Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Revenue as a measure for expenditure : Ming state finance before the age of silver Grass, Noa 2015

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


24-ubc_2016_february_grass_noa.pdf [ 2.28MB ]
JSON: 24-1.0223467.json
JSON-LD: 24-1.0223467-ld.json
RDF/XML (Pretty): 24-1.0223467-rdf.xml
RDF/JSON: 24-1.0223467-rdf.json
Turtle: 24-1.0223467-turtle.txt
N-Triples: 24-1.0223467-rdf-ntriples.txt
Original Record: 24-1.0223467-source.json
Full Text

Full Text

REVENUE AS A MEASURE FOR EXPENDITURE:  MING STATE FINANCE BEFORE THE AGE OF SILVER by NOA GRASS A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (History) THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver) December 2015 © Noa Grass, 2015ii Abstract   This dissertation explores fiscal policy during the first half of the Ming dynasty. Assuming a constant state of financial crisis caused by an ideological refusal and institutional inability to increase revenue, it identifies aspects of financial administration that contributed to the durability and resilience of the state. It first analyzes the principles of early Ming financial administration as reflected in the founding administrative text, Zhu si zhi zhang. The chapter devoted to the Ministry of Revenue focuses on the management of local resources through the timely and accurate flow of funds and information throughout the realm and along the administrative hierarchy. Based on evidence of standardized annual revenue reports, this dissertation argues that those principles were applied with relative success throughout most of the fifteenth century.   Next, it identifies the practice of commutation in tax collection and official payments as the main fiscal policy that enabled the Ming to abide by its principle of keeping expenditure low while avoiding financial default. Commutation served as a partial tax remission that enabled taxpayers to convert the grain they owed to a money or commodity at a favourable rate. It also alleviated the physical and financial burden of transportation. But as the state came to depend on fixed silver payments, financial administration transformed from a system that was focused on managing local resources to one that was geared to maximizing revenue in the  iii political centre. Finally, payments to officials, soldiers, and princes were affected by commutation. Despite their different social status, these groups were all treated as servants of the state and were managed according to the fiscal principle of measuring expenditure according to revenue. Throughout the fifteenth century payments were partially commuted to scrip and as a result salaries and stipends were greatly diminished. Nevertheless, particular policies and practices maintained a minimal degree of remuneration. And strategies employed by members of these groups in order to better their material condition illuminate the relationship between the state and its servants, as well as their place in local society.    iv Preface  This dissertation is my original, previously unpublished work. I prepared the tables that appear in Appendix D using data from published historical sources. Unless otherwise indicated, translations from the Chinese sources are my own. v Table of Contents  Abstract ............................................................................................................................... ii Preface ................................................................................................................................ iv Table of Contents ................................................................................................................ v List of Tables ................................................................................................................... viii List of Figures .................................................................................................................... ix Notes on Usage ................................................................................................................... x Acknowledgements ............................................................................................................ xi Dedication ........................................................................................................................ xiii 1 Introduction: Ming Fiscal Policy in Historical Context .............................................. 1 1.1 Distinction Between Private and Public Revenue ................................................ 6 1.2 Indirect Taxation .................................................................................................. 9 1.3 Money................................................................................................................. 13 1.3.1 Government Credit...................................................................................... 19 1.4 Financial Administration and Expertise ............................................................. 21 1.5 Another Approach to Financial Administration ................................................. 24 1.6 Distinguishing between the Fifteenth and Sixteenth Centuries ......................... 30 1.7 Accounting ......................................................................................................... 33 1.8 Tax Commutation ............................................................................................... 37 1.9 Exposition of the Chapters ................................................................................. 41 2 The Structure and Role of the Ministry of Revenue ................................................. 57 2.1 Legal Foundation of Financial Administration .................................................. 58 2.2 Political Origins of Ming Financial Administration .......................................... 67 2.3 The Structure of the Ministry of Revenue .......................................................... 73 2.3.1 The Board of Population ............................................................................. 75 2.3.2 The Board of Expenditure ........................................................................... 82 2.3.3 The Board of Money ................................................................................... 84 2.3.4 The Board of Granaries ............................................................................... 85  vi 2.4 Annual Revenue Reports .................................................................................... 88 2.5 Sixteenth Century Developments ....................................................................... 96 3 The Role of Tax Commutation in State Finance .................................................... 102 3.1 Scrip ................................................................................................................. 108 3.2 Silver ................................................................................................................ 129 3.3 Cotton and Silk ................................................................................................. 156 3.4 The Annual Revenue Reports and Ming Financial Administration ................. 163 4 Civilian and Military Salaries ................................................................................. 166 4.1 Government Land ............................................................................................. 168 4.2 Salary Charts .................................................................................................... 172 4.3 Commutation of Salaries into Scrip ................................................................. 176 4.4 Commutation into Sappan Wood, Pepper, and Cloth ...................................... 185 4.5 Silver ................................................................................................................ 191 4.6 Sending Money Home ...................................................................................... 194 4.7 Paying Local Officials ...................................................................................... 199 4.8 Ming Salaries in Comparative Perspective ...................................................... 202 5 Military Finance ...................................................................................................... 207 5.1 Rations .............................................................................................................. 212 5.2 Awards ............................................................................................................. 216 5.3 The Salt for Grain Policy ................................................................................. 220 5.4 Military Colonies .............................................................................................. 229 6 Princes ..................................................................................................................... 239 6.1 Regular Income ................................................................................................ 247 6.2 Commutation of Grain Allocations .................................................................. 252 6.3 Princely Estates ................................................................................................ 256 6.4 Expansion of Princely Estates .......................................................................... 260 6.5 Other Expenses of Princely Households .......................................................... 268 6.6 Hereditary Titles ............................................................................................... 270 6.7 Implications for the State Budget ..................................................................... 272 7 Conclusion .............................................................................................................. 275 Bibliography ................................................................................................................... 287  vii Appendices ...................................................................................................................... 309 Appendix A: List of Ming Emperors .......................................................................... 309 Appendix B: Weights and Measures ........................................................................... 310 Appendix C: Commutation Exchange Rates .............................................................. 311 Appendix D: Revenue Reports ................................................................................... 312  viii List of Tables  Table A-1 List of Ming Emperors .................................................................................. 309 Table B-1 Weights and Measures ................................................................................... 310 Table C-1 Commutation Exchange Rates ....................................................................... 311 Table D-1 Yongle Reign ................................................................................................. 313 Table D-2 Yongle Reign (cont.) ..................................................................................... 314 Table D-3 Yongle Reign (cont.) ..................................................................................... 315 Table D-4 Yongle Reign: Tributes from Jiaozhi ............................................................ 316 Table D-5 Hongxi-Xuande Reigns ................................................................................. 317 Table D-6 Hongxi-Xuande Reigns (cont.) ...................................................................... 318 Table D-7 Hongxi-Xuande (cont.) .................................................................................. 319 Table D-8 Hongxi-Xuande Reigns: Tax Remissions on Government Land .................. 320 Table D-9 Zhengtong-Tianshun Reigns .......................................................................... 321 Table D-10 Zhengtong-Tianshun Reigns (cont.) ............................................................ 322 Table D-11 Zhengtong-Tianshun Reigns (cont.) ............................................................ 323 Table D-12 Chenghua Reign .......................................................................................... 324 Table D-13 Chenghua Reign (cont.) ............................................................................... 325 Table D-14 Chenghua Reign (cont.) ............................................................................... 326 Table D-15 Hongzhi Reign ............................................................................................. 327 Table D-16 Hongzhi Reign (cont.) ................................................................................. 328 Table D-17 Hongzhi Reign (cont.) ................................................................................. 329   ix List of Figures  Figure 3-1 Total Annual Tax Income Collected in Scrip (ding) ..................................... 128 Figure 3-2 Total Annual Income from Silver Mining (liang)......................................... 143 Figure 3-3 Total Annual Income in Silk (pi) .................................................................. 162 Figure 3-4 Total Annual Income in Cotton (pi) .............................................................. 162  x Notes on Usage   Chinese terms and names are rendered in pin yin except for references in other transliteration systems. Chinese and Japanese names are written surname first, unless they appear differently in English publications. Chinese terms for measure words are used except for liang, the standard measure word for silver in the Ming. I chose to use the term tael since it has become standard in English publications and in order to avoid confusion with liang as a measure unit for cloth. I use shi, a unit of volume, instead of dan, a unit of grain bearing the same character, for denoting the standard unit for grain since grain was measured by volume in the Ming. Reign dates are designated according to the equivalent of the end of the Chinese year in the Gregorian calendar. Therefore the end of one reign and the beginning of the next share the same year. Translations of official titles rely, wherever possible, on Hucker, Dictionary of Official Titles (1985).  xi Acknowledgements   This dissertation has been long in the making. It saw the joy of the birth of my daughter and the hardship of a difficult illness. Throughout this time, I have had the privilege of the guidance of my teachers and supervisors, Timothy Brook and Leo Shin. Professor Brook saw the direction in which my project is going even before I fully grasped it and encouraged me to continue along that path. Professor Shin has the gift of putting me back on track and reminding me to look at the big picture in the midst of confusing details. I was also lucky to work with wonderful colleagues who had direct impact on the development of my academic identity. I had long and fruitful discussions with Desmond Cheung and Malcolm Thompson. Many cups of coffee were shared with Laura Madokoro while working out our individual projects. Gabriela Aceves has often provided the right word of advice at the right time. The Institute of Asian Research provided a wonderful space to work and converse. I would like to acknowledge the collegiality of IAR employees Karen Jew, Carolyn Grant, Yoko Nagao, Kirstin Luettich, and Marietta Lao. The lunchtime discussions helped maintain my mental wellbeing during stressful times. I would like to send my deep gratitude to the staff of the dialysis and transplant clinics at Vancouver General Hospital. Without their care I would have not been able to finish this dissertation. Many thanks go to all of my friends in Vancouver who were always there to lend a hand and open their hearts when times were rough. Finally, I would like to thank my family in Israel for their care and concern that helped  xii me get through difficult times. And to my husband Beni and Daughter Daniela, without your constant support and limitless love this project would have been meaningless. xiii Dedication  To the memory of my mother Keren Grass,  who always kept my desk clean so I could study.  1 1 Introduction: Ming Fiscal Policy in Historical Context  On a summer day in 1390, the Hongwu emperor held a nomination ceremony for the new heads of the six ministries. Addressing the former Minister of Works and new Minister of Revenue, Zhao Mian 趙勉 (jin shi 1385), the emperor laid down his principles for conducting state finance: On this day I command you, Mian, to act as Minister of Revenue, to measure revenue in order to determine expenditure so as not to harass the people, and to always keep to public interests and avoid private interests.1   As a household advice, measuring revenue to determine expenditure (liang ru wei chu 量入為出) means not to spend more than what one has. As an ideal of governance, it is connected to the idea of maintaining the public interest and avoiding private gain, and is part of a wider principle of managing wealth locally rather than amassing it in the political centre.2 The earliest record of this idea is in the ancient Confucian classic, Li ji 禮記.3 In the chapter dedicated to the king’s rule, it describes the way in which the national budget should be prepared, summarizing income at the end of the year and                                                       1 Ming Taizu Shi lu, 明太祖實錄 [veritable records of Ming Taizu] (Beijing: Zhonghua shu ju, 1961-66): 202:1a. 2 Ming Taizu Shi lu, 135.2b-3a. The ruler, being the head of the “all under heaven,” stores wealth “all under heaven.”人君為天下之主當貯財于天下. 3 Li Ji 禮記, 5:29. (accessed: 12/18/2015). In the ancient record the phrase is 量入以為出.   2 planning the following year’s expenditure according to it. The founder of the Ming dynasty, Zhu Yuanzhang, was a former peasant educated by an elite group of advisors that promoted adherence to conservative Confucian ideals of government. Together, they attempted to shape state and society according to these principles and bring China back to what they viewd as Confucian rule after centuries of foreign domination. Zhu Yuanzhang personally experienced the abuses of a predatory tax system, in which local strong men acted as tax collectors and wielded arbitrary power over the population. He therefore established a small administrative structure that imposed minimal government intervention in the livelyhood of the common people. He kept expenditure at a minimum and devoted it entirely to the basic requirements of remunerating state servants.   But after the death of the first emperor the Ming expanded geographically and administratively. The transfer of the political centre back to the far north demanded vastly larger resources than did the first capital that nestled in the heart of the most fertile region in China. Population growth and growing social and economic complexity could not be controlled without a bigger administration. But in the choices that reformers made at key moments in the history of the dynasty, it is clear that they chose to adhere to the basic principle of keeping both revenue and expenses low rather than seeking ways to increase revenue. Instead of introducing change to the system of government finance, they made efforts to sustain the existing one. The government did not create fiscal institutions that could generate more revenue. Tax revenue did not reflect agricultural and commercial expansion. There were no official financial mechanisms of borrowing and credit. During most of the dynasty the government stifled maritime trade and did not tap into its profits. And after the demise of scrip the state did not even control the production of money.   3  In his monumental work on state finance in the Ming, Ray Huang exposed the inadequacy of state institutions to collect tax and distribute it in an efficient way.4 In his detailed account of various aspects of Ming financial administration Huang points at organizational weakness as the main characteristic of Ming government in the sixteenth century. He exposes the cumbersome process of tax assessment and collection, the atrophy of tax data, and the surcharges imposed on taxpayers in order to finance the process of payment and delivery of taxes. His conclusion that the Ming state was under-taxed overhauled the standard interpretation of the fall of the dynasty that placed the blame of the late Ming mass rebellions on heavy taxation. Still, low taxes did not mean that commoners were better off. Without sufficient funds for the proper functions of government, officials resorted to informal and at times extortionist means of extracting revenue. When faced with growing expenditure it did not have the political aptitude to initiate meaningful reform. It was therefore doomed to operate in a perpetual state of financial crisis.   As Huang states elsewhere, fiscal policy in the Ming suffered from its adherence to Confucian principles of government that was more literal than in other dynasties.5 This dissertation starts out from this assumption and asks what kind of fiscal policy did Ming administrators nevertheless conduct? And what structural or practical means where employed in order to over come the limitations of this approach to state finance? In                                                       4 Ray Huang, Taxation and Governmental Finance in Sixteenth-Century Ming China (London: Cambridge University Press, 1974). 5 Ray Huang, “Fiscal Adsministration During the Ming Dynasty,” Chinese Government in Ming Times (New York: Columbia University Press, 1969): 73-77.   4 examining Ming fiscal policy according to its own principles, this dissertation offers an alternative approach to understanding financial governance in the Ming. It first argues that the model of agrarian taxation was at its best during the fifteenth century when the political centre was able to obtain financial information from the provinces and maintain good communication along the administrative hierarchy. Secondly, it identifies tax commutation as the primary fiscal mechanism employed in order to adapt to changing economic circumstances, and demonstrates how large-scale commutation of the grain tax into silver impacted fiscal administration. The final three chapters examine the role of commutation in expenditure, concentrating on official payments to the three main groups of state servants: officials, soldiers, and members of the imperial clan.  In order to highlight the unique character of the Ming and to place it in the context of the history of fiscal policy in China, this introductory chapter examines key aspects of fiscal policy in the Ming against notable developments that took place mainly during the Song. In doing so, this discussion utilizes historical categories of fiscal systems that describe a developmental trajectory. One of the peculiarities of the Ming is that whereas the Song developed sophisticated fiscal institutions that were partially continued during the Yuan the Ming did not continue along the same path. Instead, it employed fiscal policy that was determined by institutional limitations and supported by ideological opposition to fiscal expansion. The second half of this chapter introduces the practices of accounting and commutation of revenue and expenditure as two aspects of this fiscal policy that contributed to the viability of the dynasty.  First, a distinction should be made between the terms “fiscal policy” and “fiscal state.” The term “fiscal” pertains to public revenue and was originally associated with the  5 wealth of the king, which was later identified with public revenue.6 Despite its modern ring, in its broadest sense the contemporary definition of the term “fiscal policy” can be applied to imperial China. The International Monetary Fund defines it as “the use of government spending and taxation to influence the economy.” And, it is added, the purpose of fiscal policy is usually to promote growth and reduce poverty.7 In the Ming, the promotion of growth and the reduction of poverty took place in an agrarian context. And the peculiarity of its fiscal policy is that it remained so even as the economy became more commercialized. This is why it is important to distinguish between fiscal policy and the term “fiscal state.” As it is used in historical research, the latter designates the most advanced stage of financial administration that characterizes contemporary states. Fiscal states by definition exist in developed commercial societies. Therefore, although the Ming dynasty was not a fiscal state in terms of this modern definition, it did conduct fiscal policy within its own governmental framework.  The exploration of historical change in fiscal systems goes back to Joseph Schumpeter who described the transition between the feudal and capitalist eras in fiscal terms.8 While feudal rulers of domain states extracted revenue in kind, the tax state that replaced it operated within a capitalist system in which the goal was to increase tax                                                       6 The earliest example of the term in the Oxford English Dictionary is from 1570. Accesed 12/19/15. 7 Mark Horton and Asmaa El-Ganainy, Fiscal Policy: Taking and Giving Away,” Accesed 09/22/15.  8 Josef A. Schumpeter, “The Crisis of the Fiscal State,” (1918), reprint The Economics and Sociology of Capitalism, edited by Richard Swedberg, 99-140 (Princeton, N.J.: Princeton University Press, 1991).   6 revenue by promoting and interacting with a commercial economy. This historical perspective has since been expanded to other periods and countries.9 With the advantage of numerous case studies, spanning from the Middle Ages to the early modern period across Europe, the distinction between domain and tax states has been modified so that it no longer represents a modern-premodern binary. Instead, characteristics of domain and tax states can both be found in the early modern period. Furthermore, the addition of the concept of the fiscal state provides another distinction in the degree of sophistication between financial systems.10 In the following pages elements of fiscal policy during the later period of imperial China, spanning from the Tang-Song to the Qing, are analyzed according to these categories in order to elucidate the character of Ming state finance.  1.1 Distinction Between Private and Public Revenue  An essential aspect in Schumpeter’s distinction between domain and tax states regards an institutional separation between private and public revenue. This insistence on institutional separation between public and private wealth has obscured the meaning of this distinction in the context of premodern China. Using Schumpeter’s own example, when a Chinese emperor went to war it was not his personal affair as was the case with contemporary European kings. When addressing issues of state finance officials used the                                                       9 Richard Bonney ed., Economic Systems and State Finance (Oxford: Oxford University Press) 1995; Richard Bonney ed., The Rise of the Fiscal State in Europe, c. 1200-1815 (Oxford: Oxford University Press) 1999; W. M. Ormord, Richard Bonney and Margaret Bonney eds. Crisis Revolutions and Self-Sustained Growth: Essays in European Fiscal History 1130-1830 (Stamford: Shaun Tyas) 1999. 10 W. M. Ormord, Crisis Revolutions and Self-Sustained Growth, 4-8.  7 term guo yong (國用), or ‘state expenses’ that explicitly invokes the public domain, as do the words of the first emperor stated above warning the new minister to avoid private interests (私, si) and uphold the public good (公, gong).   That said, whereas European tradition created institutional and legal boundaries between these spheres, the Chinese imperial institution embodied them both. Ironically, in imperial China there was no clear separation between public wealth and the private wealth of the ruler because such a separation would acknowledge that the emperor was a private individual with material interests. Therefore, the classic statement that the whole world is the emperor’s home does not mean that it is his private domain. On the contrary, the emperor belongs within the public domain and should refrain from exploiting it. Hongwu himself explained this in an early discussion on state finance with one of his officials who urged to increase reserves at the political centre. After the official argued that the emperor should store funds just like any other head of household, the emperor replied that the entire realm is his household, and as such his wealth is stored throughout the realm.11 As shall be discussed in the following chapter, the arrangements for supplying the capital were vague during the Hongwu reign and depended on agricultural surplus. When giving the order to build a palace treasury (nei fu ku, 內府庫) in the capital, Hongwu did so with mixed feelings, citing the accumulation of vast sums of money in the Song imperial treasury.12 He accepted the interpretation of the History of the Song blaming emperors using public revenue for personal use while their armies were                                                       11 A translation of this discussion appears on pages 56-57.  12 Ming Taizu Shi lu, 179:1a.  8 in need of funds, and thought of the imperial treasury as a potential threat to public finance.  This was only part of the truth. In the Song the imperial treasury was an important financial institution. The imperial treasury profited from sales of various items and commodities that were received through tribute. It lent funds to the state treasury and often demanded them in return. In this sense, it operated as a business enterprise that contributed to public finance.13 The Ming palace treasury too served as an emergency fund chest. Stocks of precious metals, silk, commodities such as sappan wood and pepper, as well as scrip in the early Ming were issued to buy supplies for the military and sent as aid to famine stricken households. But in the Ming the imperial treasury functioned only as a giant warehouse. As in the Song, eunuchs managed the various commodities, stored in a cluster of storehouses. Revenue officials occasionally took inventory of the storehouses, but they did not have full access to the accounts. Control was ultimately in the hands of the eunuchs.14  Although the imperial treasury functioned as a public repository, there was no legal prohibition to imperial appropriation of funds. For example, in the Song the costs of housing and stipends for imperial kin were mostly taken from the state treasury. The Ming administrative code was deliberately vague in regards to the income of the imperial family. Funds were to be derived from the leftovers of tax revenue that was retained in                                                       13 Robert M. Hartwell, “Finance and Power: The Imperial Terasuries in Song China,” Bulletin of Song and Yuan Studies 20 (1988): 18-89.  14 Huang, Taxation and Governmental Finance, 9.   9 the provinces.15 Zhu Yuanzhang followed the tradition of Confucian political culture and preferred instilling the notion of the priority of public interest by educational means rather than legal or institutional ones. The problem of course was that maintaining the integrity of public revenue became a matter of the emperor’s personality.  The difficulty to distinguish between domain and tax state in imperial China becomes apparent when considering the role of the palace treasury. In accordance with the definition of a domain state, the emperor derived his income from the economy through imperial prerogative but it was often used for supplementing public expenditure. Instead of a narrative of separation between private and public funds, statecraft thinkers distinguished between keeping funds on the local level and centralizing funds in the political centre. The means to prevent imperial appropriation of public funds was to prevent excess funds from reaching the imperial palace in the first place. As shall be discussed in chapter 3, growing difficulties to maintain large military units along the northern border necessitated a concentration of funds in the capital and undermined this ideal.   1.2 Indirect Taxation    While the Song and Ming emperors derived their wealth in a similar way, the two dynasties played radically different roles in the economy. As stated earlier, a primary distinction between the Ming and its predecessors was its focus on agricultural sources of                                                       15 See pp.84-85.  10 revenue. This prevented the state from reaping profits from commercial and industrial enterprises. In the historical literature, indirect taxation is not only a feature of an advanced tax state but also a prerequisite of a developed fiscal state. In imperial China indirect taxation first became an important source of state finance through the utilization of a state salt monopoly. Near the end of the Northern Song, revenue from salt reached half of total revenue.16 In contrast, in the late Ming, revenue from salt constituted around 6.5% of total revenue.17 The salt monopoly was structured according to the general rule of decentralization under the sole leadership of the emperor that characterized Ming administration. With no director of its own, salt production and distribution centres operated separately under the Ministry of Revenue that had no special staff dedicated to their operation. The government chose to maintain as little control as possible over the production and sale of salt. Beyond a certain quota, salt producing households were free to sell their salt to merchants, and households that turned to growing crops on the land allocated to them paid the equivalent of their tax quota in grain.   Furthermore, in the fifteenth century the government sacrificed the revenue it could have earned from a tax on salt in order to promote the circulation of scrip. By the mid fifteenth century an average of 20,000,000 guan of scrip from the sale of salt was                                                       16 Cecilia Lee-Fang Chien, Salt and State: An Annotated Translation of the Songshi Salt Monopoly Treatise, (Ann Arbour, MICH: Centre for Chinese Studies, University of Michigan, 2004): 36.   17 Fang Jun, China’s Second Capital: Ming Nanjing Under the Ming, 1368-1644 (Oxon: Routledge, 2014), 102. Still, it should be said that within its modest size, the salt monopoly did expand production during the fifteenth century from an average of fewer than 1,000,000 to between 3000,000-5,000,000 yin  (引, units of between 120 and 240 kilos). See tables D-3,6,10,13,16.  11 reported.18 In silver value the income would be around 600,000 taels. Actually, the total value was much lower since many of the notes were damaged and collected for the specific purpose of removing them from circulation. Salt was also exchanged for grain transportation to the border. This policy cut transportation expenses but did not contribute to available funds at the political centre. Revenue from salt was collected in silver only at the end of the fifteenth century. It was therefore only then that it began to contribute to the income of the central government.  The Song also expanded its revenue from commercial enterprises. William Liu contrasts over 2,000 tax stations across the Song realm with only 125 stations operating in the Ming. 19 This assessment of Ming tax stations seems too modest. According to Brook, there was a tax station in every county (1,144 in the late Ming).20 Also, as shall be discussed in the following chapters, since in the fifteenth century official and military salaries were partially paid in scrip, and scrip was collected through commercial taxes, localities probably obtained it through local commercial taxes. Furthermore, there is no doubt that commercial taxes were not a significant source of total revenue, but recent work on local finance suggests that commercial taxes may have provided an important revenue source for magistrates. Local officials could temporarily set up a tax station in                                                       18 See tabels D-6,10.  19 William Guanglin Liu, “The Making of A Fiscal State in Song China,” Economic History Review, 68:1 (2015): 67. 20 Timothy Brook, “Communications and Commerce,” Denis Twitchett and Frederick W. Mote eds. The Cambridge History of China: The Ming Dynasty Part 2 vol. 8 (Cambridge: Cambridge University Press, 1998): 675-676; Huang maintains that most tax offices closed down in the sixteenth century. See: Huang, Taxation and Governmental Finance, 232.   12 order to collect revenue for a specific purpose such as building a city wall.21 He Zhaohui provides a glimpse into the struggles of a late sixteenth century magistrate to meet the expenses of daily administration. After the central government gained more control over local funds through the Single Whip Reform, there was very little flexibility in tax collection to enable local magistrates to supplement their budgets. One of the last independent components of local revenue came from commercial taxes, which in the case of Magistrate Yuan Huang (袁黃) in Baodi County provided him with extra funds to meet unexpected expenses.22 Indirect taxation also stands at the centre of Qing state finance in the nineteenth century. Stephen Halsey analyzes the transformation of the Qing fiscal system according to the model of a “military fiscal state.” 23 In order to cope with loss of agricultural revenue during the Taiping rebellion, the state moved from direct taxation of agriculture to indirect taxation of trade.24 He Wenkai uses this conclusion as a starting point for a comparative study of the ways in which England, Japan, and China dealt with fiscal crisis. England during the Civil War, Japan during the Meiji Restoration, and China during the Taiping rebellion all faced fiscal crises. The three countries employed similar                                                       21 Brook, “Communications and Commerce,” 676.  22 He Zhaohui, “Stuggling for Balance: The Finances of Baodi County Around Wanli 20,” Ming Studies 67 (May, 2013): 54-71. 23 Jan Glete, War and the State in Early Modern Europe: Spain, the Dutch Republic and Sweden as Fiscal-Military States, 1500-1660 (London: Routledge, 2002). In this work, Glete points at developments in military technology and rising military expenses as the cause for fiscal innovation. 24 Stephen Halsey, “Money, Power, and the State: The Origins of the Military-Fiscal State in Modern China,” Journal of the Economic and Social History of the Orient, (vol. 56, no.3, 2013), 392-432.   13 methods to overcome these crises, but unlike England and Japan, which managed to transform into centralized fiscal states, the Qing’s financial system remained decentralized. He Wenkai’s definition of a modern fiscal state as one that “centrally collects indirect taxes in order to mobilize long-term financial resources from the markets.” 25 This definition adds the important aspect of centralization of tax collection. While indirect taxation of trade is conducive to centralization, He Wenkai shows that the Qing failed to do so. The Taiping rebellion disrupted commercial as well as official networks of trade and communication. The result was that the Qing had to further decentralize its financial administration, a move that prevented it from developing the institutions that supported government credit in Japan and England, among them is paper currency that will be discussed below.  1.3 Money  Throughout most of the premodern world, control over money meant control over precious metals. And one of the primary means of gaining control over the supply of precious metal was through mining and minting operations. During the Song, the mining of copper reached industrial proportions that were not matched anywhere in the world before the nineteenth century. In the Song, annual quotas ranged in thousands of tonnes. The peak was around 1070 with almost 13,000 tonnes. Excessive mining led to the                                                       25 Wenkai He, Paths Toward the Modern Fiscal State: England, Japan, and China, Cambridge, Mass.: Harvard University Press, 2013.  14 depletion of copper reserves by the second half of the twelfth century.26 In contrast, quotas for government receipts from copper mining in the Ming are registered during the first half of the fifteenth century and amount to around one tonne.27 Accordingly, the Song was minting over a million strings of coins a year in the eleventh century, while the revenue reports register a negligible quota of 3,000 strings for a short period during the Yongle reign.28   Monetization is an aspect of an advanced tax state, though it is not necessarily so. For example, in the late twelfth century, the Hungarian king Béla III (r. 1172-1196) had greater silver revenues than the kings of England and France. The majority of revenues were derived from profits from mines and mints, the salt monopoly, and customs duties. But although these were revenues derived from a developed monetary market economy, they were all derived by royal privilege: “Behind an apparently modern fiscal system lay, in fact, the might of a ruler whose power rested in his position as the greatest landholder in the country.”29 On the other hand, assuming that the capability to institute direct                                                       26 Peter J. Golas, “Mining,” Science and Civilization in China vol. 5 part 13 (Cambridge: Cambridge University Press, 1999): 88.  27 See tables D-2,6.  28 Richard von Glahm, Fountain of Fortune: Money and Monetary Policy in China, 1000-1700, 49 (Berkeley: University of California Press, 1997). For Ming miniting quotas see table D-2.  29 Gabor Barta and Janos Barta, “Royal Finance in Medieval Hungary: The Revenues of King Béle III,” in Oromond, Crisis, Revolutions, 22-37.   15 taxation necessitates a more complex taxing apparatus, Ormond finds that the first attempt towards universal taxation in England involved taxation in kind and not money.30 In this sense, the degree of monetization in an economy does not in itself testify to the degree of administrative sophistication. But, as discussed in the previous section, it does increase the degree of efficiency in concentrating tax revenue in the political centre. This happened in the sixteenth century as silver eclipsed revenue in kind. Tang Wenji estimates the silver revenue from the grain tax in 1531 at 6 million taels.31 In addition, Ray Huang estimates the annual silver revenue from non-agricultural taxes and levies in the 1570’s at around 12,000,000 taels. 32 Calculating the silver value of the grain tax income estimated for 1531 according to a contemporary average government exchange rate of 0.7 taels would yield 15,400,000 taels. This means that silver income surpassed grain income that year. As for the higher grain tax income reported in 1578, valued at 18,647,049 million taels, silver and grain income would be about equal. It can be said, therefore, that only in the sixteenth century did the Ming exhibit a high degree of monetization in state finance.                                                       30 Ormond, “The West European Monarchies in the Later Middle Ages,” Economic Systems and State Finance, 123-162. 31 Tang Wenji, Ming dai fu yi fuyi zhidu shi 明代賦役制度史 [A History of the Tax and Corveé System in Ming China] (Beijing: Zhongguo she hui ke xue chu ban she: xin huan shu dian jing xiao, 1991): 188. AThis was about 27% of an annual total of 22,000,000 shi.  32 Huang, Taxation and Govermental Finance, p.164 – annual grain tax revenue for 1578: 26,638,642; p. 214 - estimated income from the salt monopoly: 1,292,224 taels; pp.263-265 - estimations of cash revenue from commercial and other taxes: 4,634,000.   16 The ambivalence of the Ming towards silver is apparent in Qiu Jun’s writings on the subject in his late fifteenth century magnum opus, Da xue yan yi bu. In an essay on currency, Qiu Jun proposes to acknowledge the superiority of silver in a three-tire system of currency along with coin and scrip. 