Citation
Effects of Urban Water Scarcity on Sociability and Reciprocity in Cochabamba, Bolivia

Material Information

Title:
Effects of Urban Water Scarcity on Sociability and Reciprocity in Cochabamba, Bolivia
Copyright Date:
2008

Subjects

Subjects / Keywords:
Anthropology ( jstor )
Bodies of water ( jstor )
Cultural anthropology ( jstor )
Groundwater ( jstor )
Rain ( jstor )
Regression analysis ( jstor )
Scarcity ( jstor )
Social interaction ( jstor )
Water resources ( jstor )
Water usage ( jstor )

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Embargo Date:
7/24/2006

Downloads

This item has the following downloads:


Full Text












THE EFFECTS OF URBAN WATER SCARCITY ON SOCIABILTY AND
RECIPROCITY IN COCHABAMBA, BOLIVIA















By

AMBER YODER WUTICH


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2006

































Copyright 2006

by

Amber Yoder Wutich

































For Annette and Ashley















ACKNOWLEDGMENTS

A number of people have been involved in the making of this dissertation. Each

contributed, in small and large ways, to my intellectual growth, emotional health, and

physical safety. I thank each of them sincerely.

At the University of Florida, my mentor Dr. H. Russell Bernard worked tirelessly

to guide my scholarly development and encourage my intellectual independence. My

committee members, Drs. James Jawitz, Anthony Oliver-Smith, and Marianne Schmink,

contributed enormously to the conception and execution of the research. Dr. Chris

McCarty provided a working environment, at the Bureau of Economic and Business

Research Survey Research Center, where I honed my research skills. My friends from the

SOR (Students of Russ) group, including Stacy Giroux, Lance Gravlee, Mark House,

David Kennedy, Rosalyn Negron, Elli Sugita, and others, continually provide

entertainment, companionship, and guidance in anthropology's unusual milieu.

In South Florida, a loving network of family and friends has supported me and my

studies over the last ten years. My mother and sister, Annette Victor and Ashley Yoder,

have done everything imaginable to ensure my success and happiness. My mother

deserves special recognition for taking on onerous paperwork duties, as does my sister for

accompanying me during three months of fieldwork in Villa Israel. I thank my family,

Ronald Yoder, Ann Barlow, Greg Victor, Mort Victor, Lois Victor, and Kyle Victor, for

their continual support. The encouragement of our family friends, including the Cohen,

Kavanaugh, and Gomez families, has also been most appreciated.









I have always relied on one small circle of amazing women, and each of them made

extraordinary contributions toward this work. Candice Aloisi, Neha Gandhi, Kristin

Kavanaugh, Kathleen Ragsdale, Alicia Turner, and Wilda Valencia all provided warm

beds, delicious meals, advice, perspective, laughter, and-most importantly-their

friendship.

In Bolivia, I would not have become a true participant in Cochabamba life without

the love and guidance of the Valencia family. I consider the Valencias-Wilfredo,

Dominga, Willycito, Lidia, Wilda, and Richard Aguilar-to be my family, as much as if

we were born of the same land and blood. I also thank the people of Villa Israel for

accepting me in their homes and lives. I thank Susana Southerwood and Abraham

Aruquipa of Water for People and Jim Schultz of Water for People for their support in the

field. I am grateful for the research funds provided by the National Science Foundation,

IIE-Fulbright, Paul and Polly Doughty, and the Center for Latin American Studies at the

University of Florida.

Finally, I give my heartfelt thanks to Luis Fernando Amarilla Avalos for his

companionship and support during fieldwork and after. His decision to stay in Bolivia

was a turning point in my life and his.
















TABLE OF CONTENTS

page

A C K N O W L E D G M E N T S ................................................................................................. iv

LIST OF TABLES ........... ................................ xi

LIST OF FIGURES ......... ....................... .......... ....... .......... xiv

ABSTRACT. ....................................... xvii

CHAPTER

1 IN TR OD U CTION ............................................... .. ......................... ..

Two Practical Problems: Environmental Sustainability and Human Well-being ........2
Problem One: Studying Human-Environment Interactions ................................2
Problem Two: Survival and Well-being of the Urban Poor................................4
New Directions in Interdisciplinary Research on Urban Water Scarcity ............6
Toward a Cultural Anthropology of Urban Water Scarcity in Latin America.............6
Politics, Economics, and the Urban Poor in Latin America: 1960-2005 ............7
Urban W ater Distribution in Latin Am erica............................... .....................9
Understanding Human Responses to Urban Water Scarcity ...................................11
Ecological Anthropology: Classic Contributions and New Directions ..............11
Building a Theory to Test Human Responses to Urban Water Scarcity .............14
Chapter C conclusion ....................... .................... ... .... ........ .......... 16

2 BOLIVIA: A CONCISE HISTORY ........................................ ...................... 18

In tro d u ctio n .................. .................. ................... ............. ................ 18
Geography, Environment, and Natural Resources ...................................................19
A Brief History of Bolivia's Social, Political, and Economic Organization..............21
C culture and R religion ........... ................ ............................ .. .......... .. ........... .. 23
Recent Demography: Migration, Identity, and Urbanization .................................. 25
Recent Econom ic H history ........................................................................ 26
Recent Political History .................. ............................ ........ ................. 28
Bolivian Economy and Politics After 2000 ........................................................... 29
Poverty, Well-being, and Livelihoods in Contemporary Bolivia.............................31










3 S E T T IN G ........................................................................................3 4

F finding the F field Site............. .... ...................................................... .......... ....... 34
G gaining E ntry........................................................................................ 35
Landscape ................................................37
The Cadence of Com m unity Life ........................................ .......................... 39
O rig in s ...............................................................................4 1
Com m unity D evelopm ent................................................. .............................. 42
Economy ic W ell-being ................................................. .... ..... .. ...... .... 44
Economy y and Education ................................................... ... ........................ 46
Households and the Division of Labor .................................................................... 47
Language and Gender .................. ............................ ........ ................. 49
R religion and Politics ......................................................... .. ............ 51

4 F IE L D M E T H O D S ........................................................................... .................... 53

G aining Com m unity Support........................................................... ............... 54
Participant-O observation .............................................................................55
The R research Team .......................... .............. ................. .... ....... 58
The Sampling Fram e ......................... ...... .. .. .................................. 61
Interview Sampling, Informed Consent, and Compensation....................... ...64
D developing Interview Protocols ........................................ ........................... 67
M measuring W after Provision and Scarcity ....................................... ............... 71
D direct O b serve action ........ ...................................................................... ........ .. ....... .. 75
Sum m ary of Field M ethods ............................................................................... .... 78
Contextualizing the Research Design: Rethinking Community and Cooperation .....78

5 THE POLITICAL ECOLOGY OF URBAN WATER SCARCITY ..........................81

Introdu action ........................ ............ ........... ............ .. ........................... .. 8 1
Section 1-Political Ecology of Urban Water Provision and Scarcity in
C ochab am b a .................. ......... ....................................................... 82
Introduction to Cochabamba's Water Situation ...............................................82
Hydrology of the Cochabamba Valley ....................................................83
Water Resources in the Cochabamba Valley ........................................87
Power, Privilege, and Water Provision in the City of Cochabamba ...................89
Section 2-Water Resources, Provision, and Storage in Villa Israel.........................94
Local W ater R esources......................................................... ............... 94
Surface W ater Collection ............................................................................. 97
R ainw after C collection .................................................................... ............... 100
Community W ater System : Tap Stands .................................... ............... 101
Aguateros, or W ater Delivery Trucks .................................... ............... 104
W ater Loans, Gifts, and Charity................................................................ 108
Buying Water from Businesses and Neighbors.............................................. 112
W ater Storage and Purification .................................................. ................... 114
Section 3-Defining and Operationalizing the Urban Water Scarcity Concept ......116









Approach One: Absolute Measures of Water Provision .................................117
The self-adm inistered direct m measure ................................... ................. 118
The task-based w ater use estim ate ............ ................................... .........119
The overall estimate of weekly water use ..........................120
Comparing three measures of water provision................... ...............120
Discussion: What do the water provision data really mean?......................124
Approach Two: Proxy Measures of Water Security .............. .... ...............125
P precipitation ............................................... ................. 126
Physical location of households .............................................................. 127
Political and religious connections............................................................128
W ater storage capacity ........................................................................ 130
Summary: Proxy measures of water security............................................131
Approach Three: Measures of the Experience of Urban Water Scarcity ..........131
From interpretation to measurement ................................ ................... 132
D ata collection and analysis.................................... ....................... 133
B odily distress ............................................................................. 136
W ater-based tasks ................................. .... ............ .. .......... .. ...... .. 137
Buying water from the aguatero............... ..............................................137
W ater exchanges ........................................ ............................... 138
W ater conflicts ......... ............ .............. .... ...... .............. 139
Summary: Measures of the experience of water scarcity...........................140
Discussion: Defining and operationalizing the urban water scarcity
concept .... ...... ........ .............. ..... .. ....... ...... ......... 141

6 SOCIAL RELATIONS AND ECONOMIC EXCHANGES .............. ...............143

Anthropological Research on Reciprocity ........................................... ...............143
Reciprocity as a Social Foundation ............ ............................................ 144
Reciprocity as a System of Economic Exchange .............................................145
U rban R eciprocity .............................................. ................ ............ .... ..146
Sociability, Reciprocity, and the Laughlin-Brady Model................. ................147
Studying N ested Scarcity Response Cycles ..................................................... .... 148
Data on Sociability, Reciprocity, and Participation in Villa Israel...........................150
Public Social Interactions ...................................................... ............... 150
Data collection and coding ...... ..................... ...............150
D ata an aly sis ..................... ..................... ..... ..................15 1
Household-level data on Sociability, Reciprocity, and Participation..............152
R espondent selection........................................... ........................... 152
D ata collected at the household level ....................................................... 153
D ata an aly sis ............................................................... 15 5
Cultural Systems and Values are Unique ........................................ .............157
Data on Exchanges within Four Cultural Institutions ......................................159
R religion and church ............................................................................160
N neighborhood .................. .......................... .... .... ................. 163
F a m ily ................................................................................................... 1 6 5
B u sin e sse s ............................................................................................ 1 6 6
C onclu sion ..................................................................................................... 168









7 D A T A A N A L Y SIS ......................................................................... ................... 170

Section 1-Changes in Water, Sociability, Reciprocity, and Participation over
T im e .......................................................................................170
A N ote on M ethods of A analysis ............................................... .................. 170
Changes in W ater Availability over Time .............................. ................... 171
Secure access to water: Rainfall measure ...............................................171
Experience of water scarcity: Water-based task scale............................173
Household water provision: Task-based water use estimate ................175
Conclusion: Changes in water availability over time..............................178
Changes in Sociability, Reciprocity, and Community Participation over Time 179
Sociability: Changes in food sharing over time ......................................180
Sociability: Changes in visiting over time ............................................181
Sociability: Changes in public social interactions over time ...................183
Reciprocity: Changes in loans over tim e...................................................185
Reciprocity: Changes in helping over tim e...............................................185
Participation: Changes in church attendance over time ...........................185
Participation: Changes in neighborhood council meeting attendance over
tim e ..................... .............. ....... ... ......... .... ...... ........... 18 5
Conclusion: Changes in sociability, reciprocity, and participation over
tim e ........................... ................. ...... .......... ................ 18 6
Section 2-Explaining Social Interactions across Households ...............................188
Identifying Independent V ariables ....................................... ............... 188
Building the Regression M odel ...................................................................... 191
Understanding the Regression Model ...................................................... 193
Section 3-Explaining Household Social Interactions over Time...........................195
D ata A analysis Conclusions ............................................ .............................. 197

8 CONCLUSION S ................................... .. .. ........ .. .............199

Section 1- Sum m ary of Findings ........................................ ........................ 199
Section 2-Contributions to Theory and Practice.......................... ...............201
Theoretical Contributions to Anthropology ............................................... 201
Urbanism Poverty, and W ater Scarcity ................................. ............... 201
U rban R eciprocity ................ .. ..... ......................................... ... 202
Human Adaptations to Resource Scarcity.............................. ..................203
Interdisciplinary Contributions............. .... ......... ...............204
Ending Urban Water Scarcity: Six Recommendations ...................................207
Section 3-Looking Ahead in Cultural Anthropology............................................210
Moving Forward: Anthropological Research on Urban Water Scarcity ..........210
Cultural Anthropology's Future: Collaboration, Method, and Theory ............211
Collaboration ................................... ................................. 211
M eth o d ................... ............................................................. 21 2
T h e o ry .................................................................................................. 2 1 3

APPENDIX: REGRESSION MODELS .................................................................. 215










L IST O F R E FE R E N C E S ...................................................................... .... ..................233

B IO G R A PH IC A L SK E T C H ........................................ ............................................249

























































x
















LIST OF TABLES


Table page

4-1 The semi-structured interview protocol, modified from Stack 1970 .....................68

5-1 Descriptive statistics for three measures of water scarcity .................................. 122

5-2 English translations of thirty-three yes/no questions about the experience of
w ater scarcity in V illa Israel ........................................................ ............... 134

5-3 Thirty-three yes/no questions about the experience of water scarcity in Villa
Israel, written in the south Cochabamba Spanish dialect..................................... 135

7-1 Factor loadings for six potential predictors of social interactions, for summaries
of data collected over five interview periods. ......................................... ..........190

7-2 Regression models predict two kinds of social interactions, using four
independent variables. Data are means for five data periods.............................192

7-3 Regression models predict two kinds of social interactions, using two measures
of water availability. Data are means for five data periods..................................194

7-4 Regression models predict visiting and food sharing for two study periods:
June-July 2004 and August-September 2004. .............................................196

A-i Regression model predicts loans using four independent variables. Data are
m eans for five data periods. ............................................ ........................... 215

A-2 Regression model predicts helping using four independent variables. Data are
m eans for five data periods. ........................................... ............................ 216

A-3 Regression model predicts participation in Neighborhood Council meetings
using four independent variables. Data are means for five data periods..............216

A-4 Regression model predicts participation in church meetings using four
independent variables. Data are means for five data periods.............................217

A-5 Regression model predicts loans using two independent variables. Data are
m eans for five data periods. ........................................... ............................ 217

A-5 Regression model predicts helping using two independent variables. Data are
m eans for five data periods. ........................................... ............................ 218









A-6 Regression model predicts participation in Neighborhood Council meetings
using two independent variables. Data are means for five data periods ..............218

A-7 Regression model predicts participation in church meetings using two
independent variables. Data are means for five data periods.............................219

A-8 Regression model predicts food sharing using two independent variables. Data
were coded from interviews conducted during study period 1 (April-May 2004).219

A-9 Regression model predicts visiting using two independent variables. Data were
coded from interviews conducted during study period 1 (April-May 2004). .......220

A-10 Regression model predicts loans using two independent variables. Data were
coded from interviews conducted during study period 1 (April-May 2004). .......220

A-11 Regression model predicts helping using two independent variables. Data were
coded from interviews conducted during study period 1 (April-May 2004). .......221

A-12 Regression model predicts participation in Neighborhood Council meetings
using two independent variables. ........................................ ....................... 221

A-13 Regression model predicts participation in church meetings using two
independent variables. ............................... ................ ................ ............. 222

A-14 Regression model predicts food sharing using two independent variables. Data
were coded from interviews conducted during study period 2 (June-July 2004)..222

A-15 Regression model predicts loans using two independent variables. Data were
coded from interviews conducted during study period 2 (June-July 2004). ..........223

A-16 Regression model predicts helping using two independent variables. Data were
coded from interviews conducted during study period 2 (June-July 2004). ..........223

A-17 Regression model predicts participation in Neighborhood Council meetings
using tw o independent variables.. ............................................... ............... 224

A-18 Regression model predicts participation in church meetings using two
independent variables. ............................... ................ ................ ............. 224

A-19 Regression model predicts loans using two independent variables.......................225

A-20 Regression model predicts helping using two independent variables....................225

A-21 Regression model predicts participation in Neighborhood Council meetings
using tw o independent variables.. ............................................... ............... 226

A-22 Regression model predicts participation in church meetings using two
independent variables.......... .... ...................... .......... ............. ...... .... ................226









A-23 Regression model predicts food sharing using two independent variables ..........227

A-24 Regression model predicts visiting using two independent variables....................227

A-25 Regression model predicts loans using two independent variables .....................228

A-26 Regression model predicts helping using two independent variables....................228

A-27 Regression model predicts participation in Neighborhood Council meetings
using tw o independent variables.. ............................................... ............... 229

A-28 Regression model predicts participation in church meetings using two
independent variables. ............................... ................ ................ ............. 229

A-29 Regression model predicts food sharing using two independent variables ..........230

A-30 Regression model predicts visiting using two independent variables....................230

A-31 Regression model predicts loans using two independent variables. ......................231

A-32 Regression model predicts helping using two independent variables....................231

A-33 Regression model predicts participation in Neighborhood Council meetings
using tw o independent variables.. ............................................... ............... 232

A-34 Regression model predicts participation in church meetings using two
independent variables. ............................... ................ ................ ............. 232
















LIST OF FIGURES


Figure page

4-1 A map of Villa Israel land plots (from Claure, Periera, & Asociados 2001)
similar to the one used to build a sampling frame...........................................62

4-2 The first page of the water data collection tool. Artwork by Ashley Yoder, 2004..72

4-3 The second page of the water data collection tool. Artwork by Ashley Yoder,
2 0 04 .................................................................................7 3

5-1 Map of the Cochabamba Valley (from Stimson et al. 2001, p. 1100). ....................84

5-2 Seasonal variations in rainfall for the city of Cochabamba (from Vera undated).... 86

5-3 Northern Cochabamba has alluvial fans, forests, plant life, high rises, and the
residences of the city's wealthiest people. ...................................................87

5-4 Southern Cochabamba is arid, barren, and contains the residences of the city's
poorest people. ..................................................... ................. 87

5-5 Villa Israel's river A) during the rainy season, in January 2005 and B) during the
dry season, in June 2004. ............................................... .............................. 95

5-6 Villa Israel's runoff canals are dry year round, except during heavy rains..............96

5-7 The Villa Israel river during the wet season.................................. ............... 97

5-8 People in Villa Israel rely on river water for bathing ...........................................97

5-9 The riverbed is a center for social interactions where women wash and chat.........98

5-10 Children splash and play in the river during the wet season ..................................98

5-11 Wastewater is disposed of in unsealed dumping holes. Here, the hole is located
inside a housing compound. ..... ........................... ......................................99

5-12 The open area between these two bushes is used as an outdoor toilet by many
Villa Israel residents. Wastes drain directly into the riverbed below....................99

5-13 Rooftop rainwater collection equipment includes a drain from roof gutters, a
turril, and a pipe that extends to a water tank underground................................100









5-14 Tap stands are small brick columns built around a water pipe. ..........................102

5-15 Water delivery trucks, or aguateros, are the most common source of water in
Villa Israel. Note the aguatero passing through the entrance to Villa Israel.........105

5-16 People buy water from businesses like public baths, corer stores, and
restaurants ............................... .. .................................... 113

5-17 Line graph depicting average statistics for three measures of water provision,
collected across five survey periods and 76 households.....................................123

5-18 Line graph depicting monthly rainfall in the Cochabamba Valley (17.45 S and
66.09 W ), from January 2004 to January 2005. .................. ............................. 126

5-19 Line graph depicting water storage capacity among Villa Israel households........130

6-1 Relationships hypothesized by the Laughlin-Brady model................................ 148

6-2 Map of Villa Israel, with dots representing 9 of 60 randomly sampled public
places................................ ... .................................. .......... 152

6-3 Means for five measure of reciprocity, sociability, and meeting participation ......156

6-4 Means for loans that took place within four cultural institutions in Villa Israel
during August and September 2004. .......................................... ............... 159

6-5 Means for the number of loans, helping, and food sharing that evangelicals and
Catholics engaged in between April 2004 and January 2005. ............................162

7-1 Total rainfall in mm across six data collection periods, from January 2004 to
January 2005, in Cochabamba, Bolivia. (Rainfall data from SENAMHI 2005). ..172

7-2 Mean composite household scores for Guttman task-based measure of water
scarcity over four study periods, from June 2004 to January 2005, in Villa Israel.173

7-3 Total Cochabamba rainfall is plotted against the task-based water scarcity
measure for data collected between June 2004 and January 2005 .......................175

7-4 Mean estimates of household water use (per person, per day) over four study
periods, from April 2004 to November 2004, in Villa Israel..............................176

7-5 Mean Cochabamba rainfall is plotted against the household water use measure
for data collected between April 2004 and November 2004 ..............................178

7-6 Mean numbers of food sharing over five study periods, from April 2004 to
January 2005, in V illa Israel. ........................................................................... 180

7-7 Mean number of visits over five study periods, from April 2004 to January
2005, in V illa Israel .................. ............................. ........ .. ........ .. .. 182









7-8 Mean number of public social interactions over four study periods, from June
2004 to January 2005, in Villa Israel. ....................................... ............... 184

7-9 Mean number of public social interactions over four study periods, from June
2004 to January 2005, in Villa Israel. ....................................... ............... 185

7-10 A plot depicting factor loadings for three factors. ............................................190















Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

THE EFFECTS OF URBAN WATER SCARCITY ON SOCIABILITY AND
RECIPROCITY IN COCHABAMBA, BOLIVIA

By

Amber Yoder Wutich

May 2006

Chair: H. Russell Bernard
Major Department: Anthropology

Global trends like desertification, urban growth, and economic restructuring are

making water increasingly scarce-and water access increasingly inequitable-in cities

around the world. While much is being done to extend water provision systems to the

urban poor, the roots of urban water scarcity are complex and difficult to resolve. Today,

as urban water scarcity worsens, research is needed to help mitigate its impact on the

environment and human welfare.

This study is designed to engage with interdisciplinary scholarship on two practical

problems in urban water scarcity: (1) small-scale human-environment interactions and (2)

changing survival strategies among the urban poor. The study is also designed to answer

one unexamined question in anthropological theory: what are the effects of urban water

scarcity on social interactions and economic exchanges? To do so, I employ the

Laughlin-Brady model of human adaptation to resource scarcity.









Over 18 months, I performed participant-observation, direct observations, and

survey interviews in an impoverished, water-scarce neighborhood in Cochabamba,

Bolivia. The dissertation examines the political ecology of urban water scarcity in

Cochabamba, and shows how hydrological, infrastructural, and social trends combine to

create conditions of severe water scarcity for residents of the city's poorest residents. The

study contains a thorough discussion of the definition and operationalization of the

"urban water scarcity" concept, including water provision, water security, and the

culture-bound experience of water scarcity.

The data indicate that water scarcity does affect sociability during the dry season,

as predicted by the Laughlin-Brady model. When people are first hit by severe water

scarcity, they become more sociable. As scarcity worsens, however, sociability decreases.

The data also show that water security and provision are positively correlated with

sociability during the dry season, but that the relationship disappears during the wet

season. There were no significant changes in reciprocity. The findings indicate that social

responses to water scarcity mimic responses to other resource scarcities (like famine),

and point to future directions for research on and prevention of severe urban water

scarcity.


xviii














CHAPTER 1
INTRODUCTION

Global trends like desertification, urban growth, and economic restructuring are

making water increasingly scarce-and water access increasingly inequitable-in cities

around the world. At least 157 million urbanites have no access to an improved water

source, and hundreds of millions more lack adequate access to safe water. The vast

majority of people without safe water are concentrated in developing cities in Latin

America (15 percent), Africa (25 percent), and Asia (57 percent) (UN-Habitat 2003).

While much is being done to extend water provision systems to the urban poor, the roots

of urban water scarcity are complex and difficult to resolve. Today, as urban water

scarcity worsens, research is needed to help mitigate its impact on the environment and

human welfare.

Anthropological theories of resource scarcity have much to contribute to

interdisciplinary research on urban water scarcity. In this study, I examine the effects of

urban water scarcity on sociability and reciprocity-and show how the findings relate to

ongoing interdisciplinary research. The study is based on 18 months of field research

conducted in an impoverished, water-scarce neighborhood called Villa Israel located on

the outskirts of Cochabamba, Bolivia. There, my research team and I performed

participant-observations, direct observations, and survey interviews. I will dedicate later

chapters to discussing water scarcity, sociability, and reciprocity in Villa Israel. Here, I

wish to focus first on the broad practical and theoretical issues with which this study

engages.









The study's theoretical contribution is to advance anthropological theory in ways

that contribute to (1) interdisciplinary theories of urban water scarcity and (2)

anthropological knowledge of human belief and behavior. This chapter, then, is divided

into two main sections. In the first section, I examine real-world problems that motivate

social, biological, and physical scientists to study water-scarce urban environments. In

the second section, I show that urban water scarcity is a topic that bridges theoretical

issues in urban, economic, and ecological anthropology. To understand its causes and

hypothesize its effects, I examine anthropological research on the urban poor in Latin

America, the structure of urban services in Latin American cities, and human adaptations

to environmental scarcity across cultures.

Two Practical Problems: Environmental Sustainability and Human Well-being

Studies of urban water scarcity are needed for compelling environmental and social

reasons. Below I briefly examine two key problems that this study engages: (1) integrated

human-environment interactions and (2) the survival and well-being of the urban poor.

For each, I briefly discuss the relevant academic research in the problem area, and situate

this study within it.

Problem One: Studying Human-Environment Interactions

In this section, I discuss the environmental aspects of water scarcity. First, I

examine how human-environmental interactions shape urban water scarcity. Next, I

discuss how interdisciplinary collaborations have developed to examine this issue.

Finally, I explain how this study contributes to new research in human-environment

interactions.

Despite extensive efforts to improve water delivery and quality over the last two

decades, 1.1 billion people still lack adequate access to water resources around the world









(Gleick 2004). Although resolving water problems is a major international policy

priority, it will be difficult to sustainably manage water resources until we understand the

complex human-environment interactions that produce local scarcities. Ecosystems shape

the physical availability of water resources for human use around the world (Gleick,

Singh, and Shi 2001). However, human social systems overlay these natural systems, and

shape peoples' ability to access water resources. From these complex ecological and

social systems emerge human patterns of environmental exploitation. Human patterns of

environmental exploitation, in turn, have enormous impacts on ecosystems, which

ultimately affects the quality and distribution of global water resources (Daily 1997). Put

simply, cyclical interactions between the natural ecosystems and human systems are

complex, ever-changing, and the key to creating systems of sustainable water

management.

Over the last decade, many scholars in the natural sciences have become interested

in collaborating with social scientists to understand social-ecological systems and human-

environment interactions (cf. Gunderson and Holling 2002). Recently, successful

interdisciplinary research teams have been built around the study of urban ecosystems, a

new area of research that has the potential to yield generalizable theories and better

management practices for urban social-ecological systems (Collins et al 2000, Grimm et

al 2000, Pickett et al 2001, Redman, Grove, and Kuby 2004). The urban ecosystems

approach has much to contribute to the study of urban water scarcity, in which scholars

often treat the city and its water systems as analytically isolated from basin-wide

hydrological and ecological issues (Falkenmark and Lundqvist 1995).









While this study is predominately anthropological in its use of theory and methods,

it is designed to engage with outside disciplines. Without a serious discussion of Villa

Israel's wider social and ecological environment-and particularly the hydrology of the

Cochabamba valley-it would be impossible to understand urban water scarcity and its

profound effects on daily life there. This study, then, represents a modest contribution to

cross-disciplinary efforts to integrate the social and ecological sciences in the study of

urban ecosystems, especially with regard to the relationship between desertification,

urbanization, and the sustainability of human systems of water distribution and use.

Problem Two: Survival and Well-being of the Urban Poor

In this section, I discuss the effects of water scarcity on the survival and well-being

of the urban poor. A number of studies have established that water scarcity and water

degradation have serious effects on household economies and personal health. Beyond

this, however, few have examined the effects of water scarcity on families, communities,

and cultural institutions. After briefly reviewing the economic and health impacts of

water scarcity, I argue that more should be done to understand how urban water scarcity

affects social, economic, and cultural institutions among the urban poor.

In developing cities, urban water distribution generally takes a center/periphery

form, in which well-provisioned downtown give way to outskirts with progressively

fewer municipal services (Gilbert 1998). Without adequate infrastructure, the urban poor

suffer from severe water scarcity even in cities that have enough water to meet their

residents' needs. As over-abstraction (or the excessive removal) of groundwater and

desertification diminish available water resources, even independent community wells are

running dry. When people lack municipal and community water sources, they usually

rely on private water vendors that charge 4 to 100 times the price of municipal water









(Elhance 1999). The relatively high cost of water burdens the over-strapped budgets of

the urban poor. Many household heads must make decisions about whether scarce cash

will be spent on food or water, and whether precious time will be spent earning income or

trying to track down a water vendor.

While the economics of urban water distribution undermine the livelihoods of the

urban poor, water scarcity and degradation pose a direct threat to their survival. The

effects of unsafe water on health are staggering: 80 percent of illnesses and 30 percent of

deaths in developing countries are caused by the consumption of contaminated water

(Elhance 1999) through a variety of diseases such as cholera, typhoid, and dysentery. For

those who survive, water related illnesses can push at-risk families into a downward

spiral of health care costs and lost of work opportunities (Gleick 2004). In recognition of

these problems, the United Nations Millennium Development Goals aim to halve the

number of people without access to safe water and to significantly improve the lives of

100 million slum dwellers by 2015 (United Nations undated). Even if the goals are met-

which seems unlikely given the inadequacy of resources and funding currently dedicated

to the problem-between 34 and 76 million people will die of preventable water-related

illnesses by 2020 (Gleick 2004).

Water scarcity clearly does major damage to the health and economic well-being of

the world's urban poor. Beyond this, however, the effects of urban water scarcity on

families, communities, and cultural institutions have not been studied extensively. In

rural communities struck by drought and famine, an enormous amount of research shows

that families break down, communities disperse, and long-standing cultural institutions

are destroyed as people pursue survival at all costs (e.g., in studies of rural African









communities Turnbull 1972, Laughlin 1974, Cashdan 1985, Colson 1979 and reviews of

famine literature Dirks 1980, Corbett 1988, Walker 1989). Whether these findings can be

generalized to urban populations in today's political and economic conditions remains an

empirical question (cf. Ember and Ember 1992, Booth 1984)-one that is tested in this

research.

New Directions in Interdisciplinary Research on Urban Water Scarcity

In this section, I showed that urban water scarcity is a problem that bridges two

interdisciplinary research areas. New research on the political ecology of urban water

scarcity, and its effects on human social and economic interactions, would inform theory

in both problem areas. First, understanding how human systems of water distribution and

consumption work will help explain larger cycles of environmental degradation. Second,

understanding the effects of urban water scarcity on social and economic relations would

help us find the links between the direct effects of urban water scarcity (on health and

well-being) and possible secondary effects (on families, communities, and cultural

institutions). In the next section, I explain how I used anthropological research on urban

poverty, resource distribution, and resource scarcity to create new hypotheses regarding

the effects of urban water scarcity on sociability and reciprocity.

Toward a Cultural Anthropology of Urban Water Scarcity in Latin America

In the second half of this chapter, I begin to develop a research approach that tests

cross-cultural hypotheses of urban water scarcity, while situating them in their

ethnographic context. In the first section, I review changes in Latin American politics,

economy and the lives of the urban poor over the last thirty years. Recent Latin American

history helps explain how urban water scarcity emerged-and why it appears to be

worsening today. In the second section, I discuss the structure of water distribution and









scarcity in Latin American cities. In the third section, I examine anthropological research

on resource scarcity, asking how three phases of research in ecological anthropology-

classic findings, political ecology, and new ecology-can be integrated to theorize the

social effects of urban water scarcity. Finally, I pose four hypotheses regarding the

effects of water scarcity on sociability and reciprocity.

Politics, Economics, and the Urban Poor in Latin America: 1960-2005

Starting in the early 1960s, increased rural-urban migration and the buildup of huge

squatter settlements in Latin America sparked fears that the urban poor would organize

around a radical political agenda capable of toppling existing power structures. In

response to these fears, anthropologists and other social scientists extensively

documented the cooperative economic and conservative political strategies that people

used to combat scarcity. Scholars showed that the urban poor relied primarily on strong

familial relationships and reciprocal help networks to insure themselves against

unemployment, low wages, and a lack of water, electricity, and other urban services

(Stack 1974, Leeds 1971, Safa 1974, Lobo 1982, Lomnitz 1977, Isbell 1978, Halebsky

1995). Further, they found that impoverished migrants were not marginal or isolated, but

were bound up in patronage relationships with political elites (Eames and Goode 1973,

Dietz and Moore 1979, Perlman 1976, Portes and Walton 1976, Cornelius 1975, Lloyd

1979, Castells 1983, Mangin 1967) and rural communities (Isbell 1978, Buechler and

Buechler 1971, Lobo 1982, Graves 1974).

