Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

On the edge : a history of livelihood and land politics on the margins of Hà Nội Labbé, Danielle 2011

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

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata


24-ubc_2011_fall_labbe_danielle.pdf [ 3.79MB ]
JSON: 24-1.0072256.json
JSON-LD: 24-1.0072256-ld.json
RDF/XML (Pretty): 24-1.0072256-rdf.xml
RDF/JSON: 24-1.0072256-rdf.json
Turtle: 24-1.0072256-turtle.txt
N-Triples: 24-1.0072256-rdf-ntriples.txt
Original Record: 24-1.0072256-source.json
Full Text

Full Text

  On the Edge: A History of Livelihood and Land Politics on the Margins of Hà Nội       by  DANIELLE LABBÉ  B.Arch., Université Laval, 2001 M.Sc., Université Laval, 2004    A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY  in  THE FACULTY OF GRADUATE STUDIES  (Planning)      THE UNIVERSITY OF BRITISH COLUMBIA  (Vancouver)     September 2011    © Danielle Labbé, 2011  ii   ABSTRACT   This dissertation takes an historical approach to explore the territorial formation processes on the near periphery of Hà Nội, the capital city of Việt Nam. It relies on methods inspired by the ethnographic tradition to document how a small locality, named Hòa Mục, has made the shift from rural village to urban neighbourhood over the course of the 20 th  century. The analysis centres on the evolution of villagers‟ livelihoods and land strategies in relation to the ebb and flow of state regulations and territorialization projects. Secondary literature and policy documents contextualize this micro-study and position it within the wider framework of socio-political and institutional changes in Northern Vietnam. The results are presented chronologically along four broad historical stages: i) the late colonial era (1920-1940), ii) the socialist revolutionary transformation process (1940-1960), iii) the anti-American war and subsidy era (1960-1980), and iv) the đổi mới reforms (1980-2009).  By placing the periurban formation process in a longer historical context, the study shows that some territorial orders from the pre-reform period have travelled across different political-economic regimes and thus continue to influence the ongoing urban transition. This provides an important counterpoint to understandings of state policy as key determinants of urban change in contemporary Việt Nam. The discussion instead shows how local practices and norms interact with the state‟s regulatory function to shape the periurbanization process. As part of this dynamic system, the state responds in flexible ways to territorial claims from the grassroots and to emerging socio-spatial configurations on the urban edge. The case of Hòa Mục, thus indicates that the state can and does rely on systems of exceptionalism, deliberate institutional ambiguity, and the selective reproduction of informality to govern urbanization on the edge of the capital. In a context like that of Việt Nam, this suggests the need to enlarge the repertoire of what we call planning activities.  iii   PREFACE   Two chapters in the present dissertation include material from published and forthcoming research articles. Chapter 6 selectively borrows elements from a co-authored paper entitled “Understanding the Causes of Urban Fragmentation in Hanoi: The Case of New Urban Areas,” published in the International Development Planning Review (see Labbé and Boudreau 2011). I collected and analyzed all of the data presented in that chapter. The contributions of my co-author, Julie-Anne Boudreau, were limited to revisions of the prose and to theoretical discussions which are not included in this dissertation. Chapter 7 is largely based on a single-authored article entitled “Urban Destruction and Land Disputes in Periurban Hanoi during the Late-Socialist Period” to be published in the September 2011 issue of Pacific Affairs (see Labbé 2011b). I contributed all the data and did all the analysis and writing leading to this publication. The research for this dissertation was reviewed and approved by the Behavioural Research Ethics Board of the University of British- Columbia. The Certificate of Approval (minimal risk) number for this project is: H08-02563.  iv   TABLE OF CONTENTS   ABSTRACT ...................................................................................................................... ii PREFACE ........................................................................................................................ iii TABLE OF CONTENTS .................................................................................................. iv LIST OF TABLES ............................................................................................................ vi LIST OF FIGURES............................................................................................................. vii LIST OF ABBREVIATIONS ......................................................................................... viii ACKNOWLEDGMENTS......................................................................................................ix CHAPTER 1: Introduction ................................................................................................. 1 Before and Beyond Đổi Mới: Revisiting the Urban Transition in Việt Nam ..................... 4 The Periurban as “New Urban Frontiers” ............................................................................. 9 The State-in-Society ............................................................................................................. 12 Planning and Regulatory Informality .................................................................................. 14 Research Focus and Scope ................................................................................................... 16 CHAPTER 2: Bending Research Expectations into Hà Nội‟s Environment ...................... 19 Research Style Reconfigured............................................................................................... 20 Contingencies of Selection and Access .............................................................................. 22 Negotiating the Field: Sampling and Data Generation ...................................................... 27 From Analysis to Uploading Discourse into Writing ........................................................ 32 Conclusion ............................................................................................................................ 34 CHAPTER 3: The Early Urban Transition (1920-1940) ................................................... 36 A Small but Independent Commune-Village ..................................................................... 38 An Economy of Complemented Agriculture ...................................................................... 43 The Great Transformation ................................................................................................... 47 Conclusion ............................................................................................................................ 54 CHAPTER 4: Uneven Socialist Revolutions (1940-1965) ................................................ 57 The Outskirts as a Productive Belt ...................................................................................... 59 A Partial Land Reform ......................................................................................................... 61 Socialist Economic and Institutional Re-formations.......................................................... 67 Làm Nội, Làm Ngoại: Working Inside, Working Outside ................................................. 73 Conclusion ............................................................................................................................ 76 CHAPTER 5: Eating by Point and Coupons is not enough (1965-1980) ........................... 79 Legislative Unevenness ....................................................................................................... 82 Residential Land as Space of Self-Supply and Self-Help.................................................. 85 States of Emergency ............................................................................................................. 95 Conclusion ...........................................................................................................................100   v   CHAPTER 6: The New Territorial Order ....................................................................... 102 Reforming Urban and Regional Development Control Mechanisms ..............................105 Blackening Grassroots Practices, Whitening “New Urban Areas” ..................................107 Making it Happen ................................................................................................................114 Appropriating Land by Law ...............................................................................................120 Conclusion ...........................................................................................................................124 CHAPTER 7: Land for Fresh Ghosts, Land for Dry Ghosts............................................. 126 Hòa Mục under the “Policy of Working a Lot, Eating a Lot” ..........................................128 The New Urban Order Comes to the Village ....................................................................134 Urban Expansion and Land Disputes on the Rise .............................................................139 State Commitments as Tools for Resistance .....................................................................145 Conclusion ...........................................................................................................................149 CHAPTER 8: Conclusion ............................................................................................... 151 Lessons from History ..........................................................................................................152 Repertoires of Popular Agency ..........................................................................................156 Coalitions of Interests and Recombinant Territorial Orders.......................................... 159 Invisible Moral-Territorial Orders .....................................................................................160 Moving In and Around Policy ............................................................................................162 BIBLIOGRAPHY ........................................................................................................... 167 STATUTES CITED ........................................................................................................ 179 APPENDIX A: Information Sheet ................................................................................... 181 APPENDIX B: Official Research Summary .................................................................... 183 APPENDIX C: List of Interviews ................................................................................... 185 APPENDIX D: Sample Interview Questions .................................................................... 187 APPENDIX E: Chronology ............................................................................................... 191  vi    LIST OF TABLES  Table 1 Housing space produced in Hà Nội between 1981- 2008 (in m2) .....................................110    vii   LIST OF FIGURES  Figure 1 Villages absorbed in Hà Nội‟s built fabric during the 20th century ..................................... 2 Figure 2 View from the village towards the new urban area bordering it ......................................... 3 Figure 3 Location of Hòa Mục ..................................................................................................... 25 Figure 4 Map of the region of Hà Nội in 1935 .............................................................................. 37 Figure 5 Administrative map of Hòa Mục‟s immediate region in 1941 ......................................... 39 Figure 6 Periurban villages specializing in textile industry production in the Red River Delta, circa 1930 ............................................................................................................................................ 53 Figure 7 A thatched house (left) and a “level-4 house” (right) ...................................................... 92 Figure 8 Map of the districts of Đống Đa and Từ Liêm in 1972 .................................................... 94 Figure 9 Urbanized area in 1992 (left) and spatial expansion prescribed by the master plan for 2010 (right) .........................................................................................................................................107 Figure 10 Post-reform housing built privately by households .......................................................109 Figure 11 “The Manor,” a KĐTM under construction near Hòa Mục ...........................................113 Figure 12 A “State and People” neighbourhood in Trung Hòa ward.............................................117 Figure 13 New urban-styled, multi-story houses in Hòa Mục .......................................................133 Figure 14 Areas earmarked for recovery and redevelopment........................................................141 Figure 15 Communal house of Hòa Mục (left) and new the cultural house (right) ........................146 Figure 16 The new road that cuts through the village (left) and the new residential development built on its former agricultural land (right)...................................................................................149   viii   LIST OF ABBREVIATIONS  APEC:  Asia-Pacific Economic Cooperation ASEAN: Association of Southeast Asian Nations BOLUC: Building ownership land-use certificate COMECON: Council for Mutual Economic Assistance DRV:   Democratic Republic of Việt Nam FDI:   Foreign direct investment KĐTM:  Khu đô thị mới (new urban areas) KTT:  Khu tập thể (collective zones) LPF:  Land-price framework MOC:   Ministry of Construction SOE:  State-owned enterprise SRV:  Socialist Republic of Việt Nam VWP:   Việt Nam Workers' Party VCP  Việt Nam Communist Party VNĐ:   Việt Nam đồng WTO:  World Trade Organization    ix   ACKNOWLEDGMENTS  I first want to thank the people and authorities of Hòa Mục for having generously given up their time and shared the tumultuous story of their village with me. I also wish to thank Dr. Trịnh Duy Luân for having generously accepted to sponsor my research visa and to host me during fieldwork within the Institute of Sociology in Hà Nội. Phạm Quỳnh Hương, Phùng Thị Tố Hạnh and Trần Nguyệt Minh Thu deserve particular mention. Without their trust, support, and wealth of experience, this research would simply not have been possible. I also wish to thank Đặng Bảo Khánh and Trương Thúy Hằng for their unflagging support and help during interviews. The warm contact that the above-mentioned group of female researchers established with the people of Hòa Mục and local bureaucrats played a large part in the quality of the discussions and information gathered during this research.  Within the university, my supervisory committee was a source of inspiration and support during the entire research process. Michael Leaf, John Friedmann, Terry McGee, and Abidin Kusno provided me with different, though highly complementary theoretical and methodological perspectives. A lively community of students both at UBC and in the field augmented the advice and feedback I received from my supervisors. Thanks to Leslie Shieh, Clément Musil, Juliette Segard, Lisa Drummond, and Trần Nhật Kiên for having been supportive friend and critical colleagues. I also wish to thank the group “Doing Research in Việt Nam,” Vũ Tuấn Huy and Nguyễn Văn Sưu for opportunities to present early analyses while being in the field and for useful feedback. Thanks also to Lê Minh Hằng, Trương Thúy Hằng, Lương Ngọc Thúy, Trần Thùy Dương, Nguyễn Thị Diễm Hà, Vũ Quỳnh Dương and Nguyễn Thị Thanh Mai for their help as research assistants, interpreters, or transcribers. My debt of gratitude also goes to my great friends Trần Ngọc Minh and Leszek Sobolewski who generously helped me with translations.  Financial support for research and writing came from a Social Sciences and Humanities Research Council Graduate Scholarship, from the research projects “The Challenges of the Agrarian Transition in Southeast Asia” and “Informality and Governance in Peri-urban Southeast Asia: A Study of the Jakarta and Hanoi Metropolitan Region”, and from the Centre Urbanisation, Culture, Société at the Institut National de Recherche Scientifique du Québec. I am grateful to all of these organizations for their generous assistance. A final word of thanks to my friends and family; for although their contributions were often indirect, they were no less significant and are no less appreciated.  1   CHAPTER 1 Introduction        Hòa Mục is a small village located on the western bank of the Tô Lịch River, about 6 kilometres from the historic centre of Hà Nội, the Vietnamese capital. While quite close to the inner city, this area still corresponded to the city‟s rural-urban interface less than a decade ago. Today, other than ritual buildings and an organic network of narrow roads, few visible traces of rurality remain. The Tô Lịch River, which long stood as a natural border between the urban and rural administrative worlds, lost its demarcation role in the late 1990s. This happened following the construction of new bridges and large avenues, penetrating ever deeper into the capital‟s hinterland. The outward expansion of the urban built fabric, functions, and markets further accelerated following the 1997 redesignation of this zone as an urban administrative district. In a matter of a few years, a new urban landscape had engulfed the pre-existing settlements and superseded the large swaths of paddy fields that had surrounded it since times immemorial.  Although now taking place at an unprecedented pace, the integration of formerly rural places into Hà Nội is not a new phenomenon. A tight network of densely settled villages has long characterized the capital region‟s geography. A corollary of this is that, since at least the early 20th century, the city has necessarily absorbed periurban villages as it expanded into its rural hinterland (figure 1). Maps produced since the colonial era nevertheless indicate that the urban expansion process rarely wiped a pre-existing village off the map. Rather, the urban built fabric developed from within and around these erstwhile rural settlements. Still today the city‟s growth continues to encompass rather than obliterate periurban villages (Labbé 2011a). The resulting city is, as can be observed in other East and Southeast Asian contexts, a mosaic of former rural villages, spontaneous neighbourhoods, and planned redevelopment zones (e.g., Sorensen 2002; Lockard 1987; Cybriswsky and Ford 2001; Hsing 2010).  Vietnamese people now call Hòa Mục a “village within the city” (làng giữa phố). In urban residents‟ vocabulary, this expression designates the many residential areas, formerly dominated by rural populations living off agriculture, which are now engulfed in the city‟s space. Hòa Mục‟s  2   urbanity takes many forms, among which, is a new—and rapidly changing—built form. Across the imaginary boundary separating the village‟s territory from the rest of Hà Nội, one can now observe a streetscape of eclectic, multi-storey, row houses. This residential landscape is very similar to that of other spontaneous neighbourhoods located in the inner city. Only a handful of traditional, one- storey, rural houses surrounded by gardens and outbuildings and circumscribed by walls remain in the entire village. These material artefacts, inherited from a rural past, might not resist the passage into the urban world for very long as their occupiers have, in most cases, demolition and reconstruction plans in the making.  Figure 1 Villages absorbed in Hà Nội’s built fabric during the 20th century  Source: author  Outside of the old village settlement‟s limits, the millennial landscape of rice fields, polders, and canals, has also given way to urban forms. This occurred very slowly during the 1960s and 1970s, as the state recovered small tracts of land and redeveloped them into schools, army compounds, and collective housing areas. Then, between 2000 and 2003, Hòa Mục‟s peasants were forced to give up their use-rights on all that remained of the farming land that they and their ancestors cared for and  3   tilled for innumerable generations. The state—which is the ultimate owner of all land in Việt Nam—retrieved this productive agricultural area and transferred it to a former state-owned enterprise active in the construction sector. In accordance with the master plan of the capital then in force, this enterprise levelled up the fields, filled up the canals, and redeveloped this site into a so- called “new urban area.” The resulting urban neighbourhood features large avenues flanked by high-rise residential buildings, commercial and office space (see figure 2). According to both the city and the developer, this new environment represents the future of the capital city. It does so by encouraging a more “civilized urban way of life” (nếp sống văn minh đô thị); one that purposefully moves away from rural traditions, built forms, and spatial practices.  Figure 2 View from the village towards the new urban area bordering it  Source: author, 2009  Changes can be observed in other spheres of Hòa Mục‟s life. Compared to the situation about a decade ago, the village‟s population is larger, denser, and socio-economically more diverse. Between 1997 and 2009, the local population of the ward of Trung Hòa (to which Hòa Mục belongs) grew from 14,000 to 27,000 people. Two thirds of these new residents are migrants. A first wave of newcomers arrived in the early 1990s. For the most part, these were relatively well-off people from neighbouring provinces, including many retired bureaucrats. These people relocated closer to Hà Nội on a permanent basis, in many cases to facilitate their children and grandchildren‟s access to the capital‟s white-collar and governmental jobs. A few years later, another group, consisting of seasonal and temporary migrants, started to move into the village. Most of them rented  4   small rooms built by villagers next to their houses, in makeshift buildings called “nhà trọ”. These migrants are much less affluent than their predecessors. They consist of a mix of students and workers originating from various provinces in the delta. They came to the village seeking cheap accommodations outside of the unaffordable urban core, and yet at commuting distance of the city‟s universities, blue-collar, and informal job markets. Since the revocation of agricultural land rights, in the early 2000s, these renters provide former peasant households with their primary source of income.  The recent integration of the village into the city‟s administrative system, the penetration of new urban built forms, the end of farming activities, and the arrival of a large migrant population all suggest that Hòa Mục has completed the historic shift from rural to urban. Yet, for a whole segment of the population, this place is still very much a village (làng); their homeland (quê hương) and the land of their ancestors. This sense of place comes up very rapidly in discussions with native residents. It is also visible in how these people try to maintain and transmit to their heirs ritual practices and communal values inherited from the past. This does not mean, however, that villagers attempt to live in the past. Rather, Hòa Mục residents perpetuate their attachment to the village‟s history and values while practicing urban occupations, living in urban-styled homes, enjoying new cell phones and satellite TV dishes, and encouraging their children to learn foreign languages. As their parents and grandparents did throughout the last century, the villagers of today are selectively adopting, maintaining, and rejecting both elements of the new economy and culture and aspects of community and ritual life inherited from the past.  This dissertation explores the long process of adaptation and hybridization that underpinned Hòa Mục‟s shift from rural village to urban neighbourhood. These chapters tell the story of a small periurban place and its community “becoming urban” against the backdrop of the tumultuous contemporary history of Northern Vietnam, a history marked by colonial domination, struggles for independence, post-war reconstruction, socialist transformations, and market reforms. The point of taking the reader on this long historical journey is not simply to provide descriptions of demographic, economic, or built forms changes over time. My primary aim is rather to identify the origins and transformations of the practices and rules that structured particular territorial orders on the outskirts of the Vietnamese capital from the beginning of the 20 th  century up to now.  Before and Beyond Đổi Mới: Revisiting the Urban Transition in Việt Nam Friedmann (2008: 254) wrote that a planner‟s work is inevitably confronted by urban and regional dynamics that can hardly be understood except in a way that cuts across disciplines. This comment holds true for the study of the periurban which is a complex phenomenon characterized by a fluid  5   combination of social, economic, cultural, and political changes. In an effort to account for the periurban process as a whole, I took several “mining expeditions” (ibid: 254) into the literature of the disciplines of geography, anthropology, social history, and political science. There, I looked for and selectively borrowed ideas, concepts, and methodological approaches capable of illuminating the periurban processes in Việt Nam. The resulting analytical framework, which I am about to describe, integrates elements from these various fields yet with the ultimate objective of integrating them back into the planning discipline.  This dissertation is located within a large but widely scattered corpus of literature on the ongoing urban transition of Asian developing countries. In its most familiar form, this phenomenon refers to the shift from a society defined by a largely agricultural population to one in which an urban population predominates (Friedmann and Wulff 1975; Ginsburg 1990). Assessed from the vantage point of this basic definition, it seems that this transition is just beginning in Việt Nam. In 2009, official government data reported just above 26 million urban dwellers out of a total population of 85.8 million (BXD 2009).1 While this corresponds to a two-fold increase of the proportion of urban population compared with that of 1950, it still only represents 30 percent of today‟s Vietnamese national population. Demographic projections indicate that this upward trend is likely to continue for the coming decades, with half of the Vietnamese population expected to be classified as urban in 25 to 30 years from now (United Nations 2009).  The urban transition, however, involves more than a redistribution of population from rural to urban places. As noted by Friedmann (2005: xiv-xv) with reference to China, the process by which a country “becomes urban” consists of a dynamic matrix of administrative, economic, physical, sociocultural, and political changes. One of the basic challenges in studying the urban transition is to characterize how this complex process drastically alters the fabric of predominantly agrarian societies. Beyond such characterization, the study of the urban transition also calls for understanding the patterns underlying urban and regional transformations. This entails identifying the set of forces (both past and present) and actors (both endogenous and exogenous to the urbanizing territory and society), and determining how these forces and actors interact to shape the conditions, processes, and outcomes of urbanization.  In analyzing these various forces and actors, the scholarship on the urban transition in Việt Nam assigns a central role to the state and, in particular, to policies it promulgated as part of the  1 For census purpose, the Vietnamese government defines an urban place as a city, town, or district with 2,000 or more inhabitants. The urban population only includes individuals officially registered in urban places. This definition therefore excludes a large number of rural migrants permanently or temporarily living in urban areas. Source:  6   country‟s shift from plan to market. This shift refers to a series of socio-economic reforms adopted since the early 1980s and globally referred to as đổi mới (lit. “renewal”). These purportedly state- led reforms have recast the model of centralized planning that defined Việt Nam‟s socio-economic system since the 1960s. They did so by giving market mechanisms a much greater role in the allocation of goods and services yet within an economic system still officially defined as socialist in orientation.  Students of the urban transition in Việt Nam have placed the đổi mới reforms at the centre of their explanatory frameworks to explain a variety of urban phenomena. These include changes in the socio-economic and population structures of Vietnamese cities (e.g., Boothroyd and Pham Xuan Nam 2000; Ledent 2002), rural-to-urban migrations (Li Tana 1996; Gubry et al. 2002), transformations in the production and expression of the built environment (e.g., Nguyen Quang and Kammeir 2002; Pandolfi 2001b), and changes in municipal administration and governance (e.g., Trinh Duy Luan 1996; Forbes and Le Hong Ke 1996). This approach is sensible: both the academic and journalistic literature on the post-reform period suggests that recent transformations of the country‟s socio-economic system affected virtually all spheres of Vietnamese society in one way or another. The urbanization process is certainly no exception to this rule.  Yet, in trying to explain Việt Nam‟s urban transition with the reforms as a central explanatory factor, most studies build on what I believe is a problematic and somewhat misleading assumption: They suppose that the changes observed in and around Vietnamese cities since the 1990s (rural-to- urban migration, development of an urban-oriented economy, urban physical expansion, etc.) are essentially an outcome of đổi mới. Implicitly or explicitly, these studies argue that urbanization phenomena observed over the last two decades could not have happened prior to the changes brought about by the reforms. These reforms are understood as having „liberated‟ urbanization forces previously constrained under the plan. In other words, the phenomena characteristic of the urban transition in Việt Nam are seen as a societal response to state-led policy changes.  This assumption reflects a conception of the relationship between reform policies and social change that emphasizes the Vietnamese state‟s control over its various arms and, more generally, over the society and territory it governs. It builds on the idea that an authoritarian state rules the national territory and dominates society. In this view, đổi mới is understood as top-down adjustments of the national economy through macro-structural policies stipulated by the party and enacted by the state apparatus since the early 1980s. It is further assumed that the state has the capacity to effectively impose its governing rules and norms on society through the powerful and pervasive Vietnamese Communist Party (VCP) (see Womack 1992; Porter 1993; Thayer 1992; Abuza 2001).  7   An alternative conception of state-society relations in Việt Nam underpins this dissertation. As will be further elaborated below, I ground my analysis in the idea of a state that does not unilaterally dominate society, but rather has a symbiotic relation with it. In this view, socio-spatial changes result not only from centrally-devised policies, but also from ad hoc adaptations of the state‟s rules and programs at the grassroots, and from pressure and influence coming from various parts of society (including those located within the state apparatus) (e.g., Beresford 1988; Kerkvliet 1995a, 2005; Thrift and Forbes 1986; Fforde 1989).  This interpretation has important methodological and analytical implications. First, it confers recognition on the role that pre-reform circumstances and practices play today. As discussed above, a majority of authors concerned with the ongoing urbanization process in Việt Nam assume that urbanization practices observed in recent years were generated de novo in the present period as a result of state-led reform policies. A majority of studies thus describe ongoing urbanization practices as unprecedented. When authors refer to the pre-reform period in order to explain recent changes, they generally depict an urban Việt Nam under the plan, which, I presume, has more to do with the ideal-typical model of socialist urban and regional development than with actual historical reality.  This assumption is somewhat surprising considering that very few urban and planning studies have paid attention to changes happening in and around Vietnamese cities prior to the 1990s. Lack of reliable sources and difficult access to “objective” studies partly explains this a-historical aspect of the literature on recent urban changes. Indeed, between the end of the French colonial era and đổi mới, local and foreign scholars conducted virtually no field-based research:  The acceleration of the socialist developmental path in the North did not favour independent scholarly research [...] From 1965, when the war with the Americans took on an open character, up to 1976, general scholarly interests in Indochina, and in Vietnam in particular, was considerable but quickly faded thereafter. There were only a handful of social scientists who conducted field studies of a more anthropological nature during the First and Second Indochina Wars (1946-75), while virtually nothing was written in the decade that followed (Kleinen 1999: 8).  Inspired by the work of Janet Abu-Lughod (1996, 1999) on urban formations in both the developed and developing worlds, I seek to demonstrate the value of taking a longer perspective in the study of contemporary urban and planning changes. In line with this, I have intentionally let go of the prevailing assumption that practices observed in recent years are merely responses to state-led policies, and that they have no historical precedents. By putting socio-spatial transformation in a longer historical context, and by focusing on everyday practices, institutional evolution, and shifts  8   in governing practices I consciously depart from stereotypical portraits of periurban places. By extending the framework of research on the urban transition backward in time to include the pre- reform period, I instead intend to show that many contemporary urbanization practices, even those that seem to have emerged because of the new market environment (real estate transactions, housing construction and rentals, industry and commerce), find their roots in the pre-reform era.  Emphasizing mutual influences between state and society has a second major implication: It draws attention to the role of popular agency. In this dissertation, I define agency as the socio-culturally mediated capacity of individuals and groups to act. This conceptualization draws heavily on Anthony Giddens‟s (1979, 1984) theory of structuration and on the theory of practice proposed by Sherry Ortner (2006). In line with Giddens‟s work, the above definition emphasizes the idea that people‟s actions are shaped (in both constraining and enabling ways) by the very social structures that those actions then serve to reinforce or reconfigure.While acknowledging the pervasive influence of structural forces (including culture) on human intentions, beliefs, and actions the understanding of agency used in this study is different from ideas of free will and routinized practice. I instead embrace Ortner‟s (2006) view and posit that agency involves some degree of intentionality. An agent is thus understood as someone who intervenes in the world with “something in mind (or in heart)” (ibid.: 136). This is not, however, to say that agency is a straightforward synonym for resistance. An important point raised by Ortner (and relayed by Roy (2011) in her critique of subaltern urbanism) is that, by focusing essentially on resistance to the status quo or existing power differentials we are missing out on the non-oppositional forms that agency does take.  The literature on Việt Nam often neglects or makes invisible the role that both confrontational and non-confrontational forms of popular agency play in the urbanization process. This scholarship tends to depict rapidly urbanizing territories and their people as „victims‟ of the urban transition who essentially deploy defensive and adaptive tactics in the face of changes driven by external forces and state-led policies. This study questions such portrayals by focusing on the everyday, human aspect of urbanization at the local level. In doing so, I wish to open the door to the possibility that communities, households, and their members participate in a complex process of change, and that they have a role in shaping the urban transition through individual or collective decisions and actions.  Thus far, only a handful of studies have looked at aspects of the urbanization process in Việt Nam through the lenses of history or popular agency (DiGregorio‟s 2001; Thrift and Forbes 1986; Hardy 2003; Koh 2006). While limited, this scholarship interrogates the assumption that state-led reforms  9   entertain a unidirectional relationship with societal practices. This work more generally questions the actual power of state policy to effect social change in Việt Nam. What the authors listed above suggest is that whenever policies devised at the central level do not fit local needs, values, and practices, various societal groups—including those evolving within the state apparatus—have some room to manoeuvre in which they can ignore, circumvent, or adapt official norms and rules. This promotes a more complex understanding of the shifting relationships between structural conditions and forces, central-state plans and policies, customary rules and moral norms held by local communities, and the actual everyday practices of both state agents and populations in shaping the urbanization process.  This emerging scholarship on state-society relations in Việt Nam informed my decision to organize and develop this study around two major conceptual themes. First is a conceptualization of the periurban as a zone of encounter, and as recombinant socio-spatial assemblages where institutional arrangements, practices, and forms from the past are constantly redeployed and reinvented in the present. Second is the importance played by regulatory informality as a co-evolutionary process where new or hybrid socio-spatial practices regularly arise that seem to contradict the directives or values of the central state, and yet continue to exist alongside the purportedly official way of doing things. I now review each of these conceptual orientations in turn.  The Periurban as “New Urban Frontiers” Students of urbanization in countries of developing Southeast Asia signal that the urban transition is experienced unevenly across national territories with perhaps the greatest effects in the expanding spatial zones surrounding the largest cities. The literature describes these transitional zones between country and city as a theatre of rapid and fluid changes operating simultaneously at the spatial, functional, environmental, institutional, and human levels. Most authors acknowledge that the scope and speed of the changes occurring in these zones require more research attention. Supporting this interest is the view that periurban places might be insightful sites to examine, understand, and theorize the urban transition in the region (e.g., Jones 1997; Ginsburg et al. 1991; Webster 2001).  While it enjoys increasingly important currency in the literature, the term “periurban” is ill-defined. At the mere etymological level, this expression refers to areas around (peri-) the city (urban). Yet, beyond this basic definition, debates go on as to whether the periurban corresponds to a discrete spatial zone that can be precisely delineated on a map, or whether it consists of a combination of features and phenomena. Either way, the question is raised as to what characterizes periurban spaces or processes, how we can identify them as periurban (or not), and why such categorization matters (see, for instance, Adell 1999; Browder et al. 1995).  10    Attempts to generalize about the periurban struggle to account for the situated characteristics of this phenomenon, which are no less varied across national settings than they are within single metropolitan regions (e.g., Browder et al. 1995; Simon 2008). Yet, the multiplicity of features and processes underlying the periurban and its various forms across time and space do not mean that this concept should be dispensed with in toto. Nor does it mean that the phenomenon should be reduced to a set of “particular” conditions that must be closely documented, enumerated, and subjected to the operations of taxonomy. However conceptually incomplete or vague the current definitions of the periurban, and however limited the spatial and temporal foci of most case studies detailing its modus operandi, the burgeoning literature attests to the world-wide presence of this phenomenon. The diversity of its manifestations does not make the periurban a less valuable concept, but rather indicates the need for a greater degree of conceptual flexibility.  I am not, therefore, attempting here to define periurbanization as an ideal-typical territorial form or geographical space identifiable by a specific combination of socio-spatial characteristics (population composition, employment structure, land use, etc.) or processes (livelihood diversification, migratory patterns, market relations, built environment mutations, etc.). I am rather adopting a process-oriented conceptualization. Central to this is the fragmented, unfinished, and unstable character of the periurban. This invites narratives about the fluidity of changes as social agents experiment in new ways with economic opportunities, the material environment, or institutional arrangements. It is a claim that, in periurban areas, the territorial formation process has not yet arrived “at the end,” and that being unfinished matters analytically and politically.  It is in that sense that I understand and build on Leaf‟s (2008) conceptualization of the periurban in Southeast Asia as “new urban frontiers.” The term frontier is used here with reference to:  [A] place of encounter, of interaction and contestation between disparate groups, with the potential for new forms of social mixing, a place of promiscuity. But the frontier is also a discourse, implying newness and change. In this sense it is a place of hope, perhaps inevitability, a source of worry and uncertainty. [...] one also encounters the idea of lawlessness, with recourse to brute force as a principal means for the expression of power; from this we may understand institutional and regulatory weakness to be a fundamental characteristic of the frontier. Gaps open up, with ambiguity as to how they are to be filled. In the frontier‟s lawlessness, we may also see indications of its position as a geopolitical strategy, expressive of state interests and perhaps conditioned as well by market relationships. The processes of frontier formation are thus forms of territorialisation, that is, the territorial expression of state intentionality. (ibid.: 8-9)  Leaf (ibid: 9-11) goes on to identify three types of periurban frontiers:  11    i) A frontier of urbanization, in the conventional sense of outward expansion of urban functions across erstwhile rural spaces. This includes the rapid growth of population in periurban places, livelihood diversification away from agricultural activities, and socio- cultural transformation. This first frontier also refers to the space needed for a modernizing city to expand into; space for new airports, industries, waste disposal, water supply, cemeteries, parks, etc;  ii) A frontier of globalization with reference to the global flow of capital, goods, symbols, and ideas into urban hinterlands. This includes the transformation of agricultural areas into industrial landscapes under the impulse of export-oriented national development policies, and foreign direct investment. This second frontier encompasses the flow of images and ideas of newness and modernity conveyed by the media and the ever-expanding influence of urban spatial economies;  iii) A governance frontier in the sense of ongoing re-territorialisation particularly, but not exclusively, by the state. This goes beyond urban administrative redesignation processes to include the various expressions of regulatory power by political and economic elites and the ways in which this reconfigures the everyday terrain of habitation, livelihood, self- organization, and politics.  It is difficult to trace neat boundaries between the various areas constitutive of the periurban and to identify where these frontiers are located exactly. Spatiality nevertheless matters. Important factors underlined in the literature relate to a given area‟s relative distance from and ease of access to the inner city. With reference to the Chinese context Hsing (2010) distinguishes between the “near” and “far” periphery. McGee (1991: 6-7), on the other hand, subdivides what he calls mega-urban regions (or extended metropolitan region) in three spatial zones loosely defined as:  i) an administratively expanding city core; ii) an intermediate periurban zone where the components of the built environment are penetrating previously rural space; and iii) an urban hinterland in which aspects of the urban are “leapfrogging” along highways and in which there is still some co-existence or urban and rural activities.  The site on which this study focuses was on the near periphery or intermediate periurban zone of Hà Nội for most of its history but is fully integrated into the city core since the late 1990s. As will be described in more detail in Chapter 2, I chose Hòa Mục as my main study site for logistical reasons, and because I thought that this village could shed light on the process through which a place becomes urban. The spatiality of this village is important because it contextualizes many of my  12   findings. To give only one example, the enduring socio-economic linkages to the capital city which profoundly impacted the development trajectory and governance of this periurban place were made possible by the short commuting distance separating it from inner city populations and markets. The focus on a site on the near periphery also limits the scope of my findings, especially with respect to their value for understanding transformation processes in localities on the far periphery.  The State-in-Society One function of thinking about the periurban as multiple “frontiers” is to gain analytical and critical insight into the periurban as a “zone of encounter, conflict, and transformation” (Friedmann forthcoming). The orientation toward the contested and contradictory underlying the idea of “frontiers” provides thinking space for the reconsideration of how territorial claims by various social groupings interact with the state‟s planning function. This then calls for a conceptualization of the relationship between state and society in the Vietnamese context.  