33  In another essay he also acknowledges the benefit of commuting labour services into silver, another important source of revenue.34 But his opposition to silver mining exemplifies the inherent contradiction in the Ming approach towards the precious metal. Qiu mentions the various uses people make of silver as a means of conspicuous consumption, displayed in such conspicuous forms as ornaments, household utensils, and images of gods. He argues that silver mining would just add more silver that will be used for conspicuous consumption, and should therefore be kept to a minimum.35  In this section Qiu does not even mention the role of silver as money. In other words, he accepts the existence of silver as a means of payment but does not believe the state should be involved in increasing its quantity in the economy. This approach also elucidates the contrast between the Ming and its Western contemporaries. In contrast to early modern European polities that built their fortunes on silver and gold mines and sponsored expeditions to the other side of the world in search of gold and silver, there was no attempt to seek silver internationally to compensate for inadequate domestic production. Even the relaxing of trade restrictions in the second half of the sixteenth century was more a result of concern over social order than a fiscal                                                       33 Qiu Jun, Da xue yan yi bu, 大學演義補 [Supplement to the Commentary of the Great Learning] (1487, repr. Si ku quan shu zhen ben er ji, 603-666, Taiwan shang shu yin shu guan, 1971): 27:16b-17a. 34 Ibid, 31:20a-21a. 35Ibid, 29:12b-15a.   17 measure.36 Even in the late fifteenth century, when silver became crucial in financing Beijing and the border armies, Qiu Jun declared in the concluding words of his essay on government mining: “Comparing our dynasty’s taking of the people’s profit with that of former eras, its benevolence and generousness in benefitting the people are truly profound.”37   The institution of paper currency alludes to the sophistication and strength of the state system in imperial China, and distinguishes the dynasties that employed it from any other country before the modern period. During the Song dynasty and for the first two decades of its existence in the Yuan, it succeeded in increasing the money supply and expanded the resources available to the state. It served as a supplement to coin during the Song and was issued for limited periods after which it was bought back by the state using hard currency. The state further backed its scrip by allowing the notes to be convertible. Yuan scrip served as a replacement for bronze coin that had assumed a minor role in the economy. During the first twenty years of national circulation, Yuan scrip was convertible with silver, but thereafter the state ceased to back its notes.38   In the Ming, scrip and coin were intended to circulate side by side. Coin from previous ages was allowed to circulate alongside the officially minted coins due to insufficient quantities. Still the amount of coin in circulation was never enough for it to                                                       36 von Glahn, Fountain of Fortune, 114-118. 37 Qiu Jun, Da xue yan yi bu, 29:15b. 38 38 Herbert Franz Schurmann, Economic Structure of the Yuan Dynasty (Cambridge, MA: Harvard University Press, 1956).  18 be instrumental in state finance. Although paper currency existed in China from the tenth to the fifteenth centuries, its management declined over time. The Ming continued to issue scrip with no external backing. The early Ming made all the possible mistakes in trying to enforce its paper currency on the market. The government stopped minting coin, and banned silver and gold from market transactions, thus contributing to the rise in the value of these commodities against scrip. With the additional harm of excessive issuing, Ming scrip devalued beyond repair within a few decades. Thereafter, there was no attempt to mint silver even as it became the common medium of exchange. Instead, administrators were content with supervision of market prices and determination of official exchange rates.    He Wenkai’s discussion on the failure to issue paper currency in the Qing adds further insight. The attempt to reintroduce paper currency into the economy failed in the wartime conditions of the Taiping rebellion. The government created provincial offices for the dissemination of scrip and intended to support the notes by provincial stocks of specie. But provincial stocks were not sufficient to guarantee convertibility. Moreover, the dissemination of paper currency was unsuccessful. While the government paid soldiers and construction workers in paper notes, they often had to commute them into silver or bronze in order to purchase supplies or equipment in local markets. There were some local successes in issuing paper notes such as Jiangsu, where bronze coin from the  19 local custom taxes served as security. But, as mentioned above, the failure of the Qing to centralize fiscal resources left the currency with no sufficient backing.39   In comparison, the early Ming built a financial apparatus that was deliberately decentralized. Furthermore, the state never built any mechanism into its paper currency that would ensure its creditworthiness. Ming notes were never secured by any government pledge to redeem them. Finally, it could be argued that just as the economic and political conditions in wartime Qing China were detrimental for establishing national paper currency, they were not ideal in the early Ming either. China in the 1370’s was a war-ravaged country that was just beginning to recuperate from economic devastation. The trading networks and economic surplus that were needed to support a means of credit such as scrip were not yet in place. To that we should add the war of succession at the turn of the fifteenth century, which probably further diminished confidence in the political viability of the Ming. By the time the economy recovered and stabilized, paper notes were largely discarded.   1.3.1 Government Credit  As William Liu proposes, the Song management of paper currency suggests that there was an understanding that government-issued notes are essentially a form of public debt. By redeeming notes periodically, and using scrip as an additional form of currency                                                       39 He Wenkai, “Economic Disruption and the Failure of Paper Money in China, 1851-1864,” Paths Towards the Modern Fiscal State,” 131- 152.   20 alongside coin, the state was managing a system of short-term credit. In addition, the Song allowed the payment of taxes in scrip, a mechanism that could be seen as another form of state backing. This was the only active policy the Ming employed in order to encourage the circulation of its notes, but as shall be discussed in Chapter 3, this was not enough to maintain their value. And in any case the state largely backed away from a commitment to collect tax in scrip early in the fifteenth century.  Liu further argues that the salt certificate system (ru zhong 入中) of the Song operated as a form of public debt. By issuing salt certificates to merchants in return for transportation of grain, the state was indebted to them until the salt was redeemed. But more importantly, merchants were able to trade in these vouchers in special voucher stores in Kaifeng.40 The interpretation of the salt exchange system as a form of public debt is also at the centre of Winkin Puk’s discussion of the similar salt certificate policy (kai zhong, 開中) of the Ming.41 Despite the similarity between the two policies, the appearance of a form of public debt is much clearer in the Song, when the state was actively promoting trade in government credit. In contrast, the Ming response to trade in salt vouchers was to fix their price in order to prevent speculation.                                                          40 Liu, “The Making of the Fiscal State,” 68-72. 41 Win-Kin Puk, “The Ming Salt Certificate: A Public Debt System in Sixteenth Century China?” Ming Studies 61 (Spring, 2010): 1-12. This issue will be discussed in more detail in chapter 5.  21 1.4 Financial Administration and Expertise  In comparison with their predecessors from the Tang to the Yuan, the status of financial officials was greatly reduced in the Ming. The Ministry of Revenue had little capacity to initiate reform and its role was mainly supervisory. In contrast, the Song exhibited considerable fiscal organization and expertise. During its first hundred years, the State Finance Commission (san si 三司) was one of the three highest branches of government, placed above the Ministry of Revenue. It contained three separate agencies: The Salt and Iron Monopoly supervised much more than the production and sale of these commodities, for it also handled the production and distribution of military supplies, paid salaries to military officials, oversaw communications, and collected commercial taxes. A special tax bureau, literally the Office of Measuring Expenses (du zhi si 度支司) prepared annual financial reports on receipts and disbursements, and paid the salaries of civilian officials. The third agency was the Census Bureau (hu bu) and it was responsible for maintaining population records, collecting agricultural taxes (along with the wine monopoly tax that was attached to them), and handling the long-term storage of goods. These three agencies had independent directors who were not subordinate to the central government organs but answered directly to the emperor. 42 In the Ming, by comparison, the duties of these three agencies were all under the Ministry of Revenue, meaning that a much smaller staff was allocated to these duties.                                                        42 Peter J. Golas, “The Sung Fiscal Administration,” The Cambridge History of China Vol.5 Part 2 (Cambridge: Cambridge University Press, 2015): 139-213.  22  The Song employed on the local level an entire class of officials devoted to financial matters. On the circuit level, fiscal commissioners handled all financial tasks, especially commercial taxes. The closest equivalent to this post would be the Ming provincial administration commissioner (bu zheng shi 布政使) who, along with a military and surveillance commissioner, headed provincial administration. His duties encompassed civilian, and in particular financial matters.43 Further down the hierarchy, the post of vice prefect was an innovation of the Song, and came to be devoted to tax collection and other financial matters of the prefecture. The dedication of administrative posts to financial matters on the prefecture level was completely absent in the Ming. Moreover, the first emperor attempted to exclude officials from tax collection altogether by instituting the li jia (里甲) system, which organized society in tax units of approximately 110 households from which the most prominent were responsible for collecting tax and transporting it on a rotating basis. Once this sub-bureaucratic taxation system broke down, the role of collecting tax fell on the county magistrates, along with their other duties. This change meant that state agents became more directly involved in tax collection but with no additional posts to support them.   Furthermore, there is evidence that Song financial officials reached a degree of professionalization and prestige that was not possible in Ming political culture. Robert Hartwell conducted a survey of the careers of high Song officials showing that they all                                                       43 Charles O. Hucker, A Dictionary of Official Titles in Imperial China Part 1 (Stanford, Calif.: Stanford University Press, 1985): 76.   23 had extensive experience in a variety of financial posts.44 His work also made it apparent that a particular body of knowledge on economic and financial subjects was studied by imperial students and served as material in the imperial examinations. This knowledge was not lost in the Ming, but it did not have much value in official careers. First, instead of encouraging specialization, the Hongwu emperor tended to shuffle his heads of ministries around every few years, presumably to prevent any one of them gaining too much power. Second, the most prominent officials involved in state finance were not necessarily connected with the Ministry of Revenue. As shall be discussed in chapter 2, an official leading major tax reform in Jiangnan in the mid-fifteenth century served in the Ministry of Works. This was also the case in Qing government. Revenue ministers were appointed after successful careers in other branches of government. 45  Little regard for financial expertise was based on an inherent suspicion of fiscal policy. While Song statesmen distinguished between sound fiscal policy and sheer opportunist fiscalism, Ming statesmen assumed that such policies would ultimately come at the expense of the common people. The gulf between the two attitudes is apparent in the different appraisals of Liu Yan (劉晏, 716-780), the reformer of the Tang salt monopoly who turned it into an extremely beneficial fiscal apparatus. While an eleventh century writer praised his ability to increase revenue, put the dynasty on a sounder                                                       44 Robert M. Hartwell, “Financial Expertise, Examinations, and the Formulation of Economic Policy in Northern Sung China,” Journal of Asian Studies, 30, (1971): 281-314.  45 E-Tu Zen Sun, “The Board of Revenue in Nineteenth-Century China,” Harvard Journal of Asiatic Studies 24 (1962-63): 180-181.  24 economic base, and refrain from raising agricultural taxes, 46 Qiu Jun, the leading statecraft writer of the fifteenth century wrote in a treatise on the salt monopoly: Although Liu Yan was good at finances, he understood that profiting the state is of benefit but he did not understand that profiting the people is of even greater benefit. He knew how to monopolize a certain source in order to gain profit but he did not know how to spread out the taking of profit in order to gain an even greater benefit.47  Qiu Jun held the Confucian mainstream opinion that the state should not extract high profits from a particular source, but spread out the burdens of taxation equitably. Later on in his essay he argues that even if certain fiscal policies have some short-term benefit, in the long run they create an atmosphere in which the state is competing for profit with the people.   1.5 Another Approach to Financial Administration  The works cited above, and many others, use concepts from European economic history in conducting a meaningful exploration of similar dynamics in imperial China. They represent an intellectual climate that is free from the interpretation of the history of late imperial China along the trajectory of early modern Europe, previously revolving around the question of “why did China fail to reach modern Western-style capitalism?”                                                       46 The Song essayist is Zhang Fangping who served as financial commissioner twice during the mid eleventh century. A partial translation of his essay appears in Hartwell, “Financial Expertise,” 297.  47 Qiu Da xue yan yi bu, 28:9a.   25 This question has been put to rest after R. Bin Wong’s path breaking approach to the comparative economic history of Europe and China. By comparing eighteenth century China to its most advanced counterparts in Europe, Wong shows that the two ends of the Eurasian continent were largely comparable. He further argues that the reasons for the European breakthrough from the limits of premodern economic growth cannot be attributed to any failure on China’s behalf but to the special conditions that emerged in Europe. The advance of commercial capitalism owed to the development of means of long and short-term credit encouraged by debt-ridden rulers, and international trade. England’s industrial advantage owed much to its ability to tap into large deposits of coal.  To these structural differences Wong adds the significant distinction in Chinese and European early modern political economy. In China the Confucian notion of benevolent government took root. The role of the state was to promote an agrarian economy and ensure an equitable distribution of wealth. Commercial activities were tolerated or encouraged as long as they contributed to this ideal and did not create large income gaps. Moreover, it was not the business of the state to pursue wealth since that would entail taking from the people what is rightly theirs. In contrast, in a growing atmosphere of competition between European nation states, the richer a country became, the better chance it had to overcome its competitors. Questions of public welfare and equitable distribution of wealth were secondary in Europe, whereas the pursuit of wealth as a political goal was repugnant to most Chinese administrators.  The consideration of traditional Chinese financial administration according to different principles can be supported by a different theoretical approach to fiscal policy. Whereas Schumpeter’s definition of fiscal policy concerns the means by which states  26 influence the economy in order to increase tax revenue, fiscal policy is also directed at promoting public welfare. Or in other words, states use taxation in order to distribute wealth in society. The economist Richard Musgrave was the first to propose a theory of this aspect of fiscal policy, which he termed as “public finance,” in the English-speaking academic community. The relatively late appearance of works on this aspect of fiscal policy helps put in sharper relief the distinction between Western and Chinese traditions of political economy. In imperial China, ensuring public welfare was the first mandate of the state and its raison d’être. Musgrave identifies three main roles the state undertakes to distribute wealth in society. Though working within a modern democratic context, they can be applied with some modifications to Ming China.   The first role is that of allocation, the process of transferring private resources to predetermined public wants. The principle of “using the realm’s wealth for public purposes” (yin tian xia zhi cai gong tian xia zhi yong, 因天下之財供天下之用) is the Confucian equivalent of this principle.48 “Public wants” in imperial China related first to food security. This principle was applied through allocating portions of tax revenue to ever-normal granaries and special issues of famine relief. In addition, tax remissions and tax breaks were another form of allocation of funds for public wants since tax remission was a form of subsidy for struggling farmers. In the Ming this was considered as an essential part of financial administration; from 1426, tax remissions were even part of the annual revenue reports. The obligation to provide security from foreign invasion and attack was another form of allocation that took up the bulk of public revenue. But, as will                                                       48 Ming Taizu Shi lu, 179:1a.  27 be discussed below, growing military expenses in the Ming did not lead to the development of more advanced fiscal institutions.    The second role Musgrave cites is distribution, in which the state intervenes in the economy in order to diminish inequality in society. Equalizing the tax burden (jun fu, 均賦) was a common feature of tax reform and will be discussed in the context of the reforms conducted in Jiangnan in the mid-fifteenth century. Huang offers an important observation that in the Ming, officials generally preferred economic equality to development:  As Ming administrators saw it, to promote those advanced sectors of the economy would only widen economic imbalance, which in turn would threaten the empire’s political unity. It was far more desirable to keep all provinces on the same footing, albeit at the level of the more backward sectors of the economy.49  The third role is that of stabilization. In the context of a democratic market society this means maintaining high employment and controlling inflation through adjusting tax rates.50 Though unsuccessful, the Ming did attempt to overcome inflation through taxation. The Household Salt Consumption Tax of the fifteenth century was a universal tax designed specifically for this purpose. Although there was logic in the assumption that decreasing the number of notes in the market will strengthen their relative value against commodities, the Ming did not exercise enough control over its currency in order to reduce inflation. In addition, low tax rates and tax remissions were a means to maintain                                                       49 Huang, Taxation and Governmental Finance, 2. 50 Richard A. Musgrave, The Theory of Public Finance: A Study in Public Economy, New York: McGraw-Hill Book Company (1959), 3-27.   28 social order. It was tax debt that usually drove peasants to desert their lands, a problem that directly affected tax revenue and was considered a form of unemployment that threatened social stability. These methods had limited effect against social dynamics that increased inequality. Powerful and wealthy landowners were able to evade tax payments whereas growing numbers of small and poor cultivators were left with a growing tax burden.   These fiscal policies were considered to be basic principles of government in Imperial China. In following them the Ming was not different from other dynasties. But unlike their predecessors, Ming administrators kept holding to the notion that if the state extracts more revenue from the economy, it is necessarily taking from the people. Therefore, the development of more sophisticated fiscal practices, such as those discussed above for the Song, was discouraged in the name of public welfare.   Within the boundaries of classical Chinese political economy the debate on the degree of state involvement in the economy, famously summarized in the Discourse on Salt and Iron (yan tie lun 鹽鐵論), delineated two extremes of fiscal policy.51 Whereas the Song operated according to one extreme of this debate, the Ming adhered to the other. If Ming administrators had not been aware of Song policies, its seemingly anti-fiscal approach could be explained as ignorance. But since history is the one field in which Chinese literati cannot be blamed to lack knowledge, we must conclude that the                                                       51 Huan Kuan 桓寬, Yan tie lun鹽鐵論 [Discourses on salt and iron] (1st century B.C., reprint, Taipei: Yi wen, 1967). For an English translation see: Esson M. Gale, Discourses on Salt and Iron: A debate on State Control of Commerce and Industry in Ancient China, Chapters X-XIX, (Leiden: E.J. Brill, 1931).   29 abandonment of some of the more market-oriented policies of the Song was deliberate. Indeed, from the point of view of the Ming founding elite, they were correcting the errors of Song administration. What to modern eyes is the most advanced period of fiscal policy until the early modern period was perceived as a deviation from the principles of good government. Early Ming policies of agricultural restoration were extremely successful in the first decades of the dynasty but they were insufficient to support the dynastic vision of the Yongle emperor. Beyond the expensive military campaigns and maritime expeditions, moving the capital to Beijing and the huge expenses entailed in supplying it destroyed the tax system of the founding era. Moreover, a tax system that was designed for a stable agrarian population was incompatible with the dynamism and commercialism of the mid and late Ming periods.   The story of state finance in the Ming is therefore one of practical solutions to an inadequate state system. Ming administrators employed solutions that complied with the principle of determining expenditure according to revenue.  But they had to employ financial strategies that went beyond the management of a simple agricultural economy. These policies had a very modest goal; they did not intend to increase state revenue but to reduce expenses and ensure a stable flow of funds throughout the empire and to the capitals. The following section introduces the structure of financial administration as well as major fiscal policies that took place in the first half of the dynasty as manifestations of this principle.    30 1.6 Distinguishing between the Fifteenth and Sixteenth Centuries  In concentrating on the first half of the Ming, this dissertation answers the call for more research on the fifteenth century.52 Almost thirty years ago, Edward Farmer alerted that this period remains somewhat of a black hole in the study of the Ming. Compared to the dramatic events and outstanding personalities of the founding period (1368-1424) and the late Ming period (1522-1644), it appears to be relatively uneventful. Nevertheless, he argued that knowing more about fifteenth-century Ming China is crucial for understanding the trajectories that set early and late Ming so much apart.   Thomas Nimick contributed greatly to our understanding of fifteenth century governance in his discussion of the transformation of local administration. This change originated in the disintegration of the li jia system, the basic mechanism of tax collection. He finds that during this period the Ming exhibited institutional flexibility and creativity that successfully met the challenges of changing social circumstances.53 The central government managed to regain control over local administration through specially appointed prefects who acted as representatives of the central government in the local community and at the same served as defenders of local interests against the demands of the political centre. This was the administrative context of the tax reforms initiated during the Xuande and early Zhengtong periods that will be discussed in chapter 3.                                                        52 Edward L. Farmer, “An Agenda for Ming History: Exploring the Fifteenth Century,” Ming Studies 26 (Fall, 1988): 1-17. 53 Thomas G. Nimick, Local Administration in Ming China: The Changing Roles of Magistrates, Prefects, and Provincial Officials (Minneapolis: Society for Ming Studies, 2008).  31  In the sixteenth century power shifted to the two top provincial administrators, the regional inspector and the grand coordinator. In order to control the growing population they relied on the magistrate who became the centre of all local government operations and held sole responsibility over fiscal deficiencies. This shift in local administration marked a lessening in the degree of control the central government had over local financial administration: Whereas for much of the fifteenth century central officials and specially selected prefects had played a role in shaping how localities met the demands for resources, in the sixteenth century, as a result of structural changes, magistrates were generally left to work out their own solutions.54  Nimick proposes that whereas the changes made to local administration in the fifteenth century retained the ability of the political centre to reach down to local administration and maintain a certain degree of control over local finance, this ability was lost in the sixteenth century. This dissertation contributes a fiscal dimension to this trajectory and its emphasis on fifteenth century administrative viability. After the fiscal model of the Hongwu era failed, the grand secretaries worked along with emperors in order to reform local administration and taxation. These reforms met central government demands while taking into consideration local interests and abilities. Moreover, these officials improved the system of financial reporting. The continuous recording of annual revenue reports throughout the century reveals that the central government maintained access to local financial information. Although it is clear that this information was becoming less accurate with time, it was still annually updated. In the sixteenth century this type of                                                       54 Ibid, 138.   32 financial reporting was no longer recorded in the official histories, suggesting that the data was no longer available. One reason, as Nimick attests, was the new independence of the local magistrate. Another is that fixed quotas took over tax collection, thus annulling the need for updated reports. Furthermore, the political centre became more preoccupied with its own finances and the state assumed a more predatory role in its demands for tax revenues. In the sixteenth century the renowned officials Lu Kun and Hai Rui both noted that detailed audits of local accounts had turned into occasions for collection of extra revenue for the provincial administration.55  In chapter 3 this shift is further connected to the growing reliance on silver in state finance. While commutation of the tribute grain (zao liang 漕糧) tax into silver began in the early fifteenth century, this chapter argues that the transformation in financial administration occurred towards the turn of the sixteenth century. The tribute grain tax was the rice transported to Beijing from south and southeast China. Its purpose was to feed the northern capital and pay its officials and soldiers. Only in the late fifteenth century do we find clear evidence of a fixed silver tribute tax quota and a shift to silver in all official payments. To that can be added the reform in the salt monopoly that resulted in the sale of salt for direct silver payments. From a fiscal perspective then, the “age of silver” began at the turn of the sixteenth century, between the Hongzhi and Zhengde reigns.                                                          55 Ibid, 141.   33 1.7 Accounting Fiscal administration is based on accounting. As Hartwell acknowledged, it is also a prerequisite for political power. The emperor always had the power to mandate taxes and order disbursements, but it was of little use unless he knew how much was collected and where it was stored.”56 In the Ming, accounting and accountability permeated the bureaucratic hierarchy and connected periphery and centre. It was also the main occupation of the Ministry of Revenue. As shall be further discussed in the following chapter, the chapter concerning this ministry in the Ming administrative code (zhu si zhi zhang 諸司職掌, or Handbook of Government Posts) stresses the production of timely and accurate reports on government revenue and the transportation of funds.    Between the early fifteenth and early sixteenth centuries, national accounting of all major categories of public revenue was conducted on an annual basis. Recorded in the Veritable Records of the Ming 明實錄, updated information appeared on population, cultivated land, grain tax from civilian and military fields, various cloth levies, salt and tea production and sales, commercial taxes, mining output, military land, and transports of grain to Beijing. Preparing and producing such a document every year necessitated a degree of communicative centralization that could enable information to flow from local administrative units to the political centre. Despite the decentralized structure of Ming financial administration, regulations regarding bookkeeping and auditing subjected officials to constant record making and reporting that were scrutinized by higher                                                       56 Hartwell, “Finance and Power,” 79.   34 authorities. Accountability was one parameter by which local officials were evaluated at the end of their term in office.   Interestingly, national accounting was not one of the strong parts of Song financial administration. According to Golas: All the preachment about “determining expenditures by calculating income” (liang ru wei chu) cannot disguise the fact that no one in the central government from the emperor on down had in most years more than a very vague idea of just how much the government took in or disbursed. Only eleven times in the 319 years of the dynasty was the government able to complete a k’uai-chi lu (accounting register) or general account of finances and population that sought to provide that information.57  This is a harsh assessment that neglects to mention that under the Song, the process of accounting and auditing was the most sophisticated in the history of imperial China. But it does allude to an inherent weakness in the Song system that emanated from the multiplicity of accounting agencies. Two separate audit agencies were responsible for regular and random audits of government accounts. There were record keeping offices for each of the three agencies in the State Finance Commission. A special office for managing debts owed to the government kept track of debtors and enforced payment, and an office calculating tax receipts. There were also special accounting offices that handled military finance.58                                                         57 Golas, “Sung Fiscal Administration,” 140.  58 For a detailed discussion of accounting in the Song see Philip Yuen-Ko Fu, “A Study of Governmental Accounting in China: With Special Reference to the Sung Dynasty (960-1279),” PhD diss. (University of Illinois at Urbana-champaign, 1968): 175-298.    35  The problem with this plethora of accounting offices was that it was not very efficient. Their tasks often overlapped and confusion over the delegation of duties could result in neglect on all sides. When reforming a government agency, officials complained about entire decades that passed without comprehensive auditing. Interestingly, when one office was not working properly the preferred solution was often to simply create a new one. On the one hand this attests to the flexibility and independence of Song officials, but frequent reorganizations also created instability and further confusion. It also probably did not contribute to personal accountability.   In this context, it is worth mentioning Qing reactions to what it perceived to be problems in the accounting system. Madeleine Zelin describes the practice of national accounting as an innovation and an improvement of the Ming fiscal system. This is certainly true in comparison with the mess financial administration was in in the late Ming, but for the most part, it echoed early Ming practices. The obligation of every province to send an itemized report of taxes to the Ministry of Revenue, based on the accounts of counties and prefectures, is specified in the administrative code of the Ming.59 Moreover, the early Qing requirement that every transaction of funds on the local level be reported and authorized by the Ministry of Revenue was a Ming routine procedure as well.60 Finally, Zelin finds a high degree of consolidation of financial information in the Qing annual accounts with the inclusion of data on military land and                                                       59 Zhai Shan 翟善, Zhu si zhi zhang 諸司職張 [Handbook of Government Posts] (1393, reprint Guo li zhong yang tu shu guan, 1981): 2:13a-13b. 60 Ibid, 2:14a.  36 revenue from mining, items that were formally under the jurisdiction of the Ministries of War and Works. As mentioned, this information appeared in the Ming annual revenue reports as well. At least in the fifteenth century, therefore, the system of national accounting did not fall from the more advanced Song or the reform-minded Qing.61   That said, the abundance of detailed fiscal memorials and tax information from the Qing enables researches to explore the process of accounting and fiscal management on the provincial level in a depth that can not be matched for the Ming. Utilizing these sources, Helen Dunstan reveals how Qing officials balanced provincial military budgets during 1746-48, when a universal tax remission was spread among the provinces. In particular she shows how the Ministry of Revenue separated between the provinces that had the heaviest responsibilities for military subsidies so that in any given year at least one of the major subsidizers collected tax that could be at its disposal in the following year.62 This kind of planning could only be achieved through control over provincial accounts and an ability to obtain updated tax data, and thus reveals the process of Qing accounting. Lack of similar documentation for the Ming prevents this degree of exploration and may suggest that its accounting practices were less advanced. But since the reporting infrastructure and accounting principles were similar, it is plausible that                                                       61 Madeleine Zelin, The Magistrate’s Tael: Rationalizing Fiscal Reform in eighteenth Century Ch’ing China (Berkeley: University of California Press, 1984): 12-15.  62 Helen Dunstan, “About As Interesting As the Telephone Directory: Encounters with Routine Fiscal Memorials of the High Qing Period” (paper from the Fourth International Conference on Sinology “Exploring the Archives and Rethinking Qing Studies, Taipei: Academia Sinica, 2013).  37 Qing practices are comparable to those in the fifteenth century, the difference being that Qing documents are better preserved.   1.8 Tax Commutation  If the establishment of a national accounting system helped the new Ming state gain control over the economy, the practice of commuting agricultural taxes was a practical solution employed in order to cope with the inherent difficulties of taxing an agricultural economy. In its original role, it was a fiscal policy that struck a balance between local abilities and state needs. For example, in years of scarcity when the collection of the grain tax was insufficient, the state commuted tax debt to other means of payment. By fixing an exchange rate that was favourable to tax payers, commutation entailed an actual reduction while avoiding a complete loss of that year’s tax revenue from a certain region. As a more permanent institution, tax commutation alleviated the high costs of transportation to the capital. In the 1430s, when the dynasty was facing debilitating revenue shortages, commutation of the Jiangnan tribute grain into silver became official policy. From this time on, silver became more prominent in official transactions and reflected its growing prominence in the market. But with the permanence of tax commutation the flexibility of the temporary tax commutation was lost.   Tax commutation was based on the availability of other means of payment that were universally accepted. In the first decades of the Ming, scrip was a common medium of tax commutation. After it devalued, cloth and silver became the main means of  38 exchange. As substitutes for the grain tax, both depended on and contributed to the development of a market economy. Although the Ming did not actively promote a market, as Brook suggests, the economic effects of Ming political economy were on the whole positive: “Much of the economic growth in the Yuan and Ming, though more so in the Ming, was private in organization and capital, but public in terms of its operation within an infrastructure created and funded by the state.”63 Brook alludes to a physical infrastructure of roads and waterways that were constructed and maintained for official use but served commercial transportation as well. In this dissertation, taxation is interpreted as another form of economic infrastructure since it influenced the economic decisions of farmers who had to pay taxes in marketable commodities   In expenditure, commutation had especially deleterious effects on official payments to state servants. Salaries for officials and officers, rations for soldiers, and stipends for princely households were all designated in grain. Once they were commuted to scrip, they were greatly reduced as a result of devaluation. Frequent adjustments to official salaries and military rations did not improve income significantly, though they probably managed to maintain a ration of one shi of rice a month for officials and officers. Soldiers’ rations often fell under one shi. Less known is that grain stipends for princes were partially commuted to scrip too. This situation continued until the late                                                       63 Timothy Brook, The Troubled Empire: China in the Yuan and Ming Dynasties (Cambridge, Mass: Harvard University Press, 2010): 109. A more detailed discussion appears in Confusions of Pleasure: Commerce and Culture in Ming China (Berkeley: University of California Press, 1998).  39 fifteenth century when all official payments were commuted to silver. But even then, commutation did not amount to the full original grain quota.    Tax commutation was responsible for the most significant fiscal reform of the fifteenth century, the collection of the Jiangnan tribute grain tax in silver. As a response to difficulties in paying military salaries for officers in Beijing, an edict was issued in 1436 in the name of the Zhengtong emperor approving the commutation of the tribute grain tax into silver. Though initially announced as a temporary policy, tax commutation became permanent and by the sixteenth century an annual quota of 1,000,000 taels was standardized in the tax registers.   In the highly monetized economy of the Song, commutation of the grain tax was not particularly significant. And suggestions to commute large amounts of grain raised concerns over disturbing the market. In 1070 State Finance Commissioner Wu Chong (吳充, 1021-1080) proposed commuting 2,000,000 shi from the grain tribute of Jiangnan and Huai’an circuits in order to fund the border armies. Repeated over five years, he calculated, the money could purchase enough supplies for three border circuits. Wang warned that this kind of commutation essentially means purchasing grain from the market. On such a scale it would bring down grain prices in those circuits and cause a shortage in coin. Instead he advised the commissioner to stick to the ever normal granary system (chang ping fa, 常平法) and limit grain purchases to districts that were experiencing high grain prices.64 The possibility of having individual taxpayers conduct                                                       64 Tuo Tuo (脫脫), Song shi (1345, reprint Beijing: Beijing zhong hua shu ju, 1977): juan 175.   40 the commutation on their own did not even come up in this discussion. Perhaps such an uncontrolled method was unthinkable in the northern Song. Wang Anshi was worried about disrupting local markets and decided against a large-scale commutation of the tribute grain tax.   Qing officials expressed similar reservations. As Jane Kate Leonard shows, when a suggestion was made to convert the Zhejiang grain tribute tax to silver in order to fund a huge reconstruction project along the Grand Canal, the Daoguang emperor dismissed the suggestion after initially endorsing it. The reasons for declining the suggestion revolved around different scenarios of disturbance to the market and the smooth operation of the canal. There were two options for funding the maintenance project: The first was buying grain from the Jiangnan markets in return for silver. This raised the concern that a purchase on such a large scale would affect grain prices. The other option was to have peasants perform the commutation themselves and pay their tax dues in silver. But top Qing officials did not trust local magistrates to refrain from manipulating exchange rates for their benefit. In addition, the Grand Canal enterprise relied on a complex relationship between government and private sources of funding and service. Shaking this system too profoundly might harm these relationships and create resentment.65                                                        65 Jane Kate Leonard, “The State’s Resources and the People’s Livelihood (Guoji minsheng): The Daoguang Emperor’s Dilemmas about Grand Canal Restoration, 1825,” To Achieve Security and Wealth: The Qing Imperial State and the Economy, 1644-1911 Jane Kate Leonard and John R. Watt eds. (Ithaca, NY: Cornell East Asia Series, 1992): 47-74.  41   These cases raise interesting questions regarding the Ming. Did commutation of grain into silver disturb Jiangnan grain markets? Did it cause a shortage in silver? These questions are not pursued here, but as far as official commentaries can be trusted, there was no serious disturbance to Jiangnan markets as a result of this practice. If any, this procedure was concurrent with a boost to Jiangnan’s economy. In comparison with the seemingly smooth implementation of this policy in the Ming, the concerns of Song and Qing administrators seem unfounded. On the other hand, both the Song and the Qing were experiencing currency shortages that may have deterred administrators from authorizing government purchases of that scale.      