Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Shequ construction : policy implementation, community building, and urban governance in China Shieh, Leslie L. 2011

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

Item Metadata

Download

Media
24-ubc_2011_spring_shieh_lesliel.pdf [ 4.53MB ]
Metadata
JSON: 24-1.0071650.json
JSON-LD: 24-1.0071650-ld.json
RDF/XML (Pretty): 24-1.0071650-rdf.xml
RDF/JSON: 24-1.0071650-rdf.json
Turtle: 24-1.0071650-turtle.txt
N-Triples: 24-1.0071650-rdf-ntriples.txt
Original Record: 24-1.0071650-source.json
Full Text
24-1.0071650-fulltext.txt
Citation
24-1.0071650.ris

Full Text

SHEQU CONSTRUCTION: POLICY IMPLEMENTATION, COMMUNITY BUILDING, AND URBAN GOVERNANCE IN CHINA by  LESLIE L. SHIEH  B.Sc., Cornell University, 1998 MCP, University of California, Berkeley, 2000    A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in  THE FACULTY OF GRADUATE STUDIES  (Community and Regional Planning)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)  March 2011  © Leslie L. Shieh, 2011   ii ABSTRACT  China’s nationwide Shequ (Community) Construction project aims to strengthen neighbourhood- based governance, particularly as cities wrestle with pressing social issues accompanying the country’s economic reforms. This policy has produced astounding outcomes, even though it is implemented through experimentation programs and the interbureaucratic document system rather than through legislation. It has professionalized the socialist residents’ committees and strengthened their capacity to carry out administrative functions and deliver social care. Thousands of service centres have been built, offering a range of cultural and social services to local residents.  This research addresses how the centrally promulgated policy is being implemented locally and what its impacts are in various neighbourhoods. The lens of community building is used to explore how the grass roots organize themselves and how they are defined and governed by the state. The research thus seeks to analyze the impact of Shequ Construction, not through measuring outcomes against the intentions set out in policy documents, but through considering the wider, sometimes unforeseen, implications for other processes going on in the city. Based on fieldwork in Nanjing, the chapters explore the meaning Shequ Construction has in four areas of urban governance: 1) fiscal reform and decentralization of public services, 2) suburban village redevelopment, 3) community-based social service provisioning through the emergent nonprofit sector, and 4) role of homeowners’ association under housing privatization and neighbourhood inequality.  By examining the interaction of Shequ Construction with a diverse set of policies, this research demonstrates how policy becomes interpreted during the course of implementation by local agencies as they contend with realities on the ground; and conversely how the Shequ policy alters the course and outcome of other policies and projects simultaneously unfolding. Furthermore, the perspective of policy interactions sheds light on the policy-making process in China. In presenting the Chinese experience, this dissertation seeks to contribute to the broader planning discourse on the function and appropriation of community building as a means of urban governance.     iii PREFACE  A version of chapter 6 has been accepted for publication. Shieh, Leslie (forthcoming). “Awaiting Urbanization: Urban Village Redevelopment in Coastal Urban China,” in Arif Dirlik and Alexander Woodside (eds) Global Capitalism and the Future of Agrarian Society. Boulder, CO: Paradigm Publishers. A version of chapter 7 has been accepted for publication. Shieh, Leslie (forthcoming). “Nongovernmental Organizations in Contemporary China: Development of Community- based Social Service Organizations,” in Huhua Cao and Jeremy Paltiel (eds) China in the Twenty-First Century: Multidisciplinary Perspectives from Students in Canada and in China. Ottawa, ON: University of Ottawa Press. The research undertaken was approved by the UBC Behavioural Research Ethics Board (certificate number H06-04074).   iv TABLE OF CONTENTS  Abstract .......................................................................................................................................... ii  Preface ........................................................................................................................................... iii  Table of Contents .......................................................................................................................... iv  List of Tables ............................................................................................................................... viii  List of Figures ............................................................................................................................... ix  Acknowledgement.......................................................................................................................... x  1. Introduction: Policy Implementation, Community Building, and Urban Governance ...... 1  Community Building in China’s Urban Transition ..................................................................... 1  Bureaucratic Hierarchy and Cellular Units .................................................................................. 6  Shequ Construction Research ...................................................................................................... 9  Danwei and shequ .................................................................................................................... 9  Shequ policy objectives and content ...................................................................................... 11  Urban development and shequ ............................................................................................... 14  Policy Implementation in China ................................................................................................ 15  Research Questions: Shequ Construction in China’s Urbanization ........................................... 18  Fiscal reform .......................................................................................................................... 18  Urban village redevelopment ................................................................................................. 19  Nonprofit sector in community-based service delivery ......................................................... 19  Interest-based community ...................................................................................................... 20  Dissertation Outline ................................................................................................................... 20  2. Research Design: Methodology, Fieldwork Process, and Methodological Issues .............. 23  Introduction ................................................................................................................................ 23  Overview of Research Design ................................................................................................... 24  Selection of Methods ................................................................................................................. 28  Initial fieldwork ..................................................................................................................... 29  Refocused fieldwork .............................................................................................................. 30  Access and Sampling ................................................................................................................. 33  Sources of Data and Collection ................................................................................................. 37  Shequ researchers .................................................................................................................. 37  Bureau officials ...................................................................................................................... 38  Shequ and urban villages ....................................................................................................... 39  Social organizations ............................................................................................................... 40  Documents ............................................................................................................................. 41  Conditions of Fieldwork and Research Limitations .................................................................. 42  Positionality: On being a foreign researcher of Chinese descent .......................................... 42  Official approval and sampling bias ...................................................................................... 43  Note on Referencing Fieldwork ................................................................................................. 44  3. A Historical Overview: Urban Community in China .......................................................... 45  In Other Worlds and In Other Words: Community and Shequ ................................................. 45  Morality and Elite Activism in Imperial Streets and Wards ...................................................... 49   v On Self-Governance: Early Political Thoughts and Debates ..................................................... 53  Republican Modernity and Traditional Continuity .................................................................... 56  Birth of the Residents’ Committee ............................................................................................ 61  Legal Standing of the Residents’ Committee ............................................................................ 67  The Appearance of Shequ: Modern Term, Old Concept ........................................................... 68  Observing China and a Chinese Perspective ............................................................................. 72  4. Shequ Jianshe: China's Community Construction Policy Agenda ..................................... 74  An Experiment’s Experiment .................................................................................................... 74  Changing Conditions: 1980s to 1990s ....................................................................................... 75  State enterprise restructuring and danwei-based welfare reform ........................................... 77  Rising unemployment and urban poverty .............................................................................. 80  Urban social assistance programs and trends towards welfare socialization ......................... 82  Management of the floating population ................................................................................. 84  Party building and keeping in touch with the masses ............................................................ 85  From Services to Construction: Two Decades of Shequ Experiments ...................................... 86  Phase 1: Shequ services (1983 to 1992) ................................................................................ 87  Phase 2: Greater experimental autonomy (1993 to 1997) ..................................................... 88  Phase 3: Shequ services to Shequ Construction (1998 to 2000) ............................................ 89  A Look at Document 23: Advancing Shequ Construction Nationwide ..................................... 91  1. What are the governmental rationales? .............................................................................. 93  2. What is the objective of Shequ Construction? ................................................................... 94  3. What is being constructed? ................................................................................................ 96  4. Who will be doing the construction? ............................................................................... 100  5. Where does Shequ Construction stand among state priorities? ....................................... 104  Post Document 23 .................................................................................................................... 104  5. Fiscal Decentralization and Shequ Reform: Nanjing Implementation ............................ 107  Viewing the Local State through Shequ Construction ............................................................. 107  Fiscal Reforms and Intergovernmental Fiscal Relations ......................................................... 109  Decentralization and Spending on Social Services and Welfare ............................................. 113  Municipality-District Relations ............................................................................................... 114  Shequ Construction in Gulou District, Nanjing ....................................................................... 117  Shequ Residents’ Committee Election .................................................................................... 119  Shequ Residents’ Committees and the Shequ Party Branch .................................................... 122  Mounting Responsibilities and Evaluation Standards ............................................................. 125  Shequ Budget: A Gulou District Experiment .......................................................................... 132  Shequ Residents’ Committees as Extrabureaucrats ................................................................. 134  6. Shequ Construction And Urban Expansion: Nanjing's Urban Village Redevelopment 136  Awaiting Urbanization ............................................................................................................. 136  Integration: Framing the Urban Village Phenomenon ............................................................. 137  Nanjing’s Urban Village Redevelopment ................................................................................ 139  Redevelopment through Dissolution: Land Acquisition and Relocation ................................ 142  Dissolving the village collective .......................................................................................... 145  Relocation and social integration into urban neighbourhoods ............................................. 147  Redevelopment through Integration: Conformity of Interests ................................................. 150  Co-existence of villagers’ committee and shequ residents’ committee ............................... 151  Standardization and the Village Shequ Construction Program ........................................... 152   vi Negotiating Urbanization ......................................................................................................... 155  7. Nonprofit organizations in Shequ Construction: Role of Minfei Organizations ............. 157  The Shequ Service Industry ..................................................................................................... 157  Social Organizations and the Emergence of Minfei Organizations ......................................... 158  Definition of social organizations ........................................................................................ 158  A new category: minfei organizations ................................................................................. 162  Sunrise Senior Care Services: A Model of Collaborative Service Provision .......................... 165  Community college for the elderly ...................................................................................... 167  Social insurance ................................................................................................................... 168  Government contracts .......................................................................................................... 169  Shequ as the Intermediary Layer in Urban Elder Care ............................................................ 170  A New Type of Seniors’ Centre: Lateral Partnerships at the Grass Roots .............................. 174  Understanding Shequ Construction through Minfei Participation ........................................... 177  Corporatist third sector ........................................................................................................ 178  Lateral organizational relations ........................................................................................... 180  8. Homeowners’ Associations in Shequ Construction: Their Incorporation into the Shequ Structure .................................................................................................................... 182  Informal Community Construction .......................................................................................... 182  Emergence of Homeowners’ Associations .............................................................................. 184  Housing reform and the emergence homeowners’ associations .......................................... 184  Incidents of homeowners’ resistance ................................................................................... 187  Relations between Homeowners’ Associations and Residents’ Committees .......................... 189  Regulations on Homeowners’ Association – Residents’ Committee Relations ...................... 193  Reaction at neighbourhood level ......................................................................................... 194  Reaction at local policy level ............................................................................................... 195  Debate over Incorporating Homeowners’ Associations into the Shequ .................................. 197  The case for incorporation ................................................................................................... 200  The case against incorporation ............................................................................................ 201  Shequ Governance and Neighbourhoods as Civic Spaces ....................................................... 203  9. Conclusion: Shequ Construction as a Window on Urban Governance ............................ 207  At Policy Junctures .................................................................................................................. 207  Interbureau Memoranda in Chinese Policymaking ................................................................. 211  Building a Harmonious Society through Shequ Construction ................................................. 214  Conformity to (urban) standards .......................................................................................... 215  State-minfei relations ........................................................................................................... 215  Production of inequality ...................................................................................................... 216  A Point of Inarticulation .......................................................................................................... 217  Government through Community in and beyond China .......................................................... 221  Shequ Construction and Chinese Urban Planning ................................................................... 225  Interdisciplinary shequ research .......................................................................................... 226  Shequ and sustainable cities ................................................................................................ 227  Community-based nonprofit sector engagement ................................................................. 228  A Decade of Shequ .................................................................................................................. 228  References .................................................................................................................................. 230    vii Appendices ................................................................................................................................. 255  Appendix 1: Shequ governing organization research survey .................................................. 255  Appendix 2: Shequ Jianshe policies and key events ................................................................ 257  Appendix 3: Summary of Nanjing third shequ election procedure, 2006 ............................... 266  Appendix 4: Nanjing Gulou District shequ evaluation measures ............................................ 267  Appendix 5: Nanjing urban shequ and rural village standardization measures ....................... 272    viii LIST OF TABLES  Table 2.1 Overview of research design  .................................................................................. 27 Table 2.2 Phases of research  .................................................................................................. 32 Table 4.1 Employment statistics for years of concentrated SOE restructuring  ...................... 82 Table 5.1 Revenue and expenditure shares by level of government  .................................... 112 Table 5.2 Expenditure categories and shares by level of government, 2003  ....................... 113 Table 5.3 Socioeconomic statistics of Nanjing districts and counties, 2006  ........................ 118 Table 7.1 Legal categorizations of Chinese social organizations  ......................................... 159 Table 7.2 Registered popular organizations, 1988 to 2007  .................................................. 162 Table 7.3 Types of minfei organizations, 2005  .................................................................... 163 Table 7.4 Proportion of beds in different welfare facilities  .................................................. 172 Table 8.1 Responses to survey question 14  .......................................................................... 198 Table 8.2 Responses to survey question 12  .......................................................................... 199   ix LIST OF FIGURES  Figure 1.1 China’s administrative structure  ............................................................................. 6 Figure 2.1 Geographical location of Nanjing  ......................................................................... 26 Figure 2.2 Map of Nanjing  ..................................................................................................... 36 Figure 3.1 Ming Dynasty map of bridges, streets, and fang  ................................................... 52 Figure 3.2 Drawings from Nanjing’s Capital Plan (1929)  ..................................................... 60 Figure 4.1 Timeline of key events  .......................................................................................... 76 Figure 4.2 Shequ’s basic organizational structure  ................................................................ 103 Figure 5.1 Shequ centre building types  ................................................................................ 130 Figure 5.2 Facilities inside shequ centres  ............................................................................. 131    x ACKNOWLEDGEMENT  My doctoral studies and dissertation could not have been completed without the support and friendship of many people. First and foremost, I would like to thank the members of my doctoral committee. Michael Leaf, as my supervisor, has always given me the space and time to explore and make my own discoveries. John Friedmann has shared my enthusiasm for shequ and guided me through the research and writing process. He has often reminded me that good writing is just as important as good ideas. My interest in China’s urbanization came from a field studio to Quanzhou, Fujian Province led by Dan Abramson, who since that first trip to China has continued to help me make sense of the complexities of Chinese society. My warm thanks go to the shequ residents’ committee members and social organization leaders for welcoming me into their neighbourhoods and organizations and for sharing their time, stories, and experiences. The dedication and innovation with which they approach their work and the daily challenges they face motivated me to think about Shequ Construction as more than a state-led project. I am also grateful to professors at Nanjing University and Nanjing Normal University. Government officials in the Planning Bureau and Civil Affairs Bureau have facilitated my research by helping me understand the contents of Shequ Construction and introducing me to shequ residents’ committee directors. The Chinese professors and officials persistently challenged me to think from a Chinese perspective, and for that, I am particularly grateful. Though I have not mentioned their names, as the research protocol required that their identities be kept anonymous, they are very much appreciated. I would like to acknowledge the financial support from the Province of British Columbia, Brahm Wiesman Memorial Scholarship, Stephen D. Hassenfeld Fellowship, and the Fukien Chinese Association. I benefited greatly from my time as a student and then as a junior research fellow at the Nanjing-Hopkins Center. The professors and librarian helped me navigate sources and made introductions on my behalf. The directors and staff, in DC and Nanjing, provided much support, from arranging my visas to making the Center a home away from home. I have made  xi lifelong friendships at the Centre. I am particularly thankful to my Centre friends for being there during the ups and downs of fieldwork. I also want to thank the external and university examiners for their constructive comments on an earlier version of this dissertation: Dr. David Bray, Dr. Gavin Shatkin, Dr. Amy Hanser, Dr. Timothy Cheek, and Dr. Diana Lary. Their written reports and questions during the final examination were valuable in the completion of this final version and for thinking about future research directions. At various stages of the writing process, I received valuable feedbacks from members of the UBC China Studies Group and participants of the 2008 UBC Wall Summer Institute. I also want to thank Leslie Prpich, Danielle Labbe, Silvia Vilches, and Mark Shieh, who although are not researchers of China, read drafts of my chapters, provided feedback, and pressed for greater clarity. I am blessed to have the warm friendships of many classmates at SCARP who have accompanied me through the happy, confusing, and disappointing moments of this long journey. Finally, and above all, I am indebted to my family who has provided encouragement and support on a daily basis. As I write these final words of acknowledgements, I feel immense gratitude to the many people who made this long endeavour possible. It has not been a lonely pursuit. Many have contributed to this dissertation. Its limitations are my own.  1 1. INTRODUCTION Policy Implementation, Community Building, and Urban Governance  Community Building in China’s Urban Transition Enfolded in what observers have termed China’s “urban transition” (Panell 1995, 2002; Friedmann 2005), the “great urban transformation” (Hsing, 2010), and “urban revolution” (Campanella 2008) that capture the scale and pace of the country’s urban development are extraordinary lived experiences and enormous pressures to govern. My research seeks to explore the implications of China’s rapid urbanization for notions of community. More specifically, it asks how community is being reconstituted in the reform era to address pressing social issues. In their vision of community, officials have focused on the outcomes and potentials of a recent policy initiative, Shequ Construction, that seeks to create neighbourhood units by revitalizing the socialist residents’ committee (jumin weiyuanhui). Set up under statute during the Maoist era, the residents’ committee’s basic functions consisted of liaising between residents and the government, disseminating official policies, assisting with local policing, and mediating neighbourhood quarrels. Its members were usually housewives and retired elderly, appointed by officials but unpaid. In the mid 1980s, as market reform policies unfolded, cities were faced with tighter fiscal budgets, a growing urban population, the demise of work-unit-based social welfare, and rising unemployment. At the neighbourhood level, the antiquated residents’ committees strained to keep up with the social changes and greater demand for social assistance. The central government saw in their efforts the potential for residents’ committees to carry some of the increasing burdens on local governments. The administrative area under the jurisdiction of a residents’ committee is referred to as shequ,1 or “neighbourhood community” in official parlance. The reform of the residents’ committee is officially referred to as Shequ Jianshe or “Community Construction.” Renamed  1 The Chinese word shequ (pronounced shè qū) is commonly translated into English as “community.” The term is composed from the characters for “social” (she) and “district” (qu), referring to a collective identity in a defined space. Its origin is explored in greater depth in the next chapter.  