Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
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
http://creativecommons.org/licenses/by-nc/3.0/us/
This dissertation is online at: http://purl.stanford.edu/hr313dw9240
© 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 Attribution-Noncommercial 3.0 United States License.
ii
I certify that I have read this dissertation and that, in my opinion, it is fully adequatein 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 adequatein 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 adequatein 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 adequatein 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 adequatein 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 inUniversity Archives.
iii
iv
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 bureau-
contracting – where contracting takes place within the state bureaucracy. In a bureau-
contracting 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
v
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.
vi
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
vii
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
viii
Zhijian, Zhu Jinwei, Wu Yuanyi, Feng Shanshu for helping me to arrange interviews. Kuo Tai-
Chun 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
ix
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.
x
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
xi
TABLE OF CONTENTS
Acknowledgements
List of Tables
List of Figures
CHAPTER 1 –
Introduction
1
CHAPTER 2 –
Bureau-Contracting: The Industrial Organization of Chinese Bureaucracies
27
CHAPTER 3 –
The Structure of Cadre Rewards
65
CHAPTER 4 –
Dual Fiscal Incentives
79
CHAPTER 5 –
The Political Economy of Bureaucratic Adaptation
104
CHAPTER 6 –
Conclusion
125
Appendix A: List of Local Government Offices
Appendix B: The Formal Cadre Wage Scale
Appendix C: List of Interviews
xii
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
Transfers
121
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
1
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
2
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 ideal-
type. 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 pre-
bureaucratic organizations. In particular, I underscore the economic character of their non-
Weberian 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.
3
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.
4
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).
5
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 semi-
officials. 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).
6
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.
7
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.
8
“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.
9
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
10
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,
11
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.
10 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 inter-
meshing of personal loyalties and impersonal duties in manager-worker relationships.
12
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.
13
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
14
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, rent-
scrapping and dividend-collecting. He argued that corruption in East Asia took the forms of rent-
scrapping (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
15
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 bribe-
taking. 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 dividend-
collecting 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
16
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 neo-
patrimonialism 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.
17
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 bureau-
contracting.
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.
18
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.
19
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).
20
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.
21
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 self-
financed 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 rent-
seeking 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).
22
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 high-
powered 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,
23
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.
24
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 market-
supporting 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
25
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 bureau-
contracting 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.
26
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
27
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 error-
correction 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.
28
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-
29
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
30
(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.
31
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).
32
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
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.
Market-Based Hierarchy-Based
Ma
rke
t-B
ase
d
Hie
rarc
hy-
Ba
sed
FINANCIAL RELATIONS P
ER
SON
NE
L R
ELA
TIO
NS
Certain forms of
public
enterprises
Private
Contracting
Bureau-
Contracting
Public
Bureaucracy
33
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,
34
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
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.
Conventional Models of State Organization
MODEL 1
(Big Bureaucracy, No Contracting)
Bureaucracy
MODEL 2
(Small Bureaucracy, Extensive Contracting)
B
C
C
CC
C
China’s Model of State Organization
B
EB
EB
EBEB
EB
35
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.
36
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 sub-
national 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.