During the 1970s, the dynamics of urban land invasion and political patronage

became established, and more squatters came to the city expecting to acquire improved

basic services through clientalistic relations with the state. The more city outskirts and the

demands of their residents grew, the less capable the state became of supporting new









squatters (Castells 1983). When economic growth slowed across the region, Latin

American elites shifted to loan dependence to support social programs (like the extension

of basic services). By the early 1980s, most Latin American nations lacked the foreign

exchange to make loan payments and faced default (Weeks 1995). International

financiers arranged emergency aid packages that required compliance with neoliberal

economic reform programs. These structural adjustment programs called for the

privatization of state industries and free trade reforms including the acceptance of

expanded foreign investment, an end to subsidies, and a reduction in tariffs. The collapse

of national economies and the policies designed to correct it resulted in decreased wages,

unemployment, and cuts in state services and subsidies for Latin America's poorest

people (Safa 2004).

After the crisis of the 1980s, research began to indicate that cooperative economic

strategies that once protected livelihoods in shantytowns and squatter settlements were in

decline (Eckstein 1990). Gonzalez de la Rocha and her collaborators developed the

"poverty of resources" approach to explain changes in survival strategies among the

urban poor in Mexico. They argued that survival-oriented exchange networks only

function when people have enough goods, income, and labor available to invest in

reciprocal relationships. The restructuring of the economy and deterioration of labor

markets have profoundly eroded poor households' asset bases, undermining their ability

to sustain cooperative help networks (Gonzalez de la Rocha and Gantt 1995, Gonzalez de

la Rocha 2001, 2004). Moser also explained that global trends in infrastructural decline,

job loss, and economic crisis caused a shift from reciprocal community-level survival

strategies to preservationist household-level survival strategies among the urban poor in









Ecuador, Hungary, the Philippines, and Zambia. In a recent symposium on 30 years of

scholarship on Latin America's urban poor, leading scholars agreed that survival

strategies had been seriously undermined by recent economic trends (Safa 2004,

Gonzalez de la Rocha 2004, Ward 2004).

Latin America's cities share a history of rural migration, shantytown formation,

political patronage, economic crisis, and changing survival strategies among the urban

poor. These trends all contributed to the emergence of urban water scarcity-and help

explain why it seems to be worsening. In the following section, I explain in greater detail

how urban water distribution systems (and exclusion from them) are structured in Latin

American cities today.

Urban Water Distribution in Latin America

From a theoretical perspective, resource scarcity is commonly viewed as having its

origins in a shortage of supply, an excess of demand, elite resource capture, and

ecological marginalization (Homer-Dixon 1999). In semi-arid Latin American cities, all

four of these factors play an important role in creating municipal water distribution

systems that exclude the poor. However, many municipalities suffer neither from a

shortage of supply nor an excess of demand-they lack the political will or funding to

extend municipal systems to far-flung settlements where the urban poor are concentrated.

In that case, the inequitable distribution of entitlements, that is, the ability to access

resources (Sen 1999) probably best explains why some people have access to adequate

water resources in cities and others do not.

Power relationships in the city determine the extent of municipal water services,

rationing requirements, and the quality of water distributed to various neighborhoods

(Swyngedouw 2004). In most city centers in the developing world, for instance, water is









delivered through municipal systems directly into households. On urban peripheries,

however, entire neighborhoods often lack access to municipal water distribution systems.

In Latin America alone, 24 million people are estimated to have no urban water service

whatsoever, and many more have inadequate or unsafe service (UN-Habitat 2003). In

cities, the opportunity to collect drinkable surface water is severely restricted by water

pollution and public ordinance. Households usually buy low-quality water from a cistern

truck, which is operated by a private vendor. Private vendors typically charge 10 to 20

times the fee charged by public utilities, and people living in marginal urban areas pay

between 10 and 40 percent of their incomes to acquire water in this way (Marvin and

Laurie 1999, Swyngedouw 1997). As a result, the geographic marginalization of people

on the urban periphery deepens their economic marginalization (Satterthwaite 1998).

When rural migrants first began to move en masse to Latin American cities, they

too lacked municipal water services. After a period without services, living conditions

normally catalyzed groups of squatters to negotiate for community water sources-

whether through municipal connections or local wells. While the urban poor have always

been pushed to the worst available land, there had been more surface water available and

higher water tables also facilitated groundwater abstraction. After the collapse of state

patronage systems and funding for basic services, squatters' ability to tap state sources

for infrastructural improvements was severely limited. While non-governmental

organizations (NGOs) have filled in the funding gap to some extent, population growth,

over-abstraction, and climate change make it increasingly difficult to find water sources

in the areas where squatter settlements are located. Additionally, as the urban poor

withdraw from cooperative community networks, it becomes more difficult for them to









exert political pressure, self-organize to provide services, and maintain existing

community infrastructure.

While water scarcity can hit anywhere, it currently tends to be concentrated among

the urban poor, and especially among the impoverished residents of shantytowns and

squatter settlements. In Latin America, environmental, political, economic, and social

trends determine how water scarcity is distributed across urban populations.

Understanding Human Responses to Urban Water Scarcity

While this project is rooted deeply in urban and economic anthropology, it draws

primarily on ecological anthropology to find testable theories of human response to urban

water scarcity. In the first section below, I briefly review how ecological anthropologists

from three stages of research have all contributed to our understanding of resource

scarcity, and to some extent of urban water scarcity. In the second section, I argue that a

long-neglected theory of human response to resource scarcity-when updated with new

theoretical perspectives-proves an excellent framework for testing the effects of urban

water scarcity on human reciprocity and sociability.

Ecological Anthropology: Classic Contributions and New Directions

Fifty years of research in ecological anthropology has taught us a great deal about

how humans respond to changing ecological conditions and resource availability.

Cultural materialists and neo-functionalists demonstrated how competition for scarce

resources shaped the presence, function, and evolution of cultural traits (cf Vayda 1969,

Rappaport 1974, Harris 1979). Ecological anthropologists also explored how humans

respond to environmental stress through studies of demography, adaptive strategies, and

coping mechanisms (cf Vayda and McCay 1975, Laughlin and Brady 1978). While these

scholars made enormous contributions to our understanding of human-environment









interactions, they later came under attack in cultural anthropology for misunderstanding

basic biological principles, blinding themselves to the ways in which systems of political

and economic domination create non-functional adaptations, and making overly-

simplistic ecological determinist explanations of complex phenomena (Hallpike 1973,

Friedman 1974, Diener et al. 1980). In the 1980s, political ecologists began to examine

how ecological, economic, and social systems shape human-environment interactions (cf.

Durham 1979, Schmink 1982, Stonich 1989, 1993). The shift toward political ecology

was widely regarded as a move away from active borrowing of ecological terms and

concepts that characterized earlier research in ecological anthropology (Orlove 1980,

Little 1999). Over the last 20 years, political ecologists have developed new models for

how anthropologists can produce rigorous ethnographic work that examines the complex

interrelationships between ecological, economic, and social systems.

Negotiating the difficult theoretical terrain between anthropology and ecology

appears to have been a fruitful exercise for ecological anthropologists, as some are now

emerging as leaders in new interdisciplinary collaborations on social-ecological systems

(Redman 1999, Pavao-Zuckerman 2000, Abel and Stepp 2003, Stepp et al 2003). Cultural

anthropologists and archaeologists have been welcomed in interdisciplinary

collaborations on the study of urban ecosystems, where they lend expertise on social,

economic, and political issues (Wali et al 2003, van der Leeuw and Redman 2002). To

date, anthropologists' work on urban ecosystems combines political ecologists' concern

with class, gender, ethnicity, and power (cf Durham 1995) with the ecologists' need to

understand how human history, values, and desires figure into ecological outcomes. Early

results from two major urban ecology projects in which anthropologists participate (the









NSF Central Arizona Phoenix Long Term Ecological Research program and the Field

Museum's Chicago project) indicate that these collaborations, while difficult, are yielding

more complete understandings of urban human-environment interactions and

management needs.

While cultural anthropologists have yet to publish on water issues in urban

ecosystems, they are developing innovative approaches to the study of urban water

scarcity. For instance, authors in an edited text on Water, Culture, and Power: Local

Struggles in a Global Context use political ecological approaches to examine how

ecological and social factors combine to create local conflicts and water scarcity

(Johnston and Donahue 1998). In his study of "suffering from water" in a Mexican town,

Ennis-McMillan explained how people experienced bodily distress over the unjust

allocation of scarce water in a community where local hydrology, overpopulation,

poverty, managerial and technical negligence, and price gouging combined to create

water scarcity (2001, 2006). These studies demonstrate how anthropologists' analyses of

culture, politics, economy, and power relationships can contribute to scholars' overall

understanding of urban water issues. However, as Durham explained, future

anthropological research should go further-by explicating the causal links between

social processes, forming hypotheses, and testing the hypotheses with statistical analysis

(Durham 1995). Using reproducible scientific methods facilitates collaborations across

disciplines, as well as the accumulation of cross-cultural knowledge within anthropology.

When participant-observation is paired with scientific methods, research findings can be

particularly compelling.









Building a Theory to Test Human Responses to Urban Water Scarcity

I propose using one long-neglected anthropological theory-Laughlin and Brady's

1978 model of human adaptation to resource scarcity-to test the socio-economic effects

of water scarcity. Laughlin and Brady hypothesized that sociability and reciprocity

underwent a series of predictable changes during cycles of resource scarcity (like

seasonal scarcity or drought cycles). When people experience initial periods of scarcity,

they tend to unify to find cooperative solutions to the problem. People tend to increase

resource sharing and enlarge cooperative groups to include socially distant groups. As

scarcity worsens, however, people lack the resources and energy to sustain this kind of

enhanced sharing and cooperation. Instead, they withdraw from the enlarged sharing

network, narrowing their group participation and limiting reciprocal sharing to people

with whom they have stronger, more trusting, and more durable social ties.

As people withdraw from cooperative relationships, Laughlin showed, they often

switch from exchanges based on positive to generalized reciprocity or from generalized

to negative reciprocity (1974). With people increasingly looking for benefits for

themselves or their narrowed cooperative groups, competition increases and cleavages

form between groups. Cooperative goal-making drops off, and the likelihood that the

community will adopt cooperative economic or political tactics to deal with scarcity

decreases. Laughlin and Brady suggested that each society, then, has a structure that

governs the making and maintenance of solidarities, and that withdrawal of social

relationships is ordered by that structure in scarce times, as is the re-establishment of

those ties when times of plenty return.

The model was tested by respected anthropologists at the time it was proposed (cf.

Dirks 1978 with West Indian slave societies, Lomnitz 1978 in a Mexican shantytown,









and Turnbull 1978 among the Ik of Uganda). Today, it still provides a valuable

framework for understanding human adaptive cycles. While the Laughlin and Brady

volume may include out-of-date references to ecological theory, ecologists have been

able to build new theories out of the old ones, and anthropologists should do the same.

Political ecological approaches can be used to explain how, where, and when urban water

scarcity emerges in complex societies. Additionally, anthropological research that draws

on "new ecology" can help clarify how concepts like nested hierarchies, nonlinearity,

surprise, and transformation can be integrated into Laughlin and Brady's work (Abel and

Stepp 2003, Scoones 1999, Redman 2005). These concepts can help us explain when and

why social systems do not follow predicted patterns. Together, the Laughlin and Brady

model, political ecology, and ecological anthropology provide a framework for

understanding how urban water scarcity is created by human societies, and how it affects

social relationships within affected groups.

To test the relationship between urban water scarcity, sociability, and reciprocity, I

developed four hypotheses using information I received from three development agencies

working in Villa Israel in 2002. I was told that water scarcity was present in Villa Israel

throughout the year, and became so severe during the dry season that fourteen children

under the age of one died from water-related illnesses between April and July of 2002

(Trujillo 2002). Based on this information, I concluded that people probably experienced

moderate water scarcity during the wet season, and suffered from severe water scarcity

during the dry season months of May, June, July, and August. For this study, then, I

tested the following four hypotheses:









Hypothesis 1: Household heads participate in more social interactions during the

wet season than during the dry season.

Hypothesis 2: Household heads participate in more reciprocal economic exchanges

during the wet season than during the dry season.

Hypothesis 3: Household heads with more water participate in more social

interactions than those with less water.

Hypothesis 4: Household heads with more water participate in more reciprocal

economic exchanges than those with less water.

Chapter Conclusion

This study is designed to engage with interdisciplinary scholarship on two practical

problems in urban water scarcity: (1) small-scale human-environment interactions and (2)

changing survival strategies that affect the resilience of households, communities, and

cultural institutions. The study is also designed to answer one unexamined question in

anthropological theory: what are the effects of urban water scarcity on social interactions

and economic exchanges? To do so, I employ the Laughlin-Brady model of human

adaptation to resource scarcity-as updated by recent anthropological work in political

ecology and new ecology.

The study provides a test of the relationship between urban water scarcity,

sociability, and reciprocity in Villa Israel. It also contributes one data point to any future

cross-cultural study of the Laughlin-Brady model in water-scarce urban communities. It

uses ethnographic and reproducible scientific methods to test four hypotheses over 18

months of field research in Villa Israel. It also situates the findings in the wider

environmental, cultural, political, and economic realities of Bolivia, Latin America, and

the world.









Having examined the global trends that shape studies of urban water scarcity in this

chapter, I will now situate the study in social and environmental realities of Bolivia

(chapter 2) and Villa Israel (chapter 3). Then, in chapter 4, I discuss the study itself, and

the field methods that I used over 18 months in Villa Israel. In chapter 5, I examine the

political ecology of urban water scarcity in Villa Israel using data from the field research.

In chapter 6, I describe the social interactions and economic exchanges that took place in

Villa Israel using data from the field research. In chapter 7, I test the relationship between

urban water scarcity, social interactions, and economic exchanges in Villa Israel. Finally,

in chapter 8, I discuss how the findings relate to interdisciplinary theories of urban water

scarcity and anthropological theories of human belief and behavior.














CHAPTER 2
BOLIVIA: A CONCISE HISTORY

Introduction

When Laughlin and Brady predicted that responses to resource scarcity would be

similar across cultures, they cautioned that the shape those responses take could be

distinct in each ethnographic context (1978). In this study, I show how Laughlin and

Brady's theory can be applied to a water-scarce shantytown-an ethnographic context I

chose because people's circumstances there are similar to those found in squatter

settlements and shantytowns across cultures. While I believe Laughlin and Brady's

theories are generalizable, I recognize that historical events play an important role in

shaping culture, belief, and action. As a result, this chapter is designed to convey a

concise history of the relevant geographic, environmental, demographic, cultural,

economic, and political trends that shape life in Bolivian squatter settlements today.

I begin with a discussion of Bolivia's physical environment. This section provides

an orientation to the country's complex geography and ecology. Next, I discuss elements

of Bolivian history that still influence the country today-in particular, pre-Columbian

forms of social organization, Spanish and criollo dominance, the exploitation of natural

resources, and the rise of powerful workers' unions. Then, I examine the distribution of

culture, religion, and demography, with an eye to explaining how the country has been

transformed by modern historical events. After that, I discuss major changes in

economics and politics since the 1980s, examining the new climate of neo-liberalism,

indigenous activism, and political controversy in Bolivia. The unrest that began with









Cochabamba's 2000 Water War is now unfolding in a spectacular series of political

developments that may affect the entire region. Finally, I examine how the urban poor are

affected by the new neoliberal economy. This last section will lead directly into a

discussion of life among the urban poor in one of Cochabamba's most impoverished

squatter settlements, Villa Israel.

Geography, Environment, and Natural Resources

Located in the center of South America, Bolivia has been landlocked since it lost

access to the Pacific Ocean to Chile in 1884 (Kluck 1989). On the eastern edge of its

Amazonian region, Bolivia shares a long border with Brazil. To the southeast, the arid

Chaco scrublands straddle Bolivia and Paraguay. To the direct south, Bolivia borders

Argentina. To the southwest, Chile borders Bolivia along the Cordillera Occidental of the

Andes. To the northwest, Bolivia shares Lake Titicaca and its surroundings with Peru.

Running roughly from west to east, Bolivia has four major geographic regions: the

Altiplano, highland valleys, Amazon, and Chaco. The Altiplano region is a high Andean

plateau that runs along the western edge of the country. Plain-level altitudes range from

4,200 to 4,400 meters (Kluck 1989). Located in the Cerrada river basin, the Altiplano

receives between 100 and 500 mm of annual rainfall (Van Damme 2002, Mattos and

Crespo 2000, FAO 2005). While known for its inhospitable climate, the Altiplano has

always been Bolivia's most populous region, and is currently the home of 41 percent of

the population (Library of Congress 2005).

To the east of the Altiplano, the central highland valleys have a temperate semi-

tropical climate, moderate elevation, and fertile soil. They contain major cities such as

Cochabamba, Sucre, and Tarija, and 29 percent of the Bolivian population (Library of

Congress 2005). To the north and east of the highland valleys, Amazonian rainforests are









situated in the tropical lowlands. This area encompasses one of the world's most

biodiverse regions and 60 percent of Bolivia's landmass (Swaney 2001). Together, the

highland valleys and Amazonian rainforests are located in the Amazonas river basin,

which receives between 300 and 500 mm annual rainfall in the valleys and between 1700

and 2200 mm of rainfall annually in the Amazon.

In the far southeastern lowlands, the Chaco is an arid, sparsely-populated

scrubland. The Chaco (along with the southern tip of the highland valley region) is

located in the La Plata basin, and receives less than 100 mm of annual rainfall (Van

Damme 2002, Mattos and Crespo 2000, FAO 2005). Bolivia's combined lowland

regions-the Amazon and Chaco-contain 30 percent of the population, which is

growing rapidly due to rapid economic development in Santa Cruz (Library of Congress

2005).

Within Bolivia's four climactic regions, there are 16 large eco-regions and 198

ecosystems. At least 22,000 superior plant species and 2,342 vertebrate species have been

discovered in Bolivia, and new research constantly identifies more species (UNDP 2002,

Bojanic 2001). In addition to its biological wealth, the country has great mineral wealth.

Bolivia contains significant deposits of tin, gold, and zinc, as well as South America's

second-largest natural gas reserve and fifth-largest oil reserve (Library of Congress 2005,

EIA 2005). Because population and consumption have historically been low in Bolivia,

human impacts on natural resources have been relatively limited (Ibisch 1998). Even so,

recent large-scale land conversions have caused vegetation loss, erosion, and soil

salinization. As a result, desertification has developed in arid areas, as has flooding in









tropical areas (Van Damme 2002, Bojanic 2001, UNDP 2002, Paredes and Aguirre

2000).

Politically, Bolivia is divided into nine departments. The La Paz department

includes the capital city of La Paz, Lake Titicaca, the headwaters of the Rio Beni, the

north Altiplano, and the Amazonian highlands. To the south, Oruro is a small department

containing the famous San Jose mine and Lakes Poop6 and Uru Uru. Bolivia's

southeastern department is Potosi, which contains the salt flat Salar de Uyuni and the

once silver-rich mountain of Cerro Rico. The highland valley departments are

Cochabamba, Chuquisaca, and Tarija. The northernmost of the Amazonian lowland

departments is the Pando, which contains the Rio Beni. To its south is the Beni

department, which contains the Rios Mamore and Guapore. Finally, Santa Cruz is an

enormous department in the southeast corner of Bolivia that encompasses both rainforest

and scrubland, the Rio Paragua, and Laguna Concepci6n (Kluck 1989).

A Brief History of Bolivia's Social, Political, and Economic Organization

At one time, population in the area now known as Bolivia was clustered primarily

on the fertile edges of Lake Titicaca, in the western slopes of the Andes. Pre-Columbian

communities were structured around kinship-based groups called ayllus (Wagner 1989).

The ayllu was the unit through which natural resources, productive activities, and social

obligations were organized. The entire system, while not egalitarian, was based on

reciprocal exchanges called ayni and labor rotations called mit'a (Moseley 1992).

The first major regional civilization, the Tiwanaku, dominated the area from about

600 to 1200 AD. After a long period of regional power struggle and Aymara dominance,

the Quechua-controlled Inca empire gained control of highland Bolivia in 1450 (Library

of Congress 2005). Building on the basic Andean socio-political unit, the ayllu, the Incas









instituted successful development programs in agriculture, mining, and lowland

colonization (Moseley 1992). The colonization program prompted a major change in

Bolivia's population distribution, as large numbers of Quechua immigrants were sent to

the highland valleys to establish settlements loyal to the empire (Wagner 1989).

In the 1530s, Spanish conquistador Francisco Pizarro and his troops conquered and

dissolved the Incan territory. In 1544, a large vein of silver was discovered in Potosi, and

Bolivia became a major source of income for the Spanish Crown. Appropriating the mit'a

concept, Spanish colonizers required that indigenous highland men spend every sixth

year working in the mines. The Spanish also created the encomienda, a peonage-based

agricultural system in which the Crown gave colonists control over indigenous land and

laborers (Wagner 1989). Over the next century, millions of native people were

conscripted into mining and agricultural work, especially in Potosi and Oruro, creating

major cities and population growth in those areas.

In the 1700s, Spanish dominance went into decline as the easy-to-mine mineral

veins were tapped and organized indigenous resistance grew. A series of indigenous and

criollo revolts ensued from 1780 until independence was won in 1825 (Wagner 1989).

After independence, wealthy criollo landowners continued to control Bolivia's economic

assets, and peasants continued to labor in the peonage system. The next century of

Bolivian national politics was marred by inept leadership, corruption, failed wars,

territory losses, and economic mismanagement.

In 1952, mine workers led a revolt, the April Revolution, to reclaim control of the

government. The Nationalist Revolutionary Movement (MNR) enacted sweeping reforms

including the nationalization of mines, land redistribution, universal suffrage, and









expansion of education. The MNR also institutionalized miners' and peasants' unions,

which participated in decisions about natural resource distribution and use. This was the

first step taken in 400 years toward ceding control of Bolivia's natural resources to

indigenous groups. The miners and peasant unions established the organizational

groundwork for popular protest coalitions that remain powerful in Bolivia today.

The MNR controlled Bolivia until 1964, when a democratically-elected president

was overthrown in a military coup. Eighteen years of political instability followed, in

which the MNR and military coup leaders struggled for control of the country. Since

democratic elections were reinstituted in 1982, there have been no coups in Bolivia

despite serious internal economic and political instability.

Culture and Religion

Despite a relatively small population of about 9.0 million people (World Bank

2005), Bolivia has a very diverse cultural life. About 62 percent of Bolivia's population is

of native descent. Of these, 31 percent say they are Quechua and 25 percent say they are

Aymara (INE 2004). People of Quechua and Aymara origin traditionally occupied the

highlands, although a half-century of economic crises, political turmoil, and migration

has blurred such ethnic boundaries. Another 35 percent of the population is mestizo.

After that, much of the remaining population is of Chiquitano, Guarani, Mojefio, and

European descent, while 1 percent, living mainly in the La Paz department, is of African

descent.

The national religion of Bolivia is Roman Catholicism. Many sources claim that

Bolivia's Catholic population is very high-95 percent (cf. Library of Congress 2005,

CIA 2005). According to the 2001 national census, however, 78 percent of the Bolivian

population is Catholic, and 20 percent report being Protestant, evangelical, or belonging









to some other Christian sect (INE 2003a). Others estimate that only between 60 and 70

percent of Bolivia's population is Catholic (cf. Bureau of Inter-American Affairs 1998),

and by all accounts Protestantism is on the rise, as it is throughout Latin America,

especially in countries with large indigenous populations (Dow, in press). Whatever the

statistics, the Catholic Church still dominates much of the cultural life of Bolivians,

through numerous fiestas for saints or virgins, charity work in local communities, and its

role in weddings, baptisms, and funerals. Despite the dominance of the Catholic Church,

Bolivian Catholic beliefs, rituals, and symbolism are interwoven with Quechua and

Aymara ones. The most salient example of this is the fusion of the Virgin Mary and

Pachamama, the earth mother, in Bolivian Catholicism. Throughout Bolivia, people

practice Catholicism and indigenous religions concurrently, consulting both priests and

yatiri (curers) for help with different ailments.

Since the 1980s, Protestant missionary groups have become increasingly common

in Bolivia. In 1999, religious groups totaled 400, most of which were Protestant

missionary groups including Baptists, Pentecostals, Mennonites, Seventh-Day

Adventists, and various evangelical groups (Bureau for Democracy, Human Rights, and

Labor 1999). In general, all religions are tolerated, although discrimination against

religious minorities, particularly Protestants, persists. For instance, Bolivian Protestants

tend to be excluded from participation in community events and governance, often

because they refuse to participate in important social activities such as drinking alcohol,

dancing, and fiestas celebrating Catholic saints (cf. Buechler and Buechler 1971,

Goldstein 2003).









Recent Demography: Migration, Identity, and Urbanization

Since the agrarian reforms of the 1950s, temporary and permanent migrations have

become an increasingly common part of Bolivian lifestyles. Prior to the 1950s, many

peasants were tied to the land through the hacienda system of labor obligations to

landowners. After the hacienda system was abolished in 1953, families began to send

young men and women to labor seasonally in the lowlands, to mine in the highlands, and

to market year-round in city centers like Cochabamba (Kluck 1989). In the 1980s,

economic crises caused an enormous migration of highlanders to urban squatter

settlements and lowland colonization areas (McFarren 1992). After highlanders flooded

to the cities in the 1980s, urban settlement became dominated by rural-urban migrants in

the 1990s.

Over the last fifty years, highland migrants to the valleys and lowlands have

maintained close contact with the relatives they left behind in the Altiplano. Peasant

families establish exchange ties through migrating members with urban vendors and

producers of non-local goods all across Bolivia. Migrants, in turn, usually continue to

contribute to their home communities by dancing in homage to a virgin, hosting migrants,

and employing urban networks or Spanish skills to solve community legal or political

problems (cf. Buechler and Buechler 1971). In this way, multilevel ties between urban

and rural communities have been maintained, and can be used to sustain impoverished

urban and rural populations with infusions of goods, labor, and food during crises.

While highlanders often maintain relationships in their home communities, they

occupy an uncertain role in Bolivian society. As a group, they have come to be known as

cholos-a class of people who represented the integration of indigenous ancestry and

culture with European culture (Weismantel 2001). Cholos normally speak accented









Spanish and adopt some mestizo cultural practices, while maintaining other indigenous

beliefs and practices (such as language use, dress, food preparation, and the use of

indigenous medicine). Cholos often are often socially isolated; in towns they continue to

be disparaged as peasants, while in their villages, they are considered to be urbanites. In

the new context of citizenship movements emerging in Bolivia after 2000, cholos are left

exposed on the national political scene, lacking the claims to rights and resources

associated with both conventional urban power structures and new indigenous rights

groups (Jackson and Warren 2005).

Today, 62 percent of the Bolivian population is urban, and urbanization continues

to rise at a rate of 3.6 percent per year (Library of Congress 2005). Bolivia's population

is becoming increasingly concentrated in four cities: La Paz, El Alto, Santa Cruz, and

Cochabamba. In the 2001 census, 77 of Bolivia's urban population resided in these four

areas (INE 2002). As I explain in the next section, profound changes in Bolivia's political

economy during the 1980s and 1990s led large numbers of people to leave their homes in

the Altiplano to seek new lives in the country's four major urban areas (Mattos and

Crespo 2000).

Recent Economic History

In 1984, a devastating cycle of debt, trade deficits, and hyperinflation that crippled

Bolivia's banks and markets led the government to enact a sweeping IMF-sponsored

structural adjustment program (Nash 1992). Despite extensive grassroots opposition,

outsiders considered the plan to be an astounding success when the Bolivian economy

began to function again nearly immediately after its enactment (McFarren 1992). Then, in

1994, a new reform called "capitalization" privatized 50 percent of Bolivia's saleable

national industries-oil and gas (YPFB), electricity (ELFEC), telecommunications









(ENTEL), airlines (LAB), and trains (ENFE) (Kohl 2003). Many Bolivians believed,

based on the Bolivian government's claims at the time, that the capitalization program

would cause foreign companies to bring enormous infusions of capital to create jobs,

increase economic growth, and help struggling Bolivian industries compete

internationally (Kohl 2002).

While privatization and capitalization did improve the performance of Bolivia's

economy and national industries, it was a bitter disappointment for many Bolivians-

particularly in highland mining communities. The privatization and closure of national

mines in 1986 disassembled the country's core industry, dispersing 27,000 ex-miners

mainly into urban squatter settlements and the lowland coca-growing regions (McFarren

1992, Nash 1992), while capitalization did not create new jobs for displaced workers. In

fact, many key industries suffered a net loss in jobs after capitalization, as companies

imported materials and even workers rather than developing Bolivians' capacity to

provide the necessary materials domestically (Kohl 2002). In addition, the Bolivian

government lost hundreds of millions of dollars in revenues after the privatization of the

YPFB (Kohl 2002), which led to cut-backs in jobs and social programs through the

country. As a result, many Bolivians lost a substantial portion of their social safety net in

the 1980s and 1990s, including public services, income security, and market regulation of

prices for essentials.

The economic and social disruptions of the 1980s undermined grassroots activism

in Bolivia (Whitehead 2001). The 1985 privatization and closure of national mines, in

particular, disassembled Bolivia's core activist networks, dispersing ex-miners into new

regions and professions. The dispersal of miners' unions was widely considered a major









step backward in indigenous people's fight to regain control of Bolivia's natural

resources. A political reform enacted alongside the Capitalization Law, however, revived

the old activist networks, and enabled grassroots groups to reinvent themselves as new

social movements.

Recent Political History

In 1994, the Popular Participation Law (LPP) was passed. The LPP was designed to

make Bolivian democracy more representative and redistribute national funds and

responsibilities at the local level. As part of the reorganization of local-level Bolivian

politics, the LPP created Grassroots Territorial Organizations (GTOs) based on

recognized civil society organizations such as pre-colonial indigenous groups (like

ayllus), urban neighborhood councils, and rural campesino unions (Kohl 2003). The

decentralization and indigenous rights reforms passed in 1994 gave new political

resources, access, and legitimacy to territorial grassroots organizations (Kohl 2003, Van

Cott 2000, Whitehead 2001). The compulsory development of GTOs unified people on

the local level, educated them in political procedures, and linked them to sponsoring

political parties (cf. Medeiros 1990), and helped them gain democratic representation in

the national parliament for the first time (Van Cott 2000). The LPP, then, created a

political opportunity for disenfranchised groups, like indigenous, urban poor, and

campesino groups, to seize control of tiny pieces of municipal budgets-and to work

toward larger goals.

Now, after the austerity, capitalization, and decentralization reforms, five major

regional political interests can be identified in Bolivia. First, in La Paz, the Aymara

heartland, Aymara separatists agitate for territorial sovereignty. A second regional

interest encompasses all of the western highlands (La Paz, Oruro, Potosi, and









Cochabamba departments). There, politically active urban poor, campesinos, and

indigenous peoples demand various economic and political reforms, and exert influence

through highly visible protest tactics like regional blockades. Third, in the Yungas and

Chapare, coca growers constantly find new ways to circumvent U.S.-backed coca

eradication programs. Chapare is now the stronghold of the Movement Toward Socialism

(MAS) party. With the support of coca growers and other disenfranchised political

interests, MAS leader Evo Morales has risen to remarkable fame and influence in Bolivia

and abroad-and was elected Bolivia's first indigenous president in December 2005.

Fourth, in the Amazon, eastern indigenous groups spearheaded the March for Life and

the drive for expanded indigenous rights in the 1990s. Fifth, lowlanders based mainly in

Santa Cruz and known as Cambas (as opposed to highlanders, or Kollas) are Bolivia's

final major interest group. Camba-dominated Santa Cruz area has its own culture, in

which residents are generally mestizo, speak a distinct Spanish dialect, and reject

highland values and beliefs. It also has political and economic interests and goals quite

distinct from those of highland and indigenous activists.

Bolivian Economy and Politics After 2000

The rivalry between Cambas and Kollas is arguably the most important political

division in Bolivia today. After the privatization of the national mines in the 1980s, many

highlanders settled in eastern lowland colonies (Kluck 1989). Since then, major natural

gas discoveries and Bolivia's fastest economic growth have been located squarely in

Camba territory. While Santa Cruz prospers, the highlands have become increasingly

impoverished and politically unstable. In the 2001 census, Santa Cruz was the only

department in all of Bolivia in which less than half of its residents (38 percent) were

impoverished. Potosi, once the heart of the Spain's colonial mining empire, had become









Bolivia's poorest department, with 80 percent of its residents living in poverty (INE

2003b).

Many Cambas have come to see the Kolla regions as an economic, political, and

cultural drag on their success, and have initiated a growing movement for regional

autonomy. Kollas, outraged by the racist undertones in Camba political rhetoric, retort

that Santa Cruz's success has been built on Kolla labor and are fighting to nationalize

lowland gas reserves. At the core of the dispute are two distinct visions of Bolivia's past

and future. Cambas, encouraged by financial successes in the coca and natural gas

sectors, argue that Bolivia's future is in neoliberal economic reform and participation in

the globalized economy. Kollas, drawing on the historical memory of centuries of mining

booms and market collapses, argue that only a welfare state can protect its citizens from

the risks of having a national economy based on the export of raw materials and cash

crops.