1.9 Exposition of the Chapters  Chapter 2 establishes the relationship between the legal foundation of the dynasty and the character of its financial administration. As part of the ambitious endeavor to reshape Chinese society after centuries of foreign rule, the Hongwu emperor presided over a project of legislation and codification of government regulations. In his work on early Ming legislation, Farmer defines the role of these compilations in Ming society as that of a constitution.66 These texts were constitutional in the sense that they laid out what were intended to be permanent arrangements of power. Subsequent emperors could not change them but merely introduce additions in response to changing circumstances.                                                        66 Edward L. Farmer, Zhu Yuanzhang and Early Ming Legislation: The Reordering of Chinese Society Following the Era of Mongol Rule (Leiden: E.J. Brill, 1995): 10.  42  The Zhu si zhi zhang was written as a technical manual for the central government. It was modeled after the Tang liu dian but its content is original to the Ming. It is organized by chapters dedicated to each office and agency in the central government: the six ministries, the Censorate (du cha yuan 都察元), the Office of Transmission (tong zheng shi si 通政使私), the Grand Court of Judicial Review (da li si 大理司), and the Five Chief Military Commissions (wu jun du du fu 五軍都督府). Unlike the legal codes, the Minister of Personnel wrote this government manual, though the emperor probably closely monitored the work. This text served as the foundation for financial organization for the rest of the dynasty. An analysis of the chapter on the Ministry of Revenue in the Zhu si zhi zhang reveals that the main goal of Hongwu-era financial administration was to create centralized control over the accurate and timely flow of information and supplies, while avoiding centralizing financial power in the hands of any man or agency besides the emperor.  This compilation forms the basis for the two extent editions of the Da Ming hui dian 大明會典 (Collected statutes of the Great Ming).67 The first edition was completed in 1503 and after some revision printed in 1511 and presented to the Zhengde emperor. It is therefore referred to as the Zhengde edition. The role of the Hui dian was to add later imperial decrees as legal historical precedents (shi li事例) to the original constitution. It                                                       67 Li Dongyang東陽李 ed., Da Ming Hui dian 大明會典 [Collected Statutes of the Ming Dynasty] (1511, reprint Ying yin wen yuan Si ku quan shu vol. 375-76, Shanghai: Shanghai gu ji chu ban she, 1987). For a reprint of the Ming edition see Da Ming Hui dian, (Tokyo: Kyûko shoin, 1989). An earlier edition that was completed in 1479 and a later one, completed in 1550, are not extent. The edition of 1587 is the most used by scholars: Li Dongyang 東陽李 and Shen Shixing 申時行 eds. Da Ming Hui dian大明會典 [Collected Statutes of the Ming Dynasty] (1587, reprint Hua wen shu ju, 1964).  43 also collected relevant articles from the Ming code and other Hongwu era compilations and entered them at the beginning of each section. As shall be discussed in the following chapter, these texts were not only neutral administrative records but represented an affirmation of government agendas. Changes in the structure of the Hui dian attest to a decline in the status of the Zhu si zhi zhang in the sixteenth century.   Routine financial work was reported in the Ming shi lu 明實錄 (Veritable Records of the Ming). Upon the death of an emperor a special committee made of top court officials assembled the various materials accumulated during the former reign period. Though arranged in chronological order, it is much more than the annals of an emperor. The Shi lu served as a tool for administrators and reflected the agenda of its compilers.68 If we assume that the information these committees chose to include was based on their importance, the consecutive recording of annual revenue reports throughout the fifteenth century attests to the place financial administration held in government work.    These reports also affirm that Hongwu’s emphasis on accounting and auditing did not disappear after his death. The annual updating of information on tax revenue suggests that an infrastructure of gathering and transferring information existed throughout the fifteenth century. Interestingly, while the reports align with the intentions of the founding emperor, they begin only in the reign of his son. This can either mean that the compilers Hongwu Shi lu did not have access to such information, or that the officials compiling the history of the Yongle emperor attached greater value to the preservation of financial                                                       68 Xie Guian 謝貴安, Ming Shi lu yan jiu 明實錄研究 [A Study of the Ming Shi lu] (Taipei: Wen lu chu ban she yin hang, 1995).   44 information. In the introduction to the Yongle Shi lu, the editing committee outlined its editorial principles, in which they explicitly stated that data on various sources of revenue should be recorded at the end of every calendar year.69 Thus the recording of such information testifies to its being an integral part of the management of the realm.   In all probability, the revenue reports were a summary of the annual tax income reports. This does not necessarily mean that the income reported was of actual receipts. Rather, it was probably a summary of total owed taxes. As such, the data cannot be used to assess actual income but it can follow general trends in tax collection and provide a sense of how Ming administrators managed the national budget. Nevertheless, the fact that during the first half of the fifteenth century most of the data was updated annually reveals that the Ming did not yet use a fixed quota system. The revenue reports contain only a standard list of items and their quantities. Beyond that they do not reveal much about the process of accounting or the reasons for including certain items. Their simple and direct title does follow the basic principle of financial administration in the early Ming period: This Year All Under Heaven (shi sui tian xia是歲天下), meaning the income of the entire realm and not just the emperor’s.   In the last decades of the fifteenth century data gradually turned into fixed quotas to the degree that there was no longer any connection between changing economic circumstances and financial data. In the Jiajing reign (1522-1567) the reports ceased to appear. After an unsuccessful attempt to reinstitute them during the Longqing reign                                                       69 “Xiu zuan fan li” 修纂凡例(Editorial Principles), Ming Taizong Shi lu 明太宗實錄 (Veritable Records of Ming Taizong).  45 (1567-1572), they disappeared from the imperial record altogether. Taken as a measure for the ability to centrally collect and process data, the demise of the revenue reports indicates change in the nature of financial administration.   Chapter 3 proposes a fiscal reason for this change. It discusses tax commutation as both a temporary and permanent feature of tax collection. The commutation of the tribute grain tax into silver, known as Gold Floral Silver (jin hua yin金花銀), introduced the most significant transformation in the Ming economy. The shift to permanent silver payments fixed grain quotas and made annual updates redundant. This chapter explores available sources in order to trace the origin and development of this institution. It also follows the emergence of the term Gold Floral Silver and its possible origins, and connects it to a deeper shift in the nature of Ming state finance.   This chapter also uses information from the revenue reports in order to examine other sources of non-grain tax income. As mentioned earlier, although paper currency had little value in the fifteenth century, it did not disappear from circulation so quickly. It continued to be collected in various commercial taxes and used in official salaries. Cotton was another common option for tax commutation and served as an alternative to silver. Though there is evidence that paying taxes in cotton spread during the fifteenth century, the revenue reports do not indicate a rise in tax revenue. Silk was produced in the imperial workshops for imperial use, but was also registered in the revenue reports. Interestingly, a doubling in annual quotas of silk cloth is noted around the same time Gold Floral Silver was initiated. This suggests that silk was a part of the commutation policy as well.   46  The next three chapters address the main avenues of government expenditure. The Ming dynasty controlled and governed the realm through a civilian and military bureaucracy, a standing army, and a large number of princely courts that were related to the ruling house through kinship ties. These three groups, though differing greatly in status and function, had one thing in common: They were entitled to fixed, periodical payments. In budgeting its expenses, the imperial court directed its regular expenditure at these three groups, and by that gave material shape and substance to what it imagined as the state. In their different capacities, officials, soldiers and princes ran the machinery of the state and articulated its presence in society. Although bound to their vocations either through social convention or hereditary obligation, their identity as servants of the state manifested itself through the symbolic and tangible bind of standardized and regular payments.  The Ming never came close to the official salary charts in paying these three groups. Instead, the entire period before the commutation of payments into silver is characterized by frequent adjustments and manipulations to grain allocations. These three chapters explore how the state managed the commutation of salaries and how it tried to compensate for the meager remuneration of its servants. The management of expenditure is explored through memorials, imperial decrees, and biographies that appear in the Shi lu, Hui dian, and the official history of the ming, Ming shi 明史.70                                                       70 Zhang Tingyu 廷玉張, Ming shi 明史 [History of the Ming] (1739) (repr. Beijing: Zhong hu shu ju,1974).  47  Chapter 4 discusses the salaries of military and civilian officials in the central and provincial governments. Official salaries were extremely low due to commutation into scrip. From the very beginning of the dynasty, official salaries were calculated in grain but were paid mostly in other means. These substitutes were never sufficient because their exchange rate was lower than market rates. Nevertheless, a monthly salary was the reality of most serving officials who did not rise in rank high enough to be able to translate status into material gain. This is the reality the imperial court had to address in order to maintain some kind of semblance of a client-patron relationship between the bureaucracy and the imperial court. As part of this effort, this chapter argues that the court attempted to maintain a minimum level of rice allocations and examines a mechanism that helped officials support their families while away from home.    Chapter 5 considers military finance. Commutation of military salaries was determined by their size. While top military officers were subject to the same commutation as civilian officials, regular soldiers were entitled to rations (yue liang月糧) that were never high enough to be partially commuted. During the fifteenth century these rations, small to begin with, further diminished and contributed to widespread desertion. Unlike official salaries, military rations were supposed to be largely derived from agricultural fields. Since the produce of most military fields was never sufficient, special transports of grain and other supplies were needed to supplement reserves. This problem was especially grave on the northern border, and serves as the background to the large-scale commutation of the grain tax into silver.   48  This chapter also examines the role of the salt-for-grain policy (kai zhong fa 開中法) in supporting border armies. It argues that in the Hongwu reign it had a primary fiscal role in supporting the reconstruction of north China, and follows its deterioration as a result of the collapse of the military colonies.  The kai zhong is also a good reflection of the Ming relationship with the commercial sphere. As discussed in an earlier section, the goal of this system was to reduce expenses and provide a stable flow of grain and other commodities to border areas. It was not intended to increase revenue.   Finally, this chapter offers a fiscal interpretation to the demise of the Ming military. The system of military finance in China was tied to the registration of hereditary military households and military land. But unlike societies with military elites that assume the character of a feudal, military caste, military households were essentially a fiscal unit. Just as civilian households were required to pay tax and render labour services, soldiers were attached to military guards and were required to work on the military fields of their guard, pay taxes on them, and render military service to the state. In this light, the system of military colonies, in which soldiers were supposed to rotate among training, campaigning, and agricultural work was essentially another form of tenancy. Tenant soldiers were obliged to pay rents that were even higher than those of civilian government tenants. And just as government tenants fled from their fields when the tax burden was high, so did soldier tenants. The strategy to restore military colonies was similar to that of the tax reforms of the mid-fifteenth century, but the results differed; lowering tax rates encouraged the privatization of military land and its disappearance from military registers.   49  Chapter 6 discusses the imperial clan. In theory, princes constituted a class of their own, endowed with fiefs on which tenants worked the land and paid rent. In reality they lived off a fixed allocation of funds, granted to every princely household according to rank. Princely courts were allocated an official staff to preform administrative, educational and household tasks. They were also endowed with military guards as part of their task as peacekeepers of the region in which they resided. But these were not their private armies and the funds to maintain them, even if collected locally, were centrally administered. There is no fixed term that is used for these allocations but supplies (gong ying 供應) appears in the princely code.71 Unlike soldiers who were given food rations, and officials who were granted basic salaries, princely households were entitled to large amounts of grain and other items in order to enable them to conduct a lifestyle fitting for nobility as well as to remunerate their large staffs.  These funds were to be collected annually in grain. But, as with official and military salaries, about half of these allocations were soon commuted to scrip. As a result, the annual allocations for the lower ranks of the nobility were quite meager. As the historical record shows, it seems that only the two highest ranks of nobility (out of eight) actually owned land. Princes often attempted to expand their domains beyond the territories that were allocated to them, but they mostly did so within regular official                                                       71 Zhu Yuanzhang 朱元璋, Huang Ming zu xun 皇明祖訓 [Ancestral Instructions of the Great Ming] (1395) (reprint Jinan: Qi Lu shu she, 1996). The Hongwu emperor compiled the Ancestral Instructions for his decedents. It contains general instructions regarding the administration of princely courts, their ritual roles, appropriate conduct, legal status, and granting of noble titles.   50 channels. The biographies of Ming princes in the Ming shi reveal the negotiating process that took place between princes, local officials, and the court to manage these tasks. Finally, this chapter finds that the phenomenon of princes as large landowners was more widespread in the sixteenth century and closely related to imperial favour. It is argued that princely households burdened state finance not so much in the expansion of their domains but in the demographic growth of members of the lower nobility that were entirely dependent on local administration for their livelihood. The discussion of state expenditure is limited to these three groups because they were the only avenues of expenditure that entailed fixed sums and regulations. It is therefore in these avenues that state administrators could manage and manipulate some kind of a budget. In contrast, the expenses of the imperial household were never budgeted and therefore could never be managed by officials. In addition, large-scale public works such as waterway maintenance, construction works, or maintenance of land transportation routes were not regularly budgeted. The funding for them was usually under the supervision of the Ministry of Works that had its own power to requisition labour and supplies from the population. It also made special requests for additional funds from the Ministry of Revenue. At this point it should be acknowledged that this dissertation is not an economic, monetary, or financial history of the Ming. Unlike these disciplines of economic history, fiscal history addresses aspects that fall into these fields only as far as they concern the ways in which the state managed revenue and expenditure. In this sense, fiscal history is important to our understanding of the state more than the economy. There is a vast body of work on Ming economic history that deals with fiscal issues such as population and  51 land registration, taxation, transportation of funds, military finance, local finance and its relationship to the political center, and others.72 This dissertation builds on this foundational work while restricting its scope to fiscal policy at the political center. Moreover, it concentrates on subjects that were not at the forefront of research before; commutation of grain in revenue and expenditure, and national accounting.  The fact that the grain tax was often commuted to other means of payment is well known but has not been systematically explored in the context of a comprehensive budget                                                       72 In the economic history of the Ming some of the standards include the work of Liang Fangzhong on various aspects of the land tax. His works are conveniently assembled in several collections, among them is Liang Fangzhong jing ji shi lun wen ji 梁方仲經濟史論文集[A Collection of writings on economic history by Liang Fangzhong] (Beijing: Zhonghua shu ju, 1989). Among other interests, Wu Jihua contributed valuable insights on the tax structure of the important rice-producing prefectures in Jiangnan that will be mentioned in chapter 3. Fujii Hiroshi has produced a detailed examination of fiscal land registration that will be mentioned in the following chapter. Hoshi Ayao (星斌夫) detailed the complex operation of the grain tribute system in Mindai sōun no kenkyū 明代漕運の研究[A study of the Ming tribute grain system] (Tokyo: Nihon gakujutsu shinkōkai, 1963). Mark Elvin translated this work into English under the title The Ming Tribute Grain System (Ann Arbor: University of Michigan, Center for Chinese Studies, 1969).  In military finance Wang Yuquan produced the definitive work on the Ming military colonies. Yu Zhijia elaborated on the registration system of soldiers in the Ming military guards. And Terada Takanobu explored military finance and the impact of merchants on military finance along the northern border. These works are mentioned in chapter 5. Wang Yuquan and Zhang Xiaoyu’s work on princely estates, and Zhang Dexin’s research the population of the princely clan inform chapter 6. Peng Xinwei’s comprehensive monetary history of China is the standard reference work in the field. Works on Ming account books include Liang Fangzhong’s exposition to the Wanli kuai ji lu, discussed in chapter 2. Hiroshi Iwami’s discussion of the local account books of the late Ming is translated to English: Iwami Hiroshi, “An Introduction to the Shandong jing hui lu”, State and Society in China, ed. Linda Grove and Christian Daniels (Tokyo: University of Tokyo Press, 1984): 311-334. Guo Daoyang wrote a general history of accounting in China that is mentioned in the following chapter. In fiscal history Bian Junjie 边俊杰 reemphasized the impact of political corruption on fiscal policy in Ming dai de cai zheng zhi du bian qian明代的财政制度变迁[A Study of the Transition in the Financial System of the Ming] (Beijing: jing ji guan li chu ban she, 2011). Accounts of fiscal policy on the local level include Cheng Liying 程利英, Ming dai bei zhi li cai zheng yan jiu 明代北直隶财政研究 [a study of the financial administration of the northern metropolitan region in the Ming], (Beijing: Zhongguo she hui ke xue chu ban she, 2009), and He Zhaohui’s article, discussed in this chapter.  52 that includes revenue and expenditure. Terada Takanobu discussed the commutation of grain transports to the northern border, and the annual emissions of silver to military garrisons in his work on the expansion of trade in north China.73 Similarly, Nishijima Sadao pointed at the commutation of the grain tax into cotton cloth as a factor in the expansion of the cotton industry in south China.74 The two works mentioned above look at commutation from the point of view of market expansion, and are limited to certain regions of China and particular sectors of the economy.  Finally, commutation of the grain tax has been most widely discussed in the context of the Single Whip reform.75 The combination of the grain and labor tax into one silver payment signaled a transition to a monetized tax system, or as Liang termed it, a “modern tax system.”76 In contrast, this dissertation looks at the practice of tax commutation before the transition to silver and situates it in a broader context of Ming governance. It shows on the one hand that commutation was a regular part of tax collection before the advent of silver. And on the other hand, it equates the move to fixed                                                       73 Terada Takanobu 寺田隡信 Sansei Shōnin no kenkyū: Min-dai ni okeru shōnin oyobi shōgyō shihon 山西商人の研究: 明代における商人および商業資本 [A study of the merchants of Shanxi: Merchants and commercial capital in the Ming]. (Kyoto: Tōyōshi-kenkyū-kai, University of Kyoto, 1972). 74 Nishijima Sadao, “Early Chinese Cotton Industry,” Christian Daniels and Linda Grove eds., State and Society in China: Japanese Perspectives on Ming-Qing Social and Economic History, (Tokyo: University of Tokyo Press, 1984). 75 For an English translation on Liang Fangzhong’s seminal work on the subject see: The Single-Whip Method of Taxation in China, translated by Wang Yuquan (Cambridge, Mass.: Harverd University Press, 1956).  76 Ibid, 1.   53 silver quotas in the tribute grain tax with a deeper change in fiscal administration that actually undermined the earlier logic of tax commutation.  Economic historians have used the annual revenue reports as a reliable source to gather statistics on population, land and tax revenue but have not cited them as evidence of national accounting.77 In fact, the more meaningful work on Ming accounting regards the national and local account books that developed in the late Ming period, well after the disappearance of the annual revenue reports.78 In the following chapter, I suggest that the annual revenue reports are in themselves a testament to a functioning system of national accounting that conforms to the principle of managing resources locally, and that their disappearance in the early sixteenth century is a result of the changing nature of financial administration.  Finally, the goal of this dissertation is to imagine a budget within which Ming administrators operated. It therefore gives equal importance to tax collection and public expenditure. By grouping together three very different social groups that were all in some way on the state’s payroll, this dissertation reveals how similar solutions were employed in order to overcome budgetary limitations and suggests that from a fiscal point of view, princes, officials, and soldiers were all state servants that where supposed to abide by the                                                       77 For example Otto Berkelbach van der Sprenkel, “Population Statistics of Ming China”, Bulletin of the School of Oriental and African Studies, No. 2 (1953), 289-326. And Liang Fangzhong, Zhongguo li dai hu kou tian di tian fu tong ji 中国历代户口田地田赋统计 [A comprehensive account of population, land, and land taxes in Chinese history] (Shanghai: Renmin chu ban she, 1980): 185-199.  78 For example Liang Fangzhong, “Shu ji ping lun: Ming Wanli kuaiji lu” 书籍评论:明万历会计录 (Book Commentary: The Account Book of Ming Wanli), Zhongguo jindai jingji shi yanjiu 3.2. (1935): 292-298. And Iwami Hiroshi’s discussion on the Shandong jing hui lu.  54 principle of measuring revenue in order to determine expenditure. Furthermore, the focus on official salaries and stipends presents a contribution to scholarship that has focused largely on tax revenue and paid less attention to expenditure. And the works that do deal with salaries do not do so in the context of fiscal policy.79   Monetary history deals with the production and circulation of currencies with or without relation to the state. In the economic history of Europe that has set the standards of the discipline, money came to be equated with metallic currencies. This dissertation expands the definition of money to anything the state used in commuting the grain tax. From the first decade of the Ming, grain could be substituted for cotton and silk cloth, bronze coin (also referred to as cash), scrip, or silver. In addition, the commodities sappan wood and pepper were used in salary payments. All of the above could be defined as money because they were of a standard size and weight, made out of durable material, and were accepted as means of exchange and a store of wealth.   Within this broad definition there are distinctions that should be acknowledged. First, cloth could serve as a currency but at the same time was used for its material. Silver too could be considered commodity money since silver had a variety of uses such as religious statues, household utensils and personal ornaments. To a lesser extent, coin too had a material value beyond its function as currency. Copper was used for the production of agricultural utensils and weapons. Finally, sappan wood and pepper were commodities                                                       79 The regulations for the payment of official salaries throughout imperial China are explored in Huang Huixian 黄惠贤 and Chen Feng 阵锋, Zhongguo feng lu zhi du shi, 中国俸禄制度史 [A History of the salary system in China] (Wuhan: Wuhan da xue chu ban she, 1996). Works on princely income include:  55 that were used and consumed, but their durability and standard measures made them suitable to be used as means of payment. Scrip was the only currency that had no use or value beyond the implicit promise of the state to honor it. In commutation of taxes and salaries, the exchange rates of these means of payment were determined according to one shi of grain (the implicit reference is usually to rice), which renders it a unit of account. But in determining prices the silver tael was often used as the standard unit of account.  Looking at these currencies from a fiscal point of view offers a different interpretation of their role in state finance. In contrast to the evidence of their circulation in the private market, silver seems to have taken over only at the turn of the sixteenth century, and scrip remained in government circulation until the last decade of the fifteenth century, long after their perceived demise as a viable currency in market transaction.  Finally, as financial history deals more broadly with the accumulation and investment of wealth in a society, fiscal history is limited to the decisions made on the level of the state. While scholarship has greatly expanded our knowledge of fiscal policy on the local level this project addresses the political centre.80 As such, it is limited to discussions and decisions made on the highest levels of the government, and interprets these decisions according to contemporary principles of political economy and as a respons to changing economic conditions. The sources that inform this study were produced by the inner group of central-government officials, and reflect their approach to financial administration. In addition, this dissertation recognized that official                                                       80 By the local level I mean the structure of local taxation and its transformation as studied in the posts of tax captains (li zhang 里長), the li jia 里甲 system,  and the Single Whip reform.   56 compilations necessarily contain the agenda of its compilers. This is not viewed as an obstacle to understanding the realities of the time but as an opportunity to reflect on the ways in which officials used these compilations to initiate reform or encourage policy. Although using a large body of quantitative data, there is no attempt to quantify state income or expenditure. Rather, this dissertation concentrates on the meaning of their production in the context of Ming state finance, and follows broader changes in tax figures in order to trace policy change.      57 2 The Structure and Role of the Ministry of Revenue   The Ministry of Revenue (hu bu 戶部) was one of the six ministries of the central government. It presided over the collection of tax revenue, disbursement of payments, transportation of supplies, maintenance of reserves, and handling the book keeping of these vast operations. As such, it acted as the headquarters of an empire-wide bureaucracy that reported financial information and forwarded it up the administrative hierarchy. The staff of the ministry was confined to the capitals and was surprisingly small. Under the minister served two vice ministers.81 Between five and ten officials worked in each of the thirteen offices processing documents from the thirteen provinces. A records office employed a proofreader and several clerks. An office that supervised the management of scrip had three workers. And two workshops, one for producing the plates fused for printing paper notes and a printing office manned by three workers each. Then there were storehouse and granary officials that usually worked in pairs, though in some storehouses there were up to six workers. This was the entire salaried personnel of the ministry.82   This structure is original for the Ming and it reflects the ideas of the first emperor about government. The following chapter discusses the legal and historical context of the foundation of the Ministry of Revenue in order to explain its particular character. A                                                       81 (Zhengde) Ming Hui dian, 3:4a.  The Yongle emperor nominated a censor to supervise the imperial granaries in Beijing and Tongzhou.  82 Zhai, Zhu si zhi zhang, 1:6a-7b.   58 description of the roles of the ministry reveals that it was mainly an information-processing office. It received information from the provinces and forwarded requests for famine relief, additional funds, or tax remissions to the emperor. The orders regarding any of these requests always came from the emperor. In addition, as the institution of the Grand Secretariat acquired political power in the fifteenth century, its officials rather than the ministers of revenue stood behind national tax reform initiatives. Since the Ministry of Revenue operated as an accounting office, the annual revenue reports stand out as its main achievement throughout the fifteenth century. The characteristics of these reports are discussed here and the decline in the quality of information is tied to changes in financial administration. Finally, this chapter ends with an exposition on the crisis in financial management and national accounting in the sixteenth century.    2.1 Legal Foundation of Financial Administration  In February of 1381 the Hongwu emperor voiced his opinion on fiscal policy in the following conversation: A minister close to the throne suggested: “The state should manage its finances in order to relieve state expenses [guo yong].” He spoke with much knowledge. The Emperor replied: “Heaven and Earth create wealth in order to nourish the people, therefore whoever holds the position of ruler has the obligation to use it in order to nourish the people. If even while minimizing waste and taxing lightly, one still worries about hurting the people, how much more, when tax is heavy, could there not be resentment?” The minister continued: “From the Son of Heaven to the common people there is no one who could create a state without storing.” The emperor replied: “The ruler manages wealth differently from the common man. The common man plans for a single household, and therefore his wealth is stored in one  59 household. The ruler is responsible for All Under Heaven and so he stores the wealth All Under Heaven. How can he block people’s [path to] nourishment and surreptitiously seize their wealth? In the past, Emperor Wu of the Han employed the likes of Dongguo Xianyang and Kong Jin to tax them heavily.83 They exploited the people and seized their wealth, and there was suffering all over the land. Shenzong of the Song employed Wang Anshi for managing wealth.84 It lead to the rise of petty people and All Under Heaven was in turmoil. This should be guarded against.” The speaker was ashamed and terrified. From this time on there was no more talk about wealth and profit [財利 ,  cai li].85  The Hongwu emperor equated fiscal expertise with heavy taxation, and heavy taxation meant bad government. While the anonymous minister entreated for a sounder financial policy that would increase the available resources at the centre, the emperor replied that resources should for the most part be kept in their localities. In this sense Hongwu was a “commoner emperor” who even on his throne held the point of view of the commoner. He was suspicious of any ideas to increase state revenue, and believed that the central government should have as little possession over the wealth of the dynasty as possible. In this conversation it becomes clear that maintaining a decentralized and weak financial administration was a way to prevent the accumulation of wealth in the                                                       83 Dongguo Xiangyang (東郭咸陽 ) and Kongjin (孔僅 )were heads of the salt and iron state monopolies during the reign of Han Wudi (141-87 BCE). 84 Emperor Shenzong (神宗, r. 1068-1086) appointed Wang Anshi (1021-1086) as vice Chief-Councilor, a post that enabled him to carry widespread reforms in the economy, such as the Green Sprouts Act that offered peasants government loans at lower interest rates than private lenders. See: Dieter Kuhn, The Age of Confucian Rule: The Song Transformation of China, (Cambridge, Mass.: Harvard University Press, 2009): pp. 54-55. 85 Ming Taizu Shi lu, 135:2b-3a.  60 political centre. It would take more than a century for court officials to suggest reforming financial administration away from this model.  In an otherwise highly centralized administration that could penetrate down to the level of the individual household, the emperor’s aim was to prevent privileging interests of the political centre over those of the localities, and to prevent those with political power from gaining economic power as well. He tried to do so by maintaining centralized control over a decentralized system of financial administration, a system in which no one person, besides the emperor, had total knowledge or control over the nation’s finances. And no one, besides the emperor, had the final say on how to allocate resources.   Decentralization, though, did not mean abandoning uniform standards. When the regulations for financial administration were finalized they were a manifest of standardization, punctuality and exactness. These regulations are the subject of the chapter on the Ministry of Revenue in the administrative code of the Ming, Zhu si zhi zhang. Legally, the status of this compilation was not as firm because Zhu Yuanzhang did not write it. Langlois points out that Ming law was connected to Zhu Yuanzhang’s person. In Ming political culture, the words uttered by the founding emperor could not be corrected or changed because they provided “a solid rock upon which to rest the myth of legitimate, fair, and just rule.”86   There is evidence that early Ming legislation retained its authority after the death of the founding emperor, but its literal application is less clear. Tam Kai-chai examined                                                       86 Langlois, “Ming Law,” 176.   61 legal texts from the Jiajing period proving that the Da gao (大誥, Grand Pronouncements) and the Da Ming lü ( 大明律, Great Ming Code) were used in order to define crimes and fix punishments. He also finds that throughout the Ming possession of a copy of the Da gao served to reduce punishment.87 Timothy Brook, on the other hand, finds that although the Da gao continued to be revered after the Hongwu reign, it was not used in court or as a means of popular education.88 This kind of reverence with no actual legal substance may also be the case in regards to extra punishments and warnings Zhu Yuanzhang declared in publicly displayed placards; Hok-Lam Chan has found that the placards issued by the Hongwu emperor, containing extra-legal punishments to violators in an array of categories were still hung in government offices during the Jiajing reign.89 These were extremely harsh punishments for relatively minor offences such as execution for playing football in military camps. Although punishments were not necessarily carried out, officials continued to hang these placards in public places.   Although the words of the Zhu si zhi zhang were not the emperor’s, in its technical and formulaic writing it is part of a long tradition of administrative texts beginning with the Zhou li (周禮, Rites of Zhou), the government blueprint attributed to                                                       87 Tam Kai-chai 家齊譚, “Ming Taizu Yu zhi da gao zai Hongwu chao yi hou xing yong qing kuang xin tan” 明太祖御製大告在洪武朝以後行用情況新探 [A Reexamination of the Legal Authority of Zhu Yuanzhang’s Grand Pronouncements After the Hongwu Reign], Journal of Chinese Studies 47 (2007): 73-91.  88 Timothy Brook, Jerome Bourgon, and Gregory Blue, Death by a Thousand Cuts, 118-119 (Cambridge, MA.: Harvard University Press, 2008). 89 Hok-lam Chan, “Ming Taizu’s “Placards” on Harsh Regulations and Punishments Revealed in Gu Qiyuan’s Kezuo zhuiyu”, Asia Major, 22.1 (2009): 13-39.  62 the Duke of Zhou, that most probably reflects Qin political culture.90 Following Schaberg who argues for the constitutional status of the ancient text, I would like to suggest that Zhu si zhi zhang played a similar role in the Ming. Schaberg uses ‘constitution’ in the ancient Greek sense of politea, or system of government, and views the Zhou li as a rational technical vision of the total management of the resources needed to sustain the realm.91    The compilation of the Hui dian in the late fifteenth century serves as an affirmation that Zhu Yuanzhang’s legal project did not disappear after his death. And the best indication that his administrative code continued to be at the centre of government administration is its place in the expanded administrative guide Da Ming Hui dian. This compilation was organized around the Zhu si zhi zhang and other Hongwu-era legal texts, and structured according to the central government offices and court agencies.92 In addition, the editors included subsequent decrees that served as legal precedents. Arranged according to a clear hierarchy, each sub-section begins with a quote from Zhu si zhi zhang. Then follow quotes from other Hongwu-era texts and finally later imperial decrees chronologically arranged. By doing this, the compilers reiterated the constitutional status of the original administrative code but also acknowledged historical                                                       90 Jin Chunfeng 春峰金, Zhou guan zhi cheng shu ji qi fan ying de wen hua yu shi dai xin kao 周官之成書及其反映的文化與時代新考 [A New Consideration on the Historical and Cultural context of the Production of the Zhou guan] (Taibei: Dong da tu shu gong si, 1993). 91 David Schaberg, “The Zhouli as Constitutional Text”, Statecraft and Classical Learning: The Rituals of Zhou in East Asian History, ed. Benjamin A. Elman and Martin Kern (Leiden: Brill, 2010): 33-63. 92 These include compilations such as Grand Pronouncements (Da gao) of which four collections were printed between 1385-1387, Ancestral Instructions (Huang Ming zuxun) from 1396, and Da Ming lü (The Great Ming Code) printed in 1397.  63 change. This arrangement gives the sense that the Hongwu-era law was viewed as constitution whereas imperial decrees served as amendments to the law in response to changing circumstances. The Hui dian was a final authority in cases of uncertainty about regulations. For example: In 1520, a traveling censor reported that every year a certain county in the southern metropolitan region sends cloth to the Ministry of Rites in Nanjing, but according to the Hui dian, it should be sent to Beijing. The emperor ordered to follow the old rule.93  Even when procedures had deviated from the original plan, in policy discussions emperors continued to refer to it either as a model to follow or as a model that unfortunately cannot longer be followed. In such a discussion, Hongzhi demonstrated knowledge of the text when he remarked: “Even though the Zhu si zhi zhang contains a regulation for coin-minting stations in each province, this has long been discontinued.”94 A second edition was prepared during the Jiajing reign but never published. The only part remaining is the emperor’s forward. A third edition was published during the Wanli reign (1573-1620).95 In the Wanli edition the hierarchy of what we could call                                                       93 Yuan Ruiqin (原瑞琴), “Da Ming huidian xingzhi kaolun” 大明会典性质考论(The Nature of Da Ming huidian),  shixueshi yanjiu 3 (2009). For a more thorough discussion comparing the two extend editions see also the author’s Da Ming huidian yanjiu 大明会典研究 (A Study of the Da Ming huidian, Beijing: Zhongguo shehui kexue chubanshe, 2009). 94 Ming Xiaozong Shi lu, 29:12.b. 95 The Wanli edition is conspicuously absent from the siku quanshu, but is far more accessible in modern reprints. Shen Shixing (申時行), Da Ming huidian (1587, repr, series: guo xue jiben congshu, Taibei: Huawen shuju, 1964). The annotators of the title in the Imperial Library Si ku quan shu knew of the Wanli edition but apparently had no accsess to a copy. They also claimed that it was a flawed work that reflected the flawed government of the time, and therefore was inferior to the Zhengde edition. See ji yun 記昀, Si ku quan shu zong mu ti yao四庫全書總目提要 [An annotated catalog of the Imperial Library] vol. 2 (1782, reprint 1933): 52-53.   64 ‘constitutional’ texts and imperial decrees disappears. In the guiding principles for the revision, the authors acknowledge this change: The Hui dian was previously arranged with Zhu si zhi zhang first and the historical precedents later. But the regulations are from the twenty-sixth year of Hongwu and so should be considered as Hongwu era regulations.96  Accordingly, the other Hongwu-era compilations receive the same treatment: Excerpts from the Da Ming lu are written under “first year of Hongwu”, those from the Da gao as “twelfth year of Hongwu” and so on. Though not erasing them, by making this editorial change, the authors rejected the special status of Hongwu administrative regulations.97 Although the editor-in chief of this edition, Zhang Juzheng (張居正, 1525-1582), died before its completion, as grand secretary he nominated an editing committee of like-minded officials. It is therefore very likely that such a fundamental change in the nature of the Hui dian would not take place without his permission. From what we know of his statecraft philosophy, this decision reflects his conflicting attitude towards government reform. On the one hand, Zhang argued that statecraft should be brought back to the simplicity and decisiveness of the Hongwu era and its institutions. On the other hand, he believed that government should be relevant to the time and that no regulation is eternal.                                                        