2 the shequ residents’ committee,2 the mass organization of self-management that came into existence in the 1950s now governs a larger jurisdiction, is composed of a younger and more professionalized staff, and is charged with providing specific social services. Experiments with Shequ Construction began in the 1980s as community social services (shequ fuwu) to provide care to society’s most vulnerable members, particularly the elderly, the handicapped, and the thousands of workers laid off from dissolved or bankrupted state-owned enterprises (SOE). Initial successes in selected pilot sites in various cities led the Ministry of Civil Affairs (MCA) to undertake further experiments with restructuring the residents’ committee to provide more extensive governance functions at the neighbourhood level. After two decades of debate and experiments, on November 19, 2000, the Central Committee and State Council endorsed the first formal document concerning Shequ Construction: Memorandum from the Ministry of Civil Affairs on Promoting Urban Shequ Construction throughout the Nation (hereafter Document 23). This memorandum did not signify that the shequ construction process had finished, but rather that debates and experiments had reached certain conclusions to move ahead with a nationwide policy. As this research demonstrates, in its implementation many arising issues remain to be resolved. Almost a decade has passed since the promulgation of the central policy document, and its outcomes are impressive. In cities across China, thousands of neighbourhood service centres have been built and offer a range of services, from registering unemployment to providing welfare services to organizing cultural activities. There are also less obvious and more difficult to measure outcomes, such as the impact of numerous self-organized social groups (choirs, book clubs, dance groups, etc.) and the earnestness with which members of the residents’ committees approach their underpaid responsibilities to care for those in need in their neighbourhoods. These direct outcomes represent one facet of the impact Shequ Construction has on cities. From another perspective, this research seeks to understand the impact of the Shequ Construction policy program, not through measuring outcomes against the intentions set out in the policy document, but through analyzing the wider, sometimes unforeseen, implications  2 For better readability, throughout the dissertation the “shequ” in the renamed “shequ residents’ committee” is dropped. The full phrase is only used to emphasize and distinguish between the old residents’ committee and the new shequ residents’ committee.  3 for other processes going on in the city. It examines how the policy confronts other initiatives undertaken by social actors and other state agencies, which, at first blush, may seem to have little to do with Ministry of Civil Affairs’ shequ work. By articulating Shequ Construction with a diverse set of initiatives, it asks how the policy is being reinterpreted in the course of implementation as local agencies contend with realities on the ground and how the recent state-led emphasis on shequ reform has altered the outcome of other urban governance projects that are concurrently being undertaken. The set of policy interactions examined in this dissertation is outlined later in this introductory chapter. Beyond the specifics of the shequ policy, the broader urban planning question this dissertation seeks to explore is the role of community building in the governance of an urbanizing society. I am primarily concerned with urban governance as the coordination between central directive and local implementation and as mechanisms of social control and social service delivery (Wu 2002, 1072-3). My aim is not to present a Chinese model of community building. Rather, through the lens of community-building, I hope to begin to understand how the grass roots organize themselves and how they are defined and governed by the state, particularly during times of tremendous social and political change. As a planner educated in the North American planning tradition, I believe the importance of asking these questions in the Chinese context privileges sociopolitical processes in shaping the urban structure and challenges some fundamental assumptions about community in Western planning thought. Visitors to China will unequivocally note American forms of spatial community: suburban new towns, gated housing estates, and neighbourhoods segregated by income and increasingly by class. Indeed, China’s capitalist economic development has initiated a vibrant debate over whether there is evidence that Chinese cities are converging with Western urbanization patterns (Dick and Rimmer 1998; Ma and Wu 2005; Huang 2006; Lin 2007). However, as those opposing the convergence thesis have argued, we must look beyond these similar outward features suggesting a convergence of capitalist urban form and examine the sociopolitical forces and policy mechanisms that have created them. To privilege social processes removes the assumption that the practices of the Third World would necessarily converge toward the “superior norms” of Western cities (Ma and Wu 2005, 12). In thinking cross-culturally about the notion of community in China,  4 attention to process is particularly important as community has a central place in Western social thought and in American history of social action. Relative to community development in the United States and Canada, much less is known about how China’s new urban political economy has altered approaches to neighbourhood governance. In the North American context, the concern over community was born out of a time of accelerated urbanization. Western urban studies of the early twentieth century (for example, Tönnies 1887/2001; Simmel 1908/1988; Park, Burgess, and McKenzie 1925/1967; Perry 1929; Wirth 1938/1970) demonstrate a period when social scientists sought to make sense of capitalism, modernization, and the impacts these changes had on traditional ways of life. The community-building movements that proceeded have sought to address urban problems of unemployment, poverty, rootlessness, and social segregation.3 Similar concerns have arisen in contemporary Chinese society, particularly now, as the social ramifications of three decades of rapid economic development are becoming apparent. Because a shequ discourse is only just beginning to form within Chinese social thought, Western concepts of community are often used to discuss China’s shequ movement.4 But these theoretical concepts grew out of a particular place and time; while they are influential in how we make sense of social change, they are not universal. In the country’s transition from a socialist planned economy to a market oriented economy, governance challenges and the policy instruments used to respond to them are both socialist and capitalist in nature. A critical understanding of neighbourhood governance and community building must necessarily be situated within China’s new urban political economy, social issues, and policy mechanisms. To planners in North America, the idea of neighbourhood residents engaged in mutual help and empowered to self-govern is compelling (Talen 2000). Neighbourhood planning has become an important strategy through which cities address some of their most pressing problems from poverty to class conflict (Keating, Krumholz, and Star 1996). It speaks to  3 I am referring here to movements of the twentieth century, namely the settlement house (Chambers 1963, chapter 5; Kraus 1980), garden city (Howard 1902/1946; Perry 1929), and advocacy planning (Davidoff 1965; Arnstein 1969; Friedmann 1987, chapter 6). Each espoused particular ideals of community to be realized in practice. 4 For example, in a Chinese textbook on community development planning theory and practice, Zhao and Zhao (2003) begin with Tönnies’ concepts of gemeinschaft and gesellschaft and continue with various Western approaches in sociology and urban planning. While they discuss Chinese and Western approaches throughout the book, there is no explicit connection made between them.  5 qualities of democratic spirit, social capital, and associational life that are quintessential in Western notions of community (Putnam 2000). When transplanted to the Chinese context and given the legacies of Maoist thought and restrictive political freedom in China today, the idea of mutual help and self-governance is more likely to be said to resonate with socialist utopianism, engineering self-reliant communities to create social order and harmony. However, as I argue in this dissertation, the Shequ Construction program needs to be examined neither as an empowerment-based nor a dogmatic approach to community building. The complexity of the shequ movement results from the coexistence of an authoritarian state, market forces, and an emerging civil society. The state has sought changes in multiple directions, resulting in contradictions among policies and in central policies being used to achieve disparate ends to serve local interests. To more fully grasp and evaluate the impact the policy has had in over two decades of experimentation, as well as some of the potentials and challenges that lie ahead, this dissertation examines the interaction between Shequ Construction and other processes unfolding simultaneously. The subject of study, then, is not the policy program itself, but the points at which it confronts other forces, be they state-led interventions or grassroots projects. In this dissertation I have chosen not to use the English translation community, but to keep the romanized Chinese term shequ when referring to the post-reform community-building project. As chapter 3 will expand on, the concept of community is ladened with cultural values and history. In adopting the romanized shequ, I seek to dissociate connotations and differentiate the differing notions of community. Specifically, shequ is not an interest-based community but an administrative, spatial community. And, shequ construction speaks to the building of residents’ committees’ capacity to govern. The rest of this introductory chapter is divided into five sections. The first places the residents’ committee within China’s administrative context. The second section provides an overview of Shequ Construction research and the dominant themes in the current literature. In the third section, I discuss the policy-making process in China and why examining policy interactions is an appropriate approach for analyzing policy implementation. Following this, I discuss the set of policy interactions explored in this research. I conclude with a brief outline of the dissertation.    6  Bureaucratic Hierarchy and Cellular Units The Chinese polity is both hierarchical and cellular. In cities, a shequ unit comprises between 1,000 and 3,000 family households and forms the base of urban governance (figure 1.1). About 46% of China’s 1.3 billion people currently live in cities, and the majority of them live under the jurisdiction of one of more than 80,000 residents’ committees across the country that act as intermediaries between the state and its urban citizenry (NSB, Office of Social and Technological Statistics 2007, table 9-20).5                                                                            5 In 2008, China’s urban population was 607 million, or 46% (National Statistics Bureau 2009, table 3-1). Statistically, urban population is defined as the resident population in towns and cities by household registration. Resident population includes those who hold local and temporary household registration. Unregistered migrants who reside in towns and cities are not enumerated. State Council (ministries and commissions) Provincial Government Municipal Government District Government County Government County-level City Government Street Office Shequ Residents’ Committee Villagers’ Committee Township Government Street Office Shequ Residents’ Committee Villagers’ Committee Township Government Central level Provincial level Prefecture level County level Township level Grassroots level (not part of state organ Figure 1.1 China’s administrative structure The hierarchy represents the basic administrative structure. There are exceptions. For instance, as a city’s boundary expands, at the urban periphery, towns can fall under the jurisdiction of districts, and villages can fall under the supervision of street offices. Also, the administrative hierarchy of centrally administered cities and autonomous regions are slightly different.  7 Proceeding from this basic building block, typically in large cities 10 to 15 shequ units are grouped under the supervision of a street office (jiedao banshichu).6 Depending on size, generally 8 to 10 street offices report to a District People’s Government. Urban districts (shixiaqu), county-level cities (xianjishi), and rural counties (xian) fall under the supervision of a Municipal People’s Government. At the next level up, Provincial People’s Governments oversee the municipalities and rural prefectures in their jurisdiction. With higher levels of government supervising those beneath, the ultimate authority lies with the State Council and central ministries and commissions at the top. Within this administrative structure, Chinese citizenry is organized by a system of household registration (hukou). A person’s permanent hukou is the officially recognized place of residence – a specific city, district, shequ, county, and village. Permanent hukou records identify people as either agricultural or non-agricultural, which have more to do with resource allocation than actual occupation. Non-agricultural persons, including urban workers, state cadres, state farm workers, and their dependents, are broadly considered urban hukou holders. Whether they actually live in cities or work in rural areas, they are entitled to state-provided welfare. For those living outside their place of permanent residence, a temporary hukou is required to rent housing and seek local employment. The system is intended to control mobility by tying a person’s eligibility for services and right to pursue livelihood activities to a specific locality (Chan 1994; 2009; Wang 2005). The state-provided welfare is distributed through one’s work unit or residents’ committee. Working with local public security bureaus, residents’ committees manage and keep details of the household registrations of those living in their jurisdictions. Residents’ committees have a somewhat paradoxical relationship with higher levels of government. Under the Chinese Constitution, residents’ committees, considered “self- governing mass organizations,” are not administrative organs of the state (article 111). Theoretically, they are to be as self-sufficient as possible, electing representation and managing the day-to-day neighbourhood affairs. At the same time, they are mandated to  6 It is important to note that a shequ sometimes encompasses both the street office and residents’ committee. Under statutory provisions, the street office, considered the dispatch office of the district government, is the “grassroots organ of state power.” The residents’ committee under its charge is “the mass organization of self-management at the grassroots level” (Chinese Constitution, Article 111). In this dissertation, unless otherwise noted, shequ is used to refer specifically to the jurisdiction of the residents’ committee.  8 provide assistance to government bureaus and offices as required, becoming in the process an extension of the administrative apparatus. Composed of mainly retirees and women in their 50s, residents’ committee members possess no formal coercive powers over their constituents, but rather are to serve, oversee, and care for them. During the Mao era, they were instrumental in mobilizing political campaigns, conducting struggle sessions, and reporting information about residents, such as pregnancies, undocumented visitors, and suspicious activities. While residents’ committees are less politicized today, an important aspect of neighbourhood work still consists of enforcing government regulations and maintaining social order in the neighbourhoods, earning committee members the title “granny police” (Benewick and Takahara 2002; Pan 2006). They continue to watch for adherence to the one- child policy, to register new residents, to mediate disputes between spouses and neighbours, and to report anything and anyone arousing suspicion. Prior to the recent shequ reforms that sought to improve their working conditions, committee directors and members were meagrely compensated, if at all. In essence, they held what were considered “the most menial positions of general leadership in urban China” (Whyte and Parish 1984, 212). During the Mao era, the workplace mattered more in the lives of Chinese urbanites than the residential neighbourhood (Lu and Perry 1997). In cities during the early period of Communist rule under Mao, residents’ committees supplemented the all-providing work unit structure (danwei). Socialist ideology had encouraged the right to welfare through work, and the state provided subsidies through work units rather than direct welfare programs. Every able-bodied urban worker was assigned a job at a work unit that provided its workers with employment, housing, goods, and welfare services and also managed several aspects of their lives, from family planning to the education and job allocation of their children. Endeavouring to exemplify self-sufficiency, large work units such as factories, hospitals, and universities included workplaces, housing, canteens, schools, and health clinics within the walls of the danwei compound. As chapter 3 will discuss further, residents’ committees staffed by retirees and housewives acted as an auxiliary, serving those who lived outside the danwei compound in the city’s traditional quarters and those who, for various reasons, were not assigned to a danwei, such as the unemployable and the handicapped.  9 Shequ Construction Research The work unit’s diminishing centrality in the life of urbanites and the need to strengthen the capacity of residents’ committees provide the context in which the shequ discourse is situated. This section reviews the discussions in the recently expanding shequ literature on the shift to a shequ-oriented social life and the objectives of the shequ policy. It also examines the implication that the broader processes of urban development have for the study of shequ. Danwei and shequ Shequ Construction is typically framed against the backdrop of the earlier danwei-based society. When people I met in Nanjing learned that I was researching Chinese community, many proceeded to share their memories of danwei life. The social life within the work unit compound is what people remember most; for them it represents their quintessential idea of community. The lao san jie7 – the generation that came of age during the Cultural Revolution – who had grown up living by Mao’s ideologies and many of whom were laid off in the market economy, describe this socialist past as a simpler time with nostalgia. One’s dependency on the work unit and the social relationships within it made the danwei one’s community.8 The danwei institution that had dominated urban social life became increasingly difficult to maintain in the post-Mao market economy that emphasized productivity, competition, and efficiency. The “breaking of the iron rice bowl,” as danwei reform is referred to colloquially, was by no means instantaneous. Throughout the 1980s, the danwei-based welfare institution remained basically unchanged (Gu 2001). Moreover, managers sought to protect workers’ benefits and circumvented disciplinary procedures that were being implemented under new “scientific” management practices to increase productivity. The longterm interpersonal relationships between managers and workers, who were colleagues as well as neighbours, undermined the modern individual-oriented labour practices (Bray 2005, 164). However, after over a decade of various initiatives to restructure the planned economic system, losses  7 This term translates literally as “the three old classes” and refers to those born in the years 1954, 1955, and 1956. This cohort, singled out for the Cultural Revolution’s impact on it, was in middle school when the political and social upheaval began. With their education interrupted, many could not participate in the post-reform economy that emphasized skills and credentials. 8 Walder (1986, 13) termed this “organized dependence,” where society was made stable and governable through extraordinary job security and benefits in exchange for compliance and acceptance of the system.  10 in the state sector created great pressure for enterprise restructuring. Policy shifts aimed at increasing the efficiencies of state-owned enterprises in the early 1990s eventually severed the cradle-to-grave relationship between workers and their danwei, after which became contractual. For the first time, managers could dismiss workers and hire through a competitive process. Within a span of five years, from 1997 to 2001, more than 25 million danwei jobs were lost.9 And, as chapter 4 chronicles, this restructuring of employment relations was swiftly followed by housing and welfare reforms to unburden work units of their social service responsibilities. The danwei, with its social welfare layers peeled off, now refers simply to one’s place of employment. More specifically, the term is used chiefly by workers in the state sector. As opposed to a universal welfare system, welfare had only been provided through the grassroots unit to which one belongs, be it the danwei, villagers’ committee, or residents’ committee. Following reform policies that permitted failing state-owned enterprises to dismiss workers and declare bankruptcy, the number of those relying on state welfare increased with the rise in unemployment. Local governments became responsible for redundant workers who had exhausted their unemployment benefits and no longer qualified for social relief from state agencies (Wong 2001, 47). And, in their limited capacity as agents for the local state, residents’ committees were expected to shoulder some of the responsibilities. However, the old residents’ committees lacked the skills and resources both to handle the increasing demand for welfare services and to ease the fears and feelings of uncertainty that threatened social stability. The current shequ movement is situated within this context of the bygone danwei-based society, and is part of the call for urbanites to reorient themselves from a “danwei person” to a “person of society” (danwei ren zhuanxiang shehui ren) relying on self, family, and neighbourhood. Nevertheless, the collectivist governmental logic of the danwei system persists in the urban structure and spatial practice (Bray 2005, 166). Recent danwei research draws attention to the construction of shequ as an alternative form of collectivization in the post-Mao era, where feelings of togetherness, belonging, and dependency are to be formed in places of residence instead of the workplace (Bray 2005, 181-90; Lu 2006, 20; Hurst 2009,  9 Figures vary between sources. This estimate is based on official figures and its calculation is discussed in chapter 4.  11 135-37). Taking up where the danwei research leaves off, shequ studies begin against the backdrop of revitalizing the residents’ committees so that place-based communities could manage the social service provisions formerly provided through work units. This is particularly evident in case studies of shequ reform in industrial regions that were heavily impacted by the closures of state-owned factories and plants, such as China’s northern rust belt (Tian and Qi 2005). Shequ policy objectives and content The shequ policy is premised on the principle that residents’ committees can and should play an important role in urban governance and be part of the solution to the unprecedented social challenges accompanying the country’s transition to a market economy, in particular the swelling unemployment, escalating demand for welfare provisions, and significant rural to urban migration. Setting the basis for the ensuing initiatives around neighbourhood reform, in December 1989 the State Council passed the Law on the Organization of the Urban Residents’ Committees (hereafter 1989 Organic Law), recognizing the committees as managers of public affairs and providers of services rather than merely as keepers of social order (Choate 1998, 11). Throughout the 1990s, as chapter 4 will discuss at greater length, various municipal and district governments across the country experimented with ways to approach neighbourhood reform given each locality’s socioeconomic conditions and fiscal resources. The experiments culminated in several models that used the shequ structure in a range of governance areas to fill gaps in the administrative bureaucracy (Benewick and Takahara 2002; He 2003; Derleth and Koldyk 2004; Benewick, Tong, and Howell 2004). Each local model differed, for instance, in how the shequ jurisdiction was determined, what types of services and functions were delivered through the neighbourhood unit, and how much decision-making power and responsibility the residents’ committee had. Elements and lessons learned from the experiments became incorporated into the Document 23, greatly influencing the paths other cities took in implementing their own shequ reform. The 2000 policy document effectively defines what a shequ is supposed to be and what Shequ Construction entails. The Ministry of Civil Affairs defines a shequ as “the collective social body formed by those living within a defined geographic boundary” (MCA 2000).  12 Examined in depth in chapter 4, the so-called construction movement speaks to the building up of three areas. The first is with regard to neighbourhood-based social services (Chan 1993; Wong 1998, chapter 6). Document 23 specifically mentions that “with enterprises shedding social functions and the state transferring out service functions, urban shequ will be required to assume the majority of [these functions]” (MCA 2000). Depending on the needs in their local context, residents’ committees are typically engaged in three types of services: 1) free services to the elderly, the poor, the young, and the disabled; 2) cultural and recreational programs; and 3) convenience services such as the operation of corner stores, bicycle stands, canteens, and newspaper stalls to provide everyday services closer to home as well as to generate revenue to support administrative and activity costs (L. Wong 1998, 128). Second, Shequ Construction has to do with enhancing the administrative authority of residents’ committees in realizing the reform-era concept of “small government, big society.”10 The repeated reference to “self-governing” (zizhi) in Document 23 does not pertain to political independence but to self-organization for working collaboratively with government agencies in the delivery of social services and the resolution of local concerns. The downloading of administrative responsibilities and service provision by the state onto neighbourhood units is not a form of political decentralization. Its primary purpose is to push communities to create their own service provision network, thus lessening their dependence on the government (Shieh and Friedmann 2008). A dominant line of questioning in shequ research has been on the appropriate role that the shequ institution should have in urban governance. The literature, encompassing various disciplinary lenses, can largely be divided into two schools of thought. In one strand of shequ research, residents’ committees are assumed to be necessary to carry out important government functions that, given the country’s systemic restructuring, are not fulfilled by any other agency (Wang 2003; Xu 2005; Pan et al. 2006). The emphasis of these studies is on seeking improvements within the existing bureaucratic structure, such as clarifying the legal standing and the functions of residents’ committees; defining the relationship between the neighbourhood Party branch and  10 The concept of “small government, big society” (xiao zhengfu da shehui) first originated in reports written to guide Hainan Province’s experiment with administrative reform. Liao Xun, one of the principle authors, envisioned a reduction of government and Party organs and an enlargement of the role of social organizations. The concept received national attention following the ninth National People’s Congress to launch an institutional reform campaign in 1998 that aimed at streamlining the bureaucracy, redefining core functions of the state, and shedding functions that should be taken care of by society (Brodsgaard 2009, 84).  13 the residents’ committee; and institutionalizing elections of residents’ committee members who have in the past been recruited or appointed by street office officials. Taking a more critical stance toward the neighbourhood reform, others scholars contend that the measures taken to build the capacity of the residents’ committee are more politically driven to restore state legitimacy in the wake of massive SOE layoffs than to effectively deliver social services (Wong and Poon 2005; Yan and Gao 2007) and to regain social control than to offer meaningful expressions of mutual help (Chan 1993; Read 2000; Bray 2006). Rather than freeing residents’ committees from the grasp of government bureaus, the concern is that this new emphasis by all levels of government and the resources poured into their reform have institutionalized what is supposedly a grassroots organization. The third area, referred to as the construction of a “socialist spirit,” demonstrates that Shequ Construction has as much to do with building the grassroots capacity of the Chinese Communist Party (CCP) as with the functions of the residents’ committee (Kojima and Kokubun 2002, Bray 2006, 535). Among many new challenges, the Party recognizes the need for popular legitimacy to continue its rule. One of its many strategies in this regard has been to broaden its membership base and enliven existing local Party committees and branches (Shambaugh 2008, 135). With a neighbourhood Party branch providing the core leadership, Shequ Construction is linked to the efforts of Party Construction (dang jian). Under the planned economy, the majority of Party members belonged to the CCP branch organized within their state-owned work units. At present, an increasing number of Party members are rendered “homeless” as a result of state-owned enterprise bankruptcies, unemployment, mobility between workplaces, and growth in private enterprises which do not have to establish a branch. For the central committee, the shequ structure presents a way to support the growing number of branchless members as well as to grow its membership and elevate the influence of Party leadership in neighbourhoods (Kojima and Kokubun 2002). Despite their different approaches toward the study and assessment of the policy content, the current shequ research has most importantly opened up a broader discussion as to the various directions the policy should and could take. As such, Shequ Construction remains open ended and an unfinished project.  