Joel Migdal‟s (2001) “state-in-society” approach provides a useful starting point to analyze the workings of the postcolonial, socialist state in Việt Nam and that of its planning agencies. In his analysis of Third World states, Migdal emphasizes the contradictions that regularly arise between the state‟s image of wholeness and the day-to-day governing practices of its various agents and institutions:  The state-in-society model focuses on this paradoxical quality of the state; it demands that students of domination and change view the state in dual terms. It must be thought of at once (1) as the powerful image of a clearly bounded, unified organization that can be spoken of in singular terms, as if it were a single, centrally motivated actor performing in an integrated manner to rule a clearly defined territory; and (2) as the practices of a heap of loosely connected parts or fragments, frequently with ill-defined boundaries between them and other groupings inside and outside of the official state borders and often promoting conflicting sets of rules with one another and with “official” Law. (Migdal 2001: 22)  By combining these two paradoxical sides of the state, Migdal‟s approach questions the state‟s ability to turn rhetoric into effective policy. It draws attention to the cross-purposes of the activities of state agents and institutions. The “state-in-society” approach does not yet reduce the state to a mere collection of predatory officials seeking personal benefits through public functions. Migdal‟s conceptualization rather calls for heightened attention to situations where various parts of the state ally with one another and with groups outside to further their goals. From there, we can move on to explore what sets of rules are promoted through these coalitions and networks and how these rules either reinforce or thwart the state‟s own official laws and regulations (ibid: 21).   13   Another important dimension of the “state-in-society” approach is to highlight the interactions between social groupings and the state and to their “mutually transforming” relations. This approach gives prominence to the dynamic evolution of social groupings located both within and outside the state apparatus. It encourages the study of these groupings‟ behaviours, of the shifting alliances that they form with each other, and how these change over time. In this view, the state is a contradictory entity and a social construction that entertains a “mutually transformative” relationship with society. In this contingent relationship, the state induces social changes while at the same time being transformed by society.  By drawing on Migdal‟s conceptualization of the state in Việt Nam, I wish to emphasize the notion of planning as a relational activity occurring in a common playfield. As pointed out by Gainsborough (2010) the various individuals and institutions that constitute the Vietnamese state do not always move in the same direction, work together, or sing the same hymn. The fragmented power of the state and its penetration by what we might call private interests sheds light on what often appears as a lack of coherence between the interventions of various parts of the state apparatus (local state, central institutions, agricultural cooperatives, SOEs, etc.) in the day-to-day management of grassroots practices on the urban edge.  As will be made evident in the following chapters, the “planning” process (be it labelled as such or not) in the region of Hà Nội is not limited to this distinct arena within state bureaucrats and experts seek to orient the development of cities and regions. As part of a complex ecology of actors, the people that we conventionally call “planners”—those state agents in charge of carrying out planning functions—operate in uneasy, unstable interrelationships with other actors and sources of societal power. Their actions and decisions are shaped by the territorial claims of ordinary people. These actions and decisions are also influenced by various social groupings who penetrate the more “porous” reaches of the state bureaucracy and who pressure (through discourse or actions) political elites to orient territorial policies in specific directions. In this context, the boundary between public office and the private activities of state agents is often blurry and this makes it difficult to distinguish the regulated from the regulators. 2   As will be shown in chapters 4 to 7, various social groups and forces have been active in shaping the territorialization of periurban Hà Nội since the beginning of the 20th century. These include local populations, political and economic elites, agents of the state operating at all scales (from the commune to the national level), state-owned enterprises, etc. These groups draw on various sources of authority, from pre-colonial customs and traditions, to discourse about national unity, progress,  2 See Jean Oi‟s (1989, 1995) for a similar analysis of the local state in China.  14   or modernity. The story told in this dissertation is that of these recombinant coalitions of interests between these various actors, of how these shifted over time, and what particular socio-spatial arrangements they created or supported along the way. This is also the story of the uneven and changing capacities of the individuals and social groupings involved in these coalitions to act and influence the course of change, and the patterns of collaborations or conflicts emerging between them over time.  Planning and Regulatory Informality In unpacking the shifting coalitions of interests responsible for the urbanization process in periurban zones, this dissertation revisits the question of regulatory informality and that of the state‟s role in the production and reproduction of this phenomenon. The literature generally conceptualizes the informal in relation to the formal. From this viewpoint, the central feature of informal urban practices such as casual employment, land squatting, unrecorded land subdividing and transactions, etc. is their occurrence outside of formal institutional frameworks. These frameworks are those in which the state intervenes (or is supposed to) to regulate processes and outcomes according to a set of enforceable legal rules (see, for instance, AlSayyad 2004; de Soto 1989, 2000; Sanyal 1988).  This dissertation calls for a reconsideration of this state-formality equation. Formal state institutions are in fact oblivious to the role that informality plays “formally.” Recent studies have explored a variety of situations where states benefited from governing practices that seem to contradict their regulatory function. This includes the deliberate formulation of ambiguous regulatory frameworks (Ho 2001; Gainsborough 2010), the temporary lifting of regulations, or the selective retreat of the state‟s policing power from specific economic sectors or geographic areas (e.g., Leaf 2008; Roy 2009c; Ong 2006; Yiftachel 2009a, 2009b). Making sense of these governing practices requires that we move beyond attempts to identify whether particular practices are “legal violations” stricto sensu. We rather need to look at informal transactions as expressions of social relations, and more specifically of state-society relations (Tabak and Crichlow 2000; Leaf 2005: 94).  In the following chapters, I bring supporting evidence for the idea that states (including planning authorities) do not seek to extend the reaches of their formal regulatory authority at all times, across all sectors of the economy, over all societal practices, or geographical areas (e.g., Portes et al. 1989; Crichlow 2000). Following Ananya Roy (2009c), I suggest that informality results not only from government tolerance to resolve potential social conflicts or to promote patronage but can also be an intrinsic element of local cultures of governance (including planning). In this view, urban informality is not necessarily a social process developing outside the purview of the state as a form of popular resistance or insurgency against public powers (e.g., Lefebvre 1991; Holston 2007; Roy  15   2009b). The conditions for the production or reproduction of informal practices are instead made possible by state interventions. It is in this sense that urban informality might be understood as the expression of an alternative form of state control.  Building on Roy‟s (2011) recent proposition, I use two main concepts to shed light on the role of the state in urban informality. First is the idea of “zones of exception.” Put forward by Aihwa Ong (1999, 2006) in her study of transnationalism, citizenship, and neoliberalism, this concept emphasizes spatial fragmentation as an instrument of territorial governance. Ong posits that, while competing with multiple sources of power, nation-states—with their supposed monopoly over spatial planning—do play important roles in structuring territorial orders. One of the ways in which this state control over space is exercised is through the creation of a system of non-contiguous, graduated spaces within which different populations “are variously subjected to political control and to social regulation by state and non-state agencies” (Ong 1999: 219). Ong calls “zones of exception” these spaces where policies are unevenly enforced or where regulations are temporarily lifted.  The concept of “gray space” complements this analysis by making evident the state‟s flexible use of its regulatory power across time and space. 3  Developed by Yiftachel (2009a, 2009b) with reference to the contemporary Israeli planning regime, “gray spaces” are defined as:  [D]evelopments, enclaves, populations and transactions positioned between the „lightness‟ of legality/approval/safety and the „darkness‟ of eviction/destruction/death. Gray spaces are neither integrated nor eliminated, forming pseudo-permanent margins of urban regions which exist partially outside the gaze of state authorities and city plans. (Yiftachel 2009b: 250)  The concept of “gray space” is closely associated to a governing practice that Yiftachel calls “gray spacing.” This refers to the manipulation of plans, policies, and regulations by political and/or economic elites to “whiten” (legitimize/authorize) or, alternatively, to “blacken” (delegitimize/criminalize) different spatial practices or configurations occurring in gray spaces. As conceptual lenses, “zones of exception” and “gray spaces” contribute to explain how, while being fragmented, penetrated by private interests, and in competition with other sources of power, the state remains pivotal in shaping the process and outcomes of the periurbanization process in Hà  3  An earlier statement on periurban “gray zones” can be found in McGee (1991: 17) with reference to his formulation of the extended metropolitan regions: “[EMRs] are to some extent "invisible" or "gray" zones from the point of the view of state authorities. Urban regulations may not apply in these "rural areas," and it is difficult for the state to enforce them despite the rapidly changing economic structure of the regions. This feature is particularly encouraging to the "informal sector" and small scale operators who find it difficult to conform to labor or industrial regulations.”  16   Nội. This is not because the state is an authoritarian force that stands above society, or because it exercises its regulatory coercive powers forcefully. It is rather because the state is composed of agents who are an integral part of the society they govern. Similar to Migdal (2001: 20), I observed above that “various parts of the state ally with one another, as well as with groups outside, to further their goals.” This embedding of state agents within various coalitions of interests contributes to shift the rule of access and control over material and immaterial resources during the urbanization process. It is, therefore, as part of these coalitions that various arms of the state can sway the balance in favour of particular interests. As will be illustrated in this dissertation, in some, but certainly not all cases, these interests happen to be synonymous with the public good.  Research Focus and Scope The purpose of this study is to advance us towards a better understanding of the periurban formation process that has been unfolding on the edge of Hà Nội since the colonial era. As discussed above, the approach taken in this study consists of exploring the various ways in which individuals and groups (including those evolving within the state apparatus) have met and experimented—in more or less organized and coordinated fashions—with the restructuring of institutional arrangement, market relations, spatial practices, and other phenomena responsible for the periurbanization process. The general questions that this dissertation seeks to answer are: Who participates in creating these shifting socio-spatial arrangements on the periurban edge of Hà Nội? Using what means? With what degree of control or power over the resources and institutions shaping the urbanization process? And, with what intentions, if any?  Providing general answers to these questions is a vast project. A wide array of physical, social, economic, institutional and environmental change is inherent to the periurbanization process. An exploration of the full range of practices, decisions, and ideas, and of all the societal groups involved in each of these spheres of change, is obviously beyond the scope of this study, conducted by a single foreign researcher with limited resources and time. Based on a critical review of existing literature on periurban changes in the Red River delta region and on data collected during preliminary fieldwork, I have decided to focus my exploration on the possibilities of (and limitations to) livelihood strategies afforded by periurban populations in relation to the ebb and flow of state regulations.  A sub-element of livelihood changes explored in this dissertation is was I call “land strategy.” I define this concept as the various ways by which communities, households, and individuals appropriate, use, secure, and exchange lands on which they have some form of entitlement in order to generate, sustain or improve their current and future well-being. These “well-being  17   improvements” include the economic-materialistic as well as the social, political, and cultural dimensions of communities, households and individuals‟ lives.  Livelihood and land strategies are closely connected—land having been one of the key resources underpinning subsistence activities on the edge of Hà Nội throughout the 20th century. This is made evident in the cases, for instance, of the diversification of agricultural production by farmers, negotiation and exchange of land use-rights with public and private actors, land subdivisions, contestations of land use-right compensations in the context of large public or private developments, private interventions on the residential built form, etc.  The rationale for this focus on livelihood and land strategies is twofold. First is the observation that, in deploying such strategies, local populations participate in several aspects of the periurbanization process including: the redefinition of the local economy, the formation of land and housing markets, and the shaping of the physical landscape. Second, is the fact that livelihood and land are two areas that have been the object of state regulations since independence. These two areas are therefore insightful lenses into the territorial formation of periurban Hà Nội.  On this basis, the specific research questions guiding this study are:  (i) What were the livelihood and land strategies deployed by periurban communities, households, and individuals since the beginning of the 20 th  century? (ii) How were these strategies devised and implemented within local socio-cultural and institutional contexts? (iii) How did these strategies intersect with official land use controls and land development plans and projects put forward by the state prior to and during the reform period? (iv) What do these dynamics and interactions reveal about state-society relations in the metropolitan formation process around Hà Nội?  The rest of this dissertation is structured as follows. Chapter 2 discusses the methodological approach, data collection and analysis procedures used in this study. This discussion also serves to highlight the mutually constitutive relationship between my practice of fieldwork in Hà Nội and the representation of findings in this dissertation.  The subsequent chapters are organized in chronological order. Chapter 3 provides historical background on Hòa Mục‟s geography, demography, and institutional structure during the late- colonial era. It also analyzes a period of unprecedented socio-economic change that began in the 1920s, showing how rural-urban linkages developed under the influence of new market relations.  18   Chapter 4 analyzes the early years of the socialist transformation process (1940-1965). I explain how the flexible and uneven implementation of state policies (Land Reform, agricultural collectivization, nationalization of industry) at the city‟s periphery gave villagers some room for manoeuvre within which they maintained earlier rural-urban linkages and livelihood structures. Chapter 5 (1965-1980) explores regulatory informality as a state governing practice and its implication in the reproduction of practices responsible for an early in situ village urbanization phenomenon. Chapter 6 steps away from Hòa Mục to analyze the formulation of pro-urban orientations at the national and regional governmental levels during the reform period (1980-2009). I discuss the manifestations and impact of these new orientations with regard to new urban planning mechanisms, models of urban developments, and shifts in land legislation. Chapter 7 returns to Hòa Mục to analyze how these changes unsettled the moral territorial order established during the pre- reform era and how groups of villagers responded to some of these changes through open forms of contestation.  19   CHAPTER 2 Bending Western Research Expectations into Hà Nội‟s Environment        This research project began with a fairly open plan, informed by ten years of doing research, teaching, and consulting in Việt Nam, two pre-fieldwork trips to Hà Nội (in 2004 and 2005), and a review of the literature on fieldwork in socialist contexts (de Soto and Dudwick 2000; Nordstrom and Robben 1995; Kurti 1999; Heimer and Thogersen 2006). I tried to design a flexible research plan, keeping in mind the challenges typically faced by students doing research in Việt Nam: requests for research visas and official clearances to access research sites and governmental data, control and censorship over sampling and the content of interviews, and negotiations with local institutional hosts and assistants in an increasingly commoditized research environment (see Scott et al. 2006; Turner 2010). Despite all these precautions, upon arriving in Hà Nội, I soon had to adjust, circumvent, and sometimes bend my plans in modes I had not expected. Doing fieldwork became something of a dynamic experiment, which I tried to safeguard—as best I could—with critical reflexivity and accountability.  This chapter recounts this story. It grounds this dissertation by documenting how I generated and analyzed data and by reflecting on the impact of my positionality on this research (England 1994; Rose 1997; Robertson 2002). Beyond exposition of methodologies and introspection, I also want to render the circumstances surrounding the research process explicit and point up the numerous people that made an imprint—purposefully or not—on the material and ideas from which this dissertation is built. Following Reid-Henri‟s (2003) proposition, I wish to account for the bundle of logistical, ethical, political, practical and human challenges into which I was embedded during my time as a researcher in a cross-cultural and transitional socialist context. In making explicit how I coped with such challenges, the role played by the events, people, and conditions encountered in the field, and the “empirical drifts” taken by this project, I hope to show the mutually constitutive relationship between the practice of fieldwork and the production of knowledge.   20   This chapter begins with a discussion of the research approach I adopted in the field. It then moves on to discuss the challenges associated with my selection of Hòa Mục as my main research site. A description of these steps in the research process reveals the sensitivity of the land question around Hà Nội. I then discuss issues of sampling and data generation, and describe how my project shifted from responding to the local authority‟s censorship towards dealing with villagers‟ concerns. I also discuss the key role of research assistants during the interview phase, and the problematic issues of disengagement and reciprocation. This chapter concludes with an overview of the process through which data and field experiences were analyzed and uploaded into discourse in this dissertation.  Research Style Reconfigured A combination of research objectives, feasibility constraints, and personal inclination led me to announce in my research plan that I was to focus on a single case. From Canada, the single case study seemed to be a realistic strategy for an independent, foreign researcher (i.e., not part of a large research project) within a limited timeframe. An additional limitation was that I brought no funding into a research environment where foreign scholars are increasingly seen as a potential source of supplementary revenues. 4  In this context, the focus on a single site allowed me to maximize resources and time available for actual field research.  In my pre-fieldwork proposal I had also announced that my study was to draw on qualitative procedures inspired by the ethnographic tradition (I very much drew on approaches promoted by Stake 1995 and Lofland 1995). I wrote that I was “to subject myself to the contingencies that play upon individuals, [to] penetrate their past and present life circumstances, [and to] pick up the complexity of their decisions and actions.” While this approach was well intentioned, the constraints I faced in the field did not allow for its materialization. I was given limited access to do research in the village I chose to study and I could not live there. 5  The arrangements I had with my host institution: the Vietnamese Institute of Sociology at the Vietnamese Academy of Social Sciences and the local authorities further prevented me from conducting formal interviews alone. In addition, my command of the Vietnamese language, if functional, meant that I needed the assistance  4 The basic salary of Vietnamese researchers at a public research institute can be doubled or even tripled through participation in research projects for which researchers are paid according to the nature and scope of their involvement. Foreign-funded projects are seen as one such source of supplementary revenue. In the same way, independent foreign researchers, including PhD students, are commonly asked to pay a monthly rate (up to a few hundred dollars) for their affiliation with local host institutions. I shall mention here that this was not my case. I am grateful to the Institute of Sociology for having graciously hosted me. 5 To my surprise, while local authorities accepted my request to live in the ward, it is villagers who turned me away. Most of them thought that accommodation available in the village (aimed at students and migrant workers) was not up to foreigners‟ standards. They also wished to avoid the burdensome procedures associated with renting to foreigners; i.e, getting a formal authorization from the local People‟s Committee, registration with the police, and levying of a special tax on foreign renters.  21   of an interpreter.  6  What I had called an “ethnographic approach” in my proposal turned out, in practice, to be more accurately describable as “qualitative research methods.” Although some planners might have done what can properly be considered ethnographic work, I am not one of them. My use of the ethnographic approach extended somewhat beyond metaphor, but not as far as a methodology: semi-structured interviews remained my stock-in-trade, and the long term, personal engagement with the people of Hòa Mục I had hoped for was not possible in these conditions.  My experience mirrors Luong Van Hy‟s (2006) view according to which ethnographic fieldwork is not yet possible in socialist Việt Nam. This does not mean, however, that students of urbanization in Việt Nam should altogether abandon approaches aimed at getting a face-to-face, direct, and, if possible, intimate knowledge. I instead agree with Bunnel and Maringanti (2010) that “ethnographic engagements with urban spaces and lives, the use of languages other than English and the kinds of cultural competencies that are usually associated with area studies training” (ibid: 417) are necessary to counteract “metrocentricity.” By this, the authors mean that bringing ethnographic approaches to urban and planning research is a counterpoint to the tendency for metropolitan-based scholars to keep the research focus within the realm of the “English-speaking urban worlds in which [they] are able to operate comfortably and effectively, or else to topics that are amenable to research at a distance through the collection of secondary data” (ibid: 418).  Forsaking engagement with the particular and the local means keeping many urban regions and cities, such as those found in Việt Nam, “off the urban and planning research map” (Robinson 2002; see also Roy 2009a). From this viewpoint, culturally attentive, „local‟ research is better than nothing, even when it challenges established (Western) methodological wisdom. As the Canadian anthropologist Jean Michaud (2010) writes, refusing to adapt to the methodological constraints imposed by socialist contexts would indeed entail that “a large proportion of the world‟s population would remain outside of our scope” (ibid: 223). Michaud instead argues for creative and productive adaptations of methodological approaches to the restrictive local research conditions characteristic of socialist Asia.  I adopted this approach in the field. Throughout the year I spent in Hà Nội, distinctions of distance, race, class, education, and language remained important barriers separating me from the place and people I studied and maintained my outsider‟s position. Constraints imposed by local authorities and the sensitivity of the topic I studied also limited the scope of my participation in a milieu, a  6 I have been learning the Vietnamese language since 2001. By the spring of 2009, when I began conducting interviews in Hòa Mục, I could have simple conversations, ask interview questions and comprehend most of the responses but still needed the support of a research assistant especially during in-depth interviews.  22   sense of which I could only partially and imperfectly internalize. I nevertheless tried to make a virtue out of necessity. Whenever I could, I criss-crossed the village by foot, bicycle, or motorbike, taking pictures and mapping the activities and rhythms of daily life. I hung out at cafes and tea stalls, roamed through local markets, and dropped by temples and pagodas, chatting in my clumsy Vietnamese with whomsoever was passing by. At the invitation of villagers, I also attended ritual activities (e.g., the end of summer celebration, mid-autumn festival, etc). I never became integrated into the village society, but my observations and informal conversations helped me discover ongoing issues, potential interviewees, and upcoming activities in which I could take part.  Contingencies of Selection and Access I came to Hà Nội having established that the unit of study for this research would be a “natural village.”7 By focusing on this small unit of analysis, I hoped to document in some depth the evolving structure, processes, and outcomes of villagers‟ practices within the wider context of the urban transition in the Red River Delta. I wanted to understand how local livelihood and land strategies developed from the colonial era to the đổi mới era. More specifically, I wanted to document the changing role of state regulatory powers and local institutions in the production and reproduction of local practices and institutions over time.  Hurdles of Site Selection The question as to how to select a particular village came up during my first meeting with the researchers at the Institute of Sociology. My first selection criterion was logistical: I needed to commute by bicycle on a daily basis between the city center, where I lived, and the chosen village. My research site therefore had to be located within a radius of no more than ten kilometres from the historic city. That still left me with about fifty potential settlements, all of them located within the fringe of intense socio-spatial transformations surrounding Hà Nội (see figure 1 in previous chapter). How was I to pick one?  The plan I had made while still in Canada was to visit several villages in different periurban areas accompanied by a research assistant. At each site, I would gather basic socio-economic data and conduct a few exploratory interviews with the local authorities and inhabitants to sketch a rough portrait of the locality. I hoped to evaluate the “suitability” of each site for the purpose of my research and then select the one that maximized what I could learn about land and livelihood strategies and the periurbanization process. This scoping phase was also expected to provide me  7 I borrow the expression “natural village” from the China scholarship where it is used to designate a long- standing rural settlement, distinct from the unit of government called the “administrative village.” In the Vietnamese context, the natural village (làng) differs from the rural commune (xã) by the absence of a formal governmental structure and responsibility.  23   with basic comparative data from various villages that I could later use to contextualize the case selected for this study.  This approach did not fly very well with my Vietnamese collaborators. Entering each site required formal authorizations from local ward authorities. My colleagues knew from experience that this process would be costly and easily use up my entire time in the field. The “cost” issue was related to the practice, widely applied by local researchers, of giving “envelopes” to local officials in order to get clearance to research sites. This practice is institutionalized to the point that specific rates informally apply for officials holding key positions in local administrations (chairmen and vice- chairmen of People‟s Committees or agricultural cooperatives, heads of mass-organizations, etc.). It is, of course, subjectively and ethically challenging to determine when and where it is appropriate to compensate poorly paid state employees for their time and assistance in the field. What I found even more problematic was to impose my Western ethical views about this question to my local collaborators. For this reason, when it came time to ask the local authorities for access to Hòa Mục, I decided to follow the local practice within the limits of my research budget.  I abandoned my initial selection strategy but used alternative sources of information to get a qualitative sense of the mosaic of periurbanization trajectories in the Red River Delta‟s villages. I relied on notes from pre-fieldworks trips (summer 2004 and winter 2005) during which I visited several periurban villages around Hà Nội. During the main period of fieldwork, in 2009, I also benefited from the help of foreign and Vietnamese researchers and friends who generously shared their data, took me for visits, and even let me interview local people in the different periurban villages and districts they studied.  I thereafter considered the possibility of selecting a research site based on government-produced data. This alternative avenue presented another set of obstacles: although good-quality population censuses and living standards surveys are conducted on a regular basis in Việt Nam these data were only available to me in aggregated form at the district level and above (city, province, etc.).  Since a single periurban district encompasses at least ten natural villages, such data was of little use to my site selection process. Another problem is that the natural village stopped being a formal administrative unit in Việt Nam after independence. Getting socioeconomic data at the village level implied a reworking of raw data. This proved an unrealistic project not only in terms of the time it involved but also because I could not access un-aggregated data during my time in the field.     24   The Empathic Researcher It is in the face of these limitations that I finally decided to focus my research on one of the villages I had visited during a 2004 pre-fieldwork trip. The village of Hòa Mục (figure 3), which was to become my main research site, had been pointed out to me by a Vietnamese friend who then worked as a journalist in Hà Nội, and generously acted as my research assistant in her free time. One day, she took me to a village which she thought would be of interest to me because it had recently featured in local newspapers as the site of protests against forced land acquisitions.  Our first stop was at a tea stall along the Tô Lịch River. There, we chatted with an old lady selling green tea, cigarettes, and candies to passersby. Sitting across the low wooden bench on which our two cups sat, she was quick to mention that the village had experienced major transformations in recent years. As the conversation developed, I nevertheless saw that understanding what was happening there, and how local people perceive recent urban changes was not going to be a straightforward business:  Old lady: They have taken our land. Me:  What land? Old lady: All the land. It belongs to the state, you know? They have policies, they do whatever they want and we, villagers, can‟t do anything. We, the native people of this village, “have short throats, narrow throats” (thấp cổ bé họng). Me: What do you mean by that? Old lady: I mean that we have no say in the matter. Our voices cannot be heard up there. Nowadays, in this village, it‟s difficult; the high-rise buildings behind us, they throw a shadow on this village. They are beautiful from outside but seen from inside the village, it‟s another story. The shadow of the city eats the village, little by little: the fields, the gate, the communal house, the people. It‟s a hungry shadow [laughing]. Me: So the new buildings around the village affect you a lot? Old lady: Of course they do, what else would you think? But the new high-rise buildings make the whole area more beautiful, it looks urban now. Before, it was all dirt roads, thatched houses, ponds, paddy fields. It was always very damp, it was dark at night... And now, the new supermarket, so big, so well-organized. Me: So you go there to shop? Old lady:  Oh no! Never. It‟s far too expensive! Me: But you still like them, the new high-rises, the supermarket? Old lady: Of course I do, they make the face of the village look more beautiful (bộ mặt làng đẹp lên).   25   We had other conversations with villagers during the following visits I made to Hòa Mục. I was intrigued by the paradoxical opinions about urbanization expressed by local people, such as the seemingly contradictory views of the old lady in the above excerpt. During these visits, I also learnt that this place had once been a craft village, that it had a surprisingly large number of ritual buildings, and that its agricultural land had been requisitioned to build a flagship residential redevelopment.  Figure 3 Location of Hòa Mục  Source: author  All of these features sharpened my curiosity. But what really drew me to Hòa Mục was a full afternoon spent with elderly men and women who revealed to me their concerns about the future of their fatherland (quê hương) in the face of land acquisition, demolitions, and new construction associated with the city‟s rapid expansion. At the end of that afternoon, these elderly villagers asked me if I could write a piece denouncing the threat faced by their village during the urbanization process and publish it abroad, in a foreign newspaper.  26   I explained to the villagers that I was not a journalist but a PhD student. I then suggested ways I could maybe help in that capacity. One of them was, if possible, to come back to study their village and tell their stories through presentations at local conferences and seminars attended by local scholars and professionals who have some influence on policy-making. 8  It is therefore not only because I assessed Hòa Mục as a “good” case-study that I selected it but also, and perhaps more importantly, because I was sincerely touched by the stories these old villagers shared with me and by their request for help. Without knowing it, I had become a morally engaged planning scholar.  Controlled Entrance I knew accessing a village that had recently been the site of land conflicts would not be straightforward, but gaining access to Hòa Mục was more complicated than I had anticipated. It did eventually prove feasible, but not without a cost. The conditions under which I was allowed to conduct this study made me more dependent on my institutional host than I had planned. I also had to cope with local authorities‟ restrictions on the sample of interviewees, and censorship of the questions I planned to ask them. A few months into the interview process, I was asked to leave the site due to emerging political tensions discussed in more detail below. The local authorities‟ attitude towards my research revealed in very concrete ways how sensitive the land issue can be in periurban villages of Hà Nội.  I decided that in order to facilitate access, I would introduce the project to the People‟s Committee of the ward, as being led by the Institute of Sociology. This decision aimed at downplaying my role as a foreigner in the field. It reframed my project in ways that went beyond its new official description on paper. In the process, my institutional host moved from an advisory to a collaborator status. It was agreed that for the first two phases of interviews I would be accompanied by one of the senior researchers affiliated with the Institute. 9   These same senior researchers also strongly advised me to introduce my project to the local administration as being about the history of urbanization in Hòa Mục with a focus on the villagers‟ role in this transformative process, rather than as a study about villager‟s access and control over land. This second recasting of my project aimed at drawing the authorities‟ attention away from the  8 In fact, at the end of 2009 I was interviewed by a journalist from Le Monde Diplomatique about urbanization in Việt Nam. I then used the case of Hòa Mục to illustrate challenges faced by periurban populations during the current outward expansion of the city. See Monthéard (2010). 9 This arrangement affected my original research plans in various ways, for instance, in terms of the rhythm of interviews. Following textbook recommendations, (e.g., Kaufmann 1996; Grawitz 2001), I wanted to limit myself to one interview per day, so as to have time to quietly write notes after each of them, and to revise the list of questions to be asked of the next respondent. To accommodate my research collaborators, I nevertheless agreed to schedule up to three interviews per day during the first two phases of the research.  27   sensitive land issue (see Official Research Summary in Appendix B). Self-censorship went one step further when ward bureaucrats asked for higher-level official clearance from the district of Cầu Giấy and Hà Nội‟s Department of Foreign Affairs (Sở Ngoại vụ). Obtaining these higher-level authorizations required me to submit the list of questions I was to ask villagers. In an attempt to smooth out this process, I removed from my interview schedule all questions related in one way or another to land or conflicts in the village. By April 2009 I could finally enter the village. Compounding the multiple self-censoring phases that preceded, I was, by then, wondering what was left of my original research project.  Negotiating the Field: Sampling and Data Generation Three conditions set by the local authorities for allowing me and my local collaborators access to the village sharpened this questioning. These conditions were: 1) that the leader of a local mass- organization act as a guide and introduce us to interviewees; 2) that we only interview people on lists drafted by the ward authorities; and 3) that we financially compensate the guide and interviewees for their participation in the study. 10   Secondary literature suggests that foreigner researchers‟ access to research sites is commonly subject to such conditions. In some cases, this goes so far as to require that the researcher carry out all interviews in the presence of a local government official or policeman (O'Rourke 2004: 53-4; Scott 2001: chapter 1). These rules nevertheless appeared very constraining to me: the people listed by the ward authorities were most likely to be members of model-families which had not been involved in land conflicts or other forms of resistance. I also feared that these people might not want to talk about the informal land strategies and other extra-legal activities I wished to document.  Interview Phase One: Coping with Elderly People The reality of the field took me yet again in an unexpected direction. On our first visit to Hòa Mục, in April 2009, my colleagues were guided to the house of the chairman of the local Elderly People Association (Hội người cao tuổi). The People‟s Committee of Trung Hòa had appointed this man to be our guide for this study‟s first phase which focused on elderly people. The chairman and his wife frowned upon the list of interviewees produced by the ward that we handed to them. This list, they told us categorically, would not do: this person was too sick to answer questions, this other one was near-deaf, these two were wife and husband and I should interview them together, this other was  10 I agreed to give the leader of the mass-organization the equivalent of 50 cents CDN for each introduction to a local family, $2.50 to village residents, and $5.00 to officials who participated in this study. These small amounts appeared to me as symbolic gestures of appreciation for the time interviewees were going to spend answering my questions. Some village residents, especially elderly people, refused to take the envelope we insistently offered them after the interview.  28   away visiting his grandchildren in another province. Our official sampling of interviewees, they added, needed to be revised and they would take care of it.  We therefore started interviews on that day at the house of a neighbour suggested by the chairman of the Elderly People‟s Association. Decisions about who to interview next were made one at a time during the following three weeks of this first phase (from end of April until mid-May 2009). Since the chairman was open to suggestions, I started to make special requests for certain types of respondents. In this way, I tried to balance out the sampling of interviewees in terms of genders, occupational backgrounds (agricultural, non-agricultural), place of residence in the village (older areas, recent extensions), and place of origin (native or not). Appendix C provides a description of the sample for this and the two subsequent phases.  My original plan had been to rely on life-histories for this first phase of interviews. Logistical constraints once again reoriented my methods. Since the study was now officially led by the Institute of Sociology, interviews had to be led by the senior female researcher who graciously accompanied me in the field. It was agreed that I would sit next to her and be introduced as a “student-assistant.” Since I was not in a position to guide the conversation, I quickly abandoned the idea of asking broad, open-ended questions about the interviewee and his/her family‟s life. Such general questions, when we tried to ask them, also tended to elicit answers about how things “ought to be” rather than about what actually happened in the village in past decades. In these conditions, I decided to use a semi-structured interview schedule, which I adjusted as new topics and issues emerged. (A sampling of interview questions used during this phase and the following ones is provided in appendix D.)  While I sat next to my senior Vietnamese colleague, I actively listened and sometimes asked supplementary questions to probe or clarify points raised by interviewees. 11  I also took advantage of my position, out of the action, to take notes about the house in which we sat and about the interviewees‟ non-verbal reactions to specific questions. If my status as a white, female, Canadian created obstacles in trying to access the village, it generated much curiosity during interviews. Once we were done with our questions, respondents and their relatives often started their own questioning, which I answered with great pleasure. As all my research collaborators were female, the questions asked by the women we interviewed often flowed into personal matters regarding my  11 All interviews were conducted in the house (villagers) or at the workplace (local bureaucrats) of the respondents. Interviews lasted between one and two hours. Past fieldwork experiences indicated that it is difficult to ask Vietnamese respondents to sign informed consent forms. For this reason, at the beginning of each interviews, I provided informants with an information sheet about the project (see sample in Appendix A) and then asked them for oral consent to participate in the study.  29   marital status, my parents and siblings, my place of residence in Hà Nội, my command of the language, etc. The men we interviewed also expressed curiosity about me. Some of them also engaged me in discussions on the differences between the management of infrastructure, demographic growth, and urban development in Việt Nam and Western countries.  The objective of this first phase of interviews was to comprehend the complexities of decision- making about livelihood and land strategies by putting them in the wider context of individuals‟ lives. In my research proposal, I had written that I was to relate these strategies to “key-changes in individuals‟ lives, to the village‟s institutional life, and to changing structures of land ownership and control.” As mentioned above, as we started the first phase of interviews, I wondered how I was to do this without asking questions about land.  This problem dissolved during the first interviews: while my questions focused on family and village life, the issue of land was invariably brought up at one point or another by the interviewee. Elderly people had much to say about the role land played in their lives; they could not talk about livelihood or housing without broaching the issue. Respondents often brought the conversation around to questions of the interdependence between livelihood strategies and access to land as a form of natural capital. Interviewees also spoke of the tensions created by opposing viewpoints as to how the village‟s land ought to be used, and about how they think that ongoing land mismanagement will affect their households‟ economic future. At first, we simply let the interviewees talk about these issues. Then, I carefully introduced indirect and later more direct questions about land strategies in relation to the institutions controlling the disposition of land at the village and commune/ward levels.  Interview Phase Two: Ward Officials and Leaders A second phase of interviews took place during the first half of June 2009. I then met eleven governmental officials working for the People‟s Council, People‟s Committee, agricultural cooperative, and mass organizations. These interviews were initially aimed at collecting basic socioeconomic information about the ward (land area, population change, territorial administration, etc.) and its governmental institutions. I wanted to chart how these changed during the administrative shift of this place, in the late 1990s, from a rural commune (xã) to an urban ward (phường). I also wished to explore the official discourse on urbanization held at the local level.  These interviews fulfilled this objective in part, but important documents remained out of reach. Accessing cadastral maps, in particular, proved impossible, revealing once again how sensitive the land question is in the region of Hà Nội. There are several ongoing litigious affairs about land  30   ownership and property delimitations across the city, some of which date back to forced land and housing acquisition during the first years of the DRV government (see Gillespie 2009). As explained by an official in charge of the cadastral reform at the municipal level, maps, old and new, are seen as potential pieces of evidence that can be used by citizens against the authorities in these conflicts. For this reason, they are generally kept inaccessible to the public (personal communication, 23, November 2009). Yet, this phase of interviews provided unexpected material. Many of the officials we interviewed expressed critical viewpoints on state policies, especially regarding forced land acquisition. This uncovered the existence of tensions about development policies within the state apparatus.  