96 “Guiding Principles of the Revision” (chong xiu fan li), (Wanli) Da Ming Hui dian, p. 22, 1a-b. 97 Alternatively, Yuan Ruiqin interprets the dating of Hongwu era compilations as a sign of greater systemization. See: Yuan Ruiqin, “Zhang Juzheng yu Wanli Da Ming Hui dian zuan xiu” 张居正与万历大明会典纂修 (Zhang Juzheng and the compilation of the Wanli Hui dian), Jiangnan da xue xue bao 12.2 (March, 2013): 46.   65 So, in order to return to the spirit of the early Ming, it was necessary to discard the blind faith to its institutions that were no longer applicable to current times.98   Since Zhang was a political reformer, we should consider this revision of the Ming Hui dian as a major part of this reform. In a political culture that revered every word of the founding emperor as sacred, and was often crippled by the inability to diverge from his plan, this editorial change signaled the abandonment of even lip service to the Ming administrative code, and an open acknowledgement that it was no longer relevant to contemporary government. Not everyone agreed. In 1579, Vice Censor in Chief Zhang Lu (張鹵, 1523-1598), who openly opposed Zhang Juzheng’s reforms, published a new edition of the collected Hongwu-era legal and administrative compilations titled Huang Ming zhi shu.99  Whether or not this was done in direct opposition to the compilation of the Hui dian cannot be gleaned from the preface to the compilation, but the re-appearance of the Da Ming ling, Da gao, and Zhu si zhi zhang suggests that Zhang Lu was making a stand regarding the administrative and legal standards of his contemporaries.    The Wanli edition of the Hui dian put an official seal to a departure from the original plan. In this sense administrative guides should be viewed not only as records of                                                       98 For a discussion of Zhang Juzheng’s thought see: Robert Crawford, “Zhang Juzheng’s Confucian Legalism”, State and Society in Ming Thought, ed. William Theodor De Bary, New York: Columbia University Press (1970), pp. 367-414. For a discussion of Zhang Juzheng’s efforts in reforming financial administration and the political climate of the time see: Ray Huang, 1587: A Year of No Significance (New Haven: Yale University Press, 1981).  99 Zhang Lu, Huang Ming zhi shu 皇明職書 (Administrative Compilations of the Great Ming, 1579, reprint Taibei: Chengwen chubanshe, 1969), 1a-b.  66 government organization but also as embodiments of government reform. It may also be said that the timing of the compilation becomes crucial for understanding important junctures of administrative reform. The Zhengde and Wanli editions of the Hui dian are both evidence of official activism. If late fifteenth-century reformers reiterated early Ming institutions in the face of deteriorating standards, sixteenth-century reformers tried to free administration from its Hongwu-era constraints. If this is the case, how was this revision related to reform in financial administration? As shall be discussed in the following chapters, while the task of the Hongwu emperor was to maintain decentralized control over financial administration embodied by the principle of ‘storing wealth among the people,’ Zhang Juzheng pushed for further centralization of state finance. This transformation in financial administration occurred long before Zhang’s time, but he was the person that recognized it and further pursued it for the benefit of the state.  Because it covers a longer period of the dynasty, the Wanli edition of the Hui dian became the standard edition used by scholars of Ming statecraft. And since it diminishes the importance of Zhu si zhi zhang for government administration it is easy to forget that there was in fact a plan for organizing financial administration. In the following pages I return to this neglected text and treat it as the basis for the institutional dimension of financial administration in the early and mid-Ming periods. Furthermore, this chapter places its compilation in the political context of the early Ming. It argues that the creation of the Ming Ministry of Revenue, its function and its limitations, are a product of political crisis as much as it is a manifestation of the personality and ideology of the Hongwu emperor.   67 2.2 Political Origins of Ming Financial Administration The structure of Ming government was finalized in the Zhu si zhi zhang, after a period of experiment and change. The final version was a response to several political scandals that shook the central government during the first two decades of the dynasty. It also reflects the ideology of the founding elite and the personality of the first emperor, who insisted on centralizing power in his own hands. The first government reform was initiated in 1381 after Grand Councilor Hu Weiyong (胡惟庸, d. 1380) was charged with conspiring to assassinate the emperor. As a result the emperor permanently abolished the position of grand councilor. Under the new government structure, ministers reported to the emperor directly and their authority to make policy decisions was curtailed. Soon after this upheaval, the court issued new government guidelines.100 This was probably done in haste since they were mostly copied from the Tang liu dian 唐六典 (Six Statutes of the Tang) but the particular design for financial management already showed discrepancies from the Tang model. This is most evident in the guidelines regarding the Board of Expenditure in the Ministry of Revenue.101 Literally meaning Board of Measuring Expenditure (度支部, du zhi bu), it became the most powerful financial agency during the second half of the Tang owing to its authority over accounting and the distribution of funds. The Tang liu dian illustrates                                                       100 Ming Taizu Shi lu, 130:5a-8a. 101 Translated as Revenue Section in the Tang and Bureau of General accounts in the Ming. See: Hucker, Dictionary of Official Titles.   68 its role in these words:  The Expenditure bureau (i.e. Board of Expenditure) is responsible for administering payments and state expenses. [according to] the changing numbers of tax revenue, the adequacy of manufactured goods, [and] the profits from land and water routes, it calculates annually what was sent out and pays what is needed.102  This paragraph was copied word for word from the Tang original except for the sentences highlighted above. The sentences highlighted here were omitted from the Ming version first because the early Ming government did not collect taxes from ‘land and water routes,’ or custom duties, and secondly because in the Ming, the Board of Expenses was not authorized to make annual calculations on a national basis or decide how revenue will be allocated. During the following decade the emperor’s suspicions of his officials only grew. The case of the pre-stamped documents (kong yin an 空印案) of 1382 was not in itself a clear-cut corruption case, but it was treated as such by the emperor. Once completing a transport of some kind, officials would check the amount received against the amount ordered and if it matched they would stamp the order document. If the amounts did not match, transporters had to turn back and make the trip again with the proper amount. In order to avoid being sent again to correct the numbers and save an extra trip, transporters were equipped with pre-stamped documents that could be filled in after all transportations were completed to ensure the numbers matched. But since this practice also had the                                                       102 The translation is from Twitchett, Financial Administration,100-101. For the original see: Li Linfu 李林甫, Da Tang liu dian (8th century, repr. Taibei: wen hai chu ban she, 1962), 3:43b.  69 potential to falsify reports, the emperor judged this procedure harshly and executed the officials involved in it.103 A tally system then replaced the former one in all government communications. This system used stub-books with detachable copies that ensured documents were not tampered with after being sent out of a government agency.104  As if to confirm his greatest fears, in 1385 the emperor discovered a nation-wide corruption case led by Vice Minister of Revenue Guo Huan 郭桓. Officials and commoners connected to agencies in the central government, and various administrative offices in most provinces were implicated with issuing false receipts for transportation of the grain tax, embezzling scrip and silver, money laundering, and outright stealing of grain from granaries containing military reserves. According to the emperor, the revenue loss was as high as twenty-six million shi of grain, about a full year’s income. Almost immediately after the discovery of the case, the emperor issued the first collection of the Da gao, in which he recounted some of the crimes connected to the case, and issued warnings to officials and commoners.105  During the next several years he issued three installments, one of them dedicated to soldiers. The emperor expected every household in the empire to hold a copy of these compilations and treat them as the law.                                                        103 Zhang, Ming shi, juan 94, 2318-19; Wu Han 吳含, Zhu Yuanzhang juan 朱元章卷 [Biography of Zhu Yuanzhang] (1965) (repr. Hongkong: Long men shu dian, 1973): 166-167. 104 For the imperial decree see: Huang Ming zhao ling 皇明詔令 [Imperial Decrees of the Ming Dynasty] (1539),(repr. Si ku quan shu cun mu cong shu, Jinan: Qi lu shu she, 1996), 2:24b-25b.  105 On the severity of punishments for financial crimes in the Da gao see Timothy Brook et al. Death by a Thousand Cuts, chapter 4 (Cambridge, MA.: Harvard University Press).. Also Lin Tai-yung, Ming T’ai-tsu and his Administration of Justice, PhD dissertation, University of Minnesota, 1979. Lin offers a detailed account of the Guo Huan case and its relation to the compilation of the Grand Pronouncements.   70 In the sections pertaining to financial procedures, the emperor addressed the corruption cases attributed to Guo Huan, as well as more mundane administrative problems that harbored potential corruption. His attention to minor misconduct is particularly interesting because it reveals his insistence on clarity and timeliness in government work. For example, in response to a complaint forwarded by the Minister of Revenue and his staff regarding the heavy workload, the emperor ordered an investigation that found that the 143 documents supposedly dated to a single day’s work had actually accumulated over two weeks. Only six of them were actually received on the day in question.106 In another case, the emperor lashed out at his officials for chaos in the annual commercial tax reports, forwarded individually from 2,437 local administration units. Some documents bore seals that were placed upside down or were missing a seal altogether. Some were missing names of the people taking part in the transaction while others did not separate tax income figures item by item. He concluded his public admonition of the careless officials with an overt death threat that must have been truly terrifying: “The souls of the dead are not yet gone, and new ones appear.” 107 But then he expressed his magnanimity by willing to pardon those who admitted their faults. A couple of months after Guo Huan was tried and executed, the emperor ordered the Ministry of Revenue to inscribe in stone the revenue and expenses of the year in the main hall of the office.108 This public display suggests that an accounting of public                                                       106 Ming Taizu, Yu zhi da gao chu bian (1385), repr.: Xu xiu si ku quan shu, shi bu, zheng shu lei vol. 862, Shanghai: Shanghai gu ji chu ban she, 1995-2002, article 67, 38a-38b. 107 Ming Taizu, Yu zhi da gao xu bian (1386), article 51, 43b-44b. 108 Ming Taizu Shi lu, 174:2a.  71 revenue and expenditure took place, but fixing the numbers in stone gives the impression that the emperor intended to fix the economy in a static state, as if the ultimate goal of the Ministry of Revenue was to maintain the same budget from that year on. This fits with Hongwu’s general vision of the economy: Fixing the population in their localities and professions and preventing both revenue and expenditure from growing.  Most of all, it reflects the emperor’s demand for complete transparency. This may not have seemed an excessive reaction on behalf of a ruler who had lost all confidence in the integrity of his officials. To him, lack of clarity was a sign of foul play. On the other hand, the Hongwu period was one of the darkest periods in the history of Chinese public life. Mass executions of officials and commoners, confiscation of property and exile plagued the administration for more than a decade. Few officials rising to high positions in government emerged unscathed. It was in this political context that organization of government was finalized. Zhai Shan (翟善, c. 1370-1451) was one official who managed to earn the trust of the emperor. In his capacity as Minister of Personnel, he supervised the compilation of Zhu si zhi zhang. The chapter dealing with the Ministry of Revenue outlines a compartmentalized agency functioning mainly as a giant communication centre, processing and verifying data as well as the transportation of funds and supplies throughout the empire. Owing to the circumstances of the early Ming and to the personality of the founding emperor, administration was, in principle, devoid of personal agency. The task                                                         72 of officials was to maintain a smooth flow of funds and information by recording, inspecting, and verifying every aspect of economic activity that was related to public revenue. In this sense, the financial apparatus created in the early Ming was close to Weber’s definition of a rational bureaucratic state. In developing his thesis, Weber referred to China as contrasting model.109  In his understanding, the Chinese imperial state operated according to ethical and magical assumptions on top of existing clan and commercial networks. The imperial bureaucracy was not a rational one because it was not manned by specialized trained men but by gentlemen. In contrast, the bureaucratic state, which resembles a capitalist enterprise, “rests primarily on calculation and presupposes a legal and administrative system, whose functioning can be rationally predicted, at least in principle, by virtue of its fixed general norms, just like the expected performance of a machine”.110 Weber contrasted the impersonal, machine-like modern state against the irrational state that rested on personal ties, honored traditions, and individual judgment. In the beginning of the Ming, Hongwu tried to crush these features by force and create an administration that would leave little place for personal ties and individual judgment. The following section demonstrates this through a closer look at the structure and roles of the ministry of revenue.                                                         109 Max Weber, “The rational State and Its Legal System,” Essays in Economic Sociology, edited by Richard Swedberg, 116-119 (Princeton, N.J.: Princeton University Press, 1999). 110 Ibid, 110.  73 2.3 The Structure of the Ministry of Revenue The general structure of the Ministry of Revenue follows that of the Tang as well but the functions of each department are different. A minister and two vice-ministers headed an office divided into four boards: The Board of Population (min bu 民部) dealt with the registration of population and land for tax collection, along with the administration of famine relief. The Board of Expenditure (du zhi bu 度支部) handled government payments. The Board of Money (jin bu 金部) supervised and reported all non-agricultural revenue. And the Board of Granaries (cang bu 倉部) supervised the maintenance of grain and fodder stores and shipments outside the provinces. Initially, the boards were separate offices with their own staff of a director, vice-director, and secretary. But in 1390 he replaced them with twelve provincial bureaus (qing li si, 清吏司) that handled the responsibilities of each of the former boards. Although these bureaus bore the names of the provinces they supervised, they were not part of the provincial administration and resided in the capital. Their task was to handle the documents moving back and forth from the capital to the provinces. Initially, each office was manned by a director, and vice-director. During the Zhengtong reign between three and four secretaries were added to each office. In some cases another vice-director was added to the staff with specific responsibilities. For example, a second vice-director in the Jiangxi office was responsible for military provisions in Xuanfu.111 The actual paperwork was done by ranked and unranked clerks: thirteen head clerks, another forty-one clerks (ling                                                       111 (Zhengde) Ming Hui dian, 3:4a-6b.   74 shi, 令史) and over a hundred unranked clerks (dian li, 典吏).112 Every bureau was issued an official seal, which implies they held a certain degree of executive power in their respective jurisdictions.   Since the Metropolitan Region (zhi li 直隸, or regions, after the establishment of dual capitals in the fifteenth century) had no provincial administration, the regular administration of the finances of prefectures, military guards, and tax agencies within it were divided between the bureaus. For example, in addition to handling the financial administration of the province, the Shanxi bureau was responsible for eleven military guards and battalions, one prefecture, two sub-prefectures, two granaries, and a salt distribution agency in the two metropolitan regions. Furthermore, against any geographical logic, the finances of the most important prefectures in the Nanjing region were divided between different bureaus. So while the Shanxi bureau dealt with Suzhou prefecture, the Henan bureau kept the records of neighboring Songjiang and Fengyang prefectures. 113  It is hard to understand the reason for this cumbersome arrangement other than further preventing the consolidation of financial information in one particular office. Some degree of consolidation is recorded by the time the Wanli edition of the Collected Statues was published when all Nanjing metropolitan prefectures were placed together under the jurisdiction of the Sichuan bureau.114  Following is a description of each of the four boards as it appears in the Zhu si zhi                                                       112 Ibid, 6:5a. 113 (Zhengde) Hui dian, 16:3a-5a and 10b-11a. 114 (Wanli) Hui dian, 14:8b-9a.  75 zhang. Although the boards were turned into departments in each provincial bureau, the discussion here will treat them as separate offices for the sake of clarity. Though it may be that tasks were allocated to clerks within each bureau according to the four categories, it might as well be the case that they all handled all of them.   2.3.1 The Board of Population The term min bu is translated here as Board of Population since it dealt primarily with matters such as census taking, famine relief, marriage and popular legal education.115 Population and land censuses were first and foremost elements of a fiscal apparatus. The basic fiscal unit was the individual household. Households were taxed according to the amount of land and other forms of property they owned, and preformed labour services according to the number of fit adult males. The census was thus the basic tool used to determine tax quotas. The quality of information on land and population for the early Ming surpassed those of the Song and Yuan dynasties.116 A national census was conducted every decade and recorded population and property in what became known as the Yellow Registers (huang ce 黃冊), yellow being the color of the binding of the imperial copy. The Fish-Scale Map Registers (yu lin tu ce 魚鱗圖冊) mapped the plots of land in each locality. These documents were stored in what was probably the largest                                                       115 Zhai, Zhu si zhi zhang, 3:93a-154b. 116 The Ming census also serves as the earliest source of reliable information for the study of historical demographic patterns. See: Ho Ping-ti, Studies Population of China (Cambridge, Mass.: Harvard University Press, 1959), 3. And  Dwight Perkins, Agricultural Development in China 1368-1968 (Chicago: Aldine Publishing Company, 1969).  76 archive in the world on an island in Xuanwu Lake at Nanjing.  The first census of the territory under Ming control was announced in 1370. In the proclamation, the future characteristics of the Hongwu-era administration were already apparent. It entailed the first application of the tally system: each household received a blank form to fill out, which contained two duplicates that could be torn along a seal. One half was sent to the ministry while the other was retained by the household and was presented in case of inspection.117 Households that tried to avoid registration were registered as military households, and officials that were caught reporting false numbers were decapitated.  Once the regulations for the Yellow Registers were established in 1391, the Ministry of Revenue remained nominally in charge of the census, but its staff did not participate in census inspections or in archive maintenance. In charge of the census were two investigating censors, an official from the Revenue Scrutiny Office (hu ke 戶科) and four secretaries from the Ministry of Revenue. 118 Students from the imperial academy carried out the tasks of population surveys and archive maintenance. Students usually                                                       117 The text of the proclamation was transcribed into a local gazetteer along with the registration form, as filled by one family in the county. This is described in Arthur W. Hummel, “A Census Report for the Year 1370 A.D.”, Annual Report of the Librarian of Congress for the Fiscal Year Ended June 30, 1940, (Washington: United States Government Printing Office, 1941), 158-159. A photocopy of the declaration and registration form appears in George Sarton, Introduction to the History of Science, vol. 3, part 2, (Baltimore, 1948), 1268-1270. 118 The revenue scrutiny office was an office outside the ministry of revenue in charge of inspecting its procedures. Zhang Wenxian finds that in the fifteenth century, a special commission took over the duties of the previous committee. See: Zhang Wenxian, “The Yellow Register Archives of Imperial Ming China”, Libraries and the Cultural Records 43.2 (2008).  77 served for one year before continuing to their first official post.119 The deteriorating quality of the data can be attributed to diminishing personnel. The first emperor assigned 1,400 students for the task. This number fell to 800 in 1442, and 350 in 1493. One could argue that the original number was not enough to keep track of the population to begin with, but by the end of the fifteenth century it was hardly enough even to maintain the archive. It is also probable that the number of personnel dwindled because updated population and land statistics were no longer important for financial management. By the sixteenth century, the Yellow Registers became a repository of fixed quotas that had little relation to demographic and economic reality. On the one hand this was a sign of weakening control over local administration. By settling for fixed quotas, the state obtained revenue well under the full tax potential of the economy.120  On the other hand, this was a blessing for local officials who were left with some leeway in tax collection. By under-reporting the figures of registered households and cultivated land, officials could rely on a larger pool of revenue in order to meet extra expenses and demands from the central government. One method was to continue reporting an original household and not its descendants. This way more households were sharing the tax burden of one.121 For this purpose, local officials created their own local registers called  Complete Records of Tax and Labour Service (fu yi quan shu, 賦役全                                                      119 Bin Xie 彬謝, Nanjing hu bu zhi 南京戶部志 [Gazetteer of the Nanjing Ministry of Revenue] (1550),  5:10b-11b. 120 Huang, Taxation and Governmental Finance, 46-47. 121 John W. Dardess, A Ming Society: T’ai-Ho County, Kiangsi, Fourteenth to Seventeenth Centuries, (Berkeley: University of California Press, 1996):74-76.  78 書,) that were not reported to the throne. The information in these registers may have been more accurate than that of the Yellow Registers, but they were essentially the same kind of compilation, containing fixed tax quotas that were used as a basis for taxation.122  Land was registered in two main categories: government land (guan tian 官田) and private land (min tian 民田). Government land was further divided into a number of categories according to their fiscal obligation. The main difference between private and government land is that the first was subject to taxation and could be bought and sold, while the second was worked by tenants who paid rent and could not be traded. Originally, the income from private land was the primary source of local revenue and the income from rent financed the capitals and the central government. But these distinctions blurred in time. It should be noted that the Board of Population was responsible for collecting data on land registration but not on the actual grain tax. But it did supervise the transportation of sericulture taxes, paid in bolts of cloth. Once these shipments arrived in the capital, it was up to the Population Bureau officials to verify the shipping documents before entering them in the imperial storehouses.  Perhaps as part of their role to register the population, revenue officials were also given the task of enforcing the social prohibition to marry within the same surname as well as other aspects of the law.123 This was apparently done by posting placards in places of interaction between revenue officials and tax captains or corveé labourers. In                                                       122 Iwami Hiroshi, “An Introduction to the Shandong jing hui lu”, State and Society in China, ed. Linda Grove and Christian Daniels (Tokyo: University of Tokyo Press, 1984): 311-334. 123 Zhai, Zhu si zhi zhang, 3:108b-109a.  79 this capacity, officials that were responsible for fiscal control were also responsible for social control. Since revenue officials had demographic information, and since they came into contact with commoners in tax payments and labour services, they were in a position to supervise the legality of marriages. Moreover, there may be a fiscal reason for this extra diligence. Marrying within the same surname could disguise the creation of a new household. It is unlikely that many of the commoners could read the statutes and regulations. Besides officials conducting learning sessions, it is hard to imagine how this task could have been carried out. The Department of Population also oversaw special allocations of funds. In the wake of the nation-wide embezzlement case, strict procedures were devised to ensure control over the transportation of funds. Once an allocation of extra funds from one granary to another was approved, revenue officials had to fill in a stub book that had two identical copies, one of which remained in the Ministry of Revenue and the other sent to the receiving office. The information to be filled in on this form contained the date of issue, the name of the official issuing the shipment, a receipt number, and the name of the provincial or prefectural office requesting the funds. Once the shipment was completed the official had to write a report on the manner in which it was carried out. This information was later used to investigate the granary or storehouse officials when their time came to change office. A clear account record would result in a promotion. In this capacity, revenue officials were also responsible for other important functions such as the routine inspection of military reserves of grain and fodder as well as the condition of roads for transporting military supplies, and the allocation of funds in cases of natural disaster and famine. Revenue officials were to travel to the area affected  80 and prepare an itemized report including the nature of the disaster, the number of households affected, their names, their property and tax obligations. Using this information a memorial was sent to the throne with an estimate of the aid administered locally and an estimate of extra aid, whether in grain or scrip, for the approval of the emperor. The reaction to famine was therefore twofold; first immediate and local, and in the case this was not sufficient, officials applied for extra aid directly from the court. The task of revenue officials was to verify the size of aid needed and officiate the allocation of extra funds. Finally, a special register of the entire process was compiled and archived for future reference.  The final task of the department was bookkeeping, or accounting (kuai ji 會計). This is the only mention of accounting throughout the entire chapter of the Zhu si zhi zhang. It does not refer in any way to total tax revenue but is applied to the maintenance of granary reserves. Still, this passage is interesting because it reveals the accounting techniques of the time. The department was required to produce two reports: the first containing data on the previous year and the second containing a projected budget for the following year. In the first report all stored grain, beyond three years of military reserves was to be calculated according to “existing sum” (xian zai 見在), ”deficiencies” (bu fu 不敷), and “surplus” (yu sheng 餘剰). The second report detailed expected income and expenditure for the coming year. The categories in this report were “remainder from previous years” (jiu guan 舊管), “new revenue” (xin shou 新收), “disbursement” (kai chu 開出), and “net sum” (shi zai 實在). These reports were supposed to be jointly reported  81 every year to the Ministry of Revenue.124  The accounting procedure that is described here is the “four-column” (si zhu 四) accounting system that was practiced in China at least since the Tang and was the most sophisticated accounting procedure of the time.125 In terms of budget making, it is clear that the knowledge and expertise of former dynasties was retained in the Ming. Granary officials kept the accounts. Then, assuming that the information was forwarded to the local magistrate and on to the provincial administration without hindrance, top revenue officials at the centre could produce a similar report at the national level. It is important to note that in this case, accounting procedure is applied in order to manage supplies and is not used in the context of revenue. In other words, there was no institutional arrangement or requirement to aggregate tax data from various sources in one document. For one, aggregate tax income would be impossible to calculate since tax was collected in different commodities and monies. There was therefore no concept of total national revenue that could then be distributed according to a government plan. On the other hand, since grain was the main source of income and expenditure, it is understandable that the level of grain supplies was almost equal to total revenue.                                                        124 Ibid, 3:109a-109b. 125 Guo Daoyang郭道扬, Zhongguo kuai ji shi gao 中国会计史稿 [A Draft History of Accounting in China], 2 vols., (Wuhan: Zhongguo cai jing jing ji chu ban she, 1982); Chen Shimin, “The Rise and Fall of Debit- Credit Bookkeeping in China- History and Analysis”, The Accounting Historians Journal 25.1 (June, 1998): 73-92; Y. L. Zhao, “A Brief History of Accounting and Auditing in China,” Accounting and Auditing in the People’s Republic of China (Dallas, TX: Centre for International Accounting Development, 1987): 165-191.   82  2.3.2 The Board of Expenditure Government expenses were organized under seven categories, each with their own specific procedures: grants and awards, military salaries, military salt rations, miscellaneous payments, government salaries, granary allocations (for traveling officials), and travel provisions for soldiers.126 Every payment was conducted through a similar procedure of verification using stub-books. The main role of officials was to communicate requests for funds, verify them, and follow up on the completion of the transport. In other words, officials in the Board of Expenses dealt with the paper work of transporting funds. The term “awards and grants” (shang ci 賞賜) essentially meant any form of monetary or non-grain irregular payment. Grants could be issued to meritorious officials or foreign dignitaries, but they were especially important in supplying the military with cloth and uniforms. Initially, grants were issued in scrip, but as early as the Hongwu reign other commodities that often replaced scrip appeared, such as Indian pepper and sappan wood.127  Grants were issued from the palace treasury (nei fu 內府). This was a general term for a number of storehouses that collected vast amounts of scrip, cloth, silver and gold, ivory and pearls that came from commercial taxes, the sericulture tax, converted grain tax, tribute levies, and money from fines and confiscated property. Although                                                       126 Zhai, 3:111b-127a.  127 These commodities will be discussed in chapter 3.   83 commonly regarded as the private treasury of the emperor, it had an important public function in a system that made no distinction between these two capacities.128 Beyond palace consumption, the palace treasury paid out grants and awards, issued funds for famine relief, and sent emergency funds to the military. By accommodating special requests from the provinces to the throne, the palace treasury could supplement funds that the local system of granaries and storehouses lacked. In this sense, it played an important role in public finance. By involving the emperor directly, the allocation of funds could bypass the regular avenues of the bureaucracy and provide a swift response to pressing needs anywhere in the empire, serving as a ready depository for immediate action in case of emergency.  The proper procedure for issuing grants for the military began with a formal request to the Ministry of Revenue. The ministry then assigned officials to verify the names and numbers of the soldiers in question with their own records, and calculate the amount of funds requested. Once the information was deemed to be correct, a memorial was sent to the palace treasury where the numbers were again verified. Then functionaries from the Board of Population were sent to carry out the shipment. Originally, this means that the jurisdiction of the Board of Expenses ended once the payment was authorized. This would maintain a segregation between the officials extracting the funds from the palace treasury and those who transported them. But once these departments came under the same office, there was no clear separation between the                                                       128 Liu Ying 刘穎, Ming dai nei cheng yun ku shi tan 明代內承運庫試探 [The Study of nei cheng yun ku in Ming Dynasty], M.A. thesis, (Shandong University, 2009): 12-13.  84 tasks.  Monthly request forms detailing the number of officers and officials in the various capital offices and agencies were sent to the ministry for approval. In the case of monthly grain rations for soldiers, although the grain was procured from a local granary, the request had to be approved in the capital. A request was first made to the Board of Expenses, which prepared duplicate documents and sent them to the requesting unit and the supplying granary. The documents could later be matched in order to ensure that what left the granary tallied with what the unit received. A signed copy was finally stored in the ministry’s archives to be used in granary inspections. The same procedure took place for the allocation of monthly salt rations to soldiers. Salt rations for soldiers outside the capital were paid in paper notes at a rate of 100 wen per jin of salt. Salaries for civilian officials outside the capital were paid in scrip as well at a rate of two and a half guan per shi of rice. Issuing miscellaneous payments was another standard function of this office. The examples for such payments in the Zhu si zhi zhang include payments for building projects undertaken by the Ministry of Works or the production of military uniforms. Again, the requesting agencies had to submit a request detailing the funds and supplies they intended to use. These were cases in which local administrative units could apply for funds from the central government in addition to their own budgets.  2.3.3 The Board of Money The Boards of Granaries and Money were responsible for supplying the court.  85 The Board of Money specialized in money and money commodities that as a rule went to the Palace Treasury. These items were derived from commercial and monopoly taxes (ke cheng 課程), and fines (fa zang 罰贓). It deposited all the tax that was paid in money in the appropriate imperial storehouses. Like the Board of Granaries, it verified the amounts transported, but had no access to the funds themselves. Therefore it operated as another accounting agency. The revenue it monitored included proceedings from the salt and tea monopolies, commercial taxes, fines, and income from government silver mines. Money was collected in paper notes, gold, silver and coin. Bolts of silk products were also collected as a kind of money. The new code of 1393 differed from the previous one in one particular aspect: whereas formerly officials from the Board of Money also handled tribute arriving on foreign ships, the only legal avenue for foreign trade, this function was terminated, probably as part of the new restrictions on foreign trade. The sensitivity in the transportation of such high value items is reflected in the extra precautions taken to prevent embezzlement. Money and precious metals had to arrive at the imperial storehouse with “the seal unbroken,” meaning that the chests were sealed and shipped intact to the capital. Once again, the role of the revenue officials was confined to following the transaction as supervisors, verifying its proper completion, and recording it for future reference.  2.3.4 The Board of Granaries  The Board of Granaries handled the calculation and recording of the grain tax, as well and the transportation of tribute grain tax from the provinces to the capital or to  86 other destinations.129 Every year the tax income was calculated. Locally retained funds that included military reserves, official salaries, student stipends, and welfare reserves were distinguished from surplus funds, which were reported to the court and shipped, either to the capital or to military units. Tribute tax from the directly administered prefectures was handled in the same way. First, grain for military, administrative, educational, and welfare purposes was retained. The text states that “in case there is more than is needed” arrangements should be made to transport them through designated tax captains.130 The way in which it was distributed reflects the intent of the first emperor to store for “All Under Heaven” because only what was left after calculating official and military salaries, welfare funds and an additional two-year reserve for military guards, was transported to the capital. In other words, there was no separate budget for the central government. In preparing grain for shipment, officials first sent a report on available funds, subtracting two years’ worth of military pay, official salaries, student stipends and welfare funds. Only the extra funds were reported to the ministry. Officials from the Department of Granaries received tax captains that shipped grain to the capital and handled the procedure of depositing the supplies in the imperial granary. After that they recorded the completion of the transaction. These officials stood at the very end of a long chain of transactions. The Tax Captain (liang zhang 糧長) supervised the Community Head (li zhang 里長) who supervised the Tithing Head (jia shou 甲首). Finally, the                                                       129 Zhai, Zhu si zhi zhang, 3:139b-154b. 130 如有多足用. Ibid, 3:140.  87 tithing head supervised the individual households.131  It seems that the Tax Captain actually made two trips. The first was to report the extra tax income in the area under his jurisdiction. Then he would receive a shipment document to take back with him. In the second trip he shipped the funds to their designated destination. After the capital moved to Beijing, tax captains continued to obtain their certificates from the Nanjing Ministry of Revenue.132 This was done so that the Ministry of Revenue could be able to verify that the amount arriving in the capital tallied with the original report.  The only point of interaction between government officials and commoners was on the level of the tax captain who supervised the transportation of the tax. This means that the main role of the Granary officials was to verify the reports on retained and extra tax income and to assign and check the shipment documents. More inspection was done once the shipment had entered in the respective granaries and storehouses. The storehouse managers had to report every new acquisition to the scrutiny office, which would check all the documents. Only once the receipts tallied with the shipment document of the Tax Captain would he be allowed to go back home.133                                                       131 For a discussion of the relationship between the tax captains and the li jia system see: Martin Heijdra, The Socio-Economic Development of Rural China During the Ming, PhD Diss. (Princeton University, 1994): 104-151. 132 Liang Fangzhong 梁方仲,” Ming dai liang zhang zhi shu yao” 明代糧長制述要 [On the tax captain system of the Ming dynasty], Ming Qing Shi lun cong, (Hubei: Hubei ren min chu ban she, 1957. Reprint: Liang Fangzhong jing ji shi lun wen ji fu bian, Henan: Zhong zhou gu ji chu ban she): 123   133 Zhu si zhi zhang, 3:141a-141b.  88  2.4 Annual Revenue Reports The annual revenue reports represent a record of accounting on a national level that took place before the sixteenth century. The Ming Shi lu contains these reports for the years between 1403-1522. The information is organized in lists of income from various taxes and state monopolies. Other than the declaration that these reports pertained to the income of the entire realm (tian xia), we have no information on what the numbers actually indicated. Moreover, it is not clear why the editors of the Shi lu decided to place these reports at the end of the last day of the calendar year. Since Shi lu were compiled from daily court diaries the inclusion of the reports likely indicate that they were presented to the emperor at the end of every year. On the other hand, unlike most entries, there is no indication of an audience with the emperor, no mention of the presenters, nor any reaction from the emperor. Moreover, the standardized manner in which the data is organized suggests that it was a written report based on a fixed template. In this case, it could be that the written form had been either attached to the court diaries, or subsequently incorporated into the Shi lu from a file of annual reports. The first Shi lu complete revenue report looked like this: 是歲天下戶千一百四十一萬五千八百二十 九口六千六百五十九萬八千三百三十七人稅粮三千一百 二十九萬九千七百四石布帛一十萬五千四百二十六匹絲 棉三十七萬九千二百一十五斤綿花絨十六萬二千二百四 十九斤課鈔五百六十萬六千八十七錠金五十兩銀八萬一百八十五兩銅二千四百二十三斤鉄七萬九千八百六斤鉛 六萬二千四十二斤硃砂千八百五十五兩海肥二十六萬六 千七百五十四索茶百六十七萬八千一百七十斤鹽百二十 九萬二千八百六十二引屯田子粒二千三百四十五萬七百 九十九石馬三萬七千九百九十三匹 This Year throughout the realm: Households- 11,415,829, Adults- 66,598,337  89 persons, Grain Tax- 31,299,740 shi, Hemp and Cotton- 115,426 pi, Cotton and Silk- 379,215 jin, Raw Cotton- 162,249 jin, Commercial Taxes- 5,606,087 ding, Gold- 50 liang, Silver- 80,185 liang, Copper- 2,423 jin, Iron- 79,806 jin, Lead- 62,042 jin, Cinnabar- 1,855 liang, Shells- 266,754 string, tea- 1,678,170 jin, Salt- 1,292,862 yin, Grain from Military Land- 23,450,799 shi, Horses- 37,993 pi. 134    We can only assume why these reports were recorded in the last day of every year. I find it hard to believe that the emperor was engaged in a national budget meeting for the coming year on the eve of the Lunar Festival. But the reports may have played a certain ritual function. During the Yongle reign the reports appear right after an imperial visit to the ancestral temple. In some cases these two entries are actually coupled together as one occurrence. Perhaps Yongle was showing his dead father that the empire was in good shape. For example, on the last day of his first year on the throne, which was also the first year of the revenue reports in this format: 壬寅享 太廟    是歲天下戶千一百四十一萬五千八百二十九口.135  Leaving space after “ancestral temple” may have been enough for denoting a separation between the two events. It could also be that the scribe simply forgot to enter one, since punctuation marks do appear before subsequent reports. Still, the highly ritualized nature of the last day of the year must have given special meaning to the revenue reports as well.   