in scope and quality as a dissertation for the degree of Doctor of Philosophy. Jean Oi, Primary ... Approved for the Sta
STATE, MARKET, AND BUREAU-CONTRACTING IN REFORM CHINA
A DISSERTATION SUBMITTED TO THE DEPARTMENT OF POLITICAL SCIENCE AND THE COMMITTEE ON GRADUATE STUDIES OF STANFORD UNIVERSITY IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY
Yuen Yuen Ang November 2009
© 2010 by Yuen Yuen Ang. All Rights Reserved. Re-distributed by Stanford University under license with the author.
This work is licensed under a Creative Commons AttributionNoncommercial 3.0 United States License. http://creativecommons.org/licenses/by-nc/3.0/us/
This dissertation is online at: http://purl.stanford.edu/hr313dw9240
ii
I certify that I have read this dissertation and that, in my opinion, it is fully adequate in scope and quality as a dissertation for the degree of Doctor of Philosophy. Jean Oi, Primary Adviser
I certify that I have read this dissertation and that, in my opinion, it is fully adequate in scope and quality as a dissertation for the degree of Doctor of Philosophy. David Laitin
I certify that I have read this dissertation and that, in my opinion, it is fully adequate in scope and quality as a dissertation for the degree of Doctor of Philosophy. Beatriz Magaloni-Kerpel
I certify that I have read this dissertation and that, in my opinion, it is fully adequate in scope and quality as a dissertation for the degree of Doctor of Philosophy. Jonathan Rodden
I certify that I have read this dissertation and that, in my opinion, it is fully adequate in scope and quality as a dissertation for the degree of Doctor of Philosophy. Alberto Diaz-Cayeros
Approved for the Stanford University Committee on Graduate Studies. Patricia J. Gumport, Vice Provost Graduate Education
This signature page was generated electronically upon submission of this dissertation in electronic format. An original signed hard copy of the signature page is on file in University Archives.
iii
ABSTRACT
Why and how has China succeeded as a developmental state despite a seemingly rents-ridden bureaucracy? Following conventional wisdom, “Weberian” bureaucracies are an institutional precondition for development, especially in interventionist states like China. However, my research finds that China’s fast-growing economy has not been governed by a purely salaried civil service. Instead, Chinese bureaucracies still remain partially prebendal; at every level of government, each office systematically appropriates authority to generate income for itself. Such a bureaucratic form normally invites predation and hinders capitalism. Yet, in China, state bureaucracies have been collectively the backbone of economic growth and resilient authoritarian rule.
My study unravels the paradox of “developmentalism without Weberianness” by illuminating China’s unique path of bureaucratic adaptation in the reform era – labeled as bureaucontracting – where contracting takes place within the state bureaucracy. In a bureaucontracting structure, the state at each level contracts the tasks of governance to its own bureaucracies, assigning them revenue-making privileges and property rights over income earned in exchange for services rendered. Contrasting previous emphases on the prevalence of illicit corruption in China, my study shows how and why bureaucracies in this context are actually authorized by the state to profit from public office. My research draws on interviews with 165 cadres across different regions and governmental sectors, as well as new statistical evidence.
Specifically, I identify two factors that constrain arbitrary and excessively predatory behavior among Chinese bureaucracies. First, I argue that bureau-contracting represents a remarkably rigorous system of rents management. Agencies are not free to extract as they please; rather, they must win and comply with revenue-making policies awarded by the state. Additionally, I show that with new budgetary instruments in place, financial authorities have increased information and control over the use of bureaucratic funds. Second, narrow departmental interests are mediated by local developmental incentives that have remained even after the iv
1994 fiscal recentralization reform. Using regression analyses of previously unavailable budget data from Shandong province, my analysis demonstrates that individual local cadres benefit only in the short-term from extracting rents for their departments, but gain more in the long-term by promoting businesses and expanding the formal tax base. In short, the combination of an incentive-compatible fiscal design and increasingly sophisticated instruments of oversight have sustained an otherwise unorthodox structure of governance in China.
From a comparative institutional perspective, I see bureau-contracting as a species of hybrid organizations that fuse high-powered contracting with internal authority structures. To use the language of firms, bureau-contracting is to franchising as Weberian bureaucracies are to directly owned firms. The success of organizations like these lies in maximizing the benefits of entrepreneurial incentives and minimizing opportunism simultaneously. In the case of the Chinese state, bureau-contracting powerfully motivates cadres to self-finance a part (or even all) of their costs of office, thus reducing the government’s financial burden and allowing the ruling party to run an expansive state apparatus and patronage network. At the same time, by constraining rents extraction and preserving local growth incentives, the bureaucratic structure evolved co-existed with rapid market development under single party rule.
In a phrase, bureau-contracting presents a high-powered but opportunistic alternative to the Weberian ideal-type. The Chinese experience suggests that “market-compatible” bureaucratic institutions need not necessarily conform to – and may even diverge significantly – from standard Western models, at least at early stages of development. Through a micro-level view into the workings of bureaucracies in China, my observations challenge prevailing assumptions of bureaucratic organization and notions of corruption in the comparative literature. Relevant to policy-makers and students of political economy, my findings also inform our understanding of incentive design in developing and reforming governments.
v
ACKNOWLEDGMENTS
I remember feeling, during my first field trip to China in the summer of 2006, that I was standing at the bottom of a mountain, gazing up, and wondering how I could ever get to the top. As a foreign student, a parent of two children, completing this dissertation has not been easy and could not have been accomplished alone. It is a great privilege, especially on Thanksgiving, to acknowledge the kindness, generosity, and guidance of teachers, friends, and family, without which my journey would have been impossible.
My first debt of gratitude is owed to my advisor, Jean C. Oi. I had taken a leave of absence from the political science program, unsure if I would return and unsure about what to do with myself in general. Jean called, emailed, and urged me to return. I still recall her words, “Chiku (eat bitterness) a little and you’ll have lots of options later on.” It is thanks to her insistence and devotion as an advisor that I have the privilege today of having an academic career. Although this career can be “bitter,” sometimes, I cannot imagine another vocation that could be more rewarding and worthwhile. I thank Jean for prodding me onto this path.
As a mentor, Jean has inspired me with her ferocious tenacity to do research. She imparts by example a respect for the subjects whom we study, be they officials or farmers, through a genuine interest to listen to their stories. Methodologically, Jean has taught me the importance of uncovering complex day-to-day realities through painstaking fieldwork. Analytically, she has taught me not to judge the behavior we observe, but rather to ask why people act the way they do. Her influences are every where evident in this dissertation.
Many thanks are owed to the other members of my committee. Alberto Diaz-Cayeros has been an indispensable source of support ever since my return to graduate school. Alberto injected sparks into an embarrassingly confused project I presented in my third year, encouraging me to look into the political economy of public spending. His ideas inspired my data collection efforts in China and got me hooked on budgetary politics. Jonathan Rodden has listened patiently to my hours of “rambling.” Although he claims to know little about China, his responses have never vi
failed to make me see China in fresh light. I am much indebted to David Laitin’s insights and teaching. David literally identified for me the forest in the trees. I had been so lost at one point that David actually wrote me an abstract of what my dissertation should be – that abstract showed me the light. Last but not least, as a distinguished scholar and mother of three children, Beatriz Magaloni has been a role model to me. Beyond the dissertation, Beatriz has advised me throughout a stressful job-hunting process. I am grateful to Beatriz for her academic and personal advice.
My heartfelt appreciation also goes to the chair of my committee, Andrew Walder. Andy has been more than a committee chair. He has been very much a member of the committee. Over the years, he has offered incisive comments on my drafts and presentations that helped shape my project. Andy has a quiet yet powerful manner of giving encouragement. His support is cherished deeply.
Although not on my committee, Karen Jusko has been extremely generous in sharing her time and feedback. She has read my papers, attended presentations, and served on last-minute notice on my reading committee, each time offering detailed and valuable comments. I also thank Jonathan Bendor for his sharp insights and his intriguing course on organizational theories that got me interested in studying bureaucracies.
I wish to say a special word of thanks to my undergraduate advisor, Eve Grace. Learning political theory from Eve at Colorado College was a liberating experience. Eve’s classes planted in me the desire to go to graduate school. Although, to her disappointment, I did not in the end study political theory, I feel truly fortunate to have read some of Tocqueville’s and Rousseau’s philosophy. Having been raised in a rote-learning environment in Singapore, these classics of the Western tradition gave me ideas I did not have before. My interest in bureaucratization and state power, the basis of my dissertation, had stemmed from Tocqueville’s Democracy in America.
My project is based on substantial field research, which could never have been accomplished without the assistance of professors and friends in China. I thank Yang Yan, Bian Huimin, Qiao vii
Zhijian, Zhu Jinwei, Wu Yuanyi, Feng Shanshu for helping me to arrange interviews. Kuo TaiChun and Xiao Meng connected me to friends in Beijing during my first field trip. Nick Hope shared useful insights into China’s economic system and was always generous in making introductions. I have learned a great deal about China from conversations with Bian Huimin, Han Chaohua, Yuan Weishi. These scholars have survived tumultuous times in the country, and I deeply admire their courage and ideals.
My appreciation also goes to the officials and rank-to-file cadres who have shared their time and stories with me. To ensure their anonymity, I cannot name them individually. Some interviewees were extremely bright and insightful. Some have truly worked hard to serve public interests. I thank each of them for teaching me about China, institutions, and power through their real-life experiences.
Besides establishing contacts, the success of my fieldwork depended on having financial assistance. I have been fortunate to receive grants from the Center for East Asian Studies, Institute of International Studies, and the Graduate Research Opportunity Fund at Stanford University, as well as the 1990-Institute and Overseas Young Chinese Foundation. The Andrew Mellon Foundation/American Council of Learned Societies awarded two fellowships to support my dissertation write-up and post-doctoral research. At Stanford, the Escondido Family Fund provided four years of housing subsidy and Bing Nursery offered scholarships to my children. These had been a critical source of financial aid for my family.
On a personal note, there are many friends whom I must thank. I had returned to graduate school in the winter of 2004 alone, seven months pregnant with my second child. My husband was working in Singapore, and my older son, Justin, was cared for by my mother-in-law in Taipei. After my baby girl was born, I had to tend to my two children while juggling work in graduate school, until my husband joined us in 2005. Many friends came to my aid during that difficult time: Xuehua Zhang & Jon Otto, Doug Kerr & Christina Gwin, Vicki Sherman, Brian Goldsmith, Shawn Gaines, Mee Smuthkalin, Charlotte Lee, and our “Saturday volleyball friends” at Stanford, Jessica, Zhu-Zhu, Vincent, Xiaobing, Ju Bin, et al. These friends have taken me on grocery
viii
shopping trips, babysat while I studied for the field exams, sent my son to school, delivered medicine when I was sick, and so on. I am touched by their unconditional acts of kindness.
At Stanford, I have been lucky to enjoy the comradeship of colleagues, including Woramut Smuthkalin, Joo-Joun Jung, Chao-Chi Lin, Martin Dimitrov, Kay Shimizu, Charlotte Lee, Alex Kuo, Chris Chan, Xiaojun Li. My colleagues have provided much-appreciated company, veteran tips on fieldwork, and feedback on my project at every stage of dissertation research and writing. Yongshun Cai has played an influential role in encouraging me to publish my field paper and eventually to return to graduate school. I thank Xuehua Zhang for being a great friend and an informal mentor. She introduced me to the nuts-and-bolts of doing fieldwork, made contacts for me, and lent a listening ear to my complaints. At the political science department, I wish to thank Jeanette Lee Oderman for her compassion and especially her help during my transition back to Stanford. Eliana Vasquez and Chandelle Arambula also deserve much credit for keeping scatter-brain students like I on the alert.
Beyond Stanford, family friends and teachers have contributed crucial moral and practical support. My family and I are deeply grateful to Reverend Daxing for his profound wisdom, teachings, and medical care. Mr. Westgate, my husband’s high-school teacher, has been a teacher and a family elder to us. His humor and honesty are always much appreciated. Heartfelt thanks goes to Hsien L. Chen for his hospitality and generosity. Fiona Ng, who has been like a big sister, has helped me numerously during my stay in Hong Kong. Yang Yan has been incredibly kind to me. Shawn Gaines lifted my spirits whenever I was down. Although we had lost touch twice, he magically showed up again when I needed his support and words of inspiration.
Finally, I give thanks to my family. My parents, both of whom had only attended high school, worked hard to give me an education. My parents-in-law have provided financial support during our years in school and showered loving care upon our children. My two children, Qi-Xuan (Justin) and Qi-Mei (Jamie), fill my life with joy. I thank them for being so gracious and sweet in bearing with my constant absence, long work hours, and bad moods during dissertation writing.
ix
My husband, Chia-Yu Tang, gave up his career so that I may pursue mine. He left his job in Singapore and became a full-time dad in Palo Alto, enabling me to work on this dissertation. He has been an amazing father to our children, especially when I am away from home for weeks, even months. It is little known to others that he is also an amazing coach with a dual background in arts and business; he debates hypotheses with me, designs my web site, and edits my powerpoint slides. Above all, Chia is the beacon of my life. His qualities of kindness, patience, and quiet courage are priceless.
As this dissertation is built upon my husband’s sacrifices and love for the family, I dedicate my work to him.
November 26, 2009 Shanghai, China
x
TABLE OF CONTENTS
Acknowledgements List of Tables List of Figures
CHAPTER 1 –
1
Introduction CHAPTER 2 –
27
Bureau-Contracting: The Industrial Organization of Chinese Bureaucracies CHAPTER 3 –
65
The Structure of Cadre Rewards CHAPTER 4 –
79
Dual Fiscal Incentives CHAPTER 5 –
104
The Political Economy of Bureaucratic Adaptation CHAPTER 6 –
125
Conclusion
Appendix A: List of Local Government Offices Appendix B: The Formal Cadre Wage Scale Appendix C: List of Interviews
xi
LIST OF FIGURES
Figure 1.1: Bureau-Contracting and Weberian Bureaucracy Compared
22
Figure 2.1: Bureau-Contracting: Fusing Markets and Hierarchies
31
Figure 2.2: Varieties of State Organization
32
Figure 2.3: Organization Chart of Typical Local Government
35
Figure 2.4: ‘Rules of the Game’ – A Simulation of Budgeting Procedures and Outcomes
63
Figure 4.1: Dual Fiscal Incentives
83
Figure 4.2: Decomposition of Public Spending
88
Figure 4.3: County Spending Structure
88
Figure 4.4: County Revenue Structure
88
Figure 5.1: Local Revenue and Spending Before and After the 1994 Fiscal Reform
113
Figure 5.2: Change in Composition of Extrabudgetary Revenue, 1980-2003
114
Figure 5.3: Change in Public Employment Size, 1955-2003
116
Figure 5.4: ‘Eating budget’ – Cadre Rewards Consume Tax Revenue and Fiscal
121
Transfers
LIST OF TABLES
Table 2.1: Examples of Functions and Fiscal Status of Extrabureaucracies
38
Table 3.1: Composition of Local Cadre Rewards, 2004
66
Table 3.2: Distribution of Cadre Rewards Across Provinces, 2004
71
Table 3.3: Personnel and Administrative Spending Per Cadre in Windy County in 2007
71
Table 4.1: Tax Categories After the 1994 Fiscal Reform (Tax-Sharing System)
89
Table 4.2: Descriptive Statistics of Cadre Rewards, 2001-2005
90
Table 4.3: Descriptive Statistics of Income Streams, 2001-2005
91
Table 4.4: Determinants of Cadre Rewards in Shandong Counties (2001-2005)
94
Table 4.5: Simulation of Short-Term and Long-Term Effects on Benefits Per Cadre
97
xii
CHAPTER 1
INTRODUCTION
“Capitalism and bureaucracy have found one another and belong intimately together.” Max Weber, Economy & Society, p. 1465 “Can we identify the kind of transitional administrative system that exist today in reality? I think we can, but to do so we shall have to employ some new words and concepts which cannot be found in the standard literature on public administration.” Fred Riggs, Administration in Developing Countries, p. 10
Over the past centuries, state institutions have launched on an inexorable process of natural selection, converging upon a common organizational form – the bureaucracy. As Max Weber famously observed, the bureaucracy is a new organizational species. Contrasting pre-modern institutions of governance, modern bureaucracies are legal-rational. They are rules-bound, hierarchical, meritocratic, politically neutral, and the most basic of all, salaried. Civil servants receive regular wages from state budget allocations and are barred from exploiting public office for personal profit. By virtue of its discipline, efficiency, and reliability, the legal–rational bureaucracy is superior to early patrimonial administrations, featuring fiscal arbitrariness, personal loyalties, and compensation of officials in kind rather than in money. As is well known, the Weberian bureaucratic form is a product of and necessary complement to the unique conditions of modernity: industrial capitalism and centralized authority. Emphatically stated, “its 1
rise and expansion has everywhere had ‘revolutionary’ results” and lies “at the root of the modern Western state” (Weber 1968, 1002; 223).
Weber’s thesis has had a profound influence on social science theories and public management best practices. Present-day notions of state institutional qualities stem from the Weberian idealtype. Among policy-makers, there is a near-universal consensus that in order to have state capacity, it is necessary to first build “effective” administrations along standard Weberian principles. “Separate public office from private property.” “Introduce entrance examinations.” “Establish the rule of law.” “Eliminate corruption and nepotism.” “Require reporting.” “Impose hierarchical oversight.” “Delineate functional jurisdictions.” Efforts to reform public administrations aim unanimously to bring about higher levels of rationalization and to rid patrimonial practices that had long ago been abolished in the developed Western hemisphere.
In light of our conventional wisdom, China in the reform era presents a puzzle. Since the beginning of market reforms in 1979, China has achieved record-breaking development and modernization under the authoritarian rule of the Chinese Communist Party (CCP). In explaining the political basis of growth, many experts have characterized China as a developmental state.1 At the local levels, in particular, communist officials had and still continue to perform a crucial pro-market role. They plan, build, coordinate, attract investments, and manage businesses. And yet – here lies the puzzle – my study finds that China does not have a standard Weberian bureaucratic structure, which according to conventional wisdom, is necessary for strong, developmental states.
Instead, my study finds that Chinese bureaucracies have until this day retained features of prebureaucratic organizations. In particular, I underscore the economic character of their nonWeberian quality – bureaucracies in China systematically appropriate the powers of office to generate income for themselves. In the Weberian language, the Chinese administration can be described as prebendal insofar as public offices can be exploited for departmental or personal gain. As Weber saw it, prebendalism is dangerous because it invites predation and corrodes 1
We will discuss this literature in more detail in the later sections. Some representative works include Wade and White 1988; White 1988; Blecher 1991; Gore 1998; Oi 1999; Walder 1995; Duckett 1998; Howell 1993; Blecher & Shue 2001. 2
hierarchical control. To students of politics, prebendalism equates “official deviance” (Lu 2000b), “legalized corruption” (Tarkowski 1988), “[subversion] of the rule of law for personal gain” (Van de Walle 2007), and “incessant pressures on the state and the consequent fragmentation of state power” (Joseph 1987). Given the prebendal qualities of the Chinese bureaucracy, it is perhaps not surprising that many view the administration in China not as developmental, but as rents-ridden.2
It is not uncommon to find prebendal bureaucracies qualities in present-day developing countries. Such bureaucracies, for example, have persisted in and ravaged impoverished parts of sub-Saharan Africa (for example, Joseph 1987 on Nigeria; Evans 1995 on Zaire). What is new, however, is to find state bureaucracies with prebendal qualities governing a high-growth economy in a developmental setting, and even more unusually, with a Communist party at the helm of power. Following Weberian wisdom, if “the development of markets is obstructed, the use of money primarily consumptive, and the development of capitalism impossible” under patrimonial and prebendal administrations (Weber 1968, 238), then how could China defy the norm? What explains the incongruence between reform China’s bureaucratic structure and economic outcome? How could the same set of bureaucratic agents be developmental and entrepreneurial on the one hand and predatory on the other? In short, why and how has China succeeded as a developmental state despite a seemingly rents-ridden bureaucracy?
The aim of this study is not merely to unravel a case of Chinese exceptionalism. Rather, in explaining the divergence of Chinese state infrastructure from European and more recent East Asian experiences, I hope to shed new light on some fundamental questions about the institutional and political bases of development and modernization: What kind of bureaucratic forms best facilitate state capacity building and market development? Are rents for the ruling class always incompatible with development and reforms? If not, under what conditions are they compatible?
2
This literature, which we shall also later discuss, is too long to be exhaustively listed here. Xiaobo Lu wrote a comprehensive study about organized corruption in China (2000a; 2000b). Minxin Pei (2006) viewed China as a quintessential “decentralized predatory state.” See Manion (2004), Sun (2004) and Wedeman (2004) for recent empirical studies of official corruption. 3
My study suggests that the Chinese experience could offer a structural alternative to the Weberian ideal-type. I call this alternative bureau-contracting. As a form of state organization, bureau-contracting is a double-edged sword: it powerfully incentivizes governmental agents to maximize revenue, which may be taxes or rents, but it also encourages opportunistic behavior. An understanding of how the Chinese melded and wielded this sword points us to a broader understanding of the problems of incentive design and rents management, pertinent to reforming economies and developmental autocracies everywhere. Pairing a neo-institutional framework with new sources of empirical evidence, both qualitative and quantitative, this study will illuminate the mechanisms of bureau-contracting, as well as the processes that shaped them. The following sections will survey the existing literature and consider competing explanations for the Chinese paradox. Then I will present the analytical tools employed in this study.
Bureaucratic Structure, Development, and State Capacity
Good governance matters for development and social well-being. A critical element of good governance is the quality of bureaucracy. 3 As Geddes puts it, “All abilities [of states] depend on the existence of effective bureaucratic organizations. If one wants to understand states as actors, one needs to look at their bureaucratic innards” (1996, 14). The focus on bureaucratic competence prompts these questions: What constitutes “effective” bureaucratic organizations? What connects these organizations to state capacity and desirable economic outcomes? Weber’s monumental work will be discussed in more detail below.
The Weberian Thesis in Brief
Weber divided institutions of governance into two ideal-types and historical periods: patrimonial bureaucracies of pre-modern states and legal-rational bureaucracies of modern states. Patrimonial bureaucracies had two definitive features: personal loyalties and prebendal financing. Person-to-person clientelism governed patrimonial regimes. Instead of performing 3
In econometric analyses of the relationship between institutions and development, bureaucratic quality has become a standard measure of state institutional quality (Mauro 1995; Knack and Keefer 1995; La Porta et al 1999). 4
official duties in an impersonal capacity, officials and retainers typically pledged personal allegiance to their political patrons. At the same time, feudal lords rarely paid public servants regular wages for their services. Instead, they assigned prebends (or benefices) for personal exploitation.
Essentially, prebends were licenses for officials to extract rents from public office or state assets as “forms of maintenance.” They included the rights to lease land, collect fees, or conduct monopoly trade (Weber 1968, 235; 966). This method of administrative financing was known as prebendalism. Closely associated with prebendalism was the practice of tax farming. In many early states, such as the Roman Republic, rulers contracted the task of tax collection to entrepreneurial agents, who offered the highest bid for the right to collect taxes (Levi 1988). Tax farmers handed over a contracted share of revenue to the ruler or paid a royalty upfront and then kept the remaining collections as profits.4
Patrimonialism was the most important kind of administration before the emergence of the modern bureaucracy. Over centuries, as political authority centralized and demands for direct tax collection grew, rulers abolished prebendal practices and tax farming, replacing them with salaried civil services (Tilly 1993; Levi 1988). Modern bureaucracies possess five of the following traits absent in the past: (1) strict adherence to impersonal rules and duties; (2) established hierarchy of offices; (3) stable career paths with ordered promotion based on merit; (4) emphasis on technical expertise and fixed jurisdictions; (5) a fixed salary paid in money. These features lend modern bureaucracies “indubitable technical superiority” and “stringency of discipline” over their pre-modern predecessors (Weber 1968, 223; 983).
At the heart of the Weberian thesis is that claim that bureaucratic structures have economic consequences. 5 Patrimonialism obstructs market development, while modern bureaucracies complement capitalism. Different from traditional forms of capitalism (e.g. barter trade), free market capitalism required large capital investments and routinized organization of inputs and 4
In certain places, tax farmers were private businessmen, but in others, they were local elites and semiofficials. 5 It should be noted that Weber did not offer a causal argument of bureaucratic competence leading to growth. As Weber wrote, “The mere fact of bureaucratic organization does not tell us unambiguously about the concrete direction of its economic effects” (1968, 989). 5
outputs to serve mass markets. Market capitalism “is altogether too sensitive to all sorts of irrationalities in the administration of law, administration, and taxation, for these upset the basis of calculability” (Weber 1968, 240).
As Weber saw it, patrimonial administrations were inimical to industrial capitalism for several reasons: predation, fiscal arbitrariness, and appropriation. As prebendal officials could generate private income from public office, they had incentives to maximize extraction. Consequently, this provoked regular spells of over-taxation and popular rebellion. Further, extraction in patrimonial regimes was rarely based on laws and rules. Instead, it rested on the whims of lords and officials, making demands for payment unpredictable to entrepreneurs and subjects. Finally, as benefice-holders enlarged their coffers and gained independence, the power of the central ruler risked disintegration. These conditions posed a constant threat to the expansion of modern markets. In Weber’s view, although legal-rational bureaucracies do not cause growth in and of themselves, they provide an essential administrative environment for market capitalism to thrive. Legal-rational bureaucracies allow capitalists to “count on the… rational, predictable functioning of legal and administrative organs” to make long-term investment decisions (Weber 1968, 1095).
The advantage of legal-rational over patrimonial bureaucracies stems from one primary factor – the transformation of the official from a benefice-holder to a salaried employee. Patrimonial officials lived primarily on prebends, not wages; although they performed public duties, they were entrepreneurial agents whose monetary interests were often at odds with the professed goals of office. However, when the prebendal bureaucrat or tax farmer becomes a fully salaried employee, the terms of exchange between state and bureaucrat are altered. In Weber’s words (1968, 959): Entrance into an office… is considered an acceptance of a specify duty of fealty to the purpose of the office in return for the grant of a secure existence. [Simultaneously then] office holding is not considered ownership of a source of income, to be exploited for rents or emoluments in exchange for the rendering of certain services.
6
It is certainly possible that a salaried official could exploit his office for personal rents even after being paid – however, doing so is no longer a right, but a crime. As Rose-Ackerman noted in the contemporary context, “If civil service employment is well paid, corrupt officials suffer real pain if they are caught and forced out” (1999, 78). Paired with meritocratic recruitment, hierarchical controls, and processes of professionalization, salaried bureaucracies can perform in an unprecedentedly disciplined and efficient manner.
In brief, Weber’s central point was that bureaucratic predation is especially disruptive to modern capitalism. A legal-rational bureaucratic structure supports market development by removing the threats of arbitrary and excessive extraction. This point is echoed numerously in recent political economy literature emphasizing the importance of limiting governments and committing state actors to preserving private property rights (North & Weingast 1989; Weingast 1995).
Empirical Evidence of the Weberian Thesis
The Weberian thesis has inspired volumes of qualitative and quantitative studies. Case studies by political scientists highlight the role of state bureaucracies in promoting late industrialization and growth. These authors stressed the “internal characteristics of the state” as a key determinant of state capacity and effective policy implementation (Stepan 1978; Haggard & Kaufman 1989; Haggard 1999). Cross-national quantitative analyses by economists found “state institutional quality” positively associated with investment and growth rates (Mauro 1995; Knack & Keefer 1995).6 In recent years, a group of sociologists tested the Weberian thesis more systematically (Evans and Rauch 1999; Rauch and Evans 2000; Henderson et al 2007). They created a “Weberianness scale” based on survey measures of two Weberian qualities across countries, meritocratic recruitment and the presence of a predictable career ladder. Confirming part of the Weberian thesis, their analyses across 35 developing countries from 1970-1990 find
6
Drawing on subjective cross-country risk assessments, measures of state institutional quality have included quality of bureaucracy, expropriation risk, repudiation of contracts by government, red tape, corruption, etc. 7
“Weberianess” significant in predicting growth. 7 Collectively, these authors confirmed the economic impact of state structures and called to improve public institutions along Weberian principles.
Although Weber did not envision an interventionist role for governments,8 some two decades ago, the Weberian perspective took an interventionist spin in the developmental state literature. Inspired by the experiences of newly industrialized economies (NIEs) in East Asia, the literature attracted a huge following in academic and policy circles. The seminal book of the developmental school was Chalmers Johnson’s MITI and the Japanese Miracle (1982). Attributing Japan’s economic take-off in the 1970-80s to pro-active governmental economic policy, Johnson “constructed a Weberian ideal type of an interventionist state” (Woo-Cumings 1999, 2). Among the features of a developmental state are the establishment of national priorities for development, the use of economic interventionist tools (e.g. tax breaks and preferential policies), productive private-public partnerships, and cultivation of strategic industries.
According to the developmental state literature, a crucial ingredient behind effective state intervention are elite, meritocratic, politically insulated bureaucracies, as detailed in the cases of Japan (Johnson 1982), South Korea (Amsden 1989) and Taiwan (Wade 1990). Evans qualified that for development to occur, state ties with the business class matters too (Evans 1995). As this condition may fall prey to state capture and possibly corruption, he added that “internal bureaucratic coherence should be seen as an essential precondition for the state’s effective participation in external networks” (Evans 1995, 154). Professional bureaucracies structured along Weberian principles, it is claimed, are inclined to pursue collective goals of national development, while prebendal bureaucracies are easily torn asunder by self-seeking bureaucrats (Evans 1995). As such, developmental scholars argued that state intervention could work if countries build competent bureaucracies and get interventions right (Wade & White 1988; Campos & Root 1996; Root 1996; Appelbaum and Henderson 1992; Onis 1991; Castells 1992; 7
East Asian countries (Korea, Taiwan, Hong Kong, and Singapore) showed a strong correlation between high Weberian scores and rapid growth, while sub-Saharan African countries showed a reverse pattern (Evans and Rauch 2000). 8 As Evans (1995) pointed out, Weber had a neo-utilitarian view of government, that is, the role of government is only to correct market failures and provide public goods. 8
Rodrik 1995). By the 1990s, even the World Bank joined the bandwagon, advancing a twin formula of developmental success that combined “a reputable civil service” with “selective interventions” (World Bank 1993; World Bank 1997).
The developmental discourse took an abrupt turn in 1997, when the Asian financial crisis struck. Although the crisis has largely discredited the developmental state model, it did not disprove the Weberian thesis. To the contrary, many critics blamed the crisis on weakened bureaucratic quality (Segal and Goodman 2000; Henderson 1998; Pempel 1999). As Kang (2002) countered, the South Korean bureaucracy was not politically neutral, as earlier claimed (Amsden 1989); instead, it was susceptible to money politics. Others argued that it grew difficult to sustain meritocracy and bureaucratic insulation as political conditions changed in the 1990s (Haggard 2000; Heo and Kim 2000; Yusuf 2001, 25). The crisis triggered calls to overhaul the corporate and state governance structure in East Asia. As Schneider wrote, “interventionist states require reforms to make them more Weberian” (1999, 297). In short, to many observers, it was not Weberianness, but rather the lack of Weberianess, that contributed to the 1997 meltdown.
As Pearson (2005) argued, the fallout of state interventionism in East Asia has not eliminated the role of governments in the economy. Rather, she believes that the developmental state has merely given way to the regulatory state model. While the developmental state picks winners, the regulatory state creates a level playing field and enforces rules fairly on market players (World Bank 2002). Yet, regardless of whichever model prevails, the organizational precondition for “good governance” remains the same – a law-abiding, rational, salaried administration, as Weber had described a century ago.
To summarize, the conventional wisdom holds that effective bureaucracies possess certain characteristics, as outlined by Weber. Weberian bureaucracies do not produce growth in and of themselves. Nonetheless, a legal-rational bureaucratic structure is almost universally accepted as a precondition for state capacity and sustained development. Of course, not all bureaucracies meet the Weberian standards. After all, what Weber provided was an ideal type. Even so, as Evans asserted, “it only takes a very rough approximation of the Weberian ideal type to confer
9
advantage” (Evans 1995, 64). Hence, the desirability of the Weberian model of administration has been taken for granted. The real concern has been how we may realize the model.
The Chinese Anomaly: Developmentalism Without Weberianness
Having established the conventional wisdom, we now turn to consider the Chinese anomaly. As early as the 1980s, China was seen as a “socialist developmental state” and “socialist guided market,” grouped alongside capitalist developmental states in East Asia (Wade and White 1988; White 1988). Underscoring the pro-growth and entrepreneurial traits of Chinese governmental agents, scholars have employed varied labels, including “local developmental state” (Blecher 1991), “bureaucratic entrepreneurs” (Gore 1998), “local state corporatism” (Oi 1992; 1999), “market-oriented agents” (Walder 1995), “entrepreneurial state” (Blecher 1991), “state entrepreneurism” (Duckett 1998), and “market facilitating state” (Howell 1993).
More specifically, developmentalism, broadly defined as state involvement in the economy, in China has taken several models. First, in the classic developmental state model, governmental leaders, from central ministries to village heads, are seen to have taken active measures in building a market-friendly environment. Such measures included macro-economic planning, infrastructural investments, mobilization of funds, risk and capital pooling, creation of specialized agencies, provision of incentive packages, coordination across agencies and firms, and even the supply of advertisement services (for examples, see White 1988; Oi 1992; Oi 1995; Oi 1999; Walder 1996; Blecher 1991; Blecher & Shue 2001). In doing so, it was said that local state bureaucracies “consciously aided and abetted the emergence of a greatly expanded private enterprise sector” (Blecher & Shue 2001, 370). Second, in the corporatist model, local governments ran publicly owned township and village enterprises (TVEs) like diversified, dynamic corporations, with governmental executives acting simultaneously as a corporate board of directors (Oi 1999; Walder 1995; Walder 1998). In a manner reminiscent of administrative guidance in Japan, leaders selectively targeted enterprises for development. As noted, “resources and breaks are directed to those enterprises that can generate the most benefit for the corporate good” (Oi 1999, 119). Third, in the entrepreneurial model, individual state agencies directly participated in the market by operating risk-taking, profit-making businesses, 10
ranging from large companies hiring several hundred employees to dance halls and restaurants (Blecher 1991; Duckett 1998). Collectively, these authors argued that local officials’ abilities to make long-term decisions, forge alliances with firms, and run productive businesses of their own contributed to China’s economic dynamism.
And what kind of bureaucratic structure does a local developmental state like China have? It needs to be acknowledged at the outset that the Chinese bureaucracy does possess some important Weberian qualities. The CCP enforced a strict hierarchy of offices and an elaborate system of written files, reporting, and cadre assessment since state establishment. Being a communist system, China boasts a dense network of party, state, and social organs, reaching down to previously untouched levels of society. Following the launch of market reforms, central leaders promoted meritocracy in the bureaucracy. Consequently, Chinese officials have become increasingly educated (Lee 1991; Walder 2004). On top of that, mandatory cadre retirement was introduced at Deng’s initiation (Manion 1993; Lee 1991, Chapter 10), creating a relatively stable career and exit path in the officialdom. Most significantly, the civil service law (gongwuyuanfa) was promulgated in 2005, providing a legal basis for a formal civil service establishment.9 To some, China’s bureaucracies are little different from competent bureaucracies in East Asia (World Bank 1993).10
However, the features and developments discussed above belie a fundamental violation of Weberian principles – that is, Chinese bureaucracies are not purely salaried. Rather, my research finds that they remain partly prebendal, or to use a colloquial term, “self-financed.” Indeed, several studies before mine have noted and criticized the patrimonial, prebendal qualities of the Chinese bureaucracy. Previous studies described such qualities as “Leninist patrimonialism” (McCormick 1990), “neo-patrimonialism” (Wei-Arthus 2000), and “patrimonial bureaupreneurs” (Lu 2000b). 11 Going further, Xiaobo Lu asserted that China’s reforms “have not produced ‘party
9
We will discuss the limited scope of the civil service reform in more detail in Chapter 2. For example, an influential report published by the World Bank, The East Asian Miracle: Economic Growth and Public Policy, praised China and the East Asian NIEs collectively for “producing strong bureaucracies” that supported pro-growth policies (1993, 178). 11 Another study related to this literature is Andrew Walder’s Communist Neo-traditionalism (1986). However, Walder did not see neo-traditionalism as necessarily corrupt. His emphasis was on the intermeshing of personal loyalties and impersonal duties in manager-worker relationships. 11 10
technocrats’ who are procedural-and-rule oriented, but rather neo-traditional quasi-bureaucrats who approximate prerevolutionary local officials” (Lu 2000a, 233).
Indeed, detracting from the developmental school are those who highlight the corrupt and predatory side of the Chinese bureaucracy. Empirical studies report that official corruption has grown more intense and widespread since 1979 (Root 1996; Sun 2004; Manion 2006; Wedeman 2004; Lu 2000a). One author goes so far as to claim that China epitomizes a “decentralized predatory state” where “the ruling elites are unaccountable and immune from punishment for wrongdoing [and] consequently unconstrained from adopting predatory policies and practices” (Pei 2006, 12). Naysayers predicted that corruption will lead China towards “long-term stagnation” (Pei 2006, 214) and eventually collapse (Chang 2001; Goldstone 1996).
Such critical, pessimistic perspectives certainly capture part of the truth, as do glowing developmental accounts of the Chinese state. However, the former leaves unanswered the nagging puzzle of why, if the Chinese bureaucracy were patrimonial and inept, could rapid development have occurred under the leadership and intervention of CCP officials. The juxtaposition of developmental behavior alongside bureaucratic predation demands further explanation. We consider existing approaches to this paradox in the next section.
