Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Vietnam decollectivizes : land, property, and institutional change at the interface Scott, Steffanie 2001

You are currently on our download blacklist and unable to view media. You will be unbanned within an hour.
To un-ban yourself please visit the following link and solve the reCAPTCHA, we will then redirect you back here.

Item Metadata


831-ubc_2001-715345.pdf [ 30.73MB ]
JSON: 831-1.0090738.json
JSON-LD: 831-1.0090738-ld.json
RDF/XML (Pretty): 831-1.0090738-rdf.xml
RDF/JSON: 831-1.0090738-rdf.json
Turtle: 831-1.0090738-turtle.txt
N-Triples: 831-1.0090738-rdf-ntriples.txt
Original Record: 831-1.0090738-source.json
Full Text

Full Text

V I E T N A M D E C O L L E C T I V I Z E S : L A N D , P R O P E R T Y , A N D I N S T I T U T I O N A L C H A N G E A T T H E I N T E R F A C E by STEFFANIE SCOTT B A , Simon Fraser University, 1993 M A , University of Guelph, 1995 A THESIS S U B M I T T E D IN P A R T I A L F U L F U L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F D O C T O R O F P H I L O S O P H Y in T H E F A C U L T Y O F G R A D U A T E STUDIES Department of Geography We accept this thesis as conforming to the required standard T H E U N I V E R S I T Y O F BRITISH C O L U M B I A May 2001 © Steffanie Scott, 2001 In presenting t h i s thesis i n p a r t i a l f u l f i l m e n t of the requirements for an advanced degree at the U n i v e r s i t y of B r i t i s h Columbia, I agree that the Library s h a l l make i t f r e e l y a v a i l a b l e for reference and study. I further agree that permission for extensive copying of t h i s thesis for s c h o l a r l y purposes may be granted by the head of my department or by h i s or her representatives. I t i s understood that copying or p u b l i c a t i o n of t h i s thesis for f i n a n c i a l gain s h a l l not be allowed without my written permission. Department of The U n i v e r s i t y of B r i t i s h Columbia Vancouver, Canada Abstract This dissertation addresses the multifaceted process of decolectivization in Vietnam—the shift from colective to household production and the alocation to households of long-term land-use leases. The fieldwork-based study aims to outline the institutional changes within this process and assess their implications for livelihood vulnerabilty, particularly in terms of ethnic and gender diferences. Two case studies from Thai Nguyen province in the northern midlands of Vietnam highlight the diverse outcomes of and responses to decolectivization. The reconfiguration of property rights created competion over access to resources, with land conflicts over inheritance emerging at the intra-family level and conflicts over ancestral lands at the inter-household and inter-ethnic level. There are six broad conclusions that can be drawn from this analysis. First, interpreting decolectivization as institutional restructuring emphasizes the multiple and interelated dimensions of changes underway—in property rights, the organization of production, scales of decision making, discourses of development, new stakeholders, and various forms of informal institutions. Second, the analysis points to frequent gaps between national policy and on-the-ground practice and to the need for greater atention to complexity in social processes. Third, in reestablishing the household as principal production unit, decolectivization and property rights restructuring in Vietnam have afected mariage and inheritance trends and, in turn, household and kinship relations. Fourth, these processes of institutional change can be linked to new paterns of access to land and related resources, thereby shaping new paterns of vulnerabilty. Fifth, these paterns of vulnerabilty are mediated in part by formal institutions, exemplifed by the loss of some support services formerly provided to farmers by agricultural colectives. And lastly, informal social instiutions are a further factor mediating new paterns of livelihood vulnerabilty. Social networks operate diferentialy and can lead to discrimination for some women and ethnic groups. i i i Table of Contents Abstract ii Table of Contents iii List of Plates vi List of Figures • vii List of Tables viii List of Maps x Acknowledgements xi List of Acronyms xiii Glossary and Units of Measurement xiv Chapter One. Positioning the Author: Experience and Method 1 Paths to Vietnam 2 Siestas, World Cups, and Other Contingencies 4 Making Sense of Statistical Data 6 Making Inroads and Unearthing Data 8 Formalities of Local Fieldwork 13 The Commodification of Research: Expectations and Incentives 16 The 'Typical' Case: Resistance to Acknowledging Process and Difference ...17 My Interpreter and I: Interpreter-Researcher Relations 22 Synthesizing Complexity: Research Approaches and Data Analysis 25 Chapter Two. Study Aims and Research Design 28 Aims of the Study 30 Organization of the Dissertation 31 An Overview of Methods 35 The Place of the Local: A Multi-level Analysis 36 Studying Up, Down, and Around: Multiple Methods and Data Sources 37 Site and Case Study Selection 41 Chapter Three. Conceptualizing Institutions, Property Relations, and Vulnerability 45 Why Instiutions? 45 Linking Instiutions to Livelihood Vulnerabilty 48 iv From Vulnerability to Social Capital 52 Ethnicity, Social Capital, Land and Identity 57 Gendering Social Capital and Collective Action 59 Gender, Property, and Land Rights 62 Post-socialist Agricultural and Property Rights Restructuring 70 The Doi Moi Reforms in Vietnam 79 New Discourses ofthe Market Economy and Development 83 Chapter Four. Land, Institutional Change, Rural Governance, and Development in Vietnam 88 Land Relations and Instiutions in Vietnam 88 Symbiosis of Household and Collective Sectors 94 The Seeds of Decollectivization 98 Exercising New Land Rights 104 Landholdings and Differentiation 106 State Capacity for Implementation and Enforcement 109 Reconfiguring Colective and Cooperative Instiutions 114 Disintegration of Mechanisms for Cooperation 120 Shifting Systems of Welfare and Poverty Reduction 127 Governance and Rural Finance: Taxation and Rural Credit 132 The Reinforcing of Micro-level Social Capital 135 Social Networks and Access to Information 140 From New Instiutions to New Identities: Perceptions of Land and Property 142 Chapter Five. Gender, Land, and Rural Livelihoods in Vietnam 147 Vietnamese Gender Relations 148 Vietnamese Family and Kinship Relations 152 Socialism and Gender Equality in Vietnam 156 Reworking the Family-Household through Decolectivization 159 Collectivization 163 Phases of Decollectivization 165 State-issued Land Allocation 169 Market Land Transfers 170 Inheritance 170 Diversity among Female-Headed Households in Vietnam 173 Vulnerabilty, Household Headship, and Entilements to Land: A Case Study 179 A. Female heads with resident husband 181 B. Female heads with husband absent 182 C. Female heads abandoned by husband 183 D. Widows, divorced or separated female heads 184 E. Unmarried female heads 184 F. Special cases of vulnerable male-headed households 185 Implications and Interpretations 186 Ambivalent Implications of the 'Reworking of the Household' 188 Chapter Six. Livelihood and Ethnicity in Upland Vietnam and Thai Nguyen Province 190 The Making of Vulnerable Livelihoods in Upland Vietnam 191 Characterization of a Diverse Region: Thai Nguyen Province 202 The Making of a Frontier Territory: Histories of (Resettlement, Ethnicity and Land Relations 209 Land Relations and Collectivization of Agricultural Production 215 Land Use, Land Allocation, and Rural Landholdings 221 Deforestation and Agro-forestry Programs 227 Chapter Seven. Reconstructing Ethnic Identities and Property Relations through Decollectivization 232 From Policies to Practices: Local Autonomy and Adaptations of National Policy 233 What was Colectivization, de jure and de facto? 236 Alocation and Ancestral Land Claims: A Case Study 238 Local Responses 243 Implications and Interpretations 248 Chapter Eight. What Comes Next? Defining the New Rules of the Game 250 References 260 a. Statistical Data on Thai Nguyen Province 260 b. Statistical Data on Vietnam •. 260 c. Newspaper, Magazine, and Journal Articles Published in Vietnam 261 d. Reports and Books by Vietnamese Agencies or Researchers Published in Vietnam 261 e. Reports by International Agencies Published in Vietnam 262 f. Films 263 g. Other Sources 263 Appendix: Statistical Narrative on Thai Nguyen Province 274 Rewritng the Map: Administrative Shufling 274 Demographics, Sex Ratios, Urbanization, and Ethnicity 276 Colectivized Agricultural Production and State Agricultural Enterprises 279 Agricultural Production Trends and Use of Forest Products 280 Development Indicators and Rural-Urban and Regional Disparities 286 Education, Health, and Recreation 287 GDP, Economic Growth, and Employment by Sector 289 vi List of Plates Plate 1. Rounding Up the Wagons on an Excursion with Friends 3 Plate 2. Interview in a Multi-generational Household 25 Plate 3. Discussing Local Legends over a Cup of Tea 25 Plate 4. Women with Shoulder-poles Carying Goods for Sale 83 Plate 5. Mechanization and Large-scale Agriculture as Shown on a 200 Dong Bill 93 Plate 6. Fences as Markers of Private Property 144 Plate 7. Postcard: 'Being Ready for the Batle to Protect the Revolutionary Base' (1966) 150 Plate 8. Protecting Herself from the Sun, An Urban Woman Sports a Motorbike 150 Plate 9. A War Invalid Leading a Literacy Campaign among Ethnic Minorities (painting from National Art Museum) 196 Plate 10. Site of a Resetled Hmong Vilage in Thai Nguyen Province 200 Plate 11. Postcard Image of 'Uncle Ho' with Female Labourers on an Agricultural Colective in Dai Tu District, Thai Nguyen Province 218 Plate 12. Evidence of a Growing Market Economy: A New Rural Store 231 Plate 13. Deforested Slopes Cleared by Dao People Who Were Not Alocated Paddy Land .245 Plate 14. Women Bringing Crops by Bicycle to a Recently-established Commune Market 281 Plate 15. Hmong Women Seling Gourds in a Nung Vilage Market 281 Plate 16. New Tea-drying Technology in Thai Nguyen Province 283 List of Figures Figure 1. Institutional Development for Economic Transiton 47 Figure 2. The Articulation of the Three Modes of Production. 71 Figure 3. Evolutionary Theory of Land Rights 78 Figure 4. Distribution of Total Land and Irrigated Annual Cropland by Region in Vietnam, 1992-93 108 Figure 5. Existing Distribution Structure: Channels of Non-Wood Forest Products from Cho Don District to Provincial and Hanoi Markets 230 V l l l List of Tables Table 1. Number of Interviews Classified by Type of Interviewee 40 Table 2. Components of a Rural Livelihood 49 Table 3. Asset Vulnerabilty Matrix 53 Table 4. Spatial Dimensions and Levels of Mediation for Analysis of Vulnerabilty 57 Table 5. A Trade-of Matrix of Alternative Policy Stances for Addressing Underdevelopment 72 Table 6. Results of Agricultural Decolectivization and Distribution of Farm Land in Central and Eastern Europe and Former Soviet Republics 74 Table 7. Average Farm Size by Organizational Structure in Central and Eastern Europe and Selected Former Soviet Republics 75 Table 8. New Trajectories of Development through doi moi 82 Table 9. Stages and Scales of Colectivization in Vietnam 92 Table 10. Chronology of Land Policy Reforms for Decolectivization 100 Table 11. Average area of land per capita of the Poor and Nonpoor in Rural North and South Vietnam, 1992-93 (m2) 107 Table 12. Average area of land per capita in Rural Vietnam by Region, 1992-93 (m2) 107 Table 13. Agencies Dealing with Land Alocation and Upland Development 113 Table 14. Reconfiguration of Household Status and Functions through Property Rights Reforms in Agricultural Production 126 Table 15. Production-related Functions of Various Instiutions in the Periods of Colectivization and Decolectivization 127 Table 16. Ambiguous Discourses: Interpretations of 'Household Reworking' for Intra-household and Gender Relations 162 Table 17. Demographic Statistics of Types of Female-Headed Households in Vietnam 175 Table 18. Proportion of Female-Headed Households Disaggregated by Age of Female Head (percent) 175 Table 19. Characteristics of Female-Headed and Uxorilocal Households Surveyed 182 Table 20. Poverty Rate by Region in 1993 and 1998, and Gini Coeficient 194 Table 21. Poverty Characteristics by Region in Vietnam, 1992-93 194 Table 22. Widening Rural-Urban Income Gap in Vietnam: Rural GDP per capita as a Percent of Urban GDP per capita, 1990-94 195 Table 23. Population Density, by District, Thai Nguyen province, 1955-present 208 Table 24. Population and Ethnicity by District, Thai Nguyen province, 1989 210 Table 25. Population of Ethnic Groups across Districts, Thai Nguyen province, 1989 211 IX Table 26. Land Tenure in Thai Nguyen province, by District, c. 1930 216 Table 27. Agricultural Colectives, by Number of Households, Bac Thai province, 1975-1980 219 Table 28. Agricultural Colectives, by Land Area, Bac Thai province, 1975-1980 219 Table 29. Number of Colectivized Households (ho tap te), Thai Nguyen province, 1955-present 220 Table 30. Area of Land and Alocated Land, by Land Type, Thai Nguyen Province, 1996 (hectares) 225 Table 31. Number of Agricultural Households by Size of Agricultural Landholdings, by District, Thai Nguyen province, 1994 227 Table 32. Greater Freedom or Responsibilty? Consequences of Economic and Property Rights Restructuring 257 Table 33. Number of Communes per District, Thai Nguyen province, 1955-present 275 Table 34. Area of Districts, Thai Nguyen province, 1955-present 276 Table 35. Population of Females and Males in Urban and Rural Areas, Selected Years 278 Table 36. Assets of Rural and Urban Households Compared, Thai Nguyen province, 1995.287 List of Maps Map 1. Vietnam, Showing Thai Nguyen Province 42 Map 2. Map of Field Sites in Thai Nguyen Province 44 Map 3. Map of Gross Domestic Product Per Capita by Province, 1993 193 Map 4. Administrative Map of Northern Vietnam, with Inset of Districts of Thai Nguyen Province 203 Map 5. Map of Bac Thai Province (1960s-1997) and the Later Split into Thai Nguyen and Bac Kan Provinces 204 Map 6. Topographical Map of the Area Surounding Thai Nguyen Province ..207 Map 7. Map of Population Density by District, Thai Nguyen Province, 1996 208 Map 8. Distribution of Population by Ethnic Group, by district, Thai Nguyen province, 1989 (three classes equal interval) 212 Map 9. Map of Forest Land as a Percent of Total District Land Area, by District, Thai Nguyen Province 223 xi Acknowledgements Throughout the preparation of this dissertation, I have become increasingly appreciative of the many forms of support ofered to me by my supervisor, Tery McGee, and my commitee members, Trevor Barnes, Geof Hainsworth, Pat Howard, and Maureen Reed. Catherine Grifiths did great work with her creative cartographic skils. Maureen Reed, Christine Veileux, Van Nguyen-Marshal, Maija Heimo, and Dawn Curie ofered hard-to-find feedback on various papers. I would also like to acknowledge Stan Barret, a former mentor, whose fascinating qualitative methods course gave me a greater sense of the craft of ethnography. Although my interviewees go unnamed in the dissertation, I thank them for their openness and wilngness to make time for me in their busy schedules. For helpful commentaries as wel as logistical assistance, I thank Ha Huy Thanh, Vu Tuan Anh, and Tran Thi Que at the Instiute of Economics, and Pham Xuan Nam and Dang Anh Phuong of the National Centre for Social Sciences and Humanites. I acknowledge Ms. Dung (Director of the Department of Agriculture in Dong Hy district, Thai Nguyen province) and Nguyen Khanh Quae (former Rector of the Agro-Forestry Colege of Thai Nguyen University) for helping with the necessary authorizations to facilitate my fieldwork. Mark Hawkes of the Association for Research and Environmental Aid and Pascal Bergeret, Piere Bal, and Marie Melac of the Programme Fleuve Rouge / Groupe de Recherche et d'Echanges Technologiques provided invaluable insights into local dynamics in the Thai Nguyen region. I extend thanks to my interpreters for their wilngness to endure motorcycle rides and their patient eforts to explain and make sense of what might have seemed strange and complicated questions and responses during interviews. My good friends in Hanoi and the students I met in Thai Nguyen brought great pleasure and insights to my fieldwork experience. Thanks also to Helen Booth for the stocks of Orangina she brought up periodicaly from Hanoi and for the use of her kitchen at the Guesthouse in Thai Nguyen—a great respite. Thanks to al my Vietnamese and Canadian coleagues in the 'Localized Poverty Reduction in Vietnam' program for the opportunity to atempt to put some of my ideas into practice after I completed my fieldwork. I want to particularly acknowledge Do Thi Binh, Tran Thi Van Anh, Dang Bich Thuy, Nguyen Van Minh, Nguyen Hoang Trang, and Leonora Angeles for their stimulating discussions. xii I would like to extend an apology to friends, family, and coleagues for my frequent silences, a consequence of my self-imposed pressure to move along in the long and sometimes lonely process of writing-up. Thanks to my e-family for their tolerance and for keeping me entertained and connected during the long stretches between visits. Thanks to Tod for providing statistical advice in a moment of crisis. Sobre todo, agredezco a Francisco por su fe en mis capacidades y su buena energia. The International Development Research Centre (Young Canadian Researchers' Award), the Asia-Pacific Foundation of Canada (Travel Grant), the Ford Foundation's Northwest Regional Consortium for Southeast Asian Studies (Graduate Students' Field Research Grant), and the Centre for Southeast Asia Research at the University of British Columbia (Hampton Grant) generously provided funding for two research trips. Despite the many people who contributed in various way to the construction of this dissertation, I alone take responsibilty for the contents and any erors within it. X l l l List of Acronyms ADB Asian Development Bank FAO Food and Agriculture Organization GDLA General Department of Land Administration HEPR Hunger Eradication and Poverty Reduction program IPM Integrated Pest Management MARD, DARD Ministry of Agriculture and Rural Development (national), Department of Agriculture and Rural Development (provincial) MOLISA Ministry of Labour, Invalids and Social Afairs NGO Non-governmental organization NTE New Instiutional Economics PRA Participatory Rural Appraisal SFE State forest enterprise UNDP United Nations Development Program VAC Vuon (vegetable garden), Ao (fishpond) and Chuong (pig sty): a model of integrated production and nutrient recycling xiv Glossary and Units of Measurement Collective and cooperative: In this dissertation, I reserve the term cooperative to refer to 'new service cooperatives' that have been newly created or formed from converted colectives after land for agricultural production became alocated to households in the late 1980s. I use the term colectives to refer to the organizational entiies established through successive stages from the early 1960s to the late 1980s. This distinction is important since, for the International Labour Organization, a cooperative must have voluntary not compulsory membership and must be autonomous from state control. These are two characteristics of the new service cooperatives but not the former agricultural production colectives in Vietnam. Having made this distinction, I realize that in common usage in English and Vietnamese the terms colective and cooperative are considered largely synonymous. Doimoi: Literaly meaning 'new change' or 'renovation,' doi moi refers broadly to the open door policy and market economy orientation adopted in 1986 at the Sixth Congress of the Communist Party. Kinh (or Viet): The majority ethnic group in Vietnam, comprising approximately 86 percent of the population. Level of administration: There are four levels (cap) of administration in Vietnam: the central level (cap trung uong), the province (tinh), the district (huyen), and the rural commune or administrative vilage (xa). Other distinctions include the urban commune or ward (phuong) and the commune town (thi tran). The later is usualy the capital of a district. What are termed communes in Vietnam should be distinguished from people's communes in China. The later are more akin to what I cal production colectives. Patrilocal, matrilocal, uxorilocal, and neolocal residence: Patrilocal (or virilocal) residence refers to the patern of residence after mariage in which a wife moves to live in her husbands'  (father's) natal home or vilage. This is common among Kinh people (particularly for the first few years after mariage) and many other ethnic groups in Vietnam. In uxorilocal residence, a woman's husband comes to live with her parents or in her native vilage upon mariage. I prefer the term uxorilocal over matrilocal (residence folowing the mother's line) for my analysis since matrilocality is not a cultural trait of the Kinh people and is generaly only associated with various matrilneal ethnic groups of the Toy Nguyen (Central Highlands) region. By contrast, in Vietnamese, cases of uxorilocal residence are refered to (from the husband's perspective) simply as o re, meaning 'at the in-laws.' A final form of residence patern is neolocal, in which the couple resides in a new area distinct from the hometown of either husband or wife. A further distinction in residence paterns relates to endogamous vs. exogenous mariage (i.e., mariage between people of the same or diferent vilages. This is discussed on page 155. Units of measurement: One sao in most parts of northern Vietnam is equal to 360 m . One mau is equivalent to ten sao or 3,600 m2 (0.9 acres). Currency conversion rate: In 1998 the conversion rate for US$ 1 was approximately 14,000 dong. 1 Chapter One. Positioning the Author: Experience and Method Some dissertations or books open with a series of vignetes for readers to gain a flavour of the social and geographical seting of the subsequent research study. I instead ofer this chapter that situates me as author in the intelectual and fieldwork seting of the study. My rationale for including this chapter is to provide readers with a context to unearth who was behind the analysis and what was involved in carying out ethnographic fieldwork in Vietnam and in puling the analysis in the remainder of the dissertation together. Conducting the field research for this dissertation posed numerous chalenges, the discussion of which at times I felt could easily constiute a thesis in it own right. It is my contention that these processes and experiences are as valid and significant a research subject as the dissertation topic of decolectivization itself. Too frequently, the writing process smoothes over and eliminates the many ambiguites of field research and field notes. The task of the researcher, many would argue, is to present a 'seamless,' convincing and thoroughly researched study that engenders confidence in the results. Yet, fieldwork is by nature a rather chaotic process. There are many gaps between the clean and somewhat mechanical procedures laid out in methods textbooks and the 'messiness' of fieldwork practice. Very few publications examine fieldwork practice in Vietnam (but see Ambler 1998 and Christoplos 1995 for some exceptions). Given the fast pace of change in Vietnam, conditons and procedures for carying out fieldwork are bound to experience similar shifts. Thus, the documenting of fieldwork processes portrayed here might serve as a basis for comparison of changing 'fieldwork possibilties' in the country, including issues of entry, access and commodifcation, as Kurti (1999) ofers for Romania and Hungary (see also De Soto and Dudwick 2000). Through an examination of various logistical and methodological chalenges, the discussion in this chapter reflects on issues of positonality—of myself and others. I open this expositon of my fieldwork odyssey by situating myself as researcher on a longer intelectual journey of experiences and interests that led me to Vietnam in the first place. I then recount some of my fieldwork encounters as a foreign researcher negotiating Vietnamese language, geography, administration1, and much more. I touch upon issues of data access, data quality, 1 Later in this chapter I describe some characteristics of institutional culture that posed challenges to the realization of my research objectives. However, in making these assertions I want to underline that Canadian bureaucracies could be equally, if not more, impenetrable—:as colleagues working for Canadian organizations in Vietnam would 2 research strategies, the commodifcation of research, the chalenges of moving beyond 'the typical case' in discussions with officials, interpreter-researcher relations, and the complexites of interpretation and analysis. Paths to Vietnam Many paths have led me to this research in Vietnam. In high school, I developed an interest in global inequalites and development. At the age of 18,1 spent a year in Thailand studying language and culture, during which time I became drawn to understanding the roots of uneven development, particularly in the Isan, Thailand's rural northeast region. Upon entering university, I folowed programs in Chinese Studies, Latin American Studies, and Geography, which most closely matched my interests in development studies. I took courses on development theory and joined field schools to China and Nicaragua, which alowed me the opportunites to explore recent experiences in land reform and socialist rural development policies. I also pursued courses on Mexico's rural restructuring during a time of landmark shifts in land reform, as the communal ejido lands began to be privatized under the neoliberal reforms of then-President Salinas. During my master's degree program, I spent six months in an indigenous area of Mexico, turning my atention to the impacts of macro policies of structural adjustment on localized poverty reduction eforts. In the face of cutbacks in state agricultural extension services and paralel economic and property rights reforms, non-governmental organizations (NGOs), such as the one I studied, were emerging to promote techniques for sustainable agriculture and marketing of organic produce. I analyzed how, in a context of ethnic, cultural, and class diferences, the NGO project staf struggled to merge scientific and local knowledge and sensitivity to gender issues. Through an analysis of project dynamics, I presented a critique of participatory development discourses showing how, despite its apparently inclusive participatory discourse, power relations and uneven flows of information were implicit within the project's structure. I further included