37
Figure 2.3: Organization Chart of Typical Local Government
Party Committee
Social Organizations
Pa
rty Co
mm
ittee
Se
creta
riat
Org
an
izatio
n D
ep
artm
en
t
Pro
pa
ga
nd
a D
ep
artm
en
t
Stra
teg
y D
ep
artm
en
t
Po
litical a
nd
Lega
l Affa
irs Co
mm
ittee
Disc
iplin
ary C
om
mitte
e
Co
mm
un
ist You
th L
ea
gu
e
All-W
om
en
’s Fe
de
ratio
n
Wo
rkers’ U
nio
n
Co
mm
erc
e U
nio
n
Fed
era
tion
for th
e H
an
dica
pp
ed
Asso
ciatio
n o
f Scie
nc
e
Government Office
De
velo
pm
en
t an
d R
efo
rm C
om
missio
n
Trad
e C
om
missio
n
Ed
uca
tion
Bu
rea
u
Scie
nce
& Te
ch
no
log
y Bu
rea
u
Eth
nic A
ffairs O
ffice
Pu
blic
Sec
urity B
ure
au
Civ
il Affa
irs Bu
rea
u
Leg
al A
ffairs B
ure
au
Fina
nc
e B
ure
au
So
cial Se
curity
Bu
rea
u
Co
nstru
ctio
n a
nd
Pla
nn
ing
Bu
rea
u
En
viron
me
nta
l Pro
tec
tion
Bu
rea
u
Tran
spo
rtatio
n B
ure
au
Wa
ter A
ffairs B
ure
au
Ag
ricu
lture
Bu
rea
u
Fam
ily P
lan
nin
g B
ure
au
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
38
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 tax-
for-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.
39
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.
40
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)
Organize Public Services
(youchang/wuchang gongyi)
Runs Businesses
(yingyexing)
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)
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)
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).
41
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.
42
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, risk-
taking 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.
43
hamper market development and feed corruption. Another reason for the demise of bureau-
operated 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 extra-
bureaucracies 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.
44
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)
Receive state
funding
Entirely Self-
funded
TOTAL
Core bureaucracy
(jiguan danwei)
8,926,450 96,733 9,023,183
(19.4%)
Extrabureaucracies
(shiye danwei)
35,199,835 2,312,272 37,512,107
(80.6%)
TOTAL
44,126,285
(94.8%)
2,409,005
(5.2%)
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 fully-
funded 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.
45
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 within-
budgetary 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;
46
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 extra-
budgetary 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 2007-
129). 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,
47
“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).
48
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
49
• 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 2008-
146).
50
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 “red-
headed 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.
51
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).
52
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
53
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
54
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.
55
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
56
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 present-
day 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.
57
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 2007-
151).
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 key-
point 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.
58
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 “rule-
abiding” 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
59
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.
60
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.
61
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 extra-
budgetary 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.”
62
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
63
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.
64
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.
65
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 bureau-
contracting 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
66
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.
67
Figure 2.4: ‘Rules of the Game’ – A Simulation of Budget Procedures and Outcome
Within-Budgetary
Funds
Extra-Budgetary
Funds
Local Taxes
Deficit of 10 mil Y; owed
wages, lower benefits
Surplus of 5 mil Y; more bonuses
or benefits for staff or rollover
funds to next year
Breaks even
Revenue sent to
upper levels
Revenue retained
at local level
Budgeted Funds =
100 million Y
(for administrative
expenditure)
Pooled extra-
budgetary funds
Department B
(allocated 30 mil Y)
Department A
(allocated 50 mil Y)
Department B
(allocated 20 mil Y)
5 mil Y
25 mil Y
50 mil Y
35 mil Y
15 mil Y
20 mil Y
50 mil Y
50 mil Y
Actual
Expenditure Unmet Needs Departmental
Funds
100 mil Y
Department A’s Funds
Department B’s Funds
Department C’s Funds
Fiscal Transfers
Mobilizable
Funds
68
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 self-
financed income, as I have analyzed in Chapters 2 and 3. Factor (1) results in wide inter-
governmental 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.
69
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 sub-
provincial) 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.
70
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 well-
documented 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.