The Chinese Paradox and Competing Explanations
How can we reconcile contrasting views of the Chinese state bureaucracy, at once developmental and predatory? To be clear, the purpose of this study is not to explain the occurrence of growth per se. There are already many theories on China’s growth.12 Rather, our objective is to explain how markets can flourish even when a crucial prerequisite – a rational,
12
Economists and political scientists have offered many different explanations on China’s spectacular growth. Political scientists have focused mainly on the role of the state and China’s strategy of economic reform without political democratization. Ma (2000) provides a useful summary of this literature. One strand of this literature debates the contribution of decentralization to China’s economic success (Oi 1992; Oi 1999; Montinola, Qian & Weingast 1995; Cai & Treisman 2006). Modifying the previous literature, Chapter 5 of this study, “Dual fiscal incentives,” identifies the intersection of inter-governmental (between levels of government) and intra-government (within each level of government) fiscal decentralization in reform-era China. 12
non-predatory bureaucracy – appears lacking. This study also explores why a late industrializing state with explicit developmental goals, like China, detracted from well-trodden institutional paths.
One approach to explaining the paradox above is to maintain that China does have an effective bureaucracy, notwithstanding its imperfections. As I have earlier discussed, the Chinese bureaucracy is not entirely non-Weberian. Noting the advancement of meritocracy in the reform age, Lee (1991) claimed that Chinese officials have transformed “from revolutionary cadres to technocratic bureaucrats.” Supporting Lee’s view, David Li argued that “the massive retirement program has radically changed the human capital of the Chinese bureaucracy… being better educated in almost all cases, [the younger officials] were also generally more competent than their predecessors” (Li 1998, 394). Similarly, Li Cheng points to the “meteoric rise of Chinese technocrats” (Li 2001, 35), writing that “amidst widespread corruption, there were talented and highly motivated technocratic bureaucrats and bureaucrat-politicians… who had an increasing impact in the 1980s and 1990s” (7). The advancement of meritocracy and functional expertise in the Chinese bureaucracy certainly present significant advantages. Yet, I counter that these features do not overcome the threats of predation and arbitrariness in a partially prebendal system. Historically, China was governed by an elite stratum of scholarly officials, but who were nonetheless susceptible to bureaucratic excesses and corruption. If perverse incentives and opportunities avail, technocrats will just as likely as unqualified cadres abuse their power for private gain.
A second approach would argue that development occurred in China despite a corrupt, patrimonial bureaucracy. As Lu pointed out sharply, “those who perceive the economically overachieving Chinese state as developmental have overlooked an important factor in the literature on the capitalist developmental state: a disciplined and capable bureaucracy” (2000a, 288). So why is China both a developmental and predatory state? Lu’s answer was that China had developed an involutionary bureaucratic regime that failed to rationalize and thus remained indefinitely patrimonial. In his view, the patrimonial character of the Chinese bureaucracy is the fundamental cause of widespread corruption in the reform era (2000a; 2000b). Thus, the author concluded, “We may now better understand the contradictory roles of the Chinese state: the 13
state’s developmental goals and policies are compromised by its inability to maintain a disciplined bureaucracy, even though some of its goals have been achieved” (Lu 2000a, 289).
Although Lu has observed a critical paradox, his answer falls short of addressing the puzzle. To claim that the Chinese state has only achieved “some of its goals” is a serious understatement. Lu’s formulation furthermore fails to explain why the Chinese state had been unusually successful at achieving “some of its goals” (some!) and yet unsuccessful at the goal of bureaucratization. If China’s bureaucracies were backward and undisciplined, as claimed, then how could China emerge as the fastest-growing economy, in contradiction to Weber’s predictions? Asked differently, how does patrimonialism in China compare to patrimonialism elsewhere? Why have patrimonial states like Zaire failed miserably (Evans 1995), while China saw a different fate?
A third approach can be located in the so-called “East Asian paradox,” where high levels of corruption accompanied high growth in the 1980s-90s (Wedeman 2002; Rock and Bonnett 2004; Campos 2001). This literature argued that development can coexist with predatory state actions if the latter were relatively benign or sufficiently contained. Some claim that corruption in East Asia was less harmful than corruption in other under-developed states (Wedeman 1997, 2002a, 2002b; Sun 1999, 2005). Wedeman (1997), for example, distinguished between looting, rentscrapping and dividend-collecting. He argued that corruption in East Asia took the forms of rentscrapping (e.g. self-seeking manipulation of regulatory powers) and dividend-collecting (e.g. kickbacks to officials in return for favorable treatment) rather than plunder. On a similar note, others contend that corruption in East Asia is comparatively predictable, with payees usually getting what they pay for, and was thus less harmful to the economy (Campos et al 2001; Hutchcroft 1997).
In this literature, it is further argued that structural conditions in East Asia favored restrained predation. The centralization of power, often in the hands of one dictator, ensured predictability and limited corruption, as was the case of Indonesia under Suharto (MacIntrye 2000; 2001; see also Rock and Bonnett 2004; Kang 2002). Furthermore, regulatory power was concentrated in a small core of elite agencies, approximating centralized corruption, less destructive and more 14
predictable than decentralized corruption (Bardhan 1997). On the business end, East Asian economies were dominated by mega-conglomerates, such as the Korean chaebols. Kang (2002) argued that when the business community is united and strong vis-à-vis a coherent state, like in the case of South Korea, excessive predation or state capture can be mitigated.
Although the East Asian literature is useful in understanding the differential impact of predatory state actions, its applicability to China is limited. The East Asian literature centered on grand money politics between mega-firms and elected politicians, especially in the form of bribetaking. While grand corruption also exists in China, the primary issue that affects daily business conditions is decentralized predation, especially at the local levels. As Kellie Tsai, who has studied private entrepreneurship in China, related, “In any given week, the typical factory owner may be approached by dozens of different agencies requesting seemingly random user charges, surcharges, and contributions for local projects” (Tsai 2004, 15). Notably, such “petty” predation was absent in the East Asian NIEs, which offered investors market environments free of harassing bureaus. Furthermore, unlike its East Asian counterparts, China has a mammoth, multi-tiered administration, with individual offices at each level generating their own pots of funds, and each having distinct regulatory powers in their jurisdictions. China also has an extremely diverse business community, a dizzying mosaic of private, state-owned, and collectives firms without a common voice. Taken together, the fractured structural conditions in China defy a neat “mutual hostage” situation, which would predict a limited and equitable distribution of rents (Kang 2002). Thus, the Chinese paradox calls for a different framework of analysis.
My approach builds on some insights of the existing arguments. I agree that there are different forms of predatory actions, some more harmful than others. I also agree that extraction is less harmful if predictable and controlled. However, in reference to Wedeman’s typology, the existing literature does not explain when authorities would choose rent-scrapping and dividendcollecting over looting, especially if looting yields more immediate benefits. Furthermore, the existing literature cannot tell us how patterns of rents-taking become stabilized in a massive state bureaucracy like in China, where power is not concentrated in the hands of one dictator or
15
a few pilot agencies, but is highly decentralized.13 In other words, we need to account for how particular state policies and institutional design condition extractive behavior. We also need to examine mechanisms of rents management in complex organizational settings.
My interpretation and empirical findings will depart from previous conceptions of neopatrimonialism and corruption in significant ways. I submit that the Chinese bureaucracy is not a standard Weberian bureaucracy seen in developed and developmental states. But I also submit that it is not a typical patrimonial bureaucracy. In the next section, I develop a neo-institutional theoretical framework to explain what I think the Chinese bureaucracy represents.
The Bureau-Contracting State
Why has China succeeded as a developmental state despite a seemingly rents-ridden bureaucracy? My answer to this question comes from an identification of the nature of China’s bureaucracy and its development in the recent decade. My findings, based on recent fieldwork and statistical testing, lead me to characterize China’s unique path of bureaucratic adaptation as bureau-contracting. As the label indicates, bureau-contracting combines the terms “bureaucracy” and “contracting.” Bureau-contracting denotes a system of contracting within the state bureaucracy. In terms of physical structure, the term bureau-contracting also describes a bifurcated bureaucracy that is unique to China, composed of a core elite administration (jiguan danwei) and a periphery of extrabureaucracies (shiye danwei), bound by a thick familial relationship.
To students of public administration, the label “bureau-contracting” is paradoxical because bureaucracy and contracting are ordinarily alternative modes of organization (Moe 1984). Governments either contract public services to private suppliers or establish their state-funded public agencies. In developed systems of governance, public and private entities are strictly
13
Some may rightly disagree that power is decentralized in China. Indeed, personal power is still dominant in the Chinese political system. Here, by “decentralized,” I mean that individual agencies can exercise administrative power in their respective domains. Often, businesses in China find themselves having to deal with not just one agency, but multiple ones. 16
divided. Where mixed entities exist, they are the exception, but not the rule.14 Contrastingly, in China, bureaucracies are governmental units with party-appointed personnel, but economically, I found that they behave to varying extent like entrepreneurial contractors in two fundamental ways. (a) They may generate “self-financed” income through the provision of services to the party-state.15 (b) They can rightfully retain and use collected funds to finance jobs, benefits, and administrative costs for their own offices. In short, the state at each level assigns individual offices revenue-making privileges and de facto property rights over income earned. Framed differently, it is as if the Chinese government “contracts” the tasks of governance to its public bureaucracies, exchanging prebends for services, thus creating the paradox of bureaucontracting.
To be clear, the features of bureau-contracting, as described above, were not implanted overnight based on some leader’s grandiose design. Neither do I claim that these features are permanently static. Rather, as I see it, and as we will discuss at length in Chapter 5, the prevailing structure of bureaucracy is descended from a long historical legacy of prebendal governance. And as we speak, these features are undergoing what I view as a relentless path of institutionalization. In other words, I shall claim that while much of the Chinese state has not – as yet – transformed into a standard Weberian bureaucracy, it is remarkably capable of adapting and “legal-rationalizing” prebendal elements of governance into its formal bureaucratic organization. While Weber has described for us one set of bureaucratic rationality, we can observe in contemporary China, a different set of rationality, so to speak.
An understanding of the Chinese state as a bureau-contracting model removes us from the binary lens of China as either a classic Weberian bureaucracy or an unfettered corrupt regime, both of which, I contend, leads us to impartial views of the complex realities. Further, my framework allows us to identify and test the mechanisms that may allow local state-led development to occur despite the predatory risks inherent in a neo-prebendal bureaucratic structure. I underline two of such mechanisms below. 14
I thank Scott Kennedy for pointing me to the existence of mixed public-private entities in the United States in comparison to China. 15 By self-financed income, I mean revenue earned by and belonging to individual offices at each level. Self-financed income is part of the non-tax revenue (feishui shouru) in China. Not all non-tax revenue is extrabudgetary. I will explain further in Chapter 2. 17
Managed rents
I argue that bureau-contracting constitutes a system of rents management. Before I proceed, let me define rents in the context of this study. I use rents in the prebendal sense of the word. As Weber defined, prebends are “life-long assignment to officials of rent payments… in compensation for the fulfillment of real or fictitious duties of office” (Weber 1968, 967). When the state awards an office the right to generate private income through the exercise of public duties, that income functions as rent payments in lieu of budget allocations. In this definition, self-financed income earned by and belonging to individual state agencies are a form of rent payments. Often, as seen in China, bureaucracies generate funds for themselves by imposing excessive regulation or quasi-monopoly rights (see Chapter 2), thus creating socially inefficient rents in the classical economic sense (Krueger 1974; Buchanan et al 1981; Bhagwati 1983; Tollison & Tullock 1988). Importantly, bureaucratic rents-extraction in China is qualitatively different from the kind of lawless plunder described in places like Zaire (see Evans 1995; Wedeman 1996).
In my account of bureau-contracting, I emphasize two features that function to manage rents provision for the bureaucracy. First, public agencies and providers require what I call “policy awards” from the state (awarded at the central and local level) to earn revenue, e.g. policies sanctioning the exaction of a particular fee. From the late 1990s onwards, a comprehensive institutional framework and norm of earning self-financed income according to – rather than in violation of – state policy has evolved. Second, I show how financial authorities developed enhanced information and control over bureaucratic funds. My research finds that, following budgetary reforms in the mid-1990s, it has become standard practice for individual bureaucracies to turn their income into state bank accounts, instead of stashing them away illegally, but to retain property rights over the use of those funds. The rules of making, submitting, and spending self-financed income are more rigorously enforced today than during the “go-go” years of the 1980s and early 1990s. In short, we observe bureaucrats in China still behaving in entrepreneurial ways but increasingly following standardized rules of procedures. I discuss both of these features in Chapter 2.
18
In a utopian state, rents are economically harmful, but in the real world, rents are politically indispensable. Rents, firstly, are crucial for autocracies to reward and unite political elites (Magaloni 2006). Secondly, as some argued, creating rents for political interest groups may be a necessary cost to pay for realizing otherwise painful reforms and producing greater good in the long run (Shleifer and Treisman 2000; Stiglitz 2000; Zhuravskaya 2007). Thirdly, creating a privileged governing class is central to any hegemonic party’s power to co-opt society. In the case of China, as Lee noted, “the sole channel of upward mobility for ambitious individuals has been through the bureaucracy – a channel that the party-state has easily controlled through its prerogatives over the personnel management of cadres” (1991, 4; emphasis added). The combination of a relatively well-rewarded cadre corps and partial meritocracy allows a hegemonic party not only to satisfy the existing elites, but also to induce citizens to aspire membership in the bureaucratic strata. 16 Finally, from the perspective of individual leaders, rents sustain patronage machines. As Shirk (1993) pointed out, although China is not an electoral system, political leaders run extensive patronage networks that feed on the distribution of jobs, perks, and market access.
Put differently, once politics are taken into account, the problem of development is not to eliminate rents but to regulate rents provision so as to minimize its costs and to stabilize patterns of extraction.17 Much of the existing political science literature has identified strategies for “splitting the spoils” within a small coterie of powerful political elites (Magaloni 2006; Shleifer and Treisman 2000). However, the systematic distribution of rents, from leading officials
16
The dilemma of citizens under autocratic rule is well illustrated in a blog session upon which I chanced. A recent graduate had asked whether he/she should join the land bureau or a large private company on a public blog. The poster received many replies urging him/her to take the position at the land bureau. One blogger (A) responded, “I noticed a strange phenomenon. Whenever we speak of civil servants or corruption, we grind our teeth in hate. Yet as soon as someone posts a message like this one, everyone urges the poster to join the civil service.” Another blogger (B) replied, “There is nothing strange about this. We don’t hate civil servants, we hate the system. The reason is clear: civil servants have too much grease.” Blogger A retorted, “Hypocrite! If you hate the system, then why is it that when the opportunity comes, you want to benefit from the system. You are merely looking for an excuse.” The blog exchanges were accessed at http://bbs.cz.soufun.com/1820170502~0~304/727074_728265.htm. 17 The problem, as compellingly summed up by Shleifer and Treisman (2000), is a “paradox of rents.” They wrote, “Transforming stakeholders from opponents to supporters of reform often requires the creation of rents by the government that these stakeholders can be offered in exchange of their support. An obvious paradox arises. The goal of reform is to reduce rents and rent seeking” (9). 19
down to street-level cadres, in massive state organizations is a much less straightforward and under-examined problem.
In a communist system like China, the cadre corps constitutes a distinct interest group (Oksenberg 1967). In the revolutionary sense, cadres are the leaders of the masses (Lee 1991). Today, cadres are composed of a group of over 40 million (46 million by 2007) functionaries serving in party-state organs, including about 500,000 leading officials, who constitute the heart of the political system.18 As Walder (2004) asserted,
The political elite of 500,000 cannot rule the country unless it can retain the obedience of 40 million state cadres… If the elite maintains the discipline of state bureaucrats and the allegiance of party members, it can withstand challenges from other groups in society, even in periods of economic hardship and social upheaval. If, however, challenges from other groups stimulate a defection of the Party membership and parts of the state bureaucracy, the elite is in real trouble (197). But how can the political elite retain the obedience of 40 million cadres? Shirk’s (1993) alludes to the establishment of rents and patronage machines. However, as she pointed out, “we as yet have little empirical information about how these local industry-based political machines work in China” (1993, 189). Building on Shirk’s observations, my research aims to identify mechanisms for a systematic, arms-length distribution of rents. Yet, different from Shirk’s emphases, my focus is not on elite politics or person-to-person clientelist exchanges between public and private actors. Rather, my focus is on the organized allocation of privileges at all levels of the socialist hierarchy. I bring attention to the contracting relationship between state and bureaucracy and between sub-components of the bureaucracy.
Dual Fiscal Incentives
I further contend that narrow departmental interests to extract rents are mediated by local developmental incentives. Several earlier studies have underlined the contribution of fiscal contracting and local fiscal incentives to China’s economic growth (Oi 1992; Oi 1999; Montinola,
18
The “leading officials” are cadres at the chu rank and above. 20
Qian & Weingast 1996). From 1979 onwards, each level of government in China contracted with the next lower level of government, requiring the latter to submit only an agreed share of revenue. At the same time, local authorities were given unprecedented autonomy over retained funds and increased responsibilities for expenditure needs, meaning that budget constraints were hardened. As Oi (1992; 1999) argued, locally retained revenue constituted an economic surplus that motivated local officials to grow businesses and collect more taxes. Montinola, Qian and Weingast (1996) concluded that fiscal contracting “induced a strong positive relationship between local revenue and local economic prosperity for all provinces and cities, thus providing local officials with an incentive to foster that prosperity” (64).
While the contribution of fiscal contracting to China’s developmental success has received much attention, I would argue that surprisingly little is known about a simultaneous process of contracting that has taken place within each level of government. Just as local governments needed incentives from upper levels to pursue growth, bureaucracies at each level also needed motivation to self-finance in a new entrepreneurial climate. Following reforms, it was not only the local governments that enjoyed rights to surpluses in the form of retained taxes.19 I emphasize that individual bureaucracies too had rights to surpluses in the manner of selffinanced income, which formed each locality’s non-tax revenue. Just as local governments had autonomy to expend extrabudgetary funds for the locality’s needs, each office had the right to use the pots of funds generated to pay for cadre wages, benefits, and administrative costs.
A key – but yet unexamined – implication of the fiscal structure described above is that dual fiscal incentives are at play. On the one hand, local governments collectively had incentives to grow the economy and increase taxes. On the other hand, individual bureaucracies that compose local governments had incentives to extract rent payments for the benefit of their own departments and personnel. Thus understood, a critical question arises as to whether rentseeking incentives resulting from bureau-contracting might overshadow developmental incentives provided by inter-governmental fiscal contracting, particularly in light of the fiscal recentralization policy of 1994. In my analysis, I hypothesize that local cadres benefit tremendously in the short-term from increasing self-financed income, but they benefit in the 19
The term “rights to surpluses” come from Oi’s Rural China Takes Off (1999). 21
long-term only from increased taxes, which in turn depends on the collective capacity of local governments to attract investments and sustain productive market activities. In Chapter 4, I report supporting evidence for my hypothesis in a regression analyses of an original budget dataset.
To sum up, I argue that it is the combination of an incentive-compatible fiscal design and external instruments of rents management and budgetary oversight that have supplied highpowered incentives and at the same time checked bureaucratic predation in China. My findings offer
an
empirically-tested
explanation
for
the
paradoxical
coexistence
of
local
developmentalism and appearances of widespread bureaucratic predation.
The Franchising Logic of Bureau-Contracting
From a comparative institutional perspective, bureau-contracting is a species of hybrid organizations that fuse high-powered incentives with internal authority structures. What do I mean? Consider an example from the business world – franchising. Say I am a small business owner of several restaurants. Each restaurant is operated by a manager whom I hire for a fixed salary each month. So long as they are happy to keep their jobs, the managers will perform satisfactorily. But my managers are unlikely to work extremely hard, take initiatives, or bear risks because they receive the same payoffs regardless of their efforts. To motivate my managers, I contract with them to take only a share of the restaurant’s revenue and allow them to keep whatever remains; any losses incurred will also have to be borne by them. In this new arrangement, my managers are no longer salaried employees; they have become independent owners.
As they have a residual claim, franchisees are highly motivated to maximize profits. It is this feature of high-powered incentives that unleashed the entrepreneurial energies of franchisees and made some franchised firms like McDonald’s extraordinarily successful. However, franchising has its costs compared to direct ownership. Once my managers become franchisees, they become independent. If I instruct them to make burgers a standardized way, they may not follow my orders as they did before, unless I can credibly threaten to terminate the contract. Yet, 22
even more insidious than open defiance is opportunism. The managers-turned-franchisees may try to “cheat” our agreement by maximizing their own profits at the expense of my chain’s reputation. For example, they may use cheaper, low-quality ingredients when making burgers or add water to milk shakes. Each franchisee will save money for itself by cheating, but the negative externalities will be spread across the chain. If I fail to control such opportunistic behavior, customers may soon lose confidence in the brand name and the franchise will go out of business.
Thus understood, the central problem of franchised organizations is how to unleash agent entrepreneurism and control opportunism simultaneously. Resolving this problem lies at the heart of successful franchising, as illustrated in the McDonald’s experience. “[Ray Kroc, the founder] used franchising to unleash the power of operators who have an ownership stake in this businesses. Though he demanded adherence to strict operating standards, he also freed franchisees to market their service as they saw fit, and he motivated them by given them an opportunity – unheard of in franchising – to become rich before he became rich” (Love 1996, 6). In a phrase, “The fundamental secret to McDonald’s success is the way it achieves uniformity and allegiance to an operating regimen without sacrificing the strengths of American individualism and diversity” (7).
The long preamble aims to illustrate an unusual idea that forms the framework of my analysis, namely that the problem of bureaucratic organization in China is analogous to the problem of corporate franchising. In both organizational forms, contracting is conducted within a hierarchical structure. In organizations like these, incentives are high-powered but the risks of opportunism (or strategic self-seeking interests) are also high. In the context of the reform-era Chinese bureaucracy, the central organizational problem is to inject entrepreneurial incentives into the state bureaucracy but at the same time to constrain opportunistic risks arising from bureaucratic involvement in the market. The structure of bureau-contracting is a compound of mechanisms evolved to address this central problem of state organization.
23
Bureau-Contracting vs. Weberian Ideal Type
As Weber noted, ideal types “after all, are to be considered merely border cases which are of special and indispensable analytical value, and bracket historical reality which almost always appears in mixed forms” (1968, 1002; emphasis added). Bureau-contracting is an ideal-type, and my documentation of the Chinese case illustrates how one significant variety of this ideal-type may exist in reality. The organizational logic of the ideal-type suggested by the Chinese case may lend us comparative insights into divergent paths of state institutional development.
Bureau-contracting is a high-powered but opportunistic alternative to the Weberian ideal-type. Returning to the corporate analogy in this study, bureau-contracting is to franchising as Weberian bureaucracies are to directly owned firms. Just as franchised corporations may work as well as non-franchised firms, a bureau-contracting structure may perform a marketsupporting role as Weberian bureaucracies do, depending on the economic and political goals of the leadership, as well as on the design and enforcement of disciplinary mechanisms and incentives. Figure 1.1 summarizes the organizational parallels.
Figure 1.1: Bureau-Contracting and Weberian Bureaucracy Compared
My conceptualization of bureau-contracting as an ideal type builds on Weber’s approach in two respects. First, I do not claim that a bureau-contracting structure causes development, just as 24
Weber did not argue that legal-rational bureaucracies would lead to growth. Rather, the argument here is one about “institutional complementarities,” to borrow a phrase from the authors of Varieties of Capitalism (Hall & Soskice 2001). Second, bureau-contracting as an ideal type captures a trajectory of development. In describing the ideal type of a legal-rational bureaucracy based on Germany’s experience, Weber projected the course of European bureaucratic formation. As he maintained, the modern bureaucratic structure “is everywhere the product of a long development” (Weber 1968, 957). Similarly, I maintain that the bureaucontracting model was not consciously imposed by leaders in Beijing at a specific point in history. Like the Weberian bureaucratic form, it too is the product of an evolutionary process and policy choices. In Chapter 5, I trace the circumstances and choices that have shaped this process of institutional change.
Data & Methodology
My research methodology combined “thick interpretation” during an early stage of research with empirical testing in the later stages. The dissertation involved nine months of fieldwork in China and 165 interviews with officials and rank-to-file cadres across provinces and governmental sectors from the period of 2006-2009. I conducted most of my research at the local levels (particularly county and city governments) in Shandong, Tianjin, Jiangsu, and Sichuan, with some central-level interviews conducted in Beijing. To my best ability, I heeded the advice of Fenno to “soak and poke in the data.” In addition to transcribing interviews verbatim, I took notes of day-to-day observations and casual conversations. In general, I sought to immerse myself in the habitat of my subjects and develop a realistic set of lens that could help me see how bureaucrats in China performed their work. In Appendix C, I describe my interviewing strategy and interview targets.
My primary sources of qualitative data are interview and field notes. Where appropriate, I use policy documents, Chinese-language secondary literature, and media reports to supplement my analysis. I believe this purposive choice of data has helped me to draw qualitative conclusions different from some prevailing views of the Chinese bureaucracy. Many studies relating to bureaucratic financing, entrepreneurism, and rents have fallen under the scope of corruption. 25
These studies tend to draw heavily on secondary and media reports because, understandably, it is difficult to gain access into bureaucracies in China. But, while media reports can be extremely useful supplements, they tend to be incomplete and sometimes sensational. Relying on secondary material may lead one fall into urban legends and miss oft-complex realities. In my field research, I tried simply to uncover the mundane, tedious rules that govern bureaucratic behavior (thankfully, I have a taste for the mundane!). At the end of the day, I believe it is mundane, tedious rules that help us to understand what really matters to bureaucrats. Following the collection of qualitative data, the next challenge was to find quantitative evidence to test the implications of my institutional narrative, specifically that of dual fiscal incentives. For this purpose, I created and analyzed a previously unavailable public budget dataset from Shandong province from the period of 2001-2005. This is a unique “internal” (neibu) data source. This dataset allows us to measure patterns of administrative financing and revenue streams among county governments in Shandong with unprecedented though still imperfect precision. Besides the Shandong dataset, I also employ descriptive statistics on public spending and employment from other “internal” data sources, compiled by the ministry of finance, primarily the Local Public Financial Statistics (Difang Caizheng Tongji). To my knowledge, the dissertation contains one of the first systematic analyses of public spending data in China.
Outline of Dissertation
The rest of the dissertation proceed as follows. Chapter 2 introduces the neo-institutional concept of bureau-contracting, building on and modifying existing theories of the firm. By disaggregating the formal organizational structure and budgeting processes, I show how each bureaucratic unit owns and has autonomous use over all or part of the self-financed income it generates. I also emphasize the function of state-enacted “policy awards” that authorize bureaucracies to collect revenue for self-financing. The analysis distinguishes the prebendal logic of self-financing in China from conventional views of corruption.
Chapter 3 charts the structure of cadre rewards. I decompose cadre rewards and explain why we observe such large disparities across regions and across offices. I argue that these patterns reflect an unusually high-powered incentive scheme, but one that is different from incentive 26
schemes seen in private firms and regular public bureaucracies. The analysis highlights the coexistence of two sets of fiscal incentives, one motivating bureaucratic actors to maximize tax revenues for local treasuries, and the other motivating them to maximize departmental surpluses. The strategies for maximizing fiscal gains in these two respects could potentially be at odds with each other.
Chapter 4 proceeds to evaluate a theory of dual fiscal incentives systematically. Specifically, I test the relative strengths of developmental and rents-collecting incentives. Employing an errorcorrection model on a county-level budget dataset, the analysis finds confirmation of my hypothesis that developmental incentives dominate rents-collecting incentives in the long term. This chapter underlines the combination of an incentive-compatible fiscal design and institutional mechanisms of rents management. Chapter 5 examines the political economy of bureaucratic formation in reform-era China, that is, why the Chinese bureaucracy has come to look the way it is today. I begin by discussing the historical institutional continuities in China’s administrative structure from imperial times to the pre-reform period. Then moving to the reform era, the chapter traces the emergence of entrepreneurism among Chinese bureaucracies, which contributed to initial growth efforts but created new problems of control. It then looks at how fiscal and administrative reforms introduced in mid-1990s increased pressures for bureaucracies at all levels to self-finance, but at the same time, institutionalized processes of self-financing. The end product of the reforms approximates a bureau-contracting model, rather than a salaried Weberian bureaucracy.
Finally, in concluding, I consider the theoretical implications of my study, particularly with regard to the design of second-best institutions and the relationship between development and rents. I emphasize the analytical value of alternative models, as the Chinese case offers, in rethinking some basic questions about institutional quality and corruption. While not advocating the Chinese experience, I speculate on the conditions that allow bureau-contracting to germinate and possibly take root in other contexts. Finally, I discuss remaining questions for future empirical research.
27
CHAPTER 2
BUREAU-CONTRACTING: THE INDUSTRIAL ORGANIZATION OF CHINESE BUREAUCRACIES
The main government building in the seat of Blossom County was acquired by a proud new owner: the local land bureau. The landmark property that beamed energy-guzzling laser lights from the rooftop, cost 230 million yuan to construct, and housed over twenty bureaus (including the land bureau itself), was sold through the state asset management company. “I don’t understand why the land bureau would want to buy the entire building,” one cadre mused. “All its employees can barely fill five floors. What would they do with the rest of the property?” (AI 2007-112). While the use of assets was shrouded in mystery, it was plain that the land bureau was flushed with funds. Interestingly, its funds had not come from budget allocations. As the land bureau chief was said to have once boasted, “We don’t need money from the finance bureau. We have enough money of our own to last ten years” (AI 2007-112).
The wealth of Blossom county’s land bureau may be exceptional, but the rights that public bureaucracies in China exercise over their “own money” are not. As my brief encounter reveals, Chinese bureaucracies are distinct from bureaucracies of most governments in their literally entrepreneurial character – they can own and spend the income they earn. Not unlike firms, they may suffer deficits or enjoy huge surpluses.
Chapter 1 identifies the industrial logic of bureaucratic organization in China. I develop the construct of bureau-contracting, situating it in the universe of state organizational forms. Following the theoretical framework, I illustrate three of the core elements of bureau-
28
contracting in the Chinese context. A key finding in this analysis is that the state at each level assigns individual bureaucracies de facto property rights over self-financed income generated.
Bureau-Contracting in Neoinstitutional Perspective
To illustrate how public bureaucracies may behave in part like private contracting firms, I develop a neo-institutional construct – bureau-contracting. This section first summarizes standard transaction costs theory, which explains why private and public organizations take the form they do, and then modifies it to accommodate private-public organizational forms like in the Chinese case.
Standard Transaction Costs Theory
In the new economics of organization, contracts are the starting point of institutional analyses. Ronald Coase’s (1937) classic theory of the firm asks why some transactions take place in the market between firms and others within a firm. Phrased differently, Coase puzzled over the relative value of market transactions over internal authority structures. The answer proposed by Williamson (1974) is transaction costs – that is, difficulties posed in completing transactions. Such difficulties arise from two factors. The first is bounded rationality. Humans cannot possibly foresee and design contracts ex ante to address all possible contingencies in an exchange; there is always an element of unpredictability. The second factor is the assumption of opportunism. Individuals may seek to benefit from an exchange at the expense of the other party through deception or concealment. Opportunism is, as Williamson had defined, “a lack of candor or honesty in transactions… [or] self-interest seeking with guile” (1974, 9).
In the neo-institutional paradigm, the mode of transaction chosen depends on one’s calculations of transaction costs. When it is costly to organize a market transaction, it is preferable to perform the transaction internally, administered by hierarchical rules. For example, if a firm lacks information about a service supplier or believes it has difficulty enforcing a contract, it would choose to organize the service itself through vertical integration, making the suppliers of the service the firm’s own managers. In Williamson’s words, “considering the risks that simple 29
(or incomplete) contingent claims contracts pose, the firm may decide to bypass the market and resort to hierarchical modes of organization” (1974, 9).
There exists an inherent trade-off between market transactions and authority structures. On the one hand, market transactions supply high powered incentives because efficiency gains flow directly to the transacting parties, whereas authority structures supply only low powered incentives as employees benefit indirectly from the firm’s financial gains.20 On the other hand, market exchanges pose a greater risk of opportunism, whereas authority relations allow more control and predictability. It is certainly true that even within a hierarchical structure, opportunistic risks exist. In a classic agency problem, an employee may shirk responsibilities assigned by the boss. However, in market exchanges, opportunistic risks tend to be even higher than within an organization either because information is lacking, or more simply, transacting parties do not have direct control over each other, unlike in an employer-employee relationship. Adding to our previous point, then, organizational leaders choose between markets and hierarchies depending on their assessment of the risks and costs of opportunism and the need for high-powered incentives.
To illustrate, contrast the franchising and direct ownership model in the fast food industry. Franchising approximates a market-based transaction. A franchising firm contracts to a franchisee the right to use the company’s brand name and operational software in exchange for a royalty payment or share of profits. Direct ownership is a hierarchical model. The firm opens outlets and hires managers to operate them. McDonald’s adopted franchising because the company’s founding executives aspired to build a massive chain and had do so quickly to capture market share. In contrast to McDonald’s, In-N-Out, a much smaller Californian burger chain, runs on a direct ownership model. Its salaried managers may not be as highly motivated as franchisees to maximize profits; however, managers are easier to monitor and control than franchisees.
With few exceptions, transactions costs theory has not been applied to analyzing public bureaucracies. When it has, the issue has been framed as a binary choice that governments face 20
These indirect benefits typically come in the form of promotions or wage raises. 30
in delivering public services through the private sector (i.e. high powered incentives) or the public bureaucracy (i.e. low powered incentives) (Moe 1984; Donahue 1989; Acemoglu et al 2007). As an organizational theorist states:
[The government] can create its own bureaucracy, or it can engage in marketlike transactions by contracting with private actors who promise to provide services at a stipulated price (there is no prevailing market price) to be paid by the government (Moe 1984, 759). In theory, governments choose between contracting or creating a bureaucracy depending on the transaction costs involved. Certain services, like policing, are delivered through the public bureaucracy because it poses a high moral hazard risk. Other services, like garbage collection, are more likely to be contracted to private firms through a competitive bidding process to increase the efficiency of service delivery.
In standard models of public administration, delivering services through the public bureaucracy is by definition a low-powered incentive scheme because public servants are not supposed to capture private gains from public transactions. If they do, it is considered corruption and subject to punishment. Even when bonuses are introduced, they are awarded based on some objective measures of individual performance (Frant 1996), but not on a particular agency’s financial “output.” The anomaly in China, as I will show, is that its public bureaucracies can generate real surpluses that accrue to their own coffers and not to the state treasury. Furthermore, each bureau’s financial performance determines actual cadre payoffs.
Bureau-Contracting: Standard Theory Modified
Transaction costs theory has been applied to either private or public organizations.21 It has not, however, been employed to analyze public-private entities. To push neo-institutional theory into the realm of mixed organizations, we need to consider multiple dimensions of transactions. These dimensions can be divided into financial and personnel relations. I illustrate this approach in Figure 2.1. Financial relations concern whether a government funds public service providers 21
The literature on transaction costs is huge. For a review of its application to theories of political organization, see Moe (1984). 31
with high-powered contracting rights (i.e. market-based) or low-powered fixed budget allocations (i.e. hierarchy based). Personnel relations concern whether the relationship of a service provider vis-à-vis the government is that of a contracting party (i.e. market-based) or an employee (i.e. hierarchy based).
Figure 2.1: Bureau-Contracting: Fusing Markets and Hierarchies
Market-Based Hierarchy-Based
PERSONNEL RELATIONS
FINANCIAL RELATIONS Market-Based
Hierarchy-Based
Private Contracting
Certain forms of public enterprises
BureauContracting
Public Bureaucracy
The two dimensions of financial and personnel contractual relations create four configurations of governmental organization. The top left quadrant is private contracting, which involves purely market-based financial and personnel transactions. Private contractors are entitled to profits earned from the provision of public services. The bottom right quadrant represents public bureaucracy, which are staffed by public employees whose salaries are paid through budget allocations. The top right corner capture certain kinds of state enterprises, in which employees are not directly appointed by the government but are paid fixed wages regardless of the firm’s performance. Oil companies in Mexico approximate this organizational form. Managers of these companies face weak incentives to be entrepreneurial because they receive the same wages whether or not more profits are earned.22 Finally, the bottom left corner is bureau-contracting, which fuses market-based administrative financing with hierarchical personnel control. In a
22
I thank Alberto Diaz-Cayeros and Beatriz Magaloni for raising this point. 32
bureau-contracting model, the service providers are state appointed employees, but simultaneously, like private contractors, they have rights to profit from the provision of services to the government.
From a financial perspective, bureau-contracting deviates from conventional models of public administration insofar as public bureaucracies are assigned property rights over income earned. As we will see in the Chinese case, each bureaucracy can legitimately retain all or a share of income generated, in addition to receiving some budget allocation (and in some cases none at all). Such additional income is generated in the course of conducting governmental functions, e.g. regulation or public goods provision. In this sense, these bureaucracies behave partly like contractors, or, in the historical sense, benefice holders.
In terms of personnel management, agents in a bureau-contracting model are public employees, not private actors. In the Chinese context, the CCP exercises a strict organizational principle of “party governs cadres” (dangguan ganbu). “Cadres” refer broadly to the functionaries serving in the party-state organs, numbering over 40 million. China operates a unified system of cadre management that encompasses all bureaucratic personnel, whether salaried or non-salaried. All cadres possess bureaucratic rank (jibie), which determines one’s privileges and standing in the hierarchy and allows the party to transfer cadres from place to place. The party directly appoints cadres on the nomenklatura list who are at the chu rank and above. A rigorous system of hierarchical personnel management distinguishes bureau-contracting from pure contracting and prebendalism in early states.
