capacity to resist it. Thus their tax-paying power was further undercut, and
the vicious cycle went on.
It would be erroneous to assume that all Ming bureaucrats were unaware
of this situation. One military circuit intendant wrote: 'If we must add
a hidden charge, why not make it a formal one?' 194 The magistrate of
Huai-jou county in North Chihli accused those who advocated keeping
the local administrative budget small of being 'most unkind', and 'doing
186 The land tax—(ii) Tax administration
a serious wrong to our people'.195 In 1586, Chang Tung, a supervising
secretary, made the same contention in a memorial to the emperor.196

The fallacy of'*preserving the wealth within the people''

Because of the inadequate level of its revenue, the Ming government could
provide few services for its people. Even in water control, over which
Chinese governments were supposed to be actively concerned, its record in
the sixteenth century (7, i) was actually far less substantial than it appeared.
In general, water-control projects on a large scale were carried out only
after major flood disasters. The main purpose of such projects was to keep
the Grand Canal open to traffic, rather than to improve irrigation. Hsiieh
Shang-chih revealed that in his native Ch'ang-shu county in the Yangtze
delta, there were seldom as many as three good harvests in a decade. The
problem, he argued, could be solved simply by making a small tax in-
crease197 in order tofinanceimproved irrigation. This solution would have
met with the approval of Kuei Yu-kuang who, in a work on water control
in Soochow prefecture, quoted appreciatively the views of an earlier writer
who had questioned the wisdom of the government's 'go easy' policies.
Failure to raise tax revenues to launch constructive projects, he maintained,
amounted to denying people the opportunity to improve their livelihood
and to enrich themselves. Thus the supposedly 'benevolent' adminis-
trators would end up handing out charity to their starving subjects.198
The government also adopted a passive attitude to the matter of sup-
porting farm prices, which had the effect of artificially increasing tax rates.
While the transportation commissioners under the T'ang and Sung (2, i)
had utilized tax revenues to trade, the Ming policy of tax collection effec-
tively withdrew a significant portion of hard currency from circulation,
the deflationary effects of which greatly intensified the hardships of small
taxpayers. Wen-shang county in Shantung reported in 1576 that when the
tax deadline came immediately after the harvest, the price of wheat dropped
from its normal level of 0.52 taels per picul to 0.37 taels per picul, and
barley from 0.4 taels per picul to 0.25 taels per picul. It took three months
for the prices to return to their normal levels. It so happened that the
county had a service levy account that could be held in the treasury for
several months. The magistrate, by effecting a commutation in reverse,
collected the payment in grain at a conversion ratio designed to increase
grain prices by 10 per cent; the benefit went to the taxpayers. By disbursing
the same account after the prices had recovered, he was able to pay the
hired laborers draining the Grand Canal ahead of time without charging
them interest on the advance payment. From these maneuvers he made
a profit of 210 taels of silver for the treasury.199 This device was by no
means new and had in fact been widely used under the Sung. But as illus-
 4, V A final analysis of the land tax system 187
trated in this case, it was only feasible when the administrator had some
fiscal surplus at his disposal, which could only come from taxation. The
county magistrate could undertake such a venture because the service levy
account, 3,400 taels of silver in all, could be temporarily put aside. Owing
to the tax quota system and its failure to control money and credit, the
court in Peking denied itself the opportunity to provide such a public
service. Its policies benefited only the moneylenders and the pawnshops.
More significantly, the only positive effect of the low level of taxation
was to guarantee the income of the landowners; its benefits were not
passed on to the primary producers. This was pointed out by Supervising
Secretary Nien Fu in thefifteenthcentury200 and the same observation was
made by Ku Yen-wu in the seventeenth century.201 Modern scholars, when
discussing 'agrarian exploitation' in the late Ming, often point to a few
large landowners, ignoring the fact that the taxation system made exploit-
ation feasible at all levels. The land leases and contracts that have survived
confirm that rent amounting to 50-60 per cent of the main crop was
collected from very small properties, five mou or less, by landlords whose
economic and social status was not greatly different from that of their
tenants. Some household slaves were also able to invest their savings in
land, thus gradually emerging as marginal landowners.202 Above this
latter group there were innumerable small landowners. Hai Jui, who is
regarded by many modern scholars as a champion of the peasantry, dis-
closed in his letters that he himself owned property consisting of ninety-
three small pieces of land scattered over an area of several miles, on which
a total basic tax payment of 1.28 piculs was assessed. The total acreage of
the property can be worked out from the general rate of assessment on
Hai-nan island, to have been about forty mou. This kind of land tenure, as
he himself admitted, made tax administration difficult.203
Though sixteenth-century writers always showed strong sympathy for
such small and marginal landowners, little was ever said about their
tenants. The biographies of Ming bureaucrats show that a great many of
them came themselves from such small, landowning families, who in
a sense personified all the traditional Confucian virtues. Undoubtedly,
they were often able to take full advantage of the traditional channels of
social mobility, since the low level of taxation left them with farm surpluses
that could easily be further built up through frugal management.204 When
their contemporaries spoke of taxation, their sense of justice was usually
tempered by this special sense of the social value of the landowning class;
they were not concerned purely over economic justice in the modern sense.
Though in China these small and marginal landowners were never dis-
possessed by enclosure acts and turned into an urban proletariat, their
survival was bought at a heavy economic cost. The large quantity of un-
used labor in these households certainly represented a more substantial
188 The land tax—(ii) Tax administration
loss to the economy than did that of the large landowners, who were far
fewer in number. It was indeed basically because of the former that the
land tax rates could not be adequately readjusted upwards. The admini-
stration was undoubtedly handicapped not only by its inability to impose
progressive taxation on the large landowners, but also by great concern for
the interests of small-scale proprietors. Regardless of misappropriation,
taxation in traditional China simply did not generate sufficient revenue to
enable the government to carry out all its functions properly, let alone to
stimulate any economic breakthrough.
 5
The salt monopoly

It was perhaps in its management of the salt monopoly that the Ming
government demonstrated most clearly its ineptitude in business admini-
stration. The monopoly system was not expandable and for most of the
sixteenth century and early seventeenth century its projected income
stagnated at a fixed level. Even the price schedule, settled in 1535, was rarely
revised thereafter. The administrators of the monopoly never knew how
much of its revenue was derived from the excise tax and how much from
governmental control over production.
The major weakness of the monopoly was that it applied the methods
and principles suitable for governing a crude agrarian society to the
management of an ambitious business enterprise. Owing to the incom-
petence of its administrators the system proliferated into numerous sub-
systems with the result that the single commodity involved, salt, was
divided for administrative purposes into seven or eight different categories.
Even in the sixteenth century the periodic failures of the monopoly's
operations appeared in some respects to resemble modern business cycles.
The incompetence was not entirely due to ignorance, however. In the
sixteenth century, and even in the late fifteenth century, several Ming
statesmen had pointed out the inadequacies of the system and suggested
possible remedies. These suggestions were entirely ignored, for the salt
administration was only one element in the fiscal system which continued,
according to Liang Fang-chung, to be patterned on 'the Hung-wu model'. 1
Its limited capacity and lack of adaptability made any general reform
virtually impossible.

I THE MECHANISM OF THE SALT MONOPOLY

Organization at the national level
The salt monopoly had no general director. It was supervised by the
minister of revenue but no specialized central office was ever set up to take
charge of its operations. Only in 1575 did Minister of Revenue Wang
[189]
190 The salt monopoly
Kuo-kuang order that all the accounts submitted by the regional operating
agencies be processed by the Shantung bureau in the ministry; until then
reports had been handled by whichever regional bureau was appropriate
to their place of origin.2 This procedural change, nonetheless, affected only
the bookkeeping. Though the Shantung bureau from now on handled all
the paper work, its director did not become a salt administrator and overall
responsibility for the system still rested with the minister of revenue. It was
moreover still essential that the emperor's approval be obtained for all
decisions.
The operating agencies in the provinces consisted of six salt-distribution
commissions and eight distribution superintendencies (see Table 13). A
commission controlled a major productive center and a superintendency
a minor one. These offices in general did not operate across provincial
boundaries; the only exception was the Liang-Che commission whose
jurisdiction extended over Chekiang and areas of South Chihli. There
were two superintendencies in Kwangtung and four in Yunnan.3
The lack of integration in the salt administration can in part be explained
by diverse production methods, the variable quality of the salt, regional
prices, and transportation conditions. For instance, one of the superinten-
dencies in Kwangtung was on the mainland and the other on Hai-nan island;
the distance between them made coordination difficult. In the northern
sector of the Liang-Huai region, salt production was carried out by means
of the sun-drying process which required little capital investment but
yielded lower-quality cheaper salt. In the southern sector salt was obtained
by boiling the brine, a process which yielded large quantities of better-
quality salt but involved higher costs. In certain areas of Shantung, brine
was obtained in thefirstplace by washing salt-saturated soil; it then had to
be carried over twenty miles inland to be boiled, apparently because of
a shortage of fuel near the beaches. The whole process was most un-
economical. In the Ho-tung region of Shansi, salt was obtained from a lake
twenty miles long and two and a half miles wide, the waters of which were
so saturated with salt that in the summer months it crystallized naturally,
and could literally befishedout by the workers.4 In Szechuan and Yunnan
where salt was obtained by mining, the development of new wells required
heavy investment and involved considerable financial risk.5 Salt being
a bulky commodity, its price was largely determined by transportation
costs. Those areas of production which were linked by water transport to
major population centers thus had an advantage over the others. It would
have been impossible for the Ming government to work out a uniform
price structure applicable to the whole empire and to coordinate salt
production accordingly. Instead it imitated the practice of previous
dynasties and assigned to each production center its own exclusive area of
distribution, the borders of which, in general, corresponded to provincial
 5,1 The mechanism of the salt monopoly 191
boundaries.6 Carrying salt across the border became a felony. There was
thus not one uniform salt monopoly but several, operating on a non-
competitive basis. Their administration inevitably became somewhat
compartmentalized.
In theory all the salt distribution offices were subordinate to the pro-
vincial administration and in the Ta-Ming Hui-tien they are listed under the
provincial administrative commissions.7 In practice, however, it was the
office that received the proceeds from a productive center that always came
to exercise effective control over it. In the sixteenth century the central
government had considerable direct control over the Liang-Huai, Liang-
Che, Shantung, and Ch'ang-lu commissions. The Ho-tung commission
was under the joint control of the central and the provincial government.
Central government supervision over other productive centers was only
nominal. The Kwangtung superintendency was apparently subordinate to
the local prefects,8 and the Ling-chou superintendency in Shensi was
actually managed by army officers.9
The central government exercised its control by issuing official regula-
tions, either as general orders to all agencies or as specific instructions to
one only, by appointing personnel to administer the agencies, and by dis-
patching salt-control censors periodically to tour the areas of production. 10
The Liang-Huai region, which produced by far the largest revenue for
Peking, was always under close scrutiny. A residential office for the salt-
control censor was established at its major center, Yang-chou. The censor
served only a one-year term but it was rare for the appointment not to be
filled.11 Special censors were also appointed, though less regularly, to the
Liang-Che, Ho-tung, and Ch'ang-lu commissions. In the other centers of
production this surveillance was carried out by the censors in charge of the
tea and horse trade, water control, military operations and coastal defense.
It was exceptional for one official to be appointed to take charge of
several commissions. In 1560 Yen Mou-ch'ing was ordered to supervise all
the distribution commissions apart from that in Fukien and in 1568 P'ang
Shang-p'eng was put in charge of the Liang-Huai, Shantung and Ch'ang-lu
commissions.12 Their functions will be described later, but it should be
noted that they served short terms and that their offices were eliminated
after they left.
Despite the interim nature of their appointments, the censors still played
a significant role in the salt administration, for it was they who devised its
standard operating procedures. On important matters they made re-
commendations to the emperor and on minor matters they gave orders
directly to the agencies operating under their surveillance. Other salt-
control legislation was frequently proposed by the supervising secretaries in
Peking. The executive officers of the operating agencies were by comparison
rather insignificant, they simply carried out the routine without making an)'
192 The salt monopoly
notable innovations to it. Their lack of influence and prestige was felt by
Ming officials to be one of the fundamental weaknesses of the salt admini-
stration, another being the fact that the salt censors' terms of office were
too short and few ofthem had any intimate knowledge ofthe administration.13

The distribution commission

A salt distribution commission had three or four branch offices, each of
which controlled a number of production fields (yen-cttang), not more than
twelve, but seldom less than five. The salt fields on the coast occupied
a narrow strip of territory usually less than ten miles in width.14 In the
Liang-Huai region, they were cut off from the general populace by canals
and creeks and virtually formed a restricted zone. The branch office of the
commission maintained law and order in this territory, kept the waterways
open to traffic, carried out water-control projects and handed out relief to
destitute salt producers. In other words it acted as a kind of local govern-
ment. Its lowest level agency, the field office, collected salt from the saltern
households and stored it until it was dispatched. The commission also ran
a number of checking stations (p'i-yen-so), strategically located on the major
waterways leading away from the productive center. The Shantung com-
mission had only one of these, however.15 All the salt shipped from the
center of production had to be unloaded at one of these stations for official
weighing and inspection.
The commission office was well staffed by Ming standards, though it still
had too few administrative personnel for its highly dispersed organization.
The Liang-Huai commission, the largest regional office, employed about
sixty officials and slightly over a hundred lesser functionaries and runners.
Nonetheless when spread out over three branch offices, thirty field offices,
and two checking stations, the result was only one official at each place.16
The Shantung commission had no lesser functionaries in its branch offices
and clerical duties thus devolved on the administrative officials.17
The monopoly's operations hinged on the control of productive labor.
Saltern households, once officially registered, remained so in perpetuity,
and in theory were not allowed to change their profession or residence.
Each able-bodied male in such a household was assessed as a ting, and was
required in the early years of the dynasty to turn in 3,200 catties of salt
per annum; the government awarded him 1 picul of husked rice for every
400 catties.18 Saltern households were exempted from the ordinary service
levy, and were allowed to gather fuel from specially reserved 'grass land'.
It was thus expected that the salt-producing ting would expand accordingly.
This planned development did not take place in practice. In most areas the
salt-producing ting at first remained at their prescribed level and then later
declined. Production by a distribution commission or superintendency was
 5,1 The mechanism of the salt monopoly 193
bound by a fixed quota. Even in the sixteenth century the internal organi-
zation of a typical distribution commission still remained basically un-
changed from that of the early years.
Saltern households were registered once every five years and their num-
ber of ting re-assessed. About 100 or 120 households formed one group
(t'uari). The administrators appointed several village collectors (tsung-ts'ui)
from these households to serve in annual rotation, rather like the rural tax
expediters in the land tax administration (4, i). The fuan in fact bore
a striking resemblance to the ordinary li-chia.19 Occasionally imperial
decrees were issued that vacancies in the ranks of the salt-producing ting be
filled with criminals released from prisons. There were also proposals that
ordinary households be drafted as salt producers.20 Nonetheless there is no
evidence that the composition of the labor force was ever appreciably
affected by such measures.

The distribution procedure

Despite their title, the operating agencies had no transportation facilities
for salt distribution. Their stocks were either sold to wholesale dealers or
bartered for grain delivered by the salt merchants to frontier army posts.
In either case the merchants had to come to the salt fields to receive the
salt. The barter system, known as k'ai-chung, was originally developed
under the Sung. The ministry of revenue authorized the frontier military
governors to accept grain or animal fodder from the merchants in exchange
for a granary receipt or ts'ang-chiao, in the form of a stub-book. The mer-
chants presented this receipt at the salt distribution office for repayment in
salt. But the granary receipt was not directly cashable; salt could only be
handed over on the presentation of a salt license. The distribution com-
mission had to wait until it had received the stubs from the frontier and
until most of the merchants had checked in before it could proceed to
prepare the licenses.21
The license, called yin, was also a unit of weight. In other words, a license
of one yin authorized the bearer to transport 1 yin of salt. The standard
yin was 400 catties, but even in the Hung-wu reign lesser yin of 200 catties
were issued. The weight of the yin thereafter varied from one region to
another, and at different times. It was thus another expandable and con-
tractible fiscal unit. For a greater part of the sixteenth century the yin issued
in the Liang-Huai region was 550 catties, but was reduced to 430 catties
in 1616. In the Liang-Che region the yin varied from 350 to 300 catties.
Only in the area under the jurisdiction of the Ho-tung commission did the
yin remain at 200 catties.22 Official salt had at all times to be accompanied
by the yin license otherwise it automatically became contraband.
The distribution commission had no authority to issue its own yin
194 The salt monopoly
licenses. All the plates for printing licenses were kept in the office of the
Nanking ministry of revenue. For each sale therefore the commission had
to dispatch an agent to the southern capital to present the granary receipts
and the stubs to the ministerial officials so that the exact number of licenses
could be printed.23 No office was allowed to keep licenses in reserve for
future use. Most commissions distributed licenses only once a year though
the Liang-Huai commission, because of its large salt production, printed
licenses more often. From 1568 there were several proposals that the
licenses be printed in advance,24 a measure that seems to have been put into
effect in the Liang-Che region in 1574.25 Even so, the records of the Liang-
Huai region for 1616 indicate that licenses were still printed only when the
granary receipts were on hand.26
Once the licenses had been obtained from Nanking, the distribution
commission filled in the holders' names on the blanks and directed them to
one of its subordinate salt fields to encash them for salt. All salt fields were
classified in one of three categories, upper, middle, or lower. The lower
fields produced less good-quality salt and were further from the checking
stations, thus involving extra transportation costs.27 One of the salt fields
of the Shantung commission is said to have been 600 miles from its only
checking station, which presented considerable difficulty to the salt mer-
chants.28 In principle, no merchant was to receive his stock entirely from
upper or lower fields.29
The field office, when it released salt on presentation of the yin, cut off
a corner of the license. The merchant then transported the stock to the
checking station and reported to the distribution commission that he had
completed the delivery, whereupon the commission office cut off a second
corner of the license. The salt having been impounded, the merchant had
to wait until the total salt collected at the checking station reached the
prescribed level before the official inspection could be conducted. In the
Liang-Huai region in the sixteenth century the prescribed level was 85,000
yin, which by then was 45.75 million catties, or approximately 31,200 short
tons.30 When this level was reached, the distribution commission requested
the salt-control censor to authorize the inspection and weighing. The
weighing was necessary because in the sixteenth century the license holder
was also authorized to purchase directly from the salt producers a certain
amount of salt over and above the official issue (5, m). The inspectors were
usually local prefectural judges and assistant magistrates, temporarily
drafted by the salt censors. When the bags of salt were eventually weighed
and the additional duties paid, the third corner of the license was cut off
and the merchant could proceed to ship the salt to a designated port,
decided by the commission office, not the salt dealer. The overall distri-
bution followed a master plan which prescribed exactly how many yin of
salt should be consumed by the population of each prefecture.31 On arrival
 5, II Governmental control and manipulation 195
at the designated port the merchant reported to the local officials. On the
completion of his sale the license, with three of its corners already cut off,
was surrendered to the nearest prefectural or county government, whose
officials cut off the remaining corner and returned it to the ministry of
revenue to be checked against the original issue.32 Even under the most ideal
conditions it must have taken approximately two years for a merchant to
complete a transaction, and the records indicate that it could takefive,six
or even more.

II GOVERNMENTAL CONTROL AND MANIPULATION

Registered salt producers
In a society characterized by a certain degree of social mobility it was
difficult to keep the saltern households as salt producers for generation
after generation. The government aggravated the problem by defrauding
the saltern households of the grain payments due to them. On its adoption
of paper currency in the fourteenth century the payments were converted
into government notes, and thus effectively dwindled to nothing as the
value of the notes fell. By the fifteenth century the government was thus
no longer able to finance the manufacture of salt, and instead permitted
the producers to sell their surplus salt directly to the licensed merchants.
At the same time the saltern households were encouraged to reclaim public
land, the taxes on it being either exempted or reduced. By then, however,
many saltern households had already left the coast, while others had con-
verted their grass land allocation into rice paddies. In contrast to its
attitude to the hereditary military families, the Ming government never
made any attempt to force those households now living outside the salt
fields to return to their homesteads or compel them to resume the trade in
which they were legally registered.
The emigration of salt producers occurred most notably in the Liang-
Che commission, where the fertile land attracted reclamation efforts.
Accepting the situation, the government continued to register these house-
holds as salt producers but required them to pay 6 piculs of grain for each
ting instead of 3,200 catties of salt. This grain was used to subsidize the
remaining producers on the coast who in their turn incurred extra obli-
gations in making up the totalfieldquota. Even this latter arrangement was
not carried out. The remaining salt producers were effectively self-support-
ing and the grain subsidies, apparently not generally distributed even in the
earlyfifteenthcentury, were gradually cut off altogether in subsequent years.
The grain collection in fact became a poll tax and the remaining land taxes
levied on the grass land were then designated as fang-chia (marshland
payment, 3, iv). The income from both was credited towards the com-
7-2
196 The salt monopoly
mission's salt revenue. Once this trend was established it was never to be
reversed.33 In the sixteenth century, as will be seen, only a small fraction of
the income from the Liang-Che region was actually produced from the sale
of government salt.
Another development was the attrition of the ting, which took place in
all production centers except Liang-Che. In 1529 the Liang-Huai com-
mission registered some 23,100 ting, in contrast to some 36,000 ting in the
fourteenth century.34 The Shantung commission reported in 1581 that it
had about 20,000 salt-producing ting as opposed to 45,220 ting in the early
years of the dynasty.35 The lower number of ting cannot be interpreted as
an actual shrinkage in the labor force. Throughout the Ming period there
was never any manpower shortage in salt production. The lower number,
the result of under-reporting, merely meant less salt under government
control. The actual labor force engaged in production must have increased
significantly, if contemporary reports that manufacturing capacity was
steadily expanding are to be believed. Several Ming officials surmised that
the amount of salt produced in the Liang-Huai region in the late sixteenth
and early seventeenth centuries might have been three or four times that
produced in the Hung-wu period.36
Those registered salt producers that still remained in the sixteenth cen-
tury were not only fewer in number but also generally impoverished; con-
temporary sources depict them as struggling for mere subsistence. A hand-
ful of them, however, were growing affluent. Known as hao-tsao, or
'powerful saltern households',37 they hired help and induced other regi-
stered producers to work for them. Ever since the sale of surplus salt had
been permitted, such producers had learned that through efficient manage-
ment, involving both legitimate and illegitimate transactions, they could
turn the manufacture of salt into a lucrative business despite the govern-
ment take. It was not unusual in the late sixteenth century for a saltern
household to register 30 or more ting; very likely as many again of its
workers were never registered at all.
The social evolution of the salt-producing population can be detected
from the periodic ting assessments of the Shantung commission. The
45,220 ting registered in the early years of the dynasty had been drawn from
13,570 saltern households, an average of slightly more than 3 ting per
household. The 1581 report registered some 20,000 ting belonging to 2,700
households, an average of more than 7 per household.38 From the founding
of the dynasty up to 1486, only one individual from all the saltern house-
holds registered in Shantung managed to pass the provincial civil service
examination; from that date up to the end of the Ming, seventeen persons
from the same family background achieved this distinction. Furthermore,
from 1496 onwards, though saltern households were now fewer in number,
twelve of their members succeeded in qualifying for the chin-shih degree at
 5, II Governmental control and manipulation 197
the metropolitan examinations; one of these was Kao Hung-t'u (d. 1645),
who served as minister of revenue and grand-secretary under the Southern
Ming.39 It was clearly impossible for all the registered salt producers to
remain manual laborers.
There is thus some evidence that a new 'gentry class' was rising among
the mass of the saltern households. This social mobility, though desirable
in itself, was nevertheless a serious handicap to the salt administration,
because the monopoly system was entirely dependent on the direct ex-
ploitation of statutory labor. The emergence of a new entrepreneurial
group among the registered producers themselves tended to cut potential
government income. It also provided stiff competition for the independent
producers in that the latter also had to sell their surplus salt for a living.
The government, in collecting salt from the ting, had to fix the flat rate
payment according to the capacity of the independent producers, who
were by now largely destitute. The payment, partially converted to silver,
was thus steadily reduced throughout the sixteenth century.
Though the rates varied from one center of production to another, and
sometimes even from one field to another, they were never anywhere near
the 3,200 catties of salt that had been assessed in the early years of the
dynasty.* In 1581 the ting assessment in the salt fields of Shanghai county,
subordinate to the Liang-Che commission, was 1.6 taels of silver.40 In most
of the Shantung salt fields, it was 0.6 taels. In both the Ch'ang-lu and
Shantung centers, some of the registered producers actually paid in cotton
cloth and peas,41 though in the official records the payments were re-stated
as so many yin of salt so that they could be integrated into the general
accounts. The same thing happened with the grain payments exacted from
the former producers who had turned landowners. In Jen-ho county, under
the Liang-Che commission, the ting payment dwindled from the original
6 piculs of grainf to 1.44 taels of silver, and finally settled at 0.333 taels.42
Thus the land tax exemptions of the nominal saltern households far ex-
ceeded their ting payments, which aroused the indignation of many local
gazetteer writers, some of whom asserted that the rents which these nominal
salt producers collected from the grass land were by themselves sufficient
to cover the ting assessments. This explains the unusual increase in the
number of ting in the Liang-Che region. The 74,446 ting registered here in
the Hung-wu period expanded to 165,574 by the late sixteenth century:
more than the total ting of all the other commissions combined.43 A high
percentage of these ting was claimed by the non-producers who had found

* In the southern sector of the Liang-Huai region in the late sixteenth century 3,200
catties of salt would have been worth about 5 or 6 taels of silver, when sold directly by
the producers to the licensed merchants.
t 6 piculs of husked rice was normally worth 3.6 taels of silver in Chekiang in the late
sixteenth century.
198 The salt monopoly
that the advantage of being assessed with more ting was that the accom-
panying poll-tax entitled them to a large land tax exemption.44
The amount of salt submitted by each ting actually working in the field
is not accurately known. A state paper dated 1567 indicates that in the
Liang-Huai region where the ting were graded, their payments ranged
from 600 catties to 4,600 catties of salt.45 There is much evidence that
actual collection fell a long way short of this desired standard.
The takeover of the salt-producing labor force by commercial interests
was a later development; Ping-ti Ho has dated the emergence of 'factory
merchants' in the Liang-Huai region to the mid-Ch'ing period.46 There is,
however, some evidence that even before 1600, salt merchants were making
inroads into the actual manufacturing process by making loans to the salt
producers. 47 Even the official records of the Shantung commission suggest
that some merchants, apparently using the registered producers as a cover
actually hired outside workers and engaged in production themselves.48
In the Ho-tung commission, where the gathering of the partially crystal-
lized salt from the lake water required little capital investment, the adminis-
trators likewise found that the use of a statutory labor force was inefficient.
There the operation was seasonal. The salt dissolved once the warm weather
was over and rain, an unpredictable factor, could also cause complications.
The 17,548 ting had to be called from twelve different subprefectures and
counties49 and it was difficult to assemble all of them in time for the annual
'harvest'. Moreover they worked so inefficiently that one salt-control
censor asserted that one hired worker was 'worth more than 10 ting9.50

Control over production

In the regions where salt was produced by the sun-drying process the only
way to control production was to police the beaches. The drafted salt
patrols, however, frequently conspired with the producers in manufacturing
contraband salt and sharing the profit.51 In the late Ming most of them
were required to seize a fixed quota of contraband (3, m), an irrational
regulation which reveals how ineffective the control must have been.
In the areas where salt was obtained by boiling and evaporating the
brine, the government attempted to exercise control over the equipment
needed. Iron plates were officially issued for this purpose. Those used in the
Liang-Huai region were huge; in 1527 they were said to weigh 3,000 catties
and to cost 26 taels of silver a piece. They seem to have been widely used in
the early years of the dynasty, as the 1527 report indicated that by that
time 321 of them had been damaged.52
These plates probably measured anything from 200 to 400 square feet
and were inconvenient to use, easily-cracked and too costly to replace. The
salt obtained in this way was said to be black in color and inferior to that
 5, II Governmental control and manipulation 199
produced in smaller caldrons. It seems that after 1552 few salt plates were
actually replaced as the registered saltern households had by then pre-
vailed on the authorities to bring in several caldron manufactures from the
south of the Yangtze; these operated shops at Pei-t'a-ho, a riverside town
next to the salt fields. P'ang Shang-p'eng, who in 1568 was in charge of
three salt commissions, argued that the presence of caldron manufacturers
in a salt-producing district encouraged the production of contraband salt,
as they made the necessary equipment available to everyone. In a memorial
to the emperor, P'ang revealed that some saltern households possessed as
many as ten caldrons, each of which could produce 200 catties of salt in
twenty-four hours. Such a manufacturer was thus capable of producing
400 short tons annually, bringing in an income of approximately 1,200 taels
of silver. It was these large households, P'ang concluded, who were the
major source of contraband salt. He then proposed to close the caldron
shops, impose controls on the manufacturers of salt-producing equipment,
mark all existing caldrons and register all the masonry tanks constructed
to store brine. Unauthorized tanks were to be sealed off.53 This proposal,
even if carried out, apparently did not have much effect, as later writings
indicate that conditions in the salt fields were much the same as those
described by P'ang. The use of large iron plates was never revived. The
whole episode does show, however, that officialdom would not hesitate to
resort to the most restrictive means in order to protect the monopoly.
In the Ch'ang-lu region, the authorities issued the saltern households
with caldrons of several different sizes, though their actual dimensions are
not disclosed in the sources. A suggestion was made at one time that the
required salt payment be assessed in accordance with the size of the
caldrons,54 but there is no evidence that it was ever put into effect.
Proposals of this kind were impractical because the government salt was
collected from individual ting. If the necessary level of revenue was to
be maintained it was therefore essential to support the small-scale, inde-
pendent producers. In the late sixteenth century these small producers
evaporated their brine using bamboo trays pasted over with paper, and fire-
proofed with alkalis, a crude means of production that illustrates the extent
of their destitution.55 The government failed to aid them. Nor did it make
any attempt to limit manufacturing by the larger households. Even though
these households produced contraband salt, they were always assessed
a fixed number of ting and thus contributed to government revenues. To
restrict their production would have narrowed down the source of income
even further. Yet assessment on the basis of productivity was quite a differ-
ent concept, and one that would have required new legislation and organi-
zational changes. The government in the late Ming was in no position to
take such a drastic step. Even rights to the official grass land, which the
authorities were obliged to protect for the sake of the smaller producers,
200 The salt monopoly
were by this time monopolized by the large households. They converted
some of these weed-grown properties into agricultural land and simply
took possession of the remainder, denying others the right of entry.56
Salt manufacture was basically a rural industry and since the government
exercised so little control in the countryside it was unable to exploit all
possible sources of revenue. The salt commissions, treating large and
small producers alike, were not doing any better than the local officials
collecting land taxes, who had to deal with both large and small landowners.
Only the Ho-tung commission seems to have exercised any effective
control over production. The salt lake there, ever since Sung times, had
been surrounded by a wall, which in 1474 was rebuilt by one of the salt-
control censors. It was 13 feet high, over 45 miles in circumference and
broken only by well-guarded gates. In 1476 another censor increased the
height of the wall to 21 feet. When the Ch'ing took over they found it still
intact.57

The role of the salt merchants

In principle the salt revenue was to be used only to subsidize the frontier
army posts. To overcome the problem of transporting such a large amount
of revenue from the interior to the frontier regions the government devised
the barter system, whose operations necessitated enlisting the services of
merchants.
The Ming government never proclaimed any general policy in its
dealings with the salt merchants, nor did the salt offices publish any guide-
lines. Most of the detailed procedures were established by individual
officials on the basis of current needs and circumstances. Their general
practices did fall into established patterns.
Though officials conducted business with merchants, they never felt
that the government was committed to any contractual relationship with
them. Underlying the relationship was the concept that the state was
supreme and that individual subjects were obliged to perform services to it.
Businessmen were expected to make a profit, and their voluntary partici-
pation in government activity was most desirable. Nonetheless, when there
were no profits to be made and no volunteers to handle the salt, officials
felt completely justified in drafting merchants to perform the duty, just as
they called for services from the general population. On certain occasions
merchants were actually expected to sustain a loss in doing business with
the government; they probably viewed such losses in a sense as fees for the
privilege of staying in business.
Prices, delivery procedures, deadlines, and penalties for non-fulfillment
of obligations were all unilaterally determined by the government. All
matters of any significance were decided in Peking with the emperor's
approval, though regional administrators and censorial officials usually
 5, II Governmental control and manipulation 201
made recommendations. Sometimes the salt merchants were consulted
before these recommendations were submitted, but they never had any
opportunity to bargain with government offices. Tenders from merchants
to purchase government salt at specified prices were frowned on. In 1518
and 1526, groups of salt merchants did contrive to have such offers
accepted by the court, large transactions being involved on both occasions.58
It seems, however, that they were influential financiers who had powerful
connections in Peking and were thus able to win imperial favor for their
proposals, to the chagrin of the bureaucrats. On both occasions the mini-
stry of revenue in fact protested, demanding that such merchants be
arrested and punished.
The government was under no obligation to pay indemnities to the
merchants when it failed to carry out its promises, though it did on rare
occasions distribute small consolation payments to salt dealers who had
suffered serious losses. When the terms of the transaction were too favor-
able to the merchants, the salt officials used this as a pretext to tax their
profits later. In 1527, for instance, one frontier governor, considering that
the rate of barter decided in Peking was too favorable to the merchants,
imposed a 'fine' on them. In this particular case nevertheless the salt
merchants (aided by censorial officials), appealed successfully to the court
so that the fine was subsequently refunded.59 In short, the treatment that
the merchants received depended on the sense of fair play of the officials.
Though merchants might at times befloggedby the frontier governors and
forced into business transactions with the government60 there were also
occasions when they found the terms offered quite lucrative, and competed
for the opportunity. It is true that when merchants were ill-treated the
censorial officials were obliged to protest but they did so more out of
a feeling that harsh treatment of imperial subjects conflicted with the ideal
of governmental benevolence, rather than out of a concern for individual
injustice.
It would be a mistake to assume that merchants were discouraged by
such arbitrariness on the part of the government or that all of them suffered
from it. Like the saltern households, the salt dealers were also from a wide
range of backgrounds. Some of them had little capital and could un-
doubtedly be easily forced into bankruptcy. But there were many others
who built up considerablefinancialstrength through continuous operation.
The very unpredictability of governmental regulations in itself created
numerous opportunities for making quick profits. Corruptible officials
could always be bribed; honest competition was in fact the exception
rather than the rule. Frequent alterations in procedure inevitably resulted
in many loopholes of which shrewd merchants were quick to take advantage.
While Confucian bureaucrats were generally prejudiced against mercantile
interests, they also realized that it would be injudicious to ignore the
202 The salt monopoly
interests of the salt dealers completely, because without them the monopoly
system would be unworkable. The salt administration throughout the
sixteenth century was tending to become a sort of oligopoly in which a few
salt merchants benefited at the expense of the rest. By the end of the cen-
tury this was quite clear. In 1616 the ministry of revenue dispatched an
official to investigate the administration of the Liang-Huai commission.
The investigator reported back to the minister, Li Nii-hua, that within each
group of salt traders there were always' several' operators who had amassed
large fortunes by manipulating governmental regulations to their own
advantage. 'If we were to subject them to the full vigors of the law', he
said, 'they would have no defense.' Yet in his final analysis he concluded
that to weed out these few merchants would mean the ruin of the salt
monopoly.61

Delivery priorities and deficit financing up to 1500

Ming writers give the impression that the salt administration reached per-
fection in the early years of the dynasty and that decline set in during its
middle period. This generalization is not entirely invalid but it does not
reveal the whole story. During the early period the regulating powers of the
central government were more effective and the organizational weaknesses
of the system, though steadily worsening, had not yet come to light.
Nonetheless all the unsound features of the salt monopoly system existed
from the beginning. These consisted basically of lack of support for the
salt workers, inefficient and widely dispersed administrative offices, in-
adequate distribution facilities, and the compulsory services demanded of
the merchants. In essence the operation was inadequately financed and
serviced by the government.
Defaulting on salt payments to the merchants in these barter trans-
actions started early. As early as 1429, there existed salt licenses issued
twenty-seven years earlier which had not yet been cashed.62 Thereafter the
government habitually failed to honor its promissory notes. In the fifteenth
century delays of over thirty years were common.63 An imperial decree of
1488, admitting that some salt merchants had died by the time the trans-
action was to take place, pronounced: 'If the salt license holder is
deceased the salt payment may be received by his parents, brothers living
in the same household and sharing a common kitchen, his wife (providing
she has not remarried), or his sons or grandsons (providing they are not
adopted) upon verification by the authorities. Uncles, nephews, and con-
cubines are not to be admitted as beneficiary payees.'64 This ruling is an
indication of the tone of the administration.
Some of this constant delay was due to the fact that the salt monopoly
was burdened with deficit financing. Since the government had no regular
 5, II Governmental control and manipulation 203
means of obtaining credit, the disposal of official salt became a convenient
device for receiving advance payments, though the fact of advance sale was
never explicitly admitted. It was accomplished by postponing current
delivery obligations, offering the currently available stock for sale at
a higher price, giving new buyers priority over previous purchasers in
cashing their promissory notes and delaying the delivery dates of earlier
buyers. These procedures were first adopted in 1440. The annual salt
production was classified into two categories, 80 per cent being designated
a s ' regular stock' (cKang-ku yen), and the remaining 20 per cent a s ' reserve
stock' (ts'un-chiyeri).65 The former was for normal circulation and the latter
ostensibly to be stockpiled for emergency use, such as urgent military
requirements. But no sooner had this classification been created than the
reserve stock was also made available for bartering. Because its delivery
involved no waiting, the reserve stock appeared more attractive to the
dealers and its sale was an immediate success. In 1449 the court increased the
reserve stock to 60 per cent and reduced the regular stock to 40 per cent.66
What the government in Peking failed to realize was that the adverse
effects of these actions on later operations far exceeded their immediate
benefits. The merchants, who had to take into account the loss of interest
on the funds thus tied up, could no longer absorb the regular stock at
normal prices or at the regular barter rate. Moreover when the reserve
stock came to constitute a large percentage of the annual production, its
buyers soon discovered that there was little advantage in it. There were
still delays at the production centers because the distribution offices could
rarely collect sufficient salt from the producers to fulfill their quotas; the
saltern households remained unpaid and the number of ting dwindled.
Theoretically, there was no market price for government salt. As long as
all the salt was produced under the monopoly system its administrators
could name their own prices. But in reality the higher the price the greater
the trade in contraband salt. In the end the officials were compelled
to lower the price of government salt, which resulted in a reduction of
revenue.
Lack of coordination between different offices created further unnecessary
difficulties for the salt merchants. It often happened that merchants hold-
ing granary receipts for regular stock would be told at the production
center that only reserve stock was available. Sometimes the situation was
reversed; though the dealer might be entitled to high-priority reserve
stock the salt field had received instructions to release its salt only as
low-priority regular stock and not to convert it to the other kind.67 In
1471 a salt-control censor touring the Liang-Che region reported that
there was a waiting period of over ten years even for reserve stock.68
While the salt dealers were waiting, it frequently happened that the salt
stocks began to decompose, to the advantage of contraband dealers.69
204 The salt monopoly
Throughout the greater part of the fifteenth century, in theory all the
government salt was collected from saltern households and released by the
distribution agencies to the merchants under the terms of the barter agree-
ment. Actual practice was different, however. The salt field seldom had as
much salt on hand as it claimed. Many salt merchants were not in the least
annoyed to hear that their granary receipts could not be honored. What
they wanted were the salt licenses. Armed with these papers they could
make purchases directly from the producers. Though strictly speaking it
was a kind of contraband, this salt was legitimized by the licenses. The extra
payment was still less than the interest accumulated over such a long wait-
ing period. The distribution centers of course welcomed such action by the
merchants as it meant they would have more deliveries to their credit on
the official record. Sometimes they induced the merchants to pay a small
fee in the name of charity to the saltern households and secured a portion
of official salt for them.70 Another practice that salt dealers frequently re-
sorted to was to bribe the officials to keep the salt licenses intact so that
they could be re-used. The court was fully aware of such abuses and an
edict issued by the Hung-chich Emperor in 1488 enumerated them all in
detail.71
In the late fifteenth century bartering for government salt remained
profitable only for those merchants who had the capital, patience, connec-
tions, and ingenuity to maneuver the situation to their advantage. When
a transaction was announced at a frontier army post, some merchants were
willing to pay a fee to the intermediary agents, usually friends and relatives
of the governor, to secure the trading privilege.72 Eventually all these extra
costs were passed back to the government, which had to reduce the price
of salt. In 1476 on the nothern frontier a standard yin of salt had a barter
value of only 0.15 taels of silver.73 In 1474 a 200-catty yin from the Ho-tung
commission had a barter value of only 0.05 taels, clearly less than the cost
of producing it.74
In 1489 it was decreed that the barter system be temporarily suspended
and that all salt be sold for cash at the salt fields. When the distribution
center was unable to supply the authorized amount, the merchants were
allowed to make it up by purchasing directly from the producers. In sub-
sequent years when Yeh Ch'i was minister of revenue (1491-96), he followed
the same policies.75 Under his administration the average price of the
200-catty yin of salt was raised to 0.4 taels. The salt administration there-
fore entered the sixteenth century in reasonably good order.
 5, III Administrative cycles in the sixteenth century 205

III ADMINISTRATIVE CYCLES IN THE

SIXTEENTH CENTURY
The blocking of the yin
The stability achieved by the end of the fifteenth century was largely illu-
sory. The fundamental problems of the salt administration, involving
organization, finance, production control, policing of contraband activities,
and the registration of saltern households remained unchanged. The struc-
ture of the monopoly system was fragile, it was exceedingly vulnerable to
abuse and its maladies were recurrent.
Its operations throughout the sixteenth century fell into recurrent ad-
ministrative cycles. In the more serious cases of failure, as occurred in the
1520s, 1560s and around 1600, the whole system was near to complete
breakdown.76 At such times official salt was in short supply in the remoter
provinces yet over-stocked in the salt fields and as a result revenues were
completely cut off. Contemporary writers describing the situation declared
that 'the yin was blocked', and attributed the crises to moral decay and the
machinations of evil men; a remedy could only be provided by virtuous
officials. From an objective standpoint, though it may be true that the
crises were triggered off by unethical conduct, the moral issues were always
of far less importance than the basic institutional weaknesses.
The major weakness was in the financing of the system. It was generally
recognized by its administrators that production in the salt fields actually
exceeded the official quota and that part of it was circulated as contraband.
Yet there was no way to put all this salt under governmental control other
than by financing its manufacture directly. At times the salt officials were
tempted to create a new category of salt outside the regular quota, to be
financed by other methods. Dividing up officially-marketable salt into
different categories inevitably produced disparities in the system. The
financial backing for the new category enabled it to compete for supplies
and markets on more favorable terms than the regular quota, which tended
to be driven out of existence. The wider profit margins of the new category
also made it extremely attractive to influential persons. All this led to
greater and greater abuse of the system and the breakdown of the monopoly.
In the sixteenth century the market for quota salt, that is, the salt which
was distributed through the yin licenses, was considerably narrowed.
Increased dealing in contraband salt went on in districts adjacent to the
areas of production, in mountainous regions where government salt could
not compete with it, and in the border areas between the distribution terri-
tories assigned to two different centers of production. In recognition of
this situation the government institutionalized the so-called 'ticket salt'
(p'iao-yeri), which was to be circulated exclusively in these districts. A ticket
206 The salt monopoly
was no more than an informal license issued by a commission or a super-
intendency on an ad hoc basis for a fee, bypassing all the normal procedures.
It is not known how early this system started but it was definitely recog-
nized by the court from 1529.77 Though it was a limited operation designed
to combat the contraband salt and to produce only a small income, the
ticket salt nevertheless represented more competition for the licensed salt
and further reduced its outlet.
Bottlenecks at the checking stations were another serious handicap to the
system. From the administrators' point of view it was necessary to release
the wholesale salt en bloc to prevent the salt merchants competing with
each other and causing price fluctuations. It was impossible, however, to
speed up this procedure in an effort to meet supply and demand conditions.
The ways in which the merchants reacted to it make this clear.
Only twenty years or so after Yeh Ch'i left office in 1496, the barter
system was revived. Its proponents argued that the system stimulated
agricultural production in the frontier regions. As long as the merchants
had to transport grain to the frontier army posts in order to obtain salt
from the interior, they maintained farms in the remote north, the pro-
duction of which helped to stabilize local grain prices. There is no actual
evidence that the policy was ever effective, but the arguments sounded
convincing enough to keep the barter system alive until the end of the
dynasty. The salt merchants therefore adopted various measures to cope
with it, their basic solution being trade specialization.78
From the late fifteenth century two kinds of salt merchants emerged. The
frontier trader supplied grain to the army depots and subsequently sold his
granary receipts to the residential merchants at the center of production.
The latter applied for the yin license and obtained the salt. This division of
function was necessitated by the fact that even in the sixteenth century,
when the service of government agencies had improved, the waiting period
for the salt was still normally at least three years and often nearer eight or
nine years. The frontier trader certainly could not make regular deliveries
of grain and at the same time wait for the salt to be issued. The residential
merchant took on in effect the dual function of a financier and an export
agent. Eventually the residential merchants also ceased to distribute salt
to the inland market. As soon as their stock was released from the checking
station they sold it to another group of merchants known as inland dealers.79
The government's efforts to regulate the trading between these different
kinds of salt merchants were unavailing. Any disputes among them over
the splitting of the profits could deadlock the entire operation which could
only work with all its components intact.80 Official delays created un-
certainties which enabled one group of merchants to gain at the expense of
the others. The residential merchants enjoyed an advantage here because of
theirfinancialstrength. As the only available buyers, they could dictate the
 5, III Administrative cycles in the sixteenth century 207
price of granary receipts to the frontier traders; as the only suppliers of salt
they had a similar hold over the distributing merchants, or inland dealers.
In the sixteenth century, strangely enough, the merchants instead of com-
plaining about the incessant delays in the issue of government salt, some-
times actually sought to prolong them further. The inland dealers, in
particular, were anxious to wait for retail prices to increase. There were
reports that they bribed the administrative officials to slow down the
processing of the salt, and that they even paid interest to the residential
merchants for the delay, expecting to recoup it through selling the salt at
higher prices.81
The crisis of the 1520s
The crisis of the 1520s can be traced back to the decision made in 1489 to
allow producers to sell salt privately to the merchants. Ostensibly this did
not harm the monopoly system but merely made up for the government's
defaulting on official salt payments. In practice, however, it meant that
the merchants paid twice over for the same stock, once to the government
and again to the salt producers. The authorization was intended merely to
recognize the existence of an established practice but actually once private
sale received imperial approval it began to be greatly extended in scope.
Soon the salt dealers were overloading their carriers with surplus salt
(yu-yeri) over and above the licensed limit. The officials at the checking
stations simply collected a payment on it known as 'surplus salt money'
(yii-yen yiri), which possibly originated in the fifteenth century as a fine, but
in the sixteenth century became an excise tax.82 Theoretically the surplus
salt was the amount that the registered saltern households produced extra
to official requirements and it was sold after they had submitted their
regular payments to the field offices in fulfillment of their service obli-
gations. In practice, as might have been expected, much of the total pro-
duction began to be siphoned off through this new outlet. The fact that the
salt was in this way obtained by purchase made it much more efficient than
the official system, in which the salt was collected as a poll tax.
Because the surplus salt was now competing with the quota salt it meant
that the stability attained by the system in 1500 was precarious indeed.
In 1503, the emperor's in-laws came into the picture with a request that
they be allowed to finance the manufacture of salt. They undertook, if
granted salt licenses, to purchase surplus salt from the producers in order
to make up the accumulated arrears in government salt payments. This,
they claimed, would be no burden to public finance as they would actually
be paying the state for the privilege. The emperor, perhaps unaware of the
implications of the proposal, readily approved it.83 Under his successor,
the Cheng-te Emperor, such authorization was made frequently. The
emperor himself, realizing that the sale of salt licenses was a lucrative
208 The salt monopoly
source of revenue, used them tofinancehis own personal projects, such as
the manufacture of imperial satins and the dispatching of a eunuch mission
to Tibet in search of the living Buddha. It appears that once the licenses
were secured, their influential bearers could use the authorization over and
over again, sometimes using the same license over many years.84 The
difference between official salt and contraband thus became blurred. In
1508 and 1514 the government made unsuccessful attempts to attract
legitimate merchants by steadily lowering the price of official salt.85 This
led to the 'blocking of the yin' in the 1520s.
After the crisis the government tightened up the control of licenses and
also established rigid rules to govern the purchase of surplus salt. On each
yin license was now entered a fixed amount of both official and surplus
salt. The salt merchant was to obtain the former by purchase or barter
from the government and then to secure the latter from the producers on
his own initiative. No-one was allowed to make private purchases without
first trading with the government, but on the other hand no-one who traded
with the government could ship the official salt away without making
a private purchase. The two kinds of salt actually had to be packed in the
same bag in order to be cleared at the checking station.86 The compulsory
inclusion of the surplus salt was designed to give the saltern households
a legitimate outlet for their extra production and to prevent it falling into
the hands of contraband dealers. The complication was that the salt
merchant had to make three kinds of payment. First he had to obtain the
granary receipt by barter or purchase in order to acquire the license and the
official salt. Next he had to make the additional purchase of surplus salt,
and thirdly he had to pay in the surplus salt money, essentially an excise
tax, at the checking station. This procedure remained in force right up to
the end of the dynasty.

The crises of the 1560s and 1600s

The man generally held responsible for the crisis of the 1560s was Yen
Mou-ch'ing, an associate of the notorious Grand-Secretary Yen Sung (in
office, 1542-62), and described as extravagant and corrupt.87 In the official
documents on his mismanagement of the salt monopoly there is, however,
no mention of his personal integrity, or lack of it, with respect to pecuniary
matters.88
When Yen Mou-ch'ing was salt-control censor in charge of five distri-
bution commissions, from 1560 to 1562, the Ming court was still in des-
perate need of funds to cope with the raids of Altan Khan on the northern
frontier and the piracy problem in the south. Yen therefore made un-
precedented demands on the monopoly system. The saltern households
were dunned for their payments, the salt patrols were required to produce
 5, III Administrative cycles in the sixteenth century 209
fixed quotas of contraband, and the merchants were pressed to make com-
pulsory purchases. Yen thus produced an unusually large amount of salt
without making price readjustments, and in one year was able to squeeze
2 million taels of silver from the five salt commissions under his super-
vision. After he left the whole system was paralyzed. When P'ang Shang-
p'eng took over in 1568 he found that over 5 million yin of salt were still
impounded awaiting final clearance by the Liang-Huai commission, a
process that was expected to take four years.89
Yen's critics also held him responsible for the introduction of the so-
called 'cost-paid salt' (kung-peng yen), which had actually been created in
1553, seven years before he took office. Under this system the Liang-Huai
commission allocated 82,000 taels of silver to purchase additional salt from
the saltern households, at an average price of 0.2 taels per yin. This salt
was then sold publicly, surplus salt and excise tax being added to each yin.
A gross profit of 300,000 taels was anticipated on this transaction. After
Yen left office, the cost-paid salt was condemned as another of his oppres-
sive measures and was eliminated in 1565.90
Yen's failure cannot be attributed to any single cause. Over-production
was certainly not the only reason for it. In 1550 the ministry of revenue
estimated that the government in fact took only 40 per cent of the total
Liang-Huai production and that the other 60 per cent had fallen into the
hands of contraband dealers. In 1558 the governor of Kiangsi reported
that in three prefectures in his province the only salt available was contra-
band.91 Even P'ang Shang-p'eng, on taking over, complained that the
major problem was illegal sale. If the government wanted to widen the
circulation of official salt, there should thus have been a large market for it.
P'ang's own writings do hint at the true explanation for this. It was not
that there was no scope for expanding the monopoly but that expansion
was prevented by the nature of its management. Most of its services were
already operating under extreme pressure and had reached their limit.
Every additional pressure exerted from the top was thus passed on down-
wards to be absorbed ultimately by the weakest element in the whole
system, the poor saltern households. In a memorial to the emperor, P'ang
admitted that even in regular salt collection the village collectors (tsung-
ts'ui) were extracting extras from the producers.92 There is no evidence
either that the payments allocated for the cost-paid salt were ever entirely
distributed to them. The salt merchants who purchased surplus salt from
the salterns sometimes also victimized the producers.93 On the other hand
when the merchants were weak, they too fell victims to the village collec-
tors.94 The nature of the taxation system thus made it impossible to have
any regular market for legitimate salt in the rural areas.
P'ang further revealed in a letter that the local officials appointed as salt
inspectors at the checking stations sometimes took as long as two months
210 The salt monopoly
to report for duty and that he was attempting to cut this delay to three
days.95 To speed up the issuing of salt from the production centers, he secured
the emperor's permission to have some consignments of salt inspected on
the boats bypassing the checking stations. Yet all these measures had only
limited effects. Issuing too much salt from the production centers could
depress its retail price, which the government was obliged to protect. Any
fall in prices moreover affected its future revenue by causing further
'blocking of the yin\9Q
The price of official salt was determined by the barter rates, the purchase
price of the surplus salt, and the excise tax known as surplus salt money,
the last item being a substantial one. None of these could be reduced. The
barter rates and excise tax were already being counted as budgetary items
and therefore appropriated in advance. The price paid for the surplus salt
constituted the only legitimate income of the primary producers. Its
reduction would therefore compel the saltern households to enter the
contraband trade and cause a scarcity of legal salt. Though P'ang did not
mention the interest lost during the period of waiting for the issue of govern-
ment salt, these too obviously affected the final price of the salt. As P'eng
Hsin-wei has shown, under the Ming the minimum monthly interest
charged was 2 per cent.97
There was little hope of the official salt cutting into the contraband
market because its price could not compete. By the late sixteenth century
the contraband trade was firmly established in certain regions and the
contraband quotas assigned to the salt patrols were in reality a rather hap-
hazardly administered variety of excise tax. Under these conditions the salt
administration could no longer force the merchants to trade in these areas.98
It should be noted that even when P'ang was making vigorous efforts to
clear the existing stockpile, in general salt was still issued from the pro-
duction areas in accordance with the established annual quotas. Minor
readjustments were made only when these would not affect the desired
level of retail prices in the consuming districts. It was therefore necessary
to limit the new system of inspecting the salt aboard the barges because the
more tonnage went through this channel the longer the salt impounded at
the checking station had to wait to be released. P'ang feared that some of
this stock might have to wait eight years instead of the originally estimated
four.99
The salt merchants made advance payment in grain for the government
salt, that is, for the 5 million yin impounded at the checking station and the
other million yin on its way to the station from the salt fields. Before the
stock was released from the checking station they had to pay the excise.
Some salt merchants could raise the additional funds needed to pay the
tax immediately, while others had to wait for the completion of one trans-
action and the release of their capital before the excise on the next con-
 5, III Administrative cycles in the sixteenth century 211
signment could be paid. Considerable amounts of mercantile capital were
tied up in the salt held at the checking stations.100 It is also evident that
money supply was a problem. All excise payments had to be made in cash.
Demanding that the small merchants pay the excise immediately would
force them to borrow at exorbitant interest rates, which might ruin them
completely. The government never financed the distribution of the salt and
never contemplated doing so. It could not raise the necessary capital and
had no effective way of obtaining credit.
One frontier governor and some salt merchants did suggest that the
government enlarge the quota of salt that it released, but P'ang Shang-
p'eng objected to this on the grounds that it would lower the retail price of
salt.101 The increased production quota would then become a permanent
burden to the distribution commission and lead to the recurrence of all the
problems of production and procurement that have already been described.
In short, the monopoly system, hampered as it was by inadequate
government investment, and by continual mismanagement, was quite un-
able to utilize effectively all the resources available to it. Although these
were so abundant as to create the illusion of unlimited development
potential, the system had little capacity to exploit them. Inexperienced
administrators, tempted by the seeming possibilities, sometimes tried to
force the system to yield larger revenues, but these attempts were only
successful in the short term and always led to a complete breakdown of the
system in the long run. This was the basic reason for the salt monopoly's
administrative cycles.
It was also necessary to treat government salt as a single category. Any
attempt to classify it into different categories and to establish priorities
among them tended to create imbalances which gradually impaired the
workings of the entire system. Even P'ang Shang-p'eng, in most respects
a brilliant administrator, erred on this last count. In order to expedite the
process of inspection he allowed certain consignments of salt to clear the
checking station without the usual formalities, a practice which became
customary after his departure. The new category of salt, known as 'river
salt' {ho-yeri), subsequently provided stiff competition for the usual
'dockyard salt' (tui-yeri).102 By the 1570s merchants dealing in the latter
were in such financial straits as to bring about a minor crisis in the salt
administration. The situation was serious enough to be reported to the
emperor, and in consequence the river salt was apparently discontinued
after 1578.103 The basic principle underlying the salt administration, that
the more efficient elements in the system should give way to the less
efficient in the interests of parity, applied in fact to the entire fiscal admini-
stration of the Ming.
The last sixteenth-century salt administrator to act in defiance of this
principle was a eunuch, Lu Pao, who in 1598 was commissioned by the
212 The salt monopoly
Wan-li Emperor to take charge of the Liang-Haui commission. Lu was
quite successful in the early years of his appointment; by reviving the
category of high-priority reserve stock and raising the regular quota, he
managed to increase the salt revenues considerably.104 By 1606, however,
the over-worked machinery once again ground to a halt, resulting in
a sharp rise in retail prices in the remoter provinces, excessive stockpiling
of salt at the centers of production, and a severe curtailment of revenue
(see the following section).

IV REVENUES, SALT PRICES, AND

THEIR EFFECTS ON CONSUMERS
Barter rates and excise schedules
An important turning point in the history of the salt administration came
in 1535, when it was decreed that the official salt, disposed of under the
barter system, and the surplus salt, on which an excise duty was collected,
should both be packed together in the same bag, in order to prevent the
latter from competing with the former. The grain thus obtained by barter
was permanently allocated to the army posts on the frontier, as part of the
annual payments due to them from the imperial government. The proceeds
from the excise duty were also delivered to these army posts. But as the
ministry of revenue wished to retain control over them in the interests of
flexibility and as a source of emergency funds, they went through to
Peking. This formula persisted thereafter. The complex nature of the salt
allocation can be illustrated from the case of the Liang-Huai region.105
Each yin license entitled its holder to the following:
250 catties of government salt, to be issued to the merchant at the production
center. Payment in grain to the value of 0.5 taels of silver was to be made by the
merchant at an army post. No further financing was required at the salt field.
265 catties of surplus salt, to be purchased by the merchant directly from the
producers. The government authorized the sale but did not guarantee the delivery.
An excise duty of 0.65 taels of silver was collected at the checking station.
35 catties extra allowance for packing and crating.
Each bag of salt when cleared at the checking station therefore weighed
550 catties, from which the government collected a total income of 1.15
taels of silver, including both the value of the barter and the excise
duty.
This basic formula, with variations, was carried out in the four com-
missions under the direct control of the central government. In practice
readjustments were made to suit the nature of the regional production and
market conditions. The overall weight of the yin, the proportion of official
to surplus salt and the barter and excise rates, all differed from one com-
5, IV Revenues, salt prices, and effects on consumers 213
mission to another. There were even price differentials within the territory
of a single commission. These details are included in Appendix C.
The rates in general remained stable. In the Liang-Huai region for
instance the schedule remained in force for the greater part of the sixteenth
century. When P'ang Shang-p'eng was in charge in 1568, he temporarily
reduced the weight of each bag to 485 catties. The excise on the surplus salt
was also proportionately reduced, so that merchants who had already made
the initial payment to the government for the impounded stock needed to
raise less cash to pay the excise tax. This was one of the measures devised
to clear the congestion at the checking stations;106 as soon as the situation
returned to normal the permanent schedule was restored. In 1616 the
weight of the bag was increased to 570 catties but the estimated value of
the grain bartered for it was still 0.5 taels of silver, and the excise on the
surplus salt was increased only to 0.7 taels.107 The modification was thus
a minor one.