Interview Phase Three: Focus on Land Issues The third phase of interviews took place from August to November 2009. During this last phase, the control of the ward‟s authorities on my activities somewhat relaxed: The vice-chairman of the local People‟s Committee allowed me to conduct interviews in the village alone with an assistant/interpreter I had hired. It was also agreed that, this time, I was not to follow a predetermined list of interviewees but rather proceed through a snowball sampling technique. These new conditions not only permitted me to broach more sensitive topics but also to approach some of the villagers who actively resisted expropriations and land redevelopments projects in recent years.  The questions I asked during this last phase were developed around livelihood- and land-related issues that had emerged out of the first two phases of interviews. This focus on issues was not a fixation on problems or failures. Following Stakes (1995) I rather hypothesized that the nature of social systems becomes more transparent during their struggles. I wanted to explore how villagers strategized in the face of constraints and problems. As such, the topics discussed during this phase of interviews were related to competing notions of property rights, authority over land controls, and institutions responsible for the provision of infrastructure and other public goods. These issues were not simple and clean, but rather intimately connected to political, social, historical, and especially personal contexts which I tried to explore with interviewees.  Twenty residents of Hòa Mục were interviewed during this last phase. I had originally planned to vary the demographic and socioeconomic characteristics of my sampling. I hoped to meet younger residents, male and female, recent and not so recent rural migrants, suburbanizing dwellers and temporary residents, members of local mass organizations, agriculturalists and people working in other sectors within or outside the village‟s limits, etc.   31   This proved more difficult than expected. Logistically, it was a lot easier to talk to retirees and women, who tend to stay home during the day, than to people belonging to other categories. Newcomers were the most difficult to access. This is in part because these younger residents generally work outside the village during the day and have family obligations in the evenings. It is also an outcome of the snowball technique. This sampling method indeed made it difficult to establish connections outside native villager‟s circles. As a result, half of the respondents who participated in this phase were over 55 years old and over two-thirds were native of Hòa Mục. This phase still allowed me to diversify my sampling. I met with households who, over the last decades, have relocated from the village to the inner city and vice-versa. I interviewed four heads and vice- heads of local resident groups (tổ dân phố). And, most importantly, I had long conversations with, and obtained key-documents from, a group of people who have actively resisted forced land acquisition since the late 1980s.  This last interview phase ended abruptly at the beginning of November of 2009. I was obliged to leave the field upon the request of the local People‟s Committee. I was told by a local official that I had enough data as it was and that my study was over. This came as a surprise to me. Earlier in the year, the Vice-chairman of the People‟s Committee had indeed approved my research schedule, which extended until the end of the December. We had also agreed that I was to prepare a public exhibition on the history of the village to thank villagers and local authorities for their participation in my study. In late November I understood why I had been asked to leave. A phone call from a villager informed me that a new expropriation process had begun in the village. The local authorities probably did not want me to nose around the area anymore. Once again, the sensitivity of my research topic had shaped fieldwork activities.  Secondary Sources During my time in the field, I gathered official government reports/statistics, books, reports, and historical documents about urban and periurban development available in local publishing houses, research institutions, universities, and public libraries. This material relates broadly to urban development, industrialization and economic development policy. It includes texts and plans that describe the state‟s intentions for the periphery of Hà Nội, and various legislative documents related to land and housing in Việt Nam.  My Caucasian-Western status was helpful here: an “outside” identity and citizenship from a country that was not involved in the Việt Nam War frequently opened doors that would otherwise be closed to those within the local social hierarchy (something my local colleagues repeatedly pointed out to me). This meant easier access to senior politicians, officials and academics. My architecture  32   diploma from Hà Nội‟s National School of Civil Engineering (obtained in 2001, following a semester spent on exchange at that institution) also facilitated access to Vietnamese built environment professionals, including those people responsible to design roads and new housing estates in the vicinity of Hòa Mục.  I collected archival records, aerial photographs, and a large set of old and recent maps of the region of Hà Nội. I also kept a newspaper clipping file with articles related to ongoing urbanization issues published in French, English, and Vietnamese. Some of the key-resisters in the village also let me photocopy their archives of newspaper clippings about recent land conflicts in the village and the correspondence they kept with different parts of the party-state apparatus during these disputes. I did a photographic survey of Hòa Mục to document recent physical transformations. I also collected literature, music, TV series, popular tales and proverbs as entry points into the analysis of social norms, representations, experiences, and perceptions of urban changes on the outskirts of Vietnamese cities.  From Analysis to Uploading Discourse into Writing I accumulated an overwhelming amount of data generated during fieldwork. Added to a body of documents, maps, photographs, and videos, I ended up with hundreds of hours of interview transcripts. With a few exceptions, the 42 interviews I conducted with villagers and officials were tape-recorded (I took written notes for the others). Because I wanted to capture all that was said and also how it was said, I had the recordings transcribed verbatim in Vietnamese, and then translated in their entirety by university students hired in the field. Upon my return to Canada in January 2010, I revised these full Vietnamese and English transcripts with the help of native Vietnamese speakers to ensure that meanings, expressions, and idioms, were translated as accurately as possible.  Data reduction: Coding and Memos The first step in analyzing this material was to „reduce data‟ into more manageable formats and sizes through iterative phases of coding. Before, during, and after fieldwork, I also wrote research notes (or “memos”) to elaborate about the various categories emerging out of this “data reduction” process and to clarify assumptions about what I saw, heard, and experienced in the village. I later used these memos as “theoretical write-ups” on the correspondences, patterns, and relationships I observed between the topics, categories, and themes identified in the literature or during the coding process.  To paraphrase Strauss and Corbin (1998), I used theoretical memos to exhaust momentary ideations as they took form while I inspected, interpreted, and coded data, reflected about methodological  33   procedures and about my personal fieldwork experiences. The early interpretations that emerged from going back and forth between data and interpretation were shared and refined through discussions with colleagues and friends in the field, at local seminars and conferences,  and through the writing of summary articles early on upon my return to Canada (Labbé 2011b; Labbé and Boudreau 2011; Labbé and Musil forthcoming).  Reciprocation and Representation Gathering primary data in Hòa Mục was a process suffused with a constant tension in assimilating individual experiences into an academic research process. The interview process in particular required that I break down „big‟ questions—questions that drive a research project and which have relevance in an academic context—into smaller ones through which local people could hopefully relate to my purpose. There would have been little point in talking directly about land legislation, institutional arrangements, state-society relations or territorial formation processes. Instead, a translation was required between languages and from the academic to the daily world. I needed to formulate questions to villagers about their past and everyday lives that could be easily answerable and yet would speak to larger issues.  During the analytical and writing phases that followed the fieldwork period, the opposite question arose: how to present the experiences of people with whom I interacted in terms of an academic dissertation? The first half of this “translation” of field experience into writing concerns the incorporation of research themes into a broader theoretical framework. If the theorization process is not always straightforward, it is at least well-charted in research methods textbooks. The other half of the translation process concerns the moral commitment toward the people I studied. Conversations with villagers periodically reminded me of the reasons that originally brought me to Hòa Mục and of the fact that my recording of the stories of urbanization in this village could go beyond academic purposes. The following excerpt, taken from the very end of an interview with a 70 year old native villager, is an example of such reminders:  Me: I thank you for participating in this interview. I have a small gift for you [handing out envelope]. Man: No gift. I‟m not taking any gift. Ladies, you keep this envelope and go drink something [with the money in it]. Me: I have listened to the story of many people in this village of Hòa Mục. I am touched by what the people of your village do, getting up and fighting, protesting against [state] decisions. But I cannot help you very much. I sometimes go to the city and I talk about the families and people of your village, not only wanting more money, but also wanting to protect their community, protect their ritual monuments...  34   Man: You ladies don‟t worry about that. The people here don‟t want to move out of their ancestor‟s land. We want to rebuild our house. We don‟t want to be the losers in all this [...] You don‟t need to worry about it. When you tell this story to others, you help us people like that, by talking (bằng lời nói).  Beyond the issue of reciprocation also lies the question of how to bring along the care and sensitivity put in the acts of listening and recording the voice of local people into academic representations. How to prevent the lived process of fieldwork to fade into memory and experience; to be lost in the distilling of interview transcripts, codes, and memos? In the following chapters, I have tried to use quotations 12  and description of the place where fieldwork was conducted, to soften this process of dehumanization and separation.  Conclusion This chapter shows that, though never an insider, I became part of what I studied in the sense that the places and people of the field interacted with me as much as I did with them. This narrative documented the repeated instances when I had to bend my original research plans in modes that were somehow acceptable to me, my research collaborators, the people I studied, and their local government.  Through this tale from the field, I wished to highlight how the successive drifts of this study into unplanned territories impacted its conceptual and methodological developments. I also wanted to make visible the marks left on this research by collaborators and friends who gave a hand whenever they could, and others who, in collaborating with me, straddled the employee/friend distinction. These people opened doors and gave me access to data I would not have been able to access alone. They listened to and gave their opinions on my puzzles, ethical dilemmas, negotiations of relationships, and preliminary analyses. Vietnamese colleagues, interpreters, and translators in particular played the role of cultural brokers, highlighting for me important information in interview transcripts, providing explanations to cultural references, and deciphering idioms and other subtexts they thought I might otherwise miss. The opinions and voices of fieldwork collaborators certainly shaped this study by allowing me to make sense and, ultimately, to conceptualize what I saw, heard, and sensed in the field.  Feedback and discussions with colleagues continued after I left Hà Nội. These post-fieldwork email exchanges and conversations played an important role as I started to analyze and interpret data and  12 In what follows, all interviewees are referred to using a pseudonym to ensure confidentiality. Quotations from interviews appear in two forms: first they are extracted directly from the interviews; and second, paraphrased information compiled from the interviews is utilized. The latter method is used where the precise way in which people have said things is not so important.  35   to write this dissertation. If I did not have full control over my research activities while being in the field, this dissertation ultimately still rests on my coping strategies and decisions to allow reorientation of my methodologies and focus. My role and responsibilities only grew after coming back to the predictable and comfortable Canadian environment, where most of the interpretation and writing process took place. As such, what follows results as much from what Henri-Reid (2003: 2) calls the “joint work between the researcher and the field” as it does from my ability to capture social interactions, record them in a coherent fashion, and to aggregate them in meaningful ways and from my own interests and positionality.  36   CHAPTER 3 The Early Urban Transition (1920-1940)        The village of Hòa Mục (lit. human concord) belongs to the so-called “Mọc area.” This small area, on the western bank of the Tô Lịch River, groups together five villages that share common rituals and guardian spirits that are related to each other in the Vietnamese pantheon. 1  Hà Nội historians indicate that this area, at the crossroads of two important precolonial transportation routes (the Tô Lịch River and the Lai King mandarinal road, see figure 4) was inhabited for over a thousand years (Trung Hiếu March 14th, 2010, 2010). The Revolutionary History of the Ward of Trung Hoà, similarly indicates that a first settlement was established in this area by a group of generals serving the last Hùng emperor around 300 BC (VCHĐB 2005: 7).  When exactly Hòa Mục was founded and how it developed throughout the precolonial era and most of the colonial period is nevertheless unclear. Due to this shortage of historical information, this chapter recounts the story of the village starting in the 1920s. This period corresponds to the earliest personal memories of the fifteen elderly people I interviewed about the village‟s life during the colonial period. The accounts of these septuagenarians and octogenarians are complemented by those of younger native villagers who shared with me what they know about their parents and grandparents‟ livelihoods during “the old days” (ngày xưa). Besides this life history material, this chapter also builds on studies, produced by French colonial, Vietnamese and foreign postcolonial scholars, about rural transformations in the Red River Delta during first half of the 20 th  century.  The purpose of this chapter is twofold. First, it aims to situate Hòa Mục in the broader history and geography of the Red River Delta during the colonial period, and thus provide the reader with background for understanding changes during the postcolonial era. Second, it explores the  1 In Northern Việt Nam, village‟s guardian spirits (thành hoàng, also translated as “tutelary deities”) have historically been incarnated by land deities, ancient mythology characters, military heroes, or ancestors of local noble families (Nguyen Thu Chi 1993a: 146-7). Hòa Mục‟s guardian spirits are three siblings (Phạm Thị Uyển, Phạm Miện and Phạm Huy) who helped the emperor Mai Hắc Đế oust Chinese occupiers from Hà Nội at the end of the 8th century.  37   periurbanization process under colonial rules, a question that has received little research attention so far.2  Figure 4 Map of the region of Hà Nội in 1935  Source: Adapted by the author from Hanoi 1935, Service de Géographie de l‟Indochine  The latter objective underpins this chapter‟s focus on the intense period of change between the early 1930s and the mid-1940s. During this decade and a half, new market relations associated with the colonial presence penetrated in northern Việt Nam and combined with local circumstances to foster an occupational thickening process. The people of Hòa Mục then initiated a first phase of industrialization and urbanization. This transformative process provides a counterpoint to portraits of Northern Vietnamese villages as “timeless” places which, up to the eve of the August Revolution (1945), were “not yet developed” (chưa phát triển) because of structural inabilities to escape a conservative, subsistence-oriented, agrarian economy (e.g., Nguyen Thu Chi 1993a: 58). By the  2 Philippe Papin‟s (1997) doctoral dissertation is an important exception, although much of it is unpublished as of yet.  38   early 20 th  century, Hòa Mục was not a place where people hid behind bamboo hedges, or indistinctly kept intact traditions from times immemorial. Villagers were not the plain, straight, and conservative individuals portrayed in the literature, and the community on the whole was certainly not essentially inward looking and traditionally oriented. Rather, the people of Hòa Mục initiated new activities that profoundly and very rapidly transformed the village and its relationship to the city and the outside world.  The discussion is organized as follows. In the first section, I sketch a portrait of the village‟s demographic, territorial, administrative, and lineage structures at the closing of the 19 th  century. I then turn to a description of villagers‟ livelihood strategies during the first decades of the 20th century. This discussion serves to outline the structural conditions that allowed for the rapid development of a thriving cottage industry of weaving and garment making in the village. This occupational shift is the object of the third section. There, I explain the origins of this new economic activity, its development, and impact on the village‟s space-economy. I conclude this chapter by discussing the theoretical and historical implications of this short but intense transformative period for the understanding of the periurban formation process under colonial rule.  A Small but Independent Commune-Village Hòa Mục during “Western time” (thời Tay) or “French occupation time” (thời pháp thuộc), as villagers call the period that goes from the late 19 th  century to the Japanese occupation of Tonkin in 1940, was a small place in both demographic and spatial terms. A census conducted by the colonial authorities in 1923 reports a village population of 203 (Hà Nội Mới January 25th, 2007). While this figure should be regarded cautiously, 3  it nevertheless provides a useful comparative basis. Most importantly, it suggests that Hòa Mục was far below the average village population of 910 inhabitants reported for the Red River Delta during this period (Gourou 1965 [1936]: 233). A map drawn by the French towards the end of the colonial era similarly indicates that Hòa Mục occupied a small territory (figure 5). Based on this 1941 document, I estimate that the village‟s territory covered only about 65 hectares (180 mẫu4), of which about five hectares were devoted to settlement and the rest to agriculture. This places Hòa Mục in the bottom-tier of the Red River Delta villages in terms of land area (Gourou 1965 [1936]: 15).   3 The French colonial state, like the dynastic state preceding it, was at pains to obtain accurate population data from local authorities. 4 The traditional units used to measure land in Northern Việt Nam are the mẫu (3,600 m2) and sào (360 m2).  39   At the beginning of the 20 th  century, Hòa Mục belonged to the canton (tổng) of Dịch Vọng, in the district (huyện) of Hoài Đức and province (tỉnh) of Hà Đông. This positioned the village in the “rural” administrative world. Contrary to the concession of Hà Nội, where the French could legitimately impose their institutions and rules, the province of Hà Đông (like all other “rural” provinces in the Protectorate of Tonkin) was only indirectly administered by the colonial power. Outside of Hà Nội and of the port of Hải Phòng, the Huê Court acted as an intermediary between the French colonial state and local populations. The local mandarinate was the administrative arm of the colonial state‟s power in rural provinces. There, it enforced a version of the last dynastic code (the Gia Long Code or Code Annamite, as the French called it) that was progressively reformed by the colonizers.  Figure 5 Administrative map of Hòa Mục’s immediate region in 1941  Source: Adapted by the author from the Plan de la Ville de Hanoi et du District de Hoan-Long (sic), Service du Cadastre du Tonkin  40   A particularity of Hòa Mục during the precolonial and colonial periods was its double status of “village-commune” (làng-xã) (VCHĐB 2005: 8).5 The meaning of this status is complicated by the various terms used to refer to socio-political and territorial-administrative entities in Việt Nam (xã, làng, thôn, châu, trang, trại, etc.) and by their changing definitions over time and across space (usages and meanings vary between Northern, Central, and Southern Việt Nam). This chapter is not the place to comment extensively on this issue; such discussion can be found elsewhere (cf., Papin 1997: 33-34). For the purpose of this dissertation, I will limit myself to a clarification of the meaning of the terms xã and làng which are most relevant for understanding the case of Hòa Mục.  At the risk of oversimplification, the xã or “commune” can be said to correspond to the “political space” of rural populations while the làng or “village” 6 corresponds to their “life space.” (Such distinctions echo the idea of “administrative” versus “natural” village found in China studies).  The political-administrative organ responsible to allocate the state‟s demands among local households and to regulate their local activities in the name of the monarch or colonial power operated at the scale of xã. 7  This organ was known as the “Council of Notables” (hội đồng kỳ mục). It was led by a group that consisted of: the first and second notables (the tên chỉ and thứ chỉ), a group of men occupying communal administrative functions (the lý dịch), former mandarins and literati (the quan viên and tư văn), elderly (the lão) and the so-called “rich and honourables” (the phú quí). The group of the lý dịch was the administrative arm of the council. It included three essential positions: a mayor (the lý trưởng) holding the strategic function of liaison between the Council and the local mandarinate; one or two assistants charged with executing his decisions (the phó lý), and a man responsible for local public works (the hương trưởng) (Papin 1997: 528).   5  A local tale situates the origins of this administrative status around 1425. During this period, Lê Lợi (later known as Emperor Lê Thái Tổ) reportedly rested with his troops in a village temple. One night, the deity worshiped in that temple appeared to him as an omen about his future winning of important battles against Chinese occupiers. As the story goes, the new emperor granted Hòa Mục the status of village-commune and freed its population from taxes and corvée in appreciation for this local deity‟s contribution to his victory. 6 The word “xã derives from the Hán (Chinese) compound term "xã tắc" (lit. God of the soil and grain). There is conversely no Hán equivalent for the word "làng," suggesting that this concept is indigenous to Việt Nam (personal communication with Vinh Sinh, 19 February 2004). To avoid confusion, and despite their odd fit in the Vietnamese context, I have chosen to stick to common translation (inherited from the French) of xã by “commune” and of làng by “village.” 7 The status of the commune and of the village was attested by the dynastic state in various ways. These included an official sanctioning of commune and village-level charters (hương ước), approval of Council of Notables‟ composition by the mandarinate, and official recognition of the village‟s guardian spirit (see Marr 2004, Boudarel 1991, Phan Huy Le 2006).  41   In principle, members of this council were selected within and elected by a group called the “registered inhabitants” (dân hang xã). This group included all tax-paying male villagers, between 18 and 50 years old (or 60 in some communes) whose families lived in the commune for at least three generations. While in theory anyone was eligible to sit on the council, in practice the valuing of scholarship, past office, age, and wealth, excluded the youngest and poorest segments of the population. Further limiting entry to the council was the fact that elections were only held on a regular basis for the administrative arm of the council (the lý dịch). Other positions were held in perpetuity, and only replaced through elections in cases of death or resignation. Although there was much local variation, the composition of the communal council tended to be dominated by older men who either belonged to one of the dominant local lineages, benefited from social status in the community, had landed properties, or a mix of these (Nguyen Duc Nghinh 1987; Papin 1997: 536- 54).  The Council of Notables derived its collective authority from the extensive powers over the administration of the commune‟s internal affairs vested in this institution by the central state. During the colonial period, communal councils‟ responsibilities included the allocation of land and head taxes, the administration of the military draft, and the selection of coolies for public works. The notables also selected members of the lý dịch, adjudicated internal disputes, ensured the maintenance of internal security, regulated ritual ceremonies and public festivities, saw to children‟s education, and took care of indigents (Ory 1894: 17-18; Nguyen Thu Chi 1993a).8  The jurisdiction of a given Council of Notables could encompass the territory and population of one or more villages (làng). By village, I mean a defined perimeter of housing, surrounded by agricultural land (paddy fields, grazing lands, lakes and ponds), which is exploited by a community bonded by kinship and neighbourhood ties (Nguyen Khac Tung and Bui Xuan Dinh 1993). The term “village” also refers to a place that has historically been of greater socio-cultural significance to its inhabitant than the commune (Nguyen Thu Chi 1993a; Malarney 2002; Papin 1997). To this day, rural populations continue to identify themselves as hailing from a specific village, not from a specific commune. The local collective life was, and still is, deeply rooted in lineage relationships,  8 In 1921, the French attempted to dislodge this local oligarchy and to administer rural communes more directly. They devised an attempt to implement a reform of local administrations that downplayed the role of Councils of Notables in local affairs in favour of a new elective administrative body called the “Council of lineage.” This reform, which only succeeded in producing competing authorities in rural communes, was abandoned in 1927 and followed by the reestablishment of Councils of Notables (Cao Van Bien 1993: 283; Nguyen Thu Chi 1993a: 114; Phan Huy Le 2006: 33).  42   associational life, and ritual activities that focus on the village community rather than on that of the commune:  Each lang, for example, has its own communal house (dinh) that houses the altar of the village guardian spirit. This structure serves as the primary focus for local collective ritual life with a detailed set of bi-weekly and annual rituals. Allegiance to the communal house is strong as residents of one lang will often not perform rites in other communal houses. The xa has no corresponding sacred structure, nor does it have a schedule of regular rituals. The lang has also historically been the critical unit for contracting marriages, as shown in the preference for village endogamy. Marriage within the xa was acceptable, and certainly not as problematic as marrying out of the xa, but the most sought after marriage was with someone from the same lang. Each lang also had its own distinctive characteristics in its ritual practices, such as the time of its feast or the ceremonial order, that distinguished it from surrounding lang. Finally, local moral discourse is constructed with reference to the lang and its value system. Families speak of the tinh lang or „spirit of the village‟, a constellation of values that emphasizes equality, solidarity, reciprocity and affective relations between residents. Others evaluate a person‟s relative moral worth according to their ability to live up to this code. Significantly, there is no such system for the xa. (Malarney 2002: 13-14)  From this description of the làng and xã, we see that Hòa Mục‟s double status of “village- commune” implied a congruence of political world and life world in the same place. This status gave the village community a degree of independence that was not afforded by the majority of Red River Delta villages, grouped under the same commune. Parochialism and power struggles within a multi-village Council of Notables was indeed a common feature of the local political life in the region, one that could easily hamper change and innovation (Papin 1997). This is not to say that Hòa Mục was devoid of internal tensions but rather to underscore its relative freedom to make decisions and act independently of neighbouring village communities.  In Hòa Mục, this socio-spatial unity is conjoined to the historical pre-dominance of the “Lai” lineage (hô hang—sometimes translated as “clan”). By the early 1930s, the four branches (chi) of this lineage encompassed two-thirds of the village‟s population, the remaining third being divided between the two other main lineages: the Nguyễn and the Phùng.9  This overwhelming numerical dominance of the Lai translated into its monopolizing of political power. A villager belonging to the  9 The folk history of Hòa Mục explains this kinship structure in reference to the near decimation of the village at the beginning of the 16th century B.C. As the story goes, the village was wrongly accused of having sheltered one of the wives of a dethroned emperor. Its population was massacred, upon order from the new monarch, as a punishment for this crime. Fortunately, three native Hòa Mục villagers, respectively belonging to the Lai, Nguyễn and Phùng clan, survived this mass killing. The folk story claims that they thereafter slowly repopulated the village and that this explains why for a long time it had only a small population and three lineages.  43   first branch of the Lai lineage explains: “The generation of my grand-parents and that of my great- grand-parents almost all worked as dignitaries (chức sắc) in this village. Members of the Lai lineage ruled this village for a long time because in the past, many of them were from Confucian families (nhà nho giáo), educated family.”  As will be discussed later in this chapter, the Lai lineage proved instrumental in transforming the village economy. A favourable position in the socio-political landscape of Hòa Mục allowed the members of this lineage to take advantage of new wealth accumulation opportunities that opened up at the end of the colonial era. Before I turn to this issue, let me describe the socio-economic situation and livelihood of the local population at the eve of this period of profound changes.  An Economy of Complemented Agriculture Up to the beginning of the 20 th  century, the main pillar of Hòa Mục households‟ subsistence was wet rice cultivation. Elder Vinh, 10  a 70 year-old villager remarks: “In the past, fields were the basis; if you had a family, you also had a field (có nhà, có ruộng).” For the local population, “having a field,” meant cultivating two crops of inundated rice per year.11 This activity nevertheless fell short of meeting most households‟ basic subsistence needs. As elder Vinh further notes:  “Even if you worked very hard, the fields were not enough to feed your child; you also had to do other things in order to survive.” The socio-economic and moral framework within which villagers engaged in these “other things” is important for understanding the occupational “thickening-up” which later happened under the French.  Wet Rice Agriculture: Open Economy, Moral Economy By the beginning of the 20 th  century, the region to which Hòa Mục belonged boasted nearly 10 inhabitants per hectare; a human density among the highest in the Red River Delta region and nearly double that on the island of Java at the same period (i.e., 4,3 inhabitants/hectare) (Papin  10 In this study, the title “elder” stands for the Vietnamese person-referring forms bác (older uncle/aunt) and ông/ bà (grand‟ma/grand‟pa). This is meant to identify the large group of 70 years old and above villagers who participated in this study. I do not use other common fictive kinship, person-referring forms (sister, brother, uncle, etc.) to refer to younger interview participants but instead indicate their age. 11 In Hòa Mục, as elsewhere in the Red River Delta region, the basic rice crop, known as the 10th month crop or seasonal harvest (vụ mùa), was reaped around December after transplanting in July-August. A second rice crop, called the 5th month crop or Spring harvest (vụ chiêm) was grown on the same fields, using faster maturing seeds. This second crop was usually reaped in May-June after transplanting in January. Some villagers grew a third, so-called “Winter crop” of manioc, batatas, taros, or yam in the gap between these two rice crops on higher grounds next to the village settlement.  44   1997: 359). This high population density translated into high labour availability, especially during the slack periods characteristic of the wet rice cultivation calendar. In the 1930s, Gourou (1965: 570-72) estimated that the average Red River Delta household required no more than 125 labour days to attend to its paddy fields. During an average year this left greater than 200 days during which households could devote their attention to activities associated with agricultural intensification such as the improvement of irrigation and flood protection infrastructure or inter- cycle vegetable farming. 12  As hinted by elder Vinh, outside of the labour intensive transplanting and harvesting periods, households could dispatch members in pursuit of complementary income outside the village and outside of agriculture. 13   Villagers call the additional activities they engaged beyond agriculture “side-occupations” (nghề phụ) or “extra-work” (việc làm thêm). The use of the word “phụ” (lit. help, secondary to, minor) is important. It captures the subordinate position of these activities within the household economy. Hence elder Thanh describes the village‟s economy at the beginning of the 20th century in these words: “First, the whole village tilled (cày cấy); that is, each family tilled and cultivated rice (canh tác lúa). Men and women could do extra work (làm thêm) to improve their lives. However, the main occupation was always agriculture. Cultivating rice always came first.” In those days, elder Thanh further explains, “only dirt poor families (bần cố nông), those that did not have an inch of field (không có tý ruộng nào), turned to extra work as their main source of livelihood.”  The idea of agriculture as the most important component of the traditional rural economy reflects a Confucian vision of the proper social order promoted by dynastic states since the late-Lê dynasty (1428-1788). This vision attached greater value to agriculture. It defined this occupation as a virtuous pursuit while stigmatizing trade and commerce (Reid 1993: 62). The traditional Vietnamese theory of class stated that society consists of a four-tiered prestige-stigma hierarchy (tư dan). At the apex were the educated scholar-officials (sỹ); below them were the farmers (nông); lower yet were the artisans and handicraft specialists (công); and at the bottom were the traders (thương). “If agriculture, which Confucian literati associated with the idea of the “root” was the focal point, commerce was nothing but an activity linked to the “branch,” despicable, demeaning,  12 Spare time was also spent in various local associational, family and village ritual activities, banquets and festivities of all sorts (see Gourou 1965 [1936]: 270-2). 13 For further discussion on the “openness” of wet-rice agriculture to intensification and non-agricultural activities under conditions of high labour availability in different parts of Asia, see among others: Geertz (1963); Bray (1986); Gourou (1984).   45   and which needed to be pruned, from time to time, to ensure society‟s stability” (Papin 2001: 164, my translation).  As noted by Woodside (1988: 31), these theories only partially matched the Red River Delta region‟s social reality. During pre-colonial times, few inhabitants of the delta depended strictly on handcraftsmanship and trade to make a living. Most of the population engaged in these activities as a complement to agriculture. In practice, the distinction between the three bottom rungs of the Confucian social ladder was therefore often blurred. Yet, the overall prestige-stigma structure seems to have remained salient at the moral level. In his study of the delta‟s rural economy during the 1920s and 1930s, Gourou (1965) thus notes that in most villages—even those that had a long artisanal tradition—the practice of non-agricultural activities was seen as a demeaning occupation. Villagers‟ behaviours, the French geographer writes, reflected their assessment of agriculture as the “right livelihood”: [I]f he could, [the peasant] would give all his care to cultivation and abandon industry which is essentially an occupation for poor people seeking a complement of resources” (ibid: 521, my translation).  Livelihood activities outside of wet rice cultivation are also traditionally understood by peasant households as occurring within the framework of household subsistence rather than that of wealth accumulation. Well into the colonial era, Gourou (1965 [1936])  noted that when a household managed to generate wealth through non-agricultural activities, for instance artisanal production, it did not reinvest it into the development of this activity. New wealth was more likely to be used to acquire agricultural land, seen not only as the foundation of a more virtuous life but also as a greater base of economic security for the household and as a potential source of rental income. New wealth was also more likely to be invested in the local prestige economy through the holding of banquets, land donations to village institutions, the purchase of honorific titles on the Council of Notables, construction or renovation of ritual buildings, etc. Wealth could also flow into the education of male children. Such expenses were preferred as they provided households and entire families with avenues for social mobility for generations (Nguyen Thu Chi 1993b; Nguyen Khac Tung and Bui Xuan Dinh 1993; Malarney 1998).  The reworking of this moral system under new economic conditions of colonialism (which will be discussed in the next section) is perhaps one of the most important—and yet least acknowledged— legacies left by the French in Tonkin.   46   Men Chop Wood, Women Catch Crabs The particular mix of side-occupations traditionally practiced in Hòa Mục gave rise to the local adage “men chop wood, women catch crabs” (đàn ông bổ củi, đàn bà bắt cua). The “wood- chopping occupations” (nghề bổ củi) took two distinct forms. The first involved groups of men who left the village during the slack periods of the agricultural cycle to find work in Chinese-operated wood mills (xưởng củi) located on the banks of the Red River, upstream of Hà Nội. There, villagers cut timber (probably rafted on the river from the forest areas surrounding the delta) into firewood which was then sold in the city.  In its second form, wood chopping involved the hiring of groups of villagers by urban residents to cut trees in the city. Here we see how Hòa Mục‟s spatiality, and in particular its location at a half-a- day‟s walking distance from the city, made this secondary activity possible. As a native villager explains: “The citizens of my village had an occupation of sawing trees. That is due to the village‟s periurban (ven đô) situation. Some elderly organized groups to go wood chopping. All across the city of Hà Nội, people needing to take a tree down—be they French, Vietnamese or Chinese—hired our ancestors.” Both types of work are referred to as làm thuê (lit. labour hire) because they involved the selling of villagers‟ labour against remuneration.  Next to these male-dominated occupations, women and children caught tiny crabs and picked shellfish (mò cua, bắt ốc) in the inundated paddy fields surrounding the village. An elderly woman recalls that she followed her mother in the fields to catch crabs up to 1945, by which point this side- activity was disrupted by anti-colonial guerrillas. 14  Women also took advantage of the courtyards and ponds surrounding their houses to complement their diet. Virtually all the elderly I talked to, even those who grew up in the poorest of families, told me that their mothers and grandmothers used their residential plot for vegetable and fruit gardening, to raise fowl (chickens, ducks, geese), and the occasional pig. Numerous ponds and the proximity of the Tô Lịch River also allowed villagers to grow an aquatic species of morning glories (rau muống). As with the crabs and shellfish, households consumed most of their production with women occasionally selling surpluses at local markets.   14 The catching of small seafood and fish was (and is still) a common practice across the Red River Delta; one that has long contributed a welcome source of protein to the peasants‟ frugal diet (see Gourou 1965 [1935]: 574).  47   During the first decades of the 20 th  century, very few Hòa Mục households were able to move beyond the use of income from non-rice cultivation activities to cover daily expenses. Such income, if marginal, nevertheless often spelled the difference between meagre subsistence and abject poverty. Cash incomes derived from wage work and trading activities allowed villagers to buy basic necessities they could not produce on their own at nearby markets (e.g., salt, fish sauce, agrarian tools) (Nguyen Khac Tung and Bui Xuan Dinh 1993). It also allowed villagers to participate in the increasingly commoditized economy of Indochina (Murray 1980; Dang Phong 1998: 24). This tradition of complementing agricultural activities with side-occupations further provided the backdrop against which changes in the village‟s space-economy unfolded at the end of the colonial era.  The Great Transformation In the early 1930s, a new side-occupation took root in Hòa Mục. According to villagers, it all began when a man named Hai Đằng, originating from a village located in the southern part of the delta, married the daughter of Mr. Lai Bá Diệp, a native who then occupied the prestigious function of village mayor (lý trưởng) and was about to be promoted to second notable (thứ chỉ). According to elderly villagers, Mr. Đằng learned weaving (nghề từ dệt) and knitting (nghề từ dệt kim) techniques during a trip he made to China. As the story goes, Mr. Đằng and Diệp acquired a few weaving looms (máy móc) and opened the first weaving workshop of Hòa Mục. Thereafter, wealthy households and landowners—most of whom belonged to the Lai family—followed their example. These people bought looms and opened private workshops across the village and, in their wake, profoundly transformed the village‟s socio-economic system.  The Rise of the Weaving Craft The weaving craft spread surprisingly fast across the village. Villagers estimate that, by the early 1940s, only a decade after Mr. Đằng and Diệp had started production, about 20 different workshops ran in Hòa Mục. By then, one villager noted, “the whole village weaved.” Each of these 20 workshops was staffed with two or three members of the owning household and one or two fellow villagers, related or not, to the owners. Young native boys and girls learned the new craft from their parents or employers at the age of 12 or 13. Women who married into the village were also trained to work in the cottage industry. The demand for workers nevertheless rapidly outstripped the small local workforce. As a result, by the early 1940s, workshop owners sourced extra labour from the nearby villages of Cổ Nhuế and La Phù.   48   The workshops of Hòa Mục produced cotton undershirts for adults (áo may ô) and sleeveless shirts for children (áo ba lỗ) along with woollen sweaters (áo rét) and socks (dệt bít tất). The village integrated all production and distribution steps, from spinning and dying cotton thread and wool fibre, to weaving, cutting, assembling, transporting, and retailing the finished products. A man whose family owned a weaving workshop explains: “We bought the thread and the products to dye it, weaved the thread into fabric, cut and then sewed this fabric into shirts. Then, when all was done, my mother took it downtown to sell.” As was common in the Red River Delta during this period, it is most often women who ensured trading (bán buôn) (Chi Huyen Truong 2001; Malarney 1998). These traders were generally wives or sisters of workshop owners who took their household‟s production to the city to sell it to individual customers and wholesale agents (đại lý).  Socioeconomic Relationships Reconfigured Not all families operated a workshop within this rising cottage industry. As one villager explains: “your family had to have the right conditions to do it; only then could you open a weaving workshop.” Families that did not own a weaving loom could nevertheless subcontract specific segments of the production process such as reeling fibres or cutting and dying the fabric. This is the case of this woman: “I just spun fibre. I took it home with me to do the spinning. I had a small machine to do it by hand. I was the only one doing this job in my family: My husband was a worker, he chopped wood outside of the village, and the kids were still too small to work.”  But subcontracting was not the norm. For poorer villagers, involvement in the weaving industry more commonly meant working for hire (làm thuê) for other families. A woman in her 70s whose parents weaved for hire recalls: “My parents were very poor. They had no farmland whatsoever. They could only work for others and earn a miserable living […] they didn‟t have the money to buy a machine [to weave]. Of course they went to work for hire, what else could they have done? Their lives were like that.”  Villagers belonging to underprivileged classes, like the parents of this woman, typically worked for wealthier members of their kinship group or for neighbours. This established new socio-economic relationships across the village. Previously, agricultural work and side-occupations operated mainly within the household. In this basic economic unit, all could draw on a common income. Moral principles based on gender and age determined resource and labour allocation. As mentioned in the discussion of wood-chopping occupations, waged labour was not new to Hòa Mục villagers. Yet,  49   the textile industry contributed to establish wage labouring, as a socio-economic relationship, more firmly by introducing employer-employee relationships across and within kinship groups.  Despite their growing importance in the village economy, commodity production activities did not dislodge agriculture. These activities rather resulted in the reconfiguration of households‟ resource allocation and labour management patterns within an increasingly diversified economy. This movement is especially salient in the case of wage workers and subcontracting households. Both groups tended to shift labour to maintain agricultural production while maximizing income from the cottage industry.  