In terms of physical structure, picture two ideal-types of state organization, as seen in Figure 2.2. Model 1 represents an extreme example where all services are provided by an oversized state bureaucracy. Socialist economies like the former Soviet Union come to mind. Under socialism, the state collects taxes and then redistributes revenue to pay wages to salaried officials. Model 2 represents another extreme of a small state bureaucracy coupled with extensive contracting of public services, which is typical of liberal market economies like in the United States. Model 3 illustrates the Chinese model, which diverges from the ideal types of Model 1 (big bureaucracy,
33
no contracting) and Model 2 (small bureaucracy, extensive contracting). The Chinese model is neither Model 1 nor 2 but a fusion of the two.
Figure 2.2: Varieties of State Organization
Conventional Models of State Organization MODEL 1 (Big Bureaucracy, No Contracting)
China’s Model of State Organization
MODEL 2 (Small Bureaucracy, Extensive Contracting)
EB EB
C C
Bureaucracy
C
EB
B
B
EB C
EB
C
Having introduced bureau-contracting in a neo-institutional framework, I now proceed to explain how it works concretely in the Chinese context. The four core elements are:
I.
A bifurcated state structure composed of core bureaus and extrabureaucracies
The state bureaucracy in China is composed of two basic components: bureaus (jiguan danwei) and extrabureaucracies (shiye danwei). In principle, bureaus (or literally translated as administrative
units)
are
responsible
for
administration
and
regulation,
while
extrabureaucracies (or literally translated as service units) deliver public services. In every governmental sector and at every level of government, a parent bureau manages a group of extrabureaucracies and has micro-control over the latter’s operation, finances, and personnel appointment. In reality, extrabureaucracies perform a range of tasks delegated to them by their respective parent agency. Besides public services provision (e.g. in education and health care), such tasks may involve enforcing administrative rules, collecting fees, and operating purely commercial activities. Extrabureaucracies are most appropriately seen as the contracting arm of the core bureaucracy. 34
II. Bureaucracies generate self-financed income beyond budget allocations
In addition to receiving varying sums of budget allocations that fund basic wages and administrative costs, Chinese bureaucracies – by which, I refer collectively to core bureaus and extrabureaucracies – also generate what I dub self-financed income. Such income comes from several sources, e.g. fees and fines, user charges, and profits from commercialized operations. Self-financed income belong to the category of non-tax revenue in China. Some are classified as extrabudgetary funds.
III. Bureaucracies are empowered by policy awards to generate self-financed income
In order to generate self-financed income, e.g. to collect a particular fee, bureaucracies in China require “policy awards” from the state that authorizes them to earn revenue. Policy awards may be granted by the central ministries or by the local (usually provincial) governments. These are mostly official documents that detail the rights of particular departments in collecting or generating funds for a stipulated purpose. It is useful to think about policy awards as contracts awarded by the government to public bureaucracies to provide certain services in exchange for privileges to profit from office. I also extend the term policy awards to denote the clientelist relationship between core bureaus and extrabureaucracies. In generating self-financed income, extrabureaucracies typically need market protection from their parent agencies.
IV. Bureaucracies exercise de facto property rights self-financed income
Finally, I emphasize that individual bureaucracies exercise de facto property rights over revenue earned. Self-financed income is deposited into the state treasury system. Individual offices then apply to the finance bureau to spend their accumulated funds. Much of these funds go towards providing staff wages and benefits.23 Therefore, the more self-financed income is raised, the more benefits a bureaucracy can provide for its members. In this manner, each bureaucracy retains the rights to appropriate self-financed income, while the finance authorities have improved oversight over bureaucratic spending. My documentation here describes practices 23
I report systematic evidence for this claim in Chapter 4. 35
that have developed since the mid-1990s, when modernizing budgetary reforms were introduced. In the sections below, I shall discuss each of the elements in further detail, drawing extensively upon field interviews and observations. Before I proceed, let me briefly introduce the formal organizational structure in China.
The Formal Organizational Structure
The Chinese state is composed of five formal levels of government and three sets of hierarchies that are replicated at each level. The five levels are center, province, city, county, and township.24 The three sets of hierarchies are party, government, and military. The vertical bureaucracies are known as lines (tiao) and the horizontal bodies as pieces (kuai). The intersection of vertical and horizontal authorities forms a matrix structure (Lieberthal 2003). Following the conventional literature, the state bureaucracy in this study refers only to party and governmental organs, excluding the military and state-owned enterprises.25
China is termed a “party-state” because each level of government has a parallel set of party and government organizations. At the central level, the leading party organization is the central party committee, while the government bureaucracy is headed by the state council. At the subnational levels, the head of party and head of state corresponds to the local party committee (dangwei) and government office (zhengfu bangongshi). As Lieberthal stressed, “the party always exercises ultimate authority over its government counterpart” (2003, 172). The party exercises ultimate authority through the power of appointment. The organization department, a party organ with offices at every administrative level, appoints officials. The party secretary, who leads the party committee, is always the “first-in-command,” while the head of the governmental apparatus is typically the “second-in-command.”
24
As of 2000, China has 31 provincial-level governments (including Beijing, Shanghai, and Tianjin, which are municipalities of provincial status), 660 cities, 2,461 counties, and 44,867 townships (Lieberthal 2003, 180-4). 25 Bureaucratic organizations should not be confused with state-owned enterprises, although the latter is often seen as part of the “state.” In any government, public bureaucracies do not produce goods to sell on the market; they are engaged in public service functions. Moreover, whereas SOEs are businesses that pay taxes to the state treasury, public bureaucracies should in theory be fully funded by budget appropriations. 36
Figure 2.3: Organization Chart of Typical Local Government Party Committee Social Organizations
Association of Science
Federation for the Handicapped
Commerce Union
Workers’ Union
All-Women’s Federation
Communist Youth League
Political and Legal Affairs Committee
Strategy Department
Propaganda Department
Organization Department
Disciplinary Committee
Party Committee Secretariat
Government Office
Family Planning Bureau
Agriculture Bureau
Water Affairs Bureau
Transportation Bureau
Environmental Protection Bureau
Social Security Bureau
Construction and Planning Bureau
Finance Bureau
Legal Affairs Bureau
Civil Affairs Bureau
Public Security Bureau
Ethnic Affairs Office
Science & Technology Bureau
Education Bureau
Trade Commission
Development and Reform Commission
Figure 2.3 illustrates a typical formal organizational chart of a county government. The county is headed by the party committee, followed by the government office. Generally, party organs control organizational, personnel, and propaganda functions. Governmental organizations are responsible for administration, economic regulation, and social services provision. Organs like the organization department and the disciplinary committee fall under the party committee. Also included in the party apparatus are state-directed social organizations, such as the communist youth league and national workers’ union. These organizations, though “social” in name, are conduct political and ideological work for the CCP and are therefore part of the party apparatus. Governmental organs include the finance bureau, civil affairs bureau, education bureau, legal affairs bureau, development and reform commission, and so forth. Appendix A provides a full list of county offices.
The formal organizational structure is replicated from the central to county levels. Each local office shares a “professional relationship” (yewu guanxi) with the corresponding line authority at the next higher level and a “leadership relationship” (lingdao guanxi) with the local government. The upper line authority provides guidelines on administering a particular sector. For example, the ministry of education issues guidelines on fee collections that percolate through the hierarchy to the local education bureaus for implementation. However, it is the local government, headed by the Party Committee and Government Office, which appoints 37
officials to the respective offices and manages their finances. In most governmental sectors (xitong), with the exception of a few vertically-managed offices, horizontal/leadership relations dominate vertical/professional relations (Lieberthal & Oksenberg 1988).26
At the county levels and above, there exists fairly clear boundaries between offices; each office administers a particular sector and is headed by a distinct group of executives. Township governments, the lowest-tiered formal administration, have a simpler administrative structure, with fewer offices that each take on multiple functions. In addition, townships have a much smaller staff size than county and city governments. From the 1980s onwards and especially after the 1994 fiscal reform, township governments, like bureaucracies at the upper levels, relied heavily on extrabudgetary financing, in particular the exaction of fees and surcharges. Starting in 2000, the central government introduced rural tax reforms, beginning with the taxfor-fee reform (feigaishui) (Yep 2004; Kennedy 2007). In 2005, the state council announced plans to abolish the agriculture tax nationwide. My analysis that follows, describing the behavior and organization of individual offices, applies primarily to levels of government at the county and above.
Element (I): Bureaucracies and Extrabureaucracies
Existing analyses of the Chinese bureaucracy have focused either on local governments as a homogeneous whole or various offices that appear on formal organizational charts. But it is useful to disaggregate the bureaucracy and look inside it. Each office is composed of two composite parts: a small core bureaucracy and a sprawling periphery of extrabureaucracies. Literally translated from Chinese, these organizations are “administrative units” (jiguan danwei) and “service units” (shiye danwei). About 80 percent of public employment in China is concentrated in the extrabureaucracies. Thus, shiye units are rightly described as “a big shadow of the Chinese state” (Lam and Perry 2001, 20).
26 26
This is except in the case of “vertical management” (chuizhi guanli) (Mertha 2005). Since 1998, several offices ordered by the central government to report to the upper-level administrative authority (tiao) instead of the local government (kuai). Examples of vertically-managed offices include the customs office, commerce bureau, weather offices. The majority of offices in China are still locally managed. See Appendix A for a list of local governmental offices. 38
Administrative units are those bureaucracies represented on formal state organizational charts, as seen in Figure 2.3. As their name suggests, these bureaucracies perform the tasks of administration and regulation. Administrative units include party organs (e.g. organization department), governmental organs (e.g. finance bureau, legal affairs bureau, education bureau), policing departments, as well as courts and procuratorates. Functionally, most administrative units in China are not substantially different from state agencies found elsewhere. In this analysis, I will refer to the administrative units as “core bureaus” or sometimes “parent bureaus.”
Compared to the jiguan danwei, the shiye danwei is an ill-defined and oft-misunderstood entity. The term shiye danwei has been translated variously into “business units” (Barnett 1967), “institutional work units” (Cheng 2001; Wei-Arthus 2000), “government-funded not-for-profit organizations” (Yang 2004), “a galaxy of public service providers alongside core government” (OECD 2006, 81), “nonprofit and nonregulatory” units, and “public service units” (World Bank 2005). I choose to translate shiye danwei analytically as extrabureaucracies.27 This translation captures the principal feature shared among all shiye danwei – they are attached and subservient to a core bureau. Each core bureau supervises a group of extrabureaucracies and exercises micro control over the latter. To illustrate, Table 2.1 lists the bureaus and extrabureaucracies in a typical local (provincial, city, and county) government. For the remaining analysis, my use of the term bureaucracy/office/department implies a collective unit of core bureau and extrabureaucracies.
In theory, extrabureaucracies should not have regulatory powers or profit-making motives (Cheng 2001). But it is important to separate theory from reality. As Lam and Perry aptly described, extrabureaucracies “only provide services to their administrative bosses” (2001, 27). Those services may be public or private in nature. In practice, extrabureaucracies may administer, organize public services, run businesses, or all of the above (see Table 2.1 for
27
The term “extrabureaucracies” was earlier used by Chinese historians to describe local elites and tax farmers who performed state services but who were not formal civil servants (Rankin 1993). I review the historical literature in Chapter 5. 39
examples).28 Extrabureaucracies include conventional public service providers like public schools and public hospitals. They also include amorphous entities engaged in administrative enforcement and semi-or-pure commercial activities, e.g. price assessment center and driving schools. Table 2 in Appendix A lists the extrabureaucracies under selected agencies in a local government.
Table 2.1: Examples of Functions and Fiscal Status of Extrabureaucracies (Shiye Danwei) Administrative Functions (xingzheng guanli) Party member center (full) Cement office (self) Office for the reform of construction materials (self) Office for legal verification of documents (self) Labor market center (full) Agency for the inspection of construction safety (self) Environmental inspection team (full) Price assessment center (self)
Organize Public Services (youchang/wuchang gongyi) Public schools (full) Public hospitals and clinics (partial) Disease prevention center (full) Old folk services center (self) Natural gas company (self) Street lights management bureau (partial) X park management bureau (self) Sports center (partial) Sports association (full)
Runs Businesses (yingyexing) X city conference center (self) (run by the trade bureau) Center for driving instruction (self) (by transportation bureau) X city resort (self) (by housing bureau) Car rental fleet (self) (by reception bureau) Translation services center (self) (by foreign service bureau)
Limited Scope of the Civil Service System
Now that we have drawn the divide between the core bureaucracy and extrabureaucracies, we may better understand the scope and implications of the Chinese civil service system. The Chinese government enacted the civil service law in 2005, thereby establishing a formal civil service. The law defines the functions of civil servants, guarantees them the right to receive state-budgeted wages, and delineates rules on recruitment, assessment, promotion, dismissal, as well as other legal rights and obligations.
28
Take for example the legal assistance center, a shiye danwei under the legal affairs bureau. The legal assistance center provides public services in the form of legal assistance for the poor. At the same time, it exercises administrative power by requiring law firms to perform pro bono services, an essential condition for licensing approval. To earn additional income, some legal assistance centers may offer litigation services for a charge (AI 2007-35). 40
While the civil service law presents a major breakthrough in China’s bureaucratization efforts, one would be mistaken to think that China has successfully established a regular salaried bureaucracy with a piece of legislation. The civil service law remains limited in scope. It applies only to those cadres with an official administrative position (xingzheng bianzhi), personnel who usually hold leading positions in the bureaucracy. Only these cadres are titled civil servants (gongwuyuan). The remaining employees are known as service-unit employees (shiye danwei renyuan). Nationwide, I was told that there were only about 600,000 civil servants out of 48 million public employees in 2005 (AI 2007-138). 29 The central government imposes a strict quota and entry requirements on the civil service, including competitive entry examinations.
Extrabureaucracies, State-Owned Enterprises, and Bureau-Operated Companies
Extrabureaucracies should be distinguished from state-owned enterprises (SOEs), collective enterprises, and township and village owned enterprises (TVEs). The latter are state-owned firms involved in production and are supposed to generate revenue. They do not receive budget allocations from the government (although they may receive financial bailouts when losses are incurred); instead, they are required to surrender their profits to the level of government at which they operate to finance public expenditure. In contrast to the SOEs and TVEs, extrabureaucracies are public organizations engaged in services provision. In theory, they should, like public providers elsewhere, receive regular budget allocations from the state and not pursue profits.
Extrabureaucracies are not the same as bureau-operated companies, although the two are sometimes confused in the literature.30 Bureau-operated companies proliferated in the 1980s and early 1990s (Duckett 1998). Shortly after the central leadership signaled the green light to “get rich,” agencies and extrabureaucracies plunged into a mad rush to set up their own corporate entities, variously named “jingji shiti (economic subsidiaries), gongsi (companies), or 29
The total number of public employees is reported in the Local Public Financial Statistics (Difang Caizheng Tongji), published by the Ministry of Finance. 30 For example, Duckett observed that the ‘property companies’ within the real estate management bureau in Tianjin “have existed for years as ‘administrative companies,’ meaning that they are non-profit making units that function as part of the state administration” (1998, 107). Given her description, the company she was referring to was probably a shiye unit, rather than a bureau-operated firm. 41
sanchan (tertiary sector undertakings)” (Lin & Zhang 1999, 205). Lin and Zhang (1999) gave these entities the fitting title of “backyard profit centers.” Examples of bureau-operated firms cited in the literature included restaurants, canteens, dance halls, convention centers, printing shops, small trading companies, and department stores (Duckett 1998; Blecher 1991; Park et al 1996; Lin and Zhang 1999). The particular phenomenon of bureaucracies going into business is commonly labeled state entrepreneurism (Blecher 1991; Duckett 1998; Duckett 2001).
Importantly, the kind of entrepreneurism seen in “bureaucrats in business, Chinese-style” (Duckett 2001) should be distinguished from entrepreneurism under bureau-contracting. First, bureau-operated businesses were a new product of the reform era. These companies belonged to the non-state sector, even though they were invested by party-state organs at the township levels and above. As Duckett pointed out, “they have been neither planned nor anticipated by central government policy makers” (2001, 23). Extrabureaucracies, in contrast, were an essential part of the state even under the command economy. They have been existed, alongside core agencies, since the 1950s (see Chapter 5 on the state structure under Maoist rule). Second, all extrabureaucracies operate only in the services sectors regulated by their parent bureaus, whereas corporate entities could be engaged in any business and sector, including in manufacturing.31 As Duckett noted, “sometimes the businesses [established by state organs] were related to the administrative work of the parent department, but many conducted unrelated business” (1998, 26, emphasis added). The distinction drawn here is analytically important because, as Duckett observed, bureau-operated companies were “profit-making, risktaking businesses” that could make losses (1998, 14). Extrabureaucracies, on the other hand, do not normally incur losses. Most of them enjoy competitive advantages or even monopoly rights vis-à-vis non-state services providers.
Although bureau-operated businesses prevailed in the earlier stages of reform, most have gone bust or been divested in the post-1990s period (Yang 2004). One reason for the widespread closure of these businesses was central state policy. The former premier Zhu Rongji made it a priority during his tenure to divest party-state organs of their businesses, as they were seen to
31
As emphasized in Chapter 3, extrabureaucracies are the contracting and not just revenue-making arm of the core bureaucracy. 42
hamper market development and feed corruption. Another reason for the demise of bureauoperated businesses was increased market competition. State bureaus have few strengths compared to private entrepreneurs in running businesses beyond their exclusive regulatory fields.32 As Hubbard (1995) rightly predicted, bureau-operated businesses will decline over time “as market development reduces the competitive advantage of government agencies in business” (350).
Yet, although bureau-operated businesses have largely evaporated from the market, we would be mistaken to think that entrepreneurism in the Chinese bureaucracy is a thing of the past. Bureaucrats in China remain highly entrepreneurial. In my interviews with them, Chinese officials are constantly strategizing for new ways to finance their offices.33 I would argue that from the mid-1990s, monetary strategies in the bureaucracy have shifted from running backyard profit centers to bureau-contracting. I would argue that generating revenue through extrabureaucracies and by extracting monies in accordance to state policies has become the dominant and smarter strategy for bureaus to “get along and ahead” financially. In the next section, we proceed to introduce these monetary strategies. The documentation that follows is based on the most recent institutional developments.34
Element (II): Budget Allocations vs. Self-Financed Income
Both in theory and in practice, Chinese bureaucracies are not purely salaried. In theory, core bureaus (jiguan danwei) should be fully state-funded units, while extrabureaucracies (shiye danwei) are divided into three fiscal categories: fully-funded, partially-funded, and self-funded. Fully-funded (quane bokuan) units receive full state funding for their basic budgetary needs; partially-funded (chae bokuan) units receive some financial subsidies; and lastly, self-funded
32
Lacking management expertise, many agencies could not generate sustainable profits from restaurants, dance halls, and small factories. For example, many factories operated by public schools have closed down because they were operated by school principals and teachers who were not professional entrepreneurs (AI 2007-123). 33 To cite another example, in a case study of China’s party schools, organizations that are supposed to provide ideological training for cadres, Lee (2009) reports that money-making remains a high priority for the leadership. 34 I conducted fieldwork from 2006 to 2009. 43
units (zishou zizhi) receive none. Table 2.2 shows examples of extrabureaucracies by funding type.
Table 2.2: Cross-Tabulation of Public Employees, 2003 (Percentage over total indicated in parenthesis)
Core bureaucracy (jiguan danwei) Extrabureaucracies (shiye danwei) TOTAL
Receive state funding 8,926,450
Entirely Selffunded 96,733
35,199,835
2,312,272
44,126,285 (94.8%)
2,409,005 (5.2%)
TOTAL 9,023,183 (19.4%) 37,512,107 (80.6%) 46,535,290
Source: Local Public Financial Statistics
Budget allocations usually cover only essential budgetary needs, not benefits and frills. As one finance bureau official expressed colorfully, “budget allocations are used to deliver coals during snow, not to add blossoms to silk” (AI 2008-154). Hence, being “fully-funded” does not mean that all of a unit’s expenses are covered by state revenue. Typically, “full state funding” covers basic operational costs and staff wages.35 Thus, even units that are fully-funded in name, like public schools and regulatory agencies, are not relieved of financial stress. Furthermore, it is not unusual for units that are fully-funded or partially-funded in name receive no budget allocations in fact. For example, I encountered a city-level tourism bureau that was supposed to be a fullyfunded regulatory agency, but was in fact entirely self-financed until 2001 (AI 2007-108). Similarly, a metallurgy institute was entitled to partial state subsidies, but was forced to find its own means when the metallurgy bureau, its parent agency, was abolished (AI 2007-111).
For bureaucracies in China to give out staff bonuses, allowances, and benefits not financed by the local government, or to buy cars and construct new office buildings, they would have to find their “own money,” as we saw in the example of Blossom county’s land bureau. The “own
35
Basic cadre wages are standardized across the country and hence low in well-off places. Chapter 3 explains the wage structure in further detail. 44
money” refers to revenue earned by and belonging to individual bureaucracies. For short, I label them self-financed income.
Sources of self-financed income
There are several different sources of self-financed income. (i) Administrative and regulatory fees, (ii) fines, (iii) revenue from governmental foundations (jijin) that require contributions from businesses, (iv) user charges from services provision (shiye shouru), (v) profits from commercial activities (yingye shouru). Examples of shiye income are school fees, medical charges, and ticket sales from performances. These are revenue collected from providing governmental services. Yingye income may include revenue from leasing a piece of state property. In practice, the line between shiye income and yingye income is blurred.
Self-financed income, non-tax revenue, and extrabudgetary funds
Self-financed income are a part of non-tax revenue category. The two major categories of revenue in the Chinese fiscal system are tax revenue (shuishou) and non-tax revenue (feishui shouru). Non-tax revenue is sometimes synonymous with extrabudgetary funds, but a distinction needs to be made between the two. Non-tax revenue is classified as withinbudgetary funds (yusuan shouru) if they are deposited in the state treasury account (guoku zhanghu) at each level of government and managed through regular budgetary procedures. Non-tax revenue is classified as extra-budgetary funds (yusuanwai zijin) if they are deposited in the extrabudgetary special account (yusuanwai zhuanhu) and are not managed through regular budgetary procedures, e.g. the budget does not have to be submitted to and approved by the state legislature.
A common source of confusion in analyzing Chinese fiscal data is that budgetary categories have changed over time and constantly. It is not surprising to find that budgets are reported differently across provinces and levels of government. For example, at Blossom county, finance officials no longer speak of within-budgetary versus extra-budgetary funds. To them, the appropriate terms are tax revenue and non-tax revenue (AI 2007-115; 2007-116; 2007-117; 45
2007-118). However, in the finance yearbooks published by the ministry of finance and th7 Shandong financial bureau, the budget is still divided into within-budgetary funds and extrabudgetary funds (for example, see Caizheng Nianjian in the various years).
For non-experts, it suffices to know that self-financed income collected by individual bureaucracies at each level of government are a subset of non-tax revenue. They may be classified as within-budgetary or extra-budgetary funds, depending on the rules periodically issued by the ministry of finance. It should also be noted that not all non-tax revenue belong to individual bureaucracies. Some sources of non-tax revenue belong collectively to the local governments, e.g. township and village self-raised funds and proceeds from leasing state land.
Interestingly, it appears that all bureaucracies make some self-financed income, even in localities with well-endowed finances. When I asked a local finance official whether there were any purely state-financed bureaus in the county, he responded (AI 2007-115):
“Pure” ones? Almost none… Organizations in the judicial system [e.g. courts] should be pure… but wait… actually, even they are not completely pure. Well, then are those in the party committee and local governmental office; their financial needs are guaranteed… But then again, those are not entirely pure too. So how do bureaucracies generate self-financed income? Importantly, the central and local governments have an interest in seeing that the bureaucracies can be sufficiently self-funded without resort to extortion or illegal activities. Hence, the state assigns bureaucracies policy awards – that is, revenue-making privileges. As one official expressed, “Even if [the state] does not give them [the bureaucracies] money, it will give them an avenue to make wealth” (AI 2007129). In the next section, we will discuss what these avenues are and how they function.
Element (III): Policy Awards
As a popular saying in China goes, “the state may not be able to grant money, but it can grant policies” (geizhengce bugeiqian). What does this mean? A local official fittingly explained, 46
“Giving policies and not money happens when the government wants to get something done but budgetary funds are insufficient… Whether it is to give money or policies, the objective is to serve the people… But of course we cannot rule out the possibility of individual departments seeking their own interests” (AI 2008-153). To put it simply, policy awards are privileges assigned by the government to various departments to generate funds in lieu of budget allocations. Importantly, such funds are generated in the name of public services provision, as the quote above emphasized; they should therefore be distinguished from illegal monies (Lu 2000a, 2000b) and profits earned from business operations unrelated to party-state functions (Duckett 1998).
Policy awards can come from three sources: i) the central governmental authority (usually the State Council), (ii) central-level Ministries, and (iii) local governmental authorities. Central-level ministries and provincial governments enact policy awards that will apply at all levels of government. Sub-provincial governments can then “employ [policy] decisions by upper levels with flexibility” (AI 2008-155). For example, the Anhui provincial government allowed public schools to admit students on a tuition-paying basis, but the condition was that fee-paying students could not compose more than 70 percent of the student population (AI 2008-153).36 A local official explained, “Privately admitted students pay more fees… This is one kind of policy… Revenue collected from this policy goes to the public schools, but it needs to be within a zone approved by the upper levels” (AI 2008-153). In another example, the Jiangsu provincial government allowed tax bureaus to earn commissions from tax collections. The county governments may then adjust the commission rate based on local economic conditions (AI 2007-112). It appears then that policies devised by upper level authorities powerfully shape local bureaucratic behavior. Although many depict local agents as persistently defiant of central policies, local authorities reflected that they are in fact “constantly waiting for instructions from above” (AI 2007-52).
I extend the concept of policy awards to capture revenue-making ties between core bureaus and extrabureaucracies. Core bureaus and extrabureaucracies share a special kind of clientelist
36
Tuition-paying students were initially capped at 70 percent of the student population; the ratio has recently been reduced to 30 percent (AI 2008-154). 47
relationship. In the original anthropological sense, clientelism is “an exchange relationship… a special case of dyadic (two-person) ties… in which an individual of higher socioeconomic status (patron) uses his own influence and re- sources to provide protection or benefits, or both, for a person of lower status (client) who, for his part, reciprocates by offering general support and assistance, including personal services, to the patron” (Scott 1972, 92).
In the Chinese state, the core bureau is like a patron and the extrabureaucracies the clients. The two entities share a hierarchical relationship. Core bureaus exercise absolute control over their extrabureaucracies: the former appoints leading positions and mid-level cadres in the latter; revenue from the state is allocated by the supervising agency to extrabureaucracies; and the core bureau assigns tasks to its subsidiaries (Lam and Perry 2001, 35). Simultaneously, the two entities share an exchange relationship. Often, parent bureaus provide their extrabureaucracies advantageous market access, while extrabureaucracies share their revenue with the parent agency.
Appropriately, as relations are so tightly enmeshed, leaders of extrabureaucratic units often refer to their supervising agency in familial terms as “father,” “mother,” or “mother-in-law” during my interviews with them. Uniquely, however, the clientelist relationship described is not shared between individuals, but rather between organizations in the same hierarchy. That is to say, even if a particular bureau chief were replaced, the core bureau and extrabureaucracies would continue to share the same patronage ties. That implies that the relationship between core bureaus and extrabureaucracies is not the result of individual manipulation, but rather an enduring feature of the Chinese state structure.
In the following sections, I will give examples of four major policy awards. Taken together, they illustrate how bureaucracies in various sectors generate self-financed income in accordance to state policies, as well as the clientelist relations between core bureaus and extrabureaucracies in self-financing.
•
Policies sanctioning the exaction of fees and fines
•
Monopoly privileges 48
•
Permission for public providers to sell their services and make profits
•
Permission for public units to form and generate revenue from spin offs
Policies sanctioning the exaction of fees and fines
Bureaucratic predation has been a long-standing problem in China since reforms began. State agencies were infamous for the “three arbitraries” (sanluan): arbitrary taxation, arbitrary fines, and arbitrary expropriation (Wedeman 2002, 492). Lu observed that agencies typically exacted payments “without clear legal mandates” (Lu 2000a, 276). In township and village governments, sanluan was a major source of peasant burdens and protests (Bernstein & Lu 2003). These earlier studies rightly described problems that prevailed in the 1980s and 1990s, particularly after fiscal recentralization in 1994 reduced local governmental revenues.
The good news is, in the last decade, many rationalizing institutional improvements have been made in the Chinese administration, as documented in Yang’s (2004) study on central reformers’ efforts to “remake the state.” To be clear, fee-and-fine taking has not been abolished in China; in fact, fees and fines remain an important source of self-financed income for state bureaus. But, as I observe during recent fieldwork, fee-and-fine taking has become increasingly rules-based, regulated, and even quite sophisticated, including in the townships. This is especially so following the enactment of the administrative licensing law (xingzheng xukefa) in 2004, which provides a national legal framework for the collection of regulatory fees. As these local officials reveal:
Nowadays fee-taking by various departments is very strictly regulated. The price bureau and respective departments have to issue a fee taking license (shoufei xukezheng), indicating the fees that can be taken and the ones that cannot (AI 2008-154). The [township] governments have undergone tremendous change. Nowadays arbitrary fee taking and arbitrary fines are gone; fee taking activities have been regularized… In some instances, public bureaus, for example, the electricity bureau, even have to enlist the assistance of courts in collecting fees (AI 2008146).
49
As local officials explained, policy awards sanctioning the collection of fees and fines range from “soft” to “hard” (AI 2007-127; AI 2007-128). According to the administrative licensing law, the hardest policies are central and local licensing (xuke) provisions. Licensing provisions must have a clear legal justification for collecting fees and fines. Licensing provisions require a one-year probation period before they are deliberated at the central or local state legislature (i.e. the People’s Congress). These provisions become permanent only after they are passed. The next softer set of provisions are assessment (shenpi) provisions. These are provisions not yet passed by the legislature but are still legally valid. Last but not least, the softest policies are “redheaded documents” (hongtou wenjian), named after the bright red department stamp on the letterhead of official Chinese documents. These are rules issued by regulatory agencies without higher-level or legislative endorsement. “Some departments issue red-headed documents, asking for this or that matter to be approved by them. This is fine, but it cannot be brought on the table because it does not have an explicit legal basis” (AI 2007-128).
The best example of a hard policy award backing the collection of fees and fines is the notorious case of the steamed bun offices in Zhengzhou city of Henan province. Zhengzhou city established a steamed bun office at the city level and five in each of the county governments below. And what do steam bun offices do? City authorities articulated a need to establish specialized offices to manage steam buns production. The city government issued Provision No. 93, titled “Temporary Provisions on Zhengzhou City’s Steamed Buns Production and Sales Management.” The provision was passed by the city government’s 14th People’s Congress and signed by the mayor.
Article 1 proclaimed that “this provision is established in accordance with our country’s laws and regulations, as well as our city’s practical conditions.” Its lofty goals were “to strengthen the production and sales management of steamed buns, to promote their mechanized production, to guarantee the health of our people, and to make the lives of our people more convenient.” The rest of the provision proceeded to spell out two basic functions of steamed bun offices: to issue a steamed bun production permit and to fine producers for not possessing the permit. The legal provision carefully detailed the amount and purpose for which each fee or fine is imposed.
50
According to the provision, one could be fined from 3,000 to 20,000 yuan for producing steamed buns without a permit.
The story of Zhengzhou’s steamed bun offices usefully illustrates the network of rules that authorize fee collections. In order to exact fees and fines, local offices require endorsements from multiple formal institutions. To legalize fee and fines collection, the local leadership must enact and pass a state provision through the legislature. To establish a specialized office and hire employees, the establishment office has to give consent. The price bureau has to issue a fee taking license. Without these policy awards, the steamed bun office would not be in business.37
Monopoly privileges
When we think about public monopolies, big monopolies like telecommunications and oil come to mind. However, small-scaled public monopolies have been overlooked. Many extrabureaucracies in China thrive on having considerable market advantages or monopoly rights in the supply of public and commercial services.38 Some examples of such services are utilities supply, greening, tour agencies, media publications, private security, administration of examinations,
and
environmental
impact
assessment.
Extrabureaucracies
feed
on
“administrative protection” (xingzheng baohu) provided by the parent bureaus. Such protection may come in the form of explicit legal provisions or through the covert exercise of political influence (for case studies, see Lin and Zhang 1999; Lam & Perry 2001).39 I illustrate my findings below.
37
The steamed bun office is not an isolated case. There was also the notorious case of the watermelon offices. See “Steamed bun offices, watermelon offices… are all troublesome offices,” Xinhua, August 4 2006.” 38 Some bureau-operated businesses also enjoyed monopolistic advantages (Lin and Zhang 1999). But, as Yang (2004) reported, many of these businesses, especially those belonging to the military and law enforcement organs, have been forcibly closed by the central government. 39 Other studies also reported examples of administrative bureaus extending monopoly privileges to extrabureaucracies. The personnel ministry and local personnel bureaus grant the talent exchange centers the exclusive right to process visa applications and the examination centers the lucrative business of administering professional examinations. The environmental research institute under the environmental protection bureau is one of the few organizations authorized to conduct environmental impact assessment, and “due to its special relationship with the bureau, it naturally enjoys an unrivaled advantage in the competition for business” (Lam & Perry 2001, 34). 51
Heating supply is one example of a public monopoly service. Heating services for residential and commercial properties must be centrally supplied. According to the “The Tianjin City Centralized Heating Management Regulations,” issued by the city government, the local heating offices plans and organizes the supply of heat. In one particular county, the construction bureau (the parent bureau) described the heating office as its “the greasiest” extrabureaucracy. The heating office collects 20 yuan for every square meter of property for which heat is supplied. When asked about the profits of the office, the bureau official responded nonchalantly, “Yes [they can make hundreds of thousands]. Even 10 million dollars a year is not a problem” (AI 2008-144). The interviewee added that the forestry bureau, with its subordinated greening offices, is just as “greasy” because “[the service of] greening roads and highways belongs entirely to them” (AI 2008-144). Another described the monopoly situation as follows, “The forestry bureau regulates greening. It can say ‘you cannot touch this tree,’ but then they can touch it themselves” (AI 2007-106).
Even in the tourism business, extrabureaucracies commanded market advantages. In Jiangsu province, a city-level tourism bureau was an entirely self-funded agency up until 2001. Since its establishment, the tourism bureau had lived on remissions of revenue from its extrabureaucracies, which included tour agencies, hotels, car, and boat rental companies. Up until the late 1990s, public units in the tourism sector were profitable because there were plenty of new market opportunities and limited competition simultaneously. As the head of the tour agency explained:
We still continue to share revenue with our parent bureaus… We give it to them directly, and then they refund it to us… There is some degree of monopoly involved. For example, the tourism bureau approves tour agent permits [author’s note: and by implication, it can decide to whom to give these permits]. Sometimes, administrative agencies exercise “soft” influence, rather than hard regulatory power, in supporting their extrabureaucracies’ business. I give an example from the education sector. In a county of Jiangsu province, the education bureau had recently established a youth science center, a science expo for children. The county state budget paid only one-third of the center’s costs. The remaining income came from fee-charging extracurricular courses offered by Center. The Center charged 200 yuan for class and paid its instructors 50 yuan an hour, less than the 52
market rate. Surprised by the low fees and teachers’ pay, I quizzed the director during dinner about his pricing strategy. The director explained that his plan was to offer high-quality courses at a low price.
Still, another question remains: why would good teachers choose to work at his center if the pay was not competitive? The director responded that he had already “made a verbal pact with the public school principals to send their best teachers” to the Center (AI 2007-121). “[The teachers] have to give the principals face, right?” In fact, I was told that at another nearby county, public teachers were mandated to teach at the local Youth Center for only 20 yuan an hour and were barred from offering classes at home (AI 2007-122). “We are more flexible,” said the director, “the teachers can teach at our place and build up their reputation, and then they can hold private lessons.” The vice-director chimed in, “The teachers gain both reputation and money (mingli shuangshou).” Beaming with pride, the center director unveiled his true business strategy:
The Culture Center [another extrabureaucracy under the Cultural Bureau] and the Joyful Tots Education Company hate us to death! Our two thousand students all came from them! Right now, I can depress the prices, and when these companies disappear, I can raise the prices.” The education bureau head, who was also at the banquet, added that “Mr. Li [the director] has a hard purse” and “the state treasury behind him” (AI 2007-119). That means, should the director need aid, the education bureau would not abandon him to his own means. Hearing these words, executives of the youth center immediately toasted the bureau head, thanking him profusely for his support. The hidden script is that without the bureau’s backing, the school principals would not have offered the center their best teachers. In fact, without the bureau’s patronage, many things could not have been done at the center.
Permission for public providers to sell their services and make profits
Public service providers, like public hospitals and schools, are supposed to provide essential social services free or affordably. Unfortunately, that has not been the case in China’s public health sector. In the post-Mao period, the public health care system has been financed by the 53
strategy of “giving policies not money.” Under socialism, health care was heavily subsidized by the government. The year of 1985 marked a turning point. The State Council approved and issued a report by the Ministry of Health, entitled “Notice on the Policies of Health Care Reform.”40 The report emphasized the need to incentivize health care providers and to push them to become “financially self responsible.” It wrote, “the prices charged for the provision of health care services have been too low. This is not advantageous for the development of health care services… [henceforth] public providers may set fees to reflect their costs.” In effect, the 1985 document gave public hospitals and clinics the green light to charge consumers and make profits.
After granting health care providers the right to charge fees, the state began to cut back on budgetary allocations for the health care sector. In Jiangsu province, the proportion of public wages funded by budget allocations fell from 60 percent in 1985 to 31 percent in 1988. At the national level, state budget allocations as a share of total revenue in the health care sector dropped from one-third in 1980 to one-quarter in 1990.41 In recent years, the reliance of public health care providers on user fees has grown even sharper (World Bank 2006). For example, a city-level hospital in Chongqing I had interviewed was “weaned” of all state funding and became entirely self-funded by 2007 (AI 2007-32). Ironically then, this was a major public service provider with no public funding.