Projected income from the salt monopoly and its allocation

Under normal conditions the salt monopoly produced a fixed income. The
extras created by Yen Mou-ch'ing and Lu Pao, along with other additional
income such as the profit from the cost-paid salt, were always listed as
separate items. The total cash income, excluding the grain delivered to the
frontier under the barter system, is summarized in Table 13.108
One striking feature of the account in Table 13 is that the proceeds from
the four commissions directly controlled by the central government were
all stated in round figures, that is as fixed quotas that their administrators
had to meet. In other agencies the income and allocation of funds were less
regular.
Fixed quotas were laid down for these four commissions in the late
sixteenth century when their administration began to deviate from estab-
lished procedures and their accounts could no longer be precisely audited.
The most drastic departure from standard practice took place in the
Liang-Che commission, which had very little official salt at its disposal. One
Ch'ing source indicates that by 1566 the production center had completely
ceased to collect salt from the registered producers and that all collections
had been converted to silver payments.109 Few of the registered saltern
households still actually produced salt and conversely, few of those who
did were officially registered as such (5, n). Most of the commuted salt
payments were undoubtedly derived from agricultural land. In the 1560s
the San-chiang salt field had on its registers 4,530 salt-producing ting who
owned a total of 155,550 mou of cultivated land.110 In Hai-yen county
there were complaints that the burden of salt revenue cost the district over
3,500 taels a year and that this really should be counted towards the land
214 The salt monopoly

TABLE 13. Cash income from salt monopoly, as of 1578

Annual
income
Agency (tls.) Allocation

Liang-Huai SDC 600,000 M/R

Liang-Che SDC 140,000 M/R
Ch'ang-lu SDC 120,000 M/R
Shantung SDC 50,000 M/R
Fukien SDC 24,545 2,344 tls. retained for local defense, 22,201
tls. M/R
Ho-tung SDC 198,565 76,778 tls. to Hsuan-fu Command; 74,259 tls.
to other army posts but credited towards the
land taxes of Shansi; 43,133 tls. as subsidies
to imperial clansmen in the province; 4,395
tls. M/R
Ling-chou SDS 36,135 to three adjacent army posts
Two SDSs, Kwangtung 15,968 4,790 tls. retained for local defense; 11,178
tls. M/R
Szechuan SDS 71,464 to Shensi
Four SDS, Yunnan 35,547 M/R
Total, all agencies 1,292,224 308,903 tls. retained for local defense,
delivered directly to the army posts, or
allocated for other expenses. 983,321 tls.
M/R

SDC = salt distribution commission SDS = salt distribution superintendency

M/R = delivered to the ministry of revenue

taxes.111 By contrast, the salt producers in these two fields submitted only
1,250 taels as salt payments.
The Liang-Che commission's production quota of 444,769 yin of salt
was all available for barter. When a frontier trader produced his granary
receipt, the commission simply paid him 0.2 taels of silver for every 200
catties of official salt to which it entitled him. The frontier trader then sold
the license with its remaining rights to a residential merchant, who went to
the salt field to make the actual purchase of salt. The commission office
eventually collected the excise tax from this second dealer.112 As a result
it took two or three years to complete what should have been annual
transactions.
In Liang-Che and elsewhere the accounting was further complicated by
such matters as the sale of salt tickets, extra collections from merchants
for canal dredging and poor relief, advance payments against the excise
duty, confiscation of salt in excess of the yin allowance, delivery of a com-
pulsory quota of contraband by the salt patrols and arrears of the ting
payments. As the central government could not be expected to keep track
of all these details it demanded instead that each agency submit a fixed
5, IV Revenues, salt prices, and effects on consumers 215
annual payment. Up to 1600 these payments were in general delivered to
Peking and the few shortages that occurred were always reported to the
emperor. In 1562 the total was reported as 1,323,811 taels; 113 in 1568 as
1,268,435 taels;114 and, as late as 1602, as 1,151,519 taels.115 It was thus
always close to the total shown in Table 13. When a distribution commis-
sion was unable to meet its quota the easiest way to make up the short-fall
was to exact loans from the salt merchants, or else to stiffen the penalties
to which they were subject (see below).
The contributions made by the other agencies outside the control of the
central government were less dependable. During and after the wo-k'ou
campaigns, provincial officials in these areas came to rely on the salt
revenue to finance the increased military expenditure they entailed. Once
they had intercepted the funds the central government had little means of
finding out how they administered them. In 1568 the ministry of revenue
reported that Kwangtung province had not applied for yin licenses for
three years.116 In an informal source dating from the early seventeenth
century it is estimated that a superintendent in one year of office could
make a personal income of 30,000 taels, though admittedly it might cost
him 3,000 taels to get the appointment in the first place.117
In Fukien many of the official salt fields had either ceased to produce
salt or else produced it at uncompetitive prices. The actual manufacturing
now went on outside the original monopoly system under regulations
devised by the provincial officials.118 This regional autonomy apparently
spread to Yunnan and Szechuan. After 1590 these southwestern provinces
were engaged in suppressing internal rebellions and fighting frontier wars
with the Burmese and thus had to turn to the salt monopoly for extra
funds. There is no evidence that they could at the same time make any
contributions to the imperial treasury.119 Nonetheless the income produced
by these agencies had never been large according to the official records;
and its loss did not seriously affect the total projected revenue. Technically
speaking, as the income intercepted by the provinces was still spent on
military defense, there was no conflict with established practice. Because
the accounting system never required that the expenditures of the central
government be kept completely separate from those of the provinces there
was no need to deduct these omitted deliveries from the total proceeds of
the salt monopoly. Only from occasional figures can it be sensed that the
allocations of the total income were arbitrary and not determined by general
guidelines.
The salt administrators of the late sixteenth and early seventeenth cen-
tury often referred to the total revenue as being worth 2 million taels. No
accurate accounts are available as they merely gave the income in round-
figure estimates. After comparing the various reports these are summarized
in Table 14.120
216 The salt monopoly

TABLE 14. Estimated annual income of the salt monopoly, ca. 1575-1600

tls.
Delivered to the ministry of revenue, in silver 1,000,000
Delivered to army posts in kind, valued in silver 500,000
Delivered directly to army posts by salt-administration 220,000
agencies, in silver
Intercepted by the southern provinces, in silver 280,000
Total 2,000,000

Higher salt prices and lower state revenues

According to the production quota of 1578 the total amount of salt under
the control of all the distribution agencies, including both official and
surplus salt, was over 846 million catties, or around 560,000 short tons. 121
If the total revenue was about 2 million taels, the income from each short
ton of government salt would be 3.54 taels. This is quite close to the 1535
rate in the Liang-Huai region, where every 515 catties of salt yielded an
income of 1.15 taels, that is, 3.345 taels per short ton.
These rates were high but not excessive. In Ming China, per capita
annual consumption of salt was usually reckoned as 10 catties.122 Each
consumer thus had to pay the government 0.024 taels for his annual salt
supply. Not all of this amount was indirect taxation, for it included the
cost of manufacturing the salt, which the government passed on to the
consumers. Some Ming officials were surprised to read in the histories that
under the T'ang the Liang-Huai region alone produced an annual income
from salt of over 6 million strings of copper cash.123
The Ming salt monopoly produced far less revenue but still caused
serious distress to the people, especially during its periodic administrative
crises. In 1527 the retail price of salt in Nanking was somewhere between
25-30 taels per short ton.124 In Hukwang during the crisis of the 1610s
a small package of salt weighing a mere 8 catties sold for 0.3 taels, that is,
about 56 taels per short ton. As this rise coincided with a fall in grain prices
in the province, the cost of salt to the average consumer was about half the
amount he spent on husked rice. In some districts this essential commodity
disappeared entirely. 'Filial sons were unable to provide their parents'
dishes with even a little savor of salt.'125 The immediate cause of the salt
shortage was an attempt at price control by the governor, who had set an
upper limit of 0.09 taels per package and had this officially engraved on
stone to confirm it.126 The fundamental reason for the scarcity nevertheless
still rested in the nature of the monopoly operation.
Much of the detailed information about the salt administration in the
5, IV Revenues, salt prices, and effects on consumers 217
late Ming comes from a single document, written by Yuan Shih-cheng,
a vice-director of the ministry of revenue. Yuan was assigned in 1616-17
to investigate and find solutions to the problems of the Liang-Huai region.
His subsequent report to the throne, divided into ten parts andfillingover
200 pages in the modern reprint, exposed numerous types of mismanage-
ment127 and offered many explanations for the high prices. The report also
established that while the consumers paid exorbitant prices, the state's
revenue actually declined.
Though Yuan was referring to conditions in his own time, when the
effects of Lu Pao's mistakes were still felt, the situation he described must
have developed over a longer period. The decrease in revenue appeared
first in the grain supplies bartered at the frontier army posts. When the
granary receipts became no longer immediately convertible into salt pay-
ments, their trading value dropped. This had happened even during P'ang
Shang-p'eng's administration. P'ang mentioned that in about 1570 a gran-
ary receipt for 1 yin, which cost the merchant the equivalent of 0.5 taels of
silver on the northern frontier, fetched only 0.54 taels in Yang-chou. This
meant that when operating costs were taken into account the frontier trader
actually sustained a loss. P'ang also revealed that the frontier governors, in
recognition of this hardship to the merchants, were lowering the official
barter rates. While each yin had a prescribed value of 0.5 taels, they
accepted a grain payment valued at 0.42 taels. They were forced to make
this concession because otherwise they could not press enough merchants
to take part in the system.128
By 1616 when Yuan conducted his investigation, the barter system had
deteriorated beyond recognition. The granary receipts sometimes had no
market value at all. Many of the frontier merchants turned in the unsold
receipts to the ministry of revenue and begged to be excused from further
service.129 Some frontier governors, instead of calling on the merchants to
provide grain, simply issued the granary receipts as pay to the soldiers.
Even the army granaries were in ruins. In the border region a granary
receipt for 1 yin might be sold for 0.07 taels, or about 12 per cent of its
official value. This, as Yuan put it, was only' slightly better than using them
to cover jars'.130
At Yang-chou there were still 'several hundred' residential merchants
engaged in the salt trade but only 'a few' substantial households were
speculating on the granary receipts. The value of the receipts, rather like
bond issues, increased sharply when they 'matured'. But not everyone
could take part in the business as it tied up capital and involved a certain
amount of risk. Yuan reported that these speculators were buying the
receipts at 0.2 taels per yin and selling them at 1 tael apiece.131 The latter
price was of course absorbed in the minimum price that the inland dealers
passed on to the consuming public. The former price of 0.2 taels, on the
218 The salt monopoly
other hand, represented the maximum that the frontier army posts actually
received as supplies.
The operations of the Liang-Huai region were about three years behind
schedule, but it was always difficult to tell which year's backlog was
currently being handled. In some cases the frontier posts neglected to send
in their stub-books and in others merchants who had secured the granary
receipts and registered with the commission subsequently deserted because
they were unable to pay the excise duty in advance as required by the
administrative office. When it came to allocating the quota for the current
year's delivery the salt officials devised a complicated scheme, designed to
uphold the principle that granary receipts issued over several years must
be honored simultaneously so that those traders who had already paid the
excise duty in advance could all receive their stock. At the same time the
commission had to go on absorbing new issues of granary receipts so as to
maintain the income from the excise duty at the desired level. As the
proceeds from the current year's excise duty were never sufficient to cover
the remittance to Peking, it was necessary to demand further advances
from the merchants. The commission thus bore some resemblance to
a modern treasury office, redeeming some of its old bond issues at the same
time as floating new loans. But the scheme worked in a rather irregular
fashion. Some merchants even after paying the advance taxes three times
had still not received any salt.132 At the time of Yiian's investigation, the
commission was in debt to the residential merchants to the tune of over
4 million taels.133 All these extra expenses incurred by the merchant, in-
cluding bribes made to the clerks and officials to advance their priorities,
were simply passed on to the consumer in the form of higher salt prices.134
Though the Liang-Huai region nominally produced 175 million catties of
official salt the actual collection from the saltern households was only half
that. The officials at the salt fields therefore reduced each merchant's share
while the merchants made up the short-fall through extra purchases.135
Contraband dealing had always been profitable on account of the high price
of official salt, and the contraband dealers could now outbid the merchants
by offering higher prices to the producers.136 This, combined with the fact
that the merchants also had to pay numerous 'extras' to obtain clearance
by the checking station,137 pushed up the price of salt still higher.
The final packing of the salt, meticulously done according to official
specifications, was quite costly, and the penalties for overweight were stiff.
It appears from one source that if the sack was overweight by as little
as 5 catties the fine was calculated on its total weight and not just the
extra amount.138 When the salt was sold by the residential merchant to the
inland dealers at the dockyard, however, the large sacks had to be undone
and repacked into smaller bags to meet the requirement of the retailers.
The inland dealers who carried the salt to markets in Hukwang and
5, IV Revenues, salt prices, and effects on consumers 219
Kiangsi had to surmount yet one more obstacle in the shape of a checking
station situated on the north bank of the Yangtze opposite Nanking. This
may originally have been established to prevent some of the inland dealers
arriving at their destination earlier than others which could affect local salt
prices, but in effect it was simply another bottleneck that could impose
delays of several months.139
Official salt, when accompanied by yin licenses, was supposed to clear all
inland customs stations duty-free but there was actually no guarantee that
it would not be taxed again. The officials in charge of the customs stations
could always impose duties on the carriers, exact personal 'contributions'
from the merchants and collect other kinds of service charges. The list of
customary payments {ctiang-li, 4, v and Appendix B) for Shun-an county
in 1561 indicates that the magistrate levied 0.1 tael of silver on every 100
yin of official salt passing through the county, and 1 tael on every 100 yin
actually sold in the county. Similar payments were made to the assistant
magistrate, the prefectural officials, and the lesser official functionaries.140
Though the total transit taxes thus imposed were low, amounting to only
about 0.012 taels per short ton, they were repeated in every county. In 1600,
when the eunuch Lu Pao was in charge of the Liang-Huai commission he
complained to the emperor that another eunuch, Ch'en Feng in Hukwang,
was taxing the salt that he had already released.141
The cost of manufacturing salt was always negligible. In Fukien, where
the sun-drying process was used, salt in around the year 1600 could be
produced at a cost of 0.25 taels of silver per short ton and was made
available to the nearby towns at 0.5 taels per short ton.142 Yuan Shih-cheng
reported that in Liang-Huai the producers were selling surplus salt to the
merchants at 0.3 taels for a 150-catty barrel, or 3 taels per short ton. 143 By
the time the residential merchants handed over the same stock to the inland
dealers, the price was rarely less than 9 taels per short ton. In the cities of
the interior the normal price was generally around 15 taels per short ton. 144
At this level a laborer's annual requirement of salt would cost him four
days' wages. When the price soared three or four times higher than this, as
happened in Hukwang in the 1610s, it is clear that salt was no longer within
the reach of the common people.
In sum, while the monopoly benefited a few unscrupulous merchants and
corruptible officials, and caused hardship to millions, it provided only very
small revenues for the state, and these revenues were actually even smaller
than they appeared in the official accounts. It also made the common
people shoulder the major burden of the government's deficit financing.
The fact that the government froze large amounts of funds in salt trans-
actions actually fostered generally high interest rates elsewhere, as it made
capital more scarce.
220 The salt monopoly

The settlement of 1617

The settlement of 1617, carried out on the recommendation of Salt-Control
Censor Lung Yii-ch'i and master-minded by Yuan Shih-cheng, has been
described as the most radical change in the salt monopoly since its creation
during the Former Han dynasty.145
The basic problem was that it was impossible, for reasons already
described, to readjust the barter rates, excise duty, total number of yin and
quantity of salt in annual circulation. The solution devised by Lung and
Yuan was to grant a franchise to those salt merchants who had already paid
the excise duty in advance so that they could themselves recoup the debt
that the government owed them.146
In the southern sector of the Liang-Huai region by this time, merchants
were said to have paid excise duty in advance on 2,600,000 yin of salt. After
weeding out a number of merchants on the grounds that they had 'in-
sufficient capital', the commission recognized its indebtedness on 2,000,000
yin. The accredited merchants were organized into ten syndicates (kang),
each representing a claim of 200,000 yin. Every year thereafter nine of the
syndicates were entitled to trade with the government for the current
production of salt, while the remaining syndicate received the payment of
salt already owed to it. This rotating schedule was to clear the total debt
in ten years. The payment on previous commitments was, however,
obtained not by increasing production but by taking out 36 catties of salt
from each sack in the current issue. Nor was the payment made in full.
According to the preliminary proposal the amount was set at a mere
142 catties of salt per yin instead of 570 catties. But the merchants by this
arrangement secured exclusive rights to trade with the government, each
with a permanent quota in proportion to his claim. The cancelled debt
might be a small price to pay for the privilege of the franchise. For the
government it meant an almost instantaneous cancellation of its debt to
the merchants. Within ten years even the nominal debt was also eliminated.
A similar formula applied in the northern sector of the Liang-Huai
commission, where the rotating syndicate system took fourteen years to
eliminate the old yin.
The barter system continued so that merchants were still required to buy
granary receipts from the frontier traders, but with the return to the
practice of handing over salt immediately in exchange for the receipts the
value of the latter was stabilized. The government accordingly announced
that the trade value of a granary receipt at Yang-chou was 0.55 taels.
The drastic move of granting franchise to the merchants was in fact not
entirely unprecedented. In 1591 the Shantung commission had listed the
names of more than a hundred residential merchants as being 'good
traders' who paid their excise dues on time. Some fifty merchants were
 5, V Responsibility for the failure 221
blacklisted as 'bad traders' and banished forever from the production
center. The residential merchants took turns in reporting daily to the
commission office,147 apparently to comply with its demands for mis-
cellaneous services. Attempts by the distribution agencies to regulate the
conduct of salt merchants had an even earlier origin, for the Liang-Huai
commission had issued a set of rules concerning their housing and clothing
as early as 1529.148 The late Ming system whereby a group of salt mer-
chants was granted exclusive rights to trade provided a useful precedent
for the operations of the Co-hang at Canton under the Ch'ing.

V RESPONSIBILITY FOR THE FAILURE

Official corruption
While no historian writing on the late Ming can possibly overlook the
question of official corruption, it is doubtful whether its further investi-
gation would add a great deal to our understanding of the reasons for the
failure of the monopoly. This study has already demonstrated clearly that
the salt monopoly in the sixteenth century cannot be regarded as a rational
institution, the operation of which had been perverted by dishonest ad-
ministrators. The inherent lack of integrity of the system was itself sufficient
to breed dishonesty. In the sixteenth century a man's name was tainted
immediately by the fact of his appointment as a salt official.149 When in
about 1580, a local vice-prefect was promoted to be acting commissioner
in the Ch'ang-lu commission, a friend of his sent him a letter of condo-
lence.150 In 1616 all six commissioners were indicted for irregularity.151
In 1623 it was claimed in a memorial to the throne that the salt adminis-
tration offices were simply refuges for 'degraded officials and those on
probation'. 152 In other words corruption was not only clearly in evidence
but was expected as a matter of course.

Fundamental causes of the failure

The fundamental causes of the monopoly's failure must be traced to the
beginning of the dynasty. Indeed the basic organization of the salt fields
under the Ming was quite similar to that under previous dynasties, but the
style of administration showed marked differences. As was shown earlier,
fiscal administration under the Ming in all its aspects never evolved away
from the rigid model set up by Hung-wu. The emphasis was always on
restraint rather than on expansion.
The entire fiscal establishment was strongly influenced by the economic
thought of the dynasty's founder, that is, by a strong sense of frugality and
the conviction that profit was in itself evil. Mercantile interests were thus
222 The salt monopoly
inherently in conflict with those of society and the state and had to be
curbed as far as possible. At the same time the state had to refrain from
'enriching itself, because any gain to the government automatically
meant a loss to the people. Though this orthodox sense of economy was
respected by the emperors and statesmen of previous dynasties, it seems
never to have been so faithfully followed as by the Ming. Hung-wu
emphasized repeatedly the need to reduce state expenditures for the benefit
of taxpayers and condemned the financial experts of previous dynasties
like Sang Hung-yang of the Han, Yang Yen of the Tang, and Wang An-
shih of the Sung, who had introduced fiscal reforms with a view to in-
creasing state income.153 It was not simply a coincidence that the Ming
never produced a reformer of comparable stature.
Though the Hung-wu Emperor was undoubtedly sincere in his concern
for the welfare of his subjects it is clear that the long-term effects of his
policies in many ways turned out contrary to his intentions. The salt
administration in the sixteenth century still retained all the basic features of
the fourteenth-century system, such as the registration of saltern house-
holds, collection of ting payments, control of grass land, establishment
of restricted areas, the barter system, and strict control of the yin licenses.
A fact often overlooked by the historians is that the salt officials were
never provided with the resources to do the job. From the very beginning
of the dynasty the government failed to develop its potential economic
power and resorted instead to political control as its basis of governing. It
thus persistently neglected to build up the minimum financial strength re-
quired to operate its fiscal machinery, an attitude which might be described
as'doing business empty-handed'.In this situation the extensive comman-
deering of services from the general population was unavoidable. The salt
administration therefore never became a public service but remained merely
a source of revenue. In the later years of the dynasty the officials could not
command sufficient unpaid service to meet even the needs of its chain of
revenue offices.
The failure to provide the monopoly system with adequate capital in-
vestment and service facilities has already been mentioned. Though its
administrative offices controlled huge resources and manpower and
dominated a vital sector of inter-provincial commerce, they had no
significant operating budget. Their office expenses were derived from local
requisitions in conformity with the usual practice. Right from the beginning
these revenues were always allocated even before they were collected, and
though they were internally-generated it was never permitted to use them
for investment. The only possible exception is the creation of the category
of cost-paid salt from 1553 to 1565 (5, m), yet the capital that financed the
purchase of the cost-paid salt was still not derived from regular income,
but from fines levied on merchants whose consignments were found to be
 5, V Responsibility for the failure 223
overweight at the checking stations.154 The imposition of this system
required the lower-level salt officials to forcibly produce the desired
amount, thus leading only to more irregularity.
The administrative offices had no ships of their own and could not
provide any kind of transportation. To obtain funds for poor relief and
the dredging of canals and rivers they had to resort to pressurizing the
merchants.155 In the early seventeenth century Yuan Shih-cheng reported
that some salt fields in the Liang-Huai region were no longer profitable
because the canals were not properly dredged. The merchants found that
the cost of transporting the salt from these fields actually exceeded its
purchase price.156
It is uncertain at what point the generally low level of state income made
it necessary to burden the salt monopoly with the task of deficit financing.
But with forced loans and delay in discharging governmental obligations,
a public debt was created without being recognized as such. Still devoid
of means to pay it, the court simply put pressure on the salt administrators,
who transmitted it to the merchants; the latter in turn passed the burden
to the consumers in the form of higher salt prices. Inevitably the effects
of mismanagement were cumulative.

Unnoticed proposals for reform

It is clear that by 1500 or so the monopoly system was already showing
signs of decay. The fundamental problem was that there was a persistent
short-fall in salt payments by the ting, or registered producers. Once the
government decided to allow the surplus salt to be sold privately, it would
have been sensible to abandon the ting assessment altogether. Rather than
maintain the fiction of a government monopoly it might have been better
simply to impose an excise duty on all salt production. One grand-secretary,
Ch'iu Ch'iin (in office, 1491-5), actually put forward such a proposal in his
Ta-hsiieh Yen-i-pu, a work which is said to have been read by the Hung-
chih Emperor,157 but there is no evidence that Ch'iu ever argued the case
publicly in his official capacity.
The barter system ran into similar difficulties, and was proving un-
profitable and unworkable by the fifteenth century. Because Yeh Ch'i had
the courage to suspend it he was condemned thereafter as an evil minister.
Some of his critics wilfully distorted the facts in order to accuse him of
advancing the interest of the Huai-an merchants.158 The outcry was such
that the barter system was revived. Three-quarters of a century later P'ang
Shang-p'eng also realized that making cash payments for salt was far
better than the inefficient and wasteful barter system, but though he re-
vealed his opinion in a private letter he withheld it from the emperor.159
Though it is easy to criticize the Ming government's ineptitude in busi-
224 The salt monopoly
ness administration, it must be admitted that fiscal reform in the sixteenth
century was not a simple matter. The tax regulations andfiscalprocedures
established by the founding emperor were backed up by considerable
military and political power, but by the middle of the dynasty this was
already in decline. Any drastic revision offiscalprocedures might speed up
this process, resulting in complete loss of control rather than improved
efficiency. The government's lack offinancialstrength was also in itself an
obstacle to any major reform.
Ch'iu Ch'un's proposal, while rational, would have had the disadvantage
of eliminating the service obligations of the salt-producing ting entirely,
and the collection of a universal excise duty would have been easier on
paper than in practice. By the sixteenth century landlords were registered
as saltern households while many actual salt-processers remained un-
registered; the contraband market had become unassailable, granary
receipts simply represented opportunities for commercial speculation, and
the administrators of the southern salt commissions used the revenues just
as they liked. It is thus extremely doubtful whether the central government
would still have been capable of introducing a major reform. The need to
maintain administrative uniformity was another restraining factor. The
ting was a universalfiscalunit of direct taxation, applied to all sections of
the population including the hereditary military families. The possible con-
sequences of emancipating only the salt-producing ting were manifold; it
was quite likely for instance that many of the population would have
claimed to be salt manufacturers in order to take advantage of the exemp-
tion.
The granary receipts, until they became completely worthless, still
provided the frontier governors with the privilege of issuing their own
letters of credit. The numerous reports of mismanagement indicate the
existence of strong vested interests in the frontier region, interests which
undoubtedly resented the abolition of the barter system, In short, the
government's reluctance to undertake any significant reform was deter-
mined by a variety of motives, both rational and irrational, concealed and
overt. The only way for officials in the late Ming to reach a consensus was thus
in the end to agree to abide by the 'fundamental laws of the dynasty', the
original regulations of the Hung-wu Emperor.
 6
Miscellaneous incomes

The miscellaneous incomes described in this chapter and in Table 15 com-

prise all state revenues other than the land taxes and the salt revenue.
Though in describing any other fiscal system it might seem absurd to group
together such items as the maritime tariff and the incense fees collected from
the worshippers at national shrines, the revenue from government mining
and the kitchen expenses account of the court of imperial entertainments,
this curious sort of classification does reflect the style of the Ming fiscal
administration. Governmental finance in the sixteenth century did not come
to terms with the general economic trends of the times. The fiscal apparatus
was too rigidly constructed to be able to exploit effectively the possibility of
generating staple revenues from industry and commerce; it merely clung
instead to a number of administrative incomes and what might be termed
'nuisance taxes'. It was easier to continue to collect revenues from these
sources because they were closely connected with the existing governmental
organization. Though the income produced from such sources was limited,
or even marginal, it was assured, and therefore played a larger part in state
finances than did revenues from the more modern sector of the economy.
As these anomalies are matters of historical fact they leave the financial
historian little choice in his ordering of the subject. Though it might in
theory be possible in the interests of simplification to eliminate some of the
less essential items and to reclassify the rest, it is likely that the disadvantages
of this would outweigh the possible benefits. It would obscure the nature
of each item of income, create discrepancies in fiscal terminology and, more
significantly, obscure the workings of the administration. The historian
would thus in effect be taking on the role of a fiscal reformer.
One of the basic characteristics of the Ming system was that it derived
numerous small incomes from very diverse sources. These accounts were
never integrated and a number of revenues were actually disbursed before
auditing. In order to preserve accuracy it is necessary therefore to cite the
miscellaneous incomes item by item, even though some of them can be
described only very briefly.
8 [225] HTF
226 Miscellaneous incomes
The list presented here is not absolutely complete, because the lack of
standardization in accounting methods made it possible for some items
to be split up under different names, but it nonetheless covers all the
incomes worth mentioning. Several of the items have been referred to
earlier in connection with the land tax. As they were so small some counties
simply discontinued their collection and allowed them to be covered by
the land taxes (3, iv). But this was not a universal practice, and the fact
that they never completely disappeared as separate items justifies their
listing in this chapter.

I REVENUES FROM COMMERCE AND INDUSTRY

(a) The inland customs duty
Under the Ming there were three kinds of taxation that could be imposed
on commercial traffic on inland waterways. The tonnage (cWuan-chiao)
was levied on the carrier, paid by the shipowner, and collected by the
ministry of revenue, the assessment being based on the width of the ship.
The local business tax (shang-shui), levied on all merchandise transported
on land or water, was paid by the merchant and administered by provincial
officials. The forest produce levy (chou-mo cKou-fen) was levied only on
shipbuilding materials, and was controlled by the ministry of works.
Originally it was paid in kind and levied simply on bamboo and timber but
in the later part of the dynasty the collection was extended to cover every
item that might conceivably be useful to government dockyards, including
hemp, nails, lime, charcoal, and fung oil. The actual payment was, how-
ever, converted to silver.
The origin of the inland customs houses can be traced to the four tonnage
stations established in 1429 to collect toll payments on the Grand Canal.1
By the sixteenth century there were seven of them, located at Pei-hsin-kuan
near Hang-chou, Hsii-yeh near Soochow, Yang-chou, Huai-an, Lin-ch'ing,
Ho-hsi-wu, and Chiu-chiang. Each was run by a secretary seconded for
a year from the ministry of revenue.
It was only gradually that the tonnage stations started to take over the
administration of the local business tax at the ports as well. This happened
at Pei-hsin-kuan in 1511,2 and at Lin-ch'ing shortly thereafter, but up to
1530 the stations at all the other ports still handled only the tonnage.3 By
1569, however, the takeover was complete and all the stations were put
under the joint control of the ministry of revenue and the provincial
government. The ministry's officials made the assessments and handed the
tax bills to the shipowners and merchants, while the provincial officials
received the payments and delivered the tax proceeds.4 Since the collection
covered both the ship and its cargo the stations effectively functioned as
 6,1 Revenues from commerce and industry 227
TABLE 15. Miscellaneous incomes, ca. 1570-90
Under the jurisdiction of:
C*ntP0csT\f o f
income Ministry of revenue Ministry of works Ministry of war Ministry of rites

Revenue from (a) Inland customs (g) Forest

commerce and duty produce
industry (b) Local business levy
tax (h) Government
(c) Maritime tariff mining
W) Store franchise
fees
(e) Excise on wine
and vinegar
(/) Stamp tax on
real estate
transfers
(0 Fish duty
Administrative 0') Sale of rank (o) 'Common (p) Incense
incomes (*) Ecclesiastical post fees at
license fees money' national
(/) Payment for shrines
* rationed
salt'
(rri) Commutation
of punishments
(«) Profits from
minting money
Commutation (?) ' Speed- the- (r) Artisan(«) Horse (y) Calendar
of services delivery payment payment paper
and supplies money' (s) Reeds (v) Commuta- (z) Kitchen
tax tion of service
(0 Material capital fees due
supplies guard to the
to the fourduty court of
bureaus
(w) Commuta- imperial
tion of entertain-
personal ments
attendance
(x) Savings from
postal service
Non-cash incomes: (ad) income from tea and horse trade; (bb) other items not listed.

customs houses and were formally recognized as such. In reality matters

were far more complicated than this terse description implies.
The Pei-hsin-kuan office, for example, after taking over the administration
of the local business tax virtually came to constitute a taxation authority
in itself. It controlled a network of toll stations on both land and water
extending over several counties around the city of Hang-chou. Its revenues
included the tonnage, local business taxes, registration fees on boats some
8-2
228 Miscellaneous incomes
of which might never even have to pass through the stations, and even
regular payments from the residential merchants in the cities. Its juris-
diction extended to both wholesale and retail trade. Toward the end of the
sixteenth century its total annual income was 34,975 taels of silver, only
6,318 taels of which was derived from the tonnage. 5 The situation at other
stations was similar.
The total proceeds from these seven stations, augmented by the collection
made at the Ch'ung-wen gate at Peking, formed a separate 'system' for
accounting purposes. The revenue was generally referred to as ch'uan-chiao,
thus creating the false impression that it was all derived from the inland
customs duties. In reality it constituted that portion of the business tax the
management of which was directly supervised by the ministry of revenue.
The station at the Ch'ung-wen gate was not of course situated at a port.
In the late sixteenth century the general practice was that the northbound
cargo destined for the capital paid part of the business taxes at Lin-ch'ing
and Ho-hsi-wu, and the remainder at the gate on arriving at Peking. In-
corporating the Ch'ung-wen gate station into the system was thus not
illogical, though its revenues did also include a substantial amount of petty
cash collected on goods transported overland by mules and carts.6
The duties collected by these stations were in general very light but the
published schedule needless to say did not cover all the extras that the
administrators could impose on the merchants, and its management was
far from uniform. Despite widespread reports of corruption,7 there were
some honest officials whose integrity on occasion even impressed European
observers.8 One major cause of mismanagement was administrative incon-
sistency. The operation of the stations entailed both the publication
of collection schedules and the fixing of a tax quota at each port.
While the former implied fixed collection rates applicable to a variable
volume of commerce, the latter required the station officials to guarantee
a fixed level of tax income. This inconsistency originated in the early years
of the dynasty.9
As no collection schedule appears to have survived, it is only from
scattered entries in several sources that one can get any idea of the official
rates. These were remarkably comprehensive. One Ch'ing gazetteer indi-
cates that the schedule published by the Lin-ch'ing station under the Ming
covered 105 double pages and incorporated 1,900 articles.10 The rates were
also surprisingly reasonable. At Huai-an in the early seventeenth century,
boats in transit were divided into two classes for the purpose of collecting
the tonnage dues. Neither class is clearly defined in the sources, but one
paid 0.058 taels and the other 0.029 taels.11 As the width of the boat in
either case was 5 feet, implying a capacity of no less than 15 tons, the pay-
ment was negligible. It seems however that the toll was repeated at each
passage of the customs station and that unladen vessels were probably not
 6,1 Revenues from commerce and industry 229
exempted from it. At Pei-hsin-kuan in the late sixteenth century the customs
duties, which in reality constituted a part of the local business tax, were
said to range from 0.2 to 3 per cent ad valorem. The lowest rates were on
straw mats and tin-foil, and the highest on brass.12
The main disadvantage from the merchants' point of view was that
cargoes in transit over long distances were taxed repeatedly. Officials could
also impose unscheduled collections under the guise of 'voluntary con-
tributions'. In 1583 it was pointed out by a censor that northbound cargo,
after paying 60 per cent of the business tax at Lin-ch'ing and the other
40 per cent at Ho-hsi-wu, then had to pay the entire scheduled tax all over
again at the Ch'ung-wen gate, where the collectors simply disregarded the
receipts from the other stations.13 At some ports the collectors imposed
a duty on the merchandise unloaded from the carrier while on its way to the
warehouse and then repeated the same duty when the same cargo was re-
loaded onto another carrier. This practice gradually became customary, so
that eventually the double payment was assessed on all cargo in transit,
including that which never left the original carrier.14 In the late sixteenth
century a shipload of about 500 piculs of white grain en route from Che-
kiang to Peking was required to pay duties, taxes, and fees at twelve different
places. The total amount came to 70 taels of silver, roughly 20 per cent
ad valorem, which was still not prohibitive. Nonetheless white grain was
a primary commodity delivered by taxpayers at the service of the emperor.15
Up to the mid sixteenth century the annual quota for each station was
fixed on the basis of the actual level of collection in some previous year
-not necessarily the highest, but whichever was considered the most
reasonable. Normally the collectors had no difficulty in meeting the quota.
Though in theory it might be expected that an official whose collection
consistently exceeded the prescribed quota would have a better chance of
promotion, in practice this was not so. In an age when virtue was regarded
as more important than efficiency, an official who broke earlier collection
records was liable to be criticized for being self-seeking, abusing the tax-
payers, and creating unnecessary difficulties for his successors. Throughout
the Ming period no official ever won public esteem for attempting to
increase tax revenue. From the contemporary sources it can be seen that
most of the bureaucrats assigned to the customs houses were content
merely to fulfill their quotas. It is actually asserted that one minister of
revenue, Chou Ching (in office, 1496-1500) gave lower ratings in his per-
sonnel evaluation to those secretaries who had delivered more customs
duties than they should.16
When in the later sixteenth and early seventeenth centuries state ex-
penditures rose enormously, the quotas of the customs stations were in-
creased, in general by an arbitrary amount. This resulted in dislocation,
increasing malpractice and short-falls in collection.
230 Miscellaneous incomes
The low efficiency of the customs stations was not surprising. Their
operations were quite uncoordinated and no director general was ever
appointed to organize them systematically. There was thus no official with
any professional interest in their long-term development. The ministerial
officials carried out their duties merely as interim assignments while the
provincial officials regarded the collection of the duties as a financial
burden to their districts. The prevailing anti-commercial ideology pre-
vented the bureaucrats from promoting the interests of the merchants and
the expansion of their business activities. No attempt was ever made to
facilitate the circulation of goods, whether daily necessities or luxuries,
within the network of inland ports. In the collection of dues no distinction
was made between staple goods transported in large volume and minor
items, simply because the concept of a local business tax was never dis-
carded. The contemporary sources on the customs houses reflect the same
attitude. In their descriptions of articles in transit, the emphasis is always
on farm products transported over short distances, such as hogs, straw
mats, bean cake, pickled vegetables, and preserved fruits, and less attention
is paid to staple goods such as textiles and porcelain.17 In 1526 it was
actually necessary to issue an imperial edict forbidding the customs houses
to levy duty on the food and fuel stored aboard the carriers.18 In 1643
Minister of Revenue Ni Yiian-lu reported that a customs declaration made
at the Ch'ung-wen gate might include as many as 3,000 items belonging to
scores of different merchants. 'If a single piece of gauze or a skirt was
omitted from the declaration, a fine was imposed on the entire list.' Ni
also declared that in front of a customs house ' somewhere in the south' he
had seen a stone tablet on which was engraved the warning that the penalty
for omitting even one item of merchandise from the declaration was the
confiscation of half the ship's cargo.19
The customs houses were well staffed. The Huai-an station in the late
Ming, according to one Ch'ing account, had a full-time staff of 12 officials
and 14 lesser functionaries, plus 11 associated officials and lesser function-
aries, and 121 scribes.20 The office was thus almost comparable in size to
the ministry of revenue. But from the assignment of duties it can be seen
that most of the personnel were engaged in working out and recording the
assessment of dues in minute detail. Like all Ming government offices the
customs houses were supported by local requisitions. The records of the
Pei-hsin-kuan station indicate that many of its scribes were unpaid. Its
operations also required the posting of unpaid patrolmen in the rural areas
and the appointment of civilian tax agents to make quarterly collections
from the 1,200 ships and 3,500 residential merchants in the district.21 In
short, the operations of the customs stations were limited by the essentially
agrarian nature of the economy, and they were ill-adapted to the per-
formance of more sophisticated financial tasks.
 6,1 Revenues from commerce and industry 231
TABLE 16. Collection quotas of eight inland customs stations, 1599-1625
(tls. of silver)

Station 1599 1621 1625

Pei-hsin-kuan 40,000 60,000 80,000

Hsii-yeh 45,000 67,500 87,500
Yang-chou 13,000 15,600 25,600
Huai-an 22,000 29,600 45,600
Lin-ch'ing 83,800 63,800 63,800
Ho-hsi-wu 46,000 32,000 32,000
Ch'ung-wen gate 68,929 68,929 88,929
Chiu-chiang 25,000 37,500 57,500
Total 343,729 374,929 480,929

The collection quotas of the customs stations in the late sixteenth and
early seventeenth centuries are listed in Table 16.22 There appears to be no
information available on the volume of staple goods in transit.
The late sixteenth-century quota figure of 340,000 taels or so was only
a guideline. The Pei-hsin-kuan station in 1600 reported that its actual
collection was over 42,000 taels, and in 1611 that its quota had been in-
creased to 49,700 taels.23 Sometimes the collection fell below the target
figure. Chao Shih-ch'ing, who was minister of revenue from 1602 to 1610,
reported that the total collected from the eight stations in 1597 was
actually 325,500 taels. In 1601 the total income fell to 266,800 taels, some
76,900 taels short of the quota.24
The revenue could be disbursed for a variety of purposes, such as sub-
sidizing the army units near the collecting stations, minting copper coins,
and providing famine relief and the expenses of palace construction.
Sometimes one station made its funds available to one particular agency
while the others sent their collections elsewhere. At times the court also
instructed the stations to collect part of the revenue in copper coins, to be
submitted to the Kuang-huei Treasury inside the Imperial City (1, i). In the
1570s it was decided that all the income should henceforth be delivered to
the T'ai-ts'ang Treasury, under the jurisdiction of the ministry of revenue,
to become part of the supplies sent annually to the northern frontier army
posts. Even after this date unscheduled deliveries made on an ad hoc basis
still continued.25 The accounts of the T'ai-ts'ang Treasury for the year 1580
show that it received only 162,299 taels from the customs houses, about half
the projected quota. 26

(b) Local business tax (shang-shui)

The local business tax in Ming times was in many respects a forerunner of
the likin under the Ch'ing in that it was characterized by minimal rates,
232 Miscellaneous incomes
broad coverage, and duplication of collection. In general, however, the
income it provided was too small and too scattered for it to be regarded as
a major source of revenue. After the stations established at major commer-
cial centers to collect the tax were annexed by the inland customs houses,
the shang-shui became even less important. In the early years of the dynasty
there were more than 400 stations collecting local business taxes but by the
early seventeenth century only 112 remained in operation. The rest had
become unprofitable and had been closed.27 In 1568 the ministry of revenue
reported that at one particular station labor costs amounted to about
400 taels a year while it netted only 110 taels in taxes.28
In the sixteenth century the remaining stations were run either by pre-
fectural or county governments. The low rates of collection can be seen
from the operations of Pei-hsin-kuan, described earlier, where they ranged
from 3 per cent to less than 1 per cent ad valorem. The levy was universal,
and not even peddlers or peasants who carried bales of produce to the
cities for marketing were exempted from it. In general there was only one
official at each station. As the drafted patrolmen and scribes were not only
unpaid, but were obliged to produce a fixed quota of contraband, they
were forced to rely for their living on bribes and extortion from passing
merchants.29 When the quotas were increased these abuses all too easily got
out of control.
The fact that the quotas of the business tax stations had been initially
fixed in paper currency and then converted to silver made the collection
meaningless, except in a few districts where some readjustments had been
made. For example in Fen-yang county, Shansi, the local business tax in
1609 still yielded 6,606 taels of silver, a worthwhile sum.30 But in Ching-hua
county, Chekiang, a rather prosperous area, the quota was listed in 1578
as less than 7 taels, and it was frankly acknowledged in the local gazetteer
that the collection had long since been suspended. The amount was more
than made up with commuted payments from the patrolmen.31 Other
districts filled their quotas with the proceeds of the land taxes.
This does not mean that merchants in these districts were not taxed at
all. Hui-chou prefecture in South Chihli was noted for its wealthy mer-
chants. The local business taxes of the entire prefecture, fixed at 94,565 kuan
of currency in 1482 and never revised thereafter, had a value of less than
30 taels of silver by the late sixteenth century. But because they paid hardly
any business taxes the merchants of the prefecture were frequently asked
to make substantial contributions towards the cost of construction work in
Peking.32
After the wo-k'ou campaign the provincial authorities, particularly in
Fukien and Kwangtung, established new collecting stations outside the
existing shang-shui system, mainly located at bridgeheads and ferries. These
stations prescribed their own rates of collection and delivered the proceeds
 6,1 Revenues from commerce and industry 233
directly to local defense units as military supplies. The central government
exercised no effective control over them whatever.33 Similar stations were
set up on the northern frontier. The one established in 1547 at Kuang-ning,
beyond the Great Wall, and the one re-established at Shan-hai-kuan in
1550, were of considerable international significance as they controlled the
trade route to Manchuria and Korea. They were run at first by army
officers and later by eunuchs, but seem to have left no records.34 All that is
known is that the annual income of the Shan-hai-kuan station around 1570
was 4,000 taels.35
The tolls collected at the city gates of Peking, other than the Ch'ung-wen
gate, should also be included in the category of local business taxes. The
gates were controlled by the eunuchs, who acted as collectors. The greater
part of the toll was paid by villagers bringing in farm produce to the city.
The eunuchs surrendered a fixed quota, usually paid in copper coins and
worth 5,000 taels a year, to the court of imperial entertainments. The rest
they pocketed for themselves.36
In brief, the local business tax during the sixteenth century steadily dis-
integrated. Part of the system fell into disuse; part survived under the con-
trol of provincial administration, and yet another part was transformed
into a kind of tax farming. It is clear that the universal undifferentiated
collection resulted in general inefficiency. The system was also handicapped
by the fact that the operating budgets of the tax stations was quite in-
adequate. In 1576 the governor of Kiangsi reported that the annual
collection quota of the station at Chang-shu-chen was 172 taels of silver.
After he reorganized its operations the proceeds increased ten-fold.37
Evidently much of the tax potential within the system was being wasted.
According to the Ta-Ming Hui-tien, as of 1578 the local business tax
yielded a total of 150,000 taels of silver per year, on average slightly over
100 taels per county.38 These funds were combined with the proceeds of
the land taxes and earmarked as a part of either the retained or the trans-
ferred revenue of their respective districts.

(c) The maritime tariff (fan-p'o cWou-feri)

Throughout the Ming period international trade was never regarded as
a primary source of state income, a tradition that had been established from
the beginning of the dynasty. Foreign trade was limited to the purchase
and sale of goods by tributary missions during their periodic visits. Both
Hung-wu and Yung-lo, in order to demonstrate their magnanimity to the
peripheral states, repeatedly exempted the goods brought in by visiting
embassies from the payment of tariffs, for instance in the case of the
Korean embassy in 1370 and the embassy from Brunei in 1403.39 Though
the Ming government reserved the right to purchase up to 60 per cent of
234 Miscellaneous incomes
the embassy's goods, it did not often exercise it. When it did, the payment,
termed 'awards', exceeded the value of the merchandise several times over,
and was compounded by the cost of the entertainment lavished on the
personnel of the embassay.40
In the late fifteenth century foreign trade did go on despite the tributary
theory. The eunuchs in charge of the maritime trade superintendencies,
some of the civil officials, and the army officers in charge of coastal defense
must have made enormous profits from the tributary embassies, regardless
of whether these were real or pretended. The court in Peking knew little of
their dealings. When in 1493 the governor-general at Canton reported that
the reception of frequent tributary missions was becoming burdensome to
counties under his jurisdiction, the ministry of rites, after checking its own
records, replied that the only tributary missions sent in the past six years
had been one each from Siam and Champa. 41
After this the ministry of rites suggested that in future it should not
insist on the arrival of foreign tributary missions in pre-arranged years,
a proposal with which the emperor concurred. In any case the central
government had already given de facto recognition to the limited trading
privileges of the tributary states at Canton, since their missions had no
other business there but trade. Trade was formally opened in 1509, when
the provincial officials had worked out a set of regulations to control it.42
The privilege of trade was granted to persons from the tributary states,
even when they were not accompanied by an embassy. The tariff was set
at 30 per cent, to be paid in kind and collected by the provincial officials
under the supervision of the governor. Luxury items obtained in this way
such as ivory and rhinoceros horn were delivered to the emperor, while
staple goods like pepper and sapanwood were sold in the market. In 1510
the sale of the items collected in the two previous years grossed 11,200 taels
of silver. With the emperor's approval, these funds were retained in the
province as local military supplies.43 After 1517 the flat rate tariff was
reduced to 20 per cent - still to be paid in kind.44
The liberalization of international trade was interrupted in the early
sixteenth century by the clashes between Portuguese seafarers and Ming
armed forces on the China coast, and in 1523 by the fighting at Ningpo
between the Ouchi and Hosokawa factions from Japan, who both claimed
to represent the Ashikaga Shogunate.45 In 1529, however, Canton was
again thrown open to foreign traders, and 'barbarians' were even
allowed to enter the city. In 1556 the trade was put under the supervision of
a vice-circuit intendant on coastal defense (hai-tao fu-shih) who organized
groups of Cantonese, Fukienese, and Hui-chou merchants to act as official
agents for the foreign traders.46 The revenue must have been considerable
because in 1529 Governor Lin Fu, on petitioning for the re-opening of the
trade, declared: 'In less than a month, the foreign goods [collected] in the
 6,1 Revenues from commerce and industry 235
warehouses will be worth tens of thousands of taels of silver.'47 But the
exact value of the tariff subsequently collected is unknown.
From 1557 the Portuguese secured the right to reside in Macao. In 1564
P'ang Shang-p'eng reported that nearly 10,000 of them were living there,
and that during the trading season there were more than twenty Portuguese
ships at anchor. Every year the Portuguese paid the provincial authorities
a lump sum of 20,000 taels to cover the tonnage and in addition, 500 taels
to the magistrate of Hsiang-shan county to cover the rent of Macao.48 One
contemporary source reveals that their customs payments were nominal:
'They offered customary payments (cKang-li) to the maritime trade super-
intendent and the county magistrate who made the inspections... Not
more than 10 per cent or 20 per cent of their goods were declared and paid
duties.'49 Even in the last decade of the sixteenth century the annual income
from the maritime tariff at Canton remained at the general level of 40,000
taels.50 The profits made by the Portuguese, on the other hand, must have
been enormous, for according to H. B. Morse: 'As late as 1612 it was
categorically stated that the license of the viceroy at Goa authorizing a ship
to trade at Macao was worth £25,000 to the ship for one voyage.'51
In 1567 the Ming court also authorized trade at Yiieh-kang in Fukien,
near modern Amoy. Details of the license fees, poundage and tonnage
levied at this port have survived,52 though as the main concern of this dis-
cussion is the general level of income, it is possible here to cite a few
illustrative examples.
The tariff was paid by Chinese merchants trading overseas, and on im-
ports only. The total annual number of ships was fixed at eighty-eight on
afirst-come-first-servedbasis. A ship sixteen feet wide, that sailed to Brunei
and farther west, paid a tonnage tax of 5 taels of silver, plus a license fee
of 3 taels. Sculptured ivory objects paid an import duty of 1 tael per 100
catties, and raw ivory half that amount. Pepper paid 0.25 taels of silver per
100 catties, roughly 1 per cent ad valorem.53 Since ships returning from the
Philippines usually carried no imports, they paid a duty of 150 taels per
vessel per voyage; this was later reduced to 120 taels.
The revenues collected at Yiieh-kang amounted to 6,000 taels in the first
year. This had increased to 29,000 taels per year by 1594, when Toyotomi
Hideyoshi's invasion of Korea prompted the authorities to close the port.
In 1615, when the port was re-opened, the projected annual income was
27,087 taels. In that year the rates were reduced, in general by about 15 per
cent. All the proceeds were retained in Fukien as local military supplies.
The customs duties were allocated to the army and the tonnage dues to the
naval forces.54
This brief sketch covers the entire maritime tariff collected in the six-
teenth century. The ministry of revenue never played an active part in the
operation but the collection did lessen its difficulties in securing military
236 Miscellaneous incomes
funds. The low level of income provided by the tariff has never been ex-
plained, but it is likely that the authorities were afraid of imposing higher
rates lest these induce sailors, shipowners, and traders to turn to contra-
band trade and join forces with the pirates. The officials also doubtless
wished to leave plenty of scope for making irregular collections on their
own accounts.55

(d) Store franchise fees (meng-fan shui), (e) Duties on wine and vinegar
(chiu-ts'u shui), and (/) Stamp tax on real estate transfers (fang-ti cKi shui)
These items were collected along with the business tax by the local tax
stations, or else by the county government in areas where there were no tax
stations. The franchise fees were levied on street stalls as well as permanent
shops. The duties on wine and vinegar were inherited from the Sung,
which had imposed a state monopoly on distilling. The stamp tax was
fixed at 3 per cent of the purchase price of the real estate or of the mortgage
payment.
Every county had a quota for each of these items, fixed in terms of paper
currency during the early years of the dynasty. In most districts the re-
venues virtually disappeared after the original quotas were converted to
silver. Local gazetteers quite commonly listed the total annual revenue from
the three items as 3, 4, or 5 taels of silver. The actual collection, however,
might be a different matter. The duties on wine and vinegar were generally
disregarded, as the Ming never managed to exercise any effective control
over distilling, but the other two items were of some significance in the large
cities.
The collection from the shopkeepers of Peking, the administration of
which was divided between Ta-hsing and Wan-p'ing counties, yielded
10,641 taels in 1579.56 The stamp tax in Wan-p'ing county alone exceeded
2,000 taels per year because property values in the capital were higher.57
In the late sixteenth century, according to the magistrate of Wan-p'ing,
a house in Peking might be mortgaged for as much as 7,000 taels.58
In the city of Huai-an, in South Chihli, the franchise fees were paid
quarterly in lump sums by the seven major trade guides. As the flour mills,
distilleries, and noodle shops had no guild organizations the magistrates
appointed collectors for them. There is no information on the total revenue
collected. The management of the stamp tax in this prefecture also showed
some variations. In the early seventeenth century mortgages were exempted
from taxation, and on property sales the tax rate was fixed at 1.16 per cent
of the purchase price.59
In Ch'ang-chou prefecture, South Chihli, it seems that the stamp tax
had been discontinued for some time when the local officials in the late
sixteenth century attempted to revive it. Apparently it had not been
 6,1 Revenues from commerce and industry 237
possible to carry out the collection effectively because of the 'protests and
opposition' it aroused.60
These scattered cases seem to indicate that when paper currency fell into
disuse in thefifteenthcentury the central government never made any clear
attempt to readjust the revenues that had originally been fixed in terms of
paper money. In some areas where the revenue from such items as these
had become indispensable, the local officials evidently made some sort of
readjustment on their own initiative, and their ad hoc arrangements in time
became customary practice. Where no such readjustments were made the
taxation was abandoned and the revenues made up from the land taxes.
Some time before 1602 the ministry of revenue made a belated attempt to
revive the stamp tax on real estate in the hopeful expectation that it could
raise 100,000 taels of silver from this source.61

(g) The forest produce levy (chou-mo cKou-feri)

The Ming-shih lists 13 stations set up to collect the forest produce levy.
There were 2 near Nanking, 5 near Peking, and 1 each at Cheng-ting in
North Chihli, Wu-hu (T'ai-p'ing prefecture) in South Chihli, Sha-shih
(Ching-chou prefecture) in Hukwang, Hang-chou in Chekiang, Lan-chou
in Shensi (now Kansu), and Kuang-ning Wei in Manchuria.62 The list is
not complete, for there was in fact another at Ch'ing-chiang-p'u near
Huai-an, and another at Pao-ting in North Chihli.63
The stations were by no means of equal importance. Several of them,
including those on the outskirts of Peking, collected only timber and
bamboo for the manufacture of palace equipment and utensils. The Lan-
chou and Kuang-ning stations were located on the frontier and though
their operations may have been of considerable importance they are un-
fortunately undocumented. The stations of most interest are thefivelocated
on major water routes where there was a large volume of commercial
traffic, including forest produce: namely those at Wu-hu, Sha-shih, Hang-
chou, Huai-an, and Lung-chiang, the last-named being close to Nanking.
By the sixteenth century only this last station still continued to make its
collection in kind in order to provide materials for the imperial furniture
factory in the southern capital. The other four collected payments in silver
which were used to finance shipbuilding in the government dockyards.
Sometimes the funds were diverted tofinanceother projects undertaken by
the ministry of works.
The forest produce levy and its collecting stations arose from Hung-wu's
attempts to make the ministry of works self-sufficient in the matter of
supplies. The most essential items, timber and bamboo, were therefore kept
separate from ordinary taxation. Though it is indicated in several early
works that the collecting stations did not come under the control of the
238 Miscellaneous incomes
ministry until 1471, what actually happened was that in that year Minister
of Works Wang Fu obtained the emperor's permission to send ministerial
officials to manage the Sha-shih, Wu-hu and Hang-chou stations on a one-
year basis.64 Until then the collection at these stations had been carried out
by provincial officials on the ministry's behalf. The Huai-an station had
always been supervised by the ministry of works official in charge of water-
ways in that region, and the Lung-chiang station had always been super-
vised by the Nanking ministry of works.
When the payment was made in kind, the tax generally took 20 per cent
of the timber in transit.65 Payments in silver were more reasonably assessed,
taking into account such factors as the kind of timber, the size of the logs
and their place of origin. The rates varied from approximately 5 to 10 per
cent ad valorem,66 but that at Huai-an was always lower than the rates at
other stations, partly because all the timber arriving there had already been
taxed at least once before. The Huai-an office, which taxed all commercial
traffic on the Grand Canal, imposed a flat rate of 3.33 per cent ad valorem
on all shipbuilding materials.67
Though the rates were reasonable, the fact that collections were made
repeatedly caused further hardship to the merchants. In 1608 timber
coming from Szechuan had to pay transit taxes in three different provinces
before reaching the lower Yangtze, in addition to the levies at Sha-shih,
Wu-hu, and Nanking. The cost of cedar and pine wood sent to Peking from
the southern provinces was increased at least 50 per cent by taxes imposed
en route. The Lung-chiang station was notorious for holding up shipments
of timber, owing to its rule that assessments could only be made on 100 rafts
at a time. Another problem for timber merchants was compulsory forced
purchase by government agencies, which usually took the best of the con-
signment.68 The price of timber in Peking was further increased by high
transportation costs on the Grand Canal; sometimes it was so exorbitant
that many shipowners found it was to their advantage to dismantle their
vessels at the northern terminal of the canal rather than repair them. The
masts in particular fetched very good prices.69
The imposition of the forest produce levy could also be fairly flexible.
The stations which collected the tax in silver were assigned an annual tax
quota, and in years when there was a large volume of commerce this was
easilyfilled.This was the case when Ho Tsun was in charge of the Sha-shih
station in the decade after 1510. Once the quota was met he offered the
timber merchants a general reduction on the scheduled rates.70 Shao Ching-
pang, who was in charge of the same station in the 1520s, adopted an even
more striking solution. After making up the annual quota in three months
he apparently closed down the office and for the rest of the year allowed
the timber merchants to pass through the station tax-free.71 In the 1560s
Yang Shih-ch'iao established an honor system at the Hang-chou station,
 6,1 Revenues from commerce and industry 239
and based his tax assessments on the merchants' own declarations.72
Significantly, these three officials all won high praise from the traditional
historians.
The Huai-an station went to the opposite extreme. This office was the
major supplier of funds for the Ch'ing-chiang-p'u dockyard and by the late
sixteenth century was quite unable to meet its assigned tax quota. The solu-
tion adopted here was to extend the scope of the collection, first by widen-
ing its coverage to include all shipbuilding materials, and then by taxing
shipping itself. The 3.33 per cent rate of collection was applied to all
civilian ships in transit, being calculated on the original cost of the ship.
There is no indication of when this started nor of how permission was
obtained. Even before 1600, the tax assessment took into account the ship's
hull, mast, anchor, and other fittings, and the process thus developed into
a kind of inspection and registration of shipping. Whenever a ship was sold
in the Huai-an area the buyer likewise had to pay the office 3.33 per cent
of the purchase price.73 These practices evidently survived the dynasty, for
the 1778 edition of the Huai-kuan Tung-chih indicates that even under the
Ch'ing the forest produce levy at Huai-an, still managed by the ministry
of works, was collected on shipping as before. After the payment was made
the office branded the lintel of the main cabin of the vessel and gave the
owner a certificate.74 The term' forest produce levy' therefore began to take
on a wider meaning and came to compete with the tonnage dues which
were still levied by the customs house in the same port, under the juris-
diction of the ministry of revenue.
There is no unified account of the various forest produce levy stations,
and never any mention of the actual level of the quotas imposed. The
collection was further complicated in the late sixteenth century by the
continual palace construction in Peking and the vast procurement of timber
which it required. This certainly affected the operations of the stations. For
instance when it was discovered by a ministerial official in 1596 that
44,000 logs were being shipped by merchants to Peking, instructions
were issued to the stations to intercept and tax this cargo en route for the
latest palace construction project. Payments were once more to be made
in kind and in addition purchases were made from the shipment.75 It was
even claimed that 'because of the enormous taxes, many timber merchants
left the business'.76
Information on the income obtained from this source is scattered and
fragmentary. In 1489, the Ch'ing-chiang-p'u dockyard was recorded as
having received payments from four stations - Sha-shih, Wu-hu, Hang-
chou and Huai-an - totalling 28,670 taels.77 In 1484 Hang-chou had an
annual income of 23,000 taels78 and in 1525, the Wu-hu station produced
a revenue of more than 20,000 taels. One year the collection is said to have
exceeded 37,000 taels.79 The revenue at Huai-an in 1608 is estimated to have
240 Miscellaneous incomes
been approximately 11,500 taels.80 Contemporary sources suggest that for
most of the later sixteenth century the collection quotas remained largely
unchanged.81 It is therefore probable that the projected income of the four
stations was about 75,000 taels. The annual income of Lung-chiang station
near Nanking, which collected its revenues in kind, must have been roughly
equal to those of the other stations and was thus very likely worth 20,000
taels or more.

(h) Government mining (k'uang-yiri)

At the beginning of the dynasty mining was strictly monopolized by the
state. Government iron production in the 1370s averaged 8 million catties
a year, or approximately 5,300 short tons. In 1395, the year the monopoly
was abolished, government reserves amounted to about 25,000 tons. Iron
production thereafter carried on privately was to be taxed at one-fifteenth
of the output, paid in kind,82 though there is no evidence that this scheme
of taxation was actually put into effect.
From the early fifteenth century the government concentrated its own
iron production at Tseng-hua, a town on the Great Wall northeast of
Peking. In 1509 the iron works there produced about 320 tons of pig iron,
140 tons of wrought iron, and 5 tons of steel. After 1529 production de-
clined to about two-thirds of the previous level but it was still sufficient to
meet the needs of the ministry of works. By 1581 the iron ores in the area
were exhausted and production was suspended.83
In the late sixteenth and early seventeenth centuries the government
collected each year between 300 and 450 tons of iron from Chekiang,
Fukien, and Kwangtung, though very little from other provinces. From
time to time it also purchased iron from private producers.84
It seems that three provinces which supplied the government did not
exercise any control over production. In Ch'ao-chou prefecture, Kwang-
tung, there were approximately sixty iron furnaces. Their tax payments,
collected in silver and amounting to only slightly more than 1,000 taels,
were allocated in advance as provincial military supplies.85 In Fukien,
which made the largest deliveries to Peking, the situation was much the
same. In Ch'iian-chou prefecture, Fukien, the iron quota was filled by
means of an extra collection on the land taxes. Chang-chou prefecture
made an annual collection of 290 taels or so from the district's thirty
privately-owned iron furnaces. This revenue was likewise allocated to the
provincial military authorities. 86 Undoubtedly the iron that these provinces
delivered to Peking was purchased with funds derived from other sources.
The government did not control the supply of copper either. As it was
uneconomical to transport the copper mined in Yunnan to the coastal
areas, most of the copper sold in the southeastern provinces came from
imports, mainly from Japan.87 In addition the government each year
 6,1 Revenues from commerce and industry 241
obtained about 4 tons of copper from the fish duty (see below, (/)). The
ministry of works is known to have purchased up to 24 tons annually on
the market.88
More important to the government was silver mining. In the late Ming
there was considerable controversy over the wisdom of promoting this,
ostensibly for reasons connected with traditional geomantic practices. In
the seventeenth century when Adam Schall von Bell, the court's Jesuit
adviser, proposed to improve the government's shaky finances by reviving
silver mining, his suggestion was opposed by several influential officials
including Minister of Revenue Ni Yuan-lu. They argued that mining would
raze houses and disturb graves, and 'damage other geomantic features'. 89
One problem that the Ming never solved was that of security, in parti-
cular the control of the mining labor force. This arose basically from the
nature of government mining which, from the Yung-lo period onwards,
was never run on business lines. The government was never willing to
allocate sufficient funds to develop it systematically and instead resorted
to a practice known as chia-pan, whereby the officials and workers at each
mine had to meet a fixed production quota. 90 The government's demand
for metal increased rapidly, even before the mines had been fully opened,
and it therefore did not permit the output to be used for re-investment.
When the output failed to meet the prescribed quota the latter could not be
reduced without the emperor's permission. Frequently the shortages had
to be made up by the local administration with funds derived from other
sources. Government mining thus in effect led to a new form of taxation.
All of this was noted in the seventeenth century by Sung Ying-hsing, the
author of the celebrated technological treatise, Tien-kung K'ai-wu.91
The miners themselves were hastily recruited from migrant laborers. It
was therefore difficult to control them by the principle of group responsi-
bility, the only effective means of control that the government exercised
over the local population, because they had no roots in the locality. Such
a large labor force assembled in a rural area was always felt to represent
a threat to security. Even the mobilization of corvee labor to carry out
huge water-control projects caused some concern to officialdom. With
mining labor the risks were greater. The miners got very low wages and
were given no severance pay if the project failed. Being displaced persons
with some skill in manufacturing weapons, they were apt to take to
banditry. The reason why the Ming government had some success in
operating the iron works at Tseng-hua was because they were partially
under military control.
The problem could not be solved by granting mining rights to the
general populace. While limited numbers of iron works could be operated
by private enterprise under governmental supervision, to permit the private
working of silver mines was incompatible with the ideological principle
242 Miscellaneous incomes
that forebade excessive exploitation of natural resources by private in-
dividuals. Most of the silver mines, moreover, were situated in desolate
areas, such as the border region between Kiangsi and Chekiang. To organize
industrial labor in such isolated regions and to provide it with adequate
services and support, particularly in view of the poor transport facilities of
the time, was likewise beyond the capacity of the Ming entrepreneurs.
In the fifteenth century silver mining was on occasion undertaken by
dubious characters who had neither the capital nor the organizing ability
to carry it out properly. These projects usually ended in armed rebellion by
the miners. One historic lesson that the Ming government never forgot was
provided in the mid fifteenth century by the case of Yeh Tsung-liu who,
after failing as a mining entrepreneur, became a brigand leader in the
border region between Kiangsi and Chekiang. It took the government five
years, from 1444 to 1449, to put down the rebellion.92 Silver mining was
banned in the area thereafter. In one of the areas affected by Yeh's
rebellion, Shang-jao county, Kiangsi, the whole mining area was sub-
sequently closed and all the mines sealed off. The inhabitants were re-
settled elsewhere and their farmland abandoned. The small amount of land
taxes thus lost was made up by extra collections elsewhere. Stone and brick
walls were built to cut off the roads leading to the area.93 Fears of a re-
currence of this kind of disorder proved a constant hindrance to the
government's efforts at silver mining in the sixteenth century.
Several attempts were made during this period to increase the revenue
from silver mining; each was short-lived and had undesirable repercussions.
In the 1500s the eunuch official Liu Chin94 assigned quotas for silver pro-
duction to Fukien, Szechuan, and Yunnan. When the authorities in
Chekiang pleaded that the province's silver deposits were exhausted, they
were then ordered to hand over 20,000 taels of silver from the funds derived
from fines and commutation of punishments (see (m)) instead.95 After
Liu's downfall in 1510 all the quotas were ended. Only the Yunnan silver
mines apparently went on being worked; they were closed on government
orders in 1521.96
In the 1550s the Chia-ching Emperor sanctioned further efforts to
promote silver mining, in order to provide funds for palace construction.
New mines in addition to those in Szechuan and Yunnan were opened up
in Shantung, Honan, Shansi, and North Chihli. Early in 1559 it was decided
that all mineral resources should be thrown open to private exploitation,
on payment of 40 per cent of the production to the government.97 Before
this policy had time to show results the miners in the Chekiang-Kiangsi
border again turned to large-scale banditry, and in 1566 attacked and
seized the city of Wu-yiian in South Chihli.98 The court therefore in 1568
renewed the previous prohibition and also placed the whole mining area
under stringent surveillance. In the three provinces of Chekiang, Kiangsi,
 6,1 Revenues from commerce and industry 243
and South Chihli, orders were issued, and engraved on stone, prohibiting
anyone from entering the area. To aid the provincial security forces to
maintain the blockade, the government issued a publication containing
maps showing details of pathways and mountain passes, and other
strategic information." Shantung province similarly published a list of all
its gold, silver, copper and tin mines, noting that all of them had been
'firmly sealed off', and that some were guarded by troops.100 The situation
remained largely unchanged until the Wan-li Emperor gave permission to
resume mining in the 1590s (7, m).
Although there are no adequate statistics to show the government's in-
come from mining, its upper limits can be estimated with some accuracy.
In 1548 Minister of Works Wen Ming (in office, 1547-9) reported that the
total income from mining in that year was 62,030 taels, a figure which may
have reflected a general quota.101 Even in 1557, when mining was being en-
couraged, the total amount of silver made over to the government by Shan-
tung, Shansi, Honan, Szechuan, and Yunnan provinces, and by Pao-ting
prefecture in North Chihli was only 48,271 taels.102 Though the figures for
the southeastern provinces are unknown they are unlikely to have exceeded
this sum. In 1601 Wan-li's eunuch supervisors made sixteen deliveries of
'revenues from mining', which totalled 110,210 taels.103
The only area where government mining activities showed some signs of
expansion was Yunnan. Its governor reported that whereas up to 1594 the
annual income from mining was 52,722 taels it increased thereafter to
83,600 taels. In the last decade of the century when the province was en-
gaged in a border war with the Burmese leader Nanda Bayin, most of this
revenue was intercepted by provincial officials to pay for military supplies,104
and little of it was ever delivered to Peking.105 Taking this into account it is
probable that the total annual revenue from government mining in the
sixteenth century was less than 150,000 taels, and that for many years it
was well below this.
The revenues from mining were delivered to the Chieh-sheng Treasury
which was controlled by the ministry of works. It was thus not subject to
auditing by the ministry of revenue.

(0 Fish duty (yu-k'o)

In theory the fish duty was paid by fishermen. In areas where fishing was of
considerable importance fish-duty stations (ho-p'o so) were established
under the jurisdiction of the prefectural or county government.106 In all
other areas the prefects and county magistrates collected the payments,
which could be made either in grain, by-products of fishing such as glue,
shipbuilding materials, or silver. In 1382 when the system was first organized
252 stations were established. In 1578 there were still 139 stations, about
244 Miscellaneous incomes
half of which were in Hukwang province; the others were located in South
Chihli, Chekiang, Kiangsi, Fukien, Kwangtung, and Yunnan.107
Even in the fourteenth century the system was impractical. The fisher-
men, who moved about from place to place, found it easy to evade the
duties and the local authorities could do little about it.108 By the late
sixteenth century the fish duty was collected in a great variety of ways. In
Yung-chou prefecture, Hukwang, for instance, there were six counties and
one subprefecture. By 1571 only three of the counties continued to collect
the duty from fishermen while the other four made it up from the land
taxes.109 In general it seems certain that more counties adopted the latter
approach than the former.
The fish-duty stations collected fish-glue, hemp, iron, copper, varnish,
fung oil, and vermilion.110 Shipbuilding materials were included on the
grounds that they were essential to fishermen. The stations thus functioned
rather like customs houses in the minor inland ports. Items in kind could
be commuted when the ministry of works had an abundant stock of them.
The accounts sometimes confused even the officials appointed to administer
them. The fish-duty quota for 1578 is given in chapters 25, 36, and 200 of
the Ta-Ming Hui-tien. The payments were so fragmented and the deliveries
so scattered that the local officials had to be reminded to follow in detail
the instructions of Hui-tien.111 The total revenue and its allocation are
summarized in Table 17.
On the assumption that each picul of grain on delivery was worth
0.5 taels, the total revenue from the fish duty must have been just over
58,000 taels per annum. It was thus a sizeable item, comparable to the
revenue from the maritime tariff.

II ADMINISTRATIVE INCOME
(j) Sale of rank (k'ai-na shih-li)
The Ming sources are usually reticent concerning revenues derived from
administrative incomes because of their unethical nature. Sale of official
rank, for instance, was referred to ask'ai-nashih-li, literally'contributions
according to precedent'. These irregular revenues were nonetheless sub-
stantial, and in the sixteenth century the government could not easily have
done without them.
Though the sale of official rank was frequent it was never formally
institutionalized. The general practice was to grant the contributor an
honorary title, such as that of imperial university student or army officer.
He was thus exempted from the service levy and the exemption was some-
times extended to more than one member of his household. He usually
also received the cap and belt appropriate to the purchased rank, but under
 6, II Administrative income 245

TABLE 17. Collection offish duty, 1578

Collected in grain, combined with the land tax account, 31,966 piculs
and administered by the ministry of revenue
Collected in silver, delivered to the Nanking 11,000 tls.
ministry of revenue
Collected in silver, in lieu of shipbuilding materials, and 18,900 tls.
delivered to the ministry of works
Collected in paper currency and copper coins and delivered 6,000 tls.
to Kuang-huei Treasury, to the value of
Collected in materials useful to the ministry of works, 7,000 tls.
to the value of

normal circumstances was not entitled either to a salary or an actual

official post. There were exceptions to this general rule, and it was on
occasion possible for clerks to obtain promotion without examination,
civilians to become unpaid clerks in certain offices, qualified students to be
given priority for appointment, and suspended officers to be reinstated,
all on payment of a fee.112 The sale of rank was supervised either by the
ministry of revenue or by the ministry of works, some co-ordination
being exercised by the ministry of personnel. The revenues were used
by provincial officials, frontier governors, or ministerial officials in
charge of construction projects to provide emergency famine relief, obtain
urgent military supplies, and more often by the late sixteenth century, to
finance the building of mausoleums and palaces. The contributions were
usually made in cash, but at times payment in kind in such items as horses
and bricks was also accepted. The amounts varied widely. For a lesser rank
20 taels might be sufficient. Some late sixteenth-century tomb inscriptions
of wealthy merchants strongly suggest that they had been forced to purchase
the ranks of battalion and guard commanders,113 presumably for sub-
stantial sums. Hsu Cheng-ming in 1575 stated that the title of imperial
university student cost 350 taels.114 In 1596 the contributions solicited to
finance a major palace construction project were set in the range of 500 to
1,000 taels.115
There is no extant official account of the sale of rank, and it is unlikely
that any was ever published. But the level of the income was occasionally
revealed in both official and private papers. In 1508 it definitely exceeded
430,000 taels,116 and in 1565 Minister of Revenue Kao Yao (in office,
1560-7) reported that that year it amounted to 510,000 taels.117 Chang
Chii-cheng revealed in a personal letter that between 1570 and 1580 the
state obtained some 400,000 taels a year from this source. Evidently this
level was maintained continuously and the proceeds counted as a regular
budgetary item. Though Chang personally loathed the practice, he stressed
that the revenue it yielded was indispensable.118 The total amount of
246 Miscellaneous incomes
400,000 taels a year was in fact larger than any of the revenues derived from
industry and commerce.
These benefits were obtained at a price. Few of the so-called imperial
university students actually attended the university. They merely took
advantage of their official status to obstruct justice and evade taxes in their
home areas (3, i). Towards the end of the dynasty there might be as many
as 1,000 of these imperial university students in a prosperous county, and
local administration was considerably hampered by them.119 By this time
it was common for unqualified persons to purchase official posts, clerkships,
etc. in provincial and local governments. They obtained these sinecures,
such as that of 'honorary seal custodian' (chih-yiri), or 'unpaid super-
numerary' (cWen-cKai), purely for the sake of the opportunities of illicit
income.