In the late 1990s, public opinion began to turn against capitalization and other

neoliberal reforms (Kohl 2002)-particularly among highlanders who had suffered the

worst of mining reforms in the 1980s. After the famous Cochabamba Water War in 2000,

a series of increasingly violent protests broke out across the country, mainly over the

privatization of water and gas resources (Schultz 2005). The protests have been massive

and draw from nearly all sectors and interests in highland society. To date, the protesters

have succeeded in nullifying Bolivia's business deals with foreign investors, forcing the

resignation of two presidents, and calling for a new constituent assembly to rewrite

Bolivia's constitution. After Evo Morales' accession to the Presidency in January 2006,

he appointed a government cabinet drawn from the leaders of Bolivia's major social









movement. The new government will surely work to secure re-nationalization of natural

resources, the reinstitution of a welfare state, and more formal political power for

indigenous people. Even so, powerful interests are still at work to preserve neoliberal

reforms and growth-oriented policies. In the fight to define Bolivia's political and

economic future, there is no clear winner-yet.

Poverty, Well-being, and Livelihoods in Contemporary Bolivia

While the Bolivian protests are exciting to watch and theorize about, the grinding

misery of everyday poverty affects far more people-and is the topic of this study. The

great outcry over economic reforms in Bolivia today might seem to imply that standards

of living are have fallen sharply in the country over the last decade. Indeed, many

scholars have shown that-across Latin America-austerity packages, structural

adjustment, and neoliberal reforms were followed by the absolute breakdown of

livelihoods and cooperative survival strategies among the urban poor (cf Safa 2004,

Moser 1996, Gonzalez de la Rocha 1995). By some measures, however, quite the

opposite appears to be true, at least for Bolivia.

In Bolivia's 1976 census, 86 percent of the Bolivian population was considered to

be impoverished-that is, they had one or more of five basic needs unsatisfied (housing

materials, housing space, water and sanitation, health, or education). By 1992, the

number of impoverished Bolivians had dropped to 71 percent. In the most recent national

census, in 2001, 59 percent of Bolivians, or 4.7 million people, were classified as

impoverished using the same measure (INE 2003b). Looking at Bolivia's impoverished

people in four ranked groups, the two poorest groups shrank between 1992 and 2001,

while the two moderately poor groups grew in size (INE 2003b). Non-governmental

organizations and development groups are clearly making progress toward meeting the









infrastructural, health, and educational needs of Bolivia's poorest citizens (cf. World

Bank 2005).

There is, then an apparent paradox between successes in measures of development

and Bolivians' growing intolerance for the direction of the economic reforms. The

paradox is easily reconciled, however: at the same time that material conditions are

improving, livelihoods are dwindling. As of 2003, the fraction of jobs in industry (29.1

percent) and agriculture (14.4 percent) was shrinking. The only sector of the Bolivian

economy that was growing was services, at 56.5 percent of the GDP (World Bank 2005).

Thus, the result of economic downturns and neoliberal restructuring is job growth in the

mostly-informal service sector, where benefits are few, underemployment is common,

and many are self-employed (Library of Congress 2005). Unemployment was estimated

at 8 to 10 percent of the Bolivian workforce in 2002, and self-employment at 65 percent

(Library of Congress 2005). Infrastructure, education, and health care are basic

necessities and are improving in Bolivia, but these advances do not stop the everyday

need to buy food, water, and other essential market goods.

With an economy that is increasingly market-based and deregulated, people's

survival is tied to their ability to raise cash income. In urban areas, where people have

little access to subsistence goods to offset poverty, 39 percent of the population is

impoverished (INE 2004). People living in Bolivia's squatter settlements tend to make a

living providing services in the informal market, as vendors, maids, day laborers, and taxi

drivers. With less disposable income circulating in the Bolivian economy, the middle

classes have cut back their spending on the non-essential goods and services that many of

the poorest urbanites provide (Gill 1994). In addition to the ways that their jobs are









affected, squatters are more vulnerable than are other impoverished urbanites to changes

in resource distribution and infrastructural development (Eckstein 1990). For instance,

squatters live far from the city centers, and rely on long-distance transport to get to work.

Sharp rises in gas and transportation costs can make it difficult or impossible for them to

get to work, school, or wholesale markets in the city center.

In the next chapter, I discuss how people in Villa Israel have come to live and

survive in one Cochabamba shantytown.














CHAPTER 3
SETTING

Finding the Field Site

In the summer of 2002, I made an exploratory trip to Bolivia to talk to development

specialists and hydrological engineers about the possibility of conducting water scarcity

research there. Two hydrologic engineers from a water-delivery NGO called Water for

People, Susana Southerwood and Abraham Aruquipa, expressed a strong interest in the

research project I proposed to them. Susana suggested that I consider working in a 20-

year old Cochabamban squatter settlement called Villa Israel. She explained that,

although Water for People had been involved with Villa Israel for almost ten years, the

community's complex social composition and tendency toward conflict had made it

difficult to get things done. Susana said that she would be interested in any insight I could

add into what was happening, socially, politically, and economically in the community.

I was intrigued by Susana's description of the community, and decided to go see it

for myself. I was staying in a hotel in downtown Cochabamba, and had no idea where

Villa Israel was or how one could get there. The community did not appear on any

official map of Cochabamba. Not knowing what else to do, I found a taxi driver who had

heard of the place and asked him to take me out there. I was nervous about what I would

find-especially after hearing Susana's description of the conflictive community. All of

the ethnographies I had read about Andean communities rejecting and even attacking

foreign researchers flashed through my mind during the long taxi ride into and beyond

the impoverished south side of the city.









Finally, the taxi driver pulled off the highway, careened into and back over the top

of a ditch, and announced that we had arrived. As we clattered into the community, I

looked out over a vast dry expanse of land cluttered with dusty houses. There were no

trees, no plants, and no water in sight. There was only a handful of people mulling

about-mostly highland indigenous women known as cholitas, dressed in petticoats and

full knee-length skirts calledpolleras, lacy white tops, and long braided black hair. There

was no possible pretext for me to be there-no businesses, restaurants, no foreigners of

any sort whatsoever. I asked the taxi driver if there was any bus or taxi line that serviced

the community. He said that there was a bus that came around every once in a while, but

there was not one vehicle in sight. Realizing that to present myself to the community as a

clueless intruder looking for a way out was probably not the most auspicious entree into a

field site, I told the taxista that I would pay double-an exorbitant amount of money-for

him to take me back to the city.

That summer, I took refuge in the safe havens of government offices and NGO

headquarters, learning all I could about Cochabamba's water problems and what was

being done to solve them. After meeting with about ten different development

organizations, I felt confident that I had found the right city for my research. I knew,

however, that the success of the project rode not on institutional support, but on the

acceptance and collaboration of the community members themselves.

Gaining Entry

On my second trip to Bolivia, in the summer of 2003, I knew I needed a plan to get

started in the community. Once again, Susana proposed a solution. She was supervising

work on Villa Israel's water system, and invited me to go along with her and a friend

while they checked on the workers' progress. While we were there, she introduced me to









several well-known men in the community. As we walked around, we ran into a large

group of women who were building a cement-floored sports court. The women asked us

to help them by paying for 20 bags of cement, and Susana coordinated the emptying of

our pockets. She then elbowed me along with the women, saying-this is your chance,

now go with them and get to work.

I went with two of the women to buy the cement, and they asked me to stay for

lunch. After Susana and her friend had left, I was invited to a feast of steak, salad,

potatoes, and Coca-cola. Eating alone, I was embarrassed to be treated to such a delicious

meal while 35 women stood around watching me. I ate slowly and talked endlessly,

answering every question I could draw out of the shy women about myself, my family,

my hometown-anything to keep them smiling. When the meal finally ended, I asked the

women if I could come back the next day and help them with their work, since I had

some free time and was there to learn more about their lives.

The women tried to discourage me, saying that I could just sit in the shade of the

courtyard's single tree if I wanted to see what they were doing. They confided that their

work consisted of mining large boulders from an area that also served as an informal

latrine. A few women argued that the work was far too nasty to do voluntarily. Of course,

any qualms I felt about working in an open sewer were nothing compared to the prospect

of trying to start participant-observation elsewhere in the seemingly-deserted community.

I felt very fortunate indeed to be welcomed so easily into Villa Israel.

The work was brutal, as we spent weeks in pairs, lugging boulders, and mixing,

shoveling, and spreading cement. A tenuous camaraderie developed during breaks in

which the women complained, told jokes, or sang with me. I slowly realized that there









were two subgroups among the women, and people were watching closely to see which

one I favored. If I carried rocks or shared lunch with one group too many times in a row,

I could be sure that the others would snub me. At the time, I did not realize how rare it

was to see Villa Israel residents' joys and hostilities out in the open.

After a few months of participant-observation, I found my footing in Villa Israel. I

used my role as informal mascot of the women's work group to gain wider acceptance in

the community. I slowly advanced my research program, moving from participant-

observation to surveys to direct observation, and I learned to accept that Villa Israel

residents' regard for me would always tilt between affection and resentment. As I became

a confident ethnographer, I learned how to explain to curious new visitors what to expect

on their first trip out.

Landscape

A trip to Villa Israel normally starts in the city of Cochabamba. There are only two

transport lines that serve the community, the number 35 bus line and the 101 taxi-trufi

line. The number 35 bus line is a large line that runs the north-south length of

Cochabamba, from high in the foothills of Tunari Park, snaking through downtown and

the city's enormous marketplace, and ending at the outside edge of Villa Israel. The bus

is slow to load and progress, comes about every 20 to 30 minutes, and runs only from

about 6:00 am to 10:30 pm. The 101 taxi-trufi line also originates in Villa Israel,

following the 35 route until it hits the Prado (main street) uptown, and then turns left to

end in a western barrio called Villa Mexico. It features a set of speedy taxis designed to

hold five people, but can pack in up to twelve people when some are shoved into the

hatchback. The 101 line runs at 5 to 15 minute intervals nearly 24 hours-although few

drivers want to head out of the relative safety of the downtown area between 9 pm and 5









am. Although it is slow and uncomfortable, the 35 bus is the choice of most Villa Israel

residents because (at the time of research) it costs 1 boliviano ($ 0.12), while the taxi-

trufi costs 1.50 bolivianos ($ 0.19).

On their way to Villa Israel, the 35 and 101 reach the southern edge of the market

and turn onto the Pan-American Highway. From this jumping-point at the edge of the city

proper, the ride takes about 20 minutes in taxi-trufi and 45 minutes in bus. During the

drive south, highly-populated barrios with internet cafes and restaurants gradually give

way to patches of unused land, rustic corn-beer bars known as chicherias, and adobe

housing. About halfway through the ride, one must pass a large slaughterhouse, and a

putrid stench hangs over the area and wafts into the passing vehicles. Finally, just as the

settlements appear to end completely, the bus passes a spiffy brick health center and one

last Catholic church. With a sharp turn right, the bus heads though a field and lurches

down into a usually-dry riverbed (the one I thought was an inconveniently-placed ditch).

Pulling up out of the riverbed, a Villa Israel sign appears over the road's entrance to the

community. Finally, the bus makes a wide left, and heads up Villa Israel's main

thoroughfare.

The view of the community that the bus affords is unspectacular. Looking out over

the landscape, a new visitor immediately takes in the overwhelming grayness of the

terrain. A few squat hills, dotted with scrubby trees, climb up behind the community.

Dust and litter blow along the curbs. The dry riverbed is dotted with feces, and clouds of

filthy sand occasionally gust over the bank's edges, engulfing passersby. Skinny dogs

dominate the streets, running in packs whose complex social organization appears to rival

that of the community itself. Most houses are walled in, and every street-side is lined









with an uneven barricade of thorns, chicken wire, adobe, concrete topped with glass

shards, metal fencing, and brick.

Later, one learns to pick out the community's landmarks in the dull and

unwelcoming landscape. At the head of the main street sit the community's two most

imposing structures: the Christian Evangelical Union (UCE) church and the community

land prospector's three-story house. The bus stop (just behind the UCE church) and the

taxi stop (at the side of the riverbed) are the hubs of Villa Israel's economy; they provide

work to many men and a few intrepid women, the community's only reliable vendors of

prepared food, and its lifeline to food, work, and education in the city. The open fields are

actually the heart of Villa Israel's social life, and come alive on Sundays when teams-

ranging from uniformed city-wide competitors topollera-swinging women's teams to

groups of industrious children-assemble to face off in the community's five soccer

fields, basketball court, and volleyball court. In time, one might even discover that the

patches of green trees tucked into the hills' crevices indicate where Villa Israel's two

wells suck at tiny underground pools of water.

The Cadence of Community Life

Daily life in Villa Israel starts at about 4 am. Before dawn, wives start to cook or

clean, prepare for work, and boil the morning tea. A bit later, husbands get up, and

quickly eat breakfast and get ready for work before rushing off between 5 and 6:30 am.

From about 4:30 am, the community water guard starts his morning rounds, methodically

blowing his whistle, opening a pump, and looking on as women, strong children, and a

few men fill up their families' buckets and gasoline cans with 20 to 40 liters of water.

The guard continues his rounds, closing each pump and padlocking its door in turn, until

9 am. Between 7 and 8 am there is a frantic rush all around the community, as women









line up to get to the cancha and send their children running out of front doors to buy

bread or make it to school on time.

After the frenetic early morning, life in Villa Israel drags to a slow pace by mid-

morning. Those who have not left the community to work stay close to home, doing

morning chores or attending their cottage industries. By 10 am, mothers or older sisters

must begin a 2-hour preparation process to make a typical stew. Laboriously soaking,

washing, and rinsing ingredients using just one or two buckets, some women work all

morning to ready the midday meal. Few people go out, and the streets are empty and

quiet.

At noon, there is a brief burst of energy when the children are released from school,

and mothers or older sisters serve up lunch for the family. From 3 to 6 pm, those who can

snatch a bit of precious leisure time to rest or nap. Everyone else keeps busy with errands

and housework. At 6 pm, there is another rush to the comer stores, this time to buy meat

or bread for the evening dinner. From 6 to 8 pm, children go out to play in the streets and

soccer fields. From 8 to 9 pm, wage-earning women and men begin to return home, and

dinner is served. At nighttime, from about 8 to 11 pm, families spend time together,

usually gathered around an old television. By 11 pm, the community is quiet and

everyone is at home and sleeping-except members of Villa Israel's two gangs, relatively

tame groups of young people who wander the riverbed or hang out in the soccer fields

until about 2 am.

Villa Israel's social offerings are meager during the week, as community members

focus almost exclusively on families and work. A few times during the week, people

gather in the evening to go to church or attend a community meeting. Only on Sundays









does the community come out from their walled-in houses to socialize. By late Sunday

morning, the soccer fields, community market, churches, and food stands fill up with

people enjoying their only day off from work.

Origins

To outsiders, Villa Israel leaders try to portray the community as a group of humble

evangelical Christians who came to Cochabamba from the highlands in search of a better

life. One visit to a meeting of the conflictive Junta Vecinal (Neighborhood Council) is

enough for any visitor to recognize that the real story is more complicated. The

community was indeed settled, nearly 20 years ago, by an evangelical Christian land

prospector hailing from Huanuni, Oruro-an impoverished mining district perched high

in Bolivia's Andean altiplano. Recruiting people to pioneer his squatter settlement in the

usual way, through networks of relations and contacts, the prospector found family

members, evangelicals, and Huanuni residents.

Quickly, however, word about Cochabamba's latest barrio got out, and all kinds of

people came to buy, rent, and squat in the community. People arrived from

Cochabamba's other barrios, the rainforests, the mines and the countryside, and as far

away as Argentina. In 2000, about 1712 people were estimated to live in Villa Israel

(Claure Periera & Asociados 2001). By the time that my research team and I conducted

our first random survey of 101 community residents, in April 2004, Villa Israel had

become a Bolivian melting pot. Today, the population has a wide range of origins,

experiences, identities, and beliefs-which makes it an interesting, complex, and

sometimes problematic place.

The majority of Villa Israel's adult population (92 percent) was born in the

impoverished highland departments of Potosi, Oruro, and La Paz. A few were born in the









tropical coca-growing regions of Cochabamba's Chapare (1 percent) or the La Paz

rainforests (2 percent). Another 3 percent of adult residents came from Santa Cruz, the

lowland region whose residents are despised by highland Bolivians. About 75 percent of

adult residents were born in the countryside, in the departments of Potosi (26 percent),

Cochabamba (20 percent), Oruro (18 percent), and La Paz (10 percent). The rest were

born in highland mining communities (13 percent) or in the cities (10 percent).

Community Development

While many Villa Israel residents were born in farms or small pueblos in the

altiplano countryside, not all come directly from their hometowns to Villa Israel. Those

who did tend to see their move as evidence of upward mobility. Some told us that, living

in Villa Israel, they acquired steady cash incomes, proximity to schools and health care,

and access to cosmopolitan diversions for the first time. Although they recognize the

challenges of life in Cochabamba's far-flung shantytowns, 49 percent of those polled said

life in Villa Israel compares favorably to the poverty and sadness of life in the

countryside. Those who came to Villa Israel after living in an urban downtown, the cash

bonanza of the coca-growing regions, or the security of the pre-1985 mining

communities, tend to see life in Villa Israel as harsh, intolerable, or humiliating. Of those

polled, 27 percent said that life was worse in Villa Israel than where they lived before,

while another 25 percent said it was the same.

Life in Villa Israel, of course, has improved enormously since the first settlers

came-when there were no transportation, communication, or social services at all. With

funds from aid agencies, the government, and community members, residents built

evangelical churches, an elementary school, a water distribution system, water runoff

channels, a foot bridge, a rock-paved street and marketplace, street curbs, a concrete









sports court, and a health clinic (which they share with the neighboring community,

Ch'aqui Mayu). The community government consists of four committees-

neighborhood, market, water, and school-and is charged with maintaining old projects

and initiating new ones. The community has four major social programs: Compassion (an

evangelical enrichment program for pre-school and school children), PLANE (a

government-funded program for women's employment and community construction), the

Mother's Club (a group of mothers dedicated to learning new skills), and continuing

learning classes for literacy, handicrafts, and English. The evangelical churches also have

informal charities, in which they mobilize parishioners to donate, intervene in family

problems, give advice, cook, and pray for church members.

Although the community has initiated an impressive array of projects over its brief

history, many of them were corruption-ridden, poorly executed, or both. In one infamous

incident, leaders touted a scheme to buy land in a nearby community and pump its water

to Villa Israel, and then ended up "losing" community funds totaling at least $10,000. In

the face of political debacles such as that one, interest in cooperative ventures has

plummeted. In 2004, four of the eight community groups-the water committee, PLANE,

the Mother's Club, and the continuing learning classes-were defunct. When people do

get together to work on a project, it rarely goes well. For example, the street curbs and

runoff channels began to crumble just months after they were built. A cycle of official

corruption and popular apathy undermines volunteerism and cooperation in the

community. As one woman-a savvy, long-time resident of Villa Israel-explained, "We

gave our trust and money to the leaders, and we can't get them back. Now we joke that,









around here, only the fools still attend meetings and give their money to the

Neighborhood Council."

In spite of the community's problems, there are those who still believe in

participation and progress. Many people would like to see Villa Israel look and act more

like the barrios in downtown Cochabamba. A major rift has emerged between those who

want the community to invest in paved roads, urban sanitation, and water delivery

infrastructure, and those who believe that people should leave the community as it is. The

rift is deepened by classism, racism, and real economic differences among community

members. Those who want modernization and progress bitterly blame the peasants for

their lack of education and disinterest in achieving a respectable middle class lifestyle.

Those who oppose community projects argue that one would have to be blind to the

political and economic realities of life in Villa Israel to believe that mobility to the

middle classes is still within reach.

Economic Well-being

Indeed, many of the residents of Villa Israel are in an economic situation that is

obviously dire. Many housewives have to feed their families on 5 bolivianos ($0.62) a

day. Those in the worst conditions manage to live on 2 bolivianos ($0.25) a day-getting

by with bread, tea, and bone soup. Despite the abject poverty in which many Villa Israel

families live, the community has remarkable economic diversity. Some residents own

three or four lots of land, and live off the income from their stores, taxis, or rental houses.

Others own tidy brick houses of two and three stories, with enormous private water tanks

that insulate them from the local shortages.

One of the most important elements of economic well-being in Villa Israel is home

ownership. When economic crises hit, home owners are able to take in renters or working









family members, intensify cottage industries, or leave the community temporarily to

work elsewhere (Moser 1996). Renters, however, are restricted in their use of housing

facilities as capital, and risk ending up dispossessed if they miss a rent payment or leave

their belongings unattended. About half of Villa Israel residents own their homes (53

percent). About another third (29 percent), share, borrow, or house-sit someone else's

house. After that, 18 percent rents a house or a room. Villa Israel residents, then, are split

almost evenly between high and low levels of housing security.

Economic security, of course, is more complex than housing security alone. A 2001

survey of the community estimated that the average annual household income in Villa

Israel was 5222 Bs, or about $652.75 per year (Claure Periera & Asociados 2001). While

this figure supports what little data I have household income, it does not tell us much

about the range of economic well-being Villa Israel. Since Claure et al. did not explicate

the methods they used to calculate average annual income, I was unable to replicate their

work. In my own study, I found that it was quite difficult to get good estimates of

household income because earnings varied daily, various people contributed to the

household economy, and people were very reticent about income levels.

After 14 months of research in the community, I devised a 4-point scale to assess

economic well-being at the household level. I used four criteria to rank households: (1)

access to food, water and medicine, (2) steadiness of work, (3) ownership of housing, and

(4) access to capital such as a taxi, store, or inventory. The research team used the scale

to rate 75 randomly-selected households, after conducting interviews and observation in

each household over an 11-month period. We found that the distribution of wealth in

Villa Israel was fairly normal. Households in the category we called "poorest"-those









that regularly lacked food, water, or crucial medicines-were 11 percent of the sample.

Those households we called "low-stable"-in which at least one member had steady

work, but they still struggled to maintain access to basic necessities-were 25 percent of

the sample. "High-stable" households-those with steady work, secure access to basic

necessities, and a house or capital-were 52 percent of the sample. The households we

called "wealthiest"-those with steady income, at least one house, and capital-were 12

percent of the sample.

Economy and Education

The most common sources of income in Villa Israel are the transport lines and the

Cochabamba marketplace. Poor young men are hired by taxi owners to drive in three-

hour shifts on the 101 line. Middle-income men drive their own taxis, while the

wealthiest families hire drivers to drive their vehicles for them. At the end points of the

101 and 35 lines, women vie to sell breakfast, brunch, lunch, and late-night snacks. A few

lucky workers are employed by the lines themselves to punch timecards and control the

flow of vehicles.

In Cochabamba's downtown, the marketplace is a world apart. An enormous multi-

block section of southern Cochabamba dedicated to commerce, the cancha (a word

meaning field, now used to denote the market for Cochabamba residents) teems with

thousands of precariously-rigged stalls, blanket-displays, and wandering vendors. The

competition is fierce, and there are often 50 or more stalls lined up in a row offering

exactly the same goods at the same prices. Villa Israel's vendors usually participate in the

lowest rung of market prestige, selling fruit or vegetables from blankets and circulating

wheelbarrows, away from the rows of legally-installed stalls. Profits are minimal

(between 5 and 20 bolivianos a day, or $0.63 to $2.50), but so is the capital required-









just enough to buy a day's worth of produce at the peasant market and the blanket or

wheelbarrow to display it.

Those who do not work in transportation or the market tend to work in

construction, as maids, or in household industries like leatherworking, tailoring, weaving,

cooking, or raising animals. Extremely few Villa Israel residents have college educations

or work in white-collar jobs. In fact, I knew of only two people with college educations,

both teachers, living in Villa Israel, and both moved out before the end of the study.

Despite the low level of education among Villa Israel residents, most adults value

education highly-even if in practice they lack the time and resources to facilitate their

children's educational success. In Bolivia's current economic climate, however, a college

education does not provide much insurance against unemployment and poverty. The

Cochabamba job market is increasingly flooded with underemployed young

professionals. New graduates from unconnected families-those who lack relatives and

godparents to facilitate their access to government and business jobs-have great

difficulty navigating the job market. When Villa Israel's first generation of college

students graduates, in about seven years, most will be have to contend with this

disadvantage when they enter the job market.

Households and the Division of Labor

The household is the heart of resistance against poverty and privation. In poor

communities, households are generally organized to maximize the ratio of production to

consumption (Gonzalez de la Rocha and Gantt 1994, Gonzalez de la Rocha 1995). For

this reason, household composition tends to change when times get bad. Married children

join their parents in joint households with multiple adult income earners. Children are

sent out to work. Some mothers remain in Villa Israel while the fathers support the family









from afar, while other mothers prefer to support their children alone without involving

the father.

In Villa Israel, the majority of households (55 percent) consist of a mother, father,

and underage children. A quarter (25 percent) of households is of the joint type,

containing multiple adult income earners with or without children. Just over a tenth (11

percent) of the households consists of mothers raising children alone. A few more

households (6 percent) consist of one or both grandparents raising grandchildren. Finally,

in the original study sample, there were two households with a single adult living alone

and a household of only underage children.

Nearly all adults (87 percent) work in the cash economy. Most Villa Israel fathers

are rarely at home, and work full-time in the city, cancha, or 101 taxi-trufi line. All single

mothers and most married ones participate in income-generating activities as well. Some

work full-time, in PLANE construction projects, as maids in the city, or selling produce

in the cancha. Others work part-time or seasonally taking in washing, cooking food,

knitting, weaving, or selling produce at the market. Women stay home from one to six

days a week, to concentrate on housework, cooking, and taking care of children. Men

who work full-time usually have Sundays off, and are often out-playing soccer,

attending meetings, or visiting-on their day off

In households with working mothers, responsibility for housekeeping and childcare

normally falls to the eldest sister, or occasionally an eldest brother. These teenagers are

expected to buy food, fetch water, cook, clean, and take care of their siblings. Some also

have part-time jobs, and most are enrolled in school. Many teenage household heads

marry early or run away to escape the responsibilities. Those who stay at home explain









that they cannot abandon their younger brothers and sisters. Younger siblings, too, have

household responsibilities; they make runs to the corner store, fetch water or hunt down

the water delivery truck, and usually pitch in with some of the childcare, cooking, and

cleaning. Some children also work from the age of 8, selling ice cream in the community

or plastic bags in the cancha.

The general patterns of labor division described above hold across most Villa Israel

household types, except one. In two-parent households with home industries-corner

stores, leather workshops, a chicheria, or a bakery-the division of labor tends to be

much more flexible. Mothers and fathers rotate household responsibilities, working

together in the business, housework, and childrearing. The children are normally less

burdened with housework, and have more time to pursue their studies. Such couples also

tend to have a rapport that is visibly better than that of other couples in the community.


Language and Gender

Like most Bolivians, nearly all Villa Israel residents are of indigenous descent, and

most speak an indigenous language. In addition to Spanish, Bolivia has two major

indigenous languages: Quechua and Aymara. Quechua is the more common of the two,

and is spoken mainly in the departments of Potosi, Cochabamba, and parts of Oruro.

Aymara is spoken mainly La Paz and parts of Oruro. Although most Villa Israel residents

speak a second and even third language, 90 percent speaks Spanish proficiently. Of the

10 percent who do not speak Spanish proficiently, 9 percent speaks Quechua, and 1

percent speaks Aymara. All of the people who speak indigenous languages exclusively

are women.









Villa Israel Spanish is highly accented, and employs a number of Quechua words.

Language can be a real source of discrimination for Villa Israel residents outside of the

community, where native Spanish speakers often look down on people who speak with

indigenous accents. Most children, however, speak unaccented Spanish, as well as fluent

Quechua or Aymara. Villa Israel parents place a high premium on having their children

succeed economically in urban Cochabamba. As a result, many actively encourage their

children to integrate in mainstream mestizo society, and tend to avoid inculcating them in

indigenous values and practices. The most obvious example of this is that, in Villa Israel,

not one daughter is being raised de pollera (in indigenous dress); all are de vestido (in

Western pants and dresses).

Although language and dress can be a source of grief for Villa Israel residents

outside of the community, cultural differences do not cause many problems at the

community level. Villa Israel's important power players are all men who, like most Villa

Israel men, speak fluent Spanish and dress in Western clothes. Community meetings are

conducted bilingually in Spanish and Quechua, as there are few, if any, monolingual

Aymara speakers responsible for representing their households. In community

organizations, language plays little role in excluding household representatives (usually

men) from participation and decision-making.

Among women, language-related transgressions can be a source of rancor. Women

working in community work groups or fetching water at public water taps tend to clump

together according to their native language, talking and coordinating among themselves

in Aymara or Quechua. When they feel left out, Quechuas accuse Aymaras of gossiping,

backstabbing, and other offenses, and vice-versa. And indeed, the language-segregated









groups do engage in such behavior. Despite the resentment that these situations create,

many women adapt to the linguistic barriers, moving adeptly between Aymara and

Quechua groups. For instance, one woman, who came to Villa Israel as a monolingual

Spanish speaker ten years ago, told me that she quickly learned to speak Quechua and

Aymara out of necessity. Linguistic and cultural differences-although not one of the

central problems-add to the overall discord in the community.

Religion and Politics

After the development debate, the other major clash in Villa Israel is over religion.

Relations between evangelicals and Catholics are so tense there that many people do not

associate across religious lines. Because of Catholics' dominance in popular society,

evangelicals blame them for a host of perceived Bolivian social problems, including

laziness, corruption, alcoholism, and the disintegration of the family. Many Catholics are

outraged by what they call the hypocrisy of evangelicals-some of whom engage quite

visibly in corruption, drinking, and family abuse-and have grown to regard evangelicals

with intense suspicion and even hatred. Evangelicals, for their part, have been warned by

their pastors about the corrupting influences of Catholics, and tend to avoid all non-

evangelizing contact with them.

Because of the dominance of evangelicals in community life, most residents

believe that the Villa Israel contains few Catholics and non-affiliates. And although

evangelicals do dominate the populace (58 percent), there is a large minority of Catholics

(34 percent) and people with no reported religious affiliation (8 percent). While

evangelicals have established 14 different churches in Villa Israel, Catholics split their

attendance between two churches, both outside the community. Catholics, then, are rather

invisible and disorganized in comparison to the evangelicals. As a result, Catholics see









themselves as a persecuted minority, prevented from participating fully in community

decision-making. Evangelicals, meanwhile, are secure in their dominance of community

politics, but feel threatened by the arrival of new Catholic community members and the

re-conversions of evangelicals back to Catholicism.

The religious composition of Villa Israel's residents does appear to be changing.

While evangelicals report living in Villa Israel, on average, 6.1 years, Catholics and non-

affiliates have lived in the community just 3.4 and 3.8 years, respectively. Based on the

life histories people gave us, I believe that the original settlers of Villa Israel tended to

already be evangelicals or quickly converted once they moved into the community. In

time, some of the converts converted back to Catholicism, and new community members

have moved in. Evangelicals have tried to stem non-evangelical political participation by

disenfranchising non-pioneer residents from the political process and by rejecting

Catholic-funded development projects. Today, the nearly-60/40 split between

evangelicals and non-evangelicals makes challenges to the traditional evangelical

community power structure more viable than ever before.














CHAPTER 4
FIELD METHODS

The field research was designed to yield data on social interactions, economic

exchanges, and water scarcity. The research proceeded in two general phases. During the

first phase, I did participant-observation, hired and trained a research team, made a

sampling frame, and developed interview and observation protocols. The first phase laid

the groundwork for the second, so that I was prepared to complete the first round of

interviews before rainwater reserves ran out in June. During the second phase of research,

my research team and I conducted five two-month cycles of semi-structured interviews

with 76 randomly-selected households. We also conducted 1,986 randomized

observations of 60 public places in Villa Israel.

The research design has several strengths. First, participant-observation produced

data with high internal validity about local cultural institutions and practices (Kirk and

Miller 1986). Second, the household interviews allowed me to document change by

collecting repeated measurements of household characteristics over time. Third, the

observations produced data with high external validity about social interactions in Villa

Israel. Finally, the use of three forms of data collection-participant-observation,

household interviews, and direct observation-enabled me to check the results of each

method against the other, facilitating identification of patterns and problems in the data.

In this chapter, I describe how I implemented each phase of the research design in

the field. In doing so, I have two goals: first, to report the extent to which I followed

established protocols in conducting the field research and, second, to explain how those









protocols were adapted to the ethnographic context. The organization of the chapter

follows the temporal progression of the fieldwork; each section builds on the one before

it and lays the groundwork for the next-just as the stages of field research did. I begin

with a discussion of my re-entry into the field in January 2004.