71
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
PJCP OTHERS TOTAL
WAGE
PAYMENTS
SUB-TOTAL 7012502
(46%)
3638241
(29%)
10650743
(38%)
Basic Wages 2651908 1442553 4094461 (38%)
Allowances 2482715 918154 3400869 (32%)
Bonuses 981856 645980 1627836 (15%)
Social security 183090 179496 362586 (3%)
Others 712933 452058 1164991 (11%)
ADMINISTRATIVE
BENEFITS
SUB-TOTAL 7253048
(48%)
8196625
(66%%)
15449673
(56%)
Office expenditure 506178 552040 1058218
Printing 98981 232221 331202
Utilities 289217 169854 459071
Postage 247669 130962 378631
Heating 137585 101362 238947
Transportation 791389 384535 1175924
Travel 358778 218716 577494
Conference 125891 258913 384804
Training 148431 162313 310744
Entertainment 242032 315562 557594
Benefit fee 153438 111464 264902
Labor fee 72883 167824 240707
Rental 41808 57384 99192
Asset management 38754 61840 100594
Office furnishing 329888 328356 658244
Vehicles 482164 215830 697994
Others 3187962 4727449 7915411 (51%)
SUBSIDIES
SUB-TOTAL 807524
(5%)
520369
(4%)
1327893
(5%)
Subsidies for pre-retirement 42989 18898 61887
Subsidies for retirement 256467 176490 432957
Health subsidies 50542 40368 90910
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.
72
Housing subsidies 276024 155614 431638
Scholarships 824 566 1390
Others 180678 128433 309111
TOTAL 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 in-
kind 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. 57
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.
73
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.
74
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
75
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)
Mean Std. Dev Min Max
PJBJ 4863.98 4079.04 595.26 21132.19
Others 3989.36 3240.00 116.49 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.
76
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 Personnel and Administrative
Spending Per Cadre (Yuan)
Commerce Bureau 126,035
Cultural Market Regulation Team 107,323
Party School 104,123
Public Security Bureau 94,406
Disciplinary Committee 91,148
Personnel Bureau 87,439
Finance Bureau 81,162
National Development & Reform Commission 75,585
Cultural Commission 72,930
Environmental Bureau 70,553
Relics Preservation Center 62,089
Civil Affairs Bureau 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 intra-
bureaucratic 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.
77
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).
78
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 high-
powered 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.
79
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 profit-
maximizing 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
80
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 sky-
rocketed 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.
81
A third implication of our analysis of cadre rewards is that the coexistence of two sets of high-
powered 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. High-
powered 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.
82
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
83
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 bureau-
contracting. 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 long-
term 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).
84
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.
85
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
86
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
87
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
CENTRAL
PROVINCES
SUB-
PROVINCES
Inter-governmental
contracting
Intra-
governmental
contracting
Intra-
governmental
contracting
Intra-
governmental
contracting
3
A B C
A B C
A B C
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,
88
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 bureau-
contracting. In doing so, I distinguished self-financing in the Chinese context from state-
corroding 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.
89
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, self-
financed 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
90
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.
91
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 sub-
provincial 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.
92
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.
93
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 4.2: Decomposition of Public
Spending
Figure 4.3: County Spending Structure
83%
9%
8%
76%
14%
10%
74%
16%
10%
74%
16%
10%
74%
15%
11%
02
0,0
00
40
,00
06
0,0
00
80
,00
0
Pu
blic
Sp
en
din
g (
10
00
0 Y
ua
n)
2001 2002 2003 2004 2005
Cadre benefits Acquisitions & repairs
Social benefits
Figure 4.4: Revenue
Structure
44%
32%
24%
37%
29%
34%
33%
30%
32%
35%
33%
32%
40%
31%
29%
020,0
00
40,0
00
60,0
00
80,0
00
Am
ount of in
com
e (
10000 Y
uan)
2001 2002 2003 2004 2005
tax revenue fiscal transfers
self-financing
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
94
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)
Central Taxes Local Taxes Shared Taxes
1. Customs tax
2. Consumption tax and
value-added tax
collected by the customs
authority
3. Consumption tax
4. Central enterprise
income tax
5. Income tax of local
banks, foreign banks,
and non-bank financial
institutions
6. Commerce tax, income
tax, profit and urban
management &
construction tax
collected from railway
companies, banks, and
insurance companies.