14 Urban development and shequ Much of what is known about Shequ Construction comes from the analysis of policy content and what has been actualized on the ground in neighbourhoods across the country. As the policy programming becomes more concrete, concerns over outcome have led government agencies to sponsor research for developing performance evaluations to gauge impacts (discussed in chapter 5). In evaluating outcomes, an under-researched area is analyzing the policy at work. By this I mean examining the policy and implementation not in isolation but within the messy contingencies of urban China, asking: How and in what ways does the promulgation of Shequ Construction shift ideas and courses of action in other policy areas? How does the shequ institution facilitate or hinder the ability of other initiatives to respond to pressing social needs? In other words, if the construction of neighbourhood units has been a necessary part of urban restructuring, we still do not know how it has articulated with broader processes of the country’s urban transition, such as the urbanization of the countryside and the growing involvement of market actors. Jennifer Robinson (2004) in Ordinary Cities makes the compelling argument that in studying fragments of social life and localized concentrations of flows and networks, urban studies research has made cities extraordinary by singling out successes and failures of urban development projects. For instance, the global cities approach focuses on a city’s linkage to primarily economic networks, and the developmentalist approach emphasizes a city’s poorest elements (10). To reverse the insular attention on these exceptional fragments, she argues, is to bring the city back into view – to consider the city as integrated systems. Thus, rather than focusing on policy initiatives in isolation, be it attracting foreign investment or alleviating poverty, one must think across elements. The ordinary cities approach has informed my shequ research in two ways. First, it brings attention to how policy analyses often examine interventions in isolation, focusing on specific parts of the city that development programs were intended to address. Second, it lays emphasis on how certain cities have been made extraordinary by their successful shequ experiments, such that shequ models are informally named after the city in which they were pioneered. What would an analytical approach to Shequ Construction look like if the city is brought back into view? Even though shequ units are being constructed as the base-level governing institution in cities nationwide, relatively little is known about how they engage  15 with other processes occurring in the city and across elements of urban life involving the state, civic, and market sectors, and in the urban core as well as at the periphery. This dissertation takes as its starting point that all shequ are ordinary – that is, following the ordinary cities thesis, they are all diverse and complex. The local Shequ Construction implementations are thus shaped by processes beyond the policy and the four walls of the shequ compound. Adopting this perspective, the dissertation focuses on the interconnections between the shequ policy and other urban processes simultaneously unfolding and thus creating the dynamics of the shequ. Policy Implementation in China In examining shequ policy at work in articulation with other initiatives, this research engages the broader discourse of policy making in China. Observing the interaction among policies is important in the Chinese context because the Chinese state is seeking sweeping reforms in several areas simultaneously to reorient the world’s most populous country toward greater integration in the global economy (Lampton 1987, 11). Moreover, as I discuss below, China’s distinct policy-making process and gradualist approach to reform make examining policy interactions a particularly appropriate approach for analyzing policy implementation. To begin, the Chinese state limits access to the policy formulation process, which often occurs behind closed doors, and governing elites do not always disclose the principal objectives of policies (Lampton 1987, 5-6). For instance, it is difficult to discern whether the purpose of instituting local elections is to establish democratic procedures, as the mass line asserts, or to remove conservative cadres who oppose reform policies from their posts (ibid., 7). Thus, measuring implementation success through the congruence of declared intentions and outcomes can lead to multiple interpretations. In addition, China’s policy process is protracted, opening the way at various points for revisions as policies confront ongoing societal developments and agendas. This point is made by Lieberthal and Oksenberg (1990) in an early study of China’s policy-making process in the post-Mao era. The authors assert that it is unrealistic to view a policy as a single decision on a major issue. In actuality, a series of decisions occurs and important differences exist among them: the policy’s initial formulation among leaders; the initiative’s announcement; allocation of funds; and development of construction schedules, regulations, and concrete  16 measures to make the initial decision produce the desired outcomes. For a policy to move through each of these steps in the drawn-out policy-making process requires negotiation, bargaining, and consensus seeking among 28 ministries and commissions and among central ministries and their local bureaus (26). As such, the process tends to produce directives that are ridden with ambiguity and contradiction and are prone to unintended consequences in implementation by local officials. This structural ambiguity has been characterized by scholars in various ways. Heilmann (2008) argues that the discrepancies between policy intentions and outcomes are not necessarily counterproductive. They are expected in China’s experiment-based policy- making process. The author observes that conventional understandings of the policy process, particularly in liberal democracies, hold that policy analysis, formulation, and embodiment in legislation should precede implementation. However, in China’s transition from planned to market economy, which has relied heavily on policy experimentation to guide its restructuring, the reverse is true: innovations happen through implementation first and drafting of laws and regulations second (9). Despite a tendency for the process to become arbitrary, volatile, and vulnerable to the short-term interests of local elites, it functionally promotes a dynamic central-local interaction with a continuous interplay between local- condition-driven initiatives and central sponsorship (10). This, Heilmann maintains, is one of the distinctive characteristics of China’s policy-making process. From another perspective, recent studies of China’s land politics have shown that land development processes are dominated by unprecedented discretionary powers enjoyed by lower-level officials that often produce contradictions between the centre’s intention and local governments’ policy execution (Ho 2001; Lin and Ho 2005; Hsing 2006). These contradictions are outcomes of the regime’s gradualist approach to economic reform. Given the unknown destination of the country’s socialist transformation and for the system to work, rules of land development carry a certain level of “deliberate institutional ambiguity” and underdefined authority so that they can be enforced and revised depending on changing circumstances to which the state needs to react (Ho 2001). While permitting responsive institutional fixes, the system also gives rise to contestation, circumvention, and selective implementation by local governments, contributing to the inconclusive nature of state projects (Lin and Ho 2005, 414-5).  17 China’s institutional ambiguity has also been interpreted as a phenomenon of trapped transition (Pei 2006). Pei argues that the gradualist economic reform under an authoritarian regime may have allowed leaders to respond to respond to conditions and “grow out of the plan” (26), but gradualism has also allowed ruling elites to maintain their control in lucrative high-rent sectors (44). Consequently, the situation is one in which ruling elites have an interest in a semireformed system that favours economic liberalization but are less inclined to support political reform. In this context, policies remain subservient to power relations as well as party discipline. Instead of moving toward an even more open economy and society, the system may become trapped by decentralized state predation and deterioration of governance (43). These studies demonstrate that the Chinese policy process is protracted, can be constructively experimental and ambiguous, and is subjected to local discretions and predation. Rather than conceived of as an engineered blueprint for institutional change, shequ reform is better conceptualized as a series of interbureaucratic memoranda generated through local experiments, negotiations between agencies, and responses to contingencies. Understanding this policy-making context, an alternate approach to analyzing shequ reform, then, is to go beyond the plan and examine the effects of a policy at conjunctures. In such an approach, the focus is not on outcomes but on observing the bearing one policy has on another in the implementation process as a way of teasing out contradictions and unforeseen consequences (Li 2007, 28). This framework of conjunctures, as Tania Murray Li (2007) puts forth, intertwines analysis of governmental schemes with analysis of social histories, practices, and processes. Her book, The Will to Improve, examines interactions among policies in the Indonesian highland over a period of two centuries, underscoring the contradictions as one improvement scheme yielded to another. She argues that policies do not pursue one dogmatic goal. In most cases, there are both hidden motives of gaining domination, legitimacy, and profit, and positive intentions of mitigating harm and promoting beneficial development. What is crucial in examining programs’ effects is not to rush to identify hidden motives, but rather to consider outcomes as occurring at conjunctures where policy intersects with memories, cultural ideas, and longstanding practices and struggles on historically configured terrain. Therefore, policy  18 programs are not merely interventions drawn up by technocrats but part of multiple forces articulating together at a point in time and in a particular place (ibid., 28). This dissertation draws on this line of thinking that takes a more operative approach. I do not examine Shequ Construction by itself looking for congruence between intentions and outcomes but rather consider it at work in its interactions with other government programs, local initiatives, and informal practices. In breaking down the state-directed construction of neighbourhood communities to bridge gaps in control, this research places the policy in articulation with other reform projects. Each chapter explores a particular conjuncture, examining how shequ reform interacts with other initiatives unfolding in urban neighbourhoods. These encounters demonstrate how Shequ Construction responds to ever- changing conditions and contingencies, and is reworked by local officials in the implementation process. Research Questions: Shequ Construction in China’s Urbanization Based on fieldwork undertaken in Nanjing, Jiangsu Province, this dissertation examines Shequ Construction at four crossroads, each involving multiple agencies and each allowing a different perspective through which to view the central directive. Fiscal reform The first policy I examine is fiscal reform and the division of responsibilities between subnational levels of government. I question the impact fiscal decentralization has on the capacity of urban district governments to experiment with and implement social programs and welfare provisions. Local government (difang zhengfu), in the Chinese usage of the term, refers to all subnational levels, from provinces and municipalities down to rural townships and urban street offices. Studies of lower levels of government have predominantly been in the rural context, with a particular focus on the critical role of the county government and villages in bringing about rural industrialization (Oi 1995; Walder 1995; Blecher and Shue 1996) and improvements to social conditions and quality of life (Guldin 1997). Urban studies have drawn attention to the economic decision-making autonomy of municipalities (Yeh and Wu 1996; Chung 1999; Ho 2001; Ho and Lin 2003; Lin 2007; Hsing 2010). Under decentralization measures, regulatory authority over urban land rests with municipal governments. These studies have examined the strategies municipalities have taken to guide  19 land and infrastructure projects to attract investments as land-lease sales and rents have become the most important sources of fiscal revenue. F. Wu (2002) observes that less attention has been given to lower levels of city governments, namely the district and the street offices, as local agents of development, not to mention social change. In the context of Shequ Construction implementation, urban districts and street offices, linked to a more complex hierarchy and hence less independent than their rural counterparts, have nevertheless been pivotal agencies in determining the outcome on the ground. Shequ Construction has become a major project for district governments. I question the ways in which the implementation of the policy program also seeks to build the district capacity and authority. Urban village redevelopment Over a four-year period from 2005 to 2009, Nanjing municipal government planned to requisition 71 villages on the city’s edge. Discussions of village redevelopment have been dominated by the politics of land policies (Yeh and Wu 1996; Xie, Parsa, and Redding 2002; Lin and Ho 2005). However, the redevelopment process also entails the relocation and social integration of erstwhile villagers, and thus is as much a social phenomenon as a land use issue (Leaf 2007). Shequ Construction has a significant but overlooked role in facilitating the state’s efforts to transform villages and assimilate villagers. First, the policy has both positive and negative impacts on how villagers adapt to the urban way of life in their reassigned neighbourhoods. Second, a recent initiative under experiment in Nanjing is to extend Shequ Construction from urban neighbourhoods to rural villages, beginning with those in the city’s immediate hinterland. As these sites undergo community construction, the shequ policy’s articulation with urban village redevelopment initiatives raises many as yet unasked questions of its rationale and appropriateness as a means of bringing about social order. Nonprofit sector in community-based service delivery The explosion in the number of social organizations is a significant social outcome accompanying China’s urban transition. The third policy crossroads this research examines is how the legal recognition of nonprofit social service organizations fits into the Shequ Construction scheme, particularly in view of the fact that service delivery is an integral component of the policy program. At their juncture in the neighbourhood social space, both shared and conflicting interests exist between nonprofit organizations and the residents’  20 committees. Despite the recognition of nonprofits and support for them from higher levels of government, their relationship with the shequ institution remains undefined. A major thread within the literature on China’s urban social organization centres on their vertical relationship with the state at the top, assessing their degree of autonomy from the state and the implications for the development of a civil sphere (Saich 2000; Ma 2006; Lu 2008; Zheng and Fewsmith 2008). While the number and types of social organizations have increased in recent years, the lateral relationship among organizations has received relatively less attention. The interaction between nonprofit organizations and residents’ committees offers insights into the competition and collaboration among neighbourhood-level institutions. Interest-based community The last conjuncture I examine is the meeting of Shequ Construction, as the formal representation of a neighbourhood community, with the informal production of “community” as represented by homeowners’ associations. The opposition between the modern states’ ordering of space and its contestation by social practices has spurred a great deal of interdisciplinary research and theoretical debate on power and place making (Lefebvre 1991; Massey 1994; Scott 1998). Rather than using a discourse of domination and resistance, my examination of the interaction between residents’ committees and homeowners’ associations seeks to make sense of their co-existence. Economic reform measures, together with enlarging social spaces, have given rise to the formation of interest-based communities within the place-based administrative shequ. In this case, housing privatization has led middle-class homeowners to organize to protect their property interests (Read 2003; Cai 2005). While homeowners do not oppose the demarcation and assignment of shequ jurisdictions for all urbanites, homeowners’ associations represent an alternative form of grassroots governing organization that is being established. Questions remain as to how, with the implementation of Shequ Construction, these definitions of community have been rendered by various levels of government, and how homeowners’ associations challenge or assist the work of residents’ committees in responding to the diversity of interests and needs. Dissertation Outline Chapter 2 discusses the research design, the data collection process, and methodological issues. To situate the current study in a historical continuum, chapter 3 is a historical overview of urban community in China with a focus on the Jiangnan region (lower Yangtze  21 delta) from the late imperial era to the end of the Cultural Revolution. I illustrate that the concept of community has its origins in Chinese traditions of self-management, social control and welfare provision, rather than as a sociological concept. Chapter 4 begins in the early 1990s, at the height of danwei restructuring and consequent layoffs. It presents an overview of the socioeconomic conditions at the time and the pressures on the Party-state to take actions by experimenting with neighbourhood reform. I discuss the improvements and interventions sought through an in-depth examination of the contents of Document 23. Whereas chapter 4 focuses on concerns at the national level, chapter 5 presents the perspective from the district level. With Shequ Construction efforts dependent on district- level financing and leadership, it examines the impact of fiscal decentralization on Shequ Construction implementation through the experiences of Nanjing’s Gulou District. In the three chapters that follow, I turn to the crossroads at which the policy encounters initiatives at the neighbourhood level, drawing attention to the divergent interests of various local actors who are involved. Chapter 6 examines the articulation of the shequ policy with municipal village redevelopment plans. Through the case study of a community seniors’ centre established by a local social enterprise, chapter 7 examines the involvement of the nonprofit sector as another agent in the neighbourhood arena created by the state’s welfare socialization agenda. Chapter 8 examines the working relationship between the homeowners’ association and the residents’ committee and questions the rationale behind the recent government initiatives to incorporate the former into the shequ institution. The concluding chapter discusses the theoretical issues that have emerged from this research. First, I reflect on the analytical approach of policy conjunctures and the new insights the framework has shed on the nature of Shequ Construction as an urban governance project and the unique policy-making process embedded in China’s interbureaucratic document system. Second, bringing together the individual case studies, I consider the manner in which Shequ Construction has been deployed by local officials in each of the local initiatives. I argue that while in implementation the policy is subjected to local discretion, it has nevertheless been consistently construed to harmonize diverging interests, aligning the interests of villagers,  22 nonprofit leaders, and homeowners with those of the Party-state. Third, I discuss two issues that have remained with me throughout the research: the policy’s disconnection with migrants’ welfare and the debate on the neoliberalization of Chinese forms of governance. Last, I conclude with some thoughts on the impact of Shequ Construction on Chinese planning practice and research.  23 2. RESEARCH DESIGN Methodology, Fieldwork Process, and Methodological Issues  Introduction This dissertation is based on findings from fieldwork conducted in Nanjing, Jiangsu Province between February 2007 and January 2008. Prior to fieldwork, I spent a year as a student at the Hopkins-Nanjing Center which greatly informed my research and in some ways could also be considered part of fieldwork. The intent of the year was to gain a deeper understanding of China’s social issues as they are problematized and discussed by Chinese scholars. I took classes taught by Chinese professors at the Center with other foreign students as well as audited graduate seminar classes with Chinese students in the sociology department. During my year there, under the supervision of a Nanjing University professor, I undertook an independent study of the city’s urban village redevelopment. My focus at the time was to understand how the urban village phenomenon emerged in Nanjing and to contrast it with that I observed in Quanzhou, Fujian Province where I had conducted field research two years prior. As part of this study, I conducted some interviews and participant observations in Rivertown Village, which became one of my dissertation research fieldsites. I also visited three villages that have developed profitable village enterprises in the city’s two counties, exhibiting a contrasting, in situ form of urbanization (Zhu 2000). In addition to the preliminary groundwork and building fundamental skills for navigating Chinese sources and independently managing field research, the year informed my dissertation prospectus in another significant way. I came to appreciate that contending with cultural difference involves not only dealing with differing social norms, but also confronting institutional assumptions, such as the different ways governance is understood and community planning is practiced. Building on these experiences, the research approach was guided by what I had identified as two important issues for conducting this community research cross-culturally. The first was the challenge of understanding Shequ Construction from the ground-up through local agents’ interpretation and implementation of the policy. The second was the challenge of establishing an analytical frame to examine policy within  24 the Chinese context. I developed a phased fieldwork approach that allowed immersion and flexibility to progressively focus the research. This chapter presents a detailed account of my research process and methodological issues I encountered. It begins by discussing my research design and methodological considerations. Then, it details my field research, specifically how I accessed and gathered data. The concluding section reflects on conditions of fieldwork and issues of positionality and bias. Overview of Research Design The objective of my study is to better understand the impact of the shequ policy program through its interactions on other policy initiatives. The concept of policy conjunctures does not take an evaluative approach towards Shequ Construction. It permits policy to be analyzed within local contexts and understood through the ways it adapts and responds to arising circumstances. The advantage of the framework is that the outcomes are analyzed within local conditions. Furthermore, as decision makers have been evaluating the policy through attainment of prescribed standards based on policy intentions, questioning how policies interact provides a new lens through which to view Shequ Construction. Through four policy conjunctures, this research endeavours to understand how the central directive to construct neighbourhood units becomes interpreted by local agencies in the course of implementation and integrated in various areas of urban governance. Conversely, by articulating Shequ Construction with a diverse set of policies, my research also investigates how the shequ discourse alters the content and the implementation course of other policies. It leaves open the possibility that in the interaction with other policies Shequ Construction has different effects than those intended. The research design can be classified as an embedded case study (Yin 2003, 42-43). Different from holistic case studies that examine individual communities as a whole, the analysis is focused on the interactions of policies within the larger community units. This research specifically examines Shequ Construction through its articulation with four policy initiatives in different communities. At the first policy conjuncture, Shequ Construction is understood through fiscal decentralization. I chose one of Nanjing’s districts as the larger case within which I would question the role of the district government in formulating implementation plans given China’s decentralized welfare system. The second conjuncture places Shequ Construction in the city’s urban village redevelopment process. The field sites are two  25 villages undergoing redevelopment and an urban core shequ dealing with the integration of landless villagers. The third conjuncture examines Shequ Construction and the official recognition of local service-providing nongovernmental organizations. The field sites are an organization engaged in shequ-based elder care and the shequ in which it is operating. The fourth conjuncture brings housing reform and the emergence of homeowners’ association to bear on the implementation of Shequ Construction. The field is a neighbourhood with an active homeowners’ association and in a district experimenting with integrating homeowners’ associations into the shequ governing structure. See table 2.1. I have chosen to situate the study in Nanjing because its districts and neighbourhoods were, and continue to be, sites of Shequ Construction policy experiments. The capital of Jiangsu Province, the city sits at the lower reaches of the Yangtze River about 300 kilometres west of Shanghai (figure 2.1). Under the jurisdiction of the Nanjing Municipal Government are 11 districts and 2 counties. At the grassroots level, this city of 7 million is organized into 799 neighbourhood residents’ committees and 587 villagers’ committees (2008 figures; Nanjing Statistical Yearbook 2009, table 1-1, 3-1). The old capital has stayed out of the research limelight on reform era China. Urban studies have primarily focused on Beijing and the coastal cities of tremendous economic growth like Shanghai, Shenzhen, and Guangzhou. Connected to Shanghai by highway and express trains, Nanjing sits in the shadow of the nation’s financial growth pole. While it has benefited from economic reform, it is not the main target of the central state’s preferential economic policies. Similarly, while not as hard hit by the restructuring of state-owned factories as northern cities in China’s rust belt, as one of the country’s major industrial cities, it also faces the challenges of unemployment and loss of social welfare provisions. Moreover, geographically sitting on the north-south divide, Nanjing responds to and reflects the enterprise and openness of Shanghai and Sunan cities (southern Jiangsu) that are linked to global capital flows, and the conservatism of the more insular and impoverished Subei cities (northern Jiangsu) and central provinces of Jiangxi and Anhui. In these contexts, Nanjing shares the experience of many Chinese cities struggling to formulate working solutions and to balance new financial burdens in the provision of social services.    26  Figure 2.1 Geographical location of Nanjing The capital of Jiangsu Province, Nanjing sits at the lower reaches of the Yangtze River, about 300 kilometers west of Shanghai.  ●  Wuhan ●  Shanghai ●  Nanjing ●  Chengdu ●  Hong Kong ●  Chongqing ●  Beijing  27  Table 2.1 Overview of Research Design Conjuncture Research Questions Field Site Sources of Data Fiscal decentralization What are the impacts of fiscal decentralization on Shequ Construction implementation by district governments? Nanjing Gulou District open-ended interviews; policy documents Urban village redevelopment How has Shequ Construction been interpreted by local agents to facilitate the redevelopment process and the integration of villagers into the city? 2 urban villages undergoing redevelopment (Rivertown and Willow); a shequ where some of the city’s villagers have been relocated (White Blossom) open-ended interviews; participant observations; policy documents Nonprofit sector development How has the official recognition of nonprofit organizations been incorporated into Shequ Construction? What roles do local nonprofit organizations have in shequ-based social service provisioning? a shequ with nonprofit providing social services (Nanjing New Village); a nonprofit organization (Sunrise) open-ended interviews; participant observations; policy documents Housing reform and emergence of homeowners’ associations What part does homeowners’ association play in Shequ Construction? How are homeowners’ associations being incorporated into the shequ institution? shequ with an active homeowner’ association (White Blossom) open-ended interviews; participant observations; survey; policy documents    28 Selection of Methods In recent years, critical policy analyses in the field of urban planning have opened a discussion on the limits of empiricist methods in planning research and have drawn attention to process as the subject of analysis. Planning research has gradually shifted away from a one-sided emphasis on objectivity and statistically valid sample size. Scholars in the field have advocated for broadening the language of planning to include, in addition to technical expert knowledge, other ways of knowing, such as local knowledge gained from talking, listening, and sharing with members of localities in which planners work (Sandercock 1998; 2005). Communicative planning theory has directed inquiries toward studying subtle and invisible power relations and values that shape policy making and the effects of planning practices on the ordinary, everyday world (Forester 1989; Innes 1995; Healey 1996). These studies demonstrate how interpretive approaches bring to light an alternative, complex understanding of municipal policymaking. Particularly influential on my thinking has been the ideas of “transactive planning” (Friedmann 1973; 1987) and the “ethnographic present” (Holston 1998). Instrumental to the planning field’s epistemological shift, transactive planning emphasizes mutual learning between actors and planners through dialogue and engaging in action itself (Friedmann 1987, 402). Anthropologist James Holston (1998) critiques modernist planning as having failed to recognize the conflicts and paradoxes by transforming the unwanted present by means of an imagined future. He writes of an insurgent urbanism that introduces into the city “new identities and practices that disturb established histories” (48) and “reveals a realm of the possible that is rooted in the heterogeneity of lived experiences, which is to say, in the ethnographic present and not in utopian future” (53). Linked with transactive planning and applied to planning research, the concept of the ethnographic present is not advocating that planners become ethnographers. Rather, it is about attention to lived experiences and modes of sense-making and problem-solving on the ground as people confront pressing realities. More than struggles and resistances, the ethnographic present privileges the multiple ways, both formal and informal, in which things get done. Importantly, it is about sharing the present world with participants, as opposed to studying from above, surveying with a bird’s eye view.  29 While the research question determines the research method, the reverse is also true: the type of data sought and the ways of gathering data also critically influence the types of questions asked (Pader 2006, 173). My research design adopted primarily methods of in-depth interviews and participant observations. Behind these techniques of qualitative research was the intent to achieve some level of understanding through the fieldwork approach – through immersion and personal involvement in the ongoing social activities for the purpose of research (Wolcott 1995, 66). With the intention of capturing the ethnographic present, I planned an extended period of fieldwork so that the research would be open to unanticipated events and be driven by findings and bottom-up analysis. The research was iterative rather than step-wise, responding to analyses of findings on the ground rather than pre-established protocol. This is not to say that the research was not systematic, but that there was a back-and-forth between data collection and analysis. Insights, unanticipated situations, and fieldwork realties informed new questions and new ways of conceptualizing field findings. When new issues became apparent, I reworked the research so that it became progressively focused (Stake 1995, 9). I sought to adopt the ethnographic sensibility of accepting new questions and unknown situations that could never have been conceived of before looking (Pader 2006, 174). The result of this was a phased and iterative process where new understandings and discoveries continued to shape and focus the research question and design. The 12 months of fieldwork can be divided into an initial fieldwork period on urban village redevelopment and a refocused fieldwork period on Shequ Construction policy interactions. Table 2.2 summarizes the main objectives and outcomes of each phase. Initial fieldwork The research began as a peri-urban village study that sought to analyze how the city incorporates communities on its fringe and how villagers anticipate and negotiated their urbanization. The impact of Shequ Construction as an integrative mechanism was a facet of the redevelopment process I was investigating. However, on the ground, the policy played a larger and more interesting role than I had expected. In one of the villages I was studying, its villagers’ committee coexisted with a residents’ committee, raising questions for how the two entities collaborated and whether this was a transitional or permanent solution. In interviews with district-level civil affairs bureau officials, the policy program was repeatedly mentioned.  30 Sometimes this was done to contrast the “model” shequ with what urban villages were not and sometimes, I suspected, to change topics to one that they were more interested in. I had also forayed into a resettlement housing project where villagers were relocated after land requisition. In my conversations with the shequ director, I was shown local initiatives of Shequ Construction but they seemed to have little to do with integrating the villagers into neighbourhood life. What I found particularly interesting was the fact that Shequ Construction’s application to urban village redevelopment was the specific outcome of local circumstances, beyond the intention of the policy program designed by policymakers in Beijing. The context of redevelopment drew attention to the ways the shequ policy program was being implemented to alter social practices in villages to better integrate them with the city. In the increasingly diverse Chinese society, exhibited through sociospatial segregation, what particular meaning did Shequ Construction hold in other types of neighbourhood? Just as the residents’ committee had to confront the existence of villagers’ committee in urban villages, what other institutions challenged or collaborated with the residents’ committees in constructing governable neighbourhoods? With these new questions, I returned home from fieldwork for a month. Taking what I had observed of Shequ Construction on urban village redevelopment, I refocused the research on the implementation of the shequ policy in various neighbourhood types, of which the urban village would be one type. I returned to Nanjing for another five months. The working proposition was that shequ typology determined Shequ Construction implementation and I had planned to select two additional shequ of different types to examine the policy’s interpretation in each context. Refocused fieldwork The subsequent shift in research frame from shequ typologies to policy conjunctures came as a result of further research-site realities. I began with fourteen shequ of various typologies – six danwei compounds, four market housing projects, three urban villages, and a resettlement housing project. Each neighbourhood became a particular site – each had its own characteristics, history, and resident composition, and was guided by a director with his or her own leadership style. In further selecting particular neighbourhoods to focus on as case studies, I had to determine which could help me better understand the nature of Shequ Construction and issues of its implementation. I was confounded with choosing between  31 neighbourhoods of the same type – for example, what made one danwei-type shequ more worthy as a case than another. In reflecting on the implementation of the policy program in different neighbourhood types, the important realization was that while there were differences, what stood out was the important ways Shequ Construction mattered only when it engaged with other neighbourhood projects. For instance, what I found interesting in one danwei-type shequ of mainly retired workers was how a nonprofit organization was engaged to provide elder care. A significant insight that came out of this inquiry was that social service provisioning, which was at the heart of Shequ Construction, was simultaneous being realized by state initiatives that fostered the development of local service-providing nonprofit organizations. These initial reflections led me to refocus the research on points of policy conjunctures as opposed to cases of neighbourhood type. This interpretive framework was arrived at through rethinking what was to be learned through the cases. My research was not guided by an intrinsic interest in a particular neighbourhood. Rather, the neighbourhoods provided the context in which to examine policy impact. Making this distinction between intrinsic and instrumental interest in cases established more strategic criteria for site selection (Stake 1994, 237-8). My basis for choosing neighbourhoods was the presence of other local initiatives that complicated neighbourhood governance and so thus could advance my understanding of Shequ Construction’s impact on the ground. Using this interpretive framework, the research sought to draw attention to the meanings and issues local agents have attached to Shequ Construction. The central question became, not what Shequ Construction has achieved, but the ways in which the policy is utilized in addressing problems that matter to local communities. This shifted and clarified the research design, now focused on how arising governance issues are dealt with by local agents through Shequ Construction. It broadened the examination of the policy beyond solely being the domain of residents’ committees and included the role of other neighbourhood-based social organizations such as villagers’ committees, nonprofit organizations, and homeowners’ associations.  32  Table 2.2 Phases of Research Phases of Research Objectives Outcome  1. Prior to fieldwork Hopkins-Nanjing Centre (1 year certificate program)  To gain deeper understanding of social issues as they are discussed by Chinese scholars To learn skills for conducting field research in China  Took social science classes Conducted independent study on Nanjing’s urban village redevelopment Interviewed planning bureau officials on land development process and redevelopment plans Visited different urban villages Conducted interviews and participant observations in Rivertown Village Taught at migrant school   2. Initial Fieldwork: Urban Village Redevelopment February 2007 to July 2007  To examine the urban village redevelopment policy and its implementation on the ground To determine potential urban villages as case study sites   Interviewed Nanjing University professors on land development process Interviewed planning bureau officials on land development process and redevelopment plans Conducted library research on Nanjing’s urban expansion and urban village redevelopment Began researching Nanjing’s experience with Shequ Construction from documents and discussion with Nanjing University and Nanjing Normal University professors Observed impact of Shequ Construction in village redevelopment Revisited Rivertown Village to conduct interviews and participant observations Conducted interviews and participant observations in Willow Village Taught at migrant school   3. Revise Research August 2007  To revise research plan  Refocused research on Shequ Construction implementation in various shequ types Amended research ethics application   4. Refocused Fieldwork: Policy Interactions September 2007 to January 2008  To continue with observation of urban village redevelopment through Shequ Construction lens To determine additional case study sites according to revised research  Interviewed civil affairs bureaus Surveyed various types of urban shequ Interviewed shequ researchers Conducted interviews and participant observations in Nanjing New Village and White Blossom New Village Conducted library research on Shequ Construction Taught at migrant school Worked with Sunrise Senior Care Services  33 Access and Sampling During the fieldwork period, I was affiliated with Nanjing University as a visiting scholar. This work unit (danwei) affiliation gave me access to its libraries. Furthermore, when introducing myself as a researcher to officials and communities, a local affiliation lent credibility and legitimacy. As those who have lived in China know, the first questions asked when meeting someone for the first time is where (i.e. which danwei) are you from and who (i.e. which danwei) sent you. In addition to having the affiliation for research logistics, I found that access depended on making contacts and following leads and being part of the university facilitated meeting people with similar research interests. My strategy was to visit and construct preliminary profiles of as many shequ and urban villages as I could before determining which would be appropriate for in-depth studies. Professors I met were instrumental in helping me make contacts with government officials and shequ directors. My entry into the majority of the shequ I surveyed was through three districts’ civil affairs bureaus. At the end of my interviews with civil affairs bureau officials, I inquired whether I could visit the shequ in their jurisdiction. They listed a few shequ that were, in their view, typical or demonstrated certain achievements. In one district, an intern accompanied me on initial visits; and in the other two districts I made calls to the shequ directors and interviewed them on my own. I also accompanied two university research teams on their visits to shequ. As they were not studying Shequ Construction, but conducting household surveys, I observed but did not interfere with my own questions. I saw how they conducted their field research and their interactions with their interviewees, such as their use of language and handwritten notes. After these visits, I returned on my own to talk to the shequ directors myself or to spend some time observing activities in public spaces. Entry into urban villages was more difficult as redevelopment was contentious and formal introductions were unlikely, especially for a foreign scholar. I began by visiting urban villages that had been in newspapers or in research papers I had read. I would introduce myself as a visiting scholar who wanted to follow up with what had been reported. I found that since they had been the subject of research attention, the villagers’ committee members were neither welcoming nor troubled by my presence. In addition, my volunteer work as an English teacher at schools for migrant children also took me to urban villages.  34 From the eleven shequ and three urban villages I surveyed, two shequ and two urban villages became sites for the in-depth study of policy interactions. Their selection was based on the presence of local initiatives impacting Shequ Construction implementation that I observed as well as the rapport I was developing with the community and with the nonprofits working within them. In the two shequ, I made repeated visits over roughly a four month period. They are both located in urban core districts. The first, which I call Nanjing New Village,11 is a typical danwei-type shequ where the housing was built by a work unit for its workers. The majority of the residents are the original occupants, many now retired. In recent years, the aging neighbourhood has seen an increase in newcomers as more residents rent or sell their apartments and move to live with their adult children. Here, I wanted to observe how Shequ Construction implementation and the emergence of nonprofit shequ-based social service organizations interacted. The second shequ, which I call White Blossom New Village, is one of the largest in Nanjing with over 6,000 households. Built in the early 1990s, it was one of the last few danwei-built welfare housing complexes. It is composed of residents from various work units and relocated villagers from one of the earlier redevelopment efforts. The director of White Blossom faced the challenge of meeting the needs of residents with diverse socioeconomic backgrounds and with varying expectations of neighbourhood management. The diversity of this shequ allowed me to examine two issues. First, housing privatization gave rise to the formation of homeowners’ association in China. One compound in this shequ had an active homeowners’ association and presented the opportunity to understand how, under Shequ Construction, the residents’ committee and the homeowners’ association would share governance responsibilities. Second, in another compound lived relocated villagers and the unsightliness of their vegetable gardens was a persistent issue. This raised the question of how standards of conduct, supported by Shequ Construction implementation plans, had impacted the social integration of the elderly erstwhile farmers into urban neighbourhoods.  11 The names of people and places are pseudonyms. For the urban shequ, I kept the suffix of “new village” (xincun) which was a common practice for naming urban residential housing compounds in Nanjing during the Maoist era (Interview, NJU history professor, 17 March 2007). I am not certain whether it was the practice elsewhere in China. I also could not find sources to suggest connotations or references to rural village or rural communal way of life.  35 I chose two urban villages at different stages in the redevelopment process. What I was observing in one helped make sense of what was happening in the other. The first, which I refer to as Willow Village, is a village on the edge of the urban core that has recently become an urban shequ. With a villagers’ committee co-exiting with a residents’ committee, it presented a unique case to examine the impact Shequ Construction has on the village’s redevelopment process. I was teaching in a school for migrant children in a village I called Rivertown. As I regularly visited, I continued to document the development process there. In contrast to Willow Village, a large portion of the collectively owned rural land had already been requisitioned. With the majority of the villagers relocated, rural migrants now occupy the dilapidated housing left behind. Its villagers’ committee will remain in place as the administrator until all of the villagers have received their nonagricultural hukou status and until the land negotiation process has concluded. Figure 2.2 locates the two shequ and two urban villages in Nanjing. I present the characteristics of these four communities in greater detail in the analysis chapters in which they are discussed.      Figure 2.2 Map of Nanjing Willow Rivert Nanjin White  Village own Village g New Villag  Blossom New e  Village Urban Core D Suburban Di Counties Railroad Main roads & District/Coun 36 istricts stricts  highways ty boundaries  37 Sources of Data and Collection Data for this study were collected from six main sources: interviews with municipal and district bureau officials, discussions with Chinese shequ policy researchers, site visits to villages and shequ, involvement in social organizations, a survey of homeowners, and documentary materials. In this section, I briefly discuss how data were collected from each of these sources. Generally, my fieldwork data were recorded in three ways. I had a notebook with my handwritten notes from interviews and observation. I had a computer file where I wrote up the interview notes into text that described the setting, what was asked, what was said, and my impressions. I had a second computer file which served as a field journal in which I would record discussions, observations, insights, and reflections from the day’s visit to neighbourhoods, involvement in activities, and informal conversations. I chose not to tape record my interviews for several reasons. From my past fieldwork experiences in China, I found that tape recorders made interviewees self-conscious; and the majority, especially those who were older, expressed concern over their necessity and consequence. When I accompanied Chinese students to their shequ visits, I observed that they also did not use a recorder but took handwritten notes. I further considered the fact that I was introduced to several shequ through civil affairs bureau. To shequ directors this might raise concern that what was said could be shared with higher level officials. Instead, I asked permission to take handwritten notes. After interviews and neighbourhood visits, I promptly detailed conversations and impressions. Therefore, there are not many direct quotations in the dissertation. Interviewees were told that the purpose of the interview was for my dissertation research and were promised confidentiality. I felt that this promise was appropriate as the interviews were open-ended to allow interviewees to share what they felt was most important. My notes used coded identifiers and broad descriptors (e.g. civil affairs bureau official). The names of people and places in the dissertation are pseudonyms. Shequ researchers Through a snowball referral method, I was able to identify and locate other scholars conducting research on Shequ Construction in Nanjing as well as those who were not examining the policy but working in shequ neighbourhoods. I had significant exchanges with four professors during the fieldwork period. They offered tremendous insights and practical help. Our discussions were open-ended and conversational. I would talk about some of the  38 preliminary analysis I was making and pose questions I had on Shequ Construction policies from my readings of Chinese sources as well as my interpretations of what I was observing on the ground. Survey of homeowners Through the assistance of one professor with whom I had been discussing property law and the emergence of homeowners’ associations, I had the opportunity to conduct an informal survey on homeowners’ understanding of shequ governing organizations. Because the law remains vague with regard to the legal standing of homeowners’ associations, he suggested that I pose a few questions in survey form to one of his classes of mature students to see what they would do as homeowners. The class of 43 provided an interesting survey group because, not only were they all middle-class homeowners, but many were mid-career government officials who had an extensive knowledge of the law and an understanding of what would be deemed the appropriate channels for conflict resolution. To ensure confidentiality, the questionnaire did not ask for identifiable personal information. It was designed to take five minutes prior to the start of class and participation was voluntary. Appendix 1 is a translation of the set of questions. The survey, not meant to be statistically valid, served to uncover new questions for thinking about the relationship among shequ governing organizations. Bureau officials My interviews with government officials were aimed at understanding Shequ Construction and the policies with which it interacted. With municipal planners and land management bureau officials, my questions focused on Nanjing’s urban expansion, land development process, and the urban village redevelopment plan. I interviewed three municipal planners who were familiar with the city-wide urban village redevelopment plans. The two land management bureau officials I interviewed were involved with land acquisitions. I was aware that protests related to redevelopment were politically sensitive and therefore my questions did not touch upon villagers’ reactions. Instead, they centred on the broader process, such as the regulatory mechanisms, policy rationales, and implementation procedures. I interviewed five officials at three district-level civil affairs bureaus – two urban core district bureaus and one suburban district bureau. Both urban core districts were named national Shequ Construction pilot sites and therefore were recognized as being on the forefront of  39 experimenting with the policy’s implementation. I chose a suburban district because none of Nanjing’s five suburban districts were name national pilot sites and they were also where the majority of urban villages slated for redevelopment were located. Our interviews covered shequ-related policies and the registration of nonprofit organizations which is under their supervision. The interview process in each of the three districts was similar. I met with the civil affairs bureau director first and afterward they introduced me to their subordinates in the bureau’s Shequ Construction Office. In this context, semi-structured interviews provided the flexibility to probe initial answers, emphasize certain experiences, and allow interviewees to talk about what they perceived to be most important. Furthermore, it allowed for them to ask me questions. They were mainly interested in hearing about my experiences in community development in Taiwan, Canada, and the United States. In these comparative exchanges, they reflected on their work and discussed what they believed to be central in China’s experience. The differences in how district officials talked about the shequ project illustrated various local interpretations of the initiative and highlighted how factors such as local socioeconomic conditions, fiscal constraints, and willingness of the leadership to experiment can influence policy implementation. Shequ and urban villages In the urban villages and shequ, I adopted methods of open-ended interviews and participant observations. Usually, I was usually met by the shequ director, but sometimes by the vice- director and the shequ party secretary. The exception was Rivertown Village, where, perhaps because it was in the process of being dissolved, during the times I visited I found the office closed or staffed by only one or two members of the villagers’ committee. For my initial visit and interview, I prepared guiding questions to begin the interview and to refocus if I felt the conversation strayed off topic. The topical questions also served as the basis of the community profiles I was generating. I asked about their organizational structure, such as what their members and staff do and which subcommittees and small groups have been formed. I also inquired about the nature of their work – who do they service most, how do they spent their day, and what are some of the changes that have occurred since the implementation of Document 23. The third set of questions concerned their relationships with other neighbourhood-based organization, such as with the shequ party branch and nonprofit organizations. The last set of questions was about their relationship with higher levels of government – the street office and municipal and district bureaus. With these four categories  40 of questions in mind, I conducted the interviews in a relatively open-ended manner. I allowed the interviewee to drive the conversations and my questions were answered in various ways throughout the conversation. The interviews were often close to two hours. Our conversations were usually interrupted by residents who stopped by to say hello or had a question or concern. Thus, I was also able to observe the directors’ interactions with residents. I made repeated visits to several shequ and urban villages. I asked follow-up questions to members of the residents’ and villagers’ committee, observed activities in the common spaces, and talked with residents. These questions were more situation-driven. Sometimes the directors would invite me to shequ events, such as neighbourhood clean-up, donation drive, and children’s day celebration. After each visit, I recorded my activities, the informal conversations I had, and my interpretations as field journal entries. Social organizations In addition to visits and participation in shequ and village activities, I also had the opportunity to volunteer with two local nonprofit organizations in these communities. I was placed as a volunteer English teacher at a school for migrant children in Rivertown Village through a charity organization. While I welcomed the opportunity to teach one afternoon a week as a break from field research, teaching allowed me to engage with the everyday life in the village in a different role. I also volunteered with a nonprofit organization, Sunrise Senior Care Services, working on community-based elder care in Nanjing New Village. A Nanjing University sociology professor with whom I had been discussing the policy had invited me to talk about Canadian senior care homes with Director Pan, a social entrepreneur engaged in community-based care centres. About a month after I began working with Director Pan, the opportunity arose for her to establish a centre in Nanjing New Village, where I had already begun conducting initial site visits as a danwei-type neighbourhood. I worked closely with one of her interns. In addition to attending regular planning meetings, we met once a week to undertake various tasks related to the opening of the new centre, such as designing programs and posters. Occasionally Director Pan and I would meet informally in her office over tea to discuss new ideas she had for this new centre and future projects.  41 The opportunities to volunteer were not anticipated at the beginning of the research but arose through people I met. The experiences offered tremendous insight into the role of the shequ system in responding to pressing social issues. They pushed me to think about Shequ Construction beyond the policy document. Rather than seeing whether the policy accomplished its said objectives, working with the teachers and students at the migrant school and the staff at Sunrise called attention to a perspective from outside the bureaucracy. Documents These qualitative findings were supported by published materials and government policy documents and reports. My Nanjing University library card gave me access to university and municipal libraries. I found it frustrating to navigate the library system as much is not available on open shelves but must be requested and then retrieved by staff. With the assistance of one particular librarian, I was able to locate secondary sources. Particularly useful were the local gazetteers, yearbooks, and archived newspapers. Local gazetteers12 and yearbooks are compendiums of newspaper articles, interview vignettes, policy documents, and excerpts of speeches related to particular topics. I realized that recording events through gazetteers and yearbooks has been a longstanding Chinese practice. District civil affairs bureaus also compiled compendiums of their shequ construction efforts. As these were printed in-house and less widely available, I was given copies of these and promotional materials by district officials I interviewed. In addition to those compiled into gazetteers, I located and accessed government documents and reports through searchable databases at libraries and online searches of government bureau websites. Copies of national and local statistical yearbooks were available at Nanjing University and municipal libraries. For some years, national and local statistical yearbooks were available from the websites of the National Statistics Bureau and the Nanjing Municipal Statistics Bureau. In addition, I also relied on the statistics yearbooks published by individual bureaus, such as the Civil Affairs Statistical Yearbook.  