In any case, the regularity and meticulousness of the reports indicate that the Ministry of Revenue did in fact recover from the administrative blow dealt to the central                                                       134 Ming Taizong Shi lu, 26:7b. There is no demarcation between the visit to the ancestral shrine and the revenue reports also in 1407: Ming Taizong Shi lu 62:5b. 135 Ibid, The spaces are in the original text.   90 government after the abolition of the prime minister’s office, and that despite its compartmentalized structure, it acted as a central agency that was able to process large amounts of data and present it in one unified report. The term tian xia is important because it announces the intention of the compilers to encompass the total revenue of the realm. And since the reports were compiled annually they had to rely on annual updates from the provinces. The provincial administration commissioner dealt with the finances of a province, but grand coordinators were even more involved with financial issues and often communicated with the throne regarding deficiencies and sent memorials with constructive suggestions. At least during the fifteenth century, the degree of control prefects had over local administration and their cooperation with the grand coordinator and regional inspector are reported to have been rather high.136   Moreover, The kind of data that is reported suggests that it was a tool for preparing the budget for the coming year. Since the categories listed do not contain special items appropriated by the court for palace consumption but monies and commodities that were used for public purposes, it could be said that although there was no separation between public and private income, the reports were used as a tool for state finance, guo yong, rather than an account of the emperor’s wealth.   Finally, the consistency of the reports manifests the principle of clear and timely updates on tax revenue. This was the main task the first emperor delegated to his revenue officials and may therefore be seen as a logical outcome of the administrative structure                                                       136 Nimick, Local Administration, 51-60.  91 that was established in the previous reign. On the other hand, this kind of report suggests the ability to centralize financial data in the hands of people outside the emperor. During the Hongwu reign, the recording of financial information was inconsistent. For 1385-86 only grain and horses are listed. Population and land statistics appear every decade upon completion of the decennial census, and occasionally, data on revenue from the sale of monopoly tea and salt appeared, as well as income from silver mining.137 In 1394 a more standardized and fiscally oriented format appear. The first part contains data on total grain reserves, and tax revenue in scrip, precious metal, and cloth. The second part contains a list of minerals and commodities: mercury, saltpeter, alum, lead, iron and salt, but also tea, silk floss, and cowries.138 During the short Jianwen reign (1399-1402) no reports were recorded.  The correlation between the improved quality of financial data and the new reign suggests that the responsibility for the change lay in the personnel of the Ministry of Revenue. No individual is credited for this operation but throughout the Yongle reign and until 1430, Xia Yuanji (夏原吉, 1366-1430) dominated the Ministry of Revenue. He began service in the ministry under the Hongwu emperor and rose through its ranks after making the right choice and siding with the future Yongle emperor. In addition to becoming Minister of Revenue, he became one of the emperor’s most trusted officials. The history of the Ming extols him for his role in re-organizing the tax and corveé system                                                       137 Ming Taizu Shi lu, 206:5b. 138 Ibid, 230:6b.  92 and overseeing the reconstruction of the region of Zhexi after a flood.139 But it is another biographer, Qiu Jun, who describes him as the person who held the financial management of the dynasty in his hands for close to forty years. Under his good management the state did not know scarcity.140  Huang’s opinion that “in the Ming period no hu-pu shang-shu ever rose to become a Cochrane or Colbert,” pointed at the inability of revenue ministers to initiate policy. But within the limits of what they could do, Xia Yuanji was precisely such a person.141   Intentionally or not, the reports are another representation of the principle of Expenditure according to Revenue since there is no data on expenditure. Moreover, it is not clear whether the data is a result of the accounting method described above. At least for the grain tax, it most likely represents a sum of total owed taxes rather than actual receipts. For example, in 1425, Minister of Revenue in Beijing, Guo Zi warned that the total grain tax revenue for the last three years amounted to under 23,000,000 shi.142 In contrast, the annual revenue reports recorded a total of 90,000,000 for those three years. Therefore, it is most likely that the numbers indicate anticipated rather than actual income.   Another important feature to bear in mind is that the tax categories could                                                       139 Zhang Tingyu, Ming shi, juan 149, 4150-4155. 140 Jiao Hong 焦竤 1541-1620, guo chao xian zheng lu 國朝獻徵錄 [Records of Contributions to the Dynasty] Si ku quan shu cun mu cong shu, (Jinan: Qi lu shu she, 1996), vol. 101, 415. 141 Huang, Taxation, 11. 142 Taizong Shi lu, 250:7a-7b.  93 represent units of account as the actual items indicated. As shall be discussed in the next chapter, portions of the grain tax were often substituted with other means of payment, even before the official commutation of the tribute grain tax into silver. And while there are some categories of tax commutation, they do not cover the more important aspects of this practice. This is clear from the fact that silver is never reported as commuted tax. This reality limits any attempt to assess the actual income from tax revenue at any given time, since revenue could be higher or lower depending on the commutation rate. The generality of the information doesn’t enable us to assess how comprehensive it is either. Nevertheless, there is evidence that the Ministry of Revenue not only expected timely and accurate data but had the means of checking it too: The finances of the entire realm are reported every year to this Ministry. Those who send false reports should be arrested and punished. The data sent on grain this year from Shaanxi provincial administration and Yangzhou prefecture was not only late but wrong. 143   Despite their limits the annual revenue reports are still a unique source of economic data in an era that was relatively silent regarding financial matters.  Another important aspect is the fact that data was updated annually. This suggests that unlike Hongwu’s desire to set the annual budget ‘in stone’ and unlike the fixed quotas of the latter half of the dynasty, throughout most of the fifteenth century, fiscal accounting may have been more attuned to economic reality. This assumption is corroborated by the relative quality of the data. In comparison with the other Ming                                                       143 Ming Yingzong Shi lu, 74: 9a.  94 compilations the annual reports stand out as the most accurate. As van der Sprenkel has shown, beyond clerical errors that are easily explained, the data on population is consistent and reliable. 144 It was also updated annually. Only after the revenue reports ceased to be reported in the Jiajing reign, figures appear once every decade. The reports are a more reliable source of information regarding land statistics as well. Previously, scholars relied on the Hui dian that lists 8,804,623 qing of taxable land for the year 1391, but only 4,292,310 qing for the year 1501, and a similar number for the Wanli period. This huge discrepancy led Ming commentators as well as modern scholars to conclude that between the fourteenth and the sixteenth centuries the dynasty lost control over half of its cultivated land.145 Fujii Hiroshi convincingly challenged this assertion by showing that the higher figure must be a clerical error.146 By collecting land statistics from prefectural and provincial gazeteers, he proved that wrong figures were perpetuated in the administrative codes of the dynasty. In the annual revenue reports land figures remain around 4,000,000 qing throughout the fifteenth century with one exception; for the Hongzhi reign (1488-1504) land figures jump to eight and even 9,000,000 qing. Since the Hui dian was under compilation at the time it may be that the                                                       144 Van der Sprenkel, “Population Statistics of Ming China,” 289-326. 145 Wang Qi 王圻, Xu wen xian tong kao 續文獻通考[General history of institutions and critical examinations of documents and institutions continued], Si ku quan shu cun mu cong shu ser. 3 v. 187 (Ji’nan: Qi Lu shu she, 1995): 36:14a-b.  146 Fujii, Hiroshi藤井宏. "Mindai dendo tōkei ni kansuru ichikōsatsu." 明代田土統計に関する一考察 [An investigation of Ming land statistics]. 1943, 1944, 1947, Tōyō Gakuhō 30, no. 3 (1943): 90-123; 30, no. 4 (1944): 60-87; 31, no. 1 (1947): 97-130. This finding has been noted by Martin Heijdra, “The Socio-Economic Development of Rural China during the Ming,” in The Cambridge History of China vol. 8, eds. Denis Twitchett and Frederick Mote (Cambridge: Cambridge University Press, 1998): 444-445.   95 data was copied from it, which further attests to the quality of the reports in this period. The error was corrected during the next reign and the number dropped back to around four million.  But other than this correction, the nature of the reports changes during the Hongzhi-Zhengde reigns. In some categories there are round numbers that suggest fixed quotas. In others, figures are simply copied from one year to the next. Whether the system was moving to fixed quotas or whether there was simply lack of current information, in the turn of the sixteenth century they ceased to represent economic reality and reflected a deeper change in the nature of financial administration. Fixed quotas in certain categories appeared already during the Yongle reign and they are connected with a loosening of government control over certain aspects of the economy. For example, during the Yongle reign the annual income from cinnabar remained fixed for several years at a time: 20,080 liang between 1406-1410, 1,579 liang for the years 1412-1413, and 1,504 liang between 1417-1424. During those years income from lead was fixed at 20,780 jin, and copper at 2,128 jin.147 Moreover, the quotas for these items dropped significantly in the following years and by the second half of the fifteenth century disappeared completely. A similar process occurred for iron quotas: Between 1436-1464 the 74,583 jin of iron were reported repeatedly after which the category disappears from the records.148 Throughout the Chenghua reign the fixed quota for cotton wadding was 105,000 liang, and other cloth categories remained the same for several years at a time. But it was                                                       147 See table D-2. 148 See Table D-11.  96 only during the Hongzhi period that all the categories besides the categories of land, population and grain, and fodder were practically identical throughout the reporting period. In the Zhengde period the reports become completely meaningless when even land and grain tax figures remained unchanged. During the Longqing reign (1567-1573) annual reports reappeared but again, the data is copied from one year to the next. From this time until the end of the dynasty no similar reports are recorded. Since the whole point of reporting tax figures annually is to update information, the growing repetition of information can be used as a measure for a deteriorating administration precisely in the realm in which it was formerly relatively strong: in the communication of financial information between locality and centre.  Within the confines of Ming financial administration, the ability to present clear and updated information to the emperor was key to the successful management of the empire. Without official channels throughout which information and funds could flow, management of the empire would be impossible. In order to maintain these channels, ongoing cooperation between local and provincial officials, between different ministries, and between officials and the emperor was necessary. The appearance of the revenue reports suggests that until the last decade of the fifteenth century this was the case. Once the infrastructure that supported the accumulation of financial data deteriorated, officials had to readdress the principles of accounting.  2.5 Sixteenth Century Developments  Li Chengxun 李承勛 (jin shi 1493) who served as Grand Coordinator for  97 Liaodong in the beginning of the Jiajing reign and witnessed the chronic lack of supplies on the north eastern border, opened his memorial to the throne with a sarcastic comment, saying that while good financial administration takes income as a measure for expenses and bad financial administration does the opposite, the Ministry of Revenue is doing neither. He added that in recent years, although accounting is part of the ministry’s duty in name, in practice there is no actual measuring of expenses according to revenue.149  He was echoing Qiu Jun 邱濬 (1421-1495) who a few decades earlier suggested the historical origin of the problem hi described:  Our Dynasty abolished the [post of] Grand Councilor and had the Ministry of Revenue manage the population, land and taxes of the realm. This however dispersed these tasks among subordinate offices, which ended up simultaneously managing these obligations. The tasks are numerous and complex but there is no office to consolidate them. Your Minister requests that, as in ancient times, in addition to the directorship of a finance minister (ji xiang 計相) and the assistance of the minister of revenue, another position be added [at the level] of Minister that will consolidate all national finances.150  Qiu Jun pointed at a central problem in Ming financial administration- the empire was operating without an accountant. And indeed, in the last two decades of the fifteenth century, the revenue reports show no meaningful accounting work. Qiu Jun                                                       149 Chen Zilong 陳子龍, Huang Ming jing shi wen bian 皇明經世文編 [A Compilation of Statecraft Writing of the Ming Empire] (1638),100:9b-10b. Reprint: Si ku jin hui shu cong kan, (Beijing: Beijing Chu ban she, 2000), vol. 23, 308. 150 Qiu, Da xue yan yi bu, 24:6b.   98 acknowledged the failure of the Hongwu vision of national finance and suggested to emulate past dynasties in placing a chief finance official.   The Jiajing emperor was the first to acknowledge the need for national accounting. He ordered the compilation of a national account book that would, for the first time, give equal weight to expenditure. This new emphasis on expenditure also indicates that it was a growing concern. In order to manage it better the emperor ordered a return to the accounting procedures of the Tang and Song dynasties. In 1543 he issued a decree for the production of a national budget emphasizing the long neglect of annual expenses: In the twenty-second year of the Jiajing reign (1543), the Emperor confirmed the petition of the Ministry of Revenue to follow the practice of former dynasties’ national accounts (國計, guo ji). In the end of each year it is to calculate stipends for princely households, salaries for holders of noble titles, government salaries and grain rations, monthly military grain rations for border and inland units, and tribute grain to the capital comprising actual grain and converted grain, with a distinction between collected and uncollected tax. The total amount of the realm’s income and expenditure should be written down in a simple and convenient manner that can be posted on a placard and presented to the throne.151  The first actual result to emerge from these initiatives was the Da Ming kuai ji lu  (大明會計錄, Account Book of the Great Ming), allegedly presented to the Jiajing emperor, but it is not extent. In 1576 Minister of Revenue Wang Guoguang began compiling the                                                       151 Wanli Ming Hui dian, juan 24, p. 626.  99 Wanli kuai ji lu, which was presented to the emperor in 1581.152 It contains detailed lists of regular expenses alongside annual income. In the preface, both the Tang and Song examples are mentioned as predecessors of this genre but no other similar compilation in the Ming, suggesting that the compilers considered it the first of its kind during their dynasty.   In his work on the history of accounting in China, Philip Fu suggests that the purpose of the account book was to warn the Wanli emperor of the dire financial strait the empire is in, rather than invigorate or reform financial administration.153 Ray Huang maintains that this was the apex of Zhang Juzheng’s reforms in Ming accounting. In the years before the completion of the account book, Zhang Juzheng abolished government and military posts in order to cut expenses. He ordered audits of tax and labour service quotas in all counties, and presided over a nation-wide land survey. The information was gathered in the Wanli account book and served as the basis for the fiscal chapters of the Wanli Hui dian.154   Even if this was part of a serious attempt to reform financial administration, the mechanism of tax and data collection was too damaged to permit meaningful change. Though the Wanli edition of the Hui dian lists expenses alongside revenue figures, the quality of the information is questionable. The editors of the Hou hu zhi (後湖志,                                                       152 As mentioned on p. 51, a preliminary analysis of this text appears in Liang Fangzhong, “Shu ji ping lun.” 153 Fu, “A Study of Governmental Accounting in China,” 150-151.  154 Huang, Taxation and Governmental Finance, 296.   100 Gazetteer of the Yellow Registers Archive) preferred to include population statistics from the Zhengde rather than the Wanli edition. Similarly, the editors of the eighteenth-century imperial library, Si ku quan shu, did not even include the Wanli edition for a similar reason: In the short preface to the Zhengde edition they explain that Wanli-era administration was so broken down that the Hui dian was a badly written text. Why did it take so long for such an account book to appear in the Ming? Since it has been established that there was no lack of motivation or ability in financial administration, we may assume that this was an application of the idea of Revenue as a Measure for Expenses. According to this logic, there is no need to count expenditure because it depends entirely on revenue. In the sixteenth century, the naïve idea that expenditure can be disregarded was no longer an option for administrators. By giving equal weight to revenue and expenses administrators took a more realistic approach and acknowledged that in order to survive expenditure must be managed.  Zhang Juzheng’s reform concentrated on tightening control over tax collection and reducing corruption and fraud in the bureaucracy. His aim was to strengthen the political centre and in that he succeeded marvelously. Before his death, the granaries in Beijing had grain enough for nine years. And the national silver treasury contained an emergency store of four million taels. Tax collection and auditing were more efficient, and tax evasion was prosecuted with more vigor. But, as Huang mentions, Zhang never attempted to carry out a reform in governmental finance. Local budgets were not increased so under the new audits officials were more vulnerable to state demands. There were also other opponents to these reforms; landlords naturally did not want to pay higher tax, students at the imperial academy resented the cut in official posts, and officials who  101 refused to pay travel expenses now denied from them. Zhang’s initiatives survived only as long as he remained in power.155  As mentioned earlier, Zhang appreciated the organizational capacity of the early Ming, especially its emphasis on the collection of tax data and auditing of accounts. He of course lived in a time when it was impossible to return to the system of “storing wealth with the people.” Instead, his reforms further strengthened the ability of Beijing to tap into local revenue. The localized, commodity-based model of state finance disintegrated early in the fifteenth century. The reforms that followed managed to restore stability and solvency to the realm. These reforms acknowledged the changes in Ming economy and utilized them. The replacement of grain taxes with silver and cloth accommodated local economic patterns and facilitated the transportation of supplies. Finally, the substantial amounts of silver that entered China in the sixteenth century catered to the growing demands of both market and state. Zhang Juzheng managed to accommodate state demands for a while through the traditional means of fiscal retrenchment and better auditing.  The following chapter returns to the policies that introduced money into state finance and contributed to the spread of silver in the Ming economy. It explains the role of tax commutation in state finance and follows its role from a means of negotiation between local and central demands, into a tool for strengthening the fiscal capacities of the centre.                                                       155 Ibid, 298-301.   102 3 The Role of Tax Commutation in State Finance   Tax commutation was the main source of money income for the Ming. Unlike the Tang and the Song before it, the Ming did not rely on indirect taxation, and so it did not enjoy the benefits of greater centralization and efficiency in tax collection. There was also the benefit of collecting tax in money. Money taxes are easier to assess, and easier to transport. They of course depend on the availability of money, and are aided by an active monetary policy. After the An Lushan rebellion the Tang dynasty expanded its monetary revenue sources and even attempted to institute a direct tax in cash. But, as Twitchett argues, there was not enough coin to institute a direct monetary tax and most of the levies were made in cloth.156 In the Song the industrial capacities of copper mining and coin production were matched by the scale and sophistication in which it taxed its commercial economy.   Ming China did neither. It did not have an adequate supply of money nor was it inclined to tax commercial transactions on a large scale. It could even be argued that before the late fifteenth century even the salt monopoly did not function as a commercial source of revenue. Salt was a vehicle for aiding grain transportation to the northern border, and a means of collecting excess notes and removing damaged notes form circulation. In neither capacity did the salt monopoly contribute directly to monetary                                                       156 Ibid, 42. The liang shui (兩稅) tax was made of a land tax assessed in grain, and a households tax assessed in silver.   103 revenue. In contrast, the salt monopoly, along with other commercial taxes, constituted over half of the annual income in the Tang and Song dynasties.   The benefit of commutation is that it does not necessarily involve money and therefore does not depend on a monetary economy. Before the An Lushan rebellion, tax commutation took place in order to save transportation costs. There is evidence that in Jiangnan the grain tax was paid in hemp cloth. In other regions grain was substituted for silk, salt, or cash.157 In the Tang, household production of cloth was widespread and built into the regular tax system.158 In the Ming too, cloth, probably cotton, often served as a substitute for grain and was available through local production. In these cases, commutation into cloth would not necessarily involve purchasing it on the market, though state demands did contribute to the expansion of a cloth industry. In more monetized economies money became the main substitute for grain and entailed some interaction with a market. In the Yuan, scrip was the official means of commutation, and required some kind of market interaction.159 But the reason for the commutation, as the History of the Yuan mentions, was to reduce transportation costs. These two examples                                                       157 Twitchett, Financial Administration, 30-31. 158 Francesca Bray, Technology and Gender: Fabrics of Power in Late Imperial China (Berkeley: University of California Press,1997): 205-215. 159 Yuan shi, 93:7.b.   104 show that in the logic of tax commutation there was no distinction between money and commodity. The importance was that it was light and of high value.160  In the early Ming the means of commutation were interchangeable. There is some indication that during the Hongwu reign there was a preference for scrip, but there was usually an option between several high-value commodities. After the demise of scrip silver became the most convenient means of commutation in its dual role of money-commodity, circulating in an economy that did not regulate currency. Unlike scrip, silver enabled the Ming to rely on a commodity that was high in value and easy to transport, and at the same time a stable and durable store of value. Since it was widely accepted as a means of exchange, it also reflected the growing reliance on an agricultural market economy.  In this chapter I argue that the Ming never fully relied on taxation in kind and was involved with a market economy from its inception. Furthermore, as the biggest economic player in the Ming economy, the state promoted a market economy through its tax policies on a nation-wide scale. This chapter analyzes the purpose of tax commutation, its scope in tax collection, and assesses its impact on the circulation of currencies and the production of commodities through the examination of scrip, silver, silk and cotton.                                                        160 The term that was used in this context is qing qi (輕齊). Schurmann follows Liang Yansheng in translating it as “light sending:. See: Herbert Franz Schurmann, Economic Structure of the Yuan Dynasty, (Cambridge, MA: Harvard University Press, 1956): 83, n.18. An alternative term is or qing huo (輕貨).  105 In the daily management of the empire, official grain tax quotas had little relation to actual receipts. The full quota was often not collected and the portion that was collected was not entirely in grain. The most common substitutes for grain in the first half of the dynasty were scrip and cloth. Tax commutation was initially applied as a manifestation of the principle of Revenue as a Measure for Expenditure, because its aim was not to increase revenue but to lower the costs involved in the collection and transportation of agricultural taxes. Money and money commodities were easier to carry; when roads or waterways were impassable for heavy carts or boats, tax could be still be delivered using lighter means of transportation. It was also a way to expedite the transportation of funds in case of an emergency. Commutation provided an alternative form of payment when grain was scarce. Also, though less reported, local officials also faced the problem of too much grain. After bumper harvests petitions were made to the court to substitute the grain tax for other items in order to avoid amassing quantities of grain in the government granary that would eventually rot.  During the Hongwu and Yongle reigns, paper currency was the main means for commutation of the grain tax. In the Xuande reign scrip was more strictly confined to commercial taxes. This was part of another role commutation played in the Ming, that of promoting the circulation of its scrip. Taxation was a way to collect scrip from the market and control the quantity in circulation. In complete contradiction to any fiscal reason, the dynasty collected taxes and paid salaries in scrip long after it devalued and largely disappeared from the market. Commercial taxes and the salt tax were collected in scrip until the last decade of the fifteenth century, when they were commuted to silver. This was also the time when scrip disappeared from official payments. Although the revenue  106 reports continued to register the data in ding units of paper currency, this was probably a mere convention by that time.  The entrance of silver into public finance was part of specific tax reforms that intended on the one hand to alleviate the tax burden on individual taxpayers and on the other to solve the logistical difficulties of supplying the military in the far north with grain from the southeast. The Hongwu emperor may have prohibited the use of silver in commercial transactions but never avoided using silver in public finance. Although there are no cases of widespread tax commutations, the history of his reign registers several large transportations of silver as aid to the military. That silver was fairly common in the mid fifteenth century is evident by the existence of silver payments for labour service. Instead of performing their regular service at the local yamen, men could pay the amount of silver needed to hire a day worker instead. This was an option for people who could afford it. Though commutation of labour service served as an extra source of revenue, it was not widespread before the sixteenth century.161  The organized and widespread commutation of the grain tax into silver was first announced in 1436, when the emperor agreed to collect the tribute grain tax in silver and ship it in ingots to Beijing. This policy, later known as Gold Floral Silver, signaled a shift in the practice and purpose of tax commutation and in the degree of monetization in state finance. It also signaled the withdrawal from the early Ming ban on silver. The A detailed                                                       161 Yukio Yamane, “Reforms in the Service Levy system in the Fifteenth and Sixteenth Centuries,” Christian Daniels and Linda Grove eds., State and Society in China: Japanese Perspectives on Ming-Qing Social and Economic History (Tokyo: University of Tokyo Press, 1984): 279-310.  107 discussion of early attitudes towards silver and the political maneuvers that led to the Gold Floral Silver policy will be discussed below. In the following section deals with the official policy regarding paper currency, and its role in state finance.   Cloth had a dual capacity in public expenditure both as clothing material and as a means of payment. In the beginning of the Ming, the Hongwu emperor ordered every landowner to plant mulberry, hemp, or cotton. From this a tax quota was set on each of these items. Households that did not plant any of the three had to pay in the finished product, measured in roles of silk, a bolt of hemp, or of cotton.162 This was another agricultural tax that was to be collected in kind. But when silk and cotton substituted for the grain tax, they also functioned as a currency. In north China cotton and silk cloth could be used for clothing and as currency for purchasing grain and horses. Cotton, in addition to being sent to the north to clothe the military and pad the winter uniforms of the soldiers, also served as a currency for purchasing grain for the troops. Higher valued silk was exchanged for horses on the border.   Since the revenue reports contain tax quotas and not actual receipts, the existence of commutation is difficult to discern. Nevertheless, the information does reflect the more significant changes in fiscal policy during the fifteenth century. Along with evidence from the court diary and administrative compilations, the importance of commutation in the fiscal management of the dynasty becomes clear.                                                         162 (Wanli) Ming Hui dian, 17:41a-41b.   108 3.1 Scrip  First issued in 1375, the Great Ming Circulating Treasure Note (大明通行寶鈔,Da Ming tong xing bao chao) was a direct continuation of its Yuan-dynasty predecessor. It was conceived as a high denomination currency, representing bronze coin that circulated along with it and that was used in smaller transactions. Issued in six denominations, the highest represented a string of one thousand coins, also called guan (貫). The other five notes represented between one hundred and five hundred coins. The official exchange rate of the one-guan note was initially set at one tael of silver and one shi of rice. As scrip depreciated a new unit of account, representing five one-guan notes, began to appear in tax income reports. The term ding (錠) was the Yuan silver unit of account. This suggests an attempt to attach scrip to silver, but sinc scrip was not officially convertible in the Ming, the term probably designates the alleged value of five guan of scrip. Thereafter scrip was accounted both in guan and in ding.163   In order to promote its scrip as the primary currency the Ming attempted to demonetize precious metals, especially silver that circulated widely in the Yuan. It prohibited its use in trade, and although the court held silver in reserve, it did not use it in order to back the value of its notes. Saying that, it should be noted that silver and gold were in fact convertible with scrip, but this in practice was a one-direction exchange,                                                       163 Since silver was never monetized in the Ming, ding pertained only to scrip while liang, the unit of weight of a standard tael, became the common designation for silver. The paper ding in the Ming continued to represent five-guan notes and depreciated at the same rate. Peng Xinwei 信威彭, Zhongguo huo bi shi 中國貨幣史 (Monetary History of China) (Shanghai: Ren min chu ban she, 1958), 469.  109 intended to collect precious metal from the market. Commoners could exchange their precious metals with the government but there is no indication that the government exchanged silver for scrip. But even if it did, the quantity of metal in reserve was so limited that it could hardly support a silver-based paper currency.164 Early Ming administrators may have picked this idea up from Yuan policies. In the 1280’s the government prohibited purchasing goods on the market with silver.165   Not only were silver and gold banned. Initially, Zhu Yuanzhang minted sovereign coin that circulated alongside coin from earlier dynasties and was intended to serve as small change alongside scrip. The decision to institute paper currency was possibly a result of insufficient output of copper and production of coin. Moreover, the Ming shi states that in the early years of the dynasty merchants were still accustomed to using Yuan-dynasty paper currency. 166  But as the value of coin against scrip rose, imperial mints were shut down. Copper was scarce in the early Ming and the closure of the mints further reduced the supply of coin. Hongwu’s policy regarding minting coin was extremely fickle. After establishing imperial mints in every province he started shutting them down. The mint in Fujian closed down upon imperial order in 1375, and the rest of                                                       164 Wu Han 吳含, “Ming chu she hui shen chang li de fa zhan 明初社会生产力的发展” [The Development of Productivity in Early Ming Society], Li shi yan jiu, no. 3 (1955): 53-84. 165 Nancy Shatzmen Steinhardt, “Currency Issues of Yuan China,” Bulletin of Song Yuan Studies 16 (1980): pp. 59-81. The author is not convinced about the degree of enforcement of this ban. One such prohibition from 1262 states: “ By imperial order, in all private markets were gold and silver are used to pay for cash and commodities, only scrip should be used instead” (敕私市金銀應支錢物止以鈔為准). See: Yuan shi, 5. 166 Song, Yuan shi, juan 81, 1962.   110 the provincial mints closed in the following year.167 In 1377 provincial mints reopened for production, and in 1393 they were shut down again out of concern for the burden this imposed on the people.168 Despite periods in which the state returned to minting coin, counterfeiting became so widespread that debased private coins completely replaced the official currency.169   Unlike the insistence on promoting its scrip, the Ming never made any serious effort to promote the circulation of its coin. This was in part a result of the conviction that in order to promote scrip the use of coin should be discouraged. Furthermore, due to the scarcity of copper, the state never managed to control its currency and offer an adequate replacement for counterfeit coin. That said, coin was collected in commercial taxes. Officially, it constituted one third of the tax quota. The specific arrangement in this case is not clear. Did merchants pay their allotted tax both in scrip and in coin, or were officials responsible to make sure that one third of their quota was is coin? In any case, the total income would be negligible. In the revenue reports commercial taxes are always designated in scrip so there is no way to ascertain how much of it was in coin. Furthermore, coin was never designated as a specific means for tax commutation (although it was an option during the first decades of the Ming), nor was it used in regular official payments. There was probably simply not enough of it. In other words, it was never perceived as an important part of state finance.                                                        167 Ming Taizu Shi lu, 101:2.a; 106:7.b.  168 Ibid, 112:4.a; 229:1.a.  169 Von Glahn, Fountain of Fortune, 97.   111   In comparison, the Song dynasty circulated scrip alongside currencies that acted as a measure of value and could back its fiat currency. Bronze coin served as the metal backing for scrip. Every two years the state would buy notes off the market using coin and issued new notes. It therefore managed the circulation of its scrip as a form of short-term debt that would be redeemed within designated time periods.  Between 1260 and the 1280’s, the Yuan enabled the free exchange of notes to silver and gold. Special government offices called Price Stabilization Depots were established for that purpose. But from the 1280’s silver and gold were prohibited in trade and scrip devalued continuously.170   Rather than following the example of the Song, the Ming repeated the Yuan ban on trade in silver. This probably only undermined the circulation of scrip because the scarcity of silver supported its high value. In addition, the inundation of the market with notes during the Hongwu reign was disastrous. In 1385 35,000,000 guan, were printed.171 In comparison, the quantity of paper money printed annually during the yuan did not often surpass 1,000,000 ding.172 In 1390, around 95,000,000 guan were poured into the economy in famine relief efforts and disbursements of emergency military funds.173 For example, in 1376, the emperor issued 500,000 ding of paper notes along with over 80,000                                                       170 Schurmann, Economic Structure, 135. 171 Yuzhi da gao xu bian 25a-26a.  172 Schurmann, Economic Structure, 141-143.  173 Huang, Taxation, 69-70.  112 bolts of silk to the capital guard and the standing army around Nanjing.174 Two years later, over 17,000 soldiers drilling in Henan were granted 45,000 ding.175  In response to reports of famine in 1389, the emperor sent an emergency sum of 5,360,000 ding to help more than 214,600 households from Qingzhou Prefecture in Shandong to purchase food.176 If the intention was that every household should get an equal share, then about 125 guan were distributed to each family. The value of a one-guan note in the late 1380’s was about 200 coins. In normal conditions this sum would purchase around 60 shi of rice and a larger amount of cheaper grains. In normal times this amount could last a family two years. But considering high grain prices during famine, it is likely that the seemingly generous aid was granted with consideration given to higher than normal grain prices.177 On the other hand, it is hard to imagine that such a sudden influx of money in the prefecture would assist in lowering prices, especially before sufficient grain arrived at the prefecture.   In these abrupt and large injections of currency into rural economies the first emperor demonstrated a profound misunderstanding of the nature of fiat currency. By sending large amounts of the currency to an agricultural population he ignored the fact at that time there was little one could do with money. By all accounts, Shandong’s                                                       174 Ming Taizu Shi lu,197:7b. 175 Ibid, 121:3a. 176 Ibid, 188:3a. 177 Dieter Kuhn, The Age of Confucian Rule, 215. This calculation is based on Kuhn’s assessment that one shi of grain could support a family of five for ten days.   113 agricultural economy was just recovering from the devastation of war, and it is unlikely that a viable commercial market existed in Qingzhou at the time. In other words, the quantities of notes that entered Qingzhou prefecture had no market in which to circulate.   Large issues of scrip were frequent especially during the reign of the first emperor, but continued throughout the Yongle reign too. Even if they did relieve distress in the short run, the influx of money pushed down the value of the notes. Excessive issuing with no external backing or parallel means of collecting notes back from the market could only led to inflation. Records of the exchange rate between paper currency and coin or silver indicate that within a decade of its inception, the one-guan note fell to under a half of its original value, and by the turn of the fifteenth century was exchanged for less than a hundred coins. During the Yongle era this trend continued and by the Hongxi reign (r. 1424-1425), a one-guan note was worth no more than a few coins. 178  The odd thing with the story of paper currency in the Ming is that it continued to circulate despite its being practically valueless. It continued to be disseminated in the economy through government expenditure and collected as tax until the final years of the fifteenth century. Moreover, in its first decades of circulation it was an acceptable                                                       178 Peng Xinwei created a chart of exchange rates between scrip and coin and silver based on various reports. Some of the information is deduced through reports of commodity prices. The usefulness of the information is limited since the value of silver and coin fluctuated against commodity prices too but the general trend is clear. From an initial exchange rate of one guan of scrip to a thousand coins and one tael of silver, its value dropped to 200 coins and one fifth of a tael in 1386. In 1407 an exchange rate of one guan per 12 coins was reported, but six years later it rose back to 47 coins. This is one of three cases in which the value of scrip was reported to have risen. The other two occurred in 1487 and 1525, but in these late periods scrip was no longer an actuality. As for the 1413 report, the increase is not significant but it may indicate that the currency gained some degree of confidence in it. Peng Xinwei, Zhongguo huo bi shi, Shanghai: Shanghai renmin chu ban she, 1958. P. 465.  