Gaining Community Support

Before starting the field research, I wanted to get general approval for the project

from the local governmental authority, the Junta Vecinal or Neighborhood Council.

During my pre-dissertation research in 2003, I noticed that many community members

were troubled to see a gringa in Villa Israel. I hoped that winning the Neighborhood

Council's approval would help me overcome community members' qualms. After my

return to Villa Israel in January 2004, I obtained permission to do research during a

meeting of the board of the Neighborhood Council.

From my first meeting with the Neighborhood Council, it was clear that community

leaders thought I would be useful to them. Within a week, I was given a seat of honor at

board meetings, was asked to find funding for various community projects, and was

approached by a national political party representative who wanted a foothold in Villa

Israel. At Neighborhood Council meetings, my research team and I were allocated time to

talk to Villa Israel's household heads. After we explained our research process (and that

participating households would receive monetary compensation) in one meeting, two

community leaders gave fiery speeches arguing that households should participate and

donate their earnings to the Neighborhood Council. At that moment, looking around at

the skeptical faces of household heads, I realized that being closely associated with the

Neighborhood Council could harm the project. In the weeks that followed, people told









me stories about community leaders' self-interest, corruption, and unpopularity. I decided

to put some distance between my project and the local government.

Once I realized that closeness with the Neighborhood Council might compromise

my position in the community, I decided that I would have to explain the project's

origins, methods, and goals to potential participants by myself, person by person.

Because many community members are suspicious of Americans (I discuss this further in

the last section of this chapter), it was difficult to win the community's support for the

project. Using my role in the women's work group as a conversation-starter, I explained

who I was and what I wanted to do in Villa Israel to everyone that would talk to me.

Slowly, I built grassroots support for the research project during the participant-

observation phase of the research.

Participant-Observation

In January 2004, I began living in Villa Israel and doing participant-observation.

As I mentioned in the last chapter, only a handful of visible social interactions transpire

in Villa Israel. Because people generally stay inside walled housing compounds, it is

difficult to casually meet people. Between January and May of 2004, I developed four

general strategies to build relationships in Villa Israel: spending money, helping,

attending get-togethers, and visiting.

After I moved to Villa Israel, I got to know people by patronizing local businesses.

In Cochabamba, however, vendors sometimes consider shopping around a serious affront

(that implies the vendor overcharges or has inferior products) once a commercial

relationship is established, so this tactic must be used cautiously. Carefully spreading my

business around, I bought bread from one bakery in the morning and another in the

evening, eggs and meat from one corner store and jam from another, peach juice from









one booth and soda from another. In this way, I got to know about twenty merchants and

their families. Since businesses are important gathering-places where news is exchanged,

I also learned about what was going on around Villa Israel by listening in on other

customers' conversations.

The best way to get access to large groups, before anyone knew me, was by

offering to help out. Workers on community projects, like water line construction, tree

planting, and bridge building, were always willing to welcome another laborer into their

ranks. Also, there is a large unmet need for childcare in Villa Israel. I offered to teach

free English classes and played games with unsupervised children in the evening. As a

result, I got to know many workers, children, and parents.

Once people knew me, the fastest way to find groups of people in Villa Israel was

at Sunday meetings. Every Sunday, people converge on three major soccer fields in the

community. People also assemble in Neighborhood Council meetings and in church. In

each of these settings, families and friends gather in small groups to gossip, eat fried tripe

or puffed rice, and reflect on the main event. While these informal social groups represent

rich sources of information, they are unreceptive to intrusions from strangers. For this

reason, I was generally able to collect data in Sunday meetings only from the people I

already knew well.

After I had been in Villa Israel for a few months, people seemed to be more

comfortable with seeing me around. A few women made friends with me and invited me

into their homes to eat, or to show me some weaving they had been working on, or to ask

for advice on a project. In Villa Israel's patios, kitchens, and bedrooms, I learned how

women wash laundry, cook meals, divide labor within households, and interact within









families. I also learned a few precious details about Villa Israel's history, controversies,

and long-standing rivalries.

Despite my successes in gaining access to many areas of life in Villa Israel, my

interactions with people were generally not rich. Informal conversations were filled with

empty pauses, repetition, and platitudes. When I tried to dig deeper, people often

answered "asi no mas", saying, basically, that's just how it is, and looked at me blankly if

I tried to ask more questions. I found that people were quite good at evading my queries,

and I wondered why getting a good informal conversation going was such a challenge.

After some reflection, I concluded that I had trouble building rapport because-in

addition to being a gringa-I was an unmarried student with no children. At 26, I had

little in common with other people my age in Villa Israel. As a result, many adults were

not interested in talking to me about their marriages, children, or other important topics.

Despite the challenges that participant-observation presented for me, it was a

fruitful research period. I developed four successful tactics to get to know and spend time

with people (spending money, helping, attending get-togethers, and visits). I took notes

on social relations, economic exchanges, water acquisition and use, and other elements of

daily life I observed. The participant-observation data became the foundation for all the

later stages of research, and helped me adapt data collection protocols for Villa Israel.

By the end of March 2004, I reached the point of diminishing returns for

participant-observation data. Although I began interviewing in April, I continued doing

participant-observation when interviews and direct observations were not scheduled. By

May, data entry and coding began to consume nearly all of my free time. Additionally,

there was a burglary in the room I rented in Villa Israel, in which the intruder entered my









computer and deleted the project data from the hard drive. All of the data were backed up

on a CD, and I was easily able to recover the deleted data from the hard drive. However,

it seemed clear that the project data were neither secure nor confidential in Villa Israel.

For security reasons, I moved the project computer and data to a site downtown, far from

Villa Israel. Once I began to split my time between the community and the computer, the

intensive participant-observation research phase ended.

The Research Team

While I was able to do much of the participant-observation work alone, I needed a

well-trained research team to help me complete the rest of the planned research-

sampling, interviewing, and direct observations. During the summer of 2003, I began to

look for research assistants. I wanted people with an interest in the project, a solid

education in research methods, and a familiarity with Cochabamba's squatter settlements.

I also wanted the team to reflect the demographics of Villa Israel as much as possible, in

terms of sex, religion, age, and language.

After interviewing several candidates, I decided to hire close friends and teach

them to do anthropological research. My two main assistants were a married couple,

Wilda and Richard. Both were college students in their mid-twenties, grew up in squatter

settlements, and came from Quechua-speaking families. Wilda is a Baptist woman, and

Richard is a Catholic man. Together, they only represented part of the range of

demographics I needed. I hired Wilda's father, Don Willy, as a full-time Quechua

translator, and her mother, Dofia Dominga, as a part-time Aymara translator. These two

team members were Baptist, in their fifties, had no college education, and lived in a well-

known Cochabamba squatter settlement. They helped round out the age, ethnic, and

socio-economic demographics the team lacked. The last team member was Magda, a









talented high school student who did less-sensitive direct observation work on a part-time

basis. Madga was the only Villa Israel resident who worked on the project's data

collection.

I paid each team member according to the time investment he or she made in the

project. Base salaries were set at a fair market price, and were renegotiated when team

members had additional health or transportation expenses. Base salaries were regularly

augmented with gifts, vacations, and other bonuses. I also provided educational training,

such as the opportunity to publish, do data analysis, study English, and attend

conferences. When I applied for research funding, I was not aware of the local custom of

paying a double salary to workers in the month of December. At the end of the study, I

gave researchers the project equipment (television, computer, printer, etc.) to compensate

for not offering them the expected double salary in December.

In March 2004, I began a one-month training program for the full-time assistants,

Wilda, Richard, and Don Willy. Together, we planned every phase of the research, each

using our expertise to improve the project design. I explained the proper way to get oral

informed consent, create and translate interview protocols, and design a randomized

observation schedule. The research team, in turn, taught me how to make the research

more understandable and acceptable to participants in Villa Israel. For instance, the

research team suggested that, instead of explaining that I was conducting a research

project about the effects of water scarcity on social interactions and economic exchanges,

I should say that I was "writing a book about life in the community, and especially the

lack of water here". The research team also helped me use local vocabulary to translate

the interview protocols, rather than the formal Spanish terms with which I was more









familiar. For instance, we decided to use the informal slang term plata (silver) instead of

formal terms dinero (money), sueldo (wage), and salario (salary) in the interviews. We

also used Quechua words, like mich 'a, to replace Spanish words, like tacaho (cheap,

miserly), that were less familiar to survey participants. While Don Willy did not need an

education in research methods in his role as translator, we found it helped him explain the

project to wary Quechua-speaking participants. Don Willy's ability to understand and

allay potential participants' misgivings about participating in an American-led scientific

research project was particularly valuable. Having a team that was trained in research

methods helped me adapt the research design to local knowledge, language, and political

circumstances without compromising the project's integrity.

During the month-long training period, we also developed collaborative approaches

to planning, research, and teamwork that became quite important in later stages of the

research. The project had a heavy workload of interviews, observations, and coding that

had to be completed every two months over a ten-month period. The work was

physically, mentally, and emotionally exhausting. As the dry season progressed, field

conditions worsened. We got sick with fatigue, headaches, vomiting, and diarrhea.

Perhaps worst of all, some community members suspected us of posing a threat to

community well-being throughout our time in Villa Israel. We learned to be ready for

street-side snubs, shouted accusations, dog attacks, and various other affronts when we

ventured out in the community. We discovered that our best defenses were wicked senses

of humor, conviction in our work, and-most of all-a very strong sense of team unity.

One of the greatest strengths of the research project was its excellent field research

team. I chose people whose talents and temperaments I knew well, and trained them to do









anthropological work. The demographic composition of the research team helped them

gain acceptance among the research participants. In return for their labor, I offered the

team members fair compensation, educational opportunities, and the ability to contribute

to the research design. Over time, we developed a collaborative work style that improved

our resilience to difficult field conditions and the quality of the data we collected in Villa

Israel.

The Sampling Frame

The first task we tackled during the training period was creating a sampling frame.

In January 2004, I asked the Neighborhood Council for a map of the community. I was

given a map of empty land plots that was used for land sales before Villa Israel was

settled. While it was out-of-date and somewhat inaccurate, the land sale map did give us

a good foundation from which we could build a sampling frame for Villa Israel.

First, I wanted to define the boundaries of the sampling frame-what land would

be considered part of Villa Israel for this study. Using the saleable land plots as a general

guide, we found unmapped streets and constructions at edges of the settlement.

Officially, the Neighborhood Council does not consider these "out of bounds"

constructions to be part of Villa Israel. The boundaries defined by the Neighborhood

Council, however, are highly politicized and are contested by the landholders themselves.

As a result, I concluded that political boundaries should only be one of several factors

used to determine the boundaries of the study. I also considered the existence of physical

barriers between the land and the community, the physical proximity of the land to Villa

Israel's main street, and the area's access to Villa Israel's economic and social

landmarks.







62




""- .I :I -

k% --XI I-f
: .r "i-- I:' "ZL---'-.. ; '
." .. -'.- .' I .-



F e aV\e n t r l i, A 2001)
si m"li t the o uet l s- f T








Aft- r de'idgh -d-wolb'n i .-- 'rj 't'-map'-i I ,f V Ir- I
---. ', -- '
--' ., .". '.,-,.' ; .....A-'. -- .. R I. .. ..7 -- --_ _
,--.- N.----- cU.--- ." t








th interesi mia to to sampling for....draet ob The lal map.-a
did not include housing constructions. Also, beyond a few elds labeled "green spaces",
--I- !-,'t
--





Figure 4-1. A map of Villa Israel land plots (from Claure, Periera, & Asociados 2001)
similar to the one used to build a sampling frame. The actual map was a
photocopy of a wall-sized map supplied by the Villa Israel Neighborhood
Council.

After deciding which land would be included in the project's map of Villa Israel, I

had to decide how to use the map. As I explain below, I wanted to sample households for

the interviews, and to sample public places for the direct observations. The land sale map

did not include housing constructions. Also, beyond a few fields labeled "green spaces",

there were no public places identified on the land sale map. It was clear that the research

team would have to do additional mapping before we did any sampling.

It was not difficult to create a systematic approach to mapping public places. I

determined that any indoor location-a church, school, or business-was not truly public,

since access would always be restricted for some people. Streets and open fields

(including sports fields) could be considered public, since anyone could enter them









without restriction. After mapping the streets, I decided to number every intersection and

endpoint (where a street ends) in the community. The observation sample, then, was

based on a regular grid of intersections and endpoints around which Villa Israel's various

public places are situated. Since sports fields always bordered streets and intersections,

they were included automatically. Using the original map with extended boundaries, we

determined that the community had 60 intersections and endpoints.

Drawing a sample for the interviews was much more complicated. Ideally, I would

have drawn a random sample of Villa Israel's households, but could not find an

appropriate sampling frame. I considered conducting a census to determine the location

of households. However, since community members were still quite ambivalent about the

research, any attempt to do a census would have been poorly received-and would have

yielded poor-quality data. Ultimately, I decided to draw a random sample of occupied

structures, since structures could be observed unobtrusively.

Starting again with land sale map, we plotted Villa Israel's constructions. I decided

that, to save time, we would eliminate constructions that clearly were not occupied from

the interviewing sampling frame. However, housing constructions in various stages of

completion-without doors, windows, complete roofs, or electricity-could be occupied,

or not. Wilda and Richard suggested we use the following criteria to code a construction

as occupied: the construction has a roof and at least 3 walls (does not need windows)

AND (1) has an electrical hook-up or electrical wire run from another residence, OR (2)

has a visible water storage area, OR (3) has curtains, toys, or other visible items that

indicate inhabitation. If we were unable to determine the above, we tried to find a

neighbor to consult. If, using this rubric, we were unable to determine if a construction









was occupied, we conservatively coded it as such and included it in the random sample.

Based on these coding rules, we estimated that there were about 415 occupied

constructions in Villa Israel.

By mid-March, the research team had created its own map of Villa Israel. Starting

with an out-of-date map depicting saleable land plots, we redefined community borders,

identified 60 public places, and located 415 occupied constructions in Villa Israel. The

modified map of public places and occupied constructions served as the master frame for

the random samples we drew during the research. Although new constructions were built

and old ones abandoned during the time we worked in the community, our random

samples represent the community as it was in March 2004.

Interview Sampling, Informed Consent, and Compensation

For the household interviews, we drew a random sample of 110 of Villa Israel's

415 occupied constructions using a table of random numbers (from Bernard 2002). Once

we identified a randomly-sampled occupied construction, we made an appointment to

speak with the household members. We defined a "household" as a group of people who

shared living quarters, expenses, and a cooking pot. We discovered that, occasionally,

more than one household occupied a construction that we had selected in the random

sample. If one of the occupants was the owner, we selected his or her household for the

study. If not, we asked the occupants to decide among themselves who should participate.

Although I would have preferred to randomly sample the households within the

construction to determine a participating household, intra-household sampling was not

something that community members were willing to participate in during the early stages

of the research.









On our first round of contacts, we found that 13 of the constructions in the random

sample (12 percent) were actually not inhabited. The remaining sample contained 96

eligible occupied constructions. We attempted to contact adults at each of the 96

occupied constructions up to 14 times. From the 96 occupied constructions, 73

households (76 percent) agreed to participate in the study. Another 23 households (24

percent) refused to participate or were never available to speak with us after 14 attempted

contacts. Here, I do not separate out "refusals" from "never available" because Villa

Israel residents used avoidance as a way to communicate refusal to participate.

Once we had identified a household that was interested in participating, we sought

informed consent. In most cases, we attempted to get informed consent from all adult

household members, since data would be collected on everyone in the household. We

told people that I was from the University of Florida, and received academic grants to

work in Villa Israel. We explained that I was writing a book about people's daily lives,

and especially about how people were affected by water scarcity. We told them that we

would compensate them for the time they spent talking to us, since we knew they might

have to miss work. We also told them that they had the right to refuse to answer questions

or withdraw from the study at any time, but that we would like to interview them five

times over the course of a year.

The compensation we offered was important to participants. We gave each

household 20 Bolivianos (or $2.50 USD at 2005 exchange rates) for each round of data

collection in which they participated. This sum represents average compensation for a

day's work among Villa Israel residents. To be interviewed, some participants had to take

off a full or half day's work. Without reimbursement for this lost income, the neediest









participants would not have been able to take part in the study. Participants also saw the

offer of compensation as a sign of respect. By recognizing that participants' time and

knowledge are valuable, we showed participants that we valued them. The combined

effects of economic and emotional incentives explain, in part, why the quality and

quantity of information we received in interviews was so much better than what I got

from participant-observation.

For the interviews themselves, we spoke only with people who we defined as

"household heads", that is, people responsible for the acquisition and distribution of

household goods. Of the 102 people identified as household heads, 69 percent were

women and 31 percent were men. Of the women, 59 percent headed households alone

and 41 percent headed households with another person. Of the men, 19 percent headed

households alone and 81 percent headed households with another person. It is important

to note that, in several cases, joint-headed households were headed by parent-child pairs,

siblings, and other non-married teams.

In March and April of 2004, the research team identified interview participants.

First, we located the occupied construction that was randomly selected from the sampling

frame. We approached the occupants, selected households, and explained the research

process to household adults. We also offered fair compensation for household

participation. The participation rate for the first interview was 76 percent of randomly

sampled occupied constructions. Because we conducted interviews with people who met

our definition of household head, the majority (69 percent) of people we interviewed

were women.









Developing Interview Protocols

The interviews were designed to produce testable data on social interactions,

economic exchanges, and water provision. The protocol had to be carefully constructed,

so that it produced accurate and reliable data. I also wanted to contribute to the literature

on the urban poor, and searched for a semi-structured survey protocol that could be

reproduced in Villa Israel. Stack's study (1974) of survival strategies among the urban

poor contains such a protocol.

The research assistants and I translated the Stack protocol directly into Spanish, and

from Spanish to Quechua. We ran a series of test interviews, and determined that some of

the Stack's categories and prompts did not translate well into the Andean context. Over a

period of several weeks, we drafted and tested modifications to Stack's categories and

prompts. Most of the changes removed references to Stack's ethnographic context

(African-Americans in the 1960s) and added references to life in an Andean squatter

settlement. For instance, in the set of questions about social interactions (which Stack

called "Daily Life"), we added categories that probed for interactions at soccer games,

community work groups, and other activities particular to life in Villa Israel. Similarly, in

the set of questions about economic exchanges (which Stack called "Finances"), we

added questions about marketing, water fetching, and household economic activities.

Unlike Stack, I planned to quantify and compare social interactions and economic

exchanges. That made improving participants' recall crucial, as I needed them to give us

reliable estimates of the number of times they engaged in certain activities over the

previous week. The participants were asked to recall activities over a week-long period

because test interviews showed that other standard time periods-a day or a month-

produced idiosyncratic and unreliable data. We found that most people accessed their










Table 4-1. The semi-structured interview protocol, modified from Stack 1970.

Semi-structured interview protocol (Modified version of Stack 1974)

DAILY LIVES
My aim is to learn how people spend their time from the moment they wake up in the
morning until they go to bed at night.
a. Describe a typical day in great detail.
(Probe repeatedly using the following categories -- learn who they visit, which relatives they
see daily or weekly, what they do for each other, whether they exchange goods and
services, and how these exchanges are arranged.)
Researchers-find out what identity groups Ego has in common with each person they
name-religion, kin, paisano, neighbor, work, politics, etc.
b. Who do you see or visit each day, each week? Which relatives? Which neighbors?
(Name relationship.)
c. With whom did you spend your day?
d. Where and with whom do you eat breakfast, lunch, dinner? (Probe.)
e. What housework do you do (shopping, scrubbing, cooking, dishes, etc.)? What other
work do you do? With whom do you work?
f. With whom do you enjoy spending time each week?
g. With whom do you participate in group activities (church groups, soccer or basketball,
paid work, community work projects, etc.)?
h. Who do you sit or visit with normally at neighborhood activities-soccer games,
church, neighborhood committee meetings, other events?

FINANCES
Everyone has a hard time making it on the money they get and has to get some help from
others. The aim is to try to figure out how people make it. This gets very complicated
because some people live together, others eat together, and others share their income.
a. Who is living in this house right now? (List relationships.) Who contributes to the
finances of the
household? How do they contribute (rent, utilities, etc.)?
b. Who fetches the water? Who drinks or uses it? Who helps pay for it?
c. Who ate in the household in the last week? Which meals? Who paid for the food and
cooked?
d. Try to learn the source of income of everyone who contributes to the household.
e. What did you do for someone else this week? Did anyone help you out? (Probe for
people outside the household. Get all the information you can about these people.)
f. Did you give anything (goods/services) to any of the individuals listed in e?
g. Did you receive anything (goods/services) from the individuals listed in e?
h. Did you trade food, money, child care, or anything else with anyone this week? With
whom?
i. What else should we ask you about these relationships?

THE ACQUISITION OF GOODS
(Elicit the names of all the items -furniture, pictures, radios, etc.-in each room in the house
that were acquired in the last week. Ask the following questions about each item.










Table 4-1. Continued.

a. Give a physical description of the item.
b. Does it belong to anyone in the house? Who?
c. Was the item in anyone else's home before? Whose?
d. Was it a gift or a loan? Who loaned or gave it to you?
e. Is it home-made? Who made it?
f Where did it come from? Was it bought at a store? Where? Who bought it?
g. Who will it be given to or loaned to?
h. What else should we ask you about it?

CHATTING AND GOSSIPING
a. How do you keep up on what's happening with people you don't see very often (in,
outside Villa)?
b. Who do you chat with?
c. How much time did you spend chatting yesterday? On most days? How much do
people chat with each other, in general, in Villa Israel? (Get examples.)
d. How many people chat together at a time?
e. What do people chat about? (Get examples.)
f How fast does information spread? If you tell a friend something, how long does it take
for your (sister/cousin/other relative named earlier) to hear about it?
g. Are there people you don't like to chat with? Who-why? (Get examples.)
h. Do people gossip very much in Villa Israel?
i. How much do people believe the gossip?
j. What kind of people gossip the most in Villa Israel? What do they talk about?
k. When people gossip about each other, why do they do that?
1. Can you think of a time when gossip really changed the way people in Villa think about
someone? Is there someone people tend to avoid now, because of something they heard
from others? What happened?

CONFLICTS
a. Have you ever seen people in Villa Israel get into arguments with people outside their
family? What happened?
b. When people have trouble with people outside their families, what is it usually about?
What usually happens? What are the consequences? Do other people get involved? (Get
examples.)
c. When people have conflicts outside the family, do other people in Villa usually find
out? What do they do when they find out?
d. Have you had any trouble with anyone outside your family in the last week or so? What
was it about? Did anyone else get involved? How was it resolved? (Probe as much as
possible. Ask about more conflicts).
e. When people have trouble with people inside their families, what is it usually about?
What usually happens? What are the consequences? Do other people get involved? (Get
examples.)
f When people have conflicts outside the family, do other people in Villa usually find
out? What do they do when they find out?
g. Have you had any trouble with anyone inside your family in the last week or so? What
was it about? Did anyone else get involved? How was it resolved? (Probe as much as
possible. Ask about more conflicts).









memories easily when we asked about the types of interactions they had engaged in, such

as visits, work, eating, casual encounters, meetings, and having fun. For people who had

trouble remembering, we developed additional sets of probes based on the following

native categories: relationships (family, friends, neighbors, etc.), places (home, on the

street, at work, in the market, etc.), meals (breakfast, brunch, teatime, etc.), community

activities (Neighborhood Council meetings, soccer games, etc.), and days of the week.

Interviews were conducted with household heads five times over a 10-month

period, from April 2004 to January 2005, and took an average of 65 minutes. Like Stack,

we began each interview by asking people to describe what a normal weekday was like

for them. We then used this baseline understanding of the participant's lifestyle to modify

the interview protocol to the participant's particular social and economic situation. Next,

we asked about the participant's household economy-who was living with them, who

contributed money and labor, how they divided responsibilities, where and with whom

they worked. After that, we asked people about the visits they made and received, the

meals they ate, and what they did with their free time in the community. Throughout, we

constantly probed to determine the number of social interactions and economic

exchanges in which participants had participated over the last week.

During the first round of household interviews, from April to May 2004, we

interviewed 102 household heads. From June to July, we repeated the interviews with 94

household heads. In August and September, we interviewed 93 household heads for the

third time. We conducted the fourth interview with 91 household heads in October and

November. Finally, in December and January, we conducted the fifth and final interviews

with 96 household heads. In total, we conducted 476 semi-structured interviews in Villa









Israel. Participating household heads were lost from the sample because they traveled,

moved away, had an illness or disability, or later withdrew from the study.

After each interview was conducted, we recorded and coded it. Because

participants were not comfortable with the use of recording devices, we took extensive

notes during the interviews. We used our notes to create two-page records of the

interviews after they were finished. We coded the interviews on 90 variables: 10

variables about the interview process, 10 demographic variables, 36 variables about

social relations, and 34 variables about economic relations. The variables that are relevant

to the theoretical question under study here will be analyzed in the chapter on social

interactions and economic relations.

Measuring Water Provision and Scarcity

Ideally, any study of water scarcity could begin with a simple measure of how

much water people have coming into their homes. However, when water distribution

systems are highly complex, as in Villa Israel, it can be difficult or impossible to track

how much water people acquire. In such cases, innovative, locally-appropriate

approaches to estimating water provision must be developed.

To begin, I did extensive participant-observation and ethnography to understand

how households obtain and use water. People in Villa Israel acquire water from five

sources: freelance water trucks, the community tap-stand system, rainwater, the river, and

other community members. They store water in three kinds of containers: underground

water tanks, metal 200-liter drums, and 20-liter buckets. People use water in thirteen

ways: (1) face washing, (2) hair washing, (3) tooth brushing, (4) bathing, (5) making

breakfast, (6) making lunch, (7) making dinner, (8) making beverages, (9) dishwashing,

(10) toilet flushing, (11) bathroom cleaning, (12) washing laundry, and (13) mopping.








I designed a data collection tool to determine the degree to which each household

used these water sources, water storage containers, and water resources. Because the tool

would be used by researchers and participants, many of whom cannot read, it had to be

illustrated. I brought an artist to Villa Israel, and asked her to observe how people use

water there for two days. I then asked her to draw a very simple depiction of each of the

water tasks, containers, and sources. The sketches were then lined up together on a piece

..,nr, .i, t', Ic jHq r. I, ,d I r Ad ad -..dadr? r .. Pa aD 4r ddAn el ,*JIJ p C.l. i
S.alhdes la anudao e a, dgu) iuna de estas activida.ds? Mra con


djlJ U_ U U u tj' L' L.
i'j~w-7


Figure 4-2. The first page of the water data collection tool. It depicts six water-use tasks
(face washing, hair washing, tooth brushing, bathing, making breakfast,
making lunch), two kinds of water containers (10 L buckets given to
participants and cups), and five water sources (river water, tap stand, rainfall
collection, aguatero, neighbor). Artwork by Ashley Yoder, 2004.











'U U U LJU ^L, U I*
-I lull IJJL, a : I 1/ 2,J4- -
t a s k s ( m a k i n g d -m, b e-e d i s h a- s h g .le t fu h i -n g _-



Z L L:_i W
bhm--eni-- washing laundr, and m, wt c ain


fn C: '


Figre w4- .. -Th .....' e aee tool. I' e ct er

buckets and draw an "X" over the correct water source. Artwork by Ashley
j 'SI' + I, _' 1I- .














Yoder, 2004.
"water measures, a self-administered direct measure and a task-based estimate of water

participants' houses, gave them a ten-liter bucket, and asked them to measure and record

all of the water they consumed for the thirteen tasks. We showed them the chart, and
oFipre. This two-,page hrt bce the basc fma t we ue t collect t eI atar- two
tassI ( d inner, ,, beveg ,d, toilt f ....n,

2 0 0 ..... ......















use.
ThFigure firs4-3. The second page of the self-administered dirata collection tool. Itof depicts six water use, was

collected for each household on one randomly-assigned day mopping), water containe firs, andfour
intervfive water sources. Before asked respondent was to color in the correct number of
bparticipants' houses, gave themdraw an "X" ovten-liter buthe correct water source. Artwork by Ashleyord

all of thpaper. Thiser two-page chaonsumed for th became thirteen tasks.ic format we ushowed to collect the data for two, and





all of the water they consumed for the thirteen tasks. We showed them the chart, and









explained that they should record the amount of water their families used and where the

water came from by coloring in the appropriate drawings on each of thirteen tasks. We

scheduled the household's interview for the day following the water measurements.

During the interview, we reviewed participants' water charts, verifying that all responses

were correctly recorded.

The second measure, the task-based estimation of weekly water use, was collected

directly after we verified data for the self-administered measure. Once people's memories

of water use had been jogged, we asked them how often and with what quantities of

water each member of their household performed each of the thirteen water tasks on the

list in the last week. Household heads gave very thorough descriptions of household

water use on each of the thirteen tasks. Giving a typical answer, one woman reported that,

in her seven-person household, four people bathe one time a week with 10 liters of water,

and three people bathe two times a week with 10 liters of water. In each interview, we

probed until participants reported their household water use on each task with this level

of detail.

After recording the task-based estimation, we collected data on a third measure, the

overall estimate. Because most household heads consider water quite expensive and

difficult to acquire, I hoped that they would be able to recall how many times they had

acquired it over the last week. Also, since water containers are visible, I thought that

household heads would monitor provision levels much in the same way that I watch the

gauge on my truck's gas tank dropping each day. To get the overall estimate, we asked

the question "how much water do you believe your household consumed in the last

week?" directly, without the use of additional prompts or probes.









In addition to the water use measures, we also collected data on the experience of

water scarcity. Based on data from participant-observation and interviews, I created a list

of 33 questions about the experience of water scarcity. The English and Spanish versions

of these questionnaires are presented in the chapter on water scarcity. For each question,

we asked if the household head had the water-related experience in the last week, or not.

We asked household heads the same 33 questions over four interview cycles. These data

form the basis for the water experience scale, which I discuss in greater depth in the

chapter on water scarcity.

Because we could not directly measure the amount of water coming into

households, I devised four different measures of household water provision and scarcity

(self-administered direct measure of daily water use, task-based estimate of weekly water

use, overall estimate of weekly water use, and 33 questions about the experience of water

scarcity). Data on water storage space and access for each household were also collected.

These six measures were all created from participant-observation data, and were adapted

to the specific needs and circumstances in Villa Israel. In the chapter on urban water

scarcity, I will discuss each of these measures in greater detail.

Direct Observation

In my original design for this research, I argued that observation data should be

collected to test the quality of recall-based interview data. For each interviewing cycle,

three hours of observations were to be done in every participating household. The

observations would have yielded reliable data that could be compared with recall-based

interview responses about water use, social interactions, and economic exchanges.

During the informed consent period, however, it became clear that most participants









would not tolerate household observations, and those who did would modify their

behaviors so extensively that the data would be useless.

Instead of household observations, I decided to do direct observations of social

interactions in public places. While these observations would not bear directly on the

interview data, they would provide complementary information about social interactions

and economic exchanges. Moreover, the observations could be conducted unobtrusively,

so we could record observations with a reasonable degree of confidence in the

authenticity of observed person's behavior (cf. Altman 1974, for a thorough discussion of

validity and reliability in direct observation).

Observations were carried out in four 42-day cycles over 8 months. This schedule

worked in tandem with the two-month interviewing cycles; for each 60 days of

interviewing, we did 42 days of observations. Once we had the sampling frame of 60

public places (intersections and endpoints), we used a random number table to select 42

sets of 12 public places. For each of the four observation cycles, we repeated the same

sets of 12 observations over the 42-day period.

Each of the 42 interviewing days was assigned a randomly selected time drawn

from 12 daylight hours (between from 7 am and 7 pm). Observations of the 12 randomly-

selected public places, then, took place during a randomly-selected hour-long block (e.g.,

from 7 am to 8 am). All 12 observation points had to be observed within a 60 minute time

span during the assigned observation hour.

Three interviewers and the part-time observer were trained to conduct observations

following a standardized method. Each observer carried a large green notebook that

contained grids for entering observations, locations of intersections and endpoints, the









randomized observation schedule, and a list of 40 activity codes (drawn from participant-

observations). The observer was to walk to the points he or she was assigned in a circular

direction, starting from the place where my rented room was located. If the observers

were questioned by a curious onlooker as to what they were doing, they were instructed

to say that they were "making appointments for the interviews"-an activity that required

a notebook and was often combined with observations.

When observers arrived at an observation point, they were instructed to look in all

directions to a half-block distance, or to half a river-bed, or to half a field; they were not

allowed to record everything that occurred within their field of vision. Because each

observation point only extended halfway to the next point, no two observation points

overlapped or shared the same public space. Observers proceeded to the center of an

observation point, turned 360 degrees to look at everything within view, observed for up

to 5 seconds, and continued walking. When there was a large crowd at an observation

point, observers were allowed to sit down nearby to record, but the initial observation

should only have lasted 5 seconds. The 5 second rule was instituted to make sure that

observers recorded interactions and activities that occurred over a standard snapshot of

time.