Forced to finance themselves, public health care providers turned to their most important source of revenue: selling drugs and medication. Public hospitals could mark up the price of drugs by 15 percent. Why not improve medical services and charge higher fees instead? The answer is that the central government did not authorize charging high fees for medical services. And why not? As one interviewee explained, the formal wages of white collar workers, including medical providers, are still suppressed in China. “This has to do with communist ideology. If a doctor makes thousands of dollars more than an ordinary worker, how is that acceptable?” (AI 2007-46). 40
“Notice on problems related to health care reform policy” (Guowuyuan pizhuan weishengbu guanyu weisheng gaige ruogan wenti baogaode tongzhi), Report by the Ministry of Health and approved by the State Council, Issued April 25 1985. 41 “1985: The Year Health Care Reforms Began,” Zhejiang Health Bureau Website, http://www.zjwst.gov.cn, accessed on July 1, 2008. 54
The particular combination of incentives and restrictions produced a perverse strategy of revenue maximization: maximizing the volume and sale price of drugs. This strategy is described by some health care experts as “provider-induced overconsumption” (Zhang and Gu 2006) or more popularly known as “using drugs to feed public health” (yiyao yangyi). As a result of this revenue-maximizing strategy, Chinese patients have found health care prohibitive, not primarily due to expensive services, but because doctors too often prescribed more medicine than needed.
Public hospitals not only charge fees and profit from selling drugs, they enjoy an implicit degree of protection from competition. Private hospitals in China are few compared to public providers. 42 There are two reasons for the lackluster development of private health care providers. One is difficulties faced in obtaining required licenses, and two, unlike public providers, private providers are typically not included in the national health insurance system. Consumers are unwilling to seek medical services in private hospitals if their expenses cannot be reimbursed. The head of a private hospital in Beijing lamented:43
Private hospitals, like public hospitals, carry the mission of saving patients. We provide the same quality services and suffer the same burden of owed payments as public hospitals. But we do not enjoy the state’s beneficial policies. And we even have to pay taxes. These have sharpened the challenges of survival and development for private health care providers. In fact, local health officials may have interests fending off not only private providers, but also public providers from other well-off Chinese provinces. Fearing competition, health bureau officials in a Jiangsu city barred leading public hospitals from Shanghai from setting up a branch in their territory. A local official conceded that it was no wonder that the quality of medical services has lagged behind the city’s economic prosperity (AI 2007-104). Permission for public units to form and generate revenue from spin-offs
42
Nationwide, the number of private service suppliers (minban feiqiye) are one-tenth that of public providers (World Bank 2005). 43 “Hospital executives urge government to create health assistance fund,” Zhongguo Qingnianbao, January 15 2008 55
With the enactment of the nine-year compulsory education law in 1986, public elementary and middle schools in China should be fully state-funded and are therefore not allowed to collect tuition. But in most localities, there are insufficient funds to finance all public schools. On top of that, high-ranking “key-point schools” (zhongdian xuexiao) face pressures from the leadership to build new campuses, and local governments provide limited financial support for these “luxury” goals. Confronting these circumstances, a set of policy awards emerged in the late 1990s that authorized public schools to earn additional revenue without having to collect tuition.
The first strategy for revenue-making is to collect “premiums,” i.e. additional payment for premier education services. Such premiums take two forms. First, the “school leasing fee” (jiedufei), paid by students who are not formally registered as residents of the locality.44 Second, the “school selection fee” (zexiaofei), paid by students who seek to gain admission into a particular school, but who lack a sufficiently high entrance examination score or live outside a designated geographic range. These premiums run from several thousands to tens of thousands a year. For example, a key-point elementary school in Chengdu city of Sichuan province charged a “selection premium” of 45,000 yuan for six years of study, a substantial sum for the average household. Yet, even at this high price, parents “elbowed to gain admission” into the key-point schools (AI 2007-44).
While it appears that the policies on premium collections are usually local knowledge, they are sometimes publicized by the local government. In Hefei city of Anhui province, the city government issued a public statement that set a ceiling on the collection of premiums and tuition by public schools.45 Elementary and middle schools may collect 400 and 600 yuan respectively each term in “school leasing fee.” Public high schools may collect up to 3,500 yuan for “school leasing” and 3,000 yuan for “school selection” each term. The statement also listed the titles of the documents issued by the price bureau approving the collection of those fees. A second strategy of revenue-making is for key-point schools to spin-off private schools. These have been called “restructured schools” (gaizhi xuexiao), “privately run, publicly assisted
44
It is still not easy for ordinary citizens to transfer their household registration status (hukou) in presentday China. 45 “Hefei City Education Bureau’s List of Fee Collections,” Hefei Education Network (Hefei Jiaoyu Xinxiwang), http://www.hfjy.net.cn/, accessed on April 21, 2008. 56
schools” (minban gongzhi), “a school within a school” (xiaozhongxiao), “private schools run by famous schools” (mingxiaoban minxiao). And what are they? In short, an interviewee explained, “[public] schools that charge fees!” She added, “the nine-year compulsory education policy forbids public schools from charging fees, but these schools are exempted from this policy” (AI 2007-44).
Here is an example (AI 2008-148). Mega High is the best public school in a county of Tianjin. In 1998, Mega established Stellar School, a private school admitting middle and high school students at a charge of 3,000 yuan each year. With its affiliation to Mega, the spun-off private school enjoyed a steady recruitment. The familial relationship between the two schools was evident. The director of Stellar was the principal of Mega. Stellar shared its revenue with Mega. The two schools shared the same teachers, who receive budgetary allocations from the city government for their wages. They also shared the same campus, with Stellar fenced up to separate the two schools symbolically. Lastly, they shared the same boss – the education bureau. “As you know,” an interviewee summed up, “China’s so-called private schools are different from those abroad. There are private schools abroad, but all of ours have a state character” (AI 2007151).
Political intervention lies behind the hybrid strategy of school financing. The ability of key-point schools to generate revenue is premised upon an unequal distribution of resources across public schools in a district. The more unequal the quality of local public schools, the more parents are induced to pay premiums to enter into good ones. The local education bureau arranges for keypoint schools to pick the best teachers and students before others (AI 2007-44). The advantage of key-point schools allows them to attract a huge demand for admission, which leads in turn to over-population and demand for larger campuses, a situation that one principal ironically calls “a virtuous cycle” (AI 2006-23).
Local governmental leaders also pressured school leaders to conduct lavish remodeling projects, as they can claim credit for investing in public education. Thus, key-point schools are often forced to take on sometimes unsustainable loans. An executive in a key-point school revealed, “We took on a loan of 125 million, with interests of 20 million, which must be repaid by 2010. 57
This is where it is unreasonable! ‘Build model schools! Create a conducive environment for education!’ [so we were told] It was the education bureau that made us take the loan, but the loan is made in the school’s name” (AI 2008-148). It is no wonder then that even a foreign visitor would have few difficulties distinguishing lavish key-point schools from ordinary ones. Unintentionally, then, premium school financing feeds into education inequality and vice versa.
The ‘rule-abiding’ quality of self-financing
To summarize, the key take-away of our discussion on policy awards is the surprisingly “ruleabiding” quality of bureaucratic self-financing, as I have observed based on the most recent developments. This observation is surprising in light of popular perception and previous scholarship on the prevalence of arbitrary extraction in China. That said, I am far from asserting that the existing system developed is perfect. I am sure there are still instances of bureaucracies collecting revenue in violation of state policies. Illicit acts of corruption by individual officials also remains a glaring problem. Instead, what I wish to underline is the trajectory of institutional change in the Chinese administration. From a historical perspective, policy awards are like prebends that feudal lords assigned to unsalaried officials and tax farming elites as compensation. While Chinese bureaucrats have yet to transform fully into salaried Weberian civil servants, they are certainly becoming more attentive and faithful to the rules of prebendalism, so to speak.
Element (IV): Property Rights Over Self-Financed Income
All public bureaucracies collect fees and service charges; in this respect, the Chinese case is unexceptional. In standard practice, revenue collected by public bureaucracies is turned into the state treasury and public employees are paid regular wages through annual budget allocations. Salaries for public employees are usually fixed based on rank, and public wages are openly published in developed countries. In formulating budgets, departments report the number of personnel and the compensation that each is entitled, and then expect to receive at least enough budget allocations to cover personnel costs. Public employees should not receive more
58
pay if they collect more revenue. In short, public bureaucracies do not normally own any income generated.
In China, however, revenue generated by public bureaucracies is turned into the state accounts and managed by financial authorities, but they retain the right to appropriate those funds to finance off-budget workers, wages, benefits, and other expenditure needs. In this way, the revenue earned by a particular department and the payoffs of its employees are pegged. To explain how this works, I provide a micro-level view into the budgeting rules and processes, based on extensive interviews with central and local finance officials (AI 2007-111; 112; 113; 114; 115; 116; 117; AI 2008-138; 139; 140; 141; 147; 153).
Units of Negotiation
At the local level, each core bureau and its extrabureaucracies form a collective unit of negotiation (kou) in the budgeting process. The core bureau represents its extrabureaucracies. In budgeting terms, the core bureau is known as the first-level fiscal unit (yiji danwei) and the subsidiaries the second-level fiscal units (erji danwei). The parent bureau vets the budgets of its subsidiaries and then files a consolidated budget request to the finance bureau. The finance bureau only negotiates with the bureaus, not the extrabureaucracies. It then allocates a certain budget to the parent bureau, and it is normally up to the latter to distribute the budget among its subsidiaries. Evidently, the parent bureaus command significant allocative power over its subordinates.46
Determining basic budget allocations
Next, the finance bureau has to determine budget allocations for each collective unit. The finance bureau begins by considering the official fiscal identity of the bargaining party, i.e. whether its units are fully, partially, or self-funded. The finance bureau also evaluates the number of bianzhi (official positions). The bianzhi refers to the number of positions that are 46
In one case study, Wang (2004) found huge inequality in the distribution of funds across public schools in one county. The local education bureau had significant discretion over the distribution of state budget allocations. 59
approved by the state and eligible to receive state funding.47 Simply put, the bianzhi is like a personnel budgeting instrument. The local establishment office (bianban) sets a bianzhi for every unit. Those employed outside of the bianzhi quota are considered non-official (bianwai). Normally, the wages of these workers are paid with the hiring units’ own funds and will not be budgeted.
During negotiations, the finance bureau will then adjust actual budget allocations based on the availability of self-financed income in a particular unit. “If a new leader arrives in this unit, and he is good at cracking his brains to find ways to generate revenue, then our budget will have to be adjusted” (AI 2007-114). As further explained, “Say we [the finance bureau] are supposed to fill the entire cup. But if this cup has a tiny pipe [of financial resources] attached to it, then we don’t have to fill the whole cup” (AI 2007-114). That is, if a unit has a flow of external funds (e.g. from renting an office building), it will discount the finance bureau’s evaluation of its need for budget allocations.
When the bargaining process concludes, the finance bureau will make budget allocations for two basic items, (i) salaries according to the “number of heads” (rentoufei) and (ii) operational expenditure (bangong jingfei). In the final step, budgeting officials will evaluate special budgetary requests for capital spending, such as the purchase of equipment or construction of new property.
Budgeting Procedures Following the Budget Reforms
In this section, I discuss how individual offices appropriate self-financed income following budget reforms in the mid-1990s. The launch of capitalist reforms unleashed unprecedented opportunities for bureaucracies to accumulate wealth. From the 1980s to early 1990s, bureaucracies could expend their income with little or no supervision. This resulted in
47
This is except in the case of “entirely self-funded” (zishou zizhi) bureaucracies. These bureaucracies are also allocated a bianzhi quota to control hiring. However, their positions do not receive budget allocations from the government. Instead, they are financed by the bureau’s revenue collections. 60
extravagant spending and cadre bonuses, contributing to public perceptions of widespread corruption.
Fortunately, unfettered bureaucratic discretion did not last indefinitely. From the mid-1990s onwards, the ministry of finance launched an ambitious program of budget reforms, introducing new measures to strengthen control over the collection and disbursement of public revenue. Such control was progressively extended to the category of non-tax revenue and extrabudgetary funds.
“Separating revenue from expenditure.” One of the most crucial elements of the reforms was a new fiscal management policy known as “separating revenue from expenditure” (shouzhi liangtiaoxian), introduced in 1998. Simply explained, the policy required all public units to turn revenue collected into the state accounts at the level to which they belong, and the finance authorities would then disburse the funds according to “plan.” This budgeting procedure creates one “line” for revenue and another “line” for expenditure.
Centralized treasury management.48 Following the ‘separating revenue from expenditure’ policy was the treasury management reform (guoku gaige), initiated in 2000. Piloted at the central level and gradually extended to the sub-national governments, the reform aimed to establish a treasury single account (TSA). Traditionally, the bank account system in China was fragmented not only between levels of government but also between departments at each level. Bureaucracies would set up individual transitory accounts (guoduhu) to deposit funds. 49 As these accounts were not linked or centrally managed, it was extremely difficult for finance authorities to monitor monetary flows. In implementing the reform, finance authorities took the first important step of abolishing all transitory accounts. 50 Instead, public units were required to
48
I discuss the treasury reform in detail in a separate paper, "Centralizing treasury management in China: the rationale of the central reformers." Public Administration & Development 29, no. 4 (2009): 263-273. 49 It was standard practice under the traditional treasury system for state units to open transitory accounts to deposit revenue, as a consolidated treasury account system had not then been created. Such accounts should be distinguished from so-called “small treasuries” (xiaojinku) or slush funds, private bank accounts created by officials and departments to deposit illegal monies (Wedeman 2000). 50 As a result of the reform, it has become more difficult than before for offices to set up unauthorized “small treasuries.” 61
submit all revenue directly into a consolidated treasury account, which included a separate account for extra-budgetary monies.
Some observers have been optimistic about the results of the reforms, but perhaps overly so. In an earlier study, Yang (2004) claimed, “In practice, with the emphasis on the separation of revenue and expenditure, government agencies or institutions that collect fees and levies no longer take in the funds themselves… Such a mechanism can generally ensure that those who collect funds do not simply spend it on themselves. In fact, as the reform was fine-tuned and as the funds from levies and funds became incorporated into government budgets, the agencies or offices collected the funds generally lost the right to dispose of the funds collected (240, emphasis added).” I fully concur with Yang that the budget reforms have considerably enhanced fiscal management and anti-corruption control. However, contrary to his claims, my research finds that the measures have not removed bureaucracies of their ownership over earned proceeds, as I shall explain below.
Refunding Self-Financed Income
Following the budget reforms, the prevailing practice has been for the bureaucracies to turn collected revenue into the state accounts and then to receive a full or partial refund of the funds from the finance bureau. The term “refund” is a direct translation of the Chinese word fanhuan, which local officials use. However, by refund, I do not mean that the finance bureau physically returns cash to the respective offices (bear in mind that they no longer have individual bank accounts in places where the treasury single account has been established). Rather, “refund” in this case refers to an internal accounting practice. Each office submits a formal request to the finance bureau to appropriate its income for certain purposes (e.g. buying a new car). After the request is approved, the finance bureau will pay for the item from that particular office’s “pot” of funds.
Rate of Refund. The “refund” rate varies from place to place and from unit to unit. In Blossom county, all public units received a uniform 70 percent refund (AI 2007-114; AI 2007-111; AI 2007-119; AI 2007-113). In a poorer county in Tianjin, the rate varied from 50 to 100 percent (AI 62
2008-139). The public security bureau, a quintessentially powerful agency, enjoyed a full refund. Generally, unused funds can be rolled over to the next fiscal year (A1 2007-114; 115; 116; 117). In this manner, self-raised income constitutes, in a real sense, the “surplus” of Chinese bureaucracies. Retained funds are pooled and then reallocated through regular budgeting procedures.
Limited spending discretion. With stricter budgeting measures in place, Chinese bureaucracies have less discretion than before in using the funds they earn. Only following approval of requests for spending by the finance bureau can departmental funds be appropriated. Further, as part of the treasury reform, reformers implemented a direct payment mechanism (jizhong zhifu). The new measure empowers the treasury disbursement center at each level to pay staff wages and make purchases on behalf of spending units. This mechanism prevents agencies from spending funds for unauthorized purposes. In addition, many local governments imposed a ceiling on the disbursement of cadre bonuses. For example, in Blossom county, bonuses were capped at 30,000 yuan per cadre per year (AI 2007-119).51 Nonetheless, although bureaucracies have less spending discretion than before, they have not in fact lost ownership of their proceeds.
My findings are paralleled in some recent research and commentaries. In Zhang’s (2008) field research on agencies’ enlistment of court assistance in collecting fees, she was told by a local judge in Hubei province,
Although it is said that revenue should be separated from expenditure, in practice, each unit still takes care of itself... the finance bureau will not give you additional funds. Each unit’s expenditure is still based on its income.52 Similarly, a commentary in the Southern Daily observed,
“What is ‘separating revenue from expenditure’? Those departments that collect fees and fines turn their revenue over to finance; the finance bureau manages them as part of the extra-budgetary funds and then refunds them, allowing departments to dole out wages, benefits, and operational monies. 51
Restrictions like these, however, resulted inadvertently in the proliferation of in-kind benefits to replace cash payments, as we will later see in Chapter 3. 52 Interview with County Court, Hubei, 11.2005, by Xuehua Zhang. 63
Many people see ‘separating revenue from expenditure’ as an anti-corruption reform. In actuality, it has legalized the collection of fees and fines that are unjustified in the first place.”53 Critically, in my interviews, budgeting officials emphasized that individual offices retain the “right of appropriation” (shiyongquan) even after the ‘separating revenue from expenditure’ policy. The obvious implication is that bureaucracies have an incentive to generate revenue for their offices. Further, the processes of making and spending self-financed income have become increasingly institutionalized.
A Simulation of Budgeting Rules
To summarize the rules of the game, Figure 2.4 presents a stylized model of the budgeting processes in a local government with Departments A, B, and C. Each department is composed of a core agency and its subsidiary extrabureaucracies. The “mobilizable funds” (kezhipei zijin) of the local government for budgetary distribution come from three main sources:
a) Locally retained within-budgetary revenue (i.e. tax revenue); b) Fiscal transfers and grants from above; c) Pooled funds from the extra-budgetary account.
The finance bureau allocates the “plate” (panzi) of mobilizable funds to finance the basic budgetary needs of the county’s public units. Imagine that the finance bureau allocates 100 million for administrative expenditure, granting 50 million to A, 30 million to B, and 20 million to C. If the minimum needs of A, B, and C are 100 million, 50 million, and 35 million respectively, each department would face a shortfall of funds as follows: 50 million for A, 20 million for B, and 15 million for C. The next step for A, B, and C is to submit a request to the finance bureau to cover their shortfalls with retained self-raised income.
53
“Resolutely abolish the system of self-raised funds,” Southern Daily News, March 29, 2007. The author of this article is a researcher at the Central Party School. 64
Department A has just enough funds to break even for the year. Department B has a surplus of 5 million (self-raised income exceed unmet needs), which means it can provide additional benefits or bonuses for its employees. Department C does not have enough self-raised funds to meet its minimum administrative expenditure; hence, it must cut back on benefits or even owe wage payments. This stylized example shows that the amount of self-raised income made by each bureaucracy is linked to its employees’ payoffs.
The unique-ness of the Chinese administration becomes apparent if we consider textbook descriptions of public bureaucratic organization, for example, as Terry Moe wrote (1984, 763):
The typical bureau receives a budget from governmental superiors and spends all of it supplying services to a nonpaying clientele. Regardless of the agency's performance or how it changes over time, the results are not reflected in an economic surplus accruing to bureau heads, and this major incentive for the efficient monitoring of employee behavior fails to operate. Incentive plans that give employees a share of the "profit" in partial payment for their effort (e.g. in locating welfare fraud or cost overruns) are also ruled out. The absence of a residual and its implications for internal efficiency, therefore, would seem to be factors that immediately set public bureaucracy apart from private organizations and shape our expectations about their relative performance.
In the Chinese case, each offices “financial” performance (i.e. capacity to generate self-financed income) is directly linked to the employees’ payoffs. In this case, the bureaucracies have the potential to accumulate not just a slack, savings from budget allocation, but a real economic surplus.
Conclusion
Modifying theories of the firm, this chapter develops a hybrid state institutional model labeled bureau-contracting. We examine the empirical features in the Chinese case. In a bureaucontracting structure, contracting takes place within the bureaucratic hierarchy at each level of government. As seen in China, public bureaucracies are not purely salaried; most of them raise a part or even all of their own funds. Abiding by an intricate set of state rules, individual offices can earn, retain, and use so-called self-financed income. Critically, the funds examined in this 65
study are not illicit monies or bribes. The more self-financed income is produced, the more wealth a bureau amasses. In extreme cases, such as the land bureau in Blossom county featured in the introduction, offices may grow rich enough to acquire the property of their administrative bosses.
66
Figure 2.4: ‘Rules of the Game’ – A Simulation of Budget Procedures and Outcome Revenue sent to upper levels Local Taxes
Within-Budgetary Funds
Extra-Budgetary Funds
Department A’s Funds
Department B’s Funds
Department C’s Funds
Revenue retained at local level
Pooled extrabudgetary funds Mobilizable Funds
Fiscal Transfers
Budgeted Funds = 100 million Y (for administrative expenditure)
Actual Expenditure
Unmet Needs
Departmental Funds
Department A (allocated 50 mil Y)
100 mil Y
50 mil Y
50 mil Y
Breaks even
Department B (allocated 30 mil Y)
50 mil Y
20 mil Y
25 mil Y
Surplus of 5 mil Y; more bonuses or benefits for staff or rollover funds to next year
15 mil Y
5 mil Y
Deficit of 10 mil Y; owed wages, lower benefits
67 Department B (allocated 20 mil Y)
35 mil Y
CHAPTER 3
THE STRUCTURE OF CADRE REWARDS
We cannot study political institutions without studying incentives. And we cannot study incentives without mapping the structure of financial rewards facing officeholders in a political system. Incentives for governmental actors have featured prominently in existing analyses of development. As Oi underscored, “for China the issues were not whether its bureaucracy was capable of generating economic growth but whether it had the incentive to do so” (1999, 6). But, despite a widely shared emphasis on incentives, we have not seen to date a comprehensive documentation of the actual compensation structure of Chinese officials. Unless we know how much and in what ways bureaucrats are financially rewarded, we cannot appreciate the full scope of incentives that drive bureaucratic behavior.
Filling this empirical gap, this chapter identifies the structure of cadre rewards in China. The level of cadre rewards depend on two factors: (1) economic development and tax revenue collections in a particular locale, and (2) the ability of individual bureaucracies to generate selffinanced income, as I have analyzed in Chapters 2 and 3. Factor (1) results in wide intergovernmental disparity and factor (2) results in wide inter-bureaucratic disparity in cadre rewards. In other words, cadres may receive widely variant payoffs depending on which locality they are in and which office in the locality they serve. Taken together, these patterns suggest that the Chinese bureaucracy is governed by a uniquely high-powered incentive scheme that motivates bureaucratic actors to maximize revenue – be it revenue for local state budgets or for departmental budgets. However, in a public bureaucratic organization, the goal of revenue maximization can be at odds with the goals of social welfare provision and long-term economic development. A key implication of the analysis is that conflicting fiscal incentives exist in the Chinese bureaucracy. 68
Descriptive statistics in this chapter are based on internal (neibu) budget datasets, compiled by the Ministry of Finance and local finance bureaus, which break down governmental spending by types of costs, rather than broad functional sectors. These unique datasets present a valuable opportunity to measure and examine “cadre rewards” concretely.
Patterns of Cadre Rewards
One strategy for measuring “cadre rewards” is to disaggregate categories of public spending. There are three categories of current expenditure in China: personnel spending, administrative spending, and subsidies for individuals and families. From each of these categories, we can extract spending items that provide compensation or benefits for public employees. For analytical and descriptive purposes, we measure cadre rewards as the sum of wage payments, subsidies, and administrative benefits.
In the parlance of political economy (see Magaloni, Diaz-Cayeros & Estevez 2007), wage payments and subsidies are equivalent to “private goods provision,” i.e. they are direct monetary benefits for individual cadres. Administrative benefits represent “club goods provision,” i.e. they are spending on benefits shared among employees in bureaucracies. To be sure, not all of administrative spending goes towards benefiting cadres individually or as a group; some spending items serve work purposes, e.g. purchasing specialized equipment and materials. Hence, in estimating the level of administrative benefits, I take caution to exclude those spending items not likely to function as in-kind benefits for staff members.
Table 3.1 shows the aggregated structure of cadre rewards at the local (provincial and subprovincial) level of two groups in 2004. (a) “PJBC”: public security bureaus, judiciary bureaus, courts, and procuratorates, (b) “Others”: extrabureaucratic units in the tax, statistics, finance, auditing, commerce, assets management, tourism, overseas Chinese affairs, labor, customs, discipline, agricultural development sectors, as well as in the people’s congresses, people’s consultative committees, social organizations. Although not all party and governmental organizations are included, this sample covers a wide range of bureaucratic units and thus provides a fairly representative view of the whole bureaucracy. 69
Combining the two groups, several features deserve further comment. First, wage payments composed about one-third of total cadre rewards. In this category, only 38 percent were basic wages (jiben gongzi). Basic wages are standardized across the country (see later section). Allowances made up the second largest category at 32 percent, and bonuses followed in the third place at 15 percent.
Second, in addition to the wage payments discussed above, cadres also received some additional cash income from subsidies. This category composed 5 percent of total cadre rewards. Among them, subsidies for retired personnel were the largest share (33 percent), followed by housing subsidies (32 percent).
Third, administrative benefits took up the lion’s share of cadre rewards, equaling 56 percent of the total.54 Interestingly, about half of administrative spending fell under an undefined line item – “others” [what are these?] – that cost 79 billion yuan in 2004, about three-quarters of total wage payments. We do not know from the data what “others” specifically includes. As a central finance official explained, “The saying goes: ‘others’ is a basket, everything is stuffed in it (qita shi yiluokuang, shenme dou wanglizhuang)” (AI 2008-147). I was told that finance officials would ask departments for details on “other” spending only if the size of expenditure is glaring; otherwise, this category is a black box.
Based on anecdotal evidence, the “others” spending category could include frequent purchases of gifts for staff members, a phenomenon observed during my field research and welldocumented in the Chinese media. For example, in one county that I visited, cadres regularly received gifts from their departments, including seafood, cosmetics, skincare products, visits to salons, and even shopping certificates. Another major staff benefit is department sponsored or subsidized vacations. Free wining and dining is a common perk, and state-funded cars are deployed for private use. As one cadre puts it, “I guess this can be considered a kind of grey income” (AI 2007-111). The proliferation of in-kind benefits (and bribes) has even spawned an
54
I have excluded from this category spending items that are likely to serve as office expenditure rather than as collective benefits for cadres, e.g. spending on special materials (e.g. testing equipment, etc.), spending on books and research materials. 70
unusual business of “gift recycling” (lipin huishou) across China. One can exchange gift items for cash at gift recycling shops. Bureaucrats were reportedly frequent patrons of these shops. 55
Table 3.1: Composition of Local Cadre Rewards, 2004
WAGE PAYMENTS
PJCP
OTHERS
TOTAL
7012502 (46%) 2651908 2482715 981856 183090 712933
3638241 (29%) 1442553 918154 645980 179496 452058
10650743 (38%) 4094461 (38%) 3400869 (32%) 1627836 (15%) 362586 (3%) 1164991 (11%)
Office expenditure Printing
7253048 (48%) 506178 98981
8196625 (66%%) 552040 232221
15449673 (56%) 1058218 331202
Utilities Postage
289217 247669
169854 130962
459071 378631
Heating Transportation
137585 791389
101362 384535
238947 1175924
Travel Conference
358778 125891
218716 258913
577494 384804
Training Entertainment
148431 242032
162313 315562
310744 557594
Benefit fee Labor fee
153438 72883
111464 167824
264902 240707
Rental Asset management
41808 38754
57384 61840
99192 100594
Office furnishing Vehicles Others
329888 482164 3187962
328356 215830 4727449
658244 697994 7915411 (51%)
SUB-TOTAL
807524 (5%) 42989 256467 50542
520369 (4%) 18898 176490 40368
1327893 (5%) 61887 432957 90910
SUB-TOTAL Basic Wages Allowances Bonuses Social security Others
ADMINISTRATIVE BENEFITS
SUBSIDIES
SUB-TOTAL
Subsidies for pre-retirement Subsidies for retirement Health subsidies 55
An article from Xinhua reported that the gift recycling business in Harbin city declined because a new restriction against gift-making in the government (“Harbin officials are disciplined; gift recycling business declines,” Xinhuanet, Feb. 19 2005). Another article described civil servants as regular clients at gift recycling shops (“Gift recycling revealed,” Xinkuaibao, Feb. 15 2006). Some of the items brought for “recycling” may be legitimate gifts from one’s department but they may also be bribes. 71
Housing subsidies Scholarships Others TOTAL
276024 824 180678
155614 566 128433
431638 1390 309111
15073074
12355235
27428309
Source: Author’s calculation from Local Public Financial Statistics.
The composition of cadre rewards reviewed above reflects a pattern that is already well-known to bureaucratic insiders – China has a sub-moneterized bureaucratic compensation system. Many privileges of office come in the form of administrative benefits, rather than wages. This pattern reflects a common saying that cadres receive “low wages, multiple allowances, and abundant benefits.”56 Such a compensation structure results from ceilings placed by local governments on disbursing departmental income as cash payments to public employees.57 To circumvent the ceilings, wealthier bureaus are compelled to spend surpluses by giving out inkind benefits.
Formal vs. Extra-Formal Cadre Wages
To understand what determines the level of cadre rewards, it is helpful to first understand the monetary wage system for Chinese cadres. There are two cadre wage systems coexisting in China: a formal (zhidunei) and an extra-formal (zhiduwai) system. In the formal system, the central government sets a standardized wage scale for civil servants and extrabureaucratic employees that is uniformly implemented across the country. The formal wage scale presented several problems, as pointed out by Xu Songtao (2007), the former minister of personnel management. First, formal cadre wages are generally low in China, especially in comparison to other occupations requiring equivalent skills (Xu 2007, 214). Second, the wage scale is highly compressed. Although levels of development ranged widely across regions in China, basic wages do not reflect local disparities. Hence, while the formal basic wage level may be attractive in poor and rural areas, they are pitifully insufficient in well-off regions. Further, the wage differential between officials of different ranks is kept very low. Based on the official pay scale, 56
“What has wage sunning revealed?” Liaowang, Oct. 15 2007. This ceiling appears to vary from place to place. For example, at Blossom County, bonus payments to cadres were capped at $30,000 a year. 72
57
the highest-ranked local official (a Grade 1 mayor) is entitled to 190 yuan in wages per month, while a Grade 1 administrative support personnel gets yuan (see Appendix B). As Xu pointed out, wage compression in the formal scheme is much higher in China than in other countries. Third, wage increments are kept low.58
Appendix B shows the formal wage scale for civil servants. Formal wages consist of four segments: basic wages (jiben gongzi), function wages (zhiwu gongzi), rank wages (jibei gongzi), and length-of-service wages. Combining basic, function, and rank wages, the chairman of the CCP (the paramount head of state) is entitled to a formal salary of 1190 yuan per month – the maximum in the country. On top of that, one yuan is added to a cadre’s payroll for every year of accumulated service. In addition to these wages, the central government authorized the disbursement of local subsidies. According to the pay scale, the maximum sum for local subsidies was 300 yuan per month. To place these sums in perspective, the average wage of an urban worker in 1993 per year was 3,371 yuan (or 280 yuan per month), while the average level of consumption per month was 206 yuan per urban resident and 65 yuan per rural resident.59 Based on the formal compensation scheme, a city mayor would be poorer than the average urban worker. Clearly, formal cadre wages could not keep up with steeply rising costs of living in China, especially in fast-growing cities. They were also incompatible with the education level and social status of the cadre class.
In the reform era, a highly (perhaps even unrealistically) egalitarian formal public pay scheme was coupled by dramatically expanded opportunities for local governments and departments to accumulate wealth. The marriage of these two conditions necessitated the emergence of an extra-formal compensation system. What determines the level of extra-formal compensation? First, tax revenues collected and retained by local governments, which are highly correlated with local industrialization and economic development. Second, self-financed income earned by individual bureaucracies that compose each level of government.
58
In recent years, the central government has tried to increase the formal wages of civil servants using fiscal transfers. 59 Zhongguo Jingji Nianjian (China Economics Yearbook) 1994, China Economics Press, p. 760. 73
Tax revenues and intra-governmental disparity
Generally, the level of economic development and tax revenue collections determines the amount of state budget allocations and the average level of cadre compensation in a particular locale. As Oi asserted, “increased revenues in local government coffers directly translated into lucrative (and legal) economic rewards for local officials” (1999, 49). The correlation between fiscal revenue and cadre payoffs, it is believed, provided a crucial positive incentive for local officials to pursue local economic growth. In my field research, I have indeed observed a close association between local developmental outcomes and cadre rewards. Drawing on retained taxes, rich localities can afford to pay generous subsidies, allowances, and bonuses for meeting targets, on top of formal wages.
Here, I give an example from Tianjin city (AI 2008-135; 136). Lake county and Hills county were geographically adjacent counties. However, Lake county was more developed and collected more taxes than Hills county. Consequently, the average cadre in Lake county enjoyed a much higher level of compensation than those from Hills county. An official from Hills county complained bitterly (AI 2008-136): Our salaries are about half that of Lake county. The basic wages are the same, but the difference lies in the allowances and subsidies… Since basic wages are uniform, the local governments cannot change them. But there is another part of compensation that is based on the finances of local governments. For example, Lake county gives out a transportation subsidy, but other counties do not. In addition, they have bonuses for meeting targets. And why is the difference between localities so large? Because we have no money. The story of Hills and Lake is unexceptional. Much anecdotal evidence tells us that cadre rewards vary widely across provinces, cities, counties, and even townships in the same county. Such anecdotal evidence is supported by descriptive statistics in this study. Table 3.2 compares the total monetary value of cadre rewards across provincial governments in China in 2004, using the same set of numbers discussed in Table 3.1. Unfortunately, we cannot estimate the value of rewards per cadre, as we do not know the number of cadres in these groups. Nonetheless, the statistics available indicates a very wide variation in the level of cadre rewards across provinces. As expected, we observe a very strong correlation between cadre rewards per capita and tax 74
revenue per capita (correlation = 0.96). The description here provides only a rough national view. In Chapter 4, we will discuss the variation in cadre rewards at the county level and test the determinants systematically.
Table 3.2: Distribution of Cadre Rewards Across Provinces, 2004 (unit = million yuan)
PJBJ Others
Mean 4863.98 3989.36
Std. Dev 4079.04 3240.00
Min 595.26 116.49
Max 21132.19 15595.19
Self-Financed Income and Inter-Bureaucratic Disparity
Not only is there a wide gap in cadre rewards across local governments, there is also tremendous variation in cadre wages and benefits across bureaucracies within the same locality. I have been fortunate to obtain a rare dataset that records spending by departments in one county government. Table 3.3 below shows personnel and administrative spending per cadre across several bureaucracies in Windy county, located in a Northern municipality, in 2007. 60 As we can see, spending per cadre ranged from 28,390 yuan in the civil affairs bureau to 126,035 yuan in the commerce bureau.
What explains the huge variance above? The variance is unlikely the result of unequal levels of education. Recruitment standards are similar across bureaus and better educated cadres do not fetch higher formal salaries. Far more likely, then, the variance reflects the varying capacity of bureaucracies to raise funds and their de facto ownership over revenue collected. The civil affairs bureau administers poverty relief, provides welfare services for the old and handicapped, and oversees village elections. The commerce bureau, on the other hand, regulates business and issues licenses. Based on their functions, it is no wonder that the commerce bureau can afford to spend much more on cadre wages and benefits than the civil affairs bureau. These descriptive
60
This particular dataset does not break down personnel and administrative spending by sub-categories, so we do not know how much of personnel spending is composed of basic wages as compared to allowances and bonuses. 75
statistics also support my point in Chapter 2 that bureaucracies have not in fact lost their rights to dispose of self-financed income even after the post-1990s budget reforms.
Table 3.3: Personnel and Administrative Spending Per Cadre in Windy County in 2007 Office Commerce Bureau Cultural Market Regulation Team Party School Public Security Bureau Disciplinary Committee Personnel Bureau Finance Bureau National Development & Reform Commission Cultural Commission Environmental Bureau Relics Preservation Center Civil Affairs Bureau
Personnel and Administrative Spending Per Cadre (Yuan) 126,035 107,323 104,123 94,406 91,148 87,439 81,162 75,585 72,930 70,553 62,089 28,390
Source: Budget report published by Windy County Finance Bureau
Colloquially, wealthy bureaus are known as “greasy offices” (youshui yamen) and the ones less well-off as “clear offices” (qingshui yamen). Local officials are of course well aware of the variance in cadre payoffs across bureaus. As one official mused, “Compare the archives office to the construction bureau – even a fool knows! [their gap in income] What can the archives office do with those archives?” (AI 2007-128). The media is replete with anecdotes of intrabureaucratic income disparity. In one striking example from Guangxi province, it was reported that a local tax bureau paid a generous spring festival allowance of 18,300 yuan to each employee, while the other offices, lacking “grease,” could afford no more than a few hundred dollars.61
Once we understand the egalitarian constraints of the formal wage system, it becomes clear why few bureaucracies, even fully-funded public units, would not be content with budget allocations that provide only “basic” needs. Basic wages are insufficient for officials to sustain a 61
“What has sunning wages shown? The benefits of civil servants move towards transparency,” Liaowang, Oct. 15, 2007. 76
reasonable level of consumption and social status in a capitalist age. Formal rules of reimbursement for work-related expenses are similarly obsolete in light of the rising costs of living. For example, a central-level bureaucrat lamented that the official reimbursement for lodging during official travel was only 50 yuan a day (AI 2006-11). Seasoned travelers in China will know that 50 yuan can hardly get one a half-night rest in the cities today. “I am already not used to that quality of life,” said the bureaucrat (AI 2006-11). As Blecher and Shue pointed out, “In today’s China, local state actors must compete for authority with other rising social forces in a newly materialistic culture. When municipal officials take important business prospects out to dinner these days, they had better not arrive at the restaurant in banged-up jalopies and cheap shoes” (2001, 389).