7. Retained profits of
central state-owned
enterprises
8. Export refund of foreign-
invested enterprises
1. Commerce tax
2. Local enterprise income
tax
3. Retained profits from
local state-owned
enterprises
4. Personal income tax
5. Urban land use tax
6. Fixed assets tax
7. Urban management &
construction tax
8. Real estate tax
9. Vehicle and vessel tax
10. Stamp tax
11. Slaughter tax
12. Agriculture tax
13. Agriculture special
product tax
14. Arable land use tax
15. Deed tax
16. Inheritance tax
17. Land appreciation tax
18. Revenue from leasing
state land
1. Value-added tax
Central 75%
Local 25%
2. Resource tax
Petrol tax belongs to
central; the remaining
to local governments
3. 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.
95
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
Mean Std. Dev Min Max
Cadre rewards (million yuan) 447 244 71 1479
Cadre rewards per capita (yuan) 755 447 178 4010
Cadre rewards per cadre (yuan) 29,222 16,990 8,970 150,305
Change in cadre rewards 58 93 -357 450
Change in cadre rewards per capita 99 177 -716 954
Change in cadre rewards per cadre 3,789 6,620 -14,843 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.
96
Table 4.3: Descriptive Statistics of Income Streams, 2001-2005
Mean Std. Dev Min Max
Tax revenue per capita (yuan) 435 402 13 4270
Fiscal transfers per capita (yuan) 352 273 0 2483
Self-financed income per capita (yuan) 316 172 31 1756
Change in tax revenue per capita (yuan) 17 231 -1199 1541
Change in fiscal transfers per capita (yuan) 21 149 -779 767
Change in self-financed income per capita (yuan) 31 91 -475 582
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 error-
correction 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 long-
term 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.
97
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.”
98
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 self-
financed 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.
99
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)
DV= D. Cadre Rewards per Cadre DV= D. Cadre Rewards per Capita
Model (1) Model (2) Model (3) Model (1) Model (2) Model (3)
L. tax revenue
per capita
3.997***
(1.013)
16.742***
(2.516)
13.969***
(2.716)
0.155***
(0.024)
0.542***
(0.060)
0.526***
(0.066)
D. tax revenue
per capita
12.073***
(1.108)
19.475***
(1.942)
18.188***
(1.973)
0.332***
(0.027)
0.540***
(0.046)
0.531***
(0.048)
L. self-financed
income per
capita
-6.188***
(1.603)
4.034
(4.041)
3.864
(4.015)
0.126**
(0.050)
0.479***
(0.103)
0.483***
(0.104)
D. self-financed
income per
capita
25.483***
(2.164)
25.460***
(2.804)
25.605***
(2.867)
0.849***
(0.055)
0.941***
(0.068)
0.949***
(0.071)
L. fiscal
transfers per
capita
-0.357
(1.065)
4.062
(2.683)
-1.808
(3.143)
0.258***
(0.032)
0.445***
(0.067)
0.421***
(0.079)
D. fiscal
transfers per
5.649***
(1.431)
7.863***
(2.066)
4.680**
(2.320)
0.348***
(0.038)
0.430***
(0.051)
0.420***
(0.058)
100
capita
Controls
L. public
employees (in
1000)
-0.279***
(0.082)
-0.932**
(0.383)
-1.128***
(0.392)
-0.003
(0.002)
0.008
(0.009)
0.007
(0.009)
D. public
employees
-1.631***
(0.262)
-1.167***
(0.328)
-0.954***
(0.333)
-0.001
(0.007)
-0.009
(0.008)
-0.008
(0.008)
L. population (in
10000)
53.202***
(19.168)
434.587***
(88.311)
394.431***
(88.058)
0.268
(0.442)
2.007
(2.095)
1.870
(2.134)
D. population 10.765
(37.788)
-176.650***
(56.463)
-172.595***
(55.810)
-5.459***
(0.995)
-3.847***
(1.382)
-3.789***
(1.388)
L. urban share
of population
-0.803
(10.953)
73.076
(45.863)
45.024
(46.540)
0.013
(0.276)
-1.483
(1.117)
-1.566
(1.152)
D. urban share
of population
26.562
(20.306)
-31.653
(42.575)
-8.102
(43.069)
0.061
(0.518)
1.207
(1.037)
1.250
(1.070)
L. GDP per
capita (in 10000
Y)
1056.001***
(281.238)
2161.292***
(578.929)
1548.202**
(603.871)
-6.122
(6.722)
18.861
(13.777)
16.870
(14.828)
D. GDP per
capita
-214.397
(413.255)
-1181.3**
(491.689)
-801.781
(497.619)
25.643**
(10.454)
-2.074
(11.828)
-0.574
(12.245)
Lagged DV -0.014
(0.034)
-0.445***
(0.051)
-0.482***
(0.518)
-0.237***
(0.039)
-0.636***
(0.048)
-0.637***
(0.049)
County Effects? No Yes Yes No Yes Yes
Time Effects? No No Yes No No Yes
Constant 3369.017
(969.230)
-13326.25
(7912.835)
575.176
(8881.792)
54.00
(22.98)
-293.38
(192.26)
0.637
(0.049)
N 556 556 556 556 556 556
R2 0.50 0.59 0.60 0.60 0.68 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
101
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
102
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
Change in Revenue
Per Capita
Short-Term Effects Long-Term Effects Total Effects on
CADREW per cadre
Scenario I TAXREV = $153
SEFLIN = $28
$2,754 + $728 =
$3,482
$4,437 + $0 =
$4,437
$7,919
Scenario II TAXREV = $0
SEFLIN = $181
$0 + $4,706 =
$4,706