a gender analysis of women's diferential access to information and broader entilements in the project. Arriving at the University of British Columbia (UBC) for my PhD, I found a group of faculty—including my advisor, Tery McGee—engaged in ongoing colaboration with Vietnamese scholars at the National Centre for Social Sciences and Humanites (NCSSH) in Hanoi. Turning my atention to land issues in a country in which close to 80 percent of the occasionally remind me. 3 population is based in rural areas, I made an initial trip to Vietnam in 1997. Led by the inexhaustible Geoff Hainsworth, then-Director of the Centre for Southeast Asia Research, I joined a group of U B C grad students in a study tour on rural resource management and sustainable livelihoods. I was struck by the parallels between rural institutional restructuring underway in Vietnam and what I had observed in Mexico. Intrigued by the diversity and swift pace of change in the country, I returned the following year for 12 months. I set out to make sense of the multifaceted process of decollectivization—the shifting of property rights from collective production to individual production—and what it implied for new patterns of vulnerability and social differentiation, particularly in terms of ethnicity and gender. M y research was greatly facilitated by having mastered an intermediate level of communication skills in Vietnamese—I had studied the language in Vancouver since 1996 and I later took one month of intensive language training at the Institute of Linguistics in Hanoi. My language abilities continued to improve as I shared conversations, green tea, and motorcycle rides, and occasional northern Vietnamese delicacies such as eel soup and spicy snails, with Vietnamese friends. Plate 1. Rounding Up the Wagons on an Excursion with Friends The field research for this dissertation was carried out through institutional affiliations with the Centre for Southeast Asia Research (University of British Columbia), the Institute of Economics (part of the National Centre for Social Sciences and Humanities, Hanoi), and the Agro-Forestry College at the University of Thai Nguyen, 80 km north of Hanoi, where I stayed on my return visit in 1998. From there I made frequent trips to rural villages and government offices to interview male and female farmers and local officials about their perceptions of the 4 reforms which were reorganizing rural life. Unlike parts of the Red River Delta near Hanoi, the highlands, in general, and Thai Nguyen province, in particular, might be considered as 'neglected geographies,' having been the subject of little detailed social research in Vietnamese, French, or English. Paralel to my analysis of local people's strategies and use of networks to achieve their objectives, the shaping of the research project can be traced to networks among researchers themselves to gain access to certain people and information. In this way, the value of social networks or institutional connections to facilitate this research project should not be overlooked. In saying this, I turn back on myself some of the same conceptualizations of social capital that I use on my research subjects. Despite the chalenges detailed below, I was a researcher with relatively privileged institutional afilations that provided me with leters of introduction and aranged for important interviews and access to other channels of information. Without such social capital—defined simply as the norms and networks that enable people to act colectively (Woolcock and Narayan 2000)—it would have been even more chalenging to obtain authorizations and accomplish what I did during my fieldwork. Siestas, World Cups, and Other Contingencies Before detailng some of the methodological chalenges encountered in my fieldwork, I wil outline a number of the logistical constraints. First was the poor means of communication in Vietnam, even within the capital city, Hanoi. I estimate that fuly 70 percent or more of the messages that were sent to my first email account in Vietnam never reached me. Similarly, more than half of the posted mail sent to me never arrived. Phone messages in hotels were sometimes not passed on, and faxes could be equaly unreliable. Such circumstances forced me to accept a degree of isolation and the impossibilty of dependable international, not to mention domestic, communication. Poor transportation infrastructure was a further impediment to field research. There was 'relatively' good bus service to and from a few major tourist destinations around the country, but to travel anywhere of the tourist route, even 90 km north of Hanoi to Thai Nguyen province, the options were limited: a rambling four-hour train ride, once daily; a slow, crowded bus ride renowned for pickpockets armed with razor blades; or a risky 2-hour ride on a rented motorbike. Although rented motorbikes and motorbike-taxis are the most rapid and relatively cheap form of travel in Vietnam, countless Vietnamese (and foreigners) have lost their lives traveling this way—or have come perilously close to doing so. I usualy opted for the train, but I 5 was later refered to a driver in Thai Nguyen whom I occasionaly hired to make the trip by private car. It sometimes seemed to take a miracle to coordinate al the logistics for a given field trip: the motorcycle taxi or hired car, the interpreter, the authorized appointment, and conducive weather. Beyond communication and transportation, a third logistical chalenge was the element of contingency—the France 1998 World Cup soccer tournament being a classic example. This event seemed to have captivated the nation. Children knew by heart the official theme song, Ricky Martin's "La Copa de la Vida/The Cup of Life" and would often chant its chorus line, "Ole, ole, ole...!" The month-long event usualy entailed two games scheduled in the middle of each night. Throughout the month, it was hard to find people motivated to perform their job or atend to others' requests. One of my interpreters as wel as the driver I occasionaly hired both suggested that I find alternatives as soccer would be their priority that month. A related chalenge for long-term work in Vietnam is the Tet (lunar new year) holiday. Officialy lasting only for a few days, in practice it meant that ofices and other facilities could be shut for two weeks or even a month. As for more everyday logistics, nearly al ofices, banks, research centres, libraries, museums, and publishers closed for a two- to three-hour lunch break while workers took a siesta. This posed dificulties for coordinating visits to several ofices during one morning or afternoon. And during day-long field visits in rural areas, it was sometimes hard for urbanite interpreters to remain awake after lunch. A fourth constraint was the poor conditon of documents—disintegrating, dusty, and insect-ridden—in many public institutions, including national libraries, universities, and policy institutes. Minimal maintenance and high humidity threaten these colections, including some materials of great historical significance. Investment and operating costs are lacking for improving the conservation of such research materials. Beyond these factors, living in Vietnam, one learns to befriend insects and adapt to periodic losses of water or electricity, frequent floods, 39°C heat, and 100 percent humidity— the later producing mildew in clothing, papers, and sometimes computers. But thanks to many geckos, mosquitoes were removed from my room without the aid of insecticides (too bad they would not do the same thing for the giant cockroaches). At moments when I tired of Vietnamese television channels for entertainment, there were sometimes more exciting spectacles to be found elsewhere, such as when a giant spider was strategicaly transported up my kitchen wal in Thai Nguyen by an ambitous team of ants. One requirement for adapting to life in Hanoi or 6 Ho Chi Minh City includes developing the skil  to negotiate crowded streets and sidewalks. Although I have heard that improvements have since been made, during the period of my fieldwork trafic regulations seemed to be rarely enforced. To cross a street one sometimes does beter closing one's eyes and proceeding at a steady pace; the drivers of passing motorcycles are amazingly adept in calculating their movements to avoid pedestrians. On an interpersonal level, in any public place I continualy faced a succession of personal questions from curious passers-by. This included being repeatedly asked when I planned to have children. My partner and I were even directed to visit a temple that is frequented by couples who go to pray for fertility. In other setings, I came to expect the frequent invitations—even by strangers who approached me in a restaurant—to empty a cup of rice wine ("cham phan cham!" or "100 percent" was the local equivalent of "botom's up"). Other tests for fieldworkers included becoming accustomed to hair or flies in their food and 'toilets' that were nothing more than a floor from which liquid drained into a hole in the corner of a smal room. But compensations for such hardships were also abundant, including indulgences in Vinamilk ice-cream and yogurt, fresh fruit and bread rols, and tasty Vietnamese dishes, not to mention many memorable friendships. Making Sense of Statistical Data Concerns regarding data quality in Vietnamese official statistics are many. Indeed, any findings drawn from official data in Vietnam need to be questioned for their validity. Data were frequently inaccurate, inconsistent, unavailable, inadequate, or misleading. Possible explanations for discrepancies in the data are many. In additon to measurement and recording errors, the data may have been measured at diferent points during the year or before or after administrative boundary changes had been introduced. In most cases, statistical yearbooks for Thai Nguyen province only covered periods of one, three, or five years. This made longitudinal analyses awkward due to the need to consult so many books to trace changes in a single variable over time. Moreover, yearbooks were not available for the pre-1955, 1966-1972, and 1983-1985 periods. Some statistical yearbooks spanned five-year periods but the dates for individual tables of data were not specified. In using old yearbooks, one faces the further complication of ilegible statistics: the paper turns brown and the ink is often too faded and patchy to be visible, particularly on a photocopy. Regarding problems of data analysis, Beresford and McFarlane (1995: 53) observed that 7 The study of Vietnamese regional problems is rendered especialy difficult by the paucity of data. In contrast to the situation in China, where provincial statistical yearbooks are regularly published, provincial level time series are available only in the case of some agricultural production statistics for Vietnam. The rest are fragments. At times the statistical data that one is able to piece together is so disjointed and incomparable across time periods or regions as to make them virtualy useless. Analysis was further complicated by the continual shifts in administrative boundaries (as outlined in the Appendix), rendering consistent comparisons over time for the same province, district, or even commune2, impossible. Two maps for adjacent areas were sometimes incompatible, with contours marked at diferent intervals or with diferent soil classification systems used. Discussion later in this chapter outlines problems of data projections being substiuted for real data and of commune-level data being thrown out once they had been aggregated. In such circumstances, the data often raise more questions than they answer. I aim to make these problems of interpretation transparent to the reader, rather than simply omiting inconsistencies that appear to contradict overal trends. As mentioned in Chapter Six, in instances of discrepancies between sources, or within the same source, both or al were usualy listed for readers to appreciate the complexites and shortcomings of the data and to draw their own conclusions. Given the plethora of problems outlined above, I draw conclusions rather tentatively. I hope that even if the absolute figures are somewhat inaccurate, the figures can be used to deduce broad trends and compare relative orders of magnitude. When I speculated that the accuracy of statistical data and techniques for colection must be improving in recent years in Vietnam, one researcher from the National Centre for Social Sciences and Humanites suggested to me that the opposite could equaly be true: state salaries were relatively higher in the past and local cadres who are responsible for record-keeping are now more motivated to seek supplementary employment, so they may be less conscientious in carying out tasks within their state-sector positons. Moreover, with the increased autonomy of households in the boom of the household economy, and the decline in influence of colectives , it is much harder to keep tabs on the demographics, economic activities, and particularly income data of households. This points to the need to revise approaches to data colection that are oriented to a centraly planned rather than an emerging market economy.4 2 As explained in the glossary, a commune is a level of administration comprising one or more villages. 3 See the glossary for a distinction between collectives and cooperatives. 4 The Vietnam Living Standards Survey, with technical assistance from the World Bank, fills some of these gaps. 8 Making Inroads and Unearthing Data One day early in my fieldwork, I aranged (via the Agro-Forestry Colege) a 2 pm appointment to visit the Land Administration Ofice in one district of Thai Nguyen. I arived in my hired car with an interpreter who worked in the Colege's International Cooperation Ofice. The hour came and went and we later heard, after making some phone cals, that the oficer had a sudden meeting and would be busy al afternoon and that there were no other representatives available in his ofice to meet us. Another staf member in the ofice suggested that we talk with a forestry extensionist who could perhaps answer some of our questions. Yet, after a few minutes of discussion, the extensionist explained that he was busy and had to leave and suggested we visit another ofice nearby, the Forest Protection Division. After two phone cals, he reported that no one was available there. I asked if it was possible to visit a Land Administration representative from one of the nearby communes or just make some household visits but I was told that, no, I needed to have authorization from the People's Commitee, and al the members were at a meeting out of the district for the afternoon. I asked if they had any data available on land statistics for the district. Alas, for that I should have specifed what specific data I required in my leter of introduction from the Colege, and sent it with one week's advance notice. And no, they did not have any maps. Giving up hope for accomplishing any more in the district that day, I decided to make a trip to the provincial library in Thai Nguyen City. In the first room, I asked the receptionist where I could find materials on economic and historical issues about Thai Nguyen. She said she had nothing there, but I could ask upstairs. I asked if there were any maps available. She replied that they might have been misplaced after doing renovations. Upstairs (where the magazines were located), I was told they only had information (on agriculture and other issues) for the country as a whole but nothing for Thai Nguyen province. She suggested the nearby provincial museum might have some information. I found a few other cards in the catalogue for books from 1990 about colectives and economic development, but on the advice of the young librarian, I decided to check instead in the museum (behind the library), since she indicated that they had books and maps there. When I asked about maps, the librarian checked the card catalogue and found one card indicating a provincial map dated 1968 but warned that it perhaps no longer existed. After explaining my objectives to two other women in the main library ofice, we made our way to the ofice of the vice-director of the museum. The vice-director said he could not help me because he did not have information about Thai Nguyen but only about Bac Thai (the 9 former name of the province)! Finaly, we were taken to the exhibiton rooms (which were unlocked especialy for us), and two young guides undertook to tel us about the archaeology of the region. When I asked about the history of land tenure, they replied that research was curently being conducted on this but the results could not yet be released. In one of the last exhibiton rooms, there was a cabinet displaying several books on the history of the region. I asked where I might locate copies to read, and the guides suggested going to the library. I asked if any copies were for sale: no. When asked about where to locate information on vilage-level land tenure in the pre-revolutionary period, they suggested the National Library, or the Social Science Library in Hanoi—although in Hanoi I had been directed to inquire in Thai Nguyen. On another occasion, I asked my interpreter, a lecturer from the Agro-Forestry Colege, if there weren't any economists at the Colege researching the economic history of the province, but she said she did not know. I had previously asked the Director of the Faculty of Forestry, a 20-year veteran of the Agro-Forestry Colege, if he knew of any studies on the economic history of Thai Nguyen province, but he did not. I had earlier visited the 'Scientific Library' of the Agro-Forestry Colege of Thai Nguyen University and inquired about maps of Thai Nguyen province. I was told they had none. Since it is not possible to browse in Vietnamese libraries, I requested any information they had on Thai Nguyen's history, economy, or agricultural production and was told that they only had general information but not about Thai Nguyen province. I asked about magazine articles and was given the same response. I returned to the provincial library two days later to check the Thai Nguyen province section of the periodical card catalogue for myself. I selected a few articles but was told that they were from old newspapers that were no longer available. I identified a card for a 1970 map but was told the map no longer existed—nor did they have any other maps. I requested one introductory book on Bac Thai province that I'd seen in the museum display case, which was located for me. It turned out to only contain photos glorifying the province's industrial development, making only a single-sentence reference to ethnic groups. I asked the librarian if they had any other information on the population of ethnic minorites in Thai Nguyen. She disappeared and finaly returned with a dusty and disintegrating 1976 book introducing ethnic groups for Vietnam as a whole but with no maps or statistics disaggregated by province. Nor did they have any other materials on agriculture, economy, history, or about a famous landlord in Thai Nguyen, Madame Nguyen Thi Nam. The same week, I returned to Hanoi and went to the National Archives, arriving at 3:30 pm. The gatekeeper asked if I could come back the next day, as the ofices would be closing at 10 four. I returned the next morning and after presenting my leter of introduction from the Instiute of Economics, my backpack was sequestered and I was alowed into the main ofice. There I was told that the room and building that stored the information I was looking for was under renovation and would not be accessible until the folowing year!51 asked if there wasn't any other material that could be consulted this year, and the librarian said yes, but only general information, not relating to Thai Nguyen province. I said fine, so was then asked to fill in two copies of a 'Request to Work in the Archives' form, to be completed in French. During the previous conversation the librarian had been addressing me in French, while I responded in Vietnamese. When I proceeded to write the date as '98/04/28' I was informed that I had to write it in reverse order, '28/04/98,' and was given a fresh form. When I wrote the number one as a single stick-line, I was told that I had to write it the French style, with a hook and base. When I wrote 'PhD' as the degree I was pursuing, I was told I had to write 'these du doctorat.' After I at last completed the form in satisfactory fashion, I was told I could return in one week to receive the authorization to consult archival materials. On my third visit, however, I was told once again that I'd have to come back next year. Maps seemed especialy difficult to come by in Vietnam. Perhaps this is atributable to the political sensitivity around the use of maps, or perhaps to the absence of regional studies specializations and general shortage of resources for creating and disseminating maps. In response to my request for a commune map, in one instance, a commune-level Land Administration oficer sent me to the district-level Land Administration ofice, where I was in turn directed to seek out the provincial maps ofice, which I was never able to track down. A doctoral student I met had used tracing paper to copy maps in the archives, but after this was detected most documents she was given had the maps removed from them. She was later permited only to copy certain sections of maps coresponding to the one district in which her fieldwork was concentrated. In the district itself, she spent two afternoons sketching a map from the wal in the People's Commitee meeting room since, in most cases, these were the only accessible district-level maps. When I asked one of my interpreters about maps of Thai Nguyen, suggesting that one be posted in the Colege's Ofice of International Cooperation, she claimed that "we don't need a map because we have it in our heads." She had previously lent me a copy of her master's thesis—from a foreign university—that contained a map of a district adjacent to Thai Nguyen province where she had done her fieldwork. When I asked where she acquired the 5 In addition, no archival documents could be accessed from either the 1945-54 period or post-1975 (housed in 11 map, she replied that it was from a student of hers who had worked for a French NGO in the area. She added that the French were not very 'open' in sharing information. I resisted making any comparisons with the Vietnamese. Paradoxicaly, it was easier to find good quality topographical and administrative maps of Thai Nguyen province in a make-shift bookstore in a Hanoi aleyway than anywhere in Thai Nguyen province, including the university. At this same 'bookstore' I was able to buy a book published by the Instiute of Sociology that was not available for sale or reference at the Instiute of Sociology or the Institute's library itself. As another example, a particular issue of the wel-known journal Vietnamese Studies sold at this shop was available neither at the journal's publishing outlet nor at the National Centre for Social Sciences library. From the various vignetes portrayed above, it should be clear that this it was not uncommon to be sent on wild goose chases for data. This was due in particular to the lack of centralization of information and coordination between instiutions and research centres in Vietnam, making it difficult to find many materials on a given topic in a single location. Instead, it was usualy necessary to pay visits to many diferent places, or subsequent visits to the same place—sometimes by chance talking to a diferent person in a given ofice—in the hopes of eventualy finding some useful information. I ended up going in vain to many publishing houses (Statistics, Ethnology, Women's, Social Science, Education, Transportation, and Politics) atempting to locate relevant materials. I visited the Land Administration Ofice in Thai Nguyen City, where I was able to acquire data on land types but was told that for data on land transfers I would have to go to another ofice. Moreover, the oficer informed me that the provincial ofice only housed the aggregated data, so to locate data on transfers and land alocation by district I would have to visit each of the seven districts around the province. Then upon visiting the f Statistics Ofice in one district, I was told that for some data I requested I would have to go to the district-level Land Administration or Department of Labour ofice since data for the district were not centralized in a single Statistics Ofice. Moreover, in this ofice, data on land use were 'missing' for some years, so I could only access statistics for half the communes in the district. When I asked about any available data from the pre-1985 period (the communes were reshufled between districts at this point), I was told it had been thrown out. And no, the ofice had no maps of any sort. separate Archives), as these were restricted for reasons of political sensitivity. 