High-Powered Incentives: How High? What Kind?
The Chinese bureaucracy is a state organization with uniquely high-powered incentives. These high-powered incentives take two forms. Firstly, monetary payoffs for local leaders and cadres increase with economic performance and tax collections. In other words, the local officialdom, as a collective body, is a claimant to the locality’s prosperity. Secondly, within each level of government, cadres rewards are further differentiated according to each bureau’s capacity to earn self-financed income. Each bureau is a claimant to the financial surpluses generated.
Let us place the degree of high-powered incentives in China’s bureaucracy in perspective. Susan Whiting (2004) studied cadre wage differentials across township governments in a Shanghai county. Citing a study by Baker et al (1988), she related that in a typical large American manufacturing firm, workers with the lowest performance ranking received 7.8 percent less than those who ranked the highest (Baker et al 1988, 595; cited in Whiting 2004, 110). In comparison, among the townships that Whiting studied, those cadres in the lowest ranked townships, which were ranked primarily on economic performance, received a whopping 66 percent less than those in the highest ranking townships. A large share of the cadre’s compensation and income variance across townships, she added, was a consequence of performance-based bonuses. Such heavy reliance on bonuses was ironically rare in American firms (Whiting 2004, 111). 77
My analysis finds high-powered incentives even higher than in Whiting’s study. Across 139 county governments in Shandong province, I find that the average cadre in the poorest county received 86 percent less in wages and benefits than the wealthiest county in 2005 (see Chapter 4). Returning to our comparison of reward differentials across bureaus in Windy county, listed in Table 3.3, cadres in the “clearest office” (the civil affairs bureau) had 77 percent less in monetary rewards than the “greasiest office” (the commerce bureau). That is to say, if we compared Windy county to the American firm in Baker et al’s study, economic incentives are about ten times more powerful in a Chinese county government than in a purely capitalist enterprise. Economic incentives in the Chinese bureaucracy are structured in such way that bureaucratic agents are motivated to maximize revenue. Local governments want to maximize revenue for the local coffers. Each bureaucracy wants to maximize revenue for its own department.
Careful distinctions must be drawn between the drive for revenue maximization in the Chinese context and other forms of economic motivations. Firstly, the goal of revenue-maximization in the Chinese bureaucracy is different from the private desires of corrupt officials to extract as many spoils for themselves as possible in violation of the laws. Here, the organizational principle is that revenue maximization takes place within boundaries drawn by the state, as explained in Chapter 3 on policy awards. That said, my point is not that bureaucratic agents in China are obedient and passive. Far from it, it has been emphasized that opportunism induces highpowered incentives. Instead, what I mean is that Chinese bureaucracies will strive to maximize financial gains through clever manipulation of the rules of the game, but without breaking the rules explicitly.62 Additionally, central-level departments and bureaus take the more active approach of lobbying for laws that will license their revenue-making activities. This is a phenomenon known as “departmental law-making” (AI 2007-128).
Secondly, the revenue-maximizing goal in the Chinese bureaucracy should be distinguished from the profit-maximizing goal of pure private firms. For firms, profits are the bottom-line. Profits
62
Examples are discussed in Chapter 3. Public schools, given the license to collect “premiums”, concentrated resources on key-point schools to build up their attractiveness and hence capacity to draw additional payments from willing parents. Regulatory agencies typically support their extrabureaucracies in generating more income by blocking market competition through the exercise of regulatory power. 78
are revenue minus costs. Firms will seek to maximize profits by increasing sales or decreasing costs or a combination of both. However, the corporate concept of costs does not obtain the same way in a public bureaucratic setting. To elaborate, let us first be reminded that bureaucracies in China (and any where else) are engaged in the services provision, not manufacturing. In manufacturing, the physical inputs used to make products constitute the bulk of costs. In the services business, costs come mainly from labor, especially skilled labor. A profitmaximizing firm in the services industries will strive to maximize revenue and minimize labor costs. This may be done by reducing excess staff and/or increasing labor productivity to improve services provision and consequently sales.
Different from private firms, however, Chinese bureaucracies have economic incentives but they serve political functions. One of their key functions is to finance public employment. This prevents political instability that could result from cadre layoffs, absorbs displaced military personnel and university graduates, and provides patronage capital for political leaders at all levels of the hierarchy and for the party at large. Thus, bureau chiefs in China do not usually have the freedom or desire to remove redundant staff. In fact, nepotism appears rife, and anecdotal evidence suggests that new leaders often bring new personnel, including family members (Wei-Arthus 2000), to office with them to expand their personal clientelist network.
Finally, I stress that economic incentives in the Chinese bureaucracy are not structured to reward individuals based on objective measures of good performance as public servants. The kind of high-powered incentives employed in the Chinese bureaucracy must be distinguished from high-powered incentives and bonus schemes advocated in conventional theories of new public management (Frant 1996).
Implications of the Incentive Structure
Following from the preceding discussion, the implications of the incentive structure in China are apparent. Chinese bureaucracies are motivated by unusually high-powered incentives to maximize revenue. From the state’s perspective, this is a positive feature insofar as agencies and extrabureaucracies are driven to self-finance, thereby reducing governmental fiscal burden and 79
inflating cadre rewards. From the social perspective, however, the consequences can turn out badly. A research report on shiye units conducted by the development research center (DRC) of the state council summarizes the resulting problems succinctly, “the goal of maximizing organizational and personal interests has led to the neglect, if not indeed, abandonment and damage of the goal of social welfare” (cited in World Bank 2006, 22).
One major consequence of bureau-contracting has been a sharp rise in the costs of essential social services. As reported recently by the World Bank (2006), service charges have skyrocketed in education and health care. Taking 1986 as the base year, tuition and fees in education financing rose by 15 times in 1994 and increased even more rapidly afterwards to 40 times in 1999 and about 85 times in 2002 (97). In health care, out-of-pocket costs in 1992 were about ten times that in 1986, and by 2001, they had grown 40 times since the base year. Scanning the Chinese media and even the Western media, one finds numerous reports about the adverse effects that such rising costs in education and health care have had on the lives of citizens.
Another major consequence is limited market competition in the services sectors. Notably, since China’s entry into the WTO, significant efforts have been made by the central government to liberalize parts of its services sector, for example, in banking and pharmaceuticals. However, it appears that there remains a significant chunk of the services sectors that still remain either explicitly monopolized or dominated by public extrabureaucracies, especially in the litany of minor sectors, as discussed in Chapter 2. In education and health care, although the number of private providers is growing (World Bank 2006), my research finds that many of these providers were either spun off from public providers, invested by former or existing officials, or affiliated to state entities. It remains difficult for purely private entrepreneurs to enter the services sectors. In recent years, the central government has begun to make steady efforts to liberalize the services economy. For example, whereas the tourism bureau used to monopolize the tour agency business in the 1990s, competition was introduced afterwards. However, compared to the manufacturing sectors, services still remain state dominated.
80
A third implication of our analysis of cadre rewards is that the coexistence of two sets of highpowered incentives are in tension with each other. On the one hand, cadre payoffs are linked to a particular locale’s economic development and tax collections. Taxes come primarily from businesses. To collect more taxes, a government has to bring in and retain businesses. On the other hand, cadre payoffs are also set by the amount of self-financed income that each office is able to make. Given the system of rents extraction available, will local officials be tempted to relinquish developmental efforts altogether? That is, if cadres can earn additional income and benefits through their own department’s self-financing, then does it matter to them if the locale prospers and brings in taxes to the treasury? If the answer were: it doesn’t matter, then China is in trouble. Like in a feudal system, local governments will disintegrate into prebendal rackets, each for its own, and development would not have occurred. For local state-led development to work, cadres need to have a common personal stake in the locality’s prosperity. This stake, furthermore, has to exceed, or at least balance, the officialdom’s narrower stake in departmental finances. In Chapter 4, we will test if such a condition holds.
Conclusion
Monetary incentives matter because they shape bureaucratic behavior, which in turn has profound effects on economic development and social welfare provision. This chapter presents one of the first known efforts at mapping the patterns of cadre rewards in China across regions and governmental branches, using previously unavailable budget data. To summarize, we find that the Chinese bureaucracy is governed by two coexisting sets of high-powered incentives that result from the assignment of rights of surpluses to local governments and to individual offices within each level of government. In consequence, we observe in China an unusually wide disparity in cadre rewards both across local governments and across bureaucracies. Highpowered economic incentives motivate bureaucratic actors to maximize revenue; however, they can also produce unintended negative social costs. In the next chapter, we examine the tension between developmental and rent-seeking incentives more closely.
81
CHAPTER 5
DUAL FISCAL INCENTIVES
“Politics… takes place on a knife edge… [some] descend into predation while others turn developmental. On the one side, the winning political strategy is to extract wealth; on the other, it is to promote its creation.” Robert Bates, 2008, 718.
To use the words of Robert Bates, politics in China “takes place on a knife edge.” At times, the Chinese state exudes economic dynamism. Other times, it epitomizes predation. How can we make sense of this contradiction? My study has tried to new shed light on this question first by carefully qualifying the nature of “predation” in China. In the preceding chapters, while not denying that instances of brute power still exist, I have shown that Chinese bureaucracies are governed by a surprisingly disciplined internal system of self-financing, in which they are licensed to exact rent payments as a means of compensation and administrative fund-raising. The institutional patterns described are not wholly a Chinese invention. They should become familiar if we recall prebendal systems of governance that in pre-modern states.
In this chapter, I take on a further challenge to demonstrate not only that state developmental efforts and decentralized rent-seeking coexists in China, but also to explain why they have managed to remain in equilibrium thus far. My analysis centers on a product of intersection between fiscal contracting (between levels of government) and bureau-contracting (within levels of government) – I shall call this product dual fiscal incentives. As we have already begun to explore in Chapter 4, two set of high-powered economic incentives are simultaneously at work. One set of incentives come from inter-governmental revenue-sharing. All things equal, the 82
more businesses developed and the more tax revenue a local government collects, the more local cadres are financially rewarded. A second set of incentives are presented by bureaucontracting. The more self-financed income each bureau can extract, the more its employees benefit personally. Self-financed income, while not illegal monies, are essentially quasi-rent payments.
So why would local cadres not choose to devote their resources to rent-seeking and relinquish developmental efforts, if both strategies provide legal financial payoffs? My answer pertains to their time horizon. I propose that while local cadres benefit more in the short-term from extracting departmental rents than from developing local businesses, they benefit in the longterm only from expanding the formal tax base. This condition holds insofar as local governments still benefit from marginal tax increments, even after the 1994 tax reform, and that developmental opportunities are available sufficiently.
Employing a new county-level budget dataset on an error-correction model, I present evidence for a theory of dual fiscal incentives. I begin with a brief introduction of China’s fiscal structure and history. Following which, I develop the theoretical claims and observable implications. Finally, I present regression analyses to evaluate my hypotheses. In the final section, we compare endogenous versus external limits on bureaucratic predation.
Background: The Chinese Fiscal System
Since the establishment of the People’s Republic of China to recent times, China has undergone several phases of fiscal policy. Before 1979, typical of Soviet command economies, revenue collection and spending decisions in China were highly centralized. Revenue collected by provincial governments had to be surrendered upwards and then remitted according to central plans. In such way, each local locality’s spending was not tied to its takings, so incentives to increase revenue were low. This was the period known as “eating from the same big pot” (chi daguo fan).
83
From 1980s onwards, China implemented a revenue-sharing system known as fiscal contracting (caizheng baogan) or “eating in separate kitchens” (fenzhao chifan). Under fiscal contracting, each level of government contracted with the next lower level of government; the latter remits a negotiated share of revenue upwards and then keeps the rest. Revenue-sharing was a bilateral arrangement between two levels of government. As there are five formal levels of government in China, there existed four revenue-sharing layers: central-province, province-city, city-county, and county-townships.63 Fiscal contracting gave local governments more spending autonomy but also required them to balance their own budgets, thereby hardening budget constraints.
Extrabudgetary revenues were an important element of fiscal contracting. Public funds in China are composed of within-budgetary and extra-budgetary revenues. Within-budgetary revenues are formal and nationally imposed taxes, whereas extrabudgetary revenues consisted of a wide range of non-tax revenue, including local surcharges, retained profits from collective enterprises, income earned by individual agencies, fees, contributions, etc. Following fiscal contracting in the post-1979 period, local governments could keep the negotiated share of within-budgetary revenue and the entire share of extrabudgetary revenue. Simultaneously, local governments were barred from making loans or establishing their own tax categories.
Fiscal contracting provided powerful incentives for local governments to pursue development and collect more taxes. It was also during this period that rapid local industrialization and growth took off. However, the post-1979 fiscal arrangements created new problems. Firstly, regional disparities widened. Secondly, as fiscal contracting was based on bargaining, the terms of sharing and tax categories were chaotic. Thirdly and probably most importantly, the central government rapidly lost fiscal control, as the sum and share of local revenue, especially extrabudgetary revenue, soared from 1980-1993. Many commentators feared that the central government had lost political control over provincial leaders because of the latter’s growing fiscal autonomy (Wang 1995; Wang and Hu 1995).
To address the problems discussed above, the central government replaced fiscal contracting with a tax-sharing system (fenshuizhi) in 1994. This reform abolished particularistic fiscal 63
“Local governments” thus refer to the provincial and sub-provincial levels of administration. 84
contracting that defined the 1980-1993 period; instead, it reassigned taxes between the central and provincial governments by establishing three tax categories: central taxes, local taxes, and shared taxes (primarily the value-added tax or VAT). The central government also instituted central tax collection agencies alongside local tax bureaus across the country. In short, the 1994 fiscal reform recentralized revenue and simplified the tax structure.
The 1994 reform was bold but incomplete. Provincial governments were forced to submit more revenue to the central government than before, but as a compromise, tax rebates were awarded based on the amount of revenue collected in 1993. Furthermore, while the 1994 reform instituted tax-sharing between the central and provincial governments, it did not specify the terms of tax sharing at the sub-provincial levels and hence did not entirely eliminate particularistic fiscal contracting. The central government also recentralized revenue without sufficiently adjusting expenditure assignments, thus increasing fiscal pressures on local governments. Fiscal transfers took on greater significance after 1994. With more revenue, the central government distributed larger amounts of grants to provincial governments, which then decides how the transfers would be assigned to the localities in its jurisdiction.
Inter-governmental Fiscal Contracting
Many experts saw inter-governmental fiscal contracting as a crucial reform and growth strategy in China. Susan Shirk (1993) argued that Deng Xiaoping and his reformist team had “played to the provinces” in a bid to counter resistance against market reforms from central party bureaucrats. For the Chinese reformers, particularistic fiscal contracting was the dominant strategy, as it not only benefited provincial public finances, but also provided opportunities for distributing personal patronage. As Shirk puts it, particularistic contracting “gave party and government officials at every level opportunities to earn political support from subordinates in exchange for granting them generous contract terms” (1993, 16).
Other scholars focused on the developmental role of fiscal contracting. Fiscal contracting provided each level of government the right to retain within-budgetary taxes in excess of an agreed quota and all the extrabudgetary funds; in exchange for contracts, local governments 85
took on more spending responsibilities. To generate more revenue, local governments pursued industrialization and investments actively, in particular, by promoting township and village enterprises (TVEs), which provided the bulk of local extrabudgetary revenue up until the 1990s (Oi 1992; 1999). To support development, local governments also offered a whole gamut of services (Blecher and Shue 2001).
Elaborating on the role of fiscal incentives in China’s growth, Montinola, Qian & Weingast (1995) contended that China represented a case of market-preserving federalism (MPF). Under MPF, the central government granted local governments economic autonomy in their jurisdictions and imposed hard budget constraints (see also Oi 1992). As the authors saw it, this arrangement was politically stable. Fiscal contracting created a direct link between the economy and fiscal resources, thus motivating local governments to pursue growth. As argued, fiscal contracting also induced local governments to compete for investments and taxes, thus constraining governmental rent-seeking.
Empirical studies find evidence that inter-governmental fiscal incentives do matter for development. Analysts find a positive relationship between provincial revenue retention rate and market development in China (Jin et al 2005). Contrasting the Chinese case, it was found that Russian local governments did not get to keep marginal revenue increases from regional governments and hence were unmotivated to supply public goods or constrain predation (Zhuraskarya 2003). To sum up, the authors contend that fiscal incentives propelled local developmental efforts, and these actions in turn contributed to China’s rapid growth.
The Matrix Structure and Dual Fiscal Incentives
The previous literature discussed has focused on fiscal decentralization between levels of government. My account of bureau-contracting suggests that fiscal decentralization took place not only between but also within levels of government. To explain my point, let us first review the “matrix” state structure in China. I present a stylization in Figure 4.1. There are five formal levels of government, which I represent as three levels: central, provincial, and sub-provincial governments. Multiple line bureaucracies exist at the central level, depicted as A, B, and C in 86
Figure 4.1, which are replicated down the hierarchy. Each level of government is composed of a parallel set of party and governmental organizations, headed by the Party Committee and Government Office respectively. The “heads of state” at each level are equivalent to the headquarters of a multi-tiered corporation. I represent them as red stars in Figure 4.1.
Figure 4.1: Dual Fiscal Incentives
Intragovernmental contracting
CENTRAL
A
B
C
PROVINCES
A
B
C
Intragovernmental contracting
A
B
C
Intragovernmental contracting
SUBPROVINCES
Inter-governmental contracting
3
Superimposed on this matrix structure are dual fiscal incentives. First, inter-governmental fiscal contracting is a vertical exchange that produces developmental fiscal incentives. Upper level governments grant the lower level government the right to keep a negotiated portion of revenue. Here, levels of governments (as collective bodies) are the contracting parties. Second, bureau-contracting is a horizontal exchange that produces entrepreneurial rents-collecting fiscal incentives. As detailed in Chapter 2, each level of government (i.e. “the red stars”) contracts with individual offices, authorizing them to generate self-financed income and then to keep all or a share of their earnings. These arrangements imply that local governments and individual offices each have rights to surpluses. Local governments have rights to revenue surpluses,
87
composed of tax revenue (shared with upper level governments) and extrabudgetary revenue. Disaggregating the local governments, individual offices have rights to self-financed income.
Having described the matrix structure, we may now see that each bureaucracy at each level of government is faced with mixed incentives, one to maximize revenue for the local treasury, and the other to extract departmental surpluses. Facing mixed incentives, local officials in China often behave in paradoxical ways. On the one hand, local officials are always eager to develop new market opportunities and draw investments into their localities. But on the other hand, stories of agencies slapping businesses with fees, fines, and inconveniences seem to abound as well. I would argue that officials in China are neither purely developmental nor predatory but both. But how can they be both? More precisely, what keeps the predatory elements from negating developmental goals and efforts?
Critical to explaining sustained development in China is explaining why bureaucratic predation has not appeared to “runaway,” as we have seen in other corrupt and rents-ridden states. The preceding chapters have identified and detailed the institutional mechanisms of bureaucontracting. In doing so, I distinguished self-financing in the Chinese context from statecorroding forms of extraction. In this chapter, I add that endogenous limits to bureaucratic predation exist in complement with mechanisms of rents management. I call these limits endogenous because they arise naturally, so to speak, and unplanned from the structure of dual fiscal incentives. They are not consciously imposed by central or reforming authorities.
First, I hypothesize that local cadres benefit more in the short-term from extracting funds for their own departments than from increased tax revenue. This hypothesis is easy to understand. Think about self-financed income earned by individual offices as “club funds.” Club funds belong to the bureaus and can be used to finance jobs, wages, and benefits, subject to some spending controls. In contrast, tax revenues are like “common pool funds.” Retained taxes serve the entire locality; they finance both public goods and private goods for the officialdom. So increases in self-financed income will provide more cadre rewards than equivalent increases in tax revenue.
88
Second, I hypothesize that the monetary benefits from self-financing are only short-lived, whereas the financial rewards from increased tax revenue are cumulative. This hypothesis is less straightforward and harder to test empirically. As we can gather from the earlier chapters, selffinanced income supplies cadre compensation and some administrative costs, but they are generally not reinvested, and not used for infrastructural development or social goods provision. Any increase in self-financed income is therefore like a one-time bonus that expires in a single period. In contrast, to increase tax revenue collection, local governments have to undertake long-term developmental efforts (e.g. cultivate strategic industries, urban planning, develop enterprise incentive packages), attract new businesses, and help promote existing ones. Local governments may reinvest surplus tax revenue productively, e.g. building new roads and infrastructure, or spend the funds on social welfare projects, in addition to paying cadres higher subsidies and allowances. While cadres may not benefit financially from state developmental efforts immediately, the economic gains should be realized cumulatively over future time periods.
Phrased succinctly, the implications of the discussion above are as follows:
•
H1 (Rent-Seeking Incentives): Increased self-financed income provides only short-term benefits to individual cadres, but no long-term benefits.
•
H2 (Developmental Incentives): Increased tax revenues provide both short-term and long-term benefits to individual cadres.
Read together, the hypotheses state that Chinese bureaucracies do not face either rent-seeking or developmental incentives – they face both. The dual fiscal incentive structure distinguishes Chinese bureaucracies from purely rational bureaucracies that do not make prebendal income. The structure also distinguishes Chinese bureaucracies from quintessentially predatory bureaucracies that have rent-seeking incentives but not developmental incentives.
My formulation allows us to evaluate and refine previously untested assumptions in several strands of literature. Firstly, the fiscal contracting literature proposes that local growth increases 89
tax revenue, thus motivating local officials to pursue development (Oi 1992; Oi 1999; Montinola et al 1996). This proposition is premised on the condition that increased tax revenue delivers personal benefits to the local officialdom. To date, however, we have not seen empirical evidence for this assumption. H2 aims to test the strength of developmental incentives by measuring the effects of increased tax revenues on local cadre rewards.
Secondly, the rent-seeking literature on China implies that rent-seeking incentives dominate developmental incentives (e.g. Pei 2006; Lu 2000a), but fails to provide systematic evidence for this claim. If they were right, we should observe that only Hypothesis 1 holds. That is, Chinese cadres benefit only from increased self-financed income but not from increased tax revenue. Rational bureaucrats, facing this condition, will forgo collective developmental efforts and focus instead on extracting funds for their respective offices. By comparing Hypothesis 1 against Hypothesis 2, we can evaluate the relative strengths of rent-seeking and developmental incentives, and thereby adjudicate between competing claims.
Thirdly, the aforementioned literatures have all neglected the temporal dimension of fiscal incentives. Existing models specify bureaucratic behavior as if rent-seeking and developmental incentives are mutually exclusive, when in fact, my theory contends, they coexist but along different time horizons. Chinese bureaucracies face powerful short-term incentives to extract revenue for their own benefits. In the short-run, such incentives are likely more powerful than the financial rewards from increased tax revenues and developmental efforts. However, I contend that only increased tax revenues yield improvements to cadre compensation and privileges in the long term. In other words, it is those cadres in regions with vibrant business conditions and a stable tax base who can enjoy a consistent growth in income and benefits. In the next section, I proceed to discuss the data used to test the hypotheses proposed.
Data, Measurements & Descriptive Statistics
For the purpose of this analysis, I collected and constructed a county-level budget panel dataset from Shandong province in the period of 2001-2005. The original data was compiled and published by the Shandong Provincial Finance Bureau, likely for internal policy-making purposes. 90
It is therefore a highly reliable source. The data provides an aggregated line-item budget for each county government in Shandong province, allowing us to identify the amount of personnel and administrative spending. Unfortunately, it does not break down the county budgets by departments, but the existing data is sufficient for our purposes.
Budget data is not easy to obtain in China. The data that I use is classified as “internal” (neibu), i.e., not available for public view. One reason for such secrecy is that this data is far more informative than data published in the national and local statistical yearbooks. The publicly available budgets list governmental spending by broad categories (e.g. education, health, agriculture), not line items (e.g. salaries, transportation, equipment, etc.). They are either misleading or too thin for meaningful analysis. To my best knowledge, this is the first attempt to employ line-item budget data from China for empirical testing. My data also provides an unusually detailed view of how different pots of public funds are actually spent at the subprovincial levels in China.
The county is the most important level of local government in China and therefore most appropriate for our analysis. County governments account for one half of sub-national expenditure (World Bank 2002, 34) and a major portion of gross domestic production. It is also at the county level where the delivery of public services is concentrated and overstaffing in the state bureaucracies most severe (World Bank 2007; World Bank 2002, 158). Therefore, public spending in the county governments has a significant impact on local economic development and social welfare.
How representative is Shandong of the rest of China? Shandong is one of the most populous provinces, situated on the Northern coast. While Shandong is not as wealthy as the Southern engines like Guangdong or Jiangsu, it is among the fastest growing. Additionally, Shandong still has a substantial agricultural sector, and its population is composed of a Han majority. In 2005, there were 142 county governments, ranging widely in population size and economic status. 64 It is reasonable to assume that counties in Shandong are roughly representative of the counties in coastal and central China, where the vast majority of the population lives. However, we may 64
The dataset has only 139 counties, as three of them did not exist prior to 2004. 91
expect a very different political and economic dynamic in regions of Western China with a large minority population (e.g. Tibet and Xinjiang), as well as in impoverished areas dependent on fiscal transfers.65 Nonetheless, by focusing on the Shandong case as a first step in our analyses, we can control for the effects of ethnicity and provincial-level idiosyncrasies.
Measuring Cadre Rewards
For our analysis, we need to measure the material benefits that accrue to the local cadre class. To do so, I divide total public spending in each Shandong county into two major categories: cadre rewards and social benefits. Cadre rewards include wage payments, subsidies, and administrative benefits (see Chapter 3). Social benefits include public construction and social welfare assistance. I exclude a sub-category of “acquisitions and repairs” from both categories, as it was unclear from the budget description whether this spending goes towards cadre rewards (e.g. construction of cadre housing) or public goods (e.g. road repairs). Therefore, the true size of cadre rewards may be larger than estimated here.
County Spending Structure. Figure 4.3 shows the average county spending structure in Shandong from 2001-2005. Cadre rewards compose at least three-quarters of total spending, ranging from 74 percent to 83 percent.66 On average, each county government spent 447 million yuan on cadre benefits annually. The situation in Shandong appears consistent with Park et al’s view that “most local governments put first priority on meeting their wage bills, which constitute the vast majority of expenditures” (1996, 771).
County Revenue Structure. County government finances in this study are composed of three parts: retained tax revenues, fiscal transfers, and self-financed income. Unfortunately, we do not have information on the entire category of extrabudgetary revenue. Nonetheless, in this
65
A rough comparison of public employment size in Shandong counties vs. all of China’s counties finds the former representative of national patterns, but excluding remote regions (e.g. mountainous areas in the West) and provincial-level municipalities (e.g. mega districts in Shanghai). 66 We do not have comparable sub-provincial budget data across countries. From a comparative perspective, this figure may not be unusually high as governments tend to spend the bulk of public funds on personnel and administrative costs. 92
analysis, we are primarily comparing the effects of retained tax revenues and self-financed income on cadre benefits. Hence, the data available is sufficiently useful for our purposes.
Figure 4.3: County Spending Structure
11%
10% 10% 10% 8% 9%
14%
83%
76%
15%
16%
16%
74% 74% 74%
0
Public Spending (10000 Yuan) 20,000 40,000 60,000
80,000
Figure 4.2: Decomposition of Public Spending
2001
2002
2003
Cadre benefits Social benefits
2004
2005
Acquisitions & repairs
29% 24%
32% 32%
32%
31%
34% 33% 30% 29%
40%
44% 37%
33%
2002
2003
35%
0
Amount of income (10000 Yuan) 20,000 40,000 60,000 80,000
Figure 4.4: Revenue Structure
2001
tax revenue self-financing
2004
2005
fiscal transfers
Each revenue stream is composed of the following elements. First, retained tax revenues. Following the 1994 fiscal reform, local governments are entitled to a portion of shared taxes and local taxes. The bulk of tax revenues come from business activities, for example, the shared value-added tax, commerce tax, and local enterprise income tax. Hence, to collect more taxes, local governments have to attract businesses. I list the different tax categories in Table 4.1 above. Second, self-financed income is the sum of fines, fees, and revenue earned by individual 93
bureaucracies from business activities and service charges not accounted as fines and fees.67 Finally, fiscal transfers include equalization transfers and conditional aid from the upper level governments.
Table 4.1: Tax Categories After the 1994 Fiscal Reform (Tax-Sharing System)
1. 2.
3. 4. 5.
6.
7.
8.
Central Taxes Customs tax Consumption tax and value-added tax collected by the customs authority Consumption tax Central enterprise income tax Income tax of local banks, foreign banks, and non-bank financial institutions Commerce tax, income tax, profit and urban management & construction tax collected from railway companies, banks, and insurance companies. Retained profits of central state-owned enterprises Export refund of foreigninvested enterprises
1. 2. 3.
4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18.
Local Taxes Commerce tax Local enterprise income tax Retained profits from local state-owned enterprises Personal income tax Urban land use tax Fixed assets tax Urban management & construction tax Real estate tax Vehicle and vessel tax Stamp tax Slaughter tax Agriculture tax Agriculture special product tax Arable land use tax Deed tax Inheritance tax Land appreciation tax Revenue from leasing state land
1.
2.
3.
Shared Taxes Value-added tax Central 75% Local 25% Resource tax Petrol tax belongs to central; the remaining to local governments Securities exchange tax Central 50% Local 50%
Source: Thirty years of China’s Fiscal Reform (Zhongguo caizheng gaige sanshinian), p. 63.
The three revenue streams each compose about one-third of total revenue. Self-financed income is a significant part of county public finance. Average self-financed income increased steadily each year at an annual rate of 0.24 percent in 2002, 12 percent in 2003, 19 percent in 2004, and 7 percent in 2005. 68 The highest rate of increase in self-financed income in 2003 and 67
Only this portion of the extrabudgetary revenues is reported by county in the original dataset. A MOF official I have consulted with agreed that this is a reasonable estimate of revenue belonging to individual offices (AI 2008-149). 68 The reasons for the drop in percentage in 2005 are unclear. 94
2004 correlates with the sharpest drop in tax revenues and fiscal transfers, suggesting that county bureaucracies had to earn more income to compensate for lower state budget allocations.
Variation of Cadre Rewards and Income Streams
Cadre rewards vary extremely widely across the 139 counties in Shandong Province. Measured in absolute terms, they vary from 71 million to 1.48 billion yuan, a 20:1 maximum/minimum ratio. Taking population size into account, cadre rewards span from 178 to 4,010 yuan per capita. The average county cadre in Shandong received 29,222 yuan in direct and indirect benefits each year, ranging from as little as 8,970 yuan in Chengwu County to 150,305 yuan in Laoshan County. Table 4.2: Descriptive Statistics of Cadre Rewards, 2001-2005
Cadre rewards (million yuan) Cadre rewards per capita (yuan) Cadre rewards per cadre (yuan) Change in cadre rewards Change in cadre rewards per capita Change in cadre rewards per cadre
Mean 447 755 29,222 58 99 3,789
Std. Dev 244 447 16,990 93 177 6,620
Min 71 178 8,970 -357 -716 -14,843
Max 1479 4010 150,305 450 954 39,460
The rate of change in cadre rewards varied considerably as well. Mean change in total cadre rewards was 58 million yuan, but the standard deviation was 93 million yuan. Likewise, the average change in rewards per cadre was an increment of 3,789 yuan, but the standard deviation was 6,620 yuan. In the minimum case, each cadre suffered a drop of 14,843 yuan in rewards, equivalent to one-half of mean per cadre, while in the maximum case, cadres enjoyed a dramatic raise of 39,460 yuan in a single year. Such large fluctuations in compensation are unusual in most governments, as public personnel receive regular wages from state budget allocations and follow a fixed pay structure. These large shifts in annual public personnel benefits in China provide yet another piece of evidence that actual cadre payoffs are not determined by a formal compensation scale; they are directly connected to the income of local governments and individual offices in the localities. They may therefore fluctuate widely. 95
Table 4.3: Descriptive Statistics of Income Streams, 2001-2005 Mean 435 352 316 17 21 31
Tax revenue per capita (yuan) Fiscal transfers per capita (yuan) Self-financed income per capita (yuan) Change in tax revenue per capita (yuan) Change in fiscal transfers per capita (yuan) Change in self-financed income per capita (yuan)
Std. Dev 402 273 172 231 149 91
Min 13 0 31 -1199 -779 -475
Max 4270 2483 1756 1541 767 582
We also observe wide variation in the levels of income streams across counties. Tax revenue per capita ranges from 13 to 4,270 yuan per capita; fiscal transfers from 0 to 2,483 yuan per capita; and self-financed income from 31 to 1,756 yuan per capita. We also observe fluctuations in the three income streams, and with such fluctuations varying widely across the units. Mean change in tax revenue per capita was an increase of 17 yuan, about 4 percent of the mean, with a standard deviation of 231 yuan. Mean change in self-financed per capita was larger than that of tax revenue at 31 yuan, about 10 percent of the mean, with a smaller standard deviation of 91 yuan. In the next section, we proceed to examine the relationships between the shifts in cadre benefits and the shifts in revenue streams.
Regression Analysis Using the ECM
In this section, we test the implications laid out in Section III, using a single-equation errorcorrection model (ECM) on a panel dataset. My theory predicts that rent-seeking strategies provide only short-term benefits, but developmental efforts deliver both short-term and longterm benefits to the local officialdom. I operationalize this proposition by proxying rent-seeking strategies with self-financed income earned by individual bureaucracies and developmental efforts with retained tax revenues belonging collectively to county governments. Framed in operationalized terms, I expect increased self-financed income to produce a large but only transitory boost to cadre benefits, while increased tax revenues will contribute to a gradual but cumulative growth in compensation and privileges for local cadres, holding other conditions equal.
96
The ECM specification is particularly well-suited for our analysis as it enables us to estimate the parameters of the short-term and long-term effects of the explanatory variables on the dependent variable simultaneously. The basic idea of the ECM is that there exists a long-run equilibrium between two or more time series variables, but with short-lived deviations from the equilibrium.69 Effects of changes in independent variables on the dependent variable may either be contemporaneous (i.e. limited to a single period) or cumulative (i.e. extended over several future periods). Using the ECM, we may estimate and compare the temporal effects of changes in the revenue structure on changes in cadre rewards.
We measure the dependent variable of “cadre rewards” in two ways: rewards per cadre and cadre rewards per capita. Rewards per cadre measures how changes in the revenue structure affects the material payoffs of individual cadres, providing a direct indication of the incentives facing an average cadre, which is critical for assessing our predictions. Cadre rewards per capita reflects the level of spending on cadre rewards adjusted for population size. Following conventional analyses of public financial data, we also measure each of the fiscal variables (tax revenue, self-financed income, and fiscal transfers) in per capita terms.
In the analyses, we control for a number of possible socio-economic determinants of cadre rewards: GDP per capita, population size, share of urban population, and size of public employment. We do not need to include controls for minority populations, as these are negligible in Shandong province, or a dummy for the capital city of Jinan, as city employment and spending have been excluded from the data.
The ECM specification requires that we include (a) the lagged dependent variable, (b) the lagged levels of the independent and control variables, and (c) the differenced value of the independent and control variables. In addition, to control for time-invariant unobserved effects across the county governments, e.g. historical legacies or cultural attitudes toward cadre entitlement, we include county dummies in the regression. To control for unit-invariant
69
One common application of the ECM in political science is in voting theories. While there is a normally stable voting structure reflecting cleavages in society, there are also short-term fluctuations in vote support in response to transitory “shocks.” 97
exogenous shocks, e.g. abolition of the agricultural tax in 2002, we include year dummies for each year of analysis. The error correction version of the model can be expressed as follows:
∆CADREWt =
α1 + β1CADREWt-1 + β2∆TAXREVt + β3TAXREVt-1+β4∆SELFINt + β5SELFINt-1 + β6∆FISTRANt + β7FISTRANt-1 + ∑ controls + ∑ county and time dummies + errors (1)
where CADREREW = cadre rewards per cadre and per capita, TAXREV = retained tax revenue per capita, SELFIN = self-financed income per capita, FISTRAN = fiscal transfers per capita. The parameters β2, β4, and β6 estimate the short-term effects of a one-unit change in tax revenue, self-financed income, and fiscal transfers respectively on change in cadre rewards, while β3, β5, and β7 capture the long-term effects. In other words, if the lagged fiscal variables are statistically significant, then X has a long-term effect on Y; however, the long-term effect does not register fully immediately but is distributed over future time periods. Substantively, the long-term multiplier effect equals the coefficient of the lagged independent variable divided by the negative value of the lagged dependent variable (-β1). The value of the lagged dependent variable is the error correction term, which indicates the rate at which equilibrium errors are corrected. That parameter gives us the speed at which the effects of X on Y are realized over time.
Following the discussion in Section III, the first expectation from theory is that change in selffinanced income will register a larger short-term effect on cadre benefits than an equivalent change in tax revenue (Hypothesis 1). In the context of the regression, we expect both β2 and β4 to be statistically significant, and β4 to be larger than β2. We also predict that growth in tax revenue will have a long-term impact on improving cadre benefits, in addition to having a contemporaneous effect, while an increase in self-financed income will not benefit local cadres in the long-term (Hypothesis 2).
Translated into statistical terms, we expect β5 to be
insignificant.