12 In the late 1980s and early 1990s, a series of topical gazetteers was undertaken by the Nanjing Local Gazetteers Committee. The recording of local affairs in gazetteers has been a tradition since imperial times. Work on recording and compiling gazetteers was stopped during the Cultural Revolution as family histories were a better genre to depict class struggle, and revived again in the 1980s. For a discussion of the history and revival of local gazetteers, see Thogersen and Clausen (1992).  42 I read two of the local newspapers, Nanjing Daily and Yangzi Evening News, almost daily during the fieldwork period. Back issues of Nanjing Daily were available to me through the Nanjing University library. Conditions of Fieldwork and Research Limitations Positionality: On being a foreign researcher of Chinese descent As with conducting fieldwork anywhere, interactions are shaped by who I am and how I am perceived by those I meet. In China, my ethnicity, citizenship, and gender – a Chinese- Canadian female – were the most salient. Shequ is not regarded as a politically sensitive topic. The people I met were willing to discuss the policy and its local implementation, but they were, however, wary of talking to foreigners. Mostly, I sensed hesitation from shequ directors who were uncertain about interacting with foreign researchers and the potential repercussions. The fact that I had visited them with professors as well as with referrals from the district civil affairs bureau helped to ease their initial reluctance. Most residents did not seem concerned that I was from Canada, and those who did were satisfied if I said that the shequ director knew I was there. Professors and officials, having dealt with foreign visitors in the past, were mainly concerned with whether I had the appropriate papers to be conducting research in China and whether the political sensitivity of my research topic would attract unnecessary attention from higher levels. The directors of the social organizations I was involved with did not seemed troubled by my foreign status. In the increasingly open social climate, they were independent operators who were free to engage outside help. The migrant school where I taught had foreign Caucasian volunteer teachers in the past. From these reactions to my foreign status, I realized that the main concern was bureaucratic: how those at the lower levels were to be accountable for my actions if questioned by their superiors. Because my research stayed close to the ground, I felt confident that if officials at the district level were comfortable with my research, the people I was working with at the neighbourhood level would have less cause for concern. I did not think my research gave bureau officials much concern for they allowed me to visit the shequ on my own. None of the directors whom bureau officials suggested that I contact had been notified beforehand about me. Furthermore, as far as I am aware, my presence and research never did attract the attention of those higher in the bureaucracy to whom district officials were accountable.  43 In addition to being able to blend in racially, as a young woman working alone I did not attract much attention when I spent time in shequ centres and participated in activities. Shequ is a gendered environment where residents’ committee members are predominantly female. Thus they may have reacted differently toward my presence than they would have to a male researcher. The shequ and social organization directors whom I got to know well were all female and were matronly toward me. They referred to me as “xiao Shieh.” The prefix “little” added to the surname is typical of how younger colleagues are addressed in the Chinese workplace. It connotes a sense of familiarity as well as seniority. Sometimes I was asked to assist with simple tasks typically given to junior staff. For instance, several times I sat with the elderly living at Sunrise care facilities as they ate their meals. I did English homework with a young girl as her mother talked with the shequ director. I felt they became somewhat at ease with my presence because they had constructed an identity of me that they were comfortable with and that fitted the norms of Chinese society. Official approval and sampling bias While I did not encounter any great difficulties in gaining access, I nevertheless treaded with an awareness of the “culture of fear” (Yang 1994). For fear of being questioned if my actions or questions were ever misinterpreted as political, I avoided urban villages where I knew land acquisition was contentious and where villagers were overtly resisting the process. I am unsure how the implementation of Shequ Construction would have interacted with the redevelopment process in these contexts. However, had I chosen these villages, perhaps I might not have observed the integrative mechanism of Shequ Construction which led to my questioning of the policy’s articulation with the redevelopment process and the analytical frame of policy conjunctures. Furthermore, I sought official approval whenever I could. For instance, due to the hesitation I felt from shequ directors, whom were in a relatively weak position without any bureaucratic decision-making powers, I felt it was appropriate to seek introductions from civil affairs bureaus. During the initial phase of the research, I had visited and knocked on the doors of shequ directors without any prior introductions. The directors were polite but I could tell that they were uncomfortable. One director said that she would feel better if I came back with an introductory letter from my work unit. She had not dealt with researchers before, much less a foreign researcher.  44 The fact that I was introduced into the shequ by bureau officials presents some issues of bias. The shequ were demonstrative of certain shequ types or showcased certain achievements. In other words, I was not pointed to politically conflicted neighbourhoods that may have shown the implementation of Shequ Construction to be more problematic than that I observed. Had I been able to conduct research in neighbourhoods where, for example, the homeowners’ association and the residents’ committee were not simply disengaged but embroiled in disagreements, the debate that I observed may not have centred on the incorporation of homeowners’ association into shequ governance (chapter 8). And I might have asked a different set of questions. Therefore, I recognize that the workings of Shequ Construction that I discuss, such as the program’s integrative mechanism, were derived from research conducted in noncontroversial neighbourhoods and that these mechanisms may not be the only ones at work. However, the reasons why bureau officials had chosen particular shequ for me to visit did not influence my eventual selection which was based on the interactions of policies and the rapport I formed within the shequ. For instance, bureau officials had selected Nanjing New Village because it represented a well governed danwei-type shequ. In spite of this, my reason for conducting my research in Nanjing New Village was principally based on the opportunity to question the role of nonstate service organizations in providing shequ-based social services and the relationship I had with Sunrise Senior Care Services working within the shequ. Note on Referencing Fieldwork All the names of people and places in this dissertation are fictitious. Formal interviews are referenced with the institution and position the interviewees held and the date. Interviews in Rivertown (RT), Willow Village (WV), Nanjing New Village (NV), and White Blossom (WB) are noted using the acronyms in the parentheses. Interviews conducted in other shequ and villages are referenced with the position of the interviewees and the date, but not with the place names as the pseudonyms do not provide additional information. Interviews with Nanjing University (NJU) and Nanjing Normal University (NNU) professors are noted using their respective acronyms. Information from informal interviews are referenced as fieldnotes with the date and a descriptor of the person who provided the information, such as “principal of a school for migrant children.”  45 3. A HISTORICAL OVERVIEW Urban Community in China  In Other Worlds and In Other Words: Community and Shequ The term shequ is a modern construct that has been appropriated into the reform policy discourse. Behind it lie strong traditions of associational life as well as the state’s long- standing concerns with maintaining social stability and control over its population. Elements of self-help, mobilization, and constructing governable units have continuously shaped the Chinese concept of neighbourhood community. This chapter examines the state-led and social construction of community in Chinese history from the late imperial to the Maoist period. I begin, though, by reflecting on five defining notions of community in Western (predominantly North American) social thought. This inquiry demonstrates the impacts that societal changes in the late nineteenth and early twentieth centuries – and efforts to make sense of them – have had on Western understandings of community. It also illuminates some of the connotations associated with the value-laden term community which may or may not be shared in Chinese community-building practices. Community as anticapitalist: Writings about community from the early nineteenth century carried the romanticism of an earlier, premodern time when people knew and depended on one another. The concept of community was the antithesis to the commodification of human relations under capitalism, as encapsulated in Ferdinand Tönnies’ Gemeinschaft und Gesellschaft (1887/2001). Gemeinschaft (translated as community or civil society) is the idealized preindustrial community based on personal ties and mutual dependence. In contrast, gesellschaft (society) refers to the functional relationships in industrial society based on objective interests such as work. Concerns over the loss of community, social disintegration in cities, and problems of crime, poverty, and rootlessness carried into the 1920s and 1930s in the theoretical writings of urban sociologists. The most influential of these came out of the Chicago School’s urban ecology tradition. Using a positivist approach, empirical studies sought to identify variables in  46 neighbourhood environments that contributed to weakening social bonds and segregation, and to map land-use patterns with social characteristics such as income, ethnicity, and crime levels (Park, Burgess, and McKenzie 1925/1967). These analyses came to shape urban policies, giving rise to place-based development plans designed to revive depressed neighbourhoods. Programs such as those under Roosevelt’s New Deal and Johnson’s War on Poverty were caught within the confines of local determinism, concentrating on specific services but failing to recognize that micro conditions are also shaped by macro social forces and the political economy (O’Connor 1999). Community as neighbourhood: In the late nineteenth and early twentieth centuries, the theoretical linkage between social disintegration and physical environment also carried into the disciplines of planning and urban design (Howard 1902/1946; Stein 1951). Planners shared the same spatial deterministic outlook as the Chicago School sociologists, believing that a greater communal life could be achieved in cities through addressing environmental variables. For instance, Clarence Perry’s widely influential neighbourhood unit paradigm sought to increase communicability among residents and, in turn, foster a sense of social cohesion without severing ties to the larger city (Perry 1929). The key element was a neighbourhood planned around a centrally located elementary school, no more than half a mile from the furthest dwelling and accessible via an internal street system. The school would serve as the civic centre, much like the town hall. The neighbourhood unit model greatly influenced the postwar suburban developments and the new urbanist movement (Duany and Plater-Zyberk 1991; Calthorpe 1993; Talen 1999). The treatment of community as spatial units is premised on the assumption that through market choices and the accompanying social divisions, people who are alike, whether through shared values, race, ethnicity, income, or some combination of these, tend to cluster together. Within a defined spatial unit, the socially clustered residents are more likely to engage in face-to-face contact and identify themselves as members of the community. In this way, one’s neighbourhood become one’s focal point in the anonymity of the city. Community as empowerment and mobilization: Liberal activism throughout the 1950s and 1960s gave rise to a community movement centred on resisting neighbourhood-based government programming. Moving away from the deterministic approach of planners in the  47 first half of the twentieth century, planners increasingly saw themselves as instigators and promoters of community-building processes. Davidoff (1965), in “Advocacy and Pluralism in Planning,” outlines a planning approach whereby planners work in poor and underrepresented neighbourhoods to find out what they need and become their advocates in municipal government. Since the 1960s, the advocacy model has inspired planners to build community through empowering local residents to act collectively to bring about change. Some planners focused on participatory mechanisms and techniques to incorporate more of the unrepresented and give residents more control in the planning process (Arnstein 1969). Others sought to redefine planners’ role as outside the institution of government planning offices, working with local residents in community-based organizations in a process of mutual learning (Keating, Krumholz, and Star 1987). Community-based organizations have developed bottom-up programs such as mortgage lending and affordable housing. What is interesting in the American experience is that while government-led community development programs were extensive, it was resistance to them that came to define the meaning of community. Community without propinquity: Webber’s essay “Order in Diversity: Community without Propinquity” (1970) was one of the early writings rejecting the conclusion that urbanization leads to a breakdown of social fabric and that the neighbourhood necessarily constitutes urbanites’ notion of community. Webber seems to celebrate the pluralism and accessibility to information and ideas made possible by the advanced transportation and communication systems of industrialization. He writes that “the growing pluralism in American society is more than a growing multiplicity of types of people and institutions. Each person, each group bound by a community of interest, is integrally related to each other person and group” (807). Liberated from the confines of the neighbourhood, community, then, is based on social networks that extend outside a bounded local space and across metropolitan regions. It may have more to do with shared values, interests, socioeconomic status, age, ethnicity, and gender than with place. Neighbourhood relations comprise one component of an urbanite’s overall primary networks that also include multiple nonterritorial, associational communities (Wellman and Leighton 1979). Recent writings go further to emphasize that shared interest alone does not form a community. Mandelbaum (1988) describes community as what enables  48 us to distinguish right from wrong.  He writes that each of us belongs to many and overlapping communities, such as nation, city, neighbourhood, family, firm, and church, and in each we encounter a moral code and obligations that we must negotiate and balance with individual self-interest and rights. And so, most importantly, community is a sense of mutual interdependency, without which there is no need for community. Community as a myth: Examining the social conflicts in American society, Sennett (1970) writes of the “myth of community,” where our desire for a coherent, shared community legitimizes racist and classist exclusionary behaviour and urban policies. He describes the myth of community solidarity as a purification ritual in which people draw a definite set of desires, dislikes, and goals that binds them together as one being and erase all that that might convey feelings of difference and conflict (36). The paradox Sennett brings to light is that seeking cohesive community inevitably entails a process of exclusion. While appeals to community envision more local and direct control, community operates to exclude or oppress those perceived as different, and thus reproduces the exclusion that first led to community building and affirming group identity and solidarity (Young 1990, chapter 8). Young discusses a politics of difference where the notion of the public is not conceived of as a unity transcending group differences and entailing complete mutual understanding. Rather, the public is a place where “people witness and appreciate diverse cultural expressions that they do not share and do not fully understand” (241). These critiques of the ideal of community are not against the formation of social group affinities but argue that in an increasingly global and mobile world, a given place is no longer defined by a single culture but as one with multiple voices (Sandercock 1998). From the sense of loss expressed at the turn of the twentieth century to being labelled a myth by the end of the century, notions of community are continuously being defined and challenged to reflect ongoing societal changes. To understand any notions of community thus requires attention to wider sociopolitical forces. As the rest of the chapter will discuss, in the Chinese experience, senses of place are continuously shaped by matters of governance as social actors contend with sweeping changes in the political economy. This historical overview seeks to place the specific meaning of the reform era shequ in a continuum of community building in urban China. Centred on neighbourhood life in Nanjing and the Jiangnan region of the lower Yangtze River Delta, it covers several centuries, separated into  49 three main periods crossing through late imperial wards, Republican neighbourhood units, and socialist collectives espoused by Mao. It is not meant to be a historiography of shequ. Rather, my aim is to discover the changing and persisting functions of neighbourhood-based governance over time. Morality and Elite Activism in Imperial Streets and Wards By the late imperial era,13 the Jiangnan region of the lower Yangtze River Delta was already prosperous, highly urbanized, and linked through an intricate system of canals and ports. In Nanjing and the nearby cities of Suzhou and Hangzhou, textiles were being manufactured in large quantities. Nanjing had been the imperial capital during the first decades of the Ming Dynasty. After the capital returned to Beijing in 1421, Nanjing reemerged with a new identity as a cultural centre. The city came to be a centre of intellectual and artistic life with concentrations of artists, writers, book collectors, silk and brocade producers and craftsmen, and pleasure-loving elites. The old capital became a city where a “scholar-official in office could have everything but the guts of political life [in Beijing]” (Mote 1977, 152). With continued economic development and population growth, as neighbouring cities in the Jiangnan region rose in prominence, Nanjing became one city in, but not the hub of, the wealthiest region in China (Skinner 1977a; Rowe 1993). Although not in decline, Nanjing did recede into the background, falling behind Suzhou in prominence (Santangelo 1993). In this environment of flourishing urban culture, urban governance was achieved through a combination of centralized imperial control and grassroots leadership. State control reached the base level of the prefecture (zhou) and the county (xian). Below this, control mechanisms were formally in the hands of wealthy residents who were appointed as service officers to manage tax collection, public security, and labour services (Von Glahn 1991, 282). Nanjing urban residents living within the city walls were organized into fang wards; those outside the  13 The late imperial era is generally taken to refer to the Ming (1368–1644) and Qing (1645–1911) dynasties. It was during these two last dynasties that an increase in population, commerce, and agricultural production brought significant changes to urban growth and urban life. While these two dynasties spanned over 500 years, in this section I highlight the government institutions and political thoughts on urban self- governance that would later impact Republican and then Communist leaders.  50 walls were organized into xiang suburban units.14 In this fang-xiang system, based on the rural lijia tithing system, the household was the basic unit of society. Theoretically, 110 households made up one fang ward (or one xiang in the countryside).15 Some ward names were simply numbers, such as “Tenth Fang,” and “Eleventh Fang.” Other wards took on names descriptive of the population living there, with names that indicated dialect, occupation, or trade (figure 3.1). Records show that some were named “Craftsmen Fang,” “Brocade Fang,” and even “Poor People Fang” (Mote 1977, 146; Nanjing City Local Gazetteer Editorial Committee 1994a, 150). In each ward, the wealthiest ten households assumed the role of the ward commissioners, and were responsible for collecting funds and assigning duties. Residents contributed by carrying out services as directed by the ward commissioners, providing goods required by government agencies, or paying compensation in kind (Tsurumi 1984, 258; Von Glahn 1991). The required services included posts such as watchmen who, led by unit headmen, patrolled the city at night when activities and even passage on city streets were prohibited. Over time, this labour service underwent several reforms as problems arose with unfair burden on merchants (Von Glahn 1991) and abuse by wealthier residents and degree-holding elites (Fuma 1993, 52). Instead of requiring residents to do the work themselves, by the late Ming, a tax was assessed based on property ownership, and officials took the responsibility for hiring labour services (Von Glahn 1991; Fuma 1993). For many reasons, by the Qing dynasty, the fang-xiang system was weak and ineffective. First, the wards had increasingly become less of the closed system they were designed to be. The decimal system had to be continuously adjusted to accommodate population growth. Traditionally, the system segregated the urban population according to social status and occupation. In some parts of the city, wards were enclosed by a wall and had gates that were closed at night. Over time, the fang increasingly became simply streets (jie) and alleys (xiang; Nanjing City Local Gazetteer Editorial Committee 1994a, 150). Second, the attention of the  14 An exception: The Taiping Rebellion (1850–1864) was a revolt against the Qing government, and Nanjing served as its capital. During the Taiping Rebellion, Nanjing’s fang-xiang system was abolished. When the Qing forces regained control of the city, new local administrative jurisdictions were demarcated using the baojia system of policing (Nanjing City Local Gazetteer Editorial Committee 1994b, 45). 15 Perhaps due to the need to accommodate population growth, by late Ming, an additional level was added. The 110 households became a tu, and as few as one to as many as four tu made a fang. The new fang definition took on a wider jurisdiction, but the principle of household decimal hierarchy and responsibilities remained unchanged (Nanjing City Local Gazetteer Editorial Committee 1994a, 150; 1994b, 43).  51 Qing state was not on local public management, but on regional and countrywide issues of taxation, military, water control, grain procurement, and famine relief. As the population and territory of the dynastic state continued to expand, the existing bureaucracy became less able to provide necessary services (Skinner 1977b, 548). The system, particularly in rural areas, lacked funding. The volunteer lower-unit heads, appointed by the ward commissioners, tended to bear all of the organizational expenses. Third, the fang-xiang apparatus was not about effective governance in welfare terms. Corruption and exploitation, more acute in rural governance, threatened the system’s effectiveness (Kuhn 1975, 262). Because the system was not part of the state bureaucracy, no one had the authority needed to make any meaningful changes. Authority rested in the hands of the city magistrate who, following the Qing emperor’s rule of avoidance,16 not only was not from the region but faced imminent transfer after a few years. He therefore did not have accountability or vested interest in the locality’s well-being, and was also relatively insulated from the pressures to favour one local interest over another (ibid.). Nevertheless, the fang-xiang system remained relatively intact, largely because it was supported by the informal leadership of social organizations formed along lines of occupation and native place, which roughly corresponded to the ward demarcation. Over time, as the imperial state shrank from its welfare functions, the elite-managed associations increased and their activities diversified to include construction of public infrastructure, fire control, support for poor merchants and workers, and operation of care homes and welfare services (Rankin 1993, note 3). Local elites had reasons and incentives to take on these extrabureaucratic responsibilities. While more true in rural villages where lineages and kinship obligations were stronger, local elites regarded themselves both as “native sons,” charged with the responsibility of furthering and protecting local interests, and as members of the governing class invested in performing services essential to maintaining the social stability upon which their own formal prerogatives rested (Kuhn 1975, 260). Furthermore, the breakdown of the compulsory tax and corvée system created an arena for the gentry and merchant elites in which participation in charitable concerns, in line with Confucian and Buddhist values, enhanced their local standing and elevated their status (Rankin 1993).  16 Premised on the belief that officials would act in the best interest of their home region over the interest of whole empire, the rule of avoidance prohibited officials from serving in his home province and rotated officials to new positions after a short period of time to prevent attachment to the local people.    52     Figure 3.1 Ming Dynasty map of bridges, streets, and fang Drawn in the cartographic style of the time, the block labels show the names of the main city gates, bridges, streets, and fang wards. Some of the fang names indicated are numbers, such as “First Fang,” “Second Fang,” and so forth. Others are more descriptive of the population living there, with names that indicated occupation or trade, such as “Silversmith Fang” and “Brocade Fang.” Source: Hongwu Capital Map Gazetteer (Hongwu jincheng tuzhi), reproduced and published by Nanjing Press, 2006.  53 On Self-Governance: Early Political Thoughts and Debates17 Local governance during the late Qing dynasty, whether expressed as bureaucratically administered self-management at county and subcounty levels or as extrabureaucratic initiatives to provide social services, was characterized by collaboration between officials and elites. Governance rested on how best to arrange this relationship to bring about the stability necessary for the dynasty’s longevity. Traditional political thought on local self-governance considered, philosophically, how to instil strong moral character based on Confucian values and, practically, how local officials would be chosen. As China was a predominantly agrarian society, the concept of local self-governance refers primarily to the rural context. Even so, for this research’s particular concern with urban neighbourhoods, this brief discussion is pertinent not only because the turn of the twentieth century marks an important time period but also because the debates that emerged during this time illustrates Chinese reasoning toward governing the populace. The tension between establishing a modern system of administration and hanging on to Confucian ideals of morality is a thread of continuity down to the present. During the late Qing, self-governance discussions were heavily framed around the principles of two opposing governmental systems – feudalism (fengjian) of antiquity, where power was concentrated at the local levels, versus centralized bureaucracy (junxian) of late imperial times, where power was concentrated at the top. The writings by major political thinkers of the Ming-Qing transition, and later of late Qing reforms, promoted principles of feudalism in their prescriptions for change, as they grew dismayed with local officials’ disregard for the welfare of the people under the centralized system. Gu Yanwu (1613–82) and Huang Zongxi (1610–95), two important figures in Chinese intellectual history, attributed the climate of distrust and disregard for local community interests under the central bureaucracy system to the appointment of outside magistrates by an autocratic ruler. They believed that native local officials were more capable of engaging local gentries in political discussions and more willing to subordinate self-interest to community welfare. Gu Yanwu, for example, argued for more officials at the bottom than at the top. He proposed elevating the status of magistrates, as they were the officials closest to the people, and abolishing the higher  17 This section draws upon the works by Kuhn (1975), Min (1989), and Lee (1998) on the political thoughts on local self-government during the late Qing reform (1898 to 1911).  