In spite of my efforts to make observation procedures uniform, there were some

problems. It was very difficult for observers to record all the activities that went on in

groups of six or more people. Additionally, there was some variation in the amount of

detail that different observers recorded; some observers always recorded more activities

than others. Finally, there was also some variation in our comprehension of the behaviors

we observed. For instance, I was less skilled than Bolivian observers at correctly









identifying different activities. Because of this variation in activity coding, I have decided

not to use the activity variable in this analysis.

Over four cycles, from June 2004 to January 2005, we conducted 1,986

observations of randomly-selected public places during randomly selected daylight hours

in Villa Israel. The same observations were repeated in each cycle, and so there were 496

observations repeated four times over an eight-month time period. Data were collected on

nine variables, and will be analyzed in the chapter on social interactions and economic

relations.

Summary of Field Methods

The data collection for this project began in January 2004 with a five-month period

of participant-observation. In March, I trained a field research team, mapped Villa Israel,

and drew random samples of occupied constructions and public places. In April, the

research team began conducting interviews with 102 participating household heads. The

interviews produced repeated measures of social interactions, economic exchanges, and

water use, and were conducted over five two-month cycles. In June, we began conducting

observations of 60 public places in Villa Israel. The observations produced data about

social interactions in Villa Israel, and were conducted over four 42-day cycles of

randomized observations. The data were immediately coded for statistical analyses; those

analyses will be presented in the three chapters that follow.


Contextualizing the Research Design: Rethinking Community and Cooperation

As I planned the research design, I put a great deal of work into creating sampling

frames, interviewing protocols, observation schedules, coding schemes, and other

elements needed to conduct the research in a standardized and reproducible way. Soon









after I began the research project, I realized that I had not put enough work into adapting

the research design to local traditions that emphasize community and cooperation. With

the help of my research team, I adopted new approaches to community relations and

collaborative research that, ultimately, improved the project's data collection and quality.

The relationship between the research team and community members was quite

bumpy at the beginning of the research project. While some community members did

welcome us, many more were ambivalent, frightened, or hostile to the prospect of

participating an American-run research project. Many Bolivians believe that American

involvement in Bolivian domestic politics has been to blame for some of the most serious

economic and political crises in recent Bolivian history. In recent years, Bolivians living

in the Chapare-growing coca region have said that Americans in military uniforms

burned their crops, homes, and possessions, leaving them destitute. Understandably, then,

some feared that the fieldwork was part of an American intelligence project and would

lead to some new calamity for the community.

As we discussed potential solutions to this problem, Wilda argued vigorously that

we needed to do more to involve the community in the research. To start, we held a large

meeting for study participants after the first round of interviews. In June, we delivered

personal invitations to household heads, organized snacks and skits, and planned a raffle

for those who attended. The first meeting was a huge success: the schoolroom was

packed, participants were relieved to see friends and neighbors seated nearby, and people

had the opportunity to ask questions about the interview process. After the first meeting,

we found that participants were more open in interviews. During the course of the year,









we held more meetings for Villa Israel residents-and found community members were

generally very happy to participate.

I quickly learned that inclusiveness was just as important to the research team as it

was to study participants. At the team's insistence, I modified the interviewing protocol

to include team-based interviewing. We implemented weekly meetings, in which we

discussed problems that arose in the data collection process. We also created "coding

retreats" in the seventh week of each interview cycle, in which we took a week-long

break from Villa Israel to double-check our coding. The team interviewing, weekly

meetings, and coding retreats all helped improve the quality of the data we collected.

As I mentioned in the section on the research team, having a well-trained and

unified team proved invaluable to me during this project. Team members were dedicated

to the project, and suggested important and appropriate modifications to the original

research design. One of the most important sets of modifications they recommended was

the inclusion of cooperative and community-oriented elements in the research. Because

these elements had not been included in projects I had worked on in other ethnographic

contexts, I had not anticipated the need for them in Cochabamba. In the future, I will be

sure to think more carefully about local community and work traditions as I develop new

research designs.














CHAPTER 5
THE POLITICAL ECOLOGY OF URBAN WATER SCARCITY

Introduction

Urban water scarcity is a rather new topic in social science research, and few, if

any, published articles have attempted to define or operationalize the concept. While

water scarcity has been studied extensively in rural contexts, urban water scarcity

emerges out of distinct infrastructural, political, and economic systems. Theoretical

frameworks, like those posited by Sen and Homer-Dixon, help us understand how

complex social-ecological systems produce inequities in water distribution. However, we

must go further to develop a science of urban water scarcity. In this chapter, I present an

analytic framework for defining and operationalizing the urban water scarcity concept.

The chapter proceeds in three broad sections, and uses data from Cochabamba to

illustrate how the analytic framework should be applied. In the first section, I analyze the

political ecology of urban water scarcity in the Cochabamba Valley. Because social-

ecological systems underlie all human water distribution, an analysis of urban water

scarcity should begin with a thorough discussion of water resources in the region under

study. Following recent case studies of urban water scarcity (Ennis-McMillan 2006,

Swyngedouw 2004), I will also discuss how political and economic power shape the

distribution of water in Cochabamba. In the second section, I describe water provision in

Villa Israel, including the water resources, provision, and storage in the community. In

the third section, I present three approaches to defining and operationalizing urban water

scarcity. Using data collected in Villa Israel, I examine urban water scarcity with









measures of water provision, water security, and experience. Together, these three

sections combine to create a detailed ethnographic picture of urban water scarcity in

Cochabamba in general and Villa Israel in particular.

Section 1-Political Ecology of Urban Water Provision and Scarcity in Cochabamba

In this first section, I examine the political ecology of water provision and scarcity

in the Cochabamba Valley. First, I discuss hydrology and water resources in the area,

including groundwater resources, surface water resources, and rainfall. Next, I examine

the social structures that shape water distribution, such as Cochabamba's municipal water

service, urban residential segregation, and water provision systems. Although not all of

the data that would ideally be used for this kind of study were available, I was able to

piece together an extensive analysis of natural, economic, political, social, and cultural

aspects of water provision and scarcity in Cochabamba.

Introduction to Cochabamba's Water Situation

The Valley of Cochabamba has an area of about 100 km2, and contains a number

of alluvial fans, a spring zone, confined aquifers, two rivers, a lake, and a playa zone.

Despite its rich water resources, industrial and domestic pollution have seriously

degraded the surface water resources, and over-abstraction and pollution now threaten the

aquifer system. Since local water reserves alone cannot support the valley's population,

the municipality now brings in water from outside sources to support 62 percent of the

city's population. However, 38 percent of the city's poorest residents are not connected to

the municipal system (SEMAPA in Terhorst 2003). Unconnected residents use a

combination of relatively unreliable services-such as rainwater collection, private water

vendors, and community water systems-to get the water they need to survive.




Full Text

PAGE 1

THE EFFECTS OF URBAN WATER SCARCITY ON SOCIABILTY AND RECIPROCITY IN COCHABAMBA, BOLIVIA By AMBER YODER WUTICH A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2006

PAGE 2

Copyright 2006 by Amber Yoder Wutich

PAGE 3

For Annette and Ashley

PAGE 4

iv ACKNOWLEDGMENTS A number of people have been involved in the making of this dissertation. Each contributed, in small and large ways, to my intellectual growth, emotional health, and physical safety. I thank each of them sincerely. At the University of Florida, my mentor Dr. H. Russell Bernard worked tirelessly to guide my scholarly development and en courage my intellectua l independence. My committee members, Drs. James Jawitz, Anthony Oliver-Smith, and Marianne Schmink, contributed enormously to the conception and execution of the research. Dr. Chris McCarty provided a working environment, at the Bureau of Economic and Business Research Survey Research Center, where I honed my research skills. My friends from the SOR (Students of Russ) group, including St acy Giroux, Lance Gravlee, Mark House, David Kennedy, Rosalyn Negron, Elli Sugita, and others, continually provide entertainment, companionship, and guida nce in anthropologyÂ’s unusual milieu. In South Florida, a loving network of fa mily and friends has supported me and my studies over the last ten year s. My mother and sister, Anne tte Victor and Ashley Yoder, have done everything imaginable to ensure my success and happiness. My mother deserves special recognition for taking on onero us paperwork duties, as does my sister for accompanying me during three months of fieldw ork in Villa Israel. I thank my family, Ronald Yoder, Ann Barlow, Greg Victor, Mort Victor, Lois Victor, and Kyle Victor, for their continual support. The encouragement of our family friends, including the Cohen, Kavanaugh, and Gomez families, has also been most appreciated.

PAGE 5

v I have always relied on one small circle of amazing women, and each of them made extraordinary contributions toward this work. Candice Aloisi, Neha Gandhi, Kristin Kavanaugh, Kathleen Ragsdale, Alicia Turner and Wilda Valencia all provided warm beds, delicious meals, advice, perspec tive, laughter, and—most importantly—their friendship. In Bolivia, I would not have become a true participant in Cochabamba life without the love and guidance of the Valencia fa mily. I consider th e Valencias—Wilfredo, Dominga, Willycito, Lidia, Wilda, and Richard Aguilar—to be my family, as much as if we were born of the same land and blood. I also thank the people of Villa Israel for accepting me in their homes and lives. I thank Susana Southerwood and Abraham Aruquipa of Water for People and Jim Schultz of Water for People for their support in the field. I am grateful for the research fu nds provided by the National Science Foundation, IIE-Fulbright, Paul and Polly Doughty, and the Center for Latin American Studies at the University of Florida. Finally, I give my heartfelt thanks to Luis Fernando Amarilla Avalos for his companionship and support during fieldwork an d after. His decision to stay in Bolivia was a turning point in my life and his.

PAGE 6

vi TABLE OF CONTENTS page ACKNOWLEDGMENTS.................................................................................................iv LIST OF TABLES.............................................................................................................xi LIST OF FIGURES.........................................................................................................xiv ABSTRACT....................................................................................................................xvi i CHAPTER 1 INTRODUCTION........................................................................................................1 Two Practical Problems: Environmental Sustainability and Human Well-being........2 Problem One: Studying HumanEnvironment Interactions..................................2 Problem Two: Survival and We ll-being of the Urban Poor..................................4 New Directions in Interdisciplinary Research on Urban Water Scarcity..............6 Toward a Cultural Anthropology of Urban Water Scarcity in Latin America.............6 Politics, Economics, and the Urban Poor in Latin America: 1960-2005..............7 Urban Water Distribution in Latin America..........................................................9 Understanding Human Responses to Urban Water Scarcity......................................11 Ecological Anthropology: Classic C ontributions and New Directions...............11 Building a Theory to Test Human Responses to Urban Water Scarcity.............14 Chapter Conclusion....................................................................................................16 2 BOLIVIA: A CO NCISE HISTORY..........................................................................18 Introduction.................................................................................................................18 Geography, Environment, and Natural Resources.....................................................19 A Brief History of BoliviaÂ’s Social, Political, and Econom ic Organization..............21 Culture and Religion...................................................................................................23 Recent Demography: Migration, Identity, and Urbanization.....................................25 Recent Economic History...........................................................................................26 Recent Political History..............................................................................................28 Bolivian Economy and Politics After 2000................................................................29 Poverty, Well-being, and Livelihoods in Contemporary Bolivia...............................31

PAGE 7

vii 3 SETTING....................................................................................................................34 Finding the Field Site..................................................................................................34 Gaining Entry..............................................................................................................35 Landscape...................................................................................................................37 The Cadence of Community Life...............................................................................39 Origins........................................................................................................................ 41 Community Development...........................................................................................42 Economic Well-being.................................................................................................44 Economy and Education.............................................................................................46 Households and the Division of Labor.......................................................................47 Language and Gender.................................................................................................49 Religion and Politics...................................................................................................51 4 FIELD METHODS.....................................................................................................53 Gaining Community Support......................................................................................54 Participant-Observation..............................................................................................55 The Research Team....................................................................................................58 The Sampling Frame...................................................................................................61 Interview Sampling, Informed Consent, and Compensation......................................64 Developing Interview Protocols.................................................................................67 Measuring Water Provision and Scarcity...................................................................71 Direct Observation......................................................................................................75 Summary of Field Methods........................................................................................78 Contextualizing the Research Design: Rethinking Community and Cooperation.....78 5 THE POLITICAL ECOLOGY OF URBAN WATER SCARCITY..........................81 Introduction.................................................................................................................81 Section 1—Political Ecology of Urba n Water Provision and Scarcity in Cochabamba...........................................................................................................82 Introduction to Cochabamba’s Water Situation..................................................82 Hydrology of the Cochabamba Valley................................................................83 Water Resources in the Cochabamba Valley......................................................87 Power, Privilege, and Water Provis ion in the City of Cochabamba...................89 Section 2—Water Resources, Provisi on, and Storage in Villa Israel.........................94 Local Water Resources........................................................................................94 Surface Water Collection....................................................................................97 Rainwater Collection.........................................................................................100 Community Water System: Tap Stands............................................................101 Aguateros, or Water Delivery Trucks...............................................................104 Water Loans, Gifts, and Charity........................................................................108 Buying Water from Businesses and Neighbors.................................................112 Water Storage and Purification.........................................................................114 Section 3—Defining and Operationalizing the Urban Water Scarcity Concept......116

PAGE 8

viii Approach One: Absolute Measures of Water Provision...................................117 The self-administered direct measure........................................................118 The task-based water use estimate.............................................................119 The overall estimate of weekly water use..................................................120 Comparing three measures of water provision...........................................120 Discussion: What do the water pr ovision data really mean?......................124 Approach Two: Proxy Measures of Water Security..........................................125 Precipitation...............................................................................................126 Physical location of households.................................................................127 Political and religious connections.............................................................128 Water storage capacity...............................................................................130 Summary: Proxy measures of water security.............................................131 Approach Three: Measures of the Experience of Urba n Water Scarcity..........131 From interpretation to measurement..........................................................132 Data collection and analysis.......................................................................133 Bodily distress............................................................................................136 Water-based tasks.......................................................................................137 Buying water from the aguatero.................................................................137 Water exchanges........................................................................................138 Water conflicts...........................................................................................139 Summary: Measures of the e xperience of water scarcity...........................140 Discussion: Defining and operationa lizing the urban water scarcity concept..................................................................................................141 6 SOCIAL RELATIONS AND ECONOMIC EXCHANGES...................................143 Anthropological Research on Reciprocity................................................................143 Reciprocity as a Social Foundation...................................................................144 Reciprocity as a System of Economic Exchange..............................................145 Urban Reciprocity.............................................................................................146 Sociability, Reciprocity, and the Laughlin-Brady Model.........................................147 Studying Nested Scarcity Response Cycles.............................................................148 Data on Sociability, Reciprocity, an d Participation in Villa Israel...........................150 Public Social Interactions..................................................................................150 Data collection and coding.........................................................................150 Data analysis..............................................................................................151 Household-level data on Sociabilit y, Reciprocity, and Participation................152 Respondent selection..................................................................................152 Data collected at the household level.........................................................153 Data analysis..............................................................................................155 Cultural Systems and Values are Unique.................................................................157 Data on Exchanges within F our Cultural Institutions.......................................159 Religion and church...................................................................................160 Neighborhood.............................................................................................163 Family.........................................................................................................165 Businesses..................................................................................................166 Conclusion................................................................................................................168

PAGE 9

ix 7 DATA ANALYSIS..................................................................................................170 Section 1—Changes in Water, Sociabil ity, Reciprocity, and Participation over Time.....................................................................................................................170 A Note on Methods of Analysis........................................................................170 Changes in Water Availability over Time.........................................................171 Secure access to water: Rainfall measure..................................................171 Experience of water scarcity : Water-based task scale................................173 Household water provision: Ta sk-based water use estimate......................175 Conclusion: Changes in wate r availability over time.................................178 Changes in Sociability, Reciprocity, a nd Community Participation over Time179 Sociability: Changes in food sharing over time.........................................180 Sociability: Changes in visiting over time.................................................181 Sociability: Changes in public so cial interactions over time.....................183 Reciprocity: Changes in loans over time....................................................185 Reciprocity: Changes in helping over time................................................185 Participation: Changes in church attendance over time.............................185 Participation: Changes in neighbor hood council meeting attendance over time.......................................................................................................185 Conclusion: Changes in sociability, reciprocity, and participation over time.......................................................................................................186 Section 2—Explaining Social In teractions across Households................................188 Identifying Independent Variables....................................................................188 Building the Regression Model.........................................................................191 Understanding the Regression Model...............................................................193 Section 3—Explaining Household So cial Interactions over Time...........................195 Data Analysis Conclusions.......................................................................................197 8 CONCLUSIONS......................................................................................................199 Section 1—Summary of Findings............................................................................199 Section 2—Contributions to Theory and Practice....................................................201 Theoretical Contributions to Anthropology......................................................201 Urbanism, Poverty, and Water Scarcity............................................................201 Urban Reciprocity.............................................................................................202 Human Adaptations to Resource Scarcity.........................................................203 Interdisciplinary Contributions..........................................................................204 Ending Urban Water Scarcity: Six Recommendations.....................................207 Section 3—Looking Ahead in Cultural Anthropology.............................................210 Moving Forward: Anthropological Res earch on Urban Water Scarcity...........210 Cultural Anthropology’s Future: Co llaboration, Method, and Theory.............211 Collaboration..............................................................................................211 Method.......................................................................................................212 Theory........................................................................................................213 APPENDIX: REGRESSION MODELS..........................................................................215

PAGE 10

x LIST OF REFERENCES.................................................................................................233 BIOGRAPHICAL SKETCH...........................................................................................249

PAGE 11

xi LIST OF TABLES Table page 4-1 The semi-structured interview pr otocol, modified from Stack 1970.......................68 5-1 Descriptive statistics for th ree measures of water scarcity....................................122 5-2 English translations of thirty-three yes/no questions about the experience of water scarcity in Villa Israel...................................................................................134 5-3 Thirty-three yes/no questi ons about the experience of water scarcity in Villa Israel, written in the south Cochabamba Spanish dialect.......................................135 7-1 Factor loadings for six potential predic tors of social interactions, for summaries of data collected over five interview periods.........................................................190 7-2 Regression models pred ict two kinds of social interactions, using four independent variables. Data are means for five data periods.................................192 7-3 Regression models predict two kinds of social intera ctions, using two measures of water availability. Data are means for five data periods....................................194 7-4 Regression models predict visiting and food sharing for two study periods: June—July 2004 and August—September 2004...................................................196 A-1 Regression model predicts loans using four independent variables. Data are means for five data periods....................................................................................215 A-2 Regression model predicts helping usi ng four independent variables. Data are means for five data periods....................................................................................216 A-3 Regression model predicts partic ipation in Neighborhood Council meetings using four independent variables. Data are means for five data periods................216 A-4 Regression model predicts participa tion in church meetings using four independent variables. Data are means for five data periods.................................217 A-5 Regression model predicts loans usi ng two independent variables. Data are means for five data periods....................................................................................217 A-5 Regression model predicts helping usi ng two independent variables. Data are means for five data periods....................................................................................218

PAGE 12

xii A-6 Regression model predicts partic ipation in Neighborhood Council meetings using two independent variables. Data are means for five data periods................218 A-7 Regression model predicts particip ation in church meetings using two independent variables. Data are means for five data periods.................................219 A-8 Regression model predicts food sharing using two independent variables. Data were coded from interviews conducted during study period 1 (April-May 2004).219 A-9 Regression model predicts visiting using two independe nt variables. Data were coded from interviews conducted dur ing study period 1 (April-May 2004).........220 A-10 Regression model predicts loans using two independent variables. Data were coded from interviews conducted dur ing study period 1 (April-May 2004).........220 A-11 Regression model predicts helping usi ng two independent variables. Data were coded from interviews conducted dur ing study period 1 (April-May 2004).........221 A-12 Regression model predicts partic ipation in Neighborhood Council meetings using two independent variables............................................................................221 A-13 Regression model predicts particip ation in church meetings using two independent variables.............................................................................................222 A-14 Regression model predicts food sharing using two independent variables. Data were coded from interviews conducte d during study period 2 (June-July 2004)..222 A-15 Regression model predicts loans using two independent variables. Data were coded from interviews conducted dur ing study period 2 (June-July 2004)...........223 A-16 Regression model predicts helping usi ng two independent variables. Data were coded from interviews conducted dur ing study period 2 (June-July 2004)...........223 A-17 Regression model predicts partic ipation in Neighborhood Council meetings using two independent variables............................................................................224 A-18 Regression model predicts particip ation in church meetings using two independent variables.............................................................................................224 A-19 Regression model predicts loans using two independent variables.......................225 A-20 Regression model predicts helpi ng using two independent variables....................225 A-21 Regression model predicts partic ipation in Neighborhood Council meetings using two independent variables............................................................................226 A-22 Regression model predicts particip ation in church meetings using two independent variables.............................................................................................226

PAGE 13

xiii A-23 Regression model predicts food shar ing using two independent variables............227 A-24 Regression model predicts visiti ng using two independent variables....................227 A-25 Regression model predicts loans using two independent variables.......................228 A-26 Regression model predicts helpi ng using two independent variables....................228 A-27 Regression model predicts partic ipation in Neighborhood Council meetings using two independent variables............................................................................229 A-28 Regression model predicts particip ation in church meetings using two independent variables.............................................................................................229 A-29 Regression model predicts food shar ing using two independent variables............230 A-30 Regression model predicts visiti ng using two independent variables....................230 A-31 Regression model predicts loans using two independent variables.......................231 A-32 Regression model predicts helpi ng using two independent variables....................231 A-33 Regression model predicts partic ipation in Neighborhood Council meetings using two independent variables............................................................................232 A-34 Regression model predicts particip ation in church meetings using two independent variables.............................................................................................232

PAGE 14

xiv LIST OF FIGURES Figure page 4-1 A map of Villa Israel land plots (f rom Claure, Periera, & Asociados 2001) similar to the one used to build a sampling frame....................................................62 4-2 The first page of the water data coll ection tool. Artwork by Ashley Yoder, 2004..72 4-3 The second page of the water data co llection tool. Artwork by Ashley Yoder, 2004..........................................................................................................................7 3 5-1 Map of the Cochabamba Valley (from Stimson et al. 2001, p. 1100).....................84 5-2 Seasonal variations in rainfall for the city of Cochabamba (from Vera undated)....86 5-3 Northern Cochabamba has alluvial fans, forests, plant life, high rises, and the residences of the cityÂ’s wealthiest people................................................................87 5-4 Southern Cochabamba is arid, barren, a nd contains the residences of the cityÂ’s poorest people..........................................................................................................87 5-5 Villa IsraelÂ’s river A) during the rainy season, in January 2005 and B) during the dry season, in June 2004..........................................................................................95 5-6 Villa IsraelÂ’s runoff canals are dr y year round, except during heavy rains..............96 5-7 The Villa Israel river during the wet season.............................................................97 5-8 People in Villa Israel rely on river water for bathing...............................................97 5-9 The riverbed is a center for social in teractions where women wash and chat.........98 5-10 Children splash and play in the river during the wet season....................................98 5-11 Wastewater is disposed of in unsealed dumping holes Here, the hole is located inside a housing compound......................................................................................99 5-12 The open area between these two bushes is used as an outdoor toilet by many Villa Israel residents. Wastes drain directly into the riverbed below.......................99 5-13 Rooftop rainwater collec tion equipment includes a drain from roof gutters, a turril and a pipe that extends to a water tank underground...................................100

PAGE 15

xv 5-14 Tap stands are small brick columns built around a water pipe..............................102 5-15 Water delivery trucks, or aguateros are the most common source of water in Villa Israel. Note the aguatero passing through the entrance to Villa Israel.........105 5-16 People buy water from businesses like public baths, corner stores, and restaurants...............................................................................................................113 5-17 Line graph depicting average statistics for three measures of water provision, collected across five survey periods and 76 households........................................123 5-18 Line graph depicting monthly rainfa ll in the Cochabamba Valley (17.45 S and 66.09 W), from January 2004 to January 2005......................................................126 5-19 Line graph depicting water storage capacity among Villa Israel households........130 6-1 Relationships hypothesized by the Laughlin-Brady model....................................148 6-2 Map of Villa Israel, with dots repr esenting 9 of 60 randomly sampled public places......................................................................................................................152 6-3 Means for five measure of reciprocit y, sociability, and meeting participation......156 6-4 Means for loans that took place within f our cultural institutions in Villa Israel during August and September 2004.......................................................................159 6-5 Means for the number of loans, helping, and food shar ing that evangelicals and Catholics engaged in between April 2004 and January 2005................................162 7-1 Total rainfall in mm across six data collection periods, from January 2004 to January 2005, in Cochabamba, Bolivia. (R ainfall data from SENAMHI 2005)...172 7-2 Mean composite household scores for Guttman task-based measure of water scarcity over four study periods, from J une 2004 to January 2005, in Villa Israel.173 7-3 Total Cochabamba rainfall is plotted against the task-based water scarcity measure for data collected between June 2004 and January 2005.........................175 7-4 Mean estimates of household water us e (per person, per day) over four study periods, from April 2004 to Nove mber 2004, in Villa Israel.................................176 7-5 Mean Cochabamba rainfall is plotte d against the household water use measure for data collected between April 2004 and November 2004..................................178 7-6 Mean numbers of food sharing over five study periods, from April 2004 to January 2005, in Villa Israel..................................................................................180 7-7 Mean number of visits over five study periods, from April 2004 to January 2005, in Villa Israel................................................................................................182

PAGE 16

xvi 7-8 Mean number of public social interac tions over four study periods, from June 2004 to January 2005, in Villa Israel.....................................................................184 7-9 Mean number of public social interac tions over four study periods, from June 2004 to January 2005, in Villa Israel.....................................................................185 7-10 A plot depicting factor loadings for three factors..................................................190

PAGE 17

xvii Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THE EFFECTS OF URBAN WATER SCARCITY ON SOCIABILITY AND RECIPROCITY IN COCHABAMBA, BOLIVIA By Amber Yoder Wutich May 2006 Chair: H. Russell Bernard Major Department: Anthropology Global trends like desertification, urban growth, and economic restructuring are making water increasingly scarce—and water access increasingly inequitable—in cities around the world. While much is being done to extend water provision systems to the urban poor, the roots of urban water scarcity are complex a nd difficult to resolve. Today, as urban water scarcity worsens, research is needed to help mitigate its impact on the environment and human welfare. This study is designed to engage with inte rdisciplinary scholarship on two practical problems in urban water scarc ity: (1) small-scale human-environment interactions and (2) changing survival strategies among the urban po or. The study is also designed to answer one unexamined question in anthropological th eory: what are the effects of urban water scarcity on social interactions and ec onomic exchanges? To do so, I employ the Laughlin-Brady model of human adap tation to resource scarcity.

PAGE 18

xviii Over 18 months, I performed participan t-observation, direct observations, and survey interviews in an impoverishe d, water-scarce neighborhood in Cochabamba, Bolivia. The dissertation examines the poli tical ecology of urban water scarcity in Cochabamba, and shows how hydrological, infras tructural, and social trends combine to create conditions of severe wate r scarcity for residents of the city’s poorest residents. The study contains a thorough discussion of th e definition and operationalization of the “urban water scarcity” con cept, including water provisi on, water security, and the culture-bound experience of water scarcity. The data indicate that water scarcity doe s affect sociability during the dry season, as predicted by the Laughlin-Brady model. When people are first hit by severe water scarcity, they become more sociable. As scarci ty worsens, however, sociability decreases. The data also show that water security a nd provision are positively correlated with sociability during the dry s eason, but that the relationship disappears during the wet season. There were no significant changes in r eciprocity. The findings indicate that social responses to water scarcity mimic responses to other resource scarcities (like famine), and point to future directions for resear ch on and prevention of severe urban water scarcity.

PAGE 19

1 CHAPTER 1 INTRODUCTION Global trends like desertification, urban growth, and economic restructuring are making water increasingly scarce—and water access increasingly inequitable—in cities around the world. At least 157 million urbanite s have no access to an improved water source, and hundreds of millions more lack ad equate access to safe water. The vast majority of people without safe water are concentrated in developing cities in Latin America (15 percent), Africa (25 percent), and Asia (57 percent) (UN-Habitat 2003). While much is being done to extend water provision systems to the urban poor, the roots of urban water scarcity are complex and di fficult to resolve. Today, as urban water scarcity worsens, research is needed to help mitigate its impact on the environment and human welfare. Anthropological theories of resource s carcity have much to contribute to interdisciplinary research on urban water scarci ty. In this study, I ex amine the effects of urban water scarcity on sociab ility and reciprocity—and show how the findings relate to ongoing interdisciplinary research. The study is based on 18 months of field research conducted in an impoverished, water-scarce neighborhood called Villa Israel located on the outskirts of Cochabamba, Bolivia. Ther e, my research team and I performed participant-observations, direct observations, and survey interviews. I will dedicate later chapters to discussing water scarcity, sociabili ty, and reciprocity in Villa Israel. Here, I wish to focus first on the broad practical a nd theoretical issues with which this study engages.

PAGE 20

2 The studyÂ’s theoretical contri bution is to advance anthr opological theory in ways that contribute to (1) interdisciplinary theories of urban wa ter scarcity and (2) anthropological knowledge of human belief and behavior. This chapter, then, is divided into two main sections. In the first section, I examine real-world pr oblems that motivate social, biological, and physic al scientists to study water-scarce urban environments. In the second section, I show that urban water sc arcity is a topic that bridges theoretical issues in urban, economic, and ecological anthropology. To unders tand its causes and hypothesize its effects, I examine anthropologi cal research on the urban poor in Latin America, the structure of urban services in Latin American cities, and human adaptations to environmental scarcity across cultures. Two Practical Problems: Environmental Sustainability and Human Well-being Studies of urban water scarci ty are needed for compelli ng environmental and social reasons. Below I briefly examine two key problem s that this study engages: (1) integrated human-environment interactions and (2) the survival and well-being of the urban poor. For each, I briefly discuss the relevant academic research in the problem area, and situate this study within it. Problem One: Studying Human-Environment Interactions In this section, I discuss the environmen tal aspects of water scarcity. First, I examine how human-environmental interacti ons shape urban water scarcity. Next, I discuss how interdisciplinary collaborations have developed to examine this issue. Finally, I explain how this study contributes to new res earch in human-environment interactions. Despite extensive efforts to improve wate r delivery and quality over the last two decades, 1.1 billion people still lack adequa te access to water resources around the world

PAGE 21

3 (Gleick 2004). Although resolving water pr oblems is a major international policy priority, it will be difficult to sustainably manage water resources until we understand the complex human-environment interactions that produce local scarcities. Ecosystems shape the physical availability of water resources for human use around the world (Gleick, Singh, and Shi 2001). However, human social systems overlay these natural systems, and shape peoplesÂ’ ability to access water reso urces. From these complex ecological and social systems emerge human patterns of e nvironmental exploitation. Human patterns of environmental exploitation, in turn, have enormous impacts on ecosystems, which ultimately affects the quality and distribution of global water resour ces (Daily 1997). Put simply, cyclical interactions between th e natural ecosystems and human systems are complex, ever-changing, and the key to creating systems of sustainable water management. Over the last decade, many scholars in the natural sciences have become interested in collaborating with social scientists to understand social-ecological systems and humanenvironment interactions (cf. Gunders on and Holling 2002). Recently, successful interdisciplinary research teams have been built around the study of urban ecosystems, a new area of research that ha s the potential to yield generalizable theories and better management practices for urban social-ecol ogical systems (Collins et al 2000, Grimm et al 2000, Pickett et al 2001, Redman, Grove and Kuby 2004). The urban ecosystems approach has much to contribute to the st udy of urban water scarc ity, in which scholars often treat the city and its water systems as analytically isolated from basin-wide hydrological and ecological issues (Falkenmark and Lundqvist 1995).