Thus fewer families sent members to Chinese wood mills, preferring the more lucrative textile industry. In the late 1930s, a skilled weaving worker could expect a monthly salary of 12 to 18 Indochinese piastres.15 This was twice the wage offered to unskilled labourers in the wood-chopping and agricultural sectors (6 to 7 piastres). Meanwhile, many households also adopted a new labour distribution pattern that consisted in sending elderly women to till the fields and younger members (especially males) to work for hire in textile production. As this elderly woman remarks:  Families went both ways: women worked in the fields while men worked in the craft. All the old women did agriculture then. They could do most of the work alone as the fields were small. Men would only give them a hand for transplanting and harvesting. [...] Of course, both the fields and the craft were necessary; had we abandoned either we would have died from hunger.  Some households, especially those who subcontracted part of the textile production, chose to rent out their agricultural land to small landholding or landless families under tenant farming arrangements. This strategy freed supplementary labour for textile production activities. A few households also took the risk of leaving agriculture (thoát ly) altogether in order to reallocate most of their capital and labour to cottage industry production. I heard of two households who chose this strategy. This was a bold move in a context where, as Murray (1980: 488) writes “small-scale owner-occupiers only reluctantly relinquished their land—no matter what the market situation or the price of land was—because these plots served as the foundation of their social existence and economic survival, as well as that of their heirs.” The two households who parted with their  15 In the 1930s, the average exchange rate of the Indochinese piastre varied between 1.7 and 3.8 to the US dollar (Bulletin Économique de l‟Indochine 1940, cited in Dang Phong 1998: 25).  50   agricultural land belonged to the Lai lineage, suggesting that this clan might have provided a safety net and insured its members against the risks of an entrepreneurial venture.  Weaving and knitting developed based on Hòa Mục‟s historical tradition of supplementing rice cultivation with non-agricultural activities but it differed from wood chopping and crab catching in important ways. Activities associated with the cottage industry carried on year-round rather than being limited to periods of idleness in the agricultural calendar. Textile production involved a larger number of villagers than any previous side occupations. Villagers estimate that, by the 1940s, this industry involved about 90% of local households in one way or another and was the main source of livelihood of about half of the local population. Contrary to the subordinate position of previous side occupations, the cottage industry also became a morally legitimate source of wealth accumulation, and an acceptable way to reinvest surplus income.  The rise of the weaving craft undoutedly transformed the local class structure. Some elderly villagers indicated the emergence of a new social class of wealthy entrepreneurs, pointing to its inscription in physical space through a new housing type. Three of the most successful weaving families indeed rebuilt their “traditional” houses into two-storey buildings, an unprecedented housing form in the village‟s landscape. As one villager observes:  During French times, these houses towered over the village. They could even be seen above the communal house (đình). These buildings had their first floor made of bricks, and the second one made of wood. [...] Living conditions were pretty good in these houses. That‟s because they belonged to weaving families. Only they had the money to build such houses. Of course, peasants like us had no such means.  When compared to the housing conditions of the poorest households, this description of the wealthy entrepreneurs‟ houses suggests that weaving did not bring the same fortune to all villagers. During the colonial era and beyond, the majority of villagers indeed continued to live in ramshackle houses made of bamboo walls and covered by a palm-leaf roof. Many villagers explained that these houses (known as nhà tranh mái lá or nhà rơm rạ) leaked during heavy rainfalls and threatened to be taken away by storms. Further investigation is nevertheless required to understand the impact of the rise of the weaving craft on patterns of socio-economic differentiation in Hòa Mục.  Alignment of Urban and Foreign Markets The late 1930s and early 1940s witnessed the transformation of artisans into small manufacturing entrepreneurs. This process led to the relocation of many workshops to the city. A member of a  51   former weaving family recalls that this movement was, once again, initiated by Mr. Đằng and Mr. Diệp: “They both grew very prosperous (rất phát tài) and so, in the late-1930s, Mr Diệp bought a house on Hàng Phèn street [in the Old Quarter] and continued to do his business there.” Several households followed these two men, seen as the modern founders of their village‟s craft. They relocated their families and workshops to the Old Quarter and to an area that corresponds to today‟s Đống Đa district.  For some years, production activities were divided between the inner city and the village. But the movement to the city progressively took over. This is once again an indication of the manner in which the village‟s relative proximity to the city influenced its socio-economic development. Because Hòa Mục was located relatively near Hà Nội it could benefit from its urban markets and from its national and international trading networks. This allowed production to continue in the village for a while. Poor transportation infrastructure outside of the urban core, however, impeded the flow of goods, and the slow development of the electric grid became a brake on the modernization of production.  While appearing to be relatively close to downtown Hà Nội from a bird‟s eye view, Hòa Mục was in fact physically cut from the city by the Tô Lịch river which was wider then; (it remained an important barrier until the end of the 20 th  century). To reach the urban core, villagers needed to cross one of two bridges (the Paper Bridge or the Crossroads Bridge). This involved a relatively long detour to the north or south of the village. The trip to the city thus took over two and a half hours on foot and much longer if one pulled a cart full of textile or yarn. The poor quality of the roads in the countryside further impeded the flow of people and goods between the village and the city. Many segments of these roads in fact were only small trails or mere paths along rice-fields, impassable during the rainy season.  Compounding this transportation issue was the mechanization of the weaving and knitting production processes and the associated need for energy. By the 1940s, electric looms and sewing machines were integrated into the urban production process, allowing for the production of larger volumes in shorter times. Hòa Mục was not yet connected to the electric grid but the inner city had been since 1894 (Dang Phong 1998: 19). By the mid-1940s, plans were purportedly in the making to connect the village to the urban electric distribution system, but these never materialized. By then, most workshop owners had already bought electric machines and moved their production to the downtown area.  52    As a former wage worker in the textile industry observes, “After some years, workshop owners only worked a little in the village; most of them had houses out there in Hà Nội. Little by little, my village became the rear-guard (hậu phương) of the weaving craft.” Hà Nội was then seen as a more propitious (thuận lợi) place to do business (kinh doanh) than the village. A man who followed his uncle‟s family to work in the city explains: “Things were more convenient downtown. Out there, it was easy to rent or buy a house and turn it into a shop to assemble the goods. There, it was also easier to exchange goods with traders based in the city. There was no need to bring the goods from the village.”  If relocation was in most cases permanent, it did not involve the severing of ties with the village. Most families retained their landed properties in the village thus in effect becoming absentee landlords. Up to the Land Reform (in 1955), it was customary to rent one‟s house to members of one‟s extended family who still lived in the village and/or to rent out one‟s paddy fields to fellow villagers. Families living in downtown Hà Nội remained involved in the village‟s life through their continuing participation in the multiple ritual activities held within their kinship group (ancestors‟ cult, weddings, funerals, lineage and longevity celebrations), by local associations (Buddhist, scholars, age-groups, etc.), and by the village community as a whole (Lunar New Year, cult of the village‟s guardian spirit, etc.). Connections with the village were also maintained through the continuous sourcing of labour from Hòa Mục. A new pattern of pendular migrations thus emerged with increasingly specialized wage labourers commuting back and forth between the village and the city on a weekly or monthly basis.  The goods produced in the urban workshops of former Hòa Mục residents also changed and took new commercial routes. By the beginning of the 1940s, workshop owners had enlarged their trading networks considerably. Many of them then dealt with Hà Nội merchants who supplied them with yarn imported from British India.16 Some signed contracts with traders based in the city to export undershirts, socks, and sweaters to France. A man whose family had a large weaving workshop during this period explains how the response to foreign demand shaped his family‟s production:  We started to design winter sweaters (áo rét) to meet the requirement of the traders. We only produced these [sweaters] in the city because we needed the electric machines. I  16 When the French occupied Tonkin in the late 18th century, they found that native weavers were importing large quantities of cheap yarn from foreign countries, including British India (Mitchell 1942 cited in Murray 1980: 349).  53   remember that they were nicely finished on the edges, with front pockets. [...] In the winter, we signed a contract to buy cotton thread for the summer, and at the end of the summer, we signed a contract to buy wool for the winter. In this way, we produced according to the market‟s demand.  Figure 6 Periurban villages specializing in textile industry production in the Red River Delta, circa 1930   The region of Hòa Mục is encircled. One dot = 20 artisans Source: Gourou 1936: 462   54   Local entrepreneurs thus wove their industry, if only on a small scale, into the colonial world economy. Such integration into international trading networks was not unique in the region. Between the two World Wars, other Red River Delta villages developed mechanized, wage labour based, export-oriented industries. These include localities near Hòa Mục which also specialized in garment production (see figure 6) but also villages located further out in the delta that produced rattan furniture, sedge baskets, and fine handicrafts (Dang Phong 1998: 23). Illustrating the integration of local cottage industries into regional trading networks, a resident of Hòa Mục originally from the district of Phú Xuyên told me that, prior to 1945, her father and uncles travelled every other year to Indonesia to sell the mother-of-pearl buttons they produced to Chinese and Dutch traders.  As in other Red River Delta villages, cottage industries developed in Hòa Mục despite the French colonial authorities‟ lack of interest in (if not outright opposition to) local industrialization. As Murray (1980: 195) observes: “the political role of the colonial administration was largely confined to enforcing the hegemony of metropolitan monopolies against foreign intervention and indigenous competition.” The desire to safeguard French manufacturing interests translated into protectionist policies that restricted the development of manufacturing activities in Tonkin. The impact of these policies was felt in the systematic failure to provide any direct assistance to local industries, in tariffs and barriers limiting imports of raw materials originating from countries other than France (especially the U.S., Japan, and China), in banking and loan policies discriminating against manufacturing entrepreneurs, and in the exclusion of this group from local trade associations.  French and indigenous manufacturing activities nevertheless emerged stimulated by the distance to France, the abundance of low-cost labour, and the growth in domestic demand for more affordable articles than those supplied by French imports (Murray 1980: 348). In line with these observations, Gourou (1965: 535-38) concluded his seminal social geography of the Red River Delta by recommending that colonial state barriers to indigenous commodity production and trading activities be lifted and incentives be provided for the development of rural cottage industries. In this way, the French geographer argues, the delta could see a form of in situ industrialization which he deemed more appropriate to the social and economic conditions of Tonkin than modern urban industries that involved large-scale rural-to-urban migration.  Conclusion This chapter discussed the changes in Hòa Mục‟s livelihood trajectory during the colonial period with a particular focus on the unprecedented transformations that occurred on the eve of the anti-  55   French resistance war (1945-1954). The discussion revealed how a thriving cottage industry grew out of Mr. Đằng and Diệp‟s entrepreneurial initiative and how this profoundly transformed the village‟s space-economy. Beyond the diversification of the village‟s economic basis, I have shown how the textile industry reshaped local social relations, resource allocation, and mobility patterns. I have also discussed how new connections to urban and international trading networks integrated this small village within a broader society and economy.  As mentioned in the introduction, the processes and outcomes of these changes have little to do with the typical portrait of the traditional village sketched by colonial scholars (e.g., Ory 1894; Gourou 1965) and repeated in classic texts written more recently by Vietnamese scholars (e.g., Nguyen Thu Chi 1993a, b; Phan Huy Le 1993). In concluding this chapter, I wish to highlight some of the important ways in which Hòa Mục departed from the ideal-type of the “traditional village” of the Red River Delta. This discussion seeks to debunk ideas still surprisingly prevalent in academia, the mass media, and professional circles of architects and planners in Việt Nam. This exercise has more than theoretical and historical implications. It also has contemporary significance because imagined conceptions of the village are important markers that serve to assess the current periurbanization process and inform urban and regional planning propositions.  The socio-economic transformations observed in Hòa Mục shed some light on the formation of town and country under colonial rule in northern Việt Nam. This case shows that, notwithstanding the colonial era‟s administrative boundaries and differences in regulatory regimes, Hà Nội and its surrounding countryside developed a strong socio-economic interdependence. The proximity of the city, of its growing urban market and rising demand for commodities, and of its trading opportunities and linkages to regional and international commerce played an important role in the development of industrial occupations in Hòa Mục. These factors alone would not account for the extremely rapid socioeconomic transformation observed in the village during the late-colonial era.  I have argued that local characteristics of this village served as preconditions for its early industrialization. This analysis echoes elements of the “desakota” model proposed by McGee and others to account for the unique patterns of urbanization observed in various wet rice growing regions of East and Southeast Asia.17 As with desakota regions studied during the late 1980s and  17 In explaining this outcome, the desakota model emphasizes such factors as the high rural population densities of wet rice agricultural regions surrounding large cities, the shedding of labour out of agricultural production in the wake of the green revolution technological and institutional changes, and the rapid expansion of communication and transportation technologies across vast swaths of territories. These factors are deemed responsible for a rapid, region-based, transition to urban societies that still preserve some rural characteristics and relationships (see McGee 1991, 1995a).  56   1990s, the establishment and development of a cottage industry in Hòa Mục was facilitated by the combination of low land availability, surplus labour, and slack seasons typical of wet rice cultivation. These conditions underpin the local tradition of supplementing wet rice cultivation with non-agricultural “side-occupations,” and this tradition, in turn, supported the occupational thickening through the rapid adoption of new cottage-industry activities by villagers.  Still reflecting on the desakota model, the spread and deepening of commodity production and exchange in Hòa Mục did not coincide with the separation of industry from agriculture. Villagers rather engaged in non-agricultural activities while maintaining their agricultural economic base. The French historian Philipe Papin (2001: chapter 13) similarly indicates that a unique periurban zone, distinct from both city and country, took form as early as 1921 on the near periphery of Hà Nội. By then, about 40 per cent of the households living within commuting distance from the city depended on economic activities associated with the primary sector (agriculture, fishing, market gardening, silk worm raising, salt extraction, etc.). The livelihood of the remaining 60 percent combined agriculture with handicraft production and/or emerging urban professions (e.g., mechanics, house painters, decorators).  Here again spatiality is at play. Greater access to the city afforded by populations located in close proximity to Hà Nội was certainly important in establishing a historical relationship between urban core and periurban areas. The intensity and variety of secondary activities increased dramatically on the near periphery of the city presumably because of easier connections to urban markets and trading networks. Geertz (1968) suggested a similar explanation for the occupational thickening observed on the island of Java during the same period. The fact that profound socioeconomic changes occurred around Hà Nội without much contribution of colonial economic policies and in spite of the colonizers‟ territorial administrative divisions shows how interactions within a given space/territory might supersede state intentions or power.  As observed in other periurban villages of Hà Nội during this period (Papin 2001: chapter 13; Malarney 2002: chapter 2), the industrialization process in Hòa Mục contributed to a blurring of the rural/agricultural and urban/industrial distinctions and categories. It created new rural-urban linkages and a strong interdependence between the inner city and the villages surrounding it. As will be seen in the next chapter, this periurban character—founded on persistent periurban-urban linkages—did not disappear after independence. It was rather purposefully maintained by the DRV government and thus remained a distinctive trait of Hòa Mục throughout the socialist transformation process and beyond.   57   CHAPTER 4 Uneven Socialist Revolutions (1940-1965)        It is difficult to reconstruct what happened in the village in the period of upheaval that started with the Japanese coup (nhật đảo chính) against the Vichy French in 1940.1 During the anti-colonial resistance war that followed (1945-1954), French, Japanese, and Chinese in turn occupied the region to which Hòa Mục belongs. This area was also the site of under-cover guerrillas (vùng hoạt động bí mật) led by Việt Minh troops. In addition, the village witnessed a succession of catastrophes during this period, including devastating fires, famine, and floods (VCHĐB 2005: 22). Due to the intensification of battles, families began to evacuate the village in 1946. Most of them relocated temporarily to live with members of their extended family in rural areas where the war raged less violently. Few returned home before peace was settled again, in 1954.  By then, revolutionary partisans and party bureaucrats had already been working for some years to recast Hòa Mục‟s institutions into the socialist mould. During the early 1940s, the leadership of Việt Nam Workers' Party (VWP)2 reorganized the territory of suburban Hà Nội (officially still under French control) to “increase the strength of the People and of the Resistance.” As a result, Hòa Mục was merged with the neighbouring commune of Trung Kính to form the new rural commune (xã) of Trung Hòa (Bùi Thiết 1985: 268). At first, this administrative integration aimed to coordinate guerrilla activities, but by the late-1940s, the official raison d‘être of the new commune had changed. It was now presented as one of the key channels through which the newly independent Vietnamese state was to carry out its ambitious socialist transformation project (VCHĐB 2005: 8).  1 This period of upheaval began during World War II when colonial France aligned itself with the Vichy government. The colonial authorities of Indochina then authorized the Japanese to enter and use the Vietnamese territory and resources for military purposes. This collaboration with the colonial authorities ended at the beginning of 1945 when the Japanese staged a coup, overthrowing the French civil administration, disarming the French troops, and taking full control over Indochina. 2 The first Marxist-Leninist Communist Party, founded by Hồ Chí Minh in 1930, was called the Vietnamese Communist Party (Việt Nam Đảng Cộng Sản). This party was renamed Việt Nam Worker‟s Party (Đảng lao động Việt Nam) in 1951. Following the country‟s reunification this name changed again to Việt Nam‟s Communist Party (Đảng Cộng sản Việt Nam).  58   The previous chapter explored how the coming of market relations to textile production transformed local livelihoods in the village. One of the issues raised was the strong interdependence that took shape between periurban places, such as Hòa Mục, and the inner city in terms of labour, industrial production, and trade. This chapter considers the changing possibilities and limitations of livelihood in the village during the early years of the socialist transformation. During this period, which lasted from 1940 to 1965, the post-colonial state was very much inventing itself with the ideological and material resources at its disposal. This was an open-ended and largely experimental period for the new leaders. They then flexibly sought to consolidate their authority and to rebuild national bases of production destroyed by nine years of anti-colonial war and French sabotage.  I would argue that, in the near periphery of Hà Nội, these strategies involved the uneven application and manipulation by the state of its own socialist revolutionary policies. Such political use of regulatory flexibility contributed to turning periurban areas into what the anthropologist Aihwa Ong (1999, 2006) has called “zones of exception.” As discussed in the introductory chapter, for Ong, the nation-state is a key institution in structuring spatial orders. Building mostly on recent cases, and in an attempt to refine our understanding of neoliberalism, she points out that the state territorialization practices might rely on a modulation of its regulatory power across space. The result is what she calls “graduated zones” (Ong 1999: 215-7) whereas “the population in different zones are variously subjected to political control and to social regulation by state and non-state agencies” (ibid: 219).  In a very different historical period and socio-economic context, this is precisely what we can observe at the edge of Hà Nội during the early post-independence year. The geopolitical context I am about to describe has little to do with the neoliberal governing strategies discussed by Ong. Yet the core phenomenon is the same, i.e., the political manipulation of regulations across space by a nation-state seeking to establish its authority and legitimacy. In the case of periurban Hà Nội, the need to maintain and improve food supply and economic production in the early post-independence years, underpinned the repeated bending and localized adjustments of revolutionary transformation rules. While this strategy served state-building purposes, it also contradicted the VWP‟s objective to transform the entirety of Vietnamese society and economy along socialist lines.  The zone of exception that took shape on the edge of the capital city had important implications for a place such as Hòa Mục. Most importantly, it allowed villagers to carry into the socialist period elements of the market relations forged with the city during the colonial era. The continuing flow of people, goods, and ideas between the village and the city established Hòa Mục as a distinctively periurban place. The zone of exception also created institutional gaps that villagers inventively  59   turned to their advantage. One of these gaps concerns the socialist work units established in the village during the early 1960s. Villagers seeking to get the best of both the rural-agricultural and urban-industrial institutional worlds contributed to turn Hòa Mục into a hybrid place. The village entered into the so-called “subsidy era,”3 firmly sitting on the supposedly rigid divide between the rural and urban worlds that the socialist revolutionary process had sought to establish.  The discussion of these ideas is organized as follow. Based on secondary literature and policy documents, I first explore the motives underlying the unique approach to the socialist transformation process adopted on the outskirts of Hà Nội. I then turn to the actual manifestations and outcomes of this process in the particular case of Hòa Mục. The discussion explores two important episodes: the Land Reform and the integration of the village‟s population into the socialist work units system. In each case, I discuss why the DRV leaders devised or adopted comparatively soft, inclusive, and accommodating policies in periurban areas. I also show how villagers took advantage of this “zone of exception” to preserve distinctively periurban characteristics amidst profound institutional restructuring. I conclude by highlighting the marks left on the village by this decade-long process of socialist transformation and the implications for territorialisation process during the subsidy era.  The Outskirts as a Productive Belt As discussed in the preceding chapter, by the time the socialist government rose to power, Hà Nội was enmeshed in a complex system of economic exchanges with periurban villages. During the first post-war years, the leadership of the newly created Democratic Republic of Việt Nam (DRV4) was careful to protect some of these rural-urban linkages, especially those which maintained the productive capacity of periurban zones. These zones were indeed not only essential to the functioning of the capital city but also to the reconstruction of the weak, war-ravaged economy.  Feeding a Growing City The First Indochina War (1946-1955) revealed just how dependent Hà Nội was on surrounding rural areas for its food security. Throughout most of the 1940s, and during the first half of the 1950s, the whole of northern Việt Nam suffered from recurrent rice shortages and from the threat of generalized famine (Thrift and Forbes 1986: 70; Dang Phong 1999). Feeding cities proved a major  3 Northern Vietnamese generally refer to the period from 1955 to 1986 as the “subsidy era” (thời bao cấp). This expression emphasizes the redistributive aspect of the revolutionary econonomy and the central role of the state in this process. 4 The Democratic Republic of Việt Nam (Việt Nam Dân chủ Cộng hòa) was first proclaimed by Hồ Chí Minh in 1945. The new nation-state was officially recognized by the international community through the Geneva Conference of 1954.  60   challenge, especially between 1947 and 1953, when Hà Nội was cut off from the surrounding countryside by French encirclement. During these years, the lack of access to agricultural areas and the absence of grain reserves made urban residents realize how unbearable life in the city could be in such circumstances (Turley 1975: 371).  For some years, the problem of supplying food to urban residents was eased by massive evacuations to the countryside. By 1949, only 10,000 people still resided in Hà Nội, a tremendous decrease compared to the 120,000 inhabitants surveyed on the eve of the resistance war in 1943. The problem, nevertheless, resurfaced with the return of the French to Hà Nội after the end of World War II. The city‟s population then rapidly recovered and expanded well beyond its pre-war level. The population continued to increase after the DRV government regained power over Hà Nội, reaching 650,000 people by 1960 (Turley 1975: 373). 5  The subsistence of this large population depended primarily on food produced in the near periphery. Maintaining rural-urban linkages was therefore fundamental to ensure the city‟s food security and political stability.  Restoring Commodity Production Periurban zones also sheltered a large number of small-scale handicraft workshops and manufacturing firms. These enterprises were far from the socialist ideal of heavy industries. In the aftermath of the first Indochina War, they nevertheless constituted the bulk and most effective segment of the DRV‟s commodity production sector. As discussed in the previous chapter, the French only weakly developed the manufacturing sector in Tonkin; (the focus remained on extractive industries in agriculture and mining). During the war, the colonial power even dismantled or sabotaged these few manufacturing industries (Robequain 1939; Miller 1947).  By contrast, the handicraft and small-scale manufacturing sector developed by indigenous people was still thriving. Commodity production was stimulated by the war economy. Throughout the 1940s, household-based and other small-scale enterprises supplied the population in liberated areas with building materials, agricultural tools, and goods for daily use as well as the military materiel. This indigenous production was encouraged by war policies that called for the support of all producers, be they from the collective or private capitalist sectors (DiGregorio 2001: chapter 2). In the first years that followed the final retreat of the French, the handicraft and small-scale industrial sector proved to be of continuing importance. Official statistics indicate that 111,300 workshops  5 It is unclear what pulling factors motivated this migration from rural areas. Turley (1975: 373) and Logan (2000: 137) pointed to the service and supply economy which developed rapidly during the last years of the French colonial regime. The independent government planned to send part of the newly arrived migrant population back to the countryside. It is nevertheless unclear whether these policies were ever implemented or whether they simply failed to achieve any population reduction (see Nguyen Duc Nhuan 1978).  61   were active in 1955, many of which were located in periurban areas, employing roughly 300,000 artisans (Nguyen Xuan Lai 1979, quoted in DiGregorio 2001: 89). The productive role of this population was assessed by party ideologues as crucial for the reconstruction of the DRV‟s economy.  Protecting Productive Capacities During the design and implementation of revolutionary policies, pragmatic reasoning informed the implementation of “special” rules in periurban areas. Early experiments with socialist reforms in liberated zones cautioned the party organs against the aggressive application of class struggles, coercion in the redistribution of resources, and expedited agricultural collectivization. Such an approach led to long disruptions of food production in the more remote provinces where it was tested (see Trần Phương 1968: 108). This outcome could hardly be afforded in the productive belt surrounding the capital city. Malarney (2002: 25) similarly observes that periurban areas “controlled resources and businesses that officials did not want to disrupt for fear that it would create panic in the city or retard the development of North Việt Nam‟s economy.” He further quotes a 1956 policy document indicating that the role of the outskirts in the economic recovery process was “to produce directly to meet the need of the city” (ibid: 25).  Once peace was restored in the fall of 1954, the government‟s primary objective was indeed to restore and increase food production and the supply of basic consumer goods as quickly as possible. It also sought to provide work to urban and rural populations after the war disruptions (DiGregorio 2001: 88-90). At the same time, the party progressively tried to bring the northern Vietnamese population and economy within the fold of a national socialist transformation program. In the case of Hòa Mục, this process occurred through the implementation of the Land Reform, the collectivization of agricultural production, and the progressive nationalization of the textile industry. In each case, policy-makers adapted otherwise uniform programs and policies in a way that would preserve elements of the integrated city-country economy necessary for the functioning of the capital city region.  A Partial Land Reform The Agrarian Land Reform (Cải cách Ruộng đất) in Việt Nam spanned the period 1940 to 1956, when 810,000 hectares of agricultural land were reallocated to approximately two million peasant families (Moise 1976b). The class struggle and reallocation of resources associated with the reform occurred in two distinct phases. The first phase (1940-53) relied mostly on a rent reduction program and on the redistribution of communal or abandoned land to “foster the mental and material strength  62   of the peasantry, increase production, and intensify the resistance war” (Việt Nam Government Gazette 1953 quoted in Malarney 2002: 23).  A second phase went from 1953 until 1956. This phase turned the Land Reform into a direct assault on the old rural elite and on all the resources from which it drew its power. It stated aim was “to abolish the regime of land appropriation by the French colonialists, and wipe out the feudal regime of land ownership by the landlord class” (Land Law 1952, [art. 1]). This phase not only sought to redistribute land and tenancy but also draft animals, grain stocks, and agricultural tools. Its objective was to reform key social organizations and institutions traditionally controlled by political and economic elites. These included the communal Council of Notables, lineage organizations, and religious bodies. In “reforming” these institutions, the Land Reform sought to cut local elites off from “the psychological and organizational bases of their power and to destroy their prestige, and influence over the peasants” (Moise 1976b: 5).  Landlords‘ Escape Much has been written about the excesses committed by bureaucrats in the name of class struggle during the three years that this second phase of land reform lasted (e.g., Moise 1976a; White 1981). Less attention has been paid to the comparatively moderate manner in which this same reform was implemented on the outskirts of Hà Nội.  The timing of the reform on the periphery of Hà Nội contributed to make it “softer.” Indeed, reforms in the periurban zones were only started during the so-called fifth wave, which took place between January and April 1956. By the time the first reform bureaucrats arrived in periurban areas, local populations were well aware of the “leftist excesses” (i.e., bogus denunciations, errant classifications, and unwarranted executions) committed by the reformers in the more remote provinces. 6   This late implementation granted political and economic elites who still lived in periurban villages plenty of time to shield themselves and their families from the adverse impacts of the reforms, and in particular from the punitive measures associated with a landlord classification. When the first team of reform bureaucrats arrived in Hòa Mục in November 1955, most wealthy households had already left the village for the city where they were beyond the reach of the reformers. 7  Indeed, the  6 The violence with which the reform had been implemented in some localities had filtered back to the city, and was reported in the national press since at least 1954 (Moise 1976a). 7 Malarney (2002), Kleinen (1999), and Chi Huyen Truong (2001: 97) observed a similar flight of the political and economic elite during the Land Reform in three other periurban villages of Hà Nội.  63   1953 Law on Agrarian Land Reform did not extend to urban areas until 1963, by which point it was only partially implemented (Gillespie 2009: 5).  Orders received by reform bureaucrats to avoid “excesses” in periurban villages also made for a less radical reform experience. According to Moise (1976b: 202-3), the revolutionary government was then trying to persuade urban residents that the nasty rumours they had heard about the reform‟s excesses were false. He reports that groups of residents (eventually totalling 100,000 people) were thus taken to watch rather inoffensive landlord trials in periurban villages. Moise also suggests that the proximity of the national leadership compelled land reform bureaucrats to adhere more closely to official guidelines.  From what I could gather, this moderation of policy was carried out to a significant extent in Hòa Mục. Mr Vinh, a 60-year-old villager, describes the class-struggle process as follows:  In the beginning [of the reform], three people were classified as landlords, but then one lady was moved down to a lower class [category]. Then, another family was taken off the list because they had a son who had participated in the revolution. The head of this family was a fourth grade mandarin (quan tư). His son contributed to the resistance war so he was reclassified as “revolutionary landlord” (địa chủ kháng chiến). In the end, only one family was classified as real landlords. 8   The reclassification of landlords into “lower” categories mentioned by this villager illustrates the party‟s ambivalence toward local political and economic elites. It is also an indication of the kind of compromises it was willing to make to fulfill its need for legitimacy at the local level, and this, despite aggressive policies and rhetoric against landlords.  This regulatory incoherence is manifest in the creation of categories such as that of “resistance landlord” (địa chủ kháng chiến) mentioned by Mr. Vinh or that of “democratic personality” (nhân sĩ dân chủ). These categories allowed reform teams to positively discriminate towards large landowners, wealthy families, or members of the local Council of Notables who showed a favourable attitude to the party. As hinted by Mr. Vinh, such titles were granted to individuals who demonstrated that they or their children had supported the revolution by either serving in the military, as bureaucrats, or by giving assistance and shelter to bureaucrats and guerrillas during the  8 The main criterion for deciding class status during the land reform was exploitation, defined as an economic relationship in which one person appropriates wealth created by another person‟s labour. The most important forms of exploitation in the rural economy were: the renting of land, the hiring of labour to work the land, and usury (Moise 1976b: 20-21).  64   anti-French resistance war. The grandson of the man who occupied the position of village mayor (lý trưởng) at the time of the reform thus explains:  At the beginning of the land reform my grandfather was classified as a despotic landlord (cường hào địa chủ). He indeed was the village mayor (lý trưởng), but he had also served the revolution as a soldier. So, after the Rectification of Errors (Sửa sai), 9  he was reclassified as a “resistance landlord” (địa chủ kháng chiến) and didn‟t have to give away anything. Like my grandpa, most of the rich people in this village didn‟t see their land confiscated (bị thu) by the reform cadres.  Another strategy used by large landowners to escape landlord (địa chủ) or rich peasant (phú nông) classifications was to donate their properties to the government. Many did so as part of the Patriotic Emulation Movement launched in 1948. This was, in fact, a common strategy adopted by wealthy households across northern Việt Nam (Moise 1976: 160). As one elderly villager explains:  My parents were not classified as landlords, although they owned over one mẫu [3,600 m 2 ] of fields. That is because they decided to contribute it all to the state, showing their support for building socialism. They also gave up the pond in front of our old house, the cattle and everything else. In this way, we were regarded as patriots rather than landlords.  This is not to say that the land reform was not a major traumatic episode at both the social and individual levels; it very much was. In Hòa Mục, as in other villages across the DRV, the reform took place in a general climate of fear and anxiety. Even today, more than half a century after the fact, this period is still the object of bitter memories. What I observed, instead, is that the severity and harshness of the reform measures varied across time and space and that the reform‟s rules were adjusted according to different contexts and socioeconomic situations.  The Artisan Exception Thus next to the resistance landlords and democratic personalities, villagers involved in the textile sector and in other “side-occupations” also benefited from special policy measures shielding them and their families from the harshest punitive measures of the reform. Here again, we see clearly how, by selectively maintaining the complex socioeconomic structure inherited from the colonial era, the state insulated the near periphery of Hà Nội from some of the profound transformations of society and space associated with the socialist revolutionary process.  9 Starting in the Spring of 1956, the so-called Campaign for the Rectification of Errors (Sửa sai) corresponds to a retreat of the VWP from the radical excesses of the land reform that culminated in the 1956 Nghệ An rebellion (Fall 1967). The campaign was formalized by a statement of Hồ Chí Minh that asked bureaucrats to correct misclassifications and to revise the party membership‟s class-based purges so as to reintegrate individuals favourable to the revolution (see Moise 1976b: 235-50).  65   This outcome is linked to unprecedented problems posed by the implementation of the land reform in the region Hà Nội. Contrary to the predominantly agricultural areas targeted by earlier waves of the reform, many periurban households combined an agricultural base with other economic activities. In Hòa Mục, it was textile production but in other periurban villages, it could be fishing, growing flowers, basket-weaving, or rice noodle making. In other places still, villagers worked in the city as traders, clerks, labourers, or factory workers.  Many who lived on the urban edge therefore did not fit into the land reform‟s classificatory scheme and this complicated the reform‟s agenda. To solve this problem, the Hà Nội Land Reform Central Committee only redistributed land to those who were primarily engaged in agricultural production. The Committee further stated that periurban people who had “‗stable occupations‟ outside of agriculture, such as industrial workers, traders and others with pursuits that did not require landownership, could continue their professions and were not to receive land” (Hà Nội Land Reform Committee 1956, quoted in Malarney 2002: 37).  New categories were also developed by party ideologues to categorize individuals involved in non- agricultural pursuits, establish their class entitlements, and include them in a new Marxist-Leninist narrative (DiGregorio 2001: chapter 2). In line with the party-state‟s desire to maintain commodity production, special regulations were, for instance, devised to protect individuals who worked outside of agriculture. An example of this is the decision to tolerate the exploitation of agricultural property under tenant farming arrangements by artisans and other villagers having “urban occupations” outside of agriculture. Elsewhere in the DRV, this practice would most likely have led to their classification as landlords or rich peasants (Malarney 2002: 25).  The Residential Land Exception According to the official history of Trung Hòa, in 1955, the commune had 275 hectares of rice fields, a quarter of which was purportedly owned by landlords. Following the reform, this agricultural territory was redistributed among 564 farming households that each received an area totalling 720 m 2  (VCHĐB 2005: 44). Other than these agricultural lands, the residential properties of Hòa Mục households remained largely untouched. As this villager explains:  During the land reform, the fields were turned over to the agricultural cooperative, but we could keep the land for housing, we didn‟t have to give it to the state. [...] I mean the residential land (đất thổ), the land that ancestors had passed along to their children (đất của các con), the land on which we had our lineage halls (nhà thờ), that land, we could keep, all of it. Only cultivated fields (đất canh tác) were contributed to collective production.   66   As informants recall, only the household of the one “real” landlord was coerced to part with its residential property during the reform. I should nevertheless mention that, by the time the reform began, many wealthier families—especially those involved in the weaving business who had left the village for the city—had already transferred their residential properties to family members or had “donated” them to the revolutionary government. Households classified as poor and middle peasants who remained in the village kept their housing properties intact. For most of them, this represented an area between 400m 2  and 600m 2  but could reach up to nearly 1,000m 2 .  More research is needed to explain the DRV government‟s decision to implement a partial land reform, and in particular the decision to ignore residential properties during the land redistribution process. Secondary literature nevertheless allows me to put forward a few tentative lines of explanation. One of them has to do with the precolonial state tradition, entrenched in dynastic legal texts (such as the Gia Long code), to treat residences as the inalienable, private property of families (Whitmore 2009; Ngô Kim Chung 1987; Nguyen Duc Nghinh 1993). Illustrating this, pre-colonial registers referred to settlement land as “private housing and garden” (tư thổ trạch viên trì) and dynastic codes did not stipulate taxation of usufructs from these lands (by contrast to communal or private rice fields). Also in contrast to agricultural land, land designated as residential in official cadastres was considered to be under the full control of the families occupying it, and was therefore excluded from state or communal ownership forms.  One last distinct aspect of residential land relates to its fundamental role in the ritual life and beliefs of Vietnamese families. For several centuries, households and lineage branches have devoted a portion of their housing area to structures used in the worship of their ancestors (Jamieson 1993). In rural villages in particular, lineage branches are traditionally seeking to build elaborated family halls and altars devoted to the worship of their ancestors. These structures and the ritual practices that they support transform housing space into sacred space. A significant problem with the application of land reform to the entire rural territory (including settlement areas) would have been the inevitable disruption of the ritual function and symbolism of housing spaces.  