98
Results of Analyses
Table 4.4 reports the results of the regressions, which are consistent with expectations. With two-level fixed effects included in Model 3, the first differences of TAXREV and SELFIN register a significant and positive effect on CADREW per cadre, with TAXREV having a smaller coefficient than SELFIN. In the same year, a unit increase in TAXREV will increase the value of benefits for each cadre by 18 yuan, while a unit increase in SELFIN will raise cadre rewards by 26 yuan. In other words, in the short-term, the average cadre benefits about 45 percent more from increased self-financed income than from an equivalent increase in tax revenue.
Similar results hold in Model 6, with CADREW per capita as the dependent variable. The coefficient of 0.95 on the first differenced of SELFIN indicates that 95 percent of every one-yuan increase in SELFIN per capita goes towards cadre rewards. This indicates that SELFIN earned and owned by individual bureaucracies are indeed “club funds.” Almost all of such income finances private and club goods for local cadres, not public goods provision. In comparison, only 53 percent of increases in TAXREV contributed to cadre benefits, which is consistent with our earlier characterization of tax revenue as “common pool funds” for the entire locality.
Table 4.4: Determinants of Cadre Rewards in Shandong Counties (2001-2005)
L. tax revenue per capita D. tax revenue per capita L. self-financed income per capita D. self-financed income per capita L. fiscal transfers per capita D. fiscal transfers per
DV= D. Cadre Rewards per Cadre
DV= D. Cadre Rewards per Capita
Model (1)
Model (2)
Model (3)
Model (1)
Model (2)
Model (3)
3.997*** (1.013) 12.073*** (1.108) -6.188*** (1.603)
16.742*** (2.516) 19.475*** (1.942) 4.034 (4.041)
13.969*** (2.716) 18.188*** (1.973) 3.864 (4.015)
0.155*** (0.024) 0.332*** (0.027) 0.126** (0.050)
0.542*** (0.060) 0.540*** (0.046) 0.479*** (0.103)
0.526*** (0.066) 0.531*** (0.048) 0.483*** (0.104)
25.483*** (2.164)
25.460*** (2.804)
25.605*** (2.867)
0.849*** (0.055)
0.941*** (0.068)
0.949*** (0.071)
-0.357 (1.065)
4.062 (2.683)
-1.808 (3.143)
0.258*** (0.032)
0.445*** (0.067)
0.421*** (0.079)
5.649*** (1.431)
7.863*** (2.066)
4.680** (2.320)
0.348*** (0.038)
0.430*** (0.051)
0.420*** (0.058)
99
capita Controls L. public employees (in 1000) D. public employees L. population (in 10000) D. population L. urban share of population D. urban share of population L. GDP per capita (in 10000 Y) D. GDP per capita Lagged DV County Effects? Time Effects? Constant N R2
-0.279*** (0.082)
-0.932** (0.383)
-1.128*** (0.392)
-0.003 (0.002)
0.008 (0.009)
0.007 (0.009)
-1.631*** (0.262) 53.202*** (19.168) 10.765 (37.788) -0.803 (10.953) 26.562 (20.306) 1056.001*** (281.238)
-1.167*** (0.328) 434.587*** (88.311) -176.650*** (56.463) 73.076 (45.863) -31.653 (42.575) 2161.292*** (578.929)
-0.954*** (0.333) 394.431*** (88.058) -172.595*** (55.810) 45.024 (46.540) -8.102 (43.069) 1548.202** (603.871)
-0.001 (0.007) 0.268 (0.442) -5.459*** (0.995) 0.013 (0.276) 0.061 (0.518) -6.122 (6.722)
-0.009 (0.008) 2.007 (2.095) -3.847*** (1.382) -1.483 (1.117) 1.207 (1.037) 18.861 (13.777)
-0.008 (0.008) 1.870 (2.134) -3.789*** (1.388) -1.566 (1.152) 1.250 (1.070) 16.870 (14.828)
-214.397 (413.255) -0.014 (0.034) No No 3369.017 (969.230) 556 0.50
-1181.3** (491.689) -0.445*** (0.051) Yes No -13326.25 (7912.835) 556 0.59
-801.781 (497.619) -0.482*** (0.518) Yes Yes 575.176 (8881.792) 556 0.60
25.643** (10.454) -0.237*** (0.039) No No 54.00 (22.98) 556 0.60
-2.074 (11.828) -0.636*** (0.048) Yes No -293.38 (192.26) 556 0.68
-0.574 (12.245) -0.637*** (0.049) Yes Yes 0.637 (0.049) 556 0.68
Standard errors in parenthesis. * p < 0.10, ** p < 0.05, *** p < 0.01
Having examined the short-term effects, we now turn to the long-term effects. Results in Model 3 are consistent with Hypothesis 2. Lagged TAXREV is statistically significant, while lagged SELFIN is not, meaning only TAXREV has a long-term effect on CADREW per cadre. As earlier discussed, this long-term impact is likely because growing tax revenue involves taking developmental actions, such as attracting businesses and constructing infrastructure, which produce a multiplier effect on the local economy and cadre compensation. In Model 3, the cumulative effect of a one-yuan increase in tax revenue on CADREW per cadre is 29 yuan. This effect occurs at a rate dictated by the error correction term, which is 0.482. Equilibrium errors are corrected at the rate of 48 percent, leaving 52 percent of the disequilibrium shock after one year, 27 percent after two years, 14 percent after three years, and so on. Concretely, that means 100
increased tax revenue has a long-term effect on improving benefits per cadre that is realized over a period of about 10 years. The total short-term and long-term effects of increased TAXREV on CADREW per cadre are 47 yuan, far exceeding the short-term effects of increased SELFIN, which is 26 yuan.
While only increased taxes affect benefits per cadre in the long-term, both increased taxes and self-financed income lead to a higher total spending on cadre rewards in the long-term, adjusted for population size. In Model 6, both lagged TAXREV and SELFIN show statistically significant effects on CADREW per capita. However, lagged SELFIN has a smaller coefficient of 0.48 compared to 0.53 of lagged TAXREV. In substantive terms, every one thousand yuan increase in TAXREV and SELFIN per capita would contribute to a cumulative increase in CADREW per capita of 754 yuan and 832 yuan respectively over several years. Nonetheless, from the perspective of the individual cadre, total per capita increase in benefits matters less than per cadre increase in benefits. Thus Model 3 provides strong support for Hypothesis 2.
To place the results in context, we simulate the short-term and long-term effects of fiscal changes in one particular year on rewards per cadre, as summarized in Table 4.5. From 2004 to 2005, the Shandong counties saw a mean increase of 153 yuan in TAXREV per capita and 28 yuan in SELFIN per capita. We call this Scenario I. All things equal, those increases translate into a growth in CADREW per cadre of 2,754 yuan and 728 yuan respectively, totaling 3,482 yuan. Now imagine a Scenario II. The county cadres decided to slack on developmental efforts, resulting in a stagnation of tax revenue, and instead to compensate for the loss in tax growth by collecting more self-financed income that year, such as by extracting more fees and fines. In Scenario II, the counties would experience no increase in TAXREV and 181 yuan increase in SELFIN per capita. Each cadre would enjoy 4,635 yuan growth in benefits, over a thousand yuan more than in Scenario I. This suggests that cadres face strong temptations to direct resources away from developmental efforts toward rent-seeking, as the latter delivers larger immediate rewards, but only in the short-term.
Calculations change, however, once long-term effects are taken into account. In Scenario I, an increase in TAXREV of 153 yuan has a cumulative effect on benefits per cadre totaling 4,437 101
yuan, which will occur over an approximately ten-year period, while increased SELFIN has no long-term impact. The total effect of the fiscal changes is to increase benefits per cadre by 7,919 yuan. Conversely, in Scenario II, the cadres each received a larger immediate boost in rewards than they would have had in Scenario I, but over the long term, nothing more. The total effect of a one-time decision to replace tax growth by collecting more self-financed income is to increase individual cadre benefits by 4,706 yuan, three thousand yuan less than the increase in Scenario I. If such a decision were repeated over several years, the cumulative impact of a relative loss of income to the local cadres is substantial. In other words, with a longer-term horizon in view, it does not pay for local bureaucracies to behave in purely predatory or rent-seeking ways.
Table 4.5: Simulation of Short-Term and Long-Term Effects on Benefits Per Cadre
Scenario I Scenario II
Change in Revenue Per Capita TAXREV = $153 SEFLIN = $28 TAXREV = $0 SEFLIN = $181
Short-Term Effects
Long-Term Effects
$2,754 + $728 = $3,482 $0 + $4,706 = $4,706
$4,437 + $4,437 $0 + $0 = $0
Total Effects on CADREW per cadre
$0 = $7,919 $4,706
The effects of fiscal transfers deserve some elaboration. As reported in Models 3 and 6, an increase in FISTRAN has only a short-term effect but no long-term effect on CADBEN per cadre, and both a short-term and long-term effect on CADREW per capita. In both specifications of the dependent variable, the effects of FISTRAN are small compared to those of TAXREV and SELFIN. In Model 3, every yuan increase in FISTRAN will increase rewards per cadre by merely 5 yuan, compared to 18 yuan and 26 yuan in the case of TAXREV and SELFIN. In Model 6, the cumulative effect of a 1,000 yuan growth in FISTRAN compared to an equivalent growth in TAXREV and SELFIN on total cadre rewards per capita is 660 yuan, compared to 825 and 758 yuan in the two other categories. Consistent with expectations, the results indicate that fiscal transfers are the least important of the three revenue streams in affecting cadre compensation and perks.
Finally, turning to the control variables, we find unsurprisingly that the size of public employment affects CADBEN per cadre negatively in Model 3. On average 183 staff are added to the county bureaucracies each year, which would reduce rewards per cadre by a modest sum of 174 yuan. Population size shows a negative short-term effect but a positive long-term effect on 102
CADBEN per cadre. This could be that an increased population produces more demand for public goods provision in the short-term, diverting funds from cadre rewards. Nevertheless, over a longer period a larger population may contribute to the local labor force or provide more targets for rents extraction, thus benefiting the cadres. GDP per capita registers only a longterm effect on CADREW per cadre, reflecting the delayed rewards that cadres gain from economic development in their locality. The mean annual increase in GDP per capita is about 3,000 yuan. A one-time mean increase will produce a cumulative growth in benefits per cadre by 963 yuan that will occur over a decade, a 3 percent increment over the average level of benefits.
To summarize, results of my analysis support the claim that local Chinese cadres face short-term incentives to extract self-financed income for their own departments, but long-term incentives to grow the local economy and expand the tax base. Developmental efforts and in turn increased tax revenue generate multiplier effects, both on the local economy and on individual cadre payoffs. These payoffs do not accrue immediately, but accumulate gradually over time. In other words, ignoring the future, local cadres harvest more present gains from extracting departmental rents than from cultivating the tax base. However, if the local officials are farsighted enough, they will realize that neglecting market development hurts their pockets in the long-run. All things equal, those cadres who enjoy the highest compensation growth are those in localities with thriving businesses that can contribute to the state treasury.
Incentive-Compatibility vs. Instruments of Oversight
My findings add confidence to the hypothesis that endogenous limits on bureaucratic predation arise when fiscal contracting interacts with bureau-contracting. Yet, one may rightly counter that a collective action problem remains unsolved. Even if each bureau perceives its long-term interests in economic development, it still has an incentive to extract and shirk its developmental role. The costs will be widely distributed across the local government and the benefits concentrated in the particular bureau that “cheats.” Such a problem will still lead the bureaus “to ‘overfish’ in the ‘commons’ or the rental havens” (Bardhan 1997, 1325).
103
My responses to the collective action problem identified above are two-fold. First, it should be pointed out that the Chinese government has essentially exchanged one kind of collective action problem for another. Should the government choose to pool the funds collected by each bureau, without giving them a full or partial residual claim, each bureau will have an incentive to shirk its revenue-making responsibilities and leave it to others to do the job. This collective action problem would lead to an under-farming of revenue, which would be an especially serious problem in light of heightened budgetary pressures in the 1990s, as will be discussed in Chapter 5. Hence, this takes us to the second point. Second, in confronting the potential collective action problem of “overfishing in the commons,” the state in China, both at the central and local levels, has devised alternative mechanisms to cope with it.
Mechanisms of bureau-contracting. As argued, the structure of bureau-contracting presents the foundational set of mechanisms for managing rents provision. By institutionalizing the processes of exacting self-financed income, the state at each level not only obtains information about the flow of departmental funds, it also exercises control over what kinds of revenue may or may not be collected, based on policy awards formulated. By drawing departmental funds into the state account system, the finance authorities can also implement rules governing the use of those proceeds. We discuss these mechanisms below.
Mechanisms of promotion. But that is not all. The criteria used for official assessment and promotion provide another important set of mechanisms. Hierarchical personnel management is the key to personnel control in China. As several studies have shown, the two most important criteria for the promotion of local leaders are growth rates and revenue collections (Li and Zhou 2004).70 The local leadership’s almost single-minded focus on growth and taxes guarantee their attention to business conditions.
Specialized economic agencies. As Wade noted in his study of Taiwan’s developmental experience, “governing the market requires a small number of powerful policy-making agencies able to maintain the priorities” (1990, 195). Not unlike the East Asian NICs, many local
70
The adverse effects of this incentive is that local leaders tend to pursue growth at all costs, in particular, at the costs of environmental protection. 104
governments in China established specialized economic agencies to lead business development. For example, one county I visited in Tianjin created a “large projects office” to attract and manage multi-billion projects. Their officials are constantly on the move, flying across the country to meet investors and to convince them to invest in their locality (AI 2007-93). Local FDI (foreign direct investment) offices are also common. Much like the economic development board (EDB) in Singapore, these offices specialize in planning the macro business environment and bringing in foreign investments.
Diversified economic targets. At the same time, to keep the other offices in check, I have found that diversified economic targets are created in some localities. It appears that economic targets do not fall on the shoulders of local leaders alone. In the same Tianjin county, tax collection and investment targets were distributed to almost every bureau. I was told even the county people’s consultative committee (zhengxie), a non-economic organ, is given a target to bring in a certain amount of investments to the locality each year (AI 2008-144). While this does not seem to make sense in most governments, it helps local governments in China to keep different offices involved in business development directly.
Could the institutional mechanisms described above suffice without endogenous limits? That is, to what extent does it matter that each office and cadre benefit financially from long-term growth? I would argue that it takes a combination of carrots and sticks to work. Controls, no matter how tight, may slip. Giving each bureaucrat a personal stake in the local GDP is the prerequisite to China’s high-powered developmental strategy. Without the “carrots,” it would not have been possible to stimulate an across-the-board enthusiasm among cadres to pursue local economic development. It is the promise of the cumulative rewards of collective wealth that helps the officialdom to look beyond myopic departmental gains.
Conditions for Dual Fiscal Incentives
The structure of dual fiscal incentives balances on a delicate equilibrium that rests on two conditions. The first condition is that fiscal incentives remain for local governments to pursue growth and expand the tax base. In this regard, a critical questions arises is whether the 1994 105
fiscal reform had diminished such positive incentives. Fortunately, the existing evidence suggests although revenue has been recentralized since 1994, sub-provincial governments continue to retain a significant share of taxes. My data from Shandong reveals that on average the county governments retained 65 percent of total taxes collected, with a standard deviation of 11.94 percent. Further, as Oi observed, “in spite of increased payments to the center, the 1994 system, like its predecessors, leaves the localities a clearly defined residual over which they have exclusive rights” (Oi 1999, 55). In a similar vein of finding, Jin et al (2005) reported that the marginal revenue retention rate of the provincial governments is significantly associated with higher rates of non-state market development and reforms in the state sector from 1970 to 1999.
A second condition is the availability of opportunities for business development. Some locales, especially state designated poverty counties, are stuck in landlocked and resource-weak locations with little hope of breaking out of poverty traps. Even if there were pro-growth fiscal incentives available, these counties are unable to take advantage of them. Most of these poverty counties are located in the western and central regions, not the coastal areas. They rely extensively on fiscal transfers to subsist. My analysis of the Shandong case finds that selffinanced income and retained taxes are much stronger determinants of cadre rewards than fiscal transfers. But we are likely to find a contrasting pattern of bureaucratic financing and incentives in locales destined to poverty. It is in these places, I believe, where we expect to find bureaucracies behaving in purely predatory ways.
Clearly, many questions remain about the longitudinal change in fiscal incentives over three decades of reform in China. This analysis reflects only patterns in recent years, from 2001-2005. While the scope of coverage is modest at this stage, it presents the first effort to identify and test the existence of dual fiscal incentives. It is hoped that such an analysis will provide a useful foundation for future data collection and analytical efforts.
106
Conclusion
Collectively, the results of regression analyses using an error-correction model offer solid evidence for a theory of dual fiscal incentives. We find that while local cadres benefit in the short-term from extracting income for their departments, they benefit more in the long-term from growing the local economy and expanding the tax base. In other words, in equilibrium, bureaucratic rent-seeking in China, though seemingly prevalent, is not “runaway” and has not thus far over-shadowed incentives for pursuing local development. Such equilibrium can be maintained so long as upper level governments provide relatively secure and sufficiently attractive fiscal incentives for lower units to pursue growth and in places where developmental opportunities are available. Findings from this chapter bridges polarized views of China as a “local developmental” vs. “decentralized predatory” state. They identify the institutional basis for the paradoxical coexistence of local state-led development and bureaucratic rent-seeking in China.
107
CHAPTER 6
THE POLITICAL ECONOMY OF BUREAUCRATIC ADAPTATION
The preceding chapters have established the mechanisms of bureau-contracting in the Chinese context. This chapter takes a step backward and ponders the origins and development of the state structure in China, which prompts a complex set of questions. Why did China not follow the path of other developmental states and develop a fully salaried “Weberian” bureaucracy to support capitalist development when markets opened in 1979? More broadly, what were the unique conditions and goals of reform in China that shaped the structure of administration, or, in other words, what makes the ideal-type of bureau-contracting rare compared to the Weberian model? Is bureau-contracting merely a transition to the Weberian bureaucracy? Or is it self-reinforcing?
The chapters that follow do not aim to retell the history of China’s administrative evolution comprehensively and in detail. Rather, my objective is to extract the forces that shaped bureaucratic structure in the context of a late-industrializing communist state, thus treating state structure itself as a variable compound outcome. My approach, following several others who have studied bureaucracies in democratic systems (Moe 1984; Geddes 1996), maintains that bureaucratic structures in reforming autocracies are endogenous to politics and the processes of development. State institutional patterns reflect political interests, changing economic conditions, and unexpected contingencies of reforms.
108
Pre-reform institutional legacies of bureau-contracting
This section traces the historical legacies of prebendal administration during early periods of imperial rule. Interestingly, I find what seems to be the absorption and mutation of pre-modern institutional features into the Maoist state structure.
Prebendalism in early imperial times
Although China is commonly known as one of the earliest bureaucratized state in history, its bureaucratization had in fact been persistently incomplete, as some scholars have noted (Kiser and Cai 2003). Generally, the imperial governments ruled over vast territories through a small formal administration and a thick patron-client network emanating from the state and penetrating down to the lowest reaches of society. This strategy of rule was not unique to the Chinese empire, but common to large early states. The Ottoman Empire exercised a similar strategy of rule, in which state bureaucrats co-opted bandits and governed alongside an extensive chain of tax agents and sub-tax agents (Barkey 1994). The structural logic of the past has persisted in China until the present day.
China has had a long history of prebendalism and patrimonial governance. Historically, the Chinese administration was composed of a small core of formal state officials and an expansive sub-layer of semi-officials. Officials were scholars who had passed the imperial examinations and were formally appointed to office. Although the officials were entitled to a nominal salary (feng) from the state budget, the salaries were woefully inadequate. Historians have named the peripheral agents serving alongside the formal officialdom the “extrabureaucracy” (Rankin 1993), “the third realm” (Huang 1993), and “informal extension of the bureaucracy” (Hickey 1991). The extrabureaucracy was composed of two sets of functionaries. First, they included an army of unsalaried clerks and runners who conducted rudimentary but critical tasks of administration. They were otherwise known as the “talons and teeth” of the local governments (Reed 2000). Secondly, they included members of the local gentry who performed public duties and collected taxes on behalf of the state.
109
Local offices (yamen) were notoriously underfunded (see for example, Zelin 1984, Chu 1969). The imperial governments tried to maintain the myth that nominal salaries were enough to sustain the bureaucracy, so long as the officials possessed high moral integrity. The reality, however, was that the officialdom was forced to derive an oft-substantial part of its income and administrative monies by exploiting the privileges of public office as prebends – whatever earned was kept as private property. Most commonly, magistrates resorted to fee-taking (guifei and lougui) to hire extrabureaucratic personnel, run offices, provide social services, and not to forget, enrich themselves. Fees were typically collected by unsalaried clerks or outsourced to local elites who shared their extractions with the local yamen. Patronage and prebendalism went hand in hand. As one author wrote, “patronage ties linked together all levels of the bureaucracy through a network of intra-bureaucratic payments” (Hickey 1991, 398).
Earnest efforts were made in the last decades of the Qing dynasty and the Republican era to modernize state administration, but with limited success. As part of the “new policies” campaign, the Qing court tried to eliminate prebendal financing by replacing official’s income from feetaking with increased state-budgeted salaries, but the experiment ultimately failed (Hickey 1991). During the Republican period, the Nationalist government extended salaried bureaucracies down to lower levels of administration (Huang 1993, 229). Previously self-funded clerks and runners were absorbed into the state bureaucracy. However, these newly formalized agents ended up competing for control over resources with the local gentry and ended up undermining administrative capacity (Kuhn 1975). In what Duara calls a process of “state involution,” the proliferation and extractive practices of “entrepreneurial state brokers” subverted the rationalization of state institutions, leading up to the fall of the Nationalist regime (Duara 1988).
The State Structure Under Maoism
What happened to the early state structures and prebendal practices following the Communist Party’s takeover of power? Given our basic knowledge about the Communist system and ideology, it is reasonable to expect that the CCP state would have taken over responsibility of all public functions and abolished entrepreneurial, extra-bureaucratic institutions, which had been 110
a much reviled source of oppression in traditional local society. Indeed, one official study reported that upon taking territorial control in 1949, local CCP forces “annihilated the organizations of the old political power completely” (Wu et al 1998, 38). Yet, it appears that the CCP had in fact incorporated some institutional features of the past into the Communist administrative apparatus. While this claim certainly requires more historical investigation, my preliminary findings suggest a surprising continuity of the basic contours of the traditional administrative structure well into the Communist period.
Following the Soviet model in terms of organizational structure, the CCP government had since the founding of the People’s Republic of China established two separate categories of state bureaucracies: administrative units (xingzheng danwei) and service units (shiye danwei), as reflected in official documents dating back to the 1950s.71 The third major state organizational form was state-owned enterprises (SOEs) (qiye danwei). Distinct from administrative and service units, SOEs were involved in the production of goods, not administration or services provision.
What explains the creation of service units alongside administrative units? Recalling the past, a veteran official in personnel management related (AI 2007-125).
Shortly following liberation, after our cadres and military men had won the territories, they began to establish administrative agencies in the state apparatus. But there were not enough personnel. So everyone involved in social/public services in society were hauled into the service units. These individuals helped filled bureaucratic ranks, working with the administrative units. At that time, the cadre force was not large, yet a lot of work needed to be done after liberation, unlike the situation now. The official’s account suggests that shiye units under communism were functionally analogous to extrabureaucracies under imperial and republican rule – both supplemented a relatively small formal administration, allowing the state to reach the basic levels of society for governance and control. Yet, different from the early periods of governance, shiye units and personnel under communism did not exist in an intermediate realm between state and society; rather, they were identified as state bureaucracies and agents of the communist regime. 71
The term shiye first appeared officially in the “1954 and 1995 Budget Report,” presented at the first meeting of the National People’s Congress in 1955 (Zhao 2003, 10) 111
In terms of finances, state bureaucracies were divided into fully-funded, partially-funded, and self-funded units following state establishment (AI 2007-56; 2007-62; AI 2007-110). Units that were not fully state funded relied on fee collections or commercial activities. To give an illustration, one official study reported that in 1961, a total of 93 units and 2,811 public personnel in Tianjin city were funded by the collection of fees, including administrative fees (xingzheng fei) and service-business fees (shiqiye fei) (Wu et al 1998, 177). In addition, even during the pre-reform era, there existed commercialized (yingyexing) service units that earned profits (AI 2007-56; AI 2007-125). Some examples included car fleets, hotels, water supply companies, real estate companies, grocery retail centers, grain processing plants, and so forth. Such titles as plants (chang) or companies (gongsi) hints at the commercialized nature of some extrabureaucracies.72 As one cadre explained, “actually these units were a type of state-owned enterprise, except they were not involved in production” (AI 2007-125). As we will see, many of the profit-making public bureaucracies studied in the 1980-90s were in fact not new, but had existed decades before.
Official public financial statistics indicate that state bureaucracies had earned sources of income beyond budget allocations even during the pre-reform period. From 1952 to 1979, the extrabudget was composed of five categories, one of which was titled “revenue of administrative units and institutions” (xingzheng shiyexing shoufei), composing on average 21 percent of extrabudgetary revenue in the pre-reform period. As discussed in Chapter 2, virtually all governments collect fees and service charges as part of public financing; what is uncommon, however, is for bureaucracies that collect the income to own the income. It was unclear from the aforementioned statistics whether bureaucratic units had owned the revenue they collected or were only responsible for collecting revenue during the pre-1979 period. It is also unclear whether and how much discretion local bureaucracies had in using self-financed income in the pre-reform period as they later did in reform era. An official dataset compiled by the Shandong provincial finance bureau from 1979-1997 suggests that individual offices did have ownership over income collected or earned. Extrabudgetary funds were divided by management rights.
72
Earlier official documents often referred to state-owned enterprises and service units collectively as qishiye danwei. 112
Administrative and service units (shiye xingzheng danwei) were listed as one of the agents with management rights, alongside the local finance bureaus.73
Comparing the preceding discussion with the account of bureau-contracting in Chapter 2, we observe some striking similarities between the administrative structure of the past and of the present. The extrabureaucracies (shiye danwei) in post-Mao China are likened to the extrabureaucracies of dynastic China; both conduct the bulk of administrative and social service functions, but are distinct from and subservient to a small salaried civil service. Even during the pre-reform period, part of the state bureaucracy, it appeared, was not fully salaried, but had been “self-financed” through collecting fees and selling services. More importantly, individual bureaucracies appeared to have rights to dispose the income they collected, although the existence and extent of such rights require more validation. Prior to reforms, it was likely that any income made by the bureaucracies served only a subsistence purpose, but not to incentivize offices for pursue rich rewards. Incentives, however, changed dramatically when Deng and his reform-minded lieutenants took office and opened markets in 1979.
The Policy Context of the Reform Period
In March 1978, a strategy for prosperity struck a party secretary of Lin village in Fujian province named Ye. He tried to sell his reform ideas to his supervisors, the commune party cadres. The first idea was free election at the team level, and the second was a bonus system. “I suggested that future team leaders be authorized to keep a certain percentage of the surplus from the team’s production as a bonus. The team head used his discretion to reward hardworking team members with this bonus. Only by controlling this economic leverage could the team heads promote production in their teams” (Huang 1998, 137). The Commune Party Secretary threatened to fire him for his audacious proposal. Fortunately and unexpectedly for Ye, the political winds shifted in December 1978. The Chinese Communist Party held its Third Plenum of the Eleventh Party Congress, during which the new paramount leader Deng Xiaoping announced plans for reform and opening (gaige kaifang). Ye related, “My experiments in Lin village turned
73
It is unclear whether this division of extrabudgetary funds was introduced as part of the post-1979 reforms or had previously been in place. 113
out to be in line with the new policy. When the announcement of policy change was made, the Commune Party Secretary congratulated me for having saved my skin this time” (Huang 1998, 139).
The story of party secretary Ye nicely captures the radical shift in policy direction following Deng Xiaoping’s rise to power after Mao’s death. The reform-era state ideology was, as Deng had famously pronounced, “To get rich is glorious!” A frenzy of opportunities to get rich like never before swept up the entire country, from village households to provincial governments and central departments. It was this historic leap from a socialist system on the brink of collapse to a capitalist world under communist party dictatorship that set the stage for making bureaucratic agents who were uniquely entrepreneurial.
In this section, we proceed to examine how the state structure in China, building upon existing patterns, had since evolved. To understand the path of state development in reform-era China, it is necessary to divide the reform era into two distinct periods.
•
Stage 1: The years from 1979-1993 marked a period of partial reforms, during which market forces were introduced but only on the edges of a centrally planned economy.
•
Stage II: From 1993, full market reforms were launched; the central planning system was to be supplanted by a modern market system.
Stage 1 of China’s reforms was a period of high-speed growth accompanied by disordered “state entrepreneurism.” Governmental actors played the simultaneous roles of entrepreneurs and bureaucrats in the partial reform setting. While they were each highly motivated to pursue prosperity, they also exploited the free rein afforded to extract income for their own localities and departments, but to the detriment of central control and sustained market development. Stage 2 heralded an advanced stage for reforms, where fiscal and administrative reforms were implemented to recentralize vertical and horizontal state control over local governments and bureaucracies. It was during this period that state entrepreneurial behavior became more rulesbased and came closer to approximating the ideal-type of bureau-contracting. 114
Importantly, I stress that post-1993 administrative reforms, while successfully regimenting the rules and procedures of prebendal financing, did not succeed in creating a regularly salaried bureaucracy for three reasons: first, increased local budget pressures resulting from the 1994 fiscal recentralization policy; second, expansion of extrabureaucracies as a result of downsizing policies; and third, the self-reinforcing pattern of self-financing.
Stage 1 Reform: 1979-1993
From 1979 to 1993, China retained the basic framework of a planned economy. Market institutions were implemented only step-by-step and in an experimental manner, a strategy famously known as “touching the stones to cross the river.” Through the dual track pricing system, the state controlled the prices of important commodities, while allowing market mechanisms to set the remaining prices. And while state-owned enterprises were dominant, mixed ownership forms, such as township and village enterprises (TVEs), as well as private enterprises disguised as collectives, mushroomed at the local levels.
Within the framework of a planned economy, extensive decentralization and contracting took place across their socialist hierarchy. The household responsibility system decentralized production from communes to individual households, thus encouraging agricultural production. Inter-governmental fiscal contracting decentralized rights over increased revenue to lower level governments, motivating local governments to industrialize and develop. State-owned and collective enterprises contracted with the local governments; the latter provided guidance and support in exchange for profits that enlarged governmental extrabudgetary wealth. And as argued, bureaucracies that compose each level of government were similarly accorded residual rights to income generated. As vividly illustrated in the experiences of village leader Ye, reformminded leaders, from the central party committee down to villages, were eager to employ highpowered incentives to promote the pursuit of wealth in a new entrepreneurial climate.
Although the communist states had all sought to promote “entrepreneurism” and growth in the 1980s, the strategy of reform in China had been differed from that of the former Soviet Union and Eastern European countries in critical ways. In the latter case, the leadership implemented 115
political reform and economic reform simultaneously. Furthermore, those countries opted for the “big bang” approach, uprooting the central planning system and replacing it with a market system in one fell swoop. In China, market reforms were introduced through communist political institutions. And reforms were conducted only incrementally. As Susan Shirk (1993) contrasted,
Gorbachev dismantled the Communist Party’s control over the government and the society and shifted authority to government institutions, including democratically elected legislatures (5). [Deng] opted to retain the traditional communist bureaucratic polity with only minor modifications. He apparently believed that he could use local officials as an effective counterweight to the center without changing the political rules of the game (12). To overcome resistance to change from party elders, Gorbachev introduced democratization and mass participation. His strategy, however, backfired. His reforms scrambled a previously state-centered structure of rewards. Losing faith of the durability of party hegemony and eager to seize upon newly available capitalist opportunities, party-state cadres “stole the state” en masse (Solnick 1998). As one observer vividly described, “There was a fin de regime atmosphere in Moscow in the spring of 1991, and bureaucrats were lining up to jump ship before it was too late... many members of the elite were now discovering that they could maintain their privileged positions in society even without the ideology" (Dobbs 1997, 373).
Contrastingly, in China, the path to individual prosperity was tied to one’s privileged position in the party-state hierarchy. To invert Dobb’s phrase, members of the Chinese political elite could not have done without the ideology. Indeed, another of Deng’s famous quotes – “Let some get rich first” – hints at the state-dominated nature of entrepreneurship during the early stages of reform. Party-state officials were among the “first” to get rich, for although the leadership had let loose capitalist forces, private entrepreneurship remained ideologically taboo. Consequently, leaders and cadres took the lead as entrepreneurs in the 1980s and early 1990s, and in the process, most of them became the prime beneficiaries of partial reforms.
116
Local leaders mobilized capital to run collective enterprises, which led China’s initial drive for industrialization and contributed to local finances. State-owned enterprises and supervisory departments grew rich by exploiting the dual track pricing system, buying inputs cheaply at state controlled prices and then selling them on the market. Individual bureaucracies, of course, did not sit idle at the new opportunities for making wealth. Facing a rapidly expanding market, bureaus and extrabureaucracies could collect more fees, charges, and monopoly rents than before, and with unprecedented discretion over the use of self-financed income. In addition, they could even go directly into business.
In the first stage of reform, the state policies and incentive schemes in China produced remarkable results, especially in contrast to the collapse of the Soviet Union and dismal failures of reforms in Eastern Europe. From 1979 to 1993, the economy grew at an average rate of 9 percent or 7.5 percent on a per capita basis. Consumer consumption increased threefold. Household bank deposits increased from 6 percent to 40 percent. The number of people living in poverty dropped from 250 million to less than 100 million (Qian and Wu 2003, 33). This was a stage described as “reform without losers” because almost everyone, in particular the political elites, had something to gain from the changes introduced (Lau et al 2000).
However, the story was not all rosy. Massive opportunities created for officials to profit from public office back-lashed in 1989. Fed up with the proliferation of corruption and inequality, students gathered at Tiananmen Square in Beijing to protest for democratic change. The incident, as we know, ended tragically with the leadership’s decision to crackdown violently on the protestors. The Tiananmen incident dealt Deng’s reform policies a heavy blow, tainted investment confidence, and empowered the conservative faction in the party led by Li Peng and his counterparts. The three years after 1989 saw a period of retrenchment. To rekindle the flames of reform and remobilize support, Deng embarked on his famous southern tour (nanxun) in early 1992. Deng’s tour was the precursor to the endorsement of a “socialist market economy” as the goal of reform during the Fourteenth Party Congress in September 1992.
117
Stage 2 Reform: 1993 onwards
The year 1993 marked the beginning of full market reforms in China, culminating in the groundbreaking endorsement of a document entitled “Decisions on Issues Concerning the Establishment of a Socialist Market Economic Structure.” The new mantra “socialist market economy” marked a subtle but critical change in the leadership’s ideology and policy program. “Socialist market economy” had socialist as adjective and market economy as noun, that is to say, achieving a market economy was the object of China’s reforms, albeit with socialist characteristics. Interestingly, the new Chinese mantra contrasted the Eastern European slogan of “market socialism,” which had socialism as the noun and market as the adjective (Qian and Wu 2003). Before 1993, market mechanisms, though growing rapidly, remained subsidiary to central planning and public ownership. But after 1993, the leadership decided that markets and plan would have to switch places.
Following the 1993 decision, the Chinese government aimed to construct an institutional framework for a modern market system. This involved a number of comprehensive policy changes (Qian and Wu 2003). First, the central government introduced tax and fiscal reforms to clarify the tax structure and recentralize control over fiscal resources. Second, the state-owned enterprises would undergo mass privatization; consequently, the government had to build a social safety system that had previously been unavailable. In tandem with SOE privatization, the leadership resolved to promote and protect private entrepreneurship, including through the clarification of the property rights regime. Third, the bureaucratic apparatus had to be transformed along with the rest of the economy. State agencies should no longer be administrators and businesses simultaneously; in the new administrative order, they were supposed to function as impartial regulators (Yang 2004). This final task, it would appear, amounted to a major program of bureaucratic rationalization along Weberian principles.
As comprehensively reviewed in Yang’s (2004) recent study, the central government instituted a host of reforms to discipline the administration. Anti-corruption control was considerably strengthened through anti-graft laws, anti-money laundering policies, and technical but critical improvements in the state accounting, auditing, budgeting, and procurement systems. The 118
budgeting measure of “separating revenue from expenditure,” earlier discussed in Chapter 2, was introduced in this backdrop of administrative reform. Further, the central government embarked on two consecutive downsizing campaigns in 1993 and 1998, the latter being the most ambitious in China’s reform history. Administrative licensing procedures came under increased standardization through the enactment of a national legal framework. Leaders went even further to promote the notion and practice of a “service-oriented government.” One-stop administrative services centers were introduced, along with local experiments that allowed members of the public to evaluate the performance of government officials. These are just some examples of the post-1993 institutional reforms. Among them, the central government’s decision to divest the military and party-state agencies of their business operations was of particular significance.
Fiscal Reform and Local Budgetary Pressures
For the central leadership, fiscal reforms were a critical component of the post-1993 reform package. In 1994, the central government decided to replace particularistic fiscal contracting with a uniform tax-sharing system between the central and provincial governments. Although it had resolved some old problems, the tax-sharing system created new problems. The tax-sharing system clarified the terms of tax sharing between the central and provincial governments, but left the thorny issue of sub-provincial fiscal relations unclear. At the sub-provincial level, it was bargaining as usual when dividing the local and shared taxes. Worse still, the central government had recentralized revenue without adjusting local spending responsibilities. As a result, local governments generally suffered increased budgetary pressures after 1994.
Figure 5.1. below shows the changes in aggregated local governmental revenue and spending from 1980 to 2005. Before 1994, the sum of local revenue was equivalent to local spending. In the hey days of the early 1980s, local share of total revenue consistently exceeded its share of total spending by up to 30 percent. Immediately after 1994, local revenue took a nose dive. While local revenue has continued to grow in absolute level afterwards, its rate of growth has clearly fallen behind that of spending, resulting in a widening deficit. In addition, local
119
governments have had to shoulder a disproportionate share of public spending even as its share of revenue dropped over the years.