PAGE 22

4 While this study is predominately anthropol ogical in its use of theory and methods, it is designed to engage with outside disc iplines. Without a serious discussion of Villa Israel’s wider social and ecological envir onment—and particularly the hydrology of the Cochabamba valley—it would be impossible to understand urban water scarcity and its profound effects on daily life there. This study, then, represents a modest contribution to cross-disciplinary efforts to integrate the social and ecologi cal sciences in the study of urban ecosystems, especially with regard to the relationship between desertification, urbanization, and the sustainability of hu man systems of water distribution and use. Problem Two: Survival and Well-being of the Urban Poor In this section, I discuss th e effects of water scarcity on the survival and well-being of the urban poor. A number of studies have established th at water scarcity and water degradation have serious effects on house hold economies and personal health. Beyond this, however, few have examined the effects of water scarcity on families, communities, and cultural institutions. Afte r briefly reviewing the economic and health impacts of water scarcity, I argue that more should be done to understand how urban water scarcity affects social, economic, and cultural institutions among the urban poor. In developing cities, urban water distri bution generally takes a center/periphery form, in which well-provisioned downtowns gi ve way to outskirts with progressively fewer municipal services (Gilb ert 1998). Without adequate in frastructure, the urban poor suffer from severe water scarcity even in cities that have enough water to meet their residents’ needs. As over-abstraction (or the excessive removal) of groundwater and desertification diminish available water res ources, even independent community wells are running dry. When people lack municipal a nd community water sources, they usually rely on private water vendors that charge 4 to 100 times the price of municipal water

PAGE 23

5 (Elhance 1999). The relatively high cost of water burdens the ove r-strapped budgets of the urban poor. Many household heads must make decisions about whether scarce cash will be spent on food or water, and whether prec ious time will be spent earning income or trying to track down a water vendor. While the economics of urban water dist ribution undermine the livelihoods of the urban poor, water scarcity and degradation pos e a direct threat to their survival. The effects of unsafe water on health are stagge ring: 80 percent of illnesses and 30 percent of deaths in developing countries are caused by the consumption of contaminated water (Elhance 1999) through a variety of diseases such as cholera, typhoid, and dysentery. For those who survive, water related illnesses can push at-risk families into a downward spiral of health care costs and lost of wo rk opportunities (Gleick 2004). In recognition of these problems, the United Nations Millenni um Development Goals aim to halve the number of people without access to safe water and to significantly improve the lives of 100 million slum dwellers by 2015 (United Nati ons undated). Even if the goals are met— which seems unlikely given the inadequacy of resources and funding currently dedicated to the problem—between 34 and 76 million people will die of preventable water-related illnesses by 2020 (Gleick 2004). Water scarcity clearly does major damage to the health and economic well-being of the world’s urban poor. Beyond this, however, the effects of urba n water scarcity on families, communities, and cultural instituti ons have not been studied extensively. In rural communities struck by drought and famine an enormous amount of research shows that families break down, communities disperse and long-standing cultural institutions are destroyed as people pursue survival at all costs (e.g., in studi es of rural African

PAGE 24

6 communities Turnbull 1972, Laughlin 1974, Cashdan 1985, Colson 1979 and reviews of famine literature Dirks 1980, Corbett 1988, Wa lker 1989). Whether these findings can be generalized to urban populations in today’s political and ec onomic conditions remains an empirical question (cf. Ember and Ember 1992, Booth 1984)—one that is tested in this research. New Directions in Interdisciplinary Research on Urban Water Scarcity In this section, I showed that urban wate r scarcity is a problem that bridges two interdisciplinary research areas. New resear ch on the political ecology of urban water scarcity, and its effects on human social a nd economic interactions, would inform theory in both problem areas. First, understanding how human systems of water distribution and consumption work will help explain larger cycles of environmental degradation. Second, understanding the effects of ur ban water scarcity on social and economic relations would help us find the links between the direct eff ects of urban water scarcity (on health and well-being) and possible secondary effect s (on families, communities, and cultural institutions). In the next s ection, I explain how I used an thropological research on urban poverty, resource distribution, a nd resource scarcity to crea te new hypotheses regarding the effects of urban water scarcity on sociability and reciprocity. Toward a Cultural Anthropology of Ur ban Water Scarcity in Latin America In the second half of this chapter, I begi n to develop a research approach that tests cross-cultural hypotheses of urban water s carcity, while situating them in their ethnographic context. In the fi rst section, I review changes in Latin American politics, economy and the lives of the urban poor over th e last thirty years. Recent Latin American history helps explain how urban water sc arcity emerged—and why it appears to be worsening today. In the second section, I disc uss the structure of water distribution and

PAGE 25

7 scarcity in Latin American cities. In the th ird section, I examine an thropological research on resource scarcity, asking how three phase s of research in ecological anthropology— classic findings, political ecology, and new eco logy—can be integrated to theorize the social effects of urban water scarcity. Fi nally, I pose four hypot heses regarding the effects of water scarcity on sociability and reciprocity. Politics, Economics, and the Urban Poor in Latin America: 1960-2005 Starting in the early 1960s, increased rura l-urban migration and the buildup of huge squatter settlements in Latin America sparke d fears that the urba n poor would organize around a radical political age nda capable of toppling exis ting power structures. In response to these fears, anthropologists and other social scie ntists extensively documented the cooperative economic and conser vative political stra tegies that people used to combat scarcity. Scholars showed th at the urban poor relied primarily on strong familial relationships and reciprocal help networks to insure themselves against unemployment, low wages, and a lack of wa ter, electricity, and other urban services (Stack 1974, Leeds 1971, Safa 1974, Lobo 1982, Lomnitz 1977, Isbell 1978, Halebsky 1995). Further, they found that impoverished mi grants were not marginal or isolated, but were bound up in patronage relationships w ith political elites (Eames and Goode 1973, Dietz and Moore 1979, Perlman 1976, Port es and Walton 1976, Cornelius 1975, Lloyd 1979, Castells 1983, Mangin 1967) and rura l communities (Isbell 1978, Buechler and Buechler 1971, Lobo 1982, Graves 1974). During the 1970s, the dynamics of urba n land invasion and political patronage became established, and more squatters came to the city expecting to acquire improved basic services through clientalistic relations with the state. The more city outskirts and the demands of their residents grew, the less cap able the state became of supporting new

PAGE 26

8 squatters (Castells 1983). When economic growth slowed acro ss the region, Latin American elites shifted to loan dependence to support social programs (like the extension of basic services). By the early 1980s, most Latin American nations lacked the foreign exchange to make loan payments and faced default (Weeks 1995). International financiers arranged emergency aid packages that required complianc e with neoliberal economic reform programs. These structur al adjustment programs called for the privatization of state industries and fr ee trade reforms including the acceptance of expanded foreign investment, an end to subsid ies, and a reduction in tariffs. The collapse of national economies and the policies designed to correct it resulted in decreased wages, unemployment, and cuts in state services and subsidies for Latin America’s poorest people (Safa 2004). After the crisis of the 1980s, research began to indicate that cooperative economic strategies that once protected livelihoods in shantytowns and s quatter settlements were in decline (Eckstein 1990). Gonzlez de la Rocha and he r collaborators developed the “poverty of resources” appro ach to explain changes in su rvival strategies among the urban poor in Mexico. They argued that su rvival-oriented exchange networks only function when people have enough goods, income and labor available to invest in reciprocal relationships. The restructuring of the economy and deterioration of labor markets have profoundly eroded poor households ’ asset bases, undermining their ability to sustain cooperative help networks (Gonzl ez de la Rocha and Gantt 1995, Gonzlez de la Rocha 2001, 2004). Moser also explained that global trends in infr astructural decline, job loss, and economic crisis caused a shift from reciprocal comm unity-level survival strategies to preservationis t household-level survival strategies among the urban poor in

PAGE 27

9 Ecuador, Hungary, the Philippines, and Zambia. In a recent symposium on 30 years of scholarship on Latin America’s urban poor, leading scholars agreed that survival strategies had been seriously undermin ed by recent economic trends (Safa 2004, Gonzlez de la Rocha 2004, Ward 2004). Latin America’s cities share a history of rural migration, shantytown formation, political patronage, economic crisis, and cha nging survival strategies among the urban poor. These trends all contribu ted to the emergence of ur ban water scarcity—and help explain why it seems to be worsening. In the following section, I expl ain in greater detail how urban water distribution systems (and excl usion from them) are structured in Latin American cities today. Urban Water Distribution in Latin America From a theoretical perspective, resource scarcity is commonly viewed as having its origins in a shortage of supply, an ex cess of demand, elite resource capture, and ecological marginalization (Homer-Dixon 1999). In semi-arid Latin American cities, all four of these factors play an important ro le in creating municipal water distribution systems that exclude the poor. However, many municipalities suffer neither from a shortage of supply nor an excess of demand—t hey lack the political will or funding to extend municipal systems to far-flung settle ments where the urban poor are concentrated. In that case, the inequitable distribution of entitlements, that is, the ability to access resources (Sen 1999) probably best explains why some people have access to adequate water resources in cities and others do not. Power relationships in the city determine the extent of municipal water services, rationing requirements, and the quality of water distributed to various neighborhoods (Swyngedouw 2004). In most city centers in th e developing world, for instance, water is

PAGE 28

10 delivered through municipal sy stems directly into househol ds. On urban peripheries, however, entire neighborhoods ofte n lack access to municipal water distribution systems. In Latin America alone, 24 million people are estimated to have no urban water service whatsoever, and many more have inadequate or unsafe service (UN-Habitat 2003). In cities, the opportunity to colle ct drinkable surface water is severely restricted by water pollution and public ordinance. Households us ually buy low-quality water from a cistern truck, which is operated by a private vendor. Private vendors typically charge 10 to 20 times the fee charged by public utilities, and people living in marg inal urban areas pay between 10 and 40 percent of th eir incomes to acquire water in this way (Marvin and Laurie 1999, Swyngedouw 1997). As a result, th e geographic marginalization of people on the urban periphery deepen s their economic marg inalization (Satte rthwaite 1998). When rural migrants first began to move en masse to Latin American cities, they too lacked municipal water se rvices. After a period without services, living conditions normally catalyzed groups of squatters to negotiate for comm unity water sources— whether through municipal connec tions or local wells. While th e urban poor have always been pushed to the worst available land, ther e had been more surface water available and higher water tables also facilitated groundwater abstraction. After the collapse of state patronage systems and funding for basic services squatters’ ability to tap state sources for infrastructural improvements was severely limited. While non-governmental organizations (NGOs) have filled in the f unding gap to some extent, population growth, over-abstraction, and climate change make it in creasingly difficult to find water sources in the areas where squatter settlements ar e located. Additionall y, as the urban poor withdraw from cooperative community networ ks, it becomes more difficult for them to

PAGE 29

11 exert political pressure, self-organize to provide services, and maintain existing community infrastructure. While water scarcity can hit anywhere, it currently tends to be concentrated among the urban poor, and especially among the impoverished residents of shantytowns and squatter settlements. In Latin America, environmental, political, economic, and social trends determine how wate r scarcity is distribute d across urban populations. Understanding Human Responses to Urban Water Scarcity While this project is rooted deeply in urban and economic anthropology, it draws primarily on ecological anthropology to find te stable theories of human response to urban water scarcity. In the first se ction below, I briefly review how ecological anthropologists from three stages of research have all contributed to our understanding of resource scarcity, and to some extent of urban water scarcity. In the second section, I argue that a long-neglected theory of human response to resource scar city—when updated with new theoretical perspectives—proves an excellent framework for te sting the effects of urban water scarcity on human r eciprocity and sociability. Ecological Anthropology: Classic Contributions and New Directions Fifty years of research in ecological anth ropology has taught us a great deal about how humans respond to changing ecologica l conditions and resource availability. Cultural materialists and neo-functionalists demonstrated how competition for scarce resources shaped the presence, function, and e volution of cultural tr aits (cf. Vayda 1969, Rappaport 1974, Harris 1979). Ecological anthr opologists also explored how humans respond to environmental stress through studies of demography, adaptive strategies, and coping mechanisms (cf. Vayda and McCa y 1975, Laughlin and Brady 1978). While these scholars made enormous contributions to our understanding of human-environment

PAGE 30

12 interactions, they later came under attack in cultural anthropology for misunderstanding basic biological principles, blinding themselves to the ways in which systems of political and economic domination cr eate non-functional adapta tions, and making overlysimplistic ecological determinist explanati ons of complex phenomena (Hallpike 1973, Friedman 1974, Diener et al. 1980). In the 1980s, political ecologist s began to examine how ecological, economic, and social systems shape human-environment interactions (cf. Durham 1979, Schmink 1982, Stonich 1989, 1993). The shift toward political ecology was widely regarded as a move away fr om active borrowing of ecological terms and concepts that characterized earlier research in ecologi cal anthropology (Orlove 1980, Little 1999). Over the last 20 years, political ecologists have deve loped new models for how anthropologists can produce rigorous et hnographic work that examines the complex interrelationships between ecologica l, economic, and social systems. Negotiating the difficult theoretical te rrain between anthropology and ecology appears to have been a fruitful exercise fo r ecological anthropologist s, as some are now emerging as leaders in new interdisciplinary collaborations on social-ecological systems (Redman 1999, Pavao-Zuckerman 2000, Abel a nd Stepp 2003, Stepp et al 2003). Cultural anthropologists and archaeologists have been welcomed in interdisciplinary collaborations on the study of urban ecosyst ems, where they lend expertise on social, economic, and political issues (Wali et al 2003, van der Leeuw and Redman 2002). To date, anthropologistsÂ’ work on urban ecosystem s combines political ecologistsÂ’ concern with class, gender, ethnicity, and power (cf. Durham 1995) with the ecologistsÂ’ need to understand how human history, values, and desi res figure into ecological outcomes. Early results from two major urban ecology projects in which anthropologists participate (the

PAGE 31

13 NSF Central Arizona Phoenix Long Term Ecol ogical Research program and the Field Museum’s Chicago project) indicate that these collaborations, while difficult, are yielding more complete understandings of urba n human-environment interactions and management needs. While cultural anthropologists have yet to publish on water issues in urban ecosystems, they are developing innovative approaches to the study of urban water scarcity. For instance, authors in an edited text on Water, Culture, and Power: Local Struggles in a Global Contex t use political ecological a pproaches to examine how ecological and social factors combine to create local conflicts and water scarcity (Johnston and Donahue 1998). In his study of “s uffering from water” in a Mexican town, Ennis-McMillan explained how people expe rienced bodily distress over the unjust allocation of scarce water in a commun ity where local hydro logy, overpopulation, poverty, managerial and technical negligence and price gouging combined to create water scarcity (2001, 2006). These studies dem onstrate how anthropologists’ analyses of culture, politics, economy, and power relations hips can contribute to scholars’ overall understanding of urban water issues. Ho wever, as Durham explained, future anthropological research should go furthe r—by explicating the causal links between social processes, forming hypotheses, and tes ting the hypotheses with statistical analysis (Durham 1995). Using reproducible scientific methods facilitates collaborations across disciplines, as well as the accumulation of cross-cultural knowledge within anthropology. When participant-observation is paired with scientific methods, rese arch findings can be particularly compelling.

PAGE 32

14 Building a Theory to Test Human Responses to Urban Water Scarcity I propose using one long-neglected anthr opological theory—Laughlin and Brady’s 1978 model of human adaptation to resource scarcity—to test the socio-economic effects of water scarcity. Laughlin and Brady hypothe sized that sociability and reciprocity underwent a series of predictable changes during cycles of res ource scarcity (like seasonal scarcity or drought cycles). When pe ople experience initial periods of scarcity, they tend to unify to find cooperative solutions to the problem. People tend to increase resource sharing and enlarge cooperative groups to include socially distant groups. As scarcity worsens, however, people lack the re sources and energy to sustain this kind of enhanced sharing and coopera tion. Instead, they withdraw from the enlarged sharing network, narrowing their group participation and limiting reciprocal sharing to people with whom they have stronger, more tr usting, and more durab le social ties. As people withdraw from cooperative re lationships, Laughlin showed, they often switch from exchanges based on positive to ge neralized reciprocity or from generalized to negative reciprocity (1974). With peopl e increasingly looking for benefits for themselves or their narrowed cooperative groups, competition increases and cleavages form between groups. Cooperative goal-maki ng drops off, and the likelihood that the community will adopt cooperative economic or political tactics to deal with scarcity decreases. Laughlin and Brady suggested that each society, then, has a structure that governs the making and maintenance of solidar ities, and that withdrawal of social relationships is ordered by that structure in scarce times, as is the re-establishment of those ties when times of plenty return. The model was tested by respected anthropo logists at the time it was proposed (cf. Dirks 1978 with West Indian slave societies, Lomnitz 1978 in a Mexican shantytown,

PAGE 33

15 and Turnbull 1978 among the Ik of Uganda). Today, it still provides a valuable framework for understanding human adaptiv e cycles. While the Laughlin and Brady volume may include out-of-date references to ecological theory, eco logists have been able to build new theories out of the old ones, and anthr opologists should do the same. Political ecological approaches can be used to explain how, where, and when urban water scarcity emerges in complex societies. Add itionally, anthropological research that draws on “new ecology” can help clarify how concepts like nested hierarchies, nonlinearity, surprise, and transformation can be integrated into Laughlin and Brady’s work (Abel and Stepp 2003, Scoones 1999, Redman 2005). These con cepts can help us explain when and why social systems do not follow predicted patterns. Together, the Laughlin and Brady model, political ecology, and ecologica l anthropology provide a framework for understanding how urban water sc arcity is created by human so cieties, and how it affects social relationships within affected groups. To test the relationship between urban wate r scarcity, sociability, and reciprocity, I developed four hypotheses using information I received from three development agencies working in Villa Israel in 2002. I was told that water scarcity was pr esent in Villa Israel throughout the year, and became so severe duri ng the dry season that fourteen children under the age of one died from water-relate d illnesses between April and July of 2002 (Trujillo 2002). Based on this information, I concluded that people probably experienced moderate water scarcity duri ng the wet season, and suffered from severe water scarcity during the dry season months of May, June, July, and August. For this study, then, I tested the following four hypotheses:

PAGE 34

16 Hypothesis 1: Household heads participate in more social inte ractions during the wet season than during the dry season. Hypothesis 2: Household heads participate in more reciprocal economic exchanges during the wet season than during the dry season. Hypothesis 3: Household h eads with more water part icipate in more social interactions than those with less water. Hypothesis 4: Household head s with more water particip ate in more reciprocal economic exchanges than t hose with less water. Chapter Conclusion This study is designed to engage with inte rdisciplinary scholarship on two practical problems in urban water scarc ity: (1) small-scale human-environment interactions and (2) changing survival strategies that affect the resilience of households, communities, and cultural institutions. The study is also designed to answer one unexamined question in anthropological theory: what are the effects of urban water scarcity on social interactions and economic exchanges? To do so, I employ the Laughlin-Brady model of human adaptation to resource scarcity—as updated by recent anthropological work in political ecology and new ecology. The study provides a test of the relati onship between urban water scarcity, sociability, and reciprocity in V illa Israel. It also contributes one data point to any future cross-cultural study of the Laughlin-Brady mode l in water-scarce urban communities. It uses ethnographic and reproducib le scientific methods to test four hypotheses over 18 months of field research in Villa Israel. It also situates the findings in the wider environmental, cultural, political, and economic realities of Bolivia Latin America, and the world.

PAGE 35

17 Having examined the global trends that shape studies of urban water scarcity in this chapter, I will now situate the study in soci al and environmental realities of Bolivia (chapter 2) and Villa Israel (chapter 3). The n, in chapter 4, I discuss the study itself, and the field methods that I used over 18 months in Villa Israel. In ch apter 5, I examine the political ecology of urban water sc arcity in Villa Israel using data from the field research. In chapter 6, I describe the social interac tions and economic exchanges that took place in Villa Israel using data from the field researc h. In chapter 7, I test the relationship between urban water scarcity, social in teractions, and economic exchan ges in Villa Israel. Finally, in chapter 8, I discuss how the findings relate to interdisciplinary theories of urban water scarcity and anthropol ogical theories of huma n belief and behavior. .

PAGE 36

18 CHAPTER 2 BOLIVIA: A CONCISE HISTORY Introduction When Laughlin and Brady predicted that re sponses to resource scarcity would be similar across cultures, they cautioned that the shape thos e responses take could be distinct in each ethnographi c context (1978). In this study, I show how Laughlin and Brady’s theory can be applied to a wate r-scarce shantytown—an ethnographic context I chose because people’s circumstances there are similar to those found in squatter settlements and shantytowns across culture s. While I believe Laughlin and Brady’s theories are generalizable, I re cognize that historical events play an important role in shaping culture, belief, and action. As a re sult, this chapter is designed to convey a concise history of the relevant geographic, environmental, demographic, cultural, economic, and political trends that shape lif e in Bolivian squatter settlements today. I begin with a discussion of Bolivia’s phys ical environment. This section provides an orientation to the country’s complex geogr aphy and ecology. Next, I discuss elements of Bolivian history that still influence th e country today—in particular, pre-Columbian forms of social organization, Spanish and crio llo dominance, the e xploitation of natural resources, and the rise of powerful workers’ unions. Then, I examin e the distribution of culture, religion, and de mography, with an eye to explai ning how the country has been transformed by modern historical events. After that, I discuss major changes in economics and politics since the 1980s, examin ing the new climate of neo-liberalism, indigenous activism, and political controvers y in Bolivia. The unrest that began with

PAGE 37

19 CochabambaÂ’s 2000 Water War is now unfoldi ng in a spectacular series of political developments that may affect the entire re gion. Finally, I examine how the urban poor are affected by the new neoliberal economy. This last section will lead directly into a discussion of life among the urban poor in one of CochabambaÂ’s most impoverished squatter settlements, Villa Israel. Geography, Environment, and Natural Resources Located in the center of South America, Bo livia has been landlocked since it lost access to the Pacific Ocean to Chile in 1884 (Kluck 1989). On the eastern edge of its Amazonian region, Bolivia shares a long border with Brazil. To the southeast, the arid Chaco scrublands straddle Bolivia and Para guay. To the direct south, Bolivia borders Argentina. To the southwest, Chile borders Bolivia along the Cordiller a Occidental of the Andes. To the northwest, Bolivia shares Lake Titicaca and its surr oundings with Peru. Running roughly from west to east, Bolivia has four major geographic regions: the Altiplano, highland valleys, Amazon, and Chac o. The Altiplano region is a high Andean plateau that runs along the we stern edge of the country. Pl ain-level altitudes range from 4,200 to 4,400 meters (Kluck 1989). Located in the Cerrada river basin, the Altiplano receives between 100 and 500 mm of a nnual rainfall (Van Damme 2002, Mattos and Crespo 2000, FAO 2005). While known for its inhospitable climate, the Altiplano has always been BoliviaÂ’s most populous region, a nd is currently the home of 41 percent of the population (Library of Congress 2005). To the east of the Altiplano, the central highland valleys have a temperate semitropical climate, moderate elevation, and fertil e soil. They contain major cities such as Cochabamba, Sucre, and Tarija, and 29 per cent of the Bolivian popul ation (Library of Congress 2005). To the north and east of the highland valleys, Amazonian rainforests are

PAGE 38

20 situated in the tropical lowlands. This area encompasses one of the world’s most biodiverse regions and 60 percent of Boliv ia’s landmass (Swane y 2001). Together, the highland valleys and Amazonian rainforests are located in the Amazonas river basin, which receives between 300 and 500 mm annua l rainfall in the va lleys and between 1700 and 2200 mm of rainfall annually in the Amazon. In the far southeastern lowlands, the Chaco is an arid, sparsely-populated scrubland. The Chaco (along with the southern tip of the highland valley region) is located in the La Plata basin, and receiv es less than 100 mm of annual rainfall (Van Damme 2002, Mattos and Crespo 2000, FAO 20 05). Bolivia’s combined lowland regions—the Amazon and Chaco—contain 30 percent of the population, which is growing rapidly due to rapid economic devel opment in Santa Cruz (Library of Congress 2005). Within Bolivia’s four climactic region s, there are 16 larg e eco-regions and 198 ecosystems. At least 22,000 superior plant sp ecies and 2,342 vertebrate species have been discovered in Bolivia, and new research cons tantly identifies more species (UNDP 2002, Bojanic 2001). In addition to its biological we alth, the country has great mineral wealth. Bolivia contains significant deposits of ti n, gold, and zinc, as well as South America’s second-largest natural gas reserve and fifth-la rgest oil reserve (Library of Congress 2005, EIA 2005). Because population and consumption have historically been low in Bolivia, human impacts on natural resources have been relatively limited (Ibisch 1998). Even so, recent large-scale land conversions have caused vegetation loss, erosion, and soil salinization. As a result, dese rtification has developed in arid areas, as has flooding in

PAGE 39

21 tropical areas (Van Damme 2002, Bojani c 2001, UNDP 2002, Paredes and Aguirre 2000). Politically, Bolivia is divi ded into nine departments. The La Paz department includes the capital city of La Paz, Lake Titicaca, the headwaters of the Ro Beni, the north Altiplano, and the Amazonian highlands. To the south, Oruro is a small department containing the famous San Jose mine and Lakes Poop and Uru Uru. BoliviaÂ’s southeastern department is Potos, which c ontains the salt flat Salar de Uyuni and the once silver-rich mountain of Cerro Ric o. The highland valley departments are Cochabamba, Chuquisaca, and Tarija. The northernmost of the Amazonian lowland departments is the Pando, which contains the Ro Beni. To its south is the Beni department, which contains the Ros Mamor and Guapor. Finally, Santa Cruz is an enormous department in the southeast corner of Bolivia that encompasses both rainforest and scrubland, the Ro Paragu, a nd Laguna Concepcin (Kluck 1989). A Brief History of BoliviaÂ’s Social, Political, and Economic Organization At one time, population in the area now known as Bolivia was clustered primarily on the fertile edges of Lake Titicaca, in the western slopes of the Andes. Pre-Columbian communities were structured around kinshipbased groups called ayllus (Wagner 1989). The ayllu was the unit through which natural re sources, productive activ ities, and social obligations were organized. The entire sy stem, while not egalitarian, was based on reciprocal exchanges called ayni and labor rotations called mit' a (Moseley 1992). The first major regional civilization, the Tiwanaku, dominated the area from about 600 to 1200 AD. After a long period of regiona l power struggle and Aymara dominance, the Quechua-controlled Inca empire gained co ntrol of highland Bolivia in 1450 (Library of Congress 2005). Building on the basic Andean socio-political unit, the ayllu, the Incas

PAGE 40

22 instituted successful development progr ams in agriculture, mining, and lowland colonization (Moseley 1992). The colonizati on program prompted a major change in BoliviaÂ’s population distributi on, as large numbers of Quechua immigrants were sent to the highland valleys to establish settle ments loyal to the empire (Wagner 1989). In the 1530s, Spanish conquistador Francisc o Pizarro and his troops conquered and dissolved the Incan territory. In 1544, a large vein of silver was discovered in Potos, and Bolivia became a major source of income for the Spanish Crown. Appropria ting the mitÂ’a concept, Spanish colonizers required that indigenous highland men spend every sixth year working in the mines. The Spanish al so created the encomi enda, a peonage-based agricultural system in which the Crown gave colonists co ntrol over indigenous land and laborers (Wagner 1989). Over the next cen tury, millions of native people were conscripted into mining and agricultural wor k, especially in Potos and Oruro, creating major cities and population growth in those areas. In the 1700s, Spanish dominance went into decline as the easy-to-mine mineral veins were tapped and organize d indigenous resistance grew. A series of indigenous and criollo revolts ensued from 1780 until independence was won in 1825 (Wagner 1989). After independence, wealthy criollo landowners continued to contro l BoliviaÂ’s economic assets, and peasants continued to labor in the peonage system. The next century of Bolivian national politics was marred by inep t leadership, corrupt ion, failed wars, territory losses, and economic mismanagement. In 1952, mine workers led a revolt, the Apr il Revolution, to reclaim control of the government. The Nationalist Revolutionary Mo vement (MNR) enacted sweeping reforms including the nationalization of mines, land redistributi on, universal suffrage, and

PAGE 41

23 expansion of education. The MNR also inst itutionalized miners’ a nd peasants’ unions, which participated in decisions about natural reso urce distribution and use. This was the first step taken in 400 years toward ceding control of Bolivia’s natural resources to indigenous groups. The miners and peasant unions established the organizational groundwork for popular protest coalitions that remain powerful in Bolivia today. The MNR controlled Bolivia until 1964, wh en a democratically-elected president was overthrown in a military coup. Eighteen ye ars of political instability followed, in which the MNR and military coup leaders str uggled for control of the country. Since democratic elections were reinstituted in 1982, th ere have been no coups in Bolivia despite serious internal economic and political instability. Culture and Religion Despite a relatively small population of about 9.0 million people (World Bank 2005), Bolivia has a very divers e cultural life. About 62 percen t of Bolivia’s population is of native descent. Of these, 31 percent say they are Quechua and 25 percent say they are Aymara (INE 2004). People of Quechua and Ay mara origin traditionally occupied the highlands, although a half-century of economic crises, political turmoil, and migration has blurred such ethnic boundaries. Anothe r 35 percent of the population is mestizo. After that, much of the remaining populati on is of Chiquitano, Guaran, Mojeo, and European descent, while 1 percent, living main ly in the La Paz department, is of African descent. The national religion of Bolivia is Roma n Catholicism. Many sources claim that Bolivia’s Catholic population is very hi gh—95 percent (cf. Library of Congress 2005, CIA 2005). According to the 2001 national censu s, however, 78 percent of the Bolivian population is Catholic, and 20 percent report being Protestant, evangelical, or belonging

PAGE 42

24 to some other Christian sect (INE 2003a). Others estimate that only between 60 and 70 percent of BoliviaÂ’s population is Catholic (cf. Bureau of Inter-American Affairs 1998), and by all accounts Protestantism is on the rise, as it is throughout Latin America, especially in countries with large indigenous populations (D ow, in press). Whatever the statistics, the Catholic Church still dominates much of the cultural life of Bolivians, through numerous fiestas for saints or virgins, charity work in local communities, and its role in weddings, baptisms, and funerals. De spite the dominance of the Catholic Church, Bolivian Catholic beliefs, rituals, and sy mbolism are interwoven with Quechua and Aymara ones. The most salient example of th is is the fusion of the Virgin Mary and Pachamama, the earth mother, in Bolivi an Catholicism. Throughout Bolivia, people practice Catholicism and indigenous religions concurrently, consulti ng both priests and yatiri (curers) for help with different ailments. Since the 1980s, Protestant missionary groups have become increasingly common in Bolivia. In 1999, religious groups totale d 400, most of which were Protestant missionary groups including Baptists, Pentecostals, Mennonites, Seventh-Day Adventists, and various evangelical groups (Bureau for Democracy, Human Rights, and Labor 1999). In general, all religions are tolerated, a lthough discrimination against religious minorities, particularly Protestants, persists. For instance, Bolivian Protestants tend to be excluded from participation in community events and governance, often because they refuse to participate in important social activities such as drinking alcohol, dancing, and fiestas celebrating Catholic saints (cf. Buechler and Buechler 1971, Goldstein 2003).