We can surmise that the revolutionary state took into consideration these pre-existing property and social relations, and that the decision to limit residential land confiscation to landlord families 10   10 I did not meet the Hòa Mục “landlord” family that had its residential land and housing confiscated during the land reform. Yet, I met one such family during an interview I conducted in a village in the province of Bắc Ninh in July 2009. This 80-year old man told me that the suffering associated with the confiscation of the ancestors‟ altar, furniture, and house that his family had previously owned was still so great that he was, to this day, unable to talk about it. One native Hòa Mục villager, similarly observed that for a family, the worst possible outcome (bị đấu tố anh) of the land reform was the confiscation of its residential land and housing.  67   made the land reform process more socially acceptable for the vast majority of peasants. This particular treatment of residential land is important for understanding the patterns of regulatory informality that developed in the decades following the revolutionary transformation years. As will be discussed in the following chapter, local bureaucrats and the socialist law, either explicitly or implicitly, continued to treat residential land as “pseudo-private” property. Households thus retained most of the control over this resource. It is in this context that Hòa Mục‟s settlement land became the site of many of the private activities that slowly transformed the Soviet-type economy. In a spatial perspective, the development of informality was most certainly not an arbitrary phenomenon.  Socialist Economic and Institutional Re-formations The Land Reform‟s class struggle and redistribution of agricultural land was only an intermediate step in a larger socialist transformation scheme that aimed “to establish land ownership by the peasants, release the productive forces in the countryside, give a strong impetus to agricultural production and pave the way for industrial and commercial development” (Land Law 1952, [art. 1]). In the periphery of Hà Nội, the establishment of the institutional arrangements to realize these objectives was gradual. This process, which had started with the Land Reform, continued until about the mid-1960s. Once again, the policies implemented in Hòa Mục differed from those applied in both more remote rural areas and in the inner city. In a rather unique manner, the socio-economic organization of the village‟s population mixed two types of occupations and the two very distinct types of institutions governing them: the agricultural cooperative and the urban-industrial production unit.  The Early Post-War Years As discussed earlier, during the first years after peace was settled (1954), the communist leadership‟s primary objective was to restore the production of food and basic consumer goods as quickly as possible. Even later, during the period seen as the high-tide of socialist transformations that preceded the first five-year plan (1958-60), the state‟s need for legitimacy seemed to override its intention to build a “traditional” or “Soviet-type” model of socialism. On the contrary, socio- economic policies enacted during this period were characterized by a rather loose set of mandates and constraints that favoured multi-sectoral economic development, that is, the collaboration between the emerging socialist state sector and pre-existing private firms.  In the agricultural sector, this approach led to the creation of labour-exchange groups (tổ đổi công —also translated as “mutual-help groups”). Such labour-exchange groups operated in Trung Hòa between 1956 and 1960. Their functioning borrowed from pre-existing labour reciprocity practices  68   that were concentrated during the most intense period of rice transplantation and harvests. Compared to earlier practices though, labour-exchanges became mandatory, even when reciprocity cut across kinship groups, social networks, and personal affinities. As one villager remarks:  For example, when my fields were lush with rice, I could ask you and your family to come help me with the harvest. When you needed it, my family would later come help you do the same thing. Or when you needed to transplant [young rice plants], we would come [...] Also, some households didn‟t have buffalo or oxen to draw the plough, so they could ask for help from those having these things. They would have to do other things in exchange and vice-versa. [...] The policy was that labour exchange was compulsory (bắt buộc). If you just came and asked me for help, chances are that I may refuse. It‟s hard to ask others for help when you don‟t have good relationships with them, but when the group leader [of the labour-exchange group] came, you had no choice, you had to go.  Similar to a majority of peasants across the Red River Delta region (Sau 30 Nam 1990 quoted in Kerkvliet 1995a: 68), many Hòa Mục villagers remember the few years of labour-exchange groups as a golden moment in the socialist experiment. Conditions then were particularly favourable for poor peasants. Often for the first time in their lives, these people could retain the usufruct of a small area of paddy-field over which they had nearly full control. The labour-exchange group system was also appreciated because of its small scale and organization among close neighbours. It further allowed villagers to maintain mixed sources of income, resuming earlier combinations of agricultural activities with “side-occupations.” After peace was settled, many villagers started to commute again to the city as waged-labourers in the workshops of relatives and former village neighbours.  In parallel to this agricultural reorganization, the communist leadership supported the development of the household-based and small-scale enterprises operated by capitalists (tư bản). Starting in liberated zones in the mid-1940s, policies encouraged workshop-owners to voluntarily form cooperatives (hợp tác), the party‟s preferred form of economic and political participation. For a few years though, members of the “national bourgeoisie” (giai cấp tư sản dân tộc) who chose not to join cooperatives continued to be allowed (and were even encouraged) to participate in the reconstruction of the country‟s post-war economy.  During this period, the DRV government played a leading role in coordinating trade, arranging for credit, and assisting in the supply of materials and marketing goods, providing incentives and encouraging production in all sectors. Government statistics from this period reveal a general boom in the production of craft goods by small-scale producers. At the end of 1957, the craft sector thus supplied about 60 per cent of all industrial outputs and a similar proportion of the value of exports (government data quoted in DiGregorio 2001: 87-9). But this growth in light industrial production  69   and what appears as a mutually beneficial relationship between the state and the private sector was short-lived. As commodity production started to improve, contradictions between the loose set of mandates allowing for this sectoral growth and the principles deemed necessary to build a socialist economy became increasingly apparent. At the root of these contradictions was the important role played by the private capitalist sector and the selling of craft products by artisans outside of state channels.  Low Level Cooperatives and Production Brigades The solution to these ideological problems was to implement a series of more structured reforms that moved the country‟s socioeconomic system closer to a “Soviet-type” model of socialism. In Hòa Mục this shift happened between 1959 and 1962, a period marked by the establishment of the first agricultural cooperative (hợp tác xã nông nghiệp) and by the progressive nationalization of the textile industry within which many villagers were still involved as wageworkers.  Two agricultural cooperatives were established in the commune of Trung Hòa between 1959 and 1961. The first one called Tiền Phong (lit. Vanguard) grouped peasants from the villages of Trung Kính Thượng and Hòa Mục. The second one, Tân Hòa (lit. New Harmony), corresponded to the densely populated village of Trung Kinh Hạ. The creation of these institutions was accompanied by the full-fledged collectivization of agricultural production. 11  In line with the ideal of mutual benefit promoted by the party-state throughout the post-war period, villagers were “encouraged” to “voluntarily contribute” their land, livestock, and agricultural implements to the new cooperative. As Moise (1976b) notes:  The Chinese and Vietnamese Communists, although they used some coercion in collectivization, did not wish and could not afford to use massive force, as Stalin had in the Soviet Union. To the greatest extent possible, they wanted the agricultural producers‟ cooperatives to be voluntary associations of peasants who had been persuaded that they could enjoy a higher level of living as cooperative members than as individual small holders. (ibid: 14)  In the commune of Trung Hòa, the process of “voluntary” collectivization took about two years. As the former chairman of the Tân Hòa cooperative puts it: “[Thereafter] we worked together with  11 A small but significant policy exception was made with the issuance, in 1962, of a directive allowing cooperatives to allocate up to five percent of their agricultural territory to the households of co-op members for private exploitation (yet under broad guidelines issued by the cooperative regarding, for instance, irrigation). That year, each co-op member residing in the commune of Trung Hòa received 62m2 of this so- called “five-percent land” (đất năm phần trăm, sometimes also referred to as “family economy land,” or đất kinh tế gia đình). This area was readjusted down to 48m2 in 1972, due to population growth, and then back to 72m2 in 1986 in the wake of the decollectivization process.  70   public land and public buffalos, public water, public electricity, public harvests. From start to finish, all of it was public and collective.” In the process, local households lost the control they had previously enjoyed over the means of production (land, implements, and labour) and the freedom to enjoy its usufruct. After the creation of cooperatives, it indeed became mandatory to deliver all harvests (except for that from the five percent plots) to local institutions, which would then redistribute them locally according to a system of quotas and rules imposed from above. 12   Most rights over agricultural production were transferred to the cooperative and to its “production brigades” (đội sản xuất). These brigades acted as transmission lines between the cooperatives‟ central organs (the Management and Supervisory Committees) and local groups of co-op members (xã viên). As opposed to labour-exchange groups, production brigades were bound to implement plans and policies imposed by the cooperatives‟ Management and Supervisory Committees. The new sub-managerial units were invested with more resources, powers, and responsibilities than their predecessors. Each brigade consisted of a group of peasants charged with carrying out a portion of the collective production on a given land area. Brigades were put under the management of a small team that included a leader, a vice-leader, and a secretary elected from its membership. These people were ultimately responsible for meeting the output targets set by the cooperative for each crop, and transfering a proportion of each harvest to the central management of the cooperative (Fforde 1989: chapter 2; Chử Văn Lâm et al. 1992).  Reflecting its small population, Hòa Mục had only one production brigade while the two other villages that formed the commune of Trung Hòa each had four. When it was established, this brigade included between 600 and 700 male and female, individual members. This represented approximately two-thirds of Hòa Mục‟s active population. Members were selected based on their previous occupations. Adult villagers who had previously derived the largest part of their income from agriculture, including those who had been only marginally involved in the textile industry or in other side-occupations were required to become agricultural co-operators.  Integrating the cooperative had far-reaching implications for villagers and their households. In retrospect, a former co-op member describes the brigade‟s powers as extending over most aspects of work and life in the village: “They decided everything. They led and administrated production, husbandry, tilling, irrigation, fertilization, harvesting, taxation and our lives and politics; the whole thing.” In line with this statement, villagers mentioned that the Tiền Phong cooperative took away  12 According to Kerkvliet and Selden (1998), the war against the United State tamped down Northern Vietnamese villagers‟ natural resistance towards the shift from household-based to collectivized farming. These authors write that “because the mobilizations for collectivization and for the war were interwoven, villagers realized that outright opposition to collective farming could be interpreted as unpatriotic” (ibid: 48).  71   the capacity of their households to decide who among their members was to work in the fields and when. Former co-op members still talk about collectivized agriculture as the “time of working following the gong” (thời làm thuộc kẻng), by reference to the device used to summon them at the communal house (đình) in the morning and at noon for the distribution of daily tasks.  Nuancing the above quote, the literature indicates that cooperatives were not exactly “total institutions.” Their control over population‟s activity was far from perfect, due notably to the various managerial problems faced by these institutions. Local power relations and conflicting interests that occurred between local People‟s Committees, cooperatives and their brigades often thwarted the implementation of the central state‟s prescriptions. Production brigades, for instance, proved difficult to manage because of the leaders‟ tendency to serve their own interests or those of team members at the expense of the orders received from higher administrative echelons. The staffing of agricultural cooperatives and brigades with local people further opened the way for local patterns of reciprocity and kinship groups that influenced the implementation of policies at the local level (e.g., Fforde 1989: chapter 2; Chi Huyen Truong 2001).  In practice, the socialist territorialization strategy that tried to ensure full authority over rural populations by integrating the life, work, and political space of agricultural co-operators, limited the party-state‟s capacity to govern its constituencies. In many instances, these “cracks” in the managerial system of cooperatives produced a liminal space which co-op members turned to their advantage, notably to carry out activities on the margins of the planned economy (e.g., Kerkvliet 2005, these activities will be discussed in more detail in the following chapter).  The Nationalization of the Textile Industry The remaining third of Hòa Mục‟s adult population who did not join the Tiền Phong cooperative became school teachers, or bureaucrats had professional careers in the army, or continued to work in the textile industry. A majority of the non-agricultural population fell into this last group. These people had often become specialized wage-labourers during the late-colonial period. During the progressive nationalization of the weaving industry, they joined the urban/industrial workforce while continuing to reside in the village.  This nationalization process began in 1959. From that point on, the DRV government initiated a “reform of privately owned capitalist industries and trade” (cải tạo công thương nghiệp tư bản tư nhân). This included stricter policies, issued in 1962, requiring that household-based textile workshops form so-called “joint private-state companies” (công ty hợp doanh) (Thrift and Forbes 1986: 144-5). This reform entailed a partial nationalization of the manufacturing sector. As one  72   informant explains, “the state then became the co-owner of the factories. The former owners could still be the boss and get a higher salary than his employees and make a profit but he was not a private capitalist (tư bản tư nhân) anymore.” Two weaving and knitting joint-companies were thus established during the early 1960s that grouped most Hòa Mục entrepreneurs still active in garment production. This intermediary step only lasted a few years. In 1968, three years into the anti- American war, the DRV government formally merged the two joint-companies to create the “Mùa Đông-8th of March Weaving and Knitting” state company (công ty nhà nước). The latter was established in a new industrial area in the district of Thanh Xuân, less than a kilometre south of Trung Hòa.  The integration of Hòa Mục villagers into this urban-industrial structure was gradual and selective. Specialized wage-workers from the village first followed their employers through the formation of joint-companies. Then, between 1962 and 1968, many workers were compelled to return to the village because of economic difficulties. During this period, the natural successors of Hòa Mục‟ first generation of weaving and knitting entrepreneurs started to move into other occupations. The son of one such entrepreneur explains:  Little by little, workshop owners got older and died. Their children, the young ones like me, didn‟t want to take up the craft. For instance, after my grandfather died, no one in my family was really interested in that business. Most of us [children from workshop- owning families] were then getting a good education in the city; we attended Hà Nội universities. Some of us got masters‟ degrees, others took up liberal professions; some even went abroad to study. So we mostly became bureaucrats, teachers, and academics, or entered the army as officers. The weaving craft of the elderly was bound to die out. Little by little it fell into oblivion.  Villagers who were then involved in this industry as wage-workers say that, during this period, joint-enterprises also failed to keep up with the industrial modernization process. They could hardly compete with modern factories recently set up by the state. The resulting under-employment in joint-enterprises led to the reallocation of labour between the agricultural and state industrial sectors. A villager laid off during this period explains:  The government wanted the unemployed to work. So they laid off some people who had previously worked in weaving. The rule was that families with both parents and children working as state workers (làm công nhân) could stay. They were recommended for work with the Mùa Đông-8/3 factory. Those with farming parents were told to go back to agriculture. Because my parents had been farmers, they didn‟t let me be a worker anymore [...] My wife then also worked as a weaver and was asked to quit like me. So we came back home to work in the fields in 1968.   73   Many villagers described the progressive demise of the weaving industry that occurred between the 1940s and 1960s with sadness. A shared opinion among the people I interviewed is that this industry had an overall positive effect on the village, bringing wealth and prestige to some families while diversifying the livelihood of others. One woman in her 80s who weaved for wages in her great-aunt‟s workshop, thus remarked: “once the weaving craft was gone, the locals only tilled their fields and, from then on, it was the end of development for our village.”  Làm Nội, Làm Ngoại: Working Inside, Working Outside During the first decade following independence (1954), the communist government established institutions broadly aimed at facilitating the party-state‟s penetration and control over society and the economy. Borrowing from the Chinese experience, the Vietnamese leadership sought to tie the country‟s labour force to collective work units such as agricultural cooperatives, state-owned enterprises, and government agencies. These units were designed to be the only place where their members could cover all their basic needs: food, commodities, housing, schooling, medical care, etc. Work units were also expected to allow local management of professional, political, and productive activities and to give the party-state control over the social and private life of individuals.  The kind of collective work units established in the country and city differed markedly from each other. This is most evident with regard to their territorialization strategies and the manner in which they controlled their members‟ access to state-controlled resources. A defining characteristic of Hòa Mục, as a periurban place, was the establishment of both rural-agricultural and urban-industrial institutions. The co-presence of people in the village involved in these different socio-economic systems contributed to the maintenance of rural-urban linkages. It also informed the imaginative tactics that local households used to take advantage of both systems.  In rural areas, the life of agricultural co-operators was expected to develop mainly within the commune‟s administrative boundaries. There, the state progressively put in place local-level institutions whose regulatory powers embraced almost all aspects of life. These included the local agricultural cooperative, the People‟s Committee of the commune (Ủy ban Nhân dân xã), party cells (chi bộ), and mass organizations (Farmers‟, Women‟s, and Youth Unions) operating under Việt Nam‟s Fatherland Front (Mặt trận Tổ quốc).  The main responsibilities of these institutions included: managing co-operators‟ work and supplying them with the necessities of life (food, basic commodities, housing); providing the local population with schools and medical clinics; controlling population movements through the residence permit  74   system; 13  encouraging army conscription; and propagandizing socialist ideology, policies, and campaigns (phong trào). The mandate of these institutions further went into the private life of villagers, notably with respect to policies related to personal hygiene and dress codes, birth control, and the conduct of ritual ceremonies (e.g., marriage, funerals) (see Malarney 1997: chapter 1).  Villagers employed by the state outside of agriculture spent their life within a less integrated socio- spatial environment. 14  Most of these people commuted daily from their place of residence, in the village, to their place of work, in downtown Hà Nội or in newly created industrial areas on the city‟s outskirts. This distinction between place of residence and place of work meant that urban/industrial state workers also fell under the authority of two distinct sets of governing institutions. As residents of Trung Hòa, factory workers and civil servants were under the authority of the People‟s Committee of the commune for matters such as residence permits or access to housing. As opposed to co-op members though, it was their urban employers (schools, research institutes, administrative agencies, factories, etc.) that regulated their work schedules and remuneration. Industrial workers and other state employees were also generally members of the branches of the Fatherland Front‟s mass organizations established at their work place rather than those established in their commune of residence.  Bringing the periurban labour force under territorially unified collective units therefore proved more easily done in the case of agricultural co-operators than it did with regard to urban/industrial state workers. Distinct from peasants, the residential, work, and socio-political territories within which factory workers and civil servants lived and worked straddled the village and city. Reflecting this, villagers still refer to collective-agriculture occupations as “working inside [the village]” (làm nội) and urban-industrial occupations as “working outside” (làm ngoại). The latter expression serves to indicate that, while continuing to reside in the village, the population employed outside of agriculture, worked elsewhere (mostly in the city, but also in periurban industrial areas).  Another important difference between farm workers and urban/industrial state employees relates to the respective modalities of their remuneration. The income of the former was limited to unhusked paddy (thóc). A system of labour points (công điểm) determined each co-operator‟s entitlement at the pro rata of his or her contribution to the production process. The production brigade‟s secretary recorded the number of hours spent by each co-operator in the field, the nature of the work he or she  13 Similar to China, during the 50s and 60s, Việt Nam implemented a household registration system known as “hộ khẩu” to control population movements and to limit access to state services to the particular place of residence of each citizen. 14 Those villagers who pursued careers in the army are a special case. When they were not serving on the front, most of them were posted in army camps.  75   did, and other material contributions (for instance, in manure). The effort of each co-operator was then translated into labour points. After handing a fixed quota to the cooperative, production brigade leaders distributed the remaining paddy to their members in proportion to the labour points accumulated by each of them. In line with this, villagers still use the expression “eating by points” (ăn bằng công điểm) to refer to a livelihood based on collectivized agriculture.  The remuneration system in place in the non-agricultural state sector was different. People employed in urban and industrial workplaces received three forms of income: entitlements to rice rations, coupons (tem phiếu) to be exchanged for meat, sugar, fabric and paraffin,15 and a salary in cash (lương thực).16 In interviews, informants who used to work as co-operators often insisted on state employees‟ better living standards, and especially on their privileged access to “more delicious foods” such as meat through their rationing coupons. The main advantage of urban-industrial employees over agricultural co-operators was easier access to cash. This resulted both from their income package and from the common informal practice of selling ration coupons to resellers (con phê).  Reflecting on the privileged situation of state employees, a former co-op member who is now in her late-40s remarked that, “during the subsidy time, every farmer longed (khao khát) to become a worker.” She further indicated that after the transitional period of the 1950s and early 1960s, it proved quite difficult for a villager to shift from the cooperative to a state factory. Educational requirements and a lack of social relationships, she explains, prevented peasants from moving to urban occupations. Perhaps due to the limited possibilities of moving across sectors, but also because co-op members and urban/industrial state employees were entitled to different types of remuneration, villagers explained that the ideal situation was for a given household to combine “inside” and “outside” work. One woman who worked for the local agricultural cooperative during the subsidy era puts it this way:  If possible, the wife worked in agriculture and the husband as a state worker. In this way, a family could get most benefits. They were sure to have both money and rice [...] Of course, they would eat up all their rice and spend the salary before the end of the month, but they would still be better off than those families where both wife and husband had  15 In consideration of the fact that they lived in a periurban village, Hòa Mục workers and employees were only entitled to half the paraffin ration of urbanites. Periurban people were then expected to use stubble (gốc rạ) collected in the fields instead of paraffin. 16 This income varied according to categories of state employees along an alphabetic scale from A to E, with “A” corresponding to the highest income (that, for instance, of a high-level bureaucrat working in a central ministry) and “E” corresponding to unskilled factory workers. Most of the state employees living in Hòa Mục by the early 1960s fell in the lowest categories.  76   the same job. In some village families, both wife and husband did farming or were workers, but only in a few of them.  Villagers explained that this “ideal” household distribution of labour became an important consideration when parents considered marrying off a daughter or son. They thus tried to choose a partner whose job would complement that of their child. This preference reproduced the gendered patterns of occupations observed within households during the colonial period: the woman working in agriculture and the man in the urban-industrial sector. In doing so, the periurban households of Hòa Mục partially bridged the divide introduced by the state‟s economic management system between the rural-agricultural and urban-industrial sector, straddling it to their advantage.  Conclusion In this chapter, I explored the uneven socialist transformation process in the Northern Vietnamese region during the early postcolonial period (1940-1960). I provided evidence of manipulations, by the DRV government, of its own supposedly uniform socialist transformation policies to reinforce its legitimacy and to facilitate economic recovery after the anti-colonial war. The case of Hòa Mục showed that the DRV government adapted the implementation of the land reform, agricultural collectivization, and industrial nationalization policies on the near periphery of the capital. I argued that, in seeking to achieve their state building goals, the special policies and rules implemented on the near periphery of the capital constructed a graduated socialist geography across the Red River Delta. What Ong (2006) might call a “zone of exception” thus took shape at the edge of the city that allowed urban and rural institutions, rules, and processes to continue to meet, mix, and interact with each other.  The version of the revolutionary transformation process implemented on the near periphery of Hà Nội sought to re-establish basic economic and productive activities in the Red River Delta region, and the supply of basic goods to the inner city after war interruptions. The “special” policies devised for periurban Hà Nội by the postcolonial government had major implications for the development of this zone. Most importantly, they maintained market linkages and the flow of food and goods produced in the zones immediately adjacent to the city towards the urban core. For some years, these policies also encouraged small-scale producers, including those Hòa Mục families active in garment production, to resume and increase the private production of commodities. A decade or so after the end of the first Indochina war, many of the periurban-urban linkages that had characterized Hòa Mục during the colonial era had been re-established.  Between 1940 and 1960 the near periphery of Hà Nội also experienced a comparatively “softer” version of the Land Reform than other Northern Vietnamese regions. The case of Hòa Mục offers  77   supporting evidence to the thesis that the implementation of this reform on the near periphery of Hà Nội was more a matter of imposing the stamp of party legitimacy, than a campaign to eliminate the old rural elite, or to impose forced land redistribution (see Malarney 2002: chapter 2; Moise 1976b; Kleinen 1999: 95). In Hòa Mục, a late implementation and “special rules” spared many private entrepreneurs and other members of wealthy families. Under the harsher rules applied in more remote rural areas of the DRV, these people would likely have been socially stigmatized for generations, sent to jail, or even executed. Other “special policies” protected individuals who were involved in the textile industry as sub-contractors or wageworkers. Again, under the rules applied further out in the countryside, these people would have been punished for exploiting their paddy fields under tenant farming arrangements.  Perhaps more importantly, the Land Reform left almost intact the pattern of residential landholdings in the village. I have suggested that the policy decision to focus the reform on agricultural areas and to avoid redistributing residential land reflected precolonial land relations that long defined domestic spaces as private spaces. I have also hypothesized that ritual functions embedded in housing space limited the state‟s ability to reach into and redistribute the residential properties of villages like Hòa Mục. By overlooking residential areas and housing, the reform reinforced villagers‟ sense that residential landholdings, while officially nationalized in 1952, remained de facto under the control of individual households. This perception of quasi-private ownership (and its respectful acknowledgment by local bureaucrats) is important as it partly explains why housing areas became the privileged site of informal activities during the subsidy era.  The exceptional policies implemented on the edge of the city during the early post-colonial era had another important impact on Hòa Mục. They preserved its distinctively periurban, mixed socio- economic composition. In the inner city (nội thành), workers were essentially integrated into the urban/industrial, socio-economic sector. In the more remote countryside (nông thôn), the overwhelming majority of villagers lived under the management of the cooperative system. This was the case even in those villages that had historically had a long tradition of household-based commodity production. In these cases, the establishment of the local cooperative included a handicraft branch, thus allowing the cooperative to manage all production processes occurring on its territory.  On the other hand, Hòa Mục sheltered people who worked for different sets of institutions in the same place. The village population therefore lived under different territorial regimes and systems of remuneration. Local history and spatiality played an important role in making this unusual situation possible. The small-scale textile production enterprises set up and managed by natives of the village  78   offered a useful basis for the reconstruction of the DRV‟s post-war economy. The people of Hòa Mục who staffed these enterprises also constituted a useful labour force for modernizing the DRV economy. The most educated segment of the population, including many sons and daughters of weaving families, further formed a labour pool for the new socialist administrations, schools, universities, etc., and were potential replacements for Vietnamese civil servants who had worked for the French, and who were therefore regarded with suspicion by the party. Geographic location facilitated the integration of these villagers into these sectors. The people of Hòa Mục could easily join the socialist urban-industrial workforce. They did not need to be housed in collective housing units like workers coming from further afield. Indeed, despite very bad roads and the detour to reach bridges across the To Lich River, the trip from Hòa Mục to the inner city took less than an hour by bicycle, allowing villagers to commute to work on a daily basis.  It would be wrong to think that these developments occurred simply as the result of state-led, exceptional policies, and did not involve local populations as active agents of change. The people of Hòa Mục were instrumental in carrying periurban-urban linkages or patterns from the colonial era into the socialist era, and in adapting them to new conditions along the way. A telling instance of this relates to the new labour division patterns that took shape at the household level as the socialist regime established new occupational structures and institutions. In trying to optimize their life chances, Hòa Mục villagers then invented ways to take advantage of the resources these reforms had to offer. One of these strategies was to form mixed, urban-industrial and rural-agricultural couples. Villagers then did not seek to thwart the revolutionary transformation process, but rather to take advantage of an arrangement that benefited both their families, and the socialist revolutionary process.  This episode in the history of Hòa Mục sheds some light on the geographically uneven nature of the socialist transformation process across the territory of the DRV. It helps us to understand the process through which a regime of state regulatory informality took hold in and around the village that was to shape its development during the harsh years to come.    79   CHAPTER 5 Eating by Points and Coupons is not enough (1965-1980)        Bà Thư got married to a native of Hòa Mục in 1956. Following the Vietnamese custom of patrilocal residence, the 24 year old girl should have moved in with the family of her husband. But, a year before the wedding, her future in-laws were classified as “poor peasants” by the Land Reform team. This entitled them to receive a 38m 2  building that previously had belonged to a “landlord” family. The in-laws had decided to give this building to the newlywed couple, thus saving space in the familial house for their other children. Bà Thư, her husband, and their newborn daughter lived in the “landlord house” for just two years. They moved back to the in-law‟s housing compound, Bà Thư explained, because the house was too small for them and, even more problematic, it did not come with a garden. This relocation, she remarked, was far from ideal: it meant sharing limited space with her brothers-in-law and their own families, a total of 12 people in a 50m 2  house. But she still preferred this arrangement, because it meant access to enough space to raise a few chickens and grow vegetables on which her growing family had become increasingly dependent.  Bà Thư stayed with her in-laws until the mid-1970s by which time she could hardly fit her husband and seven children 1  into an increasingly overcrowded house. Like many fellow villagers, the couple then turned to the People‟s Committee of Trung Hòa, asking for a piece of so-called “population de- densification land” (đất giãn dân) on the edge of the village. The authorities took the cramped quarters of their house into consideration and decided to allocate (cấp) to them a bare plot of 240 m2 in a swampy zone south of the village settlement. It took the couple two years to build a so-called “level-4” house (nhà cấp 4, see definition later in this chapter) on this piece of land, using bricks and roof tiles they handcrafted with the help of family members. They moved into the new house in 1977. Next to the house, Bà Thư and her husband managed to set up a small vegetable garden and a shed allowing them to raise a brood of pigs and dozens of chickens yearly. This piece of land on the edge of the village rapidly became the family‟s most important asset, not only because it gave them  1 This is a typical fertility rate for peasant households during this period. For further discussion, see Quertamp 2003.  80   access to a shelter, but also, and perhaps more importantly, because it allowed them to improve the meagre subsistence derived from official jobs within the state socio-economic system.  The specific needs of Bà Thư‟s household, and the resources she and her husband could access from the late-1950s onward, of course, shaped the above story. Bà Thư‟s experience nevertheless has important similarities to those of other villagers who shared with me their stories about how their families got through the harsh years of the subsidy era (thời bao cấp), and the difficult transitional years of đổi mới. Two very basic motives characterize these various accounts. The first is the need to secure access to space in which to perform secondary-occupations (nghề phụ) essential to a household‟s subsistence. The second one is the need to provide one‟s household (and that of newly married children) with a decent shelter.  From the 1960s onward, local people sought various ways to fulfill these two basic needs. In doing so, they carried on with earlier livelihood and land practices and adopted new ones that profoundly transformed the village‟s socioeconomic, physical, and institutional landscape. In trying to improve their lot, villagers influenced the urbanization trajectory of the village in three important ways. First, they contributed to the diversification of the local economy beyond the planned state sector. Second, they densified, expanded, and transformed the village‟s built fabric. Third, they perpetuated and developed land market practices inherited from the colonial era. Taken together, these transformations are early manifestations of what we might call an informal “in situ village urbanization” process. The set of phenomena involved in this form of urbanization was to become a defining aspect of post-đổi mới transformations in Red River Delta settlements.  The notion of in situ urbanization, as understood here, differs from the conventional city-based urbanization process dominated by rural-urban migration. In situ urbanization instead refers to the progressive transformation of rural settlements and their populations into urban or quasi-urban places without geographical relocation of their residents and regardless of administrative change. Central to this process are local people‟s strategic decisions to integrate non-agricultural practices as part of their livelihoods, to build urban-style buildings, to participate in urban land markets, or selectively adopt what might be termed “urban” ways of life.2  Studies conducted during the early years of the reform have highlighted the emergence of urban- oriented economic activities, built forms, and ways of life within the confines of many Red River Delta villages (e.g., Hoang Xuan Thanh et al. 2005; Le Van Sinh 2001; Pandolfi 2001a; Rossi and  2 For further studies of in situ village urbanization in the Malaysian context see Brookfield et al. (1991) and for a discussion with regard to China, see Zhu (2002). A proposal for a planned form of in situ urbanization can also be found in Friedmann and Douglass (1975).  81   Pham Van Cu 2002; Chabert 2004; Quertamp 2003). The general impression one gathers from this scholarship is of an air of ad hoc practices. Studies describe an open-ended, bottom-up, and self- initiated process wherein periurban populations and their local governments seek to take advantage of changing socio-economic conditions. The literature presents many such experiments as occurring at the expense of state plans, norms, and regulations. For this reason the urbanization process on the outskirts of Hà Nội, and more generally in the Red River Delta region, have been variously qualified as “informal” (Rossi and Pham Van Cu 2002), “invisible” or “hidden” (DiGregorio et al. 2003), “uncontrolled” (Le Van Sinh 2001) or “spontaneous” (Leaf 2000).  This chapter explores the motives behind the livelihood and land strategies responsible for early in situ urbanization in Hòa Mục and the mechanisms that allowed for their reproduction from the mid- 1960s until the mid-1980s. The focus on this period serves to uncover the pre-đổi mới origins of urbanization practices that became the focus of studies in the early years of the reforms. It also challenges the widely held view that the urban transition is first and foremost an outcome of the shift from plan to market in Việt Nam. A detailed study of local practices in the decades preceding the reforms further interrogates the seemingly high degree of autonomy and flexibility enjoyed by local villagers such as Bà Thư with regard to their access and use of residential village land. It allows a re-examination of the pivotal relationship between informality and the seeming absence of the state‟s regulatory controls and sheds light on the state‟s political role in the reproduction of informality.  The chapter is structured as follows. The first section briefly outlines the land management system adopted by the central leadership after independence. This discussion focuses on the significant differences in the legislative treatment of the agricultural and residential land categories. It also serves to question the nature of informal practices in relation to the institutional environment into which the residential land resource developed during the subsidy era. The next section describes how Hòa Mục villagers and bureaucrats at the commune level operated on the margin of the official legislative framework and planned economy to improve their living standards, especially in terms of basic subsistence, access to productive economic space, and housing. In conclusion, I explore the coalition of interests served by these informal practices. I suggest that what appears at first glance as the state‟s inability to effectively administer and regulate residential land uses in periurban villages, is more fruitfully understood as a deliberate strategy by a central state on the verge of a major legitimacy crisis. Finally, I discuss the implications of the state‟s selective tolerance of informal practices in Hòa Mục on the formation of a local moral-territorial order. I focus, in particular, on how this reinforced the perception within households that they held quasi-private ownership rights over their residential land.  82    Legislative Unevenness As discussed in Chapter 4, from the mid-1940s to the beginning of the 1960s, the Việt Nam Workers' party set about supplanting imperial rules and colonial laws by establishing a new political and legal system across the newly independent territory of the Democratic Republic of Việt Nam (DRV). The land regime 3  that emerged during the first years of the DRV mirrored the particular route taken by the central leadership. As a complement of the land and institutional reforms discussed in the previous chapter, the socialist land regime sought to turn Vietnamese society and territory into a modern nation-state and to develop a strong economy independent of Western powers.  The Socialist Land Regime and Management System Post-independence legislation rejected the colonial notion of inalienable private rights in favour of the Marxist-Leninist tenet according to which land, as a special means of production, should be publicly owned, exploited by cooperatives, and subjected to central economic plans (Gillespie 2005: 47-83). In line with these ideas, the 1953 Agrarian Law nationalized all lands and established the principles following which private occupiers of the land borrow their rights of use from the state. The promulgation of the Second Constitution of the DRV (1959), the implementation of the Law on Land Reform (1953), and the statute establishing agricultural cooperatives across Northern Việt Nam (Circular 449/1959/TTG) entrenched the primacy of state ownership (sở hữu nhà nước— defined as inalienable lands owned by the entire people) and of collective ownership (defined as lands collectively owned by members of cooperatives).  Legal texts acknowledged the existence of individual (or private) ownership (sở hữu cá thể) up to the promulgation of the third constitution of the DRV in 1980. Until then, the law theoretically authorized occupants “to transmit land to their heirs, to mortgage it, to sell it, to give it away, etc.” and to extract goods and revenues from land through means other than labour (Law on Land Reform 1953, [art. 31]). In practice though, the ability to exercise such rights depended upon the official categorization of land. In this system, ministries charged with territorial management (i.e., the Ministries of Natural Resources, Construction, and Planning and Investment) assign a specific use to each plot of land across the territory they govern (e.g., residential, agricultural, defence, forest, etc). In theory, the rights to use, manage, benefit, control, and dispose of a given plot of land is determined by the formal category to which it belongs. This management system based on land categories is the cornerstone of what we might call a personalistic land right system. Central to this  3 Borrowing from Kerkvliet (1997), I define “land regime” as the official answers to questions such as who may use land, for what purpose and in which ways, and with what rights and obligations.  83   system is indeed a complex matrix that defines, to this day, what types of users can access which types of land, and through which allocation mechanism. This matrix also determines how land can be exploited, by whom, under what specific conditions, and for how long.  