114 replacement for the grain tax. There is evidence that commutation of the grain tax into scrip was widespread during the Hongwu reign. This practice continued during the Yongle reign though the scale of commutation is not clear. As we shall see, from the mid Xuande period this policy was reversed.      Judging by the legislative efforts of the first emperor, scrip appears to have been a regular substitute for grain. The following imperial statement is taken from the Da gao: Conversion of the Grain Tax All levies that officials are collecting in Zhexi are harming the people with a rapaciousness resembling that of wild animals. For example, in the case of the conversion of the autumn tax, the prefectural, sub-prefectural, and county officials expand the tax of two guan of paper notes for each shi of grain and devise all kinds of categories such as: A river travel fee of a hundred coins, a carriage travel fee of three hundred coins, a meal fee of a hundred coins, a granary inspection fee of one hundred coins, a reed basket fee of a hundred coins, a bamboo basket fee of a hundred coins, and a river-god fee of a hundred coins. How can these kinds of offenses against the people be pardoned?179   Zhexi is singled out here not only because of its economic importance at the time but because it was the centre of the Guo Huan affair.180 By singling it out, the emperor used                                                       179 Zhu Yuanzhang, Yu zhi da gao: chu bian. 180 Lin, Ming T’ai-tzu’s Administration of Justice, 51-52.  115 Zhexi as a model for financial administration throughout the realm. While the emperor’s outrage was directed at the tendency of local officials to make up all kinds of surcharges, he did not object to the fact that the autumn tax, the most important source of revenue, was regularly collected in scrip. So common was this practice that it was taken for granted and in itself merited no particular attention. Moreover, in this case scrip alone is mentioned as an acceptable means of payment. Therefore it may have been privileged over other means.    Another example for the acceptability of paying taxes in scrip appears in the following clause of the Great Ming Code: Rules of Paper Currency When people bring paper currencies to granaries, markets, treasuries, or tax offices to pay various taxes in lieu of paying in grain or to purchase salt or other goods, or when any yamen delivers illicit goods or fines, it is essential that personal names or marks be recorded on the backs of the currencies for later examination.181  In this regulation, substituting grain for paper currency comes across as an option that does not require any special authorization. We might even go further and assume that a clause such as this obliged officials to accept scrip when handed to them. What is                                                       181 The translation is from Jiang Yonglin, Mandate of Heaven and the Great Ming code, (Seattle: University of Washington Press): 91.  116 interesting about these two excerpts is that they are descriptive and not prescriptive. Even if it does not indicate the scope of the practice it does attest to its commonplace nature.  In other clauses of the Da gao, the emperor details some of the fraudulent practices relating to scrip and divulges what is probably the only surviving quantitative data on commutation of the grain tax to scrip: In detailing the crimes of the Guo Huan affair of 1385, he recounts that out of the Zhexi’s annual grain tax quota of 4,500,000 shi, the equivalent of 2,000,000 was collected in scrip. And out of that amount almost half disappeared before reaching the storehouse.182 The annual scrip quota printed during 1385 is revealed in another incident; close to 35,000,000 guan. The reason why we know this is that out of this quota, Guo Huan and his culprits stole 1,437,540 ding (or 7,187,700 guan) and laundered the loot by mixing it together in several storehouses with scrip collected from commercial and commuted taxes. When actual tax proceeds arrived at the storehouse, the granary officials involved in the scheme simply avoided registering it and divided the money among them.183 In another article we find the sum of collected taxes in scrip for that year: 27,545,295 guan.184 The state was undoubtedly printing large sums of scrip, but it was also collecting much. At least in 1385, the state collected in tax almost eighty percent of what it printed, meaning that it was using taxation in order to back its currency.                                                       182 Yu zhi da gao: chu bian, 10b. 183 Yu zhi da gao xu bian 25a-26a. 184 Ibid, article no. 32.  117 In other cases scrip was one of several options for tax payment. In 1376, one year after paper notes were first issued, the emperor announced a nationwide commutation of the grain tax. That year was a good year: the last campaigns against remnants of the Mongol army ended, the realm was pacified, and granaries were full. It was a time to thank the people for their toil during the years of conquest, transporting grain to the military. A large-scale tax exemption was granted, covering most provinces.185 In addition, rent from government land that was not exempted was to be collected in money. The emperor articulated that this was done for the convenience of the taxpayers, since by fixing the exchange rate of grain to money they would not be affected by changing prices.186 Again in 1386, the emperor ordered the Ministry of Revenue to calculate all surplus grain in the provinces to be collected in either silver, gold, cloth (either hemp or cotton), or scrip. This decree is probably the basis for the Wanli Hui dian account on the origins of the commutation of the tribute tax that shall be discussed below. But I am inclined to read this decree as a temporary one since it specifies “this year’s autumn grain tax along with grain from previous years.” 187  In these two cases scrip was only one option with which people or officials could pay and collect tax. But between it and the other options of coin, silver, gold, and cloth, it makes sense that scrip would be the cheapest means of payment.                                                       185 Ming Taizu Shi lu, 105:2b-3a. 186 Ibid, 105:4b-5a. 187 Ibid, 178:1.b.  118 Since government land was concentrated in Jiangnan, the commutation was really directed towards government tenants. And since tribute grain was the main source of income for the court, the government, and the army surrounding the capital city, we can assume that scrip was in turn used to purchase grain and other foodstuffs from the market. This policy must have entailed a decrease in income but it also entailed a relaxing of the tax burden that was especially high on government land in Jiangnan.188  Substituting grain for scrip continued during the Yongle reign. Specific decrees show that the practice had spread beyond Jiangnan. In 1407, Yongle specified scrip in an order to convert that year’s summer tax in Beijing as well as tax owed from previous years.189 A late Ming discussion of this decree maintains that in this case the reason for commuting tax to scrip was promoting its circulation.190 In the summer of 1416, the Grand Coordinator of Sichuan recommended collecting the previous year’s grain tax for Luzhou Prefecture in scrip and cotton due to famine.191 The opposite situation demanded official intervention too: In the same month Xi’an Prefecture requested collecting its summer tax in scrip stating that granaries were full.192 Similarly, in 1411 a request to collect the autumn tax in scrip was sent from Fujian Province, citing the same issue of a                                                       188 The issue of high tax rates in Jiangnan will be discussed in more detail in the following sections. 189 Ming Taizong Shi lu, 65:5b. 190 Chen Renxi 陳仁錫, Huang Ming shi fa lu 皇明世法錄 [A Record of the Law of the Great Ming] (1630), reprint: Si ku jin hui shu cong kan, ser. 2, vol. 4 (Beijing: Beijing chu ban she, 2000): 33:3b. 191 Ibid, 177:1a. 192 Ibid, 177: 7a.  119 bumper harvest.193 Too much grain in store would simply rot. It was therefore more logical to release the extra grain to the market by collecting the tax in money. Whether responding to deficiency or excess, officials were commuting the grain tax as a way to moderate the effects of seasonal imbalances on grain prices. Another purpose for collecting taxes in scrip was to maintain its value. Ming officials did not seem to grasp the importance of convertibility but they did understand the relationship between the quantity of notes in circulation and inflation. In 1404 Censor Chen Ying 陳瑛 , one of Yongle’s early recruits, suggested controlling the amount of scrip in the market with increased taxation:  Recently, the script system is not operating smoothly. In every case the reason is that the court is issuing too many notes and has no way to collect them, and as a result commodities are expensive and scrip is cheap.194   He submitted a plan to rescue the failing currency by imposing a new general salt tax that would be payable only in scrip. Chen calculated 10,000,000 civilian households plus 2,000,000 military households. By taxing them the state could collect more than 50,000,000 ding per quarter. He predicted that at this rate scrip would appreciate within a few months. The emperor convened a committee to discuss this proposition. It accepted the suggestion but decided to halve the proposed tax payment in order to spread the collection of scrip throughout a longer period. The emperor accepted the revised version                                                       193 Ming Taizong Shi lu, 108:3a. 194 Ibid, 33:8.a-8.b.   120 of the suggestion. This would mean that a family of five would be charged an annual sum of 42 guan. In the assumed exchange rate of the time the payment would be equivalent to about half a tael of silver. There is some indication that in the following years the value of paper currency rose a little but the general downward trend continued.  Minister of Revenue Xia Yuanji suggested another remedy to the Yongle emperor; issuing new notes bearing the Yongle reign name.195 This method was applied a few times during the Yuan. In 1287-8 Zhiyuan scrip was issued at a ration of 1:5 to the Zhongtong scrip that had been circulating since 1260. And Zhida scrip was issued in 1309-10 again at a ratio of 1:5 to Zhiyuan currency.196 The frequency of the new issues suggests that this solution was a temporary one. Interestingly, the Yongle emperor refused. For an emperor who spent vast resources in cementing the legitimacy of his reign and glorifying it, this refusal is curious. In fact, Yongle-era notes were issued after all. In the eighteenth-century numismatic work Quan bu tong zhi Yongle and Hongxi-era notes are recorded.197   Collecting many more millions of notes in the new salt tax perhaps prolonged some sort of circulation but it could not restore their value. In addition, there seems to have been real resistance to using scrip in the market place and the currency was confined to government operations. Nevertheless, there are historical examples in which taxation                                                       195 Zhang, Ming Shi, 81:1963. 196 Schurmann, Economic Structure, 135.  197 Meng Lin 孟麟 Quan bu tong zhi 泉布統誌 [Comprehensive record of money] (1833): 8.20a-21.b; 8.24a-25a.  121 was the sole backing for a successful circulation of fiat currency. In his discussion of 1690s Massachusetts, Dror Goldberg explored the case of paper currency issued by the colonial government. Forbidden by England to mint its own coin or issuing any other sovereign money, it resorted to printing notes that were otherwise inconvertible and passed them as private bills of exchange. These bills functioned in every way as money. Though unable to talk about it openly, the colonial government did allow and even encouraged tax payment in paper notes and becoming, according to Goldberg, the progenitor of the modern character of pure fiat currency.198  The relationship between taxation and circulation of paper currency is demonstrated in land-sale contracts preserved from Huizhou prefecture. As Li Ruoyu has shown, from the time Ming scrip was first issued until the Xuande period (with the exception of the short Jianwen reign), the majority of land sale contracts specify paper currency.199 These contracts had to be approved by a local official, a process that may have obliged the sellers and buyers to officially acknowledge scrip in the contract. Nevertheless, the fact that scrip disappears from contracts later on shows that the currency specified in them was not merely an empty word. Also, the fact that some                                                       198 Dror Goldberg, “The Massachusetts Paper Money of 1690,” The Journal of Economic History 69.4 (2009): 1092-1106. Based on this work, Goldberg proposes a theoretical model by which a tax-based mechanism of fiat currency can promote circulation regardless of value. Unlike convertibility, where the value of a paper note is guaranteed by a metal reserve, in the tax-based mechanism, a government can promote circulation without guaranteeing value. See: Goldberg, “The Tax-Foundation Theory of Fiat Money,” Economic Theory 50.2, (June, 2012): 489-497. 199 Li Ruoyu 李若愚, “Cong Ming dai de qi yue kan Ming dai de bi zhi” 从明代的契约看明代的币制 [The Ming Monetary System as Reflected in Ming Contracts], Zhongguo jing ji shi yan jiu 4 (1988): 39-43.  122 contracts that are dated to the same period do specify other means of payment such as cloth or silver indicates that people chose scrip over different options.  In one of these contracts the connection between tax payments and the sale of land is direct: When in 1411 Lady Huang (黃氏), the wife of Wu Xiding (吳希慶) from the eleventh township of Qimen county in Huizhou Prefecture sold a piece of land for scrip, she stated in the contract that it was “in order to raise money for payment of the grain tax.”200 Valerie Hansen finds that the need for cash was a common reason for the sale of land in Huizhou during the Yuan period. In this case, it was not only the need for cash but also the specific need for scrip that directed this particular seller. 201  From this passage we can deduce that paper currency was the common means of exchange in Huizhou, or make a more cautious conclusion that paper currency was still a requirement in transactions that involved the government, in this case drafting a certified land sale contract, and paying taxes. As explained below, there was indeed a regular commutation of the Huizhou tribute grain tax to Nanjing, at least from 1422.  Scrip disappears from Huizhou contracts abruptly after 1428. In that same year Xuanzong issued a new regulation for commutation of the grain tax (zhe shou shui liang                                                       200 Huizhou qian nian qi yue wen shu Song-Yuan-Ming bian 徽州千年契約文書-宋元明編 [One Thousand Years of Huizhou Contracts Song, Yuan, Ming Collection] (Yangzhou: Hua shan wen yi chu ban she, 1991): 67. 201 Valerie Hansen, Negotiating Daily Life in Traditional China: How Ordinary People Used Contracts 600-1400 (New Haven and London: Yale University Press, 1995). Valerie Hansen found that the need for cash was a common reason for land sales in Huizhou during the Yuan.  123 li 折收稅糧例).202 The declaration detailed several kinds of silk and cotton cloth and their exchange rates to different kinds of grains and beans. Scrip was completely excluded from the edict. Ever since the first order to convert the grain tax to money in 1376, this was the first time paper currency was excluded. Furthermore, In the summer of 1429, the Ministry of Revenue in Nanjing requested canceling the long standing permission, since 1422, in which Fujian and Guangdong Provinces, as well as Huizhou prefecture sent all their tribute tax to Nanjing in scrip and cloth. The reason given was that the imperial storehouses had an excess of both. In this case, the emperor refused to order the change as it might cause inconvenience to taxpayers.203 Still, the disappearance of scrip from the Huizhou land contracts supports the assumption that magistrates stopped accepting it. The initiative to eliminate scrip from the land tax continued in 1430 when Suzhou Prefecture petitioned to commute its overdue autumn grain tax payments into scrip. The petition was based on an imperial approval to do so made by the previous Hongxi emperor in 1425. Revenue officials quickly stepped in and presented the emperor with numbers: The commutation of 1425 pertained to a debt of 3,920,000 shi of grain that had accumulated since 1422. In addition, in 1429, as a relief measure the emperor granted permission to pay the grain tax in scrip and cloth, either cotton or silk. But as of the date of the memorial, those funds had not been fully paid. The sum amounted to 3,920,000 shi. To date, this sum had still not been fully paid. Any further allowances would                                                       202 Ming Xuanzong Shi lu, 41:5b. 203 Ibid, 55:3a-3b.  124 jeopardize state income. The emperor ordered to reconsider the request to commute the grain tax.204  Parallel to these developments, in 1429 the emperor announced that in 33 provincial capitals and commercial centres, commercial tax would be raised fivefold and paid in scrip.205 I interpret these two decrees as part of one policy intended to divert the collection of scrip from agricultural revenue to commercial revenue. The problem of excess notes was dealt with by printing less. In the summer of 1428 a halt was put on the printing of new notes.206 There is no indication as to when or if printing resumed. Also, the new tax policies dealt with the problem of damaged notes. Poor quality of paper and printing meant that they tore and faded within a short period of time. This caused a gap in value between newly issued notes and older ones. Revenue officials found that tax collectors would not accept bad notes and drove them out of circulation. From the point of view of the government, the problem was that it forced excess printing, and so it was vital to remove old notes from circulation. In earlier times government depots for replacement of damaged notes (for a small fee) were erected, but they do not seem to have been efficient.207 In order to get rid of old and damaged notes, a special tax on owners of vegetable patches and fuit groves demanded payment specifically in them. In                                                       204 Ibid, 74:1b-2a. 205 Ibid, 50:3a-3b. 206 Ibid, 44:9b. 207 These were called ‘cash offices’ (xing yong ku 行用庫). See Ming shi, 91.1962. The depots were apparently erected in every prefectural and county yamen in 1376 but were closed down, except for the capital depot in 1380.   125 addition, revenue officials were ordered to use old but still legible notes in the payment of awards and grants, while those that were too damaged were burnt.208  Taken together, these policies reveal that although the government was still committed to supporting its currency, there was an awareness that the state could not afford commuting its grain tax the way it had been doing up until then. After the Yongle reign and the short and unproductive Hongxi era, the Xuande emperor faced the task of restoring financial balance to the dynasty. The reality of Ming scrip policy was that it was subsidizing its circulation by incurring huge losses. Paper notes devalued so greatly that collecting them in tax payments was a form of tax remission. If the Yongle emperor could afford this kind of policy for a while, the Xuande emperor could not. In an attempt to separate the grain tax from scrip the Ming confined its paper currency to two main functions: The collection of commercial and monopoly taxes and official payments. As figure 3-1 shows, the expansion of commercial tax stations did lead to an increase in the tax collection but only for a short period of time.   Scrip was recorded in the revenue reports under several categories: “commuted to scrip” (zhe chao 折鈔) pertained to commutation of the grain tax. During the first two decades of the fifteenth-century, scrip was reported as commuted grain tax only four times but in large quantities, ranging between 3,000,000 and 9,000,000 ding. From the Xuande to the Chenghua reigns average commutation was under 80,000 guan. During the Chenghua period the average rose threefold, after which it disappears from the record. A                                                       208 Ming Xuanzong Shi lu, 55:10a.  126 second category was “miscellaneous commercial taxes” (za ke chao雜課鈔). This was the aggregate tax collected in the various tax and customs stations. The more significant avenue for the collection of tax in scrip was through commercial tax. Although it was supposed to be collected both in scrip and coin at a seventy-thirty ratio, only scrip is registered. During the Yongle reign, there is an annual average of 15,000,000 guan, or 3,000,000 ding being collected annually. In the Xuande reign the average rises to 30,000,000 guan and drops to around one-third during the Zhengtong reign. In the Chenghua reign the reports register a rise to an average of 27,000,000 guan. Averages of over 70,000,000 are appearing for the last two reigns, but these were already collected in silver.   The final two categories pertained to the sale of salt (yan chao鹽鈔) and tea (cha chao 茶鈔).209 Since the figures in these two categories are registered as income, they were probably not part of the salt-grain or tea-horse exchanges but either salt vouchers sold to wholesale merchants, or revenue from the salt tax. A category of tea vouchers appears from the Xuande reign and registered small revenue that turns into an insignificant quota of 189,474 guan During the Zhengtong-Jingtai reign. It disappears altogether during the Chenghua period and returns during the Hongzhi reign, though the data is confusing.210 During the Yongle reign there is no record of income from the sale of salt or salt vouchers. From the mid 1420s to the mid 1460s the average annual income                                                       209 See tables D-3,6,10,13,16. It should be said one third of the commercial taxes was collected in coin but since the revenue reports do not make this distinction I bring the entire sum in as they appear in the sources. 210 The category pertains to revenue from monopoly sales but units of tea are counted rather than cash income.   127 is close to 18,000,000 ding. There is no report of income from salt sales in the Chenghua period. Although the idea of a general salt tax was suggested to the Yongle emperor early in his reign, the revenue reports the category of the general salt tax (hu kou shi yan chao戶口食鹽鈔) only in 1488 that replaces the former yan chao. It stood at an annual average of 88,000,000 guan.211 Again, within a few years this revenue too was commuted into silver. Each guan of scrip was exchanged for 0.003 of a tael. And if indeed thirty percent of commercial taxes were collected in coin, the rate applied would be seven bronze coins for 0.01 of a tael. In this calculation the average income from the salt tax would be 424,000 taels. In the proposal suggesting this change the estimation was more modest, and closer to the figure we would get if income in coin was excluded, 223,000 taels. This sum, they suggested, could be used for military salaries.212   Two main trends are discernable throughout the fifteenth century. The first is the low figures of commuted grain tax after the Yongle reign, and the other is the enormous sums of notes collected through commercial taxes and sale of salt. Considering the exchange rates of the time it may be surmised that the main function of commercial taxes was to enforce the circulation of paper currency, since this income was not a significant part of state finance. Under the fiscal conditions of fifteenth century Ming China, removing scrip from the grain tax was a wise move. Concentrating it on commercial                                                       211 This tax was suggested to the Yongle emperor in 1404 but does not appear as a separate category in the reports until 1488. For the original decree see: Ming Taizong Shi lu, 33:5b. 212 Ming Xiaozong Shi lu, 74:9b-10a.  128 transactions made sense too, but the modest amounts attest to the failure to harness the commercial economy for the purpose of increasing revenue.    Figure 3-1 Total Annual Tax Income Collected in Scrip (ding) (Source: Ming Shi lu)010,000,00020,000,00030,000,00040,000,00050,000,00060,000,00070,000,00080,000,000140314081413141814231428143314381443144814531458146314681473147814831488149314981503150815131518 129 3.2 Silver  The first emperor’s attitude towards silver was a negative one. After conquering Shandong, his most trusted general, Xu Da (徐達, 1332-1385) traveled to Nanjing to report on the gains for the Ming army. In addition to captive soldiers, horses, grain, and salt, he also mentioned the existence of silver mines that could be exploited by the government. Hongwu replied: I know the evils of silver only too well. The profits for the government are few while the sufferings of the people are numerous. Especially these days when so many are weak and tired, how can such a heavy burden be imposed on them? Men of past had the saying “uproot tea and plant mulberry so the people can benefit”. How can you not know of it? [Xu Da] retreated in shame.213  Like tea, silver could bring profit but could not feed or clothe the people. In the emperor’s view, exploiting the land for a commodity that does not support the people’s livelihood was not the business of the state. Beyond this ethical dimension, there was also a social cost to mining operations that Ming administrators throughout most of the dynasty were not prepared to pay. The mining workforce was usually made of migrant workers who received little pay or nothing at all when the mining operation was unsuccessful. Resentment easily turned into banditry and in one case to a widespread rebellion.214                                                        213 Ming Taizu Shi lu, 31:2b. 214 Huang, Taxation and Governmental Finance, 242.  130 On the other hand, neither the first emperor nor his descendants refrained from using silver in order to expedite and supplement payments. Silver from the imperial storehouse was generally sent to the military for purchasing grain. During the Hongwu reign, significant amounts were remitted. In response to shortages, Taizu ordered the shipment of 100,000 taels of silver to Beiping (along with 100,000 bolts of cotton cloth) and 200,000 taels to Shanxi so that grain could be bought instead of being transported.215 At another time he approved the request of the Sichuan provincial administration to send 200,000 taels of silver in order to supplement official and military salaries in the coming year.216 These amounts were much beyond the meager silver quotas of the period, which were limited to several thousand taels. Where did the silver come from?  One option is that it was confiscated from surrendering rival forces. There were was also some commutation: In 1374 officials were allowed to collect tax in gold, silver, coin, and cloth in prefectures that had difficulties transporting grain due to blocked waterways.217 There was also the 1386 decree mentioned in the previous section, in which all surplus grain beyond locally retained funds were to be commuted into precious metal, cloth, or scrip and sent to the capital.218 Finally, in 1397 the emperor announced a general commutation plan for all unpaid taxes. Cloaked in a language of alleviating people’s suffering and simplifying transportation, the emperor initiated a policy for                                                       215 Ming Taizu Shi lu, 75:5b. 216 Ibid, 86:1a. 217 Ibid, 88:7a. 218 Ibid, 178:1b.  131 collecting debt. He specified exchange rates for scrip, raw and woven cotton, ramie cloth, silk and silver.219 He further stressed the regional character of commutation: “All accumulated debt throughout the realm will be commuted, according to local geographic preference to cloth, precious metals, and other such items.”220 During the Hongwu reign silver was never singled out as the only means of conversion so there is no way to estimate the amounts that reached the capital. Paying tax in silver continued to be an option even as scrip was actively promoted, although it is hard to believe that given the option to pay taxes in scrip, taxpayers would prefer to part with much more valuable silver and gold.  While officially continuing the ban on silver in trade, Yongle reversed his father’s policy on government mining. During his reign domestic production reached its dynasty-long peak. He may also have been more lenient towards silver in trade. When the guards of Jubao gate (聚寶門), one of the commercially busiest entrances to Nanjing, seized travelers that carried in their luggage a gold bracelet and several taels of silver, the Ministry of Punishments proposed sentencing them for breaking the law of commoners wearing gold and silver ornaments and for using silver in trade. But the emperor intervened saying that it is not against the law to carry such items in one’s luggage.221                                                       219 Ramie grows in south China. It is harvested three times a year. Each harvest yields finer fiber that produces finer quality cloth. This is what may have been levied here. See: Dieter Kuhn, “Textile Technology: Spinning and Reeling, ”Science and Civilization in China Vol. 5, Chemistry and Chemical Technology Part IX (Cambridge: Cambridge University Press, 1988): 30-38. 220 Ibid, 255:4a. 221 Ming Taizong shi lu, 115:2a-b.   132 The recording of such a minor incident may have served to show the emperor’s magnanimity, but it also hinted that Yongle condoned clandestine trade in silver. In his short reign, the Hongxi emperor took steps to return to the path of the founding emperor. This included moving the capital back to Nanjing, putting an end to the maritime voyages, and pulling out of the territories conquered in Vietnam. He also exhibited a greater commitment to remove silver from public use.  After the death of his father but before officially on the throne, he responded to natural disaster in the southern metropolitan prefectures by retrieving a decree from 1407 permitting grain transportation to the capital in cloth and scrip. On the other hand, he ordered the Ministry of Revenue to close the remaining paper money depots in Nanjing and Beijing. The reason was that these depots were established to “collect gold and silver from the towns people that led to market closure.” 222 This statement suggests that these deopts were selling scrip for gold and silver in addition to the role of replacing paper notes, a procedure that involved a small payment in coin.223 This fee was payed in coin and not in silver and gold. In any case, this statement acknowledges that collecting silver and gold from the market was disrupting trade and abolishing the scrip depots may have been the closest jesture to officially revoking the ban on precious metal in trade an emperor could make.  The Hongxi emperor did not live long enough to complete the first year of his reign and accomplish the transfer of the capital back to Nanjing. His successor, the Xuande emperor, decided to keep the capital in Beijing and deal with the financial                                                       222 Ming Renzong shi lu, 3, part 1:1a.  223 See p. 116.   133 consequences of that decision as well as with the financial state in which the Yongle emperor left China. The main problem was diminishing tax income from the financial core of the dynasty. As the richest rice-producing region in the empire, Jiangnan prefectures contained the highest concentration of government land. Rent on government land was four to ten times higher than tax on private land.224 And added transportation fees made the actual payment even higher. Historically, the high rate is interpreted as part of Zhu Yuanzhang’s revenge on the Jiangnan elite that supported his rival Zhang Shicheng (張士誠, 1321-1367) during the conquest of China. And indeed, Zhu had targeted large landowners. He confiscated their lands, which automatically became government land, and sent entire households to reclaim lands in frontier areas.225 Lin Jinshu has since challenged the ‘revenge’ narrative by arguing that in fact Hongwu-era tax rates were not higher than they were under the Song and Yuan dynasties, and that most of Jiangnan government land was actually inherited from previous eras.226  Wu Jihua argues that the real burden lay in the transport surcharge that added up to two thirds of the original quota to the tribute grain tax of Suzhou and Songjiang                                                       224 Lin Jinshu 林金树, “Shi lun Ming dai Su Song liang fu de zhong fu wen ti” 试论苏松二府的重赋问题 [Regarding the problem of Heavy Taxation in Su and Song Prefectures], Ming shi yan jiu lun cong, (Jiangsu ren min chu ban she, 1982): 91-123. 225 Richard von Glahn, “Towns and Temples: Urban Growth and Decline in the Yangzi Delta, 1100-1400,” The Song-Yuan-Ming Transition in Chinese History, edited by Paul Jakov Smith and Richard von Glahn,  (Cambridge, MA: Harvard University Press, 2003): 176-212. 226 Lin Jinshu 林金树, “Guan yu Ming dai Jiang nan guan tian de ji ge wen ti 关于明代江南官田的几个问题 [Some Issues Regarding Government Land in Ming-era Jiangnan], Zhongguo jing ji shi yan jiu, 1 (1988): 73-87.  134 Prefectures.227  The result was that tenants fled their fields and prefectures defaulted on their payments.  In 1430 the Xuande emperor remarked that state finances were in a mess, and the situation in Jiangnan is the worst.228 He then nominated a grand coordinator to tour Jiangnan’s prefectures and take charge of taxation. Surprisingly, he chose a relatively minor official in the Ministry of Works for the Job. According to his official biography, Zhou Chen 周忱 (1381-1453) entered official service in the beginning of the fifteenth century but never rose to a high position. All the while only Minister of Revenue Xia Yuanji recognized his abilities. Following his recommendation and that of Grand Secretary Yang Shiqi 楊士奇 (1365-1444), Zhou was elevated to Vice-Minister of Works and charged with reforming the tax system in Jiangnan.   Zhou recognized that it was the onerous tax burden that drove tenants away. He therefore initiated several tax reforms in order to address the greater inequalities in the system and accommodate tenants. Along with his main collabourator, the prefect of Suzhou Kuang Zhong, Zhou began by initiating a land survey and correcting falsities and inaccuracies in existing registers. Next, he lowered the tax quotas of each prefecture,                                                       227 Wu Jihua 吳緝華, “Lun Ming dai qian qi shui liang zhong xin zhi jian shui bei jing ji yin xiang” 論明代前期稅糧中心之減稅背景及影響 [The Background and Impact of Tax Reduction in the Financial Core of the Early Ming], Ming dai she hui jing ji Shi lun cong 明代社會經濟史論叢 [Studies on Socio-Economic Problems in the Ming Period], (Taibei: Taiwan xue sheng shu ju, 1970). See also: Mi Chu Wiens, “Changes in the Fiscal and Rural Control Systems in the Fourteenth and Fifteenth Centuries,” Ming Studies 3 (Fall, 1976): 53-69 228 Zhang, Ming shi, juan 153.  135 established a gradation in surcharges, and introduced a new commutation plan.229 Previously, transportation surcharges were calculated as a percentage of the particular tax obligation. This meant that the more tax one paid, the higher the surcharge. Zhou reversed the relationship between the original tax quota and its surcharge. For the individual household this meant that the more grain it owed, the less surcharge it paid. On a scale of five, the highest was a household that owed over five dou. It did not pay any surcharges. The lowest on the scale were those owing between one and two dou of grain. They were obliged to pay a surcharge of 1.5 dou. As an additional benefit, the three middle ranks could commute part of their surcharges into silver. This arrangement helped minimize the gap between government tenants and private landowners.  Commutation of the grain tax itself followed the same logic. Households owing six or seven dou for every mu of land could commute the entire payment into silver or cloth. Those owing four or five dou could commute half of their tax obligation. Finally, those who paid a low rate of one or two dou per mu, paid entirely in rice. The option to commute taxes into either silver or cloth catered to local economic patterns. Prefectures that had a cloth industry would pay in cloth while others could opt for silver. Zhou dealt with such nuances as well. In 1436 he petitioned the throne to commute Huizhou’s silk tribute tax to silver because silk production was low in the region and people had to purchase silk elsewhere in order to meet their obligations.230 Three years later he                                                       229 Yu Weiming (郁維明), Mingdai Zhou Chen dui jiangnan di qu jing ji she hui de gai ge 明代周忱對江南地區經濟社會的改革 (Zhou Chen’s Economic and Social Reform of the Jiangnan Region in the Ming), (Taibei: Taiwan shang wu yin shu guan, 1990): 48-60.  230 Ming Yingzong Shi lu, 30:6b.  136 petitioned on behalf of the prefecture again, this time concerning the portion of the summer tax paid in ramie cloth.231   Why should a household that had difficulties paying tax in grain find it easier to pay in silver or cloth? The answer lies in the government exchange rates. The official exchange rate Zhou proposed was one quarter of a silver tael for one shi of rice, while the market price of grain was normally almost double; 0.4 or 0.5 of a tael purchased one shi of rice. This means that paying grain taxes according to the official exchange rate entailed a reduction of fifty percent. On the other hand, grain prices fell right after the harvest so that taxes commuted during this period would not be subject to such a discount. But records of grain commutation rates indicate that they did not drop beneath 0.3 of a tael. 232 Another important point is that no quota was set for payments in silver or cloth, meaning that taxpayers could presumably choose to pay in grain if market prices favoured it. The aim of the commutation plan was to ensure tax supply to the capital by first accommodating taxpayers. It was essentially a tax reduction that was mitigated by the considerable reduction in transportation costs, silver and cloth being cheaper to transport than grain.  To be sure, the widespread availability of cloth and silver was a prerequisite for this plan. As shall be discussed in the following section, demand for cotton and its products in order to clothe the army encouraged the expansion of the cotton industry in southern China. And silver was already being used as a convenient replacement for                                                       231 Ibid, 65:10b. 232 0.4 of a tael in 1435, 0.3 in 1438, 0.5 in 1440, 0.5 in 1474. For a list of exchange rates see Appendix C.  137 labour service. Although a systematic expansion of the commutation of labour service started only in the last decade of the century, it set a precedent for a regular commutation of a tax obligation. The difference between labour service commutation and grain tax commutation was that while the first was used as a local source of finance, the latter became a regularized means of financing Beijing. Labour service demands depended on the discretion of local officials and were levied according to necessity. Since the central government did not intervene at first in this procedure, it is a good indication of the proliferation of silver in the economy outside of state intervention.   Perhaps the most important point of this policy was the exclusion of paper currency. Until the early 1430s southern provinces and Jiangnan prefectures still enjoyed the privilege of sending tribute grain tax to Nanjing in paper currency. Zhou’s reform put an end to that. This was therefore a logical conclusion to the initiative taken by the Ministry of Revenue a few years earlier. The goal of Zhou Chen’s plan was based on the traditional concept of equalizing the land tax. The option of silver or cloth payments was designed to alleviate the burden off those who had a heavier tax rate to begin with, tenants of government land. But this was also a first step towards harnessing the commercial economy of southeast China for the needs of Beijing. In 1435, the last year of the Xuande reign, the Ministry of Revenue requested commuting 80,000 shi of Huizhou’s grain tax for 32,000 taels of silver (or 0.4 taels per shi). The silver was sent to the Shanxi border garrisons of Xuanfu and Datong in order to purchase grain.233 It is the first case I found that silver was the sole means of exchange. Since the Xuande emperor                                                       233 Ming Yingzong Shi lu, 10:3a.  138 died in the beginning of his tenth year on the throne, it was actually his son who consented. As the Shi lu editors tell us, it took some convincing: After the heir apparent voiced his concerns regarding the ‘convenience’ of this plan (不便), the Minister of Revenue Hu Ying (胡濙, 1375-1463) explained the matter again and obtained his approval.234 The fact that more explanation was needed is not surprising; the heir apparent was a boy of eight at the time. The fact that this merited special mention may allude to the opinion of Shi lu editors about this procedure. As we shall see in the actual discussions regarding the policy later known as Gold Floral Silver, the emperor is portrayed as hesitant about commutation of the grain tax, and the editors voice clear opposition to it.   The option to commute the tribute tax was discussed twice in the fall and early winter of 1436. Vice-Censor Zhou Quan 周銓 raised the problem of salaries of officials serving in military guards in Beijing. As in the case of other Beijing officials, the rice portion of their salaries was stored in Nanjing. In the past messengers were sent to deliver the grain but apparently the grain was traded for other commodities and since the price of rice was cheap, it was sold at a loss and delivered only a fraction of the needed items. Zhou does not elabourate on who sold the grain, where, or whether or not this was done individually. But it is a rare acknowledgement of the way in which salary grain was managed. His proposal essentially followed that of Zhou Chen, to commute the tribute tax to silver or cloth according to local preference, and transport it to Beijing. But he also                                                       234 Ming Yingzong Shi lu, 21:6b-7a.  139 advised to first calculate expenditures and authorize the commutation according to it. It is important to notice that this proposal regarded only civilian officials who served in the military guards (ge wei guan yuan 各衛官員). This was not a large or important group. It may have been, therefore, a pilot suggestion in order to gauge the emperor’s response before suggesting broadening the practice.   