Citation
Social structure and livelihood adaptive strategies of urban farmers in response to land use change along the Yamuna River in Delhi, India

Material Information

Title:
Social structure and livelihood adaptive strategies of urban farmers in response to land use change along the Yamuna River in Delhi, India
Creator:
Cook, Jessica Ann ( author )
Language:
English
Physical Description:
1 electronic file (267 pages). : ;

Subjects

Subjects / Keywords:
Farms, Small ( lcsh )
Urban agriculture -- India -- Yamuna River ( lcsh )
Land use -- India ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Review:
The United Nations Human Settlements Programme (UN-HABITAT) recognizes community participation as an essential component of sustainable planning and development, but there are many challenges to engaging communities. Communities, often assumed to have common interests and goals, are dynamic, multi-cultural entities, and represent many voices. Social relations play an important role in how community members experience participation and representation and, consequently, engage or disengage from the process. However, the mechanisms driving interactions among people that influence beliefs and behaviors related to participation are not well understood. This dissertation used a case study of poor urban farmers facing land development to explore how power relationships impacted their planning participation and livelihood strategies. Using the Sustainable Livelihoods Framework (SLF), this dissertation applied anthropologist Eric Wolf's theories of power as a lens through which to understand the relationship between livelihood assets and livelihood strategies of Yamuna River floodplain farmers. The primary research question this dissertation addressed was: How does the social structure of the farming community situated along the Yamuna River floodplain in Delhi, India impact livelihood adaptation strategies in response to planned land use change? Using a mixed-methods case study design, I conducted semi-structured interviews with 165 households in Yamuna Khadir to understand power relations within the context of on-going land use changes. Project aims were to: (1) describe how Yamuna farmers exchanged knowledge and resources through social networks; (2) describe the beliefs and behaviors of Yamuna farmers related to land use planning; and (3) identify livelihood adaptation strategies of Yamuna farmers in response to planned land use change. Research methods included semi-structured interviews, mapping of social networks, and observation of social relations. This research linked theories, methods, and observations together in order to operationalize power. It expands on the concept of power using Wolf's theory that it is a multiple and complex phenomenon pervasive, but working differently, across multiple levels of human interaction. In doing so, it uncovers multiple ways that power relations can impact agency within a community and in the planner-public nexus. This interdisciplinary dissertation contributes new insights into how cities and planning organizations can work toward solutions to mitigate climate change impacts at the city-level (macro-level) and strengthen livelihood strategies of vulnerable and marginalized populations (micro-level). This case study supports the value of taking a bottom-up approach to identify barriers to community participation, improve community engagement, and achieve more sustainable outcomes.
Thesis:
Thesis (Ph.D.)--University of Colorado Denver.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: Adobe Reader.
General Note:
Department of Health and Behavioral Sciences
Statement of Responsibility:
by Jessica Ann Cook.

Record Information

Source Institution:
University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
911141870 ( OCLC )
ocn911141870

Downloads

This item is only available as the following downloads:


Full Text

PAGE 1

SOCIAL STRUCTURE AND LIVELIHOOD ADAPTIVE STRATEGIES OF URBAN FARMERS IN RESPONSE TO LAND USE CHANGE ALONG THE YAMUNA RIVER IN DELHI, INDIA by JESSICA ANN COOK B.A., West Virginia Wesleyan College, 2000 B.L.A/M.L.A, The Pennsylvania State University, 2 010 A thesis submitted to the Faculty of the Graduate School of the University of Colorado Denver in partial fulfillment of the requirements for the degree of Doctor of Philosophy Health and Behavioral Sciences Program 2015

PAGE 2

201 5 JESSICA ANN COOK ALL RIGHTS RESERVED

PAGE 3

ii This thesis for the Doctor of Philosophy degree Jessica Ann Cook has been approved for the Health and Behavioral Sciences Program by Deborah S. Main, Dissertation Chair John Brett, Examina tion Chair Mallika Bose Deborah Thomas Jeremy Nemeth February 28, 2015

PAGE 4

iii Cook, Jessica Ann (Ph.D., Health and Behavioral Sciences) Social Structure and Livelihood Adaptive Strategies of Urban Farmers in Response to Land Use Change Along the Yamuna River in Delhi, India Thesis directed by Professor Deborah S. Main ABSTRACT The United Nations Human Settlements Programme (UN HABITAT) recognizes community participation as an essential component of sustainable planning and development but t here are ma ny challenges to engaging communities Communities, often assumed to have common interests and goals, are dynamic, multi cultural entities, and represent many voices. Social relations play an important role in how community members experience participati on and representation and consequently, engage or disengage from the process H owever, the mechanism s driving interactions among people that influence beliefs and behaviors related to partici pation are not well understood. This dissertation used a case study of poor urban farmers facing land development to explore how power relationships impacted their planning participation and livelihood strategies. Using the Sustainable Livelihoods F ramework (SLF) this dissertation applie d anthropologist Eric Wolf's theories of power as a lens through which to understand the relationship between livelihood assets and livelihood strategies of Yamuna River floodplain farmers. T he primary research question this dissertation addressed was : How does the social structure of the farming community situated along the Yamuna River floodplain in

PAGE 5

iv Delhi, India impact livelihood adaptation strategies in response to planned land use change? Using a mixed methods case study design, I conducted semi structured interviews with 165 h ouseholds in Yamuna Khadir to understand power relations within the context of on going land use changes. Project aims were to: (1) describe how Yamuna farmers exchange d knowledge and resources through social networks ; (2) describe the beliefs and behavio rs of Yamuna farmers related to land use planning ; and (3) identify livelihood adaptation strategies of Yamuna farmers in respon se to planned land use change R esearch methods included semi structured interviews, mapping of social networks, and observatio n of social relations. This research linked theories, methods, and observations together in order to operationalize power. It expands on the concept of power using Wolf's theory that it is a multiple and complex phenomenon pervasive, but working differen tly, across multiple levels of human interaction. In doing so, it uncovers multiple ways that power relations can impact agency within a community and in the planner public nexus This interdisciplinary dissertation contributes new insights into how cit ies and planning organizations can work to ward solutions to mitigate climate change impacts at the city level (macro level) and strengthen livelihood strategies of vulnerable and marginalized populations (micro level). This case study supports the value o f taking a bottom up approach to identify b arriers to community

PAGE 6

v participation improve community engagement and ac hieve more sustainable outcomes. The form and content of this abstract are approved. I recommend its publication. Approve d: Deborah S. Main

PAGE 7

vi DEDICATION "If we need to destroy as part of our city building, we also need to heal." (Sandercock, 2003, p.222) I dedicate this work to my parents, Roger and Sus an, and to my family and friends. Their support and confidence motivated me when I need ed it. They taught me that it is less important that our actions make an impact than that we continue to act.

PAGE 8

vii ACKNOWLEDGEMENTS This research would not have been possible without generous support from the U.S. National Science Foundation Integrative Graduate Education an d Research Traineeship (NSF IGERT) Program (Award No. DGE 0654378), directed by Dr. Anu Ramaswami (2009 2013) and Dr. Chris Weible (2013 present). This research was also jointly funded by the U.S. Dept. of State and the Government of India through the Ful bright Nehru Fellowship. I am grateful for the opportunities and support provided by dissertation committee chair Dr. Deborah S. Main, and committee members Dr. John Brett, Dr. Mallika Bose, Dr. Deb Thomas, and Dr. Jeremy Nemeth. This dissertation is a te stament to the dedication of their time, encouragement, and diverse expertise. I also recognize and appreciate the scholarship and friendship cultivated by the faculty, staff, and students of the University of Colorado Denver Department of Health and Behav ioral Sciences and the Center for Sustainable Infrastructure Systems (CSIS) at the University of Colorado Denver. This research would not have been possible without a great many peop le and organizations. I thank Dr. Praveen Singh, my advisor at Ambedkar U niversity Delhi, and the administrators and staff at the Fulbright House in Delhi, India. I also thank my research assistants Harpreet Kaur and Priyanka Singh. Special acknowledgement goes to the farmers in Delhi who shared their knowledge with me

PAGE 9

viii TAB LE OF C ONTENTS CHAPTER I. INTRODUCTION................................................................................... 1 II. LITERATURE REVIEW!!!!!!!!!!!!!!!!!!!! 9 Planning for Sustainable Cities!!!!!!!!!!!!!!!!! 10 Participatory Planning: Discours e and Reality!!!!!!! .. 1 1 Participation and Representation !!!!!!! !!! !!! 1 2 Knowledge and Problem Setting!!!!!!!!!! !!! 1 6 Power Relations and Pathways to Action!!!!!!!!!!!! .. 1 9 Power Within, Power Without!!!!!!!!!!!!!! .. 1 9 The Disempowered!!!!!! !!!!!!!!!!!!! 2 5 Social Networks and Decision Making!!!!!!!!!!!!! ... 2 7 Research Gaps!!!!!!!!!!!!!!!!!!!!!!! .. 2 9 III. THEORETICAL FRAMEWORK!!!!!!!!!!!!!!!! ... 32 Political Ecol ogy Frames Social Vulnerability!!! !!!!!!! 32 Multi scale Human Environmen t Relationship !!!!!!! .. 32 Political Ecology!!!!!!!!!!!!!!!!!!!! 3 5 Vulnerability & Resilience!!!!!!!!!!!!!!!! 3 6 Unpacking the Nature of Power: Eric Wolf!!!!!!!!! 41 Power Relations and Livelihoods!!!!!!!!!!!!!!!! 44 Livelihoods as a Window!!! !!!!!!!!!!!!! .. 44 Sustainable Livelihoods Framework!!!!!!!!!!! ... 4 6

PAGE 10

ix Research Framework!!!!!!!!!!!!!!!!!!!!! 4 9 Power as a Lens!!!!!!!!!!!!!!!!!!!! 49 Social Network Theory!!!!!!!!!!!!!!!!! .. 55 Research Question!!!!!!!!!!!!!!!!!!! 58 IV. RESEAR CH DESIGN!!!!!!!!!!!!!!!!!!!!! 59 Research Site!!!!!!!!!!!!!!!!!!!!!!!! 59 Preliminary Fieldwork!!!!!!!!!!!!!!!!!! 59 Population C haracteristics!!!!!!!!!!!!!!!! 61 Method s !!!!!!!!!!!!!!!!!!!!!!!!!! ... 65 Unit of Analysis!!!!!!!!!!!!!!!!!!! ...... 6 6 Sample Selection!!!!!!!!!!!!!!!!!!! ... 67 Pilot Testing!!!!!!!!!!!!!!!!!!!!!! 68 Research Aim 1!!!!!!!!!!!!!!!!!!!! .. 69 Measures and Definitions !!!!!!!! !!!!! 70 Data Collection !!!!!!!!!!!!!!!!! 77 Research Aim 2!!!!!!!!!!!!!!!!! ......... .... 72 Measures and Definitions !!!!!!!! !!!!! 81 Data Collection !!!!!!!!!!!!!!!!! ... 83 Resear ch Aim 3!!!!!!!!!!!!!!!!!!!! .. 85 Measures and Definitions !!!!!!!!!!!!! 86 Data Collection !!!!!!!!!!! !!!!!! ... 88 Data Analysis!!!!!!!!!!!!!!!!! !!!!!!! .. 89 Data Recording and Digitization!!!!!!!!!!!!! .. 89

PAGE 11

x Coding Interviews!!!!!!!!!!!!!!!!!!! .. 89 Mixed Methods Analysis!!!!!!!!!!!!!!!! ... 90 Quantitati ve Analysis!!!!!!!!!!!!!!!!!! 92 Social Structure Livelihood Strategies Hypothesis !! ... 93 Land Use Influence Livelih ood Strategies Hypotheses... 94 Spatial Relationships!!!!!!!!!!!!!!! 95 Qualitative Data Analysis!!!!!!!!!!!!!!!! .. 97 V. FINDINGS .. !!!!!!!!!!!!!!!!!!!!!!!!! 99 Household Background Characteristics!!!!!!!!!!!!! 101 Family Size!!!!!!!!!!!!!!!!!!!!!! 101 Migrant Origin!!!!!!!!!!!!!!!!!!!!! 101 Ownership Status and Land Size!!!!!!!!!!!!! 102 T enure!!!!!!!!!!!!!!!!!!!!!!!! 103 Profess ion!!!!!!!!!!!!!!!!!!!!! ...... 103 Household Spe aker Attributes!!!!!! !!!!!!!! 105 Geographic Distribution of Household Characteristics!!!! 107 Aim 1: Social Network Characterization!!!!!!!!!!!!! 107 Power by Social Network Dimension!!!!!!!!!!! .. 108 Household Network Power !!!!!! !!!!!! ... 109 Community Network P ower !!!!!!!!!!!! .. 112 Outer Network Power !!!!!!!!!!!!!!! 116 Combining Three Dimensions of Social Network Power. 1 20 Social St ructure!!!!!!!!!!!!!!!!!!!! .. 122

PAGE 12

xi Aim 2: Land Use Change Bel iefs and Behaviors!!!!!!!!! .. 124 Knowledge of Land Use Ch ange in Yamuna Khadir!!!!! 125 Involvement in Land Use Planning !!!!!!!!!!!! 1 27 Perception of Land Use Influence! !!!!!!!!!!! 1 29 Structural Lev el of Power!!!!!!!!!!!!!!!! .. 131 Aim 3: Livelihood Strategies!!!!!!!!!!!!!!!!!! 1 35 Livelihood Indica tors!!!!!!!!!!!!!!!!!! .. 1 36 Benefits to Farming along the Yamuna ... ................................... 1 37 Livelihood Strategies!!!! .. !!!!!!!!!!!!! ... 1 38 Linking Resea rch Aims!!!!!!!!!!!!!!!!!!!! 141 Background Ch aracteristics!!!!!!!!!!!!!!! 142 Social Structure Livelihood Strategies Hypothesis!!!!! .. 1 44 Land Use Influence Livelihood Strategies Hypotheses!!! ... 14 6 Social Structur e and Behavior!!!!!!!!!!!!!! 150 Weak Social Structure: Strong Household !!!!!! 151 Weak Social Structur e: Strong Community !!!!! ... 153 Weak Social Structure: Strong Outer!!!!!!!! .. 155 Weak Social Structure: Weak Networks!!!!!!! 157 Strong Social St ructure: Strong Networks!!!!!! 159 Social Structure and Behavior Summary!!!!!! ... 161 VI. DISCUS S ION!!!!!!!!!!!!!!!!!!!!!!!! 164 Aim 1: Social Structure!!!!!!!!!!!!!!!!!!!! .. 164 Measuring Power Relations!!!!!!!!!!!!!!! 164

PAGE 13

xii Aggregating Power Relations!!!!!!!!!!!!!! .. 166 Implications!!!!!!!!!!!!!!!!!!!!!! 169 Aim 2: Planning Beliefs and B ehaviors!!!!!!!!!!!!! .. 171 Outside the Decision Making Space!!!!!!!!!!! ... 172 Unequal Access to Participate!!!!!!!!!!!!!! 173 Implications!!!!!!!!!!!!!!!!!!!!!! 175 Aim 3: Livelihood Strategies!!!!!!!!!!!!!!!!! .! 176 Maximizing Options!!!!!!! !!!!!!!!!!!! 177 Implications!!!!!!!!!!!!!!!!!!!!!! 179 Linking Aims!!!!!!!!!!!!!!!!!!!!!!!!! 179 Embedding Beliefs and Behaviors!!!!!!!!!!!! .. 179 Reinforcing Marginalization!!!!!!!!!!!!!!! .. 181 Implications!!!!!!!!!!!!!!!!!!!!!! 184 Communi ty Representation!!!!!!!!!!!!!!!!!! .. 184 Theory to Prac tice!!!!!!!!!!!!!!!!!!!!!! .. 187 Revisiting Wolf!!!!!!!!!!!!!!!!!!!!! 187 Linking Modes of Power!!!!!!!!!!!!!!!!! 190 Research Li mitations!!!!!!!!!!!!!!!!!!!!! 191 Research Design!!!!!!!!!!!! !!!!!!!! 191 Methodological Challenges!!!!!!!!!!!!!!! .. 192 Analysis: Data Limits!!!!!!!!!!!!!!!!!! 195 Future Steps!!!!!!!!!!!!!!!!!!!!!!!!! 197 VII. CONCLU SION!!!!!!!!!!!!!!!!!!!!!!!! 200

PAGE 14

xiii Sustainable Cities and Pa thways to Action!!!!!!!!!!!! 200 A Framework for Power R elations!!!!!!!!!!!! .. 200 Beliefs, Behaviors and Agency!!!!!!!!!!!!! ... 201 Final Observa tions!!!!!!!!!!!!!!!!!!!!!! 203 REFERENCES! !!!!!!!!!!!!!!!!!!!!!!!!! ... 205 APPENDIX!! !!!!!!!!!!!!!!!!!!!!!!!!!! .. 214 A. Research Quest ions Guide!!!!!!!!!!!!!!!!!! .. 214 B. Operationalizing Ego Alter Power!!!!!!!!!!!!!!! ... 217 C. Characterizing Social Networks!!!!!!!!!!!!!!!! ... 229 D. Power by Social Network Type!!!!!!!!!!!!!!!!! 243 E. Examples of Livelihood Strategies!!!!!!!!!!!!!!! .. 246

PAGE 15

xiv LIST OF TABLES Table 4.1 Summary of research aims !!!!!!!!!!!!!!!!!!! 66 4.2 Matrix for interview question guide development !!!!!!!!!!!! 75 4.3 Interview questions measuring household background characteristics !! ... 78 4.4 Interview questions measu ring social networks !!!!!!!!!!!! ... 78 4.5 Interview ques tions measuring land use change!!!!!!!!!!!!. 84 4.6 Interview questions m easuring livelihood strategies!!!!!!!!!!... 88 5.1 Household background characteristics !!!!!!!!!!!!!!!! .. 106 5.2 Social netwo rk power !!!!!!!!!!!!!!!!!!!!!!!! 121 5.3 Reasons given for not being involved in land use activity !!!!!!!! .. 129 5.4 Livelihood indicators !!!!!!!!!!!!!!!!!!!!!!!! 136 5.5 Other home and/or land in rural village !!!!!!!!!!!!!!!! 136 5.6 Livelihood strategi es by category !!!!!!!!!!!!!!!!!! ... 138 5.7 Ordered logistic regression (OLR) of demographic characteristics and research variables: social structure, land use beliefs/behaviors, and livelihood strategies !!!!!!!!!!!!!!!!!!!!!!!! .. 143 5.8 Livelihood stra teg ies by social network strength!!!!!!!!!!!!. 145 5.9 Ordered logistic regression (OLR) of planning beliefs and behaviors and research variables: social structure, livelihood strategies, and plan to stay in Delhi !!!!!!!!!!!!!!!!!!!!!!!!!!!!! 148

PAGE 16

xv LIST OF FIGURES Figure 3.1 Diagram of the human environment relationship in the context of sustainable planning and development! !!!!!!!!!!!! !!! 33 3.2 Pressure and Release Model !!!!!!!!!!!!!!!!!! !! 37 3.3 Sustainable Livelihoods Framework (SLF)! !!!!!!!!!! !!! ... 47 3.4 Structural mode of power applied to the SLF 4 9 3.5 Organizational mode of power applied to the SLF!! !!!!!!! !! .. 51 3.6 Ego alter mode of power applied to the SLF!!!!!!!!!!!!! 52 3.7 Individual mode of power applied to the SLF!!!!! !!!!!!!! 53 3.8 Individual and ego alter modes of power applied to the SLF!!!!!! 54 3.9 Simplified diagram integrating Wolf's power and the SLF as conceptualized for this dissertation research !!!!!! !!!!!! ...... 54 3 .1 0 Example of geographically plot ted social networks! !!!!!! !!!! 58 4.1 Conte xt map of preliminary fieldwork!!!!!!!!!!!!!! !!! 60 4.2 Map of research site!!!!! !!!!!!!!!!!!!!!!!!! 61 4.3 Urban agriculture with newly constructed apartment buildings in the background; beneath active c onstru ction of a new metro line!!!! !! ... 64 4.4 Another site where urban agriculture faces newly constructed apartment buildings in the background; beneath new electricity towers!!!!! !! 64 4.5 Urban agriculture "integrated" with newly constructed (waste?) wat er pipes; removed from a new "public (but unpopulated) green space! !!!!! ... 64 4.6 Pilot test interview sites !!!!!!!!!!!!!! !!!!!!!!! 69 4.7 Examples of social network maps !!!!!!!!!!!!!!!!!! .. 92 4.8 Ego alter power mediates the relationship between socia l netw orks and livelihood strategies!!!!!!!!!!!!!!!!!!!!! !!! .. 94 4.9 Organizational power moderates the relationship between social netw orks and livelihood strategies!!!!!!!!!!!!!!!!!!!! !! .. 95 5.1 Map of interview sites!!!!!!!!!!!!!!!!!!!!!! .. 99

PAGE 17

xvi 5.2 Ma p of interviewed tenant farmers!!!!!!!!!!!!!!!!! 100 5.3 Map of migrant origin!!!!!!!!!!!!!!!!!!!!!! !! 102 5.4 Distribution of land size!!!!!!!!!!!!!!!!!!!!!! 103 5.5 Diagram showing number of households in and across different professions!!!!!!!!!!!! !!!!!!!!!!!!!! !! 104 5.6 Distribution of gend er and age of household speaker!!!!!!!!! 105 5.7 Ego alter mode of power applied to the SLF!!!!!!!!!!!!!! 107 5.8 Strong/weak household network power: power to influence/access livelihood assets t hrough hous ehold social network!!!!!!!! !!! 109 5.9 Hotspot anal ysis of household network power!!!!!!!!!!!! 111 5.10 Strong/weak community network power: power to influence/access livelihood assets th rough community social networks!!!!!!! !!! 113 5.11 Hotspot anal ysis of community network power!!!!!!!!!! !!! 116 5.12 Strong/weak outer network power: power to influence/access livelihood asset s through outer social networks!!!!!!!!!!!!!! ! 117 5.13 Hotspot analysis of outer network power!!!!!!!!!!!!! ! .. 120 5.14 Social structure!!!!!!!!!!!!!!!!!!!!!!!! ! .. 123 5 1 5 Organizational mode of power applied to the SLF!! !!!!!!! !! .. 124 5.16 Distribution of land development knowledge !!!!!!!!!!! ! 126 5.17 Distributi on of land use behavior!!!!!!!!!!!!! ! ! ! 128 5.18 Crops growing post metro construction!!!!!!!!!!!!!! ! 131 5.19 Individual mode of power applied to the SLF!!!!! !!!!!!!! 135 5.20 Future loc ations where farmers will look for more land!!!!!!! ! .. 139 5.21 Simplified diagram integrating Wolf's power and the SLF as conceptualized for this dissertation research !!!!!! !!!!!! ...... 15 0 5.22 Field Image 1 !!!!!!!!!!!!!!!!!!!!!!! ! ! 151 5.23 Case of weak social structure: strong household network overlaid on simplified conceptual diagram !!!!!!!!!! !!!!!!!!!! 152

PAGE 18

xvii 5.24 Case of weak social structure: strong community network overlaid on simplified conceptual diagram !!!!!!!!!! !!!!!!!!!! 154 5.25 Field Image 2 !!!!!!!!!!!!!!!!!!!!!!! ! ! 154 5.26 Field Image 3 !!!!!!!!!!!!!!!!!!!!!!! ! ! 155 5.27 Case of weak social structure: strong outer network overlaid on simplified conceptual diagram !!!!!!!!!! !!!!!!!!!! ! ! 157 5.28 Field Image 4 !!!!!!!!!!!!!!!!!!!!!!! ! ! 158 5.29 Case of weak social structure: no strong network s overlaid on simplified conceptual diagram !!!!!!!!!! !!!!!!!!!! ! ! 159 5.30 Field Image 5 !!!!!!!!!!!!!!!!!!!!!!! ! ! 160 5.31 Case of weak social structure: all strong network s overlaid on simplified conceptual diagram !!!!!!!!!! !!!!!!!!!! ! ! 161

PAGE 19

xviii LIST OF ABBREVIATIONS UN Habitat United Nations Human Settlements Programme DDA Delhi Development Authority The Society The Delhi Peasants Multipurpose Cooperative Society SLF Sustainable Live lihoods F ramework PAR Pressure and Release Model SNA Social network analysis Delhi NCT National Capital Territory of Delhi GIS Geographic information systems NCRPB National Capital Region Planning Board Act CWG 2010 Commonwealth Games OLR Ordered logistic regression WHO World Health Organization

PAGE 20

1 CHAPTER I : INTRODUCTION "In Foucault's analysis![t]he particular form modern power takes is centreless, it is capillary. Rather than located in the state, or in capital, power is a moving substratum of force relations, local and unstable![I]f power is anchored in the micro practices of everyday life, then that is also where oppositional politics needs to begin, with a deconstruction of the power relationships built into everyday practices, and a reconstructe d, political plannin g ( Sandercock, 2003, p.70 71 ) Conflicts that arise between planners and the public during the planning and development process are often about more than just the allocation and use of land, facilities, and other resources; they are also abo ut relationships which involve personality, politics, race, ethnicity and culture ( Sandercock, 2003 ) Sustainable planning and development is the process of working toward economic, ecologic, and social sustainability 1 but a s cities attempt to achieve sustainability goals in the face of urbanization, population growth, and c limate change, the greatest challenges are not in technical solutions but in addressing power dynamics underlying social relationships within and across multiple stakeholders. Stakeholders include individuals and groups who can affect or be affected by th e problem at hand. This includes the public, which means public inclusion specifically, affected communities will influence the sustainability of solutions. 1 The World Commission on Environment and Development states that development that meets the needs of the present without compromising the ability of future generations to meet their own needs'' is the foundation for creating sustainable cities ( The World Commission on Environment a nd

PAGE 21

2 Community participation is recognized by the United Nations Human Settlements Programme (UN H abit at ) as an essential component of sustainable planning and development ( Motasim et al., 2010 ) : "Participatory planning empowers communities and results in better design outcomes that are more responsive to the diverse needs of the different urban groups. Participation also ensures the relevance of plans when faced with limited resources and can als o increase effectiveness" ( U N Habitat, 2010, p. 19 ) T here are many challenges to engaging communities however, including the potential for conflict and mistrust, lack of authority or dominant voice and achieving adequate representation ( Irvin & Stansbury, 2004 ; Maginn, 2007 ) Communities, often assumed to have common interests and goals, are dynamic, multi cultural entities, and represent many voices. Furthermore, a community member may at tend planning meetings but not feel that he or she has a voice; presence is not the same as participation. On the other hand, participation or influence may come indirectly through representation ; i ndividuals are socially embedded, and inclusion can come from the perception that actions or outcomes of the group are aligned with one's values via representation. Social relations play an important role in how community members experience participation and representation and consequently, engage or disengage from the process H owever, the mechanism s driving interactions among people that influence beliefs and behaviors related to partici pation is not well understood ( Beebeejaun & Vanderhoven, 2010 ; Brownill & Parker, 2010 ; Leach, 2008 ; Yates, 2012 ) With out addressing the issue of power among different stakeholders, frameworks like that presented by UN Habitat may fall short in practice ( Dupont, 2007 )

PAGE 22

3 Equally important to participatory planning is the need to understand why poor and marginalized populations may not perceive a benefit in engagin g with the sustainable planning and development process and, as a result, may appear indifferent to the institutional systems in which they are embedded In setting the agenda for meeting sustainability goals, planners focus on issues of climate change an d urbanization, but communities are concerned with daily activities and livelihoods. This dissertation uses a case study of poor urban farmers facing land development to explore how power relationships impact their planning participation and livelihood st rategies. Fundamentally, this research focused on human behavior embedded in multiple levels of social engagement; a n individual may have the potential to act, but it is only by understanding which actions are possible, in which contexts, and with whom, t hat action is more likely to occur. Underlying the negotiation of social relations is the concept of power. Cultural anthropologist Eric Wolf theorizes that power is an aspect of all relations among people and that it works differently at different level s. This may be why decisions made at one level can compl e ment or conflict with decisions made at a different level. Wolf conceptualize d power as gains and/or losses through relations, and suggest ed that a catalyst wa s necessary for power to manifest C ity planning and development occurring along the Yamuna River floodplain in Delhi, India provided a catalyst to observe power relations as a n underlying driver of decision making related to planning participation and livelihood strategies Delhi is the se cond largest megacity in the world home to

PAGE 23

4 approximately 25 million people, and growing exponentially 2 3 Urbanization has led to steeply rising food prices due to greater food transport distances; growing development of urban and peri urban land is push ing agriculture farther from the city. Urbanization has also resulted in extreme pollution of the Yamuna River (the primary water source for the city) and construction of unauthorized housing structures that are particularly susceptible to increasingly ex treme annual monsoon events. The Yamuna River is planned for extensive environmental cleanup and various forms of development along the banks; however, achievement of these goals calls for a restriction on agricultural uses along the floodplain an importa nt source of food and livelihoods for local farmers and adjacent neighborhoods. The Delhi Development Authority (DDA) manages the city owned Yamuna River floodplain land A lthough the DDA has a policy in place for community engagement in development proj ects such as the one planned along the Yamuna River, historically, it has fallen short on public inclusion ( Ahmad, Balaban, Doll, & Dreyfus, 2013 ; Datta & Jha, 1983 ; Rao, 2010 ) At the time of this research project, the area along the Yamuna River was in a state of flux. In 1949, the Delhi Improvement Trust allot ted more than 5,000 acres of land along the Yamuna to Delhi Peasants Cooperative Multipurpose Society (the Society). The Society parceled out the land along the Yamuna to 2 United Nations Department of Economic and Social Affairs website: http://www.un.org/en/development/desa/news/population/world urbanization prospects 2014.html accessed 8/29/14. 3 In 2012, when the proposal for this disser tation was written, Delhi was the fifth largest megacity with a population of 23 million.

PAGE 24

5 Society members by giving them the right to cultivate the land 4 In 1957, the DDA w as created in response to rapid population growth ( Delhi Development Authority, 2012 ) Tasked with ensuring that development in Delhi adheres to an approved plan, the DDA creates and approves development plans in accordance with a city level Master Plan. In 1991, the DDA became the successor of the Delhi Improvement Trust and, consequently, the authority over the land along the Yamuna. The DDA's current Master Plan for Delhi 2021 is to make Delhi a global metropolis and a world class city, where all the people would be engaged in productive wor k with a better quality of life, living in a sustainable environment ( Delhi Development Authority, 2012 ) As the amount of land available for development in and around Delhi dwindles, the open fields of the Yamuna floodplain have attracted the attention of land developers ( Prashar, Shaw, & Takeuchi, 2012 ) The DDA is tasked with prioritizing development projects. During the summer of 2011, I was part of a team tha t undertook preliminary fieldwork to explore urban agriculture in Delhi, India ( Cook, Oviatt, Main, Kaur, & Brett, 2015 ) We found that benefits and barriers fa r m ers face d in an urban context were influenced by often conflicting social, environmental, and economic dynamics. Specifically, our research revealed that many urban farmers situated along the Yamuna River floodplain were aware of impending development and plans to remove them. However, farmers varied in the extent to which they were engaged in the planning process, as well as what they would 4 There is limited documentation on the Delhi Peasants Multipurpose Cooperative Society. The background provided in this paper is based on reviews of court rulings over land disputes between the DDA and various Yamuna tenants.

PAGE 25

6 do in the event that the land was impacted by development. Based on our findings, we concluded that cities and pla nners would benefit from greater understanding of both individual and collective behaviors of farmers as they respond and adapt to city level climate change planning. Using the Sustainable Livelihoods F ramework (SLF) this dissertation applie d anthropologi st Eric Wolf's theories of power as a lens through which I examined the relationship between livelihood assets and livelihood strategies of Yamuna River floodplain farmers. T he primary research question this dissertation addressed was : How does the social structure of the farming community situated along the Yamuna River floodplain in Delhi, India impact livelihood adaptation strategies in response to planned land use change? Using a mixed methods case study design, I conducted semi structured interview s with 165 households in Yamuna Khadir 5 to understand power relations within the context of on going land use changes. Project aims were to: (1) describe how Yamuna farmers exchange d knowledge and resources through social networks ; (2) describe the belief s and behaviors of Yamuna farmers related to land use planning ; and (3) identify livelihood adaptation strategies of Yamuna farmers in respon se to planned land use change To address these aims, research methods included semi structured interviews, mappin g of social networks, and observation of social relations. For the purpose of this research, social structure was conceptualized as the exchange of knowledge and resources through household social networks. And, livelihood adaptation 5 Yamuna Khadir is an urban farming community on the Yamuna River floodplain.

PAGE 26

7 strategies encompass ed short term (i.e. seasonal, monthly) tactics to either maintain or create new ways to support a viable livelihood. The exploratory nature of the research design meant that livelihood adaptation strategies emerged a posteriori through data collection; h owever, based on findings from preliminary fieldwork, the following hypotheses were used to guide research design and analysis: Hypothesis #1: If ego alter power within the social network is strong then there will be similarity in livelihood adaptive stra tegies. Hypothesis #2: If ego alter power within the social network is weak then there will be variability in livelihood adaptive strategies (i.e. there will not be a strong, collective response). Hypothesis #3: If organizational power is low (i.e. f armers do not feel they can impact the planning process) then farmers will plan informal livelihood adaptation strategies. Hypothesis #4: If organizational power is high (i.e. farmers feel they can impact the planning process) then farmers will plan forma l livelihood adaptive strategies. This dissertation contribute s to our understanding of community empowerment by expanding our knowledge o f why some populations may be

PAGE 27

8 more challenging to engage in sustainable planning than other s More specifically, thi s research fills three important gaps First, by examining power relations both inside and outside the planning process this study uncovers multiple ways that power relations can impact agency within a community and in the planner public nexus Second, this project fills an important methodological gap: studies that measure social capital, social relations, and social structure have been limited by the challenge of operationalizing the concept of power. This study expands on the concept of power using W olf's theory that it is a multiple and complex phenomenon pervasive, but working differently, across multiple levels of human interaction. Third, this research links theories, met hods, and observations together in order to operationalize power In t he fo llowing chapters I summarize current theory and literature pertinent to this research, layout the theoretical framework guiding research design describe research methods and data analysis, detail findings, and discuss implications limitations, and next steps.

PAGE 28

9 CHAPTER II : LITERATURE REVIEW This chapter and the next draw upon literature across multiple disciplines including sociology, anthropology, geography, urban planning, and environmental sciences. I build an argument for my dissertation research t hrough a critical evaluation of pertinent theory and research in the areas of participatory planning, vulnerability and resiliency and social network research This research is primarily concerned with human behavior embedded in multiple levels of influe nce and decision making, but within the context of participatory planning, which is where I begin. The first section of this chapter addresses the importance of community participation in sustainable planning and development, and the persistent chasm be tween the rhetoric of public participation and the reality of engaging and empowering communities. Inclusion of every stakeholder is rarely possible, and the nature of representation or who is acting on behalf of whom has been an important topic in the li terature. Representation can enable agency and empowerment of excluded or marginalized groups, but power relations often bias who is deemed appropriate to participate as well as who actually shows up to participate. Power relations impact the equitable d istribution of knowl edge and create a divide between who is involved in problem setting versus problem solving an important distinction in determining the scope of possible solutions The second section more deeply addresses the concepts of power, empow erment, and those disempowered in the context of participatory planning.

PAGE 29

10 Despite efforts to shift power from planners to the public, planners still hold the power to define who qualifies as "the public." It becomes important to contrast the spaces that a re created through invitations to participate and those that people create for themselves ( Cornwall, 2008 ) In view of that, I shift focus and look at how disempowered and marginalized groups create spaces and draw power from social relations through informal means often in the pursuit of sustaining livelihoods through reliable and consistent means to meet daily needs. The third section of this chapter acknowledges that people are socially embedded, and that agency in the planning process can take a variety of forms. Power and influence impact the control of decision making and, critically, even the awareness and framing of underlying problems. Social networks provide a way to look at the role of power in the agency structure interrelationship I summarize a nascent but growing body of research employing social network analysis in the planning literature, and identify the dearth of studies that specifically measure power as a driving mechanism for social relations. Finally, I highlight the need for a theoretical framework to organize the concepts introduced through the literature so that they can be operationalized and measure d The next chapter describes development of the theoretical framework applied in this dissertation. Planning for Sustainable Cities The myriad issues and conse quences related to urbanization, population growth and climate change require prioritization that is often fraught with conflict.