12 In gathering data from some government ofices, I would explain my general research topic and then ask what related data they had. Oficials sometimes had books of statistics and files thick with data siting on their desks, but rather than volunteering anything they knew to be relevant to my topic, they would ask what specifc data I wanted. If I mentioned something specific, they often replied that they did not have that data or would ask "from what year to what year?" I would ask what data are available (or ask for "al that is available"), pointing to their stacks of statistics, but—without specificaly denying me access to the material—they would continue to ask vaguely what data I wanted. The whole thing began to feel like a strange kind of performance. When I did begin to formulate lists prior to visiting particular ofices, geting access to the specific information on my list sometimes required gently asking the same person ten times. On another occasion, I atempted to leapfrog the Agro-Forestry Colege's bureaucracy by using my leter of introduction from the Instiute of Economics in Hanoi. This came after making several requests of my interpreter (from the Ofice of International Cooperation at the Colege) to make an appointment for me to visit the provincial statistics ofice in Thai Nguyen—and after having been taken instead to the Thai Nguyen City statistics ofice. I first tried to cal the director of the provincial Statistics Ofice, but there was no answer. I then caled the 'documentalist' directly (another doctoral student had passed on his telephone number). He explained to me in Vietnamese that even with my leter of introduction, I needed to first contact the provincial People's Commitee to get authorization and a leter from them to then be able to visit the Statistics Ofice. Without my knowing, he then caled my sponsor at the Instiute of Economics to verify the authenticity of my claims. Despite these frustrations, respect and patience did at times pay off. It took me six months of persistent but polite requests through the Agro-Forestry Colege to eventualy acquire some of the provincial statistical data I was looking for. In contrast, I have heard of more than a few researchers losing their patience, thus jeopardizing any chance of future research in a particular locale Although through my fieldwork I encountered some interesting surprises that led me in new directions, I was in some ways able to achieve much less than I originaly hoped for. This can be atributed to the multiple constraints outlined above combined with my own perhaps unrealistic expectations about what data I would be able to gather. This left me with a series of unanswered questions and unrealized plans. I had intended to compare the distribution of land in a vilage context pre- and post-colectivization but was unable to colect old or new vilage-level maps—and having people sketch their own plots to try to fit together a map at the scale of a 13 vilage or hamlet did not seem viable. In other instances, I was unable to acquire statistics for multiple administrative levels (hamlet, commune, district, and province) on land transfers, or data diferentiated by ethnicity or household headship. A representative of the Forest Protection Department told me that his ofice had no data on the historical changes in forest cover in Thai Nguyen. At the Ofice of Resetlement and New Economic Zones, I was told that this ofice had no data on population change or ethnic minority population, despite being responsible for sedentarization programs involving ethnic minorities. And a representative from the Thai Nguyen provincial Women's Union told me that her organization had no provincial demographic data on numbers of female-headed households (neither disaggregated numbers of single mothers, divorcees, and widows, nor numbers of female-headed households as a whole). Nor did she have numbers of Women's Union members, nor data on women's labour and employment (for example, in agriculture, industry, and service sectors), nor on conditons of ethnic minority women, nor on women's access to credit. All of these could have been valuable for poverty targeting and other types of analyses. When I finaly asked what data they did have, she said none. Formali t ies of Loca l Fieldwork Thurston (1983) and Kurti (1999), among others, have writen of the importance of official seals of approval for conducting fieldwork in socialist countries. The formalities and bureaucracy of authorization for field research might be indicated by noting that to arange a visit to a district or commune, I was expected to cary a leter of authorization from the Agro-Forestry Colege specificaly stating the dates and places to be visited, bearing a red stamp, and signed by the rector of the colege. A copy was sent to the provincial police, the provincial People's Commitee, and the district-level People's Commitee. I technicaly should even have had a leter of authorization to visit a foreigner working for an NGO in one district. To facilitate these authorizations and contacts in the field, it was invaluable to have a university liaison. When a senior lecturer from the Colege recommended that I talk to the provincial Forest Protection Branch or Land Administration ofice, he cautioned that I should go with someone from the Colege "or they won't tel  you anything." Moreover, I could not have brought an interpreter from Hanoi, as I did mid-way through my fieldwork, without already having had a university liaison introduce me to the local authorites in the commune where I was then conducting a survey. 14 Writng on Hungary, Kurti (1999: 174) noted that "Just like the socialist state bureaucracy itself, interviews had to progress hierarchicaly from the top down." After first meeting with district officials, I was generaly accompanied or directed to the commune that had been agreed upon for household interviews. Upon my first visit to any commune, I would generaly meet with one or usualy a group of oficials and be quoted some general statistics on demographics and production levels. An often al-male entourage—from the commune People's Commitee, sometimes a district-level agricultural extensionist, and occasionaly even my curious hired driver—would then accompany me for some initial household interviews. As the district and commune authorites became more familar with my presence and my line of questioning, there was less need for these formalities, and I was permited to proceed with only my (female) interpreter to cary out further interviews. Authorization for a foreigner to actualy reside in a vilage in Vietnam is rare. For Vietnamese researchers themselves, even those doing ethnographic research, it is unusual to spend more than a few weeks in one vilage. The restrictions on movement for foreigners are not limited to researchers. NGO project staf who did not regularly live in' Thai Nguyen had to cal ahead from Hanoi to arange visits to project vilages and be accompanied by a representative from the Department of Agriculture. Even long-term NGO project staf living and working in Thai Nguyen got hassled by the police occasionaly. Some 'sensitive areas' of the region were of-limits to foreigners for miltary-strategic reasons, as were parts of the Central Highlands in southern Vietnam. And a volunteer English teacher in Thai Nguyen was unable to organize an English-speaking club without the authorization of the Communist Youth League. Despite al these comments, I do not wish by any means to paint a 'totalitarian' portrait of Vietnam. Research on what are perceived to be sensitve issues can in any country be cause for official concern and subjected to restricted access. For al the obstacles recounted here, there were at least as many exceptions and many positve experiences. Mailng boxes of photocopied documents out of the country in my case posed little concern to authorities. At other times, I would be expecting to pay for photocopies that had been prepared at my request but they would be handed over at no charge. I was invited to countless meals, sometimes in people's own homes, was ofered rides on motorcycles or in chaufeured cars, and experienced much warm hospitality. Each institutional culture has its own rationale and history. Three hypotheses might be forwarded to explain restrictions on access to information in Vietnam. These relate to 15 Confucianist authority structures, to the bureaucratic socialist system, and to the market economy, which converts information into a commodity to be hoarded and selectively exchanged.6 The first of these hypotheses links Confucianist structures to a paternalistic atitude toward 'the masses,' who are assumed to be not inteligent enough to think for themselves. Hickey (1964: xv) provided the folowing explanation of this trend: It is a feature of the Vietnamese character that an official investigation brought to them [the Vietnamese] through regular channels wil elicit from them readier and more explicit answers than wil  any unfamilar private inquirer. This remains from their long experience of Chinese and Sino-Vietnamese oficialdom and administrative techniques. Inquiring into the common weal and woe should be the strict prerogative of the sovereign, his delegates, and his advisers. Knowledge is power. All information concerning the realm should accordingly be kept 'classified' as we [Americans] might put it ourselves—made aware as we have been, thanks to the spirit and experience of the Cold War, of the possible use and misuse of a restrictive discipline. In linking aspects of the bureaucratic system to "the Vietnamese character," Hickey might be criticized for essentialism. Nevertheless, his comments serve as a valuable reminder of the persistence of historicaly embedded practices in curent day Vietnam. Instances of the second hypothesis, linked to socialist system bureaucracies, were aluded to above in the discussion of leters of authorization necessary to access individuals or information. Linked to both the first and second hypotheses, in some ways it seems as though the 'system' is designed so that no one has to take responsibilty for anything and so that information is compartmentalized and dispersed as much as possible. This would prevent the accumulation of information in the hands of any one person who might be able to see things as a whole. Telephones are often locked to alow incoming but prevent outgoing cals—though this is common in other developing countries where charges are levied on a per-cal basis. In the context of institutional transformation in Bulgaria, Staddon (1999: 205) noted that, "One is particularly struck by the amazing perseverance of institutional arrangements associated with the colapse of the state socialist development model into the present time, including labour markets, circuits of capital circulation and information exchange" (emphasis added). The little statistical data and other research materials that are available in Vietnam tend to be carefuly guarded and not readily made accessible. Even with the appropriate leter of introduction to enter a given library, much of the most useful and most recent materials may not actualy be obtainable there but instead in private ofices and locked cabinets. These may be only gradualy ofered to foreign researchers on an ad hoc basis, in accordance with how local 6 M y thoughts on these issues have been enriched in part by discussions with Michael Leaf (pers. comm., 1998). 16 oficials or scholars gauge the likelihood of foreigners to ofer them or their institution some financial gain or other tangible benefit—be it through future colaborative research or development projects or by compensation for translation and logistical assistance. Acts of assistance may thus be calculated in terms of what each party can ofer the other. This exemplifes the third hypothesis regarding commodifcation through the market economy, elaborated in the folowing section. The Commodi f icat ion of Resea rch : Expectat ions and Incentives One subjective aspect of fieldwork involves learning how and when to appropriately compensate certain individuals—for instance, local oficials and poorly-paid state employees— for their time in aranging field visits, interviews, or other forms of assistance. Crol (1994) noted that in China conducting rural research is complicated by the fact that there is now much more competion for the time and atention of cadres and farmers: time is now money. Kurti (1999: 176) noted this reality for fieldworkers in Hungary and Romania: informants' time, "which had formerly been regulated by the party, trade union or communist youth league was now under the constraints of the market and money." Kurti also reflected that while he was formerly able to work through Communist Party-afilated mass organizations to identify interviewees, geting to people in the new market context required a degree of 'seling' himself such that participants would judge it as a worthwhile use of their time. This was a harder task than he expected among young people. Some other areas, such as parts of Mexico, where I conducted research for my master's thesis, became saturated with researchers, and 'interview fatigue' set in as local people tired of being approached and asked a multiude of questions. In such circumstances, it is not surprising that local people may begin to strategize and ask for compensation for granting yet another interview. In Vietnam, now that people are more conscious of their relative poverty on a global scale, a foreign researcher is more inclined to be seen as a walking walet. On one occasion, upon visiting a particularly poor farming family, my interpreter turned to me and asked, "aren't you going to give her some money?" In other instances, I was asked for money—or more discretely, for an 'envelope'—in exchange for various services or information, including for interviews with university professors or museum researchers in their workplace. Occasionaly, I obliged. Perhaps under the impression that a researcher's budget is infinite, some researchers had been asked by state authorites to pay US$ 200 for a single map. Some foreigners had even abandoned research projects in Thai Nguyen province for this reason. One foreigner told me he 17 had paid a lot of money for what was assumed to be curent data based on government surveys, when in fact the data had been generated through projections of old figures. In another instance, upon returning to Vietnam, a Vietnamese researcher who had been studying abroad was asked to pay US$ 100 for climate data necessary for his dissertation. In such instances, the Vietnamese interpretation of a market economy has perhaps been taken too far. These responses are often the outcome of experiences with foreign projects that ofered remuneration for interpreters and other services far beyond local prices and salaries (a typical university researcher earns less than US$ 50 per month). This made it difficult for graduate students with smaler budgets to cary out research in areas touched by this monetarized 'project syndrome.' I was advised to avoid conducting field research in some areas that were becoming known as 'project districts.' This methodological constraint might be caled the cargo cult—an anthropological term used in New Guinea and Melanesia to refer to the expectation that outsiders wil bring rich cargoes of goods. This cargo mentality can constiute an obstacle to research as wel as to development projects, to the extent that it creates expectations and thus the delivery of goods and services becomes disproportionately emphasized over local capacity building or less intrusive research. In this way, such research or development projects can harbour feelings of dependency on hand-outs. Christoplos (1995: 13) discusses the related issue in Vietnam of being perceived as a donor instead of a researcher, such that when foreigners engage in rural field visits, statements by local oficials or farmers are constrained to relate only to the 'constructed needs' that they anticipate the foreigner or development project might ofer them (e.g., credit, irrigation pumps, or schoolbooks). People often assumed that I was conducting 'market research,' and my university liaison at times had dificulty explaining to local oficials that I was a researcher seeking to understand local conditons and issues rather than a businessperson about to invest in the region—much as they might have wanted me to be. The Typ i ca l ' C a s e : Res is tance to Acknowledg ing P r o c e s s and Di f ference 7 I encountered further dificulties in my fieldwork in researching about strategies, practices, 'resistance,' processes, diferences, trade-ofs, and certain sensitve topics. This posed particular problems in researching the two case studies presented in Chapters Five and Seven that focus precisely on marginal cases and practices and processes of policy implementation. I 7 The discussion in this chapter, in general, and this section, in particular, pertains largely to my experience conducting research in northern Vietnam. Southerners often speak more openly and more critically of state policy than do northerners. Thus, I would be somewhat reluctant to apply these observations to a southern Vietnamese context. 18 thus had to be rather indirect in accessing this information when it was not forthcoming or if felt it inappropriate to present as my primary research focus. In meeting with Vietnamese oficials or even researchers, I found it chalenging to move from a discussion of policies, norms, or discourses, to actual practices. Even having people acknowledge the existence of this gap—the dissonance between the de jure and de facto—was difficult in some contexts. This division between rules, on the one hand, and practices, on the other, is perhaps more cemented in Vietnam, a legacy of unpopular socialist policies combined with Confucian and mandarin heritage. From my experience, people seemed conditoned to not talk about everyday practices that may run counter to given policies or social norms—at least not with anyone doing research, which until recently in Vietnam tended only to focus on the (de jure) 'official line.' Of course, everyone knew that disputes and 'back stage' practices existed, but to openly acknowledge these strategies of coping or resistance and their incongruity with official discourses or norms was generaly not viewed as an appropriate—or at least not a commonplace—topic for social science research. Vietnamese ethnologists,8 by contrast, seemed to focus not on legitmating (de jure) state policy but on documenting socio-cultural norms (with an emphasi  on folklore or rituals, particularly among ethnic minorities). Here, too, I perceived a resistance to examining de facto circumstances, no mater how far apart this reality may have been from the circumscribed norms. For instance, time and again, I read and heard references to the custom that women did not inherit land. Yet, in my fieldwork I came across a number of cases in which they did. From her experience in China, Crol (1994: 292) discussed this methodological chalenge of breaking through "colectively constructed representations, to diferentiate social norms from social practice," in terms of a clearly defined ideology representing social structures and social processes as they 'ought to be,' how certain socio-political and economic instiutions ought to function, and how political, social, and economic relations ought to be constructed. It is thus more difficult to identify rapidly what actualy is, as opposed to what ought to be. In terms of subject mater, it may therefore only be feasible to research the less intimate and politicaly sensitve areas for which data are relatively easily atainable and less subject to normative constraints. (Crol 1994: 292) To many people, it seemed odd that I wanted to engage them in discussions about the process of land alocation. On a related note, a Vietnamese doctoral student in anthropology who was studying abroad and returned to Vietnam to conduct fieldwork told me of her dificulties in finding an adequate translation of the term 'dynamics.' These type of 'how' The Soviet-influenced higher education system in Vietnam included training in ethnology rather than 19 questions were always the ones I found hardest to elicit responses to. People would sum up a tremendously complex process of determining land alocations in a given vilage by saying simply that "two households agree on the border between their plots." My atempts to problematize categories and assumptions were often hard to get across. A further chalenge was geting at the ambiguites and variations in colectivization by locale, by land type, and over time—i.e., the local specificities of how a national policy was received and implemented. This reflected a preference to discuss only 'the representative case,' however fictional this was. I was led by local oficials to believe that the history of land tenure in district X was the same as that of North Vietnam as a whole. I was informed by university researchers that al mountainous areas had colectives in the past, that there was no deviancy or variation in experiences. To the extent that ethnic minorites were recognized by Vietnamese researchers, they were often colapsed into a category of 'the ethnics' as a homogeneous whole, without acknowledgement of any diversity in experiences, customs, or production practices between or within specifc ethnic groups. Moreover, government statistics measuring population data by ethnic group were based on paternal'ethnicity, never taking account of mixed ethnicity despite the relatively high rates of inter-mariage in some areas. Related to the above chalenges, I was often discouraged from interviewing unsuccessful farmers seen by university researchers as 'unrepresentative.' A district-level agricultural extensionist who accompanied me on one commune visit was emphatic that a poor Hmong (ethnic minority) household I visited was atypical, but when we came upon a particularly successful Kinh (ethnic Vietnamese) household nearby he said nothing about it being exceptional. Most Vietnamese researchers and oficials seemed only interested in discussing cases that were deemed 'typical,' even if these were in fact exceptional 'model' farmers or communes. Instances of marginality and diference were dismissed as being unrepresentative— a tendency compounded by a lack of understanding or appreciation of qualitative methods.9 Until recently, research by Vietnamese scholars generaly aimed to confirm policy decisions or document success stories, such as the mechanization of agriculture in a model commune in the Red River Delta. The dangers of over-generalization are captured in Christoplos' (1995: 11) characterization of Vietnamese research practice: "The quantiative data regained dominance in anthropology. 9 The majority of existing social science research in Vietnam that is based on empirical research uses questionnaire and survey data. Folklore studies aside, ethnographic research to reflect people's everyday practices and perceptions—including methods such as participant observation and unstructured interviews—seems to be slowly growing in its use for applied social science research. 20 presentations of research results. Diversity was shoved under the rug.. Agency, and the creativity of the individual informants, was forgoten in the interest of constructing a generic 'poor farmer.'" Even reports based on so-caled 'participatory' methods "almost inevitably construct an image of an 'unembedded' agriculture, caried out by 'unembedded' farmers" (Christoplos 1995: 14). In my experience, there was a general reticence among oficials and researchers alike to look beyond the model cases and acknowledge the trade-ofs or problematic aspects of given policies. In response to a question about the impact of Vietnam's socialist experience for women, one policy advisor on gender issues told me frankly that she did not care what the revolution did for women. It was often difficult to find people who would say much about the positve side of colectivization, on the one hand, or about problems of implementation of the land alocation in the 1980s and 90s, on the other. The tendency to view issues in black and white terms was exemplifed when I inquired about any possible disadvantages of the curent economic system in terms of social problems. Local oficials often said there were none and atributed this to the high respect for community norms. In a formal interview seting, it was often hard to access people's perceptions of 'back-stage' goings-on, beyond the rosy image of the wel-integrated community and society that is so frequently projected to outsiders (and, I suppose, to insiders). As one Land Administration oficer explained, "if someone moved to the area after land alocation, the commune can give them some spare land. The atmosphere is harmonious, with people helping each other." Transitons and policy implementation were frequently portrayed as smooth and automatic. Conflict and negotiation was not seen as, or at least not portrayed to be, a part of everyday life. That said, it is worth realizing just how recently critques of overlooking diversity and conflict have been levied on international planners and development analysts for their problematic depictions of farmers or communites (e.g., Scoones and Thompson 1994; Guijt and Shah 1998). Seen in this comparative light, perhaps my observations in Vietnam were not so exceptional. The conduct of Vietnamese researchers and state oficials outlined above could be linked in part to a fear of losing face and admiting mistakes. The perceived lack of critical thinking and/or expression of individual opinions can also be linked to the political sensitivities of speaking out against state policies. A commune official in an area where I was to conduct a household survey repeated twice that I could do the interviews provided that I would not ask any 'political' questions. In earlier discussions with a university representative about the selection of potential field sites, I was told that Christian vilages had historicaly had poor 21 relations with the government—so if I worked in these vilages I might have problems with the police but, on the other hand, the local people might speak to me more openly. In another context, in noting that family ties and official political positons play an important role in the distribution of products in a given district, Kiuru, Lehtonen et al. (1997: 60) wrote that "Due to the sensitivity of the issue, a ful presentation of the distribution system may cause confusion and even disturbance." These examples al point to the chalenge of researching topics that are politicaly sensitive. Other issues requiring special discretion and sensitivity ranged from land disputes and land transfers to taxes, social classes, and women's rights. Socioeconomic diferentiation was a delicate topic and people were sometimes reluctant to report their income or land size. One doctoral student told me of discrepancies in land size figures in which the figure recorded on a land-use right certificate was two to three times smaler than the area of land the student measured on the ground as actualy belonging to the given farmer. Crol (1994: 295) noted that in China, in the past, "there was less cause for secrecy, given that taxes were paid by the colective and there were fewer diferentials within vilages, where 'everybody knew everybody else's affairs.'" In the curent context, however, especialy in richer regions, research on economic activities of individual households is a much more time-consuming exercise, given that peasant households are now much more complex, autonomous and diverse economic units, less inclined to reveal the details of their economic activities, incomes and savings." (Crol 1994: 295) Land conflicts and land inheritance posed further dificulties in interviews. Responding to a question about the implications of land inheritance for women, a Women's Union representative insisted, "It's not very important." The topic of gender issues was hard to raise in many contexts in which men were present. It was simply not taken seriously and often seen as a laughing mater. Amongst a mixed group, men often disappeared when I asked questions of or about women, suggesting boredom or lack of respect for the issue and for the person responding. Reference to any form of intra-household diferences, on the other hand, met with some hostilty amongst both men and women, to the extent that it chalenged the social norms of family harmony and unity. A university