54 provincial posts, including those of governor and governor general (Kuhn 1975, 263-4; Lee 1998, 35-7). Huang Zongxi advocated a system of schools as town halls. Run by local superintendents, schools would become sites for teaching the classics and discussing politics. They would be charged not only with educating scholars but also with overseeing government officials and guarding against malpractices. He proposed that the first and fifteenth of every month be days of assembly when local elites, licentiates, and certified students would come together to participate in political discussions led by the superintendent (Lee 1998, 37-9). The debate between feudal and centralized bureaucracy reemerged among late-Qing constitutionalists. As social turmoil threatened the dynasty, the question of how much autonomy to transfer to localities became a growing concern for the imperial state. It was at this time that the term self-governance – zizhi as it is used today – was introduced into Chinese political debate from the Japanese term jichi by Huang Zunxian (1848–1905), the Cantonese diplomat and interpreter of Meiji Japan.18 The popularization of the concept of self-governance during the late Qing and into the early Republic began with the influential writings of Kang Youwei and Liang Qichao, both of whom drew on the Western local parliamentary system as well as the Chinese feudal system. They called for administrative reform, as opposed to choosing between either feudal or centralized bureaucracy. Kang believed that the foundation of a strong nation rested on the transformation of a passive and indifferent people into active and concerned citizens involved in local institutions. He proposed that residents be organized into self-governing basic units of 10,000 people, each with an elected deliberative assembly. Citizens would participate through the election of assemblymen. However, citizenship was to be determined in the traditional sense by moral qualifications, defined as those who had a respectable family background, had never committed crimes, could afford to give alms to the poor, and could pay 10 dollars worth of  18 The Japanese term jichi was in turn a translation of the German borrowing of the English word “self- government.” Yamagata Aritomo, the founder of the Japanese local self-government system, was influenced by the constitutional thought of the Prussian legal scholar Rudolf Gneist (1816–1895). In Gneist’s work on the English Constitution, he retained the word “self-governance” without translation, recognizing its culturally specific nature. For Gneist, self-government had two major functions. One, it served as the mechanism for socializing the dominant classes. Appointing local elites to honorary unpaid public service posts raised them above their own parochial interests, imparted practical knowledge of the state, and fostered devotion to national interests. Two, self-government insulated the local administration from party politics at the national level. “Local” (difang) was an added prefix in the Japanese application, which was also transferred to China. For a detailed discussion, see Kuhn 1975, 270–72.  55 tax (Kang Youwei 1974, cited in Lee 1998, 40-41). According to Min (1989, 126), “the initial proposal of having native people deal with affairs of their own provinces gradually developed into an argument for local self-government, centering around the local parliaments, with the gentry class as the core.” The unit of “self” in the concept of self-governance drew on the moral concept of “one’s own” in Neo-Confucianism. Self-governance is understood as taking care of and improving one’s own. The ideal government is not built on control over the governed, but on guiding them and improving their ability to self-govern (Lee 1998, 44). Learning and disciplining one’s self becomes the foundation to build a harmonious collective of family, community, region, and country. Confucianists believed that elaborate codes and rigorous control mechanisms were not going to be more effective than a system that enlisted the natural feelings of a man toward that which is his – his family, property, and community.19 With his ties to local society and his own future assured, he would turn all his energy on strengthening his country (Kuhn 1975, 263–64, referencing Gu 1934). Most importantly, self-governance or the decentralization of ruling power was never about the detachment from the larger imperial political order. Self-government theories in China stemmed from the principles of the feudal system. Scholars in search of a new governmental system sought institutional reform within the monarchic structure, believing that the monarch was indispensable. The monarch was the person capable of transforming the multitude of self-interests throughout the empire into a common interest. Local governance and the survival of the unified Chinese state were believed to be interdependent (Kuhn 1975, 261-8). In the waning years of the Qing dynasty, local self-government was formalized in the constitutional programs of 1908 in hopes of saving the dynasty. In 1909, the Qing government officially issued the Charter of City, Town, and Township Local Self-government. Self-governance manifested itself in the creation of a governmental body that exercised supplementary and supporting roles alongside the formal local government at municipal,  19 The debate revolving around moral laws versus positive law is longstanding between two competing traditional schools of thought, Confucianism and Legalism, in Chinese political philosophy. Confucians theorists believe that good moral rulers are just as important, if not more, than rule of law to bring about social order. Legalists emphasize law as an instrument of state power to control people whose nature is believed to be evil and selfish. In local governance systems, the legalism relies on policing and mutual control and Confucianism on self-governance and mutual aid (Dutton 1992, 21-33).  56 town, and township levels (Kuhn 1975, 276). The new self-governing bodies undertook governmental functions but they remained under the control of local governments that could dismiss their members and overturn their decisions. Inadvertently, instead of co-opting elites into the bureaucracy by formal appointment, this arrangement only intensified the existing antagonistic relations between local elites and magistrates, adding fuel to the events surrounding the 1911 revolutions that brought about the downfall of the dynastic era (Zhong 2003, 32). Republican Modernity and Traditional Continuity20 In 1927, the Nationalist (Guomindang) government formally established its capital in Nanjing. Following Sun Yat-sen’s democratic principles and program for national reconstruction, the Nationalists were committed to the transition toward a constitutional government of democratic self-rule at the county level. Sun strongly advocated building a constitutional government from the bottom up, but the structure was not one based on individual rights and popular democratic rule. It was centred on nation building and liberation from monarchy and imperialism, even if at the expense of individual freedom (Ogden 2002, 66). Of his “Three Principles of the People” (sanmin zuyi), self-governance was central to the Principle of People’s Power of Governance (zhiquan). The idea of self-governance was not centred on constructing a system of direct local government but on the role local governments has in uniting and strengthening the young republic. Sun believed that local government was a means of cultivating citizens to build a strong nation. Local government would mobilize the public, bring local initiatives in line with national objectives, and help to bring about national integration (ibid.). While followers were committed to implementing the democratic visions of the Republic’s founding father, as political modernization progressed, the system reverted to one preoccupied with social order and control. First, at the municipal level in 1931, the Nanjing municipal government established the additional level of “self-governing district” (zizhiqu) in the administrative hierarchy. However, with limited popular participation, it really sought to  20In this section, I focus on Nanjing from the time it became the capital of the Republican government in 1927 to the Communist victory in 1949.  The warlord period from 1911 to 1926 and the period of Japanese occupation from 1937 to 1945 are not covered directly because this analysis focuses on the conception of self-governance by the Republican regime.  57 formally impose a level of coordination between the basic level of formal government and the grass roots. In the cities, districts corresponded to police jurisdictions.21 Over the course of the Republic, as the urban boundary was redrawn several times and the system was both undergoing fine tuning and suffering from the disruption of Japanese occupation, Nanjing’s urban core fluctuated between five to eight districts, and the number of suburban districts changed as adjacent townships were incorporated or removed.22 By the end of the Republic, Nanjing was divided into thirteen districts – seven constituting the urban core and six comprising the suburban area. The population of the core districts ranged from about 76,000 to close to 170,000 (Nanjing City Local Gazetteer Editorial Committee 1994b, 50). The creation of self-governing districts was more for downloading responsibilities than for actual transfer of power. In Nanjing, a separate office within the municipal government, the Office of Self-government Affairs which later became the Civil Affairs Bureau, supervised the districts. Each district office was staffed by a district head, his assistant, a secretary, and obligatory labourers (fuyi). Districts pursued matters delegated by the municipal government, mainly routine matters of civil affairs and household registration. The district head was not chosen by popular election but selected by the mayor, confirmed by the Internal Affairs Bureau, and subject to the mayor’s supervision (Nanjing City Local Gazetteer Editorial Committee 1994b, 51). Even though the Organic Law (1930) stipulated that district affairs would be determined by the district residents’ general assembly and executed by the district representative council, in reality, issues were relayed to the appropriate municipal departments for decision making and resources were dependent on allocations from the municipal treasury (Wang 2001, 60). Despite the relatively powerless general assembly, the self-governing district did instil a sense of urban residency, in contrast to the past when a person’s identity was tied to a native place in the countryside. To be eligible to vote and participate in self-governance, people had to first establish proof of residency. Residents were defined as those over the age of 20 who  21 For a discussion of the rural counterpart, based on the Shanxi model of self-government, see Kuhn (1975). 22 For instance, Nanjing was initially in 1931 demarcated into 21 self-governing districts.  To conserve resources, districts were merged two years later to correspond to the 8 police districts.  Then in 1934, with the redrawing of city and provincial boundaries, the city was enlarged to encompass adjacent townships, resulting in a total of 11 districts (Nanjing City Local Gazetteer Editorial Committee 1994b, 50).  58 had lived in Nanjing for at least a year or owned residential property for at least two years, and who had registered and taken an oath. Each district would manage its own registry and, subject to municipal inspection, issue residency permits (Nanjing Self-Governance Work Report 1937, cited in Wang 2001, 76). Second, at the neighbourhood level, the pursuit of modern governance institutions did not imply a break away from traditional practices. Self-government at the subdistrict level consisted of a three-level hierarchy of fang wards, lu streets, and ling blocks, each with an elected head person. The district pyramid was constructed from the following building blocks: every five households formed a block (ling), every five blocks formed a street (lu), every twenty streets formed a ward (fang), and lastly, every ten wards formed a district (Nanjing City Local Gazetteer Editorial Committee 1994b, 46). This social structure, based on blocks rather than households, was to work in tandem with a modern street-grid system and an adapted neighbourhood unit as outlined in the Capital Plan23 (figures 3.2). Before this elaborate system integrating the imperial street-ward governing structure and modernist neighbourhood design could be fully implemented, in 1935 the Guomindang government reverted to the baojia system in both urban and rural districts (Nanjing City Local Gazetteer Editorial Committee 1994b, 46). In the unstable climate of the impending Sino-Japanese War (1937-45) and even afterwards, the baojia system, based on mutual surveillance and collective responsibility, together with the hukou household registry system, presumably provided for greater security. It is important to point out that while the Guomindang government restored the baojia system, the imperial system built on concerns of community mutuality was adapted to serve modern management of public security and placed under the police department. Where it had once offered local autonomy, the system now strengthened social order and aided state intervention (Dutton 1992, 192). Under the baojia system, which used households as the building block, ten households formed one jia and ten jia formed one bao. In the urban baojia system, a household was  23 Nanjing’s Capital Plan (Shoudu Jihua) was part of the Nationalist government’s grand state-building. The Plan, undertaken by China’s prominent architects and planners, many of whom were trained in Western universities, and foreign advisors, introduced scientific principles to Chinese city planning.  It combined modern scientific surveys and building methods with Chinese practices and aesthetics.  For a detailed examination of the Chinese and Western influences on the Capital Plan, see Cody (1996) and Musgrove (2002).  59 defined as one street number. So, even though there may have been more than one family living at the same street number, they were counted as one household, with one household leader responsible for all families at that street number. Flexibility and variations occurred in the actual implementation of the system. For instance, in Nanjing, twenty-five households formed one jia, and twenty-five jia formed one bao (Nanjing Urban District Baojia Formation Implementation Draft Plan, cited in Wang 2001, 62). Eventually, the baojia system became more formalized and elaborate. By 1947, every bao not only had a head (baozhang), but also a vice-head (fubao), administrative secretaries (baoganshi), and supervisory staff (zhidaoyuan). As the system was chiefly about surveillance, the administrative secretaries were trained and dispatched by the capital police headquarters. The supervisory staff was placed under the supervision of the special military organization (juntong tewu zuzhi). At the level of jia, in addition to having a leader, as was the practice during the use of the system in the imperial period, there were now supervisory staff and a patrol leader. Added staff at the base level was to assist the baojia leaders in carrying out administrative duties and tax collection, as well as in implementing security measures such as inspecting hukou registration, keeping an eye on neighbours’ actions and speech, training the self-patrol team, and guarding against Communist propaganda (Nanjing City Local Gazetteer Editorial Committee 1994b, 49). Self-governance remained heavily structured with the state defining its meaning and educating the citizenry on its practice. Nanjing Self-governance Work Report (1937) discussed lectures delivered by the mayor and district leaders on patriotism, national economic development, district-led low-interest loans, and charity drives (Wang 2001, 79- 81). Despite the electoral democracy ideated in the Organic Law (1930), during the short period of Nationalist rule, governing power was never transferred. It is unclear how the district self-governance would have fared and evolved as the Republic was plagued by turmoil and war, and constantly underwent administrative reforms.   60             Figure 3.2 Drawings from the Nanjing’s Capital Plan (1929) The map and sketch are from Nanjing’s 1929 Capital Plan (Shoudu jihua). Undertaken by western-trained Chinese architects and planners and foreign advisors in the 1920s, the Capital Plan sought to incorporate Western scientific planning practices to Chinese city planning. For instance, the zoning map (left) indicates the main road system and districts zoned for various land uses – parks, residential, commercial, and industrial. The sketch (right) is of a residential neighbourhood for government workers. The suburban design and the green commons are reminiscent of the neighbourhoods of the City Beautiful Movement that was gaining popularity in the West at the time. The imperial street-ward social governing structure would presumably be adapted to this modernist, Western neighbourhood design. Source: Office of National Capital Design and Technology Commission (1929). Shoudu jihua (Capital Plan). Reproduced and published by Nanjing Press (2006). Map number 52 (right) and 56 (left).  61 Birth of the Residents’ Committee When the Communist military proclaimed the birth of a new China on October 1, 1949, the country was in chaos after years of war, first against the Japanese and then the civil war between the Communists and the Nationalists. By the time Mao’s Red Army claimed victory, urban infrastructure and provisions were in shambles. For fear of greater public disorder, the local police system was retained and staffed by Communist military personnel. Civil administration functions were undertaken by local police stations and assisted by local residents who were Communist supporters (Schurmann 1968, 371-74). These measures were necessary as one of the earliest directives of the Chinese Communist Party (CCP) was to annul the baojia system installed by the Nationalist government. The Party regarded baojia leaders as elites who held control and acted as middle men between the people and the government.24 In reconstituting local government, the rationale was for the people to have direct contact with the government.  Even so, the new forms of ordering the populace strayed little from past practices. Urban districts became a formal level of government, named the District People’s Government. The district government and the corresponding CCP branch – the district Party work committee (qu gongwei) – were different in name but staffed by the same people. What would later become street offices and residents’ committees were still being debated, with cities experimenting with different models.25 According to the Nanjing Civil Affairs Gazetteer (1994b, 59-60), in 1949 the Nanjing municipal government kept the 13 urban districts delineated by the Nationalist government. Within each district, 20 households were grouped to form a residents’ small group, and every 20 small groups formed a residents’ committee headed by 3 of the 20 small group leaders. When put into practice, the form varied from district to district, with some having only small groups but no committees. A year later, in 1950, the residents’ committee was reconfigured. The municipal document Decision Concerning Strengthening Government Work and Agencies suggested that a residents’ committee be established using the jurisdiction of the local police station. In implementation,  24 According to a memorandum dated March 1, 1949, central committee members of the Chinese Communist Party called for the denunciation of baojia leaders in CCP occupied cities. Baojia leaders were to assist with public security but subjected to a public denunciation and monitoring by the government and the mass to “strip them of their past sense of superiority” (Central Archive 1989, vol. 18, 1 March 1949). 25 Schurmann (1968) provides a brief description of the residents’ committee structure in Tianjin in the early1950s.  62 each district government staffed each local police station with one to three officials responsible for residents’ affairs, such as sanitation and health, and political mobilization such as collecting donations to “Fight the Americans and Aid the Koreans” (kanmei yuanhan). The appointed officials became the director or vice-director of the residents’ committee, and the rest of the committee members were unpaid residents. Even before the last initiative could be fully experimented with, in 1952 the residents’ committee was once again reconfigured, this time according to the East China Military Government Committee’s provisional regulation Proposal for Trial Concerning the Establishment of Residents ’ Committees in Cities with over a Hundred Thousand People. The proposal stipulated that the boundary of a residents’ committee would range from 1,000 to 10,000 people. Every 10 to 30 households would form a residents’ small group. Each small group would elect a representative and together they would form the residents’ representative council. Residents’ committee members would then be elected from this council. The members would spearhead separate subcommittees for crime and safety, fire prevention, education and culture, health and sanitation, and mediation. Subsequently, Nanjing’s existing 65 residents’ committees were subdivided into 132. The municipal government allocated 12 RMB to each residents’ committee for office expenses and a monthly living subsidy of 30 RMB26 for committee directors and vice-directors, if they were not part of any production teams (Nanjing City Local Gazetteer Editorial Committee 1994b, 53–65). After years at war, thoughts concerning governance rested between giving the people decision-making power over their own affairs and being able to lead, mobilize, and absorb them into the socialist agenda. At the time, urbanites were recognized as having two types of status – youzuzhi or wuzuzhi. Those who “have an organization” (youzuzhi) were essentially productive members of society and included workers, teachers, and students. Those who “do not have an organization” (wuzuzhi) included the unemployed, street vendors, shop owners, and the self-employed. Surveys conducted in the early 1950s highlighted that wuzuzhi, dispersed in the “urban sea,” accounted for two-thirds of all urbanites (Gao and Guo 2003, 98-99). This high number, together with multiple uncoordinated organizations formed in  26 These figures have already been converted to RMB from “old” RMB used in 1952.  63 response to different mobilization efforts (such as street cleaning and policing), became a great political concern. In a report to Mao written in 1953, Peng Zhen27 advocated the creation of a mass self- governing organization (Peng 1953/1991). For him, a self-governing organization served two practical purposes in China’s industrial and socialist transition. The first was to bring together those deemed to be without an organization under the management of a state agency. Peng believed that street offices and residents’ committees were essential because in certain cities as many as 60% of urbanites did not belong to a work unit (241). However, he argued that it was unnecessary to make street offices an administrative level (as they are today) because, in the course of transition, those not yet part of the class of workers and assigned to an organization will decrease and the role of street offices will diminish. The second purpose was to unify and standardize the varying forms and responsibilities of residents’ organizations. This new self-governing organization, operating under the principle of voluntarism, would undertake affairs of collective welfare, disseminate government policies and regulations, mobilize residents, and communicate residents’ concerns to street offices – the government’s dispatch agency. Peng Zhen envisioned an organization determined through election by residents’ small groups. It is important to note that in Peng Zhen’s use of the term, self-governance (zizhi) did not preclude the grassroots organization from working on behalf of the government. And, self-representation through elections was not the same as wielding political and bureaucratic power (zhengquan). Over the next couple of years, municipal and military governments continued to propose various configurations of residents’ committees, differing in how their boundaries were to be delineated, such as whether by population or by public security jurisdiction. Ad hoc adjustments and experiments in different cities eventually culminated in the 1954 Organic Law, approved at the fourth plenum of the first Standing Committee of the People’s Congress  27 Peng Zhen (1902–1997) was a key figure in post-1949 political history. A firm believer in and enforcer of Marxist-Leninist thought, he became a seminal figure in the development of the legal system. Peng was appointed the Party Secretary (mayor) of Beijing in 1951 and the Party Secretary of the Political-Legal Commission. Peng increasingly found himself in disagreement with Mao, particularly on the role of the Party.  For instance, he insisted that the Party be subjected to legal constraints. When he fell out of favour with the Chairman, he was forced to retreat from everyday governance, and was later jailed during the Cultural Revolution. Rehabilitated under Deng, he continued to guide the establishment of China’s judicial system and to formulate the PRC legal code (Potter 2003).  64 on December 31, 1954. The Organic Law tasked the newly institutionalized residents’ committee with five main areas of work: handling residents’ public welfare; reporting concerns from residents to the local People’s Committee; mobilizing residents to respond to the calls of the government and to obey laws; leading residents in collective security and sanitation efforts; and mediating disputes between residents (PRC Organic Law 1954, Article 2). Following public security jurisdiction, there was to be a residents’ committee for every 100 to 600 households. Under the direction of each residents’ committee, there were to be not more than 17 residents’ small groups, each representing 25 to 40 households. An elected representative from each small group would make up the residents’ committee. The members of the residents’ committee would then elect one director and one to three vice-directors, with at least one person overseeing family planning (Article 3). In Nanjing, following the adoption of the Organic Law, 499 residents’ committees were created in 1955, overseeing about 89% of the city’s population (Nanjing City Local Gazetteer Editorial Committee 1994b, 60). As Peng Zhen (1953/1991) had written, it was expected that all urban residents would eventually be engaged in productive labour in a danwei work unit. And by 1957, the workplace had become the principal organizational unit in cities with over 90% of the urban workforce belonging to either a state-owned or a collective-owned work unit (National Statistics Bureau 1994; cited in Bray 2005, 101). Keeping to the form of social organization based on grassroots units it had developed during the revolutionary years in Yanan, each workplace functioned as a self-sufficient unit with relative operational autonomy within the centralized political structure. The new regime depended on each work unit to fund and organize welfare provision as well as to construct facilities for the delivery of welfare services, such as clinics and schools (ibid., 104). Working in tandem, the institution of the residents’ committee supplemented the primary workplace-based system in three crucial ways. First, residents’ committees played an important role in the daily lives of those who lived in the older sections of the city. New workplaces that were constructed at this time were enclosed settlements in suburban districts that combined factories, residences, and wide- ranging facilities exclusively for their workers, from nurseries and canteens to co-op shops and health clinics. These developments, mainly undertaken by large work units with greater resources, represented the archetypal danwei compounds.28 Thus, while the majority of  28 For a detailed examination of the danwei spatial form, see Lu 2006, chapter 4; Bray 2005, chapter 6.  