Agricultural versus Residential Land The system of land categories in use during the subsidy era was fairly simple, with few types of use and users and few allocation mechanisms. In rural areas, the most important distinction introduced by this land management system was that between agricultural land (đất nông nghiệp, also called “field land” - đất ruộng) and residential land (đất thổ cu also called “housing land” - đất nhà ở). This was a crucial distinction. To simplify a bit, it can be said that lands designated as “agricultural” in communal land-use maps were officially out of local households‟ reach, understood here as autonomous economic units. Local cooperatives acted as gatekeepers that controlled access to these lands and regulated their exploitation and usufruct through their cooperative brigades and labour points system. Land designated as “residential” in the local cadastres stood outside the purview of cooperatives.  We have seen, in the previous chapter, how residential landholdings were, for the most part ignored by the land reform and how this reinforced household‟s perception of quasi-private ownership. This implicit understanding (or expectation) was maintained during most of the subsidy era. During this period, the area sheltering the rural population remained in a somewhat indeterminate institutional zone. Control over these plots and the housing built on them was theoretically under the responsibility of communal People‟s Committees. The central state, however, provided little guidance to these local agencies on how to manage this resource. For instance, up to the passing of the 1991 Ordinance on Residential Housing, it was unclear whether households owned the houses they inhabited or not, whether they had the right to occupy these buildings in perpetuity or for a defined period of time, and whether the right to live and use residential land and housing came with responsibilities towards the state (cadastral registration, tax, etc.) (Gillespie 1998).  It was similarly unclear how residential land and housing ought to be used (or not) by their occupiers. Early party edicts required that housing‟s function be limited to accommodation, with strict prohibition of for profit usages. But, as discussed in Chapter 4, after the retreat of the French colonial power (and also during the war against the Americans), temporary policies were issued that encouraged the practice of home-based production activities as long as these complemented production of food and goods by agricultural cooperatives and state enterprises. Later, the party ideologue Gia Ben argued that the exploitation of residential space by peasants to meet their household‟s basic subsistence needs, while being a backward practice, was not in direct  84   contradiction to socialism (see DiGregorio 2001: chapter 2). Other contradictions plagued residential land regulations. Thus, while the central leadership officially banned all horizontal transactions in land, it deemed transfers between family members an acceptable practice (Gillespie 1998, 2009; Fforde and de Vylder 1996).  Institutional Ambiguities, Informality, and State Governing Practices The degree of intentionality that led to the formation of this ambiguous legislative framework is difficult to establish. The vagueness, contradictions, and general confusion of the legislation on residential land rights might, as Marie Mellac et al (2009) have argued, simply have resulted from legislators‟ oversights during the tumultuous decades following independence. But, as livelihood and land strategies developed on the residential land of periurban villages that contradicted what appears as fundamental tenets of the socialist political economy, legislators made little to no efforts to trace a clearer line between lawful and unlawful practices. This indeterminate regulatory environment did not explicitly encourage practices diverging from basic tenets of socialism upheld by the party-state, nor did it give local administrators clear guidelines on what to do about villagers‟ residential activities.  This is important with regard to how we are to understand the informal practices observed in and around Hòa Mục‟s houses during the subsidy era. As discussed in the introductory chapter, the literature generally specifies informality in relation to a formal institutional framework in which the state theoretically intervenes to regulate processes and outcomes according to a set of enforceable legal rules. This distinction is difficult to apply in a case, like that described in the previous section, where the institutional framework is underdefined. The problem is made even more complicated by the diffused responsibility to enforce this underdefined legislation, a responsibility shared between members of communal, district and provincial People‟s Committees, the bureaucrats of local cooperatives, and provincial and districal ministerial branches. An additional issue concerns the lax enforcement of this vague legislation by state agents working for these various institutions.  Migdal‟s “state-in-society” approach and recent research on the nature of informality discussed in Chapter 1 (Tabak and Crichlow 2000; Portes et al. 1989; Leaf 2005) provide avenues to think about these questions. This sholarship highlights the many situations where the state (or specific parts of it) has an interest in facilitating the production or reproduction of informality. When we look at the development of informal practices in Hòa Mục from the vantage point of their “social embeddeness,” we begin to see just how politically delicate the task was for the central leadership to strike a balance between the strict enforcement of a “Soviet-type” socialism and the dire needs of local populations for food and housing. This approach also helps to identify the kind of ideological  85   and regulatory compromises that the party-state was ready to make to preserve its legitimacy as well as some of the limits of informality as a governing practice.  Residential Land as Space of Self-Supply and Self-Help A dysfunctional shortage economy was well-established in the DRV as early as the 1960s. Considerable economic inefficiencies of the state procurement system combined with low participation incentives in collective agriculture. While domestic supply of industrial goods and food staples stagnated, population grew at a rapid rate averaging 2.8 percent annually over the period 1960-1975. This resulted in a progressive drop in the rations allocated to both rural and urban populations in Northern Việt Nam to below subsistence level (see Fforde 1989: chapter 2; Dang Phong 2004; Beresford 1988: chapter 9).  For most of the northern Vietnamese population, this translated into daily struggles to feed one‟s family. These struggles have marked Hòa Mục villagers‟ memories and they continue, to this day, to associate it with life during the subsidy era. In interviews, the expression most often used by informants to describe this period was “cuộc sống vất vả,” roughly translated as a “strenuous or miserable life.” In hindsight, many villagers attributed this “miserable life” to the planned economy‟s inefficiencies and to shortcomings of the food distribution system. Giang, a 40 years-old woman who used to farm for the Tiền Phong cooperative, thus remarked:  What was most difficult was the value of the labour points (cộng điểm). From one harvest to the next, the value always changed. For some crops, 10 points [the equivalent of an average peasant‟s workday] meant as little as 200 grams (lạng) [of unhusked paddy]. For other crops, it could go up to one kilo. But you never really knew what to expect [...] During the bad crops, even if both wife and husband and some of the kids worked in the fields; even if you added all the points together, it was not enough food for the family to eat.  The situation of civil servants and factory workers was only slightly better than that of farmers. The ration coupons (thêm phiếu) they received from their employers entitled them only to very small quantities of food. The low-level workers and bureaucrats living in Hòa Mục generally received between 13 and 15 kilograms of rice monthly. Even added to the sugar and meat obtained in exchange for coupons, the staples received through official state channels were insufficient to feed the average household. Elder Mạnh, a retiree high-school teacher, indicated: “Invariably, we ran out of [rationed] rice on the 20 th  of the month. After that, we had to look for food outside of the state‟s distribution system.”    86   Side-Occupations as Self-Supply In this context, and in continuity with the practices of their parents and grandparents during the colonial period, villagers relied on side-occupations (nghề phụ) to get by. Hòa Mục residents thus engaged in various activities outside of their official state jobs. These activities became collectively referred to as “self-supply” (tự cung) or “self-sufficiency” (tự cấp). This terminology is important. In interviews, villagers indeed opposed the idea of a “self-sufficient” livelihood to that of a state- subsidized (bao cấp) one. This language not only points to a failure of the state system but also emphasizes villagers‟ ability to redefine the contours of the local economy on their own terms.  An important characteristic of self-supply activities is spatial. These activities concentrated on the residential plots of villagers, and rarely, if at all, did they overspill onto agricultural lands. The areas of residential land available to each household after the Land Reform thus played an important role. This resource endownment determined the economic fate and living standards of many households during and after the subsidy era. Entering the subsidy era with a large residential land area indeed meant access to an asset which could be exploited by household members, and whose usufruct often made the difference between undernourishment and relative comfort.  In the 1960s, the most common form of self-supply was small-scale subsistence farming. Households planted fruit trees, tended vegetable gardens, and raised animals (mostly domestic fowl) on the land surrounding their houses. Elder Minh, a woman in her late 70s who used to work at the Mùa đông-8/3 textile factory thus remarked: “During the subsidy era, all the families planted more vegetables and raised more chickens to lift up (tăng gia) their living [standards]. If you didn‟t do that extra-work (việc làm thêm) where else would you find something to eat?” The role played by self-supply activities in household economies tended to increase over time. Over the years, villagers devoted more and more time and resources to these activities. This generally occurred at the expense of their official work as cooperative farmers, civil servants, or state-factory workers.  At the micro-level, villagers thus made up for the progressive decreases of state allocations in the form of paddy, ration coupons, and cash. At the macro-level though, the cumulative effect of “outside economy” (kinh tế ngoại) activites worsened shortages of the state procurement system it sought to palliate. Fforde (1989) writes:  Statistical estimation of the parameters of the “outside” economy (kinh tế ngoại) and its pervasive effects is extremely difficult. In the mid-1970s the income flows generated by such trade in staples seem to have been of the same order of magnitude as state- regulated trade [...] The sheer size of this phenomenon suggests that almost everybody participated in it to some extent: co-operators active on the private plots or state employees “leaking” commodities onto the free market (small-scale embezzlement of  87   state-distributed products, sale of second-hand goods, moonlighting on second jobs, petty-commodity production, and so forth). (ibid: 14)  By the 1970s, those households that had the “the right conditions” (điều kiện)—that is access to land and a large enough workforce, started to generate surpluses and engaged in petty trade. These families sold their home-based production at nearby markets or to wholesalers who criss-crossed periurban areas. These traders bought individual households‟ produce and animals and resold them in informal urban markets. Elder Mai, an ex-factory worker whose family sold large quantities of sapodilla plums and bananas throughout the 1970s and 1980s, described such commercial activities as follows:  Some people took their fruits, vegetables, and chickens directly to markets (thị trường), mostly to the Công Mộc market [a few hundred metres South of the village]. Others sold their production to wholesalers (các chủ mổ lợn) [...] Most of them [wholesalers] were natives [from the village]. Many did that year round. Besides working in the fields, they bought produce from the people here and retailed them to get an extra income. [...] Others, from Xuân Đỉnh [a rural commune adjacent to Trung Hòa] collected unripe fruits and then waited for them to ripen to sell them [...] In those days, there were a lot of people in suburban areas who produced food to be sold in the city.  When engaging in commercial activities, the main objective pursued by villagers, was to get access to cash or gold which could then either be saved for special events (weddings, funerals, New Year‟s celebrations, and other rituals) or, more commonly, used to supplement state rations. Cash typically served to purchase basic staples on the black market (chợ đen) and other goods and services that villagers could not produce on their own or obtain from the state.  The story of elder Mạnh and of his wife Vịnh illustrates this very well. Before retiring, elder Mạnh was a highschool teacher while his wife was an agricultural co-operator. After they got married, the couple received a piece of agricultural land carved out of the residential compound of elder Mạnh‟s parents. In the time available outside of their official state occupations, the couple tended a large vegetable garden, and took care of animals raised in a makeshift henhouse and pigpen they built on this land. By the mid-1970s, they raised a brood of pigs every four months as well as dozens of chickens, ducks, and geese throughout the year. Elder Mạnh described in great detail how he used to take dozens of fowl to the market on his bicycle during the last two days of the year, just before the Têt (Việt Nam‟s Lunar New Year), when the demand for meat (then a luxury food) was at its highest level in the city.  The young couple first used the money derived from their husbandry to feed their children. They later used savings from this activity to build and furnish a new house. Self-supply activities also  88   fulfilled non-material needs. Elder Mạnh and Vịnh, thus used some of the extra income from their self-supply production to send their sons to college in the city. These activities also allowed them and other families across the village to properly perform rituals and acts of mutual help. The cult of ancestors, longevity celebrations, weddings, and funerals that have traditionally bonded rural communities together indeed all call for the offering and sharing of food with family and neighbours.  The De-densification Land Policy A somewhat similar story can be told about housing. Despite stated intentions to satisfy the people‟s demand for housing, little was done by the state to shelter the population living in areas designated as administratively rural. From the beginning of the 1960s until the end of the 1990s, the provision of public housing—through the People‟s Committee and the agricultural cooperative—took three successive forms in the commune of Trung Hòa. None of these state-led initiatives responded to the local housing demand, a situation which contributed to the reproduction of pre-existing self-help housing practices and to the development of new ones.  As discussed in the previous chapter, a first set of policies was implemented during the Land Reform period. It consisted of confiscating the residential properties of landlords and then allocating living areas subdivided from these properties to families classified as “poor peasants.” Bà Thư‟s in-laws—whom we met at the beginning of this chapter—and a handful of other households received housing support in this way. The subdivision of confiscated properties did not yet suffice to meet the local demand for housing. By the early 1960s, the agricultural cooperative therefore started to distribute construction materials to the households of local co-operators. Yet, the quantities of bricks, cement, and tiles the Tiền Phong cooperative received from the state were very limited and barely covered its own needs. As a result, the construction materials distributed to co- operators were generally insufficient to build even the most basic buildings.  In 1968, a new policy was implemented that did not rely on the state‟s resources but rather on a resource available locally: developable agricultural land. This initiative is referred to as the “policy for the de-densification of the population” (chính sách giãn dân). Through this policy, planning authorities allowed local governments to convert agricultural land areas at the edge of existing settlements to residential use, to divide these areas into individual plots, and to allocate bare land to local households. This gave local authorities the means to respond to the pressing need for more residential space without relying on the faltering governmental procurement system. The “de- densification land” (đất giãn dân) distribution policies was implemented in successive waves up until 1999.  89   In contrast to the rather loose regulatory approach adopted with regard to activities practiced on residential land, there was very close surveillance by supra-local authorities‟ (district and provincial administrations) of how the de-densification land policy was locally carried out. An official from the Trung Hòa‟s Cadastral Office (cán bộ địa chính) described what this close supervision meant for him in this way:  We had to prepare a basic map indicating which area we intended to turn into residential land. Then we went to the upper-level for approval. A senior official would then come down to the village to check and confirm our choice. The architectural department had a map and they recorded the change on it. It was a major legal violation (sai luật) for an official at the commune to make this decision. 4  You have to understand that agricultural land conversion was an important matter for the authorities. As such, we got the approval from the city, not from the district. And only when we got this approval from the higher- level bureaucrats did we proceed with the distribution of the land to households.  Several factors might have contributed to the strong presence of supra-local authorities in this process. One of these concerned food security. The designation of new residential expansion zones occurred at the expense of cultivated areas for paddy. This decrease of the space under state- regulated collective production also translated into an extension of the space available for informal, residence-based production, otherwise contained within the imaginary boundaries delineated by the official system of land categories.  In line with this, I often asked villagers what prevented them from extending their private gardens or husbandry farms outside of their residential plot, and onto the agricultural residential land surrounding the village. Invariably, I was told that land officially designated as agricultural was under the strict managerial control of the cooperative. Villagers say that local authorities would (and actually did) stop all attempts to privately encroach on land under the cooperative‟s control.5  A Persistent Private Land Market Native villagers say that, at one point or another, almost every household in the village received “de-densification land.” Some families received up to three different pieces of land via this policy. If this greatly released the pressure for residential land, it did not meet all the villagers‟ needs.  Access to more residential land was especially problematic for groups that were disadvantaged by the de-densification policy‟s distribution rules. Combining socialist ideological principles with local  4 This does not mean that local-level officials did not attempt to manipulate land-use maps in ways that contradicted planning policies. See, for instance Chabert (2004). 5 Contrary to the situation observed elsewhere during this period, villagers did not extend their five-percent land beyond the area the cooperative entitled them to farm privately.  90   customs, these rules favoured agricultural co-operators with numerous children living on small residential properties or within multi-generational housing compounds. 6  Officially, factory workers and civil servants were not entitled to receive land from the People‟s Committee. Neither were the few non-native households who had settled in the village during the subsidy era. Another group seeking to access land outside of the local state allocation system included people like Bà Thư, who wanted to expand side-occupations beyond the areas they were entitled to access through the de- densification policy.  When they had the financial means to do so, these households turned to the private market. Despite official directives banning it, the private land market did not entirely disappear in and around Hà Nội (Gillespie 2009; Koh 2006: chapter 5). The cadastral official whom we have met earlier in this chapter told us: “The selling of land is a tradition that carried on from feudal times (thời phong kiến) and French occupation time (thời pháp thuộc). After the revolution, people continued to buy and sell land as usual (vẫn bình thường), even if it was forbidden (cấm).”  Such transactions did not only originate from pre-colonial and colonial times; they also proceeded according to one of many variations of the systems developed during these periods. 7  Villagers insisted on the simplicity of these procedures compared to the burdensome paperwork they are required to go through today. Back then, the buyer and seller produced a hand-written contract (giấy tờ biên tay) stipulating the nature of the transaction.8 If the seller was in possession of a notarized cadastral title (sơ dịa bạ) obtained under the French colonial regime, it was updated and passed on to the new owner. Otherwise, the seller would ask the land management official from the communal People‟s Committee or neighbours to testify to the sellers‟ ownership. The transaction was completed in the presence of a witness (chứng kiến); an official from the local People‟s Committee would frequently, but not always, be present as a witness. As a man who bought land in the village in the early 1980s remarks—quite paradoxically considering the official ban on horizontal land transactions: “Not all families went to the local authorities to get them to approve the transaction, but our family did so because we liked to do things the official way.”   6 After the war, special rules also favoured families of war veterans receiving social assistance (gia đình chính sách —lit. policy families). 7 In continuity with older practices, these procedures did not apply to the subdivision of a residential plot to settle children or its transfer to heirs. 8 The few such contracts that villagers showed me typically read as follows: I [name of the seller] have a house [description of the location and size] and land [of X m2] which I sell to [name of the buyer] for so many taels of gold, a radio, a bicycle, etc.  91   Villagers insist on the fact that few of them actually engaged in land sales during the subsidy era, especially when compared to the ebullient market that has developed in recent years. If several households were ready to part with some of the residential land they had inherited from their parents, or with de-densification land, few buyers could gather the very small amounts of gold or goods of equivalent value (e.g., bicycle, radio) asked by the sellers. As one elderly woman remarks: “We didn‟t have enough food to eat. We didn‟t have enough clothes to wear. How could we have money to buy land?” Saving enough money to buy a small piece of land therefore required years of hard work. Elder Hòa, a 74 years-old former state worker, thus recalls the long hours she spent in her garden after coming back from the factory; trying to increase pig husbandry production and to save enough to buy land and build a house for her family:  We saved money. We worked overtime and raised pigs to get more money to buy the land. My husband and our six children were used to a hard life, and to saving money: we only ate vegetables and rice gruel (cháo), sometimes only two meals a day, sometimes even just one. We kept the money to raise the pigs. Then, we sold them and saved the money. But besides what we could save, we also had to borrow more from my siblings. I borrowed a little sum from each one to avoid paying too much interest. After buying the land, we still had to save during many years to pay my siblings back.  Level-4 Houses and Brickmaking Having accessed residential land through the local state allocation policy, the private market, or inheritance, villagers still needed to build a house. The building type most of them then aspired to was called a “level-4 house” (nhà cấp 4).9 This expression refers to a very simple, one-storey building consisting of a single room defined by concrete and brick walls and covered by a flat concrete or tiled-roof (see figure 7). In hindsight, the level-4 house appears very austere. At the time, it embodied significant improvements compared to the housing conditions offered by the thatched houses (nhà tranh mái lá) that had sheltered most households since precolonial times. Seen from the perspective of this earlier housing type, the level-4 house conveyed ideas of comfort and modernity.  However, the construction of a level-4 house posed a significant problem. It required access to construction materials scarcely available (if at all) to villagers: concrete to build the wall structure and roof, metallic rods to reinforce them, bricks to fill the wall structure, and tiles for the roof. In the 1960s, these goods were only produced by state-owned factories. A small number of villagers working in the army or having connections in the state sector managed to get access to some of  9 The use of the expression “level-4” refers to a four-tiered housing classification system developed by the Ministry of Construction during the subsidy era.  92   these industrial construction materials. Their prohibitive cost on the black market kept them out of reach of most local households.  Figure 7 A thatched house (left) and a “level-4 house” (right)  Source: Reproduced with permission from Đoàn Thể Trung (left), author 2004 (right)  The solution that villagers found to build their level-4 houses took form somewhere between 1975 and 1977. It consisted in producing handmade bricks on their residential land. Villagers produced these bricks using what they call “improved kilns” (lò cải tiến) or “Korean-style kilns” (lò kiểu của Triều Tiên—this latter appellation refers to the kiln‟s supposed origins). The technique to make these kilns was easy to learn and the material used to make the bricks (soil) was widely available. Most importantly, the “improved kilns” were small enough (three by two metres on the ground and three metres high) to be built on residential plots, something that was unimaginable with the much bigger traditional kilns.  The quality of the brick so produced was nevertheless quite low, especially compared to the industrial bricks produced in state factories. One villager evocatively remarked that they were as friable as a soybean cake (bìa đậu). Yet, they provided the means to build a modified version of the level-4 house villagers aspired to:  We used lime mixed with sand to replace the cement and added salt to make the mixture more solid. If we tried our best, we could build the foundations with the sand-salt mixture and use a mix of lime and sand for the parts above. We did not use cement like nowadays. Almost no one had cement back then. For the roof, we made tiles by ourselves. [...] My relatives and family gathered together, we got the soil to make the bricks and tiles and we baked them. Then we hired local builders, people from the village, to build the house. That is how we built this house and it‟s the way our neighbours also built theirs.   93   As with other self-supply activities discussed above, families first produced bricks to fulfill their own needs. This practice rapidly developed into a commercial activity given the high demand for construction materials:  [F]irst, we wanted to build a house to establish ourselves and settle our kids in their own places (để tách con ra). Then, we sold our leftover bricks to other families from the village. But then the people from the inner city (nội thành) started to come to buy our bricks too. It all happened in a few years. Suddenly, everyone seemed to be looking for bricks. [...] That‟s how this whole brickmaking business came about. It became very profitable because the demand for housing construction in the city was really big.  During this period, demobilized soldiers returning home after having served on the Southern and Chinese fronts contributed to this rising demand. 10  Residential land acquisition for the development of brickmaking became a common tendency among ambitious brickmaking families. This contributed to the progressive renaissance of the moribund residential land market.  Villagers estimate that by the late 1970s and early 1980s approximately seven out of ten Hòa Mục households made bricks. Most of these people were agricultural co-operators. As one informant remarks: “at that time, working in the fields was a miserable job so almost all the farmers concentrated on making bricks.” Within this group, about a dozen households produced between two and three tons of bricks for trade yearly. Along with the other periurban villages that also produced bricks during this period, this must have represented a significant proportion of the construction materials available in the region of Hà Nội. It certainly also contributed to ease the housing shortage in the city and country.  Most villagers stopped producing handmade bricks around 1985. The end of this flourishing cottage industry was not their decision, but rather resulted from the implementation of a directive issued by supra-local authorities. Villagers‟ interpretation is that the municipal authorities then responded to two emerging problems. First were the environmental nuisances caused by brickmaking. The urban built fabric of Hà Nội had progressed slowly during the 1970s and 1980s and was now reaching the eastern banks of the Tô Lịch River (figure 8). Brickmaking generated smoke and this affected the city‟s air quality. The second problem related to villagers‟ breach of the implicit principle according to which self-supply was limited to residential space. Let me quote a former brickmaker on this issue:   10 The end of the American War in 1975 and the Việt Nam-China conflict in 1978 made part of the army redundant and prompted demobilisation. A second influx of demobilized soldiers followed in 1990, after Việt Nam withdrew from Cambodia.  94   When [brickmaking] started to affect agricultural land, the authorities stopped it. During the first years, we had used soil on our own land to make bricks. But as the demand blossomed, we had to dig big holes around our houses. We needed to find soil elsewhere to maintain production, so many people started to dig land in the [agricultural] fields. Because of this, the authorities banned (cấm) brickmaking altogether. [...] The problem was that, as opposed to the land around our houses, [agricultural] land was under state management; it wasn‟t ours to use.  Figure 8 Map of the districts of Đống Đa and Từ Liêm in 1972  Source: Adapted from a 1972 map produced by Cục Đo đạc và Bản đồ Việt Nam  In a matter of a few years, all brickmaking activities had stopped in Hòa Mục. This strong and expedited policy intervention suggests that the central state had the means to “police” the informal sector when it needed to. It also sheds some light on the limits of the state‟s tolerance for informal practices. In Hòa Mục, a social contract between the local population and the state according to which informal practices were tolerated as long as they remained within the domain of land  95   officially designated as residential seems to have been operative. It would seem then, that the regulatory environment within which villagers deployed informal practices, allowed them to get away with practices that did not go beyond the terms of that social contract. Informality, in this sense, was characterized by both rigid limitations (on converting or using paddy land) and flexible areas (residential land opened to all sorts of uses and abuses).  States of Emergency Many local practices described above occurred outside the planned economic development program of the state and local commune. As one informant remarked, self-supply activities, “developed by themselves (tự mở mang), in a spontaneous (tự phát) way that responded to local people‟s needs rather than to state plans.” The villagers I interviewed were fully aware that these activities contradicted the spirit of the socialist regulatory regime in important ways. Villagers thus mentioned that the breaching of the state‟s closed-market economic policy, a principle colloquially referred to as “the interdiction to set up a market across the river” (ngăn sông cấm chợ), was a criminal offence. 11  Respondents also knew very well that horizontal land transactions violated a basic tenet of the socialist land regime.  Disaggregating the State How can we explain this relative freedom enjoyed by villagers with regard to the use and disposition of residential land? Why did local agents of the state not put an end to practices that stood in obvious contradiction to basic political-economic tenets promoted by the socialist regime? Why did supra-local authorities let these practices develop without reinforcing surveillance? 12  And why was the legislative framework governing the use and disposition of residential land in periurban villages not clarified by the central leadership? Building on Migdal‟s “state-in-society” approach, and in line with the understanding of the informal as a socially embedded phenomenon, I argue that answers to these questions can be found by looking at the conjunction of interests served by the various informal livelihood and land strategies described above.  The most obvious interests served are those of the local population. Villagers say that they would simply not have survived through the harsh years of the subsidy era if they had not engaged in “self- supply” activities. In interviews, many informants insisted on the fact that to meet basic nutritional needs, their households needed to access more food than what was delivered by the government procurement system. As a bureaucrat from the People‟s Committee of Trung Hòa puts it in the case  11 Cf. Crimial Code 1986, art. 80 (Sabotaging the implementation of Socio-Economic Policies). 12 Fforde (1993: 103), for instance, notes that throughout the subsidy era, the state rarely tried to eradicate the free market through coercion.  96   of brickmaking: “[F]amilies sold their production (on the black market) even if it was forbidden because they were hungry. They were worried about starving. It was a need of the society (nhu cầu của xã hội).” Informants similarly insisted on the way that self-help housing was just about the only way out of the increasingly cramped quarters of their houses.  In most cases, villagers did not try to hide these practices from the administrative officials and party cadres responsible to enforce state regulations locally. When questioned about the attitude of local officials working for these two governing institutions, villagers say that informal practices taking place on residential land were encouraged without being directly supported. 13  Lâm, a 58 year old former state-worker, explains:  The policy (chính sách) of the authorities was to let us be. Fruit from our trees, the animals that we raised ourselves, those we could sell. No one forbade us to do these things (không ai cấm đoán gì hết) [...] In fact, the authorities encouraged (khuyến khích) us to trade fruit and vegetables so as to get some money. Otherwise how could we make a living? Life used to be very difficult you know, not like today. 14   This interpretation echoes ideas about good political leadership widely held in the North of Việt Nam. What villagers saw in their local leaders‟ passive attitude towards informal activities is the political virtue (giáo đức) of governing with empathy (tình cảm) for the difficulties and particular situation faced by local populations (see Malarney 1997 for a discussion). Yet this was probably not mere altruism on the part of local state agents. As discussed by David Koh (2006) in the context of urban Hà Nội, by ignoring (or even supporting) informal activities that benefited local people or respected their traditions, local leaders strengthened their own political position and clientelistic relationships within their communities. In doing so, local agents of the state yet also became obstacles to the penetration of the central state‟s regulatory power into society.  I could only interview two local bureaucrats from between 1960 and 1990. They each explained turning a blind eye on self-supply and self-help practices in slightly different terms. One of them explained that the subsistence of his family also depended on such informal activities. “How could I  13 Abrami and Henaff (2003: 109-13) describe more direct involvement of local bureaucrats in “informal” practices in two periurban villages located in the region of Hà Nội. Drawing on research conducted in the 1990s, the authors report various practices that facilited the formation of specialized networks and activities outside of agriculture. These include devising of a system where co-op members involved in non-agricultural activities could “buy” their working points from cooperative leaders for grain or money. 14 Secondary literature reports that villagers from other localities seeking off-farm opportunities in the pre- reform period were harassed by commune-level officials. This stunted the development of informal practices and perhaps encouraged the intensification of passive forms of resistance on the part of local populations (e.g., Chi Huyen Truong 2001; Tô Duy Hợp 1995).  97   ask my neighbours to stop selling their chickens at the market” he remarked, “when my wife and her sisters did the same thing?” The second official I interviewed provided a somewhat different rationale that emphasized issues of social acceptability. He explained that some prescriptions from the party-state were simply impossible to enforce because they went against the grain of villagers‟ needs and customs. In the case of private land and housing transactions, he explains:  In the law, it said that the selling and buying of land was forbidden, but the villagers continued this practice as usual (bình thường). It‟s like the trading of [American] dollars. Officially, only registered banks were allowed to do so. But in practice, the state management did not function and people continued to buy and sell [currency] knowing that they could. These were needs of the society (cầu của xã hội), necessities that ought to be met (tất yếu là phải có). That‟s what happens when the law does not work. Then, it leads to problems [with enforcing] interdictions (vấn đề cấm). It‟s like trying to build a one-way street when people need to go in both directions. You build it today and say that it is one-way. But tomorrow, the people will have already turned it into a two-way street.  The comments of these local bureaucrats are a testament to their own delicate political position, not only as agents of the state, but also as members of the local community. Similar to the situation in China (e.g., Chan et al. 1992; Shue 1988), local bureaucrats in Việt Nam face conflicting obligations towards central authority and local constituents which they are supposed to govern that often pull them in opposite directions. Their position requires that they balance loyalty to the party- state and desires to please higher echelons with strong ties and affinity with members of a local community and kinship group into which they are deeply embedded (Kerkvliet and Marr 2004; Fforde 1989; Vickerman 1986; Koh 2006).  As the two examples above show, during the subsidy era, local agents of the state were not separated from society. For different reasons, both of our respondents chose to turn a blind eye to informal practices. In doing so, they failed in their roles as state officials, but enhanced their standing as carrying out their perceived duties as part of the local community. This sheds some light on how the local-level party-state agency may undermine the more or less well-articulated and coherent schemes designed and directed from the top. Though useful, this “state-disaggregation” approach that examines low-level interactions of state-society relations (cf. Migdal 1994), is still insufficient to explain how new or hybrid ways of doing things could arise that contradicted the directives or values of higher-level, formal institutions. Understanding this requires that we move up one level to examine the interests of the center in letting these practices persist alongside the official way of doing things.     98   Legitimacy Matters “Empathic” governance, personalized-clientelistic relationships, and respect for local traditions by local bureaucrats were all made possible by the near total absence of directives from higher levels of government to put an end to self-supply and self-help housing practices. As discussed above, this only occurred in rare and very specific circumstances, such as the ban on brickmaking. There is no doubt that municipal authorities, ministerial bureaucrats, and members of the central leadership, all of whom were based in Hà Nội, knew what was going on in periurban villages. It certainly became very visible, such as when periurban villagers started to sell home-based farming production and bricks on the black market or when wholesalers entered the city with carts full of fruits or bricks. This then raises the question as to why the district or provincial government did not bring lower- level bureaucrats back into the fold and require them to enforce the central state‟s dictates more forcefully.  Several factors probably combined to limit the central state‟s surveillance of activities taking place on the residential land of a place like Hòa Mục. The livelihood and land strategies deployed by local people were small concerns for the central leadership of the party, preoccupied with national defence, declines in agricultural output, and a failing industrial sector. Other possible explanations proposed in the literature relate to the state‟s inability to extend its regulatory powers in certain areas of society. This problem is differently attributed to a weak administrative structure, a strong “civil society,” and to local officials deeply embedded in local patron-client relationships (Thrift and Forbes 1986: 81-3, 101-4; Woodside 1979: 318-401; Koh 2004).  But if we look at the problems that confronted the party-state‟s leadership at the scales of the greater region of Hà Nội and of the nation, we can discern other possible motives for the persistence of both ambiguous legislation and the tolerance, by the DRV leadership, of areas beyond its regulation. Ben Kerkvliet and Mark Selden (1998) provide us with a first line of explanation that emphasizes the effects of the war against the United States. These authors suggest that compared to the situation in China during the same period, the north Vietnamese state might have decided to give villagers significant latitude in their household activities within the collective framework as a mean to maintain their support for the war effort.  Defaulting on the socio-economic promises on the part of the party-state, especially with regard to food and housing, contrasted with the very concrete promises made by revolutionary leaders during the successive conflicts against the French, Americans, and Chinese. The historian David Marr (1981) reminds us that the millions of impoverished Vietnamese people who rallied behind the communist movement, and who participated in these conflicts, were not merely motivated by  99   ideological ideals and by desire for national independence. As elder Phúc, an anti-American war veteran, indicates, many believed that independence and socialist development would lead to rapid and tangible improvements of material conditions in their lives:  I was like other revolutionaries; I joined the army being quite optimistic about the country‟s future, but also about my own family‟s future. I believed that once the war was over, things would improve a lot for us, all of us. During the war, the government promised the people that after the liberation [of the South], every family was to have a refrigerator, a sewing machine, a bicycle, etc. [...] When we were still there [in the South], we saw that the Americans provided strategic hamlets with soap, food, housing. We thought that we would at least get similar things after the liberation. But we were wrong. Things actually got worse for us after the war...  The portrait of life in Hòa Mục drawn in this chapter confirms that the socialist state had a hard time meeting its promises for material improvements, a conclusion already well-established by the literature on pre-reform Việt Nam (Beresford 1988; Vickerman 1986; Fforde and Paine 1987; Marr and White 1988). Over time the planned economy‟s persistent failure to provide the most basic of necessities to villagers like elder Phúc, who made significant sacrifices either as soldiers themselves or by sending their kids to the front, could not do other than to erode the party‟s legitimacy. As noted by Kervliet (2006: 294), people could—and did—blame the war for deteriorating living conditions, but they also saw production shortcomings in their own collective units. Moreover, material conditions at the local level only kept deteriorating after the war ended in 1975, suggesting that a chronic endogenous problem was responsible for the nation-wide economic hardship.  The food and construction materials produced by villagers on residential land, not only fulfilled basic needs of villagers, but also those of urbanites. In this context, the central state adopted a two- sided approach that combined de facto tolerance of illicit practices that alleviated shortcomings of the planned economy with explicit condemnation of their non-socialist character (see Fforde 1993: 103). This echoes the argument according to which socialism relied on tacitly approved manoeuvring to make it work—if only in form—including a measured ignorance with respect to the operation of a second, or “outside,” economy.15 The somewhat contradictory attitude towards illicit livelihood and land strategies allowed the party-state to maintain the image of a strong ideological orientation and to respond to the economic hardship and survival struggles of both rural and urban populations. In a paradoxical way, that echoes Migdal‟s (2001) dual approach, the VCP struck this improbable balance by allowing various state institutions to deploy governing practices that undermined ideological tenets and laws fundamental to the communist regime.  15 As noted by Portes et al. (1989), in the context of economies regarded as capitalist, the “informal sector” is treated in official discourse as an aberration from an ideal, even though it serves much the same role in making the system work. See also Mingione (1991) and Scott (1998).  100   Conclusion In this chapter, I have described and analyzed the livelihood and land strategies adopted by Hòa Mục households after the socialist revolutionary period, and through the early years of the reforms. I focused more specifically on the disjunction between official state dictates and grassroots practices (some of which were in fact criminal offences). I showed how, with the implicit support of their local leaders, local people regularly and continuously circumvented basic socialist principles regarding the use and disposition of residential land to perform private activities. These activities, which included food production and trade, land transactions, and self-help housing production did not only shape the space-economy of the near periphery, but actually succeeded in localizing the Soviet-style economic model promoted by the DRV.  