PAGE 30

11 Cities, particularly those in developing countries, face problems of poverty, exclusion, insecurity, and environmental degrad ation, and have looked to urban planning to combat some of these ills ( Motasim et al., 2010 ) But urbanization is not just a technical challenge; it is a social transformation ( Roberts & Kanaley, 2010 ) Urban planning has traditionally been used to regulate the production and use of space often reflecting the dominant culture and driven by the affluent and/or politically powerful ( Sandercock, 2003 ) The concept of community engagement is not new in the planning lit erature; however, there is a renewed interest in why it has had such limited success and how to improve its impact moving forward ( Brownill & Parker, 2010 ; Manta conroy, 2011 ) Participatory Planning: Discourse and Reality The discourse on participatory planning has a history that dates back to the 1960s ( Brownill & Parker, 2010 ; Davidoff, 1965 ) A rev iew of t he literature revealed many benefits to engaging communities in the planning and development process, including increased democracy, mobilization of resources and energy, education, more holistic and integrated approaches, better decisions, environ mental management, and empowering people ( Heritage & Dooris, 2009 ; Holman, 2008 ; Irvin & Stansbury, 2004 ) Participation is about communities having decision making power over resources that affect the community and incorporating local knowledge in the planning process ( B ailey, 2010 ; Swapan, 2013 ) A number of typologies have been developed that describe different types of participation. P robably most well known is Arnstein's ladder of

PAGE 31

12 participation developed in the 1960s which describes a spectrum with control by authorities on one end and control by the people or citizens on the other ( Cornwall, 2008 ) The debate is not whether to include participation, but when and how as idealized views of participation are rarely reflected in practice, a point of widespread criticism ( Brownill & Parker, 2010 ; Laurian & Shaw, 2009 ; Monno & Khakee, 2012 ; Neef & Neubert, 2010 ; Sultana, 2009 ) In reality there are many challenges to successful community partic ipation includ ing the potential for conflict and mistrust, lack of adequate representation, and absence of authority or lack of dominant voice ( Irvin & Stansbury, 2004 ; Maginn, 2007 ) T wo problematic assumptions underpin participatory planning practice relevant to this dissertation. First is the assumption that community members have equal access and motivation to par ticipate. To understand the first assumption, we n eed to ask: who is participating and why? The second assumption linked closely to the first is that the public will view the problem at hand through the planner's lens disembedded from personal agendas. To understand the second assumption, it is importa nt to consider the nature of knowledge and problem setting versus problem solving. Participation and Representation First, who is participating, and why? Having a greater number of participating community members does not mean their greater influence i n the planning process. Those who self select to participate may disproportionately represent those who are in agreement with the problem at hand or conversely,

PAGE 32

13 passionately disagree and want to make their views known. Those who don't participate may do so not out of apathy, but because they don't believe their opinions will be valued, don't feel they have a right to participate, or don't understand the proposed project ( Cornwall, 2008 ; McAlister, 2010 ) This presents a challenge for planners in identifying collective voice On the other hand, p resence and inclusion are not the same as participation; a person may attend community meetings but not feel that he or she has a voic e. Empirical evidence dispels the idea that increas ing fo rmal participation options ( town meetings, public charrettes, mail in surveys etc. ) also increase s benefits to the community, and that inclusion of communities automatically leads to more democrati c and sustainable outcomes ( Illsley & Coles, 2009 ; Manta conroy, 2011 ) Simply creating an invited s pace for participation does not necessarily make conflicts disappear ; inter group conflicts can be suppressed in the participatory process when those in opposition are either not included or exclude themselves ( Beard & Cartmill, 2007 ; Holman, 2008 ; Sultana, 2009 ) The concept of a gency in participatory planni ng is central to the democratic nature of the participatory process. Agency can be defined as "what a person is free to do and achieve in pursuit of whatever goals or values he or she regards as important" ( Sen, 1985, p. 203 ) Translated in to participatory planning where communities exhibit some level of decision making power over resources that affect them, agency is not simply measured as the act of participating especially if the means or options available to participa te do not

PAGE 33

14 align with an individual's goals Furthermore it is not only through participation that agency can occur. Individuals are socially embedded, and agency can come from the perception that actions or outcomes of the group are congruent with one's values ( Peris, Farinas, Lopez, & Boni, 2012 ) Put another way, a gency can come through representation (indirect participation) when an individual feels that others are speaking or acting how he or she would. There are many reasons that i ndividuals may not be able to directly participate such as lack of time, financial limitations cultural norms or other constraints, and so representation becomes an important means for agency. But who is acting on behalf of whom? Communities are rarely uniform ; they are mor e often dynamic, multi cultural entities, with many voices ( Panelli & Welch, 2005 ; Snodgrass et al., 2008 ) Even defining community can be difficult because c ommunity boundaries and members can change and shift depending on the issue or point of focus ( Lyons, Smuts, & Stephens 2001 ) W ithout a clear understanding of the multiple voices within and across communities, we cannot rely on representation in the planning process to reflect varying perspectives and goals Beebeejaun and Vanderhoven ( 2010 ) argue that the nature of representation must be assessed p articularly a s cities begin to address large scale issues of climate change and urbanization through sustainable planning and development T he scale of impact is often beyond the confines of one community and that makes community participation even more challenging when the "community" may i nclude hundreds of th ousands of individuals or more.

PAGE 34

15 Participation at such a broad scale necessarily moves into the realm of representation, and inequities of power among participants can be considerable in large scale projects. Pickering and Minnery pro vide d an example of power inequity in a case study of metropolitan regional planning in Vancouver, Canada and South East Queensland, Australia ( Pickering & Minnery, 2012 ) Although planning authorities in each city deemed their respective projects "successful," Pickering and Minnery question ed the role of power in determini ng whose voices were actually represented. Notably, participation efforts were focused on education and information sharing with the public a top down approach where little to no decision making power was actually given to the public. Given the diversity, size, and difficulty in defining communities, full participation leading to consensus is impractical ( Melcher, 2013 ) But, power inequities often lead to representation that reflects only the dominant groups. While a ge, gender, and wealth are predictors of individual participation ( Beebeejaun & Vanderhoven, 2010 ) connections through social networks have been shown to increase representation of wom en, the young and old, and poor in the participation process ( Mdee, 2008 ) Participation and representation have the potential to allow people commonly excluded or marginalized a space for agency ( Connelly, 2010 ) This point has important implications because if representation can capture many different voices (and particularly those most vulnerable or hard to reach), then participatory planning s hould focus on improving the quality (ensuring diversity of voices) of representation rather than quantity (more voices) of participation.

PAGE 35

16 Knowledge and Problem Setting Closely tied to the challenge of representation and identifying who is included in the discussion is the issue of framing. The public has multipl e personal agendas (and goals and values) which impact the ability to consider the problem as framed through the planner's lens. This phenomenon concerns the nature of knowl edge. Expert, scienti fic, local, indigenous, and situated knowledge are examples of the multiple ways of conceptualizing knowledge. Knowledge frames how humans understand and experience the social and physical environment, and it impacts decision making and behavior. K nowle dge often varies across individuals and social groups ( Meusburger, 2008 ) Meusburger ( 2008 ) posit s that one of the primary causes of variation in the distribution of knowledge is power asymmetry. Knowledge is not equally held, yet planners have a tendency to assume that communities either share the planner's knowledge or can be given that knowledge a nd will come to the same conclusion about resource use This can lead to power struggles and conflict. Power relations impact the participatory process, but conflict per se is not necessarily a negative feature in local government and c ommunity engagemen t it may in fact lead to better solutions if it is aptly mediated through deliberation ( Brownill & Parker, 2010 ) Unfortunately, invited formal spaces for community engagement often become spaces for diffusion of conflict and legitimization of the status quo rather than places of true deliberat ion ( Eguren, 2008 ; Newman, 2008 )

PAGE 36

17 In the planning literature, methods for community engagement primarily focus on maximizing participation through a "top down" approach this is not necessarily a criticism as it is a practitioner focused approach ( Baker, Coaffee, & Sherriff, 2007 ; Mahjabeen, Shrestha, & Dee, 2008 ) However, w hile anthropology and geography discourse has acknowledged a view of cultural and geographic heterogeneity across society for quite some time, the planning literature still tends to homogenize communities as the "public" in a planner public dichotomy And, t here persists a valuing of expert (planner) knowledge over lay (public) knowledge ( Monno & Khakee, 2012 ) The expert sets the problem, invites the designated "community" to provide some pre determined level of input, and chooses the "correct" solution using expert knowledge. K nowledge heterogeneity means that community members may not perceive a benefit in participating or they might not even find the topic relevant a risk if the community is not included in problem setting A case study in rural Ohio prov ided an example of this as played out in an unsuccessful watershed planning participation effort ( Manta conr oy, 2011 ) Participation efforts during the planning process focused on motivat ing community members to participate by trying different formats for outreach ( mail in surveys etc. ) B ut unfortunately, planners neglected to fram e the topic in a way tha t was relevant to community members. Consequently despite increased effort to reach out to community members, there were low participation rates. S tudies that provide informal means of participating ( for example, unplanned interaction at a local market) suggest that within group d ynamics, or power relations, have a greater influence

PAGE 37

18 on participation than outreach format (mail in versus online survey) ( Beebeej aun & Vanderhoven, 2010 ; Daniere, Takahashi, NaRanong, & Lan, 2005 ) When planners create a forum for participation, it often stays within the realm of problem solving ( Monno & Khakee, 2012 ) The crucial mistake of limiting public influence to problem solving is that if the public is included after the problem has been defined, planners have already imposed control over the scope of possible outcomes: "P roblem setting, which is a necessary condition for technical problem solving, is not itself a technical problem. When we set the pro blem, we impose upon it a coherence that allows us to say what i s wrong and what needs changing ( Sandercock, 2003, p.66 ) Leonie Sandercock's differentiation between phronesis and techne calls our attention to the distinction between participatory methods that are inclusive at the problem setting stage versus the problem solving stage. Participatory planning recognizes the need for an approach that relies more on layperson or indigenous knowledges to begin with problem setting. Planning and development can, then, be understood as a socially cons tructed process of meaning creation and sense making of the built environment through interactions between the agency of participants ( Peris et al., 2012 ) There are many methods for shifting the nexus of power from a top down to a bottom up approach. But power relations continue to impact the process of community participation and development outcomes particularly when planning "problems" concern larger and less well defined populations. 6 6 The mechanisms through which participation occurs is not the primary focus of this dissertation, and hence not ex plored in detail. The author acknowledges a

PAGE 38

19 Power Relations and Pathways to Action Participatory planning requires a better understanding of representation and how knowledge and problem framing mold the planning process and subsequent product. To narrow the divide between participatory planning discourse and reality requires an uncovering of the ways in which power relations act as a mecha nism in driving the flow of knowledge and resources in supporting or undermining agency. In this section, I present a more nuanced description of power and further differentiate between power and empowerment. I then look at how disempowered and marginali zed groups find ways to power and sometimes empowerment through informal means. Power Within, Power Without There is little agreement in the literature about the nature or meaning of power ( Barrett, 2002 ) However, it is generally accepted that at the most basic level, p ower is a measure of one's ability to control the environment including other's behaviors, and can be potential or actualized ( Lukes, 1978 ) Power plays a key role in participatory planning, and it can occur at multiple lev els from problem setting and problem solving to defining invited spaces and who is actually heard But without a more nuanced understanding or rather without unpacking the concept of power we are limited to acknowledging power's presence without a means t o change, negate, or resist it ( Barrett, 2002 ) significant body of literature that is not mentioned here including work by Patsy Healey, David Godschalk, Judith Innes, and John Forrester among others.

PAGE 39

20 Furthermore, power in the abstract is not useful for explaining its presence; it ca n only be analyzed in the context of human interaction. Cultural anthropologist Eric Wolf ( 1990 ) suggested that we think of power relationally and regard it as an aspec t of all human relations. Wolf's important contribution to power discourse was that he pushed beyond the confines of human agency and added a structural aspect. The next chapter provides a deeper discussion of Wolf's theories of power, but in brief, he c onceptualized power as working through different modes and involving different types of relationships : power as an individual attribute (inherent) ego alter power ( between one person and another or others ), tactical or organization al powe r (control of set ting ), and st ructural power (control of context) Top down approaches to community engagement put the negotiation of power between the planner and public (organizational level of power where the planner controls the setting or invited space ) and bottom up approaches shift it into the community ( ego alter level of power) The intention of shifting power into the community is that it can create a n opportunity space for empowerment. Empowerm ent happens when impoverished or marginalized individuals or grou ps can imagin e their world differently and make that vision a reality by changing the relations of power that h ave been keeping them from reaching that potential ( Eyben, Kabeer, & Cornwa ll, 2008 ) Empowerment theory is based on a conflict model ; it assumes society is comprised of different groups that have different levels of power and control over re sources ( Heritage & Dooris, 2009 ) More than just opening access to decision making in the planning process (i.e.

PAGE 40

21 agency through direct participation or indirectly through representation), empowerment is itself a process that leads people to perceive themselves as able and entitled to be part of the decision making space ( Rowlands, 1995 ) Agency in the planning process is embodied in the individual agent but embedded in social relations ( de Haan, 2012 ) and that is where the potential for empowerment lies. Put another way, agency is the ability to act in accordance with beliefs and values, but empowerment is the ability to change the structures that inhibit action and shape beliefs an d values. The process of empowerment has the potential to have far reaching impact on creating a more equitable planning process and the long term sustainability of planning outcomes. The problem with empowerment is that p ower by definition, cannot be be stowed upon another. One can gain or lose power, but one cannot give power. The notion of power as a zero sum, or power over something or someone leaves little opportunity for empowerment. But power doesn't always manifest as a subtractive element in w hich one person gains power when another loses power; it can also manifest as a generative element in which one person gains and another person also gains or remains neutral. Generative power is exhibited as power with power to and power within ( Eyben et al., 2008 ; Rowlands, 1995 ) Planners can cooperate with communities to create conditions necessary to increase power through capacity building (related to the concepts of power to and power within ) and enabling social action (power with ) but transfer of power isn't a given ( Heritage & Dooris, 2009 )

PAGE 41

22 A considerable risk in attempting to shift power from authorities to the local level through bottom u p approaches is that it merely transfer s power from elites at a higher level to elites at a local level and does not benefit the community equally ( Khanal, 2007 ; Saito Jensen & Nathan, 2011 ) It may be a case where there is merely a shift in who has power over the issue at hand. To ensure adoption and implementation of strategies by communities, it is fi rst necessary to identify existing power relationships before engaging communities in participatory work ( Mercer, Kelman, Lloyd, & Suchet Pearson, 2008 ) But the transition between power and empowerment is unclear Lyons, et al. ( 2001 ) hypothe size d a link between participation and sustainable planning and development that works through the process of empowerment. They present ed a case study of eight development projects in South Africa that collected qualitative data on the history of each pro ject, extent of participation (number of people involved), depth of participation (how much control the community had), reasons for participati ng or not participating, and broader political, economic, and social conditions. Projects were evaluated for per sonal empowerment and community empowerment. Personal empowerment was achieved if knowledge and skills were transferred to individuals through the participatory process and then applied in other ways ( generation of power to for example, new skills that le d to new employment). Community empowerment was achieved if the community was able to carry out projects to completion ( generation of power with ) The authors concluded that development projects that emphasize empowerment programs such as skills training might not be

PAGE 42

23 successful if they are constrained by local politics and community structure (power over ) Based on a review of the literature thus far a n important limitation in the planning literature is the often one dimensional discussion of power The focus has been on who has the power to make the decision. Although I've touched upon the more nuanced dimensions of power and power empowerment in the planning literature power is typically conceptualized at the structural level, which concerns politica l, economic, and social contexts (see for example, Brownill & Carpenter, 2007 ; Kesby, 2007 ; Pelling & High, 2005 ) Foucault who theorized how knowledge and power organize society, is frequently cited While Foucault considered power a s a complex phenomenon that extended across levels beyond the structural realm into everyday activities his conceptualization of power was a s a singular phenomenon Foucault's exploration of power, while invigorating in discourse, has limited utility in explaining how power works inside and outside participatory planning. R e cognizing power relations within the participatory process is not enough to develop strategies to reduce inequalities. In other words, a cknowledging that power is present is not the same as uncovering and being able to address why or how it is present. C ollaborative planning is often influenced by power, but it is how power is negotiated that impacts the potential for empowerment T here is a need to understand how a community's social structure impacts collective capacity to participate in and influence the planning process ( Holman, 2008 ) This critical shift has started as planners debate the opportunities for democracy

PAGE 43

24 against the "subsumption of rationality to power" ( Brownill & Carpenter, 2007 ) This means creating space for more in clusive and earlier deliberation in the planning process. What Wolf ( 1990 ) provided was a more nuanced conception of power. H e acknowledged power a s a multiple and comp lex agency structure phenomenon pervasive, but working differently, across multiple levels of human interacti on. His is a useful foundation for capturing the complexities of power relations in the participatory planning process. More recent s tudies have examine d the role of power within communities as it relates to participatory planning outcomes but they are limited in scope In a qualitative study in Chicago and Johannesburg, Beebeejaun and Vanderhoven ( 2010 ) completed semi structured interviews to explore the relationships between individuals and groups within an agency structure framework to understand patterns of participation in community driven development. They found that tensions and disagreements played out through private informal interactions (through social networks). Similar findings emerged in another case study looking at the relationship between social capital and participation across the cities of Bangkok and Ho Chi Minh City. Daniere, et al. ( 2005 ) administered 1000 hous ehold surveys across 10 total neighborhoods and found that participation could effectively take place outside formal structures through informal connections (again, through social networks). The authors concluded that higher levels of informal participati on through social capital might be more effective in achieving community goals than formal means because of the

PAGE 44

25 increased level of social integration where members negotiate varying beliefs, values and norms to produce a shared way of understanding. In su mmary, power negotiation occurs inside and outside the formal planning arena, and informal connections can provide agency through representation by those actively participating in the formal arena. But we still don't know how power negotiation works in or through social relations. It would be helpful if we knew more about the flow of power across relationships and how it impacts knowledge, perceptions, and behavior The Disempowered Participatory planning begins with a shift of power from top down planne r public approach to a bottom up into the community approach By shifting power into the community, a space for agency and empowerment is created. But, defining community boundaries and who counts as a legitimate participant is still defined by the plann er ( Connelly, 2010 ) And, when planners offer a means to participate but people don't show up, it is often construed as apathy or having nothing to contribute ( McAlister, 2010 ) It is important to contrast invited spaces and those that people create for themselves ( Cornwall, 2008 ) because a gency and power empowerment can interact in ways that produce disengagement from participation A nd, particularly in the case of poor and marginali zed groups, disengagement can reinforce marginalization and perpetuate power inequality. Stigmatization and discrimination can interfere with the ability for social relations to enable poor and marginalized groups to actualize agency. Agency is

PAGE 45

26 embedded in social relations, but when social relations undermine agency, it may become necessary for marginalized groups to find other path ways to power. S ocial relations can provide a means to accessing capital' or asset s,' but the poor can't or don't always t urn to them. Nygren and Myatt Hirvonen ( 2009 ) show ed that poverty and marginaliz ation of a community of Honduras peasants regulate d how they exercise d agency and establish ed social relationships In a differ ent case study of migrants in Hong Kong, Wong ( 2008 ) illustrate d how lack of human capital and structural factors further constrain ed participation in the new host society. One such way was that the migrants coped with eve ryday hostilities by internalizing the norms of conflict avoidance. Both authors call ed out the oversimplification of power relations, and the need to examine people's daily strategies of living in a way that highlights their perceptions about obtaining r esources through social relationships In a third study, Parizeau ( 2015 ) described how a group of informal recyclers in Buenos Aires drew upon social and material assets to sustain their livelihoods. In the pursuit of a viable livelihood, the recyclers were often faced with discrimination from mainstream residents, which limit ed their ability access resources like health care, education and police services. Parizeau suggest ed that stigma represent s a vulnerability to cultivating human capital. The establishment of social relations only as a means to meeting daily needs can se verely limit the potential for poor and marginalized groups to gain access to invited participatory spaces. The potential for local mobilization is dependent on the interconnectedness of residents in s pecific geographic locales ;

PAGE 46

27 it impacts how individuals and communities place value on their environment, are compelled to act, and further expand their social networks ( Daniere, Takahashi, & Narangong, 2002 ) The poor and marginalized present a unique challenge to planners in that they are difficult to reach. But their homes and livelihoods are often tied to the same place s where sustainable planning and development projects target (described in greater detail in the next chapter), which makes them a particularly important population to include ( Kabeer, Mahmud, & Isaza Castro, 2012 ) But, while it may be tempting to focus on them as "target groups" to create opportunities for empowerment, Cornwall ( 2008 ) suggests that it is also important to recognize that the poor and marginalized do not exist in social isolation. Their carefully selected economically and socially significant relati onships may link them to those who are better off, women with men, or other potentially better advantaged individuals or groups Cornwall calls for a dynamic understanding of social networks and the institutions and dimensions of difference that matter in the pursuit of their livelihoods as "na•ve efforts" of inclusive planning may only make things worse. Social Networks and Decision Making People are socially embedded, and agency in the planning process can take different forms. Power impact s the contr ol of decision making and, critically, even the awareness and framing of underlying problems. Social networks comprised of individuals (called agents not to be confused with the concept of agency discussed earlier ) and the relations between them ( Scott, 1991 ) provide

PAGE 47

28 a way to look at the role of power in the interrelationship between agency (as discussed earlier) and structure. As planners are faced with more challenging problems and more complex communities, there is recognition of the public as agents embedded in social networks. I have summarized throughout this ch apter numerous empirical examples th at support a link between social relations and representation in the planning process ( see for example, Beard & Cartmill, 2007 ) M any studies suggest or conclude that power dynamics play a critical role in determining whether social relations function to provide the poor a nd marginalized an informal means of participating or actually increase inequality. And while there is a recent growing body of research using social network analysis in the planning literature, few studies actually measure power within social networks ( De mpwolf & Lyles, 2012 ) I briefly highlight two here One study that measur ed power in social relations was published by Holman ( 2008 ) She examined structural relationships and inefficiencies in an urban redevelopment project in England. The strength of this study was that social network data allowed Holman to look deeply into the participation process to evaluate the relat ional aspects of power among the different stakeholders and to see who was participating and who was absent or exclud ed. However, this study wa s limited to quantitative measures of social network s and only to networks of stakeholders in the redevelopment project. By focusing only on those actors involved in the project (inside the formal planning space) Holman measur e d case specific as opposed to systemic power dynamics. Furthermore, the quantitative approach t o social network analysis limited the study to a

PAGE 48

29 functional description. Such a method is useful for identifying social network weaknesses and points of entry for improvement or intervention strategies; however, qualitative measures could have provided information on why weaknesses existed. McKether ( 2011 ) use d qualitative data to analyze social network s to evaluate an effort to increase power relations in a Michigan black community. By cre ating a structured (quantifiable) social network out of qualitative data, McKether wa s able to show important relationships embedded in interview and ethnographic data. The value of McKether's approach wa s that it link ed both the why and how of power rela tions; it combine d quantitative data on the structure and function of various interacting groups within the community with a rich qualitative description of how individuals were connected and, therefore, mobilized within the social networks (e.g. friendshi p, member of formal organization). Research Gaps Inequitable power relations plague the community engagement process in sustainable planning and development. The reality is that full public participation is neither practical nor necessarily desirable. So we must rely on agency through representation. Yet representation is driven by power relations, which can lead those with the least power particularly, the poor and marginalized to pursue informal means to power outside the planning process (often as a livelihood means) ( for example, Nygren & Myatt Hirvonen, 2009 ) The literature repeatedly

PAGE 49

30 calls for more in depth exploration of power relations wit hin communities ( Beebeejaun & Vanderhoven, 2010 ; Brownill & Parker, 2010 ; Leach, 2008 ; Yates, 2012 ) Social networks are a way to look at power in the interrelationship between individual agency and structure but there is a g ap in measuring how agents use power to draw knowledge and resources through social networks how power works at different levels of social and institutional influence, and the nuanced pathways to power that create space for agency and empowerment. This dissertation fills a few i mportant gap s in the literature First, it examines power relations both inside and outside the planning process. The dynamic nature of power empowerment disempowerment is more evident among poor and marginalized groups, and thus I selected a case study of a poor, marginalized urban farm community faced with development of their land. A principal goal of my research was to uncover the ways in which power relations acted as a mechanism driving the flow of knowledge and resources in supporting or undermin ing agency within the farm community and in the planner public nexus. S econd it fills an important methodological gap: studies that measure social capital, social relations, and social structure have been limited by the challenge of operationalizing the concept of power. This study expands on the concept of power using Wolf's theory that it is a multiple and complex phenomenon pervasive, but working differently, across multiple levels of human interaction. In doing so, this dissertation expands the fou ndation for discourse related to agency and empowerment

PAGE 50

31 Third this study links theories, methods, and observations together in a coherent manner in order to operationalize power Social network approaches have had limited ability to measure how power r elations impact the participatory planning process, and could benefit from being situated in a larger theoretical framework. The following chapter details the development of a theoretical framework used to organize the concepts introduced in this chapter in a way that can then be operationalized and measured.

PAGE 51

32 CHAPTER III : THEORETICAL FRAMEWORK Polit ical Ecology F rames Social Vulnerability Multi s cale Human Environment Relationship As discussed in the previous chapter, there are numerous challenges to engaging communities in sustainable planning and development. T here is a gap in understanding power relations within and across communities and how power affects participation and who has influence in the decision making process Many modern issues cro ss social and geographic boundaries and are beyond the capacity of any one community to solve ( Dale & Newman, 2008 ) While planners may focus on issues of climate change and urbanization, communities are more often concerned with daily activiti es and livelihoods. This macro micro framing can and often does lead to a disconnect between planners and the public in problem setting and problem solving, which can in turn lead to conflict regarding best use or allocation of resources. To understand s eemingly disparate macro micro framing and potentially conflicting behaviors, it is helpful to look at human environment interactions that embed actions in the larger context of settings and system of settings At the city level, planners apply princi ples of sustainable planning and development in decisions on how to best use and allocate natural resources (for example, land for electric powered public transit) to mitigate climate change (macro level framing). At the community level, the public relies on knowledge exchanged through social relations about the availability of natural resources (for

PAGE 52

33 example, land for farming) for livelihoods and other daily activities (micro level framing). Figure 3.1 is a diagram created to illustrate macro micro human environment framing within the context of sustainable planning and development. Figure 3.1. Diagram of the human environment relationship in the context of sustainable planning and development So, how do resources get allocated and used? This dis sertation is not an attempt to prioritize city level versus community level needs or the allocation of resources in meeting those needs. R ather, I focus on the interface between planner and public the setting where power relations play out to gain insight into how or why the planning participation process succeeds or, more often, breaks down. Sustainable planning and development frequently targets areas that are degraded or prone to natural hazards These are the places poor and marginalized groups ofte n inhabit ( Kabeer et al., 2012 ) which makes them a n important group in the discourse and practice of participatory planning Poor and

PAGE 53

34 marginalized groups are often underrepresented in the participatory planning process making it even more critical to unders tand their inclusion/ exclusion. To gain an understanding of the larger picture and why marginalized groups are likely to inhabit the places sustainable planning and development frequently target i t is helpful to look to geography and anthropology for theo ries, frameworks, and models of vulnerability and resilience. I begi n with a brief overview of political ecology, a critical approach that posits that political, economic, and social forces act on populations to influence their relationship with the envir onment Continuing with a critical approach, I next turn to concepts of vulnerability and resiliency as a broad framework that lays out components and processes of the human environment relationship. Underlying a critical approach to understanding vulne rability and resiliency is the concept of power acting as a driver of the political, economic, and social forces that shape the human environment relationship. But what is power? The next section initiates a deeper discussion of power as a theoretical c onstruct, and elaborates on cultural anthropologist Eric Wolf's theories of power as a n agency structure phenomenon. I then shift from the macro scale aspects (structural) of vulnerabi lity to micro scale aspects ( opportunities for human agency) in order to expand the concept of power To focus turns to production; political ecology posits that livelihoods offer a direct window into the social environment process. This led me to the Sustainable Livelihoods Framework (SLF). As discussed in the previous cha pter, l ivelihoods, the means to meeting daily needs especially

PAGE 54

35 those outside the formal employment sector are a clue to the multiple relationships of power and resistance. Importantly, livelihoods are not limited to economic activity, but can also functio n as cultural and political activity. An important critique of the SLF is that it is theoretically neutral; it doesn't easily accommodate questions of power and polit ical ecology I address ed this theoretical gap through the application of cultural anthr opologist Eric Wolf's theories of power. By applying Wolf's theories of power to the SLF I develop ed a theoretical framework that I then use d to guide this dissertation research. Finally, I return to social network research introduced in the last chap ter, to introduce social network analysis as an appropriate and useful method of measuring power relations within the theoretical framework developed as described below. Political Ecology Critical theory posits that power distribution is an outcome of h istorical and political economic context, and can act as a guide to understanding the sources and dimensions of inequality ( LeCompte & S chensul, 1999 ) Building from the concepts of critical theory, political e cology argues that environment al changes do not affect people homogeneously; there is unequal distribution of costs and benefits, which reinforce, increase, or reduce social and e conomic inequalities ( Robbins, 2012 ) This dynamic process impacts power relationships and vulnerability can be understood as past and present patterns of domination and marginalization ( Vasquez Leon, 2009 )

PAGE 55

36 C entral to political ecology is the assumption that the environment is a social construct ( Robbins, 2012 ) For example, n atural events can have bot h positive and negative impacts : biotic systems often de pend on disturbance floods along floodplains may destroy existing vegetation but restore nutrients to the soi l but become constructed as "hazards" when they unequally burden people. To reduce the human burden of "hazards," it is necessary to increase access to resources and empower vulnerable groups ( Blaikie et al., 1994 ) Political ecology recognizes that measures and impacts of vulnerability are scale specific and can vary based on individual, community or region level analysis. Vulnerability & Resilience McLaug hlin and Dietz ( 2008 ) conceptualize vulnerability as the interrelationship of social structure, human agency, and the environment 7 W ith roots that span across theoretical paradigms defining and understanding vulnerability to environmental change is challenging ( McLaughlin & Dietz, 2008 ) Although measuring vulnerability wa s not within the scope o f this research it i s important to define this construct because poor and marginalized groups are often vulnerable. Vulnerability is broadly defined as individual and collective susceptibility to natural events and the capacity to anticipate, cope with, resist, and recover from those events ( Scandlyn, Simon, Thomas, & Brett, 2013 ) Natural events while originating in the biologic system put people at risk 7 The author acknowledges a broad body of literature on the topic of vulnerability. McLaughlin and Dietz ( 20 08 ) offer a particularly relevant definition for setting up the theoretical framework developed for this dissertation.

PAGE 56

37 becau se of human originated geographic organization. So, although this dissertation focuses on human events (urban planning activity), there are many parallels with natural events, as I will describe below. Within the natural ha zards literature, vulnerability is comprised of three components: risk of exposure, risk of inadequate capacity to cope with exposure, and risk of severe consequences or limited recovery from exposure ( Adger, 2006 ) Birkmann ( 2006 ) argues that vulnerability is a dual interaction of (1) external susceptibility and (2) internal coping and recovering Conflict and critical theory approaches posit that vulnerability is dependent on social processes. Th is structural approach to understanding vulnerability is illustrated by the often cited Pressure and Release (PAR) Model (figure 3.2) ( Blaikie et al., 1994, p. 23 ) : Figure 3.2. Pressure and Release Model ( Blaikie et al., 1994, p. 23 )

PAGE 57

38 The PAR Model illustrates the progression of vulnerability as a convergence of root causes and dynamic pressures, which in turn leads to unsafe conditions. Root causes result from the distribution of resources an d reflect the power distribution within society. It is a linear model where s usceptibility to exposure can be observed externally by understanding political economic systems and hazards originating in the biologic system. In this model, v ulnerability and hazards combine to create a risk situation. D isaster risk is the direct outcome of root causes; however, changing political economic or ecologic systems is often not a practical solution for risk reduction ( and beyond the planner's reach in most cases). Instead we must carefully consider solutions that are more than just a treatment of s ymptoms such as preventing floods rather than simply engineering levees that hold back flood s Although the effects of a natural hazard event (for example, resulting mor tality and morbidity from an earthquake) are symptoms, the PAR Model assumes that it is the event that allows for observation of causes that may go overlooked in day to day activities. In other words, it is the hazard event that brings to light underlying structural problems. A notable limitation of the PAR Model is the implied assumption that a single cause has a single effect. It does not account for the web of causation where social, ecological and economic systems overlap and feedback into one anothe r. This may be because the PAR Model does not include vulnerability concepts of internal coping and recovery at the individual level ; it puts the focus on structure and provides little space for opportunity or human agency. Despite

PAGE 58

39 limitations, it is an important early model in the development of vulnerability as a social phenomenon and in understanding the process of coping Blaikie, et al. ( 1994 ) define d c oping as the act of using available resources to achieve a culturally acceptable end, such as the case with various livelihood activities including farming marginal land. Coping occurs at the household level and is something other than what may occur thro ugh "official" disaster relief. Coping and recovery are essential components in defining a person or group as vulnerable, and are best observed when there is a catalyst (disaster event ). This brings us to the concept of resiliency. Inherent in the con cept of vulnerability is resilience. Resilience is defined as the "capacity over time of a system, organization, community or individual to create, alter, and implement multiple adaptive actions in the face of unpredictable climatic changes ( Martin Breen & Anderies, 2011, p. 2 ) I t is a complex interrelationship where increased vulnerability to external forces creates a catalyst toward adaptive action (i.e. coping and recovery or resilience). Ideally, as the vulnerable become more resilient, they become less vulnerable. In realit y, clim a t e change is inherently unpredictable, and thus the relationship between vulnerability and resilience should be understood as a process and not an end point. Vulnerability and resiliency use similar language, which makes it challenging to avoid dic hotomizing the two. V ulnerable groups are usually poor, but it is not necessarily the most poor who are most vulnerable ( Bohle, Downing, & Watts, 1994 ) There are cases where the most vulnerable may be the most resilient, and the moderately vulnerable may rely on resources that give them a

PAGE 59

40 false sense of prote ction. For example, slum dwellers may experience significantly less mortality in the case of an earthquake event than those living in poorly constructed formal high rise apartments. It may be a discrete or identifiable event that acts as a catalyst to ob serve vulnerability and resilience in action B ut characteristics of vulnerability and resilience are embedded in a person or group prior to an event, which is why the different people and groups experiencing the event may have different coping and recov ery. V ulnerability is a challenging concept to define because of complex interrelationships and feedback loops among its components. In a case study of Hispanic farmers in southeastern Arizona, Vasquez Leon ( 2009 ) used ethnographic methods to explore how social relations through network structures shape the human environmental interrelationship. She focused on livelihood activities and histories b ecause of the dependency of farmers on local hydrologic processes (droughts). She noted that "[d]espite having limited options, the critically vulnerable can generate structures that help reduce their vulnerability. The most effective strategies tend to b e collective" ( Vasquez Leon, 2009, p. 290 ) She concluded that, ironically, institutional mechanisms of buffering people fr om climate change impacts might actually have the adverse effect of discouraging the development of collective capacity. Pelling and High ( 2005 ) further argued that multi layered and dimensional social ties built through everyday social interactions may be the best way f or a community to develop and maintain the collective capacity to change and adapt in the face of climate change.