(k) Ecclesiastical license fees (seng-tao tu-fieh)

In the early years of the dynasty the collection of registration fees from
Buddhist monks and Taoist priests was not regarded primarily as a source
of government revenue. In the Hung-wu period, registration took place
once every three years, between 300 and 500 clerics being licensed on each
occasion. During the Yung-lo period, licenses were granted once every five
years to as many as 10,000 persons at a time. Thereafter, up to the mid
fifteenth century the interval was set at ten years,120 but, after the 1460s, the
proceeding became irregular and was generally resorted to by the govern-
ment to provide emergency funds for famine relief. The ministry of rites
prepared blank certificates and distributed them to the provinces and
prefectures. Purchase of these, usually for a fee of 12 taels of silver, entitled
the applicant to registration as a cleric without taking the usual religious
examinations. In the year 1485 alone it is said that 70,000 certificates were
offered for sale,121 though the revenue obtained thereby is unknown. When
the government was particularly pressed for funds, the price of each
certificate was reduced to 7 or 8 taels. On the other hand the prospective
purchaser might pay as much as 100 taels to the intermediaries in the
transaction,122 in his anxiety to acquire the ecclesiastical status that auto-
matically entitled him to exemption from the service levy.123
There are practically no records of the sale of ecclesiastical license
certificates in the sixteenth century, but Chang Chii-cheng in a memorial
of 1578 confirms that the practice continued.124 In 1508 the certificates
were twice offered for sale and on one of these occasions the sale of 15,000
certificates brought in 130,000 taels.125 If this was typical then the practice
probably contributed at least 200,000 taels a year to state finances. It seems
that the only time it was ever suspended was in 1585 when there were
genuine fears that Buddhist influence was becoming too strong for
 6, II Administrative income 247
comfort. The move was made by Minister of Revenue Wang Ling (in office,
1583-5) who obtained the cooperation of the ministry of rites in checking
the expansion of religious establishments.126
The license fees and the revenues they yielded came mainly under the
jurisdiction of the ministry of revenue, though the revenue was occasion-
ally used by the ministry of rites for its own administrative expenses.127

(/) Payment for 'rationed salt' (hu-k'ou shih-yen-chiao)

This item has already been described (3, iv). In the sixteenth century the
category of 'rationed salt' was abolished and the collection became a poll
tax or a surtax on the land taxes. The revenue was in general shared
between the central and local government.128 Many Ming officials had
qualms of conscience about collecting this item, but in 1509 a proposal by
the eunuch Liu Chin that it should be eliminated altogether was defeated
by the ministry of revenue.129 The accounts of 1578 show that the imperial
government still in theory derived 80,555 taels of silver from it, though in
practice only about half that amount was actually delivered in silver to the
T'ai-ts'ang Treasury,130 while the rest was paid to the Kuang-huei Treasury
in copper coins and paper currency. The accounts of the T'ai-ts'ang Treasury
for 1580 indicate that in that year it received 46,897 taels of silver from this
source.131
Taking into consideration the fact that some of this revenue was ap-
propriated by the provincial governments, the total annual value of this
item in the late sixteenth century was probably something like 160,000 taels.

(m) Fines and commutation ofpunishments (tsang-fa)

The term tsang-fa in the sixteenth century carried a great variety of mean-
ings and touched on many aspects of Ming administration and juris-
prudence. Tsang literally means booty or stolen goods, while fa means
a fine or an act of forfeiture. The term can be better understood through
illustration. When a man in the sixteenth century was indicted for em-
bezzling public funds, the amount he had stolen and had to make up was
not necessarily assessed on the basis of the evidence, but by an adminis-
trative decision, and the recovery of the funds might well entail confiscation
of his property. The concepts of tsang and fa were thus both involved.
When for instance it was discovered that a clerk who had been in charge
of a treasury for four or five years had on one occasion stolen 20 taels of
silver, it was held to be quite justifiable, according to contemporary legal
practice, to force him to return a sum of 1,000 taels, as it was likely that his
thefts had been going on throughout his period of duty. If the amount
could not be made up by confiscation of his personal property, the assets
248 Miscellaneous incomes
of his relatives and close friends might also be seized, on the theory that the
embezzler had transferred the stolen goods to his associates. An outstand-
ing sixteenth-century case was the confiscation in 1565 of the property of
Grand-Secretary Yen Sung, which, as was pointed out by a censor-in-chief
twenty years later, 'poisoned the whole of Kiangsi province'.132
Even in ordinary cases the fine had many implications. It naturally
covered commutation of punishment, but not quite in the conventional
Western sense of the term. One point to be noted is that the government
made little distinction between civil and criminal cases. All cases coming
to trial involved some possibility of criminal proceedings. Private suits for
damages, when sustained, usually ended with public proceedings against
the defendant. Even in what might be considered strictly criminal cases,
proceedings were sometimes initiated by a plaintiff who in effect assumed
the modern role of public prosecutor. If he lost the case he was given the
same type of punishment that would otherwise have been meted out to the
defendant. For instance, a man who unjustly accused someone else of
a crime that carried the death penalty and failed to sustain the charge was
to be exiled over 3,000 li from his home district.133
Under this system a formal complaint, once lodged, could not be with-
drawn, and in theory no case could be settled out of court. The only way
for the plaintiff to stop the proceedings was to come forward himself and
confess that the charge was false. Whether the punishment was given to the
defendant or to the plaintiff, it could in most cases be commuted to a fine.134
For minor offences the fines were light enough to appear more like court
expenses in the Western system. One of the major purposes was to dis-
courage lawsuits of any kind. Strictly speaking the fines resulting from
lawsuits should be called shu-huan (redeeming payments), but they were
also covered by the general term tsang-fa.
Another practice established during the early years of the dynasty was
that the stationery required by government offices was supplied by persons
involved in court cases. By the sixteenth century, the stationery was actually
procured with funds derived from other sources, but the collection from
prisoners, defendants, and plaintiffs in pending cases was never completely
discontinued. The payment, known as chih-shu (paper redemption),
became another component of tsang-fa.135
Fines also included penalties for administrative shortcomings and
failures. The compilation of the Yellow Book always led to penalties of this
kind. Though deliberate falsehood in the population data was not easily
detected by the higher offices, statistical inconsistencies were constantly
being uncovered. As the entire set of books for the county or prefecture
was invalidated by this kind of technical error a new set had to be submitted
at considerable expense (2, n). Penalties were imposed on the offices or the
districts guilty of negligence in order to pay the extra costs, yet the funds
 6, II Administrative income 249
thus collected were not actually used to pay for the preparation of the
second set of books but delivered instead to the higher offices and added
to the tsang-fa accounts.136 In the late sixteenth century the local officials
applied the same method in collecting fines. In some districts fines were
collected in advance even before any penalty was announced. It was quite
common for a village community or a tax agent to be 'fined' two reams
of paper though no error had been committed.137
In brief, the fiscal term tsang-fa covered many different types of incomes,
both regular and irregular, judicial and executive, payments by individuals
and by groups, and both huge and trifling sums. In the sixteenth century
practically all officials with any financial responsibility, including governors,
prefects, county magistrates, surveillance commissioners, and some army
officers could impose such fines. In 1593 Minister of Justice Sun P'i-yang
(in office, 1592-3) revealed that there were twenty-eight methods of collect-
ing fines, some legal and some illegal, but all justified by one precedent or
another.138 Since it was impossible to audit the accounts properly the
matter of legality was only relative. The real issue was whether the collection
was excessive or moderate.
Even in the early years of the century the central government required
provincial officials to surrender some of the revenue derived from this
source. The demands made by the eunuch Liu Chin in Chekiang in 1509
(see above, (/*)) are an example. In 1564 an imperial decree further ordered
that all offices deliver 40 per cent of the tsang-fa funds to the ministry of
revenue and 40 per cent to the ministry of works, and that the remaining
20 per cent be kept by the local government to provide for famine relief.139
Apparently there was a quota system. In 1567 the tsang-fa funds previously
administered by the ministry of justice were likewise turned over to the
ministry of revenue.140 The accounts of the T'ai-ts'ang Treasury for 1580
show that it received 128,617 taels of silver from various agencies.141
The total collection across the empire should thus have exceeded 300,000
taels. In 1569 Ho Liang-chiin estimated that the entire revenue would
purchase 700,000 piculs of unhusked rice,142 which would also put it at
between 200,000 and 300,000 taels.
The 20 per cent of the revenue retained locally was not in fact used to
provide for famine relief, as was admitted in a conversation between Chang
Chii-cheng and the Wan-li Emperor in 1581.143 Until the imposition of
melting charges on tax payments provided local officials with a new source
of personal income, tsang-fa undoubtedly remained essential to their
support. Even so, not all these funds were pocketed by the officials. Some
utilized the surplus to cover any fiscal deficits of their offices, while others
used it to finance projects of public benefit, such as the publication of
educational materials.144
250 Miscellaneous incomes

(n) Profits from minting {chu-cKien li-yuri)

The question of minting has already been discussed at some length (2, iv).
In the late sixteenth century it was continually asserted that the purpose of
the operation was to make a profit, and that when this failed to materialize
it was suspended. The data are by no means complete but on the basis of
several different descriptions, the general situation can be summarized as
follows: The production of 10,000 copper coins was reckoned to cost
between 14.4 and 14.9 taels of silver. This did not include labor costs
because the workers were not specially hired but were permanently on
payroll. The market value of the coins in Peking fluctuated between 450
and 700 to the tael. The latter rate approximated to the production cost.
Government-minted coins were released at the rate of 500 or 550 to the
tael, making a profit of between 25 and 40 per cent of the original invest-
ment. This profit was not converted to silver. Instead the copper cash were
used immediately to pay government construction-workers and some official
salaries, thus directly reducing the government's financial obligations.145
Little is known about the ministry operation of 1576, in which 100
million coins were produced, and the copper-silver exchange rate cannot
be ascertained.146 The records of the 1596 operation are more complete. On
this occasion the mint produced 82,800,000 coins at a profit of approxi-
mately 30,000 taels of silver but as the rate of exchange became un-
favorable thereafter, no further production was undertaken.147
Conditions in the provinces, as opposed to those at the capital, varied
widely. The profit margins were determined by local copper prices and the
market-acceptability of the coins minted. The Huai-an prefectural mint in
South Chihli made a 40 per cent profit in the early seventeenth century,148
and the Shansi provincial mint is reported to have made an annual profit
of about 100 per cent in the 1620s. Its total profit over a decade was
117,090 taels of silver.149 Some of the large profits of provincial mints were
obtained by squeezing the merchants. In a memorial written about 1580
it was revealed that the officials forced the local merchants to supply them
with materials at below cost-price and subsequently forced another group
of merchants to accept the minted coins at a rate higher than the market
exchange rate.150

(o) The ' common post money' (chuang-p'eng yin)

The common post money originated from the principle that army personnel
should accept financial responsibility for the equipment loaned to them.
Frontier soldiers had to keep their horses fit for service over a period of
fifteen years. If a horse died before this, unless killed in battle, its rider and
his officers were responsible for making good the loss on the basis of the
 6, II Administrative income 251
number of years it had yet to serve. Since individual soldiers were unlikely
to have sufficient savings to pay for the losses, the army units deducted
part of their pay in advance to provide a sinking fund, rather like collecting
group insurance premiums. Sometimes the officers also made contributions
to the fund. This in the fifteenth century gradually evolved into the in-
stitution known as the common post money.151 Around 1477 it was officially
recognized by the government and soon fixed rates were prescribed for its
collection.152 In 1492 a censorial official reported that the general practice
was now for each cavalryman to contribute 0.3 taels of silver half-yearly.153
By the end of the fifteenth century it was said that the common post
money was being mismanaged and was used only to provide extra funds
for the senior officers. In 1506 the ministry of war secured the emperor's
permission to transfer the collected funds to the Ch'ang-yin Treasury. It
thus came under the control of the court of imperial stud, which was
responsible for replacing the horses.154 Thereafter the common post money
became a regular income of the ministry.155
In the sixteenth century this revenue was derived from another source
besides the original one, namely the income from small-scale farming
projects developed by the various frontier commands.156 The central
government was thus taxing the army posts while its main function was
to see them properly supplied.
The common post money was mixed with the other revenues of the
ministry of war and no separate account has been formed for it. In 1568
the Shansi office of the court of imperial stud indicated that its annual
collection was 19,060 taels.157 If the other commands collected similar
amounts the total must have been in the region of 50,000 taels. Owing to
increasing corruption in frontier-army administration in the late sixteenth
century, it is unlikely that the actual collection ever came up to the desired
level. One report reveals that even in 1595 at some army posts the cavalry-
men were still required to pay for their lost mounts.158

(p) Incense fees at national shrines (hsiang-shui)

Incense fees were always included in the accounts of the central govern-
ment, presumably because the income, though small, was reliable and
therefore valued. The two major shrines constantly mentioned in the
sources, namely T'ai-shan in Shantung and T'ai-ho-shan in Hukwang, were
nominally under the control of the ministry of rites, which was responsible
for the maintenance of the temples there. In practice the fees were collected
by the eunuchs residing at the shrines. At T'ai-ho-shan the eunuch custo-
dians delegated the collection to the army officers on guard at the shrine.
From contemporary descriptions it seems that the collection from
worshippers at these two places must have exceeded 40,000 taels a year.
252 Miscellaneous incomes
The collectors at T'ai-ho-shan, who in effect operated under a tax-farming
system, delivered 'several thousand taels' a year to the nearby Chiin-chou
battalion as pay for the soldiers. This amount was said to constitute only
40 per cent of the total collection.159 The collectors at T'ai-shan, even after
delivering an unspecified amount to the provincial government, still
offered to make an annual payment to the ministry of revenue's T'ai-ts'ang
Treasury. Throughout the late sixteenth and early seventeenth centuries
this was always listed as 20,000 taels.160

Ill CASH INCOME FROM COMMUTATION OF

SERVICES AND SUPPLIES
(q) 'Speed-the-delivery money9 (cWing-chi yiri)
The commutation of services and supplies has already been discussed at
some length and needs no elaborate explanation here. From the distri-
bution of funds it has been shown how the various government agencies
retained their earlier sources of income, originally paid in kind, and simply
converted them into separate cash accounts. This system of departmental
management of revenues lasted throughout the dynasty.
Another significant item handled by the ministry of revenue, besides the
commuted portions of the regular land taxes and the service levy, was the
ctting-chi yin. The term cKing-chi literally means 'lightly handed-over'.
That portion of the land taxes designated as tribute grain bore heavy sur-
charges, partly in order to pay the cost of transporting it through locks on
its way to Peking. In the early fifteenth century these surcharges, as well as
the tax payments proper, were paid exclusively in grain. From 1477 part
of the surcharges was commuted to silver, the total collection from the
entire consignment of tribute grain being 450,000 taels which was still used
to pay for the transfers.161 This arrangement accorded with the maxim:
'When the load is lightly handed-over, the goods will easily reach the
destination.' But the silver payment, hereafter referred to as the 'speed-the-
delivery money', was not handed over by the taxpayers to the soldiers on
transportation duty, but was collected as a lump sum by the various
counties and delivered to Huai-an, where the Grand Canal commissioner
distributed it to the transportation corps as the grain boats came through
the city. There were probably several reasons for this procedure. If the
soldiers received silver payments directly from the taxpayers they might
argue over the weight and quality of the silver and might also be tempted
to spend the money before it was needed. Making the payment at Huai-an
on the other hand would offer an additional incentive for the grain boats
to arrive there early.
As things actually turned out the procedure merely made it easy for the
 6, III Cash income from commutations 253
government to intercept the funds and use them for other purposes, for by
the early sixteenth century the cash payments were beginning to be with-
held from the soldiers. How this started is not at all clear. The official
explanation was that because the Grand Canal had by then been immensely
improved, the cumbersome procedure of transferring the grain through
locks was no longer necessary and the funds thus saved could be used
instead for the maintenance of the waterway.162 In practice the funds were
very seldom spent on canal maintenance. One private account of around
1511-12 indicates that many senior army officers in command of the grain
fleet were pocketing the speed-the-delivery money due to the soldiers and
that influential persons in Peking were also involved in the scandal. Under
the threat of investigation, one of the guilty officers surrendered some of
the proceeds to the government, declaring them to be surplus funds.
Others followed suit. Soon this practice of handing over the cash to the
treasury became official.163 In 1522 the commander-in-chief of the trans-
portation corps appealed on behalf of the soldiers for the payment to be
distributed as before. As it had by now become customary procedure for
the government to divert the funds to itself, the request was refused by the
ministry of revenue.164 Throughout the sixteenth century the general
practice was for 70 per cent of the speed-the-delivery money to be sur-
rendered to the government and the remaining portions to be distributed
to the soldiers.165 In 1558 when military expenditures reached a high point,
half of the speed-the-delivery money was submitted to the T'ai-ts'ang
Treasury.166
In the late sixteenth century when the situation was more settled, the
greater portion of the funds was allocated to pay the administrative ex-
penses of the granaries around Peking. The granary superintendent reported
in 1579 that the six major depots under his jurisdiction had a combined
expenses budget of 218,971 taels. Most of the money was spent on the
short-distance haulage of grain beyond the terminal of the Grand Canal.
In the same report he said that he expected an annual surplus of 119,088
taels from the speed-the-delivery money, which would be handed over to
the T'ai-ts'ang Treasury.167

(r) Artisan payment (chiang-yin)

The artisans working for the manufacturing and processing plants in
the palace compound included carpenters, blacksmiths, tailors, leather-
workers, etc. In accordance with the regulations laid down in the early
years of the dynasty, these duties were performed without pay by persons
registered as artisans in the respective trades, who in this way fulfilled their
service obligations. Some of them reported for duty annually, and others
at two, three, four, or five-year intervals. The service, performed in Nanking
254 Miscellaneous incomes
or Peking, usually lasted for three months. Travel expenses were provided
by the workers themselves. Though the government paid them no wages,
food allowances were provided.
After 1562 they were not permitted to answer the service calls in person,
and instead all their service obligations were commuted to silver in the
form of one annual payment. According to the 1578 account there were
142,486 such statutory laborers throughout the empire.168 Their payments,
termed 'artisan payments', were collected by the local officials and de-
livered in lump sums to the Chieh-sheng Treasury which was controlled by
the ministry of works. The total revenue was fixed at 64,117 taels. Some
districts, as has already been noted, simply added these payments to the
general service levy accounts instead of collecting them from individual
artisans (3, iv).
The same procedure was applied to other government establishments
outside the two capitals, such as the charcoal factory at I-chou and the
shipyard at Ch'ing-chiang-p'u (2, i), except that in these cases the payments
were delivered directly to the manufacturing establishments without being
handled by the ministry.169 In theory at least the wages of the hired workers
in government establishments were paid by the registered artisans in the
same trades.

0) Reeds tax ilu-k'd)

The reeds tax provided an annual income of 25,500 taels of silver. It was
administered by the Nanking ministry of works and collected on agri-
cultural land along the Yangtze river in Hukwang, Kiangsi, and South
Chihli. In the fifteenth century the collection was made in kind, the reeds
being used as fuel. In the sixteenth century the payments were commuted
to silver and persons who took possession of the waste land on the river
banks and islands thus became rate-payers. Though the collection was in
effect another variant of the land taxes, the acreage and the revenues in-
volved were not integrated with the regular land taxes. The ownership of
the land was left ambiguous, being neither private nor public. Further-
more, owing to topographical changes the original assessment of acreage
did not correspond to conditions in the late sixteenth century. As by that
time large tracts of reed-growing land had already been converted to rice
paddies there was a general feeling that the rate-payers were under-taxed.170
In 1597 the Wan-li Emperor arbitrarily ordered that the quota be raised
to 200,000 taels.171 As it is unlikely that the payments of individual land
holders could have been suddenly increased by eight times, it is probable
the county magistrates had to raise the funds by reallocation of existing
revenues.
 6, III Cash income from commutations 255

(t) Supplies to the four bureaus (ssu-ssu liao-chia)

This item, which provided 500,000 taels a year, was the largest account
administered by the ministry of works. The collection covered all the
provinces and prefectures except Kwangsi, Yunnan, and Kweichow (3, iv).
In the late sixteenth and early seventeenth centuries, the ministry was
authorized to make further collections in the name of 'river-control
expenses' (ho-kung) and 'contributions to construction projects' (chu-
kung).112 These two items were regarded as irregular. The former was
assessed as a surtax on the land taxes in the northern provinces, while the
latter was a percentage added to many existing revenues.

(u) Horse payment (ma-chia)

This item has already been described in some detail (3, n), and was the
largest account administered by the ministry of war. A report of 1588 put
it at approximately 370,000 taels,173 while in 1629 the collection quota was
stated to be 429,537 taels.174

(v) Commutation oj'capitalguardduty (pan-chun che-yin); (w) Commutation

of personal attendance (tsao-li che-yin); (x) Savings from postal service
(i-cKuan yin).
These three items, administered by the ministry of war, were all insigni-
ficant in the sixteenth century. The first two simply had little growth
potential while the last item, though it could have been of some importance,
was never integrated into any general account.
Capital guard duty was performed by soldiers from Shantung and
Honan. In theory they were to reinforce Peking's defenses in the spring and
autumn when nomad incursions were most likely. In reality as soon as the
soldiers reported for duty they were set to work on construction projects.
Payment was sometimes accepted in lieu of personal service. In 1540, for
instance, 46,000 soldiers failed to report for duty; they were ordered to pay
a 'fine' of 1.2 taels each. The total payment should therefore have been
55,200 taels.175 Later in the century discipline seems to have been
somewhat tightened up but the commuted payment remained a regular
item in the ministry's accounts. Those subject to tsao-li, or personal
attendance, were no longer required to serve in person from the fifteenth
century onwards and the commuted payments were used to supplement
official salaries in the capital (2, i). Since in 1487 the total tsao-li quota was
already 7,342 men, it is probable that by the sixteenth century the number
had reached at least 10,000.176 At 12 taels per man per year, the total
revenue should have been no less than 120,000 taels. Most of it, however,
was distributed directly to the officials. The only revenue of this type that
256 Miscellaneous incomes
actually accrued to the ministry of war was savings on salaries from unfilled
official posts, which was a variable item and never a significantly large sum.
Savings from the postal service, on the other hand, were quite substantial.
When Chang Chii-cheng insisted on a reduction of postal services in the
1570s most districts complied, reporting that the saving on expenses would
result in their postal-service collections being reduced (7, iv).177 All the
evidence indicates that the reductions were never made and that in most
places collection at the previous rate continued. These savings must have
constituted the greater part of the treasury reserves in the provinces. In
1598 when the central government ordered the provinces to deliver their
reserves to the capital, they were described as ' savings from postal services
accumulated since the early Wan-li period'. 178 The actual amount cannot
be ascertained.
(y) Calendar paper (li-jih); (z) Kitchen service fees due to court of
imperial entertainments (kuang-lu-ssu cKu-liao)
Three special agencies subordinate to the ministry of rites maintained
separate expense accounts which were never combined with other revenues.
These were for the botanic and mineral drugs of the imperial academy of
medicine, calendar paper for the imperial bureau of astronomy, and kitchen
service expenses for the court of imperial entertainments. These items
appeared in the regular fiscal reports, along with the revenues from the
li-chia account in all the provinces and prefectures.
Only a small quantity of the drugs levied from the general population
was commuted to monetary payments in the sixteenth century. The drugs
amounted in total to around 80 short tons, but their cash value was
never worked out, though many local accounts listed the cost of their
shares of contribution. Even in the fifteenth century the imperial calendar
was printed in over a half million copies, but the cost of the paper could
not have been very large either, compared with that of other supplies.179
The most significant item was the kitchen expenses.
In the fifteenth century the court of imperial entertainments employed
between 7,000 and 9,000 cooks (2, i). As it had to be able to serve daily
banquets for 15,000 persons and also distribute wine and meat in large
quantities,180 its expenses became a source of constant concern to the
government. The court obtained rice, salt, wine, and utensils from govern-
ment depots and its cooks were paid by the ministry of revenue, usually in
grain. In addition it received a small amount of cash from the eunuchs in
charge of the city gates of Peking (see above, (&)). Nonetheless its annual
cash expenses were still enormous, between 260,000 and 400,000 taels in the
late sixteenth century.181 These expenses were supplied by the provincial
and local governments and remained a separate item of revenue. It was
larger than that from mining and the maritime tariff combined.
 6, IV Non-cash income 257

IV NON-CASH INCOME
(aa) Income from the tea and horse trade (ch'a-ma)
Though the frontier trade whereby the Chinese obtained horses from the
nomadic tribes in exchange for tea remains a topic of considerable interest,
and its investigation will undoubtedly yield much useful information
on sino-barbarian relations, it must be pointed out that the fiscal im-
portance of the trade was small. Its contributions to the government's
finances, while of some significance, was by no means crucial. There is no
doubt that in the sixteenth century the trade was conducted almost entirely
for the sake of the revenues it provided. As has recently been pointed out
by Morris Rossabi, there is little substance to the claim that the Ming
maintained control over the tea trade as a means of pacifying the nomads
on their western borders.182
The trade was bound up with the Ming tea monopoly. At the founding
of the dynasty, tea was put under strict government control on the same
lines as the salt monopoly with a similar system of yin licenses and checking
stations (p'i-yen-so) encircling the principal areas of production. The
monopoly on tea was, however, much less effective than that on salt,
because its production was too scattered to enable the government to
exercise direct control.
According to the system devised during the Hung-wu period tea was
divided into two categories. That grown in all provinces other than Shensi
and Szechuan was circulated for domestic consumption, but had to be
accompanied by a yin license. In Shensi and Szechuan the government
obtained tea from three sources. A certain number of tea plantations were
government owned, and worked by army personnel. The authorities took
80 per cent of the tea produced while the other 20 per cent was used to pay
the soldiers. On all private production of tea a 10 per cent tax was levied in
kind. The government also purchased more tea from private growers, paying
for it in paper currency. Any tea left over after this could be sold on the
market on the purchase of a yin license as in other provinces. Furthermore,
no household was allowed to store more than a month's supply of tea.183
The government by these means obtained 1 million catties of tea a year
in Szechuan and 26,862 catties in Shensi. All of it was traded for horses
with the nomadic peoples on the western frontier, generally referred to by
Ming officials as 'fan tribes'. In the sixteenth centuryfifty-sevenof them
were listed as participating in the trade. They seem to have represented
many different ethnic groups184 but most of them were probably either
Tibetans or Turkish-speaking peoples from Kokonor and the present
Sinkiang. In the early years of the dynasty they were tributary vassals of
the Ming and were issued with golden tokens to use as credentials at the
9 HTF
258 Miscellaneous incomes
trading posts. In theory, trade for the tribal chiefs was both a privilege and
an obligation. The Ming government established four tea and horse-trading
offices {cWa-ma-ssu) in Shensi and one in Szechuan, and at one time managed
to obtain more than 13,000 horses annually through the system.185
The greatest difficulty in the early years of the system was transportation.
While most of the tea was produced in Szechuan, the trade was carried on
in north Shensi. In Ch'eng-tu 210,000 catties of tea had to be destroyed
in 1425 as it had rotted while awaiting transportation to the frontier.186
After 1435 the government experimented with the so-called k'ai-chung
method, which bore some resemblance to the barter system adopted by the
salt monopoly (5, i). In this case the merchant was not required to invest
any capital apart from the cost of transportation. At Ch'eng-tu he received
100 catties of tea from the government warehouse and on delivering it to
one of the trading offices in Shensi was given a promissory note for about
2,000 catties of salt, cashable in the Liang-Huai or Liang-che region.187
This scheme appears to have had no great success. Sometimes the tea was
transported northwards by the army188 but owing to the rapid decline of
the wei-so system this solution was equally ineffective. The cost of transport
over difficult mountainous terrain in any case made the whole operation
uneconomical.
In the late fifteenth century the government virtually lost control over
the trade and its procurement of tea declined substantially. No attempt
was made to tax the tea that was being grown in increasing quantities by
the population of Shensi, and contraband trade flourished.189 In 1490,
however, the government adopted a different approach to the tea and
horse trade. There was thenceforth less emphasis on government pro-
duction and procurement, and an attempt was made instead to tax the tea
merchants in the frontier region. Private traders were encouraged to pur-
chase tea from the producers, and the government issued them with road
passes. On arriving at the trading offices the merchants surrendered 40 per
cent of their stock to the government and kept the rest for private trade.
This procedure, also somewhat confusingly known as k'ai-chung, was put
into effect only sporadically. Owing to the lack of a capable administrator
in Shensi and the general decline in leadership at Peking, the operations of
the frontier trading-posts went on declining for at least another decade.190
The revival of the official tea and horse trade was said to be due to the
efforts of Yang I-ch'ing (1454-1530, minister of revenue, 1510-11), who
arrived in Shensi early in 1503 as vice-censor-in-chief and superintendent
of horse administration. From that position he rose to be governor of the
province and later governor-general. He remained in Shensi for eight years,
being recalled to Peking in 1510. Yang put an end to the contraband trade,
increased the taxes on the local tea plantations, and required the tribal
leaders to resume the trade as before. Yet the basic policy as laid down in
 6, IV Non-cash income 259
1490 was still to tax the merchants. In his report to the emperor in 1504
Yang described his plan as follows: Merchants were authorized to pur-
chase tea from producers and then deliver their entire stock to the trad-
ing offices. From the government sale one-third of the proceeds was
paid to the merchants while the other two-thirds were appropriated out-
right by the trading offices.191 This official tax rate of 66.6 per cent was the
highest ever recorded in Ming times, and no doubt met with some resistance
from the merchants. After 1506 the amount to be surrendered to the
government was reduced to 50 per cent and the merchants were allowed to
sell the other 50 per cent freely in the frontier region.192 The purpose of the
high tax rate was not simply to accumulate more tea but, as Yang admitted
in his memorials, to increase the value of the commodity as well. This could
only be achieved by limiting the quantity for export and deliberately adding
to the cost of the commodity.
His solution essentially was to relax some of the restrictions on the
domestic market and tighten the controls on the frontier. The trading
offices operated rather like customs houses and the 50 per cent taxation was
in reality an export duty, which took the form of a protective tariff in
reverse. Nothing is known of the private trade of the tea merchants;
presumably they too bartered tea for horses, as the sources insist that no
other merchandise of any significance could be obtained from the bar-
barians,193 while constantly mentioning 'contraband horses'.194 Since they
paid no duty on their imports the merchants should have been able to make
a profit even after surrendering half of their stock to the government. Yang
I-ch'ing's regulations, with few modifications, survived to the end of the
century. In 1571 a scheme was initiated whereby each tea merchant divided
his stock into two parts and the trading office drew lots for which part it
would take.195
The tea and horse trade in the sixteenth century thus required hardly any
investment by the government. In 1507 Yang revealed that he had accumu-
lated some 450,000 catties of tea at the trading-posts, sufficient for two
years' trade.196 Though the total tax levied on the tea plantations in Shensi
had grown from 26,289 catties in 1503 to 37,195 catties in 1505, and
50,965 catties in 1506,197 the amount was small; it is clear that the great
part of Yang's stock must have come from taxation on the tea merchants.
Trade with the frontier tribesmen was conducted once every three years.
The horses were valued in three different categories but the average rate of
exchange remained about 70 catties of tea per horse, much the same as in
the Hung-wu reign. In the early part of the sixteenth century the govern-
ment obtained between 3,000 and 4,000 horses a year by these means.198
Despite the unsettled situation on the northwest frontier, the trade
actually expanded in the late sixteenth century. The three-year trading cycle
was abandoned in 1536 and thereafter official trading was conducted on an
9-2
260 Miscellaneous incomes
annual basis, starting in the sixth lunar month and going on over a period
of about sixty days.199 Throughout most of the century there appears to
have been no shortage of tea. Rather, there were constant complaints from
officials that too much tea was being stockpiled. The volume of private
trade also increased and contraband dealing was never totally eliminated.
Nonetheless even at this time, according to a memorial of 1571, the trading
offices still procured 6,500 horses a year.200 From an entry in a local gazetteer
it appears that from 1588 the annual number was set at 12,000 horses. From
about 1600, the level began to decline and in the seventeenth century the
trade was drastically reduced to 3,040 horses each year.201 There were
probably a number of reasons for this. Rossabi, who feels that the decline
set in rather earlier, points out that in the early sixteenth century the horse-
trading tribes received harsh treatment from the Mongol chief I-pu-la and
the Moslem leader Mansur, and that their trading activities could have
been permanently affected by this.202 The eunuch tax-collectors dispatched
by Wan-li in the 1590s could also have affected the situation, as their
excesses are known to have impeded inter-provincial commerce.
The decline of the horse trade on the frontier undoubtedly contributed
to the economic depression of Shensi in the early seventeenth century. The
trade in its heyday was extremely complex, involving officials, merchants,
tea growers, army officers and soldiers, and its importance cannot be
measured simply by the figures in the official records. In 1589 the surveil-
lance commissioner of Shensi stressed that it was essential to the province's
local economy.203
From an overall view of trading policy in the sixteenth century it can
be seen that the major problems were the over-supply of tea on the domestic
market, the rather limited outlet on the frontier, the fixed rates for official
trading, and much profiteering by officials at all levels. In the last quarter
of the century, Hukwang replaced Shensi as the major supplier of tea. In
1577 Altan Khan's request that the northern frontier be opened for trading
in tea was refused by the Ming court on the grounds that once the Mongols
had direct access to Chinese tea they would use it as a means of exerting
control over the western tribes. Such a link between northern and western
barbarians would hardly be in Chinese interests. Over-production in
a non-expandable market forced the ministry of revenue in 1585 to consent
to reduce the tax rate at the trading offices to 30 percent.204 As late as 1595
the regional inspector in Shensi still requested that Hukwang tea be ex-
cluded from the frontier trade on the grounds that it was of inferior
quality.205 The real reason for his request, however, was probably the
widespread local resentment against the Hukwang product, the lower price
of which gave it an advantage in an overcrowded market.206 In the end
a compromise solution was adopted; Hukwang tea was not banned but an
additional inspection station was set up to control its quality.
 6, IV Non-cash income 261
In the 1570s and 1580s the government obtained some 10,000 horses each
year through the trading stations. As each horse on average was valued at
10 taels of silver the annual revenue from the trade was worth 10,000 taels.
This was a sizeable item yet its financial significance should not be exag-
gerated. The horses were delivered exclusively to four army posts, namely
Ning-hsia, Ku-yun, Yen-sui, and Kan-su,* though some studs were re-
tained at the pasturage office (Yuan-ma ssu) in Shensi for breeding pur-
poses.207 In the late sixteenth century this office was reported to be
extremely inefficient, and it apparently had not managed to increase the
number of horses in its charge. At most it simply served as an intermediate
collecting agency before the horses were assigned to the army commands.
Its operations had no connection with the court of imperial stud in Peking.
Nor should the tea and horse trade be confused with the 'horse markets'
(ma-shih) on the northern frontier and in Manchuria. The latter were
operated by the frontier governors-general with funds provided by the
court of imperial stud. They purchased goods attractive to the local tribes-
men and traded them for horses. This had nothing to do with taxation,208
and in the late sixteenth century the operation usually cost the government
about 300,000 taels a year. In 1594, in preparation for a campaign against
Hideyoshi in Korea, the central government sent 550,000 taels to the Liao-
tung command alone for purchasing horses.209 In other words this was
simply an item of expenditure.
As the tea grown in Szechuan was not traded on the north-western
frontier, its contribution to governmental finance was of limited value. It
was officially reported in 1542 that taxable tea in Szechuan amounted to
more than 5,000,000 catties, quite apart from that produced in the remain-
ing government plantations, yet in 1578 the provincial officials altogether
collected less than 20,000 taels of silver in tea revenues. Of that amount
14,367 taels were delivered to the Nanking ministry of revenue and
between 1,500 and 2,000 taels to Shensi province. The rest of the funds,
together with 158,859 catties of tea collected in kind, was retained in the
province to subsidise local administration.210 In the middle of the century,
the governor of Szechuan admitted that there was much trade going on in
contraband tea.211
In all other provinces taxation on tea became virtually a dead letter.
Most of them still adhered to their regional tax quotas assessed in paper
currency, and when these were converted into silver at the sixteenth-
century rate, the total collection was reduced to a derisory sum; in Yunnan
17 taels, and in Chekiang about 6 taels.212 One contemporary writer sum-
marized the situation as follows: 'Tea growers in the interior are unaware
* Kan-su here was the name of one of the northern army posts, not to be confused with
the later Kansu province. Similarly, the Shan-si command was not identical with
Shansi province.
262 Miscellaneous incomes
that the commodity is under governmental monopoly. Even some fiscal
officials do not know that such a law exists.'213

(bb) Other items not listed

The total value of goods delivered in kind to the imperial capital, as has
already been mentioned, was between 4 and 5 million taels of silver per
year (3, in). Essentially they represented different forms of tax payments.
For instance items which usually came in large quantities, such as cotton
cloth, silk fabrics and wax, were derived from diverse sources, partly from
commutation of the regular land taxes, partly from li-chia requisitions, and
partly from cash purchase by the government. These supplies should more
appropriately be listed on the warehouse inventory rather than the tax
ledger, for rather than increasing units of taxation, they actually used up
some of them. They should in fact be regarded as expenditures but owing
to the inconsistency of the accounting system were sometimes listed as
revenues.

V SUMMARY OF MISCELLANEOUS INCOMES

Estimated revenues
The estimated revenues from the miscellaneous incomes are presented in
Table 18, with the figures rounded up to the nearest thousand taels of
silver. Items from which the annual revenue is unknown are omitted, but
this is unlikely to affect the overall picture significantly.
In terms of the actual tax burden on the population, it is more likely that
these figures are under-estimates. There was certainly a discrepancy
between reportedfiguresand actual collection, owing to duplication of tax
payments, unauthorized collections, and disbursement of tax funds by the
collecting offices prior to official delivery and auditing. On the other hand,
in terms of actual state income, the figures may well be over-estimates,
especially in the case of revenues from industrial and commercial sources.
It is known that the collection of inland customs duties at times did not
come up to the specified quota, and government mining was actually
suspended for a number of years. Only in the closing years of the sixteenth
century did the ministry of revenue make a serious attempt to enforce the
stamp tax on real estate transfers. The table is by no means statistically
consistent in that it includes both general estimates and documented
figures, funds retained by the provincial officials as well as those delivered
to the imperial government. The grand total, however, represents an
estimate of the maximum amount of revenue that the government expected
to obtain from this wide variety of sources.
 6, V Summary of miscellaneous incomes 263

TABLE 18. Estimated annual proceeds from miscellaneous incomes, ca.

1570-90 (tls. of silver)

Revenues from commerce and industry Administrative incomes

Inland customs duties 340,000 Sale of rank 400,000
Local business tax 150,000 Ecclesiastical license fees 200,000
Maritime tariff 70,000 Payment for * rationed salt' 160,000
Stamp tax on real estate 100,000 Commutation of punish- 300,000
transfers ment and fines
Forest produce levy 75,000 Common post money 50,000
Government mining 150,000 Incense fees 40,000
Fish duty 58,000
Sub-total 943,000 Sub-total 1,150,000
Commutation of services and supplies
Speed-the-delivery money 338,000
Artisan payment 64,000
Reeds tax 25,000
Supplies to the four bureaus 500,000
Horse payment 400,000
Kitchen service fees 360,000
Sub-total 1,687,000 Grand total 3,780,000

NOTE: the items under the land taxes, the service levy, and the salt revenue are not
included.

Critical observations on miscellaneous sources of income

The general feeling that when taxable items proliferate, the level of taxation
must be high does not hold true for the collection of miscellaneous incomes
under the Ming. Though the items were numerous the total revenue that
they produced was comparatively small. The 3.78 million taels of silver,
which was the maximum that could be collected even under the most ideal
conditions, seems unrealistically low when considered in terms of the fiscal
requirements of the empire. As an illustration, between 1570 and 1590
many army recruits were paid 18 taels of silver a year.214 The total revenue
from the miscellaneous incomes, even if collected in full, was only sufficient
to pay the basic wages of 210,000 such recruits, quite apart from their
clothing and equipment and the operating expenses of military commands.
Considering the immensely broad coverage of this category of taxation,
its level might in fact be considered quite inadequate.
The proliferation of taxes is not in itself a sign of 'backwardness'. The
most backward feature of the Ming fiscal administration was its lack of
organization and emphasis. The history of miscellaneous incomes shows
that the taxation system had certain consistent and long-standing weak-
nesses. The initial mistake was to adopt the principle of universality in
taxing industry and commerce. Modest rates of taxation were imposed on
264 Miscellaneous incomes
all manufacturing labor and commercial goods without exception. In im-
posing business taxes the administrators took practically no account of the
capital investment of the merchants, their profit margins, the potential of
their enterprises, the scale of their operations and their trade routes.215
A similar attitude was revealed in the management of government mining,
in which the miners were expected to fulfill individual production quotas
just as the salt producers were. The quota system led to the decline of all
the revenues on which it was imposed. Having failed to obtain sufficient
revenue from those sources the government attempted in the fifteenth
century to generate revenue from its own operations, by selling official
ranks and so on. This was a simple expedient but one that in the long run
had undesirable consequences, for the numerous tax exemptions resulting
from the sale of rank and the collection of license fees later led to the
degeneration of fiscal administration. The withholding of soldiers' pay
reduced the efficiency of the army and was in part responsible for the
expansion of the military budget in the late sixteenth century. The extensive
imposition offinesencouraged corruption and debased the quality of local
government. To this list of blunders was added in the sixteenth century
the ad hoc commutation of services and supplies, which led to further
fragmentation offinancialresources.
Though many writers have criticized the Ming for over-taxation they
have done so largely on moralistic grounds. Their main concern has been
to expose the greed of the tax collectors and the hardship caused to in-
dividuals, rather than to probe the basic causes inherent in the nature of
the tax system. They have created the impression that the major problem
was over-taxation whereas in fact the difficulties were more likely due to
under-taxation. It should be pointed out that the 3.78 million taels of silver
obtained from miscellaneous incomes represented 17 copper cash per capita
if apportioned equally on the entire population, reckoned to be 150 million
or so by the late sixteenth century.216 Naturally it cannot be claimed, in
view of the general practices of Ming officialdom, that abundant tax
revenues would automatically have guaranteed honest administration, but
insufficient revenues did definitely worsen it.
Though in practice over-taxation did occur in certain areas this was
clearly not the case in others. In 1607 the governor of Honan objected to
the fact that while food vendors in the streets were heavily taxed, pawnshops
paid only nominal fees.217 Many cases have already been cited in which tax
potential was either wasted or ignored. In short, it was beyond the capacity
of the administrative system to impose a generally high level of taxation,
and to accuse it of this is to give undue credit to the administrators.
In accordance with acceptedfiscalprocedure, a large amount of the total
revenue was retained by the provincial governments and other service
agencies. This was clearly necessary, for it would have been impractical
 6, V Summary of miscellaneous incomes 265

TABLE 19. Reported annual revenues from miscellaneous sources of income

of the Tai-ts'ang Treasury, ca. 1570-90 (tls. of silver)

Item Maximum Minimum

Inland customs duties 340,000 160,000

Stamp tax 100,000 —
Fish duty 11,000 11,000
Payment for 'rationed salt' 80,000 45,000
Commutation of punishment and fines 170,000 128,000
Incense tax 20,000 20,000
' Speed-the-delivery' money 119,000 —
Tea tax from Szechuan 14,000 —
Total 854,000 364,000

for the ministry of revenue to call in all the revenue and then send some of
it back to its point of origin. But if the quantity of revenue under the direct
control of the imperial government is an indication of its ability to impose
fiscal policies, then its power to do so in the late sixteenth century was
clearly limited. Table 19 shows the revenues that the T'ai-ts'ang Treasury
actually obtained from miscellaneous sources of income.
The accounts of the other ministries were little different from this. In
1569 a supervising secretary auditing the accounts of the Chieh-sheng
Treasury of the ministry of works reported that its annual revenues
fluctuated between 700,000 and 800,000 taels.218 These figures are consist-
ent with those in Table 18, including the commuted payments of artisans,
part of the fish duty, part of the commuted payments of punishments and
fines, some mining funds, and the supplies of raw materials to the four
bureaus. This last item alone amounted to 500,000 taels in all. In 1621 the
total revenues of this treasury were actually only 567,400 taels.219 The only
significant miscellaneous item in the revenues of the ministry of war was
the horse payment - in reality a variant of the land taxes, which fluctuated
between 370,000 and 430,000 taels. All the other items were negligible.
 7
Financial management

Denis Twitchett in his study of financial administration under the T'ang

describes three distinctive stages of development. In the first stage
the administration remained fairly crude and unsophisticated, while the
second was characterized by increasing functional specialization of the
fiscal offices. During the third stage continuing institutional specialization
gave rise to tensions between the new specialized authorities and the
regular organs of the central government.1 Financial administration in the
sixteenth century by contrast presents no such clear-cut divisions. The
Ming administration as a whole was not dynamic enough to produce
similar institutional changes.
For the sake of convenience financial management during these 100 years
can be periodized as follows:
Phase one: 1501-21. At this time there was no effective leadership in
Peking, a continuation of the situation prevailing in the late fifteenth
century.
Phase two: 1521-41. An initial improvement in financial affairs was
followed by further deterioration. For a time the accession of the Chia-
ching Emperor seemed to have ushered in a new era. The new monarch,
who came from a lateral branch of the imperial family, had little connection
with vested interests in the capital. Palace supernumeraries were dismissed
en masse, the eunuchs brought under control and restraints imposed on the
growth of aristocratic estates. Yet no genuine institutional reforms were
undertaken and Chia-ching's record of egocentric pursuits ultimately put
him in the same class as his predecessors.
Phase three: 1541-70. A prolonged series of financial crises occurred,
mitigated only by uncoordinated readjustments by the provincial officials.
The sequence of events was initially triggered off by military failure,
starting with the loss of the Ordos to Jinong in 1541. The government,
forced to take whatever ad hoc measures it could, was under constant and
severe pressure. The millions poured into the military establishment proved
unavailing. In the 1550s Altan Khan's horsemen raided the suburbs of
Peking while the wo-k'ou pirates were marauding along the southeast
[266]
 Financial management 267
coast. The final disaster came in 1557 when three palace halls together
with the ceremonial gates in Peking were burned to the ground. These
buildings were of such enormous symbolic significance that their re-
construction had to be started immediately, despite all other pressing
financial problems.
Phase four: 1570-87. This period, the most successful of the century
from the government's point of view, can be called the era of Chang Chii-
cheng. By 1570 Altan Khan had been pacified and the pirates were no
longer a serious problem. When in 1572 this astute statesman became the
chief decision-maker in Peking some order was at last imposed on financial
affairs, and the government's treasuries began to fill up. Though Chang
died in 1582 the benefits of his administration lasted for a few years
longer.
Phase five: 1587-1600. The achievements of the preceding period
were once more wiped out by fiscal irresponsibility. The enormous
expense of three major military campaigns in the 1590s emptied the
state coffers. The chaotic situation was worsened by the Wan-li Emperor,
who sent eunuchs to the provinces as tax-collectors, bypassing the civil
officials.
This outline shows a succession of administrative cycles rather than
progressive growth and development. Crises were met by emergency
measures and periodic financial pressures were followed by temporary
relief. Though the administration was relaxed, tightened up, and then again
relaxed, there was little attempt at innovation. The history of the period
was made by personalities rather than by institutions.
It cannot be denied, however, that the increased use of silver in taxation
during the sixteenth century was an important development. Though in
theory it should have made the government's financial resources more
mobile and thus facilitated fiscal planning, in practice this did not happen.
It has been demonstrated at considerable length in the foregoing chapters
that the inflexibility of the Ming system prevented it from realizing many
of the potential benefits of the change. The Single Whip Reform was not
pushed to its logical extreme of a single land tax. The service obligations of
the taxpayers were never totally abolished, nor were the tax accounts
integrated. In the following sections it will be shown that this tendency to
adhere to the earlier style of administration was even more pronounced at
the top level of government.
In view of this, there is little point in describing the events of the entire
century with equal thoroughness. Such a chronological survey would only
show the same story repeating itself. On the basis of the periodization
given above the general features of the administration will first be examined,
with the focus largely on the operations of the ministry of revenue. There-
after detailed attention will be given to two specific topics: the effects of
268 Financial management
the enormously increased military expenditures in the years after 1550, and
the financial reforms carried out by Chang Chii-cheng between 1572 and
1582. These fiscal measures in times of war and peace should be sufficient
to illustrate developments in the century as a whole. The improvised nature
of these adjustments, which fell short of rationalizing the system, may well
explain the re-occurrence of the minor administrative cycles briefly
mentioned above.

I THE MINISTRY OF REVENUE IN THE

SIXTEENTH CENTURY
The ministry as an operating agency
At first sight, the operating capacity of the ministry of revenue seems to
have expanded significantly in the sixteenth century. Before 1500 the
T'ai-ts'ang Treasury, which was under its direct supervision, had rarely
handled more than 1 million taels of silver a year. Up to the mid sixteenth
century the level was still said to be close to 2 million taels, but in the 1550s
the record figure of 4 million taels was actually exceeded. Though the rate
of expansion subsequently slowed down, a level of approximately 4 million
taels was maintained.
This seeming expansion is actually illusory. The treasury's revenues
were simply increasing in proportion with the growing tendency to collect
taxes in silver, and there is no evidence that the functions of the ministry
were really being enlarged. The increased volume of revenue in the late
sixteenth century, some 4 million taels a year, comprised only about
12 per cent of the empire's total tax proceeds. The treasury, moreover,
actually retained little of its cash income; most of the revenues were
immediately delivered to the frontier army posts. Smaller sums were used
to pay for the salaries of general government officials, the wages of the
capital garrison, and the maintenance expenses of several palace agencies,
the most important of which were the imperial stable and imperial zoo. In
addition the T'ai-ts'ang Treasury functioned as a receiving agency of the
crown. As a rule the Gold Floral Silver and rents from the palace estates
merely passed through its office.
The limited financial resources under the ministry's control in fact
revealed the weakness of the system, which was not a matter of choice.
Once revenues were delivered in silver the government tried its best to
prevent unnecessary movement of funds. To avoid a situation in which tax
payments were delivered by the southern provinces to Peking and then sent
all the way back to the south again, the ministry of revenue was repeatedly
forced to authorize the southern governors to retain the tax revenue they
themselves collected even though this resulted effectively in a loss of
7,1 The ministry of revenue in the sixteenth century 269
control.2 Without establishing regional branches in the provinces, the
ministry simply put itself at the end of the supply line.
In addition the fiscal practice whereby every item of income was shared
out among many different disbursing agencies also hindered the ministry
from expanding its operating capacity. Because in the early years of the
dynasty it had served more as a general accounting office than as an
operating agency, it had missed the opportunity to set aside a substantial
portion of the tax revenues as its own reserves. By the sixteenth century
there were budgetary deficits at all levels of the government's operations.
When additional funds were required the ministry, instead of appropriating
all the income from any single revenue office, usually drew small quantities
of funds from diverse sources, effecting a general cutback of all existing
allocations and diverting the amounts thus saved to meet the new
fiscal need. Obviously it was not easy to build up huge sums by these
means.
At the same time other ministries retained their fiscal autonomy. This
was confirmed by the institution of the collection of 'horse payments' by
the ministry of war and that of the 'supplies to the four bureaus' by the
ministry of works. In these circumstances the ministry of revenue held that
its own fiscal responsibilities merely consisted in providing government
personnel with food and clothing. Even when it came to supporting the
army posts on the northern frontier the ministry regarded itself principally
in the role of quartermaster general. Financing the procurement of war
horses, firearms and armor was clearly beyond its jurisdiction. On other
matters it also attempted to evade its responsibility as far as possible.
In 1547 when the Hsiian-fu and Ta-tung military commands requested
funds for building the Great Wall, the ministry argued that such con-
struction work was the responsibility of the ministry of war which should
therefore supply the funds. Only on the emperor's personal intervention
was a compromise reached whereby the ministry of war paid two-thirds
and the ministry of revenue one-third of the necessary funds.3 This method
became general practice in the late sixteenth century. When in 1598 a naval
force was mobilized in Chekiang, the ministry of war supplied funds for its
banners, equipment, hiring of ships and bounties for the recruits while the
ministry of revenue provided food and travel allowances. The soldiers' pay
was also provided by the two ministries, the ministry of revenue paying
70 per cent and the ministry of war the rest.4

The revenues of the Vai-ts'ang Treasury

In the primary sources the treasury revenues appear contradictory and
confusing. For instance its annual income in 1549 is recorded in the Shih-lu
as 2,125,355 taels5 and in 1558, it is again described as 'in the vicinity of
270 Financial management

TABLE 20. Reported disbursement of the T'ai-ts'ang Treasury, 1551-7

{tls. of silver)

1551 5,950,000
1552 5,310,000
1553 5,730,000
1554 4,550,000
1555 4,290,000
1556 3,860,000
1557 3,020,000

TABLE 21. Reported revenues of the Tai-ts'ang Treasury, 1567-92 (tls. of

silver)

1567 2,014,200
1577 4,359,400
1583 4,224,700
1585 3,676,000
1587 3,890,000
1589 3,270,000
1591 3,740,500
1592 4,723,000

TABLE 22. Reported deficits of the Tai-ts'ang Treasury, 1583-90 (tls. of

silver)

1583 2,301,000
1584 1,180,000
1585 548,000
1587 2,030,000
1589 190,000
1590 324,500

2 million taels'. Yet the same source indicates that in the preceding seven
years the treasury actually disbursed the sums shown in Table 20.6
These figures of disbursement simply seem to be out of proportion with
the reported revenues. Undoubtedly, during the middle years of the century
revenues had increased, but deficits often occurred too. There is no system-
atic exposition of the data in the sources. Annual receivables over a number
of years have been located in separate memorials to the emperor, and are
presented in Table 21. 7
Quotations for deficits are even rarer; only in the 1580s were some cited
in memorials. They are shown in Table 22.8
Thesefigures,by themselves, do not reveal the whole truth. The fact is that
the Ming officials never developed a consistent accounting system. Often
7, I The ministry of revenue in the sixteenth century 271
they confused receivable revenues with what was actually received. The tax
arrears of a previous year collected in the current year might be included in
either year's account or, sometimes, omitted by both accounts. In the same
way, they made no distinction between a budgetary deficit and an actual
one. The quoted deficit was as a rule calculated from the projected revenues,
not from the received items. An example is the case of 1584. When the
indicated deficit appeared, the court decided to commute into silver pay-
ments 1.5 million piculs of tribute grain, 102,410 rolls of cotton cloth and
45,522 rolls of silk fabrics, which otherwise would have been collected in
kind.9 According to the commutation rates then in force these items should
have been worth a total of some 1.7 million taels. Thus instead of incurring
a deficit of 1.18 million taels as shown in Table 22, in that year the T'ai-
ts'ang Treasury should actually have produced a surplus of over 500,000
taels.
In order to investigate the incoming funds actually handled by the
treasury it is necessary to visualize the standing practice of the time. Firstly,
ever since the early sixteenth century the T'ai-ts'ang Treasury had held
a number of items from tax proceeds which in due course came to constitute
its own permanent quota. These items, in theory neither reducible nor
transferable, formed the central core of its revenues. They included the
items shown in Table 23. 10 The revenues quoted in 1549, 1558, and 1567,
of approximately 2 million taels, referred to these basic items.

TABLE 23. Estimated basic revenues of the T'ai-ts'ang Treasury, ca. 1521-90
{Us. of silver)

Commuted payments from regular land taxes 250,000

Accessory surtaxes of the land taxes (3, iv):
horse fodder 340,000
'farmland silk' 90,000
'poll tax silk' 10,000
hemp cloth 38,000
Cash payment derived from salt monopoly (5, iv) 1,000,000
Minimum level of income from miscellaneous 364,000
sources (6, iv)
Total 2,092,000

Secondly, the treasury's revenues were not significantly affected by tax

increases. Throughout the sixteenth century major land taxes increases
were ordered only twice, and on each occasion for one year only. In 1514
1 million taels was added to the annual collection to finance the construc-
tion of the Ch'ien-ch'ing palace.11 In 1551, owing to the urgent need for
military funds, the southern provinces were required to pay 1,157,340 taels
as an emergency surtax.12 Other increases made in 1592, 1598, and 1599
were retained by the southern provinces (7, iv). When the increases affected
272 Financial management
the service levy, the proceeds were not allocated to the treasury. Any extra
collections made from the salt revenue usually caused deficits in the
following years (5, m). Any raising of the quotas of the various miscel-
laneous sources of income could increase the treasury's funds only
fractionally.
Thirdly, the government had no means for deficit financing. Occasionally
local governments in an emergency were forced to collect taxes in advance,
and the salt administrators could demand advance payment, but these
measures in no way affected the operations of the T'ai-ts'ang Treasury.
Fourthly, any budgetary deficit of the treasury could be covered either
by the delayed incoming items, that is, the revenues which were listed in the
accounts as the quotas of previous years but actually arrived in the current
fiscal year, or by further commutation of taxes delivered in kind in Peking,
chief of which was the tribute grain. Any amount that was not covered by
these means constituted an actual deficit that could only be eliminated by
withdrawing treasury reserves, or sometimes by subsidies from the emperor,
which were derived from his personal income.
Viewed in the light of these procedures the accounts of the treasury,
though complicated, are by no means incomprehensible. They developed
in the following stages:
(a) Up to about 1541 the annual income of the treasury was in fact no
more than 2 million taels, and might even have been slightly less. The
revenues comprised only the basic items.
(b) In addition to these basic items, the treasury's income was later
augmented with additional items. After Yang T'ing-ho eliminated large
numbers of palace supernumeraries the annual quota of 4 million piculs of
tribute grain produced a surplus of 1.5 million piculs (2, i). After 1541 this
granary surplus was made available for conversion into silver payments,
the commutation being ordered fairly regularly although not without
interruptions. In 1541 1.2 million piculs were commuted; this apparently
went on until 1546 when Chia-ching reversed his decision and called a halt.13
Subsequent records, though incomplete, suggest that the commutation was
rarely suspended. The three sets of vital financial statistics appearing in the
Shih-lu in the mid sixteenth century show that in 1542 the commuted
tribute grain was 1,385,884 piculs, in 1552 it was 1,667,163 piculs and in
1562 it was 1,367,389 piculs.14
The annual conversion of around 1.5 million piculs of surplus grain into
silver was expected to produce an additional income for the treasury of
slightly over 1 million taels. It seems that after 1541 the actual income of
the T'ai-ts'ang Treasury was close to 3 million taels if not above that level.
But most official documents counted the commuted income as a non-
recurring item and omitted it from the reported revenues.
(c) The subsidizing of the treasury from the monarch's personal income
7, I The ministry of revenue in the sixteenth century 273
started in 1543 under the pressure of Jinong's occupation of the Ordos.
After a court conference the Chia-ching Emperor consented to give up the
Gold Floral Silver and the rents of palace estates to the ministry of revenue
for the duration of the emergency, which was originally expected to last
for five years.15 The arrangement in fact continued until 1558.16 The landed
properties designated for the support of the imperial stable and imperial
zoo had been taken over by the ministry earlier, an arrangement that had
since become permanent.17 These items altogether added some 1.3 million
taels to the treasury's revenues, pushing its total income over the 4 million
mark. In reality, however, the value of these items was considerably lessened
by the fact that after taking them over, the ministry of revenue was also
required to pay the army officers in Peking and to maintain the stable and
the zoo.
(d) In the 1550s the need for funds became acute. An emergency sur-
charge on the land taxes, mentioned earlier, was ordered in 1551, while
the salt revenue was expected to provide another 300,000 taels annually
(see 'cost-paid salt', 5, m). The sale of ranks, commutation of punishments,
and collection of fines were all stepped up. The effectiveness of these fund-
raising measures was short-lived,18 the reasons being partially explicable
by Table 20 as will be shown below. From 1553 onwards the southern
provinces had to use the funds to finance their own local defense against
the wo-k'ou pirates, and after 1557 the funds were partially retained by the
provincial officials to provide materials for palace construction.
The high point of the financial crisis came in 1558 when Minister of
Revenue Fang Tun (in office, 1552-8) submitted a seven-point fund-raising
program.19 A supervising secretary proposed a five-point program20 and
another five-point program was put forward by the ministry of revenue in
Nanking.21 A group of capital officials jointly recommended twenty-nine
methods of raising funds, all of which were approved by the emperor.22
None of these proposals and programs envisioned any kind of master plan,
but merely attempted to extract small transferable items from existing
revenues, dispose of such government properties as could readily be
converted to cash, and call for extraordinary contributions from the general
population. The Ming-shih labelled all of them 'petty and disgraceful'.23
The nature of the recommendations only reflected the extent of the despair.
Nevertheless, there is reason to believe that in the early years of the
decade, possibly up to 1553, the annual amount of silver received by the
treasury actually exceeded 5 million taels. Without an income of this size
it would not have been able to disburse the sums shown in Table 20. It is
unlikely that the imperial government had any substantial reserves at the
onset of the crisis,24 for in 1554 when a total of 889,000 taels was withdrawn
from its reserve vault, the remaining deposit amounted to less than 1.2
million taels.25 The smaller sums disbursed after 1553 do not indicate that
274 Financial management
the crisis was over but simply that the treasury had no more funds to
distribute.
These observations seem to support the belief that during the 1550s the
annual silver income of the T'ai-ts'ang Treasury was initially pushed up
to over 5 million taels and then declined to a level of less than 4 million
taels.
(e) After 1570 the main purpose of the financial administration was to
return to normalcy. The personal income of the emperor reverted to the
palace and the emergency fund-raising measures were largely terminated.
Even so the average annual revenues remained at around 4 million taels
(Table 21). Again this was not a sign that state income had expanded.
From 1573 a number of tax consignments that had previously been de-
livered by the population in several northern provinces directly to the
frontier posts were re-routed to go through the T'ai-ts'ang Treasury - a
procedural change which gave the ministry of revenue some say in the
annual distribution of these funds.26 The 1580 accounts show that 523,800
taels came from this item.27 An undated summary in the Ming-shih indicates
that some years later the sum had increased to 853,000 taels.28 The com-
mutation of the surplus tribute grain was no longer carried out every year
without fail, but it was never suspended for long. In 1578 the entire quota
of tribute grain from south of the Yangtze was commuted for a year29 while
in 1584 a total of 1.5 million piculs was converted to silver payments.30
Demands for back taxes also boosted the total revenues to some extent.
Thus it is quite plausible that with a basic income of 2 million taels the
treasury could actually register an annual cash income of approximately
4 million taels.
The essence of the story is that the income of the T'ai-ts'ang Treasury,
the principal fiscal operating agency of the empire, was derived from local
manipulations within a very rigid fiscal framework. For a century, whether
in peace or war, the nature of the income had changed little and the total
revenue ceiling remained virtually unadjustable. That the treasury came
to handle larger sums was mainly owing to the switching of delivery pro-
cedures, the conversion of payments in kind to cash, and borrowing from
the crown. The range for maneuver was always limited. Though outwardly
the treasury revenues had quadrupled from 1 million to 4 million taels, the
net increase was still only 3 million taels of silver, a rather small sum in
view of the empire's enormous fiscal needs.
In the foregoing chapters it was estimated that the land taxes, after
absorbing a number of miscellaneous items, had a total value of some
21 million taels (4, in). The collection of the service levy should have pro-
duced around 10 million taels (4, in). The total proceeds from the salt
revenue were officially reckoned to be worth 2 million taels (6, v). A few
more items cannot be accounted for, such as the 'military supplies'
7, I The ministry of revenue in the sixteenth century 275
collected by the southern governors and the internally generated funds of
the northern army posts for which there are no adequate data. The fore-
going is also an optimum estimate, for these figures include many un-
collectible and some duplicated items. These 37 million taels might still
have been insufficient to finance the administration of the empire, not to
mention the lavish ceremonial expenditures that diverted a substantial
amount of revenue to non-productive uses. For the ministry of revenue to
control only 4 million taels of this amount might thus seem dangerously
inadequate.

Disbursement in Peking
The amount of silver disbursed in Peking varied owing to a number of
factors. Part of the salaries of capital officials and the entire pay of the
Peking garrison was supposed to be paid with tribute grain. But in certain
years the grain payments could be temporarily commuted to silver,
causing the cash account to appear larger than usual. This definitely
happened in 1578 and possibly also in 1567.31 After the Gold Floral Silver
reverted to the crown the emperor resumed the obligation of paying the
army officers in Peking.32 While these funds came from his privy purse,
they were not noted in the accounts, which gave the impression that the
payments came from treasury funds. From 1578 the Wan-li Emperor
demanded that the ministry of revenue supplement his personal income
with an additional 200,000 taels per annum known as 'purchasing money'
(mai-pan fei)P This sum was regularly delivered by the T'ai-ts'ang
Treasury but was not so listed in the accounts. In other words, the total
disbursement was inflated by the pay of the army officers and deflated by
the omission of the purchasing money.
Apart from these distortions the accounts were simple and fairly stable.
The Ming-shih contains a summary of the accounts, dated around 1580,34
given in Table 24. Both the total sum and some of the listed items are very
close to the figures quoted in other sources for the period before 1580, and
for the following years, even after 1600.35 In Table 24 an effort has been
made to adhere to the form of the original summary. The inclusion of the
pay of the army officers and the exclusion of the mai-pan fei can be con-
sidered to have cancelled each other out.
These accounts basically constituted a payroll and a remarkably small
one at that. With the exception of a few holders of aristocratic titles such
as dukes and imperial princesses who each received somewhere between
600 and 1,000 taels a year, the so-called pay and salaries were no more
than pocket money for their recipients. On average a civil official received
about 10 taels, an army officer less than 5 taels, and a soldier less than
2 taels a year. The late Ming payments followed the criteria established
in the early fifteenth century. Although they were partially supplemented by
276 Financial management

TABLE 24. Cash disbursement in Peking by the T'ai-ts'ang Treasury,

ca. 1580 (tls. of silver)

Stipends of holders of aristocratic titles 16,000

Salaries of civil officials 44,000
Pay of army officers 268,000
Cash payments to soldiers and artisans 206,000
Pay of metropolitan police 50,000
Subsidies of soldiers called to capital duty 50,000
from the provinces
Winter uniforms for soldiers 84,000
Supplies for imperial stable and imperial zoo 148,000
Horse fodder for the capital garrison 16,000
Cash payments to attendants at service agencies 13,000
Total disbursement 895,000

the distribution of the tribute grain, the grain allowance was still a food
ration and rarely exceeded 1 picul per man per month. The practice where-
by officials received payments in lieu of the personal attendants allowed
for their offices continued (tsao-li, 2, i and 6, m). Nonetheless the cash
payment, 144 taels a year for an official of ministerial rank, was clearly
inadequate in terms of the standard of living in the late Ming. In the
sixteenth century most central government officials relied for their living
expenses on the 'gifts' of provincial officials. This explains why Hai Jui
called the year when the provincial officials finished their three-year terms
of appointment and reported to the capital 'the year that officials in Peking
collected their rents'. 36
The ministry of revenue was not required to provide the administrative
expenses of the various offices. For this the imperial government in Peking
relied on the districts round about, in the same way as a county govern-
ment did. All stationery used in the offices at the capital was provided by
Shun-t'ien prefecture.37 Unpaid attendants were called from Honan,
Shantung and North Chihli. During the metropolitan civil service exami-
nation of 1592 the county magistrates of Wan-p'ing and Ta-hsing not only
supplied all the materials and labor required but also rented additional
utensils and furniture for the occasion at their own expense.38 Some of the
horses used in the capital were actually provided by distant provinces. As
late as 1610 Kiangsi province still provided the college of translators
{Hui-fung kuan) with eight-two horses.39 When the service obligations
were commuted to silver the payments were made directly to the recipient
bureau or agency without going through the treasury.
After deducting the 895,000 taels of silver disbursed in Peking the
T'ai-ts'ang Treasury should have been left with slightly more than 3 million
taels of its annual revenues (assuming a total income of 4 million taels as
 7, II Inter-provincial administration 277
in the late sixteenth century). This sum was appropriated by the northern
army posts and will be discussed under military expenditures in section m
of this chapter.