$0 + $0 =
$0
$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
103
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 long-
term 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 far-
sighted 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).
104
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.
105
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
106
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 self-
financed 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.
107
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.
108
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.
109
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.
110
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 fee-
taking 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
111
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)
112
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 extra-
budget 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 extra-
budgetary 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.
113
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.
114
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 rules-
based and came closer to approximating the ideal-type of bureau-contracting.
115
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, reform-
minded leaders, from the central party committee down to villages, were eager to employ high-
powered 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
116
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.
117
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.
118
Stage 2 Reform: 1993 onwards
The year 1993 marked the beginning of full market reforms in China, culminating in the ground-
breaking 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
119
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
120
governments have had to shoulder a disproportionate share of public spending even as its share
of revenue dropped over the years.
Figure 5.1: Local Revenue and Spending Before and After the 1994 Fiscal Reform
After 1994 Fiscal Reform
-40
-20
020
40
0500
1000
1500
2000
2500
3000
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.”
121
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).
Figure 5.2: Change in Composition of Extrabudgetary Revenue, 1980-2003
0100
200
300
400
500
Extr
a-b
udgeta
ry r
evenue (
bill
ion y
uan)
19
80
19
81
19
82
19
83
19
84
19
85
19
86
19
87
19
88
19
89
19
90
19
91
19
92
19
93
19
94
19
95
19
96
19
97
19
98
19
99
20
00
20
01
20
02
20
03
Bureaucracies 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
122
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
123
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).
Figure 5.3: Change in Public Employment Size, 1955-2003
010
20
30
40
50
Num
ber
of
Public
Em
plo
yees (
mill
ions)
12
34
56
Public
Em
plo
ym
ent
Per
Capita
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
124
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.
125
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.
126
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 pre-
determined 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).
127
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
128
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
129
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.
Figure 5.4: ‘Eating budget’ – Cadre rewards consume tax revenue and fiscal transfers
0
10,000
20,000
30,000
40,000
50,000
60,000A
mo
un
t o
f S
pe
nd
ing
/In
co
me
(1
00
00
Yu
an
)
2001 2002 2003 2004 2005
tax revenue and fiscal transfers 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
130
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
131
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.
132
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 bureau-
contracting.
133
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 bureau-
contracting 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
134
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, rule-
following 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.
135
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 second-
best 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
136
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
137
“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
138
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.
139
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
140
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
141
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 non-
democratic regime like China, the interest groups who know and can manipulate the inner
workings of bureaucracy are none other than the bureaucrats themselves.