The observation that neither free markets nor private transactions were ever fully eliminated in Việt Nam, and this even during the most thorough period of socialist transformation, is nothing new in itself. The literature on the subsidy era shows that an orthodox model of socialism (such as that which could be found in the Soviet Union during the 1930s or China in the 1950s) never held complete sway in Việt Nam. Western scholarship further indicates that non-socialist activities—not officially sanctioned, but tolerated in practice, by the state—not only persisted, but also accounted for no small part of economic life during the subsidy era (e.g., Beresford and Fforde 1997; Fforde and de Vylder 1996; Dang Phong 2004).  The specific contribution of this chapter to this literature is threefold. First, I documented the particular conditions and manifestations of private practices within the context of a periurban village. Second, I presented data challenging the idea, widely re-circulated in the literature, according to which the đổi mới reforms actually generated urban changes in the rural zones surrounding the Vietnamese capital. I am not denying that such changes accelerated considerably beginning with the reforms. What my analysis instead indicates is that many of the phenomena responsible for the early in situ urbanization process pre-existed these reforms. The following chapter will further show that many of the residential-based, non-agricultural practices crafted by local households prior to đổi mới became economic springboards, allowing these same households to “leap” into the reform era.  Third, and perhaps most importantly, I suggested that this outcome was made possible by the state‟s flexible responses to multi-faceted, societal and legitimacy crises. Added to the almost incessant warfare, a general decline of living conditions characterized the subsidy era. For the overwhelming majority of the population, this decline was a tangible outcome of chronic inefficiencies in the  101   formal economic system put in place by the regime. Such economic and social upheaval made the pursuit of a rational-modern territorialization strategy by the DRV leaders particularly difficult. Under these conditions, I hypothesized that the party-state‟s need to preserve its legitimacy largely overrode and outweighed attempts to extend modern-rational bureaucratic controls over people and places. Similar to the approach adopted during the early socialist revolution period (cf. Chapter 4), the post-colonial state preserved its legitimacy by relying on systems of exceptionalism and on localized adaptation (and even lifting) of formal rules. In the particular context of the subsidy era and of Hòa Mục, such state governing practices took the form of deliberate policy ambiguity, selective enforcement of land management (with marked difference for residential versus agricultural land categories), and decentralization of regulatory power to local authorities.  Such informalization of the central state‟s regulatory function and policing powers served a coalition of interests. It encouraged periurban populations to improve their lot outside of the formal institutional system, and to provide their families with the basic necessities of life (especially food and shelter) through the exploitation of local resources. This, in turn, allowed local agents of the state to safeguard communal and kinship relationships while appearing to villagers as virtuous, “empathic” political leaders. At yet a higher level of the state apparatus, regulatory flexibility in the management of residential lands further allowed the central leadership to stay true to Ho Chi Minh‟s declaration: “[i]f the people are hungry, it is the fault of the party and government; if the people are cold, it is the fault of the party and government; if the people are sick, it is the fault of the party and government” (quoted in Gillespie 2005:46). Understood in this manner, informality is not indicative of a weak state. It is rather a tool of governance allowing the state to serve multiple constituencies and preserve its legitimacy in times of crisis.   102   CHAPTER 6 The New Urban Territorial Order        The party-state faced an unprecedented legitimacy crisis after the country‟s reunification (1975-76). By then, major defects of the Vietnamese economic system manifested in low efficiency of collectivized agriculture and industry, high levels of waste, slow economic growth, and even recession. Galloping inflation, scarce goods, and falling salaries translated into growing social misfortune. While General Secretary Lê Duẩn pushed for a stricter implementation of neo-Socialist policies, the national economy continued to decline steadily, culminating in the devastating famine of 1978. By that point, people‟s faith in the system was seriously dwindling. The VCP leaders‟ attribution of economic hardship to difficulties with the implementation of the socialist model (e.g., natural disasters, international reactionaries, poor awareness of lower administrative echelons, etc.) increasingly failed to convince the masses. It had become evident that economic problems had to do not with the implementation of the socialist model of development but with the model itself (Dang Phong 2004; Fforde and de Vylder 1996; Beresford 1988: chapters 9-11).  The post-war economic crisis and associated erosion of party-state legitimacy underpin the series of policy shifts which came to be known as đổi mới.  Starting in 1979, reforms made increasingly significant departures from the dogmas of the traditional socialist model of development. They progressively altered previously untouchable principles of the Soviet-style model of socialist development. These included reforms of the public ownership of the means of production, collective labour processes, centrally planned economy, fixed prices system, central state monopoly of foreign trade, and comprehensive leadership of party organizations in production units. 1  This profoundly transformed the regulatory environment within which periurban villages like Hòa Mục developed. Yet this was not a uniform or linear process. Following recent analyses (Abrami and  1 The literature on the reforms still debates whether đổi mới resulted from an ideological change in the leadership of the VCP. During the 1990s, analysts tended to associate the reforms with the overcoming of the party‟s conservative faction by economic reformers based in Hồ Chí Minh City (e.g., Porter 1993; Abuza 2001). More recently, authors like Martin Gainsborough (2010) have argued that the reforms might have had less to do with the contrasting ideological positions of specific VCP factions and more with individual‟s personalities, money, and shifts in political protection and patronage.  103   Henaff 2003; Fforde 2005; Gainsborough 2004), I divide đổi mới regulatory changes into two broad stages, which, I argue, had very different effects on the urbanization of Hà Nội‟s outskirts.  The first stage goes from 1979 to the early 1990s. Due to conflicts within the state, this was an extremely murky period. Reform policies “zigzagged back and forth between the old and the still undefined new, with no blueprint to follow, varying in pace and pattern from one economic sector to another and one region to another” (Kerkvliet 2006: 291). An influential line of analysis put forward by the economist Adam Fforde (1989, 1993) contends that, for the most part, the VCP then institutionalized existing endogenous changes, expressed these in terms that suited political realities and a conservative ideological opposition, and then got them embodied in policy. On the edge of Hà Nội, this first stage of the reforms translated into policy experiments and the legal sanctioning of a large array of practices deemed illicit under the then current plan. The latter encompasses those households‟ private production and trading activities, residential land transactions, and self-help housing described in Chapter 5 (for further discussion, see Pandolfi 2001b: chapter 3).  The scholarship on the periurbanization process in the early đổi mới years rarely mentions this normalization aspect of the reforms. It rather focuses on the discretionary nature of the local administration and the breaching, by households, of new regulations (e.g., Pandolfi 2001a; Le Van Sinh 2001; Leaf 2002). Yet, for about a decade and a half, the regularization of households‟ activities connected with the continuing administrative autonomy at the local level and state-backed development programs. 2  This allowed periurban villagers to build on pre-existing livelihood and land strategies and experiment with new wealth-generating activities. In some—but certainly not all 3—localities, this resulted in impressively rapid and innovative endogenous growth. Well-known examples include: the transformation of traditional craft villages into dynamic industrial clusters, the adoption of higher-value added production methods in agriculture, and development of specialized services (e.g., Rossi and Pham Van Cu 2002; van den Berg et al. 2003; Abrami and Henaff 2003:109-11). Illustrating this new economic context, a villager from Hòa Mục called the first stage of the reforms: “the policy of working a lot, eating a lot; working a little, eating a little” (chính sách làm nhiều, ăn nhiều; làm ít, ăn ít).   2 A good example of this is the VAC (and acronyme standing for Garden-Pond-Pig/Henhouse) development program that encouraged the valorization of residential and other non-agricultural lands in rural areas through a mix of gardening, fish-farming, and livestock breeding activities (see Chabert 2004: 161-2). 3 The reform also created a more competitive economic and labour environment within which some households and communities struggled to find a niche. This resulted in socio-economic differentiation both within and across rural localities (e.g., Luong Van Hy and Unger 1998; Tô Duy Hợp 1995; Akram-Lodhi 2004).  104   The possibilities for periurban households and local authorities to experiment with post-reform economic opportunities nevertheless started to close off during the reforms‟ second stage. This period began in the 1990s. Once the turmoil of the war and reunification was over, and once the economic situation was somewhat stabilized, the VCP leadership set about building a more stable regulatory environment. As made evident by discussions in previous chapters, campaigns to extend the state‟s control over people, space, and economy have historically come in waves in socialist Việt Nam. In this sense, the policy re-orientation of the mid-1990s bears some resemblance to the state-building and socialist-revolutionary campaigns of the late 1950s.  This time though, reforms sought to strengthen the state monitoring power over economic activities, to impose order over informal practices, to give the country‟s integration into regional and international economies a greater role in both the public and private sectors, and to regain managerial control over lower-level administrative bodies (Abrami and Henaff 2003; Gainsborough 2010). During this second stage, the policy-making arena was also penetrated by emerging “state business interests.” Borrowing from Gainsborough (2010), I define these interests as those “linked to state enterprises and/or bureaucratic institutions of the party-state which were at the forefront of moves to exploit commercial opportunities that emerged during the reform years” (ibid: 34; see also Fforde 1993). Companies with new state business interests concentrated in sectors that witnessed rapid growth during the 1990s: real estate, foreign trade, banking and finance, retail trade and tourism. As will be discussed below, real estate development is one of the areas that attracted state business interests early on because of the new opportunities for land and housing marketization created by reform policies.  This chapter focuses on this second stage of reforms, and in particular, on a subset of regulatory shifts related to urban development observed during the 1990s. It shows that throughout this ideological and regulatory transition, flexible use of regulatory powers remained an important element in the toolbox of state territorial governing practices. But, since the 1990s, these practices have taken new forms and, as will be discussed in more detail below, are taking two somewhat contradictory directions. On the one hand, the new planning regime seeks to delegitimize regulatory informality at the grassroots level. On the other hand, it supports regulatory exceptionalism in the management of developable periurban land, both to achieve state goals, and for the benefit of emerging state business interests in the real estate sector.  Understanding these changes requires that we step away from the story of Hòa Mục for a moment and look at policy-making processes that occurred at the national and provincial levels. I will therefore attempt to open up the “black box” of Vietnamese policy-making, an area avoided by  105   local scholars because of its political sensitivity, and an area that is difficult for foreign scholars to access. Apprehending post-đổi mới urban development and land policy formulation as well as the role of real estate economic interests in these processes is a difficult task. Policy transition in Việt Nam is marked by considerable tensions and conflicts between different parts of the state and by economic relationships occurring in the grey zone between legality and illegality. 4  As obscure as these processes may be, I believe that they deserve more research attention to make intelligible the political and administrative context within which the urban transition operates.  This chapter is a foray into trying to do just that. It begins with an exploration of the shift in the VCP‟s discourse about the role of urbanization in national development. I then move to analyze changes in urban and regional planning control mechanisms that followed from this more favourable view of urban development, including the formulation of new expansionist master plans for Hà Nội. Next, I discuss the concomitant introduction of a new urban space production model, called “New Urban Areas,” and the state‟s flexible use of regulatory power to encourage its materialization. The conclusion explores how reforms of the land legislation facilitated land redevelopment by real estate capital at the expense of the periurban farmers‟ interests.  Reforming Urban and Regional Development Control Mechanisms The formal acknowledgment by the state of urbanization as a positive development force signalled the end of the metropolitan containment approach that had prevailed in Việt Nam since independence. Between 1954 and 1986, the government concentrated its effort on limiting urban growth. A key objective was to slow down the expansion of large urban centres. The strategies promoted to achieve these goals included the redistribution of population away from large cities, the control of rural-to-urban migration through the residential permits system, and encouraging the development of small towns (Thrift and Forbes 1986; Nguyen Duc Nhuan 1978).  A New System of Cities A more favourable attitude towards cities translated into a restructuring of these approaches and to considerable revision and development of urban and regional planning mechanisms. In 1990, the Council of Ministers adopted new definitions of urban places. 5  The MoC subsequently formulated a double urban administrative hierarchy to control urban growth spatially, economically, and  4 See Gainsborough (2003) for further discussion with reference to Hồ Chí Minh City. 5 The definition of an urban place in use during the 1970-1989 period used the following three criteria: i) at least 2,000 inhabitants; ii) a majority of the population engaged in non-agricultural activities (especially industry); ii) population densities higher than in rural places; and iv) presence of infrastructure for transportation, water, electricity, and other urban facilities (Việt Nam National Institute 1992, quoted in McGee 1995b: 256). For discussion of previous definitions used since 1954, see Thrift and Forbes (1986).  106   demographically. This system first classifies urban agglomerations on a one-to-five scale according to their population size (from type 1, corresponding to one million inhabitants to type five, corresponding to places with less than 100,000 inhabitants). Overlaid on this system is a hierarchical classification of cities into three “administrative classes” that determine their relative political and economic autonomy. 6   The regular production of urban master plans (every five years) is the chief planning tool used by the MoC to work out the details of this administrative system of cities. Since the early 1990s, these plans have promoted four processes of urban growth and change: 1) the infilling and densification of existing urban areas; 2) new construction in the urbanizing peripheries of large cities; 3) the establishment of satellite cities and new towns outside of primary wet rice growing areas; and 4) the reclassification as urban of rural communes that meet urban criteria in terms of population size, employment characteristics, and existing infrastructure (MoC 1999).  The Capital City‘s Urban Expansionist Plans Like Hồ Chí Minh City, Hà Nội is a “Type 1” urban place in terms of population size and a “Class 1” national city directly under the central government‟s management. Planning the urban growth of these two cities, which started to receive migrants and foreign investment early on during the reforms, rapidly became a priority for the central government. In Hà Nội, this translated into the release, in 1992, of a new masterplan for the 2010 horizon.  This document recommended unprecedented administrative and physical expansions of the city unto periurban areas (figure 9). It proposed the redesignation of rural districts (huyện) at the edge of the urban core into new administrative urban districts (quận).7 The implementation of these prescriptions began in 1995 with the designation of the urban district of West Lake (Tây Hồ), to the north of the historic city. This first urban administrative extension annexed a group of villages that had witnessed rapid in situ urbanization during the early years of đổi mới (Leaf 2002: 26-28).  A different planning logic underpinned the subsequent administrative designation of the districts of Green Spring (Thanh Xuân) and Paper Bridge (Cầu Giấy—to which Hòa Mục now belongs) in 1997, and of Yellow Plum (Hoàng Mai) in 1999. Rather than catching up with the urbanization  6 Class 1 refers to national cities under central government management, class 2 refers to regional cities under provincial management, and class 3 refers to district towns, market towns, satellite cities and “urbanized” rural settlement areas. In 1999, this urban system consisted of 3 national cities, 47 provincial cities, and 508 district towns, for a total of 14.7 million people. 7 Similar to the Chinese administrative-territorial system (Ma 2004), the territory of the province of Hà Nội includes both urban (quận) and rural (huyện) districts.  107   process, the annexing of these three large areas to the west and south of the city sought to provide developable space for the planned expansion of urban functions. The 1992 master plan recommended an increase of urbanized areas from 9 to 27 percent of the administrative territory of Hà Nội (which then covered 920 km2). The expansion process favoured by the master plan called for massive conversion of agricultural land to urban uses in newly designated urban districts at the edge of the city. This had major implications for the already densely settled urban periphery and would affect approximately 700,000 people (Quertamp 2003: 214).  Figure 9 Urbanized area in 1992 (left) and spatial expansion prescribed by the master plan for 2010 (right)   Source: Cerise 2009: 242 - reproduced with permission from the author  Blackening Grassroots Practices, Whitening “New Urban Areas”8 The 1992 master plan (and its 1998 revision) promoted a new model of periurban space production referred to as “New Urban Areas” (khu đô thị mới, hereafter KĐTM). These are large-scale redevelopments featuring commodity housing, public services, along with commercial and office  8 This section builds on a paper co-authored with Julie-Anne Boudreau entitled “Understanding the Causes of Urban Fragmentation in Hanoi: The Case of New Urban Areas” published in 2011 in International Development and Planning Review, vol. 33, no. 3, pp 273-91.  108   space. The introduction of this model by planning authorities went hand-in-hand with new discourses and regulations that delegitimized decentralized urban space production practices, and in particular, the small-scale, privately initiated, and largely informal activities responsible for much of the periurban transformations since the early years of the reforms.  I analyze this evolution through Yiftachel‟s (2009b) concept of “gray spacing.” Coined with reference to Israel‟s planning regime, “gray spaces” refer to grassroots practices, geographic zones, or spatial configurations that develop under deliberately ambiguous or weakly enforced regulatory frameworks. According to Yiftachel, by maintaining specific social groups in a relatively high degree of uncertainty these “gray spaces” facilitate the flexible use of formal regulatory power by political and/or economic elites. As explained in the introductory chapter, the author thus calls “gray spacing” the governing practices that consist in “whitening” (i.e., legitimizing, authorizing) practices, people, or spatial configurations that develop in gray spaces, or in “blackening” (i.e., delegitimizing, criminalizing) them. To understand how such governing practices came to play in post-reform Hà Nội, I shall go back in time and briefly review the evolution of state policy discourse related to urban space production and, in particular, housing supply.  The State and People Work Together As briefly discussed in Chapter 4, the position adopted by the DRV after independence is encapsulated in the 1961Central Committee Resolution No 9, declaring that: “[L]and and housing in the cities and towns concerns everyone and is part of the socialist revolution” and that “[T]he state must satisfy the people‟s demand for housing.” In practice, starting in 1954, housing policies in Northern Việt Nam were essentially geared toward satisfying the needs of state bureaucrats, blue- collar workers, and the military. The DRV government launched two operations to respond to the residential needs of these families. A first policy requisitioned and redistributed the private housing stock that already existed in the city, forcing owners to welcome new households into their domestic space. A second policy aimed at generating housing on a larger scale. Launched in the late-1950s, it consisted of building so-called “collective zones” (khu tập thể, hereafter KTT) on the outskirts of the existing city (Lacoste et al. 2002; Logan 2000: chapter 6).  Insufficient funding throughout the subsidy era did not allow the state‟s construction programme to satisfy the growing demand for urban housing. By 1980, only 30% of the state‟s employees based in the capital city had benefited from subsidised housing (Lacoste et al. 2002: 7). New households formed during this period had little choice but to share already existing housing located in the old city and in KTTs. In 1981, the average population density in the urban core was 18,000 persons/km 2 , reaching 33,000 persons/km 2  in the old merchant city (Phạm Khánh Toàn 1991 quoted  109   in Pandolfi 2001b: 92). Since the central state continued promising all citizens the “right to housing,” the lack of investment in housing production became a major source of dissatisfaction among an urban population that had doubled between 1960 and 1989 (from 460,000 to 900,000 inhabitants) (Quertamp 2003: 272).    The housing shortage reached levels that threatened the legitimacy and stability of a regime already weakened by the economic crisis briefly outlined in the introduction of this chapter. In 1987, the state recognized the insufficiency of its resources and withdrew from its subsidized housing programme. In line with many other “first stage” policy reforms, the state opted for decentralizing some of its power. The urban housing question was therefore reframed in terms of a policy known as the “State and People Work Together” (Nhà nước và Nhân dân Cùng Làm) whereby, for the first time since independence, private stakeholders were formally invited to produce their own housing.  Figure 10 Post-reform housing built privately by households  Source: Reproduced with permission from Vincent Bertholon  The “State and People” policy encouraged various state institutions (army units, SOEs, mass organizations, ministries, universities, etc.) to get involved in housing redevelopment operations. It authorized these organizations to create construction branches through which they could be allotted developable land on the rural-urban interface. Between 1988 and 1992 these small scale, construction branches proliferated rapidly. In principle, their role was limited to the redevelopment of land allotted by the state in the form of serviced-plots or housing, which would then be  110   redistributed to state employees. In practice, the “State and People” policy became the entry point for bureaucrats and other state employees into the effervescent, post-reform, residential land and housing market (see Pandolfi 2001b: 374-5).  Table 1 Housing space produced in Hà Nội between 1981- 2008 (in m2)  1981* 1985* 1991* 1995* 2000 2005 2008 State and institutions** 40 000 (89%) 145 000 (91%) 45 000 (31%) 75 000 (31%) 82 000 (30%) 716 000 (61%) 578 000 (44%) Households  5 000 (11%) 15 000 (9%) 100 000 (69%) 165 000 (69%) 425 000 (69%) 635 000 (39%) 1 187 000 (56%) Total 45 000 160 000 145 000 240 000 507 000 (100%) 1 351000 (100%) 2 100 000 (100%)  * The data available for the period 1981 to 1995 is approximate. Data for this period groups together housing built by public national and municipal organisations and by construction companies who benefit from private or foreign capital. ** This category includes housing produced by the central and municipal state, state institutions that developed land allotted to them and state companies in joint venture with foreign investors. Source: HSO (2009: 215) and Pandolfi (2001: 315)  The decentralized approach to housing production also encouraged the development of grassroots practices, which, as discussed in Chapter 5, had remained active in and around Hà Nội under the plan. The “State and People” policy also opened the way to new informal practices such as the modifications of KTT buildings, and the encroachment on public spaces or on periurban agricultural land (Koh 2004; Geertman 2007; Pandolfi 2001b: chapter 8). Meanwhile, periurban local governments continued to convert agricultural land to residential use and to distribute individual plots to villagers under the “population de-densification” policy (chính sách giãn dân) discussed in Chapter 5. This brought an additional supply of developable periurban land into periurban housing and land markets. Summarizing the intense real estate activities that flourished during this period, the French architect Christian Pedelahore writes:  An irrepressible movement is catching hold of all social strata. There is a frenetic will to extract a profit, at forced march, from all urban and urbanizable land. Individuals, private enterprises, and a large number of enterprises of all sorts are thus physically brought cheek by jowl. The valorizing of the city‟s surrounding lands is the great business that occupies, almost daily, each member of the contemporary society. (Pedelahore 2006: 20, my translation)  The “State and People” policy undoubtedly met its primary goal: it resulted in a substantial increase of housing in and around the Vietnamese capital city (see table 1). Within this rather open-ended  111   and loose policy environment, the city‟s pre-colonial, colonial, and socialist areas were also mended and woven back together. This happened through the restoration of old buildings, the rehabilitation of the imagery of the French city, the spontaneous conversion and transformation of KTTs into vibrant mixed-use neighbourhoods, and the endogenous development of new housing styles that drew on the city‟s historical forms (figure 10).  Perceived Inadequacies of Decentralized Urban Space Production Despite all of these benefits, local professionals, planning authorities, scholars, and foreign donors considered decentralized urbanization a less than optimal solution to housing production and urban development. These groups criticized the built environment resulting from the “State and People” policy for its inability to reach desired human densities, to contain urban sprawl, and to protect periurban arable land. More specifically, they pointed out how this form of urban development fails to provide basic urban services (schools, parks, etc.), hinders the expansion of transportation infrastructure, and the development of commercial zones. In a recent interview with the Vietnamese press, a United Nations representative thus stated:  Informal development and, to a certain extent, “formal” development without appropriate infrastructure provision and development controls have produced negative effects on people's lives and the environment. In reality, a high proportion of households living in the fringe areas still suffer the difficulties of clean water and electricity shortages and lack the hygienic facilities of sewage and drainage. They also face problems of congestion and poor traffic accessibility as well as problems in social conflicts, social evils and income disparities (Ian Howie, quoted in VET 2007)  A well-respected member of the Vietnamese architect-planner community whom I interviewed about these issues in 2009 similarly remarked:  [A]t first, the government found that the solution to the housing problem was to give land directly to people who would then use their own capital to build a house on it. At that time, the state didn‟t have the capital but the people did and they could build housing by themselves. But after a while, we saw that this wasted a lot of land. Individually, households could only build 5-storey houses. The population was already very large. If we continued to let people build individually, it would require a very large territory and it would take a long time to shelter an ever-growing population.  In the press, in conferences, and through reports addressed to planning authorities, Vietnamese built environment professionals also insisted on the haphazard and chaotic physical outcomes that resulted from privately initiated residential production. They emphasized how such a built environment blemishes the face of the capital city of a nation that wishes to project an image of modern and orderly development (PHC 2000; Nguyen Ngoc Quang 2004; Đinh Đức Thành 2001).  112    Another problem was land speculation. Beginning in the 1980s, state institutions and rural local administration distributed developable land to their employees and constituents at much lower prices than (black) market values. This created incentives, especially among state employees who could acquire more than one parcel of land and then resell them to non-state employees, for these people to act as de facto real estate brokers (Geertman 2008). This problem grew in proportion in a context characterized by high pent-up housing demand and a lack of alternative investment opportunities.  After more than 30 years of centralized socialist economy, households lacked the necessary knowledge to be able to invest in economic activities such as manufacturing and industrial production. Moreover, the industrial sector has remained dominated by SOEs and foreign capital, leaving little room for smaller private domestic investors. Households have further tended to stay away from the domestic banking system. Illustrating this tendency, a representative of the World Bank in Hà Nội estimated that in 1998, less than a quarter of households‟ savings entered the banking system (cited in Pandofli 2001: 8). Conversely, households‟ investment in land and housing and rampant speculation became a well-known phenomenon. This contributed to periods of “land fever” (sót đất). Hoàng Thị Lịch (1999), for instance, reported that land values in the Hà Nội region quadrupled between 1991 and 1992. 9   State Recentralization The “State and People” collaboration program was abandoned in 2000,10 giving way to a policy framework in which the “New Urban Areas” (KĐTM) model of development occupied a central place. Planning authorities introduced this periurban space production model not only as a better way to channel and organize the capital city‟s urban expansion but also as a “cure” for the multiple problems caused by the decentralized urban expansion processes. Official statements from planning authorities, relayed by the media and by the built environment literature, 11  insisted on the importance of repositioning the state and its regulatory powers to the center of the urban space production process. This discourse underscored the capacity of KĐTM redevelopment projects to produce more orderly, functional, and accessible neighbourhoods on the city‟s edge.   9 For a discussion of these issues in other Southeast Asian contexts, see Evers and Korff 2000: chapter 5. 10 SOEs and other institutions under state authority nevertheless continue, to this day, to allocate land for purposes other than those originally planned, including residential development. Such practices are regularly criticized by the national press (e.g. Hai Bang 2009; VNS April 24, 2009; VNS July 12, 2009). 11 Such as Tạp chí Kiến trúc Việt Nam, the official magazine of Association of Vietnamese Architects, or Tạp chí Xây dựng, a magazine published by the MoC.  113   Figure 11 “The Manor,” a KĐTM under construction near Hòa Mục  Source: Reproduced with permission from Vincent Bertholon  New urban development controls supported this discourse. These, for instance, required higher occupation densities through the construction of high-rise apartments that were supposed to occupy at least 60 percent of the built surface in KĐTMs. Regulations also required the production of “complete” neighbourhoods that included schools, parks, daycare facilities, and other public services. To supersede spontaneous forms of urban development prevalent in and around the capital city, planning authorities further promoted the construction of good quality housing units located in modern environments, and affordable to a large segment of the urban population. To ensure accessibility, public policies required that 30 to 50 percent of new housing units be accessible to low-income households (gia đình thu nhập thấp).12 The KĐTM model of urban development also attempted to renew the image of state-sponsored housing production by distancing itself from the failing KTTs built in previous decades (see figure 11).  Yet, like its socialist predecessor, the KĐTM model of urban development is part of a social experiment whereby the Vietnamese state, planners, and architects seek to engineer a new city (and citizenry) through the physical and social structuring of space. KĐTMs are one of the strategies  12 In Việt Nam, this expression refers essentially to civil servants, military personnel and families of veterans receiving social assistance (gia đình chính sách). The “low-income” category therefore does not apply to the disadvantaged population living in rural areas, those who work in the informal or private sectors, and illegal migrants. In principle, the homebuyers‟ plan for these low-income households was to be guaranteed during the project approval process by the setting of state-controlled selling prices for apartments and serviced plots.  114   within a larger city-building project aimed at reshaping Hà Nội into a place that projects a global image of order (trật tự), modernity (hiện đại), and rapid economic development.13 The role of the “New Urban Areas” in this project is evidenced by the municipal by-laws governing the realization and use of these new spaces. These regulations seek to produce neighbourhoods devoid of a variety of traditional configurations and practices observed in the urban core such as vernacular housing types, mixed usages of residential buildings, extension of private and commercial activities onto sidewalks, street vending, and temporary markets. 14   In this, we see that the KĐTMs have inherited a culture of “blueprint,” high-modernist utopian planning similar to that observed across various postcolonial settings (e.g., Holston 1989; Lico 2003; Bray 2005). The Vietnamese planning culture builds on the premise that the state is the agent of national unity, human progress, and reason, and assumes extreme governmental power to remake society and space geared to an end-state vision. This planning teleology conversely rejects the idea that the state is but one element among a larger ecology of actors in the space production process. It opposes the agency and legitimacy of the state and of its architect-planner agents in reinventing periurban space to what Shaktin (2011) calls “actually existing urbanism.” This approach to planning thus conceives of the transformative actions and intentions of actors operating outside of master planning as antithetical to modernization and economic advancement (for further discussion with reference to Hồ Chí Minh City, see Harms 2011).  Making it Happen The implementation of the expansionist master plans for Hà Nội and KĐTM extensions called for a profound revision of the socialist land regime. The ensuing evolution of the legal and institutional frameworks governing urban development at the national and provincial levels reflects a shift in the state‟s governing practices akin to what David Harvey (1989) has termed “urban entrepreneurialism.” Similar to the various geographic settings to which Harvey applies this notion, urban entrepreneurialism emerged out of economic hardship. In Việt Nam, this shift was associated with a tightening of budget expenditures allocated by the center, very limited fiscal tax bases, and difficulties in attracting foreign capital and channelling it into projects prescribed by the master plan. In the early 1990s, these factors contributed to encouraging the transformation of the rules of  13 For further discussion with regard to Hà Nội and other Vietnamese cities, see Logan (2009), Douglass and Huang (2007), and Waibel (2006). 14 A variety of campaigns aimed at building “civilized urban ways life” (xây dựng nếp sống văn minh đô thị), at improving urban order (trật tự đô thị), and encouraging the construction of “cultured” neighbourhoods (khu phố văn hóa) reinforce this orientation. For further discussion, see Koh (2006: chapter 4), Drummond (2004:162-7), Harms (2009b), and Leshkowich (2005).  115   the periurban redevelopment game so as to accommodate the kind of domestic “state business interests” discussed earlier in this chapter.  As will be seen below, this not only involved the legal sanctioning of real estate markets for non- agricultural land categories, but also the creation of exceptional conditions for real estate state- business interests. It sheds light on a governing strategy which, drawing again on the work of Aihwa Ong (1999, 2006), I analyze as the creation of a “zone of exception.”  In Chapter 4, I applied this concept to the spatially uneven application of socialist revolutionary policy by a newly independent nation-state seeking to establish its legitimacy and authority. In Chapter 5, I relied on the same set of idea to discuss the informal livelihood practices of villagers during the subsidy era. Here, I show how the post-reform entrepreneurial state altered regulatory institutions to bring order into decentralized urban production processes and to facilitate the realization of its master plan. Marked by considerable tensions and conflicts within the state itself, this process created a new zone of exception, which gave state-business interests privileged access to developable periurban land.  Mirages of Foreign Investment At the outset, the main constraint on the realisation of KĐTM extensions was of an economic nature: the financial means of the post-war Vietnamese state simply did not match its ambitions. The political scientist Laurent Pandolfi (2001b: 407-16) recounts how, in 1993-94, the People‟s Committee of Hà Nội asked the Vietnamese Institute of Architecture and Planning (under the authority of the MoC) to produce detailed plans for the new urban areas outlined in the master plan. During the same period, the municipality sought financial partners, particularly foreign, able to inject the capital necessary to realize these plans. Planning authorities expected foreign investors to form joint ventures with SOEs in the construction sector. In this arrangement, the state‟s contribution would be limited to providing low-cost developable land, and to financing the main facilities outside the perimeter of the projects.  The MoC and People‟s Committee did not foresee the gap between their urban development plans and foreign investors‟ financial criteria for investment. While they promoted the development of affordable housing areas in the southern part of the city, foreign investors wanted to develop tertiary activities (services, shops, luxury hotels) in the downtown, and high-end residential areas to the North of the city, on land less prone to floods, and closer to Hà Nội‟s international airport. Foreign investors also required land ownership guarantees that contradicted the Vietnamese legislation then in force. As a result, most of the foreign investors approached during the 1990s refused to invest in residential projects outlined in the master plan. Only one foreign developer (Ciputra from  116   Indonesia) implemented a residential project as a joint venture with one of the city‟s public construction companies. The proposed gated community was geared toward the wealthy expat clientele and is testimony to the discrepancy between the expectations of foreign investors and the urban development ambitions of the Vietnamese planners during this period.  Reluctant Domestic Capital Pandolfi (2001b: 410-28) convincingly demonstrates how the shortfall of foreign capital in the development of new urban areas—aggravated by the 1997 Asian financial crisis—and difficulties to channel it into realizing the master plan forced the Vietnamese state to rely on its internal strengths. The state‟s response consisted in trying to use SOEs to carry out the capital city‟s planned residential developments. The implementation of this backup solution proved challenging. At first, most SOEs were simply not interested in building “New Urban Areas” and other urban infrastructures planned for the city. Enticing these domestic economic organizations to make the master plan happen required major transformations of the state‟s plans for Hà Nội and of the institutional framework governing land redevelopment. Relying on a combination of first and second-hand information gathered in the literature and in the field, I now attempt to reconstruct how this process occurred.  Considering how lucrative the real estate sector had become during in the post-reform years, it is difficult—at least at first glance—to make sense of SOEs‟ reluctance to engage in the state‟s urban development projects. Understanding this problem requires that we put it back in the socio- economic context of the early, post-subsidy era years. From discussions with members of the architect-planner community of Hà Nội, I learned that a major obstacle to the realization of “New Urban Areas” by SOEs related to the urban form that architects working for the MOC, and for the city‟s planning department, wanted them to take. As mentioned above, development controls required that 60 percent of these redevelopments be devoted to high-rise residential buildings.  The opinion of SOEs on the social acceptability of this new urban form probably echoed that of the Vietnamese architects I spoke with in the early 2000s. The vast majority of these professionals then believed that, for cultural reasons, Hanoians would never be willing to move into high-rise towers. 15  This assessment was notably informed by the case of the first residential high-rise in Hà Nội. This building remained empty for several years. Accustomed to living in poorly maintained collective  15 A telling instance of this widespread view was provided to me in the spring of 2001. I spent this period as a foreign-exchange student at the National University of Construction in Hà Nội. I was working then on a studio project that consisted in the design of a (fictive) redevelopment for one the inner city‟s industrial sites. During this period, virtually every Vietnamese faculty member I consulted with about my project told me that for cultural reasons Vietnamese people would never be willing to live in high-rise buildings.  117   housing buildings, urban residents feared that elevators would break down, or that water would not reach the higher floors. SOEs in the construction and real estate sectors were certainly aware that the population of Hà Nội then expressed a marked preference for neighbourhoods realized through the “State and People” policy consisting of four to five storey buildings, single-family houses and row houses (see figure 12).  Figure 12 A “State and People” neighbourhood in Trung Hòa ward  Source: author, 2009  Added to this social acceptability problem were the financial arrangements that the state had devised for the realization and marketing of KĐTM housing. These state corporations then operated in an environment characterized by the severing of state subsidies and progressive “equitization” (cổ phần hóa).16 However, many of them were already engaged in profitable residential production activities through the “State and People” policy. Through this policy these companies could market their products according to informal market values, but now the state not only wanted these enterprises to produce less easily marketable urban environments, but also to commercialize them according to state-controlled prices which were much lower than informal market values that could  16 This expression refers to the partial privatization of SOEs. It involves a restructuring of the enterprise‟s ownership structure according to which the government typically remains the main shareholder. Gainsborough (2010: chapter 5) thus suggests that the “privatization” of SOEs does not reflect a disengagement of the state from the economy, but rather represents new forms of indirect interventionism.  118   be reaped through the “State and People” policy. In a period characterized by an explosion of informal residential production, and multiplication of small construction enterprises, nothing indicated that large investments required by KĐTM could bring the kind of short-term returns that many SOEs were looking for. These difficulties are behind a series of regulatory changes that mobilized a whole complex of forces and social agents, along with latent real estate capital.  The State and SOEs Work Together The informal and often simply illegal way in which institutional arrangements related to land and urban redevelopment have evolved makes any exploration of their content and stakeholders very difficult. The policies and legislative changes that resulted nevertheless allow us to surmise that “state business interests” in the construction sector, founded upon the commercialization of SOEs, and their client and patrons, played an important role in transforming the institutions and mechanisms that govern the production of periurban space in Việt Nam.  During the first half of the 1990s many of these large SOEs, previously operating under the management of the MoC or of provincial governments, were restructured into so-called “general corporations” (Tổng công ty).17 These are conglomerates inspired by the Korean chaebol model through which the post-reform state sought to consolidate national industrial activities. The creation of general corporations meant that these enterprises depended directly on the Prime Minister‟s Office. As such, they benefited from considerable management autonomy, especially with regard to ministerial or provincial-level institutions (Gainsborough 2010: chapter 4). The fiscal and employment contributions of these heavyweight SOEs in the region of Hà Nội further reinforced their negotiating power at the local level. We can therefore surmise that, in trying to make the realization of a given “New Urban Area” project happen, the MoC and Hà Nội government‟s capacity to manipulate the relevant actors was therefore limited (Pandolfi 2001b: 376-7).  