researcher once went as far as to suggest that my research on the gendered implications of land alocation could be ilegal, but I assured her it had the endorsement of the National Centre for Social Sciences and Humanites. Her response is indicative of the common view among state oficials that land alocation could not be a gender issue. Moreover, it reflects how such oficials fail to recognize any gap between policy and 22 practice and how they see little value in studying the implementation of a policy, as distinct from the policy itself. Given the plethora of methodological and logistical constraints detailed above, I developed certain strategies to negotiate my research objectives. To compensate for missing or hard-to-access data, I turned to alternative means and methods, located other types of informants, and drew on documentary sources. To avoid certain sensitivities, I 'repackaged' my research topic depending on the context. I often presented my project to oficials by explaining that my focus was the development of the household economy in the doi moi period in Vietnam, downplaying the historical and ethnic components. Rather than referring to 'problems,' 'conflicts,' or 'disputes,' I sometimes used indirect questioning or observation or chose less controversial language by asking about 'difficulties' in implementation of land policies ('chalenges' was harder to translate). In place of wealth, income, diferentiation, and land conflicts, I refered to resources, wel-being, and land use. Much information could be inferred about colectivization and decolectivization by using biographical methods of listening to people's life histories and experiences under diferent policies and historical periods. These various anecdotes related important aspects of setlement history, old landlords, war experiences, administrative changes, and environmental changes. I was often able to draw useful insights as wel from popular jokes, which appeared to be an avenue for indirectly venting critques of the economic and political system. Urban residents joked, for example, about the double meaning of the standard acronym for socialism, XHCN (xa hoi chu nghia), which in the past they alternately coined as meaning xep hang ca ngay: 'queue for goods al day long.' My Interpreter and I: Interpreter-Researcher Relat ions The literature on ethnographic fieldwork has become increasingly sensitve in recent years to issues of the researcher's positionality. In my discussion earlier, I refered to some of the expectations that local people may have of a foreigner in Vietnam. Yet, most accounts of researchers working in cross-cultural contexts brush aside the positionality of the field assistant or interpreter—to the extent that one is necessary. This is a serious omission, since factors such as age, gender, regional and class background, and prejudices such as atitude towards women or ethnic minorites can play a tremendous part in shaping interactions between the researcher, interpreter, and research subjects and the nature of the data obtained. In my fieldwork in Thai Nguyen, I had a total of four interpreters—two who worked with me for only a couple of days, 23 and two for longer periods of time. The potential chalenges of working with a male interpreter on one occasion became obvious when he invited his buddies from the Department of Agriculture to a lunch I was to pay for, ordered large quanties of beer, and later began drinking rice wine with a group of men at the next table in the restaurant—encouraging the driver to do the same, against his wil. He was then (not surprisingly) reluctant to start work punctualy after lunch and later cracked jokes about 'the women's movement' while I was interviewing the leader of a local Women's Association. One of my longer-term interpreters was a female instructor at Thai Nguyen University. Having studied in Cuba in the 1970s, she was more fluent in Spanish than English, so we communicated with each other mainly in Spanish. She had the advantage of being wel-known and respected by many oficials in the district-level Departments of Agriculture, since in many cases she had been their instructor during their studies at the provincial Colege of Agriculture and Forestry. She was older and her children had nearly completed high school so she had few family responsibilties and thus more time available to work with me. My other interpreter was a young woman from Hanoi who was completing a degree in English. Although wiling to work hard, she was rather inexperienced and became intimidated during some interviews with district officials. Despite the chalenges, my relationship with both these interpreters was friendly and mutualy supportive. Many people in Vietnam considered it hard to convince female Vietnamese researchers or interpreters to join research teams going to rural areas of the country. They may have young children at home and be unable to spend extended periods away from family or their husbands may be reluctant to alow them to go. Moreover, working in dirty, 'rough,' and non-air-conditoned rural areas had a stigma in the eyes of many people, particularly for those who had atained the status of researcher or interpreter. I remember teling some Vietnamese friends in Hanoi about where I was carying out my research only to be met with the response, "Why would you want to work there!" The quality of interpretation and translation services in Vietnam is often not high, and my interpreters had no formal training as such. In these circumstances, my own understanding of Vietnamese was helpful to check the quality of interpretation. I was often able to make clarifications, identify omissions, and ask questions myself. There were times when I faced problems of interpreters answering my questions themselves rather than asking the person being interviewed. The interpreter sometimes could not understand my rationale for asking the same question to diferent people in subsequent interviews and would remind me that I had already 24 received an answer to that question. One interpreter discouraged me from asking questions about inheritance or polygamy and (jokingly) sighed and roled her eyes at the detailed questions I asked during interviews. Another limitation among interpreters or research assistants was their lack of social science training, in terms of both analytical and methodological skils. I mentioned above my interpreter's lack of appreciation (or tolerance) of my line of questioning on some occasions. Another problem was one interpreter's dificulty un-gluing her eyes from the survey questions during interviews, and her inabilty to act naturaly and make conversation before initiating an interview. An additonal factor that was noted by Crol (1994: 295) as complicating research in post-reform China was "a decline in the use of common definitions and greater variation in the meaning atributed to terms in common usage." Although this was not an obvious problem for me, there were reams of new terminology associated with the Communist Party and government that are relatively unfamilar to Vietnamese who left South Vietnam before 1975. This was an issue for one Vietnamese person who worked with me in Canada to translate some documents. A final point to add to this discussion is the problem of representation and 'othering,' in relation to ethnic minority people in Vietnam. Atitudes of Kinh, Tay, and Nung people towards other ethnic minorities, particularly the Hmong, were frequently condescending, patronizing, and fraught with misunderstanding. In a rural market, when I initiated a conversation with a group of Hmong women, the surounding crowd of Nung on-lookers began to jeer at the women in an insulting tone. When I expressed my interest in interviewing a Hmong person to one Nung commune official, he sent me of with his assistant to a nearby hamlet where the assistant proudly presented me to a man, exclaiming, "here's a nguoi Meo (Meo person)," as if he were showing of a museum exhibit or some kind of merchandise (Meo is an old and somewhat derogatory name for Hmong people). In another instance, after visiting a Hmong household, my Kinh interpreter joked with a district-level extensionist about the poor conditon of the house, the Hmong woman's iliteracy, her inabilty to specify how much land she had, and her son's incompetence in obtaining a good price for seling a bird that he had trapped. Arriving afterwards at a relatively new house of a clearly wel-of family, the interpreter exclaimed, "This must the home of a Kinh or Tay. What a nice house this is." The interpreter further criticized Hmong people for their constant dependence on state programs, always asking for project hand-outs. Another university instructor I interviewed complained that Hmong people rejected eforts of the state extension service to donate cows to them, implying that Hmong were stupid and closed-minded. In so doing, he overlooked the inappropriateness of catle to the 25 Hmong production system, as the cattle could compete with and damage the Hmong's traditional swidden crops. Plate 2. Interview in a Multi-generational Household Synthesizing Complexity: Research Approaches and Data Analysis I came to researching Vietnam out of an interest in agricultural policy and land reform in the context of socialist development models, but I soon realized that socialism represented a tiny 26 blip in the long historical experience of Vietnam's development. During my first months in Vietnam, I interpreted the socialist experiences of this country to a large degree independently of the context in which they were embedded. I frequently overlooked the multiple influences and syncretism, the mixing of Marxism, Confucianism, Taoism, Buddhism, ancestor cults, and 2000 years of nationalist struggles for the defense of the nation. All of these are intricately entwined in a complex and long-term perspective that has shaped the national identiies of many Vietnamese, compared to the historical vision of a single war, a scant three decades in the past, that many North Americans have in mind when visiting the country. The longer I stayed, the more I began to resist simplifications, and instead to appreciate and slowly untangle the many layers of historical and cultural influences that have come to shape contemporary Vietnam and its people. Without an ethnographic approach, much of this richness would likely have remained invisible to me. Apart from the chalenges of carying out my fieldwork—the primary subject of this chapter—another series of struggles faced me in seting about the process of data analysis. This entailed synthesizing the complex aray of field notes and other bits of information colected over 14 months of fieldwork. Analysis and writing is not an automatic, mechanical process leading in a direct and linear fashion from data colection. Few textbooks provide a satisfactory guide to reflect my experience of working with qualitative data. Data analysis and writng—for qualitative and ethnographic research in particular—cals for much patience, imagination, and careful synthesis. Much of the process of reworking interview text to conform to a presentable and workable format felt as though I were struggling to contain a giant octopus. The most chalenging aspect of writing this dissertation was achieving to my satisfaction an adequate meshing of reams of so-caled 'raw data' from diverse sources with a relevant (but certainly not ready-made) conceptual apparatus, in order to lend a broader signifcance to otherwise un-integrated observations, comments, and images from interviews, conversations, observations, statistics, reports, or newspapers. I use the term 'raw data' in quotes to emphasize the extent to which ethnographic field notes are already to some degree conceptualy informed. As noted by Emerson, Fretz, and Shaw (1995: 167), "theory enters in at every point, shaping not only analysis but how social events come to be perceived and writen up as data in the first place." Rather than being grounded in field research and paying atention to local specificity, the ungrounded theory in too many dissertations ends up overburdening, de-centering, or misrepresenting experiences of 'reality.' In some cases, it is writen even before the fieldwork is 27 conducted, with minimum revisions made after completion of the fieldwork. To avoid this, it has taken a considerable amount of time and efort for me to explore diferent themes and ordering schemas and to identify overlaps between bodies of data and theory, without forcing the later onto the former. The craft lies in 'listening to the data' to build a conceptual framework for the analysis and in achieving a constant interweaving of field data with theoretical ideas. In leting topics emerge from my recorded field notes, I had at diferent times some 50 indexing codes used to file my notes—photocopied and cut up by hand—into recipe boxes. Many more sub-categories were added later when I reached the stage of fine-tuning the classifications after they had been put onto the computer. Despite al I have said about avoiding a predetermined analytical framework, the role of the researcher in filtering and conducing the direction of the analysis at al stages of the project should be acknowledged from the outset. The same set of field notes could be writen up in an infinite number of forms. Ethnography is a craft—and an underused method among human geographers (Herbert 2000). Its practice cals for a good degree of judgement and a fine line must be drawn as the research proceeds in alowing the scope of the analysis to evolve and adapting it to the context at hand. 28 Chapter Two. Study Aims and Research Design One of the major historical events occuring in the contemporary world since the 1980s was the colapse of socialist systems and their experiences of adjustment that involve the reintroduction or expansion of a market economy. In moving away from the security of the 'iron rice bowl' economy and facing greater exposure to market fluctuations, this adjustment process imposes many dificulties. This dissertation focuses on some aspects of this adjustment process in Vietnam, namely those concerned with property rights and decolectivization and their uneven consequences for livelihood vulnerabilty, with particular atention to the impact on gender and ethnic relations. The study demonstrates the diversity of experiences of decolectivization, especialy when the analysis is brought down to the micro level. This underlines the need for atention to complexity in the analysis of social processes. The term 'interfaces' that appears in the title of this dissertation is borowed from Norman Long's discussions of the interfaces of structure and agency in his actor-oriented perspective discussed later in this chapter. Land reform and property rights issues have been long debated in Vietnam to determine efective rural development strategies to meet both eficiency and equity objectives. Land policy reforms implemented in Vietnam since the 1980s suggest a new orientation in land management and production organization and new roles for the state, markets, cooperatives, and agricultural households. Decolectivization marks the shift from primarily colectivized production to household management of agricultural land. The 1993 Land Law represented a definitive step in decolectivization, ensuring farming households the rights to produce and market their own produce through private channels and to transfer, exchange, inherit, mortgage, and lease long-term land-use rights. Alocation of land to households reinstated the household as the principal production unit. Agricultural land—an important asset and often a key axis of diferentiation—has been relatively evenly alocated through decolectivization. Although China has undergone a somewhat paralel process of decolectivization as in Vietnam, the pace of socio-economic diferentiation there has been more accelerated (Luong and Unger 1999). Given the many paralels between recent economic restructuring in developing countries and former socialist countries, my analysis situates Vietnam not strictly within a lens of other emerging 'post-29 socialist'10 trajectories but within a broader context of the neoliberal agenda shaping global economic processes. Both developing countries and former socialist countries have been subjected to a common 'market panacea' message (Spoor 1997). In relation to land issues, this agenda has translated into the end of state-led agrarian reform, reflecting a trend in countries as diverse at Mexico and Vietnam to move away from colective forms of land management {ejidos in Mexico and agricultural production colectives in Vietnam) and towards the alocation of individual title or long-term land-use rights (e.g., DeWalt and Rees 1994; Szelenyi 1998). Land and its redistribution have long been considered as central elements of welfare provision and poverty reduction eforts in additon to being a 'moral imperative' to redress injustices of colonialism. The 1980s and 1990s witnessed increasing debate over the appropriate functions of states and markets in land management—in developing country and 'transitional' (or post-socialist) economies alike. Some policy analysts in Vietnam are doubtful whether any form of land redistribution or alocation wil in fact aleviate poverty given that smal-holder production lacks economies of scale or the scope to apply 'modern' technologies (pers. comm., Ha Huy Thanh). From this perspective, the policy of egalitarian land alocation wil  not be efective in stimulating economic development for the population at large. This move is seen as too generous to poor farmers and as preventing an accumulation of land in the hands of those who can beter manage it. By extension, state-led land redistribution is no longer considered a useful mechanism to target the poorest sectors of the rural population and aleviate rural poverty. Within the broader set of economic and institutional transformations underway in Vietnam, the property rights reforms in the agricultural sector represent some of the most significant changes. Of the total employed labour force, 67 percent is engaged primarily in agriculture and forestry activity (General Statistical Ofice Yearbooks 1996-98). Poverty is highly concentrated in rural areas, where 76 percent of the population officialy resides. The 1997/98 Vietnam Living Standards Survey revealed that just nine percent of urban dwelers but 45 percent of rural dwelers lived in poverty. Vietnam had a population in 1999 of 76 milion people. Yet, the land area to feed this population is relatively limited. Vietnam's agricultural land area of 7.4 milion hectares comprises only 22 percent of the country's surface area, with rice cultivation occupying over half this area. The population density averages 214 persons per square kilometer and the average area of cultivated land per capita is just 0.11 hectares— comparable to that of Sri Lanka and Bangladesh and one-third the size of China or Thailand I use this term in quotation marks since in Vietnam the country is still considered to have a socialist orientation, 30 (ADB 1997). These figures suggest why it is critical to take account of agriculture, rural development, and land issues within Vietnam's overal development strategy. Aims of the Study The overal aim of this dissertation is to analyze the institutional changes reflected in processes of decolectivization and land alocation and to emphasize the interplays of gender, household, kinship, ethnicity and agro-ecological diferences in shaping new paterns of livelihood vulnerabilty. This general aim is divided into three objectives. The first is to analyze decolectivization in Vietnam (particularly the property rights reforms via land alocation and certification) in terms of the restructuring of formal and informal institutions. This includes production organization and management, scales of decision making, service provision and social safety nets, new cooperative forms, and the reworking of relations between state, market, and rural households. This analysis is used to trace new forms of rural governance and the changing path of development implicit in Vietnam's decolectivization experience. The second objective of this study is to identify gaps between the dejure national policies of decolectivization and de facto experiences of implementation at the vilage level. It thus highlights the interplay of national policy with local responses based on socio-cultural, ethnic and gender dynamics. The study emphasizes variations in colectivization and decolectivization temporaly and spatialy and the interfaces of changes at various scales, from the macro and national to the micro and regional/ocal scale, based on an analysis of dynamics of decolectivization in Thai Nguyen province. The third objective is to demonstrate the relations between the institutional shifts associated with decolectivization and emerging paterns of livelihood vulnerabilty in Vietnam. The analysis is disaggregated by region, ethnic group, and gender, through two case studies in the ethnicaly diverse province of Thai Nguyen. To achieve the third objective, I seek to (a) identify institutional gaps in service provision associated with the loss of colectives as mechanisms for coordination and cooperation; and (b) identify changes in entilements to land and other resources and the forms of social capital that mediate this, taking account of both social norms (such as inheritance), and kin and ethnic networks. albeit with a market economy. 31 Organization of the Dissertation The dissertation is organized as folows. The opening chapter ofered readers a flavour of my experiences in carying out ethnographic fieldwork in Vietnam, including aspects of data access, quality, and interpretation, and broader considerations in adapting my research strategies in the field. The remaining part of the curent chapter on study aims and research design outlines some key elements of the research methods—the integration of multiple scales and methods of analysis, data sources, and site and case study selection—that guided my study. Land—embedded with its qualities as a store of wealth, a source of colateral, and a marketable good—is a key asset and social safety net to reduce vulnerabilty for rural inhabitants (Elis 2000). This is especialy true in much of northern Vietnam, particularly in the northern highlands in areas where there is limited livelihood diversification beyond agriculture. While focussing in particular on the productive asset of land, this study examines the related but less tangible assets of social capital and household-kinship relations to understand how they mutualy afect entilements to land. Rather than viewing poverty in static and purely economic terms, my analysis focuses on vulnerabilty to beter reflect dynamic social processes that produce diferential entilements to land. Various types of formal and informal instiutions are shown to mediate access to assets and activities for sustaining livelihoods. Agricultural restructuring, the market economy, and privatization al require certain instiutions in order to reduce risk and thereby reduce vulnerabilties. Chapter Three lays the conceptual foundation for my subsequent empirical analysis, justifying the analytical focus on institutions, vulnerabilty, kinship, gender, and social capital. Recent studies suggest that macro-level social capital and colective action tend to be low in post-socialist societies due to the termination of many formal organizational structures (e.g., agricultural colectives). My analysis examines the extent to which this is true in the context of the decline in colectives as coordinating mechanisms for agricultural production. It further examines whether less formal social capital at the meso- and micro-levels have been reinforced in some setings (e.g., kinship-lineage or ethnic groups). The study of institutional change in Vietnam, including both colectivization and decolectivization, should not be considered merely in terms of formal or legal institutions. Many practices—relating to sale of produce or implementation of land alocation policies, for example—occur despite, rather than because of, existing formal institutional structures. This point was underlined in the preface to Fforde's (1989: xi) study of cooperator resistance in 32 Vietnam: "knowledge of formal or legaly constiuted social structures does not necessarily tel much about underlying reality." Fforde (1989: 4) emphasized the importance of understanding formal structures first but then seeing through them to grasp 'the real meaning.' This is the rationale for my continuing reference to both formal and informal instiutions and their • interrelations. Anthropologists often refer to a 'front stage' and 'back stage' in a given seting. They suggest that only after considerable time conducting fieldwork do the initial representations that are portrayed officialy, and particularly to outsiders, give way to more a fine-tuned understanding of local dynamics and everyday practices. My study seeks to capture both of these aspects. Chapter Three closes with an overview of doi moi (or 'renovation')—the economic restructuring process implemented since 1986—and its reflection in discursive shifts heralding entrepreneurship and market competion. Chapter Four lays out the institutional context for my analysis. A brief historical overview of land relations in Vietnam underlines the degree of symbiosis of colective and household sectors in agricultural production through the phase of colectivization. A subsequent section details the evolution of land policy and property rights reforms and their realization in practice. As property rights and production are reorganized, new forms of rural governance begin to emerge. The remainder of the chapter outlines how the new discursive shifts are reflected in a range of new institutional forms. New actors for rural development are brought on board as colectives are dissolved or reformed. Local groups coalesce to organize and demand services such as transporting and marketing produce and providing inputs or agricultural extension information. NGOs, often in colaboration with mass organizations, substiute or complement state-led services and welfare provisioning. This later part of Chapter Four further identifies how new paterns of vulnerabilty in Vietnam are emerging in the form of institutional gaps created by the decline in some services provided by agricultural colectives. One implication of the changing status of households vis-a-vis colectives and the state is that farmers now have the opportunity to make production choices themselves. But a new handicap shaping household diferentiation is the technical knowledge to efectively exploit the land that