142
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.
143
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).
144
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.
145
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).
146
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)
147
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)
148
Table 2: List of Extrabureaucracies Under Selected Bureaus
Bureau/Administrative Unit
“Supervising Department”
Extrabureaucracies/Service Units
Construction bureau 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
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
149
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
150
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 Basic
Wages
Functional wages by grade Basic wages (jiben gongzi) and functional wages
(zhiwu gongzi) combined
1 2 3 4 5 6 1 2 3 4 5 6
Mayor 40 190 165 150 140 130 120 230 205 190 180 170 160
Vice mayor 40 150 140 130 120 110 100 190 180 170 165 150 140
Ju/Chu Chief 40 130 120 110 100 91 82 170 160 150 140 131 122
Vice Ju/Chu Chief 40 110 100 91 82 73 65 150 140 131 122 113 105
Ke Head 40 82 73 65 57 49 42 122 113 105 97 89 82
Vice Ke Head 40 65 57 49 42 36 30 105 97 89 82 76 70
Ke Officer 40 49 42 36 30 24 18 89 82 76 70 64 58
Support Staff 40 42 36 30 24 18 12 82 76 70 64 58 52
Source: Xu 2007, p. 197
151
Table 2: Wages by Function and Grade
Date of Issue: November 1993
Unit: yuan/month
Function Wages by Function (zhiwu) Wages by Grade
(jibie)
Basic
(jiben)
Wages
Length of
work
(gong
ling)
wages
1 2 3 4 5 6 7 8 Grade Wage
One yuan
added for
every
year of
service
Chairman
Vice-Chairman
Premier
480 550 630 1 470 90
2 425 90
Vice Premier
Member of the
State Council
400 460 520 580 3 382 90
4 340 90
Minister
Vice Minister
330 380 430 480 530 5 298 90
6 263 90
Si Chief
Ting Chief
270 315 360 405 450 7 228 90
Vice Si Chief
Vice Ting Chief
118 143 168 193 218 243 8 193 90
Chu Chief
County Head
144 174 204 234 264 294 9 164 90
Vice Chu Chief
Vice County Head
118 143 168 193 218 243 10 135 90
11 111 90
Ke Head 96 116 136 156 176 196 216 12 92 90
Vice Ke Head 79 94 109 124 139 154 169 13 77 90
Ke Executive 63 75 87 99 111 123 135 147 14 65 90
Support Staff 50 60 70 80 90 100 110 120 15 55 90
Source: Xu 2007, p. 206
152
Table 3: Regional subsidies for various “hardship” regions
Date of Issue: February 2001
Unit: yuan/month
Region (1) Region (2) Region (3) Region (4)
Average subsidy 43 86 172 300
Provincial level and above 100 200 -- --
City level 80 160 320 560
County level 60 120 240 420
Ke level 47 95 190 330
Ke and below 40 80 160 280
Source: Xu 2007, p. 203
153
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.
154
Table 1: Interviews by Location
Central government 18
Anhui 4
Beijing 2
Chongqing 1
Guangdong 8
Guangxi 1
Hunan 2
Jiangsu 28
Shandong 15
Shanghai 1
Sichuan 41
Tianjin 44
Table 2: Interviews by Level of Government
Central 18
Provincial level 14
City level 27
County level 88
Township and villages 18
Table 3: Interviews by Governmental Sector/Department
Sector/Department No.
Administrative Office 4
Agriculture 2
Arts & Culture 3
Civil Affairs 13
Construction 3
Court 3
Education 23
Environment 12
Finance 22
Health 7
NDRC (National Development & Reform Commission) 7
Party Organs 6
Personnel 21
Research 3
Township and village governments 17
Others 19
155
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.
156
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 rent-
seeking 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.
157
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 hyper-
growth . 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.
158
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 central-
local 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.
159
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.
160
—. 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): 681-
712.
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.
161
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.
162
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.
163
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.
164
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.
165
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.