PAGE 43

25 Recent Demography: Migration, Identity, and Urbanization Since the agrarian reforms of the 1950s, te mporary and permanent migrations have become an increasingly common part of Bo livian lifestyles. Prior to the 1950s, many peasants were tied to the land through the hacienda system of labor obligations to landowners. After the hacienda system was abolished in 1953, families began to send young men and women to labor seasonally in the lowlands, to mine in the highlands, and to market year-round in city centers li ke Cochabamba (Kluck 1989). In the 1980s, economic crises caused an enormous migr ation of highlanders to urban squatter settlements and lowland colonization areas (McFarren 1992). After highlanders flooded to the cities in the 1980s, urban settlement became dominated by rural-urban migrants in the 1990s. Over the last fifty years, highland migr ants to the valleys and lowlands have maintained close contact with the relatives they left behind in the Altiplano. Peasant families establish exchange ties through migrating members with urban vendors and producers of non-local goods all across Bolivia. Migran ts, in turn, usually continue to contribute to their home communities by dancing in homage to a virgin, hosting migrants, and employing urban networks or Spanish skil ls to solve community legal or political problems (cf. Buechler and Buechler 1971). In this way, multilevel ties between urban and rural communities have been maintained, and can be used to sustain impoverished urban and rural populations with infusions of goods, labor, and food during crises. While highlanders often maintain relati onships in their home communities, they occupy an uncertain role in Bolivian society. As a group, they have come to be known as cholos—a class of people who represented the integration of indigenous ancestry and culture with European cultu re (Weismantel 2001). Cholos normally speak accented

PAGE 44

26 Spanish and adopt some mestizo cultural pr actices, while maintaining other indigenous beliefs and practices (such as language use, dress, food prepara tion, and the use of indigenous medicine). Cholos of ten are often socially isolated ; in towns they continue to be disparaged as peasants, while in their villag es, they are considered to be urbanites. In the new context of citizenship movements emer ging in Bolivia after 2000, cholos are left exposed on the national political scene, l acking the claims to rights and resources associated with both conventional urban pow er structures and new indigenous rights groups (Jackson and Warren 2005). Today, 62 percent of the Bolivian populati on is urban, and urbanization continues to rise at a rate of 3.6 percent per year (Library of Congress 2005). Bolivia’s population is becoming increasingly concentrated in four cities: La Paz, El Alto, Santa Cruz, and Cochabamba. In the 2001 census, 77 of Boliv ia’s urban population re sided in these four areas (INE 2002). As I explain in the next section, profound changes in Bolivia’s political economy during the 1980s and 1990s led large nu mbers of people to leave their homes in the Altiplano to seek new lives in the c ountry’s four major urban areas (Mattos and Crespo 2000). Recent Economic History In 1984, a devastating cycle of debt, trade deficits, and hyperinfla tion that crippled Bolivia’s banks and markets led the govern ment to enact a sweeping IMF-sponsored structural adjustment program (Nash 1992). Despite extensive grassroots opposition, outsiders considered the plan to be an astounding success when the Bolivian economy began to function again nearly immediately after its enactment (M cFarren 1992). Then, in 1994, a new reform called “capitalization” priv atized 50 percent of Bolivia’s saleable national industries—oil and gas (YPFB), el ectricity (ELFEC), telecommunications

PAGE 45

27 (ENTEL), airlines (LAB), and trains (E NFE) (Kohl 2003). Many Bolivians believed, based on the Bolivian government’s claims at the time, that the capitalization program would cause foreign companies to bring enor mous infusions of capital to create jobs, increase economic growth, and help struggling Bolivian industries compete internationally (Kohl 2002). While privatization and capitalization di d improve the performance of Bolivia’s economy and national industries, it was a bi tter disappointment for many Bolivians— particularly in highland mining communities. The privatization and closure of national mines in 1986 disassembled the country’s core industry, dispersing 27,000 ex-miners mainly into urban squatter settlements and the lowland coca-growing regions (McFarren 1992, Nash 1992), while capitalization did not cr eate new jobs for displaced workers. In fact, many key industries suffered a net loss in jobs after capitalization, as companies imported materials and even workers rather than developing Bolivians’ capacity to provide the necessary materials domestical ly (Kohl 2002). In addition, the Bolivian government lost hundreds of millions of dollars in revenues after the privatization of the YPFB (Kohl 2002), which led to cut-backs in jobs and social programs through the country. As a result, many Bolivians lost a subs tantial portion of their social safety net in the 1980s and 1990s, including public services, inco me security, and market regulation of prices for essentials. The economic and social disruptions of the 1980s undermined grassroots activism in Bolivia (Whitehead 2001). The 1985 privatiz ation and closure of national mines, in particular, disassembled Bolivia ’s core activist networks, di spersing ex-miners into new regions and professions. The di spersal of miners’ unions was widely considered a major

PAGE 46

28 step backward in indigenous people’s fight to regain control of Bolivia’s natural resources. A political reform enacted alongs ide the Capitaliz ation Law, however, revived the old activist networks, and enabled grassroo ts groups to reinvent themselves as new social movements. Recent Political History In 1994, the Popular Participation Law (L PP) was passed. The LPP was designed to make Bolivian democracy more representa tive and redistribute national funds and responsibilities at the local level. As part of the reorganization of local-level Bolivian politics, the LPP created Grassroots Terri torial Organizations (GTOs) based on recognized civil society organizations such as pre-colonial indigenous groups (like ayllus), urban neighborhood councils, and rural campesino unions (Kohl 2003). The decentralization and indigenous rights reforms passed in 1994 gave new political resources, access, and legitimacy to territo rial grassroots organizations (Kohl 2003, Van Cott 2000, Whitehead 2001). The compulsory de velopment of GTOs unified people on the local level, educated them in political procedures, and linked them to sponsoring political parties (cf. Medeiros 1990), and help ed them gain democra tic representation in the national parliament for the first time (Van Cott 2000). The LPP, then, created a political opportunity for di senfranchised groups, like in digenous, urban poor, and campesino groups, to seize control of tiny pieces of municipal budgets—and to work toward larger goals. Now, after the austerity, capitalization, and decentralization re forms, five major regional political interests can be identified in Bolivia. First, in La Paz, the Aymara heartland, Aymara separatists agitate for territorial sovereignt y. A second regional interest encompasses all of the wester n highlands (La Paz, Oruro, Potos, and

PAGE 47

29 Cochabamba departments). There, politically active urban poor, campesinos, and indigenous peoples demand various economic a nd political reforms, and exert influence through highly visible protest tactics like re gional blockades. Third, in the Yungas and Chapare, coca growers constantly find ne w ways to circumvent U.S.-backed coca eradication programs. Chapare is now the st ronghold of the Movement Toward Socialism (MAS) party. With the support of coca grow ers and other disenfranchised political interests, MAS leader Evo Morales has risen to remarkable fame and influence in Bolivia and abroad—and was elected Bolivia’s firs t indigenous president in December 2005. Fourth, in the Amazon, eastern indigenous gr oups spearheaded the March for Life and the drive for expanded indigenous rights in th e 1990s. Fifth, lowlanders based mainly in Santa Cruz and known as Cambas (as opposed to highlanders, or Kollas) are Bolivia’s final major interest group. Camba-dominated Santa Cruz area has its own culture, in which residents are generally mestizo, speak a distinct Spanish dialect, and reject highland values and beliefs. It also has poli tical and economic interests and goals quite distinct from those of highl and and indigenous activists. Bolivian Economy and Politics After 2000 The rivalry between Cambas and Kollas is arguably the most important political division in Bolivia today. Afte r the privatization of the na tional mines in the 1980s, many highlanders settled in easter n lowland colonies (Kluck 1989) Since then, major natural gas discoveries and Bolivia’s fastest economic growth have been located squarely in Camba territory. While Santa Cruz prospers the highlands have become increasingly impoverished and politically unstable. In the 2001 census, Santa Cruz was the only department in all of Bolivia in which less th an half of its reside nts (38 percent) were impoverished. Potos, once the he art of the Spain’s colonial mining empire, had become

PAGE 48

30 Bolivia’s poorest department, with 80 percen t of its residents living in poverty (INE 2003b). Many Cambas have come to see the Kolla regions as an economic, political, and cultural drag on their succe ss, and have initiated a gr owing movement for regional autonomy. Kollas, outraged by the racist undert ones in Camba political rhetoric, retort that Santa Cruz’s success has been built on Kolla labor and are fighting to nationalize lowland gas reserves. At the core of the disput e are two distinct visi ons of Bolivia’s past and future. Cambas, encouraged by financia l successes in the coca and natural gas sectors, argue that Bolivia’s future is in ne oliberal economic reform and participation in the globalized economy. Kollas, drawing on the historical memory of centuries of mining booms and market collapses, argue that only a welfare state can protect its citizens from the risks of having a national economy base d on the export of raw materials and cash crops. In the late 1990s, public opinion began to turn against capitalization and other neoliberal reforms (Kohl 2002)—particularly among highlanders who had suffered the worst of mining reforms in the 1980s. After the famous Cochabamba Water War in 2000, a series of increasingly viol ent protests broke out across the countr y, mainly over the privatization of water and gas resources (Sc hultz 2005). The protests have been massive and draw from nearly all sectors and interest s in highland society. To date, the protesters have succeeded in nullifying Bolivia’s business deals with foreign investors, forcing the resignation of two presidents, and calling fo r a new constituent assembly to rewrite Bolivia’s constitution. After Evo Morales’ ac cession to the Presidency in January 2006, he appointed a government cabinet drawn from the leaders of Bolivia’s major social

PAGE 49

31 movement. The new government will surely work to secure re-nationalization of natural resources, the reinstitution of a welfare st ate, and more formal political power for indigenous people. Even so, powerful interests are still at work to preserve neoliberal reforms and growth-oriented policies. In th e fight to define Bolivia’s political and economic future, there is no clear winner—yet. Poverty, Well-being, and Livelihoods in Contemporary Bolivia While the Bolivian protests are exciting to watch and theorize about, the grinding misery of everyday poverty affects far more people—and is the topic of this study. The great outcry over economic reforms in Bolivia today might seem to imply that standards of living are have fallen sharply in the c ountry over the last decade. Indeed, many scholars have shown that—across Latin Am erica—austerity packages, structural adjustment, and neoliberal reforms were followed by the absolute breakdown of livelihoods and cooperative su rvival strategies among the urban poor (cf. Safa 2004, Moser 1996, Gonzalez de la Rocha 1995). By some measures, however, quite the opposite appears to be true, at least for Bolivia. In Bolivia’s 1976 census, 86 percent of th e Bolivian population was considered to be impoverished—that is, they had one or more of five basic needs unsatisfied (housing materials, housing space, water and sani tation, health, or education). By 1992, the number of impoverished Bolivians had dropped to 71 percent. In the most recent national census, in 2001, 59 percent of Bolivians, or 4.7 million people, were classified as impoverished using the same measure (IN E 2003b). Looking at Bolivia’s impoverished people in four ranked groups, the two poorest groups shrank between 1992 and 2001, while the two moderately poor groups gr ew in size (INE 2003b). Non-governmental organizations and development groups are cl early making progress toward meeting the

PAGE 50

32 infrastructural, health, and educational needs of BoliviaÂ’s poorest citizens (cf. World Bank 2005). There is, then an apparent paradox between successes in measures of development and BoliviansÂ’ growing intolerance for the direction of the economic reforms. The paradox is easily reconciled, however: at th e same time that material conditions are improving, livelihoods are dwindling. As of 2003, the fraction of jobs in industry (29.1 percent) and agriculture (14.4 percent) was shrinking. The only sector of the Bolivian economy that was growing was services, at 56.5 percent of the GDP (World Bank 2005). Thus, the result of economic downturns and neo liberal restructuring is job growth in the mostly-informal service sector, where bene fits are few, underemployment is common, and many are self-employed (Library of Congress 2005). Unemployment was estimated at 8 to 10 percent of the Bolivian workfor ce in 2002, and self-employment at 65 percent (Library of Congress 2005). Infrastructu re, education, and h ealth care are basic necessities and are improving in Bolivia, but these advances do not stop the everyday need to buy food, water, and other essential market goods. With an economy that is increasingly market-based and deregulated, peopleÂ’s survival is tied to their ab ility to raise cash income. In urban areas, where people have little access to subsistence goods to offset poverty, 39 percent of the population is impoverished (INE 2004). People living in Bolivia Â’s squatter settlements tend to make a living providing services in the informal market as vendors, maids, day laborers, and taxi drivers. With less disposable income circulating in th e Bolivian economy, the middle classes have cut back their spending on the nonessential goods and serv ices that many of the poorest urbanites provide (Gill 1994). In addition to the ways that their jobs are

PAGE 51

33 affected, squatters are more vulnerable than are other impoverished urbanites to changes in resource distribution and in frastructural development (E ckstein 1990). For instance, squatters live far from the city centers, and rely on long-distance transport to get to work. Sharp rises in gas and transportation costs can make it difficult or impossible for them to get to work, school, or wholesale markets in the city center. In the next chapter, I discuss how people in Villa Israel have come to live and survive in one Cochabamba shantytown.

PAGE 52

34 CHAPTER 3 SETTING Finding the Field Site In the summer of 2002, I made an exploratory trip to Bolivia to talk to development specialists and hydrolog ical engineers about the possibili ty of conducting water scarcity research there. Two hydrologic engineers from a water-delivery NGO called Water for People, Susana Southerwood and Abraham Aruqui pa, expressed a strong interest in the research project I proposed to them. Susana suggested that I consid er working in a 20year old Cochabamban squatter settlement called Villa Israel. She explained that, although Water for People had been involved with Villa Israel for al most ten years, the community’s complex social composition a nd tendency toward conflict had made it difficult to get things done. Susana said that she would be interested in any insight I could add into what was happening, socially, poli tically, and economically in the community. I was intrigued by Susana’s description of the community, and decided to go see it for myself. I was staying in a hotel in dow ntown Cochabamba, and had no idea where Villa Israel was or how one could get th ere. The community did not appear on any official map of Cochabamba. Not knowing what else to do, I found a taxi driver who had heard of the place and asked him to take me out there. I was nervous about what I would find—especially after hearing Susana’s descri ption of the conflictiv e community. All of the ethnographies I had read about Andean communities rejecting and even attacking foreign researchers flashed through my mind during the long taxi ride into and beyond the impoverished south side of the city.

PAGE 53

35 Finally, the taxi driver pu lled off the highway, careened into and back over the top of a ditch, and announced that we had arrive d. As we clattered into the community, I looked out over a vast dry expanse of land cluttered with dusty houses. There were no trees, no plants, and no water in sight. Th ere was only a handful of people mulling about—mostly highland indi genous women known as cholitas dressed in petticoats and full knee-length skirts called polleras lacy white tops, and long braided black hair. There was no possible pretext for me to be there—no businesses, restaurants, no foreigners of any sort whatsoever. I asked the taxi driver if there was any bus or taxi line that serviced the community. He said that there was a bus that came around every once in a while, but there was not one vehicle in sight. Realizing th at to present myself to the community as a clueless intruder looking for a way out was proba bly not the most auspic ious entre into a field site, I told the taxista that I would pay double—an e xorbitant amount of money—for him to take me back to the city. That summer, I took refuge in the safe havens of government offices and NGO headquarters, learning all I could about Co chabamba’s water problems and what was being done to solve them. After meeting with about ten different development organizations, I felt confident that I had f ound the right city for my research. I knew, however, that the success of the project rode not on inst itutional support, but on the acceptance and collaboration of the community members themselves. Gaining Entry On my second trip to Bolivia, in the summe r of 2003, I knew I needed a plan to get started in the community. Once again, Susana proposed a solution. She was supervising work on Villa Israel’s water system, and i nvited me to go along with her and a friend while they checked on the workers’ progress. While we were there, she introduced me to

PAGE 54

36 several well-known men in the community. As we walked around, we ran into a large group of women who were building a cement-fl oored sports court. The women asked us to help them by paying for 20 bags of ceme nt, and Susana coordinated the emptying of our pockets. She then elbowed me along w ith the women, saying—this is your chance, now go with them and get to work. I went with two of the women to buy the cement, and they asked me to stay for lunch. After Susana and her friend had left, I was invited to a feast of steak, salad, potatoes, and Coca-cola. Eating alone, I was emba rrassed to be treated to such a delicious meal while 35 women stood around watching me I ate slowly and talked endlessly, answering every question I could draw out of the shy women about myself, my family, my hometown—anything to keep them smiling. When the meal finally ended, I asked the women if I could come back the next day a nd help them with their work, since I had some free time and was there to learn more about their lives. The women tried to discourage me, saying that I could just sit in the shade of the courtyard’s single tree if I want ed to see what they were doing. They confided that their work consisted of mining large boulders from an area that also served as an informal latrine. A few women argued that the work wa s far too nasty to do voluntarily. Of course, any qualms I felt about working in an open se wer were nothing compared to the prospect of trying to start participan t-observation elsewhere in the seemingly-deserted community. I felt very fortunate indeed to be we lcomed so easily into Villa Israel. The work was brutal, as we spent week s in pairs, lugging boulders, and mixing, shoveling, and spreading cement. A tenuous cam araderie developed during breaks in which the women complained, told jokes, or sa ng with me. I slowly realized that there

PAGE 55

37 were two subgroups among the women, and peopl e were watching closely to see which one I favored. If I carried rocks or shared lu nch with one group too many times in a row, I could be sure that the others would snub me At the time, I did not realize how rare it was to see Villa Israel residents’ j oys and hostilities out in the open. After a few months of partic ipant-observation, I found my footing in Villa Israel. I used my role as informal mascot of the women’s work group to gain wider acceptance in the community. I slowly advanced my re search program, moving from participantobservation to surveys to direct observation, and I learned to accep t that Villa Israel residents’ regard for me would always tilt be tween affection and resentment. As I became a confident ethnographer, I learned how to explai n to curious new visito rs what to expect on their first trip out. Landscape A trip to Villa Israel normally starts in the city of Cochabamba. There are only two transport lines that serve the commun ity, the number 35 bus line and the 101 taxi-trufi line. The number 35 bus line is a large lin e that runs the nor th-south length of Cochabamba, from high in the foothills of Tunari Park, snaking through downtown and the city’s enormous marketplace, and ending at the outside edge of Villa Israel. The bus is slow to load and progress, comes about every 20 to 30 minutes, and runs only from about 6:00 am to 10:30 pm. The 101 taxi-trufi line also originates in Villa Israel, following the 35 route until it hits the Prado (mai n street) uptown, and then turns left to end in a western barrio called Villa Mexico. It features a set of sp eedy taxis designed to hold five people, but can pack in up to tw elve people when some are shoved into the hatchback. The 101 line runs at 5 to 15 mi nute intervals nearly 24 hours—although few drivers want to head out of the relative safety of the dow ntown area between 9 pm and 5

PAGE 56

38 am. Although it is slow and uncom fortable, the 35 bus is the choice of most Villa Israel residents because (at the time of research) it costs 1 boliviano ($ 0.12), while the taxitrufi costs 1.50 bolivianos ($ 0.19). On their way to Villa Israel the 35 and 101 reach the sout hern edge of the market and turn onto the Pan-American Highway. From this jumping-point at the edge of the city proper, the ride takes about 20 minutes in taxi-trufi and 45 minutes in bus. During the drive south, highly-populated barrios with inte rnet cafes and restaurants gradually give way to patches of unused land, rustic corn-beer bars known as chicheras and adobe housing. About halfway through the ride, one must pass a large slaughterhouse, and a putrid stench hangs over the area and wafts into the passing vehicles. Finally, just as the settlements appear to end completely, the bus passes a spiffy brick health center and one last Catholic church. With a sharp turn ri ght, the bus heads though a field and lurches down into a usually-dry riverbed (the one I thought was an inconveniently-placed ditch). Pulling up out of the riverbed, a Villa Israel si gn appears over the roadÂ’s entrance to the community. Finally, the bus makes a wide left, and heads up Villa IsraelÂ’s main thoroughfare. The view of the community that the bus affords is unspectacular. Looking out over the landscape, a new visitor immediately takes in the overwhelming grayness of the terrain. A few squat hills, dotted with sc rubby trees, climb up behind the community. Dust and litter blow along the curbs. The dr y riverbed is dotted with feces, and clouds of filthy sand occasionally gust over the bankÂ’s edges, engulfing passersby. Skinny dogs dominate the streets, running in packs whose complex social organization appears to rival that of the community itself. Most houses are walled in, and every street-side is lined

PAGE 57

39 with an uneven barricade of thorns, chicken wire, adobe, concrete topped with glass shards, metal fencing, and brick. Later, one learns to pick out the community’s landmarks in the dull and unwelcoming landscape. At the head of the main street sit the community’s two most imposing structures: the Christian Evangeli cal Union (UCE) church and the community land prospector’s three-story house. The bus stop (just behind the UCE church) and the taxi stop (at the side of the riverbed) are the hubs of Villa Israel’s economy; they provide work to many men and a few intrepid wo men, the community’s only reliable vendors of prepared food, and its lifeline to food, work, a nd education in the cit y. The open fields are actually the heart of Villa Is rael’s social life, and come alive on Sundays when teams— ranging from uniformed city-wide competitors to pollera -swinging women’s teams to groups of industrious children—assemble to f ace off in the community’s five soccer fields, basketball court, and volleyball court. In time, one might even discover that the patches of green trees tucked into the hills’ crevices indicate wher e Villa Israel’s two wells suck at tiny underground pools of water. The Cadence of Community Life Daily life in Villa Israel st arts at about 4 am. Before dawn, wives start to cook or clean, prepare for work, and boil the morni ng tea. A bit later, husbands get up, and quickly eat breakfast and get ready for work before rushing off between 5 and 6:30 am. From about 4:30 am, the community water gua rd starts his morning rounds, methodically blowing his whistle, opening a pump, and looking on as women, strong children, and a few men fill up their families’ buckets and gasoline cans with 20 to 40 liters of water. The guard continues his rounds, closing each pum p and padlocking its door in turn, until 9 am. Between 7 and 8 am there is a fran tic rush all around the community, as women

PAGE 58

40 line up to get to the cancha and send their children running out of front doors to buy bread or make it to school on time. After the frenetic early morning, life in Villa Israel drags to a slow pace by midmorning. Those who have not left the comm unity to work stay close to home, doing morning chores or attending their cottage i ndustries. By 10 am, mothers or older sisters must begin a 2-hour preparation process to make a typical stew. Laboriously soaking, washing, and rinsing ingredients using just one or two buckets, some women work all morning to ready the midday meal. Few pe ople go out, and the streets are empty and quiet. At noon, there is a brief burst of energy wh en the children are released from school, and mothers or older sisters serve up lunch fo r the family. From 3 to 6 pm, those who can snatch a bit of precious leisure time to rest or nap. Everyone else keeps busy with errands and housework. At 6 pm, there is another rush to the corner stores, this time to buy meat or bread for the evening dinner. From 6 to 8 pm, children go out to pl ay in the streets and soccer fields. From 8 to 9 pm, wage-earning women and men begin to return home, and dinner is served. At nighttime, from about 8 to 11 pm, families spend time together, usually gathered around an old television. By 11 pm, the community is quiet and everyone is at home and sleeping—except member s of Villa Israel’s two gangs, relatively tame groups of young people who wander the riverb ed or hang out in the soccer fields until about 2 am. Villa Israel’s social offerings are meag er during the week, as community members focus almost exclusively on families and work. A few times during the week, people gather in the evening to go to church or attend a community meeting. Only on Sundays

PAGE 59

41 does the community come out from their walle d-in houses to socialize. By late Sunday morning, the soccer fields, community market churches, and food stands fill up with people enjoying their only day off from work. Origins To outsiders, Villa Israel leaders try to portray the community as a group of humble evangelical Christians who came to Cochabamba from the highlands in search of a better life. One visit to a meeting of the conflictive Junta Vecina l (Neighborhood Council) is enough for any visitor to recognize that th e real story is more complicated. The community was indeed settled, nearly 20 y ears ago, by an evange lical Christian land prospector hailing from Hu anuni, Oruro—an impoverished mi ning district perched high in Bolivia’s Andean altiplano Recruiting people to pioneer hi s squatter settlement in the usual way, through networks of relations a nd contacts, the prospector found family members, evangelicals, and Huanuni residents. Quickly, however, word about Cochabamba’s latest barrio got out, and all kinds of people came to buy, rent, and squat in the community. People arrived from Cochabamba’s other barrios, the rainforests, the mines a nd the countryside, and as far away as Argentina. In 2000, about 1712 peopl e were estimated to live in Villa Israel (Claure Periera & Asociados 2001). By the tim e that my research team and I conducted our first random survey of 101 community residents, in April 2004, Villa Israel had become a Bolivian melting pot. Today, the popul ation has a wide range of origins, experiences, identities, and beliefs—which makes it an interesting, complex, and sometimes problematic place. The majority of Villa Israel’s adult population (92 percent) was born in the impoverished highland departments of Potosi, Oruro, and La Paz. A few were born in the

PAGE 60

42 tropical coca-growing regions of Cochabamba ’s Chapare (1 percent) or the La Paz rainforests (2 percent). Another 3 percent of adult residents came from Santa Cruz, the lowland region whose residents are despised by highland Bolivians. About 75 percent of adult residents were born in the countryside, in the departments of Potosi (26 percent), Cochabamba (20 percent), Oruro (18 percent) and La Paz (10 percent). The rest were born in highland mining communities (13 percent) or in the cities (10 percent). Community Development While many Villa Israel residents were born in farms or small pueblos in the altiplano countryside, not all come directly from their hometowns to Villa Israel. Those who did tend to see their move as evidence of upward mobility. Some told us that, living in Villa Israel, they acquired steady cash in comes, proximity to schools and health care, and access to cosmopolitan diversions for th e first time. Although they recognize the challenges of life in Cochabamba’s far-flung sh antytowns, 49 percent of those polled said life in Villa Israel compares favorably to the poverty and sadness of life in the countryside. Those who came to Villa Israel after living in an ur ban downtown, the cash bonanza of the coca-growing regions, or the security of the pre-1985 mining communities, tend to see life in Villa Israel as harsh, intolera ble, or humiliating. Of those polled, 27 percent said that life was worse in Villa Israel than where they lived before, while another 25 percent said it was the same. Life in Villa Israel, of course, has im proved enormously since the first settlers came—when there were no transportation, communi cation, or social services at all. With funds from aid agencies, the government, and community members, residents built evangelical churches, an elementary school a water distribution system, water runoff channels, a foot bridge, a rock-paved street and marketplace, street curbs, a concrete

PAGE 61

43 sports court, and a health clinic (which they share with th e neighboring community, Ch’aqui Mayu). The community government consists of four committees— neighborhood, market, water, and school—and is charged with maintaining old projects and initiating new ones. The community has f our major social programs: Compassion (an evangelical enrichment program for pr e-school and school children), PLANE (a government-funded program for women’s employ ment and community construction), the Mother’s Club (a group of mo thers dedicated to learning new skills), and continuing learning classes for literacy, handicrafts, and English. The evangelical churches also have informal charities, in which they mobilize pa rishioners to donate, intervene in family problems, give advice, cook, and pray for church members. Although the community has ini tiated an impressive array of projects over its brief history, many of them were corruption-ridden, poorly executed, or both. In one infamous incident, leaders touted a scheme to buy la nd in a nearby community and pump its water to Villa Israel, and then ended up “losing” community funds totaling at least $10,000. In the face of political debacles such as that one, interest in cooperative ventures has plummeted. In 2004, four of the eight comm unity groups—the water committee, PLANE, the Mother’s Club, and the continuing learni ng classes—were defunct. When people do get together to work on a project, it rarely goes well. For example, the street curbs and runoff channels began to crumble just months after they were built. A cycle of official corruption and popular apathy undermines volunteerism and cooperation in the community. As one woman—a savvy, long-time resident of Villa Israel—explained, “We gave our trust and money to the leaders, and we can’t get them bac k. Now we joke that,

PAGE 62

44 around here, only the fools still attend m eetings and give their money to the Neighborhood Council.” In spite of the community ’s problems, there are t hose who still believe in participation and progress. Many people would like to see Villa Israel look and act more like the barrios in downtown Cochabamba. A major rift has emerged between those who want the community to invest in paved ro ads, urban sanitation, and water delivery infrastructure, and those who believe that pe ople should leave the co mmunity as it is. The rift is deepened by classism, racism, a nd real economic differences among community members. Those who want modernization and progress bitterly blame the peasants for their lack of education and di sinterest in achieving a respectable middle class lifestyle. Those who oppose community projects argue that one would have to be blind to the political and economic realities of life in Vill a Israel to believe th at mobility to the middle classes is still within reach. Economic Well-being Indeed, many of the residents of Villa Isra el are in an economic situation that is obviously dire. Many housewives have to feed their families on 5 bolivianos ($0.62) a day. Those in the worst conditions manage to live on 2 bolivianos ($0.25) a day—getting by with bread, tea, and bone soup. Despite the abject poverty in which many Villa Israel families live, the community has remarkable economic diversity. Some residents own three or four lots of land, and live off the in come from their stores, taxis, or rental houses. Others own tidy brick houses of two and three stories, with enormous private water tanks that insulate them from the local shortages. One of the most important elements of ec onomic well-being in Villa Israel is home ownership. When economic crises hit, home owners are able to take in renters or working

PAGE 63

45 family members, intensify cottage industrie s, or leave the comm unity temporarily to work elsewhere (Moser 1996). Renters, howev er, are restricted in their use of housing facilities as capital, and risk ending up dispossessed if they miss a rent payment or leave their belongings unattended. About half of Villa Israel residents own their homes (53 percent). About another third (29 percent), share, borrow, or house-sit someone else’s house. After that, 18 percent rents a house or a room. Villa Israel residents, then, are split almost evenly between high and lo w levels of housing security. Economic security, of course, is more complex than housing security alone. A 2001 survey of the community estimated that th e average annual household income in Villa Israel was 5222 Bs, or about $652.75 per year (Claure Periera & Asociados 2001). While this figure supports what litt le data I have household inco me, it does not tell us much about the range of economic wellbeing Villa Israel. Since Clau re et al. did not explicate the methods they used to calcu late average annual income, I was unable to replicate their work. In my own study, I found that it was quite difficult to get good estimates of household income because earnings varied daily, various people contributed to the household economy, and people were very reticent about income levels. After 14 months of research in the comm unity, I devised a 4-point scale to assess economic well-being at the household level. I us ed four criteria to rank households: (1) access to food, water and medicine, (2) steadin ess of work, (3) ownership of housing, and (4) access to capital such as a taxi, store, or inventory. The research team used the scale to rate 75 randomly-selected households, after conduc ting interviews and observation in each household over an 11-month period. We fou nd that the distribution of wealth in Villa Israel was fairly normal. Households in the category we called “poorest”—those

PAGE 64

46 that regularly lacked food, wa ter, or crucial medicines—we re 11 percent of the sample. Those households we called “low-stable”—in which at least one member had steady work, but they still struggled to maintain access to basic necessiti es—were 25 percent of the sample. “High-stable” households—those wi th steady work, secure access to basic necessities, and a house or capital—were 52 pe rcent of the sample. The households we called “wealthiest”—those with steady income at least one house, and capital—were 12 percent of the sample. Economy and Education The most common sources of income in Vi lla Israel are the transport lines and the Cochabamba marketplace. Poor young men are hired by taxi owners to drive in threehour shifts on the 101 line. Middle-income men drive their own taxis, while the wealthiest families hire drivers to drive their vehicles for them. At the end points of the 101 and 35 lines, women vie to sell breakfast, brunch, lunch, and late-night snacks. A few lucky workers are employed by the lines themselves to punch timecards and control the flow of vehicles. In Cochabamba’s downtown, the marketplace is a world apart. An enormous multiblock section of southern Cochabamba dedicated to commerce, the cancha (a word meaning field, now used to denote the market for Cochabamba residents) teems with thousands of precariously-rigged stalls, bl anket-displays, and wandering vendors. The competition is fierce, and there are often 50 or more stalls lined up in a row offering exactly the same goods at the same prices. Vill a Israel’s vendors usually participate in the lowest rung of market prestige, selling fruit or vegetables from blankets and circulating wheelbarrows, away from the rows of lega lly-installed stalls. Profits are minimal (between 5 and 20 bolivianos a day, or $0.63 to $2.50), but so is the capital required—

PAGE 65

47 just enough to buy a day’s worth of produce at the peasant market and the blanket or wheelbarrow to display it. Those who do not work in transportati on or the market tend to work in construction, as maids, or in household i ndustries like leatherwor king, tailoring, weaving, cooking, or raising animals. Extremely few Villa Israel residents have college educations or work in white-collar jobs. In fact, I knew of only two peop le with college educations, both teachers, living in Villa Israel, and both moved out before the end of the study. Despite the low level of education among Villa Israel residents, most adults value education highly—even if in prac tice they lack the time and re sources to facilitate their children’s educational success. In Bolivia’s cu rrent economic climate, however, a college education does not provide much insurance against unemployment and poverty. The Cochabamba job market is increas ingly flooded with underemployed young professionals. New graduates from unconnected families—those who lack relatives and godparents to facilitate their access to government and business jobs—have great difficulty navigating the job market. When V illa Israel’s first generation of college students graduates, in about seven years, most will be have to contend with this disadvantage when they enter the job market. Households and the Division of Labor The household is the heart of resistan ce against poverty and privation. In poor communities, households are generally organized to maximize the ratio of production to consumption (Gonzalez de la Rocha and Ga ntt 1994, Gonzalez de la Rocha 1995). For this reason, household composition tends to chan ge when times get bad. Married children join their parents in joint households with multiple adult income earners. Children are sent out to work. Some mother s remain in Villa Israel whil e the fathers support the family

PAGE 66

48 from afar, while other mothers prefer to support their children alone without involving the father. In Villa Israel, the majority of households (55 percent) consist of a mother, father, and underage children. A quarter (25 percent) of households is of the joint type, containing multiple adult income earners with or without children. Just over a tenth (11 percent) of the households c onsists of mothers raising children alone. A few more households (6 percent) consist of one or bot h grandparents raising grandchildren. Finally, in the original study sample, there were tw o households with a si ngle adult living alone and a household of only underage children. Nearly all adults (87 percent) work in the cash economy. Most Villa Israel fathers are rarely at home, and work full-time in the city, cancha or 101 taxi-trufi line. All single mothers and most married ones participate in income-generat ing activities as well. Some work full-time, in PLANE construction projects as maids in the city, or selling produce in the cancha Others work part-time or season ally taking in washing, cooking food, knitting, weaving, or selling pr oduce at the market. Women stay home from one to six days a week, to concentrate on housewor k, cooking, and taking care of children. Men who work full-time usually have Sundays off, and are often out—playing soccer, attending meetings, or visiting—on their day off. In households with working mothers, responsibility for housekeeping and childcare normally falls to the eldest sister, or occasionally an eldest brother. These teenagers are expected to buy food, fetch water, cook, clean, a nd take care of their siblings. Some also have part-time jobs, and most are enroll ed in school. Many teenage household heads marry early or run away to escape the responsibil ities. Those who stay at home explain

PAGE 67

49 that they cannot abandon thei r younger brothers and sisters. Younger siblings, too, have household responsibilities; they make runs to the corner store, fe tch water or hunt down the water delivery truck, and usually pitch in with some of the childcare, cooking, and cleaning. Some children also work from the ag e of 8, selling ice cream in the community or plastic bags in the cancha The general patterns of labor division desc ribed above hold across most Villa Israel household types, except one. In two-parent households with home industries—corner stores, leather workshops, a chichera or a bakery—the divisi on of labor tends to be much more flexible. Mothers and fathers ro tate household responsibilities, working together in the business, housework, and ch ildrearing. The children are normally less burdened with housework, and have more time to pursue their studies. Such couples also tend to have a rapport that is visibly better than that of ot her couples in the community. Language and Gender Like most Bolivians, nearly all Villa Israel residents are of indigenous descent, and most speak an indigenous language. In a ddition to Spanish, Bolivia has two major indigenous languages: Quechua and Aymara. Quechua is th e more common of the two, and is spoken mainly in the departments of Potosi, Cochabamba, and parts of Oruro. Aymara is spoken mainly La Paz and parts of Oruro. Although most Villa Israel residents speak a second and even thir d language, 90 percent speaks Sp anish proficiently. Of the 10 percent who do not speak Spanish profic iently, 9 percent speaks Quechua, and 1 percent speaks Aymara. All of the people who speak indigenous languages exclusively are women.