farmers now control. Gaps in technical knowledge and entrepreneurial know-how are becoming increasingly important in shaping vulnerabilty in Vietnam's market economy. Social networks in turn often shape access to knowledge. The recent atention by political ecologists to issues of uneven access to natural resources has some relevance for my study. However, because I have not concentrated my analysis on environmental changes, or environmental consequences of the institutional changes 33 in the Vietnamese countryside, I decided against adopting a political ecology approach. I am also sensitve to critques of political ecology suggesting that many analyses impose preconceived theoretical agendas and over-privilege political factors at the expense of a sound analysis of the multiple natural and human-induced causes of biophysical changes (Vayda and Walters 1999). The later point cals for an analysis which many social scientists, myself included, are unqualifed to perform. Moreover, although political ecology models traverse micro- and macro-scales of analysis, Zimmerer (1991: 443) found them lacking in that they fail to "setle the contradiction between the primacy atributed to the individual in human and cultural ecology and the structuralist framework of political economy." Chapter Five takes up the intersections of gender, land, and rural livelihoods in Vietnam, highlighting the gender-diferentiated impacts of changing property rights and land privatization. Through an analysis of kinship- and gender-based norms and networks, it is argued that a lack of social capital can disadvantage some types of female-headed households. Inheritance and household headship paterns are shown to shape livelihood vulnerabilty, particularly through a lack of entilements to land. The reordering of the household and its renewed status as a unit of decision-making and alocation of labour—responsibiltes formerly taken by agricultural colectives not households—have served to reshape intra-household relations of authority. Shifts in intensification of land use and diversification of production bring about concomitant paterns of diferentiation between households and workloads within households, as people's use of use of time, space, and resources is restructured. These shifts bring about new forms of vulnerabilty at the inter- and intra-household level. Inheritance of land—an important customary instiution—is reemerging in the decolectivized countryside, bringing about new intra-family divisions along gender lines. Analyzed this way, inheritance is seen as a key factor in explaining diferential entilements to land. In the later part of the chapter, a vilage-level case study of household headship, residence paterns, and inheritance is presented to ilustrate these trends. Chapter Six opens with an examination of regional and ethnic disparities, practices of high modernism, and the changing of the social geography of Vietnam. This occured through the resetlement of two main groups during the colectivization period, one being Kinh from the Red River Delta, and the other, shifting cultivators from higher altiudes into valey areas of the midlands. The remainder of the chapter provides a characterization of the northern uplands region and of Thai Nguyen province. This begins with an examination of the population dynamics of its development as a frontier territory and site of in-migration. Next, drawing on 34 provincial statistics, I demonstrate the process of colectivization from the 1950s to 1970s and its gradual disintegration in the 1980s. Folowing from this, the household economy experienced significant expansion in the 1990s. A discussion of contemporary land use and the alocation of agricultural and forest land folows. This chapter further highlights the curent features of rural land tenure and rural-urban disparities. A key emphasi  throughout my study is on the localized outcomes of policy processes, akin to Fforde's (1989: 4) observation that "while central authorites determine the form of rural institutions, local practice may be free to determine their content." Gaps between policy declarations and the practices of implementation at the local level have a long history in Vietnam. These gaps are highlighted in my analysis of the periods of colectivization and decolectivization. The problems of periodization, or clearly demarcating these diferent institutional regimes, is ilustrated by the significant symbiosis or co-existence of the private/household economy within the former colective system and by the uneven and at times intermitent experiences of colectivization. Each locality had its own idiosyncrasies in not folowing the norms of national policy—however clear these norms may or may not have been to begin with. Chapter Seven examines how local autonomy shaped the diverse processes and outcomes of colectivization and decolectivization. A case study demonstrates the role of ethnic identiy as a basis for ancestral land claims and diferential land access, in turn shaping paterns of livelihood vulnerabilty. This analysis can be seen as reflecting "the signifcance of the concept of locality for analyzing how global trends are translated into place-specifc processes" (de Haan 1997: 176). From this case, I discuss how institutional changes are shaped by and in turn shape new cultural, ethnic, family, and territorial identities, and identiies based on regional autonomy; identities. I argue that not only formal institutions, such as laws or established organizations—but also customary ones, based for instance on shared kinship or ethnic identiy—shape new livelihood vulnerabilties in rural Vietnam. This chapter demonstrates how ethnic identiy can serve as a basis for ancestral land claims and how the operation of social networks can exclude outsiders in an ethnic group. It further ilustrates how Kinh are not always the most advantaged group. Through land alocation, Tay, Nung, and San Diu groups, among others, claimed land that would have otherwise been shared also with Kinh and Dao in-migrants. Other experiences of the new land policies include Hmong and Dao shifting cultivators facing increasing enclosure of lands and resources that they formerly depended upon through common property regimes. This is particularly critical with new classifications and 35 alocations of forest lands. Policy design to date has not been suficiently flexible to develop culturaly sensitve alternatives for such scenarios. A principal contribution of the dissertation is to traces the multiple contours of policy trade-ofs and new paterns of vulnerabilty in light of the institutional reshufling in contemporary Vietnam. This reshufling resulted in new opportunites but also in a decline and privatization of some colective goods and services formerly available to farmers. Compensating for and responding to new circumstances and opportunites, those with a favourable combination of material and social assets are able to take beter advantage of the situation and get ahead. The concluding chapter of the dissertation draws together the diverse threads of this analysis, underlining the signifcance of the empirical trends documented through the study. An Overview of Methods Relatively few foreign or Vietnamese scholars have conducted long-term local-level fieldwork in Vietnam. Even fewer have focused on the northern midlands and highlands region of the country. In this context, I set out to make a modest contribution in a field that lay stil  relatively unexplored. I sought to make sense of the complexites of agricultural land relations in the local implementation of land alocation and decolectivization, as lands formerly managed by colectives came under control of individual households. The study makes both empirical and analytical contributions, based on original data colection and on a synthesis and reconceptualization of existing studies. Given the limited number of empirical studies available on some of the issues addressed in this dissertation, my approach was often inductive and exploratory. Determining the appropriate questions to be asked was frequently a large part of the chalenge, even if time and logistics did not permit answering them as fuly as I would have liked during the period of fieldwork. The dissertation might thus also serve to define an agenda for further research on processes of decolectivization as derived from these preliminary trends and observations. Consistent with recent ethnographic approaches, my study of land policy interprets the actions and responses of local people as engaged subjects rather than as passive recipients of policy. This 'actor oriented approach' (cf. Long 1989b; Long 1997) centres on elucidating diferent perspectives and practices among actors and on the interplay of structures with agency. Such an approach is particularly crucial for studies in Vietnam where many of the institutional adaptations in land policy were established in response to changes already underway on the ground (Fforde and de Vylder 1996). 36 The P lace of the Loca l : A Multi-level Ana lys is Human impoverishment and social relations of vulnerabilty reflect relationships that are complex, multilevel and multidimensional. Understanding these relationships cals for interdisciplinary research combining institutional, economic, environmental, demographic, and socio-cultural processes. Interdisciplinary studies are often about diversity. The metaphor of 'border crossings' is common to such studies, suggestive of much of the disorientation that accompanies venturing into foreign territories and disciplinary domains. This may be accompanied by a sense of being overwhelmed with an infinite number of interrelated issues and perspectives to consider when one aims at a 'holistic' or integrative analysis. There is a temptation to try to account for every possible factor and to avoid criticisms of "but what about.?" Managing an integrated multi-level, multi-dimensional, multi-method analysis is fraught with chalenges, at times provoking self-doubt on the part of the researcher. Are al these 'multiples' a product of the post-modern fetish for pluralism and diversity? Am I going overboard in an infinitely diferentiated analysis? How does one adequately combine an analysis of macro trends with an ethnographic (hermeneutic) approach? I hope my study wil serve to raise questions about just how do we cross borders. My research resists simplistic categorization. It crosses many boundaries, not merely international geographic ones, but disciplinary and institutional ones as wel, moving between academic-theoretical and policy-orientations, in the process engaging diverse audiences. Academic cultures can inhibit interdisciplinary research and individual disciplines often have traditons of research that are scale-specific. For example, many sociological studies have tended to be at a more macro-scale and classic anthropological studies at vilage-level. Moreover, empirical studies have been further conditoned by discipline-specific methods, with sociology employing large-scale surveys and aggregated census data and anthropology employing ethnographic methods such as participant observation, with a smaler number of cases. Geographers often employ hybrid methodologies to address issues of multiple scales. The tension between large-scale generalizations and the uniqueness of place lies at the heart of geographical analyses. Yet, frequently insuficient atention is given to the interplay between large- and smal-scale processes and phenomena and the local becomes subservient: The local context [is seen as] structured by general processes, which deprive the local of its singular character. .. .The local is either used as a source for sampling data, or seen as a spatial expression of general tendencies, without a history and an identiy of its own. (de Haan 1997: 153-154) 37 The decline of community and vilage studies is due to their perceived failure to acknowledge the social disintegration within communites, on the one hand, and the relation between communites and the broader social and economic context, on the other. While myopic studies overlook the broader context of events, macro studies decontextualize the specificity of locality, masking the richness of local-level detail. The later often disregard the 'situatedness of practice' and reflect little of the lifeworld of individuals (de Haan 1997: 153-154). One issue I muled over throughout my fieldwork and the elaboration of my analysis is the interrelation between the multiple scales of vulnerabilty and outcomes of the economic reforms. I toyed with how to encapsulate the geographies and spatial ramifcations of vulnerabilty. In particular I sought to move beyond the identification of vulnerabilty solely at the level of a spatial unit (e.g., a commune, or a remote region), as is common in Vietnamese measures of poverty. Selecting a single level of analysis—such as one community or ethnic group—would have given too partial a view of the multifaceted implications of decolectivization. My study analyzes households within their agro-ecological production system, embedded within broader economic and social systems and formal and informal institutional structures in which they engage in everyday life. Such an actor-oriented approach erases macro and micro distinctions, understanding how local processes are integrated into regional, national and global systems. There is a growing interest among geographers in socio-spatial dynamics and the production of space and scale, linking the production of localities to the production of scales. Marston (1999: 3) suggests that scale maters to geography in order "to understand processes that shape and constiute social practices at diferent levels of analysis" and the relations or interaction between scales. The organization of chapters in this dissertation reflects the atention to scale in my analysis. Chapter Four presents an overview of institutional change at the national level, while Chapter Six presents the regional context. In addressing the emerging paterns of vulnerabilty at diferent scales, the two case studies presented in Chapters Five and Seven address the relevance of scale by iluminating how vulnerabilty plays out at the levels of family-household, region, and ethnic group. Studying Up, Down, and Around: Multiple Methods and Data S o u r c e s The title of this section makes reference to Laura Nader's (1982) article (originaly writen in the 1970s) on perspectives gained from 'studying up.' Nader denounced anthropologists for their lack of research on the middle and upper classes, in comparison to the 38 relatively abundant literature on the poor and disadvantaged, and on ethnic minorities. She chalenged us to consider What if, in reinventing anthropology, anthropologists were to study the colonizers rather than the colonized, the culture of power rather than the culture of the powerless, the culture of afluence rather than the culture of poverty? (Nader 1982: 457) Studying up as wel as down, Nader noted, forces us to turn many conventional research approaches on their head, to ask questions in reverse, and to examine resistance to change among the privileged instead of only among the poor. Doing this ultimately produces beter theoretical and empirical analysis. Although applied to a US context, studying 'at home' within the discipline of anthropology, Nader's ideas might apply equaly to research internationaly and in other disciplines. In adopting an institutional analysis for my study, I atempt to incorporate some of Nader's perspectives on studying up. I do this by drawing on a pool of interviewees that includes not only poor farmers, but state officials, researchers, and development project staf in order to understand the making of new social relations and networks in contemporary Vietnam. A further argument made by Nader is for an eclectic aray of methods. No single research method is comprehensive; al have trade-ofs. In light of this, my study emphasizes methodological complementarity. I employed a variety of methods that would alow me both the scope to understand broad trends as wel as the specificity to observe exceptions to the rule, including macro- and micro-level divergences in policy implementation and experiences diferentiated by ethnic group and gender. Although complemented with participant observation and documentary research, the study's core method was unstructured and semi-structured interviews. The rationale for this was to promote free-flowing discussion and alow for a more diverse range of issues to emerge spontaneously without having to ask directed (or mis-directed) questions. In so doing, I sought to avoid the problem of imposing categories or topics when people would not otherwise construct their reality in the way I do—although undoubtedly I remained guilty of this at times. This may be as subtle as asking if there are any disadvantages of land alocation when local people did not generaly think in such terms. The 341 interviews that I conducted over two periods of fieldwork range from the national to household level. The interviewee types are listed in Table 1 and can be further grouped as folows: 1. 69 unstructured and semi-structured interviews with vilagers of Kinh, Nung, Tay, San Diu, Hmong, and Dao ethnic groups in 13 communes across six districts of Thai Nguyen and neighbouring Bac Kan provinces. This included a household survey of 25 households in one 50-household vilage of primarily Nung ethnicity in Dong Hy district, Thai Nguyen 39 province. The survey used random sampling of one-third of the vilage households, supplemented with purposive sampling to encompass al cases of female headed and uxorilocal1 households. Given the structured set of questions on household characteristics, no separate interviews were conducted for husband and wife. 2. 38 unstructured and semi-structured interviews with rural households in Cao Bang, Ha Giang, Lao Cai, Ha Noi, Bac Ninh, and Nam Dinh provinces in northern Vietnam, Quang Binh and Thua Thien Hue provinces in central Vietnam, and Can Tho, Long An, and Binh Phuoc provinces in southern Vietnam. 3. 106 unstructured and semi-structured interviews with hamlet, commune, district, provincial, and central-level authorites in various government ministries and departments, nearly half of which were in Thai Nguyen province. These ministries and departments included Agriculture and Rural Development; Labour, Invalids and Social Afairs, General Department of Land Administration, Forestry Protection units, the Instiute for Agricultural Planning and Projections, Ofice of Resetlement and Fixed Cultivation, Agricultural Extension ofices, Statistics ofices; Cooperative Association, People's Commitee and People's Council, and a variety of mass organizations (Vietnam Women's Union, Farmers' Association, Gardener's Association, Veteran's Association, Youth Association). Other local informants included health clinic atendants and teachers. 4. 113 unstructured and semi-structured interviews with Vietnamese and foreign researchers and staf of international development agencies and NGOs. 5. 15 unstructured and semi-structured interviews with other people, including shopkeepers and university students. In additon to these interviews, the folowing sources were drawn upon: 1. Books, reports and historical documents colected from government sources, national research institutes, publishing houses, international agencies, and university and public libraries, to complement and coroborate findings from interviews (27 libraries and resource centres were consulted). 2. Quantiative economic and demographic data from national and provincial government ministries and agencies. 3. Maps (administrative, cadastral, topographical, and land-use type). 4. National and local newspaper articles in English, French, and Vietnamese. 5. Notes from five weeks of workshops on 'Concepts and Methods for Localized Poverty Reduction in Vietnam,' that I co-faciltated through a colaborative university project with the National Centre for Social Sciences and Humanites at universites in Thai Nguyen, Vinh, Hue, Dalat, and Ho Chi Minh City. 6. Informal conversations and observations: gossip, oral histories. 7. Content and discourse analysis of museum exhibits (from 16 museums across the country), photos, postcards, pamphlets, proverbs, popular expressions, literature, and music. These often unexpected data sources ofered insights into social norms, representations of gender, experiences of societal change and popular perceptions of economic systems. 1 1 Refer to the glossary for a definition of these terms. 40 Table 1. Number of Interviews Classified by Type of Interviewee Type* 1997 1998 Total" Rural households*** in Thai Nguyen and Bac Kan provinces, of which: Kinh households Nung households Tay households Hmong households Dao households San Diu households Household survey (not included above) 2 11 1 14 1 3 7 2 3 25 69 Rural households outside Thai Nguyen province 33 5 38 State officials and community representatives in Thai Nguyen province Commune or hamlet level District level Province level 17 16 4 12 49 State officials in other provinces Commune or hamlet level District level Province level 16 13 20 49 Central-level officials 4 4 8 Researchers Vietnamese (National Centre for Social Sciences and Humanities; universities and research institutes; museums) From Thai Nguyen University and provincial museums Foreigners (and overseas Vietnamese) 24 18 2 17 2 7 70 NGO or international development agency representatives Vietnamese Foreigners (and overseas Vietnamese) 10 6 7 20 43 Others (e.g., shopkeepers, students) 15 15 T O T A L 138 201 341 * Instances of joint positions (e.g. academic researchers and NGO staff) were classified as per the principal capacity in which the interviewees' experience and ideas were being solicited. ** Instances of repeat interviews were counted no more than twice. *** Households of mixed ethnicity were counted as per official Vietnamese reporting, according to the husband's or father's ethnic group. Although useful to lend breadth to my analysis, not al data I colected were analyzed exhaustively for this dissertation, given the constraints of time and space. Interview data and documentary sources were analyzed by developing an evolving series of codes based on both research questions and categories emerging from the data. New categories were developed when 41 recuring themes were identified. These various data types provide the backbone to the analyses in Chapters Four through Seven. The analysis of provincial statistical data is presented in Chapter Six and in the statistical narative in the Appendix. Data were colected during two principal periods of fieldwork in May-July 1997 and January-December 1998. In 1997 I participated in a study tour on rural resource management and sustainable livelihoods organized by Geofrey Hainsworth through the Centre for Southeast Asia Research at the University of British Columbia. In 1998,1 was hosted by the Instiute of Economics and Thai Nguyen University's Agro-Forestry Colege, the later being my home base for field work in Thai Nguyen province. I studied Vietnamese for one and a half years in Canada, and one month intensively in Hanoi, in additon to the practice I gained during my fieldwork. From these experiences, I atained an upper intermediate level of conversation, reading, and writing in Vietnamese. Although I was accompanied by an interpreter during most of my formal interviews, I was able to ask questions and understand about half of the communication. Site and C a s e Study Select ion This study bridges macro, regional, vilage and household levels, using Thai Nguyen province—located some 85 km north of Hanoi, in the northern midlands of Vietnam—as the seting for the largest part of the fieldwork (see Map 1). In relation to the more numerous studies of the Red River Delta, fewer studies have examined the midlands and highlands regions of northern Vietnam, highlighting the diversity and unevenness in experiences in the implementation of both colectivization and decolectivization. The midlands region, and Thai Nguyen province in particular, seemed to be a prime seting to reflect this diversity, given its ethnic compositon and agro-ecological variations. 42 Map 1. Vietnam, Showing Thai Nguyen Province Map 2 shows the location of my research sites (indicating commune names) within Thai Nguyen province. Within the province of Thai Nguyen, the selection of research sites (and specifcs of the research problem) were identified by their relevance to the issues I wished to examine. But as most ethnographers appreciate, this is also guided to an extent by the availabilty of, and access to, adequate data (as discussed in Chapter One). The criteria for selecting given districts and communes as research sites in Thai Nguyen province were complicated by the problem of finding 'representative' sites, not to emphasize a 'typical case,' but rather the diverse outcomes of the reforms. Particularly grounded in an analysis of ethnicity, the study documents various forms of contestation of national policy to explain localy divergent outcomes. Through my review of secondary sources and in combination with primary data colection, the study established a set of conditons to account for diferential local outcomes of 43 land alocation. These include, inter alia, traditons of customary land tenure and mutual assistance, ethnic compositon, historical experience of land concentration and colectivization, history of population migrations, population density, agro-ecological conditons (proportion of agricultural vs. forest lands, valey vs. upland production systems, paddy rice vs. slope-land crop and swidden production, extent and tradition of home gardens, etc.), degree of market integration and access (remoteness), land quality, and extent of land and labour markets. Such variation in local conditons contributed to diferent structures and outcomes of alocation of agricultural and forest lands, particularly in terms of being more or less egalitarian, encouraging more or less efective protection of forest resources, or reflecting to a greater or lesser degree of reclaiming of ancestral land. Given these considerations, plus my emphasi  on ethnicity, I sought sites that would reflect a range of conditons in midland and highland provinces rather than delta conditons. For this reason, I de-emphasized the deltaic southern districts in Thai Nguyen. Through this approach, I was able to contrast various sites in the study and seek instances in which the implementation of land policy had been subjected to 'negotiation' between the national and local (province, district, commune and vilage) levels. 