The process through which these large SOEs gained preferential access to developable periurban agricultural land and obtained exceptional powers to commercialize it was gradual and characterized by overlapping and often contradictory policy decisions. The degree of intentionality in the creation of this “zone of exception” is therefore unclear. Below is an attempt to interpret the sequence of changes leading to this outcome.   17 According to Laurent Pandolfi (2001b: 376), the General Corporations active early on in Hà Nội‟s real estate sector included the Hà Nội Construction Corporation, Vinaconex, Licogi, and the Thang Long Bridge Construction Company.  119   The first signs of an institutional reorganization of the periurban space production and urban development sectors appeared around 1991-92. The main objective at this time seemed to be to bring order to the booming post-reform construction activities. These early institutional changes included the release of the new master plan for Hà Nội, a document accompanied by a directive from the People‟s Committee that formally banned informal periurban agricultural land conversion. Another important change was the establishment, in 1991 of Chief Architect Offices in both Hà Nội and Hồ Chí Minh City (Nguyen 1996). Typical of the second đổi mới stage, the creation of this agency aimed at recentralizing the state‟s regulatory powers. The new urban planning agency was placed directly under the authority of the Prime Minister, and thus above city-level departments and, in some instances, also above the Chairman of the People‟s Committee of the city. In theory, the mandate of the Chief Architect was (and is still) to formulate urban development plans and regulations. In practice though, as an official from Hồ Chí Minh City remarked, this agency was established “to stop city institutions to do whatever they liked” (quoted in Gainsborough 2010: 58).  Another important policy measure adopted in 1992 sought to rein in the flurry of small-scale construction branches that had multiplied within state institutions since the adoption of the “State and People” policy in 1987. This consisted in limiting the allocation of developable land at the periphery of the city to only the larger SOEs traditionally involved in the construction sector.  In parallel, a small group of these large developers was also selected to carry out “New Urban Area” redevelopments under special legal conditions. Six years prior to the reforms that entrenched these measures into the national legislation, this small group of SOEs was authorized to redevelop periurban land earmarked in the master plan, and to sell serviced plots and commodity housing to individuals according to state-stipulated price lists. To facilitate these redevelopments, the city authorities also relied on a “gray spacing” strategy of maintaining periurban agricultural land under a somewhat under-defined ownership status. To this end, the People‟s Committee of Hà Nội sent out a directive to the cadastral offices of all periurban communes, asking them to “suspend” the land titling process required in the 1993 revision of the Land Law. In less than a decade, 90 percent of Việt Nam‟s farming households had been issued agricultural land titles. The remaining ten percent included the periurban population of Hà Nội, intentionally kept in a “gray space” so as to facilitate the state‟s recovery of the land they farmed.  This state-backed monopoly in the construction sector ended in 1998. While short-lived, it nevertheless left its mark on Hanoi‟s real estate sector. Up to 2005, the vast majority of new urban areas built in the Vietnamese capital were still funded and carried out by this small group of former SOEs.  120   Appropriating Land by Law The implementation of the “New Urban Areas” model of development quickly raised legislative issues. These problems had already emerged after the country‟s opening to foreign investments. During the late 1980s, foreign investors were required to form joint ventures with domestic enterprises to operate in the country. These partnerships typically involved a Vietnamese SOEs that would contribute capital in the form of land-use rights. The Soviet inspired property system still in force during the 1980s, and the associated prohibition of horizontal land transfers, made this practice complicated (see Magennis and Nguyễn Tấn Hải 1992). As rules of access to periurban land redevelopment were progressively relaxed, and entry into this sector opened up to a larger number of economic actors, developers and buyers asked for clearer, more stable, and more easily marketable property rights.  From Regularization to Commoditization In line with other first stage reforms, legal changes during the first half of the 1990s were mostly concerned with regularizing the private land market practices that had persisted throughout the subsidy era. This movement started by acknowledging individual and households‟ ownership of private housing (1991 Ordinance on Residential Housing and 1992 Constitution). The 1993 Law on Land later recognized residential land allotments in perpetuity (provided that the occupiers not infringe the law, and that use-rights not be recovered by the state through its powers of eminent domain). As noted above, this document also required that land documents be provided to all land users and that residential land and housing be combined under the same titling system (the so called “building ownership land-use certificate” or BOLUC). It also granted possessors of residential land titles the rights to lease, inherit, and mortgage use-rights and housing, and to receive compensation in case of recovery by the state (1993 Law on Land and Decree 60).  The acknowledgment of real estate activities and shift towards pro-urban development legislation came with the 1998 revision of the Law on Land. This revision openly sought to facilitate the redevelopment of agricultural land on the periphery of Vietnamese cities by real estate enterprises (public, private, domestic, foreign, or any mix thereof). It opened this market to a larger number of players, and terminated the state-backed oligopoly discussed above. From then on, any registered real estate developers could build, commercialize, and retain profits from the renting and selling of housing and apartments. The promulgation of this law also allowed for a barter system known as “land against infrastructure” (đổi đất lấy hạ tầng) which permits the allotment of permanent residential land-use rights over developable lands to domestic and foreign construction companies in exchange for infrastructure provision (mainly regional roads but also public facilities). This urban  121   development mechanism allows local governments, especially at the provincial level, to encourage land redevelopment on their territory while limiting their contribution in capital.  Decree 22, also passed in 1998, redefined the legislation to smooth out land clearance operations for urban and industrial development projects. Three years later, the 2001 revision of the Law on Land reinforced this orientation. It decentralized land conversion powers at lower levels of the administration and thus encouraged local development initiatives in the industrial and urban sectors. Between 2001 and 2009 (when these powers were recentralized) provincial authorities could convert up to 200 hectares of farmland to non-agricultural uses. 18  New legal texts also relaxed the modalities of access to residential, industrial, and commercial lands for a variety of users, including foreign enterprises and individuals. This came along with an enlargement of individuals, households, and economic organizations‟ rights over these land categories, including the right to use residential titles as collateral or capital contribution in construction projects and development of residential property (Decrees 181 and 188 of 2004. For further discussion, see Mellac et al. 2009).  ―Tấc đất, tấc vàng‖ – A Piece of Land, a Piece of Gold The evolution of institutional arrangements discussed above serves a new coalition of interests that coalesced around increasingly profitable real estate activities. As will be illustrated in the next chapter, this does not happen without creating social tensions. At the heart of what has become a heated, nation-wide societal debate, is the manner in which the state flexibly turned around regulations originally crafted to protect the interests of farming households to deploy its eminent domain and land acquisition powers. 19   Many of these disputes concern the differential treatment of land categories, in particular the differentiation between agricultural and non-agricultural uses. In the wake of the 1980s decollectivization process, a first series of legislative reforms clarified farming households‟ rights over agricultural land. This process reflected compromises between different schools of thought about land ownership. It also integrated preferences expressed at the grassroots level as to how agricultural land ought to be governed. This included Northern Vietnamese peasants‟ preference for institutions promoting the equitable distribution of agricultural lands at the local level and protections against landlessness and other adverse effects of the free market (Nguyen Van Suu  18 For a discussion of the unintended consequences of this decentralization of power over land conversions in relation to the recent administrative expansion of Hà Nội, see Labbé and Musil (forthcoming). 19 Kim (2011: 499) cites a source according to which 85 percent of all complaint letters reported in the country are concerned with land, housing, and compensations. Illustrating the gravity of this situation, the British journalist Bill Hayton (2009: 41) writes that: “land disputes continue to be the hottest political issue in Việt Nam, far more corrosive to the legitimacy of the Communist party than calls for multiparty democracy.”  122   2004; Kerkvliet 2006: 294-5). The legislation related to the agricultural land category was thus articulated so as to safeguard the basic socialist public ownership tenet and to maintain households‟ continued access to a resource that has historically been conceived of as a peasant‟s most important means of subsistence.  Many legal provisions served the latter purpose. Since 1988, the successive revisions of the Law on Land have consistently prescribed allotment of agricultural land to users free of charge (all other land categories are subject to fees). The documents also imposed a ceiling on agricultural landholdings to prevent excessive land accumulation. They further stipulated severe punitive measures to deal with illicit conversion of farming land to non-agricultural purposes. As a cornerstone of communist egalitarian principles, agricultural land is allotted for a defined period of two to five decades at the end of which it is redistributed among local farming households to account for local socio-economic and territorial changes (demography, socio-economic activity, farming land area, etc). 20  Finally, while the 1993 revision of the Law on Land granted farming households the right to exchange, transfer, lease, bequeath, and mortgage agricultural land-use rights, it also imposed tight controls on these transactions. For example, transfering agricultural land requires administrative approval and supporting documentation from three layers of administration at the communal and district levels.  These legal prescriptions, originally aimed at protecting the possessor of agricultural land-use rights have recently turned against them. This is most obvious with regard to the evolution of compensation rules in the case of recovery for redevelopment purposes through eminent domain. The 1988 Resolution 10-CT/TW on “Renewing agricultural management” required that peasants be compensated with farming land so as “to carry on living normally” if their land was requisitioned for state purposes. By the late 1990s, it became increasingly difficult to apply this measure. Population densities are high and land is scarce in urbanizing areas of Việt Nam. To resolve this problem, the 1993 Law on Land required that land developers compensate farming households in cash following the so-called Land-Price Framework (hereafter LPF).  On the edge of Vietnamese cities, this compensation mechanism has since become the object of much contention. Despite successive adjustments that brought the official land pricing formula closer to market values, land-users complain that the LPF systematically underestimates the value of agricultural lands. The authorities justify this situation by pointing to special measures that were put in place in the 1980s to prevent the formation of a free agricultural land market. Chief among  20 The current Law on Land prescribes a period of 20 years for paddy fields and up to 50 years for perennial crops. In the region of Hà Nội, the first rice land allotment period goes from 1993 until 2013.  123   these measures is the principle following which use-rights over agricultural land are merely lent by the state to farming households to provide them with a stable livelihood. Since there is officially no market for agricultural land, the LPF calculates land values using an income-based pricing method that only accounts for the user‟s loss in annual earnings for a given acreage during the number of years left in the allotment period.  In addition, official spokespersons point to provisions in the 1992 Constitution [art. 23] stipulating the eminent domain rights of the state: “for reasons of security and national defence and for the national interest.” In 2007, Decree 84 [art. 34] specified that the “national interest” circumstances under which land-use rights can be recovered include the implementation of projects “for important economic development, for residential areas, or for economic development in urban areas and rural residential areas.” This broad definition of “national interest” has opened the door for the redevelopment of periurban land. As noted by many analysts, low agricultural land prices, institutionalized through the LPF system, further facilitate the implementation of these projects. The valuation method used by the state also contributes to the maintenance of high levels of domestic capital inflow and foreign direct investment into the real estate sector (Pandolfi 2002; Han and Kim Trang Vũ 2008; Kim 2009a; Mellac et al. 2009).21  A Dual Land Pricing System Another source of tension relates to the Land Pricing Framework‟s selective application to agricultural land and the coexistence of this official pricing system with quasi-private land markets. This is what Truong Thien Thu and Perera (2011) call a “dual-land pricing system.” The LPF is indeed only used in land transactions between the state and land users such as allocation, renting, taxation, and compulsory acquisitions. It does not apply to horizontal transactions between use- rights possessors, such as transfers or leasing.  Yet land conversion regulations make vertical transactions involving the state unavoidable: to be redeveloped, agricultural land-use rights first need to be recovered by the state. Only then can it be reclassified to a much higher value as largely inalienable settlement land and then be reallocated to a developer at many multiples of the original compensation price. Commenting on this land redevelopment mechanism, DiGregorio (2009) writes:   21 According to the Ministry of Planning and Investment (MPI), total FDI absorbed into the real estae sector for the period 2006-2008 was four times that of the total FDI in this same sector over the two preceding decades. In 2009 alone, the real estate sector attracted $7.6 billion US, over a third of all FDI in the country for that year (VietnamNet October 14, 2009).  124   This process, common now in and around major urban areas, is an exceptionally regressive form of taxation that has as its moral hazard a virtually unlimited capacity to create revenue for the state and private fortunes for the deal makers as long as the urban growth machine remains propelled by investment. Needless to say, in disputes, the state is generally on the side of developers, most of whom are state corporations or state corporations in joint ventures with foreign and domestic corporations. (ibid: 7-8)  In this context, farming households across Việt Nam are questioning the nature of the sacrifice they are asked (or forced) to make in the name of national developmental goals. These interrogations are echoed in press stories of peasant dispossessions. They are also starting to appear in reports written by Vietnamese researchers who ask questions such as the following:  Land recovery for economic development means recovery of agricultural land, then this land is reallocated to investors or enterprises; this is the manifestation of the government downplaying the role played by agriculture in the country and economy. The question that this raises is: when this land is used for agricultural development, social security, or environmental protection, is it not also serving the objectives of economic development? (Le Duc Thinh et al. 2009: 49)  Conclusion This chapter explored the new urban order that planning authorities and real estate capital have sought to impose on the edge of Hà Nội during the second stage of the reforms. I have argued that, starting in the 1990s, policy changes sought to reaffirm the party-state‟s regulatory authority and to bring order to the informal socio-economic experiments that had developed during the subsidy era and the first years of đổi mới. With regard to urbanization policy, this includes governing practices that emphasize the role of a strong regulatory state and that facilitate the involvement of state business interests in the construction of “civilized cities.” The resulting territorial order is eroding many of the pre-existing social relationships that have benefited the periurban population since independence.  Changes in the urban and regional planning control mechanisms in Việt Nam and the formulation of new expansionist master plans for Hà Nội in 1992 and 1998 are testimonies to the state‟s attempt to regain control over the urbanization process and to reposition planning institutions at the center of the urban development process. The formulation of the KĐTM model of urban development provides further evidence of these intentions. I have argued that this model developed under a new urban entrepreneurial government and new “gray spacing” (Yiftachel 2009b) of governing practices that delegitimized pre-existing, self-help housing production.  The creation of the new periurban neighbourhoods planned by the state also involved the creation of a “zone of exception” (Ong 2006). At first, the state‟s intentions in doing this were probably simply  125   to bring order into the activities of the multitude of construction enterprises that had developed under the “State and People‟s” policy. But the new rules governing entry into the periurban land redevelopment sector and rights over land and commercialization ended up serving a small group of state-connected and state-derived real estate development companies. This state-backed oligopoly also provided incentives to get reluctant domestic capital involved in bringing about the large scale developments prescribed by the city‟s master plan.  The manipulation of the legislation governing periurban land redevelopment offers another source of analysis and explanation. In this chapter, I have argued that, in a matter of about a decade, the state‟s regulatory power moved from efforts to shield agricultural landholders from market forces, to the implementation of mechanisms that make for easier appropriation of developable, periurban land. This was made possible by the Vietnamese tradition of governing agricultural and non- agricultural lands under distinct sets of rules. Similar to the urban/rural land divide in China (Yeh 2005; Hsing 2010), the land category system in Việt Nam now serves to facilitate domestic and foreign investment into an emerging real estate sector.  What some observers perceive as an evolution of Việt Nam‟s land regime towards the “rule of law” (Mellac et al. 2009: 71, 80) might therefore better be understood as “rule by law.” By this, I mean the selective manipulation of the regulatory framework by both the state and real estate capital to facilitate the deployment of eminent domain and primitive accumulation practices (understood in terms of public resources seeping out into de facto private hands). Such manipulation of property rights and development controls to favour state-capital collusion is not unique to Việt Nam. Vast swaths of territory in metropolitan Jakarta, Bangkok, and Manila were similarly captured through mechanisms that allow for the interwoven interests of state and capital to combine when neither alone has sufficient power to get things done (e.g., Shaktin 2008).  This chapter also focused attention on how competing interests in land, held by different social groups (peasants, planning authorities, developers, etc) and incentive structures of these different actors in the urban development process define land governance. In the transitional society of contemporary Việt Nam, the power balance between these social groups is highly uneven. Moreover, the possibilities for underprivileged people to organize politically to further their interests, if not non-existent, are decidedly limited. Yet, as recently observed in both the Vietnamese and Chinese contexts (Kim 2011; Wells-Dang 2010; Hsing 2010), while open spaces for political expression may indeed be fewer or differently structured in authoritrian regimes, people can and do find ways to voice their opinions. I turn to this question in the following chapter, with an exploration of resistance against land appropriation in Hòa Mục.  126   CHAPTER 7 Land for Fresh Ghosts, Land for Dry Ghosts1        The previous chapter explored the new order that planning authorities and land developers sought to impose on the rapidly urbanizing periphery of Vietnamese cities since the 1990s. One of the issues raised was the rise in the number and intensity of land disputes that accompanied this shift. Such conflicts now occur on a daily basis around Hà Nội. Rural and periurban villagers typically protest against the annexation of farmland or residential expropriations by state-backed developers for the construction of residential areas, transportation infrastructure, industrial parks or recreational areas. Occasionally, resisters take these disputes to the provincial and national levels in the form of silent demonstrations in front of the National Assembly and other governmental buildings in Hà Nội and other cities. Conflicts can also take more pointed and violent forms. These events are less well reported in the media but have been discussed by foreign scholars, journalists, and human rights activists (e.g., Tran Dinh Thanh Lam August 10, 2006; Hayton 2009: chapter 2; Monthéard 2010; Kerkvliet 2006; Wells-Dang 2010). 2  For instance, Agence France Presse reported the following on November 8 th  2002:  Violent confrontations between police officers and peasants occurred in a district near Hanoi. Local officials from the district of Hoai Duc, in the province of Ha Tay, 30 kms southwest of Hanoi were trying to clear land by force for its redevelopment. Thousands of peasants gathered to oppose the authorities‟ action. A police officer indicated that some “extremists” detained two other police officers and a bureaucrat for several hours. Eight police officers were injured and hospitalized. Mme Phan Thuy Thanh, spokesperson of the [Ministry] of Foreign Affairs recognized that “some peasant families opposed the government‟s decision and that light confrontations occurred between local inhabitants and officials on duty.” In October, during a similar demonstration, 11 people, including a police officer, were injured (quoted in Mellac et al. 2009: 19-20, my translation).   1 This chapter expands on an article entitled “Urban destruction and land disputes in periurban Hanoi during the late-socialist period” forthcoming in Pacific Affairs, vol 84, no.4. 2 Kim (2011: 499-500) notes that the reporting of land disputes in Việt Nam‟s state-owned media was quite limited in the late-1990s, but then picked up rapidly in the early 2000s as newspapers started to sell advertising and became more sensationalist.  127   Popular resistance is an enduring theme in Vietnamese studies. Peasant uprisings and contentious politics reach far back into the region‟s pre-colonial history (Taylor 1983: 187-90, 211; Woodside 1988: 57). Rural unrest in central and northern Việt Nam was also an important feature of the late- colonial era (Scott 1976; Popkin 1979; Marr 1971, 1981). “Everyday resistance” (Scott 1985) later became a key theme of the scholarship on state-society relations in Việt Nam during the subsidy era. Studies building on this concept revealed the covert strategies devised by rural and urban households to navigate official policies without openly challenging authorities (e.g., Kerkvliet 1995b, 2005; Fforde 1989; Hardy 2001; Koh 2006). Some scholars have further suggested that these quiet struggles later informed key-đổi mới reforms (Kerkvliet 2005; Beresford and Fforde 1997; Dang Phong 2004).  So far, few authors have explored more organized forms of collective action under the reforms. 3  This is particularly problematic in a rapidly urbanizing area where, since the late 1990s, local populations increasingly resort to open forms of confrontation to voice their discontent and promote their interests. This lack of scholarly attention to overt conflicts leaves several questions unanswered: Who are the agents of open contestations during the urbanization process? What are their motives and on what resistance strategies do they rely? And, more importantly, what is the relationship between this contentious politics and the new territorial order, described in the previous chapter, that the state and real estate capital try to impose at the periphery of Vietnamese cities?  This chapter sketches answers to these questions based on the case of a resistance movement that took form in Hòa Mục. Native villagers have acquired a reputation as dissidents (người chống cự) due to their repeated attempts to thwart or reshape urban development projects affecting their locality. My analysis of their discourse and actions indicates that the rise in land disputes on the edge of the nation's capital results from a clash between the assemblage of social relations and state governing practices established since independence and the territorialization project that municipal authorities and land developers try to impose on periurban populations since the 1990s. I thus argue that the protests of periurban villagers dispossessed of livelihoods and land need not be interpreted in mere economistic terms as “rational responses” to hardship and deprivation, but should rather be placed within the framework of “moral economies” (Thompson 1971; Scott 1976).  Local resisters in Hòa Mục build their claim on the reformist party-state by pointing out ongoing violations of the social contract that tacitly bounded political elites to periurban people since independence. Their resistance strategy is similar to what a study on China called “rightful resistance” (O'Brien 1996). Villagers resist by highlighting gaps between what they believe are state  3  Notable exceptions include: O'Rourke 2004, Luong Van Hy 2005, and Tran Thi Thu Trang 2009.  128   values and commitments to their welfare, and the actual process and outcomes of the urban expansion. Villagers thus criticize the new pro-urban territorial order that is promoted at the expense of land relations, “gray spacing” governing practices, and local administrative autonomy that once allowed them and their local government to experiment with livelihood and land strategies. This provides the moral charter for the militant and even violent protests of disadvantaged villagers.  This chapter first outlines changes in households‟ livelihood and land strategies under the reforms. Next, I review the perceived benefits and disadvantages brought to the village by the urban administrative and physical expansion process. This discussion highlights the key role played by institutional reforms giving local administrations a stronger central mandate and by compulsory land acquisition in the emergence of disputes. I then turn to villagers‟ resistance discourse and actions. I discuss the central role played by village elders during resistance in Hòa Mục, the strategies they used, and the claims they made in their confrontation with political and economic elites. In conclusion, I reflect on the meaning of this local response to powerful de facto coalitions between the government and private developers.  Hòa Mục under the “Policy of Working a Lot, Eating a Lot” As discussed in Chapter 5, by the time of đổi mới, Hòa Mục was already engaged in its own in situ urban transformation process. As elsewhere in the Red River Delta region, the reforms accelerated and transformed this process of change. In what follows, I briefly describe shifts in local livelihood trajectories since the early 1980s. The discussion is organized following villagers‟ periodization of the post-reform era into two broad periods: before and after 1995, that is before and after the heating up of the residential land market in their specific area. 4   Conservative Agriculture and Urban-oriented Livelihood Diversification The return to family farming was progressive. 5  In the commune of Trung Hòa, agricultural land was redistributed three times: in 1985 under Directive 100, in 1988 under Resolution 10, and in 1993,  4 The land market in Hà Nội has experienced considerable temporal and spatial variations since the 1990s. According to village informants, the year 1995 corresponds to a turning point in Trung Hòa, yet one should be cautious in generalizing this periodization beyond the surrounding areas. It is also important to note that the residential land market across the metropolitan region has experienced a series of boom-and-bust cycles since the reforms. This makes it difficult to evaluate land market and land values in relation, for instance, to policy changes or external shocks (e.g., the 1997 Asian economic crisis). 5 A first wave of agrarian decollectivization occurred between 1981 and 1987, with the nationwide implementation of household contracts under the aegis of the 1981 Directive 100. During this period, cooperatives allocated land areas to farming households in exchange for a fixed output quota. Cooperatives then continued to supply inputs and allocate labour for land preparation, irrigation, and input distribution, but crop management was devolved to farming households. A second wave started with the 1988 Resolution 10,  129   under the new parameters set by the revised Law on Land. Three main rules guided these two last redistributions: 1) land was exclusively allocated to members of the local agricultural cooperative; 2) the redistribution rate was one sào (360 m 2 ) per farming member in a given household; 3) land was redistributed as equitably as possible between households in terms of soil quality, distance to the village, etc.  As discussed previously, Hòa Mục villagers have favoured the combination of agricultural and non- agricultural occupations within the same household since at least the colonial era. This preference started to change during the subsidy era as parents working as civil servants or state workers sought to orient their children into the slightly more remunerative industrial and govermental sectors. The đổi mới reforms reinforced this tendency, resulting in a progressive decline of the agricultural population. The Chairman of the Farmer‟s Union (Hội Nông dân) of Trung Hòa estimates that members of the local agricultural cooperative represented about 80 per cent of the commune‟s population in the 1980s, a figure that dropped to about 50-60 per cent in the 1990s.  Given the redistribution rules outlined above, households with fewer co-operators received limited agricultural areas during the decollectivization process. The areas distributed to the households in my sample were typically around 2 to 3 sào (720 to 1,080 m 2 ), divided into 3 to 5 non-contiguous plots. 6   Despite these very small land areas, respondents recalled being very happy to return to family farming. The multiple advantages brought by the land redistribution policy are captured by this remark by a bureaucrat from the local Women‟s Union branch:  This policy was simpler and made everything easier. It was more open, it brought us much more freedom [...] When my family could not finish its work in the field, it could delay it or get the neighbours or relatives to help. For this, the work was less strenuous. Before this, if one of us was weak, we could not get the work done. This meant fewer [work-] points and a lack of rice to eat […] Redistributing farmland based on the number of people was more appropriate to a farmer‟s way of life. After that, farmers worked very hard because they cultivated rice for their own household [...] As such, we cared for and fertilized our fields as best we could and productivity increased. We worked hard but then, we had enough rice to eat (được cái thóc lúa thoải mái ăn).  For a variety of reasons, briefly listed here, Hòa Mục farmers did not engage in the kind of agricultural intensification processes observed in other localities. Villagers carried on with a regime  which obliged cooperatives to contract out land to farming households for extended periods (15 to 40 years depending on the crop). Output quotas were significantly reduced, and then completely abandoned during the 1989-1992 break-up of rural cooperatives (for further discussion, see Grossheim 1999; Kerkvliet 2006). 6 The average land area ratios in 1998 were 231m2/person for the district of Từ Liêm compared to an average of 358m2/person for all the rural districts of Hà Nội (Chabert 2004: 62).  130   of two inundated rice crops per year, complemented by the cultivation of vegetables in ponds and on higher lands. The former farmers with whom I spoke attributed this low-level of dynamism in agriculture to three main causes. First are the very limited land ratios available to households. Second is the low elevation of agricultural land surrounding the village and associated irrigation challenges posed by the shift to crops others than wet-rice. Third is the lack of entrepreneurship and innovative spirit in the locality. As elder Đức, a 77 year old former peasant puts it: “We simply didn‟t know how to do it. People here have never planted anything other than rice and vegetables. Someone must start these movements, only then will others follow. But here, this never happened.”  Another important blockage to innovation and development in the agricultural sector—yet one not mentioned by villagers—relates to the attitude and actions of the authorities of Trung Hòa. Contrary to the situations observed in many flourishing localities of the Red River Delta, bureaucrats banned informal agricultural land exchanges in the commune. This prevented the kind of arrangements between villagers which, in other periurban localities, allowed the reorganization of landholdings into the larger contiguous plots needed to introduce mechanization and higher-value crops such as fruits or ornamental flowers (Chi Huyen Truong 2001; Gironde 2001). The officials I interviewed did not explain the rationale behind this decision. One can nevertheless surmise that, in line with the state‟s gray spacing strategy of retaining agricultural land titles, preventing villagers from shifting from seasonal (rice) to perennial or higher value crops, was aimed at facilitating the recovery of agricultural territories earmarked for redevelopment in the 1992 master plan.  The People‟s Committee of Trung Hòa also recovered and reassigned land as soon as locals stopped practicing agriculture. Farmland areas available to households thus shrank as members of their workforce shifted towards more lucrative, urban occupations. This situation contrasts with that observed across most of the Red River Delta region where farming households could diversify their economy by sending members to work in the city, or engaging locally in non-agricultural work, while retaining their alloted farmland for the 20-year period prescribed by the 1993 Law on Land. By 2000, the households I interviewed possessed, on average, only one or two sào (360-720 m 2 ) of agricultural land. Similar to the colonial era patterns discussed in Chapter 3, it was most often an elderly or a female member of the household who cultivated it, with the assistance of family members or hired labourers during transplanting and harvesting seasons.  None of the households I interviewed marketed their harvest. Grains were essentially used for family consumption. Actually, the garden-sized paddy fields of Hòa Mục barely met the basic needs of households. In hindsight though, now that the state has recovered this land, many farmers lament the loss of self-sufficiency and food security that these small plots brought to their family. Coupled  131   with the high inflation rates of the last decade (reaching nearly 30 per cent in 2008), the land seizures that accompanied the urban expansion process had a devastating impact on the economy of those households that still depended on their own agricultural production to get by. Today, these people struggle to find work, and suffer from the adverse effects of the more volatile urban economy. Giang, a 40-year old former farmer who now runs a tea stall (quán nước) on her porch, summarizes these concerns as follow:  The economy is not steady (vững). Before, all I needed was rice. I cultivated my own paddy field so I knew I would always have rice. But now, if I want to eat, I need money [...] This life is not like farming, it‟s unpredictable (không thể đoán trước được). I know for today but who knows what tomorrow will be like (hôm nay biết ngày hôm nay, mai chưa chắc đã biết). At my age, I would like to be retrained to find a stable job (làm việc ổn định) and earn more money. But for now, I am still jobless, so I run this small tea stall. It‟s not bringing much revenue in. I don‟t have to tell you, really. You can see for yourself: since the interview began I didn‟t sell a single thing, I haven‟t had a single customer. You see that and you know it all.  Even before the land recovery, farming households‟ basic subsistence required that they find supplementary work in the non-agricultural sector. Throughout the 1980s and 1990s, villagers thus built on the “side-occupations” (nghề phụ) they had practiced since the subsidy era and earlier. Many continued to grow vegetables and fruits, and to raise animals on their residential land.  By the time paddy-fields were seized (in the early 2000s), some peasant households had also diversified their livelihood basis by taking jobs in the urban sectors, especially in the petty trade and construction industry. Mr. Tựng, a 50-year old former state worker, describes this process as follows: “Back then, the main thing on the urban outskirts (ngoại thành) was planting rice and vegetables and also working for hire (làm thuê), like rural families now, going up [to Hà Nội] to work for others. Like them, we worked in the inner city (nội thành). For example we pedaled a cyclo [rickshaw], sold vegetables at the market, etc.” But the competition for jobs on the newly liberalized urban labour market was ferocious. Periurbanites, like the people of Hòa Mục, were not alone in seeking work in Hà Nội. So too did peasants from all over the Red River Delta, and hundreds of former factory workers furloughed during the big lay-offs in the state sector during the early 1990s (Beresford 2008).  In many cases, urban jobs turned out to be only temporary, leaving villagers jobless for significant periods. Another problem, as many informants pointed out, was that the purchasing power of the money so earned rapidly lost value in Hà Nội‟s high-inflation environment. Villagers with a university education and personal connections within governmental institutions generally faired better in this new economy. These people could integrate into civil service jobs or start a small  132   business on their residential land. These doors were neverthelesss closed for the majority of farmers who had low education levels, little work experience outside of the rice fields, and few connections outside of the commune. For them, the forced integration into the urban economy often translated into livelihood insecurity and social instability.  “Counting on Land” By the mid-1990s, villagers started to reorient their livelihood strategies to take advantage of the rapid renaissance of residential land and housing markets in periurban areas. The demand for residential land and housing emerged progressively across the territory of Hà Nội. Again, Mr. Tựng remarks:  For some time, we could see the land fever (sốt đất) hitting different places around us, but it only reached our village around 1995. Before that, no one counted on land (không ai tính đất) [to make a living]. But after 1995 and later in the years 2000, land began to have a commercial value (giá trị kinh doanh). Land could be sold (bán được). Life in the village then changed completely.  New livelihood strategies emerged which capitalized on residential land. One of them was trading of residential plots with land speculators as well as migrant households seeking cheaper land near the capital city. Land so traded included that which had been distributed by local authorities to villagers under the “population de-densification policy” (chính sách giãn dân). By 1999, the year of the last distribution by the local People‟s Committee, this land supply mechanism had turned into a “pot of gold” for villagers. Through it, villagers could acquire residential areas of 60 to 80 m2 for a minimal fee of 400,000 vnđ/m2 (approx $25 cnd/m2). Villagers who resold land allocated that same year, could get around 5 million vnđ/m2 (approx $325 cnd/m2). A decade later, the same land sold for about 70 million vnđ/m2 (approx $3 800 cnd/m2). The commodification of “de-densification land” by villagers thus generated unprecedented fortunes in a locality whose native population has historically struggled to meet basic nutritional needs. 7   Another significant source of marketable land was that on which villagers already lived. These larger landholdings had generally been obtained either through family inheritance or through earlier de-densification land distributions. Given difficulties in converting agricultural land to residential use outside of the “de-densification” policy, home gardens, fishponds and farm house courtyards, all considered settlement land, became primary sites of construction. Those families that managed  7 This led a small number of villagers to the conclusion that the local authorities purposefully deployed the de-densification policy for their benefit. Local bureaucrats responsible for managing population de- densification in the commune denied this interpretation. They rather indicated that they were glad to terminate this program, which due to the escalating land values had become a source of tensions in the local community.  133   to retain considerable areas of residential land during the Land Reform process are now holding highly valuable capital assets. These large residential compounds can now be subdivided and sold to generate cash, which they can reinvest in new economic activities or, as will be discussed below, used to build rental space for temporary migrants. Thus, half a century after the fact, the “partial” implementation of the Land Reform described in Chapter 4, and in particular the exclusion of residential land from this process, defines households‟ relative capacities to engage and benefit from perhaps the most lucrative economic sector in post-đổi mới Hà Nội: real estate (see Labbé and Musil forthcoming).  Figure 13 New urban-styled, multi-story houses in Hòa Mục  Source: author, 2009  Land supply—through the commodification of de-densification policy and subdivision of ancestral plots—led to population growth and changes in its social composition. The official population of Trung Hòa increased sevenfold between 1955 and 1997, from approximately 2,200 to over 14,000 inhabitants. Nearly a third of these new inhabitants were permanent migrants from the inner city and neighbouring provinces who bought land and relocated to the commune since the 1990s. The land and housing market boom also contributed to the densification of the built fabric and to the  134   emergence of urban-styled, multi-story homes that progressively came to replace the traditional single-storey, thatched and brick houses of the village (figure 13).  This transformation of the village‟s built form occurred both through the construction practices of migrant households and from villagers‟ demolition and reconstruction activities. This pattern of investment in housing upgrades is observed in other parts of the Red River Delta (DiGregorio 2009: 5). It follows a long-standing custom already noticed by Gourou (1965 [1936]) during the colonial period: “[A]s soon as an Annamite earns some money, he constructs a respectable or very handsome house, according to his resources, which will worthily shelter the ancestors‟ tablets. Successors will inherit it and only sell it in case of utmost necessity.” (ibid: 288, my translation).  Quoting figures from the Ministry of Social Affairs (MOLISA), Tô Xuân Phúc and Drummond (2009) indicate that 60 per cent of households used the compensation they received during agricultural land recovery to build a new home and fewer than three percent used it for vocational training. They note that, “as a result, many former farmers now have multi-storey villas and motorbikes for every member of the family, but are unable to meet their daily food needs” (ibid: 5). In Hòa Mục, I similarly observed how traditional patterns of investment and conspicuous displays of new wealth through the construction of large houses combined to create what the two authors cited above have called “patterns of self-produced poverty.” The result, especially among the poorest segment of the population, is increased livelihood precariousness and descent into debt.  Another livelihood strategy developed during the 1990s. Following the example of nearby localities, villagers started to build a kind of makeshift building called a “lodging house” (nhà trọ) on their residential land. These buildings are divided into rooms that accommodate students attending nearby colleges and universities, seasonal migrant workers, along with young couples unable to afford accommodation in the inner city. Bureaucrats from the People‟s Committee of Trung Hòa estimated that about 30 per cent of Hòa Mục households operated a nhà trọ in 2009. This activity has become the main source of income of state workers laid-off in the late-1980s and of farming households after the revocation of their agricultural land-use rights in the early 2000s.  The New Urban Order Comes to the Village While Hòa Mục villagers reoriented their livelihood strategies to take advantage of an emerging land market, important institutional changes discussed in Chapter 6 were taking place in policy- circles. Two new policy orientations affected local people directly. First is the designation of the new urban district (quận) of Cầu Giấy and ensuing reform of the local administration. Second is the implementation of a series of land redevelopment projects in and around the village, including  135   unprecedented penetration of state eminent domain powers into the village residential and ritual spaces. These two sets of changes unsettled the territorial order under which the population of Hòa Mục had previously functioned, and became the backdrop of the villagers‟ overt conflicts with developers and the authorities.  The Upgrade to the Urban Created in 1997, the new urban district of Cầu Giấy (lit. Paper Bridge) was carved out of the larger rural district (huyện) of Từ Liêm. Most native villagers received the news of the annexation of