II INTER-PROVINCIAL AND INTER-MINISTERIAL

ADMINISTRATION
Direction and supervision of provincial management
Because it was primarily occupied with the maintenance of the capital and
the northern army posts, the ministry of revenue exercised only superficial
control over other areas. By the sixteenth century strict observance of all
fiscal procedures was likewise no longer feasible for many other offices. It
is difficult to say whether fiscal operations in the provinces were budgeted
or not. Because of fixed tax quotas, delivery procedures, rates of pay and
salaries and numerous customary precedents, every office was restricted
by some kind of semi-permanent budget. Though some provincial officials
continued to try to follow these guidelines, not all the earlier procedures
still corresponded to conditions in the sixteenth century, as is shown by the
fact that in many districts the scheduled payments already exceeded their
total retained revenues (4, iv).
Part of the problem was created by the imperial government. Tax
remissions, sometimes arbitrarily ordered by the emperors, caused further
difficulties to provincial and local officials. In 1534 the Chia-ching Emperor
ordered that all the land taxes collectible in that year be reduced by half.
When the ministry of revenue argued that the order would only result in an
enormous deficit, the well-meaning sovereign simply directed the ministry
to find the best remedy it could.40 In 1582 the Wan-li Emperor reduced the
stamp tax in Peking by half. Petitions by the local magistrates several years
later requesting a reversion to the earlier rates were not approved.41 There
were also cases where tax payments, having been explicitly remitted, were
actually collected. In a memorial dated 1502, the ministry of revenue itself
admitted that this did occur.42 When in 1523 Sung-chiang prefecture in
South Chihli suffered serious flood damage, its local officials secured the
emperor's permission for a 50 per cent remission of the land taxes in the
disaster-stricken areas. Subsequent instructions from Peking, however,
interpreted the previous rescript to mean that the tax remission applied
only to the prefecture's retained revenue and that the transferred revenue
was still to be delivered in full. Since the transferred revenue in Sung-
chiang amounted to 1.38 million piculs and the retained revenue to only
60,000 piculs, the second order in effect cancelled the tax remission.43 In
the late sixteenth century similar cases were frequent.
As was mentioned earlier, taking small sums from tax revenues which
278 Financial management
had already been appropriated for other uses was a convenient way for the
imperial government to meet urgent fiscal need. Such diversion of funds
was frequent and inevitably caused shortages elsewhere. At times the
imperial government instructed the local officials to make up the lost
revenues at their own discretion. In reality it means that the existing
budget was not binding, and expenditures of lesser priority either cut back
or suspended. For example, in the 1530s the proceeds from the inland
customs houses were already allocated by standing orders for a number of
uses, including subsidies to nearby army units. In 1536 the tonnages
collected at Huai-an and Yang-chou were diverted to Peking for a year to
provide funds to repair the imperial mausoleums.44 This was only one of
the numerous occasions on which existing appropriations were cut back
to make way for new expenditures.
In its directions to the provincial governments the ministry of revenue
naturally stressed the delivery of transferred revenue and paid far less
attention to the revenue retained in the provinces. Personnel evaluation in
the late Ming confirms this emphasis. An official who delivered transferred
revenues promptly and in full gained public recognition but his handling
of retained revenues was an insignificant matter. Minister of Revenue
Liang Ts'ai (in office, 1528-31) tried to reverse this trend but without
success, as it simply reflected the ministry's own priorities. Some twenty
years later Minister P'an Huang (in office, 1549-50) had to admit that his
primary concern was still the transferred revenue; the retained revenue was
mentioned only once in unspecific terms in his memorial.45 This sense of
priorities, coupled with revenue shortages, undoubtedly gave rise to the
situation that a number of appropriations within the permanent budget
were largely fictitious (4, iv).
The auditing of fiscal accounts was never relaxed throughout the entire
dynasty. It was essential to the administration precisely because the budget
was not integrated; the officials collected the taxes and disbursed the funds
themselves. Considering the physical difficulties, the auditing cannot be
dismissed as ineffective. In the mid sixteenth century several high-ranking
and influential officials were repeatedly censured by censorial officials after
their misappropriations and embezzlement of public funds were brought
to light. They included Chang Ching, previously governor-general of
Kwangtung, Hu Tsung-hsien, governor-general of three provinces, and
Chao Wen-hua, minister of works and a special representative of the
crown.46 The periodic check of fiscal records by the regional inspectors was
conducted with such vigor and thoroughness that European observers
made special note of it in their journals.47
Yet, it would be unrealistic to expect this vigorous investigation to
maintain the fiscal integrity of the government personnel. At best, it
checked the more blatant forms of larceny. In the sixteenth century the
 7, II Inter-provincial administration 279
nature of some fiscal operations, such as the management of the salt
monopoly and the collection of palace supplies, made auditing virtually
impossible. Moreover, official auditing covered only the recorded items;
whereas most of the corrupt practices involved unscheduled collections,
extortion from the civilian tax agents and bribery of all kinds. The censorial
officials followed the standards of the times in exposing only those cases
which were considered excessive; no attempt was made to judge every act
according to the letter of the law. For instance, during this century the
handing out and receiving of'gifts' by the officials were no longer regarded
as misdemeanors. Relatively few persons were punished strictly on bribery
charges. Usually when an official was accused of such an offense it was
only because his donor or recipient had already been condemned for more
serious crimes. In other words, bribery charges were used mainly for
exposing political association rather than financial irregularity per se.
It is clear that towards the end of the century official corruption worsened.
This is shown not merely by the larger sums involved in the exposed
cases but by their wider extent and the general decline of moral standards.
Some of the sums mentioned in these cases were impressive, such as the
personal income of 30,000 taels enjoyed by a salt-distribution superinten-
dent in Kwangtung in the early seventeenth century, referred to earlier
(4, iv). At the same time the positions of the warehouse-receiving men in
Peking, nominally unpaid, became so profitable that they were bought and
sold at 4,000 taels each.48 By the end of the dynasty their sale was being
formalized by written deeds. Similar posts, less lucrative but also offered
for sale, were reported to exist in all provinces and in all commissions.49
Their holders, usually lesser functionaries, formed a kind of sub-bureau-
cracy within the regular administration. They became indispensable
because the officials relied on them to provide their own illicit personal
incomes; worse still, by then actual administrative practice deviated so far
from the existing and still unrevised fiscal regulations that the government
simply could not do without their irregular services.50

Execution of large-scale water-control projects

In the sixteenth century the building of the Great Wall was carried out by
the frontier governors-general, its cost being merged with military expendi-
tures. The water-control projects in the Yangtze delta, as well as the sea
walls on the Chekiang coast, were maintained by provincial and local
officials.51 The two major public works that constantly occupied the atten-
tion of the imperial government were the restitution of the Yellow river
and the maintenance of the Grand Canal. Usually the two were combined
into one project and in the sixteenth century the work went on almost
incessantly, rarely being interrupted for more than five years. The
280 Financial management
enormous and burdensome sums involved, coupled with the unpredictable
disposition of the Yellow river, definitely contributed to the general
financial uncertainty.
Major construction works were carried out in 1528, 1565, 1578, 1587,
and 1595.52 After the Grand Canal had been repeatedly silted-up and
inundated by the Yellow river it was decided in 1593 to construct a new
channel in its middle section so as to avoid the Yellow river. This project
was suspended several times owing to shortage of funds. Only in 1609,
sixteen years after the initiation of the project, was the 110-mile channel
opened to traffic.63 The management of all these construction projects had
several features in common. No detailed budget was ever compiled but
a rough estimate of the cost was made beforehand. This included only the
basic sum provided by the imperial government which was to be agumented
by local contributions and requisitions. Normally such projects came under
the jurisdiction of the ministry of works but the ministry of revenue was
heavily involved in the raising of funds. The final allocation of the funds
was left to the project director, who as a rule was given a dual appoint-
ment both as censor-in-chief and imperial commissioner responsible
directly to the emperor. Sometimes he was also made supernumerary
minister of works but this title was honorific only; it did not compromise
his autonomous position as a field director, nor did it give him more
influence with the ministry.
The financing of the 1578 project under the direction of P'an Chi-hsiin
is more satisfactorily documented than other cases and will be summarized
here for the purposes of illustration. It started in the summer of 1578 after
P'an's predecessor had died in office, and was carried on until early 1580.
Initially 50,000 laborers were mobilized though the project as a whole
seems to have involved at least 100,000 men for a little over a year. Al-
together 139 breaks in the dikes were repaired and over thirty miles of new
dikes were constructed, complete with tunnels, glacises, and stone embank-
ments. In addition 830,000 new willow trees were planted. Had all the
materials and labor been paid for in full, a conservative estimate would
have put the total cost at more than 2.5 million taels. In his final report
P'an indicated, however, that only 560,637 taels of government funds were
actually disbursed.54
In preparation for the project the tribute grain from all the southern
provinces was commuted to silver for a year. The basic payment covering
the value of the grain was delivered to Peking to defray routine expendi-
tures in the capital; only the surcharges and the'speed-the-delivery money'
were handed over to the project director. Additional funds in small amounts
came from the sale of ranks, contributions by salt merchants, and horse
payments deposited at Nanking. 55 P'an used the prefectural treasury of
Huai-an as his bursar's office.56
 7, II Inter-provincial administration 281
While the standard pay for each laborer was set at 0.03 taels of silver
per day, the records indicate that their wages were actually provided by the
local districts with funds derived from the service levy collections. Those
who were unable to pay the latter answered the service call instead.57 No
detailed accounts survive but it can safely be concluded that the funds
under centralized control in reality constituted a grant-in-aid and that
most of the financial burden was carried by the local districts. The con-
struction work therefore involved all fiscal agencies from the imperial
government down to the village communities, the only integrating factor
being thefinancialauthority of the project director. In one of his memorials
to the emperor, P'an suggested that all prefects and county magistrates be
ordered to stay in their offices for the duration of the project so that they
could be held personally responsible for the execution of procurement
orders; their deputies were to be assigned as field supervisors, and any
negligence promptly reported by the surveillance officials.58 P'an himself
held the concurrent assignment of censor-in-chief of the right.
Another feature of thefinanceprogram was that the small contribution
from the central government was again derived from many different
sources. Early in 1580, when the project was completed, P'an suggested
that maintenance work and secondary construction be continued for
several years, and that the necessary funds be provided by the salt distri-
bution agencies, the inland customs houses, and the local business tax
stations in the Hsii-chou-Huai-an-Yang-chou region. Yet his estimate of
the annual requirement was less than 30,000 taels.59
Though we know less about the construction project in other years, it is
safe to assume that the methods of raising funds and conscripting service
for them were similar. The 1528 project involved 300,000 laborers and
went on for two years, the total cost to the state being estimated as only
200,000 taels.60 The 1565 project again enlisted 300,000 workers and cost
700,000 taels.61 The actual costs must have exceeded the quoted sums
several times over. In 1584 for one minor construction project the com-
missioner of the Grand Canal proposed a budget of 216,000 taels. A
secretary from the ministry of rites challenged this on the grounds that it
would not pay even a tenth of the actual cost. This official reported from
his own personal experience that in the affected districts a single household
could be required to pay the wages offiveworkers. The recruited laborers
were subjected to extortion and harsh treatment by the foremen, and when
they absconded their home districts had to replace them with new recruits.62
Understandably, such construction projects imposed extra strains on
central as well as local government. The enormous cost was only one
aspect of the problem; the other was that both the central and local
authorities lacked the organizational capacity to execute these projects
satisfactorily.
282 Financial management

Palace construction
Palace construction went on incessantly from the time that Yung-lo moved
the capital to Peking up to the end of the T'ien-ch'i period. In the sixteenth
century the palace residences and ceremonial halls, all built of wood, were
repeatedly destroyed by fire. This happened in 1514,1525,1541,1557,1584,
1596 and 1597. The mere replacement of these buildings necessitated con-
struction work in the capital almost every decade. The most disastrous
blaze was that of 1557 which lasted over twenty hours and destroyed two
ceremonial gates and three major buildings. Their reconstruction, com-
pleted in the autumn of 1562, took five years in all.63 Even in other cases
the work rarely took less than two years. Clearly these building expenses
were too substantial to be dismissed merely as incidental items.
Building came under the jurisdiction of the ministry of works but other
ministries were also involved. For instance in 1557 and 1596 the ministries
of revenue and war were both ordered to deliver 300,000 taels each from
their treasury funds to assist in starting the work.64 In addition the ministry
of works was authorized to obtain funds from the commutation of punish-
ment, sale of rank and minting, and to retain some of the revenue from
local business taxes, stamp taxes, and the land taxes. The financial pinch
therefore affected all other governmental agencies and some of the provinces.
The complexities in fiscal transactions and supply procedures make it
difficult to assess the actual cost of each project. Some of the frequently
mentioned sums are evidently exaggerated. The T'ai-su Hall, constructed
in 1515, is said to have cost 20 million taels, though it is doubtful whether
such an enormous sum could have been raised in the early sixteenth
century. Lien-sheng Yang has shown however that the actual cost was
200,000 taels and that a careless copyist of the Shih-lu must have replaced
the character 'ten' with 'thousand', the two characters differing merely by
one stroke.65
Nonetheless the official figures do not represent the true cost either. The
200,000 taels for the T'ai-su Hall in 1515, the 150,000 taels for the Tz'u-
ning Palace in 1585,66 and the 720,000 taels for the Ch'ien-ch'ing and
K'un-ning Palaces in 159867 were calculated from the amount of cash
disbursed in Peking, largely omitting building materials and labor supplied
by the provinces.
The most costly item on the list of supplies was timber, which came from
Shansi, Szechuan, Kweichow, and Hukwang. In 1558, Kweichow alone
claimed to have provided timber worth a total of 1.38 million taels.68 This
sum may have been somewhat exaggerated since it was estimated by
provincial officials who had good reason for emphasizing the fiscal
difficulties caused by these procurement orders. In 1584 Szechuan was
authorized to retain 700,000 taels in taxes to pay for the cost of timber and
 7, II Inter-provincial administration 283
its transportation,69 a sum which was probably close to the actual expendi-
ture. Hui-chou prefecture in South Chihli, not a timber-producing district,
in 1557 was ordered by Peking to provide 86,766 logs at a cost of 129,314
taels, plus transportation costs of 41,640 taels. The estimate was made by
the ministry of works and was deductible from the district's taxes. After
the supply was delivered, however, the prefecture was instructed by the
ministry to pay its back taxes totalling 138,000 taels, plus a new assessment
of 30,000 taels to finance the construction work. The purchase thus appears
to have turned into an outright requisition.70
A more accurate estimate of the cost of timber can be arrived at with
the aid of other contemporary accounts. One sixteenth-century account
indicates that timber from Szechuan was fastened into rafts of 604 logs
each. The cost of each raft was only 148 taels but it required the labor of
forty persons over a three-year period to bring it to Peking. The trans-
portation cost was another 2,160 taels, making the average value of each
log about 4 taels by the time it reached the capital.71 For the 1596 project
160,000 logs were used, each of which was estimated to be worth slightly
less than 2 taels in the south.72 It can therefore be assumed that while
a minor project might require some 500,000 taels' worth of timber, the
expense for a major one could easily exceed 3 million taels. Sixteenth-
century travellers on the Grand Canal often mentioned that the waterway
was congested with imperial timber shipments.73 Matteo Ricci noted in
1598 that these rafts were hauled by thousands of laborers who trudged
for five to six miles a day.74 The transport costs must therefore have been
enormous, although the government seldom paid the conscripted labor
anything more than a food ration.
The glazed tiles and bricks used in the construction work were baked in
Soochow, South Chihli. No permanent factory was established but the
labor and manufacturing facilities were provided by six prefectures on the
lower Yangtze as required.75 For the 1596-8 project 1,700,000 tiles and
bricks were produced. The ministry of works subsidized this with 20,000
taels of silver, a mere fraction of the manufacturing cost.76 Transportation
costs were economized by ordering each boat hauling tribute grain to
Peking to carry an additional load of bricks and tiles, though it was again
observed by Ricci that special barges were provided to transport bricks on
the Grand Canal.77
The stone used on the frontage of ceremonial buildings was cut into
monolithic slabs each 30 feet long, 10 feet wide and 5 feet thick. In 1596
special sixteen-wheeled carts pulled by 1,800 donkeys were built to transport
them. Though the details are unclear, the project director's report indicates
that the stone took twenty-two days to arrive at the building site at a total
cost of less than 7,000 taels.78 Undoubtedly a major portion of the cost was
absorbed by the general population.
284 Financial management
The wages paid to labor in Peking were an insignificant item. The con-
struction workers were drawn principally from the capital garrison and the
artisans on the regular payroll. It was also necessary to purchase additional
building materials in Peking, quite apart from what was made available
by the palace warehouses. These were supplied by designated merchants at
officially prescribed prices. An imperial order of 1588 acknowledged that
the prices were often set below the standard market rate and that the
payments were in arrears. A supervising secretary estimated that the
government was by then indebted to the merchants to the tune of over
a half million taels. Even though the emperor instructed the ministry of
revenue to use the funds of the T'ai-ts'ang Treasury to clear its debts, it is
uncertain whether this order was ever carried out in full.79
Owing to the numerous uncertainties affecting the data it is unfortunately
impossible to make any accurate estimate of total building costs. Failing
that, it is still worth noting that the overall estimate in the Ming-shih, of
between 2 and 3 million taels of silver each year at the high point of con-
struction in the late 1550s and early 1560s,80 is consistent with other veri-
fiable administrative factors. The estimate undoubtedly covers only the
accountable funds disbursed by the central and provincial governments,
and cannot include extraordinary requisitions of materials and labor.
It would be erroneous to assume that these extra requisitions represented
a net gain to the government. In the late Ming the operating capacity of the
government was so limited that taxation had virtually reached saturation
point and any additional imposition in one area generally led to short
payments in others. In 1592 a vice-minister of revenue in charge of the
T'ai-ts'ang Treasury reported that the total arrears of the various pro-
vinces over the previous six years amounted to 7,641,100 taels.81 This was
the period when the Tz'u-ning Palace was being rebuilt, a major water-
control project was under way, and when the Wan-li Emperor had in
addition ordered the construction of his own mausoleum (7, iv).

Ill APPROPRIATION OF MILITARY SUPPLIES

Rising military expenditures
The traditional historians attributed the increase in the defense budget in
the late Ming entirely to the decline of military farming. What is now
known of developments in the sixteenth century shows clearly that this
view is no longer tenable and that many other factors have to be taken into
accounts.
Up to the late fifteenth century the soldiers' pay consisted only of a food
ration and some winter clothing. Army service was the duty of the heredi-
tary military families whose soldiers were expected to provide even their
 7, III Appropriation of military supplies 285
own basic equipment. This system became increasingly impractical in the
sixteenth century, and by around 1500 it had become a standard practice
to fill vacancies in the northern army commands with hired recruits.82
Initially each recruit was offered a bounty of 5 taels of silver and was
provided with a horse and clothing. These mercenaries of course had to be
paid regularly. In the mid sixteenth century 6 taels per man per year was
still adequate, but it went up steadily thereafter as a result of the wider use
of silver and the intensified recruiting program in the southern provinces
during the wo-k'ou campaign. When the southern troops were transferred
to the north in the 1570s the increased rates of pay were generally adopted
by all frontier army posts. By the end of the century many recruits received
18 taels per year, which became the standard rate in the early seventeenth
century.83
The number of army posts on the northern frontier increased. Before the
sixteenth century there were only seven such commands but in 1507 Otfsai's
invasion of Ning-hsia led to the establishment of the Ku-yiian command.84
In 1541 the Shan-si command was set up to check the threat from Jinong.85
In the late sixteenth century five more minor posts were elevated to the
same status as the nine major posts with regard to fiscal matters.86 All this
added to the administrative costs (Fig. 5, p. 289).
The construction of the Great Wall entailed vast expenditure. It was
first started systematically by Yii Tzu-chiin, governor-general of Yen-sui
in 1472.87 At the outset it was intended only to build up embankments and
dig deep trenches, connecting the existing strongholds with pounded
earthwork, but the construction gradually became more elaborate. Soon
it involved the building of massive brick structures with crenelated walls
and gun-emplacements. The process of fortification went on for over
a century throughout the Ch'eng-hua, Hung-Chih, Cheng-te, and Chia-
ching eras into the early years of Wan-li, that is, into the 1580s. By the mid
sixteenth century the construction costs were already extremely high. In
1546 the construction of the wall in Hsiian-fu and Ta-tung districts, where
labor was conscripted and unpaid, cost the government about 6,000 taels
a mile.88 For the building of the Chi-cheng section in 1558, the hired labor
cost 6,357 taels a mile. The accusation was made that when the labor was
conscripted the local population actually paid seven times that amount,
and that mismanagement could inflate labor costs to 44,500 taels per mile.89
Another reason for the expansion of the military budget was the use of
firearms. Although the Ming armies are known to have used firearms
in the early fifteenth century they seem to have come into general use only
at a later date, notably in the later sixteenth century when Portuguese-made
cannons were extensively copied. A standing order of 1498 attempted to
restrict the manufacture of firearms to the ministry of works; frontier posts
were not allowed to make them. In the Chia-ching period this restriction
286 Financial management
was gradually removed.90 As late as the 1560s the standard cannonballs
used by the Peking garrison were still covered with clay; from 1564 they
were made of lead and after 1568 they were cast in iron.91 In 1586 the
ministry of war sent a mission of inspection to the four frontier posts in
Shensi. Its subsequent report listed all their stocks of supplies and equip-
ment, some of which were connected with the use of firearms. Unfortunately
the inventory listed iron and lead together with rocks, making it impossible
to calculate their cost, but one post alone had a stockpile of over 2,000 tons
of them. The so-called firearms obviously included chemical-propelled
arrows and pellets of which each post possessed 2 million or more.
Evidently a more modern type of warfare had begun to change the cost
picture.92 Wheeled vehicles with firearms mounted on them were introduced
in the fifteenth century, but it was Yii Ta-yu (1503-79) and Ch'i Chi-kuang
(1528-87) who promoted their use as a defensive weapon on a large scale
in the late sixteenth century.93 The ministry of works recorded in 1609 that
each of these vehicles cost 30 taels of silver to produce.94 This too was a new
item of expenditure.
Suffice it to say that army logistics in the sixteenth century were con-
siderably different from what they had been 100 years earlier.

Military farming in the sixteenth century

There is much information on this subject but few verifiable facts. The
acreage farmed by the army and its annual returns are both listed in the
Shih-lu. Wang Yii-ch'iian has tabulated the figures,95 which can be
summarized as follows:
Between 1487 and 1504 the annual returns were always approximately 2.7 million
piculs.
From 1505 to 1518 production was stationary at exactly 1,040,158 piculs per year.
After 1519 the figures were recorded only sporadically.
From 1522 to 1571, a period of fifty years, annual production remained at
approximately 3.7 million piculs, with the exception of 1567 when afigureof only
1.8 million piculs was reported.
The reports ended in 1571. Though Wang considers that the figures for the
later period are probably inflated,96 this is in fact an understatement. In
1549 Minister of Revenue P'an Huang revealed that for over ten years not
a single frontier post had reported its farm production.97 When in 1570
P'ang Shang-p'eng was commissioned to investigate the farming program
he discovered that in many places there were production quotas on the
books but no cultivators in the fields.98 Though administrative discipline
was later considerably tightened-up by Chang Chii-cheng, even he admitted
in a private letter, possibly dated 1575, that he still needed more time to
put military farming in order.99
 7, III Appropriation of military supplies 287
Unlike the land tax data which were periodically published by the local
communities, the statistics of military farming were based entirely on
official reporting. Other surveys were compiled on the basis of observations
by senior officials but this information is generally superficial and provides
no basis for comparison.
Nonetheless it is clear that military farming was never suspended com-
pletely and the principle that each command should produce at least part
of its own food-supply was observed throughout the dynasty. In the frontier
region both uncultivated land and military manpower were always avail-
able and energetic governors-general and their civilian assistants were
constantly organizing farming programs of some kind. Yet these programs
did not follow any overall scheme laid down by the central government,
nor did individual commands attempt to formulate their own permanent
guidelines. Military farming was thus extremely diversified.
In the sixteenth century the Hsiian-fu command leased most of its land
to civilian farmers and collected rents from them, these being counted as
the proceeds from military farming.100 The Liao-tung command, whose
soldiers were mostly on active duty, enlisted other male members of the
hereditary military households to carry out its farming program. 101 Though
the prototype of military farming, in which individual households were
allocated fixed acreages of land and paid part of the produce on a per
capita basis, still existed in some areas in the sixteenth century, collection
on the basis of acreage became popular and eventually came to differ little
from regular rents or land taxes. Other commands introduced a type of
collective farming program (ying-fieri). The soldiers on active service
cultivated the land in their spare time, so that the guard units obtained the
entire crop.102 Some commands on occasion preferred cooperative farming
(t'uan-chung) whereby the soldiers, after handing over a fixed production
quota to their commanding unit, shared the remainder of the crop.103
It is difficult to ascertain when the governors-general first started to grant
permanent ownership of the land to the soldiers who cultivated it, but from
the mid sixteenth century the practice became widespread. P'ang Shang-
p'eng in his reports to the emperor repeatedly argued that such grants
should be extended in order to encourage the opening of new land.104
In 1585 the governor-general of Chi-chou converted an entire transpor-
tation battalion into a farming unit, giving each soldier twenty mou of land
and a certificate of permanent ownership.105 An army unit was thus in
effect demobilized in order to improve the grain supply.
In short, military farming, by no means systematic under the early Ming
(2, in), was much less so in the sixteenth century. The only common feature
shared by all the various types of program was that they enabled each
command to produce a part of its food supplies. Some army posts, while
remaining fiscal offices under the supervision of the ministry of revenue, in
288 Financial management
effect performed the combined functions of local government, landlord and
tax collector. The decentralization was so complete and diversities so
numerous that the aggregate figures given in the national accounts can
have been no more than rough estimates. The primary data submitted by
the different units, even if accurate, were not statistically uniform.
In the 1570s the imperial government subjected the accounts of the
fourteen northern army posts to thorough auditing. Investigating teams
were also frequently sent out. Chang Chii-cheng personally demanded that
governors-general increase their farm production.106 The annual returns
dating from between 1578 and 1579 that eventually appeared in the Ta-
Ming Hui-tien, are quite likely to have been over-estimates, but it is im-
possible to obtain more reliable data. The accounts show that the fourteen
units annually produced a total of about 1.4 million piculs of grain of
various kinds, plus 180,000 taels of silver.107 This level presumably could
only be attained in the rare intervals of peace when the military commanders
were able to assign more men to farm duty.
The situation in the interior provinces was simpler, in that it showed
a clear and consistent decline. By the late sixteenth century the scattered
plots of land that had originally been assigned to military farming, even
if they still survived, produced only a small amount of rent. The local
officials generally collected this and merged it with the regular land taxes
in order to support the local army units.108 Because the accounts were
scattered and fiscal responsibility decentralized, the civil officials handled
both the collection and disbursement of the proceeds. In some cases the
sites of army barracks were converted to agricultural land and rented out; 109
in others soldiers became landlords,110 or sold or mortgaged their allotted
land, or else made private arrangements for their dues to be paid by
civilian cultivators.111 All these made the so-called national statistics
meaningless. According to the Hui-tien the total income from military
farming in the interior provinces in 1579 was still some 3 million piculs112
but it is likely that the data were copied from earlier records.

Support of the northern army posts

Because of the different supply systems in use, the Ming army in the
fifteenth and sixteenth centuries appeared to be made up of three separate
components. The capital garrisons in Nanking and Peking were supplied
directly by the central government. (The Nanking garrison received pay-
ments from the Nanking ministry of revenue under similar conditions as
prevailed in Peking, see 7, i.) The guard units in the interior provinces were
supplied by their local districts. The northern frontier posts were partially
self-supporting but also received supplies from four northern provinces
(Shantung, Shansi, Honan, and Shensi) and from Peking.
 Manchurian frontier
I o o K §

Great Wall of China

Headquarters, frontier defense areas MONGOLIA -*"jlr*Shen-yang
HSUAN-FU

r Headquarters with fiscal accounts equalling

those of the defense areas TA-TUNG
COMMAND
COMMAND
(HSUAN-FU^Oj^p'ing
Mi-vim COMMAND
Other important army posts (TA-TUNG)! ^ (CHI-CHOU)

han-hai-kuan
5
Yung-p'ing «-^~
KAN-SU
COMMAND
(KAN-CHOUWEI) MNG-HSIA
COMMAND
(NING-HSIAWEI)

SHAN-SI
Hua-ma-ch'ih D COMMAND
(T'AI-YUAN)

KU-YUAN
I COMMAND
i (KU-YUAN)

Fib. 5. Northern frontier army posts in the late sixteenth century.

290 Financial management
In 1569 Vice-Minister of War T'an Lun stated that while the army's total
authorized strength was 3,138,300 men, in fact only about 845,000 men
could be accounted for.113 The latter estimate seems quite reasonable. It
can be further estimated that 500,000 men were serving on the northern
frontier, using at least 100,000 horses.114
In 1576 the ministry of revenue compiled an account of the annual
expenditures of the fourteen northern army posts, that included four key
items: silver, food, animal-feed and hay. The account covers twenty-one
pages of the Shih-lu115 but though it lists the quantities of each item for each
army post, it gives no assessment of their total monetary value nor any
aggregate figures. In Table 25 the figures have been added together and the
commodities converted to silver.
The revenues of the same army posts in 1578 likewise take up twenty-
eight pages of the Hui-tien11* and are condensed in Table 26.
The two accounts are very close to each other even though the primary
data were calculated by different accounting methods and submitted by
fourteen different agencies. The similarity may be partially due to coin-
cidence; since the conversion rates are only approximate, the total amounts
in both tables are probably fairly accurate only. The actual prices were
subject to regional variations and may have differed considerably from
these estimates. But it is safe to assume that the expenditures of the fourteen
posts in the 1570s were fixed at a level that on the whole corresponded to
the level of income.
Both accounts seem to have included quartermasters' items only, omit-
ting the articles delivered by the ministries of works and war. The total
annual expenditure of the order of 8 million taels nevertheless appears to
have been rather a tight budget. In 1594 when an army of 20,000 was
stationed in Liao-tung during a relatively quiet period of the Korean
campaign a soldier's basic maintenance and pay cost 2 taels of silver per
month.117 The whole army, estimated at 500,000 soldiers, would thus cost
12 million taels a year to maintain. The frontier posts in the 1570s managed
to stay within a smaller budget because the troops were serving as domestic
garrisons in time of peace and were not all paid at standard rates; in
addition the military authorities could requisition services and supplies
from the local population. These conditions, which permitted a reduced
level of subsidies to the army commands, were far from satisfactory. At
the same time the delivery of annuities at the rate of 3 million taels of silver
had already used up the disposable funds of the T'ai-ts'ang Treasury.
In the following decade the frontier governors-general repeatedly
appealed to Peking to increase their annuities, as their supplies including the
grain they produced themselves, the subsidies from the northern provinces,
and those derived from the salt revenue were rarely delivered in full. The
ministry of revenue, having no more funds at its disposal, simply turned
 7, III Appropriation of military supplies 291

TABLE 25. Major items of supply disbursed by the fourteen frontier army
posts in 1575

Disbursed items and Monetary value

quantities (reported) (estimated) (tls.)

Grain for human consumption 2,008,918 piculs @ 0.80 tls. = 1,600,000

Grain for animal-feed 1,125,080 piculs @ 0.35 tls. = 390,000
Hay 14,314,822 bundles @ 0.03 tls. = 420,000
Silver 5,908,562 tls. approximately 5,910,000
Total 8,320,000

TABLE 26. Revenues of the fourteen frontier army posts, 1578

Monetary value (estimated)

Item, quantity and source (reported) (tls.)

Annuities, from the T'ai-ts'ang Treasury, in silver 3,180,000

Subsidies, delivered by the northern provinces, 2,730,000
in silver
Proceeds from military farming, in silver 180,000
Salt revenue, delivered by merchants and 640,000
salt-distribution offices, in silver
Grain, dispatched by Peking, 350,000 piculs 0.80 tls. = 280,000
Grain, delivered by the northern provinces, 0.80 tls. = 220,000
280,000 piculs
Proceeds from military farming, including grain for 0.50 tls. = 730,000
both human and animal consumption,
1,450,000 piculs
Hay, delivered by the northern provinces and from 0.03 tls. = 210,000
military farming, 6,830,000 bundles
Total 8,170,000

a deaf ear to these appeals. Chang Chu-cheng even admitted in a private

letter that he was under great pressure to reduce military expenditures.118
Up to 1591, the central government's allocation of some 3 million taels
remained the permanent budget for the frontier commands. The Wan-li
Ku'ai-chi-lu, published in 1582, lists the total annuities delivered to the
fourteen commands by the ministry of revenue as 3,105,000 taels119 and in
1584 the ministry asserted that an extra of 300,000 taels had accumulated
over the previous eight years. But in 1587 it reported that the total amount
was again down to 3,159,400 taels and in 1591 that the total was 3,435,000
taels.120
Supplies of military establishments in the interior
The supply procedure of the army units in the southern provinces was
established during the wo-k'ou campaign. The fund-raising program
292 Financial management
carried out in these provinces in the 1550s showed the following features.
Firstly, all the funds were raised locally, not handled by the ministry of
revenue. The court either permitted the governors and governors-general
to make arrangements at their own discretion or else authorized imposi-
tions at their own requested rates. Secondly, all the extra collections were
to be kept separate from the regular revenues and audited separately.
Thirdly, the sources of income were very diverse. As they were managed
by provincial officials and military officers the total amount they yielded
was never disclosed. Fourthly, many of the new revenues and surcharges
created during the campaign, including a number of nuisance taxes, were
never subsequently abolished.
The military situation in the south was quite different from that in the
north. In the 1550s no adequate army organization existed and the entire
command had to be built up from scratch. Even the governors-general,
military circuit intendants and commanders-in-chief were commissioned
at short notice. Most of the fighting men were recruited in the field. When
Chang Ching was governor-general of the region (1554-5), the recruits
included mountain aborigines from Kwangsi, salt-smugglers from South
Chihli, and Buddhist monks from Shantung. Later on the recruits were
distinguished by the names of their native districts, being known as P'ei
troopers, Chang troopers, I-wu troopers and so on.121 Wei-so soldiers and
militiamen, on the other hand, played only a minor role. The prefectural
gazetteer of Shao-hsing, Chekiang, summarized the situation as follows:
'The wei-so regulars are called soldiers and the recruits are called troopers.
The troopers fight; the soldiers merely sit idly.'122
In the course of the campaign Chekiang enlisted 100,000 such troopers.
Even ships had to be hired. The recruiting was carried out by officers of all
ranks. Detachment commanders and their superiors were encouraged to
recruit 'standard troopers', and the generals to gather companies of
'housemen' to form a more highly-paid elite corps. This unsystematic
approach was dictated by circumstances.
In the early phase of the campaign the necessary funds were derived
principally from fi-pien, & term for which there is no adequate English
equivalent; fi meaning to lift up, andpien to organize (3, iv). The original
concept of fi-pien was similar to that of the federation of national guards
in the United States, except that in Ming China the term was used almost
exclusively in its fiscal sense, and rarely applied to personnel. In 1554 the
court ordered all the counties and subprefectures of South Chihli to defer
the service of 40 per cent of their militiamen, and that each of these men
should contribute 7.2 taels of silver to Chang Ching's war chest. The
following year each county of South Chihli and Chekiang, according to
its size, was ordered to provide 200 or 300 militiamen on the governor-
general's command. Later on the service obligation was largely met by an
 7, III Appropriation of military supplies 293
annual payment of 12 taels of silver per man. In addition, 1 tael of silver
was levied on every corvee laborer in the two provinces.123
As the campaign dragged on, the fi-pien was extended to the li-chia and
chiin-yao. Since at this time most counties still required the village com-
munities to supply raw materials and labor service in five-year cycles, the
fi-pien called those households scheduled to serve in the following year to
perform their duty in the current year. The following year's services were in
turn performed by those who had been scheduled to serve in the third year.
As the military authorities were naturally not interested in the collection of
raw materials or labor service unrelated to the war effort, all the irregular
collections were made in silver. Though fi-pien had originally been devised
as a stop-gap measure, in the course of the protracted campaign these
extra-schedule collections gradually became permanent. They were de-
signated as 'military supplies', being at first added to the service levy
accounts and later indirectly charged to the land taxes (3, iv). At the high
point of the campaign the surcharges on the land taxes in both South Chihli
and Chekiang were approximately 500,000 taels a year.124 This was the
situation that forced the local officials to undertake the Single Whip Reform.
The number of miscellaneous taxes in the southern provinces increased
enormously on account of the campaign. Monastic properties in Fukien
that had previously been tax-exempt were partially confiscated by the state
or made subject to taxation. Hilly land in Ku'ai-chi county, Chekiang,
which prior to the campaign had been only lightly taxed, was made to pay
a heavier assessment. An extra amount was likewise added to the 'roof
tax' in the city of Hangchou, the proceeds of which were allocated as
military supplies. Kwangtung province imposed a toll at its principal
bridges and a tax on the slaughtering of cows was collected in Shun-te
county. A salt transit tax was imposed along the southern borders of
Kiangsi. Fishermen in the coastal provinces were required to pay a new
tax and had to produce receipts for this before they were allowed to pur-
chase salt. As mentioned earlier, the newly-created maritime tariff in
Fukien and the old one in Kwangtung, together with the taxes on iron
mining in these two provinces, were also appropriated as military supplies
(6, i). Some existing revenues, such as the stamp tax on real estate transfers
and miscellaneous excise duties, were retained by the provincial officials to
help pay defense costs, a practice which spread to the inland provinces.
Yunnan was authorized to retain the proceeds from taxation on mining
while Szechuan derived its military supplies from the excise on tea and the
salt revenue.125
After the wo-k'ou pirates were suppressed, the militiamen in the southern
provinces were partially demobilized. In 1595 there were still 199,650 of
them under arms however,126 and in the Lung-ch'ing and Wan-li periods
there were frequent demands that they be discharged. Efforts were made
294 Financial management
likewise by the civil and military authorities to replace the recruits with
wei-so regulars, but met only limited success. During the Korean campaign
in the 1590s recruiting was once again stepped up and towards the end of
the dynasty recruits formed the backbone of the army.127
Remarkably enough this did not result in the complete abolition of the
wei-so system. Chia-hsing prefecture in Chekiang throughout the second
half of the sixteenth century maintained a mixed force of five battalions,
one of which was filled with recruits, and two each with militiamen and
wei-so regulars. The prefecture also had a naval force of 1,500 men, among
whom the recruits were classified as 'experienced helmsmen', and the
wei-so soldiers as 'assistant sailors'. As late as 1597 all the warships were
hired. These military forces were under the direction of the military circuit
intendant and their supplies were delivered by the Chia-hsing prefectural
treasury.128 It is impossible to define the status of these mixed units in the
imperial army.
It is very difficult to form an accurate estimate of troop strength and
defence costs in the interior provinces in the late sixteenth century on the
basis of the extant sources. While relevant data do exist for certain districts,
the extent of regional diversity in military matters renders them of little use
in studies at the national level. Nevertheless, in 1570 three prefectures in the
lower Yangtze area, namely Soochow, Sung-chiang, and Ch'ang-chou,
altogether supported 10,565 recruited soldiers.129 In 1573 it was reported
that the Nanking garrison, which had an authorized strength of 120,000,
actually consisted of 22,000 men.130 In 1575 Kwangtung reported that the
province actually had 30,000 wei-so soldiers under arms, as against an
authorized strength of 120,000 men.131 With the inclusion of the hired
militiamen, it is estimated that in the provinces south of the Yangtze there
should have been a total military force of over 250,000 men in the two
decades between 1570 and 1590. To keep these units ready for active
service would have cost 6 million taels of silver a year, allowing 2 taels per
soldier per month. Though this estimate is a very tentative one, it does
show, nevertheless, that the cost of supporting the army remained the
government's heaviest financial burden. There is no evidence that the
problem was ever solved. To maintain T'an Lun's estimated army of
845,000 men would have cost 20 million taels a year in military supplies
alone, that is, more than half the state's total estimated revenues.

IV FISCAL RETRENCHMENT UNDER CHANG CH0-CHENG

The record of Chang's administration
The period of Chang Chii-cheng's administration from 1572 to 1582 was an
exceptional phase in late Ming financial history. Shortly before his death
 7, IV Fiscal retrenchment under Chang Chii-cheng 295
the granaries in Peking had enough grain in stock to meet the needs of the
next nine years.132 The deposits in the old vault of the T'ai-ts'ang Treasury,
which were not to be drawn on except in an emergency, rose to over
6 million taels of silver.133 The court of imperial stud held another 4 million
taels134 and the vaults in Nanking likewise contained reserves of 2.5 million
taels.135 The provincial treasuries were also well stocked with grain and
cash. Wang Shih-hsing (1546-98), on the basis of his conversations with
local officials, recorded that in the 1570s and 1580s, the prefectural and
provincial treasuries of Kwangsi, Chekiang, and Szechuan on average held
deposits of between 150,000 and 800,000 taels.136 This forms a paradoxical
contrast to the general state of sixteenth-century governmental finance.
Chang came to power in 1572 at an opportune moment, when the peace
settlement with Altan and the gradual falling off of the wo-k'ou raids
enabled him to carry out a general policy of retrenchment. This was aimed
at a drastic reduction of government expenditures without any curtailment
of income. Under his direction, all unnecessary and less urgent government
operations were either suspended or postponed.137 The number of students
on government stipends was reduced and the procurement missions of the
palace eunuchs were stringently supervised. Provincial officials were ordered
to cut down on corvee labor, in general to one-third of the existing level.
The hostel services provided by the imperial postal system were likewise
cut to the minimum. These cutbacks were not accompanied by correspond-
ing reductions in the contributions required from the population; the
savings simply accrued to the government treasuries.138 The proceeds from
fines, confiscations, and commutation of punishments were made subject
to auditing. Tax delinquents, most of them affluent landowners, were
prosecuted with vigor and a real attempt was made to collect their tax
arrears. The sale of official rank and ecclesiastical licenses was continued.
The austerity program was also extended to the army. Since the Mongols
were temporarily pacified, reductions could be made in the frontier guards
and border patrols. This permitted savings all round and also freed more
soldiers for farming.139 Governors-general in charge of frontier posts were
requested to reduce their expenditures so as to save up to 20 per cent of the
annuities sent to them by the central government.140 Army stud horses that
were being maintained by civilian households were sold and the stabling
services that these households had performed in lieu of land taxes were
commuted to money payments (3, n). The stipends due to the imperial
clansmen were generally not paid. Only after Chang's death did some of
them dare to appeal directly to the emperor, revealing that in some cases
the payments had not been made for over twenty years.141
Chang Chii-cheng's insistence on rigorous auditing of fiscal accounts
was unprecedented in the sixteenth century. Instead of relying on the
censorial officials, Chang began the auditing at the ministerial offices.
296 Financial management
In 1572, as a first step, the ministry of revenue declared twenty-eight types
of accounts superfluous and discontinued them; another twenty-two types
were simplified and consolidated.142 Thereafter the accounts, especially
those sent in by the frontier army posts, were condensed into a much
shorter format.143 These records were still cumbersome by present-day
standards but were at least easier to examine and review. In 1579, on the
suggestion of a supervising secretary, Chang ordered all counties and
prefectures to submit their service levy accounts to Peking for review.144
The accounts of Shantung and Hukwang are known to have been scruti-
nized by the grand-secretary himself.145 After the books were returned with
the required revisions the local officials were instructed to publish them as
a semi-permanent budget. The most enormous project for gathering fiscal
data undertaken on Chang's orders was the compilation of the Wan-li
Ku'ai-chi-lu, which began in 1572 and was completed in 1582, exactly the
period when he was in office. This work was used as the foundation for the
fiscal sections of the Ta-Ming Hui-tien. The junior officials participating in
the project included Ku Hsien-ch'eng, Li San-ts'ai, and Chao Nan-hsing,
all of whom later made names for themselves.146
Chang Chii-cheng's administration involved no innovations, but rather
laid stress on administrative discipline and strict observation of tax laws.
In 1576 it still required the emperor's personal intervention to make the
provincial officials deliver their tax payments in time, but in 1581 the grand-
secretary was able to report that due to the tightening up of personnel evalua-
tion with regard to fiscal matters the tax quotas were now generally filled.147
Chang's record in eliminating official corruption was a mixed one, and
clearly subject to controversy. His greatest success was to check the ex-
tortion by warehouse receiving men in Peking, who habitually demanded
extras from the civilian tax agents delivering goods to these depots. This
involved taking on a formidable adversary, the emperor's own maternal
grandfather, the Earl of Wu-ch'ing, who had been receiving payments
from the tax agents and arranging for them to deliver tax materials of
inferior quality to the warehouses. Chang secured a roll of cotton cloth
thus checked in, which was clearly below the official standard and urged
the youthful emperor to complain before the empress dowager.148 He then
took advantage of the public scandal to replace the eunuchs and other
personnel in the supply depots. In a letter written after the incident, in
1577, he revealed that some of the most notorious extortioners had been
put to death and the demands for 'cushion money' effectively stopped.149
On the other hand Chang made no attempt to compel all officials to live
within their salaries, which would have been impossible without substanti-
ally increasing them. Such a reform was beyond the capability of the
government and consequently not expected by the public. Chang's main
objective seems to have been the prevention of the misuse of the account-
 7, IV Fiscal retrenchment under Chang Chii-cheng 297
able items of public funds. Like his contemporaries, he apparently did not
think that absolute honesty in officialdom was necessary or possible. He
himself is known to have lived in elegant style and to have had luxurious
tastes. His biographer and critic Wang Shih-cheng (1526-90) even accused
him of advancing the careers of his subordinates in return for their bribes.
Since Wang was at odds with Chang, there is an element of malice in the
criticism.150 Yet several of Chang's subordinates, including Yin Cheng-
mou, governor-general of Kwangtung and later minister of revenue, and
Wang Ch'ung-ku, governor-general of Shansi and later minister of justice,
were considered by their contemporaries as corrupt. 151 In Chang Chii-
cheng's extant correspondences the mention of 'gifts' is frequent.152

The background of the administration

A discussion of the political background is essential to any understanding
of the fiscal management in the late sixteenth century, because govern-
mental finance under the Ming was far more responsive to the power
structure in Peking than to economic conditions in the empire.
Chang chii-cheng maintained his pre-eminence through his collaboration
with the eunuch Feng Pao and through his own position as the tutor of the
Wan-li Emperor. The confidence placed in him by the Empress Dowager
Li was likewise indispensable to his career.153 But under the Ming system
it was impossible for him to assume authority overtly. Furthermore at this
period no official would dare to suggest reorganization of governmental
institutions; the introduction of drastic reforms was an invitation to im-
peachment. As senior grand-secretary, Chang's official functions were
limited to drafting rescripts for the emperor. To initiate fiscal legislations
on his own authority would clearly have been out of order.
This was why Chang was forced to maintain his position for over ten
years chiefly through the manipulation of personal relationships. Through
Feng Pao he maintained cordial relations with the empress dowager, and
through her influence he controlled the emperor. By influencing the
emperor's orders he in effect exercised the power of appointment, which he
used to place his lieutenants in key positions in and out of the court. The
reports of the secret police kept him informed of all the major events of the
empire. His administration was effected through private correspondence
with senior officials. Before proceeding to measures of any significance it
was necessary for him to urge his trusted ministers and governors-general
to submit memorials proposing the desired change; in drafting the rescripts
to them he approved his own proposals.154 In a letter to Wang Tsung-mo,
imperial commissioner of the Grand Canal, Chang said:
Your servant now serves under a young sovereign. He is obliged to be extremely
law-abiding and to go easy on the people. All ambitious policies must wait until
298 Financial management
His Majesty adds a few more years to his august age and until his precious wisdom
has further developed. Only then can suggestions be leisurely presented for his
imperial decision.155
In view of what is known of Ming court practice, this statement may
genuinely reflect Chang's opinion, though it is tinged with over-exaggerated
humility.
It is impossible to guess the nature of the 'ambitious policies' he en-
visioned. His extant letters reveal only his vigor and devotion to duty,
which do not by themselves make Chang an original thinker. The men
whom he helped to place in important positions were remarkably capable,
though not all of them were noted for financial integrity. They included
important names of the late sixteenth century such as Ling Yun-i, Wang
Ch'ung-ku, Chang Hsiieh-yen, Liang Meng-lung, P'an Chi-hsiin, Chang
Chia-yun, Yin Cheng-mou, Ch'i Chi-kuang, and Li Ch'eng-liang. The
letters that he wrote to them contain a mixture of cajolery, persuasion,
polite reprimands, and hints of career advancement. Most of the issues
discussed in them were connected with regional administrative problems,
such as the enforcement of tax laws, troop dispositions, water-control
projects, and so on, and though the grand-secretary showed himself
remarkably well-informed and his directions were usually sound, the scope
of the problems under discussion was too narrow for him to demonstrate
his genuine greatness.
Chang Chii-cheng never proposed the creation or abolition of an office.
If he ever contemplated increasing official salaries, it is not apparent in his
writings. He did at one time indicate that minting copper coins would
benefit the population, since silver was in short supply,156 but made no
attempt to carry this out when he was in office. On the contrary, his efforts
to build up silver reserves further reduced the money supply, resulting in
deflationary tendencies. Whereas Ni Yiian-lu, the last minister of revenue
under the Ming, did attempt to induce the Ch'ung-chen Emperor to abolish
the wei-so system, carry out fiscal decentralization, transport tribute grain
by sea, and build up a base in central or south China to derive income from
industrial and commercial sources,157 Chang Chii-cheng advocated no such
far-reaching measures. It would be unfair to condemn him on these grounds
however; any judgement must take into account the difficulty of his
position and the complacency and senility of the Ming court.158
Chang Chii-cheng differed from his fellow officials over the enforcement
of tax laws. He argued that it was not wrong for the state to build up its
armaments and fiscal strength,159 and that in this respect he aimed to return
to the ideals of the Hung-wu and Yung-lo Emperors.160 This was not
enough to silence his critics. Though some of their concern over Chang's
monopoly of power and the increasing harshness of his administration may
well have been genuine, it seems likely that they were largely motivated by
 7, IV Fiscal retrenchment under Chang Chii-cheng 299
personal factors. Wang Shih-cheng, never hiding his own grudge against
Chang, made the objective observation that the grand-secretary's austerity
program was resented by the officials (who regarded travel at the expense
of the post system as their rightful privilege), hated by the government
students (who had to wait longer to enter the civil service), and disliked by
the eunuchs (who saw their income from procurement missions being cut
off). His rigorous auditing of the tax accounts displeased the large land-
owners in the lower Yangtze area, who were used to keeping their payments
in arrears. In addition to all this Chang was intolerant of criticism.161
Unknowingly, he thus made himself the enemy of the entire empire.
Meanwhile the existing system, which depended on a delicate balance
between the prevailing ideology and vested interests, and was safeguarded
by the dynastic founder's instructions that his style of government should
never be altered, endured firmly enough to frustrate the most capable and
determined statesman of Ming times.

On the threshold of a major reform

Because most of his time was devoted to gathering more accurate fiscal
data, enforcing the current tax laws and filling up the treasuries, Chang
Chii-cheng was never able to introduce any fundamental reforms. Though
he took the preliminary steps that might have led to a reform, these of
themselves did not result in any institutional changes.
The only change in legislation that Chang actually carried out was the
abolition of the stabling services provided by civilian households. This
measure, which represented a deviation from the fiscal instructions of
Hung-wu and Yung-lo, aroused no criticism as it affected no vested
interests. No other institutional reform was attempted. The minting of
copper cash was still totally inadequate and the salt monopoly was never
reorganized. Though Ming-shih claims that general implementation of the
Single Whip Reform was first decreed in 1581, within the period when
Chang was in power,162 the statement is clearly erroneous. Only in 1588
did the governor of Shansi request its adoption in that province, by which
time Chang Chii-cheng had already been dead for six years.163
The extant sources show that Chang adopted a cautious attitude towards
the controversial Single Whip Reform. It must be remembered that when
the reform was first introduced in north China it was strenuously resisted
by the population and by officials of northern origin (3, in). When in 1570
it was adopted in Shantung it had soon to be rescinded in the face of local
opposition.164 In 1577 a senior supervising secretary, Kuang Mou, a native
of Shantung, even memorialized the emperor requesting the abolition of
the reform all over south China and demanding punishment for Pai Tung,
magistrate of Tung-ao county in Shantung, who had carried it out in that
300 Financial management
district.165 The imperial rescript, drafted by Chang Chii-cheng, read: 'We
have already announced that the adoption of the Single Whip is entirely
dependent on the wishes of the local population. It has never been our
intention to call for its universal implementation. The matter is hereby
closed.'166 In a private letter to Li Shih-ta, governor of Shantung, Chang
admitted that Pai Tung was a capable official and that there were few dis-
advantages to the reform but that the opposition had whipped up such
feeling against it that he himself was helpless.167 These incidents reveal how
little he was in control of events, and how far politics in the capital were
dominated by the combination of vested interests and ideology.
Chang did, however, take one step towards a major reform by initiating
a national land survey. In a letter to a provincial governor he stated:
'A land survey will benefit the small people and will inconvenience the
affluent households and those of an official background.'168
For several years he found it prudent to defer the measure, for the decree
promulgated on the Wan-li Emperor's accession in 1572 had explicitly
ruled out the possibility of a national land survey.169 Chu Tung-yun
speculates that after 1577 Chang underwent a psychological change, and
that his decision may have been hardened by the impeachment actions
against him and the growing opposition.170 The survey was first tried in
Fukien in 1578. The counties were notified that its purpose was to facilitate
their internal tax apportionment and that their tax quotas would not be
revised regardless of the outcome.171 This assurance was apparently given
to prevent the magistrates from under-reporting when pressurized by
local landowners. The whole provincial survey took one and a half years,
being completed in the summer of 1580. On 16 December 1580, it was
decreed that the survey be carried out over the whole empire.172 As far as
is known, this was the only order of its kind ever issued during the Ming.
When Chang died on 9 July 1582, the returns were still not complete.
The details of the survey are not at all clear. The imperial order entrusted
the responsibility for it to the provincial administrative commissioners,
their circuit intendants, some military circuit intendants and the local
officials. The 240-square-pace standard mou was declared the universal
unit of measurement; this was recorded in the gazetteers of Chi-chou in
Shantung, Huai-ch'ing prefecture in Honan, and Huai-yiian county in
Shensi, all of which had previously used larger measurements than this.173
According to Ming-shih curve measurement was introduced,174 but it is
impossible to verify this claim.
The survey could not be considered a success. Honan province took one
and a half years to submit its returns, which were later discovered to be
simply the old data resubmitted. Though the provincial officials were
reprimanded and ordered to carry out the survey again, the second report
was rushed through only five months after the first was rejected.175 Wen-
 7, IV Fiscal retrenchment under Chang Chii-cheng 301
shang county in Shantung had previously conducted a local survey in 1567;
its magistrate, on being notified of the national land survey, merely informed
the landowners of the new unit of measurement so that the latter could
convert their own land-holdings.176 K'ai-hua county in Chekiang also
evaded the survey, by simply adding 0.27 mou to every mou of land
previously registered.177 On those districts which did carry out the survey
Wang Shih-hsing made the following general observations. For the land
close to the cities, the surveyors used the yardsticks.* Over 20 // from the
cities, they measured with ropes but beyond 50 // even the ropes were dis-
carded. The report was thus no more than a general estimate.178
The returns of the national land survey of 1580-1 were never officially
published and the provincial aggregate figures given in the Shih-lu are in-
complete (see Appendix D). Two months after Chang Chii-cheng's death,
the survey was severely criticized. The emperor therefore ordered that in
those districts where the population was satisfied with the returns, the
officials should use the new land data as the basis for taxation. In other
districts the governors and surveillance commissioners were to take
'corrective measures'. No new survey was authorized.179 The arguments
still continued. Some officials contended that all the survey records of 1581
should be invalidated and taxation remain as before,180 while others
actually made a second survey. Ling-chang county in Honan conducted
a local land survey in 1588 and Wen-shang county in Shantung did so in
1591.181 Yet, all things considered, Chang Chii-cheng's land survey was not
a total failure. In some counties the returns of 1581 were used as the
new basis of taxation, as was the case in Shun-te county (3, i). The
failure was mainly at the national level, and a complete set of land
statistics for the whole country was not successfully compiled until the
present century.

Governmental finance after Chang Chii-cheng

Six months after the grand-secretary's funeral all his measures were reviewed
and his associates either dismissed or impeached. The nature of the post-
humous charge against Chang was never clearly stated, but from a note
appended to his collected works, published by his son in the seventeenth
century, it appears that it was suspected high treason.182 The feelings against
him were so intense that in the 1580s the local officials who had been
negligent over his land survey were praised as righteous men.183 In these
circumstances to continue his policies was out of the question.
Thus Chang Chii-cheng, despite his efforts and intentions, actually did
little to reorganize the fiscal administration. Nevertheless his retrenchment
* The stick was made in the shape of a bow, with studs on each end and was rotated like
a compass.
302 Financial management
can still be credited with having prolonged the life of the dynasty by half
a century. The so-called 'three grand campaigns of the Wan-li era', that is
the Korean war against Toyotomi Hideyoshi from 1592 to 1598, the cam-
paign against Piibei in 1592 and the suppression of Yang Yung-lun and his
Miao tribesmen from 1594 to 1600, could never have been carried out
without the treasury reserves that he built up. The government actually
began drawing on the reserves of the T'ai-ts'ang Treasury several years
earlier. Up to 1587 there were still 6 million taels of silver in the treasury's
old vault which in theory were never to be spent, and in addition 4 million
taels buried underground in the new vault. The initial withdrawals were
from the latter. In 1588 and 1589 a total of 1.75 million taels was removed,
and in 1590 another 1.06 million taels were taken. In three years the total
reserves of 10 million taels were reduced to slightly over 7 million.184 In the
1590s the government began to draw on the old vault, the Chang-ying
Treasury administered by the court of imperial stud, the reserves in Nan-
king and the provincial treasuries.185 The regular taxation system, which
badly needed overhauling, could make little positive contribution to the
war effort.
When in 1592 the government authorized Chekiang province to increase
its land taxes by 0.003 taels of silver per mou, the proceeds were retained
by the provincial officials to strengthen coastal defenses, in anticipation of
a naval invasion by the Japanese.186 In 1598 a surcharge was imposed on
all service levy accounts; the counties being required to divert between
20 per cent and 40 per cent of their militia service expenses to finance the
imperial army.187 In 1599 the governors of Hukwang and Szechuan were
permitted to increase their land taxes at their own discretion in order to
finance the campaign against Yang Yung-lung.188 Though in theory the
quotas of the inland customs duties and the salt revenue were increased,189
it was subsequently reported that the actual income fell even below the
previous level (5, iv and 6, i). It was fortunate that at the point when
treasury reserves were almost exhausted and the ineffectual nature of the
tax increases was becoming apparent the series of military campaigns also
came to an end. But it is clear that by 1600 both governmental finance and
the taxation system were in a worse state than they had been in 1572 when
Chang Chii-cheng took over and possibly even worse than they had been
in the mid sixteenth century.
Most traditional historians and some modern scholars hold the Wan-li
Emperor solely responsible for this deterioration in affairs of state.
Admittedly Wan-li, cynical and indolent, is hardly deserving of sympathy.
In 1584, at the age of twenty-one, he ordered the construction of his
own mausoleum, which was completed four years later.190 When recently
excavated it was found to contain dazzling riches.191 The monarch's greed
for wealth was legendary. From 1596 onwards he sent eunuchs and some
 7, IV Fiscal retrenchment under Chang Chii-cheng 303
army officers as 'mining supervisors' to the provinces. In theory all under-
ground mineral resources were open to exploitation by the populace; the
government took no part in extracting them but merely collected half the
proceeds.192 In practice, however, the eunuchs issued orders to the local
officials, conscripted labor at will and dispatched their personal agents to
harass the rural areas. In time they attracted crowds oflocal ne'er-do-wells
who persistently extorted money from the people on the pretext that
deposits of metal existed under their houses and in graves. When the
eunuchs were put in charge oflocal business taxes, riots broke out in many
cities. The abundant records of malpractice in the last half of the Wan-li
reign make depressing reading.193
These events should not, however, be allowed to distract attention
from the main issues. It should be stressed that many dangerous problems
were developing in public finance even before Wan-li's abuses. The
emperor indeed exercised arbitrary power, but outside the formal fiscal
structure and the governmental organization. Throughout the sixteenth
century the latter was never able adequately to mobilize the financial
resources of the empire. Chang Chii-cheng's retrenchment represented
basically a negative approach to the problem. As no effort was made to
strengthen the fiscal apparatus the benefits of the accumulated treasury
reserves could only be temporary.
Though Wan-li's self-indulgence cannot be defended, there is little
validity to the charge that he wrecked the fiscal foundations of the empire
single-handed. His behaviour reflected the dynasty's institutional weakness.
From the mid fifteenth century onwards the emperor's position became
increasingly oriented towards ceremonial functions, and had little to do
with public service. Imperial pomp and extravagance proliferated and the
50,000 eunuchs and palace women, like the civil officials, were virtually
unpaid, though their food and clothing were provided by the state. A closer
scrutiny of Wan-li's personal expense accounts reveals that some of the
charges brought against him were exaggerated.
The imperial estates, which produced an annual income of 49,000 taels,
provided the expense accounts of several dowager empresses.194 The portion
received by Empress Dowager Li was largely donated to the construction
of stone bridges outside Peking and to religious institutions.195 The
emperor's personal account consisted of the Gold Floral Silver, which
yielded a million taels a year, but some 200,000 taels of this provided the
pay of the army officers in the capital. After 1578 this was compensated for
by the so-called 'purchasing money' (mai-pan fei) demanded from the
ministry of revenue, again bringing the privy purse close to a million taels.
In addition Yunnan annually supplied the palace with 2,000 taels of gold;
in 1592 Wan-li increased the quota to 4,000 taels.196 The emperor never
travelled apart from making short visits to the imperial tombs in the
304 Financial management
suburbs of Peking, but spent most of his money on jewelry and on presents
for his favorites.197 Some of this gold and jewelry has recently been found
in his mausoleum. Another heavy item of his expenditure seems to have
been the wedding ornaments of the imperial princesses.198 When he died
in 1620 the vaults in the Forbidden City were found to contain approxi-
mately 7 million taels of silver, the bulk of which was transferred by his two
successive heirs, T'ai-ch'ang and T'ien-ch'i, to the ministries.199 Wan-li's
worse characteristic was his greed which made him hoard money as well
as spend it. In order to keep his own savings intact he often forced the
government treasury to pay his petty bills.
In their memorials to the emperor, Ming officials often criticized his
personal extravagance by citing various items of palace expenditure over
a number of years. One frequently-cited article was silk, the cost of which
often ran into several million taels.200 The imperial satin, woven to special
designs showing the wearer's rank, was tailored to a sort of dress uniform
for wear in the palace. Every year between 8,000 and 28,000 bolts of such
fabrics were ordered at an average cost of about 12 taels of silver apiece.201
The costs were deductible from the tax accounts of the districts which
produced them. At times, however, the imperial allocation failed to cover
the total expenses or else local tax collection fell short of the delivery quota,
resulting in a shortage of funds. The local officials then had to adjust the
difference either by abridging payments at some point or else by making
extra impositions on some taxpayers. In 1575 when affairs of state were
still under the control of Chang Chii-cheng 97,000 bolts of such fabrics
were ordered, to be delivered over several years.202 Though the procure-
ment program resulted in fiscal irregularities, it was consistent with the
general practice of the dynasty.
The most distorted story concerning Wan-li's personal expenditures is
connected with the wedding of his son, the later T'ai-ch'ang Emperor.
Wan-li wished to remove him from the succession, but the officials con-
sidered primogeniture to be an inviolable rule of the dynastic constitution
and bombarded the emperor with requests that the crown prince be
formally installed, that he be tutored by the Hanlin academicians, and that
his wedding date be promptly announced. Every possible attempt was
made to secure the heir apparent's right to the succession. Though the
emperor continued to put off the proposals with excuses, he had no legiti-
mate way to enforce his personal wishes or even to put a stop to the peti-
tions. In 1599, as a last resort, he ordered the ministry of revenue to collect
24 million taels of silver in preparation for the weddings of his three sons.
This was merely an ingenious subterfuge, as both the emperor and the
officials were well aware that the ministry would never be able to raise
such an enormous sum. Both the Shih-lu and Ming-shih, however, record
the demand without explaining the circumstances.203
 7, IV Fiscal retrenchment under Chang Chii-cheng 305
Though such a tale may seem inappropriate for inclusion in the present
study, it is mentioned because in recent years several scholars, including the
author of a highly-specialized study of Chinese monetary history, have all
taken it for granted that the 24 million taels were actually delivered and
spent.204
 8
Concluding observations

The Ming empire remained non-competitive with other nations, militarily

and economically; therefore it was not keenly concerned with adminis-
trative efficiency. Even when governmental institutions were degenerating,
this seldom led to an immediate crisis, for the population exhibited remark-
able tolerance of mismanagement. Unsatisfactory conditions could some-
times persist for decades or even for over a century without causing serious
alarm. Furthermore, the ad hoc, piecemeal measures adopted to deal with
problems had the tendency not to solve them, but to transfer them into
other areas. For these reasons the institutional history of the Ming dynasty
is extremely difficult to analyze. Earlier discussions of such matters as the
manipulations of the currency, the management of the salt monopoly, and
the administration of the wei-so system has provided many examples of
such complexities. The case of army logistics is most illuminating. Having
been neglected since the mid fifteenth century, it became a matter of
genuine concern only 100 years later. When a solution was devised, it did
not involve reorganization of the armed forces, but further increases of the
land taxes.
Obviously, most historical problems have deep roots. The constitutional
crisis in Stuart England cannot be understood without reference to the
Tudor Reformation, and it is only comparatively recently that historians
have established that the foundations of Meiji Japan were in fact laid
during the late Tokugawa period. It is thus essential to take a long-term
view of Ming financial history, especially as the government avoided
periodic reorganization and drew no clear-cut divisions between its depart-
ments. The fiscal apparatus is noted for its ability to grow like a living
organism.

Historians who adopt the methodology of the social scientist usually

prefer to divide their topic of study into many small segments, which are
subjected to intensive and specialized examination. No final synthesis is
attempted until all available evidence has been analyzed in minute detail.
This approach, though logically sound, is of only limited value in the study
[306]
 8,1 The risks of over-simplification 307
of governmental finance under the Ming. The narrow perspective may
involve the investigator into a greater risk than would a broad survey. The
final analysis thus arrived at may not be an accurate presentation of the
whole system, but rather that of its disjointed parts. In the same manner,
to rearrange Ming financial data is not a simple matter. The effort to bring
some logical order to what is essentially an institution of disorder may
facilitate the comprehension of the reader, but at the same time obscure the
issues about which he wishes to know. The narrator may unknowingly
assume the role of a fiscal reformer rather than that of a financial historian.
It is also impossible at present to treat the subject in exhaustive detail,
owing to the scarcity of reliable data. What is clear, however, is the general
style of the administration; all the scattered information on its various
aspects presents a consistent pattern.
This general style of administration furnishes the substance of the dis-
cussion in the following sections. First it is examined together with several
theories of Chinese historical development, and then analyzed for its long-
term consequences in Chinese history.

I THE RISKS OF OVER-SIMPLIFICATION

The dynastic cycle theory
This theory has been propounded by both traditional historians and
modern scholars. The former, preoccupied with moral issues, attributed
dynastic decline to the character of the ruler, portraying most dynastic
founders as paragons of virtue and the monarchs at the end of the dynasty
as corrupt and incompetent tyrants. Modern scholars on the whole have
replaced the ethical and personal elements in the theory with economic
factors.
The foremost proponent of the dynastic cycle theory in the present
century is Wang Yli-ch'iian, whose influential article 'The Rise of Land
Tax and the Fall of Dynasties' places particular emphasis on the Ming and
Ch'ing. In his view the Ming dynasty collapsed because 'the agricultural
economy of China was bled to exhaustion by special land taxes levied on
the peasantry'. 1
In describing the Ch'ing, Wang is even more explicit. He says:
The process of corruption may be described briefly as one in which the central
government was robbed of real wealth and power, which were transferred to
the very individuals who, as members of the ruling class, controlled the govern-
ment. They could not be possibly restrained because, while responsible as
officials and as a class for protecting the interests of the nation, they were as
private individuals the sole beneficiaries of corruption. While some of them, as
officials, understood what was wrong, the most they could accomplish as a
308 Concluding observations
class was to try to protect both the government interest and their class interest
by trying to make up for the taxes which they themselves evaded, by increased
taxation of the poor and unprivileged classes.2
The charges against the Ming are not substantiated by the rates of tax
collection which it imposed. The tax increases ordered by the government
in the early seventeenth century, in the name of 'Manchurian supplies'
and 'bandit-suppression supplies', at the very most produced a total of
21 million taels of silver a year3 and this extra revenue was not raised
entirely from the land taxes. For instance in 1623 when the state had to
meet additional military expenses of 6,668,677 taels it was expected to
produce 4,491,481 taels from increases on the land tax, while the rest of the
revenue was raised from forced savings on other expenditures, the selling-
off of government properties, and miscellaneous taxation, including a tax
on pawnshops.4 Even if the 21 million taels of silver had all been added to
the land taxes the burden on the taxpayers would have been high but not
absolutely unbearable, as Wang asserts. Inflation in the early seventeenth
century, when commodity prices in general rose by 40 per cent,5 meant that
the tax increases were to a large extent illusory. This is corroborated by the
fact that the Ch'ing subsequently retained most of the Ming tax rates.
Though Wang considers that the Ch'ing kept its promise to reduce taxes,6
a check of the fiscal records for the early part of the dynasty shows that
this was not the case.7 Taxation under the new dynasty was at about the
same level as before, that is, at approximately the same rates that Wang
asserts to have bled China's agrarian economy to exhaustion.
Nonetheless Wang's description of the late Ch'ing, quoted above, is not
entirely inaccurate and actually bears remarkable points of resemblance to
conditions in the late Ming. It should be noted here that widespread tax
evasion by influential landowners and their shifting of their tax burden to
the less affluent constituted only one of several manifestations of insti-
tutional breakdown, the fundamental causes of which went much deeper.
In the late Ming, despite the proposed tax increases, even the regular tax
quotas were never filled. In 1632 tax arrears of 50 per cent or more were
reported in 340 counties, that is, more than a quarter of the fiscal districts
of the entire empire. Moreover, 134 of these counties had actually delivered
no tax payments to the central government whatsoever.8 This situation
supports the view that tax collection under the Ming, for historical reasons,
had a definite ceiling, and that when the demand for revenue significantly
exceeded this limit it led to the collapse of the fiscal apparatus. All this
cannot fully be explained by higher rates and tax evasion. The solution is
more likely to be found in the tax quota system. From the fourteenth cen-
tury and even earlier the income from agricultural land, after regular taxes
and a minimum wage had been paid to the primary producers, was claimed
by a host of interested parties by virtue of special kinds of land tenure,
 8,1 The risks of over-simplification 309
leases, mortgages, reserved rights, share-cropping, and private transfers of
public obligations (4, n). In addition some share of it went to local officials,
office clerks and rural auxiliaries in the form of customary fees and gifts
(4, v). Since the tax rates were generally low and money was scarce, this
combination of circumstances must have created a paradise for agrarian
exploitation on all levels and on all scales. In the course of two centuries,
the tax-paying capacity of the population was thus severely undercut.
Sometimes even the share of the exploiter was so small that he himself was
reduced to a minimum level of subsistence. In general the effect of the
increased agricultural productivity was nullified by a growing population.
Though slight tax increases might sometimes effectively cut into the
agrarian profit margin, any drastic increase was bound to meet with
resistance. In addition to the powerful landowners, many different sections
of the population were involved. Tax evasion became a really serious
problem only when the general collection was held up.
The theory of the dynastic cycle thus fails to take full account of the
institutional weakness. Financial administration under the Ming was handi-
capped by the limited capacity of its apparatus and the rigidity of its
approach. The proponents of the dynastic cycle theory simply take it for
granted that at the founding of the dynasty its institutions were in a state of
perfection and that all later deviations from this ideal state were due to
corruption. Their research usually adds little to our understanding.