PAGE 60

41 Based on this brief overview vulnerability can be understood as a social construct dependent on social mechanisms ( Birkmann, 2006 ; Blaikie et al., 1994 ; Scandlyn et al., 2013 ) Underlying t he critical approach to understanding vulnerability and resiliency is the concept of power as a driver of the political, economic, and social forces shaping the human environment relationship. Yates ( 2012 ) called for research to address how knowledge, power, and skills flow informally outside government systems in the creation of adaptive capacity. Although brief, I have summarized structural aspects of vulnerability pertinent to this disse rtation research Before I shift the focus from constraints of human agency (structural aspects) to opportunities I first initiate a deeper discussion of power as a theoretical construct. Unpacking the Nature of Power: Eric Wolf In Power meets Culture an in depth review of the current status of power across the disciplines of anthropology, sociology, and psychology, Steven Barrett ( 2002 ) concludes that: First, t here is no agreement in the literature about the nature or meaning of power. Second, p ower doesn't add to analysis unless it is unpacked (because power is nearly always involved it must be unpacked to capture variabi lity ). Third, p ower in the abstract is not fruitful for explanation; it can only be analyzed in the context of human interaction and s ome would say after the fact. And fourth, a n understanding of power doesn't necessarily mean nor lead to rejection of po wer I n other words, acknowledgement of the presence of power does not mean that it can or will be negated, neglected, or resisted

PAGE 61

42 One of the early explicit definition s of power credited to Steven Lukes ( 1978 ) wa s that p ower is a measure of one's ability to control the environment including other's behaviors, and can be potential or actualized ( Barrett, 2002 ) The many concepts in the literature typically used to construct a definition of power include authority, manipulation, consensus, consent, persuasion, influence, coercion and force. In his seminal work titled "Power: A Radical View Lukes ( 1974 ) p roposed a three dimensional view of p ower First, he described power conceptualized as one dimension: power over or zero sum. It is measured as who makes the decision. He then described the idea that power can also function to reinforce barriers ove r another person or group (the second dimension) It is measured as who controls the agenda (from which the decision is made) The first and second dimensions rely on confli ct as a mechanism to create a situation where power can play out. The third dimension of power, described by Lukes as his "radical view," can be found in the prevention of conflict through agenda setting (by omitting the item of conflict from the possible options on the agenda ) The third dimension con ceptualizes power as the ability to shape perception, cognition, and preference. All three dimensions are embedded in human agency the ability or potential of an agent to act. The con tribution cultural an thropologist Eric Wolf made to the discourse on power was to push power beyond the boundaries of human agency ( Barrett, Stokholm, & Burke, 2001 ) He combined agency and structure building on the work of Anthony Giddens ( 1984 ) and Jeffrey C. Isaac ( 1989 ) who described power as the cap acity of an agent to act as constrained or enabled by his or her

PAGE 62

43 social relations. Interestingly, W olf did not portray power as everything: power is only one principal concept that is used in conjunction with ideas and social relations. Further, Wolf dis missed the conception of power as a substance or force, something that can be grasped or lost; rather arguing that power should considered as a relational mechanism and regarded as an aspect of all human relations ( Barrett et al., 2001 ) Further power relations occur through different modes and thus involve different types of relationships ( Wolf, 1990 ) As noted in the previous chapter, Wolf define d four modes of power: power as an individual attribute (individual or ego centric power), ago alter power (the ability of one person to impose power over another), tactical or organization al power (power over a setting or arena), and structural power (power over labor). Although Wolf took an agency structure approach, he was most concerned with power as it played out through social politi cal economic historical processes. He a pplied a political economy frame to power and focused his research primarily at the structural level. He believed that human action s affect the world, and th at the world, in turn, affects human action. And while a great deal of anthropology research at that time focused on the lived reality of people doing things (focus on agency) Wolf's goal was to situate human action (agency) in the larger reality of the world (structure) ( Wolf, 1990 ) In consequence, h is attention was on organizational and structural modes of power. The product of Wolf's work was an in depth examination of three case s : the Kwakiutl, the Aztecs, and Hitler 's Germany; w hich represented the extreme of human variability related to power and ideas (extreme societies) ( W olf, 2001 )

PAGE 63

44 R elationships of power are most evident when a major organizational transformation occurs (in cases of conflict), and that was how he selected his cases to investigate power at the structural level As Lukes stated, power can be found in si tuations where conflict is suppressed; however, power and conflict are intrinsically connected ( Barrett et a l., 2001 ) Therefore, power is nearly always present in cases of conflict. Wolf's structural approach to understanding how social relations produce power, which in turn shapes ideas, undoubtedly advanced how we conceptualize power as manifest across soc iety. Wolf's thesis wa s that ideas, social relations, and power are interdependent and vary according to the form of social labor This means that while Wolf endeavor ed to embed lived reality (human agency) in time and place (structure) in order to gi ve us a more nuanced understanding of power, he did not negate the notion that society is a produced through everyday activities enacted in accordance with the specific mode of production Power Relations and Livelihoods Livelihoods as a Window In an atte mpt to unravel the mechanisms of social and environmental interaction, political ecology puts the focus on production, and posits that livelihoods offer a direct window into the social environment process ( Robbins, 2012 ) In other words, s ocial environmental relations emerge and evolve through everyday activities ( Loftus & Lumsden, 2008 ) For example, Carr and McCusker ( 2009 ) looked at how the co production of land use and li velihoods are different

PAGE 64

45 manifestations of the same social process through which humans negotiate challenges in the face of various resources available to them. The authors present ed a case study that illustrates how land reform laws impact not just land u se, but livelihood decisions and spatial decision making. Set in South Africa, they describe d a case of rural black farmers forced onto the most marginalized land in overcrowded conditions. The decrease in their productivity reinforced white and state be liefs that blacks were unproductive farmers as well as changed the mindset of the black farmers themselves in their ability to be productive. Black farmers no longer trusted their own skills or knowledge and turned to white, scientific resources. By chan ging livelihood opportunities, land use changes have deep impacts at the level of the household and even on gender relations. The implication is that sustainable development must carefully consider context specific livelihood decision making processes. To reiterate livelihoods go beyond economic activity to include cultural and political activity. And, l ivelihoods, especially those outside the formal employment sector, clue us in to the multiple relationships of power and resistance. In a case study i n Buenos Aires, Whitson ( 2007 ) conducted 93 in depth interviews with informal workers to understand how power relations were reconstructed in the struggle over meaning and control through everyday activities. Informal work enabled the Buenos Aires workers to use resources in new ways and activel y or unintentionally resist the political system. Although Whitson's study was in the context of labor and economic crisis, it illustrated that

PAGE 65

46 power can be expressed in multiple ways including exploitation, subjugation, and resistance, and that these can interact simultaneously. Sustainable Livelihoods Framework The Sustainable Livelihoods Framework (SLF), which evolved from the vulnerability resiliency literature in response to rural development, was conceptualized as a framework for understanding the process of creating sustainable livelihoods (figure 3.3) ( Scoones, 2009 ) In defini ng SLF concepts, Scoones states: "A livelihood comprises the capabilities, assets (including both material and social resources) and activities for a means of living. A livelihood is sustainable when it can cope with and recover from stresses and shocks, maintain or enhance its capabilities and assets, while n ot undermining the natural resource base" ( Scoones, 1998, p. 5 ) The SLF focuses on people based on dai ly needs acknowledg ing various capabilities rather than seeing vulner able groups as homogenous ( Birkmann, 2006 ) It was developed to fill the need for a framework for bottom up participatory approaches to community development, and was indebted to Sen's work on agency and capabilities ( de Haan, 2012 ) Interestingly, the livelihoods approach, which shifts the focus from constraints to opportunities ( structure to agency), is more theoretically neutral than many previous vulnerability models ; there is space for power, but it is no t explicit.

PAGE 66

47 Figure 3.3. Sustainable Liv elihoods Framework (SLF) ( Scoones, 1998 ) adapted and printed by ( Birkmann, 2006 ) In the SLF, livelihoods are defined as the assets (the five capitals: human, natural, financial, physical and social) and activities, as well as the abil ity to access assets and activities, that determine the living gained by an individual or household ( Groenewald & Bulte, 2012 ) Bu t the definition of livelihoods can be expanded to encompass the circulation of knowledge and skills, the management of relationships, processes of inclusion and exclusion, and gaining of individual and group identity and status ( De Haan & Zoomers, 2005 ) The SLF was developed for and primar i ly used in community development practice, but, to the best of my knowledge, has not been applied in understanding the participatory plannin g process It does, however, provide a good fit for this research because: [T] he approach is people centred, in that the making of policy is based on understanding the realities of struggle of poor people themselves, on the principle of their participat ion in determining priorities for practical intervention, and on their need to influence the institutional structures and processes that govern their lives. [I] t is holistic' in that it is non sectoral' and it recognises multiple influences, multiple ac tors, multiple strategies and multiple outcomes.

PAGE 67

48 [I] t is dynamic' in that it attempts to understand change, complex cause and effect relationships and iterative chains of events'. [I] t starts with analysis of strengths rather than of needs, and seeks t o build on everyone's inherent potential. [I] t attempts to bridge the gap' between macro and micro levels. [I] t is committed explicitly to several different dimensions of sustainability: environmental, eco nomic, social and institutional ( Scoones, 1998 ) The SLF organizes key components across multiple scales that impact the process of attaining a sustainable livelihood, but it doesn't explain underlying mechanisms that drive the process. Scoones ( 2009 ) himself admits there is a need to introduce theory to the SLF. Livelihood strategies are influenced by the assets that are available to individuals or households; those with more assets have more options ( Parizeau, 2015 ) Hence a ccess to assets is a key element within the framework To reduce vulnerability, it is necessary to increase access to resources and empower vulnerable groups ( Blaikie et al., 1994 ; Scandlyn et al., 2013 ) However, it isn't just the number of assets, but rather the quality and stability and the ability (pow er) to access them when needed, which is influenced by social relations. T here is a gap in empirical research linking social relations and access to tangible and intangible assets and livelihood outcomes I n the link between access and livelihood opportu nities, power is an important and often overlooked explanatory variable, and focusing on the various layers of power would complete the concept of access in the livelihoods approach ( De Haan & Zoomers, 2005 )

PAGE 68

49 Research Framework Power as a Lens The SLF provides a framework for understanding multiple levels of influence on decision making within the context of livelihoods but it is an organizational strategy and does not have e xplanatory power; it lacks theory I ntegrating Wolf's four modes of power ( 1990 ) into the SLF however, improves the ability to explain underlying mechanisms that drive the different components of the SLF. (1) At the structural level, power organizes the setting by shaping the social field and determining which actions are even possible it gives the "big picture." At this level, power specifies the distribution and direction of energy flows. Wolf draws from Marx's idea of the power of capital to employ and allocate labor power; it structures the possible field of action ( Barrett, 2002 ) The structural mode of power is represented in the SLF as the interaction between the "vulnerability context" and the "transforming structures" (figure 3.4). Figure 3.4. Structural mode of power applied to t he SLF ( Scoones, 1998 )

PAGE 69

50 Power at this level is not situated in human agency, but in the reinforcing process that occurs between structure and agency. In the "vulnerability context" of the SLF, recall the definition of vulnerability as individual and collective susceptibility to natural events and the capacity to anticipate, cope with, resist, and recover from those events ( Scandlyn et al., 2013 ) Natural events while originating in natural system s put people at risk because of human originated geographic organization. So, power in the vulnerability context is the ability to become more resilient; the power to create, alter, and implement multiple adaptive actions in the face of unpredictable climatic changes ( Martin Breen & Anderies, 2011 ) This is where environmental resources, shocks, trends, and seasonality impose limits on the infinite possibilities of social labor and livelihood options. Structural p ower in the SLF is also a transforming structure comprised of levels of government and the private sector. This is where the system of labor is defined and where the systems organizing human activities a re defined ( human rights, legal/penalty system rituals /symbols etc ). (2) At the organizational level, power manifests as control over the contexts within which people interact. The organizational mode of power is represented in the SLF as "transformin g processes" (figure 3.5).

PAGE 70

51 Figure 3.5. Organizational mode of power applied to the SLF ( Scoones, 1998 ) Organizational p ower as a transforming process is comprised of laws, policies, culture, and institutions The organizational level is important because it sets up the relationships among people that determine how resources are allocated, who controls them, and who benefits. It is, in effect, about setting the agenda. The focus shifts more to human agency, but maintains an interaction with structure. (3) The ego alter level moves into the arena of individual human behavior through the interactions and transactions that take place between one person and another. At this level Wolf takes the Weberian view that power is the ability of a person to impose his or her will on another person or group(s). The focus is on human agency without an arena (agency not structure). The eg o alter mode of power is represented in the SLF as "livelihood assets" (figure 3.6). Unlike Wolf's treatment of the individual as the unit of focus, the SLF puts the focus at the household level. So in this framework, ego alter is translated as household alter.

PAGE 71

52 Figure 3.6. Ego alter mode of power applied to the SLF ( Scoones, 1998 ) Ego alter power, defined as the interactions a nd transactions that take place between people, is further detailed in the SLF as the influence of and access to livelihood assets in the form of human, natural, financial, physical, and social capital. Wolf's structural approach to understanding how soci al relations produce power left us without a deep understanding of how to approach the more agency oriented modes of power. Drawing from the Weberian view, Wolf's theory of power translates in the SLF as power to influence and access livelihood assets thr ough social networks (the critical mechanism for mobilizing resources beyond the household). The next section of this chapter develops the link between power and social networks in more detail. (4) Lastly, at the individual level, power is exhibited as t he potential to act; it represents a proxy for behavior because it is where inherent action exists. The individual mode of power is represented in the SLF as "livelihood strategies" (figure 3.7). Again, I note that Wolf's theory of power at this level is translated from the individual as the unit of analysis to the household in the SLF

PAGE 72

53 Figure 3.7. Individual mode of power applied to the SLF ( Scoones, 1998 ) Individual power is the inherent strength or capability of a person or, in this case, household. Presumably, in the SLF, livelihood strategies are part of a process of awareness, evaluation, and valuing of inherent indi vidual and household strengths and capabilities. Thus, livelihood strategies serve as one component of the individual mode of power. This limitation is taken up in the methods section in order to develop a more holistic method of measuring the individual mode of power Beyond the scope of Wolf's lens of power, l ivelihood outcomes" in the SLF are the product of the multiple levels of influence on human behavior and livelihood decisions In other words, l ivelihood outcomes are the end result of multiple m odes of power inf luencing the agency structure process. Power relations work differently at different levels. But, t he levels are not mutually exclusive, and the SLF helps to illustrate feedbacks among levels. I acknowledge the importance of considering all four modes of power as described by Wolf. However, the primary focus of this dissertation was on human behavior

PAGE 73

54 as embedded in multiple levels of influence. Thus, this research focused on human agency rather than structural aspects of power. I conc entrated on ego alter (access to livelihood assets ) and individual (livelihood strategies) modes power (highlighted in red and blue boxes in figure 3.8 below) but considered organizational and structural modes to gain a multi dimensional understanding of behavioral processes. Figure 3.9 is a simplified diagram integrating Wolf's power and the SLF as conceptualized for this dissertation research Figure 3.8. E go alter and individual modes of power applied to the SLF ( Scoones, 1998 ) Figure 3.9 Simplified diagram integrating Wolf's power and the SLF as conceptualized for this dissertation research

PAGE 74

55 Social Network Theory Wolf's ego alter mode of power concerns individual behavior (translated as household behavior in the SLF) through the interactions and transactions that take place between people but it lacks a method for measuring interactions and transactions Social networks are a critical mechanism for mobilizing resources (in the form of five capitals) beyond the household ( Blaikie et al., 1994 ) Social network analysis (SNA) is a promising area of research for examining social networks A core assumption of the theory behind SNA is that direct or intensive contact exposes individuals (called "agents") to better information, greater awareness, and higher susceptibility to inf luence; whereas indirect contact exposes agents to new ideas, and potential access to useful resources (e.g. livelihood opportunities) S ocial networks act to channel information and resources to particular structural locations, help create interests and shared identities, and promote sh ared norms and values ( Knoke & Yang, 2008 ) As a result of information exchange, social networks affect perceptions, beliefs, and actions through socially constructed structural mechanisms. SNA theory situates individuals within the larger community rather than seeing the community as composed of unrelated individuals, and allows for the observation and location of exchanges among individuals ( LeCompte & Schensul, 1999 ; McKether, 2011 ) Social networks are comprised of agents and relations. Agents are defined as ind ividuals or collectivities (such as households ), and relations are defined as the contact, connection, or tie between a pair of agents.

PAGE 75

56 A relation exists only if two agents maintain association. Central to SNA theory is the concept that social structures facilitate or constrain opportunities, behaviors, and cognitions ( Carrasco, Hogan, Wellman, & Miller, 2008 ) In other words, social relationships impact the creation of common meanings ( Norris, Stevens, Pfefferbaum, Wyche, & Pfefferbaum, 2008 ) and have an important influence on behavior ( Knoke & Yang, 2008 ) SNA provide s a compatible method for developing a measur e for ego alter power SNA theory draws from many disciplines including anthropology, graph theory, and management science, and has roots in sociology in assumi ng that the whole is greater than the sum of the parts ( Carrasco et al., 2008 ) There are three key assumptions in SNA theor y ( Knoke & Yang, 2008 ) The first assumption is that attributes (i.e. gender, age) are static, but relations exist at specific time place locales and can disappear or be suspended elsewhere. This means that relations are more important in understanding observed behavior than attributes. The second assumption, already noted above, is that the type of information exchanged and degree of influence depends on whether the relation is direct or indirect. The third assumption i s that agents change relation structures both intentionally and unintentionally; structures are a micro macro level process of within group and across group relations. In this way, structural relations can be understood as a dynamic process. The core theo retical problem in developing a methodology to measure a social network is to explain the occurrence of different relationships (called "ties") between agents and account for variation in linkages (i.e. the type of relation).

PAGE 76

57 Typically, social networks ar e measured quantitatively through a structur e function lens; metrics include identifying nodes and links (agents and relationships) and weighting them based on measures that include between ness, bridging, centrality, closeness, and reach ( Scott, 1991 ) A quantitative approach provides information on the distribution and connectivity of relationships and k no wledge exchange but is limited in the ability to id entify how or why power acts as a mechanism through social networks underlying beliefs and behaviors SNA is therefore, strengthened by a mixed methods approac h ( Edwards, 2010 ) Social network relationships are beginning to incorporate a spatial component because of emerging evidence of a relationship between physical proximity and relational influence. For exa mple, a case study in rural Bangladesh explor ed community characteristics and participation in water resource management ( Sultana, 2009 ) Focus gr oups and interviews revealed that nature, social factors, and spatial relations impacted how people understood and experienced community and participation. The authors illustrated that two families of similar socio economic status exposed to different geo graphic attributes one clean water ; one contaminated water may have different interests and goals related to water resource management. There is also growing evidence that physical proximity is related to social interactions (figure 3 .1 0 ); however, few studies explicitly map social networks using specific geographic locations ( Butts, Acton, Hipp, & Nagle, 2012 ; Doreian & Conti, 2012 ; Hipp, Faris, & Boessen, 2012 ; Verdery, Entwisle, Faust, & Rindfuss, 2012 ) Although not the primary goal of this dissertat ion research, geography

PAGE 77

58 be it physical context (environmental resources) or spatial proximity (distance to social relations) is an influential variable and important to consider in addressing research aims. Research Question Understanding how househol ds access resources through social networks in the pursuit of sustainable livelihoods can provide insight int o how agency works in the participatory planning process Using the theoretical framework described in this chapter, t his dissertation answer ed th e primary research question: How does the social structure of the farming community situated along the Yamuna River floodplain in Delhi, India impact livelihood adaptation strategies in response to planned land use change? The next chapter lays out the r esearch design. Figure 3.10 Example of geographically plotted social networks (Hipp, et al., 2012)

PAGE 78

59 CHAPTER IV : RESEARCH DESIGN This dissertation use d a case study design to explore the link between power relationships, planning participation, and livelihood strategies among urban farmers facing land development City planning and de velopment occurring along the Yamuna River floodplain in Delhi, India provide d a timely catalyst to observe power relations as a mechanism of decision making. This chapter first includes a description of the research site and characteristics of the target population. This is followed by a detailed description of research methods, including study questions, aims, sample selection, definitions, and data collection methods. Finally, I provide detail on data analysis, including a description of data preparat ion, cleaning, coding, and qualitative, quantitative, and mixed methods analysis. Research Site Preliminary Fieldwork The a rea selected for this research is located on the floodplain of the Yamuna River within the political boundary of the National Capita l Territory of Delhi (NCT) India. Th e area was selected based on preliminary fieldwork that explore d urban agriculture in a rapidly developing megacity ( Cook et al., 2015 ) We conducted interviews with 35 farm households across eight sites during the summer of 2011. Interview sites spanned the length of the Yamuna River as it flows through Delhi and included peri urban sites upstream and downstream of

PAGE 79

60 Delhi. There was a distance of approximately 25 miles (40 km) between the two farthest sites ( figure 4.1 ). Figure 4. 1 Context m ap of p reliminary fieldwork I estimate d approximately 2,500 farming families situated along the length of the Yamuna R iver in Delhi using geographic information systems (GIS) data Based on preliminary findings, I selected a typical case site, Yamuna Khadir, located on the eastern bank of the Yamuna River immediately west of the Mayur V ihar neighborhood encompassing app roximately 1,500 acres ( figure 4.2 ). At

PAGE 80

61 the time of this dissertation research, there was on going development of a new elevated metro line on and adjacent to the site and it had recently been impacted by new highway and earlier phases of metro construct ion as well as development to house the 2010 Commonwealth Games (CWG) Consequently, s ome households had already lost some or all of the land they farmed Figure 4. 2. Map of r esearch site Population Characteristics Findings from preliminary research provide d rationale for taking a social network approach (detailed later in this chapter ), and guide d operationalization of research concepts. We found that farm sizes varied somewhat; however, in general, the typical total land farmed by a household was 11 bigha ( slightly less

PAGE 81

62 than two acres or three quarter hectares ) 8 The typical household size was six to ten people, and both men and women actively participated in farm related activities Most families lived on the land they farmed, although some did own a hou se or property elsewhere in Delhi or a nearby village. The Delhi Development Authority (DDA) was the primary land owner. Many fa r mers mentioned that permanent residence was officially illegal, but that did not prevent them from building housing s tructures. Houses on the farms ranged from temporary reed shelters to sturdy mud houses. Despite annual monsoons, farmers were able to grow high value vegetable crops nearly year round. Multiple cropping seasons provide d a steady source of income, and f ew farmers looked f or other means of employment in between harvests Most interviewed households grew produce primarily for sale and hired temporary laborers to help at various times throughout the year. Once crops were harvested, most were sold through one or more of the many distribution options available to Delhi farmers. Most farmers utilized at least two distribution options, but a few reported having up to three or four. These characteristics indicate d that Yamuna farmers had access to a variety o f capitals; although they represented a vulnerable, low income group living on marginalized land they were not the extreme poor. Through preliminary research interviews, I found that the people living and working in Yamuna Khadir identified as a collective group based on membership 8 B igha are a commonly used unit of land measurement The size of a bigha varies by region, but in Delhi 6 bigha is approximately 1 ac re or 0.4 hectares.

PAGE 82

63 in t he Delhi Peasants Cooperative Multipurpose Society (the Society) 9 sharing a common profession as farmers and representing households of similar social status. H owever, preliminary fieldwork suggested a lack of collective c apacity to respond to government development pressure s. Yamuna Khadir farmers were at risk for land eviction, yet they had not actively mobilized or engage d in the on going planning process targeting la nd use and development along the Yamuna River. Compo unding this, these farmers face d unpredictable climatic changes as a result of increased variability and intensity of flooding events. Given this context, Yamuna Khadir offered the potential to learn and contribute new knowledge about the role and influen ce of social networks of poor and marginalized groups on their adaptive strategies when faced with land use changes that directly impact their livelihoods (figures 4.3 4.5 ). The planning process serve d as a catalyst for (i.e. set in motion) this research ; such an unusual circumstance is rare and was worth studying in depth. 9 In 1949, the Delhi Improvement Trust allotted more than 5,000 acres of land along the Yamuna to the Society. The Society parceled out the land to Society members by giving them the right to cultivate it. In 1991, the DDA became the successor of the Delhi Improvement Trust and, consequently, the authority over the land along the Yamuna (refer to Chapter I: Introduction for more details) From my understanding, Society membership is now irrelevant in determining land tenure ; however, past beneficiaries continue to associate themselves as members.

PAGE 83

64 Figure 4. 3 (Left) Urban agriculture with newly constructed apartment bui ldings in the background; (right) beneath active construc tion of a new metro line Figure 4 .4 (Left ) Another site where urban agriculture faces newly constructed apartment bui ldings in the background; (right) beneath new electricity towers Figure 4. 5. (Left) Urban agriculture "integrated" with newly construc ted (waste?) water pipes; (right) remov ed from a new "public" (but unpopulated) green space

PAGE 84

65 Methods This dissertation appl ied a mixed methods case study design to answer the primary research question: How does the social structure of the farming community situated along the Yamuna River flo odplain in Delhi, India impact livelihood adaptation strategies in response to planned land use change? For the purpose of th is research, social structure was conceptualized as household level power to influence and access livelihood assets through soci al relations L ivelihood adaptation strategies encompass ed short term (i.e. seasonal, monthly) tactics to either maintain or create new ways to support a viable livelihood. And, planned land use change related to current, on going development (including construction of an elevated metro line) on the research site. I answer ed the primary research question by addressing three specific research aims designed to explore social structure (A im 1) response to planned land use change (Aim 2), and livelihood stra tegies (A im 3) The theoretical framework described in Chapter III : Theoretical Framework guided development of research aims. Table 4. 1 summarizes research aims, secondary research questions, and data collection methods. Data were collected through sem i structured interviews, observation, GIS mapping, and photography. The interview guide is included as Appendix A and took approximately 30 45 minutes to administer.

PAGE 85

66 Table 4. 1. Summary of r esearch a ims Research Aims Secondary Research Questions Mode of Po wer Research Variables Data Collection Methods 1. Describe how Yamuna farmers exchange knowledge and resources through social networks. How is the social network of each household characterized? Ego alter Ego centric (household) social structure (a) S emi structured Interviews (b) Record: map & notes How is the social structure of the community characterized? Group (community) social structure 2. D escribe the beliefs and behaviors of Yamuna farmers related to land use planning How do farmers perce ive their influence on land use planning? Organizational & Structural Potential for action (a) Semi structure Interviews (b) Record: notes (c) Secondary data Action What is the behavior of farmers in response to land use planning? Context for action 3. I dentify livelihood adaptation strategies of Yamuna farmers in response to planned land use change. What do farmers plan to do in response to planned land use changes that impact their livelihoods? Individual Livelihood adaptation strat egies (a) Semi structured Interviews (b) Record: notes Unit of Analysis T his research focused on two units of analysis: household and community Maintaining or improving livelihoods is an individual level (household ) endeavor ; however, this research i s also concerned with the planner public nexus or organizational level (community). I considered Yamuna farmers a community because they were socially bounded as tenant farmers and past members of the Society and physically bounded within Yamuna Khadir by the Yamuna River,

PAGE 86

67 highways and metro line. However, like all communities I acknowledge pervious social and physical boundaries in the delineation of who is inside versus outside a community due to various outside influences and migration patterns. S ample Selection Emerging SNA research demonstrates that s ocial interaction is more likely among people who are closer in social status and physical proximity ( Butts et al., 2012 ; Shaefer, 2012 ) so I developed a purposeful, adjacent sampling method. My rationale was that I could compare patterns and variations based on geographic context and spatial proximity to gain a deeper understanding of hou sehold level and community level social structure. I select ed the first household through convenience by selecting the first family I encounter ed in the field. I then proceeded to interview adjacent household s until I exceeded 1 50 households, averaging 3 4 families per field day. I pilot tested this approach with t hree households in another nearby farm community to help develop the research design ( described in the next section). Pilot test findings showed variation in social networks among the three ho useholds farming similar plots, growing similar crops, and living within close proximity to each other This led me to believe I could also expect variation in the social networks among Yamuna Khadir farmers. I analyze d my data concurrently during data c ollect ion to assure variation in the research sample, but did not have to make any adjustments to the sampling scheme.

PAGE 87

68 As noted in the literature review chapter, a critique of urban planning approaches is the assumption that communities are homogenous. Th is research was designed to seek out diversity within the communit y To the outsider, t he case study site appeared to be a homogenous group of farm ers doing all the same things but the inner reality w as more dynamic. In the interest of understanding the "community," all sampled adjacent households were included in other words, I didn't exclude non agricultural households from participating. And, i n adhering to the sample selection methodology, all of the interviews occurred on the site where the househo ld farmed, lived or did some o ther kind of work. H owever, given the myriad variables that mediate and moderate decision making, and for the purposes of this dissertation, I limit ed data analysis to farming households. Pilot Testing Given the complexi ty of the research aims developed for this dissertation, it was prudent to pilot test whether collected data would be able to answer research questions. I conducted pilot interviews with three representative households farming two miles north of Yamuna Kh adir, the field site (figure 4.6 ). The pilot test interviews were ex cluded from the final research sample because they were outside the field site but were used to develop and test a coding scheme. Once dissertation interviews began, data analysis was a concurrent process and evolved as data and findings emerge d The initial pilot interviews

PAGE 88

69 and iterative process of developing the data analysis method assured that the collected data were suitable to answer research questions. Figure 4.6 Pilot test in terview sites. Research Aim 1 Aim 1: D escribe how Yamuna farmers exchange knowledge and resources through social networks. Aim 1 was designed to capture e go alter power, which deals with the interactions and transactions that take place between people. In the SLF, this mode of power relates to the influence of and access to livelihood assets in the form of human, natural, financial, physical, and social capital through social relations I answered Aim 1 through two secondary research questions:

PAGE 89

70 SRQ #1 : How is the social network of each household characterized? SRQ #2 : How is the social network of the community characterized? Measures and Definitions Social networks were characterized in terms of strong to weak household power to influence and access livelihood assets through social relations. In this dissertation, social structure is the term used to refer to social networks characterized by power in this way. Research Aim 1 required definitions for social networks, knowledge and resources in the f orm of livelihood assets, and ego alter power. Social structure was operationalized in a way that integrated these three components. Social networks Social networks comprised of agents and relations are crucial for gaining access to resources beyond the household For the purposes of this dissertation, agents were defined as Yamuna Khadir farm families (i.e. households), and relations were defined as the people that household s interacted with (including interactions within the family) Based on prelimi nary fieldwork, I created a list of generic relations to represent social network types: within the household, other farmers, laborers, vendors, landlords, industry, and other relations (table 4.2). The first assumption of social network theory is that s tructural relations exist at specific time place locales. I captured time place attributes by asking households where and when they interacted with each social relation type The second assumption of social network theory is that the type of information

PAGE 90

71 exchanged and degree of influence depends on whether the relation is direct or indirect. Direct relations are defined as being more intensive whereas indirect relations are defined as weaker and more distal contacts ( Knoke & Yang, 2008 ) Type of information exchanged was conceptualized as livelihood assets, and degree of influence was conceptualized as bonding and bridging ties: Bonding Ties A bonding relation was defined as a strong tie between immediate family members, neighbors, and friends. Bridging Ties A bridge relation was defined as a connection with a person or people of different socioeconomic and/or cultural backgrounds. Livelihood assets Knowledge and resources were conceptualized as livelihood assets in the fo rm of five capitals: human, natural, financial, physical and social (table 4.2) : Human Human capital was defined as livelihood assets related to knowledge and labor. I identified the following examples as potential human capital assets of Yamuna Khadir households: Large or extended family involved in farming activity School aged children in school Skills to do other work Knowledge to do other work

PAGE 91

72 Natural Natural capital was defined as livelihood assets related to ecosystem assets. I identified the following examples as potential natural capital assets of Yamuna Khadir households: Proximity to the river. Trees and vegetative cover. Depth to water table. Health risks. Micro climate and soil productivity. Urban proximity. Metro proximity. Financial Fi nancial capital was defined as livelihood assets related to money. I identified the following examples as potential financial capital assets of Yamuna Khadir households: Large farm (i.e. higher income from larger crop yields) Ability to buy tractor withou t a loan 10 Ability to hire a lot of laborers Other sources of income (rental property, other jobs, etc ) Physical Physical capital wa s defined as livelihood assets related to tangible "things." I identified the following examples as potential physical capi tal assets of Yamuna Khadir households: Land ownership and land characteristics Farm equipment Alternative livelihood equipment 10 In a capitalist context, access to loan funds is an important source of financial capital. However, the research population had very limited or no access to formal loans for a variety of re asons including lack of identification, poverty status, etc.

PAGE 92

73 Housing quality and quantity Animals and livestock Household items Social Social capital was defined as livelihood assets relate d to social connections. The distinction between soc ial capital and social networks is that so cial networks can be positive or negative whereas social capital is created through positive social network s I identified the following examples as potential social capital assets of Yamuna Khadir households: Sharing the same rural homeland village Sharing meals or celebrating festival and marriages with other families Long tenure and positive interaction with neighboring farmers Ego alter power A key focus of this dissertation research was to operationalize the role of power as it relate s to the influence of and access to livelihood assets through social networks Influence of and access to livelihood assets through social networks can produce gains and/or los ses but power isn't a simple equation of net gain. Wolf conceptualized ego alter power based on the Weberian view that it is the capacity of an ego (household) to impose his or her will on an alter (relation) Although he was vague in describing how ego alter power works he defined it as the interactions and transactions that take place between people The focus is on agency without an arena (structure). T his dissertation research expand ed on Wolf's definition. P ower was conceptualized as power to : as in power to access and power to influence livelihood assets through social networks. However, even the more nuanced

PAGE 93

74 conception of power as power to is limited. Wolf equates p ower with gain, but gain can take the form of immediate or delayed (potential) gain the act of investing with the expectation of a future return 11 Investment in a social network can provide direct returns, for example teaching a hired laborer a more efficient and cost saving farming technique, or indirect returns, for example increa sing the status of the group that the farmer is associated with, which in turn increases the farmer's status. Investment can also provide immediate returns in the form of exchange farmers working for each other in kind so they don't have to pay outside la borers. Thus, power to influence and access livelihood assets through social networks was measured as: investment, withdrawal, exchange or barrier. Investment Investment was defined as contributing with the expectation of a future return. It was in thi s sense, power to give (it is implied that investment wa s free of coercion ; no other person was imposing his or her will on the household ) Withdrawal Withdrawal was defined as utilizing an asset. It was in this sense, power to take. Exchange Exchange oc curred when the investment and withdrawal were concurrent (in other words, the household reaped the benefits of investment without waiting) 11 Wolf's theories of power were heavily influenced by Marx.

PAGE 94

75 Barriers A barrier to power was defined as a negative social relation that blocked access to assets. Although Wo lf doesn't discuss the concept of barriers to power, it is embedded in the concept of power as zero sum or "power over A barrier to power is not the same as lack of power but rather identifies the particular relation preventing a household from getting something they believe they could have if that relation wasn't preventing them from getting it Another way to think about barriers is that a household has the desire and the skills ( individual level power; potential to act ) but they can't act (un realiz ed power) Table 4.2 is a matrix of social relations and capitals used to guide interview question development. Table 4.2 Matrix for interview question guide development Social Relation Access/Influence Capital Human: knowledge & labor Natural: ecosys tem Financial: $$$ Physical: "stuff" Social: networks Other Farmers Hired labor Market/Vendors Industry Landlord Other relations Social Structure In summary, social relations included: the hous ehold, other farmers, laborers, vendors, landlords, industry, and other relations. Based on the definitions above e go alter p ower had three qualities: type of transaction (five capitals), direction of transaction (investment, withdrawal, exchange or barr ier ) and degree of influence (bond/bridge ). Because of the complex ity in determining

PAGE 95

76 social structure for Aim 1 s ocial relations were first combined into three dimensions : househol d, community, and outer network ; and then aggregated as social structure. Refer to Appendices B and C for tables and an in depth description of how ego alter power was operationalized. Household social network The household social network related to interactions among f amily members living together as a unit. Community soci al network The community social network related to those interactions within the farming community primarily with other farmers. It also included laborers if they were other farmers, and industry if farm inputs and/or advice came from inside the communi ty Outer social network The outer social network related to those interactions or connections beyond the farming community primarily through the landlord and vendors, but also including labor, industry, and other relations if they were outside the comm unity Household background characteristics Social networks are linked to social position, thus it was necessary to collect household background information to determine relative household and community level position. In terms of the theoretical framew ork, household background characteristics provided a proxy for existing family assets, position in the community and potential to access future assets. Characteristics included

PAGE 96

77 such attributes as education and employment and were used to determine a bas eline for each household that could be control led for in analysis of associations between social networks, land use beliefs and behavior s and livelihood strategies. Geographic location Social relations are time place dependent, and one aspect of this di ssertation was to capture the geographic location and frequency of interactions between agents and relations (households and social network members). T his research consider ed the potential influence of spatial proximity in addition to social proximity. F or example, households might have more power in their social relations with other households from the same village (social proximity) but they might also draw power from neighbors from other villages if they work ed next to each other in the field everyday (spatial proximity) Data Collection Interviews I collect ed data on household background characteristics and ego centric social networks through semi structured interview questions ( t ables 4.3, 4.4 and Appendix A Background & Part I: Social Network ). Interview questions were used to prompt families to talk about time place interactions and influence of and access to livelihood assets in regard to each social network relation type. A local Hindi English translator conducted interviews in Hindi. Interv iews were not audio recorded because of the potential for distrust by farmers which was a greater

PAGE 97

78 concern than the nominal benefit of capturing a few additional details. The interpreter was well trained and had proven she could relay what the farmers s ai d (" I say what they say not what I think they mean t to say"). During interviews, English translated household responses were recorded with pen and paper by either a second translator or m yself Households participating in interviews were compensated with a small toy for children and a family photo taken in the field and printed using a portable Polaroid Zinc photo printer. Table 4.3 Interview questions measuring household background characteristics Interview Questions Household background characteristi cs How long has your family been here? How many people are in your family? Do you live here? If not, where is your home? Do you have land any place else? Tenure H ousehold size Rent or o wn land Other home location Migrant origin Other land location T able 4.4 Interview questions measuring social networks Interview Questions Do farmers generally know each other? What kinds of things do you talk about? Prompts: Do you ask other farmers or tell other farmers what to grow, where to sell, how to get sc hemes or IDs? Do farmers generally help each other? How? Do you share laborers, tractors, borrow/lend $ When a new family comes here, how do you get to know them? Can you tell us about your experience when you first came here? Do you hire any laborers ? Prompts: How do you find out who to hire and where do they come from? Do you teach them new skills or give them anything to make your farm better? Do they tell you new things about how to make your farm better? Do they connect you with other people that can help improve your livelihood? How do you know what to pay them? Do you ever negotiate? Do you talk about anything else?

PAGE 98

79 Table 4.4 (con't) Interview questions measuring social networks Interview Questions How do you sell your produce? Prompt s: Does a vendor come or do you take it to a market? Do they give you a good price? How do you know what a good price is? Do they ever refuse to buy your produce? What do you do if that happens? Does (s)he tell you anything about how to make your farm be tter? Does (s)he connect you with other people that help improve your livelihood? Do you take out any loans? How often? When do you have to pay it back? What if you cannot pay it back or are late in payment? Do you talk about anything else? Do you pay rent to someone? (If not, talk about ownership) Prompts: Do they come here to collect? Do you know where they live? What do you talk about? Do you ask for any improvements on your property? Can you contact them if you have any issues or need something? Is the rent the same every month? What happens if you are late or cannot pay? Do they tell you new things about how to make your farm better? Do they connect you with other people that can help improve your livelihood? Are there any other people who co me here? (Prompt: family, friends, gov't NGOs, KVKs, etc.) Prompts: Where do they come from? What do you talk about? Do they tell you new things about how to make your farm better? Do they connect you with other people that can help improve your livelih ood? Do you participate in any schemes? Is there anyone else that you talk to? Note taking and memos In addition to recording interviews, either a research assistant or I took n ote s during interviews to capture observations and aid preliminary analysis of the respective household. A minimum of three memos were recorded at the end of each field day ( typically 3 4 interviews per day ). Memos and note taking

PAGE 99

80 provide d additional explanation, observations, insight, and summaries of the overall process Geo graphic location I record ed the geographic location of each interview site (the farm location of each household ) with an iPad 2 with 3G using the Bento app 12 Location and frequency of social relation interactions were recorded as notes during the intervie w. Research Aim 2 Aim 2: D escribe the beliefs and behaviors of Yamuna farmers related to land use planning. The second research aim addressed organizational power, which manifests as control over the contexts within which people interact including law s, policies, culture, and institutions For the purposes of this dissertation Aim 2 measured farmers' power over the context of planned land use change and was addressed through two secondary research questions: SRQ #3 : How do farmers perceive their in fluence on the land use change planning process? SRQ #4 : What is the current behavior of the farmers in response to the land use change planning process (e.g. have they gone to court, do they meet with other farmers)? 12 The Bento app is no longer available or supported by Filemaker.