44 Map 2. Map of Field Sites in Thai Nguyen Province 45 Chapter Three. Conceptualizing Institutions, Property Relations, and Vulnerability The pace of change has not been uniform, and the path of transformation has not been unidirectional. (Luong 1992: 186) The bases for land access, and more broadly for livelihood vulnerabilty, are in flux in the new economic and institutional context of contemporary Vietnam. My analysis interprets decolectivization in terms of shifts in formal and informal social instiutions that can produce or reduce vulnerabilty. This chapter outlines the elements of a framework for a social assessment of Vietnam's recent reforms in land policy and land relations and the associated creation of new vulnerabilties. This chapter opens by addressing the question of why a framework of instiutions is pertinent to this study. It distinguishes between formal and informal or customary instiutions and discusses their relevance to economic development. From here, the second section link an institutional analysis to the notions of poverty and vulnerabilty and, later, social capital, including its ethnic and gender dimensions. The third section outlines the relationship between land rights, property, and gender. The fourth discusses institutional restructuring, land reform, and new rural governance, examining in particular experiences of property rights reform in post-socialist or transition countries. The final part of the chapter turns to an overview Vietnam's economic reform program broadly refered to as doi moi. The discussion examines the new market economy discourses and development orientations implied by doi moi. Why Institutions? The post-Fordist development of local governance has led to partnerships between diferent actors and administrative levels, such as various business coalitons of private, quasi-public, public, and voluntary actors. .New ideological underpinnings are a consequence of the breakdown of the possibilty (and wilngness) of pursuing traditional planning and governance strategies. (Tykkylainen 1998: 326) Deregulation and wider processes of restructuring of governance around the world are leading to new institutional constelations, as indicated in the above quotation. The 1990s witnessed a surge of interest among many international development agencies in building instiutions for development, creating or strengthening "organizations and fiscal arangements to ensure eficient delivery of health, education, and family planning services, clean water, roads 46 and other infrastructure. [for] building the human and social capital necessary for the rural population to participate in and benefit from the emerging market system" (Rondineli and Litvack 1999: 9). Institutional coalitions, such as the type indicated by Tykkylainen above, represent an extension of this. Figure 1 provides a conventional framework and the various components of "institutional development for economic transition." Within this, the economic support instiutions include property rights, financial markets, labour markets, marketing and distribution systems, and legal instiutions for business. The property rights component of this wil be discussed in greater detail later in this chapter. Douglas North, the foremost proponent of new institutional economics (NIE), defines instiutions as 'rules of the game' in a society, referring to: the humanly devised constraints that structure human interaction. They are made up of formal constraints (e.g., rules, laws, constiutions), informal constraints (e.g., norms of behaviour, conventions, self-imposed codes of conduct), and their enforcement characteristics. Together they define the incentive structure of societies and specifc economies. (North 1990: 360) A related influential approach based in the analysis of instiutions is that of common property resources, exemplifed by Elinor Orstrom: "Common property analysts such as Orstrom (1990), by contrast, tend to take their theoretical grounding from game theory, looking at colective action dilemmas and focussing on the ways in which instiutions or rules can be purposively crafted to produce colective action" (Mehta et al. 2000: 14). Both of these approaches has "established firmly that instiutions mater and that local people, as wel as state governments, can successfuly manage resources through property regimes varying in scale and space" (Mehta et al. 2000: 13). Both these approaches, however, can be found lacking in their explanatory power. Leach, Mearnes, and Scoones (1999) are critical of the tendency of NIE to lump together norms, rules, and behaviour in a single understanding of institutions. This approach is further limited by its exclusive focus on formal instiutions that alone are unable to account for the complexites of resource management practices characterized by overlapping rights. NTE thus fails to consider the interactions between formal and informal institutional arangements at multiple scales. Linked to a more anthropological or sociological notion of regularized practices, Leach, Mearnes, and Scoones (1999) instead consider instiutions as regularized paterns of behaviour that emerge from 'rules in use' between individuals and groups in society. 47 Figure 1. Institutional Development for Economic Transition MACROECONOMIC REFORMS STRUCTURAL ADJUSTMENT POLICIES STABILIZATION POLICIES FOREIGN TRADE AND INVESTMENT POLICIES NATIONAL ECONOMIC AND SOCIAL DEVELOP MENT POLITICAL REFORMS DEMOCRATIZATION DECENTRALIZATION DEREGULATION ECONOMIC SUPPORT INSTITUTIONS SYSTEM OF PROPERTY RIGHTS FINANCIAL INSTITUTIONS MARKETING & DISTRIBUTION SYSTEM LABOR MARKET INSTITUTIONS BUSINESS LEGAL SYSTEM MARKET DEVELOPMENT PRIVATE SECTOR DEVELOPMENT PRIVATIZATION OF STATE-OWNED ENTERPRISES SMALL & MEDIUM -SIZED BUSINESS DEVELOPMENT INCENTIVES TO ATTRACT MULTINATIONAL COMPANIES INSTITUTIONS OF CIVIL SOCIETY INSTITUTIONS OF HUMAN RESOURCE DEVELOPMENT SOCIAL SAFETY NET POUCHES SOCIAL SERVICES SOCIAL AND PHYSICAL INFRASTRUCTURE Source: Rondinelli and Litvack (1999: 10). When analyzed in conjunction with uncertainty, instiutions need to be seen not as mere rules of the game or rigid organizations but rather as sites of social interaction, negotiation and contestation comprising heterogeneous actors having diverse goals (not al of which are material or economic in nature). (Mehta et al. 2000: 35) Instiutions are not static, but in flux, responding and adapting to ongoing changes in circumstances. This is al the more true in periods of dramatic economic reform, such as Vietnam has experienced in the last two decades. Instiutions are at the heart of shifts in property rights and ownership structures associated with the recent land reforms and decolectivization. The analysis of instiutions provides a central framework for this study in explaining paterns of diferential entilements to land between social groups. "Instiutions play a critical role in livelihood sustainabilty, both in the sense of exclusion from access to instiutions (such as credit markets), and exclusion by instiutions (such as the tenure system preventing access to 48 land for certain social actors)" (EDS 2000). Instiutions that mediate economic relations and resource management structures extend from macro to local levels. A macro-level focus on formal structures such as state policy, laws, colectives, mass organizations, banks and financial markets, should be matched by a paralel focus at a micro-level, particularly focusing on informal (or what is sometimes caled customary) institutions. The later may operate as schemes for mutual assistance (also known as moral economy or informal social safety nets) or for regulation of communal property resources. Adger (2000a: 353) describes the complex institutional arangements of "local level property rights associated with coastal resources.. [These common property resources are] complex mixes of state, private, and regulated and unregulated commons, often nested within each other and al changing and evolving over time." Customary instiutions encompass vilage-level rotating credit schemes and funeral funds as wel as more generic instiutions such as households, familes, mariage, and inheritance. Together with social norms, al of these social or customary instiutions in diferent ways play key roles in mediating access to, and control over, productive resources. References to informal or customary instiutions inevitably bring us to a discussion of culture. Rather than constiuting an independent sphere, culture and social relations are increasingly seen as intrinsic to politics and the economy and as subjects of interpretation and contestation (Hefner 1998). Culture and traditons are unevenly assimilated. Culturaly- and socialy-embedded instiutions shape the structures of agricultural organization and are themselves afected by farmers' paterns of resource use (Bery 1989). My analysis in subsequent chapters considers how economic relations, be they state-legal or market structures, are implicated by the strength and persistence of customary institutions. Such instiutions shape the outcomes of state policy at the local level and, in turn, can often explain the uneven entilements for diferent social groups. This explains my rationale for focusing on the interplay of formal with culturaly embedded structures and institutions. Linking Institutions to Livelihood Vulnerability The analysis of vulnerabilty is part of the third objective of this study. The term vulnerabilty can be applied to a wide variety of contexts and begs the question, vulnerabilty to what? My analysis revolves around livelihood vulnerabilty, the opposite of livelihood security. As described by Elis (1998), a focus on livelihoods incorporates multiple dimensions: A livelihood encompasses income, both cash and in kind, as wel as the other social instiutions (kin, family, compound, vilage, and so on), gender relations, and property 49 rights required to support and to sustain a given standard of living. Social and kinship networks are important for facilitating and sustaining diverse income portfolios.. social instiutions are also critical for interpreting the constraints and options of individuals and familes distinguished by gender, income, wealth, access, and assets. (Elis 1998: 4) My study thus adopts this operational definition of livelihood, incorporating elements of income, social instiutions and social networks, gender relations, and property rights, al of which combine to form the constraints to and opportunites for ensuring against livelihood vulnerabilty (see Table 2). Table 2. Components of a Rural Livelihood Assets: • natural capital (e.g. land) • physical capital (e.g. tools, housing) • human capital (e.g. labour, skills) • social capital • financial capital Activities: • crop output, livestock, gathering, farm wage, non-farm wage, non-farm self-employment, remittances Access mediated by: Institutions: • property rights (including common property), real markets, policies, etc. • social relations (village, ethnicity, gender, etc.) • organizations (government agencies, community associations, NGOs, etc.) Source: Adapted from Ellis (2000: 16). To situate the analysis, this section presents a conceptualization of vulnerabilty. In recent years the concept of poverty has to an extent been superceded by a series of alternative concepts: entilements and capabilties, social exclusion, marginalization, and vulnerabilty. These concepts aim, in diferent ways, to do at least two things: (1) to beter capture the multi-dimensionality of the experience of deprivation, by moving beyond simply measuring income and (2) to emphasize dynamic and qualitative processes rather than considering poverty as a static condition, quantiatively measured at a fixed point in time. In a similar vein, studies of poverty 'as a lived experience' have increasingly drawn atention to transient poverty, underlining how the resource or asset base of many people (graduate students included!) fluctuates over periods of time in the short or long term. Important contributions to the reconceptualization of poverty have been made by Amartya Sen, 1998 Nobel laureate in economics. Sen's work on poverty and famines focused atention on what he caled people's functionings and capabilities, as constrained by entitlements—the set of alternative bundles of commodites over which a person can establish command, or the range of resources at one's disposal (Dreze and Sen 1989). The earlier 50 understanding of entilements in a legal sense was subsequently expanded as Sen developed the notion of 'extended entilements' to take greater account of social sanctions in access to resources. These sanctions may be based on social conventions or customary instiutions which are not legaly codified. The connection between poverty and vulnerabilty was also emphasized by Moser and McHwaine (1997a), who demonstrated how, having a smaler range of assets, the poor may be, on the one hand, more vulnerable, insecure, and exposed to risks than the non-poor and, on the other hand, less resilent in their capacity to cope or respond to such risks and stresses. Vulnerabilty can thus be seen as comprising two components: the sensitivity, or degree of exposure to a given risk or stress, and resilience, referring to the capacity to cope and respond to shocks (Chambers 1989; Davies 1996). Davies (1996) takes this further to distinguish between risks that are diferential and household specific, such as ilness or unemployment of household members, and those that are community- or regionaly-based and livelihood-system specific, such as a drought. Others have identified a similar distinction between 'idiosyncratic shocks' afecting a single household, and 'covariate shocks' that afect whole communites (Van de Wale 1999). Vulnerabilty can be further distinguished as structural (e.g., in terms of dependency ratio, gender, or ethnicity) or proximate (such as short-erm ilness, pregnancy, or seasonal fluctuations in weather and production levels) (Davies 1996). The concept of poverty is often used to portray poor people at passive victims, located at fixed points in time. This understanding of poverty gives less atention to the potential for people's situation to improve or deteriorate in relations to a given reference point, somewhere below a determined 'poverty line.' The concept of vulnerabilty ofers a more dynamic analysis. Atention to coping strategies and responses, through the notion of resilence, emphasizes an individual's agency to a greater degree than conventional notions of poverty. Recent interest in coping strategies appears to have been born from the analytical shift from outcomes to processes (Davies 1996: 45). Yet, as Davies insisted, not al coping strategies imply resilence: such strategies are often eroneously assocated with the capacity to bounce back. She distinguishes short-erm coping from the longer-term process of adapting, which implies a more permanent change, once coping strategies no longer function. Social safety nets contribute to the aleviation of both poverty—through redistribution— and vulnerabilty—by protecting individuals, households and communites from uninsured income and consumption risks and, thereby, from faling into poverty again. Formal social safety nets include food subsidies, public works, agricultural extension, credit provision, and 51 charity organizations. Informal safety nets tend to be based on family and community networks. In a context of cutbacks in many state services and formal safety nets, informal social safety nets and customary instiutions are becoming increasingly important. As fal-back measures, these informal measures can fluctuate in efectiveness over the long or short term, being either strengthened or weakened in periods of stress. The concept of vulnerabilty has been more commonly applied to contexts of natural disasters or famine and food insecurity. Adger (1999: 251) noted that vulnerabilty tends to be "used to describe the state of exposure [e.g., to stress from environmental change], usualy associated with a geographical location rather than with individuals or social groups." Adger's analysis thus brings greater atention to both agency and power relations in shaping the social vulnerabilty of individuals and colective groups. I examine the issue of power in more detail in the discussion that folows. Ditrich (1998) outlined how spatial vulnerabilty can be deduced from three elements. First are environmental conditons, as climate and natural resource endowments constiute natural constraints for food production. Second are infrastructural and institutional constraints. These include public instiutions for regional and local development (e.g., to support production and marketing) and for social service provision. Some indicators of these constraints are the extent and quality of transportation, communication, market places, social and charitable institutions, agricultural extension services and development projects. Third, Ditrich identified processes related to economic dependence and political powerlessness, vis-a-vis political and economic centres. Indicators include the character of relationships between lowlands and mountain regions, systems of economic exploitation, and the degree of social articulation and political participation. My conceptualization of vulnerabilty emphasizes the later two dimensions. It moves beyond solely environmental constraints to address how institutional structures and power relations mediate property rights and aleviate or reinforce paterns of deprivation. This conceptualization paralels other multi-dimensional studies of vulnerabilty. For example, in the framework used by Wats and Bohle (1993) to analyze individuals and social groups experiencing multiple forms of deprivation, these authors emphasized the need for a multi-level analysis that takes account of both the micro-level shaping of entilement sets and the broader political economy context. They conceptualize three elements defining spaces of vulnerabilty, thereby broadening the focus from entilements to empowerment and political economy. In this analysis, vulnerable individuals can be seen not only as resource-poor but also disenfranchised 52 and exploited. As emphasized in the earlier discussion on gender and land rights, equal access to property rights constiutes an important aspect of empowerment. The domestic domain is an important political space where property rights relations are determined and in this sense it constiutes a space of vulnerabilty (Wats and Bohle 1993). Land is a key asset to reduce vulnerabilty for rural inhabitants, particularly where there are few options for livelihood diversification. Diferential access to land is often a key factor shaping distinct livelihood strategies pursued by rural households of diferent income levels (Elis 1998). How is it determined who gets what benefits? Empowerment approaches to vulnerabilty, which raise questions such as these, highlight the centrality of politics and a theory of power. Wats and Bohle (1993), for instance, see vulnerabilty as a political space. Property rights ensure access to land and other assets, but political rights are also central to the process by which claims can be made regarding public resources to maintain and defend entilements. The political space of vulnerabilty, for Wats and Bohle, is comprised of three aspects. First is the domestic, referring to patriarchal and generational politics. Vulnerabilty in the domestic domain is expressed through gender diferences, access to, and control over, resources. As seen later, Moser and Mclwaine (1997a), by contrast, give more emphasi  to household relations as an asset, de-emphasizing the political dimension. The second aspect of vulnerabilty for Wats and Bohle is work or production politics. Third is the public or civil space, or state politics. Formal political rights (enfranchisement) may be important in securing the promotion of entilements and being granted social security through public action. Atention to institutional structures, influencing both exposure to risk and capacity to cope, implies links to the policy environment. State policies can afect both sources of and cals on entilements. While state intervention on the sources side is widely acknowledged, through provision of production and exchange entilements, "state policies towards coping and adaptation may be less so. The state can reinforce or undermine strategies, by restricting migration, fining travelers without identification cards, impounding goods on the way to market, and pursuing policies to encourage sedentarization of rural producers more generaly" (Davies 1996: 37). From Vulnerabil i ty to Soc ia l Capi ta l To operationalize the concept of vulnerabilty, Moser and Mclwaine (1997a: 65) developed an asset vulnerabilty matrix (see Table 3). This matrix demonstrates how vulnerabilty can be aggravated or reduced through stocks of community, household and 53 individual/intra-household assets. People's wel-being depends on the amount and quality of these diferent resources. At the community level, one can identify elements that strengthen or weaken social capital. Moser and Mclwaine's (1997a: 65) definition of social capital is derived from Putnam (1993): "features of social organization, such as networks, norms (of reciprocity), and trust facilitate coordination and cooperation for mutual benefit. Social capital enhances the benefits of investment in physical and human capital." Such reciprocity exists within communites and between households based on social ties. At the household level, the matrix includes productive and material assets (such as housing, land, machinery, means of transport, and savings) and household relations (a mechanism for potential pooling of income and sharing of consumption). The central assets at the level of the individual are labour and human capital. Human capital determines one's capacity to work and the returns on one's labour, and can fluctuate depending on an individual's health, skils and education. Moser and Mclwaine (1997a) extend the analysis of vulnerabilty and colective action to transitional socialist contexts. In their case study of Angyalfold, Hungary, they found that people had no experience of organizing together for colective action in the face of crumbled instiutions led by a socialist state. They interpreted this context in terms of a lack of social capital stocks for community mobilzation. At the same time, the authors did point out other forms of social capital, particularly inter-household exchanges of smal loans and child care. Table 3. Asset Vulnerabilty Matrix Individual assets (intra-household level): • labour • human capital: health, skills and education (determining capacity to work and returns to labour) Household assets: • productive assets: land, livestock, housing, equipment, means of transportation • household relations: as a mechanism for income-pooling and sharing consumption Community assets: • common property resources: community natural resource base (grazing area, wild plants, sources of fuelwood, fodder, water) • social capital: reciprocity and trust within communities and between households based on social ties and drawing on social resources, reinforced through community organizations, NGOs, informal associations for savings and lending of credit, rice, labour Source: Adapted from Moser and Mcllwaine (1997). For Putnam, communites with a high degree of social capital are more eficient and equitable, with a greater ability to address common problems through shared trust, established 54 through 'horizontal associations.' Putnam's (1993) study contrasted these associations, common to northern Italy, to the more 'vertical' patron-client relations typical of southern Italy. Patron-client relations involve what Davies (1996: 37) identified as relations of subservience or dependence implying structural vulnerabilty for the dependents but a source of entilement for the 'asset' holder. In Putnam's view, the networks and organizations associated with enhanced social capital are considered to have, among other characteristics, voluntary membership, transparent decision-making, and the abilty to cross class boundaries. Researchers and planners are increasingly recognizing the value of the community as "an important asset to decrease vulnerabilty or increase opportunity, depending on its 'stock' of social capital: ..the networks, norms and trust that facilitate coordination and cooperation for mutual benefit" (Moser and McDwaine 1997: 65, drawing on Putnam 1993). Many so-caled 'traditional' practices have long been considered to hinder efective development, but recent conceptualizations suggest that social capital, as a product of customary institutions, and conditoned by cultural instiutions or values, can facilitate cooperation. "Across class and gender boundaries, communites rely on social capital in the form of age-old systems of resource exchange as a means of ensuring colective survival of the community" (Buenavista et al. 1994: 7). These assumptions about the mutualy beneficial nature of social capital wil be explored in more detail below. In her classic study of social instiutions in Africa, Bery (1989) argued that entilements to productive resources such as land depend on social identity. Entilements to resources frequently stem from one's position, membership, or status in social groups—a community, lineage, mariage, colective, or patron-client relation. Berry's (1989) research on social (or customary) instiutions demonstrated the rationale for farmers' preference for investment in social relations over direct productive investment, in order to strengthen their positon in a social group. This alowed them to draw on these stores of group support in times of need. The centrality of these often kin-based social instiutions as social safety nets is highlighted, Bery noted, by the definition of poverty given by local people in Ghana as being 'without kin.' Although Bery does not use the term, her analysis centres on social capital. This can take the form of networks for support, cooperation, forums and meetings, traditional events and ceremonies, and informal instiutions including fictive kinship as a means of gaining access to resources. Social capital (through kinship, ethnic or community networks) may facilitate access to labour, information, loans, or other resources, thus constiuting a form of non-tangible but productive asset that can contribute to reducing vulnerabilty. 