40
1000 1500 2000 2500 3000
Figure 5.1: Local Revenue and Spending Before and After the 1994 Fiscal Reform
0
-40
500
-20
0
20
After 1994 Fiscal Reform
1980
1985
1990
1995
2000
2005
Local government revenue (billion) Local government spending (billion) Local share of total revenue minus share of spending
How could local governments fill the deficit? Local governments (provincial and sub-provincial governments) are legally barred from borrowing. They are also required to balance their budgets. Since 1994, the central government distributed increasing amounts of fiscal transfers to local governments. However, as many experts have pointed out, China’s fiscal transfers system remains under-developed.74 In many localities, particularly below the provinces and cities, it would be difficult, if not impossible to subsist only on retained taxes and fiscal transfers. It appears then that extrabudgetary revenue, which belongs entirely to the localities, became more important to local finances, particularly at the county levels and below, after fiscal recentralization.
Yet, even as local governments needed extrabudgetary revenue, local extrabudgets were taking a hit of a different kind in the 1990s. State-owned and collective enterprises, including the TVEs, began to incur huge losses. These enterprises were no longer cash cows but a burden to local 74
Citing an earlier World Bank (2002) report, Wong (2005) described the current inter-governmental fiscal arrangements as “dysfunctional.” 120
governments. SOEs incurred losses because of persistent inefficiencies and increased market competition. Market competition had grown steadily over the 1990s. The state pricing system was largely obsolete by the 1990s, and by 1993, only 10 of product prices were fixed according to plan (Qian and Wu 2003, 36). The economy had gradually shifted from a “seller’s market” to a “buyer’s market,” as access to resources freed up beyond the state sector. Losing their advantages, SOEs soon went deep into the red. TVEs were not spared from huge losses either (Kung & Lin 2007).
Bureaucracies
2003
2002
2001
2000
1999
1998
1997
1996
1995
1994
1993
1992
1991
1990
1989
1988
1987
1986
1985
1984
1983
1982
1981
1980
0
Extra-budgetary revenue (billion yuan) 100 200 300 400 500
Figure 5.2: Change in Composition of Extrabudgetary Revenue, 1980-2003
State-owned enterprises
Local governments
Township self-raised funds
Government Foundations
Others
Widespread collapse and privatization of SOEs and TVEs left a huge gap in local extrabudgetary coffers. Moreover, profits from state-owned and collective enterprises were shifted to the within-budgetary category after 1994 (Oi 1999). Nationwide, retained profits from state-owned and collective enterprises used to be the leading source of extrabudgetary revenue in the 1980s and early 1990s (Oi 1999; Shue, etc.). Figure 5.2 shows the composition of extrabudgetary revenue before and after the 1994 fiscal reform. Although the data in the two periods are not strictly comparable because of amendments in the classification of revenue, it nonetheless suggests that dramatic changes had taken place in the composition of extra-budgets. Before 1993, retained profits from enterprises had composed about 75 percent of total extrabudgetary revenue, followed by self-financed income belonging to individual bureaucracies, and finally 121
extrabudgetary revenue belonging collectively to local governments. But from 1993 onwards, the lion’s share of extrabudgetary revenue came from the bureaucracies.
Based on the patterns described, we may gather that as local share of revenue fell, retained profits from SOEs and TVEs also tumbled and huge losses were incurred. In a double whammy situation, it was also during this period that state agencies were forced to severe themselves from the corporate offsprings that used to supply profits to the bureaucratic sector (Yang 2004). These developments implied that local offices likely faced more intensive pressures than before to self-finance.
Expansion of the Extrabureaucracies
From the mid-1990s onwards, budgetary pressures intensified and many departments were forced to close loss-making commercial operations. Simultaneously, the extrabureaucratic sector was pushed upon an accelerated track of expansion and marketization. From the 1979 onwards, China’s bureaucracy, measured as the total number of state personnel and on a per capita basis, has grown steadily and at a steeper rate than prior to reforms, despite a transition from socialism to capitalism. Notably, growth in absolute personnel size has been concentrated in the extrabureaucratic sector, while the core administration has held steady at a share 20 percent of total employment throughout the post-1979 period.
A major cause of the physical expansion of the extrabureaucratic sector was, ironically, the central government’s downsizing campaigns. The Chinese government had conducted regular downsizing campaigns from the 1950s onwards. In particular, the restructuring program in 1998, spearheaded by Zhu Rongji, was the most ambitious to date. The 1998 campaign slashed the number of central ministries from 41 to 29 and central-level personnel under the state council by half. Downsizing was extended to sub-national governments. It was reported that provincial governments had successfully reduced their number of organs by 20 percent and personnel size by 47 percent on average, and sub-provincial governments by 20 percent (Brodsgaard 2002, 375-6; see also Yang 2004, 37-38). Citing media releases, Yang reported similarly that downsizing had achieved impressive results. Guangdong province was said to have cut its staff 122
size by 49.3 percent. Hainan province trimmed 50 percent of governmental personnel and 20 percent of party personnel. Yang concluded optimistically, “the Chinese leadership has by the beginning of the new century achieved remarkable progress in streamlining and downsizing the government… in the late 1990s, the pace of government reforms, measured in terms of downsizing… picked up noticeably” (2004, 17 and 25).
0
1
Public Employment Per Capita 2 3 4 5
6
10 20 30 40 50 Number of Public Employees (millions)
Figure 5.3: Change in Public Employment Size, 1955-2003
1955 1960 1965 1970 1975 1980 1985 1990 1995 2000 2005 PE per 100 population
PE per 100 working population
Total public employment
Employment in core bureaus (jiguan)
While the results claimed are impressive, it is important not to take official statements at face value. As John Burns rightly cautioned, “previous research fails to examine the outcomes of downsizing attempts and sometimes reports changes to bianzhi targets as though they were actually achieved” (2003, 776; in reference to Brodsgaard 2002). Bianzhi, like the term shiye, is yet another Chinese term that is confusing to most outsiders and hard to translate. In my view, the closest translation is from Burns, who defined bianzhi as “all the positions that have been officially created” (2003, 776). From the central to county levels, the party-state sets and enforces targets for official public employment through the establishment offices (bianzhi weiyuanhui bangongshi). Most of the positions that fall within the bianzhi target are budgeted, i.e. they are eligible for state funding and will be considered during negotiations for budget 123
appropriations (see Chapter 2). 75 Positions that exceed the hiring quota are known as “exceeding the bianzhi” (chaobian). Personnel who are hired but not reported as part of the bianzhi are known as “outside the bianzhi” (bianzhiwai).76 In both of these cases, it is usually the hiring unit that pays the non-official personnel using its self-financed income, although there are always exceptions when local finance authorities agree to subsidize personnel costs using state funds.
To put it simply, then, the bianzhi is an instrument that allows the government to control the ideal size of bureaucracy. As illustrated in Chapter 2, it is also a tool (although sometimes an imprecise one) for budgeting personnel costs. In most governments, equivalent bianzhi institutions do not exist because the size of hiring is controlled through formal budgeting processes. The final budget passed by the legislature determines the number of personnel that each office could hire and how much they each would be paid from state revenue. However, in China, procedures for formulating budgets, and even more importantly, enforcement of budgetary decisions are still immature. Therefore, the bianzhi system provides a partial personnel budgeting function. Alternatively, in a socialist framework, we can think about the bianzhi system as a centrally controlled “plan” for distributing labor in the party-state bureaucracy. However, bearing in mind that since it is just a plan, the reality may deviate significantly from the plan.
Having explained the bianzhi system, we may now better understand how it could have been possible for the central and local governments in China to downsize their personnel by as much as half in a remarkably short time and without mention of any detailed compensation scheme. After all, the task of laying off bureaucrats is extremely politically sensitive anywhere, and not to mention, in a socialist system like China where party-state cadres are guaranteed lifetime employment. If massive layoffs caused by SOE privatization in the 1990s triggered widespread workers’ protests across the country, then why had the cadre class kept silent when 20 to 50 percent of them had reportedly lost their jobs? There are two answers: the first is that the
75
Not all of these official positions, however, are budgeted because even wholly self-funded public units are subject to a bianzhi quota. 76 This happens when units hire more personnel than they are authorized to and when units hire temporary workers like cleaners and drivers. 124
central government had cut bianzhi targets, but not the actual number of personnel; the second is that downsized personnel is systematically shifted from core administrative units to extrabureaucracies. These answers are clearly reflected in Figure 5.3. Total public personnel size increased steadily each year, and the expansion is concentrated among the extrabureaucracies.
Flexible Bianzhi. So why were downsized cadres moved from core bureaus to extrabureaucracies? One reason is that while the core administration was subject to strict hiring limits, while the extrabureaucracies had a more elastic quota. To elaborate, we return again to the bianzhi system. As we are familiar by now, there are two organizational components in the Chinese bureaucracy: core bureaus and extrabureaucracies. In tandem, there are two kinds of positions: official positions in the administrative bureaus (xingzheng bianzhi), and official positions in the extrabureaucracies (shiye bianzhi).77 The central government sets a fixed hiring quota on the former category; this quota is then transmitted from level to level. For example, the central government may set a fixed quota of 500,000 administrative bianzhi. The province of Shandong is allotted a bianzhi quota of 50,000, leaving 450,000 for the remaining provinces. Shandong then distributes this quota among its city governments, and so forth. For the extrabureaucracies, the bianzhi is not limited by a fixed quota, rather it is determined by a ratio (hebian biaozhun) that can be readjusted based on local conditions (AI 2006-15). For example, each school may be allowed to hire additional personnel for every 100 increase in the number of students served. In another example, a county finance official reflected that the xingzheng bianzhi had been frozen since 1993. To recruit new personnel, the bureau could request to add an extrabureaucratic function and hire more shiye bianzhi, whose tasks were in fact identical to colleagues positioned in the core bureaus (AI 2006-16).
Because of the relative hiring flexibility in the extrabureaucracies, core governmental departments typically conformed with downsizing mandates by reclassifying their staff from the category of xingzheng to shiye bianzhi (see Burns case study). In some instances, such reclassifications involved physical transfers of staff. In most cases, however, no physical change occurred. The downsized personnel stayed in exactly the same job, except they no longer 77
Following the establishment of the civil service system, functionaries who occupy the xingzheng bianzhi are officially identified as civil servants (gongwuyuan), who are legally protected from unjustified dismissal. 125
occupied a post in the administrative quota, but instead a shiye bianzhi. In a third possible scenario, which I found common, core bureaus “borrowed” bianzhi from extrabureaucracies; this practice is known as jiebian. For example, education bureaus “borrowed” teachers from public schools to work in the bureaus as administrators (AI 2007-42; AI 2007-121). What are the benefits of doing so? Not only would the education bureau not violate the administrative hiring quota, “the schools will have to pay for them [the borrowed personnel]” (AI 2007-44). For the core bureaus, jiebian allows them to expand recruitment without having to incur additional financial obligations.
Flexible Revenue-Making. Extrabureaucracies have more flexibility in hiring because they have more flexibility in generating self-financed income. Particularly after the post-1993 administrative reforms, the state has made it clear that core agencies are supposed to be engaged in administrative and regulatory work, not profit-making. Regulatory agencies (e.g. environmental protection bureaus, commerce bureaus, etc.) could collect fees and fines as part of their regulatory functions, but they could not directly be involved in other forms of commercial activities. Extrabureaucracies, on the other hand, are engaged in services provision, vaguely defined. Through the provision of state-related services, they could not only collect fees but also sell their services to consumers for profits often at a monopolistic advantage.
As a World Bank (2005) study reported, extrabureaucracies were increasingly “pushed into the market” in the reform era. As the study usefully, “The central measure of reform along this direction is to cut of budget funding to [extrabureaucracies] to force them to generate their own revenues to survive. ‘Contract responsibility systems’ of various forms were introduced to implement ‘budget contracting’ (yusuan baogan), thereby the government allocated a predetermined amount of fund to the [extrabureaucracies], which were then supposed to retain any surpluses or make up any funding gaps (World Bank 2005, 91).” Compared to core bureaus, extrabureaucracies have more autonomy and avenues for generating revenue, depending on their functional type. As reported by the World Bank, “when a [service unit] is completely ‘pushed into the market,’ it operates almost the same way as a Chinese SOE. This gives rise to a formal category of [service units] that are ‘managed as enterprises’ (shiye danwei qiyehua guanli)” (World Bank 2005, 92). 126
A Bifurcated Bureaucracy. The rationale for maintaining a bifurcated bureaucracy becomes clear if we recall the historical bureaucratic structure. Following the establishment of the civil service system in 2005, functionaries who occupy the xingzheng bianzhi are identified as civil servants (gongwuyuan), who enjoy a set of legal rights and obligations. Only individuals who have passed the civil service examinations and are formally appointed to office may occupy the administrative bianzhi. The modern-day civil servants in China are likened to the imperial officials of the past. They represent the elite strata of the hierarchy and occupy a special privileges status in society. This stratum must be kept unadulterated at a high quality, for the state depends on this subset of distinguished individuals to govern the country.
The extrabureaucracies, on the other hand, are like “shock absorbers” in China’s reform machinery. Shock absorbers are installed in cars to cushion the impact of kinetic energy, thereby protecting the structure of the automobile. Likewise, extrabureaucracies help to provide revenue autonomy (when local revenues were cut), flexibility in hiring (when the central government froze the “administrative” bianzhi), and labor absorption (when bureaucrats from local governmental agencies were downsized). Extrabureaucracies also allow local authorities, despite fiscal constraints, to sustain and perpetuate a clientelist network through the distribution of public jobs, contributing in part to the cohesion of party-state. Contrary to the view that the CCP has “[failed] to adapt to a changing environment in the post-revolutionary period” and is threatened by “a process of involutionary development” (Lu 2000, p. 229), we find that the regime has adapted in quite remarkable ways.
The Self-Reinforcing Logic of Self-Financing
Apart from increased local budgetary pressures and expansion of the extrabureaucratic-state, there is yet a more basic explanation for why it is so difficult to turn a partially self-financed bureaucracy into a purely salaried establishment. As highlighted in Chapter 3, official cadre wages are not only low in China, but also extremely egalitarian. Wages are tightly compressed and do not reflect varying local economic conditions, meaning the wage gap between the lowest and highest ranking personnel and those in the richest versus poorest locales is extremely small. It is self-financed income earned by individual departments and locally retained tax revenues 127
that determine actual levels of cadre compensation. That means, if the state removes the rights of bureaucracies to collect revenue for themselves, it would either have to (a) compensate them for the loss in income through increased budget allocations, or (b) cadres nationwide will be forced to accept a lower level of compensation and some wholly self-financed bureaucracies would go unpaid. Option (a) is economically unfeasible, and option (b) is politically unacceptable. To illustrate my point, consider the following budget statistics from county governments in Shandong province. In any local government, there are three major sources of revenue: tax revenue, fiscal transfers, and self-financed income belonging to individual offices. In this particular dataset, we find that each of these revenue categories compose roughly one third of total combined fiscal resources. The grey bar shows the average combined sum of retained taxes and fiscal transfers in the Shandong counties. The red bar shows the average level of cadre benefits, i.e. wages and in-kind benefits combined. If the county governments barred bureaucracies from raising their own income or denied their claims over income generated (in which case the bureaucracies would have no incentive to generate self-financed income), that would amount to eliminating one third of the county’s fiscal resources. In a situation like this, as shown clearly in Figure 5.4, nearly all of the remaining revenue in Shandong’s counties would go towards cadre compensation and administrative monies, leaving little else for infrastructural development, purchase of equipment, and public goods provision. In fact, in the year of 2002, the mean county government in Shandong will have to bankrupt itself just to maintain the original level of cadre benefits.
This example clearly illustrates to us what bureaucrats in China actually mean when they claim that budget allocations are “insufficient.” The critical question is: When are budget allocations ever sufficient? How much is objectively sufficient? In most governments, budgetary funds are allocated based on the number of personnel and the formal salaries that each personnel is entitled to in a particular office. In this regular context, budgetary funds are considered insufficient when there is not enough allocated to pay every bureaucrat his or her formal compensation. In a neo-prebendal context like China, however, the official pay scale lacks meaning because in many locales and departments, actual compensation far exceeds what is formally provided. There are certainly cases of poor departments in poor locales that have trouble paying even minimal formal wages; in this case, we may consider budget allocations 128
objectively insufficient. But in many other instances across the country, no such “objective” standard exists: cadres are rewarded based on what they can make, and offices can hire as many personnel as they can independently finance. If there is in fact no limit to an official’s actual compensation or an office’s actual size, then budget allocations are always insufficient.
Amount of Spending/Income (10000 Yuan)
Figure 5.4: ‘Eating budget’ – Cadre rewards consume tax revenue and fiscal transfers 60,000 50,000 40,000 30,000 20,000 10,000
0
2001
2002
2003
tax revenue and fiscal transfers
2004
2005 cadre benefits
My point is best illustrated in the case of Blossom county, already one of the richest in the entire country. As I was told, all bureaucracies in the county make self-financed income (AI 2007-114). This is puzzling. If Blossom county is already so wealthy, why does the government not pay its bureaucracy in full? Why not do away with self-financing, especially given the problems created? For the local officials, the answer is obvious: there is not enough money for increased budget allocations. As one official explained (AI 2007-113):
This suggestion [of a fully salaried bureaucracy] is not bad. I raised it to the former finance bureau chief once. He replied, ‘No way, that will take tens of millions!’ At that time, our finances were not good enough. Now it’s even worse. The other day the public security chief told me that that the supporting police force alone has over 400 personnel. How much do you think that is going to cost? Interestingly, when the interviewee above remarked that “now it’s even worse,” he did not mean that Blossom’s public finances were getting worse. Blossom’s tax revenues are impressive and growing every year. Instead, what he had meant by “now it’s even worse” is that personnel 129
size had increased again. The growth of formal revenue could not keep up with the growth of actual administrative costs. And actual administrative costs have kept growing because individual departments make discretionary sources of income. In other words, a system of prebendal self-financing, once allowed to take off, becomes a self-reinforcing institution.
The self-reinforcing nature of prebendal financing is not unique to contemporary China. It is an enduring problem in the governance of early states. Hickey (1991) detailed a fascinating account of the late Qing government’s efforts to introduce a comprehensive salary reform in 1909-1911. Formal salaries had been persistently inadequate in the Qing government, thus officials had to rely heavily on fee-taking to upgrade their compensation and cover the costs of running office. The objective of the proposed salary reform was to provide large increases in formal salaries in a payment known as gongfei and thereby abolish fee-taking activities. Reformers undertook painstaking efforts to induce officials to reveal their true income, promising not to punish them for hitherto reported funds, in order to determine the increases necessary to offset the officialdom’s loss in revenue from fees. Nonetheless, in the end, the salary reform still failed to eliminate fee-taking because despite significant increases in state-authorized income, the gongfei provided fell short of allowing the officials to maintain their original levels of consumption. Furthermore (and strangely), the reformers failed to consider providing salaries for unpaid yamen clerks and runners that constituted the vast sub-bureaucracy in the Qing dynasty. As Hickey explained, Qing reformers had underestimated the estimation of the gongfei schedule because “only customary fees deemed ‘legitimate’ were to be converted to the new forms of remuneration” (411). The author concluded, “By denying the legitimacy of the profit gained in office, the Qing state was defying the expectations of its officials. To meet such expectations, salary reform needed to provide not only the actual costs of holding office, but also financial rewards for the officeholders” (413, emphasis added).
Similarly, in the case of present-day China, leaders and cadres expect to sustain (or increase) their existing levels of income and benefits, and to keep positions that had been created without state funding. If the state were to deny bureaucracies of their property rights to income generated, the bureaucracies would have no motivation to self-finance. And if self-financed income falls, cadres would have decreased financial rewards and offices would have insufficient 130
operational monies. The central and local governments may literally have to empty its coffers to match the bureaucracies’ income from self-financing, and still, that may not be enough.
My identification of the self-reinforcing logic of prebendal financing calls for a reconsideration of observations in the literature that bureaucracies in China have had to collect fees and charges or run commercial operations because of “inadequate funding.” There are certainly cases in which state funding is inadequate, for example, when cadres cannot even make enough money to make ends meet. But it is important to understand that because Chinese bureaucracies are, by design, quasi-firm units that profit from public office, no amount of funding is objectively enough to numb their entrepreneurial instincts.
Thus understood, the challenge of transforming a neo-prebendal bureaucracy into a salaried bureaucracy goes beyond implementing modern budgetary measures, issuing central directives, or increasing formal salaries. The challenge is at the core a political one; it involves renegotiating a contract between the ruling party-state and its bureaucracy. The ruling-party state has to provide the bureaucracy with its expected level of material rewards, either in cash or through benefices. A salaried bureaucracy cannot easily be achieved, contrary to some expectations, just by requiring bureaucracies to hand over their proceeds to the state treasury (Yang 2004; see Chapter 2).
Conclusion
This chapter has reviewed the evolution of China’s bureaucratic structure from 1979 to the present period. Evidently, the leadership’s vision of reform has had profound effects on the trajectory of state institutional development. As we have seen, Deng and his reformist lieutenants changed the policy landscape dramatically in 1979. The reformers had constructed their policies upon an existing economic and political structure that had already been more decentralized than in the former Soviet Union (Shirk 1993), as well as a long and persistent legacy of prebendal patterns of governance, reviewed in Chapter 6. Deng’s reforms unleashed a rush of entrepreneurism among the bureaucratic class. From then, high-powered economic incentives were borne. During the first stage of reforms, they took off in unplanned ways. 131
From 1993 onwards, as part of the central state’s plan to build a “socialist market economy,” ambitious fiscal and administrative reforms were launched. These reforms, I have argued, did not in the produce a regular salaried bureaucracy. In fact, the new tax measures and downsizing campaigns in the 1990s added pressures on bureaucracies to self-finance, as well as to expand and marketize the subordinated extrabureaucracies. I have also shown why, independent of the reforms, self-financing has inflated the actual levels of cadre rewards. As such, it has proven more difficult than expected to realize an adequate compensation scheme that could allow the state to institute a purely salaried civil service. Confronting these challenges, I argue that the state in China has opted for a pragmatic approach. The approach has been to incrementally regulate opportunities for bureaucracies to profit from public office, rather than to outlaw them (or to pretend they could be outlawed). This, obviously, is still work in progress, but it appears the Chinese government is getting better at its tasks. The perfection of this approach is bureaucontracting.
132
CHAPTER 6
CONCLUSION
“Where the ‘model bureau’ does not exist, it is obviously futile to ask questions about what does exist as though it were a ‘model bureau.’ The first task is not to make this assumption, but to ask: ‘What does in fact exist?’ One may discover, of course, that what exists is not at all a bad thing.” Fred Riggs, Public Administration in Developing Countries, p. 9 “The history of the McDonald’s System is the story of an organization that learned how to harness the power of entrepreneurs – not several but hundreds of them… In essence, the history of McDonald’s is a case study on managing entrepreneurs in a corporate setting.” John F. Love, McDonald’s: Behind the Arches, p. 8
Heeding Fred Rigg’s advice in a classic written nearly 50 years ago, this study began with a simple empirical task to discover “what does in fact exist” in China in the present day, if not a “model bureau.” What I found in the end was a state organization not entirely unlike that of franchised firms like McDonald’s. In a bureau-contracting structure, bureaucrats and contractors are fused. Bureaucrats are incentivized to maximize revenue though developmental efforts and state-sanctioned strategies of self-financing. High-powered incentives in any organization, however, induce individual players to “game the system” by scrapping gains for themselves at the costs of collective interests and organizational integrity. The practical challenge in a bureaucontracting state is not simply a problem of control. Rather, the challenge is to harness the entrepreneurial energies of bureaucratic agents, but at the same time, force conformity of entrepreneurial behavior to standardized rules in a highly decentralized and massive hierarchy. Borrowing Love’s phrase, in essence, the history of China’s bureaucratic structure is a case study on managing entrepreneurs in a bureaucratic setting. Once we understand this central 133
organizational logic, the particular features of the bureaucratic structure in China, though unusual and perhaps even unthinkable in view of prevailing doctrines in public administration, would make sense. And indeed, as Riggs suggested, “what exists is not all at a bad thing.”
Development, Reform, and Second-Best Institutions
The model of bureau-contracting and my findings in the Chinese case challenge our conventional wisdom about the institutional foundations of state capacity building and market development. As Weber asserted, “capitalism and bureaucracy have found one another and belong intimately together” (Weber, 1465). According to Weber, legal-rational bureaucracies have two defining features: first, expertise and meritocracy; and second, the separation of public office from the private property of officials. While the Chinese government has advanced the first feature, it has retained and institutionalized the second.
So why has China been able to achieve state-led development without a Weberian bureaucratic structure necessary for effective governance? The arguments laid out in this study suggest several answers. Firstly, consider what we mean by “effective.” The effectiveness of an instrument rests on the objectives of its use. If we desire an honest, time-sheet-logging, rulefollowing administration that shies from infringing upon private property rights, then Weberian bureaucracies are effective. Alternatively, if we desire a fiercely motivated, interventionist bureaucracy capable of doubling as entrepreneurs in a start-up, impoverished environment where pure private entrepreneurship is non-existent or marginal, then Weberian bureaucracies may not be as effective. A uniquely high-powered bureaucratic structure was suitable for China’s initial stage of reforms, in light of partial reforms and fiscal constraints. Secondly, as reforms matured, China has been able to sustain market development without a Weberian bureaucracy because the refinement of bureau-contracting mechanisms allowed the state to construct a stable flow of rents for the cadre corps. Thirdly, stable rents need to be matched, critically, by developmental incentives that prevail so long as local governments are entitled to a significant portion of marginal tax revenue, even after the 1994 tax reform, as demonstrated in my analysis of dual fiscal incentives.
134
My story, in other words, attends to varying institutional demands at different stages of development, a concern earliest expressed by Gershenkron (1962). In the realm of institutional design, there has been a long-standing debate between two schools of thoughts. One is the “first-best institution” school of thought. This school is exemplified by the Washington Consensus and embraced by multilateral organizations like the IMF. It advocates exporting the basic institutions of Western developed countries to developing and reforming countries. Examples of first-best institutions or practices bring to mind legal-rational bureaucracies, a secure private property rights regime, independent judiciaries, regulatory framework for free market competition, elimination of trade barriers, and so forth.
However, the “second-best” institutional school rejects the procrustean approach. This school contests that institutions should be tailored to reflect local knowledge and initial conditions (Rodrik 2000; Berkowitz et al 2003). As Rodrik, one of the strongest proponents of the secondbest paradigm, expressed, “I shall argue that dealing with the institutional landscape in developing economies requires a second-best mindset. In such settings, a focus on best-practice institutions not only creates blind spots, leading us to overlook reforms that might achieve the desired ends at lower cost, but can also backfire” (2008, 100-101). On the same token, Qian (2003) argued that China’s reforms have succeeded through the adoption of a range of transitional institutions from dual track pricing to TVEs. According to Qian, transitional institutions work “because they achieve two objectives at the same time – they improve economic efficiency on the one hand, and make the reform compatible for those in power on the other” (305).
My view of the evolution of the Chinese bureaucracy builds on both the first and second-best paradigms, but with a twist. Picture China’s political leaders choosing between two institutional models: bureau-contracting vs. the Weberian ideal-type. At the initial stage of development, I posit that bureau-contracting was the first-best choice compared to a low-powered Weberian structure because it could produce more efficiency gains and political capital. However, as discussed, bureau-contracting was then at its infancy and the mechanisms had yet been perfected. At the post-1993 stage of development, as reforms deepened, bureau-contracting became the second-best choice relative to the Weberian model, as the latter is more compatible 135
with a modern market system. Ironically, it was only from this stage onwards that the government developed the capacities to realize the bureau-contracting model in its entirety. In short, what at first had been a first-best model transformed into a second-best reality because growing out from an initial mode of governance proved more difficult than later administrative reformers had wished.
Is Bureau-Contracting Transitional?
Qian’s (2003) perspective on transitional institutions points usefully to the political interests and initial constraints that condition institutional choices. In line with his view, we may consider bureau-contracting a transitional institution insofar as “they incur high costs and generate lower benefits than some alternative institutions if other complementary institutions are in place” (2003, 323, emphasis added). That is, if China had already established a free market system, an effective taxation apparatus, and had not been ruled by communists, the Weberian model would have been preferable to a bureau-contracting arrangement. However, I question Qian’s predictions that “transitional institutions can be superseded by conventional best-practice institutions when more development and reform take place” (2003, 326). Citing the disappearance of dual track pricing and TVEs, Qian asserted that transitional institutions “should not be viewed as permanent” (2003, 323).
Should we expect the Chinese bureaucracy to eventually transform into a salaried Weberian structure like the ones seen in the East Asian NICs, Western Europe, and the U.S.? As discussed, China appears in some respects to be inching towards international best practices, most significantly, through the establishment of a civil service law. But how long might it take before prebendal practices are completely replaced with regular budget allocations like in most governments? As I see it now, the Chinese administration appears to be undergoing a protracted transition.
My point is not that China may never create a salaried bureaucracy. Quite the contrary, consistent with the views of several others (Qian and Wu 2003; Yang 2004), I see Chinese policy-makers aiming to take state institutional development in a direction aligned with 136
“international best practices.” Yet, we must separate aspirations from reality. Discussions in the preceding sections have underscored the particular challenges and paradoxes confronting Chinese government in its struggle to Weberianize the bureaucracy from 1993 onwards, a period of transition from partial to full market reforms. All western developed states were at some point governed by prebendal bureaucracies; the transformation from prebendalism to bureaucratization was, as Weber emphasized, a very long process. In China, this long process is exacerbated by its historical-structural legacies, communist political system, multi-leveled administration, sheer large size, and an initial decision to unleash entrepreneurial forces among bureaucratic agents.
Facing the constraints described above, I argue that the leadership has made the pragmatic choice of regulating, rather than eliminating (or pretending they could eliminate), opportunities for bureaucratic agents to profit from public office. Long-existing self-financing practices were gradually regimented through an elaborated internal system of “policy awards.” In this way, the state compromised by allotting licenses for extracting fees, fines, and monopoly rents on the one hand, and punishing agencies for arbitrary extraction that goes against state rules on the other. Furthermore, the bureaucracies’ self-financed income was brought into the state bank accounts system. As such, the state compromised by allowing bureaucracies to retain a share of their revenue legally on the one hand, but limited discretion over the use of the income and subjecting the funds to supervision by financial authorities. Hence, over time, the Chinese state came closer and closer to approximating a bureau-contracting ideal-type than a Weberian model.
The Chinese experience reminds us of Rigg’s (1964) prismatic model of administration in developing countries. Riggs envisioned that structures of administration, like white light passing through to emerge diffracted upon a screen, can be caught captive in the middle, “imprismed.” Wisely, he probed, “May not some ‘transitional’ conditions turn out to be relatively permanent? Can we be certain, for example, that the present stage of administration in Haiti or Bolivia or Afghanistan is temporary and transitional rather than permanent and final? Or if these societies are undergoing change, cannot the same be said of England, France, and Canada?” (1964, 4). Reflecting on Chinese administration, perhaps we should reconsider the assumption that 137
transitional institutions are stepping stones to some ideal rationalized outcomes found in the West. Some transitional institutions could, as Riggs suggested, turn out to be relatively permanent.
Managed Rents, Corruption, and Compromises
Corruption is an oft-bandied about term in the study of governance. What exactly is corruption? Should we take a strictly legalistic definition (Nye 1967) or broader definitions from the public interests perspective (Gardiner 2000)? If we use the latter approach, the problem is, as Rose Ackerman pointed out, “in many societies, no such clear distinction exists [between one’s public and private roles]” (1999, 91). China is an instance of a society in which public and private spheres overlap.
My study of China urges that we distinguish state-managed rents provision from rules-violating acts of corruption, e.g. stealing public funds and taking bribes. Drawing this distinction is analytically important, because while illicit acts of corruption corrode hierarchical authority or are signs of corroded authority, certain organized schemes of generating private gains from public office may actually contribute to authoritarian state strength and even reform agendas. Bureau-contracting illustrates the latter possibility. The issue here is not one of semantics, but contextualization and institutional design.
In China, the state organizes and disciplines processes of rents extraction. In the former Soviet Union under Gorbachev’s reform initiatives, opportunistic bureaucrats “stole” the state en masse, leading eventually to organizational disintegration (Solnick 1998). By contrast, in China, the state makes the “rules of the game.” To get along and ahead in the political system, Chinese officials played by those rules than in defiance of them. To borrow Solnick’s “bank-run” analogy, the state in China continued to run the bank following reforms, and its agents lined up to collect their fair share of dividends.
138
Bureau-Contracting as an Alternative Model
Transactions are the building blocks of all institutions. Transactions, as neo-institutional economists tell us, take place either through markets or within hierarchies. Markets and hierarchies are themselves ideal-types. They bracket a whole range of mixed realities with qualities of the two extremes in varying proportions. There is a third institutional ideal-type that we have yet to explore: the union of markets and hierarchies. Alternatively, in diagrammatic terms, think about markets and hierarchies as the ends of a two-dimensional straight line, and the fusion of markets and hierarchies as projecting a third dimensional space. Some historical and existing realities have rested on this third dimensional space. Early patrimonial administrations featured prebendal financing with varying degrees of top-down personnel control. Modern franchised companies manage entrepreneurial franchisees in complex corporate organizations. The Chinese state is run by revenue-making bureaucratic agents in a communist-style authority structure.
All three organizational forms, as unrelated to one
another as they may seem, are in fact manifestations of the same organizational abstract: contracting within hierarchies.
By laying out the conceptual foundation of organizations, this study is not just about the Chinese bureaucracy. It is more fundamentally about enriching our knowledge of the “third dimensional space” of organizational forms, specifically in the realm of state governance. Instead of presuming that bureau-contracting is a “transition” to legal-rational types, it may be more useful instead to view it as an alternative to the Weberian model. Whereas the Weberian bureaucracy is a low-powered, low-risk state model, bureau-contracting is high-powered, high-risk model. Weberian bureaucracies are to directly owned firms as bureau-contracting is to franchising. Both have virtues and shortfalls. The two models may be equally effective for the purposes of governance and growth, depending on the stages of development, the goals and constraints confronting political leaders, and the specific institutional mechanisms designed and enforced.
Ideal-types, as Weber noted, have “indispensable analytical value” (1968, 1002). Following on Weber’s point, Riggs reminded us that the tendency to accept standard models uncritically arises “from the lack of alternative models” (1964, 11). “The possibility of describing 139
administrative reality in terms other than the formal administrative bureau and the criterion for efficiency arises only when alternative models become available” (12). In light of this objective, my study has used China as a valuable case for developing an alternative model of public administration, which, I hope, may help us to better appreciate neo-traditional and developing contexts where the formal bureau does not exist. In my view, the China case is especially valuable not only because of its material successes, but also because its leaders have insisted on attending to the particularities of “Chinese characteristics.” It is this highly tailored model of reform that has allowed them to forge outcomes that defy norms and straightforward description.
Pre-Conditions for Bureau-Contracting
The purpose of this study is not to advocate the Chinese experience. Nevertheless, my approach necessarily raises the question of whether there is any applicability of the bureau-contracting model outside of China. After all, the conditions in reform-era China are truly unique. Besides the case of Vietnam, no other communist party-state has launched market reforms successfully and still survives. If we were to imagine the possibility of permutating bureau-contracting, what would the underlying conditions be? In what soils can this species grow? I consider some basic conditions below.
Talented and reform-minded leadership. We cannot discuss China’s reform success without first giving credit to its talented and reform-minded leadership, most notably Deng Xiaoping. If a personality like Zaire’s Motubu had taken the helm of power in China in the late 1970s, I would think the outcome today would have been very different – and disastrous – even if the initial conditions were exactly the same. The institutional bent in the prevailing social science analyses makes it oftentimes unappetizing to accept leadership quality as an explanatory variable, especially since leadership quality is endogenous and disturbingly random.
Yet, as Glaser et al (2004) found and reminded us, poor countries get out of poverty through policies, often pursued by dictators, and not just by having conventionally desirable institutional qualities. To deny the role of political leadership in economic and institutional development is to 140
deny the role of CEOs in producing corporate successes. We need “good” leaders to make “good” institutions. Good leadership qualities bring to mind a willingness to set priorities for market development, discipline, determination, foresight, and an eagerness to set an honest example for one’s subordinates. It is fair to say, I believe, that Deng Xiaoping has exemplified these qualities.
Instruments of personnel control. The Chinese party-state’s communist system and its revolutionary beginnings led to the establishment of an impressive organizational apparatus of personnel control. The degree of penetration of the party-state apparatus into society under communist rule was unprecedented in China. Party and state organs are enmeshed; officials are appointed top-down in a unified cadre management system; and dense parallel hierarchies ensure the transmission of party ideology and state policies from Beijing down to the lowest work units. As Yasheng Huang (1999) had argued, the power of appointment is the central government’s weapon of control against local officials. Instruments of personnel control are critical to a bureau-contracting system; without them, the risk of disintegration can be prohibitive.
A non-democratic political system. Can bureau-contracting be seen in places where officials are elected competitively to office? I doubt so. Being a one-party communist regime, the Chinese government has almost free rein in the structuring of public offices. Not only do citizens have a weak voice in matters of administrative reform, they have very limited information about how bureaucracies actually work. The details of administration presented in this study are obtained through effort. While anecdotes and rumors about the bureaucracy are common, the actual institutional details are not transparent to the public or even to the scholarly community. As Terry Moe (1989) noted in the American context, citizens do not care about the arcane details of public administration, even if the details were open to them. Only organized interest groups care about bureaucratic structure, because it can be manipulated to fit their interests. In a nondemocratic regime like China, the interest groups who know and can manipulate the inner workings of bureaucracy are none other than the bureaucrats themselves.