PAGE 100

81 In addition to beliefs and behaviors of Yamuna farmers related to the land use change process, I looked at structural power. A full exploration of the structural mode of power was beyond the scope of this dissertation, but it was important to understand the historic and on going community en gagement process to contextualize this particular case for interpreting Aim 2 Wolf describes the structural mode of power as power embe dded in the settings themselves. Structural power organizes the setting within which behavior is possible Using the SLF as a guide, the structural level wa s represented by the interaction between the "vulnerability context" and the "transforming structures" (i.e. levels of government and the private sector). Measures and Definitions Organizational power deals with co ntrol over the context s within which people interact. In this research I investigate d the interface between the planning process and Yamuna Khadir farmers Secondary research questions focused on two dimensions of control: potential (perceived influence ) and act ual (behavior). The focus of the organizational mode of power is on agency, but within an agency structure dynamic. I applied concepts of agency and power as zero sum (power to and power over ) in addressing this research aim Underlying the org anizational mode of power is structure where the system of labor is defined and where the systems organizing human activities are defined For the purpose of this dissertation, the structural mode of power related to the underlying drivers of land develo pment

PAGE 101

82 Land use knowledge Knowledge acts as a baseline for interpreting perception and behavior, and was thus an important reference piece of information There must exist awareness that there is something to do before one can have the power to do it Case in point, if farmers weren't aware of any land use change, then they would be unlikely to have acted in response or have any beliefs about it. Knowledge is a starting point for agency in that it influences goals and values. Land use knowledge was d efined as information related to who, when, what and any other description of development planned for the land. Knowledge also included information related to land ownership. Land use behavior Behavior is the clear est measure of intention. For example, if a farmer was involved in a protest or rally against land development, he or she most likely intended to protest or rally against land development. Discussion was the primary source of information for Yamuna farmers in their understanding of what was h appening; thus, gaining information through discussion was a key mechanism for setting in motion action. Behavior was defined as i nvolvement in discussi ons about land development with other community members or outsiders and activities related to land use change. It is tempting to use behavior as a measure of agency; however, behavior doesn't always align with intension. Recall Sen's definition of agency as "what a person is free to do and achieve in pursuit of whatever goals or values he or she regards as important" ( Sen, 1985, p.

PAGE 102

83 203 ) If behavior doesn't match goals and values, then it doesn't fit the definition of agency. Land use perception Land use perception was key to understanding why farmers were or were not in volved in the land development process. Perception was defined as the d esire to be involved or belief in one's ability to influence land use changes Underlying this definition was the question: How did farmers weigh advantages against disadvantages of d iscussion and activity? By combining, knowledge (which influences), perception (goals and values), and behavior (what a person does or does not do), t his research aim captured the organizational mode of power M easuring agency wa s beyond the scope of this dissertation, but power at this level highlight s aspects of agency. Land Development Land development was defined as (1) the historic institutional land use change planning process to engage Yamuna Khadir farmers, and (2) the current institutional land use change planning proces s to engage Yamuna Khadir farmers. Data Collection Interviews I collect ed data through semi structured interview questions ( t able 4.5 and Appendix A Part II: Land Use Change ). See D a ta C ollection above under Research Aim 1 for details.

PAGE 103

84 Table 4.5 Interview questions measuring land use change Interview Questions Is there any development planned for this land? Prompts: Can you tell us about this who, when, what? Have you been involved in any discussions about the developme nt? With other farmers? Landlord? Authorities? What is discussed? If not, do you want to be involved in discussions? Why aren't you? Or why do you not want to be involved? Land development Data collection on land development occurred after conducting all interviews and after completing the majority of data analysis to limit researcher bias in interpreting interviews. In other words, I wanted to understand land use knowledge, beliefs and behavior of Yamuna farmers without preconceived notions Key inform ant meetings I conducted key informant meetings with city officials, NGOs, and local political advocates. Open ended topics were tailored to each interviewee and addressed the following : (1) The individual/organization's involvement or role in improving/ restoring/developing the Yamuna River floodplain/corridor; (2) The planning process to develop the Yamuna River; (3) The tenant farmers any interactions, engagement, professional viewpoints; (4) Key informant's understanding of the history of developin g the Yamuna River.

PAGE 104

85 Secondary data I reviewed newspaper articles, journal papers, and other media sources to understand the larger context of land use planning along the Yamuna. Research Aim 3 Aim 3: I dentify livelihood adaptation strategies of Yamuna farmers in response to planned land use change. This aim was designed to measure individual (household) power, defined as the potential to act (inherent power) Aim 3 captured farmer s' livelihood adaptive strategies through one secondary research questi on: SRQ #5 : What do the farmers plan to do in response to planned land use changes that impact their livelihoods? Potential to act is best measured in relation to potential change. Within the scope of this dissertation, livelihood strategies were situat ed in the context of metro construction and land development a change with the potential to significantly reduce the amount of land available for farming in Yamuna Khadir The individual mode of power represents the potential for action; it is a proxy for behavior but may not necessarily become action ( action is represented by "livelihood outcomes" in the SLF). Livelihood strategies become livelihood outcomes over time but this research investigated relational variables rather than causal or directional variables which would have require d measurement over time and beyond the scope of this dissertation

PAGE 105

86 The individual mode of power concerns the inherent strength or capability of a person or, in this case, household. Livelihood strategies are part of a process of awareness, evaluation, and valuing of inherent individual family member's and collective household strengths and capabilities. To fully capture inherent strengths and capabilities of households, I expanded the SLF concept of livelihood strategi es to include livelihood indicators, which I define below. Measures and Definitions Livelihood i ndicators Livelihood indicators provided information about household strengths and capabilities relative to the potential range of future livelihood options This measure was designed to capture whether a household was investing in the future or had other options for making money. Livelihood indicators were also designed to capture whether a household only knew how to farm and would need land or new skills t o sustain themselves if their land was taken and they could not find more land to farm. The following are definitions for different livelihood indicators: Land improvements Land improvements were an investment in future crops and instilled a sense of owne rship or stewardship for the land. Other skills This indicator captured the ability of households to diversify their livelihoods. If a family only farmed then if there was no farming they wouldn't

PAGE 106

87 have a livelihood. If the family had members who could sew, run a shop, drive, or had completed some level of education, then there were more livelihood options. This speaks directly to the concept of adaptation. Other sources of income This indicator captured diversification. If the family relied solely on selling produce, they would be entirely reliant on the availability of land, good weather, local market demands, etc. With other sources of income from running a small shop, selling milk from their cows, etc., the family could, in theory, sustain short p eriods of farm income interruption. Other work This indicator represented the dependency of the household on the land and options for moving if they were removed. If a family member drove a rickshaw, or worked as domestic help, then the household might no t need to find more farmland immediately. Also work as middlemen Middlemen bought produce from farmers and sold it in the market. This indicator was similar to "other work" in that it represented how dependent a household was on farmla nd. Households that worked as middlemen might not need to find more farmland to live on, but they would need to find more farmers. This was a unique category of interdependence. Households were opportunists and may have bee n able to find other middleman ty pe work, but being a middleman also required special knowledge and relationship building between producers and sellers/consumers.

PAGE 107

88 Other home/land Households that had other options for living or farming have different options for making livelihood decisions. Migrant ho useholds that still owned a home and/or land in their rural village might weigh the option to return to their village differently than households with nothing back in their villages; likewise for households with land or property in other areas of Delhi (or elsewhere). Livelihood strategies To capture the potential for action, livelihood adaptation strategies were defined as short term ( seasonal, monthly) tactics to either maintain or improve livelihoods when faced with planned land use change. This include d both beliefs and behaviors and could be in the form of action or inaction. Specific livelihood adaptation strategies were defined a posteriori as they emerge d through analysis of beliefs and behaviors. Data Collection Interviews I collect ed data thro ugh semi structured interview questions ( t able 4.6 and Appendix A Part III: Livelihood Strategies ). See Data C ollection above under Research Aim 1 for details. Table 4.6 Interview questions measuring livelihood strategies. Interview Questions Do you t hink that your family will farm this land next year? If yes or not sure, are you doing anything to improve your land or your family's prospects? If no, what will you do? Will you still try to improve your farm? How?

PAGE 108

89 Data Analysis Data analysis was designed to answer the primary research question: How does the social structure of the farming community situated along the Yamuna River floodplain in Delhi, India impact livelihood adaptation strategies in response to planned land use change? Social stru cture was capture d through A im 1, response to land u se change was captured through Aim 2 and livelihood st rategies were captured through A im 3 Data Recording and Digitization English translated interviews were recorded with pen and paper and typed as ra w notes in Microsoft¨ Word for Mac 2011 (Word) within a few days of each interview. Raw notes were organized using the interview guide for easy importing into Atlas.ti 6 Photos were imported to iPhoto 11 then uploaded to Evernote Premium in order to l ink them to the raw notes from the associated household. Memos were typed in Word. Interview sites (household locations) were geocoded in ArcMap 10 .0 using the .xml X Y coordinates exported from the iPad 2 Bento 4.0 app. I reviewed all data with the Hin di English translator for accuracy and completeness. Secondary data and key informant meetings were typed as raw notes with summaries in Word. Coding Interviews I used the interview guide linking theoretical constructs, research aims, and interview quest ions to provide a frame for coding interviews in Atlas.ti

PAGE 109

90 (Appendix A). The first round of coding captured household background characteristics, social networks, land use beliefs and behavior, and livelihood strategies. The second round of coding capture d concepts related to power and included measures of withdrawal, investment, exchange and barriers to influence/access livelihood assets through social networks. Mixed Methods Analysis I began by qualitatively summarizing Atlas.ti coded results (export ed as text files) for each interview question in a Word document. Summaries were organized by research aims into separate documents for: household background characteristics, social networks by each type (vendors, landlord, etc.) ( as part of Aim 1), land use beliefs and behaviors (Aim 2), and livelihood strategies (Aim 3) The detailed summaries allowed me to identify themes and categories. For example, as per Aim 2, I coded reasons households reported not being involved in the land development process. As I evaluated responses, some emergent categories were : (1) don't own/landlord doesn't own the land; (2) landlord exploits them; (3) lack of information; (4) landlord is taking care; and (5) lack of time because of work. I then created a Microsoft¨ Exce l¨ for Mac 2011 (Excel) file and categorized household interview responses as the data allowed The data I categorized in Excel was used to quantify responses. I imported the Excel data into Stata/IC 11.2 for Mac (Stata) and ran descriptive statistics (frequencies, percentages, distributions of responses) which I compared back to the qualitative summary. Through an iterative qualitative

PAGE 110

91 quantitative data analysis process, I revised the summary of interview themes and categories and identified or creat ed (1) variables appropriate for more advanced statistical and/or spatial analysis, and (2) data appropriate for qualitative explanation of findings. After summarizing household background characteristics, specific social networks by type (part of Aim 1) land use beliefs and behaviors (Aim 2) and livelihood strategies (Aim 3) I developed a method for characterizing overall household and community social network s (completion of Aim 1; Appendices B and C ). I first evaluated each social network type quan titatively and developed a scale of strong to weak social network power (i.e. 1 6; 1=strong and 6=weak). I next created social network diagrams using paper and marker to represent the social networks for each household. I marked the household in the cent er of the paper and then add ed the type and location of each interaction described during the interview relevant livelihood assets, and power notations ( figure 4.7 ). S ocial network diagrams were created by reviewing both raw and coded interview notes. The purpose of drawing social network diagrams was to compare with the quantitative social network scale. I added the quantitative social network scales to the social network diagrams to see if they matched and to adjust for any substantial divergence bet ween quantitative and qualitative characterization of the social networks. Finally, I developed a method for characterizing social networks at the household, community, and outer dimensions; bonding and bridging social networks; and social structure overa ll.

PAGE 111

92 Figure 4.7 Example s of social network map s After completing the mixed methods analysis, I returned to the field site and conducted in depth follow up interviews with 10 households to probe deeper into the responses given during the original intervie w. I purposefully selected households along the metro construction corridor because I wanted farmers who had recently been or were currently being impacted by construction (the catalyst event). I did not develop a new interview guide, but only probed dee per into original responses. Returning to the field deepened quantitative analysis, and provided insight to the qualitative analysis both described below. Quantitative Data Analysis I cleaned the quantitative data generated through the mixed methods anal ysis and imported the generated variables into both ArcMap and Stata to evaluate spatial and non spatial relationships. Non spatial relationships between

PAGE 112

93 research variables were evaluated using Stata. I ran descriptive statistics for each variable and e xplored patterns. Next, I used ordered logistic regression (OLR) to test the likelihood of relationships between demographic characteristics and social networks, land use beliefs and behaviors, and livelihood strategies. OLR is appropriate for analyzing data that has more than two categories and a meaningful sequential order. I developed the following hypotheses to act as a guide for quantitative data analysis: Social Structure Livelihood Strategies Hypothesis First, I posit ed that ego alter power (stro ng/weak) would mediate the relationship between social networks and livelihood strategies ( f igure 4.8 ). For example a social network that supported strong ego alter power would lead to similarity in livelihood strategies because of the high exchange of s imilar knowledge and resources among group members. Hypothesis #1 : If ego alter power within the social network is strong (strong social structure) then there will be similarity in livelihood strategies among households Hypothesis #2 : If ego alter power within the social network is weak (weak social structure) then there will be variability in livelihood s trategies among households (i.e. there will not be a strong, collective response).

PAGE 113

94 Figure 4.8 Ego alter power mediates the relationship between s ocial network s and livelihood strategies. Land Use Influence Livelihood Strategies Hypotheses Second, I posit ed that organizational power would moderate the relationship between social networks and livelihood strategies ( f igure 4.9 ). For example, if the social network was strong, then the level of organizational power ( strong/weak ) would become an important influence on whether livelihood strategies were to be enacted through formal or informal means. Organizational power was operationalized as farmers' perceived ability to influence land use planning and development. Hypothesis #3 : If organizational power is weak (i.e. farmers do not feel they can influence the planning process) then farmers will plan informal livelihood strategies. Hypothesis #4 : I f organizational power is strong (i.e. farmers feel they can influence the planning process) then farmers will plan formal livelihood strategies.

PAGE 114

95 Figure 4.9 Organizational power moderates the relationship between social network s and livelihood strategi es. Spatial Relationships Spatial relationships between research variables were evaluated using ArcMap. To adequately address research aim 1, I explored spatial patterns of household social structure As noted in the Social Network Theory section of Cha pter III : Theoretical Framework studies suggest that physical proximity is related to social interactions, and thus influences social networks ( for example Butts et al., 2012 ) I assumed that households near to each other would have some of the same social relations (use the same vendor, have the same landlord, etc.) and therefore have s imilar influence/access to livelihood assets. To adequately a ddress research A ims 2 and 3, I explored spatial patterns of household planning beliefs/ behaviors and livelihood strategies. Building on the idea that physical proximity has the potential to in fluence who is interacting with

PAGE 115

96 whom, and that knowledge is unevenly distributed, I assumed that households near to each other would have similar knowledge and, based on having similar knowledge, would have similar future plans I first looked at how individual variables mapped geographically to identify visual patterns. I next looked at background characteristics to evaluate if they were randomly distributed. Similar to the need to control for background characteristics in statistical models (sex, r ace, etc.), understanding if background characteristics clustered spatially was important in understanding if any spatial patterns that emerge for other variables were based on spatial distribution or background characteristics. I next identified pattern s of similarity/dissimilarity among households based on physical proximity related to the three research aims. G eo reference d social network data can be analyzed with programs including Pajek, Ucinet, R, and ArcMap; however, because this dissertation coll ected within household ( ego centric ) social network data and did not measure inter household ( dyatic or complete networks) connections it was not relevant to perform typical SNA tests such as centrality, density, cohesiveness, etc. I instead tested two a ssumptions : (1) that households near to each other would have similar influence/access to livelihood assets and therefore similar social structure; and (2) that households near to each other would have similar knowledge and similar future plans. I employed the ArcMap hotspot analysis 13 tool because it has been used to explore spatial patterns of social indicators (childhood health, crime rates) 13 The hotspot analysis tool can be run in other software including R and QGIS.

PAGE 116

97 ( Anselin, Sridharan, & Gholston, 2007 ; Chainey, Tompson, & Uhlig, 2008 ) Although I am unaware of its use in SNA research, it was a good fit for my intended analysis. Given a set of weighted features, hotspot analysis us es the Getis Ord Gi* to identify statistically significant hotspots and coldspots. It identifies clusters of points (in this case households) that are higher or lower in magnitude than one would expect to find by random chance and outputs the results as a z score. I weighted households using the Inverse Distance option, which gives more weight to points (households) that are closer in distance 14 and used Euclidean distance since Manhattan distance 15 was not relevant in this geographic context. Using the h otspot analysis tool I identified hotspots where statistically significant high values clustered geographically and coldspots where statistically significant low values clustered. Results illustrated clusters of households characterized with much stro nger or much weaker social networks. I similarly analyzed patterns of land use planning knowledge, beliefs, and behavior, and livelihood strategies. Qualitative Data Analysis In general, q uantitative analysis is limited in describing why one variable, attribute or activity might be more likely to coincide with another variable, 14 Based on the principle of least effort, which implies that people are more likely to have ties with people who are more proximate ( McLaughlin & Dietz, 2008 ) 15 Manhattan distance is the distance between two points measured along axes at right angles (such as along a street network), whereas Euclidean distance is the straight line distance between two points (often referred to as the crow flies') (ArcGIS www.resources.arcgis.com accessed December 9, 2014).

PAGE 117

98 attribute, or activity. For example, co occurrence of strong/weak social network power with certain livelihood strategies indicates that the two variables are related, but can not explain why After exploring the data quantitatively, I revisited the interview notes, code summaries, photographs, social network diagrams and memos to gain insight on why patterns may or may not have emerged through the quantitative analysis I id entified cases that illustrated the key patterns and lack of patterns produced by the quantitative analysis. I incorporated the in depth follow up interviews, key stakeholder interviews, and secondary contextual research data. This final phase of analysi s illuminated research findings.

PAGE 118

99 CHAPTER V : FINDINGS In this chapter, I present findings from my dissertation research. I first describe household background characteristics of interviewed households, and then present findings organized by the three r esearch aims Last, I summarize my final analysis to address the primary research question. A total of 165 households (figure 5.1) participated in interviews between March 7, 2013 and February 17, 2014. Figure 5. 1. Map of interview sites (note: inter view sites are not shown in the exact location where the household was interviewed in order to protect privacy) In the field, I employed an adjacent sampling method; I interviewed household s that were next to each other. But, I found that some household s

PAGE 119

100 were not practicing agriculture ; they had non farming livelihoods (rickshaw driver, maid, bicycle repair) The primary focus of this dissertation research was on urban farmers, thus I only included tenant farmers (n=121; 73%) in analysis. I excluded all landowners (n=10) and non farm ing households (n=34) (figure 5.2) because their livelihoods were different enough that interpretation of findings could be complicated by factors not anticipated or accounted for in the research design Figure 5. 2. Map of interviewed tenant farmers (note: interview sites are not shown in the exact location where the household was interviewed in order to protect privacy)

PAGE 120

101 Household Background Characteristics Family Size Household family size fell into three types: (1) sma ll family that included husband wife and kids (n= 74 ; 61 %) ; (2) small multi generational or mixed family including some adult kids who are married and a few grandchildren (n= 18 ; 15 %) ; (3) large extended family (n= 7 ; 6 %) ; 22 hous e holds ( 18% ) were missing dat a (table 5.1). In most cases, it was difficult to determine absolute household family size : s ome families were clearly nuclear (mom, dad children), but others were multi generational or included extended family members (in laws, uncles/aunts, etc.). In a few cases large extended families liv ed nearby and shar ed adjacent land, or liv ed together in one shelter on family farmed land. Two families had 20 household members and another two families had 30 40 household members. This type of co housing is not uncommon in India. N uclear family size (mom, dad, children) ranged from 1 to 10 members; 52% of families had 4 7 members (n= 41 ; 3 4 % of households were missing family size data ). The mean number of kids was four; mean number of boys was 2.3 and mean numb er of girls was 2.3. Migrant Origin The majority of migrant households were from Badaun (n= 91 ; 75 %) Another 26 migrant households ( 21 %) were from other small cities or villages Three households ( 2 %) were not migrants or at least did not migrate in the last generation (1 household (1%) w as missing migrant origin data) (figure 5.3; table 5.1).

PAGE 121

102 Figure 5. 3. M ap of migrant origin Ownership Status and Land S ize Families reported three different arrangement s in terms of land (n=116; 96%) : (1) rent (n=65; 54%) ; ( 2 ) batai 16 (n=44; 36%) ; ( 3 ) mix of rent and batai (n=7; 6%); 5 (4%) were missing data (table 5.1). Land size was calculated for 116 ( 96 %) households: 3 ( 2 %) had less than 2 bigha, 42 ( 38 %) had 2 to less than 5 bigha, 49 ( 40 %) had 5 to less than 1 0 bigha, 15 ( 12 %) had 10 to less than 15 bigha, 6 ( 5 %) had 15 to less than 50 16 Batai is translated as share cropping. The tenant and landowner share profits 50:50 rather than exchanging rent. In this arrangement, the tenant and landowner share profits and losses.

PAGE 122

103 bigha, and 1 ( 1 %) had 50 or more bigha (n= 5 ; 4 % were missing data) (figure 5.4; table 5.1). Figure 5. 4. Distribution of land size (2 bigha is approx. 0.3 acre s or 0.13 hecta res ; 10 bigha is approx. 1 .67 acres or 0.67 hectares ) 17 Tenure T enure ranged from a minimum of 2 months to a maximum of centuries (ancestral land). Tenure was categorized as the following: <1 year/monsoon cycle (n= 9 ; 7%) ; 2 3 years (n=1 2 ; 10 %) ; 4 10 years (n= 49 ; 40 %) ; 11 20 years or one generation (n= 31 ; 26 %) ; and m ultiple generations (n= 11 ; 9 %) (n=9; 7% were missing data) (table 5.1). Profession Although many families were primarily engaged in agriculture, there were different agricultural arrangements i ncluding farmers who worked as agricultural 17 The size of a bigha varies by region, but in Delhi 6 bigha is approximately 1 acre or 0.4 hectares. 3 42 49 15 6 1 0 10 20 30 40 50 60 <2 bigha 2 to <5 bigha 5 to <10 bigha 10 to <15 bigha 15 to <50 bigha 50+ bigha # Households

PAGE 123

104 laborers in addition to farming their own land Many farmers reported other jobs or sources of income 18 Figure 5.5 illustrates the range of other jobs and sources of income reported by households. It shows tha t t here were 109 households that only farmed, but two households also had family members who worked as vendors, one had a family member who worked as a maid in the local colonies, one had a family member who worked at the Yamuna Khadir Banpool school, two had family members who worked outside Yamuna Khadir as teachers, two had family members who drove rickshaws, and seven ran a small shop on their farmed land. Figure 5.5 Diagram showing n umber of households in and across different professions (note: tw o farming households had two additional sources of income) I found that jobs and sources of income were not static within the farming community they chang ed in response to local needs, supply, and adaptation. 18 It is uncommon for the poor to wo rk full time or be fully employed, which has led to households with multiple profession and sources of income.

PAGE 124

105 Over the period of time that I completed int erviews, there was an increase in the number of small shops near the on going metro construction site. Many farmers began selling small items to construction workers as a source of extra albeit temporary income. Household Speaker Attributes Age and gen der of household speakers were recorded during interviews to describe the sample and account for respondent bias due to gender or other cultural differences (figure 5.6). Gender of speakers was evenly split between men and women : w omen completed 4 9 % (n= 59 ) of the interviews and men completed 4 0 % (n= 48 ) of the interviews (table 5.1). There was a mix of both who completed 1 2 % (n=1 4 ) of the interviews. Age groups were also well distributed, although age was assessed visually rather than reported so there i s potential researcher bias. Age distribution : 3 1 (2 6 %) were <30 years old; 45 ( 37 %) were in their 30s ; 25 (2 1 %) were in their 40s ; 15 (1 2 %) were >50 years old ; 5 ( 4% ) were missing age assessment (table 5.1). Figure 5. 6. Distribution of gender and a ge of household speaker 49% 40% 12% Gender Female Male Mix 26% 37% 21% 12% Age <30 30s 40s >50

PAGE 125

106 Table 5. 1. Household background characteristics Characteristic Category Farmer s (n=121) Freq. Perc. Family Size S mall family 74 61% Multi generational or mixed 18 15% Large extended family 7 6% Missing 22 18% Migrant O rigin Badaun, Uttar Pradesh 91 75% Not migrants or did not migr ate in the last generation 3 2% F rom other small cities or villages 26 21% Missing 1 1% Ownership Status Rent 65 54% Batai 44 36% Mix of rent and batai 7 6% Missing 5 4% Lan d Size <2 bigha 3 2% 2 to <5 bigha 42 35% 5 to <10 bigha 49 40% 10 to <15 bigha 15 12% 15 to <50 bigha 6 5% 5 0+ bigha 1 1% Missing 5 4% Tenure <1 year/monsoon cycle 9 7% 2 3 years 12 10% 4 10 years 49 40% 11 20 years or one generation 31 26% Multiple generations 11 9% Missing 9 8% Profession Farmer 121 100% Non farmer n/a Missing 0 0% Gender Men 48 40% Women 59 49% Mix 14 12% Missing 0 0% Age <30 years old 31 26% 30s 45 37% 40s 25 21% >50 years old 15 12% M issing 5 4%

PAGE 126

107 Geographic Distribution of Household Characteristics I looked at spatial distribution of background characteristics to see if they clustered spatially but I did not find any relationship. Aim 1: Social Network Characterization The pur pose of the first aim was to describe how Yamuna farmers exchange knowledge and resources through social networks Aim 1 address e d the ego alter mode of power conceptualized as household influence of and access to livelihood assets in the form of the fiv e capitals (human, natural, financial, physical, and social) through social relations (figure 5.7) Figure 5. 7 Ego alter mode of power applied to the SLF ( Scoones, 1998 ) Aim 1 was achieved through two secondary research questions: ( 1 ) How is the social network of each household characterized ? And ( 2 ) How is the social network of the community characterized ? Social network s were characterized in terms of strong to weak household power to influence and

PAGE 127

108 access livelihood assets through social relations ( refer to Appendices B and C for operationalization; refer to Appendix D for findings by social network type ) A strong hous ehold social network enabled a family to access many different assets through social relations to sustain or improve their livelihoods, whereas a household with a weak network had few relations they could turn to or people who actively prevented them from getting the things they needed. A social network is not the net sum, but rather the integration of various social relations as part of a holistic network of access and barriers. Because of the complexity of aggregating household social networks, I first report findings based on three networks (representing three mutually exclusive dimensions of social relations ): household, community and outer network s I have included case examples to illustrate how these complex concepts of ego alter power manifest in real examples of access and influence Last, I summarize social structure overall Power by Social Network Dimension Social network power was dichotomized (strong/weak) into three dimensions: household, community, and outer network power. The househo ld social network related to familial ties. The community social network related to those interactions within the farming community primarily with other farmers. And, the outer social network related to those interactions or connections beyond the farmin g community primarily through the landlord and vendors, but also including labor, industry, and other relations.

PAGE 128

109 Household Network Power The household social network was comprised of the family A total of 75 (62%) households gave enough detail to evalu ate household network power (refer to Appendices B and C ) As shown in figure 5. 8 below, the majority of households interviewed (n=61; 81%) reported strong household network power : they were able to influence and access various livelihood assets through h ousehold members. Figure 5. 8 Strong/weak h ousehold network power: power to influence/access livelihood assets through household social netw ork Households were evaluated for household network power based on whether they educated their children ( power to influence or investment in human capital), had other jobs or sources of income in addition to farming ( power to access or withdra w human and financial capital), and had a large family size ( power to access or exchange social and human capital). This list is only a sample of the ways that ego alter power manifested through the household netwo rk; the full list is included in Appendices B and C. H ouseholds varied in their responses related to household network power, but the following is a n example of a household (1 of 61 ; 81% ) with strong household network power: 14 61 0 100 No Yes Strong Household Network Power Strong Household Network Power Freq. Percent Cum. No 14 18. 7 18. 7 Yes 61 81. 3 100 Total 75 100

PAGE 129

110 The husband 's uncle's son has lived here for 20 years and invited this family to come farm [ withdraw social capital ] The husband wants his kids to be educated that is why they came [ invest human capital ] He has a BA (bachelor of arts) a general university degree [ invest human capital ] He lives here with his mother, wife, and 5 k ids. His eldest daughter is in the 4th class. It is a tough situation to live and farm here, but he wants to educate his children. He taught at a school in Bandaun. He can teach up to 10 years old. He teaches here and is given a small compensation [ wi thdraw human and financial capital ] He is hopeful that he could get a job at a Delhi school ( excerpt from MV021 notes coding for household network power noted in bracketed italics ). The following is an example of a household with a weak household n etwor k (1 of 14 ; 19% ) : The brother has been here for 2 years. The brother in law has been here for 5 years. The brother came through the brother in law and the brother in law came through family/friend contact [ withdraw social capital ] There are many childr en. The two families live together in one hut. The brother has 4 children; the brother in law has 6 children. They don't go to school because they are scare d to send them across the highway [ barrier to human capital ] He says: What will they do anyway ; they'll just work in the field [ barrier to human and financial capital ] (excerpt from MV082 notes coding for household network power in bracketed italics ). An important question wa s whether households with strong versus weak household networks were phy sically close to each other. A positive finding would suggest that households might influence each other indirectly through their activities and choices For example, a household surrounded by families who sent their children to school rather than having them work on in the fields might be more inclined to send their own children to school; on the other hand, a household surrounded by families who didn't educate their children and rather had them work in the field to increase household income might think twice about sending their children to school A map illustrating spatial analysis of household network power is shown in figure 5.9 below. This map confirms that there were two areas where

PAGE 130

111 households with strong network power clustered geographically: along the main road and among households living closer to the nearby built up urban edge. These households were more likely to educate their children (often sending them to government schools in the adjacent built up neighborhood), have larger extended fa milies, and multiple sources of income. Households with weak household network power clustered along the main road closer to the Yamuna River. These households were less likely to educate their children, be comprised of smaller nuclear families, and have a single source of income (only farming). Figure 5. 9 Hotspot analysis of household network power (red indicates clusters of households with high household power; blue indicates clusters of households with low household power) (note: interview sites are not shown in the exact location where the household was interviewed in order to protect privacy)

PAGE 131

112 The measure for h o usehold networks only captured social relations within the household or family and, therefore wasn't a measure of relationship s with neighbors. The clustering of households displayed in figure 5.9 may be explained by proximity to the main road and Mayur Vihar, which has government schools and opportunities for domestic work, rickshaw pulling, etc. rather than proximity between househol ds. However, clustering could also indicate indirect influence between households. For example, a household might want to educate their children if they see their neighbors send their kids to school However, limitations of my data limit ability to expla in observed patterns in these data. Community Network Power The community network was comprised of household relations within the farming community primarily with other farmers. A total of 120 (99%) households gave enough detail to evaluate community ne twork power (refer to Appendices B and C ) As shown in figure 5. 10 below, the majority of households (n=105; 88%) reported strong community network power which meant that they were able to influence and access various livelihood assets through their rela tionships within the farming community This was primarily through relationships with other farmers but could also be with laborers, vendors, and industry people if those people were part of (living in) the Yamuna Khadir community

PAGE 132

113 Figure 5. 10 Strong/weak community network power: power to influence/access livelihood assets through community social networks Some attributes of a st rong community network were that a household gave other farmers loans (power to influence or invest in financial and social capital); borrowed money from or found land through other farmers (power to access or withdraw financial, social, and physical capit al); and talked to, helped, and built relationships or friendships with other farmers (power to access or exchange h uman, financial, social, and physical capital). Other attributes included: households that hired other farmers for labor in exchange for th eir own labor at a future date (power to access or exc hange social, financial, and human capital); selling to fixed vendors or buying other farmers' produce and selling themselves (power to exchange financial and social capital); and gaining skills and kno wledge related to agricultural practices from the educated local seed shop owner (power to influence or invest in human capital). Conversely, some barriers to household power were: not talking to other farmers, fighting, not trusting, not loaning money, a nd generally not helping other farmers (barrier to power to access or influence social, human, and financial capital). This list is 15 105 0 200 No Yes Strong Community Network Power Strong Community Network Power Freq. Percent Cum. No 15 12.5 12.5 Yes 105 88.5 100 Total 120 100

PAGE 133

114 only a sample of the ways that ego alter power manifested within a household's community network ; the full list is included in Appendices B and C The following two case s illustrate one household with a strong community netw ork and one with a weak community network It is notable that the majority of households exhibited strong community network power. This was primarily b ecause of the strong camaraderie among many farmers and willingness to help each other because they were "in the same boat." H ouseholds varied in their responses related to community network power, but the following is an example of a household (1 of 105 ; 89% ) with strong community network power : They are from Badaun [ exchange social capital ] The elder brother came and then called him to come with his family [ withdraw physical capi t al ] Both are here. Everyone lives like a family here. All the farmer s live nicely. They talk and ask all kinds of things, and they help with whatever comes up [ exchange human, social, financial capital ] When someone is gone they take care of the place for them. If they need a small amount of money they just give each o ther but if it is a big amount they ask the landlord and he gives it. They work for each other but don't hire outside labor ers. They work in cash or kind (excerpt from MV161 notes coding for community network power in bracketed italics ). The following i s an example of a household with weak community network power (1 of 15 ; 13% ) : They have been in Delhi for 2 3 years. They are from Sheharsa District. They speak to everyone but only within their family not with other farmers [ barrier to human, social capital ] They h ire laborers from outside and the money is paid by the landlords [ barrier to social capital ] They do not sel l produce only work in the farm [ barrier to social, financial capital ] (excerpt from MV086 notes coding for community network p ower in bracketed italics ). (Note: the landlord relationship was evaluated as part of the outer network reported in the next section not the community network.) Similar to household social networks, an important question relates to whether households wit h strong versus weak community networks were physically close to each other A positive finding would suggest that households

PAGE 134

115 near to each other might have some of the same social relations and, therefore, have similar influence and access to livelihood a ssets (such as use the same vendor, work for each other as laborers, talk to each other or loan each other money). A map illustrating spatial analysis of community network power is shown in figure 5.1 1 below. This map confirms that there was one area w here community network power clustered. Households with strong community network s clustered near the main road. These households had more access to and influence over a greater variety of livelihood assets through connections and interactions within the community. It may have been that close proximity increased the likelihood these families talked with one another, loaned each other money, worked for each other as laborers, or shared transportation costs to get their produce to market. However, limitati ons of my data, related to which households actually talked to each other prevent me from explaining observed patterns.

PAGE 135

116 Figure 5.1 1 Hotspot analysis of community network power (red indicates clusters of households with high community power; blue ind icates clusters of households with low community power) (note: interview sites are not shown in the exact location where the household was interviewed in order to protect privacy) Outer Network Power The outer social network was comprised of household re lations beyond the farming community primarily through the landlord and vendors, but also including labor, industry, and other relations who lived outside Yamuna Khadir A total of 118 (98%) households gave enough detail to evaluate outer network power (f igure 5.12 ; refer to Appendices B and C ) In contrast to findings above, in which more than 80% of househol ds reported strong household and/ or community network power, just over one third (n=43; 36%) reported having strong outer network power Strong out er network power meant that households

PAGE 136

117 were able to influence and access various livelihood assets through int eractions or connections beyond the farming community Figure 5.1 2 Strong/weak outer network power: power to influence/access livelihood assets through outer social networks Some attributes of a strong outer network were that a household had a positive relationship with the landlord (power t o invest or influence human and social capital); the landlord made improvement s to their land such as paying for a well for irrigation, helped them with loans, gave advice, didn't charge rent, offered small jobs, or was involved in land use discussion or a court case (power to withdraw or access human, physical, social, and financial capital); a household taught skills to laborers hired from outside the community (power to invest or influence human capital); when they sold to fixed vendors and were paid up front (power to withdraw or access financial and social capital); and, learned new skills or knowledge that improved their agricultural practice through industry relations (power to withdraw or access human capital). Some of the attributes of other social relations beyond the Yamuna Khadir community included accessing healthcare (power to withdraw or access human capital), and getting disaster relief during the monsoon floods (power to withdraw physical capital). This list is only a sample of the ways tha t ego alter power manifested within a household's outer netwo rk; the full list is included in Appendices B and C. 75 43 0 100 No Yes Strong Outer Network Power Strong Outer Network Power Freq. Percent Cum. No 75 63.6 63.6 Yes 43 36.4 100 Tot al 118 100

PAGE 137

118 H ouseholds varied in their responses related to outer network power, but the following is a n example of a household (1 of 43 ; 36% ) with stro ng outer network power: This family works on D 's batai ( D, name changed to protect privacy, is a multi generational landowner who has his land on batai with many families) [ withdraw social capital ] D tells them what to grow and how to farm, but they ge nerally discuss what is better [ exchange human capital ] They share is 50/50 for inputs and labor [ exchange financial capital ] D tells him to educate his family, to save money, and to do things that make a profit [ withdraw human capital ] This man/fami ly did not protest. He says that D is taking care of them. None of the farmers working with D protested. D has told them that he will deal with the DDA [ withdraw human, social capital ] ( excerpt from MV019 notes coding for outer network power in brackete d italics ). The following is an example of a household with weak outer network power (1 of 75 ; 64% ) : They have taken 5 bigha land on rent. He buys seeds from L (in community seed shop) or Barafkhana sometimes too [ withdraw physical capital ] His landlor ds did not help like others; they are only bothered about the money [ barrier to social, human capital ] He has not seen people from outside on a regular basis but sometimes companies come to experiment their seeds. Once some company gave him ladyfinger se eds, which were bad [ barrier to human, financial capital ] While we were talking he was preparing to go to the market to sell vegetables in Kondali near Vasundhara [ withdraw financial capital ] He goes to the market 4 5 times a week because he is otherwis e busy in the farm. He thinks that his landlord might have been compensated but not sure [ barrier to financial, human, social capital ] ( excerpt from MV129 notes coding for outer network power in bracketed italics ). Similar to household and community soci al networks, an important question relates to whether households with strong versus weak outer networks were physically close to each other A positive finding might suggest that households near to each other have some of the same social relations and, th erefore, have similar influence and access to livelihood assets For example,

PAGE 138

119 they might have the same landlord, share outside laborers, buy farm inputs and get advice from the same places, or go to the same medical centers. A map illustrating spatial a nalysis of outer network power is shown in figure 5.1 3 below. This map confirms that there were two small areas where outer social network power clustered. Households with strong outer social network s had more access to and influence over a greater varie ty of livelihood assets through connections and interactions beyond the community. There was only slight clustering of households with strong outer social networks. It is possible that area wa s owned by one of the landlords who wa s positivel y involved wi th tenants provided advice, was fair, and helped when asked. The area where households with weak outer network power cluster ed is more difficult to explain because it is an area managed by multiple landlords. Based on the data collected for this disserta tion, I am limited in the ability to explain observed patterns. I t is impossible to know if households had the same outside relations hips but future research could help to understand connections (or lack of connections) with more confidence.