The Ming dynasty as a feudal state

Since most modern scholars are Western-trained they naturally tend to
view Chinese history in terms of Western experience. Though the com-
parative method is obviously useful, it can lead to distortion, especially as
simultaneous developments in Chinese history can often be viewed as
corresponding to different stages of development in the history of the
Western world.
The Ming dynasty should probably be viewed as neither entirely 'back-
ward' nor completely 'modern'. Herrlee G. Creel, in his study of the
origins of bureaucracy in China, goes so far as to say that 'as early as the
beginning of the Christian Era the Chinese Empire showed many similari-
ties to the super-state of the twentieth century', which he characterizes as
'modern, centralized, and bureaucratic'.9 The Ming, developing in this
tradition, certainly retained most of these so-called 'modern' features. It
has already been mentioned that the ministry of revenue as early as the
fourteenth century controlled the accounts of some 2,400 offices (1, i).
The use of stub-books in business transactions can be said to have antici-
pated the use of computer cards. Yet such practices as flogging tax
delinquents to death, assigning contraband quotas to patrolmen, and
310 Concluding observations
arbitrary confiscation of personal property all seem to mark Ming China
as 'medieval'. The Wan-li Emperor's construction of his own huge
underground mausoleum, using the labor of thousands of soldiers (7, iv),
might in fact be more appropriately labelled as 'ancient'.
In the past two decades it has become fashionable for Chinese scholars
to designate the Ming as a feudal society. While it is undoubtedly possible
to assemble a considerable number of episodes from the Ming histories
which show superficial parallels with medieval Europe, in terms of insti-
tutions and organization this categorization is clearly misleading. Under
the Ming dynasty there was not a single hereditary office, apart from that
of the emperor, that carried any functional responsibilities. After Yung-lo
even the emperor's consorts were deliberately selected from families of
lesser status, rather than those with aristocratic titles.10 From the fifteenth
century onwards the prestige of the military sank to probably its lowest level
in Chinese history. In the sixteenth century neither the central government
nor the provincial authorities were able adequately to support an army and
even their warships had to be rented (7, in). The rents on palace and aristo-
cratic estates were actually collected by civil officials rather than the title-
holders, effectively reducing the latter's privileges to a mere stipend (3, n).
Ironically enough, while some modern scholars consider that the mis-
management of the tax administration was due to the survival of feudal
practices, Ku Yen-wu in the seventeenth century argued that it was due to
the absence of feudal spirit in the government structure. Ku considered
that the solution was to restore what he called 'the essential meaning of
feudalism', which included the relaxation of centralized control, the right
of local officials to select their own subordinates, more fiscal authority for
local government, and even the opening of some public offices to inherit-
ance.11 While there is no point in debating the feasibility of Ku's proposed
reforms, it should be emphasized that they were an attempt to deal with
genuine problems. In the late Ming, though the imperial government was
in theory omnipotent, in practice it was often unable to act; the local
officials could have achieved much more but did not possess the necessary
authority. This resulted in deadlock. The land-tax administration could not
cope with conditions in the rural districts and mismanagement was wide-
spread in all areas and at all levels of administration. As was shown earlier,
taxes could not be collected without applying pressure and the pressure was
generally directed to those who could offer least resistance to it. The arbi-
trary and excessive demands of the tax collectors, which some modern
scholars regarded as a feudal characteristic, in part reflected this loss of
control, and in part represented attempts by officialdom to compensate for
its own organizational weakness. Oppression was thus not a sign of
strength, but of a lack of it. It can by no means be considered the funda-
mental characteristics of any system.
 8,1 The risks of over-simplification 311
Possibly less confusion might have arisen if these scholars had limited
their discussions of' feudalism' to those practices within traditional Chinese
polity the origins of which could be traced to China's remote feudal
past. The principle of group responsibility, for instance, was adopted by
the feudal states of the first millennium B.C. when they first attempted to
create a bureaucratic administration, and was never subsequently aban-
doned. The emphasis on the service obligations of imperial subjects was
a historical survival of the practice of paying tribute. But in general these
Marxist historians show little patience with historical details. Starting
from the premise that the Ming was a feudal society, they then go on to
postulate the 'historical inevitability' of class struggle within it, concluding
with the emergence of 'budding capitalism' in the late Ming. The impli-
cations of this last view will be considered later.

Moral interpretation of history

Chinese traditional historians, not surprisingly, always tended to interpret
history in terms of moral leadership. When good men were in power,
governmental finances were naturally in a sound state. Conversely,
corruption always started with degeneration in the quality of leadership.
Since Chinese administrative theory traditionally laid more stress on the
personal ability of the officials than on the creation of specialized insti-
tutions, this approach sometimes has its uses in assessing localized, short-
term situations. Its complete disregard of all environmental and technical
questions make its limitations obvious, however.
In the previous chapter it was shown that Chang Chii-cheng, despite his
strong character and undeniable moral courage, had only limited success
in financial administration (7, iv). Though his savings on state expenditure
undoubtedly strengthened public finance in the short term, the deflationary
effect caused by his building up of silver reserves must have resulted in
considerable public distress. His methods moreover were conditioned by
political pressures; the fiscal machinery was forced to work harder without
being overhauled. Resistance was only to be expected, and probably
accounts for his own posthumous disgrace. The subsequent reaction not
only led to the abandonment of his policies but was also the beginning of
a general split in the bureaucracy.12
Though one might expect the moral interpretation of history to be
generally discredited by now, this is not the case. Some modern historians
tend to view the Confucian morality of certain individuals in terms of their
own sense of social justice. Wu Han, for instance, praises his hero Hai Jui
for' standing on the side of peasantry' in their conflict with the landlords.13
He also asserts that Hai, 'despite a hundred setbacks, still continued the
struggle to build a socialist society'.14
312 Concluding observations
In reality Hai Jui was an orthodox and rigorous Confucian who de-
manded the same austerity of his subordinates as he did of himself. As
governor of South Chihli, he waged a relentless struggle against official
corruption and the gentry's abuse of the tax laws, which undoubtedly
evidenced great personal courage and integrity. Nonetheless he can hardly
be considered a social reformer, let alone a revolutionary. In his frequently
cited memorial to the Chia-ching Emperor he greeted the sovereign with
the statement: 'All under heaven is Your Majesty's own household.' 15 His
opinion of the common people (hsiao-min) was that they 'rival each other
in greed, all being motivated by profit so that their lawsuits are unending'. 16
After his retirement, moreover, Hai compromised his own standards, as
his letters show, by accepting 'gifts' from a local prefect, a surveillance
commissioner, two military defence intendants, and two governors-general,
the last two being Ling Yun-i and Yin Cheng-mou, who were generally
regarded as corrupt.17 At least one of the gifts was large enough to enable
him to purchase a piece of real estate.
The purpose here is neither to downgrade an upright man nor to exoner-
ate official corruption, but to emphasize that institutional weaknesses in
the late Ming were too serious to be remedied by moral rearmament. Yet
Wu Han writes that H a i ' all his life was opposed to bad men and bad deeds
but never to good men and good deeds', 18 and censures anyone who has
criticized his hero including those who, though they approved of Hai's
objectives, felt that he could have exercised greater diplomacy in pursuing
them.19 This kind of attitude has already spread to the study of insti-
tutional history, as is shown by the fact that for some writers the Single
Whip Reform is above criticism (3, in). The revival of historical stereotypes
is a dangerous practice.*
It seems of extremely doubtful value to classify the personalities of
sixteenth-century Ming China with the current labels of 'liberal' and
'progressive', 'conservative' and 'reactionary'. While few of them fit
neatly into any one category, many overlap into several at once. Though
practically all of them aimed to preserve traditional social values, egalitarian
principles were by no means absent from their thinking. While they relied
on the state to protect their own personal interests, they also held that the
livelihood of the general population was of prime importance to the state.
To label them as 'good' and 'bad' leads therefore to confusion rather than
clarification.

* To make things more complicated, Wu Han was himself purged for reasons connected
with his interpretation of Hai Jui.
 8, II Ming administration in Chinese history 313

II MING FINANCIAL ADMINISTRATION AND ITS

PLACE IN CHINESE HISTORY
Centralization ahead of technology
A generalization can readily be applied to the vast range of disparate topics
discussed in this volume: Governmental finance under the Ming represented
an attempt to impose an extremely ambitious centralized system on an
enormous empire before its level of technology had made such a degree of
centralization practical. By the latter is meant both practical technology
and economic expertise. They included transportation, communication,
other service facilities, the principles of money and banking, techniques of
accounting and data processing, and even the mental attitudes of the
officials.
The financial system created by Hung-wu, of not separating the emperor's
personal income from that of the state, nor imperial from provincial
expenditures, imposed a unified administration over all the financial
resources of the empire. Such a scheme would be ambitious even for
modern times, and not unnaturally encountered technical problems from
the very beginning under the Ming.
The difficulties would not have been so acute if the scheme had been
carried out in an area of dry cereal cultivation, such as the northern loess
plan. The geographical diversity of China as a whole multiplied the
problems enormously. The Ming, relying on land taxes as its major source
of income, never produced a set of consistent land data throughout its
history of 276 years (2, n; 3, n; 7, iv); on reviewing the regional reports it
becomes understandable. Since the taxation had to cover all households,
including the lowest income group, every type of land had to be accounted
for, down to the reedy stretches along the banks of the major rivers and the
mountain forests (6, in). On Hai-nan island, for instance, palm trees were
counted in order to provide funds for the service levy.20 Fish ponds were
often too productive to be omitted, and likewise mulberry groves, which
were sometimes far more profitable than rice paddies, were not neglected.
It was not infrequent for single-cropping land to give higher yields than
the same acreage of double-cropping land in the same locality. Some
types of land provided only a modest annual return, but a reasonably
certain harvest, whereas conversely some highly productive land might be
more subject to crop failure. The possible variations are simply too
numerous to be described, but it is obvious that the odds were heavily
against any scheme of detailed universal taxation. The administration was
also badly short of technical skills. Even in the late sixteenth century, land
surveys were conducted not by professionally trained teams but by workers
from the villages who were incapable of comprehending even the pre-
314 Concluding observations
fectural standards for land classification, let alone those at the national
level. Lack of control over standards facilitated all kinds of abuses, even
before a survey was carried out.21
In Tokugawa Japan, by contrast, land taxes were as a rule assessed on the
village communities collectively, not on individual taxpayers.22 Though
the Japanese method might seem crude and backward and the Ming
system far more modern and egalitarian, in practice, owing to the absence
of effective control in China, this was not necessarily so. It is probable that
the Japanese feudal lords knew the conditions in their own villages much
better than the Ming magistrates did those in their assigned counties.
Centralization in Ming China created many other paradoxical results.
Though it might be expected that it would stimulate the development of
technology and economic theory, this did not in fact happen. It is essential
to remember that the integration affected only fiscal authority, not
fiscal responsibility, and that the actual handling of materials and goods
and manpower was still done on the lowest possible level (2, i). The fiscal
data, which were not standardized to begin with, easily deviated from
reality because they were not directly related to actual business trans-
actions. The establishment of the quota system made many of the techni-
calities even more irrelevant. Gradually fiscal logic became distorted. The
overall level of taxation and the regional tax quotas, which should have
been elastic, became fixed, whereas such concrete units as one adult male,
a mou of land, and a picul of grain, were all transformed into elastic
measurements. The administration became less and less methodical and
more artistic, as it evaded rather than solved its technical problems.

Separation of theory from practice

Another characteristic of the Ming system was its limited handling
capacity, which was a deliberate part of the founder's design. Hung-wu,
being a cautious man, preferred to create a centralized fiscal authority with
extensive rather than intensive coverage. His main concern was to prevent
any sub-system arising within this monolithic structure, rather than to
refine and improve on its operations. From the beginning, therefore, the
fiscal administration was characterized by simplicity and even crudity.
This is evidenced by such practices as assessing taxes directly on productive
labor, and failure to re-invest tax income or to build up logistical capacity
at the intermediate level. But the low level of taxation was only made
possible by the dynastic founder's policy of peace abroad and extreme
austerity at home. The army was made to produce its own food, and
village autonomy cut down the functions of local government. Under-
staffing saved administrative overhead. The tax quota system, however,
was introduced with no thought to its possible consequences.
 8, II Ming administration in Chinese history 315
Though it did reduce the tax burden on the population, it should logically
have been followed up with a series of protective measures. The creation of
a sound monetary system was essential, to enable the state to control credit
and interest rates. It would also have been desirable to impose some effective
control over property transactions and surveillance over land mortgages
and rents. Such provisions were essential if the peasants were to enjoy the
benefits of their minimal tax rates in perpetuity, but needless to say nothing
of the kind was done.
By the fifteenth century conditions had already begun to deteriorate.
No adequate monetary system was ever created and the self-sufficiency of
the army turned out to be a myth. Census-taking and periodic registration
of property bore little relation to reality, as is shown by the disappearance
of large tracts of public land (3, II). As the government had from the
beginning failed to establish control in the rural areas it was impossible to
introduce it later.
As a result fiscal practice diverged increasingly from theoretical ideals.
Tax increases were imposed in the form of surcharge readjustments, or
of local procurement orders. Officials supplemented their meager salaries
with unscheduled collections. Funds earmarked for legitimate expenses
were siphoned off to meet other fiscal needs. Though the permanent
regulations were never revised, much of the administration no longer con-
sidered itself bound by them. All this continued in the sixteenth century.
It has been argued throughout this work that, contrary to the usual view,
mismanagement was not simply a matter of over-taxation. Attention should
be directly not exclusively to palace extravagance, which was a form pf
misappropriation, but to the inadequacies of public services, the local
government budget and army logistics. When the financial history of the
Ming is examined in detail the unfortunate consequences of the govern-
ment's excessive emphasis on frugality are clear. Income and expenditure
were always in excess of the limited handling capacity of the system, and
this affected even the self-indulgence of the emperors. Since public finance
was always in disorder, the emperors saw no reason why they alone should
exercise self-restraint. When Wan-li dispatched his eunuchs as business tax
collectors (7, iv) he may well have considered that he was only keeping up
with his bureaucrats.23
The organization created by the dynastic founder, though clearly un-
workable, was so vast in extent that no thorough overhaul could be carried
out, except perhaps by the establishment of a new dynasty. Left as it was,
however, the system was obviously incapable of handling the problems of
the day. Owing to its ideological preconceptions, rigid sense of responsi-
bility, compartmented spheres of action, unrealistically low salaries, in-
sufficient office personnel, lack of information, deficiency in logistical
capacity at the intermediate level, and reluctance to invest (all connected in
316 Concluding observations
one way or another with initial under-taxation) the state was unable to
mobilize the empire's financial resources, some of which were within its
grasp. This was always the case with regard to the salt monopoly, maritime
tariff, inland customs duties, forest produce levy, and government mining.
Eventually the taxpayers had to pay more, especially to those who were
unable to resist the extra-schedule collections. Most contemporary writers
never understood that it was the low official rates that were ultimately
responsible for the increased tax burden on the populace, and only a few
realized the harm caused by under-taxation (4, v).
Though in the sixteenth century some improvements were made, they
had only limited effect. Moreover as they were neither systematic nor
thorough, they mostly had undesirable features of their own. On the
positive side, the collection of military supplies in the southern provinces,
the introduction of the Single Whip Reform and the compilation of accounts
in silver did at least bring the financial system somewhat closer to reality
and meant that some of its resources were put to legitimate use. On the
other hand these measures may have used up much of the remaining tax-
paying capacity of the population, which had not been and could not be
expanded.
Retrogression and retardation
The Ming financial administration was undoubtedly inferior to those of
some earlier dynasties. Even under the Sung, the fiscal administrators had
discovered that by gearing their policies towards an expanding economy
the state could increase its revenues without over-burdening the taxpayers.
Wang An-Shih's fiscal reform, antedating the Single Whip Reform by
500 years, undertook the task of converting the service obligations of the
populace into monetary payments. The state in turn stepped up its pro-
duction of copper coins, the potentially inflationary effect of which was
neutralized by the increased demand for money in tax payments. The Sung
frequently used its financial power to achieve its economic goals. Even
under the Yuan dynasty, land taxes were initially assessed in copper coins.
When grain was required by the state, the amounts were calculated from
the assessments, effecting a commutation in reverse. This enabled the fiscal
accounts to be integrated.24 All these devices and techniques seem to have
been completely lost under the Ming. Whereas during both the T'ang and
the Sung there was a distinct trend towards professionalism in fiscal
administration, 25 under the Ming, salt administrators were often drawn
from the ranks of disgraced officials (5, v). In many financial offices,
including the ministry of revenue, the actual work was performed by clerks
(1, i), but in the early seventeenth century even these lesser functionaries
also hired substitutes to perform their duties.26 Obviously the quality of the
administration underwent a drastic decline.
 8, II Ming administration in Chinese history 317
The inertia of the Ming administration had many social and economic
implications. One is that the service sector of the economy was seriously
retarded. Mention has frequently been made of the dynasty's failure to
work out a monetary policy. In fact it appears that the government minted
copper coins in only 40 out of its 276 years and that the average output was
no more than 200 million coins a year (2, iv). Thus at a rough estimate, the
total Ming production of copper coins was 8 billion coins, the amount that
the Northern Sung could have minted in two years. Many Ming coins are
visibly inferior to those minted by the Sung in the eleventh century. This
failure resulted in the widespread use of unminted silver in both official
and private transactions, which led to enormous inconvenience and con-
fusion. Worse still, even when the government failed, private agencies were
not permitted tofillthe gap. Up to the end of the sixteenth century there is
little mention of private financial institutions27 and the only ones in the
money market were the pawnshops.
In the area of transportation, the most the Ming achieved was to keep
the Grand Canal open to traffic. Yet despite the vast quantity of materials
and labor used in its maintenance (7, n), the waterway played a much
smaller part in vitalizing China's economy than many modern scholars
have assumed.28 Though a remarkable achievement, the Grand Canal
contained lakes, rapids, and channels full of sandbars, and at the northern
end crossed the courses of two rivers, both of which froze during the winter.
At its junctions with the Yangtze and Yellow rivers only government boats
were permitted to pass through the watergates. All other vessels had to be
unloaded and lifted over by capstans.29 In 1548 a Japanese tributary
mission spent sixteen days at the entrance to the canal before itsflotillaof
5 boats could be transferred to it from the Yangtze.30 In the middle section
of the waterway, there were 38 watergates over a distance of less than
200 miles. These gates, only twelve feet wide, had to be repeatedly closed
and opened to adjust the water level which, according to travel logs of the
late Ming and early Ch'ing, was often less than four feet in depth.31 It must
be recalled that there were around 12,000 boats carrying tribute grain on
the canal, and that they quite often took a whole year to complete the
round trip, including the time that they were frozen in the northern sector
during the winter months. If placed end to end their combined length would
actually have occupied one-tenth of the entire distance from the Yangtze
river to the northern terminal. In addition, the court operated another
fleet of 1,800 craft for haulingpalace supplies from Nanking.32 Most of their
cargo was of doubtful economic value, including such items as imperial
furniture, dragon robes, and fresh fruit and vegetables from the south.33
There is no evidence that the Ming government spent funds on road
maintenance and construction, apart from erecting the stone bridges out-
side Peking. No effort was made by these means to increase the efficiency
318 Concluding observations
of the imperial postal system, the primary function of which was to trans-
mit official documents, not to serve the public. Its supply procedures
actually discouraged local road construction programs, for it was felt that
better roads would encourage more official travel in the area. The costs of
this, and of providing hostel accommodation, all had to be met by the local
population which was reluctant to see its burdens increased. When in 1560,
owing to the wo-k'ou campaign, Shun-an county in Chekiang suddenly
became a focal point for inland traffic, the financial burden imposed by
official travel proved so intolerable that the magistrate finally decided to
suspend all travel by land. Instead travelling officials were requested to
make a detour by water, even though this was more time-consuming.34
The Ming's only achievements in construction were the building of the
Great Wall, palace halls, ceremonial gates and imperial mausoleums.
The fact that public services were starved of funds inevitably retarded
the progress of technology and of the economy as a whole. Silver mines
were actually sealed off (6, i) and the process of salt manufacture de-
generated, huge iron plates being eventually replaced with small caldrons
and then bamboo trays pasted with paper and alkalis (5, n). In the late
sixteenth and early seventeenth centuries many Ming soldiers wore helmets
made of bamboo, and armor made of paper.35
This evidence conflicts directly with the theory, vigorously promoted by
the history seminar of the Chinese People's University, that the late Ming
and early Ch'ing witnessed the 'budding of capitalism' in China.36 Ad-
mittedly during this period general political stability, population growth
and a rise in the standard of living of the upper classes, did lead to in-
creasing economic diversification and specialization in certain parts of
China, notably the southeastern provinces. But this in itself is not necessarily
indicative of incipient capitalism, which is characterized by a persistent
trend towards the accumulation of capital by industrial and commercial
enterprises. So far the theorists have only come up with isolated cases in
which individuals became affluent through the farming of cash crops or
participation in the handicraft industries. As Albert Feuerwerker has
pointed out, they have yet to substantiate their claims with more convincing
data. It must be pointed out that in the late Ming most of the service
facilities indispensable to the development of capitalism were clearly
lacking. There was no legal protection for the businessman, money was
scarce, interest rates high and banking undeveloped. Such an environment
was hardly favorable to industrial production and the efficient circulation
of goods. At the same time merchants and entrepreneurs were hindered
by the frequent road-blocks on the trade routes, government purchase
orders and forced contributions, the government's near monopoly of the
use of the Grand Canal and active involvement in manufacturing. On the
other hand the security and status of land ownership, the tax-exemption
 8, II Ming administration in Chinese history 319
enjoyed by those who purchased official rank, and the non-progressive
nature of the land tax increased the attractions of farming to the detriment
of business investment.
The Marxist concept of 'class struggle' is often applied too literally. In
fact there have been many instances, during the early stages of the develop-
ment of capitalism, of entrepreneurs entering into partnership with the
feudal authorities. In Tokugawa Japan, for instance, the daimyos licensed
the merchants, granted them monopolistic privileges, and even commis-
sioned them as their own business agents. The merchants for their part
provided the feudal lords with service facilities and business expertise, and
sometimes advanced credit. It was in other words a kind of joint enterprise,
with the feudal authorities controlling agrarian production and the
merchants in charge of marketing.37 Only in this way could farm surpluses
gradually be invested in the industrial and commercial sectors of the
economy. It is quite possible that the feudal lords had to liberalize their
trade regulations and bring them into line with commercial practice in
order to promote the expansion of the business in which they themselves
were partners.
In sixteenth-century Ming China the government had practically no
surplus to dispose of. The actual handling of its financial resources was so
fragmented that no large-scale operation run on commercial lines was
necessary. The officials were simply content to use the service facilities at
their disposal; if these proved inadequate they resorted to commandeering
rather than contracting for extra service. The only department which might
have proved an exception was the salt monopoly, yet this still relied on the
managerial experience derived from other sectors of the administration, its
primary concern being merely to fulfill its immediate quota. It exhibited
neither cost-consciousness nor any awareness of the need for long-term
planning, and at best can only be considered to have provided a model for
an early form of bureaucratic capitalism, the kuan-tu shang-pan enterprises
under the Ch'ing.
It is doubtful whether capitalism would of itself have created a more
desirable social order in sixteenth-century Ming China, in view of the
conditions experienced by the urban proletariat in other industrializing
societies. It could, however, have led to increasing production and more
efficient circulation of goods, besides promoting legal reform and techno-
logical advance. Significantly, it is in the following areas that China's
backwardness has been most pronounced: a low per capita income,
an inefficient transportation system, archaic legal proceedings and the
failure to institutionalize technology. Ming fiscal administration clearly did
nothing to ameliorate these conditions.
The inadequacy of the administration is further exemplified in the
management of government-owned industries. This touches another
320 Concluding observations
controversy. Many mainland Chinese historians, while labeling the Ming as
feudal, at the same time credit it with 'promoting productive forces'
(shen-cKan-li) by requiring artisans to work in rotation in government
manufacturing agencies (see 6, m(r)). The specialized division of labor in
these agencies is said to have resulted in technological improvements, the
effects of which spread to the private sectors of the economy.38 In point of
fact from the midfifteenthcentury onwards few of these artisans performed
their statutory duty in person but discharged it with silver payments in-
stead. The actual workers were hired substitutes. In 1562 even the remain-
ing statutory laborers also had their duties compulsorily commuted to
silver. The hired artisans working for the government were generally of two
kinds. Some of them, such as the silversmiths and jewelers in the Imperial
City, produced only luxury items; in reality this entire category of artisans
was committed to elaborate piece-work on products which had little market
potential. The other category of workers was to be found in agencies
involved in mass production. Management in these was generally de-
centralized. A factory was usually divided into several units, each under
a supervisor who assumedfiscalresponsibility for it. For instance, a furnace
chief in the government mint, in charge of a squad of workers, was issued
with a fixed amount of copper and expected to produce a fixed quota of
coins; he personally had to make up any short-fall in production (2, iv).
In these circumstances it was impossible to organize production on an
assembly-line system. Despite the mint's seemingly impressive production
quota there was little incentive to introduce labor-saving devices. The Ch'ing-
chiang-p'u dockyard near Huai-an in the same way comprised eighty-two
units, each of which had its separate housing and was under a separate
manager (2, i). Though the dockyard in its heyday produced 746 rice
carriers in one year, each unit constructed no more than 10 craft. The so-
called functional specialization of the workers actually implied no more
than their trade identification, with carpenters constructing the hull of the
vessel, blacksmiths forging the anchor, and bamboo-workers making the
sails. At most there may have been one or two specialists who did jobs like
caulking.39 The Ming officials in charge of these operations, so far from
being content with their technological progress, made many complaints
about the craftsmanship of their workers, considering them inferior to the
artisans in private establishments.40
The arsenals in Peking, also admired by some modern scholars, con-
centrated on producing ornamental weapons for the palace guards. Be-
cause of the supply problems, most of the weapons used by thefieldarmy
had to be locally procured. Ch'i Chi-kuang, when in charge of army
training in the 1570s, found that the weaponry of his own command was
not standardized, and complained frequently of projectiles that did not fit
the gun barrels, and gun barrels that tended to explode.41 The Ming
 8, II Ming administration in Chinese history 321
government's manufacturing enterprises were actually even more backward
than their critics had imagined.
The institutional resemblance of the Ming to previous dynasties is clearly
misleading. The T'ang, the Sung, and the Yuan never imposed such a rigid
fiscal structure as did the Ming, nor did their top government offices assume
so little operational responsibility as under the Ming. The Ming adminis-
tration was indeed remarkably self-denying, in that it reduced its own
operational capacity to the minimum, neglected to develop revenues from
industrial and commercial sources, and refused to consider the possi-
bility of seeking assistance from private quarters. The whole tone of the
administration was regression rather than progress.

Purposes of the financial administration

On the whole, the virtues of the financial administration were negative ones.
In traditional China the major concern was always governmental stability,
which was well served by the Ming fiscal system.
One of its notable advantages of the system was to prevent the regions
from over-developing any financial potential of their own, and thus from
challenging the central government. Thus every fiscal office in the empire
was forced to draw its revenues from innumerable different sources. This
meant that though provincial and local officials were effectively prevented
from improving the quality of their administration, they were likewise un-
able to assert their independence against the state. This system, combined
with a policy of frequent rotation of office and of preventing officials from
serving in their home provinces, worked so well that throughout the Ming
period no official or general ever attempted to raise a rebellion. Most
revolts in the Ming seem to have been led by imperial princes, discontented
peasants, or tribal chiefs. One frequent reason for their failure was that
none of them commanded a financial base substantial enough to sustain
the uprising in its early stages. The only exception was Yang Yung-lung,
whose rebellion, supplied by the Miao tribes, lasted from 1594 to 1600. The
empire's fiscal operations were so fragmented as to make them virtually
safe from capture, and the mere knowledge of this fact was sometimes
sufficient to discourage potential rebels. In 1555 Chang Ching, governor-
general in command of the armed forces of six provinces, was abruptly
arrested by imperial order on a very flimsy charge and was subsequently
put to death.42 It is doubtful whether the most powerful emperor in the
Han or T'ang could have acted with such ease and confidence.
The security of the dynasty was founded on ideological indoctrination,
bureaucratic consensus and the vigilance of the censorial officials and the
secret police. Military power was of comparatively less importance. When
in 1553 a band of wo-k'ou pirates, reported to be fewer than 100 men,
322 Concluding observations
managed to make a raid south of the Yangtze penetrating several hundred
miles inland, they met with practically no resistance.43 Though this was
partly due to the decline of the wei-so system, it also shows that the Ming
state, despotic as it might seem, in normal times could survive with very
few garrisons, particularly in the interior provinces. This apparently en-
couraged the emperors to divert army funds for their own extravagant
purposes. The situation did change somewhat after the middle of the
sixteenth century, but not drastically. The funds retained by the southern
governors as military supplies were still composed of small items from
a variety of sources (7, in).

The long-term consequences

The Manchu takeover in the seventeenth century was indeed a setback for
China, but not because, as the Marxist historians claim, the invaders
represented new forces of feudal reaction who crushed the development of
indigenous capital in its early stages.44 The major fault of the new dynasty
was to follow the old too closely, disregarding the criticisms of thinkers like
Huang Tsung-hsi and Ku Yen-wu.45
As alien conquerors who to some extent lacked administrative experi-
ence, the Manchus were simply anxious to gain effective control and to
promote a swift return to normality. They did attempt to tighten up
administrative discipline, notably by a rigorous persecution of tax delin-
quents in 1661,46 but in general they avoided major institutional reforms
particularly with regard to fiscal practices. No other major dynasty in
Chinese history had introduced fewer organizational changes. Before the
nineteenth century, the only important modifications made to the Ming
system were the permanent freezing of the ting quota as of the record of
1711, the subsequent merging of ting payments with the regular land taxes,
the institutionalization of melting charges, and the creation of supplemen-
tary official salaries to principal officials in the name of' nourishing honesty'
(yang-lieri)*1
The essentials of the Ming fiscal system, notably its monolithic structure,
undivided fiscal authority at the top in the person of the emperor, com-
bined with the residual fiscal responsibility at the lower levels, the quota
system, the criss-crossing supply lines, and the absence of a central
treasury, thus survived in China up to the present century.48 The inflexi-
bility of the Ch'ing land tax system, as described by Yeh-chien Wang, was
inherited from the Ming; apparent increases in revenue were in fact largely
the result of price inflation. When allowance is made for this, it appears
that the collection actually declined in the closing years of the dynasty.49
Though the Ming financial system had many unusual features, by the
time it had been in operation with only minor modifications for close on
 8, II Ming administration in Chinese history 323
500 years, its peculiarities were taken for granted and its social and econo-
mic consequences accepted, not unjustifiably, as being typical of traditional
China. It is rarely observed that the Ming system represents a significant
break in Chinese fiscal history. From this time onwards the main aim of
governmental finance was to maintain the political status quo, and it
ceased to exhibit any dynamic qualities.
Perhaps the most remarkable feature of the financial establishment was
its passivity. There was almost no centralized planning. Preventive measures
or readjustments were only undertaken when a crisis was actually at hand.
This complacent attitude is illustrated by the mismanagement of the mari-
time tariff right up to the Opium War.
Throughout this study more emphasis has been placed on institutional
inadequacy at all levels than on personal graft and corruption in the
bureaucracy. The latter, though abundantly documented, were of second-
ary importance, and can all too easily distract attention from the major
problem of the financial administration: its superficiality.
The survival of this system depended on continuing cultural and political
dominance over a large and self-sufficient economy which was able to dis-
regard commercial pressures and competition from outside. The sixteenth
century was, however, a turning point in world history, the period when
Western Europe was beginning its modern transformation. One of the
consequences was the arrival of Europeans in East Asia in large numbers.
Even though China's size and remoteness from the West were to delay the
final conflict for about three centuries, the first challenge to her isolation
and self-sufficiency had been made. As early as 1524 Cristavao Vierira,
himself for a time imprisoned in Canton, was talking of seizing the city
with no more than fifteen Portuguese ships and 3,000 men.50
The evidence presented in this volume, though incomplete, indicates that
many of China's more recent economic problems, such as the difficulty of
re-investing agrarian surplus in industrial production, have their roots in
the past, and are in some cases traceable to the sixteenth century. Any fiscal
policy, carried on for a long time, will inevitably affect the history of the
nation. It is therefore unwise to dismiss the financial system created by the
Ming as inconsequential. Its lack of positive features is not an indication of
its lack of overall influence.
 Works cited only by alphabetical
abbreviations

CSWP Hsii Fu-yuan f& ^ g£ et al. ed. Huang-Ming Ching-shih Wen-pien M W ® "ffi
3t II, reproduced blockprint edition Taipei, 1964.
HYWCL Chang Hsuan % H, HsUyuan Wen-chien I « g | B | , £|, reproduced from
a 1627 manuscript. Yenching University, Peking, 1940.
/CL Ku Yen-wu |g i£ K. Jih-chih-lu Chi-shih R ft] £| ^ P . W^/w-jw Wen-k'u
^^XW- edition.
MS Ming-shih IH^ j£. National Defense Research Institute edition, Taipei, 1963.
M5L M/«^ Shih-lu BJ Jf ^ . Academia Sinica reprint. Taipei. The annals of individual
emperors are identified by their temple names, i.e. T'ai-tsu ;fc |§.; T'ai-tsung -X. ^ ;
Jen-tsung iz ^ ; Hsiian-tsung 2 ^ ; Ying-tsung ^ ^ ; Hsien-tsung ^ ^ ; Hsiao-
tsung ^ ^ ; Wu-tsung g| ^ ; Shih-tsung i t ^ ; Mu-tsung fj§ ^ ; Shen-tsung ^ ^ ;
Kuang-tsung it ^ ; and Hsi-tsung ^ %.
THCKLPS Ku Yen-wu, Tien-hsia Chiin-kuo Li-ping Shu % T M M V] ffi fl- Ssu-pu
Ts'ung-Wan VBUMfil edition.
TMHT Ta-Ming Hui-tien ^ 0J # Jft. Southeast Publishing Company reprint from
the 1587 edition.

This list is repeated before the notes to the text, p. 331.

[324]
 [325] Appendix A 326
Appendixes

Appendix A. Landed properties not subject to regular taxation

Estimated
Estimated annual
Category Location acreage income Disposition

Palace Shun-t'ien, Ho-chien, (Until 1522) (Late 16th The revenues were used
estates Chen-ting and Pao- 3.7 million century) as the expense accounts
ting prefectures, mou; (there- 50,000 tlsi. of the dowager em-
N. Chihli after) 2 mil- presses.
lion mou
Princely Mainly in Shantung (ca. 1500) Unknown, Princes Hui, Ch'i, Hsing
estates and Honan, partly 1 million but un- and Heng may have
in Hukwang. Acre- mou; (early likely to received 700,000 mou
ages in two peri- 17th century) have ex- before 1505. Prince Te
pheral provinces, 3 million ceeded received 130,000 mou
Shensi and mou 100,000 tls. in the mid 16th century
Szechuan, are un- at any Prince Fu received
certain time. 2,000,000 mou in the
early 17th century.
Other Concentrated in (Until 1530) 50,000 to In theory the local
aristocratic N. Chihli, but the 4.4 million 90,000 tls. officials delivered the
estates holdings of the mou (there- revenues to the title-
Mu family in Yun- after) 2.8 mil- holders. Sometimes the
nan uncertain lion mou estates were returned
to the state.
Pastures N. Chihli Unknown (Until 1580) The revenues were de-
belonging ll,000 tls livered to the court of
to the capi- (1596) imperial stud, and
tal garrison 30,000 tlsi. sometimes partially
to the ministry of
revenue.
Pastures N. Chihli, S. Chihli, (In 1596) (ca. 1576) The 1576 account shows
belonging Shantung and Honan 3 million 92,400 tls. that 85,000 tls. were de-
to the mou livered to court of im-
court of perial stud in Peking,
imperial 7,400 tls. to Nanking.
stud The 1580 account shows
that the ministry of
revenue received
65,000 tls.
Land as- N. Chihli, Shantung (ca. 1580-90) 50,000 tls. 34,000 tls. were delivered
signed to and Honan in excess of to the ministry of re-
the im- 3 million venue; the eunuchs col-
perial mou lected 16,000 tls.
stable and
imperial
zoo

Notes 1. Sources: MS, 77/821, 82/866; MSL, Hsien-tsung, 3678, 3708; Shih-tsung,
0151, 0258, 0651, 1516, 1842, 5884; Shen-tsung, 5152, 9193; TMHT, 17/22-4, 27-30,
23/1-11, 251/2-9; CSWP, 202/7-14; Sun Ch'eng-tse, Meng-yu-lu, 35/9; Yang Shih-
ch'iao, Ma-cheng-chi, 8/2-3; Shimizu, Mindai tochi seidoshi kenkyii, pp. 15-155.
2. With the exception of the pastures belonging to the court of imperial stud, landed
properties in the above categories were generally understood to be of poor quality and
the rents were set at 0.03 tls. per mou or lower. In some cases the princes and aristocratic
title-holders assigned managers to the estates and collected the rents themselves, despite
the prohibition on this. They also sometimes encouraged the practice of 'commendation'
whereby private landowners merged their own properties with these estates, paying
a nominal rent to the title-holders in order to evade taxation. The extent of this practice
was grossly exaggerated by contemporary writers, however, and some of their arguments
are logically inconsistent. See Shimizu, tochi seidoshi, p. 404.
Appendix B 327

Appendix B. Customary fees and extra services collected by the magistrate of

Shun-an county, Chekiang, 1561

1. Added to regular tax collection

All silver payments increased by 5 per cent.
Collection of summer tax in silk fabrics; extra payable by taxpayers: 174 tls. in silver
plus 12 bolts of sample silk fabrics.
Autumn tax collection: extra payable by the rural collectors: 20 tls.
Collection of payments from registered saltern households: extra payable by collectors:
10 tls.
Tax delivery: extra payable by each deliverer: 0.5 tls.
2. Fees for special occasions, payable by each of the county's 80 communities
Checking of registered artisans and hereditary military households: 1 tl.
Compilation of the county's tax register book: 1 tl.
Assessment of chiin-yao: 1 tl.
Compilation of the Yellow Book (once every ten years): 2 tls.
3. Added to the magistrate's salary and allowances
Magistrate, rank 7a: nominal salary 90 piculs of grain per year; actual collection from
taxpayers, 180 tls.
Allowance for personal attendants, etc. (the original account does not indicate the exact
amount, but usually the magistrate was allowed 30 to 50 tls.), actual collection increased
by 200 per cent.
4. Services provided by the '//' chiefs
The 80 // chiefs to attend the magistrate's office daily, in rotation.
Presents to the magistrate from each // chief, on his first turn of duty: polished rice, 1 or
0.50 piculs; geese, chickens, fish, candles, fruit, and 1 jar of wine.
The magistrate's travel expenses and gifts to fellow-officials: to be paid by the li chief on
duty.
5. Extra collections from salt merchants
On salt in transit: 0.1 tl. per 100yin. About 50,000yin in transit annually, making a total
of 50 tls. (The rate was about 1/5000 of the value of the salt.)
On salt sold in Shun-an county: 1 tl. per 100 yin. About 7,000 yin sold each year, making
a total of 70 tls. (The rate was about 0.5 per cent.)
Notes. 1. Source: Hai Jui, Hai Jui Chi, pp. 48-9. The list has been slightly rearranged
and explained.
2. Similar customary payments were provided for the magistrate's staff and the
prefect.
3. Tours of inspection by higher officials always imposed extra financial burdens on
the county. The tour of the salt-control censor could cost it 200 tls. and that of the
provincial governor somewhere between 300 and 400 tls. The cost had ultimately to be
borne by the general population, as no funds were available to subsidize them. See Hai
Jui, Hai Jui Chi, p. 62.
328 Appendix C

Appendix C. Barter rates and excise tax on each yin of salt, 1535

All the following packages were considered as one yin:

District and Government salt Surplus salt Total package

production Estimated
area Barter Excise income
Weight value (a) Weight tax (b) Weight (a + b)
(catties) (tls.) (catties) (tls.) (catties) (tls.)

Liang-Huai
Huan-nan 285* 0.50 265 0.65 550* 1.15
Huai-pei 285* 0.50 265 0.50 550* 1.00
Liang-Che
Chia-hsing 250* 0.35 200 0.50 450* 0.85
Hang-chou 250* 0.35 200 0.45 450* 0.80
Shao-hsing 250* 0.35 200 0.40 450* 0.75
Wen-chou 250* 0.35 200 0.20 450* 0.55
CKang-lu
North 205 0.20 225* 0.35 430* 0.55
South 205 0.20 225* 0.30 430* 0.50
Shantung
Entire 205 0.15 225* 0.31 430* 0.46
province

* = allowance for crating included.

Notes. 1. Source: TMHT, 34/12. This differs slightly from the passage in MSX,
Shih-tsung, 3793. The estimated total revenue, including the bartered commodities and
excise duty in cash, was 1,166,734 tls. See Shih-tsung, 3794.
2. Fujii has compiled a similar table, in * Mindai ensho no ichikosatsu', Shigaku zasshi,
54: 7 (1943), p. 734.
 Appendix D 329

Appendix D. Partial returns of the land survey of 1581, as recorded in

Ming Shih-lu

The Shih-lu records the provincial returns in two different ways. For some provinces it
gives the actual reported acreages, and these are indicated in the following table by
asterisks after the reported acreages. In these cases I have calculated the increase in
acreage. In many cases, however, the Shih-lu gives only the increase in acreage without
recording the actual survey results. These are indicated by asterisks after the increase in
acreage; here, conversely, I have had to calculate the reported acreages.

Reported Recorded Increase in Per-

Province acreage acreage of acreage centage
(special area) (a) 1578 (b) (a-b) increase

(mou) (mou) (mou) (%)

Chekiang 48,308,192 46,696,982 1,611,210* 3.5
Kiangsi 46,261,081 40,115,127 6,145,954* 15.3
Shantung 112,734,500* 61,749,899 50,984,601 82.5
Shansi 37,313,922 36,803,922 510,000* 1.5
Honan 94,949,374* 74,157,951 20,791,423 28.0
Shensi 50,299,925* 29,292,385 20,937,540 71A
Szechuan 40,934,767 13,482,767 27,452,000* 203.6
Kwangtung 32,960,030* 25,686,513 7,273,517 28.3
Kwangsi 9,478,961 9,402,074 76,887* 0.8
Pao-ting pref. 11,467,550 9,709,550 1,758,000* 18.0
N. Chihli
11 southern 45,158,050* 36,853,886 8,304,164 22.6
districts of
S. Chihli
4 northern pre- 29,553,047 27,227,047 2,326,000* 8.6
fectures of
S. Chihli
Total 559,349,399 411,178,103 148,171,296 36.1

Notes 1. Sources: MSL, Shen-tsung, 2190, 2225, 2237, 2283, 2343, 2346, 2356, 2361,
2371, 2436, 2449, 2712; TMHT, 17/8-12.
2. The following returns are recorded in the Shih-lu but are not included in the above
table because of lack of any basis of comparison:
Ta-t'ung (military command); land under private ownership, 3,153,979 mou; land
under military farming, 4,781,104 mou.
Liao-tung (military command); land under military farming, 890,350 mou; other,
2,418,870 mou.
Hsuan-fu (military command): total 6,310,036 mou.
Kweichow province: total increase in acreage, 159,495 mou. (The land classification is
not clear.)
Hukwang province: actual acreage, 83,852,546 mou. (The entry is ambiguous as to
whether this was a complete or a partial return. The above figure moreover includes
'lakes' as a land category.)
Kan-su (military command): actual acreage, 4,599,335 mou.
For these entries, in the above order, see Shen-tsung, 2238,2276,2341,2344,2412,2482.
3. Fukien province conducted its survey before the national land survey and its returns
are not given in the Shih-lu. There is no mention of Yunnan province.
330 Appendix D
4. The complete returns of the survey appear to be permanently lost. The editors of
the Shen-tsung Shih-lu merely summarized the above figures from the documents giving
the results of the survey, and it is doubtful whether they ever saw the actual survey records
themselves. Though Minister of Revenue Chang Hsiieh-yen declared his intention to
publish the complete returns (see his preface in Wan-li K'uai-chi-lu), no reference to such
a publication has ever been found. As the whole project came under severe criticism after
Chang Chii-cheng's death it is unlikely that anyone would have dared to suggest that the
records be preserved.
5. The above table itself shows that either the 1578 acreage was unrealistic or that the
1580-1 survey was not carried out on the basis of any uniform standard. This is evidenced
by the increases of over 200 per cent in Szechuan and of less than 1 per cent in Kwangsi.
There are probably fallacies in both views. For further discussion see Shimizu, Mindai
tochi seidoshi kenkyu, pp. 563-92. Shimizu notes on p. 583 and p. 592 that he has not
seen the original text of Wan-li K'uai-chi-lu.
 Notes

Works cited only by alphabetical abbreviations

CSWP Hsu Fu-yuan f£ ^ }t et al. ed. Huang-Ming Ching-shih Wen-pien M I1B JE i t
3^C l i , reproduced blockprint edition Taipei, 1964.
HYWCL Chang Hsiian jfg W, Hsi-yuan Wen-chien Lu ffi gg fill j | £g, reproduced from
a 1627 manuscript. Yenching University, Peking, 1940.
JCL Ku Yen-wu Jg j£ J£, M-chih-lu Chi-shih El ftl $ 4g P . H^«->w Wen-Vu
n^XM edition.
MS Ming-shih \$ A- National Defense Research Institute edition, Taipei, 1963.
MSL M/>?# Shih-lu PJ3 ff g t Academia Sinica reprint. Taipei. The annals of individual
emperors are identified by their temple names, i.e. T'ai-tsu ^c jjj§; T'ai-tsung ^c ^ ;
Jen-tsung (i ^ ; Hsiian-tsung 2 ^ ; Ying-tsung ^ 5f?; Hsien-tsung M ^ ; Hsiao-
tsung # ^ ; Wu-tsung ^ ^ ; Shih-tsung ty: 5j<; Mu-tsung 9 ^ *» Shen-tsung # ^ ;
Kuang-tsung %. %i; and Hsi-tsung ^ 5j<.
THCKLPS Ku Yen-wu, Tien-hsia Chun-kuo Li-ping Shu X T H$ 12 ^Ij IS § . &«/-/?*/
Ts'ung-k'an B PP K ?'J edition.
TMHT Ta-Ming Hui-tien ~X \% # A. Southeast Publishing Company reprint from
the 1587 edition.
This list is repeated before the appendixes, p. 324.
Most other references in the following notes give short titles only. For complete
references see bibliography, p. 367.

Modern reprints of Ming source materials have new page numbers as well as the old
ones. Since the Academia Sinica reprint of the Ming Shih-lu is widely used only the
modern page numbers are cited here, in accordance with the practice of the Ming Bio-
graphical History Project. For other works, such as the Ta-Ming Hui-tien, which are
available in more than one reprint, the original chapter number and pagination are cited,
omitting the letters a and b used to denote the first and second halves of a double page.

1 Fiscal organization and general practices

1 For the Nine Ministers' Conference, see Hucker, 'Governmental Organization',
p. 65.
2 MSL, T'ai-tsu Shih-lu, 2544-5.
3 MS, 81/849; MSL, Ying-tsung Shih-lu, 0224; Wada Sei, Shokkashi yakuchu,
pp. 721-2.
4 Wu Han, Chu Yuan-chang Chuan, p. 212.
5 See Hucker, 'Governmental Organization', p. 30 for the difficulties of the grand-
secretaries in acting as intermediaries between the emperors and the bureaucracy.
6 For the palace layout see Sun Ch'eng-tse, Meng-yii-lu, 6/8-17, 56-7.
[331]
332 Notes to pages 9-23
7 For the functions of those service agencies, see ibid. 6/56-7'; TMHT, 30/1-19; Ho
Shih-chin, CKang-k^u Hsii-chih, passim.
8 TMHT, 30/2,18; MS, 79/835. More information on palace warehouses is given by
Liu Jo-yu, Cho-chung-chih, 97, 149.
9 For the operations of the Kuang-huei Treasury see TMHT, 30/5-6; Ta-Ming
Kuan-chih, 4/2431.
10 The nei-kuan-chien was the highest-ranking eunuch and at one time exercised con-
trol over all the others. His power gradually fell into the hands of the ssu-li-chien,
possibly in the early fifteenth century. See Sun Ch'eng-tse, Meng-yii-lu, 6/56.
Cf. MS, 74/778; Hucker, 'Governmental Organization', p. 25.
11 Hucher, 'Governmental Organization', p. 25.
12 Ni Hui-ting, Nien-p'u, 2/11,4/8.
13 For the reorganization of the ministry of revenue during the Hung-wu period, see
MS, 72/740-1; MSL, Tai-tsu Shih-lu, 0609, 1261, 1481, 1723, 2068.
14 See MS, 139/1747, 149/1833, 157/1904, 202/2347; MSL, Ying-tsung Shih-lu, 1674,
1684, 1904; Ta-kao, 1/73; Yen Ts'ung-chien, Shu-yii Chou-tzu-lu, lift.
15 Based on MS, Chronological Table of Seven Ministers, 111/1400-112/1455.
16 MS, 160/1929,186/2170,194/2266,214/2488,220/2543, 241/2749, 256/2897; Chiang
P'ing-chieh, Pi-shao-pao-kung Chuan, 22.
17 MS, 151/1847; MSL, Ying-tsung Shih-lu, 1786, 2024.
18 The functions of the Nanking ministry of revenue are outlined in TMHT, ch. 42 and
can also be found in Pi Tzu-yen, Liu-chi Su-ts'ao. Japanese scholars cite a work
entitled Nanking Hu-pu-chih, which I have not seen.
19 TMHT, 41/42-4; Ta-Ming Kuan-chih, 4/2431-2, 2455.
20 Ta-kao Hsu-pien, 1/181-2.
21 MSL, Shen-tsung Shih-lu, 1076.
22 CKung-chen Ts'un-shih Su-cWao, 1/100.
23 MSL, Tai-tsu Shih-lu, 1481, 1723, 2066-7, 3054; TMHT, 2/4-10.
24 See 'Chiang Ch'en' in Ni Hui-ting, Nien-p'u, 4/8-9.
25 MS, 75/795; TMHT, 2/4-5; Sun Ch'eng-tse, Meng-yii-lu 37/1-2.
26 MS, 72/743, 225/2596; TMHT, 14/1.
27 MS, 225/2595.
28 Lu Shien-chi, Jen-chen-ts'ao, chs. 1 and 2.
29 /CL 13/79.
30 Sun Ch'eng-tse, Meng-yii-lu, 25/29.
31 Ibid. 38/1.
32 The precedent was established as early as 1444: see TMHT, 21/21, 22/32-40.
33 MS, 92/968; MSL, Tai-tsu Shih-lu, 1441,1526; TMHT, 150/1-18; Sun Ch'eng-tse,
53/5.
34 For the fiscal functions of the ministry of works, see MS, 72/749-51; TMHT,
chs. 181-207; Sun Ch'eng-tse, Meng-yii-lu, 46/1-3; Ho Shih-chin, CWang-k'u
Hsii-chih, passim.
35 An example of 'local procurement orders' is to be found in Hui-chou Fu-chih, 8/4.
The cost of this particular order was deductible from the deposits of the Yung-feng
granary, which was under the control of the prefect.
36 Ho Chung-shih, Ting-chien-chi; Hsiang Meng-yiian, Tung-kuan Chi-shih, passim.
31 In 1575, the procurement orders were defended by the ministry of works: see MSL,
Shen-tsung Shih-lu, 0951-2, 0956.
38 For an illustration, see MSL, Shen-tsung Shih-lu, 10764-5, 10768-9.
39 For the 1373 system, see MSL, Tai-tsu Shih-lu, 1503. For the 1393 re-organization,
see Ta-Ming Kuan-chih, 4/2361-2; MS, 71/734, 75/803. For the number of local
districts, see MSL, Tai-tsu Shih-lu, 1149.
40 Hucker, 'Governmental Organization', pp. 44-5.
41 MS, 75/803.
42 For the functions of a subprefecture, see MS, 75/803.
43 For example, in 1590 Yunnan was ordered to deliver to Peking the funds which it
had previously been permitted to retain: see MSL, Shen-tsung Shih-lu, 4177.
 Notes to pages 24-37 333
44 THCKLPS, 32/46.
45 MS, 75/803 has a brief outline of the magistrate's fiscal responsibilities. For the
details it is necessary to refer to the local gazetteers.
46 Hsi Shu and Chu Chia-hsiang, Ts'ao-cK uan-chih, 5/12.
47 MS, 75/803; Hucker, 'Governmental Organization', pp. 45-6.
48 One local official who took this initiative was Wang I, prefect of Soochow in 1538
For Wang's administration, see Shimizu, tochi seidoshi, pp. 556-61.
49 Hucker indicates that there were 'variable numbers of vice-prefects', see 'Govern-
mental Organization', p. 44. There were seven vice-prefects in Shun-t'ien prefecture,
including those who bore different titles. See TMHT, 2/28-9.
50 Hucker, Censorial System, p. 73. In general all local offices comprised six sections.
An illustration of prefectural government is provided by TMHT, 9/15-16.
51 The list of granaries appears in TMHT, 21/14-21, 22/1-27.
52 Hucker, 'Governmental Organization', pp. 43-4. It seems that the creation of
circuits under surveillance commissioners can be traced to the Hung-wu period. See
MSL, Tai-tsu Shih-lu, 3231-2. For the list of circuits see MS, 75/799-801.
53 For the origin of military defense intendants see MS, 75/810; MSL, Jen-tsung
Shih-lu, 0298, 0301; Hucker, Censorial System, pp. 71, 78.
54 MS, 79/834; MSL, Ying-tsung Shih-lu, 0135; TMHT, 22/29.
55 TMHT, 227/13-15.
56 Fen-chou Fu-chih, 5/5, 55-6.
57 TMHT, 21/21-6, 22/29-41. See also Chang Hsiieh-yen, Wan-li K'uai-chi-lu, ch. 23
for Hsiian-fu; Chang Yii, Pien-cheng-k'ao, ch. 4 for Kan-su; and Wei Huan,
Chiu-pien-Wao, ch. 2 for Liao-tung.
58 Hucker, 'Governmental Organization', p. 9.
59 Control over the imperial clansmen was gradually relaxed. See MSL, Shen-tsung
Shih-lu, 0609, 0637-9. In 1590 they were permitted to take the civil service examin-
ations, ibid. 4162-8. Presumably the permission was put into effect in 1595. See
Ping-ti Ho, Ladder of Success, p. 22. In the early seventeenth century some imperial
clansmen were appointed as business tax collectors. See Ch'i Piao-chia, Jih-chi: the
pagination of the original text is inadequate but the entry is to be found in vol. 5,
dated 2nd day, 9th lunar month, 1643.
60 MSL, Tai-tsu Shih-lu, 0070.
61 Wei Ch'ing-yuan classifies the households into three major categories: military,
civilian and artisan: see Huang-ts'e Chih-tu, pp. 20-1. An additional category, that
of saltern households, appears in TMHT, 20/5. Eight categories, including physi-
cians, Confucian scholars, Buddhist monks, and Taoist priests in addition to the
four mentioned, appear in TMHT, 9/25.
62 Ta-kao Hsu-pien, 1/150-1.
63 Feng Ch'i (1559-1603), for example, traced his family origins to military registration.
The family established a record by producing holders of the chin-shih degree in four
consecutive generations. Feng himself rose to become minister of personnel and
minister of rites: see Feng-tsung-po Chi; also MS, 216/2506.
64 TMHT, 20/4, 104/2-6.
65 Each hunter was required to provide annually one tiger skin and nine miscellaneous
hides: see Hui-chou Fu-chih, 1/47.
66 TMHT, 104/19.
67 Ho Shih-chin, CKang-k'u Hsu-chih, 9I62-3', THCKLPS, 9/46.
68 THCKLPS, 6/4, 5, 38, 39-40, 72, 94-6.
69 Fang-chung Liang, The Single-whip Method, p. 37; Yamane Yukio, Mindai yoeki
seido no tenkai, pp. 43-7.
70 TMHT, 20/10-11.
71 See MSL, Tai-tsu Shih-lu, 1279, 1507, 1724, 2653; MS, 78/825; TMHT, 29/2-3.
72 MSL, Tai-tsu Shih-lu, 2144.
73 Ibid. 1724-5. Liang Fang-chung, Liang-chang Chih-tu, p. 42.
74 Ibid. pp. 6, 48, 54.
75 Ibid. pp. 62-3, 70-2.
334 Notes to pages 38-45
76 THCKLPS, 6/41, 83^.
77 The list of postal stations appears in TMHT, chs 145-6. In addition there were
140 transport offices of which a list appears tin TMHT, ch. 147. For details see
Su Tung-pin, I-ti Chih-tu.
78 For an example of double-cropping land paying taxes twice, see Wei Ch'ing-yiian,
Huang-ts'e Chih-tu, pp. ii-iv In the late Ming, however, many districts levied both
their summer tax and autumn grain quotas on all land owners, in which case every
piece of land paid tax twice.
79 MS, 78/823.
80 Temporary commutation of tribute grain from Chia-ting county, South Chihli,
started in 1584. Thereafter the commutation order was renewed once every three
years. In about 1596 the commutation became permanent. See THCKLPS, 6/25.
81 A yield of 2 piculs of husked grain per mou was reported in Ch'ang-shu county, South
Chihli; Shang-yii and I-wu counties, Chekiang, Chang-chou prefecture, Fukien; and
Shun-te county, Kwangtung. Except in the case of Shang-yii county, Chekiang, all
sources are dated before 1600. See Ch'ang-shu Hsien-chih, 4/13; Chang-chou Fu-
chih,5/53; Shun-te Hsien-chih, 3 /I; THCKLPS, 22/118; Ni Hui-ting, Nien-p'u, 3/13.
82 Yeh Sheng (1435-94) reported that most properties in K'un-shan county, South
Chihli, yielded 'over 4 piculs of husked rice and wheat per mou annually'. See
Jih-chi, 31/12. Wang Ao, writing about 1506, indicated that top-quality land in Sung
times paid taxes at 1.5 piculs per mou and assessed the tax rate as 30 per cent of the
yield. These two statements together suggest that around 1500 this top-quality land
actually yielded 5 piculs of grain annually. Though the nature of the grain is not
here specified, the Sung is known to have collected land taxes in husked rice. See
Ku-su Chih, 15/1. (Unit of measurement under the Sung, however, was somewhat
smaller.) In the early Ming, some landed properties in Soochow prefecture actually
paid taxes at a rate of 2 piculs of husked rice per mou: ibid. 15/6.
83 Ho Liang-chim, Ssu-yu-chai, 3/179. The productivity also varied from 3 piculs of
husked rice to 1.5 piculs per mou.
84 Ping-ti Ho, Studies on the Population, pp. 102-23.
85 MS, 78/824; TMHT, 17/13; Shimizu, shakai keizaishi, pp. 17 ff.
86 Ibid. p. 17.
87 MS, 77/819.
88 MSL, Tai-tsu Shih-lu, 2726. Cf. Chin-hua Fu-chih, 6/2.
89 Ping-ti, Ho, Studies on the Population, p. 108; Fujii, 'dendo tokei', pp. 104-5.
90 Wada, Shokkashi yakuchu, p. 48; Wei Ch'ing-yiian, Huang-ts'e Chih-tu, p. 74.
91 THCKLPS, 2/28; Fujii, 'dendo tokei', p. 105; Shimizu, tochi seidoshi, p. 462.