141
Future Research Directions
It is hoped that this study may invite further empirical research in a number of directions, particularly in the China case.
Bureau-Contracting and Corruption
First, we might ask what the relationship is between rents management and corruption, and how we can measure this relationship. My explanation has centered on the creation of a stable flow of rents to the bureaucracy via bureau-contracting mechanisms. More broadly, where institutionalized rents are available, are bureaucrats and politicians less likely to resort to illegal forms of corruption, such as embezzlement and extortion? In other words, when opportunities to make legal profits from public office are widely available, are officials less likely to steal? Unfortunately, systematic corruption data is extremely difficult to obtain anywhere and particularly so in China. As a first step, we may consider coding reports to exposed corruption cases, but bearing in mind that such data more accurately measures the rate of corruption control than actual levels of corruption. Another possible step is to conduct extensive business surveys to obtain perceptive measures of corruption.78
Regional and Temporal Variation
My work here provides one of the first systematic documentation of bureaucratic finances and cadre rewards in China. Yet, many questions remain with regard to temporal and regional variation. Will we find a similar pattern of dual fiscal incentives outside of Shandong province, which is largely representative of the coastal and central regions? I hypothesize that the patterns of cadre rewards and incentive structures will differ significantly in the Western and impoverished regions compared to the coastal areas. In the impoverished regions, few developmental opportunities are available, and local cadres are more likely to behave in a predatory manner against local residents. In the Western hinterlands like Xinjiang and Tibet, the
78
Reinikka and Svensson (2003) have written a guide on how to survey and measure corruption. They also provide a useful review of existing corruption surveys. 142
ethnic factor and succession threats should figure prominently in the distribution of cadre rewards.
Data on the distribution of public funds remains extremely difficult to obtain in China, and generally, in non-democratic and especially communist regimes. Scholars of non-democratic regimes have had to invent creative strategies to trace the distribution of resources and public funds. My study employs primarily a complete county-level dataset from Shandong province (Chapter 4) and segments of a national budget and agency-level datasets (Chapter 3). Hopefully, as more data becomes available in China, we could make progress in empirical measurements.
Fiscal and Career Incentives
Bureaucracies and money have been the subjects of this study. My focus on fiscal incentives is deliberate. The essence of bureau-contracting is the union of bureaucrat and entrepreneur. We cannot study this organizational type without a thorough examination of monetary flows and pursuits. However, for the parsimony of argument and constrained by the lack of data, I have omitted a systematic analysis of equally important set of incentives – career incentives.79
As a scholar of American politics pointed out, officeholders are motivated by two As: avarice and ambition (Schlesinger 1966). From a rational choice point of view, individuals who serve in office seek to maximize their utilities, either by amassing personal wealth or by climbing the career ladder. Wealth and power normally go together, we would think. But in China, the correlation between wealth and power is unclear. Anecdotal evidence suggests that officials higher up on the organizational chart may not necessarily make more money than their subordinates.80 In future work, more attention needs to be paid to the complicated interaction between fiscal and career incentives in the Chinese political system. 79
In recent years, some scholars have broken new ground by collecting systematic evidence about the paths and determinants of official promotion in China (Li and Zhou 2004; Landry 2008). 80 To illustrate, consider some of these examples. Provincial leaders, especially in places like Shanghai and Guangzhou, could easily be (legally) richer than central party hegemons. As I was told, local officials of the Beijing municipal government were compensated twice as much as central officials working in Beijing (AI 2006-9). In a city of Sichuan, county officials made less money than street-office (equivalent to townships) bureaucrats, who were the former’s subordinates, because the latter had well-to-do enterprises under their direct management (AI 2007-81). 143
Concluding Note
On a final concluding note, I would like to return to Weber. Weber’s theory of bureaucratization was based on European, specifically German, experiences, which could be generalized across modern states. It is seldom noticed that Weber had in fact noted China’s uniqueness compared to the European model. He wrote (1047-9):
The Chinese empire constituted an essentially different type… Because of the tremendous expansion of the empire and the small number of officials relative to the size of population, the Chinese administration was neither intensive nor was it centralized under the average ruler. Chinese officialdom did not develop into a modern bureaucracy, for the functional differentiation of spheres of jurisdiction was carried through only to a very limited extent in view of the country’s huge size… This accounts for the specifically anti-bureaucratic and patrimonial tendency of this administration, which in turn explains its ‘extensive’ character and its technical backwardness. Written a century ago, Weber had insightfully pointed to the large size, small officialdom, decentralized administration, and mixed spheres of jurisdiction – features that continue to define parts of the present-day Chinese bureaucracy. However, since Weber’s writings, dramatic and unprecedented changes have unfolded across the Chinese political and economic landscape. New, entrepreneurial elements emerged and were gradually fused into a pre-existing patrimonial framework, thus creating novel institutional forms. My study has suggested that it is not sufficient to appreciate only the Weberian trajectory of bureaucratization. Equally, we would be mistaken to view the contemporary Chinese bureaucracy as being little different from early patrimonial or Zairian-style predatory states. The unique-ness of the bureau-contracting model, I hope, may help us to unravel not only empirical puzzles about the structural bases of China’s development, but may also point us to alternative paths of modernization and state institutional building.
144
APPENDIX A
LIST OF LOCAL GOVERNMENT OFFICES
Table 1 list the offices in a typical local (provincial, city, and county) government based on information listed on official governmental websites. In China, the structure of bureaucracies established is generally replicated from the central level to the local levels. Hence, the state organizational chart is similar across provincial, city, and county governments. Some variation may exist across locations where certain local governments establish ad-hoc offices for special purposes. Offices are usually listed on governmental websites in the following clusters: (i) party organs, (ii) governmental organs, (iii) vertically-managed offices, (iii) court and procuratorate, and (iv) social organizations. With the exception of social organizations, most of the offices in Table 1 are “administrative units” (xingzheng danwei). Each functions as the “supervisory department” (zhuguan bumen) to a subsidiary group of “service units” (shiye danwei).81 In my analysis, I refer to the administrative and service units as core bureaus and extrabureaucracies respectively.
Table 2 lists the extrabureaucracies under five different bureaus. Similarly, the structure of extrabureaucracies vary little between levels of government. Interestingly, in areas I have conducted fieldwork, local governments remained reluctant to publicize the list of service units in their locality; some even regard this information as a secret. In my research, I have not come across any government website that provides a comprehensive list of extrabureaucracies. The list from Table Y was a partial list provided by the personnel bureau of a city government in Jiangsu province.
81
The social organizations have a special status because technically they do engage in administrative work or services provision. Cadres serving in the social organizations are usually given the title of “shiye personnel approximating civil servants” (canzhao gongwuyuan shiye bianzhi). 145
Table 1: List of offices listed in a typical local governmental website Party Organs Party Committee (党委) Disciplinary and inspection bureau (纪委监察局) Organization department (组织部) Propaganda department (宣传部) Strategy department (统一战线工作部) Rural affairs office (农村工作办公室) Party school (委党校) Dossier office (档案局) Retired cadres office (老干部局) Governmental Organs Government office (人民政府办公室) Development and reform committee (发展和改革委员会) Commerce and trade committee (经济贸易委员会) Education bureau (教育局) Public security bureau (公安局) Civil affairs bureau (民政局) Foreign affairs office (外事办) Legal affairs bureau (司法局) Finance bureau (财政局) Personnel bureau (人事局) Labor and social security bureau (吴江市劳动和社保局) Land bureau (国土资源局) Construction bureau (建设局) Urban management (城市管理局) Transportation bureau (交通局) Water management bureau (水利局) Forestry bureau (农林局) Water production bureau (水产局) Culture and broadcasting bureau (文化广播电视管理局) Health bureau (卫生局) Population and family planning committee (人口和计划生育委员会) Audit bureau (审计局) Environmental protection bureau (审计局) Sports bureau (体育局) Tourism bureau (旅游局) Statistics bureau (统计局) Price bureau (物价局) Grain bureau (粮食局) Letters and petition bureau (信访局) Safety and production inspection bureau (安全生产监督管理局) Legal enforcement office (法制办公室) Science and technology bureau (科学技术局) Minority and religious affairs office (民族宗教事务局) Administrative services center (行政服务中心) 146
Vertically-managed offices Borders inspection and disease control bureau (出入境检验检疫局) Drugs inspection and management bureau (食品药品监督管理局) Local tax bureau (地方税务局) Commerce bureau (工商局) Customs (海关) Quality inspection bureau (质量技术监督局) State tax bureau (国家税务局) Weather bureau (气象局) Housing funds management center (住房公积金管理中心吴江分中心) Court and procuratorate Local procuratorate (检察院) Local court (法院) Social organizations Labor union (总工会) Communist youth league (团市委) All-women’s federation (妇女联合会) Science and technology association (科学技术协会) Chinese returned from overseas association (归侨侨眷联合会) Handicapped association (残联)
147
Table 2: List of Extrabureaucracies Under Selected Bureaus Bureau/Administrative Unit “Supervising Department” Construction bureau
Extrabureaucracies/Service Units
Forestry bureau
Party school for the forestry bureau Forestry and greenery inspection team ABC city vista point management office A park management office B park management office C park management office D park management office Greenery quality inspection and management center Asset management center
Health bureau
Health and sanitation inspection center Medical personnel and technical services center Disease prevention and control center Blood center Emergency services center Health association Medical fees accounting center Party school of the health bureau Cheer city hospital
Personnel Bureau
Talent services center Transferred military personnel services center International talent exchange services center Talent examination center Talent market
Commerce and trade commission
Finance and trade information center Procurement center for electronics equipment
Construction projects quality assessment office Construction projects inspection team Construction transaction management center Construction projects quality inspection office Construction assembly management office Construction project design inspection center Railroad construction office Urban science research institute Major construction projects management office Dossier center for construction projects Office for managing relocation and moving Professional school for construction New technologies in construction company/outreach office Construction services center Construction training center Construction safety inspection center Construction information center School for construction techniques
148
Cement office Enterprise training center Energy and resource training institute Office for the reform of construction materials Small and medium enterprise services center Commerce association Development and reform commission
Information center Planning and economics training center Investment and management company
Source: Personnel Bureau of Cheer City Government
149
APPENDIX B
CHINA’S FORMAL CADRE WAGE SCALE
Table 1: Basic and functional wages of civil servants at the provincial and city level Date of Issue: October 1993 Unit: yuan/month Function
Mayor Vice mayor Ju/Chu Chief Vice Ju/Chu Chief Ke Head Vice Ke Head Ke Officer Support Staff
Basic Wages 40 40 40 40 40 40 40 40
190 150
165 140 130 110 82 65 49 42
Functional wages by grade 1 150 130 120 100 73 57 42 36
140 120 110 91 65 49 36 30
2 130 110 100 82 57 42 30 24
3 120 100 91 73 49 36 24 18
4
5 230 190
82 65 42 30 18 12
Source: Xu 2007, p. 197
150
6 205 180 170 150 122 105 89 82
Basic wages (jiben gongzi) and functional wages (zhiwu gongzi) combined 1 2 3 4 5 6 190 180 170 160 170 165 150 140 160 150 140 131 122 140 131 122 113 105 113 105 97 89 82 97 89 82 76 70 82 76 70 64 58 76 70 64 58 52
Table 2: Wages by Function and Grade Date of Issue: November 1993 Unit: yuan/month Function
Chairman Vice-Chairman Premier Vice Premier Member of the State Council Minister Vice Minister Si Chief Ting Chief Vice Si Chief Vice Ting Chief Chu Chief County Head Vice Chu Chief Vice County Head
Wages by Function (zhiwu)
1 480
2 550
3 630
4
400
460
520
580
5
Wages by Grade (jibie)
6
7
8
Basic (jiben) Wages
Length of work (gong ling) wages One yuan added for every year of service
Grade 1
Wage 470
90
2
425
90
3
382
90
4
340
90
330
380
430
480
530
5
298
90
270
315
360
405
450
6 7
263 228
90 90
118
143
168
193
218
243
8
193
90
144
174
204
234
264
294
9
164
90
118
143
168
193
218
243
10
135
90
11
111
90
12
92
90
13 14 15
77 65 55
90 90 90
Ke Head
96
116
136
156
176
196
216
Vice Ke Head Ke Executive Support Staff
79 63 50
94 75 60
109 87 70
124 99 80
139 111 90
154 123 100
169 135 110
Source: Xu 2007, p. 206
151
147 120
Table 3: Regional subsidies for various “hardship” regions Date of Issue: February 2001 Unit: yuan/month
Average subsidy Provincial level and above City level County level Ke level Ke and below
Region (1) 43 100 80 60 47 40
Region (2) 86 200 160 120 95 80
Source: Xu 2007, p. 203
152
Region (3) 172 -320 240 190 160
Region (4) 300 -560 420 330 280
APPENDIX C
SUMMARY OF INTERVIEW TARGETS
For this study, I conducted 165 interviews in China between 2006 to 2009. The individuals interviewed are central and local level cadres, ranging from civil servants in leadership positions to rank-to-file public employees working in administrative agencies and public service providers. To widen the scope of my research to the furthest extent possible, I conducted interviews in multiple locations, at all levels of government (from the central government down to the villages), and across governmental sectors. Some of the interviewees were interviewed more than once.
Because of the delicate nature of my topic, I have chosen to be doubly careful in protecting the identity of my interviewees. This study names interviews only by the identification number assigned to each interviewee, followed by the date on which the first interview was conducted. I have also chosen not to identify the county and city governments referred to in the study. Instead, for easy reference, I assign random names to them. The following tables summarize the distribution of my interviews by location, level of government, and governmental sector/department.
153
Table 1: Interviews by Location Central government Anhui Beijing Chongqing Guangdong Guangxi Hunan Jiangsu Shandong Shanghai Sichuan Tianjin
18 4 2 1 8 1 2 28 15 1 41 44
Table 2: Interviews by Level of Government Central Provincial level City level County level Township and villages
18 14 27 88 18
Table 3: Interviews by Governmental Sector/Department Sector/Department Administrative Office Agriculture Arts & Culture Civil Affairs Construction Court Education Environment Finance Health NDRC (National Development & Reform Commission) Party Organs Personnel Research Township and village governments Others
154
No. 4 2 3 13 3 3 23 12 22 7 7 6 21 3 17 19
REFERENCES
Acemoglu, Daron, Michael Kremer, and Atif Mian. "Incentives in markets, firms, and governments ." Journal of Law, Economics, and Organizations , 2007: 1-34. Amsden, Alice. Asia's Next Giant: South Korea and Late Industrialization. New York : Oxford University Press, 1989. Appelbaum, R.P., and J. Henderson. States and Development in the Asian Pacific Rim . Newbury Park, CA: Sage, 1992. Baker, George P., Michael C. Jensen, and Kevin J. Murphy. "Compensation and Incentives: Practice vs. Theory." The Journal of Finance, 43, no. 3 (1988): 593-616. Bardhan, Pranab. "Corruption and development: A review of issues." Journal of Economic Literature XXXV (1997): 1320-1346. Barkey, Karen. Bandits and Bureaucrats: The Ottoman Route to State Centralization. Itaca: Cornell University Press, 1994. Barnett, Doak. Cadres, Bureaucracy, and Political Power in Communist China. New York : Columbia University Press, 1967. Bates, Robert. "The role of the state in development ." In The Oxford handbook of political economy , by Barry R. Weingast Weingast and Donald Wittman. New York: Oxford University Press, 2008. Berkowitz, Daniel, Katharina Pistor, and Jean-Francois Richard. "Economic development, legality, and the transplant effect." European Economic Review 47, no. 1 (2003): 165-195. Bernstein, Thomas P., and Xiaobo Lu. Taxation without representation in contemporary rural China . Cambridge: Cambridge University Press, 2003. Bhagwati, Jagdish. "Directly unproductive, profit-seeking activities." Journal of Political Economy 90, no. 5 (1983): 988-1002. Blecher, Mark. "Developmental State, Entreprenuerial State." In The Chinese State in the Era of Economic Reform, by Gordon White. London: Macmillan, 1991. Blecher, Mark, and Vivienne Shue. "Into leather: State-led Development and the Private Sector in Xinji." China Quarterly, 2001: 369-393. Brodsgaard, Kjeld. "Institutional Reform and the Bianzhi System in China." The China Quarterly, 2002: 361-386. 155
Buchanan, James M. "Rent seeking and profit seeking." In Toward a theory of the rent-seeking society , by Buchanan et al. College Station: Texas A&M University Press, 1980. Buchanan, James M., Robert D. Tollison, and Gordon Tullock. Toward a theory of the rentseeking society . Texas: Texas A&M University Press, 1980. Burns, John P. "Downsizing the Chinese State." The China Quarterly, 2003: 776-802. Campos, J. Edgardo, Donald Lien, and Sanjay Pradhan. "The impact of corruption on investment: predictability matters." World Development 27, no. 6 (1999): 1059-1067. Campos, J. Edgardo, ed. Corruption: The boom and bust of East Asia. Manila, Phillippines: Manila University Press, 2001. Campos, Jose Edgardo, and Hilton Root. The Key to the Asian Miracle: Making Shared Growth Credible . Washington D.C. : Brookings Institution , 1996. Castells, Manuel. "Four Asian Tigers with a Dragon Head." In States and Development in the Asian Pacific Rim, by Richard Appelbaum & Jeffrey Henderson (eds.), 33-70. Newbury Park, CA: Sage, 1992. Chang, Gordon. The coming collaspe of China. New York : Random House, 2001. Cheng, Siwei. "Strategic Directions and Policy Implementation for Reforming China's Institutional Units." In Studies on Economic Reforms and Development in China, by Siwei Cheng. Oxford University Press, 2001. Chu, Tung-Tsu. Local government in China under the Ching. Stanford: Stanford University Press, 1969. Coase, Ronald. "The Nature of the Firm." Economica 4 (1937): 386-405. Dobbs, Michael. Down with Big Brother: the fall of the Soviet empire. Alfred A. Knopf, 1997. Donahue, John D. The Privatization Decision: Public Ends, Private Means. New York : Basic Books, 1989. Duara, Prasenjit. "State involution: A study of local finances in north china, 1911-1935." Comparative Studies in Society and History 29, no. 1 (1987): 132-161. Duckett, Jane. "Bureaucrats in Business, Chinese-Style: The lessons of market reform and state entrepreneurialism in the People's Republic of China." World Development 29, no. 1 (2001): 23-37. —. The Entreprenuerial State in China: Real Estate and Commerce Departments in the Reform Era in Tianjin. New York: Routledge, 1998. 156
Evans, Peter. Embedded Autonomy: States and Industrial Transformation. Princeton, NJ: Princeton University Press, 1995. Evans, Peter, and James E. Rauch. "Bureaucracy and Growth: A Cross-national Analysisof the Effects of “Weberian” State Structures on Economic Growth." American Sociological Review 64, no. 5 (1999): 748-65. Frant, Howard. "High-Powered and Low-Powered Incentives in the Public Sector." Journal of Public Administration Research and Theory, 6 1996: 365-381. Gardiner, John. "Defining corruption." In Political corruption: concepts and contexts, by A.J. Heidenheimer & Michael Johnston (eds.). New Brunswick: Transaction Publishers, 2000. Geddes, Barbara. Politician's Dilemma: Building State Capacity in Latin America. Berkeley. University of California Press, 1996. Gerschenkron, Alexander. Economic Backwardness in Historical Perspective. Belknap Press , 1962. Glaeser, Edward L., Rafael Porta La, Florencio Lopez-de-Silanes, and Andrei Schleifer. "Do institutions cause growth?" Journal of economic growth 9 (2004): 271-303. Goldstone, Jack. "The Coming China Collapse." Foreign Policy 99 (1995): 35-53. Gore, Lance. Market communism : the institutional foundation of China's post-Mao hypergrowth . New York: Oxford University Press, 1998. Haggard, Stepan. The Political Economy of the Asian Financial Crisis. Washington D.C.: Institute for International Economics, 2000. —. Pathways from the periphery : the politics of growth in the newly industrializing countries. Ithaca, NY: Cornell University Press, 1990. Haggard, Stephan, and Robert Kaufman. "The politics of stabilization and structural adjustment." In Developing country debt and economic performance: the international financial system, by Jeffrey ed. Sachs. Chicago: University of Chicago Press , 1989. Hall, Peter A., and David W. Soskice. Varieties of capitalism: the institutional foundations of comparative advantage. New York: Oxford University Press, 2001. Henderson, Jeffrey. Industrial transformation in Eastern Europe in light of East Asian experience. London : Macmillan, 1998. Henderson, Jeffrey, David Hulme, Jalilian Hossein, and Phillips Richard. "Bureaucratic Effects: Weberian State Agencies and Poverty Reduction." Sociology 41 (2007): 515-532. 157
Heo, Uk, and Sunwoong Kim. "Financial Crisis in South Korea: Failure of the Government-Led Development Paradigm." Asian Survey 11, no. 3 (2000). Hickey, Paul. "Fee-taking, salary reform, and the structure of state power in late Qing China, 1909-1911." Modern China 17, no. 3 (1991): 389-417. Howell, Jude. China opens its door: the politics of economic transition. Boulder, CO: Lynne Rienner, 1993. Huang, Philip C. "Public sphere/civil society in China? The third realm between state and society." Modern China 19, no. 2 (1993): 216-40. Huang, Shu-min. The Spiral Road: Change in a Chinese Village Through the Eyes of a Communist Party. 2nd Edition. Boulder: Westview Press, 1998. Huang, Yasheng. Inflation and investment controls in China: The political economy of centrallocal relations. Cambridge University Press: 1999. Hubbard, Michael. "Bureaucrats and Markets in China: The rise and fall of the entrepreneurial local government." Governance 8, no. 3 (1995): 335-353. Hutchcroft, Paul D. "The Politics of Privilege: Assessing the impact of rents, corruption, and clientelism on third world development." Political Studies XLV (1997): 639-658. Jin, Hehui, Yingyi Qian, and Barry Weingast. "Regional decentralization and fiscal incentives: federalism, Chinese style." Journal of public economics 89 (2005): 1719-1742. Johnson, Chalmers. MITI and the Japanese Miracle. Stanford: Stanford University Press, 1982. Joseph, Richard. Democracy and prebendal politics in Nigeria. Cambridge: Cambridge University Press, 1987. Kang, David. Crony Capitalism: Corruption and Development in South Korea and the Phillippines. Cambridge : Cambridge University Press, 2002. Khan, Mushtaq H., and Kwame Sundaram Jomo. Rents, Rent-Seeking and Economic Development:Theory & Evidence in Asia . Boston, MA: Cambridge University Press, 2000. Kiser, Edgar, and Yong Cai. "War and Bureaucratization in Qin China: Exploring an Anomalous Case." American Sociological Review 68, no. 4 (2003): 511-539. Knack, Stephan, and Phillip Keefer. "Institutions and Economic Performance: Cross-Country Tests Using Alternative Institutional Measures." Economics and Politics 7, no. 3 (1995): 207-27.
158
Krueger, Anne. "The Political Economy of the Rent-Seeking Society." American Economic Review 64, no. 3 (1974): 291-303. Kung, J.K., and Lin Y. "The Decline of Township-and-Village Enterprises in China's Economic Transition." World Development 35, no. 4 (2007): 569-84. La Porta, Rafael, Florencio Lopez-de-Silanes, Andrei Shleifer, and Robert Vishny. "The quality of government." Journal of Law, Economics, and Organization 15, no. 1 (1999): 222-79. Lam, Tao-Chiu, and James L. Perry. "Service Organizations in China: Reform and Its Limits ." In Remaking China's Public Management , by Peter Lee and Carlos Lo, 19-41. London: Quorum Books, 2001. Landry, Pierre. Decentralized Authoritarianism in China: The Communist Party's Control of Local Elites in the Post-Mao Era. Cambridge: Cambridge University Press, 2008. Lau, Lawrence, Yingyi Qian, and Gerard Roland. "Reform without losers: An interpretation of China's dual-track approach to transition." The Journal of Political Economy 108, no. 1 (2000): 120-143. Lee, Charlotte. Strategies of Political Control and Party Adaptation in Reform-Era China. Dissertation, Department of Political Science, Stanford University, 2009. Lee, Hung-Yung. From Revolutionary Cadres to Party Technocrats in Socialist China. Berkeley: University of California Press, 1991. Levi, Margaret. Of Rule and Revenue. Berkeley: University of California Press, 1988. Li, Cheng. China’s Leaders: The New Generation. Rowman & Littlefield Publishers, 2001. Li, David. "Changing Incentives of the Chinese Bureaucracy." American Economic Review 88, no. 2 (1998): 393. Li, Hongbin, and Li-An Zhou. "Political Turnover & Economic Performance: The Incentive Role of Personnel Control in China." Journal of Public Economics 89, no. 9-10 (2005): 1743-1762. Lieberthal, Kenneth. Governing China: From Revolution Through Reform. W. W. Norton, 2003. Lin, Yi-min, and Zhanxin Zhang. "Backyard profit centers: the private assets of public agencies." In Property Rights and Economic Reform in China, by Jean Oi and Andrew Walder. Stanford: Stanford University Press, 1999. Love, John F. McDonald's: Behind the Arches . New York: Bantam Books, 1986. Lu, Xiaobo. "Booty Socialism, Bureau-Prenuers, and the State in Transition: Organizational Corruption in China." Comparative Politics 32, no. 3 (2000a): 273-294. 159
—. Cadres and corruption: the organizational involution of the Chinese Communist Party. Stanford: Stanford University Press, 2000b. MacIntyre, Andrew. "Funny Money: Fiscal Policy, Rent-seeking and Economic Performance in Indonesia." In Rents, rent-seeking, and economic development, by M.H. Khan & Jomo K.S. Cambridge: Cambride University Press, 2000. —. "Investment, Property Rights, and Corruption in Indonesia." In Corruption: the boom and bust of east asia, by J. Edgardo Campos (ed.). Manila, Phillippines: Manila University Press, 2001. Magaloni, Beatriz. Voting for autocracy: hegemonic party survival and its demise in mexico. Cambridge: Cambridge University Press, 2006. Magaloni, Beatriz, Alberto Diaz-Cayeros, and Federico Estevez. "Clientelism and Portfolio Diversification: A Model of Electoral Investment with Applications to Mexico." In Patrons, Clients, and Policies, by Herbert Kitschelt and Steven Wilkinson. Cambridge: Cambridge University Press, 2007. Manion, Melanie. Retirement of revolutionaries in China: public policies, social norms, private interests. Princeton, N.J.: Princeton University Press, 1993. —. Corruption By Design: Building Clean Government in Mainland China and Hong Kong. Cambridge: Harvard University Press, 2004. Mauro, Paulo. "Corruption and Growth." Quarterly Journal of Economics 110, no. 3 (1995): 681712. McCormick, Barnett. Political reform in post-Mao China. Berkeley: University of California Press, 1990. Mertha, Andrew. "China's 'Soft' Centralization: Shifting Tiao/Kuai Authority Relations." The China Quarterly 184 (2005): 791-810. Moe, Terry. "The New Economics of Organization." American Journal of Political Science 28, no. 4 (1984): 739-777. Montinola, Yingyi Qian, and Barry Weingast. "Market Preserving Federalism, Chinese-Style: The political basis for economic success." World Politics 48, no. 1 (1995): 50-81. Nee, Victor, and Peng Lian. "Sleeping with the Enemy: A Dynamic Model of Declining Political Commitment in State Socialism." Theory and Society 23 (1994): 253- 296.
160
North, Douglass, and Barry Weingast. "Constitutions and commitment: The evolution of institutions governing public choice in 17th century England." The Journal of Economic History 49, no. 4 (1989): 803-832. Nye, J.N. "Corruption and Political Development: A Cost-Benefit Analysis." The American Political Science Review 61, no. 2 (1967): 417-427. OECD. "Reforming PSU: Challenges." In Governing China. OECD Press, 2006. Oi, Jean. "Fiscal reforms and the economic foundations of local state corporatism." World Politics 45, no. 1 (1992): 99-126. —. Rural China Takes Off: Institutional Foundations of Economic Growth. Berkeley: University of California Press, 1999. —. "The role of the local state in China's transitional economy." The China Quarterly 144 (1995): 1132-49. Onis, Ziya. "The logic of the developmental state." Comparative Politics 24, no. 1 (1991): 109-26. Park, Albert, Scott Rozelle, Christine Wong, and Ren Changqing. "Distributional consequences of reforming local public finance in china." The China Quarterly 147 (1996): 751-778. Pearson, Margaret M. "The Business of Governing Business in China: Institutions & Norms of the Emerging Regulatory State." World Politics 57, no. 2 (2005): 296-322. Pei, Minxin. China's Trapped Transition: The Limits of Developmental Autocracy. Cambridge: Harvard University Press, 2006. Pempel, T.J. (ed.). The Politics of the Asian Financial Crisis. Ithaca: Cornell University Press, 1999. Przeworski, Adam. "The last instance: Are institutions the primary cause of economic development?" European Journal of Sociology 45, no. 2 (2004): 165-188. Qian, Yingyi. "How Reform Worked in China." In In Search of Prosperity: Analytical Narratives on Economic Growth, by Dani Rodrik. NJ: Princeton University Press, 2003. Qian, Yingyi, and Jinglian Wu. "China's transition to a market economy: how far across the river? ." In How far across the river? , by Nicholas C. Hope, Dennis Tao Yang and Mu Yang Li. Palo Alto: Stanford University Press, 2003. Rankin, Mary Backus. "Some observations on a Chinese public sphere." Modern China 19, no. 2 (1993): 158-182. Rauch, James, and Peter Evans. "Bureaucratic Structure and Bureaucratic Performance in Less Developed Countries." Journal of Public Economics 75 (2000): 49-71. 161
Reed, Bradly Ward. Talons and Teeth: County Clerks and Runners in the Qing Dynasty. Stanford: Stanford University Press, 2000. Reinikka, Ritva and Jakob Svensson. “Survey techniques to measure and explain corruption.” World Bank Policy Working Paper 3071, 2003. Riggs, Fred Warren. Administration in developing countries; the theory of prismatic society. Boston: Houghton Mifflin, 1964. Rock, Michael T., and Heidi Bonnett. "The comparative politics of corruption: Accounting for the east asian paradox in empirical studies of corruption, growth, and investment." World Development 32, no. 6 (2004): 999-1017. Rodrik, Dani. "Getting Interventions Right: How South Korea and Taiwan Grew Rich." Economic Policy 20 (1995): 55-97. —. "Second-Best Institutions." American Economic Review 98, no. 2 (2008): 100-104. Root, Hilton. Small Countries, Big Lessons. Hong Kong: Oxford University Press, 1996. —. "Has Corruption in China Become Systemic." Asian Survey 8 (1996b). Rose-Ackerman, Susan. Corruption and Government: Causes, Consequences, and Reform. MA: Cambridge University Press, 1999. Schlesinger, Joseph. Ambition and Politics: Political Careers in the United States. Chicago: Rand McNally, 1966. Schneider, Ben Ross. "The Desarrollista State in Brazil and Mexico." In The Developmental State, by Meredith Woo-Cumings (ed.). Ithaca: Cornell University Press, 1999. Segal, Gerald, and Davis (eds.) Goodman. Towards Recovery in the Pacific Asia. London: Routledge, 2000. Shirk, Susan. The political logic of economic reform. Berkeley: University of California Press, 1993. Shleifer, Andrew, and Daniel Treisman. Without a map: political tactics and economic reform in russia. Cambridge: MIT Press, 2000. Shue, Vivienne. "Beyond the budget: finance organization and reform in a chinese county." Modern China 10, no. 2 (1984): 147-186. Solnick, Steven. Stealing the state: control and collaspe in Soviet institutions . Cambridge, MA: Harvard University Press, 1998. Stepan, Alfred. The State and Society: Peru in Comparative Perspective . Princeton, NJ: Princeton University Press, 1978. 162
Stiglitz, J. "Wither reform: ten years of transition." In Annual World Bank Conference on Economic Development , by B. Pleskovic and J. Stiglitz. Washington D.C. : World Bank, 2000. Sun, Yan. Corruption and Market in Contemporary China. Ithaca: Cornell University Press, 2004. —. "Corruption, Growth, and Reform: The Chinese Enigma." Current History 104, no. 683 (2005): 257-263. —. "Reform, State, and Corruption: Is Corruption Less Destructive in China Than in Russia?" Comparative Politics 32, no. 1 (1999): 1-20. Tarkowski, Jacek. "A centralized system and corruption: the case of Poland." Asian Journal of Public Administration 10, no. 1 (1988): 48-70. Tilly, Charles. Coercion, capital and European states, A.D. 990-1992. Blackwell Publishing, 1993. Tsai, Kellee. "Off Balance: The Unintended Consequences of Fiscal Federalism in China." Journal of Chinese Political Science 1, no. 26 (2004): 1-26. Van de Walle, N. "Meet the new boss, same as the old boss? The evolution of political clientelism in Africa." In Patrons, Clients, and Policies: Patterns of Democratic Accountability and Political Competition, by H Kitschelt and Wilkinson S. Cambridge: Cambridge University Press, 2007. Wade, Robert. Governing the Market: Economic Theory and the Role of Government in East Asian Industrialization. Princeton, NJ: Princeton University Press, 1990. Wade, Robert, and Gordon White. Developmental State in East Asia. IDS Research Reports R16, 1988. Walder, Andrew. Communist Neo-Traditionalism: Work and Authority in Chinese Industry. Berkeley: University of California Press, 1986. —. "Local governments as industrial firms: An organizational analysis of China's transitional economy." American Journal of Sociology 101, no. 2 (1995): 263-301. —. "The county government as industrial organization." In Zouping in Transition, by Andrew, ed. Walder. Cambridge, MA: Harvard University Press, 1998. Wang, Rong. "Political determinants of county government budgeting in china: a case study." Working Paper, Institute of Development Studies (IDS), 2004.
163
Wang, Shaoguang. "The Rise of the Regions: Fiscal Reform and the Decline of Central State Capacity in China." In The Waning of the Communist State, by Andrew Walder. University of California Press, 1995. Wang, Shaoguang, and Angang Hu. "Guoqing Baogao (Report on National Conditions)." Chinese Economic Studies 3 (1995). Weber, Max. Economy and Society. Edited by Guenter Roth and Claus Wittich. New York: Bedminster, [1904-1911] 1968. Wedeman, Andrew. "Budgets, extra-budgets, and small treasuries: illegal monies and local autonomy in China." Journal of Contemporary China 9, no. 25 (2000): 489-511. —. "Development and corruption: the East Asian Paradox." In Political business in East Asia, by E.T. Gomez. London: Routledge, 2002a. —. "Looters, Rent-scrappers, and dividend-collectors: Corruption and growth in Zaire, South Korea, and the Philippines." The Journal of Developing Areas 31 (1997): 457-478. —. "State predation and rapid growth: politicization of business in China." In Political business in East Asia, by E.T. Gomez. London: Routledge, 2002b. —. "The Intensification of Corruption in China." The China Quarterly 180 (2004): 895-921. Wei-Arthus, Huiying. A study of authority and relations in chinese governmental agencies and institutional work units. Lewiston: Edwin Mellon Press, 2000. Weingast, Barry. "The economic role of political institutions." Journal of Law, Economics, and Organization 7 (1995): 1-31. White, Gordon. "The role of the state in China's socialist industrialization." In Developmental States in East Asia, by Robert Wade & Gordon White (eds.). IDS Research Reports R16, 1988. Whiting, Susan. "The Cadre Evaluation system at the grass roots: the paradox of party rule." In Holding China Together, by Barry et al Naughton. Cambridge, MA: Cambridge University Press, 2004. Williamson, Oliver. Markets and hierarchies: Analysis and Antitrust Implications. NY: Free Press, 1974. Woo-Cumings, Meredith. "Introduction ." In The Developmental State, by Woo-Cumings (ed.) Meredith. Ithaca: Cornell University Press, 1999.
164
World Bank. Building institutions for markets: World Development 2002. New York: Oxford University Press, 2000. —. China: National Development and Sub-National Finance: A Review of Provincial Expenditures. Report No. 22951-CHA, World Bank, 2002. —.Deepening public service unit reform to improve service delivery. Beijing: CITIC Press, 2005. —. The East Asian Miracle: Economic Growth and Public Policy. New York: Oxford University Press, 1993. —. World Development Report 1997: The State in a Changing World. New York: Oxford University Press, 1997. Wu, Zhongjian, and Sun Jiaming (eds.). Tianjin city bianzhi annals (Tianjinshi jigoubianzhi zhi) . Tianjin: Tianjin Social Science Academy, 1998. Xu, Songtao. Looking back at 28 years of personnel reform (Huimou zhongguo renshi zhidu gaige 28 nian). Beijing: Zhongguo Renshi Press, 2007. Yang, Dali. Remaking the Chinese Leviathan: Market Transition and the Politics of Governance in China. Stanford: Stanford University Press, 2004. Yusuf, Shahid. "The east asian miracle at the millennium ." In Rethinking the east asian miracle, by Joseph E. Stiglitz and Shahid Yusuf. New York : Oxford University Press, 2001. Zelin, Madeleine. The Magistrate's Tael. Berkeley: University of California Press, 1984. Zhang, Xuehua. Enforcing environmental regulations in Hubei province, China: Agencies, Courts, Citizens. Dissertation , Interdisciplinary Graduate Program in Environment and Resources, Stanford University, 2008. Zhao, Libo. The reform of shiyedanwei: an analysis of public services development (Shiye danwei gaige: gonggong shiye fazhan xinjizhi tanxi). Jinan: Shandong People's Press, 2003. Zhuravskaya, Ekaterina. "Incentives to provide local public goods: fiscal federalism, Russian style." Journal of public economics 76, no. 3 (2000): 337-386. —. "Wither Russia? A review of Andrei Shleifer's Normal Country." Journal of Economic Literature 45, no. 1 (2007): 127-146.
165