PAGE 139

120 Figure 5. 1 3 Hotspot analysis of outer network power (red indicates clusters of households with high outer power; blue indicates clusters of households with low outer power) (note: interview sites are not shown in the exact location where the household was intervi ewed in order to protect privacy) Combining Three Dimensions of Social Network Power I have presented findings using three dimensions of household social networks: household, community, and outer networks. Because the three dimensions measured mutually exclusive attributes, most households had various combinations of strong and weak household, community, and outer network s In order to assess whether the overall social network was strong or weak I first crossed the three dimensions for each household ( table 5.2).

PAGE 140

121 Table 5. 2. Social network power (household HH, community, outer ) (n=71; 58.7%) Outer Weak Outer Strong Community Weak % Community Strong % Community Weak % Community Strong % HH Weak 3 4 2 3 2 3 7 10 HH Strong 3 4 36 51 1 1 17 24 Because of the semi structure nature of the interviews, n ot all households provided enough social network detail to determine all three of their social network dimensions. I was able to evaluate 71 out of 121 (59%) households for all three social networ k dimensions. Of the 71 one quarter (n=17; 24%) had strong power in all three dimensions. This meant, for example, that they were educating their children and/or had multiple sources of income they had good relationships with their farming neighbors, t hey had a good relationship with their landlord, and were able to access many of the things they needed (seeds, advice, good paying customers) to maintain or improve their livelihoods. However, what was more common represented by h alf of the 71 (n=36; 51% ) was for households to have strong household and community power but weak outer network power. This was primarily due to poor relationships with landlords. The remaining 18 (25%) had other combinations of strong to weak household, community, and outer n etwork s For example two households had weak household and community networks, but strong outer networks they didn't educate their children, didn't have a good relationship with other farmers, but had a supportive and helpful landlord. There were also tw o households that didn't educate their children, had a poor relationship with their landlord, but had good relationships with neighboring farmers.

PAGE 141

122 Social Structure This dissertation conceptualized s ocial structure as the exchange of knowledge and resourc es through farm household social networks A key focus of A im 1 was to operationalize the role of power as it relate d to the influence of and access to livelihood assets through social networks. Social structure wasn't a simple equation of net gain (powe r) in other words, it wa s not a simple sum of household, community, outer social networks R ather I careful ly consider ed the different aspects of household social networks based on the type of relation (farmer, landlord, vendor, etc ), the transaction (i nvest, withdraw, etc ), and the degree of influence (bond, bridge). Households were characterized across a broad spectrum, but finally dichotomized as having strong or weak social structure. In summary, 66 (55%) of households were characterized with weak social structure, and 55 (45%) were characterized with strong social structure. I have not included a map showing results from hotspot analysis because the distribution of strong/weak social structure did not cluster like the separate dimensions of househ old, community, and outer networks. Figure 5.14 below is a distribution map that illustrates households based on strong (red) or weak (blue) social structure.

PAGE 142

123 Figure 5.1 4 Social structure (strong/weak) (note: interview sites are not shown in the ex act location where the household was interviewed in order to protect privacy) Households with strong household, community and outer social networks were characterized with strong social structure; however, households with two or more bonding and bridging ties were characterized with strong social structure even if they had weak household or outer social networks. Households with a weak community social network were characterized with weak social structure regardless of the strength of other social network dimensions or ties because I observed through the analysis that the community social network was the most important group of relationships for household mobilization of livelihood assets

PAGE 143

124 Aim 2: Land Use Change Beliefs and Behaviors The second aim was to describe the beliefs and behaviors of Yamuna farmers related to land use planning Aim 2 address ed the organizational mode of power Recalling the theoretical framework, this was represented by transforming processes and was conceptualized as control over the contexts within which people interact ( figure 5.15 ) In this case, the context was defined as land use change in Yamuna Khadir. Figure 5 15 Organizational mode of power applied to the SLF ( Scoones, 1998 ) Aim 2 was accomplished by carefully addressing two secondary research questions: (3 ) How do farmers perceive their influence on the land use change planning process? And (4) What is the current behavior of the farmers in response to the land use change planning process? In this section, I summarize findings on farmers' land use knowledge related to land use planning and development, farmers' land use behavior, and l and use perceptions. In addition, I also present contextual information related to land development (structural

PAGE 144

125 mode of power) including the historic and current institutional land use change planning process to engage Yamuna farmer s. Knowledge of Land Use Change in Yamuna Khadir Households were asked if there was any development planned for the land. They were prompted to talk about who, when, and what if anything was going on. Only 5 (6%) households provided specific, detailed information, 61 (68%) made some general statements, 39 (43%) gave minimal or vague information; and 20 (22%) said they didn't know anything about development or if anything was planned (figure 5.1 6 ). The most common response was that metro would be constructed (n=51; 34%). Th e second most common response was that a park or, more specifically, a Biodiversity Park was planned (n=28; 18%). H ouseholds also mentioned the possibility of : an airport, bus station, buildings or apartments, a market, a mall, a levy, a nursery, and a pla yground. Nearly every household said that the DDA or government was involved with development of the land. A few households commented that t he MLA (local political advocate) said they didn't have to worry about the government taking the land. A few furt her clarified that whoever was in power would try to take the land, and those in opposition would try to save the land.

PAGE 145

126 Figure 5.1 6 Distribution of land development knowledge In terms of knowledge about the timing of land development, the re was con siderable variation of the 72 households that talked about it (44% of 121 households), only 31 (43% of 72) said or implied that development was occurring. Few households mentioned the location of development, but those who did pointed out that development would occur closer to the main road and along the new metro line, and away from the areas that had the greatest risk for annual monsoon flooding. In response to the question of who owned the land farmers said : the DDA, UP Irrigation Department, the Socie ty, the government, a landlord, the Yamuna River (as in, the land was owned by the river) the government doesn't own the land, and the landlord doesn't own the land. The DDA was reported by 23 (42% of 55 responding) households, the government by 10 (18% of 55) households, and 9 (16% of 55) households said that a landlord owned the land. The diversity of responses and lack of clear consensus or consistency regarding who owned the land was a n important finding. It might indicate confusion among 5 40 28 17 0 5 10 15 20 25 30 35 40 45 Specific knowledge Generic knowledge Minimal or vague knowledge Doesn't know Land Development Knowledge (n=90; 75%)

PAGE 146

127 farmers ab out who technically owned the land or it could point to the larger legal issue of land tenure pervasive across India land ownership can be ambiguous. Adding to the confusion was the issue of compensation of landowners for development of the land. Sixty t wo (51% of 121 total) households mentioned government compensation for development of the land. Of those, half (n=38; 47% of 62) said they believed that the landowner(s) had already received compensation for the land; whereas 24 (30% of 62) said they did not know if the landowners had been compensated or not. Other topics that came up related to land use change and land development included: experience of hut razing, eviction notification, and land taking. One third (n=50; 30% of 121 total) of household s said that either their own hut or huts in general had already been razed. A total of 15 households (9% of 121 total) were already notified of their eviction; interestingly, 5 households (4% of 121) reported instances where DDA already took land "here an d there," making it difficult to know who owned what. I did not find any geographic clustering of land use knowledge. Involvement in Land Use Planning Of the sample of 121 farming households, 68 (56%) talked about involvement in land use planning and de velopment. Of this subsample, 33 (49% of the 68) households said they had participated in a protest or were involved in a court case, and 13 (19% of 68) said they discuss ed land use planning with other farmers or their landlords (figure 5.1 7 ). Nineteen ( 28% of 68 ) reported no

PAGE 147

128 involvement in any activity or discussion related to land use change. Three (4% of 68) reported other behavior s: for example, one household said they had demolished their jhuggi (hut) themselves when they found out that DDA official s were coming to raze their hut so they c ould save some of the materials and make it look like it had already been demolished. I did not find any geographic clustering of land use behaviors. Figure 5.1 7 Distribution of land use behavior Of households i nterviewed, 24 (20% of 121) talked about why they were not involved in land use planning and development activities or discussi ons (table 5.3 ). T he list of reasons for not being involved implied that households possibly want ed to be involved but involvem ent might put them at risk for eviction or harassment Three households said that their landlords told them to protest, but they felt the landlords were exploiting them they thought the landlords should put themselves at risk since it was their land, not the farmers'. Another three said that no one will listen to them, and anyway "who wants to mess with the government?" 33 3 13 19 0 5 10 15 20 25 30 35 Active: protest or court Other activity Active: discussion Not involved Land Use Behavior (n=68; 56%)

PAGE 148

129 Table 5.3 Reasons given for not being involved in land use activity Reason Freq. Perc. Don't own/landlord doesn't own the land 7 6% Landlord exploits them 3 2% Lack of information 6 5% Landlord is taking care 1 1% Lack of time or has to work 1 1% No one will listen/who want's to mess with the gov't 3 2% Disrespectful to ask L andlord 2 2% Landlord compensated so can't help 3 2% Landlord didn't help when huts razed; Landlord 's hut has been razed 2 2% Confusion re: land ownership 1 1% Landlord says nothing can be done 2 2% Gender ( it is not culturally acceptable for women to speak to men ) 1 1% Total 24 20% Perception of Land Use Influence The difference between beliefs and behavior is an important distinction in determining perceived power at the organizational level. Just as lack of involvement is not necessarily an indicator of lack of desire to be involved, involvement i s not necessarily an indicator of belief in ability to influence the planning process. I evaluated 112 (93% of 121 total) households based on their perceived ability to influence land use planning and development of Yamuna Khadir. The majority (n=98; 88% of 112 ) did not believe they had any influence. Even among those involved in active protest or a court case 21 (64% of 33 ) did not believe they had any influence. One household said that the husband went to the protes t, but nothing much happened! t hey don't have land it's not their land so they really can't say anything. Another household said that the husband : g oes to the rallies. The frequency of razing has decreased but in 2014 [ she; the wife] hears that this place will be vacated. This land is

PAGE 149

130 government land and the government needs the land back now that the lease is over. Once the land is taken, we won't be allowed to farm." In the follow up interviews, one man said that he had been going to protests, but the protests had been mostly abo ut the razing of homes and not land use change or metro construction. He went on to clarify that the metro was a government thing it "has to happen." He stressed that the metro was for the "good of the people." He didn't protest the metro because it wa sn't impacting his work; he was more upset about the razing of homes. And he believed that the protests had led to less razing of homes. Notably, the man's house and farm were in the path of the metro construction, and the household had lost 1 1/2 of 8 total bigha. He said that as a result of losing some land, his land was small enough that he didn't need to hire additional laborers, but it was still big enough to support a profitable livelihood; it was a positive impact. Other households had similar sentiments: the metro took a little land from them, but it was for the good of society what they were opposed to was the razing of their homes. With the change in government, households reported a lull in razing of homes. Coinciding with those sentime nts was the near completion of the elevated metro rail and consequent receding construction activity. A year and a half after beginning household interviews, the climate had changed from a pervasive feeling of distress, anger, and fear to that of gentle r outine and agrarian activities (figure 5. 18 ).

PAGE 150

131 Figure 5. 18 Crops growing post metro construction Contrary to my assumption that households next to each other would talk to each other, share information, and, therefore, have similar beliefs and behaviors, I did not find any spatial clustering of households based on similarity of responses related to knowledge, beliefs or behaviors regarding land development in Yamuna Khadir This analysis was however, limited by the data, which did not capture which hous eholds were actually ta lking to each other. Structural Level of Power In addition to assessing household land use change beliefs and behaviors, I looked at city level planning structures A full assessment of the structural mode of power was beyond the scope of this dissertation, but it wa s essential for

PAGE 151

132 interpreting organizational power. As the capital city, Delhi has multiple layers of government that sometimes overlap and conflict. As stated in the introduction to this dissertation, the DDA was cre ated in 1957 in response to rapid population growth ( Delhi Development Authority, 2012 ) Now in it's third edition, t he DDA's current Master Plan for Delhi 2021 is to make Delhi a global metropolis and a world class city, where all the people would be engaged in productive work with a better qua lity of life, living in a sustainable environment ( Delhi Development Authority, 2012 ) However, I found skepticism across the academic literature regarding the ability of government and planning entities in Delhi NCT to equitably engage citizens ( Ahmad et al., 2013 ; Dupont, 2007 ; Kundu, 2004 ; Rao, 2010 ) In addition to the DDA, the National Capital Region Planning Board Act (NCRPB) was enacted in 1985 to "promote balanced and harmonized development of the Region, and to contain haphazard and unplanned urban growth by channelising the flow and directi on of economic growth (on which the urban phenomenon feeds) along more balanced and spatially oriented paths." 19 The NCRPB emphasizes economic growth thro ugh infrastructure development, but t here is little mention of the public in their policies except wit h the last o f six stated goals to promote sustainable development in the region for improving the quality of life." The NCRPB's Regional Plan 2021 section on i mprovement of social infrastructure ( health care facilities, schools) encourages private partic ipation in social infrastructure provision, but no other item in the plan 19 National Capital Region Planning Board website: http://ncrpb.nic.in accessed 8/27/14.

PAGE 152

133 "en courages" private participation and "private participation" is not defined so it is unclear what it means As a city interacting with the global community, Delhi has hosted a ha ndful of mega events including most recent ly, the 2010 Commonwealth Games (CWG) The aspiration to become a global metropolis and a world class city prompted a slum free Delhi 2010' campaign ( Rao, 2010 ) This led to beautification schemes and large scale infrastructur e and urban development, which resulted in demolition of poor settlements and eviction of entire communities it is estimated that 45,000 homes were destroyed in the area of the Yamuna River in preparation for CWG ( Ahmad et al., 2013 ) The DDA has policies for demolition of dilapidate d structures and resettlement of displaced people, but resettlement requires households to be registered with the city government and to hold a title deed stating they have occupied the place for a certain number of years. In the 1980s, the demand for rel ocation became too great, and the government severely limited relocation assistance for slum dwellers ( Ahmad et al., 2013 ) Then in 2000, the Supreme Court instructed the government to stop relocation programs stating: "The promise of free land at the tax payer's cost, in place of a jhuggi [impermanent structure] is a proposal which attracts many land grabbers. Rewarding an encroacher on public land with a free alternative site is like giving a reward to a pickpocket. (Supreme Court, cf. Padhi 2007:77)" ( Kundu, 2004 ; Rao, 2010 ) The normative conception of citizenship in Delhi is that of the homeowner ( Rao, 2010 ) But it is estimated that there is a shortage of 1.13 million homes. With a burgeoning population exceeding 24 million, three quarters of households live in unplanned settlements ( Ahmad et al., 2013 ) The irony is that (not unlike

PAGE 153

134 many global cities) cheap migrant labor is responsible for building much of the infrastructure that has so recently put De lhi in the global realm. R epresentation of disempowered citizens of Delhi is not merely non existent, but in fact institutionally criminalized as a result of an anti poor bias pervasive among government officials and the elite and wealthier classes ( Ahmad et al., 2013 ; Kundu, 2004 ) Delhi has a planning history that mixes neo liberal notions of citizen responsibility and paternal like ( and in some case s severe) discipline There was little to no mention of Yamuna riverbank farmers in the literature; however, key informant meetings only confirmed the sentiment in the literature. A key informant at the DDA 20 said that local communities want the metro to be constructed; that it is convenient, it is safe, it is good for the environment. He said that metro impacts are only positive. He said: On the process of engaging the community: "[The people on the land] are constantly informed to vacate the area tha t belongs to the DDA, but they continue to stay." On the process of removing people: "They are given notice before and are asked to move. In case of Mayur Vihar, they are encroachers and are not allowed to work there. Most of the l and was under lease f or 99, 5, or 10 years. The land was mostly given on lease for raising animals especially dairy activities. Since this is a flood prone area, nobod y is allowed to build in houses temporary or permanent. They are not farmers, they are mere labor er s who h ave taken land on rent or whatever Those people might be farmers in their state but are recognized as informal laborers [in Delhi]" On the requirement of urban agriculture for sustainable city: "Delhi's growing population cannot be afforded on such small scale farming, as such the farming in Delhi is diminishing with every passing day. The Yamuna Khadir farmers are not just an invisible population in the eyes of Delhi, but in the perspective of some of those with power and influence, 20 Key informant meeting held August 22, 2014.

PAGE 154

135 they represent uns avory aspects that Delhi as a world class city seems to be embarrassed by. Aim 3: Livelihood Strategies The third aim was to identify livelihood adaptation strategies of Yamuna farmers in response to planned land use change Aim 3 address ed the individu al mode of power (in this case represented by the household unit) Recalling the theoretical framework, this wa s represented by livelihood strategies and conceptualized as inherent household strength or capability ( power to act ) ( figure 5.19 ) Figure 5. 19 Individual mode of power applied to the SLF ( Scoones, 1998 ) A im 3 was achieved through one secondary research question : (5 ) What do the farmers plan to do in response to planned land use changes that impact their livelihoods? Livelihood strategies are part of a process of awareness, evaluation, and valuing of inherent individual family member's and collective

PAGE 155

136 household st rengths and capabilities. To fully capture the individual mode of power, I expanded the SLF concept of livelihood strategies to include livelihood indicators. Thus, I present findings related to inherent capabilities measured as livelihood indicators (la nd improvements, other skills, other sources of income, and other work, othe r work as middlemen, other land or home), benefits to farming along the Yamuna and last, summarize livelihood strategies. Livelihood Indicators I asked if farmers were improvin g their land and found that they were "tilling the land very nicely," fertilizing, and irrigating as necessary to improve their land so they could produce more crops to sell. Farm practices were driven by current crop maximization and not necessarily foc used on long term livelihood security. Other livelihood indicators emerged from the coded interviews related to: other skills, other sources of income, other work, and work as middlemen (table 5.4 ). I also found that 51 (42% of 121 total) households had property (land and/or a home) in their rural village or elsewhere in Delhi (table 5.5 ). Table 5.4 Livelihood indicators Indicator Freq. Perc. Other skills 3 2% Other sources of income 30 2 0 % Other work 5 4 % Also work as middlemen 3 2 % Total 38 31 % Table 5.5 Other home and/or land in rural village Other home/land Freq. Perc. Other home/land 51 42% No other home/land 16 13 % Total 67 55 %

PAGE 156

137 Benefits to Farming along the Yamuna Households noted many benefits and barriers when describing why they migra ted to Yamuna Khadir, when they talked about land development, and when they expressed reasons why they wanted to stay. Urban benefits emerged as an important motivation for migrating to Yamuna Khadir from rural villages and the desire to stay in Delhi. Many families voiced more than one benefit to living and farming in Delhi, but the main reasons reported were (1) that they earned a higher income in Delhi and (2) that there was not enough land in their homeland. It is ironic that farmers migrated to a highly urbanized, dense city to farm because their rural villages didn't have enough land. Rental prices were probably much higher in Delhi, but the net income was clearly more profitable. Farmers seemed to have much more flexibility in what they could grow, where they could sell, easier access to resources, etc. They were also able to recover quickly when a crop was destroyed by flooding or rains because growing seasons were short at two to three months for urban crops (radishes, chilies) versus six or more months for rural crops (wheat, rice). Findings related to urban versus rural benefits had a few implications for livelihood strategies: first, were the direct livelihood benefits of farming in a city (more profit, close to market); second, were the i ndirect benefits from living in a city (education, healthcare); and third, were the barriers faced in their homeland (not enough land to be profitable, not enough water for irrigation). Some migrants left their homeland because they didn't have opportunit ies (they had to come to Delhi), whereas others saw opportunity in moving (they wanted to come

PAGE 157

138 to Delhi). This distinction could have an impact on motivation to stay in Delhi or return to their homeland and, consequently, impact livelihood options. Liv elihood Strategies Households described a range of livelihood strategies in the case their land was developed and they had to move In summary farmers had a strong desire to stay on their current land as long as that was an option (all 102 responding ho useholds; 84% of 121 total), and would look for more land in Delhi only (n=36; 30% of 102), would return to their homeland only if they could not find more land in Delhi (n=26; 21% of 102), or would return to their homeland without looking for more land in Delhi (n=17; 14% of 102 ) (table 5.6 ). Some households said that they would sell vegetables if they couldn't find more land (n=1; 1% of 102), sell vegetables instead of farm (3; 2% of 102), find another place to live and not farm (4 (3% of 102), or find o ther jobs (16; 13% of 102). E xamples of specific household responses are included in Appendix D Table 5.6 Livelihood strategies by category Livelihood strategy Freq. Perc. Find more land to farm 36 30% Find more land or return to village 26 21% Retur n to homeland to farm 17 14% Find more land or sell vegetables 1 1% Sell vegetables instead of farm 3 2% Find another place to live 4 3% Other jobs 16 13% Not sure 3 2% Total 102 85%

PAGE 158

139 Some households reported specific locations where they would look for land if they had to move, whereas others were vague. The majority of migrant households reported that they would try to stay in Delhi rather than return to their homeland. Figure 5.20 is a map of locations where households reported they would look f or land in the future. These included: Badarpur border, Barola, Bhangeli, Burari, Ghaziabad, Noida, Sonia Vihar, Haryana, Shakarpur, Laxmi Nagar, Siarci, Tilpat, Trilokpuri, and Yamuna bank. Figure 5.2 0 Future locations where farmers will look for mo re land Some farmers described in detail future livelihood strategies in the case their land was developed whereas others mentioned offhandedly a probable action. I categorized 111 (92% of 121 total) households by the level of detail in

PAGE 159

140 their reported l ivelihood strategy. I assumed that households that provided greater detail in livelihood strategies might be better prepared for the future, and therefore less vulnerable. On the other hand, h ouseholds without a plan or scared of the future might be at g reater risk for homelessness and gaps in income. Less than one third (n=38; 31% of 121 total) reported detailed strategies, half (n=62; 51% of 121) gave vague strategies, 11 (9% of 121) said they were not sure what they would do if their land was develope d and they had to vacate, and 10 (8% of 121) did not say anything about livelihood strategies. Contrary to my assumption that households next to each other might talk to each other about what they would do if their land was developed and share informatio n about vacant land, I did not find any spatial clustering of households based on similarity of livelihood strategies This analysis was, however, limited by the data, which did not capture which households were actually talking to each other ; I am limite d in the ability to explain observed patterns. This dissertation was concerned with the response of farmers to human produced threat (development SLF "transforming process" ), but natural hazards (annual monsoon flooding SLF "vulnerability context" ) are a regular threat to homes, health, and livelihoods. Approximately one third of interviews coincided with an unusually long and severe monsoon season (June to September 2013). The entire Yamuna Khadir was flooded and remained submerged for weeks. In recen t years, only a few areas had been submerged and waters had receded within days. During that time, people moved to the road into government proved tents. Some people took work in construction, and some remained idle. There

PAGE 160

141 was some discussion in the int erv iew notes about shifting around after the rains ended, but when I ground sleuthed household locations early spring 2014, nearly all households had returned to their respective hut or field. This indi cates that the Yamuna Khadir had a relatively stable settlement and lends credence to the differentiation in different types of threats to livelihoods and coping mechanisms of poor populations. Linking Research Aims In this section I explore patterns and relationships based on linking research aim s in orde r to answer the primary research question: How does the social structure of the farming community situated along the Yamuna River floodplain in Delhi, India impact livelihood adaptation strategies in response to planned land use change? Through Aim 1, I c haracterized the structure of social networks, conceptualized as the exchange of knowledge and resources through social networks based on farmers' perception of power. Through Aim 2, I addressed the issue of planned land use change by analyzing data on fa rmers' knowledge, behaviors, and perceptions of land development. And, through Aim 3, I specifically focused on livelihood strategies what farmers intended to do in regards to their livelihoods in response to land use planning and development. Based on the theoretical framework of this dissertation, I hypothesized that strong social structure would be associated with similarity of livelihood strategies among households (linking A ims 1 and 3). I also hypothesized that

PAGE 161

142 farmers' perceived ability to influe nce la nd use planning and development would be associated with formal rather than informal livelihood strategies (linking A ims 2 and 3). Before testing hypotheses, I first looked at the association between background characteristics and the three research variables. Background Characteristics I looked at the likelihood of co occurrence between background characteristics (age, tenure, etc.) and the main research variables: social structure (Aim 1) land use beliefs/behaviors (Aim 2) and livelihood strat egies (Aim 3) This was to be aware of the possibility of confounding. For example, if length of tenure co occurred with specific livelihood strategies, I would have to determine if livelihood strategies were related to length of tenure or another indepe ndent variable (such as geographic location). Table 5. 7 summarizes odds ratios between background characteristics and research variables based on simple ordered logistic regression (OLR) models. Background characteristics were run as the independent vari able, separately and then in a full model, and included: small family, Badaun origin, rent (as compared to batai), land size (very small, small, medium, large, very large), male gender, and age (20s, 30s, 40s, 50s+). Social structure, land use, and liveli hood strategies were run individually as dependent variables.

PAGE 162

143 Table 5. 7 Ordered logistic regression (OLR) of demographic characteristics and research variables: social structure, land use beliefs/behaviors, and livelihood strategies Background Char acteristics (odds ratios) Research Variables Small family Badaun origin Rent (v batai) Land size Tenure length Gender (male) Age Full model Social Structure Household 0.545 0.740 1.818 0.817 11.4 00 ** 12.852* 1.473 0.862 2.896*** 2.609* 1.736 1.002 0.589 0.452 p=0.018* Community 0.769 0.933 1.761 0.577 2.414 1.696 1.355 1.368 1.246 1.320 3.000 2.148 0.943 1.212 p=0.715 Outer 1.900 2.233 0.500 0.471 0.173*** 0.158** 1.164 1.347 0.845 0.993 0.649 0.887 1.023 0.718 p=0.009** Bond 1.864 1.716 1 .932 1.139 1.537 1.781 0.816 0.890 1.325 1.288 1.261 0.785 1.5 16 1.582 p=0.454 Bridge 0.868 0.651 1.097 0.820 0.438* 0.380* 1.154 1.280 1.363 1.600* 1.129 0.713 1.241 0.946 p=0.095 Social Structure 0.406 0.276* 0.879 0.396 0.684 0.714 1.346 1.451 1.3 11 1.450 0.892 0.651 1.201 1.166 p=0.010 Land Use Knowledge 0.791 0.953 0.539 0.996 0.546 0.728 0.687 0.768 0.62 1 0.826 0.564 0.705 0.679 0.655 p=0.562 Timeframe 0.805 0.457 0.415 0.375 0.813 1.312 0.749 0.678 0.907 0.997 2.057 1.919 1.142 1.167 p= 0.857 General 1.389 1.619 0.610 1.157 0.845 1.189 0.757 0.896 0.570 ** 0.729 0.643 0.614 0.69 3 0.601* p=0.399 Discussion 2.264 3.142 0.458 0.170 0.865 1.562 0.803 2.173 0.572 0.508 0.224 0.160* 1.294 1.033 p=0.291 Activity a a 0.832 0.444 1.029 1.583 1.489 a Behavior 3.610 2.777 1.049 1.706 0.608 0.550 0.429 ** 0.518 0.870 1.122 0.721 0.777 0.801 0.887 p=0.264 Desire 0.623 0.653 4.023 1.914 0.755 0.329 1.628 2.628* 1.271 0.862 0.363 0.359 0.998 0.782 p=0.240 Influence 0.393 0.655 1.972 0.834 1.2 98 1.106 1.571 1.647 1.348 0.914 0.580 0.456 0.905 0.974 p=0.879 Livelihood Strategies Plan to stay 1.174 1.298 1.566 2.379 1.211 1.070 1.567* 1.157 1.375 1.281 2.446* 2.041 0.902 0.834 p=0.319 Plan detail 1.267 1.466 0.4 3 3 0.367 1.065 1.047 1.107 1 .188 1.027 1.288 1.293 1.622 1.029 1.159 p=0.437 p<0.05; ** p<0.01; *** p<0.001 a Model didn't work probably due to small n

PAGE 163

144 I summarize the full OLR models briefly: Households that rented were 12.9 times more likely to have strong household network po wer than households that were on batai or didn't pay rent. However, they were less likely (0.2 odds) to have strong outer network power. Length of tenure was also related to household network power: for each increase in tenure category, there was 2.6 inc reased odds that the household had strong household network power. No other background characteristics were significantly associated with research variables. Refer to table 5.8 for more detail. Social Structure Livelihood Strategies Hypothesis I hypoth esized that ego alter power (strong/weak) would mediate the relationship between social networks and livelihoo d strategies (refer to Chapter IV : Methods figure 4.8). This hypothesis link ed research A im 1 (social structure as the independent variable ) and Aim 3 ( livelihood strategies as the dependent variable ): Hypothesis #1 : If ego alter power within the social network is strong (strong social structure) then there will be similarity in livelihood strategies among households Hypothesis #2 : If ego alter power within the social network is weak (weak social structure) then there will be variability in livelihood s trategies among households

PAGE 164

145 Households with strong social structure did not differ from those with weak social structure in terms of their plan to stay in Delhi (OR=0.73; SE=0.29; p=0.429). Households with strong social structure were also no more likely to report having a detailed versus vague livelihood strategy than households with weak social structure (OR=1.00; SE=0.35; p=0.993). Both grou ps reported similar livelihood strategies at roughly the same distribution (table 5. 8 ). Examples of typical cases are i ncluded in Appendix E to illustrate the similarity between livelihood strategies of households with strong social structure and househol ds with weak social structure. Table 5. 8 Livelihood strategies by social network strength Livelihood Strategies (DV) Total (n=121) Strong Social Structure (IV) (n= 55; 45 %) Weak Social Structure (IV) (n= 66; 55 %) Future Location 94 (78%) 46 (38%) 48 (40%) Return to homeland 19 (16%) 11 (9%) 8 (7%) Look in Delhi or return home 27 (22%) 13 (11%) 14 (12%) Stay in Delhi 48 (40%) 22 (18%) 26 (21%) Future Plan 102 (84%) 48 (40%) 54 (45%) Find more land to farm 36 (30%) 19 (1 6%) 17 (14%) Find more land or return to village 26 (21%) 13 (11%) 13 (11%) Return to homeland to farm 17 (14%) 10 (8%) 7 (6%) Find more land or sell vegetables 1 (1%) 0 (0%) 1 (1%) Sell vegetables not farm 3 (2%) 1 (1%) 2 (2%) Find another place to live 4 (3%) 0 (0%) 4 (3%) Other jobs 16 (13%) 5 (4%) 11 (9%) Not sure 3 (2%) 1 (1%) 2 (2%) Future Plan Detail 111 (92 %) 50 (41%) 61 (50%) Detailed or specific 38 (31%) 17 (14%) 21 (17%) Vague or generic 62 (51%) 29 (24%) 33 (27%) Not sure 11 (9%) 4 (3%) 7 (6%) In summary social structure of households that described future plans in detail was not significantly different from households that said (or implied) very little about livelihood st rategies. I did not find a statistically significant relationship

PAGE 165

146 between social structure and livelihood strategies. However, there were cases and circumstances where social structure mattered and m ore exploration of this relationship based on quantita tive analysis is provided in the final section of this chapter. Land Use Influence Livelihood Strategies Hypotheses Second, I hypothesized that organizational power (i.e. farmers' perceived ability to influence land use planning and development) would mod erate the relationship between social networks and livelihood adaptiv e strategies (refer to Chapter IV : Methods figure 4.9). This hypothesis links research A im 2 (land use beliefs and behaviors as the independent variable ) and Aim 3 ( livelihood strategies as the dependent variable ): Hypothesis #3 : If organizational power is weak (i.e. farmers do not feel they c an influence the planning process) then farmers will plan informal livelihood strategies. Hypothesis #4 : If organizational power is strong (i.e. farmers feel they c an influence the planning process) then farmers will plan formal livelihood strategies. The majority of households (n=98; 88%) did not feel they had any influence over the planning process. Furthermore, no households planned

PAGE 166

147 formal liv elihood strategies. Based on findings, this hypothesis was not useful for explaining Yamuna Khadir household livelihood strategies since there was no variability to test Rather than consider the distinction in the formality informality of livelihood stra tegies, I found that there was a distinction between households reporting detailed versus vague livelihood strategies. I also found a distinction between households planning to stay in Delhi versus those planning to return to their homeland. Thus, I pres ent findings related to organizational power by exploring the relationship between planning beliefs and behaviors and (1) detailed as compared to vague livelihood strategies and (2) plans to stay in Delhi as compared to returning to homeland. OLR models were run for each independent variable, and then run in a model that controlled for social structure (strong/weak) and other potentially confounding background variables (table 5.9 ). Strong social structure was significantly associated with more detailed land use knowledge (p<0.01) more detailed general knowledge (p<0.01) and perceived ability to influence land use planning and development (p<0.01) Households with detailed livelihood strategies and the desire to stay in Delhi were more likely to be inv olved in activities (protests, meetings) related to land use change (p<0.05) Those planning to stay in Delhi were more likely to want to be involved in such activities (p<0.05) and those with detailed livelihood strategies were more likely to perceive t hat their activities had an influence (p<0.05)

PAGE 167

148 Table 5.9 Ordered logistic regression (OLR) of planning beliefs and behaviors and research variables: social structure, livelihood strategies, and plan to stay in Delhi Planning beliefs and behaviors (IV) Social Structure (control) Livelihood strategies (DV) Full model a Stay in Delhi (DV) Full model b Knowledge 0.338** 0.600 0.616 0.613 0.551 Timeframe 0.710 0.464 0.483 0.871 0.749 General 0.356** 0.717 0.715 1.310 1.212 Behavior 1.281 0.353* 0.3 56* 2.369 2.505 Desire 1.797 1.533 1.530 0.154* 0.189* Influence 5.547** 3.172* 3.370* 0.356 0.370 p<0.05; ** p<0.01; *** p<0.001 a Mode l controls for social structure b Model controls for social structure and Badaun origin The relationship between o rganizational power (in this case the perceived ability to influence land use planning and development) and livelihood strategies wa s in one sense clear: the Yamuna Khadir community perceive d no or minimal influence (n=98; 88% of 121 total) and they were p lanning (or not planning) informal livelihood strategies. But of the 14 households who believed they had an influence on land use development, more than half (n=8) gave detailed livelihood strategies. And, households with detailed livelihood strategies h ad three times the odds of believing they had an influence on land use planning and development. They were also more likely to have strong social structure. And they often have a landlord or someone politically connected advocating for them: When asked i f he thinks he has a say in de velopment, he mentions Subadar Subadar has been elected by the farmers in this area to represent them. He talks to the MLA. He got them Aadhar cards. He represents the farmers nicely. He helps them get IDs. He lives nea rby. If this man needs anything, he goes to Subadar. How is he elected? He is well connected to the people outside s o that is why they elected him ( excerpt from MV034 notes). He doesn't want to leave but he will if the government says they must. They will try to find land nearby or will return home. He prefers to stay. He can work in agricultural labor or construction. There are 7 members: 1 girl, 4 sons. The eldest son is in 8th class learning how to fit windows into buildings. They have a voter card, ID, and Aadhar card. In

PAGE 168

149 case all the farmers are evicted, Delhi won't get vegetables. Vegetables must be transported up to 4 days and will spoil by the time they arrive. Vegetable quality is poor at a big store; the market sells fresh vegetables. Anyone can come and see how they farm so they get high quality products. One time he used too much fertilizer and he threw the entire crop away. They also run a small shop in a brick building adjacent to their home and on the main road. The you nger so n takes care of the shop ( another excerpt from MV034 notes). The above case was not common, but represented a household with knowledge and connections. Such a household might have the potential to act as an important community liaison in the planner publ ic nexus. The more common responses were: A few of the farmers have given land to the DDA. They have taken the compensation. Then they give the land further to farmers to farm and make huts. When the DDA/police come, they don't differentiate between wh ich land has been compensated they just clear everything. They plan to build the Yamuna Biodiversity Park here. It will be constructed here and on the other side (Hasanpur Village). Whosoever is in power takes the land and those in opposition try to save the land. In 2014 the land will be gone. The older man asks: what will happen to the vegetable market? They have gone to protest at the rallies. The opposition party mobilizes the people to protest. They have a voter ID and ration cards. Initially t hey got food through the ration card, but that has stopped because they want the people to go. They considered this a green belt but have re categorized the land as undeveloped/vacant so that it can be developed. They have 2 bigha of land in Badaun among 3 brothers. Two brothers are here and one brother manages the land in Badaun. If the land goes, they will go back. This is the biggest area of land in Delhi. It will be difficult to find more land because 3 lahk farmer s will be looking for more land ( excerpt from MV076 notes). Their hut was run down during the rainy season last year. The DDA says they have bought all this land and they don't want anyone here. Even if they call the landlords, they don't show up. The husband has gone to the protests, but she doesn't go. If they protest regularly then the huts are not broken so often. The metro is being constructed on the other side of the main road and a colony [flats] is going to be constructed on this side. The landlord has told them that it migh t be flats. On May 10th [4 days from the interview], they have been told that the DDA will run down their houses so she doesn't know if she will be here. They will look for more land but they are not hopeful. They may go ho me because they have land ther e (excerpt from MV077 notes).