2 The heritage of the sixteenth century and major fiscal problems

1 In 1384 an administrative commissioner in Fukien was executed for violating this
rule: see Ta-kao Hsii-pien, 1/117-18, 119-21. Though it was not enforced in later
reigns, it is still recorded in TMHT, 173/3.
2 MSL, Tai-tsu Shih-lu, 3396, Tai-tsung Shih-lu, 0654-5, Hsiian-tsung Shih-lu, 1991.
3 In 1371 the total number of provincial and local officials was 5,488: see MSL,
Tai-tsu Shih-lu, 1176; Hucker, 'Governmental Organization', p. 70.
4 Ta-kao Hsii-pien, and San-pien, passim.
5 For the 1382 case see MS, 94/987-8; Wu Han, Chu Yuan-chang, pp. 166-8. Meng
Seng considers that the case actually took place in 1376; see Ming-tai Shih, p. 57.
For the 1385 case, see MSL, Tai-tsu Shih-lu, 2490, 2581,2631; Ta-kao, 1/26, 29-31,
54-5, 77; Hsii-pien, 1/143-5, 147-9. Wu Han suggests that these two cases claimed
70,000 to 80,000 lives: see Chu Yuan-chang, p. 159.
6 MSL, Tai-tsu Shih-lu, 2871, 2998.
7 The emperor left a list of fifteen states to which his descendants were permanently
prohibited from sending expeditions: see Huang-Ming Tsu-shiin, 3/1589-91; Wu
Han, Chu Yiian-chang, p. 154.
8 The level of taxation was not expected to rise. The emperor denounced all the
 Notes to pages 46-51 335
financial experts of previous dynasties who had increased state revenues as enemies
of the people: see MSL, Tai-tsu Shih-lu, 2141, 2681-2.
9 MSL, Hsiao-tsung Shih-lu, 3548-55, cf. Chiao-k? an-chi, 648-50 for typographical
errors.
10 Wang Chih-jui, Ching-chi-shi, p. 135.
11 Ibid. pp. 31, 62. A seventeenth-century scholar also observed that the level of
taxation was below that of the Sung: see Shen Te-fu, Yeh-huo-pien Pu-i, 2/27.
12 The quota system was given in Ta-Ming Kuan-chih, 4/2430, 2452. In 1387 the fish
duties were declared to be outside the quota system: MSL, Tai-tsu Shih-lu, 2779.
13 MSL, Tai-tsu Shih-lu, 1848.
14 Ibid. 2641.
15 Ibid. 3370. Previously the level of land taxes tended to increase. In 1381 the income
was 26,105,251 piculs; in 1385: 20,889,617 piculs; in 1390: 31,607,600 piculs; and in
1391: 32,278,800 piculs. Ibid. 2218, 2673-4, 3078-9, 3166-7.
16 Ibid. 3532; TMHT, 17/16-17; JCL, 4/46.
17 MSL, Tai-tsung Shih-lu, 1652.
18 In 1408 Governor Huang Fu was authorized to fix the tax quotas for the districts of
Annam, ibid. 1043.
19 From 1425 to 1428, it always remained close to 30 million piculs, the figure for the
latter year being 30,249,936 piculs: MSL, Hsuan-tsung Shih-lu, 1196.
20 MS, 153/1863-5; MSL, Hsuan-tsung Shih-lu, 1448, 1639-40.
21 MSL, Tai-tsu Shih-lu, 1176.
22 MSL, Ying-tsung Shih-lu, 5417.
23 HYWCL, 34/2.
24 For the Hung-wu salary schedule see MS, 82/864; MSL, Tai-tsu Shih-lu, 1182,
1598, 2061-2, 2101, 2778, 3249.
25 Partial payment in paper currency started in 1377. The rates and later revisions
appear in MSL, Tai-tsu Shih-lu, 1748, Tai-tsung Shih-lu, 0270-1, Jen-tsung Shih-lu,
0136, Hsuan-tsung Shih-lu, 2254-5, Ying-tsung Shih-lu, 0414.
26 TMHT, 39/8-10,
27 MSL, Ying-tsung Shih-lu, 2033, 2160.
28 MSL, Hsien-tsung Shih-lu, 2136, 2218.
29 Ibid. 3583.
30 For allowances made for substituting for personal attendants, see MS, 158/1910-11;
TMHT, 157/10; MSL, Hsien-tsung Shih-lu, 3940, 4917, Hsiao-tsung Shih-lu, 3945.
For their origin, see Hucker, Censorial System, pp. 267-8; Yamane, yoeki seido,
p. 110.
31 MSL, Hsuan-tsung Shih-lu, 1342-4, Hsien-tsung Shih-lu, 4686, Hsiao-tsung Shih-lu9
0452; Hucker, Censorial System, pp. 260-2.
32 MS, 183/2148-9. Cf. Ho Liang-chim, Ssu-yu-chai, 2/99.
33 MSL, Hsien-tsung Shih-lu, 1499.
34 JCL, 3/85.
35 The expenditures were partially made in paper currency. I have estimated that, on
the basis of the capacity of the two paper currency superintendencies, during the
Yung-lo period the annual production of paper notes was only between 50 million
and 100 million kuan. The purchasing power of this amount of money was no more
than 3 million piculs of husked grain.
36 MSL, Tai-tsung Shih-lu, 2341-2.
37 Ibid. 0686, 0835-6, 0936, 0988, 1128, 1435, 1482, 1545, 2267; HYWCL, 92/1; Sun
Ch'eng-tse, Meng-yu-lu, 46/63; Chiao Hung, Hsien-cheng-lu, 59/112; Ch'eng Wen-
shih, Hai-ching Cheng-ts'e, p. 85.
38 This was indicated by an imperial decree; see MSL, Hsilan-tsung Shih-lu, 1639-40.
The decree issued by Hung-hsi in 1424 virtually apologized for previous excesses:
see MSL, Jen-tsung Shih-lu, 0015-17.
39 Chang Hsiieh-yen, Wan-li K'uai-chi-lu, 24/22.
40 Hoshi, The Ming Tribute Grain System, passim.
41 MSL, Hsuan-tsung Shih-lu, 1949; TMHT, 27/30.
336 Notes to pages 51-9
42 MSL, Hsien-tsung Shih-lu, 0556.
43 TMHT, 21/21-9; Kiangsi Fu-i CKuan-shu, 'sheng-tsung' (a general summary of
provincial quota), 6-7.
44 MSL, Hsien-tsung Shih-lu, 2315, 2378; Wu Chi-hua, Hai-yun chi Yu'n-ho, p. 127;
Hoshi, soun no kenkyu, pp. 64-8. Cf. Ch'ii T'ung-tsu, Local Government, p. 140.
45 THCKLPS, 12/95.
46 MS, 78/824; MSL, Ying-tsung Shih-lu, 0293, 0414-15, 0966; TMHT, 30/1; Horii,
'Kinkagin no tenkai', p. 47.
47 The functions of transportation commissioners under the T'ang and Sung are
described in Chili Tang-shu, ch. 49; Hsin Tang-shu, ch. 53; Sung-shih, chs. 186, 327.
The most successful of these commissioners was Liu Yen (715-80). See Twitchett,
Financial Administration, pp. 90-6.
48 A total of 47,004 full-time laborers worked on the canal system, as a part of the local
corvee service. Shantung province alone provided 14,150 men: see THCKLPS, 15/9.
49 TMHT, 27/11-15; Hsi and Chu, Ts" ao-cK uan-chih, 3/2-3.
50 76/^.1/2,4/11-15.
51 MSL, Hsien-tsung Shih-lu, 2178.
52 CSWP, 108/1-5; Huang Shun, Ming-ch'en Ching-chi-lu, 22/22.
53 TMHT, 27/40-1.
54 MSL, Hsien-tsung Shih-lu, 3578.
55 Chou Chih-lung, Ts'ao-ho I-pi, ch. 8.
56 Hsi and Chu, Ts'ao-ch'uan-chih, 3/12-14.
57 MSL, Hsiao-tsung Shih-lu, 0254; TMHT, 27/51; Hsi and Chu, Ts'ao-cKuan-chih,
3/20-1, 4/2-5, 6/44-8.
58 Ibid. 1/5-9.
59 Ibid. 7/15, 25.
60 Sun Ch'eng-tse, Meng-yu-lu, 6/58. It was also stated that each agency or bureau had
only two to five eunuchs: see MSL, Hsiao-tsung Shih-lu, 0152.
61 Hucker, Traditional Chinese State, p. 11, and Censorial System, p. 24.
62 MSL, Ying-tsung Shih-lu, 2067.
63 Cheng Hsiao, Chin-yen Lei-pien, 3/137.
64 Ting I, Te-wu Cheng-chih, pp. 22-6.
65 MSLt Hsiian-tsung Shih-lu, 0143, Ying-tsung Shih-lu, 0152, Hsiao-tsung Shih-lu,
0624.
66 MSL, Hsiian-tsung Shih-lu, 1883.
67 MSL, Hsiao-tsung Shih-lu, 3664.
68 MSL, Ying-tsung Shih-lu, 1947, Hsien-tsung Shih-lu, 1909, 3934; TMHT, chs. 205-6.
69 Based on the 1468 commutation rate of 1 tael per man per month, as indicated in
TMHT, 206/3.
70 MS, 82/863; TMHT, 194/4. Also see MSL, Ying-tsung Shih-lu, 1557, Shih-tsung
Shih-lu, 2499.
71 MSL, Ying-tsung Shih-lu, 6750.
72 MS, 197/2292; MSL, Hsiao-tsung Shih-lu, 0090-1.
73 MSL, Tai-tsu Shih-lu, 1503.
74 MSL, Shih-tsung Shih-lu, 2406.
75 MSL, Ying-tsung Shih-lu, 5417.
76 MSL, Shih-tsung Shih-lu, 2407.
77 HYWCL, 34/2.
78 MS, 79/831; TMHT, 27/5, 62-3; MSL, Hsiao-tsung Shih-lu, 3549.
79 TMHT, 39/1-7, 41/9, 13, 15.
80 In addition to the capital officials and army officers, there were 100,000 soldiers in
the capital garrison. Another 150,000 artisans served the palace agencies in rota-
tion. 40,000 soldiers from Honan and Shantung reported in two groups to the
capital garrison, in the spring and autumn. There were also several thousand
Mongols on the payroll. Though there were actually numerous vacancies in the
ranks of these forces, grain payments in general were made according to the autho-
rized strength; thus a substantial amount of the grain was intercepted by the com-
 Notes to pages 59-65 337
manding officers and eunuchs. For the evidence for this estimate see MS, 90/951;
TMHT, 289/10; MSL, Hsien-tsung Shih-lu, 2415, 4069; CSWP, 36/16.
81 MSL, Hsiao-tsung Shih-lu, 3549.
82 MS, 190/2217; MSL, Shih-tsung Shih-lu, 0202, 0258, 1153, 3768. One source
indicates that the actual annual saving was 1.68 million piculs: see HYWCL, 33/25.
83 TMHT, 17/1, 4, 8.
84 Fujii, 'dendo tokei', pp. 108-10. Wada, Shokkashi yakuchu, pp. 55-6, 60-1, 68-9;
Yamane's preface in Shimizu, tochi seidoshi, pp. 6-7.
85 MSL, Tai-tsu Shih-lu, 2218, 3166-7.
86 Shimizu, tochi seidoshi, pp. 6-7, 514-6.
87 For instance, Hou-hu-chih also indicates that 1393 total acreage was 880,462,368
mou; quoted by Wei Ch'ing-yuan in Huang-ts'e Chih-tu, pp. 74-7. One piece of
evidence in favor of Fujii's theory is that in many late Ming local gazetteers, land in
the Hung-wu period is designated as 'acreage assigned for reclamation': see
THCKLPS, 7/56; Shanghai Hsien-chih, 3/2.
88 Wei Ch'ing-yuan, Huang-ts'e Chih-tu, pp. 142-51. Cf. MSL, Shih-tsung Shih-lu,
5799.
89 Ping-ti Ho, Studies on the Population, p. 18.
90 Wei Ch'ing-yuan, Huang-ts'e Chih-tu, p. 92, pp. 115-20. Cf. MSL, Hsien-tsung
Shih-lu, 4128.
91 Wei Ch'ing-yuan, Huang-ts'e Chih-tu, p. 224.
92 Lung Wen-pin, Ming Hui-yao, 2/944; Wen-shang Hsien-chih, 4/1-2.
93 Ku'ai-chi Chih, 5/2.
94 TMHT, 19/1, 6, 12.
95 From MS, 80/837-8 and Hsu Wen-hsien Tung-k'ao, 2955-8, the annual salt pro-
duction can be estimated to have been 459,316,000 catties. MSL, Tai-tsu Shih-lu,
1433 suggests that the average annual production from 1368 to 1372 was approxi-
mately 600 million catties.
96 Ping-ti Ho, Studies on the Population, pp. 12-13.
97 Ibid. pp. 23, 277.
98 Koyama, 'daitochi shoyu', p. 64. The author emphasizes that tenant farmers
partially supported themselves by handicrafts.
99 Ku'ai-chi Chih, 5/2-3.
100 Sakuma, 'kaigai shiboeki', pp. 1-25.
101 For instance, People's University, She-hui Ching-chi, pp. 221-48.
102 P'eng Hsin-wei, Ho-pi-shih, pp. 464, 466-7.
103 HYWCL, 32/6, 16; Wei Ch'ing-yuan, Huang-ts'e Chih-tu, pp. 197-8.
104 MS, 11mi', Wei Ch'ing-yuan, Huang-ts'e Chih-tu, p. 228.
105 Mentioned in THCKLPS, 1/1; Hang-chou Fu-chih, 28/6; Shanghai Hsien-chih,
3/25-6.
106 TMHT, 18/1-2; Sun Ch'eng-tse, Meng-yu-lu, 36/2, 42/72.
107 MSL, Tai-tsung Shih-lu, 0489.
108 MS, 77/820. The entry is partially based on a decree issued in 1403: see MSL,
Tai-tsung Shih-lu, 0495.
109 Sun Ch'eng-tse, Meng-yu-lu, 36/3.
110 Ta-kao Wu-cKen, 1 /appendix 9.
111 MSL, Tai-tsu Shih-lu, 0919,1089, 1203, 1290, 1442, 2510, 2620, 2771, 2777, 2781,
2902, 2998, 3184, 3377, 3470, 3559, 3591.
112 MSL, Tai-tsung Shih-lu, 0500.
113 Ibid. 2421.
114 MSL, Hsuan-tsung Shih-lu, 1224-5.
115 Wang Yu-ch'iian, Chun-tun, pp. 104-5, 210-11.
116 In the central government, military farming came under the jurisdiction of the
bureau of state lands in the ministry of works. In later years this bureau virtually
became redundant: see Sun Ch'eng-tse, Meng-yu-lu, 46/3. The lack of overall
planning can be observed from Ta-kao Wu-cKen 1/appendix 9; Wang Yii-
ch'iian, Chun-tun, pp. 114-20, 194-9.
338 Notes to pages 65-8
117 MSL, Tai-tsu Shih-lu, 3529-30. Cf. Chiao-k" an-chi, 728 for typographical errors.
Wu Han stresses that the number of servicemen was a state secret. See ' Ming-tai ti
Chun Ping', p. 157. Also reprinted in Tu-shu Cha-chi, p. 101.
118 MSL, Tai-tsung Shih-lu, 0589, Hsiao-tsung Shih-lu, 3322.
119 MSL, Tai-tsu Shih-lu, 1881-2; Sun Ch'eng-tse, Meng-yu-lu 42/10.
120 MSL, Tai-tsu Shih-lu, 1331, 2533-4, 2735, 2788, 3192, 3225, 3264-6.
121 Ibid. 1292, 3592; Tai-tsung Shih-lu, 2172; Wei Ch'ing-yiian, Huang-ts'e Chih-tu,
p. 55.
122 Shun-te Hsien-chih, 3/21; Chin-hua Fu-chih, 21/6-7.
123 TMHT, 41/4; Ta-kao Wu-cKen, 1/appendix 20.
124 TMHT, 40/1-2.
125 The last time the paper currency was distributed as unscheduled payments to the
soldiers was in the winter of 1423-4: see MSL, Tai-tsung Shih-lu, 2426, 2475.
126 Some soldiers are said to have owned 300 to 500 mou of land: see MSL, Tai-tsu
Shih-lu, 1203, 2782; Wang Yii-ch'iian, Chun-tun, p. 72. It seems that some of the
hereditary military families were originally landowners and it is possible that some
of them acquired landed property outside the military farming system.
127 MSL, Tai-tsung Shih-lu, 1478; HYWCL, 91/2; Wang Yu-ch'iian, Chun-tun, p. 191;
Shimizu, 'Chun-tun ti Peng-huai', pp. 37-8.
128 Shun-te Hsien-chih, 3/12, indicates that every soldier received 20 mou. For other
similar cases, see Wang Yu-ch'iian, Chun-tun, pp. 63-7, 70, 190.
129 MSL, Tai-tsung Shih-lu, 0495-6; TMHT, 18/12.
130 Military farming was ordered by Hung-wu prior to the founding of the dynasty,
and was repeatedly emphasized thereafter in decrees: see MSL, Tai-tsu Shih-lu,
1203, 2782, 2902, 2910, 3104.
131 For the supply procedure see Ta-Ming Kuan-chih, 4/2452; Ta-kao, 1/16-18;
MSL, Tai-tsu Shih-lu, 2791, 2998. Supplies to Liao-tung were exceptional.
132 MSL, Tai-tsung Shih-lu, 0927.
133 MS, 77/820; MSL, Jen-tsung Shih-lu, 0214; Sun Ch'eng-tse, Meng-yu-lu, 36/3.
Yung-lo might have temporarily reduced the production quota at an earlier stage:
see TMHT, 18/13.
134 MSL, Ying-tsung Shih-lu, 0014; TMHT, 18/13.
135 See Shun-te Hsien-chih, 3/14; THCKLPS, 13/71, 26/106-7; Sun Ch'eng-tse,
Meng-yii-lu, 36/4; Wei Huan, Chiu-pien-k'>ao, 1/25.
136 TMHT, 41/3-18.
137 MSL, Hsien-tsung Shih-lu, 1166.
138 MSL, Hsiao-tsung Shih-lu, 0579.
139 TMHT, 41/17.
140 MSL, Shih-tsung Shih-lu, 2066.
141 For 'troop purifying censors', see Hucker, 'Governmental Organization', p. 51
and Censorial System, pp. 75-7. An actual case is described by Lu Jung, in Shu-
yuan Tsa-chi, 1/11.
142 Ni Hui-ting, Nien-p'u, 4/22; Shun-te Hsien-chih, 3/4; Wang Yii-ch'uan, Chun-tun,
pp. 244-7.
143 MSL, Hsiao-tsung Shih-lu, 1261.
144 Ibid. 3424, cf. Chiao-k'an-chi, 587 for errors.
145 Chin-hua Fu-chih, 21/5.
146 MSL, Hsien-tsung Shih-lu, 4069, Hsiao-tsung Shih-lu, 0809, 1899, 4058.
147 In 1487, on the accession of Hung-chih, each soldier on frontier duty was awarded
2 taels of silver by the emperor, a total of 615,320 taels of silver being thus dis-
tributed: see MSL, Hsiao-tsung Shih-lu, 0095. For frontier troop dispositions see
also MSL, Hsien-tsung Shih-lu, 2109, 3908-9.
148 For early recruiting, see MSL, Hsiao-tsung Shih-lu, 3418, 3447, 3682; CSWP, 90/7,
99/7.
149 MS, 91/956. Cf. Wu Han, 'Ming-tai ti Chun Ping', p. 221.
150 MSL, Hsiao-tsung Shih-lu, 3554; Sun Ch'eng-tse, Meng-yu-lu, 35/14, 18. The
actual amount is quoted as 483,132 taels.
 Notes to pages 68-75 339
151 MSL, Hsiao-tsung Shih-lu, 3554.
152 Ni Yiian-lu, CKuan-chi, Memorials, 11/6-7.
153 Chu Hsieh, Hsin-yung Ho-pi, pp. 156-7; P'eng Hsin-wei, Ho-pi-shih, pp. 432-3;
Li Chien-nung, Ching-chi Shih-kao, p. 102; Lien-sheng Yang, Money and Credit,
p. 67.
154 These sixty-nine entries are to be found in MSL, Tai-tsu Shih-lu, 2981-3078.
155 Ibid. 3079.
156 Ibid. 3062.
157 Ta-kao Hsu-pien, 1/143-5.
158 Based on MSL, Tai-tsu Shih-lu, 2458, 2926.
159 MS, 81/849; MSL, Tai-tsung Shih-lu, 0509, 0589-90. Cf. Wada, Shokkashi
yakuchu, p. 608.
160 TMHT, 31/582. From the increased payments made to soldiers and laborers it can
be observed that the actual value of the paper notes was much less. See MSL,
Tai-tsung Shih-lu, 0623, 0691, 0836, 1657, 1681, 1723.
161 In 1425, a picul of husked rice could be purchased with 40 to 70 kuan of paper
currency. See MSL, Hsiian-tsung Shih-lu, 0175.
162 Ibid. 1313-14.
163 Ibid. 0493, 1171.
164 Ibid. 0095.
165 For franchise fees and other taxes payable in paper currency, see MS, 81/849;
TMHT, 31/4-6, 35/2-6; MSL, Jen-tsung Shih-lu, 0219, Hsiian-tsung Shih-lu,
1324-6.
166 Ibid. 1977.
167 Ibid. 2406.
168 For tax reductions 1433 see ibid. 2305-6; TMHT, 31/6-7.
169 MSL, Hsuan-tsung Shih-lu, 0151-3.
170 MS, 81/849; Feng Hsin-wei, Ho-pi-shih, p. 422.
171 MSL, Ying-tsung Shih-lu, 0224. The account in MS, 81/849 is somewhat distorted.
Cf. Wada, Shokkashi yakuchu, pp. 721-2.
172 MSL, Ying-tsung Shih-lu, 0293.
173 Ibid. 2723.
174 Ibid. 3209. The order was rescinded by Ching-t'ai in 1453: see TMHT, 31/7.
175 MSL, Hsien-tsung Shih-lu, 0533; Shen Pang, Wan-shu Tsa-chi, 57.
176 TMHT, 35/44.
177 For taxes paid in paper currency, see TMHT, 35/8, 47; MSL, Hsiao-tsung Shih-lu,
0093, 1104, 1690. On dealers buying and selling paper notes see MSL, Hsien-tsung
Shih-lu, 2471, 2680, 2971.
178 By the mid sixteenth century 1 kuan notes were counted i n ' blocks' of 1,000, which
were worth between 4.6 and 10 taels of silver to the tax collector: see MSL,
Shih-tsung Shih-lu, 1634; P'eng Hsin-wei, Ho-pi-shih, p. 441.
179 Li Chien-nung, Ching-chi Shih-kao, p. 103.
180 P'eng Hsin-wei, Ho-pi-shih, p. 441.
181 MSL, Hsi-tsung Shih-lu, 0051.
182 Ho Shih-chin, CKang-Vu Hsu-chih, 1/19.
183 TMHT, 31/8.
184 MSL, Hsi-tsung Shih-lu, 2342, 3142, 3355.
185 CSWP, 244/16.
186 MSL, Wu-tsung Shih-lu, 0081.
187 The official rates were 800:1 in 1481, 700:1 in 1527, and 800:1 in 1567: see TMHT,
31/10, 12, 13.
188 On the process of coining money, see MS, 81 /850; MSL, Wu-tsung Shih-lu, 0080-2;
Sun Ch'eng-tse, Meng-yu-lu, 38/12; Wada, Shokkashi yakuchu, pp. 748-9; Sung
Ying-hsing, Tien-kung K'ai-wu, 158-9, 166-8.
189 HYWCL, 92/18.
190 MS, 81/850; CSWP, 244/16.
191 P'eng Hsin-wei, Ho-pi-shih, pp. 425, 437.
340 Notes to pages 75-9
192 Ibid. p. 425.
193 MSX, Tai-tsu Shih-lu, 1419, 1617.
194 This productive capacity is calculated from the list in Chu-ssu Chih-chang (kung-
yii-pu), 24-5. The provincial mints had an annual capacity of 177,867,800 coins,
while the Nanking production quota was 12.8 million coins. P'eng Hsin-wei esti-
mates the annual capacity as 166,090,400 coins: see Ho-pi-shih, pp. 438-9.
195 From 981 to 1080 the Sung on average produced 2 to 3 billion coins annually. The
lowest level of output was 500 million copper cash in 981 and the highest 5,060
million copper cash in 1086: see Wang Chih-jui, Ching-chi-shih, p. 85; Peng
Hsin-wei, Ho-pi-shih, p. 281.
196 Imperial edicts mention the production of copper coins in connection with these
voyages: see MSX, Tai-tsung Shih-lu, 2267, and Jen-tsung Shih-lu, 0015-17. That
copper coins were exported by Cheng Ho is suggested by Liang Fang-chung in
'Kuo-chi Mou-i', pp. 281-2, and confirmed by P'eng Hsin-wei and Ch'en Wen-
shih: see Ho-pi-shi, pp. 442-3, and Hai-ching Cheng-ts'e, pp. 84, 86.
197 MSX, Ying-tsung Shih-lu, 5141; Ch'en Wen-shih, Hai-ching Cheng-ts'e, p. 60.
198 MSX, Hsien-tsung Shih-lu, 3663, 3689.
199 According to MSL, Hsiao-tsung Shih-lu, 3622 and Chu-ssu Chih-chang (kung-yii-
pu), 25, the annual productive capacity in Peking was only 12,830,400 coins.
200 MSL, Hsiao-tsung Shih-lu, 3622, 3644-7. Though the emperor ordered that certain
revenues be intercepted to subsidize the minting of cash (ibid. 3657) it appears that
this order was never carried out.
201 Ibid. 4005.
202 MSX, Wu-tsung Shih-lu, 0081.
203 This estimate is based on MSX, Hsiao-tsung Shih-lu, 3622.
204 Ibid. 4241.
205 MSL, Wu-tsung Shih-lu, 1130. P'eng Hsin-wei suggests that after 1505, the project
was suspended: see Ho-pi-shih, p. 425.
206 MSX, Shih-tsung Shih-lu, 1855: TMHT, 31/11.
207 MS, 81/850; MSL, Shih-tsung Shih-lu, 7119-21; Ko Shou-li, Ko-tuan-su-kung Chi
2/3.
208 Li Chien-nung, Ching-chi Shih-kao, p. 103.
209 MS, 81/849; TMHT, 194/9. The information can be traced to MSX, Shih-tsung
Shih-lu, 7063. The entry in the Shih-lu, however, does not make clear how many
coins were to be minted.
210 P'eng Hsin-wei, Ho-pi-shih, pp. 426, 444.
211 TMHT, 31/11, 194/9.
212 MS, 81/850; MSL, Shih-tsung Shih-lu, 7297-8.
213 In 1558, for instance, 61.75 million coins were minted: see ibid. 7789, 8819.
214 CSWP, 244/17. On the suspension of production in Nanking see Fu I-ling,
Shih-min Ching-chi, p. 28.
215 MS, 81/850. The memorial is reproduced in CSWP, 244/16-18 and dated 1564.
See also Wada, Shokkashi yakuchu, p. 748.
216 The imperial orders were not always consistent: see TMHT, 31/10; MSX, Wu-tsung
Shih-lu, 1585, Shih-tsung Shih-lu, 7059, 7119.
217 MSX, Mu-tsung Shih-lu, 1113-14, Shen-tsung Shih-lu, 1141, 1802-3.
218 Shen Pang, Wan-shu Tsa-chi, 56-8.
219 For example, in 1571 and again in 1574, small quantities of coins were minted: see
MSX, Mu-tsung Shih-lu, 1519, Shen-tsung Shih-lu, 1293.
220 TMHT, 31/14; MSX, Shen-tsung Shih-lu, 1130-1. The operation was largely
suspended in 1580, ibid. 1802-3.
221 P'eng Hsin-wei, Ho-pi-shih, p. 445.
222 Yang Tuan-liu, Ho-pi Chin-jung, pp. 71-2. Silver taels were actually minted in
Tibet. Also see P'eng Hsin-wei, Ho-pi-shih, p. 508. Taels had also been minted by
the Jurched Chin in 1197, ibid. p. 364.
223 Liang Fang-chung, 'Kuo-chi Mou-i', pp. 267-324, Liang-chang Chih-tu, p. 127.
224 P'eng Hsin-wei, Ho-pi-shih, pp. 461, 471.
 Notes to pages 80-96 341
225 Ku Yen-wu, Wen-chi, 1/13-14.
226 Basis of estimate: Salt revenue, 1 million; Gold Floral Silver, 1 million; tax deli-
veries by the northern provinces to the frontier, 2.5 million; miscellaneous, 0.5
million. See Tables 14, 23 and 26.
227 See Sakuma, 'Keitokuchin yogyo', p. 483. Nishijima, 'mengyo shijo', p. 274;
Fu I-ling, Shin-min Ching-chi, pp. 16-17. Also see Ku Yen-wu, Yu-chi, 13; Wang
Shih-ch'i, San-yun Chou-tsu-k'ao, ch. 2.
228 Ku Yen-wu, Wen-chi, 1/13.
229 HYWCL, 92/21.
230 Chou Hsiian-wei, Chitig-lin Hsu-chi, 2, 5.
231 Feng Hsin-wei, Ho-pi-shih, pp. 459-60.

3 The land tax—(/) Tax structure

1 Shun-te Hsien-chih, 3/1.
2 MS, 45/507; Shun-te Hsien-chih, 3/6-8.
3 Ibid. 3/9.
4 Ibid. 3/15.
5 MS, 223/2574; HYWCL, 32/9; Chiao Hung, Hsien-cheng-lu, 59/95; MSL, Shih-
tsung Shih-lu, 8181-2, Shun-te Hsien-chih, 3/23-4.
6 Ibid. 3/24.
7 Ibid. 3/26-7.
8 Ibid. 3/19-21.
9 Ibid. 3/31-2.
10 /ta/. 3/30-1.
11 /fa/. 3/21.
12 Ibid. 3/15.
13 /fa/. 3/19.
14 Chou Hsiian-wei, Ching-lin Hsii-chi, 47. The author also observes that grain prices
in that region were the lowest possible.
15 THCKLPS, 6/65.
16 Chang Chi-shun, et al. ed. CKing-shih, II, p. 1464.
17 Ho Liang-chim, Ssu-yu-chai, 3/179.
18 Ibid. 3/190.
19 HYWCL, 32/24.
20 Fu I-ling, Nung-ts'un She-hui, pp. 24-5. Cases occurring in the early Ch'ing are
included.
21 Wei Ch'ing-yiian, Huang-ts'e Chih-tu: the illustrations appear on pp. i-iv.
22 Hui-chou Fu-chih, 7/19-45; Ku'ai-chi Chih, 5/8-9.
23 Shun-te Hsien-chih, 3/27.
24 At worst there were twelve tax deadlines each month, although not all deadlines
were applicable to all taxpayers: see Ho Liang-chim, Ssu-yu-chai, 3/167, 173.
25 Ch'ang-shu county abandoned the distinction between government and private land
in 1462, Hang-chou prefecture in 1572: see Ch'ang-shu Hsien-chih, 2/33; Hang-chou
Fu-chih, 29/19.
26 An-hua Hsien-chih, 2/8; Chin-hua Fu-chih, 6/29; Chang-chou Fu-chih, 27/15.
27 Shun-te Hsien-chih, 3/9. The gazetteer indicates that the district followed the pre-
cedent of Kiangsi province.
28 Kiangsi Fu-i CKuan-shu, provincial summary, 1-2.
29 Shun-te Hsien-chih, 3/15.
30 Shanghai Hsien-chih, 3/12.
31 This estimate is based on the unedited figures in the local gazetteer. It is clear that
only 9,780 piculs of grain out of a tax quota of 45,279 piculs were paid in kind: see
Ku'ai-chi Chih, 5/1-6.
32 I have calculated that only 2,780 piculs out of a tax quota of 57,213 piculs was paid
in kind. See Lin-fen Hsien-chih, A12-5.
33 That is, an official was not permitted to serve in his own district nor, with few
342 Notes to pages 96-105
exceptions, in his native province. There were exceptions in the case of Yunnan and
Kwangsi; see TMHT, 5/14-15; also Parsons, 'The Ming Bureaucracy: Aspects of
Background Forces', pp. 175-231.
34 JCL, 3/85.
35 Kifai-chi Chih, 5/13-18.
36 Ibid. 5/17-18.
37 MSL, Shen-tsung Shih-lu, 2953-4; HYWCL, 32/27; Chang Chii-cheng, Shu-tu,
4/1.
38 Wenshang Hsien-chih, 4/4.
39 This lack of initiative on the part of local officials was criticized by Ku Yen-wu, who
considered it an inevitable result of the political system. Ku's essays are translated
and summarized in de Bary, ed. Sources of Chinese Tradition, II, pp. 611-12.
40 MS, 78/825; HYWCL, 33/7.
41 Shimizu, tochi seidoshi, pp. 460-2.
42 JCL, 4/53-6; Shanghai Hsien-chih, 3/1. The forced purchase seems to have started
in 1263.
43 MS, 71/818, 78/824; JCL, 4/53.
44 CKang-shu Hsien-chih, 2/31.
45 THCKLPS, 6/74, 7/4, 8/32; JCL, 4/53. Cf. Chou Liang-hsiao, 'Su-Sung Ti-ch'U\
pp. 65-6.
46 JCL, 4/53; Wada, Shokkashi yakuchu, p. 187.
47 JCL, 4/53.
48 THCKLPS, 6/13.
49 JCL, 4/53.
50 TMHT, 27/25-6, 28.
51 TMHT, 27/61-2; THCKLPS, 12/13.
52 TMHT, 42/40-4; THCKLPS, 6/41.
53 TMHT, 26/13-15; THCKLPS, 6/47.
54 The local grain price is assumed to have been 0.6 taels of silver for each picul of
husked rice. Other references are: THCKLPS, 6/12, 41, 67-9, 83-4, 7/32-3, 12/95;
Ming-ch'en Tsou-i, 34/656; CSWP, 438/18-19; Nishijima, 'mengyo shijo', p. 274.
55 Ku-su Chih, 15/6.
56 MS, 153/1863-4; Chiao Hung, Hsien-cheng-lu, 60/5-10; THCKLPS, 6/41, 69, 7/53;
Shanghai Hsien-chih, 3/5-6; CKang-shu Hsien-chih, 3/50; K'un-shan Hsien-chih,
2/21-2.
57 MSL, Ying-tsung Shih-lu, 2349-50, 2727, 4220, 4453, 4478. Cf. Lu Jung, Shu-yuan
Tsa-chi, 5/54.
58 Shanghai Hsien-chih, 3/18.
59 Wu-hsien Chih, 8/1, 3.
60 CSWP, 438/18-19; THCKLPS, 6/50, 79, 12/13; HYWCL, 32/18.
61 MSL, Jen-tsung Shih-lu, 0194. The rate is calculated from the data in MSL, Tai-
tsung Shih-lu, 2182, 2245, 2301, 2364, 2421. Errors have been corrected on the basis
of Chiao-k'an-chi, 111.
62 TMHT, 151/6-13.
63 MSL, Hsien-tsung Shih-lu, 0655, Hsiao-tsung Shih-lu, 1498; Yang Shih-ch'iao,
Ma-cheng-chi, 8/2.
64 MS, 92/968; TMHT, 152/7; Sun Ch'eng-tse, Meng-yu-lu, 53/1-3; Lu Jung, Shu-yuan
Tsa-chi, 4/41.
65 MSL, Hsiao-tsung Shih-lu, 0092; Sun Ch'eng-tse, Meng-yu-lu, 53/5.
66 Yang Shih-ch'iao, Ma-cheng-chi, 8/1; Sun Ch'eng-tse, Meng-yu-lu, 53/5.
67 TMHT, 152/4.
68 TMHT, 152/4-5; Sun Ch'eng-tse, Meng-yu-lu, 53/4.
69 Lu Jung, Shu-yuan Tsa-chih, 4/41.
70 MSL, Hsien-tsung Shih-lu, 1070.
71 Shen Pang, Wan-shu Tsa-chi, 44, 68.
72 MS, 92/970; Yang Shih-ch'iao, Ma-cheng-chi, 8/4; Sun Ch'eng-tse, Meng-yu-lu, 53/4.
73 TMHT, 152/6-7; Shen Pang, Wan-shu Tsa-chi, 71.
 Notes to pages 106-15 343
74 TMHT, 150/15-18; Shen Pang, Wan-shu Tsa-chi, 68-71; Hsiang-ho Hsien-chih, 5/2;
Ku-an Hsien-chih, 3/15.
75 See Shimizu, tochi seidoshi, pp. 16-90, 91-115.
76 TMHT, 17/16-17; THCKLPS, 1/61; MSL, Tai-tsu Shih-lu, 3532.
77 MS, 157/1904; MSL, Ying-tsung Shih-tu, 5488-9.
78 MS, 77/821.
79 MSZ, Ying-tsung Shih-lu, 1876-7.
80 These cases can be found in MSL, Ying-tsung Shih-lu, 1518; Hsien-tsung Shih-lu,
0402, 1065, 3561, 3678, 3708, Hsiao-tsung Shih-lu, 0629, 3924.
81 THCKLPS, 1/61; MSL, Wu-tsung Shih-lu, 1531.
82 MSL, Hsiao-tsung Shih-lu, 0629.
83 MS, 300/3366-7; MSL, Shih-tsung Shih-lu, 2369-70. It seems, however, that Chia-
ching also had personal grudges against Chang.
84 These regulations appear in TMHT, \1125-6. Frequent violations were reported:
see Shimizu, tochi seidoshi, pp. 87-90.
85 Hsiang-ho Hsien-chih, 4/2, 13.
86 Tung-cKangFu-chih, 11/4,6,8,10,13,15,17,19,26, 31, 33. See also MSL, Wu-tsung
Shih-lu, 0468, Shen-tsung Shih-lu, 4557.
87 THCKLPS, 13/51.
88 JCL, 3/63-4; THCKLPS, 13/71.
89 Honan Tung-chih, 12/6.
90 THCKLPS, 15/164-5.
91 Hai Jui, Hai Jui Chi, 73.
92 Yung-chou Fu-chih, 9/3.
93 Hang-chou Fu-chih, 7/35, 29/9-10, 31/17.
94 Chang-chou Fu-chih, 5/51-3; THCKLPS, 16/69, 86-8, 90-3; MSL, Shen-tsung
Shih-lu, 5269-70.
95 MSL, Ying-tsung Shih-lu, 6032.
96 Quoted by Yamane, in yoeki seido, p. 66.
97 MSL, Hsien-tsung Shih-lu, 0650.
98 Yamane, yoeki seido, p. 105; Friese, Dienstleistungs-System, p. 97.
99 MSL, Ying-tsung Shih-lu, 2425, 4202, 2975, 6031-2.
100 TMHT, 20/11.
101 For the origin of the militia, see Liang Fang-chung, 'Ming-tai ti Ming-ping',
p. 221. Cf. MS, 91/959. Militiamen should not be confused with local patrolmen,
who came under the category of chiin-yao.
102 MSL, Hsiao-tsung Shih-lu, 1702.
103 Hui-chou Fu-chih, 8/4.
104 Liang Fang-chung, 'Kuo-chi Mou-i', pp. 278-9, 281-2.
105 MSL, Hsiao-tsung Shih-lu, 0795; Su T'ung-pin, I-ti Chih-tu, p. 277.
106 MSL, Wu-tsung Shih-lu, 0700.
107 Chang-chou Fu-chih, 5/38.
108 Su T'ung-pin, I-ti Chih-tu, p. 287; Yamane, yoeki seido, p. 169.
109 Hang-chou Fu-chih, 31/15.
110 Su T'ung-pin, I-ti Chih-tu, p. 320.
111 Liang Fang-chung, 'Ming-tai ti Min-ping', p. 227.
112 THCKLPS, 9/48; Liang Fang-chung, 'Ming-tai ti Min-ping', p. 227.
113 Wen-shang Hsien-chih, 15/170.
114 THCKLPS, 16/59.
115 Feng Ch'i, Feng-tsung-po Chi, 51/34; Sun Ch'eng-tse, Meng-yii-lu, 35/21.
116 In 1522 eight types of unauthorized surcharge, each with its fixed rate, were added
to the tribute grain; they were payable by the army officers serving with the trans-
portation corps. See MSL, Shih-tsung Shih-lu, 0440-1. Such practices were quite
common at the time.
117 Fang-chung Liang, The Single-whip Method, p. 7.
118 Shanghai Htien-chih, 4/32.
119 Shun-te Hsien-chih, 3/24; HYWCL, 32/9.
344 Notes to pages 115-24
120 MS, 78/826. Note 5 added by Wang Yu-ch'iian on p. 42 of Fang-chung Liang,
The Single-whip Method is quite mistaken. The kang-yin in this context had nothing
to do with the salt revenue.
121 Fang-chung Liang, The Single-whip Method, pp. 56-60.
122 Such as Wang Shu and P'ang Shang-p'eng. See CSWP, 357/6-7; Wu-chiang
Hsien-chih, 10/11-12; Wu-hsien Chih, 7/7-8.
123 Cf. Fang-chung Liang, The Single-whip Method, p. 6. Note that Liang defines
H-cKai as service performed locally and yin-cKai as service performed away from
home. Though there is some evidence for this definition it is inappropriate in
a sixteenth-century context. See also THCKLPS, 6/18; Hai Jui, Hai Jui Chi, 249.
124 See Liang, 'The "Ten-Parts" Tax System of Ming', in Sun and de Francis,
Chinese Social History, pp. 271-80; also, THCKLPS, 7/9-17; TMHT, 20/14;
Hui-chou Fu-chih, 8/42-3; Hang-chou Fu-chih, 31/4.
125 Wada, Shokkathi yakuchu, p. 217, based on MSL, Shih-tsung Shih-lu, 2971. The
term had been used for some time without being precisely defined. Kuribayashi
considers that the Single Whip began to be developed fully around 1560: see
'Ichijobenho', pp. 115-30.
126 HYWCL, 32/27. This statement has been slightly rearranged here.
127 Fang-chung Liang, The Single-whip Method, p. 1.
128 J. K. Fairbank, in East Asia: The Great Tradition, co-authored with Edwin
O. Reischauer (Boston, 1958), p. 340.
129 Shen Pang, Wan-shu Tsa-chi, 49-50.
130 See the article by Feuerwerker, 'From "Feudalism" to "Capitalism"', pp. 107-8.
131 Ni Hui-ting, Nien-p'u, 4/8.
132 Hai Jui, Hai Jui Chi, 269; Hang-chou Fu-chih, 31/14.
133 Ku'ai-chi Chih, 1/5.
134 For the official standard see Shun-te Hsien-chih, 3/27; THCKLPS, 27/53. Cf.
TMHT, 20/15.
135 Hui-chou Fu-chih, 8/38.
136 THCKLPS, 22/53.
137 It was practised in Hang-chou prefecture in 1579: see Hang-chou Fu-chih, 31/14.
138 Chung-mou Hsien-chih, 2/15, 17-18.
139 Hsiang-ho Hsien-chih, 4/2-3, 19-22.
140 THCKLPS, 15/175-6; Yamane, 'ekiho no tokushitsu', pp. 221-50.
141 One of its most ardent opponents was Ko Shou-li, who served as minister of
revenue for six months in 1567: see Ko-tuan-su-kung Chi, 2/10, 3/19, 14/15, 15/11,
15/18, chia-shiln hsia, 27. Cf. Kataoka, 'tochi shoyu', pp. 148-9.
142 MSL, Mu-tsung Shih-lu, 1200-1; Shen-tsung Shih-lu, 1095, 1100, 1112, 1245, 1338,
4124.
143 Lu-cKeng Hsien-chih, 3/46; Hsiang-ho Hsien-chih, All', Huai-jou Hsien-chih, 2/6;
THCKLPS, 15/175-6.
144 In most districts some items on the service levy lists were permanently fixed, while
others could be periodically adjusted. In theory the former could not be mixed
with the latter, nor the latter items with each other. See Fang-chung Liang, The
Single-whip Method, p. 32.
145 Ku'ai-chi Chih, 7/12-13. The form is mentioned by Fang-chung Liang, in The
Single-whip Method, pp. 60-1.
146 MSL, Shen-tsung Shih-lu, 3755.
147 MS, 78/828.
148 THCKLPS, 6/61.
149 MSL, Shen-tsung Shih-lu, 4124.
150 The exemption schedules appear in TMHT, 20/19; THCKLPS, 7/20-1; Hai Jui,
Hai Jui Chi, 141-2.
151 HYWCL, 32/21.
152 THCKLPS, 7/20.
153 I-chou Chih, 3/2.
154 Shanghai Hsien-chih, 3/25.
 Notes to pages 124-37 345
155 There were 12,084 salt-producing ting in the county eligible for some form of
service exemption: Shanghai Hsien-chih, 4/8. For salt-producing ting affected by
this tax ruling, see TMHT, 20/18.
156 Shanghai Hsien-chih, 3/24-6.
157 The shih-cheng-ts'e is occasionally mentioned by local gazetteers. See, for example,
Hang-chou Fu-chih, 28/6; Shanghai Hsien-chih, 3/25-6.
158 Shanghai Hsien-chih, 3/24-5.
159 Ping-ti Ho, Studies on the Population, p. 29.
160 Wu-chiang Hsien-chih, 10/11-12; CSWP, 397/13.
161 Ko Shou-li, Ko-tuan-su-kung Chi, chia-shun-hsia, 27.
162 Chang-chou Fu-chih, 5/19-50; THCKLPS, 26/85.
163 Ibid. 21121 \ K'ai-hua Hsien-chih, 3/1-11; Shun-te Hsien-chih, 3/1-21; Sui-an
Hsien-chih, 1/48-65; Hui-chou Fu-i CKuan-shu, Hsi-hsien, 1-3, Hsiu-ning Hsien,
1-10.
164 THCKLPS, 26/63.
165 Ibid. 26/8$.
166 The extent of the variation is commented on by Wang Shih-hsing, as quoted by
Ku Yen-wu: see JCL, 3/62-3.
167 Hang-chou Fu-chih, 29/1, 31/70.
168 Yung-chou Fu-chih, 9/22.
169 Yueh-chou Fu-chih, passim.
170 Yung-chou Fu-chih, 9/11.
171 Kiangsi Fu-i CKuan-shu, provincial summary, 1-55.
172 Feng-tse Hsien-chih, 3/10, 11, 12.
173 As an example, for some palace ceremonial functions Wan-p'ing county was
ordered to provide as many as 1,000 female sedan-chair bearers: see Shen Pang,
Wan-shu Tsa-chi, 125.
174 I-chou Chih, 3/1-32; Hsiang-ho Hsien-chih, 4/1-3.
175 The data appear in Fen-chou Fu-chih, 5/3-46.
176 The quotations are taken from People's University, She-hui Ching-chi, p. 170.
177 Twitchett, Financial Administration, p. 24.
178 Wen-shang Hsien-chih, 4/14.
179 I-chou Chih, 3/25.
180 THCKLPS, 5/180-2; Ping-ti Ho, Studies on the Population, p. 31.
181 Huai-jou Hsien-chih, 2/6.
182 Ho T'ang's statement appears in THCKLPS, 13/112.
183 CSWP, 278/8, 11.
184 The ting payment, at 0.019 taels of silver per person, produced only 1,155.84 taels
from 60,834 ting: see Shanghai Hsien-chih, 3/24-5.
185 Ku'ai-chi Chih, 6/3.
186 Chang-chou Fu-chih, 5/51; THCKLPS, 16/51.
187 For fi-pien see MS, 78/826. It was also imposed in Honan, Shantung and North
Chihli: see MSL, Shih-tsung Shih-lu, 151 A, 7859-60, 8703.
188 Ho Liang-chiin, Ssu-yu-chai, 3/167, 173.
189 Chang-chou Fu-chih, 27/28-9.
190 MS, 78/823; MSL, Tai-tsu Shih-lu, 0231, 0541; Shimizu, shakai keizaishi, p.
53.
191 TMHT, 17/41.
192 Sui-an Hsien-chih, 1/49-54.
193 Hang-chou Fu-chih, 30/17.
194 Wen-shang Hsien-chih, 21A.
195 Chung-mou Hsien-chih, 2/16.
196 Hui-chou Fu-chih, 1/63-4.
197 Hui-chou Fu-chih, 7/55.
198 TMHT, 25/3.
199 The collection started in 1370, according to MSL, Tai-tsu Shih-lu, 1067. For the
regional quotas see TMHT, 25/34-8.
346 Notes to pages 137-47
200 Usually each bundle was commuted to a payment of 0.3 taels of silver. In south
China this was sometimes the price of a picul of husked rice. See note 14 above.
201 Liang Fang-chung, 'Liang-shui Shui-mo', pp. 57-61.
202 TMHT, 207/3-15; Hui-chou Fu-chih, 8/11.
203 TMHT, 289/10.
204 MS, 81/899; MSL, Tai-tsung Shih-lu, 0589-90.
205 It was 0.018 taels per person in Shun-te county and 0.0028 taels per person in
Huai-an prefecture: see Shun-te Hsien-chih, 3/18-19; Huai-an Fu-chih (1627),
12/16-17.
206 Shun-te Hsien-chih, 3/18-19.
207 Tung-cKang Fu-chih, 11/5.
208 Wang Chih-jui, Ching-chi-shih, p. 141.
209 Yung-chou Fu-chih, 9/11.
210 MS, 153/1864; Chu T'ing-li, Yen-cheng-chih, 7/69; THCKLPS, 6/120.
211 Shanghai Hsien-chih, 4/19.
212 CSWP, 357/1-6.

4 The land tax—(//) Tax administration

1 Fang-chung Liang, The Single-whip Method, pp. 52-3.
2 THCKLPS, 17/70.
3 Wen-shang Hsien-chih, 8/'61.
4 THCKLPS, 6/16.
5 TMHT, 27/35.
6 THCKLPS, 7/21; Hang-chou Fu-chih, 28/6; Shanghai Hsien-chih, 3/25-6; MSL,
Mu-tsung Shih-lu, 0355.
7 The exact date of origin of the comprehensive books is uncertain. Pi Tzu-yen, writing
in 1628, declared that their compilation had been going on for forty-five years. See
Sun Ch'eng-tse, Meng-yu-lu, 35/24. Thus it may have started somewhat earlier than
1583.
8 It was practised in Ku'ai-chi county, Chekiang, in the 1560s. See Ku'ai-chi Chi,
5/4-11.
9 Hui-chou Fu-chih, 8/20-1.
10 For these tax bills, see TMHT, 20/13-14; THCKLPS, 6/64, 66; Hai Jui, Hai Jui
Chi, 72; Ku'ai-chi Chih, 7/12-13.
11 HYWCL, 33/14.
12 TMHT, 20/14.
13 JCL, 3/79-80; Ho Liang-chun, Ssu-yu-chai, 3/184.
14 CSWP, 397/1.
15 THCKLPS, 16/34.
16 Ibid.S/56.
17 Little is known of the stone tablets but they were frequently mentioned. See Fang-
chung Liang, The Single-whip Method, p. 64; K'ai-hua Hsien-chih, 3/6.
18 Fang-chung Liang, The Single-whip Method, p. 59.
19 Ku'ai-chi Chih, 7/6-12; Huai-an Fu-chih (1573), 4/13.
20 MSL, Mu-tsung Shih-lu, 1039; HYWCL, 32/19; Yeh Meng-chu, Yueh-shih-pien, 6/1.
21 MSL, Mu-tsung Shih-lu, 1169; CSWP, 278/10.
22 Kuei Yu-kuang, Ch'uan-chi, 218.
23 Ku Yen-wu, Wen-chi, 1/15-16.
24 Chiao Hung, Hsien-cheng-lu, 17/94.
25 HYWCL, 33/17.
26 THCKLPS, 6/92, 17/70.
27 Yeh Meng-chu, Yueh-shi-pien, 6/1; Lung Wen-pin, Ming Hui-yao, 2/1018-20.
28 Ko Shou-li, Ko-tuan-su-kung Chi, 2/18.
29 THCKLPS, 6/89.
30 Ibid. 7/10-11, 18, 34-7.
31 Shanghai Hsien-chih, 4/4; THCKLPS, 6/75-6.
 Notes to pages 148-55 347
32 Shanghai Hsien-chih, 4/4; THCKLPS, 6/87. Cf. CSWP, 438/20.
33 Shanghai Hsien-chih, 4/5.
34 The statement is made by Liang Fang-chung in The Single-whip Method, p. 59.
35 CSWP, 438/20.
36 K'ai-hua Hsien-chih, 3/35-8.
37 THCKLPS, 6/89-90.
38 Ibid. 6/S1.
39 Shanghai Hsien-chih, 4/4-5; THCKLPS, 6/74, 90.
40 Shanghai Hsien-chih, 4/6.
41 THCKLPS, 6/83, 85.
42 /6/tf. 7/23, 34, 36-7.
43 Ibid. 22/10-11.
44 /W</. 15/152, 169, 178, 16/34.
45 Tung-cKang Fu-chih, 12/42-3.
46 Wang Ying-chiao, Tsou-su, 9/5.
47 Hsiang-ho Hsien-chih, A119.
48 Shanghai Hsien-chih, 4/6; Hang-chou Fu-chih, 28/6; Hsiang-ho Hsien-chih, 4/19;
CSPfT, 397/8; THCKLPS, 6/76, 91, 7/34, 9/49, 16/47.
49 THCKLPS, 6/75.
50 Ibid. 6/16.
51 CSHT, 503/1-8.
52 THCKLPS, 6/75.
53 Ibid. 7/11.
54 Ibid. 6/76-1.
55 MSX, Shen-tsung Shih-lu, 1211; C5H^P, 325/8.
56 Ku Yen-wu, tffc/i-c/w, 1/15-16.
57 C S ^ P , 397/19; Wen-shang Hsien-chih, 8/61; THCKLPS, 18/22, 22/31.
58 Yeh Meng-chu, Yiieh-shih-pien, 6/4.
59 Ku Yen-wu, Wen-chi, 1/15.
60 Kuei Yu-kuang, CKuan-chi, 217.
61 Hai Jui, # 0 / /«/ C/w, 49, 179-80.
62 HYWCL, 32/26.
63 THCKLPS, 6/89; K'ai-hua Hsien-chih, 3/32.
64 Hai Jui, ffcw / w C/w, pp. 48 ff. Cf. T'ung-tsu Ch'u, Local Government, p. 26.
65 K'ai-hua Hsien-chih, 3/30-7.
66 Ku Yen-wu, Wen-chi, 1/16.
67 Ku'ai-chi Chih, 7/10.
68 Wu-hsien Chili, 9/29.
69 rMHT, 38/7.
70 K'ai-hua Hsien-chih, 3/31.
71 flTJFCL, 32/18; THCKLPS, 6/79, 82.
72 MS, 300/3367; MSL, Shen-tsung Shih-lu, 1495.
73 THCKLPS, 6/20.
74 Ibid. 15/183.
75 Wang Ying-chiao, Tsou-su, 9/6.
76 THCKLPS, 6/82.
77 /M</. 6/21; Chang-chou Fu-chih, 5/18.
78 Chang Hsueh-yen, Wan-li K'uai-chi-lu, 4/100; THCKLPS, 25/41, 26/122.
79 When in 1519 a central government official associated with the Hanlin Academy
took possession of 52 mou of land owned by Ch'ang-shu county the magistrate
recorded his name in the local gazetteer and signed his own name below the entry.
About 1584 the magistrate of Chia-hsing entered in the gazetteer the name of a per-
son who held 1,000 mou of land in between the boundaries of two counties and paid
taxes to neither. In the gazetteer the magistrate also added that he had succeeded in
making another official landowner in a similar situation pay his taxes on his 3,340-
mou estate. See Ch'ang-shu Hsien-chih, 2/42; THCKLPS, 12/30.
80 Chang Hsueh-yen, Wan-li K'uai-chi-lu, 9/62.
348 Notes to pages 155-64
81 The Yuan tax assessment is recorded in Hui-chou Fu-chih, 7/18.
82 Chang Hsiieh-yen, Wan-li K'uai-chi-lu, 14/17.
83 Lin-fen Hsien-chih, All.
84 In 1586 for example the governor of South Chihli authorized the commutation of
more than 40,000 piculs of tax grain from Shang-hai county: see Shanghai Hsien-
chih, 3/20.
85 The option was offered by Hang-chou prefecture in 1579: see Hang-chou Fu-chih,
29/21, 36, 56.
86 CSWP, 327/12; Chang Chli-cheng, Shu-tu, 3/5, 3/21.
87 THCKLPS, 1/11.
88 Ibid. 6/79.
89 The magistrate was Wang Ssu-jen, who held office in 1610-11. The record of his
campaign was published as a book, the preface of which was reproduced by Ku
Yen-wu: see THCKLPS, 6/97-9; Sung-chiang Fu-chih, 38/15.
90 MS, 251/2846. Ch'ien Shih-sheng was regarded by his contemporaries as an honest
man.
91 Yeh Meng-chu, Yueh-shih-pien, 6/1-8.
92 CKing Sheng-tsu Shih-lu, 3/3.
93 The presence of a large number of medium-grade landowners may possibly be
corroborated by the fact that in Chia-hsing prefecture, Chekiang, there were ninety-
one distinguished clans, each comprising hundreds of households. In the late Ming
and early Ch'ing each of these clans maintained its prominence for over 200 years,
according to P'an Kuang-tan, Chia-hsing ti Wang-ts'u. For the number of tax-paying
households see Nan-chi Chih, 3/2.
94 Yeh Meng-chu, Yueh-shih-pien, 1/18.
95 Hai Jui, Hai Jui Chi, 431-2.
96 THCKLPS, 6/97-9.
97 HWYCL, 33/13.
98 THCKLPS, 6/14, 15, 24-6, 35, 61.
99 T'ang Shun-chih, Wen-chi, 9/24. Cf. CSWP, 261/8-9. P'eng Hsin-wei, Ho-pi-shih,
pp. 457-61.
100 CSWP, 397/9.
101 HYWCL, 32/22.
102 Ko Shou-li, Ko-tuan-su-kung Chi, 13/19, 14/15.
103 Yeh Meng-chu, Yueh-shih-pien, 1/18-19.
104 JCL, 4/56.
105 Wu-hsien Chih, 9/14.
106 CKang-shu Hsien-chih, 4/13. Also see ch. 1, notes 81 and 82.
107 THCKLPS, 26/85, 88-9; Fu Ming, Nung-ts'un She-hui, pp. 25, 45-6, 51-9. Cf.
Shimizu, 'Fukuken no nokakeizai'.
108 THCKLPS, 26/88-9.
109 Ibid. 26/89.
110 They are dated 1808, 1812, 1863, and 1891: see Fu I-ling, Nung-ts'un She-hui,
pp. 53-9.
111 MS, 78/825; MSL, Shih-tsung Shih-lu, 2803-6; CSWP, 366/20, 397/1; THCKLPS,
17-111.
112 MSL, Shen-tsungShih-lu, 10862-5;Ch'eng K'ai-hu, CKou-Liao Shih-hua, 11/13-17,
15/41.
113 There was apparently some justification for these readjustments, though the im-
perial order did not specify it clearly. Note that the land survey of Hukwang in 1581
reported an acreage in excess of 83 million mou. See MSL, Shen-tsung Shih-lu,
2412. Also see Appendix D above. On the basis of the unedited data it is estimated
that the taxable land in Huai-an prefecture in 1573 was 9,705,830 mou: Huai-an
Fu-chih (1573), 4/7-8.
114 MS, 78/823.
115 Teng-chou Chih, 7/5.
116 TMHT, 17/5, 10.
 Notes to pages 165-75 349
117 CKung-chen Ts'un-shih Su-cKao, 5/1.
118 For example, see Shen Pang, Wan-shu Tsa-chi, 125, 135.
119 For details, see Ch'en Jen-hsi, Shih-fa-lu, 34/35-7.
120 THCKLPS, 13/45-6, 74-5; MSL, Shih-tsung Shih-lu, 3893.
121 Chang Hsiieh-yen, Wan-li K'uai-chi-lu, 4/98-9.
122 Based on the data of Hang-chou Fu-chih, 29/1-31/70.
123 Ping-ti Ho, Studies on the Population, pp. 108-9.
124 THCKLPS, 22/118.
125 Fu Ming, Nung-ts'un She-hui, p. 288.
126 Ni Hui-ting, Nien-p'u, 3/13.
127 As quoted by Fu I-ling, in Nung-tsUm She-hui, p. 37.
128 Hang-chou Fu-chih, 29/31, 36, 56.
129 This is based on the data of Fen-chou Fu-chih, 5/3-46.
130 Fen-chou Fu-chih, 5/4.
131 MS, 41 /453; Fen-chou Fu-chih, 6/1.
132 Ibid. 5/40-1.
133 Ibid. 5/35-6.
134 Ibid. 5/32.
135 Hsu Cheng-ming, Lu-shui K'e-fan, 12; Wan-li K'uai-chi-lu, 7/79. The estimated
prices in this study are below those quoted by Ch'iian Han-sheng in 'Pei-pien
Mi-liang Chia-ke', pp. 61-2. Ch'iian's original data reflected unusual conditions
on the northern frontier.
136 An-hua Hsien-chih, 2/8-9.
137 MSL, Hsi-tsung Shih-lu, 1558.
138 Ano, 'Yosuko churyuiki', p. 66.
139 Ku-chin Tu-shu Chi-cKeng, Shih-ho-tien, 152/5/2.
140 See Ano, * Yosuko churyuiki', p. 85.
141 THCKLPS, 8/73-4.
142 For the irrigation system see Kuei Yu-kuang, San-Wu Shui-li-lu, 3/39. By com-
parison, Feng-yang prefecture, which was situated to the north of the Yangtze
and whose land was considered poor, also reported an average yield of 1 picul of
husked rice per mou: see THCKLPS, 9/13, 24.
143 Ko Shou-li, Ko-tuan-su-kung Chi, 15/18.
144 Ping-ti Ho, Studies on the Population, pp. 105-7.
145 Wen-shang Hsien-chih, 8/61-2.
146 THCKLPS, 26/88-9.
147 Chang-chou Fu-chih, 5/53. The yield has been converted from local measurements.
148 THCKLPS, 15/175.
149 K'ai-hua Hsien-chih, 3/12.
150 Ping-ti Ho, Studies on the Population, pp. 109-11.
151 The total taxable acreage was 1,495,070 mou. Total taxes were 155,537 piculs of
husked rice plus 121,447 taels of silver: see Shanghai Hsien-chih, 3/19-21.
152 The rates were as follows: On each mou of top-quality land, 0.182 piculs of husked
rice plus 0.1275 taels of silver. On each mou of medium-grade land, 0.136 piculs of
husked rice plus 0.140 taels of silver. On each mou of poor land, 0.0637 piculs of
husked rice plus 0.07 taels of silver: see THCKLPS, 6/47.
153 Ni Yiian-lu, CKuan-chi, Memorials, 8/6-7.
154 For the wartime surtaxes, see my article 'Fiscal Administration', p. 118.
155 Governor Wang Hsiang-heng mentioned in 1621 that except in the lower Yantze
region, land taxes in general varied from 0.04 taels to 0.09 taels per mou. The
lowest rates were 0.02 to 0.03 taels per mou. THCKLPS, 6/47.
156 Ibid. 18/81-2.
157 The following episode provides a basis for comparison: In 1534 the emperor
ordered that the land taxes be reduced by half for a year, which the ministry of
revenue indicated would cost the treasury 6,819,000 taels of silver. No other in-
formation was given, but presumably either the Gold Floral Silver, the tribute
grain and other tax consignments paid in kind were not included in the estimate, or
350 Notes to pages 176-87
else the retained revenue was excluded. Taking this into account it seems likely
that the total value of the land taxes was close to 20 million taels. For the 1534
order, see MSL, Shih-tsung Shih-lu, 3712.
158 TMHT, 42/37, 38, 43, 46.
159 The main features of the administration remained unchanged until the nineteenth
century: see Sun, 'The Board of Revenue', pp. 175-228.
160 The figures have been arrived at by summarizing the data given in chapters 25, 26,
27, 28, and 42 of TMHT.
161 For the 1502 summary, see MSL, Hsiao-tsung Shih-lu, 3548-55. Similar summaries
appear in HYWCL, 34/1 and 32/24.
162 The summary is derived from the unedited data in Lin-fen Hsien-chih, 4/2-5.
163 THCKLPS, 13/76.
164 MSL, Hsiao-tsung Shih-lu, 3552.
165 MSL, Shih-tsung Shih-lu, 5169.
166 Cheng Hsiao, Chin-yen Lei-pien, 1/36.
167 MSL, Mu-tsung Shih-lu, 0843,1424.
168 Hucker, 'Governmental Organization', p. 9; Wu Han, Chu Yuan-chang Chuan,
p. 216.
169 MSL, Shih-tsung Shih-lu, 8448-9; MS, 82/864.
170 MSL, Mu-tsung Shih-lu, 1039.
171 MSL, Shih-tsung Shih-lu, 5753-4, 5902-6, 8191-2, Mu-tsung Shih-lu, 0570, 0843,
0857.
172 The summary is derived from the unedited data in Wu-hsien Chih, 8/1-51.
173 MSL, Hsien-tsung Shih-lu, 3596.
174 CSWP, 369/9.
175 Chang-chou Fu-chih, 5/18.
176 CSWP, 397/16-17.
177 MSL, Hsien-tsung Shih-lu, 3596.
178 See Ho Liang-chun, Ssu-yu-chai, 3/196-7; Chin-hua Fu-chih, 8/13.
179 The combined budget was called' lien-ping-mi' (' troop-training rice') o r ' lien-ping-
yin* ('troop-training funds'): see ShanghaiHsien-chih, 3/21; Wu-hsien Chih, 1/11.
180 For instance, Ku-su Chih, 15/1; Chin-hua Fu-chih, 8/40; Hui-chou Fu-chih, 111,
7/4.
181 MS, 224/2585.
182 MSL, Shih-tsung Shih-lu, 4208, 4215.
183 Wang Ao, Chen-tse Ch'ang-yu, 1/23; HYWCL, 34/2.
184 TMHT, 39/1-7; Sun Ch'eng-tse, Meng-yu-lu, 27/5.
185 Hsieh Shang-chih, CKang-shu Shui-lun, 10; CSWP, 375/11.
186 MS, 222/2559; MSL, Shen-tsung Shih-lu, 1986, 3317, 5491, 6543, 11266.
187 Su T'ung-pin, I-ti Chih-tu, p. 439.
188 Hui-chou Fu-chih, 8/23.
189 K'ai-hua Hsien-chih, 3/31-2.
190 Hai Jui, Hai Jui Chi, 48-9.
191 HYWCL, 32/9.
192 Ni Yuan-lu, CKuan-chi, Memorials, 9/5; JCL, 3/80.
193 Lach, Asia in the Making of Europe, i, p. 769.
194 K'ai-hua Hsien-chih, 3/33.
195 Huai-jou Hsien-chih, 2/10.
196 CSWP, 438/3-4.
197 Hsieh Kuo-chen, CKang-shu Shui-lun, 3-4, 10-11.
198 Kuei Yu-kuang, San-Wu Shui-li-lu, 1/5, 7.
199 Wen-shang Hsien-chih, 8/61-2.
200 MSL, Ying-tsung Shih-lu, 0110.
201 JCL, 4/56.
202 Fu Ming, Nung-ts'un She-hui, pp. 11-13, 22, 24, 33; Hsieh Kuo-chen, Tang-she
Yun-tung-k'ao, pp. 268-9.
203 Hai Jui, Hai Jui Chi, 418, 457.
 Notes to pages 187-96 351
204 Ping-ti Ho, in Ladders of Success, shows numerous cases of upward and downward
social mobility. Fu I-ling also admits that large land-holdings were always
fragmented after a few generations: see Nung-ts'un She-hui, p. 85. Marginal land-
owners must have played an important role in the process as they represented both
the initial stage of upward mobility and the final stage of downward mobility.