55 But as pointed out by de Renzio and Kavanamur (1999) in their study of Papua New Guinea, studies drawing on Putnam and Coleman's interpretations of social capital rarely emphasize the downside of this concept. Instead it is seen as having only positve consequences. By contrast, de Renzio and Kavanamur highlight the cultural constraints, obligations, nepotism, downward leveling, and other disincentives associated with the wontok system of relationships based on common language, kinship group, geographical area of origin and religious group. They also shed light on the lack of cohesion and cooperation between diferent social groups, contributing to a lack of capacity to reach common objectives. Their analysis points to a need for scaling up, to overcome localy confined solidarities; expanding horizontal linkages to 'thicken' existing relationships; rewarding initiative and volunteerism; and cross-sectoral linkages and greater dialogue to improve state-society relations. These authors identify a range of sources of social capital, from micro to macro levels. These can include the family, kinship, household, ethnicity, family business, cooperative, religious group, social movements, cross-sectoral linkages between NGOs, civil society associations, government agencies, and universities. Renzio and Kavanamur's analysis raises the important critique that the benefits of social capital may not be equaly shared. Social capital can function as a community or an individual asset to the extent that it constiutes an element of diferentiation and diferential vulnerabilties. Social capital and customary instiutions are mediated at the level of either the community, as in credit groups, or the family, as in inheritance. They are closely bound to the notion of moral economy. Davies (1996: 37) clarifies how moral economy elements have sometimes eroneously been considered as welfarist leveling mechanisms, or as "an informal insurance system which provides a community safety net in times of stress." To respond to these potential confusions, Shields and Flora et al. (1996) distinguish two types of social capital: those that are built horizontally, referring to egalitarian forms of reciprocity, between people of similar socioeconomic backgrounds, and those built vertically, to which not everyone in a community or social group has equal access. The distinctions above go some distance in providing a finer-tuned understanding of , social capital—a term that has come to be applied uncriticaly in a wide variety of contexts. Yet, perhaps the most useful clarification to address the conceptual vagueness that generaly surounds the term is that of Woolcock and Narayan (2000). These authors distinguish between three strands of social capital: bonding, bridging and linking. 56 bonding social capital [refers to the] strong ties between immediate family members, neighbours, close friends, and business associates sharing similar demographic characteristics; bridging social capital [to the] weaker ties between people from diferent ethnic, geographical, and occupational backgrounds but with similar economic status and political influence; [and] linking social capital [to the] ties between poor people and those in positons of influence in formal organizations such as banks, agricultural extension ofices, schools, housing authorities, or the police. (Woolcock 2000, emphasi  added) This interpretation of social capital helps to make sense of the diferent nature of social capital of poor and beter-of people. Bonding social capital is often abundant among poor people. They tend to have some bridging social capital among more distant friends and acquaintances, but very little linking social capital, or 'friends in high places' to facilitate bureaucratic procedures for commercial, educational, legal or political purposes (Woolcock 2000). It is the lack of the later form of social capital that enhances their vulnerabilty. The strategy that Woolcock advocates for NGOs, firms and government agencies is thus to "mobilze bonding social capital within communites; build more extensive bridging social capital to markets; [and] enhance linking social capital to public institutions." These distinctions and strategies wil  be caried through in my analysis of decolectivization and paterns of livelihood vulnerabilty in Vietnam. In additon to the distinction between types of material and non-material assets (including social capital), an important contribution of the schema of Moser and Mclwaine (1997a) outlined earlier is the authors' atention to the multiple levels or scales of the social processes that shape vulnerabilty. Other analysts have conceptualized this slightly diferently, establishing an 'anatomy of poverty' in terms of both inter-regional and inter-household poverty (e.g., Mason 1996). My analysis integrates these two approaches, examining vulnerabilty and diferential entilements from the intra- and inter-household levels to the levels of kinship group, ethnic group, community, region, and nation. Table 4, adapted from Ditrich (1998: 154), diferentiates these levels in terms of vulnerable social groups and vulnerable spatial units. In practice, however, the two are often interrelated: vulnerabilty is more extreme among particular social groups within already vulnerable or disadvantaged localities. 57 Table 4. Spatial Dimensions and Levels of Mediation for Analysis of Vulnerability Spatial dimension: Level of mediation: • Individual and intra-household • Inter-household, kinship and ethnic group vulnerable social groups • Local community • Regional • National • Macro-regional • Global vulnerable localities and regions (spatial units) Source: Adapted from Dittrich (1998). Ethnicity, Soc ia l Capi ta l , Land and Identity Ethnicity, like kinship, is a binding force in social relations and an important ingredient of social capital. The Vietnamese have fought many wars to protect familes and communites as much as, or more than, to protect the nation. Hetne (1996: 17) notes that the ethnicity "has a primordial or ascribed quality, but it is also true that ethnic identiy is shaped by historical experiences. It is thus at the same time objective, given, and subjective, a creation." Noting the lack of fixity of any given identity, Egwu (1998: 37) cals for a "more dynamic theory to codify various aspects of ethnicity," that accounts for the maintenance, weakening, and renewal of ethnic boundaries. The belief in shared descent, based on ethnicity and kinship, frequently shapes paterns of exclusion and vulnerabilty through diferential entilements. This ambiguous role of ethnicity has been noted in numerous studies. In his study of rural ethnicity in Nigeria, Egwu (1998: 17) observes that ethnicity and kinship "provide a more enduring basis for cushioning the impact of [structural adjustment]. The objective situation, therefore, tends to reinforce the emotional basis of ethnicity." Yet, Egwu underlines the ambiguity in ethnic solidarity constiuting both a liability as wel as a basis for advancement for rural people. On the one hand, inter-ethnic conflicts can threaten development, but depending on the specifc conjuncture can also be constructive for socioeconomic development, encouraging networking for mutualy beneficial economic ventures such as mutual assistance for weddings, sharing of credit, and forming smal cooperatives. One dimension of my study addresses the links between land, territory and the 'reinvention' of ethnicity and identity. The analysis in Chapter Seven demonstrates how ethnic and kin networks are simultaneously unitng but exclusionary forces. The reformulation or renegotiation of the identiy of a given social or ethnic group can serve as the basis for colective 58 action, sometimes lying at the base of conflicts in resource use. Ethnic diferences can often explain regionaly diferentiated outcomes of reforms, in this case shaping who ends up with or without land. In the context of decolectivization in 'post-socialist' transformations, Kanef (1998) notes the emergence of new ethnic-based solidarities, what she terms a 're-activation of ethnicity.' She interprets the emerging capitalist agenda of the Bulgarian state in terms of the renewed importance of land as private property, and the way in which bonds of ancestry to the land takes on real spatial boundaries through land ownership. Her analysis of rural Bulgaria shows how restiution has emphasized the relevance of kinship and ethnicity in defining who is 'native' and who does or does not have land rights. Thus, changing property rights in the Bulgarian context highlight genealogy and ancestral links and ground descent spatialy and temporaly. Kanef demonstrated how, though the content of the relationship changed, kinship relations maintained an important positon in both socialist and post-socialist politcal-economic systems. She concluded that kinship is being used in new ways, in that there is a strengthening of ties between kin and land, especialy as a survival strategy (for food production) in hard economic times. In this way, the economic reforms can thus reinforce 'diference' and the strength of lineage and ethnic groups through networks of social capital. This phenomenon, with paralels to the Vietnamese situation, raises the issue of whether globalization erases or enhances diferences—whether it closes or opens up new spaces for expressions of identiy along ethnic and other lines. The trend for elements of local institutional structures to shape curent processes of restructuring is beginning to be recognized. "Developments in rural areas are increasingly determined by transnational processes and structures, but we also need to consider how the general/abstract are transformed into concrete local practices by the distinctiveness of local social structure and culture" (de Haan 1997: 154; emphasi  in original). Many explanations of restructuring have downplayed geographical diferences and "the inherited structures of society and local and sectoral characteristics" (Neil and Tykkylainen 1998a: 19). What Neil and Tykkylainen refer to as 'the inherited structures of society' others cal path dependency. Staddon (1999: 200) observes that to draw out localy-specific outcomes, research on 'transitional' countries needs to consider several factors: "the importance of path dependence in actualy existing transitions, the signifcance of local institutional negotiations, and also local adaptations to fluid politcal-economic contexts." Observing the degree of path dependence in conditoning local transition in Bulgaria, Staddon (1999: 205) noted that, "One is particularly struck by the amazing perseverance of institutional arrangements associated with the colapse 59 of the state socialist development model into the present time, including labour markets, circuits of capital circulation and information exchange" (emphasis added). Gender ing Soc ia l Capi ta l and Col lect ive Act ion Institutional economics, while ofering some important analytical tools to study social dynamics, remains limited in not suficiently accounting for extra-economic variables and power relations, particularly within family and gender systems (Evans 1993). The discussion below seeks to address this lacunae through linking instiutions for social capital with a gender analysis. Social capital is said to enhance people's capacity for colective action to address common problems. Analysts have been arguing for "more careful consideration of the distribution of social capital and of the kinds of social capital available to diferent groups in society," including class and generational efects. In this context, Lowndes (2000: 536) makes a strong argument for paying closer heed to gender diferences across and within other social categories. A growing number of feminist analysts raise the issue of there being diferent, gender-specific 'circuits' of social capital. In this debate, "a consideration of gender dynamics throws light upon two important issues: the distribution of diferent types (and levels) of social capital within communites, and the nature of the link between networks of sociabilty and paterns of political engagement".(Lowndes 2000). Chiding that studies to date have paid disproportionate atention to male-dominated activities, Lowndes (2000: 534) noted that in Britain, more than twice as many men as women undertook voluntary work related to sports and recreation. Women, by contrast, were more active in voluntary work in the fields of health, education and social services.  [M]en were more likely to occupy commitee posts, while women dominated in visiting and befriending activities. In the Philppines, Shields and Flora et al. (1996) noted that men tended to have greater roles in community politics and the cash economy. Women, on the other hand, had the main responsibilty for community management, including the alocation, provisioning, and management of items of colective consumption—water, healthcare, education, garbage colection, community gardens, playground construction, and involvement in community fairs, festivals or markets. What are the constraints to women's involvement in formal instiutions and participation in colective action? Discussing the context of instiutions for environmental management in India, Agarwal (2000) identified first, women's lack of awareness about the instiutions and, second, entry rules, since membership is open to only one person per household—usualy the 60 male head. Moreover, social norms tend to define the tasks that men and women can cary out. These expectations of gender relations and public interaction act as a form of territorial gendering of space. The colective action literature often views social norms positively, yet overlooks the role of gender ideology. Agarwal (2000: 302) writes that a gendered reading of social norms reflects the "dark side of social capital." Norms of acceptable female behaviour— such as soft speech and deference to men—and perceptions held by males of female abilties together constrain women's behaviour and participation within male social networks and formal meetings and lead to women's suggestions not being heeded within groups in which men predominate. Instances of men's reluctance to yield their 'territorial claims' to women might be found in the Farmer's Association of Vietnam. Agarwal found that informal group arangements beter responded to women's time constraints and, by ofering a child-friendly atmosphere at meetings, faciltated women's need to respond jointly to domestic and community responsibilties. In India, since their physical mobilty is more restricted than is men's, women have a greater need to build social capital through localized networks and everyday forms of cooperation. Agarwal noted that women's forest protection groups emerged as an extension of everyday social networking and tended to lack authority and to be sporadic and situation specific. This informal nature of women's groups contrasted with men's formal forest protection groups that had an authority structure linked to the state or vilage. Comparing the nature of colaboration in agricultural tasks, Agarwal again found striking distinctions. Women's agricultural tasks were frequently performed cooperatively through labour exchange systems, while men's tasks—ploughing, threshing, irigation—were more prone to be done alone or with few other people. House construction was one male task that involved labour exchange, but it took place less frequently than the colective tasks that women caried out. These observations and circumstances might be constructively compared to those in Vietnam where women face fewer sanctions on their mobilty in the public sphere, in markets, and in the cash economy. In additon to women's informal networks created through these and other channels, women's formal networks—or associational capacity—can be further reinforced through the Vietnam Women's Union as a potential vehicle for colective action, albeit circumscribed within a certain mandate and political alegiance. In distinguishing gendered formal and informal networks in the quasi-domestic and public spheres, Lowndes (2000: 534) reminds us that: "Feminist political theory has long focused on what should become a central issue for the social capital debate -61 that is, the relationship between the 'smal democracies' of everyday life and the 'big democracy' of political parties and organized government." Having reviewed debates on social capital and gendered social capital, I am concerned about the vague use of the term social capital. Referencing Putnam, Lowndes (2000: 533) notes that "High levels of social capital are associated with high-performing democratic institutions, and economic success." Yet, Agarwal and others document the existence of social capital among the poor, arguing that women, poor households, and other such disadvantaged groups dependend extensively on networks. What is not addressed is why, if social capital leads to economic success, do these groups remain poor and disadvantaged? It is here where the distinctions introduced by Woolcock—between bonding, bridging, and linking social capital— become most relevant. What needs to be underlined is that the networks among women and the poor (e.g., for child care or labour exchange) reflect mostly bonding social capital. This can be contrasted with the networks of some beter situated individuals, which reflect 'higher levels' of (bridging and linking) social capital to facilitate their economic advancement (e.g., by granting them information on economic opportunites). My second concern relates to AgarwaTs discussion of gender diferences in values and motivations around social networking or social capital. Agarwal noted that factors that facilitate colective action include trust, reciprocity, 'density of social ties,' and prior history of cooperation. She further observed that "moral norms, social values of empathy, and trust play an enabling role in enhancing cooperation and undercut the tendency to free ride" (Agarwal 2000: 295-6). Agarwal then turns to the identification of gender-specifc aspects of cooperation, complementarity, coexistence, competion and conflict. In her analysis of forms of women's involvement in environmental action, she observed greater cooperation among women. Such cooperation, for Agarwal, can be linked to the dependence of Indian women on social relationships with other women; their informal social networks are crucial. Agarwal takes her observation on the distinct character of men's and women's support networks a step further. Noting that in India women's networking more often crosses class divisions than do men's, Agarwal (2000: 295) draws the controversial conclusion that the lower degree of divisiveness among women's groups and "the greater permeabilty of women's networks across class lines, make. for beter prospects for group action among women." Women are thus portrayed as beter at conflict resolution and group functioning due to their everyday strategies and group interaction with other women and their fear of sanctions and 62 isolation if excluded from the group. Having fewer 'exit options' makes women less prone to free-riding and more inclined to cooperation. Rather than idealizing cooperation among women as part of women's essential 'nature,' Agarwal's analysis situates the character of women's networks for colective action within a context of women's conditons and positons in the social and economic hierarchy. Yet, what seems to be insuficiently addressed in this analysis is the extent of vertical social capital within women's networks. Are women in India realy less connected than men to local power structures, as Agarwal suggests? Does this lack of connection increase cooperation among women? These are questions that cal for further empirical and comparative study. Despite the chalenges posed by gender ideology, Agarwal remains optimistic: "these constraints are not immutable and much depends on building up women's bargaining power vis-a-vis the State, the community and the family" (Agarwal 2000: 304). She asserts that to change rules, norms and perceptions, to enhance group strength and have women's interests be expressed and heard, a critical mass must be achieved. Agarwal wonders whether women's informal (forest protection) groups could be formalized and empowered with the authority that men's groups have. She warns that neglecting gender can negatively afect analysis and policy on colective action. Losing opportunites to promote colective action among women and keeping women out of decision-making bodies, such as community resource management institutions, can lead to ineficiencies of many kinds: "rule enforcement problems, information flow imperfections, inaccurate assessments of resource depletion, problems in catching transgressors, unsatisfactory conflict resolution, non-incorporation of women's specific knowledge of species, and non-recogniton of gender diferences in tree-species preferences" (Agarwal 2000: 305). These factors can afect short- and long-term institutional and environmental sustainabilty. A further issue that has received little atention thus far in debates over social capital is the methodological implications. Lowndes (2000: 536) ofers the suggestion that "Despite its association to date with game theory and abstract modeling, the social capital debate could make more use of qualitative case studies and individual 'life histories.'" I hope my analysis wil go some distance in this direction. Gender, Property, and Land Rights "Economic analysis and policies concerning women have long been preoccupied with employment, to the neglect of a crucial determinant of women's situation, namely, the gender 63 gap in command over property" (Agarwal 1994b: 1). In part as a contribution to emerging feminist geographies, a key objective of this dissertation is to demonstrate the gendering of instiutions and to understand how shifts in institutional structures shape new paterns of access to, and exchange of, land. Land is a special form of property, linking economic, cultural, political, and legal dimensions of social life. Property relations often provide the foundations of identiy formation (Hann 1998). What is often overlooked is the gendered nature of such property relations and the different relationship women have to property—fundamentally shaped by kinship systems. Compared to men, women's entilements are embedded to a far greater degree in family and kinship structures. Patrilocal residence paterns after mariage greatly determine the possibilty for women's asset accumulation in the form of land (Kabeer 1989: 9, cited in Jackson 1996). Commenting on the complexites of measuring gendered poverty, Jackson (1996: 496) observed that poverty is defined commonly in terms of household assets and resource access, land and livestock, for example, but since patrilny is extremely common, women have widely diferent property relations to men. Thus land ownership is seldom as defining of women's socioeconomic positon as it may be of men's. Feminists have pointed out the considerable problems raised by defining a woman's class positon in terms of that of her husband or her father (Agarwal 1994b: 7). Her class positon is more open to change than that of her husband, and this can vary dramaticaly upon mariage, divorce, or becoming widowed. Secondly, women often do not own property themselves. Finaly, issues such as risk of domestic violence and responsibilties for domestic tasks and childcare cut across class boundaries. These observations cal into question land reforms and related approaches to poverty that emphasize the transfer of assets to the poor but overlook gender diferences. They "raise the question of whether the same policies to strengthen the positon of poor men can have the same impact on poor women" (Razavi 1998: ii). Razavi thus clarified that "the gender analysis of poverty is not so much about whether women sufer more from poverty than men, but rather about how gender diferentiates the social processes leading to poverty, and the escape routes out of destiution" (1998: ii). Commenting on the historical process of land enclosure in Britain and its implications for household formation, Li argued that the enclosure of fields in practice enclosed individuals within the household, provoking a reworking of gender and generational relations: According to Humphries (1990), prior to land enclosure women made substantial contributions to the conjugal economy through grazing and gathering activities on the vilage commons. Land privatization increased women's dependence upon a 64 husband's wages, and therefore disciplined women and subjected them to male authority. (Li 1996: 262) Policies in socialist and capitalist countries alike have been equaly implicated in neglecting the importance of women's rights to land. For communist parties, addressing the issue of access to land from a gender perspective was considered potentialy divisive within the overal objective of unitng proletarian interests against large landowners. In societies which underwent socialist revolutions, while private property ownership was legaly abolished, control over wealth generating property remained mainly with men; any positve efects on gender relations that could have stemmed from the change in ownership, if accompanied by gender-egalitarian mechanisms of control, thus went unrealized. (Agarwal 1994b: 7) Communist parties reproduced the notion of 'peasant' as a male category, again erecting blinders to an examination of how women's access to land was generaly mediated by men (Agarwal 1994a), be it their husband or leaders of work brigades, communes, and other levels of administration. Moreover, in the absence of wel-defined policy, women's labour intensified under socialist policies in many countries, thereby subsidizing economic development (Crol 1981). Factors perpetuating gender inequites in command over property can be social, administrative, and ideological (Agarwal 1994b), reflecting the need for change on each of these levels. Assumptions relating to women's capabilties, needs, and roles can prevent the implementation of progressive laws. Ideologies can be shaped and controled by multiple forms of media, education, and religious establishments (Agarwal 1994b: 7). Despite progressive gender-sensitve laws for land reform in South Africa, for instance, implementation faltered. This was due to the lack of guidelines on how to ensure that women are not marginalized in the land reform processes. Lack of policy direction and training - as wel as trainers - is aggravated by low levels of sensitivity to gender dynamics among officials, al of which weakens capacity to fulfil the gender goals of the land reform programme. .. .Moreover, the climate on the ground may be hostile to wel-intentioned interventions. (Walker 1998: ii) Gender ideology and power relations constrain opportunites for men and women and shape diferential entilements to property. The value of an entilements approach for an analysis of gender and poverty is captured wel by Kabeer: By encompassing both the outcomes of deprivation as wel as their underlying causes, such an approach draws atention to issues of equity and justice as wel as to basic needs and welfare. It takes us beyond an economistic focus on ownership and exchange to socialy constructed definitions of who is entitled to what and on what 65 basis.. it also shifts atention away from a static view of poverty - poverty as an end-state - to a more dynamic concern with the processes of exclusion, inclusion and marginalisation which are set in motion by shifts in the configuration of entilement relationships within which people define goals and devise strategies and which place some groups of people at an entilement disadvantage in relation to others. (Kabeer 1997: 4) My analysis interprets new land relations and property rights regimes emerging in the context of decolectivization in terms of the interplay of diverse forms of formal and informal institutions: state policy, market dynamics, and gender, household, and kinship relations. Elements of ethnicity, community, and migration and setlement history further play a role in explaining new vulnerabilties. In this analysis, gender is seen as culturaly, historicaly, and geographicaly embedded at various scales of analysis. Gender inequalities can manifest both within the family-household, ethnic group, and community, in market relations, and in the application of state policy. As demonstrated in Chapter Five, men can also be disadvantaged by particular constiutions of kinship relations (such as being sons of second wives), highlighting the relevance of studies of masculinites within gender analyses. Gender and development analyses often consider women's roles within the household, but they less frequently take into account broader kinship relations as an important context in which gender relations are situated: By and large, while looking at the situation of women either at the macro level or at the micro level, feminists have not shown a clear realization of the signifcance of varying paterns of kinship for understanding gender relations and explaining disparaities, inabilities, and exploitation. Even in the discussion of the family as the seat of oppression, the wider context of kinship has not been considered. The term patriarchy, which has been used rather indiscriminately, does not generaly indicate any relationship to specifc kinds of kinship organization. (Dube 1997: 159, n.l) Too few feminist analyses chalenge the assumption of "universal, cross-cultural and ahistorical female disadvantage. [and] thereby both ignore specific areas of male disadvantage and miss the insights that folow from examining the spaces in which women are not disadvantaged" (Jackson and Palmer-Jones 1998: 27). Addressing the complex gender and poverty interlinkages without resorting to over-simplifcations cals for a careful and diferentiated analysis of vulnerabilty among particular social groups, and particular kinship systems. Rather than talking of patriarchy in an overly generalized sense, Dube (1997) points to how the structural and cultural components of specifc kinship systems are comprised of complex institutions. She identifies three main types of kinship systems: patrilineal, matrilneal, and bilateral—in which both parents are relevant for determining kinship and claiming rights to resources. 