PAGE 169

150 Quantitative findings are generally limited in the ability to explain why or under which circumstances a behavioral relationship may be occurring. The next and final section of this chapter adds depth to the findings present ed thus far through a qualitative description. Social Structure and Behavior This dissertation employed a case study of urban farmers faced with development of their land in order to examine the relationship between social structure ( ego alter power ) an d livelihood strategies (individual power) within the context of the process of land development (organizational power) Figure 5.21 is a simplified conceptual diagram of this relationship. Figure 5.21 Simplified d iagram integrating Wolf's power and the SLF as conceptualized for this dissertation research I now present five examples of how social network power can interact with household knowledge, decisions, and livelihood strategies as manifest in everyday activities

PAGE 170

151 Weak Social Structure: St rong Household In the case of MV156, the household had strong household network power, but weak community and outer network power: Social Structure: Weak Power: (+) HH; ( ) CO; ( ) OU 21 Tenure = 8 years Land Use Beliefs & Behaviors DDA harasses them and s o does the landlord. No one knows who owns which piece of land. Don't know what the DDA wants, maybe they will construct some park. Livelihood S trategies They will rent a room somewhere if they are made to leave. Figure 5.2 2 Field Image 1 22 21 SS=Social structure; HH=Household network power; CO=Community network power; OU=Outer network power

PAGE 171

152 Th is family didn't have a very positive experience of community and they had very limited knowledge of what was going on. They didn't engage in land use development activities and will have to rely on themselves to find another place to live and work if the ir land is developed. In summary, this family exhibited weak ego alter power, limited organizational power, and limited individual level power. Figure 5.23 below is a diagram of how power manifests in this case overlaid on the simplified conceptual diag ram (figure 5.21 ) above. The interesting finding wa s that they exhibited strong ego alter power within the household and could draw on that power within to enable a future livelihood strategy albeit one that they did not deem ideal. Fi gure 5.23 Case of weak social structure: strong household network o verlaid on simplified conceptual d iagram 22 In order to protect privacy, field images are not related to the family presented in the case

PAGE 172

153 Weak Social Structure: Strong Community In the case of MV109, the household had strong community network power, but weak household and outer network power: Social Structure: Weak Power: ( ) HH; (+) CO; ( ) OU Tenure = 5 years Land Use Beliefs & Behaviors They don't know about the DDA's stand. They do know about the metro. Their hut has never been demolished. Livelihood S trategies They plan to take a place on rent somewhere, maybe work as help in the colonies. They intend to get work through word of mouth. This family had limited knowledge of what was going on with the metro ; nothing specific. They didn't engage in land use development activities, but they did have a strong community network and plan to rely on others to help them find another place to live and work if their land was developed. In summary, this family exhibited weak ego alter power, limited organizational power, and limited individual level p ower. Figure 5.24 below is a diagram of how power manifests in this case overlaid on the simplifie d conceptual diagram (figure 5.21 ) above. The interesting finding wa s that they exhibited strong ego alter power within the community and could draw on that power with others in the same situation to enable a future livelihood strategy.

PAGE 173

154 Figure 5.24 Case of weak social structure: strong community network overlaid on simplified conceptual d iagram Figure 5.2 5 Field Image 2

PAGE 174

155 Weak Social Structure: Strong O uter In the case of MV089, the household had strong outer network power, but weak household and community network power: Social Structure: Weak Power: ( ) HH; ( ) CO; (+) OU Tenure = 1 year Land Use Beliefs & Behaviors Their huts have been broken but they don't know why. They know about the DDA, metro construction, said that landlord has gotten compensation. Metro people do not interfere, and once the construction is over, they will be allowed to live there again. They have protested in the rallies. Li velihood S trategies If there is land they will sow here. If it is taken away they will go back. Figure 5.2 6 Field Image 3

PAGE 175

156 This family had knowledge of what was going on with the metro, and the exchange of ownership of the land. They engage d in la nd use development activities through rallies and protests, and they had a good relationship with the landlord who would let them farm as long as possible. If the land was taken, they planned to return to their homeland since they had a weak community net work and wouldn't be able to rely on "word of mouth" to find more land in Delhi. Having power through the landlord relationship meant better and more accurate land use knowledge; but weak community power meant they were still vulnerable if they had to lea ve their land and consequently break the landlord relation. In summary, this family exhibited weak ego alter power, some organizational power, and limited individual level p ower. Figure 5.27 below is a diagram of how power manifests in this case overla id on the simplified conceptual diagram (figure 5.21) above. The interesting finding wa s that they exhibited strong ego alter power outside the community (in this case with their landlord) and could draw on that power to increase their knowledge of planne d land change. Knowledge of planning and development meant that they could be more prepared to return hom e if they had to vacate the land

PAGE 176

157 Figure 5.27 Case of weak social structure: strong outer network overlaid on simplified conceptual diagram Weak S ocial Structure: Weak Networks In the case of MV076, the household had weak household, community, and outer network power: Social Structure: Weak Power: ( ) HH; ( ) CO; ( ) OU Tenure = 21 years Land Use Beliefs & Behaviors The counselor is doing work to k eep them from being harassed. She is protecting their huts. Whosoever is in power takes the land and those in opposition try to save the land. A few of the landowners have given land to the DDA (taken the compensation), then they rent the land to farmer s to farm and make huts. When the DDA/police come, they don't differentiate between which land has been compensated they just clear everything. They plan to build the Yamuna Biodiversity Park here. It will be constructed here and on the other side of the river. In 2014 the land will be gone. This land was considered a green belt but it has been re categorized as undeveloped/vacant so that it can be developed. They have gone to protest at the rallies. The opposition party mobilizes the people to protest Livelihood S trategies Maybe they will farm next year. They put fertilizers and irrigate the land. If the land goes, they will go back. This is the biggest area of land in Delhi. It will be difficult to find more land because 3 lakh (300,000) farmers will be looking for more land.

PAGE 177

158 Figure 5.2 8 Field Image 4 This family had weak power through their household, community and landlord relations; however, as a consequence the family seemed to take it upon themselves to figure out what was going on by going to the protests and rallies organized in response to metro construction and land deve lopment. Their knowledge sounded like it came from speeches at the rallies 3 lakh (300,000) farmers" sounds like a speech and not something the local farmers wo uld be aware of in their daily lives. Gaining knowledge this way led to the belief that they would not be able to find more land and thus they won't even try as they simply planned to return home. In summary, this family exhibited weak ego alter power, so me

PAGE 178

159 organizational power, and weak individual level p ower. Figure 5.29 below is a diagram of how power manifests in this case overlaid on the simplified conceptual diagram (figure 5.21) above. The interesting finding wa s that when weak ego alter power was combined with an attempt to increase their organizational power (thro ugh knowledge from the outside) it seemed to block their power within to seek out a better livelihood. Figure 5.29 Case of weak social structure: no strong networks overlaid on simpl ified conceptual diagram Strong Social Structure: Strong Networks In the case of MV061, the household had strong household, community, and outer network power: Social Structure: Strong SS Power: (+) HH; (+) CO; (+) OU Tenure = 8 years Land Use Beliefs & Behaviors They had 2 acres here but have lost 1/3 acre to metro construction. Three months ago the metro construction began on their land. He doesn't know if the landlord has been compensated for the land that was lost. The farmers have been having conve rsations about development for 10 15 years but now the metro is coming so he doesn't know how much longer farmers will be here. Someone from the family always participates in the rallys/protests. The lady who was recently elected

PAGE 179

160 doesn't let the DDA raze their homes so there is some protection. He never talks to the metro laborers even though the construction is going right through their land. Livelihood S trategies He has heard that there is some land in Siarci near Noida. He says that whoever has contac ts will find land here [in Delhi] and those without contacts will return to their homelands. He knows some people that might connect him to land in Badarpur. They had a tv that worked on batteries but they sent it home with all their expensive stuff when t heir huts were razed. His brother is in the 10th class. They all want him to get a good job. The farming here is fine. They try to do their best but they don't always get the best results. Delhi is the capital so there are more facilities here. Fig ure 5.30 Field Image 5 This family experienced land taken from them despite active protest and participation in rallies and serious discussions with other farmers. They had specific knowledge of land development (well, as specific as any of the farmer s

PAGE 180

161 interviewed no interviewed household s said they were actively engaged by the DDA ). Although they didn't know if their landlord had been compensated for the land, they did have a positive relationship with him Note that i t is not uncommon for Indians to refrain from discussing c ertain topics with subordinates. This family had specific strategies for their future livelihoods. In summary, this family exhibited strong ego alter power, some organizational power, and some individual level power. Figure 5.3 1 below is a diagram of how power manifests in this case overlaid on the simplified conceptual diagram (figure 5.21) above. Despite clearly havi ng to deal with loss of land, they had strong relational power through the household, community, and landlord, which may have play ed a role in their knowledge of land development, participation in rallies, and clear future plans. Figure 5.31 Case of strong social structure: all strong networks overlaid on simplified conceptual diagram Social Structure and Beh avior Summary In the first four cases of households that exhibited weak social structure,

PAGE 181

162 examining the different types of resources they did have access to through social relations illustrates how social network power can interact with household knowledge decisions, and livelihood strategies. In the first case, resources and livelihood strategies were generated within the household. I n the second case it came from the community. I n the third, the outer network i ncreased the household's knowledge at the organizational level which didn't translate into power to act, but seemed to increase their power to react. And, in the fourth case an attempt to increase the household's organizational level power (knowledge of land use change) without any underlying s ocial network power to respond to that knowledge, seemed to have an adverse affect on the household's power to develop livelihood strategies. In the last case, a household that exhibited strong social structure, they were better situated to act in a way t hat could protect their livelihoods. They were limited by the power exhibited over them at the organizational level, but they were able to draw power through other modes in pursuit of agency Each of these cases had similarities and clear differences in t he ways that households drew power through social networks to experience, understand, and act/ react to their situations. Throughout this chapter, I presented findings that demonstrated different ways social structure impacted livelihood adaptive strategie s in the context of planned land use change. By exploring collected data quantitatively and qualitatively, I showed both how and why power relations impacted problem framing and decision making related to livelihoods of Yamuna

PAGE 182

163 Khadir farmers faced with la nd development. The next chapter is a discussion of the implications of some of these findings.

PAGE 183

164 CHAPTER VI : DISCUSSION To gain insight into how power relations and social networks impact the participatory planning process, this dissertation took a s tep inside a marginalized group of poor urban farmers. I used a livelihoods approach to capture formal and informal spaces and found that marginalized households we re not equally disadvantaged. This dissertation demonstrates that households can lack powe r at the structural and organizational levels, but find pathways to power at the ego alter and individual levels in pursuit of meeting daily needs. Aim 1: Social Structure For the first research aim, I characterized household social structure of Yamuna Khadir farmers in order to address the ego alter mode of power. Social structure was conceptualized as a complex web of household social relations enabling and blocking access to knowledge and resources. I will highlight two contributions that I have ma de in addressing this aim. First, I show how power negotiation works in and through social relations and second, I demonstrate that power can be unevenly distributed among relations. Measuring Power Relations Measuring power as a complex, relational phe nomenon is challenging Developing a method and measuring power relations is a meaningful contribution to the literature because few studies actually measure power in social networks

PAGE 184

165 in the planning literature ( Dempwolf & Lyles, 2012 ) The ability to mea sure empirically how people access and influence knowledge and resources through the people they know is an important step toward understanding how connections through social networks have the potential to increase or decrease representation of poor and ma rginalized groups in the events that impact their lives. In Mdee's ( 2008 ) ethnographic study to understand how agency of individuals related to patterns of collective community participation, she found that connections through social networks could effectively increase representation of women, the young and old, and poo r in the participation process She initially intended to look at how individual attributes of age, gender, and wealth predicted formal public engagement, but instead she found that people could exhibit agency through everyday informal interactions. Base d on this and other studies ( Beebeejaun & Vanderhoven, 2010 ; Daniere et al., 2005 ) we know that power negotiation occurs inside and outside the formal planning arena, and that informal connections can provide agency through representation by those actively participating in the formal arena. But existing studies are limited in explaining how power negotiat ion works in or through social relations. I operationalized power based on three qualities: type of transaction (five capitals), direction of transaction (investment, withdrawal, exchange or barrier ), and degree of influence (bond/bridge). In doing so, I was able to measur e the flow of power between people I illustrated the relationships that poor marginalized households established that gave them access to and influence over livelihood assets (human, physical, social, natural, and financial capital).

PAGE 185

166 But, I also demonstrated how relationships could create barriers to livelihood assets something only recently addressed in the literature ( Adhikari & Goldey, 20 10 ; Eyben et al., 2008 ; Newman, 2008 ) For example, one interviewed household said that they had filled out an application through a local advocate for an identification card so that they could receive basic social services including food rations. But, their landlord who charged less for rent than many other landlords refused to provide proof of residency. Consequently, they could not submit the app lication. This example illustrates how households weighed costs (no IDs) and benefits (cheap rent) of accessing different resources through relations. By capturing the complexity of differing power relations that comprise household social structure, I de monstrate d how households negotiate d and made trade offs that in turn increase d or decrease d their ability to control the events that impact ed their lives. Aggregating Power Relations The second important contribution was in demonstrating that household p ower was unevenly distributed among relations. It is tempting to assume that a household that has power in one relationship can also draw resources from other relationships; however, this study illustrated that is rarely the case. In other words, not all relationships provide the same things be it small loans or advice on getting rid of a plant fungus. Conflict theory posits that power is acquired, negotiated, and disbursed unevenly across individuals and groups. Not all relationships are equal but

PAGE 186

167 neit her are livelihood assets For example, human capital and financial capital don't aggregate easily ( Anderson, 20 12 ) This made characterizing overall social structure difficult I slowly took a multi step, mixed methods approach in order to combine components. I carefully considered each household's power to access or influence each asset through each relation and then considered how each relationship impacted other household relationships In the end, I determined that 55% of Yamuna Khadir households could be considered as having weak social structure, and 45% could be considered as having strong social struc ture. However, through the process of aggregation, I found power to be more dynamic; it was most common for households to have some strong, some weak (absence of power) and some negative (subtractive power; power as zero sum) power relations. In review ing the planning literature, I had trouble find ing studies that looked at power as a dynamic phenomenon Rather, measures of power were often limited to one type of relation, within one context (at planning meetings between planners and the public), or at the structural level (governance policy). Power may not be equally distributed among individuals or households, but it is also not equally distributed among relationships For example, a young farming household might have a weak relationship with their landlord (and therefore pay a high rent), but might have a strong relationship with customers in the market (make a good profit, and thereby offset the cost of high rent). Because human a gency is enabled through relationships where one draws power or has influence,

PAGE 187

168 it is more revealing to look at the complex network of power relations and not limit investigation to one type of occurrence. Power may be a dynamic phenomenon, but it is not necessarily arbitrary. In their study on poverty and marginaliz ati on of a community of Honduras peasants Nygren and Myatt Hirvonen ( 2009 ) suggest that who is connected to whom was intentional. If households develop relationships to best suit their immediate livelihood need s, then they might have few interactions with those outside their daily activities and few to no connections to outsiders who might benefit them in the longer term. The fact that few Yamuna Khadir households had few interactions with people outside daily farming activities (in other words, they primarily interacted with v endors, the seed shop owner, laborers, etc. ) indicates that households might be put ting all their energy into relationships of immediate opportunity or mutual benefit. Even most friendshi ps had a purpose such as providing exchange of labor or small loans F armers established relationships carefully and many said that they helped each other when they could, but (stress ed ) that everyone was in the same situation. Power, in this case coul d manifest as power to do something ( borrow money from another household ) or as power not to do something (not lend money to another household ) Another example was the relationship farmers established with vendors: a household had the power to require t he vendor to pay in full up front or split the profit after sale in the market. The implication of putting a lot of household energy into establish ing relationships to meet daily needs is that they might miss out on developing relationships that

PAGE 188

169 could ben efit them in the longer term. And, unless approached by an outsider few households might be un likely to reach beyond day to day interactions Implications The implication of the two findings that I have highlighted that power in social networks is a me asurable but unevenly distributed phenomenon is that there are many pathways to power. And, conversely, there are many potential obstacles to power; relations can disempower or block access to resources. An important contribution to the literature is t hat I have provided a more nuanced measure of power relations by demonstrat ing not only how power can be drawn (generated) through social networks, but also how it can be blocked or taken away (zero sum). I have also suggested that in the pursuit of rela tionships that benefit daily needs, marginalized households may not reach outside the community to establish relationship s with those who might benefit them in the longer term. Everyday exchanges may cumulate into lived reality and future livelihood optio ns, but they can also limit future options if they are only meeting daily needs. In tying back to participatory planning, this dissertation picks up where many studies have left off. In the literature review, I summarized that a number of studies have beg un to explore the links between power, agency, and social networks within communities, but they do not go beyond the context of formal public participation. Beebeejaun and Vanderhoven ( 2010 ) found that tensions and disagreements often played out through private/informal interactions

PAGE 189

170 (through social networks) but they didn't m easure informal interactions. Similar findings emerged in the case study by Daniere, et al. ( 2005 ) who looked at the relationship between social capital and participation. They found that participation could effectively take place outside formal structures through informal connections ( also through social networks), but again, they didn't examine the actual exchanges occurring through informal interactions. Furthermore, Ad hikari and Goldey ( 2010 ) argue that i t is not just about the existence of social capital what is important is what social c apital means and how people use or benefit from it. This research takes the next step by examining power relations (not just so cial capital) in the informal spaces created by poor and marginalized households who are outside the formal participatory process. In doing so, I demonstrate d empirically how people use d and benefit ed from those relationships. The one implication of these findings is the need to develop relationships between marginalized households and those who might be able to provide influence or access toward meeting long term goals including secure land tenure, education, health care, and other social determinants of health that impact daily productivity and success. Yamuna Khadir farmers were outside the formal invited spaces of the city of Delhi, and unlikely to pursue relationships in th e formal city space because they might be discriminated against as illegal citi zens or squatters. In other words, it could do more harm that good for them to reach out. Which makes it particularly important for outside advocates (city

PAGE 190

171 officials, local NGOs, universities, formal citizens) to develop relationships with them. There are a number of ways that inside outside relationships might be promoted through the ego alter mode of power with the goal of increasing opportunities But, i nside outside relationships are unlikely to simply precipitate; establishment requires farmers to physically interact with outsiders at a specific time and in an actual location where a transaction (sell produce, gain advice) is expected. This could take the form of a co operative market, a farm to table partnership with local restaurants, or through agri tourism. Additional research would benefit from future inquiry into the types of relationships that could benefit farmers long term, and identification of a time, place, and the type s of resource s that could be exchanged to bring the farmers face to face with those people. Aim 2: Planning Beliefs and Behaviors For the second research aim, I captured the beliefs and behaviors of Yamuna Khadir farmers related to land use planning in order to address the organizational mode of power Drawing from the concept of agency that you have to know the goals and beliefs of an individual and his or her actions to understand if a person is an agent of his or her own will I distinguished beliefs from behavior. This is based on the notion that lack of involvem ent in the planning process does not necessarily mean lack of desire to be involved, and, conversely, involvement is not necessarily an indicator of belief in ability to influence the planning process. To better interpret findings, I also pulled in

PAGE 191

172 histor ical and current sociopolitical context to incorporate structural power. In this discussion, I first highlight that farmers acknowledged they were outside the decision making space, which may have further reinforce d their marginal status; and second, that contrasting formal and informal spaces can bring to light the process that led a group or groups to perceive themselves as outside the decision making space. Outside the Decision Making Space I found that the majority of participants did not believe they had any influence on land development Even among those involved in active protest or a court case few believe d they had any influence. So, even those who wanted to stop development on the land they farmed, and acted to stop development, in effect did not have agency because they did not believe they actually could stop development Complicating this, many believed that development had to happen. One household explained that metro construction was a government thing it "h[ad] to happen," and it was fo r the "good of the people." In probing deeper, I found that farmers expressed more anger and sadness about the razing of their homes than land development, over which they felt they had no influence. Through the act of and acknowledgement of ineffective participation (going to rallies or meetings but not believing it could make any difference), farmers reinforced their marginal status. Given the historic and continued eviction and displacement of the poor in Delhi, and inclusion in planning and governanc e often limited to elite groups, it

PAGE 192

173 was not surprising that the Yamuna Khadir community was not involved in the planning and development process. A few households had pending court cases against the DDA, but it was in request of delineation of land that w as DDA owned from land still under lease by the "landlord." Furthermore, protests and rallies were more of a peace offering than a threat or challenge to developers; gatherings were intended as a plea to the government to stop razing homes to construct th e metro, but leave them in peace farming alongside the new infrastructure. Taking a broader perspective, but requiring more inquiry, this could be a reflection of Kabeer's habits of democracy: in that the presence of a democratic government is not enough habits of democracy must be instilled in the values and behavior of all citizens ( Kabeer et al., 2012 ) Unequal Access to Participate It is clear that Yamuna Khadir farmers did not have equal access o r motivation to participate as part of the Delhi "publ ic," nor did they view the "problem" at hand through the planners' lens and disembedded from personal agendas. These two normative assumptions about participato ry planning as discussion in Chapter II: Literature Review, related to access/ motivation and pr oblem framing, are not often called into question, but are critical in determining the success or failure of public participation. This dissertation may present a case on the far end of the participatory spectrum (in that Yamuna Khadir farmers weren't ask ed to participate), but findings support the notion set forth by McAlister ( 2010 ) that those who don't participate may do so not out of apathy, but

PAGE 193

174 because they don't feel their opinions will be valued, don't feel they have a right to participate, or don't understand the proposed project. F urthermore, among Yamuna Khadir participants who attempted to have a voice in the land development process, few felt their voice had any impact. It is for this reason that Cornwall ( 2008 ) calls for a contrast in the relationship between invited spaces for participation and tho se peop le create for themselves There is also need to address the process that le a d s a particular group of people to perceive themselves as outside the decision making space. Because, even if the Yamina Khadir farmers are formally invited to participate and g ive voice to the development process occurring on the land they farm, they may not know what to say: "Societal groups whose historical experience has been marginalization from politics and decision making may not have a clear sense of their interests as a group, nor of an agenda for change. This may require the creation of spaces of their own, within which to begin a process of becoming aware of their specific circumstances as a group and articulating an agenda for action to address the specific inequities that they face. Measures for political empowerment are inadequate if they simply involve establishing quotas so that people from particular groups are officially given seats at the table because they are limited to seeking inclusion within a political sy stem that is fundamentally hostile to historic ally marginalized social groups ( Eyben et al., 2008, p. 16 ) Some Yamuna Khadir households were very vocal, even articulate, abou t on going development, but there was a lack of overall coherent voice. Farmers said they talked to each other, but I didn't get any sense of what farmers wanted as a community. My inability to create a cohesive summary may indicate that farmers did not have a clear sense of themselves as part of a community. If they were asked to participate, what would they say? How would they want to participate or be engaged? Who would represent and give them a collective

PAGE 194

175 voice? To the outsider, the most likely ca ndidate to represent the community might be one of the more powerful a farming landlord or shop owner; however, that person may not experience the same lived reality as the majority of the people. Each household exhibited complex social networks where kno wledge and resources were exchanged to a greater or lesser extent. This research captured the process of exchange, and differentiated the more and less connected. Measuring social structure can bring to light the people with the greatest potential for mo bilizing or acting as the voice of the group; however, that person may be unaware of his or her potential or high status with in the group and, furthermore, may not even be aware that they are considered part of the group. Implications One seemingly obvi ous implication based on findings from this research aim is the need for an invited formal space for this population; however, I question the feasibility in this case the government wants them removed, not invited to the table So where is the potential t o create space for the Yamuna Khadir households to generate power in their attempt to gain agency in the planning process? How can they find collective voice and who will actually voice it? Future work related to helping farmers find their voice is warr anted that could inform policies for creating a collective or communal space Through my dissertatio n findings, I found that power wa s exhibited as a zero sum at the

PAGE 195

176 organizational level. In the agency structure dynamic, farmers had no room for agenc y be cause the structure excluded them as formal member s of the public an exclusion further bolstered at the structural level. It is difficult to say whether farmers could gain power to have a voice as long as power is over them (zero sum) through government structures and the policies and processes related to land development. It is possible that structure overwhelms any chance for agency in this case. How can farmers be re defined as formal members of society? Two potential areas to pursue are (1) formal integration of urban farming into the city economic system and (2) government acknowledge ment of the land along the Yamuna floodplain as agriculturally productive Power over exhibited through the organizational mode, also leaves little space for farmers to have a say in the planning and development process. One way to create space to generate power th rough the organizational mode could be to implement policy related to food security that prescribes locally sourced food and thereby integrates local farme rs into the dialog on land use Aim 3: Livelihood Strategies For the third research aim, I identified livelihood strategies of Yamuna Khadir farmers in order to address the individual mode of power Households described a range of livelihood strategies w hen asked what they planned to do if their land was developed and they had to move. Livelihood strategies are based on the notion that households see future options. Thus, I used a few d ifferent

PAGE 196

177 approaches to capture how households perceived their shor t term future. A notable finding was that where marginalized people choose to go is not the same as where they are displaced They may choose to live in poor or insecure conditions and not have a clear sense of their future, but there are benefits to whe re they choose to live. Maximizing Options O verall, nearly every household expressed a strong desire to stay on their current land as long as that was an option and would look for more land only in Delhi, would return to their homeland only if they could not find more land in Delhi, or would return to their homeland without looking for more land in Delhi. Less than one third of farmers described future livelihood strategies in detail, whereas nearly half only mentioned offhandedly a probable action. Ther e was also a small group that pined that they did not know what they would do. Additionally, in trying to understand how households envisioned their future, I looked at how farmers treated the land through agricultural practices, why they migrated to Delh i (and how they ended up in Yamuna Khadir), how they viewed Delhi resources in comparison to resources available in their homelands, and how they responded to the annual monsoon floods. The vast majority of Indian citizens are in rural agriculture, but, over the last few decades, Indian p olicy has led to rising food prices and declining profits for rural farmers ( IARI, 2013 ; Kundu, 2011 ) Consequently, many rural farming families have moved to urban areas in search of better livelihood opportunities It

PAGE 197

178 is not uncommon for rural agriculturalists to continue to pursue agriculture after urban migration ( De Bon, Parrot, & Moustier, 2010 ) And I found that the livelihood pursu it of agriculture on the Yamuna River floodplain was not simply a stopgap measure while families looked for other urban work, but rather the end goal for most households. Urban agriculture in Delhi was a reliable, profitable way of living that allowed hou seholds to be better connected to their customers, to have access to better healthcare, to be able to adequately educate their children, and to have opportunities for alternative and temporary income (own a dry goods shop, work in construction). The fact that few households detailed future strategies does not automatically mean that they were not thinking about the future. Despite having improved their livelihoods, farmers continued to invest all their resources in their livelihoods, which may have left s carce resources to focus on other things even future livelihood plans. Based on findings, it is apparent that where marginalized people choose to live or work is not the same as where they are displaced They may choose poor or insecure conditions, but th ey choose a particular place because they see benefits too ironically, the geographic disadvantage is often superseded by the geographic advantage. And if marginalized people are displaced the new place will need to have similar "advantage s or they will lose any tenuous access to resources that they had. The implication is that displacing a "problem" population through relocation projects can in reality increase or exacerbate problems by breaking social networks and disconnecting access to resources res ources that

PAGE 198

179 may b e place specific including access to markets, hospitals, water fo r irrigation, and cheap daily labor Implications At the individual level, power is inherent in the individual ( or, as in this dissertation the household). There are an in exhaustible list of ways to building skills and self confidence, which can increase inherent power and livelihood options For example, I found that f ew farmers were literate or could speak English both of these skills severely limit ed livelihood options in Delhi. Farmers also faced a cultural barrier in the transition from a rural to urban environment and were afraid to go into "the city" because it's "for rich people not people like us." It is common for immigrants ( Yamuna farming migrants can be consi der immigrants based on the vast cultural difference s between village life and life in Delhi) to go through a period of ad justment and immersion. But, acculturation requires cultural skills building. Creating or increasing an individual or household's in herent power to act lies in building skills and self confidence. Linking Aims Embedding Beliefs and Behaviors The primary objective of this dissertation was to present a case study of urban farmers facing land development with the purpose of explori ng the

PAGE 199

180 relationship between social structure 23 and livelihood strategies. As detailed in Chapter V: Findings, I summarize the following relationships based on quantitative analysis: Regardless of strong versus weak social structure, h ouseholds reported : o S i milar variety of livelihood strategies o Similar desire to stay in Delhi o Similar amount of detail in reporting household livelihood strateg ies. Strong social structure was associated with more detailed land use knowledge, more detailed general knowledge, and perceived ability to influence land use planning and development (statistically significant). Households with detailed livelihood strategies and the desire to stay in Delhi were more likely to be involved in activities (protests, meetings) related to land use change (statistically significant). Those planning to stay in Delhi were more likely to want to be involved in activities related to land use change, and those with detailed livelihood strategies were more likely to perceive that their activiti es had an influence (statistically significant). 23 Social structure is defined as household power to influence/acce ss livelihood assets through social networks.

PAGE 200

181 Reinforcing Marginalization When I considered the varying albeit limited pathways to power Yamuna Khadir households chose, it reflected the literature on poor and marginalized groups ( for example, Nygren & Myatt Hirvonen, 2009 ) Agency and power empowerment interacted in ways that produced (mostly) disengagement from participation related to land use ch ange. Which, in turn further reinforced marginalization and perpetuated power inequality. I expected to find more examples through my fieldwork of how livelihoods acted as a window into the multiple relationships of power and resistance through cultural and political activities in addition to the obvious economic focus. I observed only a few cases reflecting studies ( for example, Whitson, 2007 ) that found marginalized people use resources in new ways within the context of labor and livelihood activities to actively or unintentionally resist the political system. I found examples were power relations were expressed as exploitation, subjugation, and resistance. Exploitation played out when a household offered a landowner slightly more rent for land alread y rented by another tenant farmer, and the landlord pressured the current tenants either for more rent or to move. Subjugation played out in some batai (share cropping) land situations in which land "owners" refused to rent because they wanted complete co ntrol over land productivity and profit, and tenants were forced into a batai situation despite wanting to rent because there was no other vacant land. This was not always just about the bottom dollar (market economy); it was a process of negotiating grea ter control over livelihood opportunities. For example, there was potential

PAGE 201

182 financial security in batai (if the crop failed, the loss was shared with the landowner and no rent was due); however, some landowners were so prescriptive of which crops to grow, that tenant farmers preferred to rent. If the landowner could gain greater control the land wouldn't necessarily go to the highest bidder but to the farmer who submitted. This also occurred between farmers who sublet their rented land. And, resistanc e played out when tenants refused to go to protests or rallies specifically because the landowner told them to; tenants felt it was the landowners' responsibility to protect the land not theirs'. In understanding livelihood assets of poor and marginalized households, the literature suggests that households may pursue relationships of opportunity but may also end up in relationships of exploitation; assets cannot be separated from vulnerabilities ( Anderson, 2012 ; Arun, Annim, & Arun, 2013 ; Parizeau, 2015 ) Building on existing research, this dissertation illus trated how power relations across multiple levels of influence co occur with assets and livelihoods. Each research aim was designed to address a different mode of power. I'll review findings: at the structural level, households had limited power power wa s exhibited over them as a zero sum in that they weren't even acknowledged as formal citizens. At the organizational level, some farmers did protest or go to court there was limited power to engage related to the process of land development, but power ove r households was more common. At the ego alter level is where households had the greatest potential for generating power they developed relationships that maximized their livelihood assets and opportunities at the household level, within the community, an d beyond the community. But, at

PAGE 202

183 the individual level, power was limited again households struggled to enable their inherent power or, rather, power within Structural and organizational levels exhibited a strong power over farmers and constrained their options. Although I found that many households had diverse skills, the fact that there was limited opportunity to use those skills translated into households with low self confidence they were disempowered. In many cases, I observed households self perpe tuate marginalization by literally throwing up their hands and exclaiming, "what can we do?" But in a few cases, households kne w the system needed to change! those households could be considered as being ready to be empowered (if a space were created for t hem to do so). Those were the households with the greatest potential to benefit from intervention. When I examined the interaction between the four modes of power o n households, I found that social network power was associated with household knowledge, decisions, and livelihood strategies. It was not a simple equation of having power at one level that could provide access to power at another level or conversely, not having power at one level that limited power at another level. But rather, power exhibi ted as a spectrum from zero sum to more generative types, which influenced which modes of power that households could draw from in order to gain knowledge, find opportunity, and engage in society and pursue a sustainable livelihood. I found that it was on going power negotiation that determined influence and access to livelihood assets. The everyday negotiations that I observed combined to produce an overall influence on household planning

PAGE 203

184 beliefs and behaviors and livelihood strategies. Implications Exa mining power at each level improved understanding of household decision making. It also illuminate d household options and where the potential barriers to gaining power reside d For example, power over whether external or internalized oppression b locked h ouseholds from formal participation or engagement with government and formal (the normative definition of) society. But it is not so straightforward to remove barriers (the particular economic and social conditions) that restrict options that people perce ive as available and legitimate. Social exclusion can be deeply engrained in culture even when the legal system makes discrimination illegal. Community Representation In tying back to the notion of participatory planning and community engagement, I w ant to return to the question: who represents the community and what community? Communities are not homogenous; they are dynamic, multi cultural entities, representing many voices ( Panelli & Welch, 2005 ; Snodgrass et al., 2008 ) Yamuna Khadir households varied in background attributes, social structure, land use beliefs and behaviors, and livelihood strate gies. Based on recent estimates that 40% of households in Delhi are self employed, three quarters live in unplanned settlements ( Ahmad et al., 2013 ) 22% live in jhuggies (informal structures) ( Prashar et al., 2012 ) and 9.84% are

PAGE 204

185 below the poverty line 24 ( Government of India, 2013 ) the Yamuna Khadir does not represent the extreme poor in fact, households might be considered typical poor Delhi residents except that they are agriculturalist s rather than the more common construction workers, rickshaw drivers, and domestic help. And in reality, more than one quarter of Yamuna Khadir households were involved in construction, drove a rickshaw, or worked as domestic help. Although location on t he Yamuna River floodplain is unique only an estimated 2,500 households live/work on the floodplain within Delhi NCT ( Cook et al., 2015 ) a city of 2 5 million people the research population was representative of many other informal and marginalized communities in Delhi. The Yamuna Khadir, while diverse, had a relatively stable population. More one third of households reported tenure of more than 10 y ears some multi generational. But, there were also more transient members. Twenty seven percent of interviewed households were not agricul turalists, but provided for or relied on the agricultural members for their livelihoods. Although not directly depe ndent on the availability of land, dependence on those who were dependent on the land meant that they also had a stake in land use change. This challenged me to think carefully in determining which households to include/ exclude, and I ultimately limited analysis to farming households to minimize the myriad variables that mediate and moderate decision making Just as 24 The monthly per capita poverty line for urban Delhi is 1,134 INR or 37.8 INR per day (approximately $20.50 USD or $0.70 per day) (Government of India, 2013). India sets it's own poverty rate. It is currentl y slightly more than half the $1.25 per day set by the World Health Organization (WHO) as the poverty line ( www.who.int -accessed 12/5/14). It is estimated that up to 60% of the Indian population lives on $1 per day or less.

PAGE 205

186 communities are not homogeneous, they do not have a definitive roster of members. This reflects findings in the literature that c ommunity m embers often change and shift depending on the issue or point of focus ( Lyons et al., 2001 ) With a clear understanding that Yamuna Khadir was comprised of multiple voices, I was able to look more deeply into the relationship between power relations and concepts of representation, knowledge, problem framing, and marginalization. The literature repeatedly calls for in depth exploration of power relations within communities to better unders tand the role of representation in the participatory planning process. And, particularly in the case of marginalized groups, how those who aren't involved either because they are not invited or because they choose not to be involved find pathways to power through relationships that give them access to assets in pursuit of livelihoods. Participatory planning cannot empower communities, but it can create space for communities to find their voice(s). Empowerment happens when impoverished or marginalized ind ividuals or groups can imagin e their world differently and make that vision a reality by changing the relations of power that h ave been keeping them from reaching that potential ( Eyben et al., 2008 ) But, how is it possible to create space that allows for change in power relations ? This dissertation took the first step in understanding what power looks like inside a particular community. I operationalized how power works through different modes in or der to make it visible. But, based on my research findings I question whether we should expect some marginalized households or groups to be able to advocate for themselves.