5 The salt monopoly

1 This statement is made in 'Ming-tai I-t'iao-pien Nien-piao' (A Chronological
Table of the Single Whip Reform), which originally appeared in Journal of Ling-nan
University, 12:1. I have not seen the article but have taken the quotation from
People's University, She-hui Ching-chi, p. 192.
2 MS, 225/2596; MSL, Shen-tsung Shih-lu, 0979, Chiao-k'an-chi, 309.
3 MS, 80/837.
4 For details of conditions in the production areas, see MS, 80/838; THCKLPS,
15/133, 17/82, 84-6, 19/93-102; Ho-tung Yen-fa-chih, 1/1.
5 CSWP, 382/4.
6 For areas of salt distribution, see TMHT, chs. 32-3.
7 TMHT, chs. 15-16.
8 THCKLPS, 17/81, 28/8; Sun Ch'eng-tse, Meng-yu-lu, 35/44.
9 THCKLPS, 18/27.
10 MS, 73/859, 75/802; Sun Ch'eng-tse, Meng-yu-lu, 35/44, 48/43. For salt-control
censors, see Hucker, Censorial System, p. 83.
11 Chu T'ing-li, Yen-cheng-chih, 9/9-14 contains a list of salt-control censors complete
up to the date of publication.
12 MS, 80/842, 227/2316; MSL, Shih-tsung Shih-lu, 8047.
13 THCKLPS, 12/42; CSWP, 475/23-9.
14 It is possible that the production fields had territorial responsibilities: see Liang-Che
Yen-fa-chih, 2/5.
15 TMHT, 32/1-2; CSWP, 475/27; Chu T'ing-li, Yen-cheng-chih, 7/42.
16 CSWP, 475/25; Chu T'ing-li, Yen-cheng-chih, 7/68.
17 Shantung Yen-fa-chih, 11/21.
18 Chu T'ing-li, Yen-cheng-chih, 4/28; MS, 80/839; TMHT, 32/4. In Kwangtung, how-
ever, each salt-producing ting provided only 1,286 catties of salt. See Liang-Kuang
Yen-fa-chih, 3/14.
19 For these village collectors, see Shanghai Hsien-chih, 4/22-3; THCKLPS, 31/29;
TMHT, 34/38; CSWP, 358/8.
20 TMHT, 34/36, 38; Ch'en Jen-hsi, Shih-fa-lu, 29/2.
21 For the granary receipts and stub-books see TMHT, 34/4-5.
22 For the different weights of the yin, see CSWP, 475/19-20; Ho-tung Yen-fa-
chih, 10/8; Shantung Yen-fa-chih, 11/7-8; Liang-Che Yen-fa-chih, 3/81-3. Also
refer to Appendix C above for the regulations of 1535. For yin license, see TMHT,
34/14-15.
23 Chu T'ing-li, Yen-cheng-chih, 7/49.
24 MSL, Mu-tsung Shih-lu, 0730; Liang-Che Yen-fa-chih, 3/84-6.
25 Ibid. 3/84.
26 CSWP, 414/6, 475/1, 14.
27 Ibid. 474/15,29.
28 Shantung Yen-fa-chih, 11/2.
29 CSWP, 475/15, 28-9.
30 CSWP, 475/5, 26; THCKLPS, 12/43-4.
31 CSWP, 475/9.
32 See THCKLPS, 31/29-30; CSWP, 476/4-5.
33 THCKLPS, 31/30; Shanghai Hsien-chih, 4/17; MS, 153/1864; Chu T'ing-li,
Yen-cheng-chih, 7/69; Ku'ai-chi Chih, 7/16.
34 Chu T'ing-li, Yen-cheng-chih, 7/70. It is also stated that Liang-huai 'originally' had
35,266 ting: see Ch'en Jen-hsi, Shih-fa-lu, 29/2.
352 Notes to pages 196-206
35 Shantung Yen-fa-chih, 14/11.
36 The estimate is based on a statement by Huo T'ao in 1528, a report by the ministry
of revenue in 1550, a memorial by Censor Yang Hsiian in 1551, and an analysis by
a salt official, Yuan Shih-cheng, dated 1616. For the sources see MSL, Shih-tsung
Shih-lu, 6420, 6575; Chu T'ing-li, Yen-cheng-chih, 7/50; CSWP, 475/11.
37 CSWP, 357/22; MSL, Hsiao-tsung Shih-lu, 0650, Wu tsung Shih-lu, 0085.
38 Shantung Yen-fa-chih, 14/11.
39 Ibid. 13/2-5; Ping-ti Ho, Ladder of Success, p. 85.
40 Shanghai Hsien-chih, 4/26-7.
41 Shantung Yen-fa-chih, 14/6, 25; CKang-lu Yen-fa-chih, 6/44, 79; Chu T'ing-li,
Yen-cheng-chih, 5/14.
42 Hang-chou Fu-chih, 31/18.
43 THCKLPS, 21/29, 31, 36.
44 THCKLPS, 22/21; CSWP, 357/1-5.
45 MSL, Mu-tsung Shih-lu, 0185. By comparison, the commuted payments in Shan-
tung varied from 0.2 taels to 0.9 taels per ting, depending on the household grading:
see CSWP, 358/2.
46 Ping-ti Ho, 'The Salt Merchants of Yangchou', p. 132.
47 THCKLPS, 22/133.
48 Ibid. 15/130.
49 Ibid. 17/81.
50 Ho-tung Yen-fa-chih, 10/4-5.
51 MS, 80/842; CSWP, 409/5.
52 For details of those plates see Chu T'ing-li, Yen-cheng-chih, 7/9, 43; THCKLPS,
12/44.
53 CSWP, 357/21-3.
54 CKang-lu Yen-fa-chih, 6/8.
55 THCKLPS, 21/31.
56 Chu T'ing-li, Yen-cheng-chih, 7/8, 10; CSWP, 474/31-2, 460/31; MSL, Wu-tsung
Shih-lu, 2208.
57 Ho-tung Yen-fa-chih, 2/1-2, 12/38.
58 MSL, Wu-tsung Shih-lu, 3154-5, Shih-tsung Shih-lu, 1400-1.
59 Ibid. 1704.
60 Cases of this kind are cited in CSWP, 414/2, 477/4.
61 CSWP, 477/14.
62 MSL, Hsiian-tsung Shih-lu, 1313.
63 MSL, Hsien-tsung Shih-lu, 4103-4, Hsiao-tsung Shih-lu, 0792.
64 TMHT, 34/9; Chu T'ing-li, Yen-cheng-chih, 5/15.
65 Ibid. 4/7; MS, 80/839.
66 Chu T'ing-li, Yen-cheng-chih, 4/8.
67 THCKLPS, 21/36.
68 MSL, Hsien-tsung Shih-lu, 1698.
69 Chu T'ing-li, Yen-cheng-chih, 7/5-6.
70 Ibid. 4/22, 29, 7/7.
71 MSL, Hsiao-tsung Shih-lu, 0403-5. Possible errors are checked against Chiao-k'an-
chi, 0519.
72 For details, see Fujii, 'ensho no ichikosatsu'; Chu T'ing-li, Yen-cheng-chih,
7/52-4.
73 MSL, Hsien-tsung Shih-lu, 2891.
74 Ibid. 2434.
75 MS, 80/840, 185/2161.
76 Sun Ch'eng-tse, Meng-yu-lu, 35/45; CSWP, 474/6.
77 For ticket salt, see Chu T'ing-li, Yen-cheng-chih, 4/7; MSL, Shih-tsung Shih-lu,
3750, 4205-6, 8714, Mu-tsung Shih-lu, 0897; CSWP, 358/3.
78 Fujii, 'ensho no ichikosatsu', Shigaku zasshi, 54:5, pp. 62-111; 54:6, pp. 65-104;
and 54:7, pp. 17-59. Cf. Ch'ung-wu Wang, 'The Ming System of Merchant
Colonization', in Sun and de Francis, eds. Chinese Social History, pp. 299-308.
 Notes to pages 206-16 353
79 MS, 80/840.
80 For details see CSWP, 357/24-5, 360/23.
81 CSWP, 360/24, 475/9.
82 It is not clear when the collection of the surplus salt money started. According to
one source it was during the Ch'eng-hua period: see MSL, Wu-tsung Shih-lu, 2878.
Wang Ch'iung used the term in Hu-pu Tsou-i, 1/16. It seems that the collection had
become common practice by 1507, if not much earlier.
83 MSL, Hsiao-tsung Shih-lu, 3593-4.
84 MSL, Wu-tsung Shih-lu, 0153, 0506, 0599, 2584, 2805; MS, 80/840; Ming-cKen
Tsou-i, 16/278.
85 MSL, Wu-tsung Shih-lu, 1019, 2308, 2805.
86 For the settlement see MSL, Shih-tsung Shih-lu, 3791-5, 5630, 6575, 6655. See also
Appendix C above.
87 MS, 308/3489-90; Hai Jui, Hai Jui Chi, 168.
88 MS, 80/842; Chiao Hung, Hsien-cheng-lu, 17/3; MSL, Shih-tsung Shih-lu, 8255,
8276, 8299, 8464, Mu-tsung Shih-lu, 0895.
89 CSWP, 357714.
90 MS, 80/841; MSL, Shih-tsung Shih-lu, 6922, 8868.
91 Ibid. 6420, 7776.
92 CSWP, 357/23, 358/8.
93 THCKLPS, 12/41.
94 CSWP, 414/21, 29.
95 CSWP, 360/24-5.
96 CSWP, 357/24, 360/22-3.
97 P'eng Hsin-wei, Ho-pi-shih, 474. The monthly rate could be as high as 5 per cent.
98 CSWP, 357/31.
99 CSWP, 357/24-5, 360/23.
100 CSWP, 357/26, 29, 360/28.
101 CSWP, 360/29.
102 River salt is said to have been started by Yen Mou-ch'ing. See THCKLPS, 12/38;
MS, 80/842. The statement that P'ang Shang-p'eng discontinued it is incorrect.
Cf. Ch'en Jen-hsi, Shih-fa-lu, 29/45.
103 MSL, Shen-tsung Shih-lu, 1687.
104 For Lu Pao see MS, 80/842, 237/2706; MSL, Shen-tsung Shih-lu, 6072, 6095, 6392,
6543, 8307. The consequences of his administration are discussed in CSWP
470/1-8.
105 MS, 80/841; TMHT, 34/12; MSL, Shih-tsung Shih-lu, 3791-5.
106 CSWP, 375/15.
107 CSWP, 474/25, 475/7.
108 TMHT, 32/1-33/27.
109 Liang-Huai Yen-fa-chih, 4/9.
110 CSWP, 357/5.
111 THCKLPS, 22/20-1.
112 Liang-Che Yen-fa-chih, 4/10; THCKLPS, 21/31; MSL, Shih-tsung Shih-lu, 6059.
113 MSL, Shih-tsung Shih-lu, 8482.
114 MSL, Mu-tsung Shih-lu, 0735.
115 MSL, Shen-tsung Shih-lu, 7149.
116 MSL, Mu-tsung Shih-lu, 0441.
117 Chou Hsiian-wei, Ching-lin Hsu-chi, 48.
118 THCKLPS, 26/94. See also MSL, Mu-tsung Shih-lu, 0720-1.
119 In 1616 it was reported that Yunnan had not delivered any funds outside the
province for over twenty years. See CSWP, 474/2.
120 In 1575 Minister of Revenue Wang Kuo-kuang estimated that the income from the
barter system was worth approximately 500,000 taels: see MSL, Shen-tsung Shih-lu,
0792. In 1616 Minister of Revenue Li Nii-hua estimated the total annual income to
be 2.4 million taels. See CSWP, 474/2. For other references see MSL, Mu-tsung
Shih-lu, 0850-1, Shen-tsung Shih-lu, 0624.
354 Notes to pages 216-26
121 This is calculated from the unedited data in TMHT, chs. 32-3.
122 Ya-tung Hsiieh-she, Jen-k'ou Wen-fi, P- 299.
123 CSWP, 476/1; Twitchett, Financial Administration, p. 58.
124 Hsiang-ho Hsien-chih, 11/11.
125 CSWP, 477/19.
126 CSWP, 477/21.
127 Yuan Shih-cheng's paper is entitled 'Ten Proposals of the Ministry of Revenue',
in CSWP, 474/1-411125.
128 CSWP, 357/24-6, 360/22-3, 27.
129 CSWP, 474/4, 475/15.
130 CSWP, 474/18.
131 CSWP, 474/10. Sometimes the speculators bought the receipts at 0.17 taels per
yin and sold them at 0.85 taels per yin: see CSWP, 475/7.
132 CSWP, 474/16-17, 23-4, 477/6.
133 CSWP, 474/22.
134 CSWP, 474/23.
135 CSWP, 474/26-7.
136 CSWP, 476/9.
137 CSWP, 474/26.
138 THCKLPS, 12/44.
139 THCKLPS, 12/44; CSWP, 360/25.
140 Hai Jui, Hai Jui Chi, 49-50, 55.
141 For the incident see MSL, Shen-tsung Shih-lu, 6534. Another instance of unauthor-
ized taxation is also recorded, ibid. 2522.
142 This rate is calculated from the data in THCKLPS, 26/67.
143 CSWP, 474/26.
144 The basis of this estimate is as follows: In the middle of the sixteenth century, the
price was about 10 taels per short ton, whereas later in the century the lowest price
was about 9.6 taels per short ton. One source gives it as 11.4 taels per short ton.
When the governor of Hukwang imposed price control, the ceiling price was about
18 taels per short ton. These prices are calculated from Chu T'ing-li, Yen-cheng-
chih, 7/40; THCKLPS, 12/48; Sun Ch'eng-tse, Meng-yu-lu, 35/49; CSWP, 477/21.
145 The statement is made by Fujii in Wada, shokkashi yakuchu, pp. 598-603.
146 For the settlement see MSL, Shen-tsung Shih-lu, 10607, 10687-8; Sun Ch'eng-tse,
Meng-yii-lu, 35/46-8; CSWP, 475/19-20, 477/1-5.
147 Shantung Yen-fa-chih, 14/7-9.
148 Chu T'ing-li, Yen-cheng-chih, 10/20-1.
149 CSWP, 475/24.
150 CSWP, 383/21.
151 CSWP, 475/24. It is interesting that Yuan Shih-cheng, who reported all these
irregularities, was himself later impeached for corruption. MSL, Hsi-tsung Shih-lu,
0179.
152 MSL, Hsi-tsung Shih-lu, 1569.
153 MSL, Tai-tsu Shih-lu, 2141, 2681-2.
154 MSL, Shih-tsung Shih-lu, 6922.
155 CS WP, 474/26, 475/24.
156 CSWP, 474/29.
157 Ch'iu Chun, Ta-hsueh Yen-i-pu, 28/11.
158 Both Fujii and Ch'ung-wu Wang consider the charges unjustified. For references
see note 78 above.
159 CSWP, 360/27.
6 Miscellaneous incomes
1 For the origins of the tonnage stations see MSL, Tai-tsung Shih-lu, 2365, Hsiian-
tsung Shih-lu, 1325.
2 THCKLPS, 21/53-4.
3 MSL, Shih-tsung Shih-lu, 2494. Cf. MS, 81/854.
 Notes to pages 226-34 355
4 MSL, Mu-tsung Shih-lu, 0862. Another source indicates that the joint operation
started in 1529: see HYWCL, 40/1-2.
5 THCKLPS, 21/51-3.
6 For the collection at the Ch'ung-wen Gate see TMHT, 35/48.
7 An assistant collector at Chiu-chiang had allegedly embezzled several tens of
thousands of taels: see HYWCL, 40/6. For other evidence of abuses see Ch'i
Piao-chia, Jih-chi, vol. 5, dated the second day, ninth lunar month, 1643.
8 Their impression was that 'the mandarins are arbitrary but honest', Lach, Asia in
the Making of Europe, i, p. 754. For similar observations see Hucker, Traditional
Chinese State, p. 82n.
9 The quota was first established in 1377: see MSL, Tai-tsu Shih-lu, 1848.
10 Lin-cKing Chihli Chou-chih, 9/2.
11 THCKLPS, 11/44.
12 THCKLPS, 21/51.
13 MSL, Shen-tsung Shih-lu, 2529.
14 Ibid. 2589; Sun Ch'eng-tse, Meng-yu-lu, 35/39-40.
15 THCKLPS, 22/16.
16 HYWCL, 92/24.
17 Lin-cKing Chihli Chou-chih, ch. 9; Tu Lin, Huai-an San-kuan Tung-chih, ch. 2;
I-ling-a, Huai-kuan Tung-chih, 7/1-29.
18 Hsu Wen-hsien Tung-k'ao, 2933.
19 Ibid. 2938.
20 Tu Lin, Huai-an San-kuan Tung-chih, 8/9.
21 THCKLPS, 21/52-3, 55-6; HYWCL, 40/1.
22 The table is based on the figures in Sun Ch'eng-tse, Meng-yu-lu, 35/42 and Hsu
Wen-hsien Tung-k'ao, 2937.
Wu Chao-ts'ui, Shui-chih-shih, pp. 175-6, lists the collection quota at the Ch'ung-
wen Gate in 1625 as 48,900 taels. Sakuma's table shows more variations: see
Sakuma, 'shozei to zaisei', p. 61.
23 THCKLPS, 21/53-6.
24 MSL, Shen-tsung Shih-lu, 7072.
25 TMHT, 35/7; Chu Kuo-chen, Yung-cWuang Hsiao-p'in, 2/40. For examples of
ad hoc distribution see MSL, Shih-tsung Shih-lu, 1413, Shen-tsung Shih-lu, 2579,
4530.
26 Sun Ch'eng-tse, Meng-yu-lu, 35/9-10.
27 Wu Chao-ts'ui, Shui-chih-shih, p. 169.
28 MSL, Mu-tsung Shih-lu, 0555.
29 HYWCL, 40/1.
30 Fen-chou Fu-chih, 5/12.
31 Chin-hua Fu-chih, 8/20.
32 Hui-chou Fu-chih, 1/64, 8/14-16.
33 MSL, Mu-tsung Shih-lu, 0441, Shen-tsung Shih-lu, 0764-5.
34 MSL, Shih-tsung Shih-lu, 5991, 6474; HYWCL, 40/3-4.
35 CSWP, 358/28-9.
36 MSL, Hsien-tsung Shih-lu, 1827, Shih-tsung Shih-lu, 1038.
37 MSL, Shen-tsung Shih-lu, 1197.
38 TMHT, 35/32-7.
39 MSL, Tai-tsu Shih-lu, 1116, Tai-tsung Shih-lu, 0447-8.
40 Ch'en Wen-shih, Hai-ching Cheng-ts'e, pp 59-60.
41 MSL, Hsiao-tsung Shih-lu, 1367-8. Cf. Hsien-tsung Shih-lu, 4590.
42 Ch'en Wen-shih, Hai-ching Cheng-ts'e, p. 107; THCKLPS, 33/60.
43 MSL, Wu-tsung Shih-lu, 1496, 2911-12; Ch'en Wen-shih, Hai-ching Cheng-ts'e,
pp. 107-8; Liang Fang-chung, 'Kuo-chi Mou-i', pp. 292-3; CSWP, 357/7-10.
44 Liang Fang-chung,' Kuo-chi Mou-i', p. 292; Ch'en Wen-shih, Hai-ching Cheng-ts'e,
p. 109.
45 Little is known of conditions in the intervening years. See Lach, Asia in the Making,
i, pp. 737, 788.
356 Notes to pages 234-42
46 THCKLPS, 29/105.
47 MSL, Shih-tsung Shih-lu, 2509; THCKLPS, 33/58-60.
48 CSWP, 357/7-10; Liang Fang-chung, 'Kuo-chi Mou-i\ pp. 298, 305.
49 Chou Hsiian-wei, Ching-lin Hsu-chi, 48.
50 Liang Fang-chung, 'Kuo-chi Mou-i', p. 305, quoting the 1601 edition oiKwangtung
Tung-chih.
51 H. B. Morse, The Chronicles of the East India Company, i, p. 9.
52 Chang Hsiieh, Tung-hsi-yang K'ao, 7/95-8. The rates have been tabulated by Liang
Fang-chung, in 'Kuo-chi Mou-i', pp. 292-3.
53 In 1589, the governor of Fukien indicated that the rates were about 2 per cent
ad valorem: see MSL, Shen-tsung Shih-lu, 3939.
54 In 1597 the port was partially open: see MSL, Shen-tsung Shih-lu, 5899; THCKLPS,
26/33-4, 99-104; also MSL, Shen-tsung Shih-lu, 4864-5.
55 Katayama, 'Gekko nijyushisho', pp. 389-419; Chou Hsiian-wei, Ching-lin Hsii-chi,
47.
56 Shen Pang, Wan-shu Tsa-chi, 92-3.
57 Ibid. 83-91. Cf. MSL, Shen-tsung Shih-lu, 2355.
58 Shen Pang, Wan-shu Tsa-chi, 91.
59 THCKLPS, 14/43.
60 THCKLPS, 7/40.
61 MSL, Shen-tsung Shih-lu, 6960.
62 MS, 81/835.
63 TMHT, 204/1-8.
64 MSL, Hsien-tsung Shih-lu, 1724; Sun Ch'eng-tse, Meng-yu-lu, 46/54.
65 TMHT, 204/1.
66 TMHT, 204/7.
67 THCKLPS, 11/43; Hsi and Chu, Ts'ao-cKuan-chih, 4/10.
68 Chou Chih-lung, Ts'ao-ho I-pi, ch. 8; Chu Kuo-chen, Yung-cKuang Hsiao-p'in, 4/79.
69 Hsiang Meng-yiian, Tung-kuan Chi-shih, 6-7, 14.
70 MS, 189/2213; HYWCL, 92/24.
71 MS, 206/2397.
72 MS, 224/2591.
73 THCKLPS, 11/43-4.
74 I-ling-a, Huai-kuan Tung-chih, 5/7.
75 Hsiang Meng-yuan, Tung-kuan Chi-shih, 4.
76 Chou Chih-lung, Ts'ao-ho I-pi, ch. 7.
77 Ch'en Tzu-chuang, Ching-chi-yen, 3/42.
78 MSL, Hsien-tsung Shih-lu, 4319.
79 MSL, Shih-tsung Shih-lu, 1232. The latter figure is quoted by People's University,
She-hui Ching-chi, p. 94.
80 Chou Chih-lung, Ts'ao-ho I-pi, 8/13-14.
81 Ching-chou (referred to as Sha-shih in the text) in 1587 produced only 14,000 taels:
see MSL, Shen-tsung Shih-lu, 3544.
82 MSL, Tai-tsu Shih-lu, 3518; TMHT, 194/16.
83 TMHT, 194/19-21; Sun Ch'eng-tse, Meng-yu-lu, 46/59-60.
84 TMHT, 194/18; Ho Shih-chin, CHang-k'u Hsu-chih, 6/91, 7B/1-8. (There are two
chapter 7's in this work. Here it is the second that is referred to.)
85 THCKLPS, 28/8.
86 THCKLPS, 26/73, 104.
87 Sung Ying-hsing, Tien-kung K'ai-wu, 231; Sun, 'Mining Industries', p. 840.
88 Ho Shih-chin, CKang-Wu Hsu-chih, 1/3-4.
89 Ni Yuan-lu, CKuan-chi, Memorials, 10/11-12; Ni Hui-ting, Nien-p'u, 4/17.
90 The term chia-pan appears in MSL, Tai-tsung Shih-lu, 2266, Jen-tsung Shih-lu, 0017,
Ying-tsung Shih-lu, 5372.
91 Sung Ying-hsing, Tien-kung K'ai-wu, 229.
92 MS, 10/82-3; THCKLPS, 23/52-3.
93 THCKLPS, 23/54.
 Notes to pages 242-8 357
94 Liu Chin's fiscal reforms deserve further study. Even the isolated and prejudiced
statements in the Shih-lu suggest a far-reaching purpose behind them: see Wu-tsung
Shih-lu, 0864, 1318, 1439, 1440, 1456, 1482.
95 JW. 0815, 0847.
96 MSL, Shih-tsung Shih-lu, 0072; TMHT, 37/33.
97 MSL, Shih-tsung Shih-lu, 7867.
98 MS, 18/130.
99 TMHT, 37/25.
100 THCKLPS, 15/138-140.
101 MSL, Shih-tsung Shih-lu, 6180.
102 Ibid. 7692.
103 For these entries see MSL, Shen-tsung Shih-lu, 6645, 6663, 6700, 6718, 6731, 6732,
6751,6782, 6839,6863,6864. Furthermore, in the two years between 1596 and 1598
five provinces, including Shantung, Honan, Shansi, Chekiang, and North Chihli,
were reported to have delivered 106,000 taels: ibid. 6059-60.
104 THCKLPS, 32/46.
105 MSL, Shen-tsung Shih-lu, AMI.
106 MS, 75/804.
107 The list appears in TMHT, 31/1-26.
108 Ta-kao, 1/55.
109 Yung-chou Fu-chih, 9/1; THCKLPS, 25/50.
110 TMHT, 200/26-38.
111 MSL, Shen-tsung Shih-lu, 1810.
112 For the sales of 1465, 1485, 1512, 1536, 1561 and 1564, respectively, see MSL,
Hsien-tsung Shih-lu, 0436, 0552, 4521, Wu-tsung Shih-lu, 1883, Shih-tsung Shih-lu,
3993, 8512, 8667.
The commonest form of purchase was the acquisition by qualified scholars of the
status of imperial university student. The title alone was almost equivalent to
* associate membership' of the civil service. Even if he neither attended the uni-
versity nor entered the bureaucracy, an imperial university student received a hat
and belt from the government and was allowed the forms of address and certain
privileges of a low-ranking official. Some purchasers did actually attend the uni-
versity while others merely used the title to secure appointments as local officials,
usually in the frontier provinces. See JCL, 6/59-60; TMHT, 5/11-16, 220/21;
MSL, Hsien-tsung Shih-lu, 1023; Ping-ti Ho, Ladder of Success, pp. 33, 46.
113 This is the impression given by several funerary inscriptions composed by Kuei
Yu-kuang: see Kuei Yu-kuang, CKuan-chi, 18/241-2.
114 Hsu Cheng-ming, Lu-shui K'e-fan, 7.
115 Hsiang Meng-yiian, Tung-kuan Chi-shih, 5.
116 MSL, Wu-tsung Shih-lu, 1005.
117 MSL, Shih-tsung Shih-lu, 8887.
118 Chang Chu-cheng, Shu-tu, 3/15.
119 MSL, Hsi-tsung Shih-lu, 2316; Liu Tsung-chou, Liu-tzu CKuan-shu, 415/5.
120 MSL, Hsiao-tsung Shih-lu, 2064.
121 MSL, Hsien-tsung Shih-lu, 4550; Sun Ch'eng-tse, Meng-yu-lu, 39/86. Cf. MSL,
Hsien-tsung Shih-lu, 2301, 2310.
122 Ibid. 4406.
123 Ibid. 3658.
124 Sun Ch'eng-tse, Meng-yu-lu, 35/31; CSWP, 325/18.
125 MSL, Wu-tsung Shih-lu, 0937.
126 MSL, Shen-tsung Shih-lu, 1818, 2896-7.
127 MSL, Wu-tsung Shih-lu, 0937.
128 MSL, Shen-tsung Shih-lu, 0307.
129 MSL, Wu-tsung Shih-lu, 1241.
130 This sum is derived from the unedited data in TMHT, 26/1-59.
131 Sun Ch'eng-tse, Meng-yii-lu, 35/9.
132 See MS, 308/3488; MSL, Shih-tsung Shih-lu, 8848, Shen-tsung Shih-lu, 2757-8.
358 Notes to pages 248-55
133 TMHT, 176/14.
134 See table of commutations in TMHT, 176/9-14, 15-18.
135 TMHT, 179/19-24.
136 MSL, Shih-tsung Shih-lu, 5199; Wei Ch'ing-yiian, Huang-ts'e Chih-tu, pp. 124-7.
137 Hai Jui, Hai Jui Chi, 49.
138 MSL, Shen-tsung Shih-lu, 4904-5; Lien-sheng Yang,' Ming Local Administration',
p. 20.
139 TMHT, 30/23.
140 TMHT, 179/18.
141 Sun Ch'eng-tse, Meng-yii-lu, 35/10.
142 Ho Liang-chun, Ssu-yu-chai, 3/161.
143 MSL, Shen-tsung Shih-lu, 2127-8.
144 Hai Jui, Hai Jui Chi, 259, 276; Ho Liang-chiin, Ssu-yu-chai, 1/3.
145 MS, 81/851; TMHT, 31/14; MSL, Shen-tsung Shih-lu, 1129-31, 1132, 2944;
CSWP, 325/22-3; Sun Ch'eng-tse, Meng-yii-lu, 38/1-16; Hsiang Meng-yiian,
Tung-kuan Chi-shih 3, 5; P'eng Hsin-wei, Ho-pi-shih, pp. 444-6.
146 TMHT, 31/14; MSL, Shen-tsung Shih-lu, 1129-31, 2944. The statement in the
Shih-lu that the rate was fixed at 1,000 copper coins per tael of silver seems im-
probable. P'eng Hsin-wei suggests that some 2-cash coins were minted and that
the actual rate was 500:1: see Ho-pi-shih, pp. 444-6.
147 Hsiang Meng-yiian, Tung-kuan Chi-shih, 3.
148 Huai-an Fu-chih (1627), 12/24.
149 Sun Ch'eng-tse, Meng-yu-lu, 38/8.
150 HYWCL, 92/20.
151 TMHT, 152/18. For details see Tani, 'Mindai no chuang-p'eng-yin ni tsuite',
pp. 165-96.
152 MSL, Hsien-tsung Shih-lu, 3046; TMHT, 152/23.
153 MSL, Hsiao-tsung Shih-lu, 1282-3.
154 MSL, Wu-tsung Shih-lu, 0470-1.
155 MSL, Shih-tsung Shih-lu, 2337; Yang Shih-ch'iao, Ma-cheng-chi, 8/3.
156 Wang Yu-ch'iian, Chun-tun, pp. 8-9.
157 MSL, Mu-tsung Shih-lu, 0716-17.
158 MSL, Shen-tsung Shih-lu, 5241.
159 HYWCL, 40/9; CSWP, 406/1-5.
160 MS, 82/866; Sun Ch'eng-tse, Meng-yu-lu, 35/10; CSWP, 406/5; MSL, Wu-tsung
Shih-lu, 2733.
161 The quota in 1578 was 445,257 taels and in 1579 it was 450,900 taels. See TMHT
27/28; MSL, Shen-tsung Shih-lu, 1871.
162 TMHT, 27/27-9.
163 Sun Ch'eng-tse, Meng-yu-lu, 37/15.
164 MSL, Shih-tsung Shih-lu, 0627-8.
165 MSL, Shen-tsung Shih-lu, 3491.
166 MSL, Shih-tsung Shih-lu, 7733.
167 MSL, Shen-tsung Shih-lu, 1871-2.
168 TMHT, 189/10; Ho Shih-chin, CKang-Vu Hsu-chih, 2/22-3.
169 TMHT, 206/4-5; Hsi and Chu, Ts* ao-cK uan-chih, 4/1-15, 6/16, 41, 7/15.
170 TMHT, 208/18-19; THCKLPS, 8/70.
171 MSL, Shen-tsung Shih-lu, 3556, 3628.
172 MSL, Shih-tsung Shih-lu, 3893, Shen-tsung Shih-lu, 5383; Sun Ch'eng-tse, Meng-
yii-lu, 35/43. In 1609 the income was approximately 1.5 million taels: see Ho
Shih-chin, CKang-k'u Hsu-chih, 2/21.
173 MSL, Shen-tsung Shih-lu, 3652, 5916. The latter entry states that the income in
1597 was approximately 400,000 taels.
174 Sun Ch'eng-tse, Meng-yii-lu, 53/8-9. According to Chu Kuo-chen in 1614 the
income was 980,000 taels: see Yung cKuang Hsiao-p'in, 2/41.
175 MSL, Shih-tsung Shih-lu, 4835, 4957.
176 MSL, Hsien-tsung Shih-lu, 4917; also see TMHT, 157/1-12.
 Notes to pages 256-64 359
177 MSL, Shen-tsung Shih-lu, 1408, 1448, 1494, 1558, 1559, 1581, 1706.
178 MSL, Shen-tsung Shih-lu, 5992; also see 3473, 5274.
179 Ch'en Shih-ch'i states that the commutation took place during the Wan-li period:
see Kuan-shou-kung-yeh, p. 124.
180 For the early collections, see MSL, Ying-tsung Shih-lu, 0060, Hsien-tsung Shih-lu,
4393.
181 For the number of cooks see MSL, Hsiian-tsung Shih-lu, 0149, Hsiao-tsung Shih-lu,
0624; TMHT, 194/4. In 1574 the court of imperial entertainments consumed
160,000 catties of salt, which indicates that its kitchens could feed 15,000 persons
daily: see MSL, Shen-tsung Shih-lu, 0654.
182 MSL, Shih-tsung Shih-lu, 7863-4, Mu-tsung Shih-lu, 0148-50, Shen-tsung Shih-lu,
0593-4, 3632.
183 Rossabi, 'Tea and Horse Trade', p. 163.
184 For the early system see MSL, Tai-tsu Shih-lu, 1300-1; MS, 80/843-4; TMHT,
37/3-5.
185 THCKLPS, 19/17-18, 35-7.
186 TMHT, 37/1; Hucker, * Governmental Organization', p. 46.
187 MSL, Hsuan-tsung Shih-lu, 0249.
188 MSL, Ying-tsung Shih-lu, 0189; TMHT, 37/8.
189 TMHT, 37/13.
190 Many tea plantations were not taxed, and as much as a million catties of tea a year
was traded as contraband: see CSWP, 115/1-10, 12-13.
191 MS, 80/845; TMHT, 37/8; MSL, Hsiao-tsung Shih-lu, 0846.
192 CSWP, 115/16-17.
193 MS, 80/845; THCKLPS, 18/98.
194 TMHT, 37/16; THCKLPS, 18/98; CSWP, 386/12.
195 CSWP, 386/7.
196 MSL, Wu-tsung Shih-lu, 0683-4.
197 TMHT, 31/7; CSWP, 115/12-13.
198 CSWP, 115/16-17.
199 MS, 80/845; MSL, Shih-tsung Shih-lu, 3968.
200 CSWP, 386/16.
201 THCKLPS, 19/36.
202 Rossabi, 'Tea and Horse Trade', pp. 159-63.
203 MSL, Shen-tsung Shih-lu, 3773.
204 For Altan's request see MSL, Shen-tsung Shih-lu, 1459-60. For tax reductions see
ibid. 2943, 3405.
205 MSL, Shen-tsung Shih-lu, 5207.
206 MS, 80/846.
207 CSWP, 385/7, 9-10, 19-20; THCKLPS, 18/86.
208 MSL, Shen-tsung Shih-lu, 3699, 4407; Yang Shih-ch'iao, Ma-cheng-chi, 8/4. See
also Hou Jen-chih,' Ma-shih-k'ao', an English translation of which appears in Sun
and de Francis, Chinese Social History, pp. 309-32; Wang Shih-ch'i, San-yun
CKou-tsu-Wao, 2/17.
209 MSL, Shen-tsung Shih-lu, 5085.
210 MS, 80/846; TMHT, 37/2, 9.
211 THCKLPS, 19/104-5; CSWP, 383/2-3
212 TMHT, 37/3.
213 Sun Ch'eng-tse, Meng-yu-lu, 35/54.
214 It seems to have become general in the late sixteenth century. The later entries are
dated 1585,1587,1589, and 1598: Wang I-o, Tsou-i, 1/18; MSL, Shen-tsung Shih-
lu, 3401, 4079, 5941.
215 These conditions were observed by Ni Yiian-lu, the dynasty's last minister of
revenue. See Ni Yiian-lu, CKuan-chi, Memorials, 6/2; Ni Hui-ting, Nien-p'u, 4/17,
27.
216 Ping-ti Ho, Studies on the Population, pp. 23, 277.
217 MSL, Shen-tsung Shih-lu, 8200-1.
360 Notes to pages 265-75
218 MSL, Mu-tsung Shih-lu, 0746.
219 MSL, Hsi-tsung Shih-lu, 0609.

7 Financial management
1 Twitchett, Financial Administration, pp. 97-123.
2 In 1568, for instance, the ministry of revenue accused the Kwantung provincial
officials of mismanaging the funds which they were allowed to retain: see MSL,
Mu-tsung Shih-lu, 0440-1.
3 MSL, Shih-tsung Shih-lu, 5961, 5976-7.
4 MSL, Shen-tsung Shih-lu, 5952-3.
5 MSL, Shih-tsung Shih-lu, 6405; CSWP, 199/44.
6 MSL, Shih-tsung Shih-lu, 7712-13.
7 MSL, Mu-tsung Shih-lu, 0332, Shen-tsung Shih-lu, 1831, 2684, 2852, 3517, 4084,
4170, 4333; Sun Ch'eng-tse, Meng-yu-lu, 35/31; CSWP, 325/18-19, 389/2.
8 MSL, Shen-tsung Shih-lu, 2684, 2853, 2921, 3517, 4084, 4333.
9 MSL, Shen-tsung Shih-lu, 2920.
10 These accounts changed very little over a long period. See MSL, Shih-tsung Shih-lu,
7201; CSWP, 198/14-15; Sun Ch'eng-tse, Meng-yu-lu, 35/8-10.
11 MSL, Wu-tsung Shih-lu, 2408.
12 MS, 78/826. The imperial edict announcing the collection indicated that it would be
discontinued when the frontier situation improved. See MSL, Shih-tsung Shih-lu,
6604.
13 MS, 202/2347; MSL, Shih-tsung Shih-lu, 5908-9.
14 MSL, Shih-tsung Shih-lu, 5315, 6891, 8482. Also see CSWP, 259/9-10; T'ang Shun-
chih, Wen-chi, 9/25; Wada, Shokkashi yakuchu, p. 323n.
15 MSL, Shih-tsung Shih-lu, 5339; TMHT, 17/23.
16 MSL, Shih-tsung Shih-lu, 7870-1.
17 TMHT, 17/26-30.
18 MS, 78/826; MSL, Shih-tsung Shih-lu, 7712-13.
19 MSL, Shih-tsung Shih-lu, 7713-15.
20 MSL, Shih-tsung Shih-lu, 7717-18.
21 MSL, Shih-tsung Shih-lu, 7719-20.
22 MSL, Shih-tsung Shih-lu, 7733-6.
23 MS, 78/826.
24 The MS states that at an earlier stage the T'ai-ts'ang Treasury had deposits of
8 million taels (see MS, 78/826), but this cannot be verified. The only positive
evidence is that 400,000 taels, a rather small amount, was withdrawn from the reserve
in 1543: see MSL, Shih-tsung Shih-lu, 5338. In 1550 Minister of Revenue P'an Huang
likewise stated that 'in earlier years' the treasury had a regular reserve of 4 million
taels: see CSWP, 198/14.
25 MSL, Shih-tsung Shih-lu, 7349, Shen-tsung Shih-lu, 1027.
26 See TMHT, 28/26, 29, 30, 33, 34, and 36, under the headings of Chi-chou, Yung-
p'ing, Ch'ang-p'ing, I-chou, and Liao-tung.
27 Sun Ch'eng-tse, Meng-yu-lu, 35/9.
28 MS, 82/866.
29 MSL, Shen-tsung Shih-lu, 1641.
30 MSL, Shen-tsung Shih-lu, 2920.
31 For the 1578 payment in silver see MSL, Shen-tsung Shih-lu, 1647; CSWP, 375/12.
In 1567, Minister of Revenue Ma Shen reported that pay and salaries had cost the
treasury 1.35 million taels that year, which shows that on this occasion the payments
were all made in silver instead of grain: see MSL, Mu-tsung Shih-lu, 0414.
32 TMHT, 31/1; MSL, Shih-tsung Shih-lu, 7871.
33 MSL, Shen-tsung Shih-lu, 1611.
34 An imporatnt work on the operations of the treasury was the T'ai-ts'ang ICao by
Liu Szu-chieh (fl. 1545-75), possibly completed around 1580 (see MS, 97/1028),
which unfortunately cannot be located. The summary of the accounts of the T'ai-
 Notes to pages 275-83 361
ts'ang Treasury in Sun Ch'eng-tse, Meng-yii-lu, dated 1580, was taken from this
account (see Meng-yii-lu, 35/8). The summary in MS, 82/866 is very close to Sun's,
and it is probable that they all come from the same source.
35 In 1578 the ministry of revenue stated that in Peking it annually disbursed between
700,000 taels and 800,000 taels of silver. In 1591 the amount was said to be 630,000
taels, but this may not include the pay of the army officers, disbursed by the
emperor. In 1600 the cost of maintaining the imperial stable and imperial zoo was
said to be 123,000 taels. In 1623 'salaries' totalled 526,633 taels. See MSL, Shen-
tsung Shih-lu, 1590, 4333, 6594; Ch'en Jen-hsi, Shih-fa-lu 36/8.
36 Hai Jui, Hai Jui Chi, 40.
37 See Shun-fien Fu-chih, 52/12-14.
38 See Shen Pang, Wan-shu Tsa-chi, 142-7.
39 Kiangsi Fu-i CKuan-shu, provincial summary, 41.
40 MSL, Shih-tsung Shih-lu, 3712.
41 Shen Pang, Wan-shu Tsa-chi, 84-5.
42 MSL, Hsiao-tsung Shih-lu, 3550.
43 Ku Ch'ing, P'ang-cK iu-f ing Tsa-chi, 1/8-9.
44 MSL, Shih-tsung Shih-lu, 4046.
45 MSL, Shih Tsung Shih-lu, 2374; CSWP, 198/22-3.
46 See MSL, Shih-tsung Shih-lu, 5824, 8092, 8222.
47 Hucker, Censorial System, 86-7.
48 Ho Shih-chin, CKang-k'u Hsii-chih, 2/9.
49 Ni Hui-ting, Nien-p^u, 4/13; Ni Yiian-lu, CKilan-chi, Memorials, 9/5.
50 Sun Ch'eng-tse described his personal difficulties as a magistrate in Meng-yii-lu,
35/12, 36/56.
51 THCKLPS, 4/32-5, 32/5, 44.
52 MS, 84/883-90.
53 MS, 85/849-902; THCKLPS, 15/43; MSL, Shen-tsung Shih-lu, 0056, 0845, 1057.
54 MSL, Shen-tsung Shih-lu, 1239; CSWP, 375/21, 376/31, 378/30.
55 MSL, Shen-tsung Shih-lu, 1559, 1647; CSWP, 375/12.
56 CSWP, 375/9.
57 MSL, Shen-tsung Shih-lu, 1188, 1662; CSWP, 375/11.
58 MSL, Shen-tsung Shih-lu, 1651; CSWP, 375/10.
59 CSWP, 376/10-13.
60 MSL, Shih-tsung Shih-lu, 2106.
61 MS, 83/879; MSL, Shen-tsung Shih-lu, 0845.
62 For a summary of the memorial see MSL, Shen-tsung Shih-lu, 2862-3.
63 MSL, Shih-tsung Shih-lu, 7604; Sun Ch'eng-tse, Meng-yii-lu, 6/9; Chao I, Nien-
erh-shih Cha-chi, 32/687.
64 Hsiang Meng-yiian, Tung-kuan Chi-shih, 3.
65 Yang, 'Economic Aspects of Public Works', pp. 194-5.
66 MSL, Shen-tsung Shih-lu, 2957.
67 Hsiang Meng-yuan, Tung-kuan Chi-shih, 3.
68 MSL, Shih-tsung Shih-lu, 7708.
69 MSL, Shen-tsung Shih-lu, 2787.
70 Hui-chou Fu-chih, 8/15-16.
71 This is based on the statement in Sun Ch'eng-tse's Meng-yii-lu, 46/63, the average
being calculated from the general estimate.
72 Hsiang Meng-yiian, Tung-kuan Chi-shih, 4.
73 Kuei Yu-kuang, CWiian-chi, 449, Pieh-chi, 6/8. He complained in 1562 that the
traffic virtually came to a standstill and that his boat as a result 'barely got out of
Chang-chia-wan'.
74 Ricci, China in the Sixteenth Century, pp. 306-7.
75 TMHT, 190/2-3; Ho Shih-chih, CKang-k'u Hsii-chih, 4/39.
76 Hsiang Meng-yiian, Tung-kuan Chi-shih, 1-2.
77 Ricci, China in the Sixteenth Century, 307.
78 Hsiang-Meng-yiian, Tung-kuan Chi-shih, 2; Ho Chung-shih, Ting-chien-chi, 12.
362 Notes to pages 284-90
79 MS, Shih-tsung Shih-lu, 1131.
80 MS, 78/829.
81 MSL, Shen-tsung Shih-lu, 4933.
82 MSL, Hsiao'tsung Shih-lu, 3418; Wei Huan, Chiu-pien-k'ao, 1/18-19.
83 TMHT, 41/16, 24; MSL, Shen-tsung Shih-lu, 1986, 5317, 5941, 6543, 11266; MS,
222/2559; Ming-cKen Tsou-i, 35/673-6; see also chapter 6, note 214 above.
84 On the establishment of the Ku-yuan Command see Wei Huan, Chiu-pien-k'ao
10/1; Sun Ch'eng-tse, Meng-yu-lu, 42/17, 19.
85 Wei Huan, Chiu-pien k'ao, 10/1.
86 The nine defense areas were Liao-tung, Chi-chou, Hsiian-fu, Ta-t'ung, Shan-si,
Yen-sui, Ning-hsia, Ku-yuan, and Kan-su: see Hucker, 'Governmental Organi-
zation', p. 63. The other five army posts were: Yung-p'ing, Mi-yiin, Ch'ang-p'ing,
I-chou and Ching-hsing. In 1576 the fiscal accounts listed all these fourteen units in
equal terms: see Shen-tsung Shih-lu, 1162-82.
87 See Yii Tzu-chiin's biography in Chaio Hung, Hsien-cheng-lu, 38/69. On fifteenth-
century construction work see MSL, Hsien-tsung Shih-lu, 2110, 3491, Hsiao-tsung
Shih-lu, 0523.
88 Calculated from a summary statement in MSL, Shih-tsung Shih-lu, 5800.
89 Calculated from MSL, Shih-tsung Shih-lu, 7840.
90 TMHT, 193/1, 3-4.
91 TMHT, 193/5.
92 MSL, Shen-tsung Shih-lu, 3249-53.
93 See MS, 212/2462, 2466.
94 Ho Shih-chin, CWang-k'u Hsii-chih, 8/84.
95 Wang Yii-ch'iian, Chun-tun, pp. 215-16.
96 Ibid. p. 211.
97 CSWP, 198/19.
98 CSWP, 358/21, 24, 359/3, 360/10.
99 Chang Chii-cheng, Shu-tu, 3/15.
100 The rents varied from 0.015 taels of silver per mou to 0.03 taels per mou: see
Wan-li K'uai-chi-lu, 23/7, 22.
101 CSWP, 358/24. This also went on in Kan-su, Ning-hsia, and even in the area near
Nanking: see Wang Yii-ch'uan, Chun-tun, pp. 52-4.
102 Ying-fien was practiced in Liao-tung, Shan-si, and Ning-hsia, certainly in 1591:
ibid. pp. 4-5.
103 See CSWP, 358/18; Wang Yii-ch'iian, Chun-tun, pp. 8-9.
104 CSWP, 358/10, 12, 14, 23, 359/16, 360/17.
105 Wang I-o, Tsou-i, 1/15-16.
106 MSL, Shen-tsung Shih-lu, 0379, 0467, 1162-82, 2152; Chang Chii-cheng, Shu-tu,
2/3, 9.
107 TMHT, 28/26-53. The sums have been calculated from the unedited data.
108 Several gazetteers suggest that this was a general practice: see for example Fen-
chou Fu-chih, 5/45-6; Chang-chou Fu-chih, 28/18; Szechuan Tsung-chih, 2/6-13;
Teng-chou Chih, 9/1. In some cases army officers collected the rent from this land.
109 Chin-hua Fu-chih, 21/5.
110 Shun-te Hsien-chih, 3/12-14. This was also a common practice. Technically speak-
ing the hereditary military household was in fact the landowner.
111 Best illustrated by an early seventeenth-century document. See MSL, Hsi-tsung
Shih-lu, 1557-60. All these must have been long-term developments however.
112 TMHT, 18/1-8. This is a rough estimate from the unedited data.
113 MSL, Mu-tsung Shih-lu, 0902; CSWP, 322/15.
114 Hucker, * Governmental Organization', p. 63, nl43 indicates a total strength of
553,363 men in the nine defense areas, on the basis of TMHT, chs. 129-30. By
comparison the ministry of revenue in 1620 stated that there were 867,946 men
serving on the frontier: see MSL, Kuang-tsung Shih-lu, 0047.
115 MSL, Shen-tsung Shih-lu, 1162-82.
116 TMHT, chs. 129-30.
 Notes to pages 290-7 363
117 MSL, Shen-tsung Shih-lu, 5143.
118 Chang Chii-cheng, Shu-tu, 2/3.
119 Sun Ch'eng-tse, Meng-yii-lu, 35/28.
120 See MSL, Shen-tsung Shih-lu, 2853, 3484, 4331. However MS, 224/2584 gives the
total as 3.61 million taels.
121 See Li Kuang-ming, Chu-k'e-chiin-k'ao, passim.
122 THCKLPS, 22/27, 33, 35, 33/109.
123 MSL, Shih-tsung Shih-lu, 3237-8, 7241-2.
124 Chin-hua Fu-chih, 8/13; THCKLPS, 33/109.
125 Chang-chou Fu-chih, 5/51-3; Ku'ai-chi Chih, 6/3-4; Hang-chou Fu-chih, 31/16-17;
Shun-te Hsien-chih, 3/22; THCKLPS, 22/36, 23/62, 76, 26/94, 28/8.
126 Liang Fang-chung, 'Ming-tai ti Min-ping', pp. 225, 231.
127 On the recruiting of soldiers for the Korean campaign, see MSL, Shen-tsung Shih-
lu, 4683, 5791, 5809, 5825, 5976, 6332.
128 THCKLPS, 22/27-8.
129 MSL, Mu-tsung Shih-lu, 1250.
130 MSL, Shen-tsung Shih-lu, 0509, 2504.
131 MSL, Shen-tsung Shih-lu, 0983.
132 See Chang Hsiieh-yen's memorial presenting the manuscript of the Wan-li
K'uai-chi-li to the Wan-li Emperor (printed as a preface to the work, i).
133 The new vault also held deposits of 4 million taels, some of the bullion being stored
underground. As it handled current accounts, not all its deposits could be con-
sidered as treasury reserves, however: see MSL, Shen-tsung Shih-lu, 3318, 3329.
The treasury reserve was steadily built up during the decade 1572-82: see MSL,
Shen-tsung Shih-lu, 0086, 0195, 0256, 0308, 0554, 0891, 1027, 1396, 1793, 1831,
2684, 2884.
134 MSL, Shen-tsung Shih-lu, 1503.
135 MSL, Shen-tsung Shih-lu, 5312.
136 The statement is quoted by Ku Yen-wu, in JCL, 5/5-6. It also indicates that the
provincial treasury at Cheng-tu alone had deposits of 8 million taels, which seems
unlikely. The smaller sums quoted seem more probable. It should be noted that the
author's statements were based entirely on casual conversation and memory.
137 See MS, 213/2479-82; CSWP, 324/1-328/31; Chiao Hung, Hsien-cheng-lu,
17/60-108. Many aspects of Chang's administration are documented only in the
Shih-lu and his correspondence. Chu Tung-yun's Chang Chu-cheng Ta-chuan
contains many technical errors on fiscal matters, notably that on the returns of the
land survey of 1580-1.
138 The cutback in the services provided by the postal system likewise sometimes
permitted a reduction of payments by the population: see MSL, Shen-tsung Shih-
lu, 1408,1448. But in general any savings were retained by the provincial treasuries:
see ibid. 1558-9, 1581, 5243, 5992.
139 For reduction of frontier patrols, see Wang Shih-ch'i, San-yun CKou-tsu-k^ao,
2/19-21, 4/1.
140 In 1579 it was claimed that this order could not be entirely carried out: see MSL,
Shen-tsung Shih-lu, 1830-1; also ibid. 0381, 0794, 1594, 2152, 2167.
141 MSL, Shen-tsung Shih-lu, 2435, 2668.
142 MSL, Shen-tsung Shih-lu, 0176.
143 For these simplified accounts see Chang Chii-cheng, Shu-tu, 2/3.
144 MSL, Shen-tsung Shih-lu, 1852-3.
145 Chang Chu-cheng, Shu-tu, 2/8, 6/2; CSWP, 328/26.
146 On the preparation and completion of the K'uai-chi-lu, see MSL, Shen-tsung Shih-
lu, 1076, 2132, 2261, and the preface of the work.
147 MSL, Shen-tsung Shih-lu, 1100, 2128.
148 MS, 300/3367; MSL, Shen-tsung Shih-lu, 1495.
149 Chang Chii-cheng, Shu-tu, 4/18.
150 Wang's statement appears in Chiao Hung, Hsien-cheng-lu, 17/96,
151 See MS, 213/2178, 222/2563, 2569.
364 Notes to pages 297-302
152 See Chang Chii-cheng, Shu-tu, 1/2, 10, 2/27, 5/6. At one point he claimed that the
provincial officials of Kwantung and Kwangsi offered him more than 10,000 taels
of silver in gifts, which he declined.
153 Tao-chi Chou, 'Li-shih, Imperial Consort', biography in the forthcoming publi-
cation of the Ming Biographical History Project.
154 On his use of secret police, see MS, 213/2480. His methods are largely revealed by
his personal correspondence, which was edited by his son, Chang Mou-hsiu, some
time after 1610.
155 Chang Chii-cheng, Shu-tu, 2/23.
156 HYWCL, 92/19.
157 See my article 'Ni Yuan-lu: "Realism" in a Neo-Confucian Statesman', in de
Bary, Self and Society in Ming Thought, pp. 415-48.
158 It appears that Chang discussed very few of his ideas in writing. Robert Crawford,
who has thoroughly examined his collected works, also comments on the ' paucity
of data': see Crawford, 'Confucian Legalism', in de Bary, Self and Society in Ming
Thought, pp. 367-411.
159 Chang Chii-cheng, Shu-tu, 5/1; CSWP, 327/2, 328/14.
160 CSWP, 326/6; Crawford, 'Confucian Legalism', pp. 373, 393.
161 Chiao Hung, Hsien-cheng-lu, 17/71, 75, 94.
162 MS, 78/827. Cf. Wada, Shokkashi yakuchu, p. 218; Sun Ch'eng-tse, Meng-yu-lu,
35/37.
163 MSL, Shen-tsung Shih-lu, 3755.
164 MSL, Mu-tsung Shih-lu, 1200-1.
165 For Kuang Mou's memorial and the rescript, see MSL, Shen-tsung Shih-lu, 1490.
For attacks on the Single Whip Reform see ibid. 1095, 1100, 1112, 1245, 1338.
166 MSL, Shen-tsung Shih-lu, 1490; Chang Chii-cheng, Shu-tu, 4/5.
167 Ibid. Aj\.
168 Ibid. 5121.
169 MSL, Shen-tsung Shih-lu, 0123.
170 Chu Tung-yiin, Chang Chii-cheng Ta-chuan, p. 279.
171 MSL, Shen-tsung Shih-lu, 1732, 2031; Ning-te Hsien-chih, 2/4.
172 MSL, Shen-tsung Shih-lu, 2050.
173 I-chou Chih, 3/1, 16; Huai-yuan Hsien-chih, 2/4; THCKLPS, 13/72.
174 MS, 77/819; Wada, Shokkashi yakuchu, pp. 67-9.
175 MSL, Shen-tsung Shih-lu, 2289, 2371-2.
176 Wen-shang Hsien-chih, 4/2.
177 K'ai-hua Hsien-chih, 3/3.
178 As quoted by Ku Yen-wu, JCL, 3/64.
179 MSL, Shen-tsung Shih-lu, 2378.
180 MSL, Shen-tsung Shih-lu, 2530, 2732.
181 THCKLPS, 13/64; Wen-shang Hsien-chih, 4/2.
182 Chang Mou-hsiu, in his note on his father's letter to the general, Ch'i Chi-kuang,
indicated that Chang was suspected of plotting to seize the throne, apparently
because messengers were observed going between his house and that of Ch'i
Chi-kuang at night. See Chang Chii-cheng, Shu-tu, 5/19. According to Wang
Shih-cheng some of Chang's followers actually tried to persuade him to take over:
see Chiao Hung, Hsien-cheng-lu, 17/92.
183 Chou Hsiian-wei, Ching-lin Hsii-chi, 30.
184 MSL, Shen-tsung Shih-lu, 3318, 3329, 4084.
185 See MSL, Shen-tsung Shih-lu, 4626, 5312, 5917, 5918, 6324, 6349, 6452; JCL, 5/6;
Chu Kuo-chen, Yung cHuang Hsiao-pHn, 2/41. By the early years of the seventeenth
century practically all the reserves were exhausted: see MSL, Shen-tsung Shih-lu,
7217, 8271.
186 MSL, Shen-tsung Shih-lu, 4722.
187 MSL, Shen-tsung Shih-lu, 5991.
188 MSL, Shen-tsung Shih-lu, 6331.
189 MSL, Shen-tsung Shih-lu, 5883, 6390.
 Notes to pages 302-12 365
190 For the construction of the mausoleum, see MSL, Shen-tsung Shih-lu, 2841, 2851,
3795-7.
191 Chung-kuo K'o-hsiieh-yuan, K'ao-ku ti Shou-huo, pp. cxxix-cxxx; the New York
Times, 21 May 1971.
192 See MSL, Shen-tsung Shih-lu, 5591, 5646, 6163. According to another source the
government took 40 per cent of the yield: see CSWP, 441/21.
193 MS, 305/3429-32; Ku Ying-t'ai, Ming-shih Chi-shih Pen-mo, 65/699-712; CSWP,
441/19-22.
194 The income from the estates belonging to the Tzu-ning, Tzu-ch'ing and Chien-
ch'ing palaces is given in MS, 82/866. Also refer to Appendix A above.
195 MSL, Shen-tsung Shih-lu, 0822, 2911, also see Chou, note 153 above.
196 MSL, Shen-tsung Shih-lu, 4663.
197 MSL, Shen-tsung Shih-lu, 0520, 1844.
198 MSL, Shen-tsung Shih-lu, 2945.
199 The exact amount of silver in the reserves and the sums transferred are not
entirely clear. Some of the transfer orders given in the Shih-lu seem to be duplicated
or overlapping: see MSL, Kuang-tsung Shih-lu, 0024,0032,0173; Hsi-tsung Shih-lu,
0052, 0211, 0231, 0242, 0418, 0767, 0773, 2415.
200 MS, 235/2690 serves as an example.
201 For more accurate accounts see TMHT, 201/7; Ho Shih-chin, CWang-k'u Hsii-
chih, 6/66; MSL, Shen-tsung Shih-lu, 0951, 0956, 2926.
202 MSL, Shen-tsung Shih-lu, 0951. For other orders and costs, see ibid. 1111, 2956,
3165.
203 MS, 21/143; MSL, Shen-tsung Shih-lu, 6192.
204 The relevant publications are: Li Kuang-pi, Ming-cKao Shih-lueh, p. 135; Kung
Hua-lung, 'Ming-tai Ts'ai-k'uang ti Fa-ta ho Liu-tu', in Pao Tseng-p'eng, ed.
Ming-tai Ching-chi, p. 121 (the article is reprinted from Shih-ho Fortnightly 1:2
(1954)); P'eng Hsin-wei, Ho-pi-shih, p. 463.

8 Concluding observations
1 Yii-chuan Wang, 'The Rise of Land Tax', p. 201.
2 Ibid. p. 202.
3 In 1641 the annual wartime surtax quota was 21,330,735 taels. See Sun Ch'eng-tse,
Meng-yii-lu, 35/17. He obtained his figure from the ministry of revenue.
4 The figures have been calculated from the provincial quotas listed in Ch'en Jen-hsi,
Shih-fa-lu, 34/1-78.
5 On seventeenth-century prices, see Yeh Meng-chu, Yiieh-shih-pien', P'eng Hsin-wei,
Ho-pi-shih, pp. 459-61. Cf. Ch'iian Han-sheng,' Ming tai Pei-pien Mi-liang Chia-ke
ti Pien-tung', pp. 49-87. The extremely high prices given in the last work were the
result of an abnormal situation.
6 Yu-ch'iian Wang, 'The Rise of Land Tax', p. 202.
7 See CKing Shen-tsu Shih-lu, 4/9; CKing Shih-tsu Shih-lu, 61/6-7, 70/31-2, 79/23-4;
Chang-sha Fu-chih, 7/2; Yeh Meng-chu, Yueh-shih-pien, 6/2.
8 CHung-chen Ts'un-shih Su-cKao, 2/72-89.
9 Creel, 'The Beginning of Bureaucracy', p. 155.
10 Hucker, Traditional Chinese State, p. 43.
11 A translation of Ku's arguments appears in de Bary, ed. Sources of Chinese Tradi-
tion, pp. 611-12. See also Yang, 'Ming Local Administration', p. 4.
12 See Hucker, 'The Tung-lin Movement of the Late Ming Period', in John K.
Fairbank, ed. Chinese Thought and Institutions, p. 133; also Hucker, Censorial
System, p. 153.
13 Hai Jui, Hai Jui Chi, 24.
14 Ibid. 25.
15 MS, 226/2602.
16 Hai Jui, Hai Jui Chi, 50.
17 These letters appear in ibid. 441, 447, 448, 449, 467, 469, 541. In most cases the
366 Notes to pages 312-23
amounts of the gifts are not disclosed, but it is mentioned that a military circuit
intendant once provided him with a boat for travel. The gift of another military
circuit intendant was large enough to enable him to purchase 'several mou of land'.
Gifts returned are mentioned in ibid. 417, 465, 474.
18 Ibid. 25.
19 Ho Liang-chun is criticized on these grounds. See Wu's preface in Hai Jui Chi, 9-10.
20 Ibid. 279-80.
21 Ibid. 190, 195, 280, 284, 289, 408.
22 Smith, 'The Land Tax in the Tokugawa Period', p. 204.
23 Wang Shih-cheng attributed Wan-li's greed to his envy of the extravagance of his
courtiers and the eunuchs. See Chiao Hung, Hsien-cheng-lu, 17/104; MS, 213/2482.
24 Liu, Reform in Sung China, pp. 4, 6, 49, 56-7; Hartwell, 'Financial Expertise',
p. 298.
25 Ibid. p. 298; Twitchett, 'Salt Commissioners', pp. 60-89.
26 Sun Ch'eng-tse, Meng-yii-lu, 25/29.
27 Lien-sheng Yang states: 'Although the use of draft or hui-p'iao existed already
in the seventeenth century, the oldest Shansi 'draft bank' cannot be traced back
earlier than 1800.' Yang, Money and Credit, p. 82.
28 See my doctoral dissertation 'The Grand Canal During the Ming Dynasty'
(University of Michigan, 1964; also available in University Microfilm), pp. 21-37.
29 THCKLPS, 11/16, 12/64, 68, 71, 73, 74, 79.
30 Makida, Sakugen nyuminki, I, pp. 244-7.
31 PreVost, Histoire Generate de Voyages, v, pp. 347-9. Ch'i Piao-chia, when travelling
on the Grand Canal, observed that the water was only one foot deep: see Ch'i,
Jih-chi, vol. 5, fourth to sixth days, ninth lunar month, 1643.
32 See Ch'i Ch'en-yeh, Nanking CKe-chia-ssu Chih-chang.
33 See Ricci, China in the Sixteenth Century, p. 307, 358.
34 Hai Jui, Hai Jui Chi, 155, 167-8.
35 Ibid. 112; MSL, Hsi-tsung Shih-lu, 0705, 0761.
36 Feuerwerker, 'From "Feudalism" to "Capitalism"', pp. 107-15.
37 Crawcour,' Changes in Japanese Commerce in the Tokugawa Period', pp. 169-202.
38 Ch'en Shih-ch'i, Kuan-shou-kung-yeh, passim; Wu Han, 'She-hui Sheng-ch'an-li',
pp. 70-1; Feuerwerker, 'From "Feudalism" to "Capitalism"', p. 108.
39 Hsi and Chu, Ts'ao-cK uan-chih, 6/44-8.
40 Ibid, passim, MSL, Wu-tsung Shih-lu, 0081; CSWP, 244/16-18.
41 Ch'i Chi-kuang, Lien-ping Shih-chi, 175, 176, 182, 199, 210; CSWP, 347/17, 21.
42 MS, 205/2378-9.
43 Kuei Yu-kuang, Cttuan-chi, 95; Ku Ying-t'ai, Chi-shih Peng-mo, 55/597.
44 See Feuerwerker's summary, in 'From "Feudalism" to "Capitalism"', p. 111.
45 The translated passages appear in de Bary, Sources of Chinese Tradition, i, pp. 530-
42, 556-7.
46 CKing Shen-tsu Shih-lu, 3/3; Yeh Meng-chu, Yueh-shih-pien, 6/1-10.
47 Hummel, Eminent Chinese of the CKing Period, p. 917; Ch'ii, Local Government,
pp. 131-2.
48 For these features see Sun, 'The Board of Revenue', pp. 175-228; Ch'en Kung-lu,
Chung-kuo Chin-tai-shih, pp. 238-9, 665-6, 678-9.
49 Wang, 'Fiscal Importance of the Land Tax', p. 842.
50 Lach, Asia in the Making of Europe, i, pp. 734-7.
 Bibliography

The abundance of source materials on this subject represents a serious problem to the
historian. The Ming-shih lists 1,525 historical works by Ming scholars and in addition
1,398 works classified as collections of private writings. Most of these works are in more
than one volume and, obviously, even the most industrious scholar can read only a small
fraction of what is available.
The primary sources used here are given in the list of abbreviated titles (p. 324) and the
local gazetteers. The chronological framework is provided by the Ming Shih-lu, the
modern reprint of which consists of 133 volumes. These annals of the thirteen emperors
are not edited in any consistent style; from the financial historian's point of view, those
for the later reigns are in general of better quality than those for the earlier reigns. The
Shih-lu does not note all matters of financial significance, its primary purpose being to
provide historical lessons in moral government. State documents are seldom reproduced
in full and an important memorial can sometimes be reduced to a single descriptive
sentence. Such statements, quoted out of context, can be misleading. Whenever possible
the original document must be located or the consistency of the statement checked
against other accounts.
Both the Ta-Ming Hui-tien and the Ming-shih have been in part reconstructed from
the Shih-lu. The former also reproduces many important accounts from the Wan-li
K'uai-chi-lu, compiled in 1582. Though the financial sections of the Ming-shih cite many
other earlier works it is never indicated whether a passage is quoted directly; normally it
is trimmed down and many details are omitted. Those confusions have been largely
eliminated by Wada Sei's annotated translation, Minshi shokkashi yakuchu, which is
much richer in detail and cross references than the original text.
The Huang-Ming Ching-shih Wen-pien, in thirty volumes, contains writings on
administration by 425 Ming authors, most of them being government officials, and about
two-thirds of them living in the sixteenth century. Their memorials and letters are
reproduced verbatim.
T'ien-hsia Chiin-kuo Li-ping-shu is a collection of passages copied from local gazetteers
and many other miscellaneous works. Most of these passages, varying in length from
a short paragraph to more than ten pages, reflect conditions in the late sixteenth and
early seventeenth centuries. The quality of the selection is admirable, even though the
work as a whole lacks organization.
Local gazetteers of Ming date are extremely varied in length and contents. Wealthy
counties were able to compile them more often and at greater length; more detail on
local conditions and customs could be included, and also official documents relating to
local government. Economically-depressed districts usually produced poor gazetteers in
which the vital fiscal statistics, presented without accompanying notes, are virtually
useless.
At present it is impossible to compile a comprehensive bibliography on Ming financial
history, as most Ming writers refused to treat fiscal administration as a specialized topic,
preferring to discuss it along with a wide variety of other matters under the general
heading of ching-chi, which really means 'management'. As a result the source material
[367]
368 Bibliography
is extremely diffuse and the present bibliography could, in theory, be enormously ex-
panded. For the sake of economy, however, the selection has been limited to works
directly cited in the text and notes, references being given wherever possible to modern
reprinted editions.
The reader is reminded that several bibliographical guides to Ming historical source
materials are now available. Extremely useful is Wolfgang Franke's An Introduction to
the Sources of Ming History (New York, 1969) which contains over 800 titles, with
evaluations and cross references by the author. Yamane Yukio's bibliographical list,
Mindaishi kenkyu bunken mokuroku (see below), contains 5,128 items in Chinese and
Japanese. The same author has also compiled a list of 680 contemporary Chinese and
Japanese publications on Ming social and economic history and appended them to
Shimizu hakase tsuito kinen Mindaishi ronso. Francis D. M. Dow has published a useful
work entitled A Study of Chiang-su and Che-chiang Gazetteers of the Ming Dynasty
(Canberra, Australia, 1969). The Ming biographical index published by the National
Central Library, Ming-jen Chuan-chi Tzu-liao So-yin (Taipei, 1965), also includes a list
of 593 Ming and early Ch'ing works.

Provincial and local gazetteers

An-hua Hsien-chih £ it U M, 1543 edition.
Chang-chou Fu-chih ft JM Iff M, 1573 edition.
Chang-sha Fu-chih ^ ?p Jff ;g, 1747 edition.
CKang-shu Hsien-chih % m % ;g, 1534 edition.
Chin-hua Fu-chih & ^ Jff ;g, 1578 edition.
Chung-mou Hsien-chih 4* $ Wk M, 1626 edition.
Fen-chou Fu-chih %ft jM Jff M, 1606 edition.
Hang-chou Fu-chih ia 'M Iff ;g, 1579 edition.
Honan Tung-chih S S I f e 1678 edition.
Hsiang-ho Hsien-chih f| J5f U ;g, 1620 edition.
Huai-an Fu-chih i ^ f M, 1573 and 1627 editions.
Huai-jou Hsien-chih ft ^ U ;§, 1625 edition.
Huai-yiian Hsien-chih -R jg J$ jg, 1605 edition.
Hui-chou Fu-chih % jf\ Jff M, 1566 edition.
I-chou Chih ?f J)M ^ , 1608 edition.
Ku-an Hsien-chih | $ l f e 1633 edition.
Ku-su Chih ift H i&9 1506 edition.
K'ai-hua Hsien-chih gg {b H ;£, 1585 edition.
K'un-shan Hsien-chih % \U Wh M, 1576 edition.
Ktfai-chi Chih # ff •£, hand-copy of 1572 edition.
Lin-cKing Chihli Chou-chih W» 'M it It ffl ^ , 1782 edition.
Lin-fen Hsien-chih ^ ^ ^ ^ , 1591 edition.
Lu-cKeng Hsien-chihffit$ $$ ;g, 1625 edition.
Nan-chi Chih ^f ^ ^ , 1540 edition. 7Ve/-A:o Wenk'u faMXW- copy.
Ning-te Hsien-chih $ ^ ^ ;£, 1591 edition.
P'eng-tse Hsien-chih &WM -g, 1582 edition.
Shanghai Hsien-chih ± m U M, 1590 edition.
Shun-te Hsien-chih )l@ ^ jR ^ , 1585 edition.
Shun-fien Fu-chih jig ^ Jff ^g, 1885 edition.
5m-a/i Hsien-chih & % % ^ , 1612 edition.
Sung-chiang Fu-chih 1& %L Jff ^ , 1818 edition.
 Bibliography 369
Szechuan Tsung-chih E9 HI $£ & 1580 edition.
Teng-chou Chih tp ffl * , 1644 edition.
Tung-cKang Fu-chih M H* Jff M, 1600 edition.
Wen-shang Hsien-chih #C ± f£ ;g, 1609 edition.
Wu-hsien Chih j% H •£, 1643 edition.
Wu-chiang Hsien-chih jg ££ $6 ^ , 1561 edition.
Yueh-chou Fu-chih § JW Iff ;&, 1570 edition.
Yung-chou Fu-chih T]C j \ \ Iff <£> 1571 edition.

NOTE: Only those gazetteers cited in the text are listed. The date of publication is
generally taken as that of the latest preface.