66 Dube (1997) suggests that the neglect of kinship in gender analyses may be due to the assumption that kinship is deemed irrelevant or immutuable. Yet, kinship exerts strong forces in determining the alocation of resources, production relations or gender ideologies. Gendered subjects are actors in maintaining and reproducing social systems. Kinship systems are not rooted in nature but constructed. They shape notions of men's and women's capabilties and entilements and also shape how gender roles are conceived and practiced. Kinship is further linked to religious beliefs, social norms, economic systems of production, and the political sphere. This conceptualization of kinship and gender thus considers both material and ideological aspects as interlinked. Compared to patrilneal paterns, bilateral inheritance paterns in Southeast Asia and parts of India are associated with greater autonomy, freedom of movement and public interaction of women, plus increased social independence and relative equality in marital relations. They are known for their active roles in income-generating activities, particularly in smal trading of agricultural produce, food items, and other goods, in both urban and rural areas. Southeast Asian women are said to have good business acumen, frequently owning and managing micro-enterprises. Their dominant presence in the economy is also related to their significant role in decision-making and control over the budget within the household. Dube (1997: 47) associates Southeast Asian women's active engagement in income-earning activities with "their freedom of association, their ability to migrate (often leaving children behind), the support of their kin, their hold over resources, and their rights over space." However, such paterns retain a number of characteristics of gender inequality (Agarwal 1994b): • they do not imply a more equal gender division of labour, particularly with respect to domestic work and child care so often considered the domain of women • they are not associated with women having sexual mores equivalent to men's • customary instiutions with juridical power were controled by men, and • the managerial authority over land in societies of matrilneal and bilateral kinship tend to remain in the hands of men, in their positons as husbands, fathers, and brothers of women Agarwal found that in systems of matrilneal inheritance, there was a greater divergence by gender between property ownership and its control to the extent that women owned property, but men efectively controled it. By contrast, in patrilneal systems there was greater convergence, with men owning and controling property. In other words, land ownership rights confer significant benefits on women, but women frequently continue to be excluded from the 67 management or control of the land they own, and from the instiutions of public authority to defend these rights. Thus, the arenas of contestation over effective land-rights for women wil therefore need to extend much beyond the courtyards of the household to encompass complex instiutions of community and state—the arenas where legal, social and political rules are made and unmade. (Agarwal 1994b: 28; emphasi  in original) In interpretign kinship and gender relations, residence after mariage is a key consideration. Residence can be diferentiated in terms of patri-virilocal, matrilocal, uxorilocal and neolocal paterns,12 and endogamous vs. exogamous mariage (mariage between two people within the same or diferent vilages). Residence paterns can have significant implications for women's accumulation of land assets and intra-household bargaining power and autonomy. "Residence is a material as wel as an ideological expression of principles of kinship.. Although the nature of the control that is exercised over women varied with household compositon, the ideology of patrilineal, patri-virilocal residence governs a woman's life" (Dube 1997: 93), thereby subjecting women more than men to particular gendered vulnerabilties. In cases of patri-virilocal and exogamous residence (in which a woman moves into a new vilage upon mariage), women are unable to inherit land and can only inherit mobile assets or must sel or cede land to male relatives upon moving out of the vilage. In addition, maried women in patri-virilocal residence in a new vilage generaly have less intra-household bargaining power than do women living in their natal communites. In contrast, in bilateral systems and instances of vilage endogamy, women retain and utilize social networks with their relatives and residence is more flexible. Linking kinship and residence paterns with women's bargaining power, Ireson (1996: 231) observed that "Lao women exercising more intra-household power are often those living in their natal communites, and old women in established and prosperous familes." In contrast, Khmu women in a patri-virilocal system are "hampered by their weak positon as in-marying spouses." Links can further be established between matrilocality and the education of girls: Ethnic Lao girls [in matrilocal residence] have been more likely than girls of other ethnic groups to take advantage of available schooling, while few Hmong girls do so [in patrilocal residence]. The schooling of Khmu and Hmong girls is of low priority for familes because (1) unmaried girls are valuable labourers and (2) any investment in a girl's education is lost to her natal household since her labor and education benefit her husband's family, not her natal family. Educational investment in girls would only pay of if educational costs could be recouped in an increased bride price. So far, education does not seem to enter bride price negotiations, and bride prices have See glossary for definitions of these terms. 68 reportedly been declining under pressure from local government, particularly during the socialist period. (Ireson 1996: 228) Land rights for women—a critical aspect for addressing poverty as wel as unequal gender relations—often tend to go unacknowledged. Shortal (1999) argues that in processes of formalizing land rights in many countries, the state tends to uphold men's customary access to land. Moreover, neoclassical economic models have tended to ofer little analysis of gender-diferentiated rights to property and "virtualy no atempt to examine the way in which gender relations shape and are shaped by the institutional environment" (Evans 1993: 25). Evans explains the rationale for this within neoclassical economics as folows: The fact that it is women with fewer claims to resources or rights to excludable property is... immaterial. What maters is the way in which an individual with a given set of endowments behaves within the sphere of exchange. (Evans 1993: 25, emphasi  in original) Thus no questions are raised about why some individuals from the outset have a diferent given set of endowments. Addressing this cals for an analysis of systematic gender-diferentiated assymmetries in bargaining power, linked to an understanding of the institutional environment, part of which is shaped by ideology. The acknowledgement of land rights for women could spur shifts in multiple domains— in the household, community, market, and at various levels of government—as Agarwal's (1994) work in South Asia has forcefuly demonstrated. Agarwal (1994a) identified three primary arguments for women to have land rights that are distinct from the rights of their husband or household. First, the welfare argument underlines how land serves as security against poverty. Direct access to land and other productive assets—not merely access mediated through a husband or other male family members—can significantly decrease a woman's risk of poverty and influence the wel-being of her children. As emphasized above in reference to residence paterns, land can further be used as a bargaining tool in both family and outside institutions. This last point highlights the relevance of colective actions and social networks among women. Second, the efficiency argument argues that land rights can make women more economicaly productive by facilitating their access to credit, technology, and information. Ignoring women in agricultural research and technology development, Jiggins (1986) suggests, can restrict agricultural output and welfare to below its potential. Beneficial environmental consequences can further result from women obtaining land rights and gaining greater determination over land use. Agarwal (1994a: 37) posits that, depending on the local division of labour for tasks of fuelwood and fodder colection, women may choose to favour planting trees 69 for these purposes rather than other crops that could be more damaging to the soil. Third, the equity and empowerment argument shows how acquiring land is central not only for improving a woman's economic circumstances in absolute terms but also for negotiating more equitable relations with men. For example, a woman is less likely to be thrown out of her house if she owns the land that her household depends upon. In this way, empowerment within the household is linked to economic equality. Having her own recognized rights to land could stand a woman in beter stead for bargaining with employers, given that she has a stronger fal-back position, with land serving as a social safety net. Development policies or programs often draw upon justifications of eficiency on the implicit premise that women constiute an underutilzed human resource. Policies deriving from such an approach have thus tended to add to demands on women in order to increase the eficiency or efectiveness of a program. An instrumentalist argument—that agricultural production wil  be improved by increasing women's participation—bypasses addressing gender inequalities. But while the eficiency argument contains implicit ideological dilemmas for those whose primary concerns are equity and empowerment, the strategic merits of exploitng this motivation should not be overlooked. Critics of the eficiency argument may not recognize the potential of meeting dual objectives through policies emerging from this line of 'eficiency' thinking. These contradictions can be used to strategicaly push forward a feminist development agenda. Entilements to land can be mediated by various institutions. These may be market-based, instiutions based on the formal legal system (or state alocations), or instiutions beyond the market and legal system, such as kinship networks, customary law, social conventions, and norms (Leach et al. 1999). An example of the later could be land rights derived from common property owned by a clan or vilage. Instiutions beyond the market and legal system were not originaly part of Amartya Sen's formulation of entilements theory. Their inclusion has been an important amendment of feminist analyses (Jackson 1998: 73). The case study in Chapter Five concentrates on inheritance and kinship-mediated arangements that influence gendered entilements to land. Distinctions between concepts of land access, land rights, and control over land are often unclear. Having access to land, either through use-rights or informal concessions granted by acquaintances, is not equivalent to holding land rights. Access to land refers to the ability to make use of it without necessarily owning it. Rights to land—legaly and socialy recognized and enforceable by an external legitmated authority—can be based on formal or customary 70 possession (Agarwal 1994b: 10). Agarwal highlights not only who owns land but also who controls it, since ownership does not guarantee control. Control of resources refers to command or decision-making over the resource and the benefits that derive from it. Agarwal further points out the tendency to equate property rights merely with legal rights, and the need to recognize the divergence between a woman's legal right and the social recogniton of it. She underlines the persistent gap between women's legal rights (through state laws on inheritance or divorce) and their actual ownership of land—the gap between de jure and de facto rights. These issues wil be revisited in the analysis in Chapter Five. Post-socialist Agricultural and Property Rights Restructuring Although little recognized, al socialist economies entailed some degree of market activity, reflecting a symbiosis of public and private systems (Ray 1996: 9). Despite atempts to limit or abolish them, private plots remained on state farms and colectives in the USSR, China, Vietnam, and elsewhere. In Poland, where 89 percent of agriculture remained privately owned, the average landholding was five hectares, compared to 5,000 in the Soviet Union. Since the New Economic Policy in the 1920s in the USSR, increasing the scope for market alocation was a convenient channel available to compensate for problems in the command economy. Yugoslavia experimented with a model of self-management and state co-ordination but with extensive markets and competion between worker cooperatives. In the Hungarian New Economic Mechanism that was introduced briefly in 1968, the state retained control of capital goods, distribution, infrastructure and agriculture, but alowed the paralel development of a market economy (Ray 1996: 121-123). To visualy capture these symbioses in past and curent periods, Varis (1998) diagrammed the articulation of three modes of production over time: the 'logic of socialism' (colective production), everyday self-suficiency (household economy), and the new market economy (see Figure 2). 71 Figure 2. The Articulation of the Three Modes of Production LEVEL L O C A L EVERYDAY SELF - SUFFICIENCY G E N E R A L LOGIC OF SOCIALISM NEW MARKET ECONOMY NOW TIME R E S T R U C T U R I N G Source: Varis (1998: 157). A more detailed matrix of the trade-ofs of alternative policy stances vis-a-vis the roles of state and peasantry, as elaborated by Post and Wright (1989), is presented in Table 5. In describing the changing relation between peasants and the socialist state in Poland, Kocik (1996) talks of a shift from 'repressive tolerance' to 'oppressive freedom.' Repressive tolerance refers to elements of political and administrative repression of peasant agriculture and atempts to subordinate this sector to the state-directed economy. Yet, especialy in situations of food shortage, peasant agriculture was tolerated because it constiuted a primary source of agricultural produce. Oppressive freedom, by contrast, reflects the situation in which new economic pressures and competion arise despite political and administrative pressures having been removed. Paralels can be drawn between Vietnam and Poland in Kocik's (1996) observation that the prospects for private farming improved as the food situation deteriorated. Given the ideology of egalitarianism, however, tolerance was limited, and private farming and landholdings were restricted. Systems of distribution were state-controled and prices fixed. Noting that "Individual farmers were the only major group maintaining a relatively independent economic base in the communist system" (Kocik 1996: 118), Kocik points out the paradox that peasants were the greatest oppositon to the state in the colective period, but in the transition to the market, they sufered the most, as they had no state safety net (as workers did) to protect them. [Therefore] the most important and largest group of private owners in the communist economy, in which family farming was treated as a residue of a market economy, has become one of the most important forces caling for state intervention within the market economy game. (Kocik 1996: 126) 72 Table 5. A Trade-off Matrix of Alternative Policy Stances for Addressing Underdevelopment Performance Criteria Policy Stance Producer power Consumer power Quantity of products Availability of products Quality of products Full employment Income inequality Regional inequality Autonomous civil society Administrative allocation ? - + - - + + + -Market allocation ? + ? + + - - - + Rapid industrialization - - ? - - + - - -Collectivized agriculture - - ? - - + + ? -Peasant agriculture ? + + + + - - - + Closed economy ? - + - - + + + -Open economy -• + ? ? ? - - - + Vanguard party - - + ? ? + + + -Source: Post and Wright (1989: 172). Eastern Europe and Russia have tended to experience more dire social and economic consequences of their restructuring programs. Kornai's recipe for democratization, privatization, decentralization, and price liberalization—in the form of 'shock therapy'—took insuficient account of the social dimensions of privatization. Folowing decolectivization in Russia, Hungary, and Bulgaria, for example, the agricultural sector became dichotomized between a decapitalized subsistence agriculture and a nominaly privatized but relatively unchanged large-scale sector of farms, often owned by their former managers (Szelenyi 1998). On the other hand, in its process of re-peasantization, Vietnam seems to have been too poor to permit such a dualism of organizational forms to emerge. It thus enjoyed the 'advantages of backwardness' and retrospect policy-making. In 'transitional' economies, changes have often taken place more rapidly in agriculture than in other sectors (Neil and Tykkylainen 1998b). Even so, the inter-country variations in outcomes of decolectivization and post-socialist land restructuring have been significant (see Table 6). Bulgaria, Czech Republic, Slovak Republic, Romania, Estonia, Latvia, and Lithuania adopted policies of direct land restiution to former owners. In Bulgaria's land restitution, to determine land alocation and land rights, less emphasi  was given to who works the land than to ancestral claims, to the exclusion of 'new' (post-1944) immigrants (Kanef 1998). Some other 73 countries folowed an indirect restitution, with compensation paid for lost land. In Hungary, vouchers for auctions were issued, enabling one to purchase land. Other countries undertook more egalitarian alocations of land formerly under state farm or colective management. Three elements appear to distinguish the agrarian sectors from one another in diferent contexts: scale of operational unit, pace of reforms, and form of property rights. Each country sought to balance these factors with social equity and degree of protectionism and intervention, versus the reign of the free market. As shown in Table 6, in some states (e.g., Latvia and Slovenia), nearly al land is now privately managed in smal-holdings, while in others (e.g., Slovak Republic, Belarus, Russia, Azerbaijan, Tajikistan and Uzbekistan) a large percent of land remains under the administration of large-scale state farms and colectives. Table 7 demonstrates the shifts through the 1990s in average size for various farm types— colective/cooperative, state, new corporate, and individual—in a number of Central and Eastern European and former Soviet republics (OECD 1999). As can be seen, the reforms in organizational structure of agricultural production have been much more dramatic in some countries than in others. 74 Table 6. Characteristics of Agricultural Decollectivization and Distribution of Farm Land in Central and Eastern Europe and Former Soviet Republics a, M-c O "53 u > a "c3 a o nershi ability S: llectiv jperati ite farr porate ms (% dividu ms (% 3 o o U o o TO on u o S3 c u c/i 10 OS Land type Trans use ri Pre-1990 1998 Pre-1990 1998 1998 Pre-1990 1998 CEECs:* Albania no priv. yes 74 - 22 20 - 4 80 Bulgaria yes priv. yes 58 42 29 6 - 13 52 Czech yes priv. yes 61 43 38 2 32 0 23 Republic Slovak yes priv. yes 69 60 26 15 20 5 5 Republic Hungary no2 priv. yes 80 28 14 4 14 6 54 Poland no1 priv. yes 4 3 19 7 8 77 82 Romania yes priv. yes 59 12 29 21 - 12 67 Estonia yes priv. yes 57 - 37 - 37 6 63 Latvia yes priv. yes 54 - 41 1 4 5 95 Lithuania yes priv. yes 61 - 30 - 33 9 67 Slovenia no1 priv. yes - - 8 4 - 92 96 Croatia no1 priv. yes 22 18 - - - 78 82 Ave. CEES 26 65 NIS:** Armenia no priv. yes 4 31 Georgia no priv. yes 7 23 Ukraine no priv. yes 7 17 Moldova no priv. yes 9 16 Belarus no priv.3 yes 7 12 Russia no priv. yes 2 11 Kyrgyzstan no publ. yes 1 25 Kazakhstan no priv.3 yes 0.2 13 Azerbaijan no priv. yes 3 6 Tajikistan no publ. no 2 4 Uzbekistan no publ. no 2 4 Turkmenistan no priv. no 0.2 0.3 Ave. NIS 4 14 *Central and Eastern European Countries **Newly Independent States 1 Poland and Yugoslavia had predominantly private ownership prior to 1990. 2 In Hungary vouchers were issued in place of restitution of historic boundaries. 3 Private ownership is limited to household plots only, with land for commercial farming state-owned. Source: OECD (1999: 62, 67). 75 Table 7 . Average Farm Size by Organizational Structure in Central and Eastern Europe and Selected Former Soviet Republics (hectares) Collective/cooperative farms pre-1990 1998 State farms pre-1990 1998 New corporate farms late 1990s Individual farms pre-1990 1998 Albania 1053 - 1588 - - 0.1 1.4 Bulgaria 4000 637 1615 735 - 0.4 1.4 Czech Republic 2578 1447 9443 521 690 5.0 34.0 Slovak Republic 2667 1509 5186 3056 1191 0.3 7.7 Hungary 4179 833 7138 7779 204 0.3 3.0 Poland 335 222 3140 620 333 6.6 7.0 Romania 2374 451 5001 3657 - 0.5 2.7 Estonia 4060 - 4206 - 449 0.2 19.8 Latvia 5980 - 6532 340 309 0.4 23.6 Lithuania 2380 - 1880 - 310 0.5 7.6 Slovenia - - 470 371 - 3.2 4.8 1990-91 1995-96 Russia 9500 8000 Ukraine 3700 3100 Moldova 2800 2000 Source: OECD (1999: 69). Decolectivization in diferent countries brought about curious paradoxes in the reconstiution of relations between state and private sector. It was politicaly popular in Vietnam, China, Bulgaria,13 and some other countries. By contrast, in Russia, East Germany, Czechoslovakia, and Hungary, peasants were not enthusiastic about privatization of farm land, and colective forms of ownership were maintained. In Estonia, when a new Farm Law posed threats to holders of private plots (alocated in the period of colective production) in the face of former owners' atempts to reclaim their lands, plot holders alied with state and colective farm leaders to resist the process (Abrahams 1996). These examples reflect contrasts in experiences of post-socialist reforms and atitudes among farmers toward private farming or land alocation. People are often more wiling to cooperate if they are not forced to do so. This discussion points In Bulgaria, in contrast to the popularity of privatization of land, the privatization of industrial enterprises was 76 to the relevance of considering local specificity in theories of the multiple paths in emerging trajectories in market socialist 'transformations.' In line with this perspective, Burawoy and Verdery (1999) argue against the concept of 'transition,' which is teleological, and instead prefer 'transformation,' which implies less linear a shift. Although I occasionaly fall back on using the notion of 'transitional' economies or countries, I acknowledge the problematic nature of this term. Neil and Tykkylainen (1998b: 313) identify a number of key factors influencing development in resource communites within societies experiencing major socio-economic transformation in the 1990s. Among them are restructuring of the private sector, in accordance with market ideology and economic deregulation; promotion of innovation and technical improvements; changing local economic policies towards new paterns of competion between localities; and sectoral shifts in the economy. One key process is privatization, particularly in 'transitional' countries. These processes of privatization and titling of land reflect a worldwide trend linked to mar