PAGE 206

187 Marginalized groups are different than the mainstream, and I suggest that we ne ed to develop new tools and ways for representing the marginalized voice. I echo previous research that calls for more information on how social networks can provide different routes of influence for those more versus less connected members of marginalize d groups ( Morgan Trimmer, 2013 ) Theory to Practice So far in this chapter, I have discussed findings based on addressing my three research aims and primary research question I have endeavored to gain insight into how power relations and social netw orks impact the participatory planning process, using the perspective of a marginalized group of poor urban farmers. Underlying this dissertation research was the challenge of conceptualizing, defining, and measuring power. Revisiting Wolf The motivatio n for this dissertation research topic was that, i n practice, inequitable power relations plague the community engagement process in sustainable planning and development. I argue it is rooted in a gap in understanding power relations within and across com munities A nd a gap in understanding how power affects participation and who has influence in the decision making process. S tudies that measure social capital, social relations, and social structure have been limited by the challenge of operationalizing the conce pt of power. This partially due to the tendency to look at power at only one

PAGE 207

188 level of influence and neglect the more nuanced ways that power can work. This is also partially due to the one dimensional focus of power, measured as who makes the d ecision who has power over This research expanded on the concept of power using Wolf's theory that it is a multiple and complex phenomenon pervasive, but working differently, across multiple levels of human interaction. Although Wolf took an agency str ucture approach in developing a theory of power, he was most concerned with power as it played out through structure. Wolf's structural inquiry undoubtedly advanced how we conceptualize power as manifest across society I adapted Wolf's theoretical ideas to a new context with a n emphasis on agency (refer to Chapter III: Theoretical Framework) I wanted to demonstrate empirically some of the ways in which power functions (and functions differently but not necessarily independently) through the more agenc y based modes of power that Wolf describes By drawing from Wolf's contribution to political economy I identified a compatible framework in the Sustainable Livelihoods Framework (SLF) to help translate power relations into everyday activities. The SLF p rovided an organizing framework to identify the various components within each of Wolf's modes of power addressing both structural and human agency approaches. However, while the SLF was designed to organize key components across multiple scales related to the process of attaining a sustainable livelihood, it doesn't explain underlying mechanisms that drive the process. It has been suggested that power is an important and often overlooked explanatory variable

PAGE 208

189 in the SLF ( De Haan & Zoomers, 2005 ) The SLF leaves space for power, but it is an organizational strategy and does not have explanatory power; it lacks theory Scoones ( 2009 ) called for theory to help explain SLF mechanisms, and Wolf provides a compatible theory for how the components of the SLF are related to each other. In turning to Wolf, I answered: what is power? And, in drawing from the SLF, I answered: where do we look for it? But, I was still left pondering: how will I know it when I see it? I had to find a way to measure it. By combining Wolf's theories of power with the SLF, I was able to op erationalize different modes of power. Furthermore, the SLF helped to conceptualize feedback loops by representing a framework for how power at the various levels might interact. By using power as a lens, I was able to operationalize the link between inf luence/access and livelihood assets based on Wolf's theory that power is a relational phenomenon. How households inf luence and access livelihood assets has been a gap in the SLF, but, while Wolf's ego alter mode of power prescribes power to influence and access livelihood assets through social relations, it is a theory not a method. Thus, I employed social network analysis (SNA), which, in addition to providing a method for measuring power relations, also benefitted from the theoretical framework. In de veloping this dissertation I careful ly considered theories, frameworks, methods, and prior studies across disciplines. Social relations may be a series of time place interactions, but how power is negotiated through influence and access to knowledge and resources is a complex web of co occurrences that aggregate in unexpected and uneven ways. Although the mix of qualitative

PAGE 209

190 quantitative and spatial methods employed to measure power relations in this dissertation allowed for a more in depth analysis tha n previous studies (for example by Holman ( 2008 ) McKether ( 2011 ) and Butts, et al. ( 2012 ) ), these multiple methods also complicated analysis a nd interpretation of findings. See limitations below for more about this persistent challenge. Linking Modes of Power Measuring power required me to seek out a multitude of time place interactions to see where, how, and what kinds of power households e xhibited. By piecing together micro activities, I learned about whom the households interacted with, which then influenced household knowledge base and understanding of the land development situation, and, ultimately influenced household beliefs and behav iors and how they envisioned their future. But this research was not a simple catalog of human interactions. And, further complicating interpretation of findings, was the physical environment. An important, but muted component of this research, in drawi ng from both political ecology and social network theory, related to geographic embeddedness. Physical context (environmental resources) and spatial proximity (distance to social relations) is an influential variable and was important but challenging, to consider in addressing research aims. I discuss this in more detail under limitations.

PAGE 210

191 Research Limitations This dissertation research aimed to uncover the ways in which power relations acted as a mechanism in driving the flow of knowledge and resour ces in supporting or undermining agency within the Yamuna Khadir farm community and in the planner public nexus. Drawing from multiple disciplines and combining theory, frameworks, and methods proved challenging for research design and interpretation of f indings. I carefully considered each step of the research process and acknowledge the following limitations. Research Design Conceptualizing, defining, and measuring relational power was a fundamental challenge in this dissertation research. I measured many instances of relational power; however, power, as a relational process situated in time and place, was very difficult to aggregate. While much effort was made to ensure content and construct validity, there are many internal and external factors in addition to power relations that shape and modify behavior. Thus, I am cautious in claiming that findings from this dissertation research exhibit external reliability. In regards to the research population: the dynamic nature of power empowerment disempow erment is more evident among marginalized groups, and therefore, I used the case study of a marginalized, poor urban farm community faced with development of their land. While the Yamuna Khadir was a unique community, and findings might not be generalizab le to other contexts, I do believe that the community is not unlike many marginalized groups threatened

PAGE 211

192 with displacement as developing cities look toward a future that is more economically prosperous, environmentally clean, and provides a better quality o f life for formal citizens. However, more research on power relations across different modes is required in other contexts and with other communities. A third notable limitation was the data collection timeframe. Livelihood strategies often, but don't al ways, become livelihood outcomes over time I, therefore, acknowledge that livelihood strategies represent potential for but not actual behavior. I investigated relational variables (strategies power within ; self confidence) rather than causal or directi onal variables (outcomes em power ment ; enabling assets), which would have require d measurement over time, and beyond the scope of this dissertation. Since the majority of households were migrants, and had already been forced to relocate, they were mindful of the possibility of moving again. Therefore, capturing livelihood strategies was a good proxy for action with this particular population. Methodological Challenges First, in operationalizing social relations, I found it challenging to develop a measure that encompassed all the possible social relations of Yamuna Khadir households. At the end of each interview, I asked households if there were any people they interacted with in addition to those discussed. Sometimes they mentioned seed or fertilizer co mpanies, volunteer doctors, or local politicians, but since those were emergent relations, I did not specifically ask every household about them. It may be proof that participants had few relations outside daily

PAGE 212

193 activities (as discussed above), or it may be that outsiders simply didn't come to mind. I had a limited amount of time that I could keep households engaged in conversation and a long list of topics to cover in the interview. Since interviews were designed to be semi structured, households had so me control over the direction of the conversation. This was beneficial when they talked about development or interactions with other farmers, but not so helpful when topics went far off track and participants expressed boredom when I tried to reign them b ack in. Second, language and cultural differences were a limitation. All interviews were conducted in Hindi, which means Hindi English translation may have inaccuracies. There was also potential for cultural misinterpretation when the interview questio ns were translated from English to Hindi and also when interpreting Hindi responses back to English. I pilot tested interview questions and did a preliminary analysis to minimize misinterpretation and any obvious biases; however, cultural translation is a notable limitation of this research. Furthermore, although the two translators were local natives and fluent Hindi English speakers, they were both from a much higher socio economic class than the participants. They expressed that there was some culture differences for them as well, which might have biased interpretation. Related to this was the issue of gender: both research assistants were female (as am I), which is likely the reason that we nearly equally interviewed men and women as household spokes persons.

PAGE 213

194 Length versus depth of interviews was a third notable challenge. During the fieldwork phase, half of all interviews were conducted with a Hindi English translator conducting the interviews and myself taking notes. This made interviews a very slow process. Although I carefully designed interviews to last no longer than 30 minutes, people got bored waiting for my translator to interpret back and forth. On the other hand, half of the interviews were conducted in Hindi with a Hindi English tran slator taking notes (I was not present), but the interviews were not as rich because I was not there to give prompts in reaction to responses. The breadth of data collection allowed me to analyze findings quantitatively and qualitatively, but I acknowledg e that interpretation and the conclusions I could draw were necessarily limited by the depth of data. I attempted to design this dissertation research to minimize or clearly identify potential limitations; however, social science research is by nature a m oving target. A great deal of time was dedicated to dealing with missing data (a result of the semi structured design of the interviews) in preparing for data analysis. In using households as the unit of analysis, it was sometimes hard to tell if the per son doing the interview was speaking for him/herself or the household as a whole And the sampling method interviewing households adjacent to each other impacted results, but I don't know if or how it biased them Additionally, this study was design to c ollect ego centric (household level) social network information. The limitation of ego centric networks is that I could not analyze which households are connected to which other households. I

PAGE 214

195 therefore had to consider spatial analysis based on physical p roximity or rather how likely neighboring households were to be similar regardless of friendship. Analysis: Data Limits A notable limitation of data analysis relates to the breadth versus depth of data. Most households talked in detail about social rel ations, provided some information or opinion about land development, but gave short responses to the question of livelihood strategies. I acknowledge that it may have been the result of participant fatigue, households being unsure of the response that the translator or I wanted, or honest lack of livelihood planning because it was draining to put limited resources into planning for an uncertain future. The act of planning ahead implies that there are options and opportunities and for poor and marginalized groups, options may be quite limited. For this particular research population, there is a legacy of one generation taking over the work of the previous generation: "my father farmed, I farm, and my children will farm." And it may in fact be the case tha t few thought beyond that trajectory; however, I am hesitant to assume this without further exploration. In developing the analysis plan, I chose a mix of methods. Quantitative analysis would have limited my ability to explain why or under which circumst ances behavioral relationships occurred. Conversely, a purely qualitative approach would have limited my data analysis to description without the ability to determine significant relationships among phenomena. Thus, I employed a mix of methods. This all owed me to more deeply explore

PAGE 215

196 the different pathways that social structure took in how households understood development of their land and thought about livelihood strategies. It allowed me to illustrate examples of how households negotiated power relati ons in the pursuit of sustainable livelihoods, and how that negotiation framed household understanding of the context of land use change. The multiple approaches to analyzing research data was time consuming and required many iterations; I acknowledge tha t analysis was not exhaustive. In regards to spatial analysis, one thing that surprised me was that while I found geographic clustering of households based on strong and weak social networks, but I did not find a stronger relationship between geographic proximity and household beliefs, behaviors or livelihood strategies. I expected households near each other to be more similar based on t he growing evidence that physical proximity is related to social interactions ( Butts et al., 2012 ; Doreian & Conti, 2012 ; Hipp et al., 2012 ; Verdery et al., 2012 ) particularly in this case where Yamuna Khadir farm plots were quite small (most ranging from 1 to 2 acres), and people worked and lived within close proximity. I selected an adjacent sampling method based on the assumption that people living and working next to each other would be more likely to talk and share ideas and opinions. Contrary to findings in the literature that suggest spatial relations influence how people understand and experience community and part icipation ( Sultana, 2009 ) I did not find any geographic clustering of land use beliefs or behaviors or livelihood strategies. I found patterns where I did no t expect to (for example, household social network

PAGE 216

197 power exhibited a strong spatial pattern even though this measure does not incorporate a between household measure), and did not find patterns where I expected to (for example, land use beliefs and behavio rs). More in depth spatial analysis may be required than was within the scope of this dissertation. I don't consider the spatial analysis presented in this dissertation necessarily to be in support or contrary to other findings in the literature, but ra ther that my findings are incomplete. More exploration of the data is necessary. For example, I could run hot spot analysis by combining or weighting variables differently. I could also add a spatial weighting to the statistical model using Stata. There may be some relationship between distance to the metro construction, distance to the main road, distance to and number of adjacent neighbors there are a variety of ways to explore spatial variations. Future Steps There are a number of potential direction s for next steps for this research. Developing a method and measuring power relations is a meaningful contribution to the literature because few studies actually measure power in social networks in the planning literature. Through this dissertation I ill ustrated everyday exchanges that cumulated into lived reality and future livelihood options, but I also found that exchanges might limit future options if they are only meeting daily needs. It was particularly challenging to aggregate overall social struc ture of households based on the multiple pathways and obstacles to power. Further

PAGE 217

198 consideration of how households weigh costs and benefits in developing relationships is warranted. Related to this, I also found that in the pursuit of relationships that h elped meet daily needs, marginalized households may not reach beyond to establish relationship with those who might benefit them in the longer term. It is also worth pursuing further inquiry into how and when relationships are established. In terms of pra ctical application, I plan to demonstrate a method for rapid assessment of power relations within a target community in order to inform the participatory planning process in managing participation, representation, and disengagement. And, I will continue t o explore social and spatial patterns in how households access and influence livelihood assets for the purpose of developing a method for identifying formal and informal routes of influence into and out of poor and marginalized communities. In regards to organizational power, I found that the majority of participants did not believe they had any influence on land development There is a need to address the processes that led the residents of Yamuna Khadir to perceive themselves as outside the decision ma king space. The first step might be to identify the scale at which households believe they have an impact how far d o they think their influence can reach? What other civic activities are they involved in? Do they, and how do they, conceptualize civic du ty? And, for migrants, how does civic engagement differ in Delhi from their nat al village? And I return to issue of creat ing space for Yamuna Khadir households in order to develop collective voice(s) and bring to light those who can speak for them. Mor e

PAGE 218

199 structural level data would also provide deeper insight on interpreting the boundaries and limitations of human agency in this case At the individual level, power is generated through the individual (in this case, the household). I suggest further exam ination of the interrelationship between skills and self confidence and the ability to establish mutually beneficial relationships. Examining power at multiple levels of influence helps to build an understanding of household decision making. It also illu minates household options and where the potential barriers to empowerment reside. One line of future inquiry might be to develop questions and testable hypotheses related to the co occurrence of different patterns of ego alter power with planning beliefs and behaviors and livelihood strategies. For example: Are households with strong community power but weak outer power (relationships outside the community) more likely to actively resist land development and have clear, detailed livelihood strategies? A final line of inquir y I suggest is for deeper exploration into the concepts of power as a spectrum from zero sum to generative. Building upon concepts of power, measures for agency and empowerment should also be introduced and tested This requires long itudinal studies to capture the process of households acting out their livelihood strategies by putting their assets into action. How does power move to empowerment? Having access to assets is not the same as actually mobilizing them when they are needed In the next and final chapter, I draw conclusions and make final observations.

PAGE 219

200 CHAPTER VII : CONCLUSION Sustainable Cities and Pathways to Action The United Nations Human Settlements Programme (UN H abitat ) recognizes c ommunity participation as an essent ial component of sustainable planning and development ( Motasim et al., 2010 ) Community part icipation has the potential to empower communities, result in better design outcomes that are more responsive to the diverse needs of different urban groups, ensure the relevance of plans when resources are limited, and increase effectiveness of outcomes. The persistent chasm between discourse and reality of participatory planning is rooted in power relations, which bias representation, knowledge, and problem framing. But there is a gap in understanding power relations within and across communities. Powe r relations impact access to knowledge and resources and, therefore, problem framing. While planners focus on issues of climate change and urbanization, communities are concerned with daily activities and livelihoods. A Framework for Power Relations Sus tainable planning and development often targets areas that are degraded and prone to natural hazards the same places marginalized groups often inhabit Grounded in geography and anthropology theories, frameworks, and models of vulnerability/resilience and political ecology, I applied Eric Wolf's theories of power to the Sustainable Livelihoods Framework (SLF), and used

PAGE 220

201 social network analysis (SNA) as my primary method of measuring power relations within the modified theoretical framework. This dissertatio n presented a case study of poor, marginalized urban farmers faced with land development In pursuing a degree in health and behavioral sciences, I am fundamentally concerned with human behavior embedded in multiple levels of social engagement ; a n individ ual may have the potential to act, but it is only by understanding which actions are possible, in which contexts, and with whom, that action is more likely to occur. Underlying the negotiation of social relations is the concept of power Thus, I explored the relationship between power relationships, planning participation, and livelihood strategies. I achieved this by looking at power as a relation al process exhibited as power over power to and inherent power to through four modes : (1) the individual m ode represented by household livelihood strategies, (2) the ego alter mode represented by social structure, (3) the organizational mode represented by land use beliefs and behaviors, and (4) the structural mode represented by historic and current city part icipatory planning activities. By combining Wolf's theories of power with the SLF and employing social network methods I was able to demonstrate how households access ed assets in pursuit of sustaining livelihoods and fill an important methodological and empirical research gap Beliefs, Behaviors, and Agency The concept of a gency is a central to the democratic nature of the participatory planning process. In terms of participation, agency occurs when

PAGE 221

202 one's beliefs and values align with one's own behavi ors (i.e. direct engagement), whereas in terms of representation, a gency occurs when one's beliefs and values align with the behaviors of the group (i.e. indirect engagement) This is important because it makes a case for the potential for representation to empower excluded and marginalized groups. While this dissertation was not intended to measure agency through representation, it made an important step forward in how to observe and measure power relations within and across communities. If power relat ions are observable and measurable, then the next step is to identify under what conditions agency occurs. Power is something that we gain access to through the people we interact with at home, at school, at work, in public. Power cannot be given to anot her person (or community), but we (as democratic states, planners, and community researchers) have a responsibility to create conditions that enable others to generate power ; to become empower ed In going back to the opening quote set forth in the introdu ction, this dissertation presents empirical evidence that power is center less; it is capillary. It is "a moving substratum of force relations, local and unstable" ( Sandercock, 2003, quoting Foucault ) Power is embedded in everyday activities and that is precisely where pl anners need to intervene. This dissertation research fills an important cross disciplinary and theoretical gap between on the ground ethnographic methods to capture the collective experience of small scale famers as they make daily livelihood decisions a nd top down interventions to mitigate climate change impacts.

PAGE 222

203 Specifically, this research produced new knowledge about the nuanced dimensions of power relations and decision making through a detailed analysis and description of the mechanisms through whic h spatially defined social networks impact Yamuna Khadir farmers' livelihood strategies when faced with planned land use change to the Yamuna River floodplain. This dissertation contribute s to our understanding of community empowerment by expanding our k nowledge on why some populations may be more challenging to engage in sustainable planning than other s I suggest that engaging poor and marginalized groups requires an alternative to the formal invited spaces typically created. Approaching those on the periphery by means of the informal spaces they create for themselves is necessarily very different to mainstream practice. I believe the research presented in this dissertation can act to guide future inter disciplinary research and provide insight to ci ties and planning organizations work ing to ward solutions to mitigate climate change impacts at the city level (macro level) and strengthen livelihood strategies of vulnerable or marginalized populations (micro level). This case study supports the value of taking a bottom up approach to identifying potential biases in community participation so that we can improve community engagement and achieve more sustainable outcomes. Final Observations It is not surprising that informal settlements, vast health dispa rities, and environmental degradation plague Delhi. A result of complex historic and

PAGE 223

204 political structures, a n anti poor bias persists despite a majority of the population being poor. However, t he interventions deployed [in Delhi] to render poverty invis ible do not diminish poverty, but create new place s and positions for poor people (Rao, 2010). We need to re link livelihoods with the larger political economic context if we are to create truly sustainable and inclusive cities.

PAGE 224

205 References Adger, W. N. (2006). Vulnerability. Global Environmental Change, 16 (3), 268 281. Adhikari, K. P., & Goldey, P. (2010). Social Capital and its "Downside": The Impact on Sustainability of Induced Community Based Organizations in Nepal. World D evelopment, 38 (2), 184 194. Ahmad, S., Balaban, O., Doll, C. N. H., & Dreyfus, M. (2013). Delhi revisited. Cities, 31 (0), 641 653. Anderson, B. (2012). Converting Asset Holdings into Livelihood: An Empirical Study on the Role of Household Agency in Sou th Africa. World Development, 40 (7), 1394 1406. Anselin, L., Sridharan, S., & Gholston, S. (2007). Using Exploratory Spatial Data Analysis to Leverage Social Indicator Databases: The Discovery of Interesting Patterns. Social Indicators Research, 82 (2), 2 87 309. Arun, S., Annim, S. K., & Arun, T. (2013). Overcoming Household Shocks: Do Asset Accumulation Strategies Matter? Review of Social Economy, 71 (3), 281 305. Bailey, N. (2010). Understanding Community Empowerment in Urban Regeneration and Planning in England: Putting Policy and Practice in Context. Planning Practice & Research, 25 (3), 317 332. Baker, M., Coaffee, J., & Sherriff, G. (2007). Achieving successful participation in the new UK spatial planning system. Planning Practice & Research, 22 (1 ), 79 93. Barrett, S. R. (2002). Culture meets Power : Greenwood Publishing Group. Barrett, S. R., Stokholm, S., & Burke, J. (2001). The idea of power and the power of ideas: A review essay. American Anthropologist, 103 (2), 468 480. Beard, V. A., & Ca rtmill, R. S. (2007). Gender, collective action and participatory development in Indonesia. [Article]. International Development Planning Review, 29 (2), 185 213. Beebeejaun, Y., & Vanderhoven, D. (2010). Informalizing Participation: Insights from Chicago and Johannesburg. Planning Practice & Research, 25 (3), 283 296. Birkmann, J. (2006). Measuring vulnerability to promote disaster resilient societies: Conceptual frameworks and definitions. In J. Birkmann (Ed.),

PAGE 225

206 Measuring vulnerability to natural hazards New York: United Nations University Press. Blaikie, P., Cannon, T., Davis, I., Wisner, B., Cannon, T., Davis, I., et al. (1994). At Risk: Natural hazards, people's vulnerability and disasters New York: Routledge. Bohle, H. G., Downing, T. E., & Watt s, M. J. (1994). Climate change and social vulnerability: Toward a sociology and geography of food insecurity. Global Environmental Change, 4 (1), 37 48. Brownill, S., & Carpenter, J. (2007). Participation and planning: Dichotomies, rationalities and stra tegies for power. Town Planning Review, 78 (4), 401 428. Brownill, S., & Parker, G. (2010). Why Bother with Good Works? The Relevance of Public Participation(s) in Planning in a Post collaborative Era. Planning Practice & Research, 25 (3), 275 282. Butts C. T., Acton, R. M., Hipp, J. R., & Nagle, N. N. (2012). Geographic variability and network structure. Social Networks, 34 82 100. Carr, E. R., & McCusker, B. (2009). The co production of land use and livelihoods change: Implications for development i nterventions. Geoforum, 40 (4), 568 579. Carrasco, J. A., Hogan, B., Wellman, B., & Miller, E. J. (2008). Collecting social network data to study social activity travel behavior: an egocentric approach. Environment and Planning B: Planning and Design, 35 961 980. Chainey, S., Tompson, L., & Uhlig, S. (2008). The Utility of Hotspot Mapping for Predicting Spatial Patterns of Crime. Security Journal, 21 (1 2), 4 28. Connelly, S. (2010). Participation in a Hostile State: How do Planners Act to Shape Public Engagement in Politically Difficult Environments? Planning Practice & Research, 25 (3), 333 351. Cook, J., Oviatt, K., Main, D. S., Kaur, H., & Brett, J. (2015). Re conceptualizing urban agriculture: An exploration of farming along the banks of the Yamun a River in Delhi, India. Journal of Agriculture and Human Values Cornwall, A. (2008). Unpacking 'Participation': models, meanings and practices. Community Development Journal, 43 (3), 269 283.

PAGE 226

207 Dale, A., & Newman, L. (2008). Social capital: a necessary and sufficient condition for sustainable community development? Community Development Journal, 45 (1), 5 21. Daniere, A., Takahashi, L., NaRanong, A., & Lan, V. T. N. (2005). Social capital and urban environments in Southeast Asia: lessons from settlement s in Bangkok and Ho Chi Minh City International Development Planning Review, 27 (1), 21 58. Daniere, A., Takahashi, L. M., & Narangong, A. (2002). Social capital, networks, and community environments in Bangkok, Thailand. Growth and Change, 33 (4), 453 48 4. Datta, A., & Jha, G. (1983). Delhi: Two decades of plan implementation. Habitat International, 7 (1/2), 37 45. Davidoff, P. (1965). Advocacy and pluralism in planning. Journal of the American Institute of Planners, 31 (4), 331 338. De Bon, H., Parro t, L., & Moustier, P. (2010). Sustainable urban agriculture in developing countries. A review. Agronomy for sustainable development, 30 (1), 21 32. de Haan, L. (2012). The livelihood approach: a critical exploration. Erdkunde, 66 (4), 345 357. De Haan, L ., & Zoomers, A. (2005). Exploring the Frontier of Livelihoods Research. Development and Change, 36 (1), 27 47. Delhi Development Authority. (2012). Master Plan New Delhi 2021. from http://dda.org.in Dempwolf, C. S., & Lyles, L. W. (2012). The Uses of S ocial Network Analysis in Planning: A Review of the Literature. Journal of Planning Literature, 27 (1), 3 21. Doreian, P., & Conti, N. (2012). Social context, spatial structure and social network structure. Social Networks, 34 32 46. Dupont, V. (2007). Conflicting stakes and governance in the peripheries of large Indian metropolises: An introduction. Cities, 24 (2), 89 94. Edwards, G. (2010). Mixed Method Approaches to Social Network Analysis Manchester, England: National Center for Research Methods.

PAGE 227

208 Eguren, I. R. (2008). Moving up and down the ladder: community based participation in public dialogue and deliberation in Bolivia and Guatemala. Community Development Journal, 43 (3), 312 328. Eyben, R., Kabeer, N., & Cornwall, A. (2008). Conceptualisin g empowerment and the implications for pro poor growth: A paper for the DAC Poverty Network : Institute of Development Studies. Giddens, A. (1984). The constitution of society: Outline of the theory of structuration : Univ of California Press. Government of India. (2013). Press Note on Poverty Estimates, 2011 12 : Government of India, Planning Commission. Groenewald, S., & Bulte, E. (2012). Trust and livelihood adaptation: evidence from rural Mexico. Agriculture and Human Values 1 15. Heritage, Z., & Dooris, M. (2009). Community participation and empowerment in healthy cities. Health Promotion International, 24 (S1), i45 i55. Hipp, J. R., Faris, R. W., & Boessen, A. (2012). Measuring 'neighborhood': Constructing network neighborhoods. Social Networks, 34 128 140. Holman, N. (2008). Community participation: using social network analysis to improve developmental benefits. Environment and Planning C: Government and Policy, 26 525 543. IARI. (2013). Annual Report 2012 2013 New Delhi, India: Indian A gricultural Research Institute. Illsley, B., & Coles, E. (2009). Putting Communities in Control: A Scottish Perspective. Planning Practice & Research, 24 (3), 363 378. Irvin, R. A., & Stansbury, J. (2004). Citizen participation in decision making: Is it worth the effort? Public Administration Review, 64 (1), 55 65. Isaac, J. C. (1989). "A Behavioral Theory of Social Structure": A Critique. Journal for the Theory of Social Behaviour, 19 (1), 131 139. Kabeer, N., Mahmud, S., & Isaza Castro, J. G. (2012). NGOs and the Political Empowerment of Poor People in Rural Bangladesh: Cultivating the Habits of Democracy? World Development, 40 (10), 2044 2062. Kesby, M. (2007). Spatialising participatory approaches: the contribution of geography to a mature debate. Environment and Planning A, 39 (12), 2813 2831.

PAGE 228

2 09 Khanal, R. C. (2007). Local Level Natural Resource Management Networks in Nepal: An Additional Burden or Agents of Change Ensuring Environmental Governance and Sustainable Livelihoods? Mountain Research and Development, 27 (1), 20 23. Knoke, D., & Yang, S. (2008). Social Network Analysis (2nd ed.). Los Angeles: SAGE Publications. Kundu, A. (2004). Provision of tenurial security for the urban poor in Delhi: Recent trends and future perspectives. Habitat Int ernational, 28 259 274. Kundu, A. (2011). Trends and processes of urbanisation in India : Human Settlements Group, International Institute for Environment and Development, and United Nations Population Fund. Laurian, L., & Shaw, M. M. (2009). Evaluation of Public Participation. Journal of Planning Education and Research, 28 (3), 293 309. Leach, M. (2008). Pathways to Sustainability in the forest? Misunderstood dynamics and the negotiation of knowledge, power, and policy. Environment and Planning A, 40 (8 ), 1783 1795. LeCompte, M., & Schensul, J. J. (1999). Ethnographer's Toolkit Volume 1: Designing and Conducting Ethnographic Research Walnut Creek, CA: Altimira Press. Loftus, A., & Lumsden, F. (2008). Reworking Hegemony in the Urban Waterscape. Trans actions of the Institute of British Geographers, 33 (1), 109 126. Lukes, S. (1974). Power: A radical view (Vol. 1). London: Macmillan. Lukes, S. (1978). Power and Authority. In T. Bottomore & R. Nisbet (Eds.), A History of Sociological Analysis (pp. 633 676). New York: Basic Books. Lyons, M., Smuts, C., & Stephens, A. (2001). Participation, Empowerment and Sustainability: (How) Do the Links Work? Urban Studies, 38 (8), 1233 1251. Maginn, P. J. (2007). Towards more effective community participation in urban regeneration: The potential of collabortive planning and applied ethnography. Qualitative Research, 7 (1), 25 43. Mahjabeen, Z., Shrestha, K. K., & Dee, J. A. (2008). Rethinking community participation in urban planning: The role of disadvantaged gr oups in

PAGE 229

210 Sydney metropolitan strategy Paper presented at the 32nd ANZRSAI Conference. Manta conroy, M. (2011). Influences on Public Participation in Watershed Planning: Why is it still a Struggle? Planning Practice & Research, 26 (4), 467 479. Martin B reen, P., & Anderies, J. M. (2011). Resilience: A Literature Review : The Rockefeller Foundation. McAlister, R. (2010). Putting the 'community' into community planning: Assessing community inclusion in Northern Ireland. International Journal of Urban and Regional Research, 34.3 533 547. McKether, W. L. (2011). Increasing power in a black community: A networked approach. Transforming Anthropology, 19 (1), 65 75. McLaughlin, P., & Dietz, T. (2008). Structure, agency and environment: Toward an integrated perspective on vulnerability. Global Environmental Change, 18 (1), 99 111. Mdee, A. (2008). Towards a dynamic structure agency framework: understanding patterns of participation in community driven developmen t in Uchira, Tanzania International Developmen t Planning Review, 30 (4), 399 420. Melcher, K. (2013). Equity, empowerment, or participation. Landscape Journal, 32 (2), 167 182. Mercer, J., Kelman, I., Lloyd, K., & Suchet Pearson, S. (2008). Reflections on use of participatory research for disaster r isk reduction. Area, 40 (2), 172 183. Meusburger, P. (2008). Chapter 2: The nexus of knowledge and space. In P. Meusburger, M. Welker & E. Wunder (Eds.), Clashes of Knowledge : Springer. Monno, V., & Khakee, A. (2012). Tokenism or Political Activism? Som e Reflections on Participatory Planning. International Planning Studies, 17 (1), 85 101. Morgan Trimmer, S. (2013). 'It's who you know': community empowerment through network brokers. Community Development Journal, 49 (3), 458 472.

PAGE 230

211 Motasim, H., Rae, K., Andersson, C., Auclair, C., Barth, B., Buhren, K., et al. (2010). Planning sustainable cities: UN Habitat practices and perspectives : UN Habitat. Neef, A., & Neubert, D. (2010). Stakeholder participation in agricultural research projects: a conceptual fr amework for reflection and decision making. Agriculture and Human Values, 28 (2), 179 194. Newman, K. (2008). Whose view matters? Using participatory processes to evaluate Reflect in Nigeria. Community Development Journal, 43 (3), 382 394. Norris, F. H., Stevens, S. P., Pfefferbaum, B., Wyche, K. F., & Pfefferbaum, R. L. (2008). Community Resilience as a Metaphor, Theory, Set of Capacities, and Strategy for Disaster Readiness. American Journal of Community Psychology, 41 127 150. Nygren, A., & Myatt Hi rvonen, O. (2009). Life here is just scraping by': livelihood strategies and social networks among peasant households in Honduras. Journal of Peasant Studies, 36 (4), 827 854. Panelli, R., & Welch, R. (2005). Why community? Reading difference and singula rity with community. Environment and Planning A, 37 (9), 1589 1611. Parizeau, K. (2015). When Assets are Vulnerabilities: An Assessment of Informal Recyclers' Livelihood Strategies in Buenos Aires, Argentina. World Development, 67 161 173. Pelling, M., & High, C. (2005). Understanding adaptation: What can social capital offer assessments of adaptive capacity? Global Environmental Change, 15 (4), 308 319. Peris, J., Farinas, S., Lopez, E., & Boni, A. (2012). Expanding collective agency in rural indigeno us communities in Guatemala: a case for El Almanario appr oach. International Development Planning Review, 34 (1), 83 102. Pickering, T., & Minnery, J. (2012). Scale and Public Participation: Issues in Metropolitan Regional Planning. Planning Practice & Re search, 27 (2), 249 262. Prashar, S., Shaw, R., & Takeuchi, Y. (2012). Assessing the resilience of Delhi to climate related disasters: a comprehensive approach. Natural Hazards, 64 (2), 1609 1624.

PAGE 231

212 Rao, U. (2010). Making the global city: Urban citizenship at the margins of Delhi. Ethos, 75 (4), 402 424. Robbins, P. (2012). Political Ecology: A critical introduction (2nd ed.). West Sussex, UK: Wiley Blackwell. Roberts, B., & Kanaley, T. (2010). Sustainable urban development in Asia. Australian Planner, 4 4 (1), 18 21. Rowlands, J. (1995). Empowerment examined. Development in Practice, 5 (2), 101 107. Saito Jensen, M., & Nathan, I. (2011). Exploring the potentials of community based natural resource management for benefiting local communities: Policies an d practice in four communities in Andhra Pradesh, India. Society & Natural Resources, 24 (11), 1142 1156. Sandercock, L. (2003). Cosmopolis Il: Mongrel Cities in the 21st Century New York: Continuum. Scandlyn, J., Simon, C. N., Thomas, D. S. K., & Bret t, J. (2013). Theoretical Framing of Worldviews, Values, and Structural Dimensions of Distasters. In B. D. Philips, D. S. K. Thomas, A. Fothergill & L. Blinn Pike (Eds.), Social Vulnerability to Disasters (2nd ed.). Boca Raton, FL: CRC Press. Scoones, I. (1998). Sustainable rural livelihoods: A framework for analysis. IDS. Scoones, I. (2009). Livelihoods perspectives and rural development. Journal of Peasant Studies, 36 (1), 171 196. Scott, J. (1991). Social network analysis: a handbook London: Sage P ublications. Sen, A. (1985). Well being, agency and freedom: the Dewey lectures 1984. The Journal of Philosophy 169 221. Shaefer, D. R. (2012). Youth co offending networks: An investigation of social and spatial effects. Social Networks, 34 141 149. Snodgrass, J. G., Lacy, M. G., Sharma, S. K., Jhala, Y. S., Advani, M., Bhargava, N. K., et al. (2008). Witch hunts, herbal healing, and discourses of indigenous ecodevelopment in North India: Theory and method in the anthropology. American Anthropologis t, 110 (3), 299 312. Sultana, F. (2009). Community and participation in water resources management: gendering and naturing development debates from

PAGE 232

213 Bangladesh. Transactions of the Institute of British Geographers, 34 346 363. Swapan, M. S. H. (2013). T he realities of participation in planning in Bangladesh: the role of institutional and socio political factors in shaping participatory planning in developing countries. Unpublished Dissertation, Curtain University. The World Commission on Environment an d Development. (1987). Our Common Future New York: Oxford University Press. UN Habitat. (2010). Planning sustainable cities: UN HABITAT practices and perspectives : United Nations Human Settlements Programme (UN HABITAT). Vasquez Leon, M. (2009). Hispa nic farmers and farmworkers: Social networks, institutional exclusion, and climate vulnerability in Southeastern Arizona. American Anthropologist, 111 (3), 289 301. Verdery, A. M., Entwisle, B., Faust, K., & Rindfuss, R. R. (2012). Social and spatial netw orks: Kinship distance and dwelling unit proximity in rural Thailand. Social Networks, 34 112 127. Whitson, R. (2007). Hidden struggles: spaces of power and resistance in informal work in urban Argentina. Environment and Planning A, 39 (12), 2916 2934. Wolf, E. R. (1990). Distinguished Lecture: Facing Power Old Insights, New Questions. American Anthropologist, 92 (3), 586 596. Wolf, E. R. (2001). Pathways of Power: Building an Anthropology of the Modern World : University of California Press. Wong, S. (2008). Building Social Capital in Hong Kong by Institutionalising Participation: Potential and Limitations. Urban Studies, 45 (7), 1413 1437. Yates, J. S. (2012). Uneven interventions and the scalar politics of governing livelihood adaptation in rural Nepal. Global Environmental Change, 22 (2), 537 546.