Other Ming and early CKing sources

(Ts'ung-shu Chi-cKeng | j H % JS editions are identified as TSCC.)
Chang Chii-cheng 3ft g IE, Chang Chiang-ling Shu-tu gg ft g£ # Ijf. Ch'iin-hsueh
Shu-she ^ ^ g fi reprint. Shanghai, 1917.
Chang Hsueh 5g §$, Tung-hsi-yang K'ao M ffi #- ^ . TOCC.
Chang Hsiieh-yen %^M, ed. Wan-UK''uai-chi-lu S i t ft ®. University of Chicago
microfilm, preface dated 1582.
Chang Yu 55 M> Pien-cheng-k'ao ^# iE^ 3f. National Peking Library reprint. 1936.
CKang-lu Yen-fa-chih g 31 §g ^ -g. 1726 edition.
Chao I f i l , Nien-erh-shi Cha-chi iJ* n A SI IB. r£CC.
Ch'en Jen-hsi ^ -t £§> Huang-Ming Shih-fa-lu H HJ tit j£ ®. Hsueh-sheng Shu-chu
* ^ • ^ reprint. Taipei, 1965.
Cheng Hsiao i[5 ^ , Cheng-tuan-chien-kung Chin-yin Lei-pien M ^S ffi <2> ^ B" II S .
race.
Ch'eng K'ai-hu II H jjfe, ed. Chou-Liao Shih-hua $ ^ Jg i | . 1620 edition.
Ch'i Chi-kuang J^ f£ 5fe, Lien-ping Shih-chi W.^ 'S ffi. Commercial Press reprint.
Shanghai, 1937.
Ch'i Piao-chia f[5 jg S , CKi-chung-min-kung Jih-chi 15 ^ S <2r H IB. Shao-hsing,
Chekiang, 1937.
Chiang P'ing-chieh ^ ^p pg, Pi-shao-pao-kung Chuan # ^ ^ £ <$. Early Ch'ing
edition.
Chiao Hung ^ ft, Kuo-cKao Hsien-cheng-lu M ^ Wi Wl M- Hsiieh-sheng Shu-chu
reprint. Taipei, 1965.
Ch'ing Sheng-tsu Shih-lu Sf !g IB Jf %k- Manchu Kuo-wu-yuan M MMBU edition.
1937.
Ch'ing Shih-tsu Shih-lu ffi i& jf§ 'g g . Manchu Kuo-wu-yiian edition. 1937.
Ch'iu Chiin ® ^ , Ta-hsueh Yen-i-pu ^ i f i i i f . Princeton University rare edition.
Chou Chih-lung J! ;£, f| , Ts'ao-ho l-pi if M — £5 • Library of Congress microfilm of
original 1609 edition.
Chou Hsiian-wei | g | j , Ching-lin Hsii-chi M W IB #E- Commercial Press Han-fen-lu
Pi-chi ffi&mffl% edition.
Chu Kuo-chen ^ m fi|, Yung-cKuang Hsiao-p'in M d /J^ no- Chung-hua Shu-chu
4« $ $ ^ reprint. 1959.
Chu T'ing-li ^ g A/:, Yen-cheng-chih H igr ^ . 1529 edition.
370 Bibliography
CKung-chen Ts'un-shih Su-cKao |ft jpji # % |jg £I>. Reprint. Peking, 1934.
Feng Ch'i #§ f£, Feng-Uung-po Chi M ^ ffi ft. 1607 edition.
Hai Jui m 3S, Hai Jui Chi M^M- Peking, 1962.
Ho Chung-shin g # <ft, Liang-kung Ting-chien-chi M K M ^ IS. 7SCC.
Ho Liang-chim M ^ ^ , Ssu-yu-chai Tsung-shuo Ti-cKao R9 £ if H 19; JS
Ho Shih-chin M ± if, Kung-pu CKang-k'u Hsu-chih X §[5 Jg£ }$ gfft],Hsiian-lan-fang
Ts'ung-shu ^ I f S f edition.
Ho-tung Yen-fa-chih g f l S Jg. 1727 edition.
Hsi Shu and Chu Chia-hsiang Jg f£ ^ ^ l i , Ts'ao-cKuan-chih fg fpL j£. Hsuan-lan-fang
Ts^ung-shu edition.
Hsiang Meng-yiian J | ^ H, Tung-kuan Chi-shih ^ 'g
Hsieh Shang-chih H ^ ®, CKang-shu Shui-lun % ^
Hsu Cheng-ming ^ j | BJ, L«-JA«I K"e-fan ffi 7X ^ M
fTw Wen-hsien Tung-k'ao II 5t K M ^ . Commercial Press reprint. Shanghai, 1936.
Huang Shim g fl|, Hung-Ming Ming-cKen Ching-chi-lu M W £ ill $2 #¥ ifc. 1551
edition.
Huang-Ming Tsu-shun M ^ |a Hi, in Ming-cKao K'ai-kuo Wen-hsien HJ IB |g §13£ g t
Hsiieh-sheng Shu-chu reprint. Taipei, 1966.
Hui-chou Fu-i CKuan-shu ft i/II K g ^ f | . Library of Congress microfilm of 1620
edition.
I-ling-a & $ pU, Huai-kuan Tung-chih M gg 86 ^ . 1778 edition.
KiangsiFu-i CKuan-shu %L H ® g ^ • • Library of Congress microfilm of 1610 edition.
Ko Shou-li M ^ ft, Ko-tuan-su-kuang Chi g #g I t & 3g. 1802 edition.
^T«-cA/« T'w-^w Chi-cKeng £ 4- H g ^ j ^ . 1934 edition.
Ku Ch'ing |g /f, P'ang-cKiu-fing Tsa-chi $ %k ^f Jl IB. Han-fen-lu Pi-chi edition.
Ku Yen-wu H jfc g|, r/h^-Zw ^n-cW ^ # ^ « . r//^-/wi Fw-c/?/ ^ ^ . 5^«-/w
Ts'ung-k'an m%lM W edition.
Ku Ying-t'ai ^ K ^ , Ming-shih Chi-shih Peng-mo S J 8 I * ^ San-min Shu-chu
H S U reprint. Taipei, 1956.
Kuei Yu-kuang g f ft, Chen-cKuan Hsien-sheng Pieh-chi f j J[[ 5fe ^£ SO M- Ssu-pu
Ts'ung-Wan edition. San-Wu Shui-li-lu H ^ 7X ^J $k> TSCC. Kuei Yu-kuang
CKuan-chi ^ M- Tzu-li Ch'u-pan-she g ^J Mi }K tt reprint. Taipei, 1959.
Li Chao-hsiang $ 03 ;1¥, Lung-chiang CKuan-cKang-chih fi yX jg J^ ;&. Hsiian-lan-
fang Ts'ung-shu edition.
Li Hua-lung $ >fb f|, Fing-Po CKuan -shu 2p jf £ § . TSCC.
Liang-Che Yen-fa-chih W:fift$i& ^ . 1801 edition.
Liang-Huai Yen-fa-chih M ?i IK ?£ & 1748 edition.
Liang-Kuang Yen-fa-chih M K IK ?£ j£. 1835 edition.
Liu Jo-yu SI S S , Cho-chung-chih S 44 ,*. TOCC.
Liu Tsung-chou 10 ^ j i , Liu-tzu CKiian-shu fiJ •? ^ g . Chien-k'un Cheng-cKi Chi
%LttIE a ^ edition.
Lu Jung ^ ^ , Shu-yuan Tsa-chi M M i i IS. TSCC.
Lu Shien-chi |M # j ^ , Jen-chen-ts'ao i g | . TSCC.
Lung Wen-pin f| ^t |^, M/«^ Hui-yao PJ # H. Reprint. Taipei, 1956.
Ming-cKen Tsou-i m E # ^ . TSCC.
Ni Hui-ting {M # F4, Ni-wen-cheng-kung Nien-p'u {S Jt JE <2r ^- f@. Yueh-ya-fang
Ts'ung-shu % g§ S ^ t r edition.
 Bibliography 371
Ni Yiian-lu #3 jt 39$, Ni-wen-cheng-kung CKuan-chi {g £ j | £ ik H. 1772 edition.
Pi Tzu-yen l | § JR, Liu-chi Su-ts'ao §J ft Sfc ^ . Library of Congress microfilm.
Shantung Yen-fa-chih lij ^ S J£ ;£. 1725 edition.
Shen Pang ft; $g, Wan-shu Tsa-chi % % $| IB. Reprint. Peking, 1961.
Shen Te-fu ft; |g ft, Yeh-huo-pien 1? » ffi, P«-i fit S t 1870 edition.
Sun Ch'eng-tse ^ ^ IS, CKun-ming Meng-yti-lu # BJ3 ^ 4& S- Lung-men Shu-chu
II F9 HI lu reproduction of a late Ch'ing edition. Hongkong, 1965.
Sung Ying-hsing 5fc M M, Tien-kung K'ai-wu % X Bg ^ . Commercial Press Jen-jen
wen-k'u A A X H edition. Taipei, 1966. Original preface dated 1637.
T0-A;tf0 ^C it, 7b-A:aa Hsti-pien |g | i , 7^-A:^ San-pien H jg, T«-A:ao Wu-cKen K £ .
All in Ming-cKao K'ai-kuo Wen-hsien.
Ta-Ming Kuan-chih ^c 0J g" $lj. In Ming-cKao K'ai-kuo Wen-hsien,
T'ang Shun-chih gf Jfi ^1, Ching-cKuan Wen-chi ¥&)\\%%.. Ssu-pu Ts'ung-k'an
edition.
Tu Lin ifc # , Huai-an San-kuan Tung-chih | g £ l ,ft ^ . 1686 edition.
Wang Ao 3£ S» Chen-tse CKang-yu M'MMM- 1748 edition.
Wang Ch'iung 3£ S» Hu-pu Tsou-i fi §P # ^ . Library of Congress microfilm.
Wang I-o 3£ — f|, Tsung-tu Ssu-chen Tsou-i fH # RH i t # It. Hsiian-lan-fang Ts'ung-
shu edition.
Wang Shih-ch'i 3E ± ^f, San-yun CK ou-tsu-k" ao H ' H I 3§. National Peking
Library reprint. 1936.
Wang Ying-chiao riE JB !^, Wang-cKing-chien-kung Tsou-su 9E ft? SB & It 88. Library
of Congress microfilm.
Wei Huan §| ^ , Huang-Ming Chiu-pien-k'ao H P^ JL ^S ^ . National Peking Library
reprint. 1936.
Yang Shih-ch'iao ^ Bf S , Huang-Ming Ma-cheng-chi JI $! JH igc IE. Hsiian-lan-fang
Ts'ung-shu edition.
Yeh Meng-chu ^ ^ S , Yueh-shih-pien M 1& M- Reprinted in Ming-CKing Shih-liao
Lei-pien PJ ?g * ^ * | | . Taipei, 1969.
Yeh Sheng M & Shui-tung Jih-chi 7X IC 0 IB. Reproduction from 1680 edition.
Taipei, 1965.
Yen Ts'ung-chien g| tS fW, »SAM-JW Chou-tzu-lu % ig JH g ^|. Chung-hua Wen-shih
Ts'ung-shu t ¥ 5 t A i t edition.
Yu Chi-teng ^ i S , 77£7*-A:H Chi-wen ft ifc *E H-

Contemporary Chinese and Japanese sources

Ano Shozo g f f H, ' Minmatsu Shinsho Yosuko churyuiki daitochi shoyu ni kan-
suru ichikosatsu' BJ! ^ M t l fl ? rX + 3E « ^ ± i t BT ^ Ki BS "t ^5 - ^ ^ ,
Toyogakuho, 44:3 (1961).
Chang Chi-shun 35 K Bt ^ «/• eds. CKing-shih g A (Taipei, 1961).
Ch'en Kung-lu j ^ ^ jjg, Chung-kuo Chin-tai-shih tf* M jft ft ^ . Revised edition
(Taipei, 1965).
Ch'en Shih-ch'i |$C ff Jg, Ming-tai Kuan-shou-kung-yeh ti Yen-chiu BJ§ f^ 'g ^ X H Kj
#F 55 (Wuhan, 1958).
Ch'en Wen-shih ^ ^ 5 , Mm^ Hung-wu Chia-ching Chien-ti Hai-ching Cheng-ts'e
m m a s w ra w tW s ^ m (Taipei, 1966).
372 Bibliography
Chou Liang-hsiao | f i t , ' Ming-tai Su-Sung Ti-ch'ii ti Kuan-t'ien yu Chung-fu
Wen-tV m ft M t& ft H fit ffl H S M B9 £g, M i f t y^/f-c/i/ii, X (1957).
Chu Hsieh & #£, Chung-kuo Hsin-yung Ho-pi Fa-chan-shih 4 » I H t f l ! f i f f i i 5 J g * .
(Chungking, 1943).
Chung-kuo Yun-ho Tzu-liao Hsuan-chi + I I M S S I I (Peking, 1962).
Chu Tung-yun ^ J | , CZra^ Chu-cheng Ta-chuan jjg g IE ^c f# (Wuhan, 1957).
Ch'uan Han-sheng ^ g| # , 'Ming-tai Pei-pien Mi-liang Chia-ke ti Pien-tung' gg ft
itm^mm^^mm Hsin-Ya Hsueh-pao, 9:2 (1970).
Chung-kuo K'o-hsueh-yuan K'ao-ku Yen-chiu-so 41 S f 4 ^ [ ^ ^ ' S ^ F ^ S § f , /fr/tf
Chung-kuo K'ao-ku ti Shou-huo f f ^ H ^ - f i f t H f e U (Peking, 1962).
Fu I-ling ff $c fg, Ming-CKing Nung-ts'un She-hui Ching-chi Pj§ ?jf g ff ft # $f p
(Peking, 1961).
Ming-tai Chiang-nan Shih-min Ching-chi Shih-fan $% jX ?X ffi TfT ^ II S 1^ S
(Shanghai, 1963).
Fujii Hiroshi ^ # ^', 'Mindai ensho no ichikosatsu' K ft H fg O - ^ ^ , Shigaku
zasshi, 54:4-6 (1943).
'Mindai dendo tokei ni kansuru ichikosatsu' f$ ft P9 ± $z I t ^ IS "t" ^ - ^ ^ ,
T ^ J ^aifciito, 30:3 (1943), 30:4 (1944), 31:1 (1947).
Horii Kazuo #j§ # -* 3£, 'Kinkagin no tenkai' i ? E I O S F A Toyoshi kenkyu, 5:2
(1939).
Hoshi Ayao I? ® ^ , Mindai soun no kenkyu PJ§ ft if M © $f ?S (Tokyo, 1963).
Hou Jen-chih & {Z £., 'Ming-tai Hsuan-Ta Shansi San-chen Ma-shih-k'ao' fJfl ft S' ^C
iJj a H i t H TfT ^ , Yenching Journal of Chinese Studies, 23 (1938). Reprinted in
Ming-tai Ching-chi W ft ® jj£ ed. Pao Tseng-p'eng ^ ^ % (Taipei, 1968).
Hsueh Kuo-chen fi ^ H, Ming-ChUng chih-chi Tang-she Yun-tung-k''ao HJ Jf ±. ^
« tt « » ^ (Shanghai, 1935).
Jung Chao-tsu § gE M, Li Chih Nien-p'u ^ gf ^ |f (Peking, 1957).
Kataoka Shibako M* [53 ^ ^ , 'Kahoku no tochi shoyu to ichijobenho' ^ ^fc <D ± iik
§f ^* £• ~ fl$ 1^ ^ , in Shimizu hakase tsuito kinen Mindaishi ronso ^ 7X ft dr ^
^ «E ± m ft * Ife m (Tokyo, 1962).
Katayama Seiichiro j t ill M — ^P, 'Gekko nijyushisho no hanran' ^ M i t RH )f$ O jx.
8L, in Mindaishi ronso.
Koyama Masaaki /J\ i_Lj IE 6S, 'Minmatsu Shinsho no daitochi shoyu ...toku ni
Konan t/e/^ chitai o chushin to shite' PJ ^ ^ ^] o -^ ± i& §f ^"* . •. # ^ K ffi
delta ft ?f ^ ^ i^ 2: L T, 5/w^oA:« z a ^ / , 66:12 (1957), 67:1 (1958).
Kuribayashi Nobuo ^ # S ^ , 'Ichijobenho no keisei ni tsuite' — - { ^ ^ ^ O ^ ^ c ^ i
-2> V^ "C, in Mindaishi ronso.
Li Chien-nung ^ $J JR, Sung-Yuan-Ming Ching-chi Shih-kao % jt 0J IM ^ A fl
(Peking, 1957).
Li Kuang-ming S A PJ, Chia-ching Yu'-Wo Chiang-Che Chu-k?e-chun-k?ao % J | ^ fg
rX 8fr ^ g ¥ ^ (Peking, 1933).
Li Kuang-pi $ 3fe S , Ming-cKao Shih-lueh BJ ^ & S (Wuhan, 1957).
Li Kuang-pi, ed. Ming-CKing-shih Lun-ts'ungfy]$1 fc ffe ^ (Wuhan, 1957).
Liang Fang-chung §£ Jj ft, 'Ming-tai Kuo-chi Mou-i yii Yin-ti Shu-Ch'u-yii' §^ ft m
R H ^ H ^ K l S t t b A , Chung-kuo She-hui Ching-chi-shih Chi-k'an, 6:2
(1939).
Ming-tai Liang-chang Chih-tu BJj ft ft 5 fjfi] g (Shanghai, 1957).
 Bibliography 373
'Ming-tai Hu-k'ou T'ien-ti chi T'ien-fu T'ung-chi' 83 ft j 3 P ffl j& R ffl E St fh
Chung-kuo Chin-tai Ching-chi-shih Yen-chiu Chi-k'an, 3:1 (1935).
'I-t'iao-pien-fa' — jgjE $H ££, Chung-kuo Chin-tai Ching-chi-shih Yen-chiu Chi-k'an,
4:1 (1936).
'Ming-tai Liang-shui Shui-mo' 83 {\ M f£ ^ @ Chung-kuo Chin-tai Ching-chi-shih
Yen-chiu Chi-k'an, 3:1 (1935).
'Ming-tai Shih-tuan-ching-fa' 83 ft + I* $§ *£ Chung-kuo She-hui Ching-chi-shih
Chi-k'an, 7:1 (1944).
'Ming-tai ti Ming-ping' 83 ft KJ J£ J^, Chung-kuo She-hui Ching-chi-shih Chi-k'an,
5:2(1937).
Makida Tairyo ft 0 i IS, Sakugen nyuminki no kenkyu jg § 7^ 83 IB O ^F 2L,
2 vols. (Kyoto, 1955).
Meng Seng £ &, Ming-tai Shih 83 ft 3£ (Taipei, 1957).
Miyazaki Ichisada Hf [Iff ft? ^ , 'Min-Shin jidai no Soshu to keikogyo no hattatsu'
83 m n ft © m m tm^m^mm, Tohogaku, 2 (1951).
Nishijima Sadao |§ g /£ ^ , 'Mindai ni okeru momen no fukyu ni tsuite' 83 ft ^ S^ H"
£ * tH O § R K ^ V^ -C, Shigaku zasshi, 57:4-6 (1948).
'Shina shoki mengyo shijo no kosatsu' £ iP ftf Jffl IS t f t « O ^ f , 7?yJ
^ArwAo, 31:2(1947).
P'an Kuang-tan ?# ^ J3, Ming-CKing Liang-tai Chia-hsing ti Wang-ts'u 0J S M ft H
i W I i (Shanghai, 1947).
P'eng Hsin-wei % \t ^ , Chung-kuo Ho-pi-shih ty M M ffi & (Shanghai, 1954).
People's University, Ming-CKing She-hui Ching-chi Hsing-fai ti Yen-chiu 83 '/f ft # If
» ^ ^ ST % (Shanghai, 1957).
Chung-kuo Tzu-pen Chu-i Meng-ya Wen-fi T'ao-lun-chi 4* M jf ^ ^ S M W ^ ^
It ffe ^ , 2 vols. (Peking, 1957).
Sakuma Shigeo fe ^ ftl] M ^ , 'Mindai kaigai shiboeki no rekishiteki haikei' 83 ft 'M
^ t l i O l i K t f , 5/rfeo^tt zasshi, 62:1 (1953).
'Mindai Keitokuchin yogyo no ichikosatsu' 8! ft ft @ H H H © — ^" ^ , in
Mindaishi ronso.
'Mindai ni okeru shozei to zaisei tono kankei' 83ft^2s^t£F§f&<^Mtifc
t <D M i%, Shigaku zasshi, 65:1-2 (1956).
Shimizu Taiji ^ 7X # 5k, Chugoku kinsei shakai keizaishi 41 p^ ffi |H; jjjj- ^- ^ g^ 3^
(Tokyo, 1950).
'Ming-tai Chun-tun ti Peng-huai' 83 ft W- Tfi W ffi S , trans. Fang Chi-sheng ^7 ffi
^ , in Ming-tai Ching-chi.
'Mindai Fukuken no nokakeizai' 83 ft jpg E O JS ^ II S , Shigaku zasshi, 63:7
(1954).
MI/U&II rocA/ J C / ^ ^ I Ar^Ar^w B J f t ± i f e S I K A ^ F 3 8 (Tokyo, 1968).
Su T'ung-pin | | [H M, Ming-tai I-ti Chih-tu 83 ft W M SO K (Taipei, 1969).
Tani Mitsutaka ^ ^ K, 'Mindai no chuang-p'eng-yin ni tsuite' ^ f t O ^ i ^ i l ^ o
V> "t!, in Mindaishi ronso.
Terada Takanobu ^ ffl |^ if, 'Mindai ni okeru hensho mondai no ichisokumen' PJ3 ft
f e i ^ l t ^ ^ l l S ^ ^ O — #J ®, in Mindaishi ronso.
Ting I T i , Ming-tai Te-wu Cheng-chih 83 ft # $ i& ?& (Peking, 1950).
Wada Sei ffl ffl ?f, Mi/w/w" Shokkashi yakuchu ffljt4f^SS,2 vols. (Tokyo,
1957).
374 Bibliography
Wang Chih-jui £ M SS, Sung-Yuan Ching-chi-shih 5fc jt ffi. 9f A (Taipei, 1964).
Wang Ch'ung-wu 3£ gt R, 'Ming-tai Hu-k'ou ti Hsiao-chang' BJ3 ft J5 • ft #? 5 ,
Yenching Journal of Chinese Studies, 20 (1936).
Wang Yii-ch'iian 3E Sfi i£, Ming-tai ti Chun-tunfljft ft S T£ (Peking, 1965).
Wei Ch'ing-yiian H 8 , Ming-taiHuang-ts*e Chih-tu gg ft 31flffftflJK (Peking, 1961).
Wu Chao-ts'ui ^ ^ ft, Chung-kuo Shui-chih-shih ^ H i& Sfl £ (Shanghai, 1937).
Wu Ch'eng-lo ig ^ fe, Chung-kuo Tu-liang-heng-shih t i S i S j f c (Shanghai,
1937).
Wu Chi-hua ^ ff ^ , Ming-tai Hai-yUn chi Yun-ho ti Yen-chiu BJ ft ?g M R 31 f»J ft #f
3^ (Taipei, 1961).
Wu Han ^ B^. CA« Yuan-chang Chuan & 7U H if (Shanghai, 1949).
'Ming-tai ti Chun Ping' 0J3 ft ft W- S , Chung-kuo She-hui Ching-chi-shih Chi-k'an,
5:2(1937).
'Ming-ts'u She-hui Sheng-ch'an-li ti Fa-chan' BJ ^] f± # fe ^ ^J ft S H, L/-^/A
Yen-chiu, III (1953).
7M-S/ZH C/?a-c/z/ ^ # f ij IS (Peking, 1956).
Ya-tung Hsiieh-she gS ^ $ frh, Chung-kuo Li-tai Jen-k'ou Wen-fi Tao-lun-chi 41 US 81
ft A • PP9 M I t f^ * (Hongkong, 1965).
Yamane Yukio |ii S ^ ^ , Mindai yoeki seido no tenkai Pj§ ft fg ^ SO g O | f |
(Tokyo, 1961).
'MindaiKahokuniokeruekihonotokushitsu'gJ^ ft ^ dt K & ft % & S O # S"
in Mindaishi ronsd; also reprinted in >wA:/ .seiV/0.
Mindaishi kenkyu bunken mokuroku O S f t ^ f i f f ^ X ^ S ^ (Tokyo, 1960).
Yang Tuan-liu 4§ S 7\, CKing-tai Ho-pi Chin-jung Shih-kao ffi ft M f£ ^ Ik A ffl
(Peking, 1962).
Yang Yii-liu ^ f- T1^, Chung-kuo Li-tai Ti-fang Hsing-cheng CKu-hua *f P S ft i& ^
ff ic [M iO, 2nd edition (Taipei, 1957).

References in Western languages

Chang, T'ien-tse. Sino-Portuguese Trade from 1514 to 1644 (Leiden, 1934).
Chou, Tao-chi. 'Li Shih, Imperial Consort', Draft Ming Biography, no. 11 (Ming
Biographical History Project, New York, 1969).
Ch'ii, T'ung-tsu. Local Government in China under the CKing (Cambridge, Mass., 1962).
Crawcour, E. S. 'Changes in Japanese Commerce in the Tokugawa Period', Journal of
Asian Studies, 22:4 (1963). Reprinted in Studies in the Institutional History of Early
Modern Japan, ed. John W. Hall and Marius B. Jansen (Princeton, N J . 1968).
Crawford, Robert B. 'Eunuch Power in the Ming Dynasty', Toung Pao, 49 (1961).
Creel, H. G. 'The Beginnings of Bureaucracy in China: The origin of the Hsien\
Journal of Asian Studies, 23:2 (1964).
de Bary, W. T. ed. Self and Society in Ming Thought (New York, 1971).
de Bary, W. T. et al. eds. Sources of Chinese Tradition. Text edition, 2 vols. (1964).
Fairbank, John K. ed. Chinese Thought and Institutions (Chicago 1957).
Fairbank, John K. and Reischauer, Edwin O. East Asia: The Great Tradition (Boston,
1958).
Feuerwerker, Albert. 'From "Feudalism" to "Capitalism" in Recent Historical
Writing from Mainland China', Journal of Asian Studies, 18:1 (1958).
Friese, Heinz. Das Dienstleistlungs-System der Ming-Zeit, 1368-1644 (Hamburg, 1959).
Grimm, Tilemann.' Das Neiko der Ming-Zeit, von Anfangen bis 1506', Oriens Extremus,
1:1 (1954).
Erziehung und Politik im Konfuzianischen China der Ming-Zeit (Hamburg, 1954).
 Bibliography 375
Hart well, Robert M. 'Financial Expertise, Examinations, and the Formulation of
Economic Policy in Northern Sung China', Journal of Asian Studies, 30:2 (1971).
Ho, Ping-ti. The Ladder of Success in Imperial China (New York, 1962).
'The Salt Merchants of Yangchou: A Study of Commercial Capitalism in Eighteenth-
Century China', Harvard Journal of Asiatic Studies, 17:1-2 (1954).
Studies on the Population of China, 1368-1953 (Cambridge, Mass. 1959).
Hoshi, Ayao. The Ming Tribute Grain System, trans. Mark Elvin (Ann Arbor, 1969).
Huang, Ray. "Fiscal Administration During the Ming Dynasty", in Chinese Government
in Ming Times: Seven Studies, ed. Charles O. Hucker (New York, 1969).
'Military Expenditures in Sixteenth-Centuiy Ming China', Oriens Extremus, 17:1-2
(1970).
Hucker, Charles O. 'Governmental Organization of the Ming Dynasty', Harvard Journal
of Asiatic Studies, 21 (1958). Reprinted in1 Studies of Governmental Institutions in
Chinese History, ed. J. L. Bishop (Cambridge, Mass. 1968).
The Censorial System of Ming China (Stanford, Calif. 1966).
The Traditional Chinese State in Ming Times: 1368-1644 (Tucson, 1961).
Hummel, A. W. ed. Eminent Chinese of the Ch'ing Period. 2 vols. (Washington, D.C
1943-4).
Lach, Donald F. Asia in the Making of Europe, 2 vols. (Chicago, 1965). The sections
concerning Ming China are reprinted in a separate volume entitled China in the Eyes
of Europe: The Sixteenth Century (Chicago, 1965).
Liang, Fang-chung. The Single-whip Method of Taxation in China, trans. Wang Yu-ch'iian
(Cambridge, Mass. 1956).
Liu, James T. C. Reform in Sung China: Wang Anshih and his New Policies (Cambridge,
Mass. 1959).
Lo, Jung-pang. 'The Decline of the Early Ming Navy', Oriens Extremus, 5:2 (1958).
'Policy Formulation and Decision-Making on Issues Respecting Peace and War', in
Chinese Government in Ming Times: Seven Studies.
Morse, H. B. The Chronicles of the East India Company Trading to China: 1635-1843.
4 vols. (Boston, 1931).
Mote, F. W. 'The Growth of Chinese Despotism', Oriens Extremus, 8:1 (1961).
Parsons, James B. The Peasant Rebellions of the Late Ming Dynasty (Tucson, 1970).
'The Ming Bureaucracy: Aspects of Background Forces', in Chinese Government in
Ming Times: Seven Studies.
Prevost, Antoine Francois. Histoire Generate de Voyages, vol. 5 (Paris and Amsterdam,
1776 et ser.).
Ricci, Matthew. China in the Sixteenth Century: The Journals of Matthew Ricci, 1583-
1610, trans. L. J. Gallagher (New York, 1953;.
Rieger, Marianne, 'Zur Finanz und Agrargeschichte der Ming-Dynastie: 1368-1643',
Sinica, 12 (1937).
Rossabi, Morris. 'The Tea and Horse Trade with Inner Asia During the Ming Dynasty',
Journal of Asian History, 4:2 (1970).
Smith, Thomas C. 'The Land Tax in the Tokugawa Period', Journal of Asian Studies,
18:1 (1958). Reprinted in Studies in the Institutional History of Early Modern Japan.
Sun, E-tu Zen. 'The Board of Revenue in Nineteenth Century China', Harvard Journal
of Asiatic Studies, 24 (1962-63).
'Ch'ing Government and Mining Industries Before 1800', Journal of Asian Studies,
27:4(1968).
Sun, E-tu Zen and Francis, John de, trans, and eds. Chinese Social History (Washington,
D.C. 1956).
Twitchett, D. C. Financial Administration under the T"*ang Dynasty (Cambridge,
1963).
'The Salt Commissioners after the Rebellion of An Lu-shan', Asia Major, 4:1 (1954).
Wang, Yeh-chien. 'The Fiscal Importance of the Land Tax During the Ch'ing Period',
Journal of Asian Studies, 30:4 (1971).
Wang, Yii-ch'iian. 'The Rise of Land Tax and the Fall of Dynasties in Chinese History',
Pacific Affairs, 9 (1936).
376 Bibliography
Yang, Lien-sheng. 'Economic Aspects of Public Works in Imperial China', in the
collection of articles by the author entitled Excursion in Sinology (Cambridge, Mass.
1969).
'Ming Local Administration', in Chinese Government in Ming Times: Seven Studies.
Money and Credit in China (Cambridge, Mass. 1952).
'Numbers and Units in Chinese Economic History', in the collection of articles by the
author entitled Studies in Chinese Institutional History (Cambridge, Mass. 1963).
 Glossary index
When an item is referred to in the same context over several consecutive pages, only the
first occurrence is cited.
Administrative andfinancialterms, and names of institutions
cKa-ma tea and horse trade, chu-cKien li-yiin f Ij profits
257 from minting, 250
cKang-ku yen % fix jjjg salt of regular chu-kung Bfo X contributions to con-
stock, 203 struction projects, 255
cKang-li ^ $! customary payments, cKu [1 tax district, 147
152, 185, 219, 235 chuan-yun-shih $$ jH {j£ transportation
Ch'ang-yin Treasury S ^ ff 20, 105, commissioner, 53
250, 302 cKuan-chiao %k # tonnage, 226
cKen-cKai ^ J£ unpaid supernumer- chuang-p'eng yin $J ^ $t 'common
ary, 246 post money', 250
cheng-tui IE ;& regular transmittance chuang-fien ££ B3 aristocratic estates,
of tribute grain, 52 106
cKi-yun j£3 M transferred revenue, 23 chiin-yao j£j ^ equalization of corvee
chia-pan f$ If mining withfixedquota, service, 6, 36, 38, 91, 110, 112, 147, 184,
241 293
chiang-yin E ^1 artisan payment, 253
chieh-hu f$ ft tax transmitting house- fan-p'o cKou-fen § JQ ft ^ maritime
holds, 38 tariff, 233
Chieh-sheng Treasury f5 fl W. 20,254, fang-ti cKi-shui fB $LMM stamp tax
265 on real estate transfers, 236
c/*7e« H one copper cash or mace, the fen-chu J | £ 'fertilizer owner', 161
weight of one tenth of a tael, 74 /« flS land tax, 38, 109
Ch'ien-ch'ing Palace $£ $f ^ 271, 282 fu-i-cKuan-shu ® #£ ^ # comprehen-
chih-shu |£ S paper redemption cost sive book on taxation and services,
collected from prisoners etc, 248 142
chih-yin £p £P honorary custodian of fu-tsung J^f 1^ prefectural chief scribe,
official seal, 246 149
chinshih *M ± doctorate, the highest
civil service examination degree, 27,196 hai-tao fu-shih $$ ^ gij ^ vice circuit
ching-chih-cKien |M M It tax account intendant on coastal defense, 234
assigned to a regional administrator hao-tsao 8E ft wealthy saltern household,
during the Sung, 139 196
chin-hua-yin ^ ?£ $| 'Gold Floral ho-kung M X river-control expenses,
Silver', 52 165, 255
c/*'m# bg 100 mou of land, 157n ho-p'o so Jpf $g §f fish duty station, 243
ctting-chi yin H H $t ' Speed-the-deli- ho-yen ffi i^ 'river salt', 211
very money', 252 hsiang-shui § IS incense fees at
cKing-yu # ft tax bill, 143 national shrines, 251
chiu-ts'u shut /g §jHf fS duties on wine hsiao-min /J\ J5; common people, 312
and vinegar, 236 hsien-tsung f& |H county chief scribe,
chou-mo cKou-fen Yl /fc ft ^ forest 149
produce levy, 226 hsien-yii J | ^ declared surplus, 152
[377]
378 Glossary index
hu-k'ou shih-yen-chiao p P Jt §j§ §& pay- kung-peng yen X ;$: 'cost paid salt',
ment for rationed salt, 138 209
hu-pu shang-shu P t\l fqj U minister of
revenue, 11, 15 U-cKai jj H service performed by the
hu-yin ft £1 household payment, 120 taxpayer in person, 116, 120
huang-chuartg ]§| jf£ imperial estate, 106 li-chia M Ep rural community for tax-
huang-ts'e g ffi Yellow Book, 32 ation, 25, 30, 37, 40, 92, 110, 112, 116,
Hui-t'ung Kuan # |U fg 276 143, 185, 256, 262, 293
huo-hao 'X $£ melting charges, 148, li-chia yin M ^ $ | community payment,
151 115
li-jih ]¥f 0 cost of calendar paper,
i S service, 34, 109 256
i-chuan W {# service for postal liang-chang H ^ tax captain, 36
stations, 38, 112, 184 lien-ping mi W. J^ 7ft surtax in grain for
i-chuan yin W W M payment for pos- local defense, 350
tal service, 255 lien-ping yin Wk ^k iS surtax in money
i-Viao-pien-fa —M& #$ }£ 'Single Whip for local defense, 350
Method', 6, 118, 131 lou-Vou >M uM. furnace chief, 74
i-tHen-san-chu - 0 H i 'one field with lu-k'o W. U reeds tax, 254
three owners', 161
Inner Ch'eng-yiin Treasury ft^IS ma-cheng Ml $k horse administration,
10, 20, 52 104
ma-shih Mi rf/ horse market, 261
jen-ting ssu-chuan A T $# ffi 'poll tax ma-ts'ao Jfs jf£ animal fodder, 137
silk', 137 mai-penfei H i | g ' purchasing money'
of the Wan-li Emperior, 275, 303
kai-tui t& ^ added transmittance of meng-fan shui ^ M M store franchise
tribute grain, 52 fees, 236
k'ai-chung §§ ^ barter for government min-chuang j ^ iff militia and its main-
salt, 193, 258 tenance, 38, 111, 112, 126, 184
k'ai-na shih-li [JH $fo V- $1 sale of rank, min-mi j ^ ^ land tax as distinguished
244 from rent on government land, 85
Kan-ho M & stub-book, 14
kang M a syndicate, 220 na-liang-ti~$faff ^ tax-paying land as
kang-yin M $1 combined payments, 115 distinguished from 'horse maintenance
kao-yu wan-ch'ing H £$ ||" tg 1 mil- land', 105
lion mou of fertile land, 157n nan-Hang pg $| consignment of tax grain
kuan Jf 1,000 copper coins or its equi- to Nanking, 101
valent in paper currency, 65, 69, 71, nung-sang ssu-chuan J | ^ |^ Ifi 'farm-
138, 232 land silk', 136
kuan-liang fung-p'an fi H M 4^J assis-
tant prefect in charge of taxation, 96 pan-chun che-yin $}£ % t/f ^1 commuta-
kuan-mi If ^t rent on goverment tion of capital guard duty, 255
land, 85 pang-chii 1$ g£ 'posted standard',
kuan-tu shang-pan % % fig if enterprise 120
of joint public and private ownership, pei-liang E=] II highly polished grain for
319 palace consumption, 100
Kuang-huei Treasury Bi M W- 10, 231, pei-ti a J^ 'untitled fields', 108
245, 247 pei-tui a ^ 'empty-handed taxpayer',
Kuang-lu-ssu cKu-liao % fil f If f\ a tax agent 161
supplies to the court of imperial p^i-yen-so #t |^ 0t checking station,
entertainments, 256 192, 257
k'uang-yin $g ^ income from govern- p'iao-yen '^l M 'ticket salt' distin-
ment mining, 240 guished from licensed salt, 205
kuei-fou M II person in charge of tax ping-hsiang ^ |[nj military supplies, 28,
silver chest, 148 38, 134, 182
K'un-ning Palace tt^g 282 p'ing-mi ^p Xt 'levelling grain', 101
 Glosary index 379
seng-tao tu-fieh ffl 8t g? Jt8c ecclesiastic tsao-li che-yin 41 |£ JFf II commutation
license, 246 of personal attendants, 48, 255
shang-shui fgj 5£& local business tax, 226, ts'ao-liang ff ft 'tribute grain', 50,
231 100
shen-cKan-li tf^M J3 'productive for- tso-pan ^Ii f&f local procurement, 19, 24,
ces', 320 35
shih-cheng-ts'e jf ?J( flft actual collection ts'un-chi yen # If ^f salt of reserve
records, 124, 142 stock, 203
shih-tuan-chin + J£ £$ 'ten-sectioned tsung-ping ft" ^ commander-in-chief,
tapestry method', 117 30
shou-tui ife 5: grain checker, 147 tsung-shu ,|t H chief scribe, 149
shu-huan it Jg redeeming payments or tsung-ts'ui $| {H general tax expediter,
fines, 248 147, 193
shu-wei-ts^e j | | M flff 'rat's tail' or roster /« chih-hui cKien-shih M f H f? ^ V
of marginal taxpayers, 111 assistant regional military commis-
ssu-ssu liao-chia jzg Wj ^ {g material sioner, 30
supplies for the four bureaus of the tu chih-hui fung-chih U fit j? IU ^J vice
ministry of works, 138, 255 regional military Commissioner, 30
sui-pan Jg We annual contributions, /Vtf/i [S a group of saltern households,
35 193
Vuan-chung ffl @ military farming of
ta-tsu-chii XM. major tax-paying the cooperative type, 287
owner, 161 /iii-^/i if ^g 'dockyard salt', 211
T'ai-su Hall X * K 282 Tung-yii Treasury ^C %$ ]$. 10, 20
T'ai-ts'ang Treasury :£ # ^ 15, 138, Vung-kuan I | fulfillment of payment
231, 247, 249, 253, 265, 268, 284, 290, obligations by a fiscal unit, 14
295 Tz'u-ning Palace ^£ ^ HT 282
fang-chia |g {if 'marsh-land payment',
139, 140, 195 we/-.?*? f|f §f military colonies, 28, 65,
ti-mou mien-hua-jung j& R& #g 7b #$ 258, 294, 306, 322
cotton wadding assessed on acreage, wen % the value of one standard copper
137 coin, 74
//-/?« ?gj f| 'extra drops' of silver
accompanying ingots, 148 ya-men Wl P5 the general designation of
Vi-pien £H #§ selected call for tax pay- any governmental office, 133
ments in advance, 134, 292 yang-lien §| JH subsidies 'to nourish
Vi-pien yin £§ $f £| funds derived from honesty', 322
the tax payments in advance, 135 yang-ma-ti W: M $L 'horse-maintenance
Vieh-i-yin jj£ #£ ^ | 'money for sub- land', 105
sidizing services', 125 yeh-chu H ^ property owner, 161
fien-lien chiin-hsien EB Si f$ $& land- yen-cKang ^§ % salt production field,
holding extending over one county and 192
prefecture, 157n yin 31 the unit of officially licensed salt,
ting T the able-bodied male as a fiscal sometimes the license itself, 193, 205,
unit, 35, 86, 92, 109, 116, 123, 193, 196, 212, 257
322 yin-cKai £H Jg service commuted to
ting-ssu-fien-liu T ffl ffl 7\ 40 % payment, 116
charged to ting and 60 % to land, yin-Vien ff ffl military farming of the
126 collective type, 287
Ving-chieh jg| ft? tax deliverers, 148 yin-fou | | HI designated agent in
tsa-pan J | $jf materials requisitioned charge of silver collection, 148
irregularly, 35 yu-Vieh gj ^ tax bill, 143
tsang-fa H fj commutation of punish- yu-k'o i t p | fish duty, 243
ments and fines, sometimes including yii-linfu-ts'e | | H fish-scale boo
confiscations, 247 or land survey records, 42
ts'ang-chiao # # granary receipt, yii-yen f£ £§ surplus salt. 207
193 Yuan-ma Ssu ^ H ^P 761
380 Glossary index
Place names
An-hua ft 170 Fukien XVI
Annam fg 47, 72 Fu-ning 127
Fu-yang 167
Chang-chou § jf\ 113, 126, 155, 161,
171 Hai-nan Island ^ ^ g, 187, 190, 313
Chang-p'ing ^ ^ 127, 134 Hai-ning ^g $ 166, 167
Chang-p'u ji$ rf 162 Hai-yen $ S 149,213
Chang-sha | | #> 171 Hang-chou }H jf\ 54, 113, 128, 149,
Ch'ang-chou 'B jf\ 57, 149, 157, 158, 166, 226, 236, 293
176, 236, 294 Ho-chien ^f [HI 106
Ch'ang-hua g ft 166, 167 Ho-ch'iu M ft 130
Ch'ang-lu g jt 191, 221 Ho-hsi-wu ^ @ i 226
Ch'ang-p'ing I, 2js 289, 362 Honan ffi ^ xvi
Ch'ang-shu g ^ 99, 144 Ho-tung M S 191,214
Ch'ang-t'ai g | | 162 Hsi-hsien & U 127
Chao-ch'ing jjl Jg 120 Hsi-hua B ^ 108
Ch'ao-chou $Hf| 113,240 Hsi-ning |§ $ 289
Chekiang j|)f ft xvi Hsiang-ho § M 107, 120, 121, 149
Cheng-chiang £g ft 54, 98 Hsiang-shan ^ ill 61
Cheng-hai § | ?§ 181 Hsiao-i # ^ 169
Cheng-ho jgt ^p 162 Hsin-hua ff ft 114
Cheng-ting M ^ 237 Hsing-hua p ft 61
Ch'eng-hai ?j£ ?g 162 Hsiu-ning ft: $ 127
Chi-chou fff JH1 300 Hsu-yeh If § 226
Chi-chou (military district) fij jf\ 285, Hsiian-fu (military district) j j Jff 214,
289, 362 268, 285, 289, 362
Chi-hsien fc U 42, 108 Hukwang ^ ]S xvi
Ch'i-men jpI5 P! 90, 185 Hua-ma-ch'ih ^ ,1 fife 289
Ch'i-yang fl5 f^ 140 Hua-t'ing ^ ^ 149, 157, 158
Chia-hsing S M 98, 166, 294 Huai-an m £ 25, 54, 145, 163, 226,
Chia-shan H # 90 230, 236, 250, 278, 280, 320
Chia-ting S ^ 100, 122, 153, 159 Huai-ch'ing if H 300
Chiang-hua ft ^ 140 Huai-jou ff m 120, 185
Chieh-hsiu ^ ft; 169 Huai-yiian '|f S 300
Ch'ien-t'ang ^ jg 167 Hui-chou Wi >H 91, 112, 120, 137,
Chihli IK ^ xvi 143
Ching-chou ^J ffl 237 Hu-chou m JM 98, 109, 166
Ching-hsing # |M 289, 362
Ching-hua ^ ^ 68, 166 i 57, 254, 362
I-chou -g, '}\
Ch'ing-chiang-p'u ffi ft ffl 25, 56, 237, I-WU ^ i§ 166
239, 254, 320
Ch'ing-chou ff j11 149 Jen-ho ^ ?P 166, 167, 197
Ch'ing-p'u W ?i 149, 158, 159
Chiu-chiang % ft 226 K'ai-feng $1 M 27, 179
Chung-mou 4* $ 120, 137 K'ai-hua H ft 127, 172
Ch'ung-te ^ ^ 151 Kan-su (military district) 1t ft 29,2
Ch'ung-wen Gate ^k 3c P^ 228 362
Ch'li-chou M iMi 153, 185 Kiangsi ft H xvi
Ch'uan-chou ^ ffl 240 Ku-an 0 ^: 106
Chiin-chou &| jMi 252 Ku-yiian gj| J^ 285, 289, 362
Ku'ai-chi -|r ft 61, 62, 91, 145, 293
Fei-hsien |IJB 130 Kuang-ning mm 233,237
Fen-chou ^9- JM 130, 168 K'un-shan [ m III 342
Fen-yang fj9" (^ 232 Kwangsi ^ I§ xvi
Feng-hua 5M b 61 Kwangtung m IK xvi
Feng-yang S PI 54,349 Kweichow i" #1 xvi
 Glossary index 381
Lan-chou M J>H 237 Shen-yang ^ Pi 289
Li-yang $g fi§ 171 Shun-an }$ ^ 108, 151, 185, 219
Liang-Che (salt administrative district) Shun-te jig ^ 84, 89, 114, 122, 160,
M $f 190, 213, 258 172, 293, 301, 318
Liang-Huai (salt administrative district) Shun-t'ien jig ZX 165, 276
M m 191, 212, 258 Soochow M 41! 98, 100, 122, 151, 158,
Liao-ch'eng Ijljp J$ 139 161, 180, 294
Liao-tung (military district) $££ j$( 29, Su-chou ffi 'M 289
289, 290, 362 Sui-an ^ ^ 136
Lin-an Ba ^ 136, 167 Sung-chiang fe in 89, 98, 122, 146,
Lin-ch'ing ffi 78 54, 226 149, 153, 155,277,294
Lin-fen B35 & 95, 155, 177, 179 Szechuan [ZH J\\ xvi
Lin-hsien HS U 169 Szu-shui M 7k
Ling-chou £ >JI| 191, 214
Ling-ling 2g [^ 140 Ta-hsing ~X p 236, 276
Ling-shih M S 130, 169 Ta-tung (military district) X In] 268,
Lu-ch'eng $g %$ 120 285, 289, 362
Lung-chiang II %L 237 T'ai-ho-shan ± ft] UU 251
Lung-hsi ft g| 127, 162 T'ai-p'ing ^ zp 237
T'ai-shan ^ Jj 251
Mi-yiin $ S 289, 362 T'ai-ts'ang ^ # 100
T'ai-yuan X ^ 168, 289
Nan-ch'ang \M &i 68 Tan-ch'eng % ^ 130
Nan-ching ^ £# 127, 162, 171 Tao-chou M4H 140
Nanking ^ g xvi, 54 Te-chou ^ ^N 54, 151
Nan-p'ing ^ ^P 162 Teng-chou fMH 164,348
Ning-hsia (military district) ¥ M 29, Ting-t'ao S p@ 172
285, 289, 362 Ts'ao-chou W jNI 172
Ning-hsiang 5p ^ 168 Ts'ao-hsien W M 121, 132, 149, 172
Ning-kuo ft g| 114, 149, 176 Tseng-hua m ft 241
Ning'po $ $c 234 Tsi-nan $f fg 27,54
Ning-yiian 3p 5s 140 Tsou-hsien £$ ^ 121
Tung-an ^ £ 140
Pao-te ft M 130 Tung-ao JfC psj 97, 149
Pao-ting ft ^ 106, 237, 243 Tung-ch'ang 3I g 107, 149
Peking it M xvi, 54, 289 Tung-t'ing Lake* mfew 165
Pei-hsin-kuan ;JfcrrBS 226 Tzu-yang ^ U I 155
P'eng-ts'e I> ^i 129 Tz'u-hsi ?M ^ 181
P'ien-t'u-kuan {g 3l HI 289
P'ing-ho ^ ^n 127, 171 Wan-ch'iian |h ^ 29
P'ing-liang q± ® 174 Wan-p'ing ^g ^ 105,119,236,276
P'ing-yao -7f 3J| 130 Wen-shang ^ ± 61,97,114,137,149,
171, 186, 301
Sha-hsien /^ |$ 162 Wu-chin St 3§ 147, 149
Sha-shih fp TfiT 237 Wu-hsien ^ |% 103, 152, 172, 180
Shanghai ± ^I 94, 124, 140, 147, 158, Wu-hu M M 237, 239
197 Wu-yiian g | 1I 242
Shang-yii ± ® 166
Shansi jj_j S xvi Yang-chou f§ 54,191,217,226,278,
Shantung ill ^ xvi 281
Shan-hai-kuan Jj M | | 233, 289 Yen-sui (military district) 285,
Shan-si (military district), 285, 289, 362 289, 362
Shan-yang 1J4 pg 25 Ying-t'ien J9g % 45, 176
Shang-jao ± ^ 242 Yii-ch'ien K ?f 166
Shao-hsing IS f| 16, 166, 292 Yii-hang ^ t/L 166, 167
Shao-wu SP IES 162 Yii-lin tm W 289
Shensi [$ S xvi Yiieh-chou g ffl 128
382 Glossary index
Yiieh-kang ft 235 Yung-ming 140
Yung-an ^c 90, 162 Yung-ning 168, 169
Yung-ch'eng %fc 108 Yung-p'ing 289, 362
Yung-chou f\ 108, 140, 244 Yunnan ft
Personal names
Altas 208,260,266 KoTun fBSfc 64
Ku Hsien-ch'eng lg ;#, j ^ 296
Chang Chia-yun 3ft & A 298 Ku Yen-wu M & K 15, 49, 80, 96, 99,
Chang Ching 5g |M 278,292,321 145, 150, 152, 185, 187, 310, 322
Chang Chu-cheng 3ft Jg IE 4, 81, 97, KueiYu-kuang U ti it 145,151,186
145, 150, 153, 154, 157, 245, 256, 267,
286, 294, 304 Li Ch'eng-liang f 298
Chang Hsueh-yen %^M 154n, 165, Li Ming ft 107
298 Li Nii-hua 15, 163, 202
Chang Tung 3ft S 186 Li San-ts'ai 296
Chang Yen-ling 3ft 2E $ 107 LiShih-ta 300
Chao Nan-hsing i S M 296 Li Wei {# 153
Chao Shih-ch'ing m^&M 231 Liang Fang-chung 37,118,121
ChaoWen-hua mX~^ 278 Liang Meng-lung 298
ChaoYing ffi » 100 Liang Ts'ai * 278
Chao Yung-hsien m ffl K 44, 181 Lin Fu # g 234
Ch'en Feng Pf ^ 219 Lin T'ing-ang ^ fil 183
Cheng Hsiao f$ ^ 42 Ling Yun-i I I 298, 312
Ch'i Chi-kuang 286, 298, 320 Liu Chin gij 4,242,247
Ch'ien Shih-sheng gg ± \ 158
:
Liu Chung-fu + ft 12
Chin Lien & M 12 Liu Kuan S 49
Ch'in Chin IS ^ 12 Liu Szu-chieh ij 9f ^ 360
Ch'iu Ch'iin flS j# 222 Liu Ta-hsia ^c S 19
Chou Ching Jl M 229 LuPao t- 211,213,217,219
Chou Hung-mo Jij #t jjg 49
Chou Sheng jfj f£ 52 Ma Shen , 12
Mou K'un 167
Fang Tun ^ 273
Feng Ch'i 1$ 159 Ni Yiian-lu jjg TU M 12, 15, 68, 119,
Feng Pao ^ 297 173, 185, 230, 241, 298
Fu I-ling 5c & 90, 157n, 162 NiYiieh U S 49
Nien Fu ^ 1 187
Hai Jui H9, 151, 159, 187, 311
Han Wen W X 12, 177 Oosai 285
Ho Liang-chim ^ ft 41, 92, 249
Ho T'ang H i t 132 Pai Tung fi It 97, 299
Hsia Yiian-chi S jt ^ 12, 71 Fan Chi-hsun 19, 27, 86, 115,
Hsu Cheng-ming ^ M. ^ 169, 245 280, 298
Hsu Chieh & (^ 78, 157n, 159 Fan Huang S 278,286
Hsiieh Shang-chih If ^ M 186 P'ang Shang-p'eng Ml 27, 119,
Hu Tsung-hsien ftg ^ ^ 278 166, 191, 209, 222, 286
Hu Ying ffl W 52 Feng Hsin-wei % \t J& 62, 77, 79, 210
Huang Fu H 52 Pi Chiang mm 12
HuoYu-hsia M Wx 181 Pi Tzu-yen # @ K 12, 15
Pubei 0. n 302
Jinong ^ H 285
Sang Hung-yang # ^ ^ 222
Kao Hung-t'u 197 Shao Ching-pang gp M ^ 238
Kao Yao M 245 Sheng Shih-hsing ft B$ fl 123, 160
Ko Shou-li t 132, 160, 171 Shih Hsing 5 1 183
KoTse M 12, 50 SunP'i-yang ffi dS ^ 249
 General index 383
SungHsun 5fc H 183 Wei Chung-hsien 1^ ;cA jlf 4
Sung Ying-hsing J | | 241
Yang Chieh % if 97
Tan Lun jg Ma 109, 290, 294 Yang I-ch'ing M 1 - ^ 258
T'ang Shun-chih If I'll ;£ 159 Yang Shih-chiao HI IHf ^ 238
Toyotomi Hideyoshi H gL ^ ^ 235, Yang T'ing-ho ii g f t l 7,59,272
302 Yang Yen i§ ^; 222
Yang Yung-lung ^ Ifi fi 302, 321
Wang An-shih 3 : ^ S 222, 316 Yeh Ch'i ^ 8t 204, 206, 222
Wang Chen 3£ S! 4 Yeh Hsiang-kao SI fnl IS 7
Wang Ch'ung-ku =£ ^ •Jf 297, 298 Yeh Meng-chu '-1 # ft 158, 160
Wang Kao £ S 12 Yeh Tsung-liu 3| m @ 242
Wang Kuo-kuang 15, 190 Yen Mou-ch'ing IP g* pp 191,208,213
Wang Lin £ ^ 12, 247 Yen Sung R g 208, 248
Wang Shih-cheng 3E lit i t 61, 297 Yin Cheng-mou Jg IE m 77, 297, 298,
Wang Shih-hsing 3E ± tt 295, 301 312
Wang Shu 3E 16 63 Yin Min ^ g 49
Wang Ssu-jen 3 : ^m u 159 Yii Sheng-hsing ^F 'fl fx 118
Wang Tsung-mo =E ^ ^(c 297 Yii Ta-yu # ^c t^ 286
Wang Ying-chiao "ffiflgiK 12, 149 Yii T'ai-su 3a ^: # 12
Wang Yii-ch'iian 3 : IS ££ 64, 286, 307 Yii Tzu-chiin ^ ^ ft 285
Wei Ch'ing-yuan Iw i ^ 61,90 Yiian Shih-cheng ^ t^ M 217

General index
The general index is to supplement the glossary index, the table of contents, and the
running heads. While some degree of overlapping is inevitable, it is not intended to
duplicate the information.

aristocracy, aristocratic title-holders, 31-2,stration, 29; authorizing tax payments

106, 153, 179, 275, 310 in silver, 52; legalizing use of precious
army logistics, 5, 28, 29-31, 34, 45, 53, 58, metals, 72; captured by Mongols, 73
63-8, 88, 126, 134-5, 177-8, 181-2, Ch'eng-hua, the Emperor, xv, 285
214-16, 231, 269, 284-94, 295, 306, 310 Chia-ching, the Emperor, xv, 7, 12, 59,
auditing, of fiscal accounts, 1, 14, 16, 89, 118, 266, 285; ordering coinage, 76-8;
278-9, 295, 299 limiting aristocratic estates, 107; pro-
Brunei, 233 moting mining, 242; commuting tribute
Buddhist, monks and monastic properties, grain, 272; surrendering crown income,
17, 33, 55, 109, 134, 246, 292 273; authorizing tax reduction, 277
Chicago, University of, 154n
bureaucracy, 4, 9, 30, 45, 183, 224, 311, Ching-t'ai, the Emperor, xv, 71
323 Ch'ing, dynasty, 45, 89, 95, 198, 200, 221,
Burma, Burmese, 24, 215, 243 230, 231, 239, 307, 308, 317, 319, 322
Ch'ung-chen, the Emperor, xv, 6, 16, 173,
censorial officials (including regional 298; putting eunuchs in charge of civil de-
inspectors, surveillance commissioners fense, 10; rejecting proposal to abolish
and supervising secretaries), 16, 27, 29, hereditary military households, 68
186, 191, 201, 261, 278, 312, 321 Columbia, University, 154n
Champa, 234 commerce, 21, 230
Cheng-te, the Emperor, xv, 4,285; abusing commutation, of taxes in kind and service
the salt monopoly, 207 obligations, 3, 18, 39, 52, 86, 93, 105,
Cheng-t'ung, the Emperor, xv, 4, 12, 71; 115-17, 143, 155, 156, 166, 197, 237,
turning granaries over to civil admini- 252-6, 265, 271, 272, 320
 General index 383
SungHsun 5fc H 183 Wei Chung-hsien 1^ ;cA jlf 4
Sung Ying-hsing J | | 241
Yang Chieh % if 97
Tan Lun jg Ma 109, 290, 294 Yang I-ch'ing M 1 - ^ 258
T'ang Shun-chih If I'll ;£ 159 Yang Shih-chiao HI IHf ^ 238
Toyotomi Hideyoshi H gL ^ ^ 235, Yang T'ing-ho ii g f t l 7,59,272
302 Yang Yen i§ ^; 222
Yang Yung-lung ^ Ifi fi 302, 321
Wang An-shih 3 : ^ S 222, 316 Yeh Ch'i ^ 8t 204, 206, 222
Wang Chen 3£ S! 4 Yeh Hsiang-kao SI fnl IS 7
Wang Ch'ung-ku =£ ^ •Jf 297, 298 Yeh Meng-chu '-1 # ft 158, 160
Wang Kao £ S 12 Yeh Tsung-liu 3| m @ 242
Wang Kuo-kuang 15, 190 Yen Mou-ch'ing IP g* pp 191,208,213
Wang Lin £ ^ 12, 247 Yen Sung R g 208, 248
Wang Shih-cheng 3E lit i t 61, 297 Yin Cheng-mou Jg IE m 77, 297, 298,
Wang Shih-hsing 3E ± tt 295, 301 312
Wang Shu 3E 16 63 Yin Min ^ g 49
Wang Ssu-jen 3 : ^m u 159 Yii Sheng-hsing ^F 'fl fx 118
Wang Tsung-mo =E ^ ^(c 297 Yii Ta-yu # ^c t^ 286
Wang Ying-chiao "ffiflgiK 12, 149 Yii T'ai-su 3a ^: # 12
Wang Yii-ch'iian 3 : IS ££ 64, 286, 307 Yii Tzu-chiin ^ ^ ft 285
Wei Ch'ing-yuan Iw i ^ 61,90 Yiian Shih-cheng ^ t^ M 217

General index
The general index is to supplement the glossary index, the table of contents, and the
running heads. While some degree of overlapping is inevitable, it is not intended to
duplicate the information.

aristocracy, aristocratic title-holders, 31-2,stration, 29; authorizing tax payments

106, 153, 179, 275, 310 in silver, 52; legalizing use of precious
army logistics, 5, 28, 29-31, 34, 45, 53, 58, metals, 72; captured by Mongols, 73
63-8, 88, 126, 134-5, 177-8, 181-2, Ch'eng-hua, the Emperor, xv, 285
214-16, 231, 269, 284-94, 295, 306, 310 Chia-ching, the Emperor, xv, 7, 12, 59,
auditing, of fiscal accounts, 1, 14, 16, 89, 118, 266, 285; ordering coinage, 76-8;
278-9, 295, 299 limiting aristocratic estates, 107; pro-
Brunei, 233 moting mining, 242; commuting tribute
Buddhist, monks and monastic properties, grain, 272; surrendering crown income,
17, 33, 55, 109, 134, 246, 292 273; authorizing tax reduction, 277
Chicago, University of, 154n
bureaucracy, 4, 9, 30, 45, 183, 224, 311, Ching-t'ai, the Emperor, xv, 71
323 Ch'ing, dynasty, 45, 89, 95, 198, 200, 221,
Burma, Burmese, 24, 215, 243 230, 231, 239, 307, 308, 317, 319, 322
Ch'ung-chen, the Emperor, xv, 6, 16, 173,
censorial officials (including regional 298; putting eunuchs in charge of civil de-
inspectors, surveillance commissioners fense, 10; rejecting proposal to abolish
and supervising secretaries), 16, 27, 29, hereditary military households, 68
186, 191, 201, 261, 278, 312, 321 Columbia, University, 154n
Champa, 234 commerce, 21, 230
Cheng-te, the Emperor, xv, 4,285; abusing commutation, of taxes in kind and service
the salt monopoly, 207 obligations, 3, 18, 39, 52, 86, 93, 105,
Cheng-t'ung, the Emperor, xv, 4, 12, 71; 115-17, 143, 155, 156, 166, 197, 237,
turning granaries over to civil admini- 252-6, 265, 271, 272, 320
384 General index
copper, copper coins, 6,72,74-9,155,231, Hung-wu, the Emperor, xv, 5, 12, 39,
233, 240, 247, 250, 298, 317, 320 42, 44, 51, 55-7, 63, 65, 67, 72, 75, 104,
corruption, official, 49, 141, 144, 152-3, 233, 246, 257; as fiscal legislator, 44-6,
179, 185, 201, 208, 221, 228, 278, 307, 183, 189, 221-2, 298-9, 314; interview-
311, 312, 323 ing tax captains, 37; issuing paper cur-
cotton, cotton cloth, 9, 101-3, 137, 142, rency, 69; confiscating landed properties,
262, 271, 296 99; authorizing permanent tax exemp-
county magistrate, as fiscal officer, 24, 97, tion, 106; promoting silk and cotton,
111, 119, 142, 144, 148, 151, 159, 179, 136
185, 236, 300, 314, 327
Creel, Herrlee G., sinologist, 309 I-pu-la, 260
imperial clansmen, 31, 178-80, 182, 214,
deficit financing, 202-4 295
dual capital system, 13
Japan, Japanese, 2,234,240,302, 306, 314,
eunuchs, 4, 8-11, 56, 58, 114-15, 208, 319
211, 219, 233, 234, 242, 247, 251, 260, joint fiscal administration, by central and
266, 267, 295, 297, 299, 302, 303, provincial authorities, 17, 23, 191
315
excise tax, 207, 210,212, 224, 328 K'ang-hsi, the Ch'ing Emperor, 89
Korea, Korean, 104, 233, 290, 294, 302
Fairbank, John K., sinologist, 118
Fang, Chao-ying, sinologist, 157n land reclamation, 18, 106, 195
farm credit, 151, 160, 187, 315 land survey, 26,42, 84,98,154,168, 300-1,
Feuerwerker, Albert, sinologist, 318 313, 329-30
foreign trade. 1, 76, 233-6, 257-61 land tenure, 37, 67, 85, 99, 106-8, 156-62,
franchise, 71, 220, 236 287, 315
Friese, Heinz, sinologist, 110 law of avoidance, 96
Fujii, Hiroshi, sinologist, 60-1 lawsuits, 97, 248, 312
lesser functionaries, 16, 26, 27, 133, 144,
gentry, 45, 97, 158, 197, 299, 312 181, 192, 230, 316
Gold Floral Silver, 10, 52, 93, 102, 180, local procurement, 24, 35, 143, 315
273, 275 lumber, 19, 108, 127, 237-40, 282-3
grain, prices, 101, 155, 156, 159, 160, 167, Lung-ch'ing, the Emperor, xv, 121, 293
170,186; production, 40-1,90,166,169,
170-2, 313 Macao, 235
Grand Canal, 1, 11, 23, 40, 51-5, 94, 101, Manchuria, Manchurian, 104, 163-4,172,
186, 226, 238, 252-3, 279, 283, 297, 317, 233, 289
318 Mansur, 260
Great Wall, 269, 285, 289, 318 merchants, 33, 127, 149, 195, 198, 200-24,
226, 228, 230, 232, 235, 238-45, 250,
Han, dynasty, 220, 321 259, 318, 319
Harvard-Yenching Library, 121 militia, 111, 114, 184, 294, 302
Ho, Ping-ti, sinologist, 41,62,125,168,198 ministry of revenue, 11-17, 154, 163, 190,
horses, government, 1, 18, 104-6, 250-1, 201, 214, 216, 226, 227, 231, 235, 243,
255, 257-61, 269, 295 245, 247, 262, 268-77, 290, 303
Hsiian-te, the Emperor, xv, 12, 57-8, 64, ministry of rites, 17, 227, 256, 281
72, 75; creating aristocratic estates, ministry of war, 17, 227, 251, 265, 269
106-7 ministry of works, 13, 17-20, 87, 227, 237,
Hucker, Charles O., sinologist, 4n, 56 239, 243, 245, 254, 255, 265, 282
Hung-chih, the Emperor, xv, 46, 58, 223, Mongol, Mongolian, 12, 68, 260, 295
285; disclosing abuses of salt monopoly, Morse, H. B., sinologist, 235
204 mortgages, 90, 236, 288, 309
Hung-hsi, the Emperor, xv, 12, assigning mou, fiscal, 41, 60, 99,105, 108, 168, 172-
civil officials to administer army com- 3; standard, 40, 300
mands, 29; collecting franchise fees, 71;
creating palace estates, 106 Nanda, Bay in, 24, 243
 General index 385
palace, construction, 19, 58,239, 255, 267, 152-4,296; exemption, 97,123-4; rates,
282-4; estates, 106-8, 273, 303, 310, 41-2, 71, 85, 89, 100, 127-8, 130, 140,
325; expenditures, 8-11, 56-9, 87, 315, 166-75, 182-7, 216, 229, 235, 236, 238,
317 259, 308 (see also commutation, excise
paper currency, 1, 69-74, 232, 261 tax, poll tax, two-tax system)
People's University, Peking, 318 taxation, progressive, 35, 87, 88, 121, 131,
poll tax, 36,87,114,133,139,192,195,198 183, 319
population, 19,26,61-3,224,248,264,309 tea, 9, 35, 57, 257-62, 265
Portugal, Portuguese, 234-5, 285, 323 tenancy, 160-2, 187
postal system, 38,88,112,184,256,295,318 timber see lumber
T'ien-ch'i, the Emperor, xv, 4, 7, 74, 282,
quota system, 39, 46-7, 51, 52, 155, 163, 304
177, 183, 195, 208, 210, 213, 215, 228, tribute grain, 40, 50, 59, 95, 100, 142, 147,
231, 232, 240, 243, 254, 255, 261, 264, 252, 271, 272, 274, 282, 298
304, 308, 309, 314, 319, 322 Twitchett, D. C , sinologist, 131, 266
two-tax system, 39
reform, fiscal, 6, 72, 96, 115, 131, 223-4,
310, 316, 322 (see also Single Whip Vierira, Cristavao, 323
Reform)
Wan-li, the Emperor, xv, 4, 20, 153, 249,
Ricci, Matteo (Matthew), 283 256, 285, 293, 297, 300; commissioning
Rossabi, Morris, sinologist, 257, 260 eunuch tax collectors, 7, 10, 212, 267,
315; authorizing surtax to finance
salaries, 8, 23, 25, 48, 59,73,175,178,230, Manchurian campaign, 163; personal
275, 277, 296,298, 303, 315, 322 extravagance, 183, 303-5; ordering
Schall, Adam von Bell, 241 mining, 243; increasing reeds tax, 254;
Shimizu, Taiji, sinologist, 60, 98 demanding 'purchasing money', 275;
ship construction, 24, 50, 53, 237, 244, 320 ordering tax reduction, 277; con-
Siam, 234 structing own mausoleum, 302, 310
silk, silk fabrics, 9, 11, 40, 57, 127, 136-7, Wang, Yeh-chien, sinologist, 322
167, 262, 271, 304 water-control projects, 18, 98, 165, 186,
silver, 6, 70, 79-81, 89, 95, 103, 142, 148, 255, 279, 298
150, 177, 241-3, 267, 295, 298, 316, 318 white book, 63
Single Whip Reform, 27, 36, 103, 112-22,
130-3, 144, 147, 166, 171, 183, 267, 293, Yamane, Yukio, sinologist, 110
299-300, 312, 316 Yang, Lien-sheng, sinologist, 282
stub-books, 14, 193, 218, 309 Yangtze delta, 40, 47, 54, 62, 81, 98-104,
Sung, dynasty, 1, 21, 42, 46, 53, 75, 91, 116, 122, 124-5, 142, 159-60, 171-2,
99, 108, 111, 155, 186, 193, 200, 236, 174, 180, 186, 199, 238
316, 321 Yellow Book, 32-3, 60, 164, 248
supply lines, criss-crossing, 5, 14, 65-6, Yellow river, 165, 168, 280, 317
176, 275, 322 Yuan, dynasty, 1, 79, 99, 155, 316, 321
Yung-lo, the Emperor, xv, 50-1, 57, 63,
Taoists, 17, 33, 246 65, 75, 233, 246, 282, 298-9; expedition
Tai-ch'ang, the Emperor, xv, 7, 304 to Mongolia, 12; creating circuit inten-
Tang, dynasty, 1, 21, 47, 53, 91, 131, 186, dants, 28-9; annexing Annam, 47;
216, 266, 316, 321 currency inflation, 70; horse procure-
tax, agents, 92, 133,145-50,193, 209,230, ment, 104
296; delivery, 36-7, 40, 101, 142-3,