65 urbanites belonged to a danwei, many did not live within a work-unit compound and commuted to work. Some lived in traditional residences subdivided and allocated by their work unit. Others lived in small housing compounds of low-cost, three to five-storey apartment blocks built by their work units or government housing offices as funds and parcels of land became available (Gaubatz 1995, 31-2; Whyte and Parish 1984, 82). Two, residents’ committees served as the assigned unit for those who did not belong to state- owned enterprises. Like the danwei, the residents’ committees functioned as self-sufficient grassroots units where the livelihood and welfare of members were taken care of within the new collective-oriented organization that was meant to replace the family unit. During the Great Leap Forward (1958-60) when the idea of communes spread to cities from the countryside, street offices and residents’ committees organized mess halls, day care centres, and small handicraft shops (Vogel 1971, 86). Many of these service-oriented and economic functions were retained after the decline of urban communes, and also reappeared in today’s Shequ Construction. Three, in addition to providing welfare and livelihood opportunities, residents’ committees also served as a surveillance mechanism as those who did not belong to a place of work were considered politically suspect. CCP organization emphasized political loyalty and productivity and made active participation in industrial labour the determinant of political status and material benefit. Where the social identity of the urban worker had in the past been determined by hometowns and trade guilds, under CCP rule it was reconstituted through one’s place of work (Bray 2005, 100). For the retired, unemployed, and self-employed who had not yet be mobilized through the workplace, the residents’ committee would serve as their unit of identification. Neighbourhood-based mobilization became particularly important in the politicized atmosphere of the Cultural Revolution (1966-76). With neighbourhoods regarded as “dead corners,” the network of residents’ committees became a mechanism through which the regime sought to maintain the political commitment of urbanites whose political participation was believed to lag behind those in factories and rural communes (Salaff 1971, 318). Residents’ committees were renamed “cultural revolutionary small groups” (wenge xiaozhu) and their functions, as set by the 1954 Organic Law, were suspended. They became more of  66 an instrument in political class struggles and assisted in public denouncement (pidou) sessions. Furthermore, the lack of municipal funds decentralized social service provision and drew residents into broader and more intense community-based activities (ibid., 314). Many people I talked to about everyday life in residential compounds prior to reform and opening remember the xiangyangyuan activities of their residential compound (literally sun-facing courtyard). The workers, cadres, and youth of each compound would mobilize residents to study the works of Marx, Lenin, and Mao; disseminate the policies of the Party-state; and look for and denounce “feudal” practices such as worshipping ancestral spirits. Legal institutions had essentially collapsed during the upheaval of the Great Leap Forward and tumultuous decade of the Cultural Revolution. Revolutionary morality and the thought of Mao Zedong was placed above all else and calls for rule of law were denounced as “rightism” (Tay 1987, 570-2). Following the death of Mao and the arrest of the “Gang of Four” in 1976, the new Party leadership sought legal reform as a first step in restoring social and political order and safeguarding economic reform initiatives. Deng Xiaoping, resuming Party leadership after being purged from his position and imprisoned for his disagreements with Mao, declared broad goals for the re-establishment of a legal system whereby the “rule of persons” had to be replaced by the “rule of law” (Lo 1992). The Standing Committee of the National People’s Congress resolved that all laws enacted during the 1950s and 1960s were to remain in effect until the passage of revisions and if they did not conflict with the Constitution and other enacted legislation (Li 1996, 331). Accordingly, the 1954 Organic Law was reinstated for the time being. In December 1989, the Organic Law on Residents’ Committee was adopted at the eleventh plenum of the seventh National People’s Congress Standing Committee. In addition to the broader context of legal reform, the renewed attention to the residents’ committee came at a particular point in time. As the next chapter discusses in greater detail, by this time the Ministry of Civil Affairs had already put forth the concept of neighbourhood-based social services that relied on strengthening the existing residents’ committee and had begun to issue circulars and to organize symposiums to exchange ideas. Furthermore, the Party leadership had enacted the Organic Law on Villagers’ Committee two years prior in 1987 and new measures concerning their urban counterpart were expected to follow suite (Choate 1997, 8). The attention also came following the Tiananmen Square demonstrations in June of the same  67 year. Party and government leaders were wary of social organizing activities, and the attention given to residents’ committees was likely responding in part to their role as watchful guardians of neighbourhood activities. Legal Standing of the Residents’ Committee The People’s Republic of China Constitution and 1989 Organic Law spell out the basic nature, organizational structure, and function of the residents’ committee. First, the Constitution recognizes residents’ committees as “mass organizations of self-management at the grassroots level,” parallel to the villagers’ committees (Article 111). The Organic Law, Article 2, further defines residents’ committees as grassroots organizations that self-manage, self-educate, and self-service (also known as the “Three Selfs” [sanzi]). Second, residents’ committees are formed with five to nine members, composed of a director, vice-director, and committee members elected by residents either through direct election or by representatives (Organic Law, Articles 7 and 8). While the committee is elected and legally not part of the state bureaucracy, the Organic Law stipulates that municipal and district bureaus and departments may request help from residents’ committees and direct them in administrative duties29 (Article 20). Aside from being responsible to the government, within the residential living space, residents’ committees are also responsible to and must report their work to the residents’ council, composed of resident representatives (Article 10). Third, with regard to their functions, residents’ committees do not have any formal powers to govern or initiate their own policies and programs. Fundamentally, they are tasked to establish committees for civil dispute mediation, public security, and public health in order to “deal with” (banli) matters of public affairs and to assist in safeguarding public order. They are to act as the communication channel between residents and the government, conveying residents’ opinions and making suggestions to the people’s government (Constitution, Article 111). In essence, residents’ committees are an implementation organ rather than a power  29 There are at least ten laws that mention specific responsibilities of the residents’ committees, some of which are not even in the Organic Law. For example, the Marriage Law (2001) stipulates that if requested by victims of family abuse, residents’ committees have the legal obligation and right to step in to help and to mediate. Another example is that under the General Principles of the Civil Law (1986), for minors whose parents have passed away and are without grandparents or siblings, the parents’ work unit or residents’ committee becomes the temporary legal guardian until one is appointed.  68 organ of the state. To carry out their functions, they depend on government funds, as well as donations from residents and enterprises within their jurisdiction, and fees from convenience services they provide to residents (Organic Law, Articles 3, 16, and 17). The Constitution and the Organic Law do not accord residents’ committees much space for self-directed development. As the following chapter will address, the recent shequ reform seeks to change some of these limitations. The Appearance of Shequ: Modern Term, Old Concept In the Chinese Constitution and Organic Law, the jurisdiction under the residents’ committee is generically referred to as “residential districts” (juzhu diqu). Until the late 1980s, the word shequ remained a modern term primarily used in academia by sociologists and anthropologists in their study of rural villages, counties, and ethnic minorities.30 In comparison with the North American concept of community, two important distinctions can be made. First, shequ is place based: It refers to a neighbourhood community when used in an urban context and to a village community in the rural context. Second, as the beginning of this chapter explores, the word community in the English describes social relations based on personal ties and mutual dependence on which sense of belonging, solidarity, and identity are built; it stands as the antithesis to the functional and market-based transactions in the modern metropolis. The Chinese word shequ does not carry this connation. In Chinese, when one refers to the normative values of identity and cultural belonging, devoid of spatiality, the more precise word collective (gongtongti) is used. To further emphasize the specificity of the word shequ that is typically used as the translation for the English word community, this section discusses in greater detail the etymology and roots of the Chinese term. In other words, the interchangeable use of shequ and community is not taken lightly and is at best an approximation. The Chinese idea of a neighbourhood community speaks to a collective identity in a defined space. The etymology of the word shequ, or the two characters she (pronounced shè) and qu (pronounced chü), reflects this practice. Chinese characters are pictographic and characters can be broken down into parts. Put together, the characters are insights into a history and  30 The popularization of the term in Hong Kong and Taiwan took a different course with the influence of colonialization and continued development of sociological research and community planning post-1949, when mainland China closed its doors to the outside.  69 traditional way of life. The character she (社) is made up of the components shi (示), meaning religious and ancestral, and tu (土), meaning earth. Together, she literally means “the spirit of the land.” In the Chinese-English dictionary written and compiled by Herbert Giles (1912), a further explanation is provided that in the worship of this spirit, every li (里) (group of 25 families) had its own spirit to whom it made sacrifices (Giles 1912, 1191). Thus, li is known broadly as one’s hometown or village and adapted more specifically as a territorial grouping of households devised by the imperial state for the purpose of governance, as in the tithing lijia system. Considering the cultural and governing practices, she means a clan, society, village, or tribe that lives together on a territory and away from other clans, societies, villages, or tribes that worship other land spirits. The character qu (區 or simplified 区) means “district” (noun) or “to differentiate” (verb). The character’s etymology depicts a box (xi 匸) containing three objects or mouths or many objects and people (kou 口). The noun qu, generically, means a place, and it can vary in size between a region (as in quyu 区域) and a locality (diqu 地区). The verb qu means “to store away” (as in qucang 区藏) and” to differentiate” (as in qufen 区分). As a noun or verb, it connotes the assigning of things, land, or people to their proper place. Dictionaries published during the early Republican era did not yet contain a definition for shequ; the word shehui (社会) was its closest equivalent and meant the gathering of people who worship the same spirits and ancestors (Giles 1912, 1191). With the establishment of Western academic disciplines in Chinese universities in the early 1900s, the word shehui became the translation for the English word society which, similarly, means fellowship or an organization of people sharing a common cultural background. And, sociology became shehuixue, literally the study of society. Until recently, the word shequ was used primarily within sociological research to mean local society or to reference a particular social research methodology in the study of rural and non- Han ethnic society. With beginnings in the 1930s as sociology was gaining recognition as a  70 field of study,31 the growth of community studies (shequ yanjiu) as a subfield came at a time when Chinese sociologists were searching for an identity and intellectual independence.32 At this time, in addition to missionary colleges bringing foreign sociology teachers to China, many Chinese students who had gone abroad to study sociology also returned home. They had studied at leading institutions, such as Chicago, Columbia, and Harvard in the United States, and the London School of Economics in England. Upon their return to China, these individuals headed sociology departments and research institutes, influencing the field’s development at home. Wu Wenzao (1901–85), appointed in 1933 as the head of the Department of Sociology at Yenching University33 (later Beijing University), was a key figure in pushing for community studies. Wu, who had just returned after receiving his doctorate in sociology from Columbia University in 1928, believed that to build a Chinese sociology, theories had to be based on Chinese realities. The sentiment at the time among Chinese sociologists was neither a dismissal of Western theories nor an argument against the applicability of Western theories to China. Up to this point, Chinese sociology students were being introduced to Western ideas. However, they were interpreting them using traditional Chinese concepts rather than on the ground in present realities (Li et al. 1987). To make sociology Chinese involved using sociological theories and methodologies to observe China’s real social life and explain the reasons for the social problems that faced the new China. For Wu Wenzao, sociology should not simply be about conducting social surveys or systematic fact gathering on a particular theme. He was attracted to the concept of community studies (shequ yanjiu) because it  31 By the mid-1930s, 17 of China’s 41 universities had a sociology department (Wong 1979). Up until then, sociology, as a field of study, was offered primarily in the curricula and research of Christian colleges established by American missionary bodies.  The Christian colleges, operating outside the jurisdiction of Chinese education authorities, brought over American professors, many of whom were missionaries.  One of the most prominent of these American missionary sociologists was John Steward Burgess (1883–1949). Upon finishing his MA in Sociology at Columbia University in 1909, Burgess went to Beijing under the sponsorship of Princeton students and alumni as a YMCA secretary. In Beijing he taught sociology and Christian ethics in the School of Theology at Yenching University. He was instrumental in establishing the Department of Sociology and Social Work at Yenching and became its department head in 1922. For more about the missionary sociologists, see Wong (1979). 32 The earliest social surveys of China were conducted by missionary sociologists in English and published by American presses.  These social surveys tended to focus on the collection of social information for the purpose of supplying background for missionary work and were thus carried out with the aim of reform and social work.  In short, there were few original works produced in Chinese by Chinese sociologists.  For summaries of these early social surveys, see Sun (1949), Hsu (1931), Fried (1954), and Guldin (1994). 33 Yenching University was a Christian university and later became part of Beijing University when the People’s Republic of China was founded in 1949.  71 emphasized the comprehensive study of a locality, examining the intersections of institutions such as industry and family (Zhu 2006). At Yenching, Wu Wenzao advocated the theory and methodology of social ecology associated with the Chicago School. Robert Park spent time at Yenching lecturing and guiding students in fieldwork. In a lecture at Yenching, Park had emphasized to his students that “a society is not a community.” In translating this talk into Chinese for a commemorative publication of his visit to China, the term shequ was the result of some musing by Fei Xiaotong, then a student at Yenching, and his fellow classmates (Lu and Peng 2005, 124). At the time, community was translated as local society (difang shehui); however, in translating this particular sentence, they were met with the obscurity of saying “a society is not a local society.” In expressing the sociospatial organization central to urban ecology studies, the character she for "associations was joined with qu for the space the social group occupied (Fei 1948/1999, 531). For Chinese community researchers in the early twentieth century, community and society were not oppositional terms of premodern versus modern, such as conceptualized by Ferdinand Tönnies (1887/2001).34 Rather, community and society were an interrelated pair. Wu Wenzao writes: “The new perspective I would like to put forth is that it is through the eyes of community that one observes society … Society is an abstract concept that describes a collective way of life; it is a term for all the complexity of social relations. Community is a concrete term that describes the everyday realities of a people in a locality; it has material foundation and is observable” (Wu 1935, 66; cited in Wang 1996, 5; translated by author). It is unclear when the word shequ began to be adopted outside sociological research and when it became colloquial. It did not begin to appear in mainstream Chinese dictionaries until the early 1980s.35 Furthermore, a brief survey of sociology and urban planning academic  34 I have discussed a particular moment in time at Yenching University when community studies were popularized under Wu Wenzao who was the head of the sociology department.  For a detailed examination of the community studies research being conducted by Chinese sociologists during the first half of the twentieth century, see Hsu (1931), Sun (1949), Fried (1954), Freedman (1962), Wong (1979), Li et al. (1987), and Guldin (1994). 35 Shequ in mainstream dictionaries is defined as a group of people who live collectively in a defined locality, share collective identity and interests, have established social relations, and are organized in an  72 journals36 suggests that the revival of sociology and the scholarly exchanges between China and the West, as well as the appropriation of the word into political discourse to refer to a neighbourhood administrative unit, contributed to the popularization of the term shequ. Throughout the 1980s and 1990s, articles used the term shequ to mean human settlements, encompassing the spectrum from rural townships to urban neighbourhoods to city-regions. During this period, an increasingly large portion of the articles in sociology and planning journals was devoted to neighbourhood-based work, using phrases such as community planning (shequ guihua) and community services (shequ fuwu). In this literature, the word shequ is used in two main ways. First, authors referencing Western grassroots community development, such as community economic development practices and the role and history of neighbourhood houses, often chose to use shequ as a translation for the English words community and neighbourhood. Second, as professors and their students in sociology and social work became engaged in Shequ Construction initiatives as policy advisors and researchers, their published findings used the term shequ as it is used in policy documents. Observing China and a Chinese Perspective This chapter’s discussion has brought the inquiry of Chinese community governance into the present. The present state-led program to construct neighbourhood communities to undertake governance functions has deep roots in the long-standing practice of the state throughout Chinese history to define self-governing institutions at the grass roots. I have further sought to disengage Shequ Construction from community development as we understand it in North American planning theory and practice through exploration of the meanings, assumptions, and histories embedded in the word shequ. Ogden (2002), in drawing the attention of Western readers to our ethnocentrism in viewing the development of democracy in China, writes: “Understanding a society’s point of view and rationale for its actions is not the same as adopting its point of view, but we must at least try to understand a society’s ‘self description,’ even if it is confused and contradictory” (29). Like democracy, acts of community building and self-governance are socially constructed and continually evolving.  association with activities and management (Ministry of Education Revised Chinese Dictionary Compilation Committee 1981, 4060–1; Cihai Compilation Committee 1988, 4136–7). 36 My brief survey of the Chinese literature is based on “shequ”-titled articles published from 1979 to 1999 in the journals Society (Shehui), Sociological Review (Shehuixue Yanjiu), City Planning Review (Chengshi Gueihua), and City Planning Academic Academic Review (Chengshi Guihua Xuekan).  My main interest is in gauging the meaning the authors implied in choosing to use the word “shequ.”  73 Five endogenous characteristics of Chinese community-based governance remain central in the contemporary articulation under Shequ Construction: 1. The neighbourhood unit is a system of ordering the population spatially by households. 2. Self-governance refers to governing functions undertaken by local leaders, whether they are self-selected elites, voted, or chosen by local officials. Hence, one determining criterion of self-governing is that the leadership comes from within the locality. 3. Grassroots organizations play a supplementary role in assisting the formal government with administrative tasks as well as keeping the bureaucracy in touch with the concerns of the people. 4. Self-governance speaks to meeting local needs through initiatives of local leaders drawing on resources available to them. 5. Grassroots governance encompasses a moral dimension that places emphasis on taking care of one’s own and living in harmony.  I begin my examination of the post-reform community policy program with the recognition that these characteristics are inherent in the Chinese approach. These social practices, shaped over time, must be taken to be the terms on which the present community-building initiatives are based. As the historian Madeleine Yue Dong (2003, 15) describes the inseparability of everyday life practices from modern planning ambitions, the past persists in the present and each era recycles practices from the past, reinventing and transforming elements that are useful for its survival in the present.     74 4. SHEQU JIANSHE China’s Community Construction Policy Agenda  An Experiment’s Experiment Shequ Jianshe, while state initiated and led, remains an experiment, open to and dependent on local pilot programs and innovations to define and shape it. This chapter’s intention is to provide a broad overview of national economic and social circumstances and the central government’s stance toward Shequ Construction. Three paradoxes complicate China’s seemingly top-down approach to community. First, because it is situated within the administrative hierarchy and integrated into the state apparatus, Shequ Jianshe is inherently top-down. The specifics of the community agenda point to it being a project to contend with social issues arising from the dismantling of the danwei-based socialist welfare system that may elevate popular discontent and threaten the Party-state’s legitimacy. However, we must also recognize that life in China today is ever-changing. The policy program, however dogmatic, must contend with the needs of an increasingly diverse and demanding society. Residents’ committees have little coercive powers over residents and must find a balance between maintaining social control and providing constituents with a greater level of service. Second, while Shequ Jianshe appears to be regulatory, it remains ideational in what it seeks to change. Policy documents concretely state what a shequ is and the makeup of its leadership and responsibilities. Nationwide, over 80,000 neighbourhood units have been created (2006 figure; NSB Office of Social and Technological Statistics 2007, table 9-20). However, aside from the material construction of community, the goals of harmonious shequ and grassroots self-governance are largely achieved through regulating the residents’ committee and at best creating the conditions that will shape habits and educate people regarding desired social values and norms. Third, like most policies in China, Shequ Jianshe relies on local experimentation. It began as a grassroots effort to provide social services to society’s most vulnerable population groups in regions heavily impacted by economic restructuring. From this perspective, the policy  75 could be viewed as the culmination of bottom-up initiatives. Through isolated local experiments, the “community construction” concept gained endorsements from higher level officials. Not only did it relieve some of the increasing burden on local governments, it was seen as a means to curb potential social unrest as unemployment rose and social services were transferred out of work units. Even in China’s authoritarian-bureaucratic context, the policy-making process is an adaptive one that depends on locally generated solutions. As Heilmann (2008, 3) argues, “though ambitious central state planning, grand technocratic modernization schemes, and mega projects have never disappeared from the Chinese policy agenda, an entrenched process of experimentation that precedes the enactment of many national policies has served as a powerful correcting mechanism.” This chapter proceeds in three sections. I begin by summarizing the pressing social issues in the early reform period that prompted shequ reform ideas and the eventual policy formulation. Next I examine the experimentation process, highlighting policy-making strategies that have embedded within them certain levels of flexibility and ambiguity. Then I turn to the contents of Document 23, the memorandum issued by the State Council that pushed shequ reform beyond pilot sites toward nationwide adoption. In this translation, I aim to show the normative values of the good society the policy prescribes as it seeks to affect individual behaviour and structure a post-Mao reform-era social order. At the same time, the document is but a stage in the experimentation-based policy-making process, setting the conditions for further decentralized experimentation through local interpretation and implementation. This crucial tension between commanding directives and local-serving adaptability is observable in each of the policy conjunctures explored in the subsequent chapters. Changing Conditions: 1980s to 1990s Local shequ experiments and political support for shequ policies were in large part a response to changing social conditions unfolding in the late 1980s and 1990s. This section outlines five growing areas of concern in cities nationwide: the restructuring of state-owned enterprises, the subsequent increase in urban unemployment, the growing need for social assistance, issues of managing rural migrants, and concerns over the diminishing presence of the Party at the grass roots. Figure 4.1 shows on a timeline the significant events and policies mentioned in this chapter. Slow down of labour reforms Initial stage of reforms Greater SOE managerial autonomy Market reforms reinvigorated and intensified Social & economic "scientific development" Deng's "southern tour" 16th Party Congress, Hu Jintao named Party secretary Rising inflation SOE Managerial autonomy to hire & dismiss workers Legal framework for SOE restructuring adopted growth of TVEs, competition with SOEs Intergovernmental tax sharing reform first of consecutive No. 1 Documents on rural reform changing policies toward rural to urban labour migration State Council housing reform decision 11th Five Year Plan (2006-10) rising inflation surge in unemployment Harmonious Society as policy direction 3rd plenum of 11th Party Congress Deng announces economic reform programs Tiananmen Square Protests Deng passes away Jiang Zemin named Party Secretary 15th Party Congress: "grasping the large, letting go the small" First Special Economic Zone established in Shenzhen Austerity campaign housing monetarization, welfare allocation of housing ceased One-child policy adopted aim to slow economic reform creation of minfei category Rural decollectivization SOEs pressured to increase labour demand 50 years of CCP rule celebrated Falun Gong practitioners' protest 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 Nanjing harmonious shequ project Qingdao Conference on shequ services Nanjing Village Shequ Construction pilots 14 state agencies jointly issue support for shequ services Document 14 reflects scientific Hangzhou Conference on Shequ Construction & harmonious development Organic Law on Residents' Committee adopted 148 districts selected as Shequ Construction demonstration sites Shequ Construction incorporated into 10th five year plan Wuhan Conference on shequ services Symposium on urban social welfare institutions Document 23 promulgates Shequ Construction nationwide increasing role of residents' committees raised 26 districts selected as Shequ Construction pilots Shequ services Greater experimental autonomy Deepening of Shequ Construction contents Shequ Services to Shequ Construction Harmonious Shequ Construction Expansion to rural villages Figure 4.1 Timeline of key events Source: information on national context and SOE reform compiled from Naughton (2003, 4-8); Fernandez-Stembridge (2003, 58); Rocca (2003, 83) S h e q u  C o n s t r u c t i o n  k e y  e v e n t s N a t i o n a l  C o n t e x t W a v e s  o f  R e f o r m P h a s e s  o f  S h e q u  R e f o r m  77 State enterprise restructuring and danwei-based welfare reform Shequ Construction is part of a series of social welfare reforms that accompanied the withering of the danwei-based socialist employment system. At the end of the Cultural Revolution in 1976, SOEs came under great pressure to create jobs for the generation of urban youth who had been sent to the countryside and were now returning home. In spite of this, increased demand for industrial production created a demand for labour, and the number of SOE employees slowly rose.37 However, with the economy being pushed into further marketization, conditions were volatile and by the late 1980s many SOEs were struggling. Rural decollectivization and the new household responsibility system produced a large labour surplus that contributed to the rise of township and village enterprises (Ma and Fan 1994; Guldin 1997; Oi 1999; Unger 2002). Consequently, not only did SOEs lose the monopoly in industrial production, but rural enterprises became a more efficient competitor. Whereas township and village enterprises were solely responsible for production output, SOEs continued to provide social services and welfare to their workers (Fernandez-Stembridge 2003, 59). Work-unit-based welfare reform proceeded slowly, and throughout the 1980s the institutional framework for urban welfare provision remained basically unchanged (Gu 2001). In the late 1980s, as China’s inflation rose and anxiety grew among Party elites that rising food prices could undermine social stability, SOEs had to contend with continued pressure from above to increase employment, despite their relatively high labour costs. Consequently, early reform efforts that for the first time allowed managers to claim a share of profits and to hire and dismiss workers according to production needs and performance did little to improve SOE efficiency and profitability. Many enterprises continued to require loans and bailouts from state-owned banks. There was growing fear that the continuing nonperforming loans would eventually bring about the collapse of the financial system. The economy could not continue to operate with the dual track system of central planning and market that was driving up inflation and was susceptible to corruption (Fernandez-Stembridge 2003, 59). A more competitive system was needed.  