PAGE 233

214 APPENDIX A : RESEARCH QUE STION GUIDE Guide for Coding Questions Questions Background Basel ine [Dominant capital: Natural, Physical] Tenure HH # Rent v. Own Other home location Migrant origin Other land location How long has your family been here? How many people are in your family? Do you live here? If not, where is your home? Do you have land any place else? Part I: Social Network Aim 1 : To describe how Yamuna farmers exchange knowledge and resources through social networks. SRQ #1 : How is the social network of each Yamuna River farm family characterized? SRQ #2 : How is the social network of the collective Yamuna River farm community characterized? Other farmers [Dominant capital: Social] Invest: Things the family gives to other farmers (information, tangible resources, etc.) Withdraw: Things the family gets from other farmers (information, tangible resources, etc.) Invest/Withdraw: What other things do you talk about? Bonds: Do farmers know each other? Do farmers help each other ? How are new families integrated? Bridges: Connections to outside resources Ice breaker: Do you celebrate [Hindi word] with other farmers? Do farmers generally know each other? What kinds of things do you talk about? Prompts: Do you ask other farmers or tell other farmers what to grow, where to sell, how to get schemes or IDs? Do farmers generally help each other? How? Do you share laborers, tractors, borrow/lend $ When a new family comes here, how do you get to know them? Can you tell us about your experience when you first came here? Laborers [Dominant capital: Human] Bridge/bond: Whom do they hire and where do they come from? Invest: Things the family teaches/gives to laborers Withdraw: Things the family gets from laborers Bridges: Connections to outside resources Power negotiation: pay Do you hire any laborers ? Prompts: How do you find out who to hire and where do they come from? Do you teach them new skills or give them anything to make your farm better? Do they tell you new things about how to make your farm better? Do they connect you with other people that can help improve your livelihood? How do you know what to pay them? Do

PAGE 234

215 you ever negotiate? Do you talk about anything else? Market/Vendors [Dominant capital: Financial] Power negotiation: price and loans Withdraw: Things the family gets from vendors Invest: Things the family teaches/gives to vendors Bridges: Connections to outside resources How do you sell your produce? Prompts: Does a ven dor come or do you take it to a market? Do they give you a good price? How do you know what a good price is? Do they ever refuse to buy your produce? What do you do if that happens? Does (s)he tell you anything about how to make your farm better? Does ( s)he connect you with other people that help improve your livelihood? Do you take out any loans? How often? When do you have to pay it back? What if you cannot pay it back or are late in payment? Do you talk about anything else? Landlord/ownership [Dom inant capital: F inancial] Power negotiation: Ability to contact landlord and rent Withdraw: Things the family gets from the landlord Invest: Things the family teaches/gives the landlord Bridges: Connections to outside resources Do you pay rent to someone? (If not, talk about ownership) Prompts: Do they come here to collect? Do you know where they live? What do you talk about? Do you ask for any improvements on your property? Can you contact them if you have any issues or need something? Is the rent the same every month? What happens if you are late or cannot pay? Do they tell you new things about how to make your farm better? Do they connect you with other people that can help improve your livelihood? Other relations [Dominant capital: Varies] Withdraw: Things the family gets from others Invest: Things the family teaches/gives others Bonds: Connections to other farmers/this land/etc. Bridges: Connections to outside resources Are there any other people who come here? (Prompt: family, friends, government, NGOs, KVKs, etc.) Prompts: Where do they come from? What do you talk about? Do they tell you new things about how to make your farm better? Do they connect you with other people

PAGE 235

216 that can help i mprove your livelihood? Do you participate in any schemes? Is there anyone else that you talk to? Part II: Land Use Change Aim 2: To describe the beliefs and behaviors of Yamuna farmers related to land use planning. SRQ #3 : How do farmers perceive their influence on the land use change planning process? SRQ #4 : What is the current behavior of the farmers in response to the land use change planning process? Knowledge of land use changes Behavior: Involvement in discussions about the development, w ith other farmers, authorities, etc. Perception: Desire to be involved and/or ability to influence land use changes Is there any development planned for this land? Prompts: Can you tell us about this who, when, what? Have you been involved in any discussions about the development? With other farmers? Landlord? Authorities? What is discussed? If not, do you want to be involved in discussions? Why aren't you? Or why do you not want to be involved? Part III: Livelihood Strategies Aim 3: To identify l ivelihood adaptation strategies of Yamuna farmers in response to planned land use change. SRQ #5 : What do the farmers plan to do in response to planned land use changes that impact their livelihoods? Short term livelihood strategies Do you think that your family will farm this land next year? If yes or not sure, are you doing anything to improve your land or your family's prospects? If no, what will you do? Will you still try to improve your farm? How? Closing Questions Insight on social network Are you willing to talk to us again if we have more questions? Is this a good time of day? Who should we talk to next and why?

PAGE 236

217 APPENDIX B : OPERATIONALIZING EGO ALTER POWER E go alter power wa s conceptualized as influence of and access to livelihood assets through social networks. It was measured based on three qualities: type of transaction (five capitals), direction of transaction (investment, withdrawal, exchange or block/negative), and degree of influence (bond/bridge). Tables in this appendix are organized by the following social relation groups : within the household, other farmers, laborers, vendors, landlords, industry, and other relations.

PAGE 237

218 Table B.1. Power to influence/access capital within the household Description Responses Invest Withdraw Exchange Negative Bond Bridge Capital H ousehold categories Small multi generational or mixed family including some adult kids who are married and a few grandchildren X X S/H Large extended family X X X S/H School age children in school No school age children in school X Some or all school aged children in school X X H Length of tenure Tenure less than one year/monsoon cycle X Rent v own Don't pay any rent. X Other forms of HH income Other income: Run a shop X X H/F Other income: Nursery or Gardener X H/F Other income: Work as Laborers X X H/F Other income: Driver or Rickshaw X X H/F Other income: vendor X X H/F Other income: maid X X H/F *S=Social; H=Human; F=Financial

PAGE 238

219 Table B.2. Power to influence/access capital through relationships with other farmers Description Responses Invest Withdraw Exchange Negative Bond Bridge Capital Topics of discussion among farmers In general farmers talk (personal, work, or in general) X X H/S Farmers don't talk, sometimes fight, or only talk to family or immediate neighbors X Farmers generally trust each other Families borrow/ lend money X X F/S People don't trust each other X This family knows a lot of people X X S People are cooperative X X S This family's relatives live here/nearby X X S Loans Farmers don't borrow or loan money X Farmers do borrow or loan money X X F/S Farmers borrow/loan money among each other Famers don't borrow or loan money X Borrow/loan with security or interest X X F Borrow/loan on good faith X X F/S Borrow/loan in exchange for labor X X F/S This individual/family gives loans to others X X F/S This individual/family borrows money from others X X F/S People don't trust each other with money X They borrow only from the landlord X X F No one has any money X No one loans money X Farmers generally help each other Farmers help a lot X X S Farmers help when they can X X S Farmers don't help X

PAGE 239

220 Table B.2. (con t) Power to influence/access capital through relationships with other farmers Description Con't Responses Invest Withdraw Exchange Negative Bond Bridge Capital Type of help or exchange farmers are involved with Farmers share produce X X P Farmers help if there are health problems X X H They share laborers X X H They work for each other X X H Farmers arrange marriages among their families X X S Farmers share news X X H Farmers eat together or celebrate festivals X X S Description of how the HH found the land they farm This family came through relatives who live nearby X X P This family came through relatives who have moved X X P This family found land through contacts X X P Mi g r ant village This family is from Badayun X X S Who is involved in discussion They discuss land use change with other farmers X X H S=Social; H=Human; F=Financial; P=Physical

PAGE 240

221 Table B.3. Power to influence/access capital through relationships with laborers Description Responses Invest Withdraw Exchange Bond Bridge Capital How laborers are compensated Pay cash for labor X H Owner/landlord pays laborers X X H They exchange labor for labor with other farmers X X H Landlord and batai residents share cost to hire laborers X H Pay in cash and kind X H Who is hired as laborers Hire other farmers only for labor X X H Hire other farmers if available or outsiders if not X X H Hire outsiders for labor X X H There are permanent laborers (i.e. batai or nursery) X X H Works as a laborer for other farmers Work as labor for other farmers X X F Do they teach/tell laborers anything Hire farmers who already know how to farm X H Hire laborers who already know how to farm X H They teach/train laborers that they hire X H Do laborers teach/tell them anything They get advice from laborers X X X H *H=Human; F=Financial

PAGE 241

222 Table B.4. Power to influence/access capital through relationships with vendors Description Responses Invest Withdraw Exchange Negative Bond Bridge Capital Where do they interact with vendors or sell their produce Sell only from nursery/farm X X F Go to market and sell themselves X X F Buy from others and sell in market X X X F Where do they interact grouped by category Only go to the market if there is extra X X F They roam around to find a vendor X X F Sell to a vendor if there is a lot X F Go to the market and sell to a vendor X X F Deliver directly to customers X X F They sell to fixed ve n dors They sell to fixed vendors X X F/S When is produce paid for They are paid up front X F/S They are paid after sale X F Other sources of income Other income: Sell milk X X F Other income: Buy/sell from vegetable market X X F Other income: Has a shop X X F Other income: Has a tractor to rent X X F Other income: Maid X X F Other income: Nursery/Gardener X F Other income: Rickshaw driver X X F *S=Social; F=Financial

PAGE 242

223 Table B.5. Power to influence/access capital through relationship with landlord Description Responses Invest Withdraw Exchange Negative Bond Bridge Capital* Location that rent is paid Tenant goes to the landlord's house X X H/S The landlord comes to the land to collect X Sometimes go and sometime come for rent X X H/S Someone else takes the rent X There is no landlord X Knowledge of where the landlord lives Know the landlord's place of residence X H/S Don't know the landlord's place of residence X Characterization of landlord tenant relationship Parent type relationship with landlord X X X S Friendship type relationship with landlord X X X S Landlord is disrespectful X Improvements that landlord makes to the land Landlord doesn't make any improvements to the land X LL helps with inputs (e.g. gives advice re: where to buy) X X H LL provided the well/pump X X P/F LL splits the costs of inputs X X F LL provides inputs X X P/F LL helps with the tractor X X P/F LL has cows X X P/F The landlord provides loans to the tenants Loans provided by another employer X X F *H=Human; S=Social; F=Financial; P=Physical

PAGE 243

224 Table B.5. (con t) Power to influence/access capital through relationship with landlord Description Responses Invest Withdraw Exchange Negative Bond Bridge Capital* The landlord helps with issues or problems LL loans money and then adjusts the profit X X F LL loans money X X F LL doesn't loan money X LL helped in a time of need X X S LL did not help in a time of need X LL provides loans with interest X X F They left the land because the LL cheated them X Pay rent for the land or hut X X P Type of rent situation Batai (no rent) or rent in kind X X P/F Rent and batai X X P/F No rent X F Characterization of the type of advice the landlord provides Absent landlord X Landlord has negative control X LL advises X X H LL has positive engagement X X H Whether the landlord teaches the tenants anything LL teaches/trains farmers X X H LL doesn't teach/train farmers LL offers other small jobs X X F Whether the tenant gets advice from the landlord Get advice from LL X X H *H=Human; S=Social; F=Financial; P=Physical

PAGE 244

225 Table B.5. (con t) Power to influence/access capital through relationship with landlord Description Responses Invest Withdraw Exchange Negative Bond Bridge Capital* Frequency with which the landlord visits the tenant LL comes everyday or frequently X X X S/H LL never comes X Compensati on LL compensated [or not sure, but think LL has been compensated] X Doesn't know if the LL has been compensated X Ownership of land Believes DDA owns the land X Believes DDA & UP Irrigation owns the land X Believes DDA & Gov't owns the land X Believes UP Irrigation owns the land X Believes Government owns the land X Believes LL doesn't own owns the land X Don't know who owns the land X Who is involved in discussion An intermediary [landlord, MLA, advocate] is involved in land use discussion. X X H/S Court case activity Landlord filed court case X X H/S Court case filed on their behalf X X H/S *H=Human; S=Social

PAGE 245

226 Table B.5. (con t) Power to influence/access capital through relationship with landlord Description Responses Invest Withdraw Exchange Negative Bond Bridge Capital* Reason not involved Reasons not involved in land use disccusion: Don't own/landlord doesn't own the land X Landlord exploits them X Lack of information X Landlord is taking care X X H/S Lack of time or has to work X Will have to leave if the gov't/DDA asks even though they own the land X No one will listen/who want's to mess with the gov't X Disrespectful to ask LL X LL compensated so can't help X LL didn't help when huts razed; LL's hut has been razed X Confusion re: land ownership X LL says nothing can be done X Gender [it is not culturally acceptable for women to speak to men] X *H=Human; S=Social

PAGE 246

227 Table B.6. Power to influence/access capital through relationships with industry people Description Responses Withdraw Bond Bridge Capital Farm inputs from L** Buy farm inputs from L** X X P Buy from L** and get advice X X X P/H Farm inputs from Barfkhana market Buy farm inputs from Barfkhana market X X P Buy from Barfkhana market and get advice X X P/H Farm inputs from Laxmi Nagar market Buy farm inputs from Laxmi Nagar market X X P Buy from Laxmi Nagar market and get advice X X P/H Location where they buy farm inputs Other location X X P Multiple sources X X X P Other relations provide agriculture resources Other relations provide agriculture resources X X H *H=Human; P=Physical **Name is omitted for privacy. L is someone inside the community.

PAGE 247

228 Table B.7. Power to influence/access capital through other relationships Description Responses Withdraw Negative Bridge Capital Are there other people who come to this area? No one else comes X Others come / Interact with others X X H/S Have ID card(s) Don't have ID cards X Have ID cards X X P/F Have applied for ID cards Have applied for ID card(s) but don't have X Barrier to getting ID card(s) X Other relations provide medical resources Other relations provide medical resources X X H Other relations provide disaster assistance Other relations provide disaster assistance X X P *H=Human; S=Social; F=Financial; P=Physical

PAGE 248

229 APPENDIX C : CHARACTERIZING SOCIAL NETWORKS This research operationalized ego alter power as influence of and access to livelihood assets through social networks. Social networks were characterized in terms of strong to weak household power to influence and access livelihood assets through social relations. In this dissertation, social structure is the term used to refer to social networks characterized by power in this way. A social network is not the net sum, but rather the integration of various social relations into a holistic network of access and barriers. Because of the complexity of measuring social structure, social networks were first assessed by relatio n type: the household, other farmers, laborers, vendors, landlords, industry, and other relations. Then, social networks were combined into three dimensions: household, community, and outer network Bridging and bonding was then considered. Finally, soc ial networks were aggregated as social structure. Power by Social Network Type Household The Household social network was based on background characteristics Not all background information was a direct measure of the social network, and only the r elevant measures were included in this dissertation. Using the theoretical framework developed for this dissertation, households were analyzed for access to livelihood assets in the form of social, human and financial capitals.

PAGE 249

230 In terms of c haracterizing the household social n etwork households could invest in human capital through education, withdraw human and financial capital through other forms of employment, and exchange social and human capital through large family size. Barriers to ho usehold power were: new tenure, no school aged children in school, and not paying any rent (squatting) (table C .1 ). Table C.1 Score guide for characterizing household social network Scale (strong=1; weak=6) Definition 1 Educating children, other employment, and large family; no barriers 2 2 of: educating children, other employment, large family; no barriers OR 3 of : e ducating children, other employment, and large family; one barrier 3 1 of : e ducating children, other employment, and large family; no barriers OR 2 of : e ducating children, other employment, and large family; one barrier 4 1 of: e ducating children, other employment, and large family; one barrier 5 1 of: e ducating children, other employment, and large fam ily; two barriers 6 1+ barriers Other Farmers Other farmers represented the primary bonding group for interviewed households. There were many exceptions to this (i.e. lando wners, the doctor who volunteered a few hours every day to sit in Yamuna Khadir shop owners, etc), but generally speaking, this social network category represented the inner social circle. In this study, other farmers provided access to livelihood assets in the form of social, human, financial, and physical capitals. In terms of c haracterizing the other f armers social n etwork households could: (1) invest in financial and social capital through giving other farmers loans; (2) withdraw financial, social, and physical capital through borrowing money from

PAGE 250

231 other farmers, and finding l and through other farmers; and, (3) exchange human, financial, social, and physical capital through talking to, helping, and building relationships/friendships with other farmers, exchanging loans or labor, having relatives living nearby, and having the sa me homeland. Barriers to household power were: not talking, fighting, not trusting, not loaning money, and generally no t helping other farmers (table C .2 ). Table C.2 Score guide for characterizing other farmers social network Scale (strong=1; weak=8) Description 1 Farmers talk, trust or help; loan; exchange; found land through contacts; from Badayun; discuss land use change 2 Farmers talk, trust or help (can include exchange); loan (or no comment); found land through contacts ; from Badayun; may discuss land use change 3 Farmers talk, trust o r help (can include exchange); may or may not loa n; either found land through contacts or are from Badayun OR Landowner; may discuss land use change 4 Farmers talk, trust o r help (can include exchange ); may or may not loan 5 Farmers loan AND/OR Found land through co ntacts and/or are from Badayun; No other c omments OR O nly discuss land use change 6 Far mers don't talk, trust or help; do loan; found land through contacts and/or are from Badayun 7 Farmers don't talk, trust or help (or no comment); don't loan (or no comment); found land through contacts and/or are from Badayun 8 Far mers don't talk, trust or help; don't loan; did n't find land through contacts; not from Badayun OR Far mers don't talk trust or help; no other comments

PAGE 251

232 Laborers Laborers represented one group that had the potential to provide bridging and/or bonding connections. Since laborers were lower in the hierarchy than farmers (who employ ed them), their bridging potential was limited; however, bonding potential was high among farmers who worked for each other. Exchange of labor involved trust (withdrawal of social capital) that the hired person would improve crops and investment in improving the crops of other farmers. Exchan ge of labor among farmers also saved money because they didn't have to pay cash for labor. If farmers schedule work on a rotating cycle, they could always be working either to till, plant or harvest. Farmers could also plan to plant different crops to r educe competition. In this study, laborers provided access to livelihood assets in the form of human and financial capitals. In terms of characterizing the labor social network households could (1) invest in human capital by paying cash for labor, hiri ng outsiders, and teaching/training laborers; (2) withdraw human capital when the landlord hires/pays for laborers, hiring other farmers for labor rather than outsiders, and hiring laborers who already know how to farm; and (3) exchange human and financial capital by exc hanging labor for labor (table C .3 ). The primary capital provided by labor was human capital, but hired labor could cost financial capital. As such, the ideal labor social network provided high human capital at low cost to financial capita l (i.e. farmers working for farmers and exchanging information).

PAGE 252

233 Table C.3 Score guide for characterizing labor social network Scale (strong=1; weak=8) Description 1 Exchange labor with other farmers, teach and learn 2 Exchange labor with other farmers, may teach but not learn 3 Landlord pays for or shares cost of labor OR Landlord pays for labor and give advice to labor 4 Hire other farmers and pay cash but also works as labor OR Hire outsiders or pay other farmers and give and get advice 5 Hire other farmers and pay cash OR Hire outsiders and teach them, may also work as labor 6 Hire outside labor and work as labor 7 Hire outside labor only, pay cash for labor, or work as labor 8 Don't hire labor because of the expense Vendors Vendors represented a social group that had the most immediate bridging opportunities and were the primary source of financial capital. These were people that were, generally speaking, on the same social and economic level as farmers. They provided a critical l ink in the food system, acting as gatekeepers between the producer and the consumer. This social network was conc eptualized as the income source; social relations that provided other, non farm sources of income were also included. In this study, vendors provided access to livelihood assets in the form of financial and social capitals. In terms of c haracterizing the v endor social n etwork households could (1) invest in financial capital if they give the produce to a vendor and expect payment after sale, (2) withdraw financial capital when they sell produce and social capital if they are paid up front, and (3) exchange financial and social capital through

PAGE 253

234 selling to fixed vendors or buying other farmers' produce and selling themselves (table C .4 ). The pri mary capital provided by vendors was financial capital. As such, the ideal vendor social network maximized financial capital. Farmers made a better profit when they sold produce themselves, but they didn't always have the human capital to do so. Therefo re, if a family sold at the market, it meant that they had enough human capital and we re maximizing their profit potential. Selling directly at the market also connected them to market prices and trends in what was for sale. If farmers didn't have a stro ng social network, selling directly at the market was ideal, whereas if farmers were well connected (and got advice on prices and what to grow), it would be better to have a good vendor and put resources into growing crops with the highest profit potential Table C.4 Score guide for characterizing vendor social network Scale (strong=1; weak=9) Description 1 Go to the market and sell self and buy sell from other farmers or at market and other sources of income (may also have some fixed vendors) 2 Go to the market and sell self and other sources of income 3 Go to the market and sell self and buy sell from other farmers or at market 4 Go to the market and sell self 5 Buy from others and sell in the market 6 Sell from the nursery/farm to fixed vendo rs that pay up front and other sources of income 7 Sell from the nursery/farm other sources of income 8 Sell from the nursery/farm to fixed vendors that pay up front 9 Sell from the nursery/farm Industry I ndustry wa s related to both tangible inputs (seeds, fertilizer) and knowledge (how much fertilizer, which pesticide to use). This social network encompassed people working both for profit and not for profit. In this study,

PAGE 254

235 industry provided access to livelihood as sets in the form of human and physical capitals. In terms of characterizing the industry social network households could withdraw physical and human capital when they purchased farm inputs and got advice. The primary capital provided by industry was ph ysical capital, but the strength of this social network was in withdrawing human capital in the form of skills and knowledge related to agricultural practices. As such, the ideal industry social network prov ided high human capital (table C .5 ). Table C. 5 Score guide for characterizing industry social network Scale (strong=1; weak=8) Description 1 Buy farm inputs and get advice from L* have multiple sources for inputs, and get advice from other relations 2 Buy farm inputs and get advice from L* may have multiple sources for inputs, and may get advice from other relations 3 Buy farm inputs and get advice from other sources, have multiple sources for inputs, and/or get advice from other relations 4 Buy farm inputs from L* (may also get from other sou rces) and get other advice 5 Buy farm inputs from L* and from other sources 6 Buy farm inputs from L* only 7 Get farm advice from other sources and no comments on where inputs are bought 8 Buy farm inputs from other sources *Name omitted for privacy. L is someone inside the community. Landlord Landlords represented a social group that had the greatest bridging potential for tenant farmers. Although there was great variability in the social economic status of landowners, owning land gave an individual greater leverage with the government and developers. Being a landowner meant that one was literally invested in the land and the future of the land. However, being a

PAGE 255

236 landowner could also mean that the value of the land was in the selling price. Thus, landlords also presented one of the greatest vulnerabil ities for tenants and for surrounding residents. In this study, landlords provided access to livelihood assets in the form of human, financial, physical, and social capitals. In terms of characterizing the landlord social network households could (1) i nvest in human and social capital through interactions with the landlord; (2) withdraw human, physical, social, and financial capital when the landlord made improvements, helped with loans or other ways, gave advice or taught, didn't charge rent, offered s mall jobs, or was involved in land use discussion or a court case; and, (3) exchange human, physical, social, and financial capital through a positive relationship, loans, paying/compensating for the land, and interacting with the landlord freque ntly (tabl e C .6 ). There were numerous ways a landlord could create blocks to household power including not interacting, being disrespectful or abusive, being compensated for the land, not owning the land, exploiting tenants, and tenants not feeling like they could ask the landlord about land use plans. The primary capital provided by landlords wa s physical capital (land), but landlords provided a bridge to financial, human, and social capital as well. As such, the ideal landlord social network provided more than just physical capital. The most important distinguishing attribute was whether the tenant reported that their landlord tenant relationship was good or bad (disrespectful). Additional aspects of the landlord social network: batai (share cropping) often tra nsferred decision making control to the landlord but buffered farmers

PAGE 256

237 from risk, on the other hand, rent gave full control over decision making to farmers but exposed them to full risk in the cases of crop failure. Batai was linked to lower productivity b ecause profit was shared (farmers were not as motivated to maximize profit); on the other hand, rent was linked to higher productivity to maximize profit over rent costs. Table C.6 Score guide for characterizing landlord social network Scale ( strong=1; weak=10) Description 1 Know where landlord lives, relationship is good, landlord makes improvements, loans money, helped in time of need, gives good advice, teaches, gives advice, comes frequently, is involved in land use change discussion, may have filed a court case 2 Relationship is good; landlord makes improvements, gives loans, or helps with issues; gives advice; comes frequently; may be involved in land use change discussion or may have filed a court case 3 Relationship is good, landlord may make improvements, gives loans, or help with issues, may not come frequently, may or may not be involved with land use change discussion or filed a court case OR Same as 2 but with negatives such as landlord doesn't own the land, etc. OR No comments bu t landlord comes frequently and is involved in land use change discussion and/or filed a court case 4 Relationship is good OR Relationship is good and comes frequently but doesn't make improvements, help, etc, and other negatives (compensation) 5 Absent landlord but may or may not be involved in land use change discussion or filed a court case OR Absent landlord has made some improvements or provided loans 6 Absent landlord OR P ay rent and no other comments or negative comments OR Landlord may come frequently and no other comments or negative comments

PAGE 257

238 Table C.6 ( con't ) Score guide for characterizing landlord social network Scale (strong=1; weak=10) Description 7 Relationship is one of disrespect and any other negatives but may or may not be involved in land use change discussion or filed a court case OR One positive comment but landlord doesn't own the land 8 Doesn't know where the landlord lives or doesn't interact with the landlord, relationship is one of disrespect, no improvements, no loans, doesn't help, absent landlord or negative control, doesn't teach or give advice, never comes, has been compensated, d oesn't own the land, and negative reason not involved in land use change discussion OR No comments or negative comments and landlord has been involved in land use change discussion or filed a court case 9 Doesn't have a landlord OR Relationship is one of disrespect and no other comments OR Pays rent but the landlord doesn't own the land and no other comments or negative comments 10 Same as 8 but owns the land and comes frequently (this is an actively negative relationship) Other Relations Other relations included anyone not already covered in the interview Other people included peripherals to livelihood sustainability including healthcare, education, and various NGOs and government officials. In this study, other relations provided access to li velihood assets through various forms of human, social, financial, and physical capitals. In terms of characterizing the other relations social network households could withdraw human, social, physical or financial capital through other relations. Huma n capital was withdrawn by interactions with others (information) and through health care. Social capital was withdrawn by interactions with others (relationship building). Physical capital could be withdrawn when households

PAGE 258

239 had ID cards and was withdraw n through disaster relief. Financial capital could be withdrawn when households had ID cards. Barriers to household power occured when "no one comes," households didn't have ID cards, or had applied for ID cards and hadn't received them, or there was a b arrier to applying for ID cards (table C .7 ). Other relations provided the potential for households to access a variety of capitals. The ideal other relations socia l network provided access to multiple capitals. Table C.7 Score guide for characterizing social network for other relations Scale (strong=1; weak=7) Description 1 Others come, have ID cards, medical resources, disaster assistance 2 Others come, have ID cards, medical resources or disaster assistance 3 Others come, have ID cards OR No one comes/no comment, have ID cards, and have medical resources or disaster assistance 4 Have ID cards, no one comes or no other comments 5 Others come (incl. medical or disaster), no mention of ID cards 6 Others come ( incl. medical or disaster), no IDs or barrier to IDs 7 No one comes, no ID cards, no other assistance Characterization of Social Networks Household social network The household social n etwork used the scale in table C .1 The strongest household social network was 1 on the scale and the weakest was 6. I dichotomized the household social network as strong (1 3) or weak (4 6).

PAGE 259

240 Community social network The community social network related to those interactions within th e farming community primarily with other farmers. It also included laborers if they are other farmers, and industry if farm inputs and/or advice came from Lalit (in community seed shop). It combined other farmers, labo r, and industry scales (tables C.2, C.3, C .5). The community social network was a composite measure comprised of: Farmers: strong to weak (1 8) o Strong = 1 5 o Weak = 6 8 Labor: strong to weak (1 8) o Strong = 1 2; 4 5 o Weak = 7 8 Vendor: outside to inside (1 9) o Inside = 6 9 Industry: strong to weak (1 9) o Strong = 1 2; 4 6 Outer social network The outer social network related to those interactions or connections beyond the farming community primarily through the landlord and vendors, but also including labor, industr y, and other relations (tables C.3 C .7). The outer social network was a composite measure was comprised of: Landlord: strong to weak (1 10) o Strong = 1 4 o Weak = 5 10 Labor: strong to weak (1 8) o Strong = 3,5 Vendor: outside to inside (1 9) o Outside = 1 4 Industry: strong to weak (1 9) o Strong = 1 4; 7

PAGE 260

241 Other relations: strong to weak (1 7) o Strong = 1 4 o Weak = 5 7 Bonding social network A bonding relation was defined as a strong tie between the household and neighbors and friends. It included other farmers, laborers if they were o ther farmers, and industry if farm inputs and/or advice came from L (in community seed shop owner ). Bonding combined other farmers, labor, and industry scales (tables C.2, C.3, C .5 ). Bonding wa s comprised of: Farmers: s trong = 1 5 Labor: strong = 1 2; 4 5 Industry: strong = 1 2; 4 6 Bonding was summed as the number of strong bonding social network types (0 3 bonds). Bridging social network A bridge relation was defined as a connection with a person or people of different socioeconomic and/or cultural backgrounds. It included the landlord, vendors, labor, industry, and other relations (tables C.3 C .7 ). Bridging was comprised of: Landlord: strong = 1 4 Labor: strong = 3,5 Vendor: outside = 1 4 Industry: strong = 1 4; 7 Other relations: strong = 1 4

PAGE 261

242 Bridging was summed as the number of strong bridging social network types (0 5 bridges ). Social Structure: Strong/Weak Strong social structure was defined as meeting the following criteria: Strong community social network and one of the following: o Stron g household social network o Strong outer social network 2+ Bonding social networks 2+ Bridging social networks OR: Strong Household, Community AND Outer Social Network

PAGE 262

243 APPENDIX D : POWER BY SOCIAL NETWORK TYPE Figure D 1. Distribution based on power in household social n etworks (high=1; low=6) Figure D 2. Distribution based on power in farm community social n etworks (high=1; low=8) Figure D 3. Distribution based on power in labor social n etworks (high=1; low=7) 0 16 45 7 0 7 0 10 20 30 40 50 1 2 3 4 5 6 Household Power 1 52 33 4 15 5 8 2 0 20 40 60 1 2 3 4 5 6 7 8 Farm Community Power 15 1 22 13 1 35 2 0 20 40 1 2 3 4 5 6 7 Labor Power Household Power Freq. Percent Cum. 1 0 0 0 2 16 21.33 21.33 3 45 60.00 81.33 4 7 9.33 90.67 5 0 0 90.67 6 7 9.33 100 Total 75 100 Farm Community Power Freq. Percent Cum. 1 1 0.83 0.83 2 52 43.33 44.17 3 33 27.50 71.67 4 4 3.33 75.00 5 15 12.50 87.50 6 5 4.17 91.67 7 8 6.67 98.33 8 2 1.67 100 Total 120 100 Labor Power Freq. Percent Cum. 1 15 16.85 16.85 2 1 1.12 17.98 3 22 24.72 42.70 4 13 14.61 57.30 5 1 1.12 58.43 6 35 39.33 97.75 7 2 2.25 100 Total 89 100

PAGE 263

244 Figure D 4. Distribution based on power in vendor social n etworks (high=1; low=9) Figure D 5. Distribution based on power in landlord social n etworks (high=1; low=10) Figure D 6. Distribution based on power in industry social n etworks (high=1; low=8) 13 7 2 1 0 16 51 4 16 0 20 40 60 1 2 3 4 5 6 7 8 9 Vendor Power 0 13 21 9 6 18 9 8 29 5 0 10 20 30 40 1 2 3 4 5 6 7 8 9 10 Landlord Power 21 35 1 3 16 14 3 3 0 20 40 1 2 3 4 5 6 7 8 Industry Power Vendor Power Freq. Percent Cum. 1 13 11.82 11.82 2 7 6.36 18.18 3 2 1.82 20.00 4 1 0.91 20.91 6 16 14.55 35.45 7 51 46.36 81.82 8 4 3.64 85.45 9 16 14.55 100 Total 110 100 Landlord Power Freq. Percent Cum. 2 13 11.02 11.02 3 21 17.80 28.81 4 9 7.63 36.44 5 6 5.08 41.53 6 18 15.25 56.78 7 9 7.63 64.41 8 8 6.78 71.19 9 29 24.58 95.76 10 5 4.24 100 Total 118 100 Industry Power Freq. Percent Cum. 1 21 21.88 21.88 2 35 36.46 58.33 3 1 1.04 59.38 4 3 3.12 62.50 5 16 16.67 79.17 6 14 14.58 93.75 7 3 3.12 96.88 8 3 3.12 100 Total 96 100

PAGE 264

245 Figure D 7. Distribution based on power in social networks with other r elations (high=1; low=7) 1 4 9 13 16 11 41 0 20 40 60 1 2 3 4 5 6 7 Other Relations Power Other Relations Power Freq. Percent Cum. 1 1 1.05 1.05 2 4 4.21 5.26 3 9 9.47 14.74 4 13 13.68 28.42 5 16 16.84 45.26 6 11 11.58 56.84 7 41 43.16 100 Total 95 100

PAGE 265

246 APPENDIX E : EXAMPLES OF LIVELIHOOD STRATEGIES Table E .1 Examples of livelihood strategies quoted from interview notes Livelihood strategies by category Example response Find more land to farm If they do not make enough profit this year, they will look for other land. They will stay as long as they can farm; they prefer it here. They will look for more land around in case they are evicted. There is plenty of land available in Delhi. They have contacts here now and will find land someplace else. They do not plan to leave Delhi; they have a better living here. They are looking for more land because the landlord told them they must vacate after this crop is harvested. They think they will be able to find more land, but they will stay here until they are evi cted. They will just roam around until they find something because they don't know anyone right now who can tell them where there is land. They don't care that much about this land and will be fine if they have to move someplace else. There is a lot of land in the peri urban areas around Delhi. Will fight for their rights he believes that as a resident of Delhi he should get land. Find more land or return to village They try to help each other find land, but if they cannot find land they will have to go back to their villages. They have been here for a long time and plan to stay. But even those who have a long tenure will return to their villages if they are evicted and cannot find more land. Not sure if they will look for more urban land or go back to a rural area. Whoever has contacts will find land in Delhi and those without contacts will return to their villages. Return to homeland to farm If the land is taken, they will go back to their village. They are very dependent of the farmers if the farmers go then they will have to go back. There is no other place to grow vegetables in Delhi so they will go back to their village.

PAGE 266

247 Table E .1 (con't) Examples of livelihood strategies quoted from interview notes Livelihood strategies by category Example response Find more land or sell vegetables If they cannot get land, they will buy produce from other farmers and sell it. They will live somewhere else and continue to sell vegetables. Sell vegetables instead of farm Just need land near farms to put a house because they only sell vegetables. Find another place to live They will take a room on rent and do some other work. Other jobs They are getting old; they are educating their kids so they can have better jobs. They will just work for anyone who needs help. He is looking for any kind of job other than farming. They will have to work as day laborers if they have to leave. Not sure Not sure. They will worry about what to do next when the time comes. He hasn't seen a place as big as this in Delhi so he doesn't think they will find more land. It will be difficult to find more land because all the farmers will also be looking for more land.

PAGE 267

248 Table E .2 Households with the strong social structure HH Livelihood Strategy Background 1 The landlord will give them land next year if the government allows it. They will work as daily laborers in agriculture or look around for oth er land if they have to leave. They have lots of contacts. Tenure >20 years; from Badaun; don't pay rent; 2 to <5 bigha; >50 years old; small HH size. 2 They will go back if their huts are broken. Tenure 3 months; from Badaun; on batai; 2 to <5 bigha; >50 years old; HH size unknown. 3 As long as there is land, they will grow. They will go back to their villages. Tenure 1 year; from Badaun; on batai; 2 to <5 bigha; age is <30; HH size unknown. 4 She is new so she doesn't know many contacts around and doesn't know where to go. They have a house in Bareilly so they can always go back there. Those who have been here some time may know where else to go to farm but they are new so they will have no choice but to return to Bareilly. Tenure 4 years; from Bareilly; on batai; 5 to <10 bigha; age is 40s; small HH size. Table E .3 Households with the weak social structure HH Livelihood Strategy Background 5 Shakarpuri, Laxmi Nagar all places near Yamuna they will go work there is their land is taken away. Tenure 8 years; from Badaun; on batai; 2 to <5 bigha; age is 30s; medium sized extended family. 6 They will g o back if land is taken away. If there is a flood, they will take a room in Trilokpuri. Tenure 7 years; from Badaun; on rent; 15 to <50 bigha; age is 30s; medium sized extended family. 7 She will probably move into th e city with her parents if they ar e asked to leave. Although she likes it, here because it is so spacious and green, cool and kids have a lot of area to play in. They had taken up a place on rent during the floods. Her mother works in the factory in the city. Her parents are also living nearby. Have taken it on rent in Sarai Kale Khan. She doesn't work bec ause her husband doesn't allow. Although she would really like to she gets bored at home. From Delhi; not a migrant; on rent; 2 to <5 bigha; <30 year s old; small HH size. 8 They will rent a room somewhere if they are made to leave. Tenure 8 years; from Bareilly; unknown rent status; 2 to <5 bigha; age is 30s; small HH size.