Voicing local interests, The touring inspector of Jiangxi province, Zhao Xin(趙新, 1381-1461) mentioned the difficulties of boats travelling from the more remote and mountainous counties in Jiangxi to deliver grain to Nanjing. The current practice was to collect tax in precious metal or cloth and sell them for grain once the commodities reached the river ports along the way to Nanjing. But when grain prices were high it was hard to meet the grain quota. If Jiangxi would be permitted to submit its grain tax in silver and cloth then the transaction costs of purchasing grain would be saved and Beijing officials would receive a more appropriate payment. He therefore second Zhou Quan’s proposal to commute rice to silver or cloth, depending on the conditions of the locality. Not wanting to deviate from traditional practice, or at least not wanting to seem as doing so, the emperor asked what did his ancestors do. Minister of Revenue Hu Ying reiterated the Hongwu commutation rates that included paper money, cotton and silk cloth, and silver and gold. He explained that this measure was first applied to aid taxpayers in Shaanxi and later in Zhejiang. The sole purpose, he emphasized, was to accommodate taxpayers. The emperor accepted the proposal. In hindsight the editors of the Shi lu did  140 not miss the opportunity to comment on the significance of this decision and add that “from this time on granary stores diminished.”235  Two months later, Grand Secretary Yang Shiqi, also holding the title of Minister of War at the time, addressed the problem of Beijing military salaries. As with the salaries of the Beijing officials, the grain was stored in Nanjing and soldiers were sent south to deliver it. In this case the grain was not sold but transported. But as soldiers were unreliable, much of the rice was squandered along the way and only a fraction of it arrived at the capital. The emperor convened a committee that suggested sending a senior official to Nanjing in order to coordinate with the Nanjing revenue officials and other authorities the sale the grain accumulated in the southern capital’s granary and transporting commodities obtained from the sale to Beijing. Once in Beijing they would be resold according to their original purchasing price for grain. The other option was to follow Zhou Quan’s proposition with some modification. The proposition included silver alone, and added that grain would be commuted only during years of high grain prices. In plentiful years, taxpayers could return to submitting grain. The emperor determined that commutation of the tribute grain tax to silver will take place for one year only and added that it shall not become a permanent regulation (bu wei chang li 不為常例). 236    The decree was made by the new Zhengtong emperor who was nine years old at the time and therefore was undoubtedly acting upon the instructions of others. Upon his                                                       235 Ming Yingzong Shi lu, 23:7b. 236 Ibid, 21:6b.  141 accession, the boy-emperor was under the direction of the three Grand Secretaries, Yang Shiqi , Yang Rong 杨荣 (1371-1440), and Yang Bo 楊溥 (1372-1446), as well as three chief eunuchs among whom Wang Zhen 王振 (d.1449) was becoming most prominent. Above them the Empress Dowager presided over the decision making process.237 It is therefore vary plausible that the people behind promoting these policies were those who stood behind the emperor’s verdict. After the death of Xia Yuanji in 1430, they continued the policies of financial retrenchment and reform that began after the death of the Yongle emperor. Zhou Chen, who was acting under the patronage of Yang Shiqi, introduced silver into the grain tax as a regular procedure in what may have been a pilot program for the introduction of silver into state finance on a large scale. Even if these developments were not as orchestrated, the continuous promotion of silver in tax collection cannot be denied. And although the emperor insisted that it be temporary, the Ming never returned to taxing entirely in kind. In this light, the recurrent hesitation of the emperor to go through with these plans may reflect the opinions of the later editors of the history rather than the people surrounding him at the time.   Despite the importance of this shift, there is surprisingly little information about actual silver income in the palace treasury during the Zhengtong reign. Moreover, there is no evidence of a fixed quota before the sixteenth century. The annual revenue reports never reported on grain commuted into silver in the same way they did for cloth and                                                       237 Denis Twitchett and Tilemann Grimm, “The Cheng-t’ung, Ching-t’ai, and T’ien-shun reigns 1436-1464,” Frederick W. Mote and Denis Twitchett eds., The Cambridge History of China- The Ming Dynasty 1368-1644 part 1, (Cambridge: Cambridge University Press, 1988): 305-342.  142 scrip. Since silver entered the palace treasury it was not considered public revenue and therefore not included in the reports. The only data regarding silver in the revenue reports concerns receipts from mining. Even at its peak during the Yongle reign, quotas reach an average of only 280,000 taels. The average during the rest of the fifteenth century falls significantly, ranging from several thousands to several tens of thousands of taels. Ironically, while the state was moving away from the control of silver mining, its reliance on it increased.238 Or perhaps it was the other way around. As the state collected more silver as tax, it had no need to spend resources on managing silver mines. As figure 3-2 below shows, there is a sharp decline and several breaks in the reports on mining receipts from the beginning of the Zhengtong reign.                                                        238 Von Glahn, Fountain of Fortune, 114. Von Glahn argues that the initial silver-mining boom contributed to the acceptability of silver in tax payments.   143   Figure 3-2 Total Annual Income from Silver Mining (liang)  (Source: Ming Shi lu) 050,000100,000150,000200,000250,000300,000350,000400,000450,000140314071411141514191423142714311435143914431447145114551459146314671471147514791483148714911495149915031507151115151519 144  Since silver receipts are only an indication of the revenue extracted from government-operated mines, they do not reflect the prevalence of silver in the market. Rather, the reports reflect a change in fiscal administration. Starting from the year 1436, the revenue reports display two new categories.  The first is titled “river-transport of rice and beans to Beijing” (zao yun Beijing mi dou 遭運北京米豆), meaning the grain tribute tax. Before this date numbers fluctuated between 1,000,000-5,000,000 shi. From this year on, the average is between 4,000,000-4,500,000 shi.239 This fits with the Ming shi account stating that in the beginning of his reign, the Zhengtong emperor set a tribute quota of 4,500,000 shi.240 The second category was “grain transports from various locations” (ge chu yun na liang mi 各處運納糧米). The difference between the two categories is first in the verb used: Cao yun refers to water transport along the Grand Canal, while yun na means transport over land. This category can either refer to transports of grain throughout the realm, but probably meant additional grain transport to the capital. The Ming shi account also adds that the Chenghua emperor set a fixed quota of 4,000,000 in 1472/3.241 But the revenue reports for this reign period never reach that figure. Perhaps the reason for adjusting the quota was due to difficulties in transportation; earlier that year revenue officials reported that out of the regular quota 1,100,000 shi had not yet reached Beijing.242 Here, again, there is no mention of silver. It is therefore fairly                                                       239 In the Chenghua reign the annual average falls to 3,500,000 and is fixed on a round four million during the Hongzhi-Zhengde reigns. See appendix D.  240 Zhang, Ming shi, 79: 1917-1918. 241 Ibid,  242 Ming Xianzong Shi lu, 110:2a.  145 certain that between the Zhengtong and Chenghua reigns a quota of between four and four and a half million shi was established. Portions of this quota were probably commuted into silver but there is no clear information on quantities.  The earliest figures of silver income are probably from the late fifteenth century. In his work on silver income during the Ming, Wang Hui relies on an estimate by Wang Qi (王圻, jin shi 1565)  who cited 2,430,000 taels during the Zhengtong reign.243 This figure is probably taken from an earlier essay on state finance made by Wang Ao (王鏊, jin shi 1475) who served as minister of revenue during the Zhengde reign.244 It is more plausible that these figures pertain to the Hongzhi reign because the point of Wang Ao’s essay is to compare the period before and after the ascension of the Zhengde emperor to the throne. 245 Due to a scribal error, “pre-Zhengtong” is written instead of “pre-Zhengde.” Moreover, the figure he cites for total silver revenue includes commutation of the salt tax and other commercial taxes that began during the Hongzhi reign. Finally, he mentions that a total of 3,209,000 shi was commuted into 814,000 taels at a rate of 4 shi per tael. This figure is close to the sum of 819,811 taels reported in the capital at the end                                                       243 Wu Hui 吴慧, “Ming Qing (qian qi) cai zheng jie gou xing bian hua de ji liang fen xi (明清 (前期)财政结构性变化的计量分析)” Zhongguo she hui jing ji shi yan jiu no. 3 (1990): 39-56.  244 Wang Yuquan ed. Zhongguo jing ji tong shi: Ming dai jing ji juan 中国经济通史:明代经济卷 [Comprehensive Economic History of China: The Economy of the Ming Dynasty] (Beijing: Jing ji ri bao chu ban she, 2000): p. 268. Unlike the authors’ claim, the figures do not pertain to the Zhengtong period. The paragraph begins with the words Zhengtong yi qian which should be Zhengde yi qian. The author is comparing between the period before and after the Zhengde era in order to criticize the rise in imperial expenditure.  245 Wang Ao 王鏊, Zhen ze chang yu, Bai bu cong shu ji cheng v. 149 (Taibei: Yi wen, 1965): 1:32.b-33.b.   146 of 1487.246 On the other hand, Wang Ao’s breakdown of the sources of tax commutation reveals that the salt and commercial taxes were not the most important. Rather, it was a commutation of “grain transports from various locations” (ge chu shui liang zhe zheng, 各處稅糧折徵)by the sum of 1,030,000 taels. This corresponds to the category in the revenue reports and confirms that at least by the end of the fifteenth century those additional land transports registered in grain were actually silver transports. Up until the Chenghua reign the grain reported under this category remains around 1,000,000 shi. But from the Chenghua reign around 4,000,000 shi are reported annually. During the Hongzhi-Zhengde period the numbers are fixed in dubiously high quotas; 15,000,000 and 11,000,000 shi respectively. It is therefore most likely that Wang Ao referred to the Chenghua-Hongzhi period, though the initiative to add another source of grain or silver to the capital started in the beginning of the Zhengtong reign.   Official quotas for commuted grain tax are first registered for the year 1502, the fifteenth year of the Hongzhi emperor. The compilers of the Hui dian entered the results of the general land and tax survey as well as updated tax quotas, in what became one of the three major account works of the Ming, the first being the land and population survey of the Zhu si zhi zhang, and the third being the Wanli Kuai ji lu. The most striking difference is that in the Hongwu survey there are no tax quotas for tribute to the capital. It is possible that since the separation between government and private land was clearer during the Hongwu period there was no need for a tax quota since government land was essentially the domain of the imperial court. Instead of fixed tribute quotas from the                                                       246 Ming Xiaozong Shi lu, 285:6a.  147 provinces, there is only a vague notion of surplus (yu 餘).247 In contrast, in the Zhengde hui dian, every province and directly administered prefecture were assigned a fixed quota of commuted grain and fodder, either in cloth or silver, and quantities of taxes in kind. This change suggests that as the boundary between private and government land blurred so the imperial court had to designate its share of the tax income.   The Zhengde Hui dian specifies a total of 3,865,500 shi of rice and wheat that was to be commuted at the rate of 0.25 taels per shi. In addition, levies of hay were commuted at a rate of 0.03 of a tael per one pack of hay (shu 束 or bao 包). The total was 146,313 taels. Finally, scrip income from the household salt tax was transported. Under the rate of one guan per 0.003 taels the total income would amount to 52,969 taels. Altogether, 4,064,782 were expected to enter the palace treasury.248  This is therefore the first confirmation of a fixed quota a million taels directed at the palace storehouse in Beijing.   Silver was commuted from the taxes of the commercial cores and centres of international trade. Most of the silver came from provinces outside the directly administered prefectures. Silver commuted from 2,474,500 shi was collected from the affluent southern and coastal provinces: Jiangxi (1,030,000 shi), Zhejiang (730,000 shi), Fujian (364,000 shi), and Guandong (400,000 shi). These sums were between one third and one half of the total autumn tax assessment of these provinces.249 The remaining sum                                                       247 See pp. 84-85.  248 (Zhengde) Ming Hui dian, 24. Nanjing received as tribute only 67,950 taels. 249 (Wanli) Ming Hui dian, 24:17b-21a.  148 of 1,391,000 shi was collected from the centrally administered prefectures of which Suzhou and Changzhou shouldered by far the heaviest burden: 796,000 and 346,000 shi respectively.250  Quotas are of course one thing and actual receipts are another. If indeed the central government intended to have one million taels shipped to Beijing every year, it seems that during the fifteenth century it had some difficulty doing so. The following excerpt is a response made by the Ministry of Revenue to the military commissioner of Shanxi, refusing a request for more silver to support the soldiers in the strategic stronghold of Xuanfu in 1451: In a full year the commuted tax silver does not exceed 700,000 taels. This year the accumulated shipments of silver to Liaodong, Shaanxi and other border regions for grain purchases, awards, fodder and horse purchases, already surpassed 970,000 taels.251   Two years earlier the Zhengtong Emperor was captured in Xuanfu by Mongol forces. In 1451 the Ming was still in a state of emergency and the military expenses were unusually high. From this communication it is clear that while military needs were approaching one million taels, in the mid-fifteenth century tax commutation did not yet reach that level.    The year-end holdings of silver at court in the late fifteenth century were not large: in 1481 officials reported an income of 678,937 taels in the imperial storehouse, 532,414 taels in 1483, and 495,420 taels in 1486. Only in 1487 was the figure close to                                                       250 Ibid, 26:2b-12b. 251 Ming Yingzong Shi lu, 205:9.a-9b.   149 Wang’s account with a sum of 819,811 taels.252 These figures do not necessarily allude to the entire commuted quota for that year. Rather, they may pertain to the funds that were retained in the palace treasury after other disbursements were made. As Wang Ao explains, during the pre-Zhengde period only a small portion entered the imperial treasury; between 100,000- 200,000 taels were reserved for the imperial household, 336,500 taels paid for officer salaries, 336,500 taels were sent to the northern border, and between 200,000-450,000 taels were kept as emergency reserve. If the figures reported pertain only to imperial and emergency reserves then they were quite sufficient.    Another question that arises when trying to assess the quantities of commuted grain regards official exchange rates. Tang Wenji finds that exchange rates fluctuated during the second half of the fifteenth century. For example, grain tax in Suzhou and other directly administered prefectures was commuted at a rate of 0.6 taels per shi in 1481, between 0.7-1 taels in 1491, and between 0.7-0.8 in 1494.  Only in 1505 a rate of 0.3 is registered. But during the subsequent Zhengde reign the exchange rate rarely drops beneath 0.5 taels per shi.253  The reasons stated for the higher exchange rates are either disaster or military emergency, meaning high grain prices or difficulties in meeting the tax quota. This might explain why commercial taxes were finally commuted into silver during the last decade of the fifteenth century.     Concentration of large amounts of silver at court transformed financial administration and the balance of power at court. Officials supported commuting the                                                       252 Ming Xianzong Shi lu, (1481) 210:10b, (1483) 235:12b, (1486) 273:7a, (1487) 185:6a. 253 Tang, Ming dai fu yi, 188-191.   150 tribute tax in order to simplify the payments for officers and soldiers in Beijing. And the Zhengtong emperor even built a special storehouse for that purpose. But silver was also by law a commodity that entered the imperial treasury. So the more the court relied on silver imports to the imperial treasury, the more access the emperor had to the precious metal. In the following pages I propose that this operational change in state finance led to a conceptual change, which can be discerned from certain discrepancies between the two editions of the Ming Hui dian in their descriptions of the palace treasury. First, it should be said that there were not supposed to have been any discrepancies at all. Just as the role of the Zhengde Hui dian was to faithfully record the Hongwu era code and add later regulations to existing law, so was the Wanli edition mandated to add the imperial edicts that were announced since the early sixteenth century until the Wanli era. But as has been demonstrated in chapter 2, this was not the case. Instead, in certain aspects the Wanli edition deviates from the first edition. Considering that the Wanli Hui dian was compiled during the tenure of Zhang Juzheng who was promoting fiscal reform, observing the discrepancies between the two editions regarding Gold Floral Silver is illuminating.    In the Zhengde edition, the chapter dedicated to the imperial treasury (nei ku 內庫) does not offer an account on the commutation of grain into silver. Only in the shi li section do we find various imperial decrees regarding temporary commutations of grain into silver and cloth. It is rather in the description of the national silver treasury254 (tai                                                       254 National Silver Vault in Hucker’s dictionary.   151 cang yin ku 太倉銀庫) that the editors acknowledge commutation of tribute tax into silver. The text is as follows: In the seventh year of Zhengtong the Imperial Silver Storehouse was established. One secretary from this ministry (i.e. Revenue) was nominated to manage it. All fodder tributes in silver from the prefectures of the southern metropolitan region such as Su and Chang were transported to the Ministry and sent to the storehouse.255   This was a special silver storehouse that was built in 1442 to store the silver that was to be either sent to the northern armies or used for purchasing grain for them. It was managed by the Ministry of Revenue and was a separate store of silver from the imperial silver depository, the nei cheng yun ku 內承運庫. Therefore, although it was part of the storehouse network of the inner court, it was supposed to demarcate a boundary between private and public use of silver transports. The Wanli edition kept the above entry describing the national silver treasury but added an entry describing the nei ku in which Gold Floral Silver was introduced as the most important item. It is here that we find the complete narrative of Gold Floral Silver as it is known today: Within the holdings of the imperial treasury the largest is Gold Floral Silver. Since the beginning of the dynasty all commuted grain was transported to Nanjing in order to pay officer’s salaries and supply emergency funds to border armies. In the first year of Zhengtong, these funds were first transferred from Nanjing to the Imperial Treasury under an annual quota of one million taels. All the surplus funds beyond officer’s salaries were available for imperial use…256                                                        255 Zhengde Ming huidian, 33:9b. 256 Wanli Ming Hui dian, 30:1a.  152 It is highly unlikely that these are simple semantic differences, a result of negligent editing. After all, Gold Floral Silver was a vital for the livelihood of the capital and its military. Furthermore, the Hui dian compilations were such monumental projects that they served as an opportunity for the officials involved in compiling them to state the agenda of their generation. In relating to the commutation of the grain tax the editors of the Zhengde Hui dian designated it as a source of public revenue alone and omitted any mention of funds entering the imperial treasury. In addition, the description pertains only to fodder. There is clearly an attempt to minimize the significance of commutation of the grain tax, or at least disregard of it being a regular procedure.   On the other hand, the editors of the Wanli edition placed Gold Floral Silver as part of private imperial wealth. Moreover, the narrative gives a sense of continuity between the Hongwu and Wanli eras when in fact they were radically different. It is true that the first emperor ordered commuting extra grain into valuable items, but there was never any insistence on silver, and the quantities cannot be compared. The claim that the Zhengtong emperor simply moved the commuted funds that were regularly shipped to Nanjing north to Beijing is misleading too. This was clearly a completely new policy. Why did the authors need to describe continuity? I argue that this was the politically acceptable way to officially acknowledge administrative change. Perhaps this was also done in order to achieve legitimacy for a controversial practice. In the end, this account justifies the appropriation of most of the commuted silver by the emperor, a clear reversal from Hongwu’s intentions when authorizing the establishment of a palace treasury. A less conspiratorial interpretation would be that more than a century after commuted silver  153 began arriving in the capital, a different reality was forgotten.257 When compiling the new edition, the editors saw a need to add a description of the imperial treasury that was missing in the former. And looking back at the history of their dynasty, they produced a narrative that made most sense to them.  Surprisingly, the term Gold Floral Silver itself does not appear before the sixteenth century, and is absent from the Zhengde Hui dian. It first appears in the Shi lu in an entry dated to November 1517, when the Ministry of Revenue presented a memorial regarding difficulties in tax collection in the prefectures of the southern metropolitan region.258 The term hua yin referred to the quality of the silver.259 The term jin hua yin appears first in Jin dynasty (1115-1234). Alternatively, it was also the term used for silver vessels and plates inlaid with gold that were among the royal utensils of emperors and imperial kin. The Wen xian tong kao tells that from the Jin and Song dynasties the emperor’s sword was of gold floral silver.260 In the Ming, decorated silver vessels called hua yin, presumably for the purity of the silver, were given to princes as wedding gifts.261 The first emperor founded a silver workshop in the palace for the purpose of producing                                                       257 The only other author that noticed the late apprearance of the term jin hua yin and to connect it to a change in the storage of silver was Horii Kazuo堀井一雄, “ginkagin no senkai”金花銀の發展 (The evolution of jin hua yin) Tōyō ji kenkyū, 5.2 (1940): 128-140. 258 Ming Wuzong Shi lu, 155:2a. 259 Zhongguo qian bi da ci dian: Song Liao Xixia Jin  中國錢幣大辭典: 宋遼西夏金編 (Dictionary of Chinese Currency: The Song, Liao, Xixia and Jin Dynasties) (Beijing: Zhonghua shu ju, 2005): 204, 299. 260 Ma Duanlin 馬端臨, Wen xian tong kao 文獻通考 (c. 1300, reprint Puban collection, Guan wu zhai, Guangxu ren yin, 1902): 117. 261 Zhi si zhi zhang, 3:25b.   154 these kinds of vessels as well as household items for the imperial family.262 This workshop also produced silver ingots during the Yongle reign. Liu Xiaoping has amassed an impressive collection of Ming silver ingots with legible inscriptions. ingots dated to the Yongle reign bear the inscription of “Silver Workshop Floral Silver” (yin zuo ju hua yin, 銀作局花銀).263  Other inscriptions include “Imperial Treasury Gold Floral Silver” (nei cheng yun ku hua yin 內承運庫金花銀).264 Therefore, the sixteenth-century term referring to the tribute tax derives its meaning from the Yongle period imperially cast silver taels, and the ornaments of the imperial household. In both usages it is clear that gold floral silver belongs to the imperial treasury as distinct from the public storehouses. The connotation this term has with imperial prerogative further strengthens the impression that tax commutation that was perceived as public revenue in the fifteenth century, came to be seen as private imperial possession in the sixteenth century.   Between Zhou Chen’s tax reform and Gold Floral Silver of the sixteenth century we find a significant change in the role of tax commutation in state finance. Starting out as a means to reduce tax and transportation costs, it served as an option for tax payment, not an obligation. When Zhou proposed his new tax scheme, he intended to resolve the problem of tax arrears by accommodating local conditions while alleviating some of the costs of transportation. This was essentially the same policy devised by the Hongwu                                                       262 Ming Taizu Shi lu, 235:1a. 263 Li Xiaoping 李晓萍, Ming dai fu shui yin ding kao 明代赋税银锭考 (A study of the tax silver ingots of the Ming) (Beijing: Wen wu chu ban she, 2013): 220.  264 Ibid, 227-28.  155 emperor in 1397, only it concentrated on cloth and silver alone. When Zhou Quan and Yang Shiqi proposed utilizing this procedure in order to pay military salaries in Beijing, they were looking for a solution to insufficient funds in the north.  Both officials dealt with the problems of collecting tax in kind, and both did so along the Ming economic model that opposed raising tax revenue. In addition, if the plan suggested by Zhou Quan was indeed followed, it did not entail a fixed quota, so that commutation depended on grain prices. But this ideal did not last long and the state moved to a quota system that was in place by the early sixteenth century at the latest. The abundance of silver in the palace treasury made these funds available for the consumption of the imperial house and the eunuchs that ran it.  Officials serving at court during the Wanli reign such as Wang Qi referred to the Zhengde reign as the time when the state stopped living within its means.265   It could hardly be said that the Ming state ceased to live within its means only in the sixteenth century when it was never able to provide adequate compensation for its officials. The problem Wang was alluding to is deterioration in the management of the finances of the realm and as a result population figures have dropped and the dynasty lost half of its registered land. As discussed in Chapter 1, this was a mistake but it served officials who were critical of financial administration to stress their point. The general deterioration in financial administration was related to the reliance on fixed quotas, since once they were established there was no need to update financial information, only to make sure that officials meet their quotas. This change is apparent in the revenue reports.                                                       265  Wang Qi, Xu Wen xian tong kao, 36:13a.   156 Starting from the Chenghua period and culminating in the Zhengde era, tax categories that were previously updated annually turned into fixed quotas. First, a category of raw cotton was fixed during the Chenghua reign and spread to all sericulture taxes in the last thirty years of the reports. By the end of the Zhengde reign, the grain tax itself was simply copied from one year to the next.   If in the first half of the dynasty there was some degree of relationship between local conditions and official reports, from the late fifteenth century the central government was no longer concerned with local conditions, or no longer able to manage them beyond its demands for revenue. As fiscal demands grew, the system of tax commutation, originally employed as a means to negotiate between state-needs and local interests, became a way to further exploit local resources.266  3.3 Cotton and Silk  Cotton and silk differed from paper notes and silver ingots in their primary capacity as commodities. The first emperor imposed a sericulture tax and directed the population to set aside farmland for mulberry trees. Cotton was levied in large quantities too. Wu Han estimates that an average of one million bolts of cloth were granted to the army every year in the early Ming.267 In the early fifteenth century, reports on income                                                       266 Nimick, Local Administration, 139. 267 Wu Han吳含, “Ming chu she hui shen chang li de fa zhan 明初社会生产力的发展” [The Development of Productivity in Early Ming Society], Li shi yan jiu, no. 3 (1955): 53-84.  157 from cotton and silk products indicate that textiles constituted the highest non-grain source of state revenue. After a slump around mid-century, the quotas rise again and during the last decades of the dynasty rival silver quotas with a value of close to one million taels, with cotton products constituting the main portion of the total. In the revenue reports, silk floss (si 絲), cotton wadding (mian 綿), silk cloth or tabby (juan絹), cotton and hemp cloth (bu布), and unspun cotton (mian hua綿花) were listed in separate categories.268 Tabby and cotton cloth were a regular feature in grants and awards to soldiers and foreign emissaries. The finer silk products for palace consumption were produced in the imperial workshops in Nanjing and other prefectures of the southern metropolitan region.  Whereas Wu Han attributed the development of the Ming cotton industry to large-scale military allocations, Nishijima Sadao looked at tax policies in order to explain the expansion of the cotton industry. This brings us back to Zhou Chen’s Jiangnan tax reforms. An excerpt from the 1512 gazetteer of Songjiang Prefecture provides more detail on the regulations for commuting grain tax into cotton: One bolt of wide, white cotton cloth was to be collected for 2-2.5 shi of rice.269 The cloth had to be of a certain weight, width, and length. If it was too light, then the cloth could be longer or wider. Moreover, he suggested that red yarn would be woven in both ends so that no one could crop pieces                                                       268 In light of the expansion of the cotton industry and its predominance in military requisitions, I assume that cotton cloth was levied in greater quantities than hemp. Therefore bu is treated as referring to cotton in the calculations made below.  269 A bolt of cotton cloth (疋/匹, pi) was 12.4 meters long. “Minmatsu Shinsho seiji hyōronshū,” 明未清初政治評論集 [A Collection of State Craft Writing from the Late Ming and Early Qing] Chūgoku koten bungaku taikei v. 57 (Compendium of Classic Chinese Literature, Tōkyō: Heibonsha, 1967-1975), 12-13.  158 of it on the way. The author of the gazetteer concludes that this method continued to take place during his time, though the exchange rate fell to one shi of rice.270 In this case the different exchange rate probably reflects a change in grain prices. That was the principle of Zhou Chen’s plan. By controlling exchange rates the government could accommodate taxpayers when prices varied. For example, in 1436 the Ministry of Revenue ordered the lowering of the exchange rate of cotton and grain: Before, because cotton was cheap and rice was expensive (we have) memorialized the emperor to allow the commutation of grain from the autumn at a rate of two bolts of cloth per shi of grain. Lately, Guangxi and other provincial headquarters are reporting that the heavy exchange rate is causing losses to the people. A return to the old exchange rate of one bolt is recommended. The emperor agreed.271   The logic is clear, when the price of grain was high, one shi of grain sold on the market for two bolts of cotton or more. But when it was cheap, a tax of two bolts became onerous.272  In contrast to what one would expect, the revenue reports do not record any rise in revenue from cotton during this period. In fact, during the Yongle reign a much higher average was reported: on average close to 780,000 bolts of cloth, and almost 290,000 jin                                                       270 Nishijima Sadao, “Early Chinese Cotton Industry,” 29. 271 Ming Yingzong Shi lu, 14:5b. 272 Nishijima, Early Cotton Industry, 31. A one to one exchange rate is mentioned for the mid-fourteenth century also in a 1605 edition of the Jiading County Gazetteer. In this account the exchange rate was one bolt of cotton cloth for one shi of rice.  159 (斤, 377,000 pounds) of raw cotton. For the next forty years the annual average of cotton cloth dropped to around 160,000 bolts, and raw cotton dropped to 240,000 jin. Only in the 1460s did the average of cotton cloth rise again, this time to more than 800,000 bolts. During the Hongzhi reign a fixed quota of over 1,000,000 bolts suddenly appears, and over 1,500,000 during the Zhengde reign.  During the reform years of the mid 1430s it is actually silk income that rises in the revenue reports. Xu Dixin and Wu Chengming have estimated that silk produced in the imperial workshops amounted to 62 percent of the total requirements, though this seems to be the case only for the first third of the fifteenth century.273 According to their estimate, in the early Ming the annual quota for all state workshops was 56,805 bolts.274 This would put the total at 94,675 bolts. During the Yongle reign silk and cotton are grouped together under one category that stands at an average of 154,000 jin. From the Hongxi reign these items are recorded separately. In the early Xuande reign the reports record an annual average of around 95,000 bolts, confirming Xu and Wu’s estimate. But from the second half of the 1430’s the figure doubles to around 190,000 bolts. In the second half of the fifteenth century the annual average is about 280,000 bolts. The                                                       273 Xu Dixin 许涤新 and Wu Chengming 吴丞明, eds., Zhongguo zi ben zhu yi de meng ya 中国资本主义的萌芽 1522-1840 [The Sprouts of Capitalism in China] (Beijing: Ren min chu ban she, 1985), 143-149. Calculations regarding quotas rely on the Collected Statutes and memorials; Dagmar Schäfer and Dieter Kun, Weaving an Economic Pattern in Ming Times (1368-1644): The Production of Silk Weaves in the State-Owned Silk Workshops, (Heidelberg: Edition Forum, 2002). 274 One bolt (pi) was 12.4 meters long. Minmatsu shinsho seiji hyoronshu 明末清出政治評論集 [Collection of Critical Essays on Late-Ming and Early Qing Statecraft], series: Chugoku koten bungaku taikei 中國古典文學大系 (Tokyo, 1971): 12-13.  160 numbers drop below 200,000 during the last two decades of the fifteenth century and continue to drop to 126,000 bolts in the first two decades of the sixteenth century.   It is tempting to conclude that the Jiangnan tax reforms first had an impact on silk, and only later on cotton. As figures 3.3 and 3.4 suggest, income in silk cloth rose significantly starting at the period of the Jiangnan tax reforms, while cotton cloth rose sharply in a few stages from the beginning of the Chenghua reign. Another way to interpret the rise in silk products would be a rise in imperial demand. But while Xu and Wu date the significant rise to imperial requisitions in the Jiajing and Wanli reigns, the revenue reports mark an increase that is evident from as early as 1435. In addition, more silk was produced by private manufacturers and not in imperial workshops that shrank in their production capacity. The imperial requisitions of silk from independent manufactures increased, but when and where is not yet clear. We can conclude that tax commutation rather than imperial requisitions first contributed to the rise in income.  Finally, there is no certainty that silk and cotton were actually collected in kind. In 1443, Zhengtong permitted regions that did not grow silkworms to substitute their silk tax with silver.275  It is therefore very probable that part of what is registered as silk is actually silver. As for the increase in cotton revenue, the tribute list in the Zhengde Hui dian confirms that by the sixteenth century, cotton surpassed silk. Counting all provinces and metropolitan prefectures, a total of 661,455 bolts of cotton and 264,441 bolts of silk were sent to Beijing every year. Interestingly, labour service in the capital was entirely                                                       275 (Wanli) Ming Hui dian, 17:42b.   161 commuted into silk, not silver. The southern capital received 200,000 bolts of cotton and only 16,452 bolts of silk.276                                                        276 (Zhengde) Ming Hui dian, juan 24.  162  Figure 3-3 Total Annual Income in Silk (pi) (Source: Ming Shi lu)   Figure 3-4 Total Annual Income in Cotton (pi) (Source: Ming Shi lu)  050,000100,000150,000200,000250,000300,000350,0001426143014341438144214461450145414581462146614701474147814821486149014941498150215061510151415180200,000400,000600,000800,0001,000,0001,200,0001,400,0001,600,0001,800,0002,000,000140414091414141914241429143414391444144914541459146414691474147914841489149414991504150915141519 163 3.4 The Annual Revenue Reports and Ming Financial Administration  The annual revenue reports are a product of early Ming financial administration. Their detailed recording in the Shi lu indicates that accounting was of primary importance to fifteenth century administrators. They also imply that an infrastructure for conducting such accounting existed. They represent the ideal of storing wealth among the people with no distinction between imperial and public wealth, and give an impression of an economy based on the collection of various commodities. The data pertains to total tax income for a certain year, collected and uncollected. And, according to the maxim of measuring expenses against revenue, there is no information on expenditure. Perhaps the main purpose of the revenue reports was to relay an image of stability and prosperity. In this also lies their weakness. The financial crises and changes of the fifteenth century are not represented in the reports, nor is the growing role of tax commutation. Since tax revenue was recorded in official units of account, there was no way to register temporary replacements of one commodity with another.   Tax commutation provided flexibility to a system of revenue collection that was not amenable to frequent change and that would otherwise be at the mercy of harvest fluctuations and of the conditions of transportation routes. It served as means to negotiate between state needs and local conditions. Commutation of the grain tax into scrip during the Hongwu reign was a way to grant tax concessions without compromising income completely. The actual reduction in grain procurement that this practice entailed could no longer be sustained during the following Xuande reign, but tax commutation was not abandoned. Rather, it shifted scrip to the commercial sector of the economy and focused on the higher-value and more stable commodities, cloth and silver, in commuting the  164 grain tax. These kinds of commutations became more and more regularized until the grain tribute tax came under a fixed quota. Although the revenue reports do suggest such a development, silver is conspicuously absent. The only hint is that of a more or less fixed quota of tribute grain to Beijing.   Similarly, if we rely on Nishijima’s assessment, state income from cotton should have risen during the 1430s. Categories of commuted tax into cotton exist in the revenue reports but in very small quantities. But since commutation into cloth was an option for Jiangnan tenants, the actual quantities may be hidden in the fixed tribute tax category.  It should be added that officials did not record tax commutation on the local level either. In the Songjiang gazetteer that recounts Zhou Chen’s tax reform, cotton tax reports do not change significantly from the Hongwu to the Zhengde reigns.277 This suggests that any additional commutations were still registered under the total sum of the grain tax.  Still, changes in tax collection can be inferred indirectly by the spread of tax categories bearing fixed quotas in the revenue reports. Repetition of the same figure throughout an entire reign suggests either fixed quotas or lack of new data. Furthermore, when a particular category disappears from the reports it may be that fixed quotas did not change from one reign to the next. During the Yongle reign the income from copper, bronze coins, and cinnabar ceased to be updated and in the following reigns disappear altogether. In the same way, from the beginning of the Zhengtong reign a fixed quota was set for iron and this category was canceled too is in the beginning of the Chenghua                                                       277 Wei Chen 威陳, Songjiang fu zhi 松江府志 [Gazetteer of Songjiang Prefecture, 1512]: 7.  165 period. This of course does not mean that these commodities were no longer collected. But unless the quota changed from one reign to the next there was no reason to continue reporting it. It seems that only the editorial committee of each reign’s veritable records could decide on such a change. That is why we find numbers repeating themselves for an entire period and then either disappearing or being changed.     The fact that fixed quotas were reported for the most important sources of revenue indicates that the purpose of the revenue reports, supplying updated information on revenue to the political centre, was no longer achieved. During the Jiajing reign the demand for better account keeping was met by the emperor’s order to produce an account book that gives equal consideration to expenditure. This was an important step towards more effective financial administration but lacked sufficient infrastructure for gathering information. The growing concern over expenditure provided the greatest motivation to return to the Tang and Song-style national account books. As shall be discussed in the following chapter, during the fifteenth century the Ming dealt with difficulties in remunerating its servants in grain by commuting portions of salaries and stipends to other, readily available, commodities and money. As greater sectors of government were moving to regular payments in silver, the flexibility that the former system provided, and perhaps its ability to mask real shortages was gone.    166 4 Civilian and Military Salaries   This chapter examines policies regarding official remuneration before payments in silver. State policies in tax collection had a direct effect on expenditure. The insistence on keeping taxation low necessarily meant keeping salaries low. The decision to substitute grain for other means of payment instead of raising tax revenue suggests that from the early Ming the management of official salaries followed the principle of taking revenue as a measure for expenditure. Official salaries certainly diminished greatly as a result of commutation, and in some cases probably hardly covered basic needs. But the court did not merely stand by while this happened. In remunerating capital officials, the court constantly adjusted official salaries by changing the size and content of their commuted portions. These were temporary solutions applied within the framework of the Hongwu salary chart that succeeded in maintaining a stable and functioning administration.   A distinction is made between officials serving in and outside the capitals. Whereas the court experimented with other means of payment instead of scrip in salaries for capital officials, it made little effort to adjust those of local officials who continued to rely on scrip supplements to modest grain payments. On the other hand, the court maintained control over local salaries mainly in the centrally administered prefectures. In the provinces conditions varied. Furthermore, it is suggested that salaries of both central and provincial officials were not nominal. Officials relied on their rice portions to support their families and were aided by the government in transferring funds across China.   