37 Naughton (2003, 13) makes the important and often overlooked point that while economic reform policies did result in public sector layoffs, it is essential to break down economic reform years to really understand the impact of the 1990s restructuring. Labour statistics show that in the first 15 years of reform, from 1978 to 1993, state-sector employment continued to grow from 75 million to 109 million employees. Industrial employees increased from 31 million to 45 million. The figures begin to decline in 1994 and drop sharply in 1998 (National Statistics Bureau 2006, table 5-4).  78 Beginning in 1993, under the new slogan “Socialism with Chinese Characteristics” that stemmed from Deng Xiaoping’s 1992 trip to southern China,38 a series of laws and policy initiatives ushered in a new wave of SOE restructuring. In essence, the term restructuring refers to revising the ownership structure of the bureaucratically run SOEs with the objective of making their operation more efficient, competitive, and financially accountable. Those that were successful were allowed to thrive and those that were bankrupt were abolished. SOEs were converted into joint stock corporations; workers and managers were offered the opportunity to buy shares, and successful enterprises were acquired or merged with failing firms. In other cases, state assets were privatized through auctions or the sale of bankrupt firms (Naughton 2003, 10-11). In the course of restructuring, reforms to the workplace-based welfare system were also initiated to shed SOEs of welfare responsibilities, significantly redefining the danwei’s traditional role in the lives of urban workers and their families. Urbanites had been largely dependent on their work unit – for employment as well as for welfare provisions, pensions, health care, and housing. The major changes comprised, first, a compulsory labour contract to govern the relationship between workers and employers which abolished the system of permanent, lifelong employment. Second, social pension and health insurance schemes transferred responsibility from state to individual, using a combination of individual accounts and social pooling funds. With experiments since the early 1990s, the state not only hoped to relieve SOEs of the burden, but also to encourage labour mobility from the state to the nonstate sector (Gu 2001, 98). Welfare benefits are now paid out to workers as a percentage of their total wages. Work units and workers contribute jointly, with workers paying into an individual account and the work unit paying half into individual accounts and half into a social pooling fund managed by the local government (Thelle 2003, 168-9). Third, in 1994 the State Council introduced a new set of housing reform policies and further promoted the selling of public-sector housing to employees. Workers were encouraged to  38 The Tiananmen Square protests of June 1989 and the state’s handling of the incidents posed significant challenges to Deng Xiaoping’s power. The conservative faction that was forming within the Communist Party increasingly questioned and criticized his reformist platform. His 1992 Southern Tour (nanxun) is regarded as a reassertion of his economic reform policies, intended to garner support in the rapidly developing southern provinces. His speeches called for bolder reform initiatives and ushered in a second bout of growth with raised targets and the development of the nonstate sector (Fewsmith 2001, chapter 2).  79 purchase their home from their danwei at a discounted price. In 1997 a new state policy sought to cease material allocation of public housing to urban employees. A compulsory housing savings system, the Housing Provident Fund, was established to facilitate housing purchases, particularly commodity housing on the real estate market. Rather than allocating housing in kind, a housing subsidy is now part of workers’ remuneration and is deposited into workers’ individual accounts (Wang and Murie 2000; Lau and Lee 2001; Thelle 2003, chapter 5). The specificities of how the housing savings system operated varied by locality, dependent on local socioeconomic conditions and living standards (Thelle 2003, 134-41). Due to a confluence of timing (i.e., the Asian financial crisis in the late 1990s), accumulated debts, outdated equipment, and mismanagement, the early restructuring efforts did not unfold as intended. SOE profits continued to decline, particularly in small- and medium-sized enterprises. About half of the SOEs posted losses in 1997 (Cai 2006, 15). Small SOEs that were still earning 20 billion RMB in profits in 1993 incurred a deficit of 20 billion RMB in 1997 (Naughton 2003, 9). Despite these realities, the central government strengthened its resolve to continue with restructuring efforts and maintained a greater tolerance for layoffs. Jiang Zemin’s address to the fifteenth Party Congress in September 1997 called for persevering with SOE restructuring: We shall convert large and medium-sized SOEs into standard corporations according to the requirements of “clearly established ownership,” well-defined power and responsibility, [and] separation of enterprise from administration …We shall also quicken the pace in relaxing control over small SOEs and invigorating them by way of reorganization, association, merger, leasing, contract operation, joint stock partnership or sell-off … We should encourage mergers, standardize bankruptcy procedures, divert laid-off workers, increase efficiency by downsizing staff and encourage reemployment projects …  (Jiang 1997; translation by Beijing Review) A major focus of the programs initiated after the Congress centred on increasing enterprise efficiency through reducing workforces and payroll costs. One example was the “Work Conference on Basic Livelihood Protection and Re-employment of Laid-off workers in SOEs” jointly convened by the CCP Central Committee and the State Council in May 1998.  80 From this initiative came sanctioned procedures through which firms could lay off workers and drafted reemployment and compensation programs to assist them (Hurst 2009, 65-66). Rising unemployment and urban poverty With SOE restructuring, the right to permanent employment was no longer an assumed benefit of urban residency status. Relationships between workers and enterprises were to be governed by labour contracts that could be dissolved by either employee or employer. Between a high rate of bankruptcy and severe streamlining to generate profits, the restructuring process dramatically shrank public sector employment, giving rise to massive layoffs and urban unemployment. Prior to restructuring, the vast majority of all urban employment was in the state sector. As summarized in table 4.1, in 1995, even as restructuring was taking place, the number of people working in the state sector decreased slightly to 59% of the total urban work force. However, following the hard line taken toward restructuring following the fifteenth Party Congress, this number rapidly shrank to 38% in 1999 and to 32% in 2001. The new category of laid-off workers (xiagang zhigong) described workers who had gone through a formal laying-off process during which they no longer worked for, but remained affiliated with, the work unit. Figures on laid-off workers vary from source to source.  Here, I use official figures, keeping in mind that the population tends to be undercounted (Solinger 2001; Naughton 2003, 15). Beginning in 1997, the Ministry of Labour and Social Security reported the number of newly laid-off workers registered with reemployment centres.39 According to national statistical figures, the cumulative number of laid-off workers from SOEs during the period 1997 to 2001 was about 25.5 million, or 18% of the SOE workforce in 1997 (table 4.1). During this period, on average, reemployment centres annually reported a  39 Reemployment centres are different from local employment centres. In 1998, after experimentation in Shanghai, the central government made it compulsory for SOEs laying off workers to establish reemployment centres. They were designed as an intermediary step to help workers adjust to their new circumstance. Registered laid-off workers would receive subsidies and training for 2 to 3 years. If they still had not managed to secure employment during this period, they would leave the centre but continue to receive unemployment insurance for up to two more years. After that, if they remained unemployed, they would fall into the care of social welfare agencies and receive minimum allowances. The costs were split three ways between the firm, the local government, and insurance organizations. For analysis of the reemployment centres, see Fernandez-Stembridge (2003, 60-64), Cai (2006, 20-23), and Hurst (2009, 66).  81 year-end figure of 6 million laid-off workers who were still looking for new positions.40 These are rather conservative numbers because they reflect only workers who had gone through the formal laying-off process. If layoffs from bankrupt and near-bankrupt enterprises and collectively-owned enterprises were accounted, the number of urban families impacted would be much higher (Solinger 2001, 684). Wages under the planned economy had been low, but urban poverty was limited to the few who were denied danwei employment. The unemployed and laid-off workers make up the majority of the new urban poor who are able and willing to work but have no jobs (Hussain 2003, 1). The official unemployment rate has remained relatively constant at 2 to 3% of the urban work force. Including the laid-off workers, the reported unemployment rate in cities is around 5 to 6%. Given the narrow definition of “laid off” and “unemployed” in official statistics, this is, once again, an underestimate of actual numbers. Surveys conducted in specific cities to better capture a more realistic account reported much higher unemployment rates. For instance, counting those who were not working and looking for work, Giles, Park, and Cai (2006, table 2) reported an unemployment rate in 2001 of close to 17% in Wuhan, over 14% in Shenyang, and over 10% in Shanghai and Xian. The nascent unemployment insurance program and pension system, riddled with problems of arrears, under funding, and narrow coverage, did little to alleviate the adverse impact of restructuring on urban workers. With economic hardship now becoming a reality for many, feelings of uncertainty, vulnerability, and anger increasingly were expressed in public demonstrations, petitions, and attacks on government buildings (Rocca 2003, 83; Cai 2006, chapter 3).   40 After 2000, the figures decline rapidly (table 4.1). Reemployment centres were deemed by policymakers as a failed initiative. At the end of 2000 no new workers were admitted and all the centres were closed by 2003. Laid-off workers were reclassified under the general category of unemployed (Hurst 2009, 66).  82  Table 4.1 Employment statistics for years of concentrated SOE restructuring  (in million persons) Year 1992 1995 1997 1998 1999 2000 2001 Employment in SOE (1) Urban employed person 178.61 190.40 207.81 216.16 224.12 231.51 239.40 In state-owned 108.89 112.61 110.44 90.58 85.72 81.02 76.40 % of urban employed 61% 59 53 42 38 35 32 Laid-off Workers From SOE at year end (2) 6.92 5.92 6.53 6.57 5.15 From SOE added this year (2)  5.62 6.19 4.45 2.34 Laid-off workers not SOE (2) 2.80 2.85 2.54 2.26 Registered unemployed (3) 3.64 5.20 5.77 5.71 5.75 5.95 6.81 Total urban unemployed 3.64 5.20 12.69 11.63 12.28 12.52 11.96 % of urban workers 2% 3 6 5 5 5 5 Note: The figures for SOE include only state-owned enterprises and exclude collectively-owned enterprises. "Total urban unemployed" is calculated from adding laid-off workers at year end and registered unemployed. "Unemployment rate" is the "total urban unemployed" as percentage of employed, laid-off, and registered unemployed persons.  Source: (1) National Statistics Bureau, China Statistical Yearbook 2006, table 5-4. (2) Ministry of Labour and Social Security, China Labour Statistical Yearbook 2000, table 8-1; 2001, table 8-1; 2002, table 2-6; 2003, table 2-6, table 2-7. (3) Ministry of Labour and Social Security, China Labour Statistical Yearbook 2002, table 2-2.   Urban social assistance programs and trends towards welfare socialization The unleashing of such a transformative restructuring process resulted in large-scale layoffs and left an immediate gap in welfare provision. In urban areas, the number of people receiving subsistence allowance through the government’s Minimum Living Standards Program (dibao) increased fivefold during the period from 2000 to 2005 (NSB Office of Social and Technological Statistics 2007, table 9-2). The hardships for laid-off workers were compounded by the lack of an universal welfare system. Even though the state slogan encouraged urbanites to transform their identity from that of a “work unit person” (danwei ren) to a self-reliant “society person” (shehui ren), welfare spending in the government sector had not kept up to create a safety net to support this transformation. As socialist ideology had  83 encouraged the right to welfare through work, the state had provided subsidies through work units rather than investing in direct welfare programs.41 Government spending on relief work was for civil affairs programs such as veterans’ aid, disaster relief, and aid for the small percentage of the unemployed who were without family members to depend on. In most years from 1949 to 1995, civil affairs expenditures (including pensions, social welfare, income assistance, and natural disaster relief) accounted for less than 2% of the total state budgetary expenditures (L. Wong 1998, 149). Despite increased demand for social assistance due to the large-scale layoffs welfare spending since 1995 has remained at around 2% to 3% of total budgetary expenditures (National Statistics Bureau 2006, table 8-4). As the next chapter discusses at greater length, the unchanged proportion of welfare spending reflects the shift in responsibility onto lower levels of government, and their dependency on off-budgetary revenues to fund social services. The retreat of the central state from welfare responsibility through separating the welfare provisioning functions of SOEs from their business activities became known as “social welfare socialization” (shehui fuli shehuihua) and “diversification” (duocengci; Thelle, 2003, 37).  Socialization transfers responsibilities from the state to society, that is, residents’ committees, social organizations, enterprises, families, and individuals. Diversification devolves responsibilities to lower levels of government. Wong (1998) sees this reduction in state provision and funding as indicators of social welfare privatization. Under the rubric of marketization and socialization, the author points out, the Chinese government has “openly espoused fee charging, community care and informal care, as well as reforms in financing and management” (155).  However, Thelle (2003) views socialization and diversification as alternate forms of collectivization because the community has assumed duties from the state, and part of the aim “seemed to be to instil a feeling of togetherness and belonging from above, maybe in recognition of the low esteem held of the Communist Party committees in many places” (173).  41 It is important to note here that welfare under the Ministry of Civil Affairs differs from social security under the Ministry of Labour and Social Security. The former provides assistance to marginalized people and oversees the development of community services for everyone. The latter is for those affiliated with a work unit (Thelle 2003, 167). Organized through work units, labour insurance benefits for workers had included occupational and nonoccupational disability pensions and medical care, retirement pensions, and medical care for dependents (Dixon 1981, appendix 4.1).  84 Management of the floating population Alongside enterprise restructuring, a separate series of policy decisions gradually loosened the restrictions placed on rural to urban migration (Solinger 1999, 45-55). The term “floating population” (liudong renkou) refers to migrants who reside outside their place of permanent residence registration without undergoing official process of registration transfer. They are typically from the countryside and “float” between cities and their home place. By the early 1990, it was estimated that between 50 and 60 million rural migrants were working outside of their townships (Mallee, 2000, 91), and the large scale rural to urban migration that had been taking place began to receive widespread attention in the urban media, particularly before and after the Lunar New Year holidays when migrants travelled to and from cities back to the countryside. Images of their massive numbers crowding the train stations gave rise to the phrase “migrant worker tidal wave” (mingong chao), which describes the scale of and sense of panic in cities toward the rural labour migration (ibid.). It was neither possible nor desirable for the Chinese authorities to block rural to urban migration as the low costs of rural labour were attractive to both state-owned and foreign- invested enterprises (Solinger 1999, 48-55). By the mid-1990s, the size of the migrant population had grown rapidly to an estimated 100 million (Chan 2008, 5). Migrant workers had become the backbone of the country’s export-led manufacturing sector, comprising as much as 70 to 80% of the workforce in industrial coastal cities such as Shenzhen and Dongguan (Chan 2009, 207). As their number continued to increase, migrants in cities, many of whom were undocumented and beyond the state’s reach, were regarded by officials as posing a threat to government control and ultimately the regime’s ruling power (Li 1998, 19). According to opinion polls conducted at the time in large cities, the migrant population were perceived by urban residents to be a critical problem for public security (Solinger 1999, 131). Over the last two decades, the state has attempted to tighten its management over the migrant population through various hukou reforms and registration policies (Wang 2004; Chan 2009). And, local urban governments have carried out “clean-up campaigns” that mobilize police to arrest and repatriate migrants without proper documents. The day-to-day management of the migrant population has been the responsibility of public security bureaus, neighbourhood police stations, and residents’ committees. It is at this lower level that the local government concentrates its efforts. The local public security bureaus are  85 responsible for hukou registrations. Every neighbourhood police station has a full-time hukou officer assigned to collect, verify, and update not only the registration of the residents in the jurisdiction, but also to record their political activities, financial status, personal friends and family relations, and hobbies. Relying on the assistance of residents’ committees, it is the hukou officer’s job to get to know those who live in the jurisdiction and to report those who do not belong there and those who threaten national or public security (Wang 2004, 124-5). Party building and keeping in touch with the masses Finally, the support for shequ reform must be situated within the political climate after the 1989 Tiananmen Square events. The demonstrations of spring 1989 were a reaction of deep social dissatisfaction. The urban industrial enterprise reforms that sought to increase competition and expand market mechanisms brought high inflation, elite corruption, a widening income gap, public angst over job security, and a split in the Party leadership regarding the direction of economic reform (Liang, Ling, and Nathan 2001, 3-18). The event was a severe test and wake up call for the leaders of the Party-state. It demonstrated their lack of awareness toward the extent of social discontentment and the diminishing presence of the CCP at the grass roots. For instance, a few weeks afterwards, in his report to the fourth plenum of the thirteenth Central Committee, then Premier Li Peng stressed the “need to make a solemn resolution to overcome the tendency of the Party and government to drift far from the masses” (Liang, Ling, and Nathan 2001, 440). At the sixth plenum the following spring in 1990, the CCP published an open directive on strengthening the deteriorating relationship between the Party and the masses. Underneath the rhetoric, there was a concern for building the Party’s capacity to influence the grass roots, reminiscent of the “mass line.”42 The directive called for cadres at the county level and above to spend time at the grass roots in order to understand the needs and difficulties of the people, propagate policies, engage in political thought work, and take part in labour (People’s Daily 1990, section 3.3; translated by author). Recognizing that  42 The mass line was a means of political mobilization developed in Yanan to strengthen the relationship between the CCP and the masses. Mao stated that “all correct leadership is necessarily ‘from the masses, to the masses’. This means: take the idea of the masses (scattered and unsystematic ideas) and concentrate them (through study turn them into concentrated and systematic ideas), then go to the masses and propagate and explain these ideas until the masses embrace them as their own, hold fast to them and translate them into action” (Mao 1943/1967, 3: 117-119). Since the Yanan period, the tactic has influenced other CCP practices, such as the danwei system (Bray 2005, 55-58).  86 many base-level Party organizations are ineffective and non-functioning, it further called on lower-level cadres to connect with and organize people (ibid., section 7; translated by author). In this context, Bray (2006) observes that in addition to emerging challenges of welfare provision and policing the influxes of rural migrants, shequ building is linked to the wider project of strengthening the CCP, or Party building (dangjian), at the grassroots (535). Aside from welfare provision, the concept of community services also served the political function of building cohesion and social stability through mutual help. The then Civil Affairs Minister Cui Naifu asserted that “to develop the work of community services is useful in regulating human relations, solving social problems, creating a harmonious social environment and realizing the guiding thoughts of serving the work of the Party centre through the work of civil affairs” (Social Security News 25 September 1987, cited in Wong 1998, 127). From Services to Construction: Two Decades of Shequ Experiments Through the 1990s, the increasing number of welfare recipients and the complexity of new programs were beginning to overstretch the capacity of staff at the district and street office levels and the untrained members of the residents’ committees. As local governments searched for solutions, the limitations and potentials of street offices and residents’ committees as providers of social welfare presented opportunities for experimentation. Innovations through two decades of localized experimental programs played a crucial role in formulating Document 23 which promulgated the Shequ Construction initative nationwide. A review of early documents suggests that organizational reforms to the old residents’ committee were not apparent at the outset. At the eighth national work meeting of the Ministry of Civil Affairs (MCA) in 1983, only the broad principle of mobilizing societal resources in providing social welfare was raised (Wong and Poon 2005, 418). Central leadership called for reform and a new line of thinking for welfare pluralization to relieve the state as sole provider, pointing toward shared responsibility among the state, work units, and individuals. The meeting confirmed and spurred various local initiatives with some involving residents’ committees. Shequ reform experiments reflect the importance of locally generated ideas in China’s policy-making process. The process involves a “point to surface” (youdian daomian) approach that begins when the central state recognizes successful local innovations and  87 solutions. Demonstration sites with more diversity in local circumstances are then selected for pilot testing. Eventually, based on research findings by advisors and on conferences where local officials exchange ideas and experiences, provisional rules and regulations are revised and a policy program is formulated and disseminated for broader implementation. Rather than being bottom-up or top-down, this process is better described as a feedback loop between local and central governments (Heilmann 2008). This ideational diffusion and its influence on the policy development path and content are observed in a wide range of reform areas, including special economic zones, administrative reform, and rural health care programs (Foster 2006; Heilmann 2008). In legislation-centred liberal democracies, potential policy impacts are usually assessed prior to enacting laws and regulations. In contrast, since the revolutionary era in China, discretionary experimentation through on-the-ground implementation has typically occurred before legislation (Heilmann 2008, 9). Below, I outline Shequ Construction’s development through the 1980s and 1990s as three phases, emphasizing the continual interchange between the centre and localities and the inseparability of bottom-up and top-down. Key documents and circulars are chronicled in appendix 2 with listings of demonstration cities and districts. Phase 1: Shequ services (1983 to 1992) After the 1983 MCA work meeting, localities took up the call to involve nonstate sectors in the provision of welfare services. In 1984, the Ministry convened a national conference in Zhangzhou, Fujian Province, bringing together civil affairs officials from various localities to share initial experiences with urban welfare reform, specifically the transferring of welfare services from work units to local governments. In 1986 MCA first raised the concept of shequ fuwu, or community services which focused on welfare delivery at the neighbourhood level instead of at the municipal and district levels (Chan 1993, 28). Street offices in Beijing and Wuhan, serving as pilots, were required to establish a range of services (ibid., 41 fn10). A year later, MCA organized the First National Symposium on Community Services in Urban Areas in Wuhan. This meeting of local officials prompted many large cities to plan shequ services and open pilot care facilities. Two years later, at the 1989 Hangzhou Conference, officials from more localities came together to exchange lessons learned, discuss challenges, and affirm the role of community-based social services in the reform-era. Even though the MCA requested municipalities to undertake community services, the existing residents’ committees lacked the skills to handle not only the increased demand for welfare  88 services but the new conditions reform policies created. By the end of the 1980s, in some cities new shequ committees had either replaced or were being established alongside existing residents’ committees (Derleth and Koldyk 2004, 750). Then, on December 26, 1989, at the eleventh plenum of the seventh National People’s Congress, the Organic Law was approved and adopted. Reflecting the experiments, it specifically called for residents’ committees to engage in shequ service activities: “Residents’ committees shall develop shequ services that provide convenience and benefits to residents, as well as initiate the development of related service enterprises. Residents’ committees shall manage their own financial affairs and no departments or units shall infringe upon the assets of residents’ committees” (Article 4, translated by author). As the next section will discuss, this directive motivated residents’ committees to operate convenience stores and fee-for- service programs, transforming residents’ committees into service providers rather than primarily keepers of social order (Choate 1998, 11). Phase 2: Greater experimental autonomy (1993 to 1997) MCA continued to hold meetings and conferences for municipalities to exchange experiences. Much confusion remained at the neighbourhood level as to what “shequ services” entailed, and much direction was needed from municipal and district civil affairs officials. Only in larger, more affluent cities like Beijing and Shanghai were plans prepared and adopted expeditiously (Wong 1998, 129). During this phase, MCA enlarged its efforts to expand community services by changing the neighbourhood administrative structure and giving shequ residents’ committees more autonomy. The first of such efforts was the 1993 policy paper Memorandum on Accelerating Shequ Service Operations. Jointly issued