167  Finally, a comparison with the salaries of Song and Yuan officials reveals that salaries were partially paid in money too, and as inflation struck they diminished in a similar way. In this sense, the problems of the Ming was not new or unique, but a feature of pre-modern Chinese economies that employed a combination of commodities and monies in state finance.   The origin of the pitiful situation of official salaries in the Ming has been discussed by Gu Yuanwu in Ri zhi lu, a text that has become an authoritative account and interpretation of Ming institutions. His argument identifies two related faults in the Ming system: The first mistake was to terminate land grants and base salaries entirely on centrally administered grain allocations. The second mistake emanated from the first one; since tax revenue was insufficient to cover salaries, scrip was paid instead. As paper notes devalued, official salaries diminished. And even when scrip was replaced with cloth and silver, the substitutes never compensated for the original grain portions.278  This chapter first discusses what Gu Yuanwu saw as the initial problem, the canceling of land grants to officials, and places it in the intellectual and institutional context of the early Ming.                                                         278 Gu Yanwu 顧炎武, “Feng lu” 俸祿 [Salaries], Ri zhi lu ji shi [Collected Notes on the Record of Daily Accumulated Knowledge, Huang Rucheng 黃汝成 ed. (1795, reprint Beijing: Hua shan wen yi chu ban she, 1990): 548-553.  168 4.1 Government Land  The complete removal of specially designated land from official salaries is an innovation of the Ming.279 In previous dynasties, official salary charts included rent allocations according to rank. This income, called zhi tian (職田) or office land, utilized government fields for the purpose of funding public projects or supplementing officials’ income.280 Although the designation of sizes of land by rank may suggest that officials collected rent, their income was calculated according to fixed payments. For example, in the Tang, officials were entitled to income from between two and a half and twelve qing 頃 of land but payment was fixed at a certain rate of grain per mu 畝.281 Song officials received payments in coin. Those serving outside the capital were entitled to supplement payment from office land, which was probably paid in cash too.282 In the Yuan, salaries included a component of rent according to rank, but were paid partially in paper money.283 In contrast, the salaries of Ming officials were not allocated from specifically designated fields but were incorporated into the general budget. This arrangement made them even more dependent on centrally administered payments than their predecessors.                                                       279 There is some indication that Zhu Yuanzhang designated specific fields in Jiangnan for official salaries but retreated from it quickly. Ming shi, juan 82, 2001-2002. 280 Shorthand for zhi fen tian (職分田). Zhou Bodi 周伯棣, Zhongguo cai zheng shi (Shanghai: Shanghai ren min chu ban she, 1981): 364. 281 Hucker, Dictionary of Official Titles, vol. 1, 36. 282 Ibid, 52. 283 Huang Huixian 黄惠贤 and Chen Feng 阵锋, Zhongguo feng lu zhi du shi, 中国俸禄制度史 [A History of the salary system in China] (Wuhan: Wuhan da xue chu ban she, 1996): 357-363.  169  In his account, Gu chooses to praise the Tang system of government fields that allegedly produced upright and hardworking officials, which remained connected to agriculture rather than being consumed by the temptations of money and office.284 But beyond historical allusions to upright officials, he does not offer concrete reasons for the superiority of one system over another. This description seems more idealic than historical. Nevertheless, his opinion offers an interesting perspective on the imperial bureaucracy of the Ming as officials seeking office for money. This description points to a very different image from that of a gentry class that regards public office as a prerogative but in no way as a profession. In his work on the philosophical and sociological characteristics of the Confucian elite before and during the founding of the Ming dynasty, John Dardess proposes to understand it as a professional group.285 Their approach to office was one of public vocation that did not include the pursuit of profit, but it did entail remuneration for applying their professional knowledge. The early Ming Confucian scholar and examination tutor Liang Yin expressed it like this: ”…I have no fields to harvest and what I stockpile are simply books. My descendants who inherit this will have no grain taxes or other obligations, yet when their studies are finished they will have something to render to superiors and to contribute to the good order of the world. My concern for my                                                       284 He mentions Tao Qian (陶潛, 365-427) who is said to have left office and a comfortable salary for the uprightness of rural life, and Ruan Zhangzhi (阮長之, 379-437) who missed the day of the annual salary payment in order to rush to his mother’s funeral.  Gu Yuanwu, Ri zhi lu, 550. 285 John W. Dardess, Confucianism and Autocracy: Professional Elites in the Founding of the Ming Dynasty (Berkeley: University of California Press, 1983).  170 descendants is thus so keen that I have no thought to set up landed property for them.286  We can be sure that this statement was written during the early Ming because Lin acknowledges that he has no grain tax obligations. This was the case during the Ming but not under the Yuan, when Confucian households had some tax obligations. Although this is undoubtedly an ideological statement that does not reflect the landed background of most holders of examination degrees in the Ming bureaucracy, it may shed light on the ideological background of the setting of official salaries separate from any land component.  The centralization of official salaries can be also viewed as part of the centralizing project of the founding period. Government land was indeed the basis for official salaries but fields were not specifically designated for salary payments. Rather, all government land was managed collectively. Government land, or fields (guan tian 官田), was a general term for several kinds of sub-categories: Manor fields (Zhuang tian 莊田) supplied the needs of princely courts. School-fields (xue tian 學田) served as local sources of revenue directed at supplying meals for teachers and students as well as covering the costs of sacrifices for Confucius.287 Military fields (jun or dun tian軍,屯田)                                                       286 Liang Yin, 梁寅(1303-90), Shi men ji 石門集 (Printed ed. 1889), 1:37a-38b. Translated in Dardess, Confucianism and Autocracy, 52. 287 Ming Taizu Shi lu, 144:2a-2b.  171 supplied military units. Although the History of the Ming does mention fields for official salaries (bai guan zhi tian 百官職田) in this list, this category disappears subsequently.288  Without the component of land grants, Ming officials can be seen as the first salary men in China. In Max Weber’s discussion of the nature of the modern bureaucracy, he argues that in state systems in which officials received their salaries in kind, it was tempting to hand over the right to tax to them, which turned them into tax farmers or landlords. This weakened the bureaucratic structure and hierarchical relationships within the state. Weber assumed that this was the case in Imperial China:  The allocation of fixed income in kind from the magazines of the lord or from his current intake- which has been the rule in Egypt and China…easily means a first step towards appropriation of the sources of taxation by the official and their exploitation as private property.289   Weber assumed that throughout imperial China, officials were paid in kind from land rent and therefore did not see their remunerations as salaries. He used the term prebendum taken from the history of the Catholic Church in order to define the Chinese system. Prebendum was remuneration for priests serving in administrative functions who were entitled to a portion of the manor’s produce. Technically, Ming officials received their salaries from rent on government land as did their predecessors, but since it did not figure as a component in the salary chart any relationship to agricultural produce was                                                       288 Zhang, Ming shi, juan 77, 1881. 289 Max Weber, From Max Weber: Essays in Sociology, Translated and Edited by H. H. Gerth and C. Wright Mills, 205 (New York: Oxford University Press, 1958).  172 completely severed. Moreover, since their salaries were commuted to money, the translation of their remuneration as salaries (feng lu俸祿), in the modern sense of the word is justified.   4.2 Salary Charts  All government officials were entitled to a standard monthly salary. In addition to civilian posts, military and civilian degree holders served in military command posts. Beneath the high level posts such as regional military commanders, the most numerous were those of guard, battalion, and company commanders. When edicts refer to ‘civilian and military officials’ (wen wu guan 文武官) they are directed at the commanding level of the military who were all ranked officials. For the first time in Chinese history, clerks were entitled to a monthly salary too. They were placed in a parallel ranking system with its own salary chart. The highest ranked clerks earned a similar income to that of the lowest degree-holding officials and they shared government positions with them. Students in the imperial academy and successful candidates serving in their first post for a three-year probation period (ban shi guan 辦事官) received a monthly income too.  Logistically, central government offices paid their officials from the imperial granaries and storehouses in the capital, while localities paid their officials from the retained portion of the tax revenue. Official salaries were part of other regular expenses such as salaries for the local military guard, educational and welfare funds and two years  173 of military reserves.290 Although it was up to the county and prefecture to budget for its workers, the court had the final say on the contents of salaries. As we shall see, the greatest distinction in the working conditions of officials was between those serving in and outside the capitals. The terms commonly used for capital officials were jing guan (京官), or inner officials (nei guan內官). This last term could also denote eunuchs, but in this context is contrasted with outer officials (wai guan 外官), or provincial officials.  Salaries for officials and clerks were fixed according to the nine-rank system that began in the Northern Wei Dynasty (386-535). In their history of official salaries in China, Huang Huixian and Chen Feng describe a general shift from payment in kind to payment in money. Beyond the land grant component, the Northern Wei paid its officials in grain and cloth. The Sui and Tang dynasties paid salaries primarily in grain with an additional payment in coin that could be commuted to cloth. In the Song the cash component of the salary became more significant though a grain and cloth component remained. Ming salaries were calculated in grain but mostly paid in scrip and other money commodities.291  As the Duke of Wu, Zhu Yuanzheng began to experiment with remuneration for his nascent administration. In 1367 he arranged for generous grants and travel allowances for his prefects and magistrates in charge of the Jiangnan prefectures and counties that                                                       290 Zhu Yuanzhang (1396), Zhu si zhi zhang 2:1.a. Other scholars believe that official salaries were paid from the portion that was transferred outside the locality. See: Bian Junjie, Ming dai de cai zheng zhi du, 67. 291 Huang, Zhongguo feng lu,   174 were recently captured from Zhang Shicheng.292 Their families were included as well as their deputies and clerical staff. The size of the grants, paid in cloth and silver, was intended to ensure that officials would not burden the local population for covering their expenses, and to secure their loyalty. A prefect received four pieces of woven silk, six pieces of thick silk, two pieces of silk gauze, and six pieces of ramie cloth. The fathers of the heads of the prefecture, sub-prefecture, and county received the same amount of cloth as their sons, while their mothers, wives and eldest sons were entitled to half that amount. With the additional allowance granted to the family members, the total income of the household tripled. The households of the deputy officials were entitled to half the sum of their superiors. This pattern repeated itself with clerks (li mu吏目 and dian li典史). Their family members received half the sum of the deputy officials. Silver for travel fare was granted to prefects, sub-prefects and magistrates in installments of fifty, thirty-five, and thirty taels respectively. Their deputies were entitled to a budget of their own, between half and three quarters of their superiors’ pay. Clerks completed the list with a budget of ten taels of silver.293   It is unlikely that this arrangement continued for very long. And official salaries and budgets were never again to be this generous. Still, they do indicate that the future emperor understood the importance of adequate compensation for his officials. Once the dynasty was established, it issued the first salary chart in 1371. The chart is identical to                                                       292 Ming Taizu Shi lu, 24:4b-5a; Yu Jideng 余繼登(jin shi 1577), Dian gu ji wen 皇明典故紀聞, (Beijing: Zhong hua shu ju, 1981): 1:11. 293 Ibid, 24:5a.  175 that in that of the administrative code of the Tang Dynasty, which served as the main template for the construction of the new government. Salaries were calculated in standard units of grain, starting from an annual sum of 50 shi of grain for rank 9b up to 900 shi for rank 1a.294 In 1381 the emperor approved a general raise: Rank 1a received an additional 100 shi down to rank 9b, which recovered an additional 10 shi annually. The raise was not evenly distributed; officials that enjoyed the biggest increase in income were ranked 3a and 3b with an additional two hundred shi.295 Beyond this sum, all officials received a bonus in scrip; 400 guan for the first four ranks down to 30 guan for rank nine. With one guan comparable to around one shi of rice, this addition was significant.  The timing of the publication of this salary chart was not arbitrary. It was part of the general reorganization of government following the execution of Prime Minister Hu Weiyong. As Wu Han interprets, along with the prime minister most high-ranking officials were dismissed or executed. And in order to obtain the complete loyalty of the new recruits Zhu rewarded them accordingly.296 In the same decree the emperor also emphasized that a clearer system of rewards and punishment should be established.297 The new salary chart was to be engraved in stone, placed publically and its implementation enforced by the revenue officials in order to “prevent confusion.” In                                                       294 Ibid, 60:8b. 295 Ibid, 130:2a-2b. 296 Wu Han, “Hu Weiyong dang an kao” 胡惟永檔案考 [On the Case of the Hu Weiyong], Yanjing xue bao 15 (1934): 163-205. 297 Hu Weiyong allegedly abused the system of rewards and punishments in order to promote his followers and eliminate his enemies.  See: Farmer, early Ming legislation, 48.  176 centralizing power in his own hands the emperor tightened his control over his servants by reiterating and clarifying the procedures for rewards and penalties. But despite this, another salary chart was issued in 1387: While the salaries of the highest ranking officials rose to an annual sum of 1044 shi, the rest were reduced to a level close to that of the first salary chart.298 This may indicate that the need to literally buy the support of his officials was over and that the emperor felt secure enough to return to the previous salary plan. Alternatively, high salaries may have been abandoned simply because the expense was too high.  These salary charts make no distinction between officials in and outside the capital. The salaries of all officials were calculated according to rank, but once rice payments began to be replaced with money and commodities, the place in which an official served was more important than rank in determining actual compensation.   4.3 Commutation of Salaries into Scrip  Huang Fulu, who served as prefect of Shunqing Prefecture (順慶府) in Sichuan in the early sixteenth century, lamented the deterioration in officials’ income. In the beginning of the dynasty, he remarks, officials were paid entirely in rice and a person of rank seven would receive around seven shi of rice every month; enough for a decent living. But thereafter, part of that portion was commuted. In the case of officials serving outside the capital, any amount beyond two shi was commuted to scrip. Coming up to                                                       298 Ming Taizu Shi lu, 185:2a-2b.; 222:6a-6b.  177 Huang’s time, scrip had depreciated so badly that the actual amount of rice that could be purchased was nowhere near the original sum.299 Huang Fulu happened to live during a time when this reality had the gravest effect on an official’s income because these were the final years of the transition to silver in the Chinese economy. In the last decades of the fifteenth century and the first two decades of the sixteenth, the court continued to pay officials in scrip or cloth but as Huang remarks, rice was purchased with silver and both scrip and cloth could be converted to only a very small amount of silver on the market.   What is surprising in this account is that even at this late stage, salaries were still being paid with scrip and that they were still being exchanged on the market. This may not have been the case throughout China. Perhaps in Sichuan the complete disappearance may have been delayed, but this account nevertheless reminds us that paper scrip did not disappear from state finance right after silver became a common currency in government transactions. Initially, commuting salaries into money was a perfectly rational idea. Paper notes are much easier and cheaper to transport than rice. As long as officials could use the notes to purchase the supplies they required, their income remained stable and the state saved transportation costs. In addition, although tax quotas were fixed, actual agricultural yields fluctuated. The availability and cost of rice changed according to the time of year and the quality of the harvest. This too made payments in money a good idea, providing the court with more flexibility in handling supplies. It also provided officials with the means to purchase other items.                                                        299 Huang Fulu 黃甫錄 (1496, jin shi), Huang Ming ji lue: yi juan 皇明紀略一卷 [Historical Sketches of the Great Ming] (si ku quan shu cun mu cang shu zi bu vol. 240, Ji’nan: Qi Lu shu she, 1995): 71-87.  178  The first introduction of paper currency into official salaries occurred in the spring of 1376, right after the first issue of the official paper currency. The Ministry of Revenue reported that beginning from the fall of that year, half of the salaries for all officials and officers in the northern provinces of Beiping, Shaanxi, and Shanxi shall be paid in rice while the other half will be made of paper notes and wheat. In the other provinces there was to be a thirty percent commutation.300 Officials in the capital were left out of the list of provinces and were probably excluded from this arrangement. Although the reason for the announced change is not stated, the greater percentage of conversion for northern officials suggests a shortage in grain or difficulty to transport it. On the other hand, the inclusion of southern officials, where grain transport shouldn’t have been that difficult, raises the possibility that the Ministry of Revenue used its power to authorize and administer salaries in order to promote the new currency that was first created in the previous year. Without a central banking system, the state did not have many avenues for entering its scrip into circulation. Disseminating the paper notes in salaries was probably the only way to equally cover all regions of the empire.  Whatever the immediate reason for this decision, it created a precedent. Ten years later, in 1386, the emperor ordered the Ministry of Revenue to convert all official salaries to scrip. The inclusive term ‘All Under Heaven’ (tian xia) seems to address officialdom at its entirety, but in the Collected Statutes this declaration is interpreted as referring to officials serving outside the capital. And indeed in 1393, when regulations were finalized in Zhu si zhi zhang, only officials outside the capital were paid entirely in scrip while                                                       300 Ming Taizu Shi lu, 104:4a-4b.  179 those serving in it received their pay from the autumn grain tax, meaning rice301. If this was indeed the case then rice portions for local officials were entirely omitted.  Salaries for officials serving in the capital were partially commuted during the Yongle reign. Following three years of bloody war, Zhu Di won the throne from his nephew. The effect of this war on imperial granaries must have been significant since in 1402, even before his enthronement as the Yongle emperor, he announced that in order to economize, certain military personnel who up to that moment earned their entire income in grain would now be paid partly in scrip. They included men serving in the military regional commissions, in provincial surveillance commissions and in princely courts. He also reiterated the 1386 regulation, declaring that salaries of officials serving on the level of the prefecture, sub-prefecture and county will continue to be paid entirely in scrip.302 Three months later he approved a new salary program, proposed by minister of revenue Xia Yuanji, that cut into the rice portions of central government officials as well.303 According to this new scheme, portions of salaries were converted from grain to scrip in a progressive manner in which the highest-ranking officials were subject to the largest conversion of sixty percent of their salaries into scrip, while ranks 7-8 received only twenty percent in paper notes. Rank 9 officials as well as clerks and soldiers were paid entirely in rice.                                                       301 Zhu si zhi zhang, 2:18b-19a. 302 Ming Taizong Shi lu, juan 12, part 1:9b. 303 Ibid, 15:1b-2a. A more detailed version of this proposal appears in Huo Tao, “guan guan feng” 關官俸[concerning government salaries], Guo chao dian lue, 35:1b-2a.   180  The aim here seems to be reducing rice payments while ensuring that all officials and soldiers would receive a basic rice stipend that, in the case of the lowest earning officials and soldiers, was their entire salary. Moreover, acknowledging the devaluation of scrip, Xia proposed an exchange rate of five guan to one shi of rice, but the emperor doubled it to ten guan. Since prices of grain are usually rendered in silver it is hard to tell whether this exchange rate resembled market rates at the time. But considering that the exchange rate of scrip to silver was not more than one qian (錢, one tenth of a tael) and at its lowest rice cost three qian, officials would have not been able to obtain their full grain allocations on the market.304   In 1411 the emperor introduced further cuts into rice portions: Now officials ranked nine who previously escaped the salary cut were subject to a 20% conversion. Another adjustment in 1418 left officials in ranks 1-5 with half of their salaries in paper notes, and the rest were subject to a 40% commutation. By the end of the Yongle reign, Nanjing officials of all ranks, now serving in what became a regency (留守, liu shou) received only 20% of their salaries in rice. Clerks, who were initially spared any salary commutation due to their low income, were divided into two categories: Those who were supporting families were entitled to a monthly salary of six dou of rice (about 60 liters), while singles had to make do with four dou. The rest was paid in scrip. At the bottom of the hierarchy were student apprentices: They earned three dou a month for new officials with or without families. In this sense we can imagine them resembling apprentices in                                                       304 For some fifteenth-century exchange rates see appendix C.  181 lucrative jobs nowadays who practically have to pay to gain experience in their future careers.  The situation of grain supplies in the capital apparently did not improve. In 1421, the emperor decreed that only 5 dou of rice would be paid from the Beijing granary every month while the rest would be paid from the Nanjing granary.305 With no arrangement to get hold of the Nanjing funds it seems that this decision was more theoretical than practical, and indeed the decree adds that those who do not wish to be paid with Nanjing rice may convert this portion too into scrip. Unranked officials (雜職官, za zhi guan) with families had to make due with six dou while singles earned 4.5 dou of rice, with the rest in scrip. In this edict the grain component of official salaries was practically reduced to a basic grain ration, with only the money component reflecting differences in hierarchy.  The idiom “bending at the waist for five dou” (五斗折腰) was a common phrase meaning that five dou of rice was a meager wage not worth serving for. It supposedly originated in a declaration of an official in the Jin dynasty that he would not serve for a salary of five dou.306 Since this idiom was popular in the Ming it is likely that it resonated with the realities of official life. It also indicates that five dou was considered a certain                                                       305 (Zhengde) Ming Hui dian, 29:13a. 306 Jin, Yong金庸. “Tan bu wei wu dou mi zhe yao” 探《不為五豆米折腰》[On the saying Refusing to Serve for Five Dou] (1958) reprint Jin Yong San wen ji 金庸散文集 [A Collection of Prose Writings by Jin Yong] (Beijing: zui jia chu ban she, 2006): 311-313. In a short essay, the martial arts novelist has argued that the meaning of the declaration as quoted in the History of the Jin is actually the opposite: the official was not willing to sacrifice his principles for the material benefits of office.   182 low limit for service. It seems that at their most harsh, cuts in rice portions rarely went below this sum.  But not all capital officials were reduced to such meager wages. Interestingly, after all other groups of officials were gradually brought into this salary reduction plan, the last group to be addressed was Mongol military officers at court. This group was prominent in the Ming court from the beginning of the dynasty and throughout the fifteenth century. They constituted about one third of the imperial guard.307 But they were also significant within the ranks of the military in the capital. Ming emperors essentially continued the Yuan practice of an imperial bodyguard and entourage. It was a multiethnic group that included a significant number of Mongols and Jurchens. As David Robinson suggests, “the capital units under the Ming, like their counterparts during the Yuan, were more closely aligned with the emperor and his interests, than with the civil bureaucracy.”308 The function of the Mongol military dignitaries at court as confidants of the emperor was probably even more pronounced during the reign of the Yongle emperor who was struggling to gain legitimacy for his reign. Their rice portions were also much higher than their Han colleagues. Their salaries were subject to a commutation too but only up to 50%.   The glaring discrepancy was pointed out by one of the capital officials: In the very end of his first year on the throne, Li Wenda, a secretary in the Ministry of                                                       307 Henry Serruys, “Foreigners in the Metropolitan Police During the Fifteenth Century,” Oriens Extramus 8, no. 1 (1961): 59-83. 308 David M. Robinson, “The Ming Court and the Legacy of the Yuan Mongols,” Culture, Courtiers, and Competition: The Ming Court (1368-1644), (Cambridge, MA.: Harvard University Press, 2008): 365-422.  183 Personnel, presented to the Zhengtong emperor a scathing diatribe against the preferential treatment of Mongol officials at court. He complained that while all Han officials, regardless of rank, received only one shi of rice every month, Mongol officials continued to get at least half of their salaries in rice. Taking a deputy commander of rank 3a as an example, from a salary that was calculated at 35 shi a month, a Han official received one shi while his Mongol counterpart enjoyed a monthly stipend of 17.5 shi. In a terminology intended to rouse ethnic tension, Li accused the emperor of “taking the food out of the mouth of his precious subjects in order to feed wild beasts.”309  So it appears that there was more grain to be found in the capital for certain privileged officials after all. While commuting salaries was an economic necessity, it was not devoid of political calculations. This was especially the case with the Yongle emperor who did not follow the principle of revenue as a measure for expenditure in his political agenda. He may have stressed the need to economize in rice, but his entire reign was characterized by ambitious projects. Moving the capital to Beijing, military campaigns in the north and in Vietnam were extremely costly. The emperor himself admitted that the extravagant maritime voyages were eating up most of the tribute income from the provinces.310 During his reign, grain tax quotas were the highest but actual receipts fell                                                       309 Ming Yingzong Shi lu, 25:10a-10b. Although the memorial is recorded in the veritable records and the positive reaction of the emperor is indicated, Chen Ziliong, the editor of a seventeenth century collection of memorials, remarks that this memorial was barred from reaching the emperor by the three Yangs who controlled the court at the time. See: Li Wenda 李文達, “Da guan zhi feng” (Salary payments for Mongol Officials), Chen Zilong ed. Huang Ming Jing shi Wen bian, Si ku jin hui shu cong kan (Beijing: Beijing chu ban she, 2000): 36:16a-18b. 310 Li Jinming, Ming dai hai wai mao yi shi (Beijing: Zhongguo she hui ke xue yuan chu ban she, 1990): 46.  184 behind. In the last years of his reign his ministers cautioned him that state supplies were in serious danger.  Scrip also continued to depreciate. Xia Yuanji suggested issuing a new paper note with the Yongle reign title on it. This adjustment had taken place several times in order to curb the depreciation of Yuan scrip and had had some success. This was an apt recommendation in light of the turmoil of the civil war and Yongle’s usurpation. The effect of political uncertainty on Ming scrip is not usually cited as one of the causes for its depreciation. But fiat currency is worthless without the stability of the political entity that issues it. During the Jianwen reign and the civil war, there was no certainty that the Ming would survive. As mentioned earlier, in contradiction to the official historical narrative, there is evidence that the Yongle and Hongxi emperors issued notes with their reign titles. There is no indication that the Xuande emperor did the same. During his reign paper currency reached its bottom price.. The one-guan note that was officially exchangeable with 1,000 bronze coins in 1376, was worth only a few coins fifty years later. Surprisingly, the court did not discard scrip completely until the last decade of the fifteenth century, and in some cases even later. Instead, it explored the use of other means of payment.    185 4.4 Commutation into Sappan Wood, Pepper, and Cloth  The first commodities to substitute for scrip in the salaries of all officials and clerks serving in both capitals were sappan wood (蘇木, su mu) and pepper.311 The choice of these two commodities is a curious one, since they were not a traditional means of official payment as were other commodities such as coin, precious metals, or cloth. Instead, they were a product of the South-Asian trade that took place as a part of tribute relations during the Hongwu and Yongle reigns. After the annexation of a region in northern Vietnam, sappan wood was also levied as tax and is listed in the annual revenue reports.  Sappan wood was a product used for medicine and as a dye for fabrics. The kingdoms of Java, Siam and Melaka regularly sent it as tribute along with pepper and other valuables such as elephant tusks and tortoise shells. On several occasions very large quantities were submitted to the court. For example, 100,000 jin of sappan wood were presented to the Hongwu emperor by a delegate from the kingdom of Champa that captured twenty boats from a pirate.312 The availability of these commodities in the palace storehouses and their high value made them a convenient means of payment and they were often used in imperial grants to soldiers. In 1371 the emperor granted gauze, brocade, and Sappan wood to a Mongol force that defected to the Ming army. 313 In 1380,                                                       311 The Hui dian record a decree from 1424, but in 1425 the Minister of Revenue stated that this policy began in 1422. Ming Hui dian, 29:9a-23a, and Ming Ren zong Shi lu, 9:7a.  312 Ming Taizu Shi lu, 84:7a. 313 Ibid, 98:2a. The interesting point here is that the loot belonged to the Ming emperor and not to the Champa king  186 he granted scrip, Sappan wood and pepper to solders from the capital military guards who were wounded in battle or fell sick during their service.314 And towards the end of that year, the emperor sent Sappan wood instead of coin to his soldiers in the capital as allowance for the winter festival.315 This and other commodities were also used as payment for labourers: In 1391, Hongwu granted close to 14,000 water transport soldiers Sappan wood, pepper and bronze coins.316 In 1419, Yongle paid construction workers with scrip, cotton cloth, Sappan wood and pepper.317   Sappan wood and pepper continued to be used in special grants, but it is their appearance in salaries that is of interest to this discussion because it created a precedent by which salaries were doubly converted, first from grain to money, and then from money to a commodity, a practice that would continue for the next hundred years. On the one hand, the court substituted the failing currency with a more viable money commodity but the trick was that the exchange rate be tied to scrip and not to the original rice quota. This means salaries may have risen a little but did not come close to the original sum. In 1424, the court set an exchange rate of one jin of pepper to 16 guan of paper notes, and one jin of sappan wood to eight guan in remitting official salaries in Beijing. These two products are always stated together with the first holding twice the value of the second.318                                                       314 Ibid, 131:7a. 315 Ibid, 134:6b 316 Ibid, 207:3b 317 Ming Taizong Shi lu, 212:2a. 318 The court used the same method when converting salaries to cloth using silk and cotton, as if they constituted a high denomination and low denomination currency.   187 This was apparently a welcome change since in the following year, the students of the imperial academy (guo zi jian 國子監) requested to be included in the program.319 In 1434 it was the turn of the Nanjing officials to join in. By this time the exchange rate stood at 100 guan for one jin of pepper and 50 guan for one jin of Sappan wood.320 This means that the court did try to adjust the exchange rate according to that of the market, but it did not issue any official exchange rate of the two commodities to grain. What the government was doing was essentially granting one jin of either commodity as a supplement for salaries and then issuing the exchange rate with scrip. Rice was by now completely out of the equation.   The Hongxi and Xuande reigns were a time of financial retrenchment. During the one-year reign of the Hongxi emperor, official posts deemed unessential were canceled, and government expenditures were cut. Also, the expensive maritime expeditions were discontinued.321 After debating the issue for several years, the Xuande emperor dealt with the problem of tax arrears by lowering tax quotas. As discussed in the previous chapter, prefectures in Jiangnan were at the centre of a reorganization and reform of the tax system.  There were also effects on officials’ income: In 1428, all rice stipends for student apprentices were stopped. And grants of cloth replaced both the scrip and grain portions                                                       319 Ming Xuanzong Shi lu, 9:8a. 320 Ibid, 114:1b.  321 Hok-Lam Chan, “Chien-wen, Yung-lo, Hung-hsi, Hsuan-de reigns,” Cambridge History of China vol.7, 182-304.   188 of court officials. In 1431, the rice for the last two months for officials in Nanjing was replaced with cloth. This procedure apparently created a precedent since in the following year cloth was used instead of all rice portions of officials in both capitals.322 The person behind this decision was Hu Ying (胡濙, 1375-1463). Although he served as minister of Rites, he received a special appointment early in 1431 to deal with matters pertaining to the Ministry of Revenue. His memorials to the throne suggest that he was appointed to look into the matter of official and military salaries.   In a memorial presented to the emperor in 1433 Hu Ying offered to replace scrip with cloth in salary payments. His explanation also provides a glimpse of a common procedure in salary payments that will be discussed in chapter 4. Instead of paying officials in the capitals, receipts for payments were issued and could be cashed in their hometowns. But the foreign soldiers had to be paid on the spot. In addition, in several of the northern metropolitan prefectures there were places were the number of officials surpassed the available stock of scrip. As a result, many were paid out from the river-boat tax along the river. This led to delays in payments. Moreover, since the guard units in the mountain passes were all paid with scrip on the spot, the result was a concentration of scrip detrimental to circulation. 323    He suggested commuting the paper money payments for these soldiers into cloth. Seventy percent of the sum in silk at a rate of 400 guan per piece, and thirty percent in                                                       322 Ming Hui dian, 29:4a. In the Hui dian the order is to convert one shi of rice, which is the rice most officials received every month.  323 Ming Xuanzong Shi lu, 100:10b.  189 cotton, at a rate of 200 guan per bolt. Here again is a double conversion that was beneficial to the court since the amount of cloth that was exchanged for paper notes was much smaller than the amount that would have been exchanged directly for rice. In addition, Hu addressed the discrepancy in the exchange rates of officials and officers on the one hand, and foot soldiers on the other. While the first group received scrip payments at an exchange rate of 25 guan per shi, soldiers were paid at a much lower rate of five or ten guan per shi. He suggested dividing the expenditure budget more equally, setting an exchange rate of 15 guan per shi for all ranks. The editors of the shi lu remark that Hu’s plan was ultimately overturned and the salaries of low ranking officials and soldiers remained very low. 324    In the following year the court returned to Sappan wood and pepper in commutations of salaries and rations in Beijing.325 The general picture that arises from these proposals is one of total desperation at court, with officials going through the imperial storehouses to find items that could be used as payment. In any case, these kinds of manipulations in commodities and exchange rates are the means by which the dynasty maneuvered through what can only be described as a time of financial crisis. As discussed in the previous chapter, this was the year when Zhou Chen was sent to the south to put the core of state finance in order.   Alternating between scrip and Sappan wood/pepper became the preferred means to deal with either shortages or excesses. In 1436 the Ministry of Revenue reported that                                                       324 Ming Xuanzong shi lu, 100:10b. 325 Ming Hui dian, 29:14a.  190 the shipment of paper notes to pay soldiers in several northern military guards was insufficient and requested to supplement it with Sappan wood and pepper.326 In the same year the assistant to the manager of the money storehouse in the imperial treasury suggested to pay officials entirely in scrip because in the last few months the imperial storehouses were short on purchased commodities (presumably sappan wood and pepper).327 The Ministry of Revenue opposed, saying that excess payment in scrip will clog the market, or in the words of the memorial, make them stagnant (zhi滯). The compromise was to divide the payment between scrip and pepper/sappan wood. It seems that this proposal managed to create a satisfactory balance between prudent management of the paper money system and availability of other means of payment since it continued for the next four decades.  In 1471, the Ministry of Revenue reported that there was not enough Sappan wood and pepper in store to pay salaries and requested exchanging them for cloth.328 By 1483 stocks of Sappan wood and pepper were completely depleted.                                                        326 Ming Yingzong Shi lu, 12.5b. 327 Ibid, 19.3b. The money storehouse was called si yao ku (司鑰庫) and was manned by eunuchs. See Ming shi juan 74. 328 Ming Xianzong Shi lu, 97:5b.   191 4.5 Silver  The shift to silver in salary payments during the second half of the fifteenth century was gradual and subject to conflicting agendas. In 1437 military officials in the capital received their Nanjing rice allowances in silver.329 But this development did not spread quickly. Only in 1450 was this regulation extended to civilian officials in the capital but with reservations:  In the first year of Jingtai, it was ordered that all civilian officials in court receive their Nanjing salary portions according to the regulation for military officers in silver. Those who do not wish to do so, are allowed (to refuse). All commissioners in chief in the prefectures will also follow this rule.330   This decree was made by the new emperor and signaled a change in the attitude towards official salaries. In 1452, silver also replaced the scrip portions of court officials. 500 guan of scrip were replaced with one tael of silver.331 The silver that was used was taken from funds earmarked for purchasing fodder in the imperial storehouse.332 This commutation rate mirrored the actual value of scrip, but this did not necessarily work in favour of officials since they received scrip according to the artificial exchange rate with grain, 25 guan for one shi. This would mean that for the equivalent of twenty shi in paper                                                       329 Ming Hui dian, 29:14b. The decree specifically designates grain tax that was collected in silver to be used for payment. 330 Ibid, 29:15a-15b. A possible explanation for officials to prefer their salary grain in Nanjing will be discussed in the following section.  331 Ming Yingzong Shi lu, 218:3a. 332 Ibid, 15.b. Part of the tribute obligations of provinces was to send silver in lieu of hay.  192 notes, an official was entitled to one tael of silver.  In the following year the emperor ordered a fixed amount of silver to be paid to civilian and military officials at court as the converted portion of their salary, 124,312 taels for soldiers and 3,589 taels for civilians.333  But once the former Zhengtong emperor came back from captivity and managed to take back the throne he reversed these policies. First he canceled silver commutations of rice portions and ordered that payments should be once again made in rice. In 1468, in what was perhaps a final attempt to revive scrip, officials in Beijing received their converted portion, 30% in paper notes and 70% coin. Only in 1480 was scrip abandoned once again for cloth,