PAGE 68

50 Villa Israel Spanish is highly accented, and employs a number of Quechua words. Language can be a real source of discriminati on for Villa Israel residents outside of the community, where native Spanish speakers often look down on people who speak with indigenous accents. Most children, however, speak unaccented Spanish, as well as fluent Quechua or Aymara. Villa Israel parents place a high premium on having their children succeed economically in urban Cochabamba. As a result, many actively encourage their children to integrate in mainstream mestizo society, and tend to a void inculcating them in indigenous values and practices. The most obvious example of this is that, in Villa Israel, not one daughter is being raised de pollera (in indigenous dress); all are de vestido (in Western pants and dresses). Although language and dress can be a source of grief for Villa Israel residents outside of the community, cultural differe nces do not cause many problems at the community level. Villa IsraelÂ’s important pow er players are all me n who, like most Villa Israel men, speak fluent Spanish and dress in Western clothes. Co mmunity meetings are conducted bilingually in Spanish and Quechua as there are few, if any, monolingual Aymara speakers responsible for repres enting their households. In community organizations, language plays li ttle role in excluding househ old representatives (usually men) from participati on and decision-making. Among women, language-related transgressions can be a source of rancor. Women working in community work groups or fetchi ng water at public water taps tend to clump together according to their native language, talking and coordinating among themselves in Aymara or Quechua. When they feel le ft out, Quechuas accuse Aymaras of gossiping, backstabbing, and other offens es, and vice-versa. And indeed, the language-segregated

PAGE 69

51 groups do engage in such behavior. Despite th e resentment that thes e situations create, many women adapt to the linguistic barrier s, moving adeptly between Aymara and Quechua groups. For instance, one woman, wh o came to Villa Israel as a monolingual Spanish speaker ten years ago, told me that she quickly learned to speak Quechua and Aymara out of necessity. Linguistic and cultural differences—although not one of the central problems—add to the ove rall discord in the community. Religion and Politics After the development debate, the other majo r clash in Villa Israel is over religion. Relations between evangelicals and Catholics are so tense there th at many people do not associate across religious lines. Because of Catholics’ dominance in popular society, evangelicals blame them for a host of perceived Bolivian social problems, including laziness, corruption, alcoholism, and the disi ntegration of the fam ily. Many Catholics are outraged by what they call the hypocrisy of evangelicals—some of whom engage quite visibly in corruption, drinking, and family abus e—and have grown to regard evangelicals with intense suspicion and even hatred. Evange licals, for their part, have been warned by their pastors about the corrupting influences of Catholics, and tend to avoid all nonevangelizing contact with them. Because of the dominance of evangelical s in community life, most residents believe that the Villa Israel contains few Catholics and non-affiliates. And although evangelicals do dominate the populace (58 percent) there is a large mi nority of Catholics (34 percent) and people with no reported religious affiliation (8 percent). While evangelicals have established 14 different churches in Villa Israel, Catholics split their attendance between two churches, both outside the community. Catholic s, then, are rather invisible and disorganized in comparison to the evangelicals. As a result, Catholics see

PAGE 70

52 themselves as a persecuted minority, prevente d from participating fully in community decision-making. Evangelicals, meanwhile, are secure in their dominance of community politics, but feel threatened by the arrival of new Catholic community members and the re-conversions of evangelical s back to Catholicism. The religious composition of Villa IsraelÂ’s residents does appear to be changing. While evangelicals report living in Villa Israel on average, 6.1 years, Catholics and nonaffiliates have lived in the community just 3.4 and 3.8 years, respectively. Based on the life histories people gave us, I believe that the original settlers of Villa Israel tended to already be evangelicals or qui ckly converted once they m oved into the community. In time, some of the converts c onverted back to Catholicism, and new community members have moved in. Evangelicals have tried to st em non-evangelical polit ical participation by disenfranchising non-pioneer re sidents from the political process and by rejecting Catholic-funded development projects. Today, the nearly-60/40 split between evangelicals and non-evangelicals makes cha llenges to the trad itional evangelical community power structure more viable than ever before.

PAGE 71

53 CHAPTER 4 FIELD METHODS The field research was designed to yiel d data on social interactions, economic exchanges, and water scarcity. The research proceeded in two general phases. During the first phase, I did participant-observation, hi red and trained a research team, made a sampling frame, and developed interview and observation protocols. The first phase laid the groundwork for the second, so that I was prepared to complete the first round of interviews before rainwater reserves ran out in June. During the second phase of research, my research team and I conducted five two-m onth cycles of semi-s tructured interviews with 76 randomly-selected households We also conducted 1,986 randomized observations of 60 public pl aces in Villa Israel. The research design has several strengths. First, participant-observation produced data with high internal validity about local cultural institutions and practices (Kirk and Miller 1986). Second, the hous ehold interviews allowed me to document change by collecting repeated measurem ents of household characteri stics over time. Third, the observations produced data with high external validity about social interactions in Villa Israel. Finally, the use of three forms of data collection—participant-observation, household interviews, and direct observation—e nabled me to check the results of each method against the other, facilitating identificat ion of patterns and problems in the data. In this chapter, I describe how I implem ented each phase of the research design in the field. In doing so, I have two goals: first, to report the extent to which I followed established protocols in conduc ting the field research and, second, to explain how those

PAGE 72

54 protocols were adapted to the ethnographic context. The organization of the chapter follows the temporal progression of the fiel dwork; each section builds on the one before it and lays the groundwork for the next—just as the stages of field research did. I begin with a discussion of my re-entry into the field in January 2004. Gaining Community Support Before starting the field research, I wanted to get general approval for the project from the local governmental authority, th e Junta Vecinal or Neighborhood Council. During my pre-dissertation research in 2003, I noticed that many community members were troubled to see a gringa in Villa Israel. I hoped that winning the Neighborhood Council’s approval would help me overcome community members’ qualms. After my return to Villa Israel in January 2004, I obtained permission to do research during a meeting of the board of the Neighborhood Council. From my first meeting w ith the Neighborhood Council, it was clear that community leaders thought I would be usef ul to them. Within a week, I was given a seat of honor at board meetings, was asked to find funding for various community projects, and was approached by a national political party repr esentative who wanted a foothold in Villa Israel. At Neighborhood Council meetings, my res earch team and I were allocated time to talk to Villa Israel’s househol d heads. After we explained ou r research process (and that participating households would receive mone tary compensation) in one meeting, two community leaders gave fiery speeches argui ng that households should participate and donate their earnings to the Neighborhood C ouncil. At that moment, looking around at the skeptical faces of household heads, I realiz ed that being closely associated with the Neighborhood Council could harm the project. In the weeks that followed, people told

PAGE 73

55 me stories about community leadersÂ’ self-i nterest, corruption, and unpopularity. I decided to put some distance between my pr oject and the local government. Once I realized that closeness with the Neighborhood Council might compromise my position in the community, I decided that I would have to e xplain the projectÂ’s origins, methods, and goals to potentia l participants by myself, person by person. Because many community members are suspicious of Americans (I discuss this further in the last section of this chapter), it was difficult to win the comm unityÂ’s support for the project. Using my role in the womenÂ’s work group as a conversationstarter, I explained who I was and what I wanted to do in Villa Israel to everyone that would talk to me. Slowly, I built grassroots support for the re search project duri ng the participantobservation phase of the research. Participant-Observation In January 2004, I began living in Villa Israel and doing participant-observation. As I mentioned in the last chapter, only a ha ndful of visible social interactions transpire in Villa Israel. Becau se people generally stay inside walled housing compounds, it is difficult to casually meet people. Between January and May of 2004, I developed four general strategies to build relationships in Villa Israel: spending money, helping, attending get-together s, and visiting. After I moved to Villa Israel, I got to know people by patronizing local businesses. In Cochabamba, however, vendors sometimes consider shopping ar ound a serious affront (that implies the vendor overcharges or has inferior products) once a commercial relationship is established, so this tactic mu st be used cautiously. Carefully spreading my business around, I bought bread from one bake ry in the morning and another in the evening, eggs and meat from one corner stor e and jam from another, peach juice from

PAGE 74

56 one booth and soda from another. In this wa y, I got to know about twenty merchants and their families. Since businesses are important gathering-places where news is exchanged, I also learned about what was going on ar ound Villa Israel by listening in on other customersÂ’ conversations. The best way to get access to large groups, before anyone knew me, was by offering to help out. Workers on community projects, like water lin e construction, tree planting, and bridge building, were always will ing to welcome anothe r laborer into their ranks. Also, there is a large unmet need for ch ildcare in Villa Israel. I offered to teach free English classes and played games with unsupervised children in the evening. As a result, I got to know many workers, children, and parents. Once people knew me, the fastest way to fi nd groups of people in Villa Israel was at Sunday meetings. Every Sunday, people conve rge on three major soccer fields in the community. People also assemble in Neighbor hood Council meetings and in church. In each of these settings, families and friends gath er in small groups to gossip, eat fried tripe or puffed rice, and reflect on the main event. While these informal social groups represent rich sources of information, they are unreceptiv e to intrusions from strangers. For this reason, I was generally able to collect data in Sunday meetings only from the people I already knew well. After I had been in Villa Israel for a few months, people seemed to be more comfortable with seeing me around. A few women made friends with me and invited me into their homes to eat, or to show me some weaving they had been working on, or to ask for advice on a project. In Villa IsraelÂ’s patios, kitchens, and bedrooms, I learned how women wash laundry, cook meals, divide labor within house holds, and interact within

PAGE 75

57 families. I also learned a few precious details about Villa Israel’s history, controversies, and long-standing rivalries. Despite my successes in gaining access to many areas of life in Villa Israel, my interactions with people were generally not rich. Informal c onversations were filled with empty pauses, repetition, and platitudes. Wh en I tried to dig deeper, people often answered “ as no mas ”, saying, basically, that’s just how it is, and looked at me blankly if I tried to ask more questions. I found that pe ople were quite good at evading my queries, and I wondered why getting a good informal c onversation going was such a challenge. After some reflection, I concluded that I had trouble buildi ng rapport because—in addition to being a gringa —I was an unmarried student with no children. At 26, I had little in common with other people my age in Villa Israel. As a result, many adults were not interested in talking to me about their ma rriages, children, or othe r important topics. Despite the challenges that participan t-observation presented for me, it was a fruitful research period. I deve loped four successful tactics to get to know and spend time with people (spending money, helping, attending get-togethers, and visits). I took notes on social relations, economic exchanges, water acquisition and use, and other elements of daily life I observed. The part icipant-observation data became the foundation for all the later stages of research, and helped me adapt data collecti on protocols for Villa Israel. By the end of March 2004, I reached the point of diminishing returns for participant-observation data. Although I bega n interviewing in Apr il, I continued doing participant-observation when interviews and direct observations we re not scheduled. By May, data entry and coding began to consume nearly all of my fr ee time. Additionally, there was a burglary in the room I rented in Villa Israel, in which the intruder entered my

PAGE 76

58 computer and deleted the project data from the hard drive. All of the data were backed up on a CD, and I was easily able to recover the de leted data from the hard drive. However, it seemed clear that the project data were ne ither secure nor confidential in Villa Israel. For security reasons, I moved the project com puter and data to a site downtown, far from Villa Israel. Once I began to split my time between the community and the computer, the intensive participant-observa tion research phase ended. The Research Team While I was able to do much of the partic ipant-observation work alone, I needed a well-trained research team to help me co mplete the rest of the planned research— sampling, interviewing, and direct observati ons. During the summer of 2003, I began to look for research assistants. I wanted people with an interest in the project, a solid education in research methods, and a familiari ty with Cochabamba’s squatter settlements. I also wanted the team to reflect the demographi cs of Villa Israel as much as possible, in terms of sex, religion, age, and language. After interviewing several candidates, I d ecided to hire close friends and teach them to do anthropological re search. My two main assistan ts were a married couple, Wilda and Richard. Both were college students in their mid-twenties, grew up in squatter settlements, and came from Quechua-speaking families. Wilda is a Baptist woman, and Richard is a Catholic man. Together, they only represented part of the range of demographics I needed. I hired Wilda’s fa ther, Don Willy, as a full-time Quechua translator, and her mother, Doa Dominga, as a part-time Aymara translator. These two team members were Baptist, in their fifties, had no college educati on, and lived in a wellknown Cochabamba squatter settl ement. They helped round out the age, ethnic, and socio-economic demographics the team lack ed. The last team member was Magda, a

PAGE 77

59 talented high school student who did less-sens itive direct observation work on a part-time basis. Madga was the only V illa Israel resident who worked on the project’s data collection. I paid each team member according to the time investment he or she made in the project. Base salaries were se t at a fair market price, a nd were renegotiated when team members had additional health or transportati on expenses. Base sala ries were regularly augmented with gifts, vacations, and other bonuses. I also provided educational training, such as the opportunity to publish, do data analysis, study English, and attend conferences. When I applied for research fundi ng, I was not aware of the local custom of paying a double salary to workers in the mont h of December. At the end of the study, I gave researchers the project equipment (television, computer, printer, etc.) to compensate for not offering them the expected double salary in December. In March 2004, I began a one-month training program for the full-time assistants, Wilda, Richard, and Don Willy. Together, we planned every phase of the research, each using our expertise to improve the project de sign. I explained the proper way to get oral informed consent, create and translate in terview protocols, and design a randomized observation schedule. The research team, in turn, taught me how to make the research more understandable and acceptable to particip ants in Villa Israel For instance, the research team suggested that instead of explaining that I was conducting a research project about the effects of wa ter scarcity on social interactions and economic exchanges, I should say that I was “writing a book about life in the community, and especially the lack of water here”. The research team also helped me use local vocabulary to translate the interview protocols, rather than the formal Spanish terms with which I was more

PAGE 78

60 familiar. For instance, we decided to use the informal slang term plata (silver) instead of formal terms dinero (money), sueldo (wage), and salario (salary) in the interviews. We also used Quechua words, like mich’a to replace Spanish words, like tacao (cheap, miserly), that were less familiar to survey participants. While Don Willy did not need an education in research methods in his role as translator, we found it helped him explain the project to wary Quechua-speaking participants. Don Willy’s ability to understand and allay potential participants’ mi sgivings about participating in an American-led scientific research project was particularly valuable. Ha ving a team that was trained in research methods helped me adapt the research desi gn to local knowledge, language, and political circumstances without compromi sing the project’s integrity. During the month-long training period, we also developed collaborative approaches to planning, research, and teamwork that becam e quite important in later stages of the research. The project had a heavy workload of interviews, observations, and coding that had to be completed every two months over a ten-month period. The work was physically, mentally, and emotionally exhaus ting. As the dry season progressed, field conditions worsened. We got sick with fatigue, headaches, vomiting, and diarrhea. Perhaps worst of all, some community members suspected us of posing a threat to community well-being throughout our time in V illa Israel. We learned to be ready for street-side snubs, shouted accusations, dog att acks, and various other affronts when we ventured out in the community. We discovered th at our best defenses were wicked senses of humor, conviction in our wo rk, and—most of all—a very st rong sense of team unity. One of the greatest strengths of the research project was its excellent field research team. I chose people whose talents and temper aments I knew well, and trained them to do

PAGE 79

61 anthropological work. The demographic compos ition of the research team helped them gain acceptance among the resear ch participants. In return for their labor, I offered the team members fair compensation, educational opportunities, and the ability to contribute to the research design. Over time, we develope d a collaborative work style that improved our resilience to difficult field conditions and the quality of the data we collected in Villa Israel. The Sampling Frame The first task we tackled during the trai ning period was creating a sampling frame. In January 2004, I asked the Neighborhood C ouncil for a map of the community. I was given a map of empty land plots that was used for land sales before Villa Israel was settled. While it was out-of-date and somewhat inaccurate, the land sale map did give us a good foundation from which we could build a sampling frame for Villa Israel. First, I wanted to define the boundaries of the sampling frame—what land would be considered part of Villa Israel for this study. Using the saleable land plots as a general guide, we found unmapped streets and constr uctions at edges of the settlement. Officially, the Neighborhood Council does not consider thes e “out of bounds” constructions to be part of Villa Isra el. The boundaries defined by the Neighborhood Council, however, are highly politicized and are contested by the landholders themselves. As a result, I concluded that political boundaries should only be one of several factors used to determine the boundaries of the study. I also considered the existence of physical barriers between the land and the community, the physical proximity of the land to Villa Israel’s main street, and the area’s acce ss to Villa Israel’s economic and social landmarks.

PAGE 80

62 Figure 4-1. A map of Villa Israel land plot s (from Claure, Periera, & Asociados 2001) similar to the one used to build a sampling frame. The actual map was a photocopy of a wall-sized map supplie d by the Villa Israel Neighborhood Council. After deciding which land would be included in the project’s map of Villa Israel, I had to decide how to use the map. As I explai n below, I wanted to sample households for the interviews, and to sample public places fo r the direct observati ons. The land sale map did not include housing constructions. Also, be yond a few fields labeled “green spaces”, there were no public places identified on the land sale map. It was clear that the research team would have to do additional mapping before we did any sampling. It was not difficult to create a systema tic approach to mapping public places. I determined that any indoor location—a chur ch, school, or business—was not truly public, since access would always be restricted for some people. Streets and open fields (including sports fields) coul d be considered public, sin ce anyone could enter them COMMUNITY ENTRANCE RIVERBED MAIN INTERSECTION DRAINAGE DITCH MAIN SOCCER FIELD UCE CHURCH SCHOOL MARKET TAXI-TRUFI STOP

PAGE 81

63 without restriction. After mappi ng the streets, I d ecided to number every intersection and endpoint (where a street e nds) in the community. The observation sample, then, was based on a regular grid of intersections and endpoints around which Villa Israel’s various public places are situated. Since sports fields always bordered stre ets and intersections, they were included automatica lly. Using the original map with extended boundaries, we determined that the community ha d 60 intersections and endpoints. Drawing a sample for the interviews was much more complicated. Ideally, I would have drawn a random sample of Villa Isra el’s households, but could not find an appropriate sampling frame. I considered c onducting a census to determine the location of households. However, since community memb ers were still quite ambivalent about the research, any attempt to do a census would have been poorly received—and would have yielded poor-quality data. Ultimately, I decided to draw a random sample of occupied structures, since structures c ould be observed unobtrusively. Starting again with land sale map, we plotte d Villa Israel’s constructions. I decided that, to save time, we would eliminate constr uctions that clearly we re not occupied from the interviewing sampling frame. However, hous ing constructions in various stages of completion—without doors, windows, complete roofs, or electricity—could be occupied, or not. Wilda and Richard suggested we use th e following criteria to code a construction as occupied: the construction has a roof and at least 3 wa lls (does not need windows) AND (1) has an electrical hook-up or electrical wire run from another residence, OR (2) has a visible water storage area, OR (3) has curtains, toys, or othe r visible items that indicate inhabitation. If we were unable to determine the above, we tried to find a neighbor to consult. If, using this rubric, we were unable to determine if a construction

PAGE 82

64 was occupied, we conservatively coded it as such and included it in the random sample. Based on these coding rules, we estimated that there were about 415 occupied constructions in Villa Israel. By mid-March, the research team had crea ted its own map of Villa Israel. Starting with an out-of-date map depicting saleable land plots, we redefined community borders, identified 60 public places, and located 415 o ccupied constructions in Villa Israel. The modified map of public places and occupied c onstructions served as the master frame for the random samples we drew during the res earch. Although new constructions were built and old ones abandoned during the time we worked in the community, our random samples represent the community as it was in March 2004. Interview Sampling, Informed Consent, and Compensation For the household interviews, we drew a ra ndom sample of 110 of Villa Israel’s 415 occupied constructions using a table of random numbers (from Bernard 2002). Once we identified a randomly-sampled occupied construction, we made an appointment to speak with the household members. We define d a “household” as a group of people who shared living quarters, expenses, and a c ooking pot. We discovered that, occasionally, more than one household occupied a construc tion that we had selected in the random sample. If one of the occupants was the owner, we selected his or her household for the study. If not, we asked the occupants to deci de among themselves who should participate. Although I would have preferred to randomly sample the households within the construction to determine a participating household, intrahousehold sampling was not something that community members were willing to participate in during the early stages of the research.

PAGE 83

65 On our first round of contacts, we found th at 13 of the constructions in the random sample (12 percent) were actually not i nhabited. The remaining sample contained 96 eligible occupied constructions. We attemp ted to contact adults at each of the 96 occupied constructions up to 14 times. From the 96 occupied constructions, 73 households (76 percent) agr eed to participate in the study. Another 23 households (24 percent) refused to participate or were never available to speak with us after 14 attempted contacts. Here, I do not separate out “refusal s” from “never availables” because Villa Israel residents used avoidance as a way to communicate refusal to participate. Once we had identified a household that wa s interested in par ticipating, we sought informed consent. In most cases, we attempted to get informed consent from all adult household members, since data would be co llected on everyone in the household. We told people that I was from the University of Florida, and received academic grants to work in Villa Israel. We explained that I was writing a book about people’s daily lives, and especially about how people were affected by water scarcity. We told them that we would compensate them for the time they spent talking to us, since we knew they might have to miss work. We also told them that they had the right to refuse to answer questions or withdraw from the study at any time, but that we would like to interview them five times over the course of a year. The compensation we offered was importa nt to participants. We gave each household 20 Bolivianos (or $2.50 USD at 2005 ex change rates) for each round of data collection in which they participated. This sum represents average compensation for a day’s work among Villa Israel residents. To be interviewed, some participants had to take off a full or half day’s work. Without reimbur sement for this lost income, the neediest

PAGE 84

66 participants would not have been able to take part in the study. Participants also saw the offer of compensation as a sign of respect. By recognizing that participants’ time and knowledge are valuable, we showed participan ts that we valued them. The combined effects of economic and emotional incentiv es explain, in part, why the quality and quantity of information we received in interv iews was so much better than what I got from participant-observation. For the interviews themselves, we spoke only with people who we defined as “household heads”, that is, people responsib le for the acquisition and distribution of household goods. Of the 102 people identified as household heads, 69 percent were women and 31 percent were men. Of the wo men, 59 percent headed households alone and 41 percent headed households with another person. Of the men, 19 percent headed households alone and 81 percent headed househ olds with another person. It is important to note that, in several cases, joint-headed hous eholds were headed by parent-child pairs, siblings, and other non-married teams. In March and April of 2004, the research team identified inte rview participants. First, we located the occupied construction that was randomly selected from the sampling frame. We approached the occupants, select ed households, and explained the research process to household adults. We also offered fair compensation for household participation. The participation rate for th e first interview was 76 percent of randomly sampled occupied constructions. Because we conducted interviews with people who met our definition of household head, the major ity (69 percent) of people we interviewed were women.

PAGE 85

67 Developing Interview Protocols The interviews were designed to produce testable data on social interactions, economic exchanges, and water provision. The protocol had to be carefully constructed, so that it produced accurate and reliable data. I also wanted to contribute to the literature on the urban poor, and searched for a semi-str uctured survey protocol that could be reproduced in Villa Israel. Stack’s study (1974) of survival strategies among the urban poor contains such a protocol. The research assistants and I translated th e Stack protocol directly into Spanish, and from Spanish to Quechua. We ran a series of te st interviews, and dete rmined that some of the Stack’s categories and prompts did not tran slate well into the Andean context. Over a period of several weeks, we drafted and tested modificati ons to Stack’s categories and prompts. Most of the changes removed re ferences to Stack’s ethnographic context (African-Americans in the 1960s) and added refe rences to life in an Andean squatter settlement. For instance, in the set of questi ons about social interactions (which Stack called “Daily Life”), we added categories that probed for interactions at soccer games, community work groups, and other activities particular to life in Villa Israel. Similarly, in the set of questions about economic excha nges (which Stack called “Finances”), we added questions about marketing, water fe tching, and household economic activities. Unlike Stack, I planned to quantify and co mpare social interactions and economic exchanges. That made improving participants’ recall crucial, as I needed them to give us reliable estimates of the number of times they engaged in certain activities over the previous week. The participants were aske d to recall activities over a week-long period because test interviews showed that othe r standard time periods—a day or a month— produced idiosyncratic and unreliable data. We found that most people accessed their

PAGE 86

68 Table 4-1. The semi-structured intervie w protocol, modified from Stack 1970. Semi-structured interview protocol (Modified version of Stack 1974) DAILY LIVES My aim is to learn how people spend their time from the moment they wake up in the morning until they go to bed at night. a. Describe a typical day in great detail. (Probe repeatedly using the following categories -learn who they visit, which relatives they see daily or weekly, what they do for each other, whether they exchange goods and services, and how these exchanges are arranged.) Researchers—find out what identity groups Ego has in common with each person they name—religion, kin, paisano, neighbor, work, politics, etc. b. Who do you see or visit each day, each week? Which relatives? Which neighbors? (Name relationship.) c. With whom did you spend your day? d. Where and with whom do you eat breakfast, lunch, dinner? (Probe.) e. What housework do you do (shopping, scrubbing, cooking, dishes, etc.)? What other work do you do? With whom do you work? f. With whom do you enjoy spending time each week? g. With whom do you participate in group ac tivities (church groups, soccer or basketball, paid work, community work projects, etc.)? h. Who do you sit or visit with normally at neighborhood activities—soccer games, church, neighborhood committee meetings, other events? FINANCES Everyone has a hard time making it on the money they get and has to get some help from others. The aim is to try to figure out how people make it. This gets very complicated because some people live together, others eat together, and others share their income. a. Who is living in this house right now? (List relationships.) Who contributes to the finances of the household? How do they contribute (rent, utilities, etc.)? b. Who fetches the water? Who drinks or uses it? Who helps pay for it? c. Who ate in the household in the last week? Which meals? Who paid for the food and cooked? d. Try to learn the source of income of everyone who contributes to the household. e. What did you do for someone else this week? Did anyone help you out? (Probe for people outside the household. Get all th e information you can about these people.) f. Did you give anything (goods/services) to any of the individuals listed in e? g. Did you receive anything (goods/services) from the individuals listed in e? h. Did you trade food, money, child care, or anything else with anyone this week? With whom? i. What else should we ask you about these relationships? THE ACQUISITION OF GOODS (Elicit the names of all the items –furniture, pictures, radios, etc.—in each room in the house that were acquired in the last week. As k the following questions about each item.

PAGE 87

69 Table 4-1. Continued. a. Give a physical description of the item. b. Does it belong to anyone in the house? Who? c. Was the item in anyone elses home before? Whose? d. Was it a gift or a loan? Who loaned or gave it to you? e. Is it home-made? Who made it? f. Where did it come from? Was it bought at a store? Where? Who bought it? g. Who will it be given to or loaned to? h. What else should we ask you about it? CHATTING AND GOSSIPING a. How do you keep up on whats happening with people you dont see very often (in, outside Villa)? b. Who do you chat with? c. How much time did you spend chatting yesterday? On most days? How much do people chat with each other, in general, in Villa Israel? (Get examples.) d. How many people chat together at a time? e. What do people chat about? (Get examples.) f. How fast does information spread? If you te ll a friend something, how long does it take for your (sister/cousin/other rela tive named earlier) to hear about it? g. Are there people you dont like to chat with? Whowhy? (Get examples.) h. Do people gossip very much in Villa Israel? i. How much do people believe the gossip? j. What kind of people gossip the most in Villa Israel? What do they talk about? k. When people gossip about each other, why do they do that? l. Can you think of a time when gossip really ch anged the way people in Villa think about someone? Is there someone people tend to avoid now, because of something they heard from others? What happened? CONFLICTS a. Have you ever seen people in Villa Israel ge t into arguments with people outside their family? What happened? b. When people have trouble with people outside their families, what is it usually about? What usually happens? What are the conseque nces? Do other people get involved? (Get examples.) c. When people have conflicts outside the fam ily, do other people in Villa usually find out? What do they do when they find out? d. Have you had any trouble with anyone outside your family in the last week or so? What was it about? Did anyone else get involved? How was it resolved? (Probe as much as possible. Ask about more conflicts). e. When people have trouble with people inside their families, what is it usually about? What usually happens? What are the conseque nces? Do other people get involved? (Get examples.) f. When people have conflicts outside the fam ily, do other people in Villa usually find out? What do they do when they find out? g. Have you had any trouble with anyone inside your family in the last week or so? What was it about? Did anyone else get involved? How was it resolved? (Probe as much as possible. Ask about more conflicts).

PAGE 88

70 memories easily when we asked about the type s of interactions they had engaged in, such as visits, work, eating, casual encounters, meetings, and having fun. For people who had trouble remembering, we developed additional sets of probes based on the following native categories: relationships (family, frie nds, neighbors, etc.), places (home, on the street, at work, in the market, etc.), meals (breakfast, brunch, teatime, etc.), community activities (Neighborhood Council mee tings, soccer games, etc.), and days of the week. Interviews were conducted with house hold heads five times over a 10-month period, from April 2004 to January 2005, and to ok an average of 65 minutes. Like Stack, we began each interview by asking people to describe what a normal weekday was like for them. We then used this ba seline understanding of the participant’s lifestyle to modify the interview protocol to the participant’s particular social and economic situation. Next, we asked about the participant’s househol d economy—who was living with them, who contributed money and labor, how they divide d responsibilities, where and with whom they worked. After that, we asked people about the visits they made and received, the meals they ate, and what they did with th eir free time in the community. Throughout, we constantly probed to determine the number of social interactions and economic exchanges in which participants had participated over the last week. During the first round of household inte rviews, from April to May 2004, we interviewed 102 household heads. From June to July, we repeated the interviews with 94 household heads. In August and September, we interviewed 93 household heads for the third time. We conducted the fourth interv iew with 91 household heads in October and November. Finally, in December and January, we conducted the fifth and final interviews with 96 household heads. In total, we conduc ted 476 semi-structured interviews in Villa

PAGE 89

71 Israel. Participating household heads were lost from the sample because they traveled, moved away, had an illness or disabilit y, or later withdrew from the study. After each interview was conducted, we recorded and coded it. Because participants were not comfor table with the use of record ing devices, we took extensive notes during the interviews. We used our notes to create two-page records of the interviews after they were finished. We coded the interviews on 90 variables: 10 variables about the intervie w process, 10 demographic variables, 36 variables about social relations, and 34 variables about economic relations. The variables that are relevant to the theoretical question unde r study here will be analyzed in the chapter on social interactions and economic relations. Measuring Water Provision and Scarcity Ideally, any study of water scarcity coul d begin with a simple measure of how much water people have coming into their homes. However, when water distribution systems are highly complex, as in Villa Israel it can be difficult or impossible to track how much water people ac quire. In such cases, innovative, locally-appropriate approaches to estimating water provision must be developed. To begin, I did extensive participantobservation and ethnography to understand how households obtain and use water. People in Villa Israel acquire water from five sources: freelance water trucks, the community tap-stand system rainwater, the river, and other community members. They store water in three kinds of containers: underground water tanks, metal 200-liter drums, and 20-lite r buckets. People use water in thirteen ways: (1) face washing, (2) hair washing, (3) tooth brushing, (4) bathing, (5) making breakfast, (6) making lunch, (7) making dinner, (8) making beverages, (9) dishwashing, (10) toilet flushing, (11) bathroom cleaning, (12) washing laundry, and (13) mopping.

PAGE 90

72 I designed a data collection tool to dete rmine the degree to which each household used these water sources, water storage contai ners, and water resources. Because the tool would be used by researchers and participan ts, many of whom cannot read, it had to be illustrated. I brought an artist to Villa Israel and asked her to observe how people use water there for two days. I then asked her to draw a very simple depiction of each of the water tasks, containers, and sources. The sketch es were then lined up together on a piece Figure 4-2. The first page of the water data co llection tool. It depict s six water-use tasks (face washing, hair washing, tooth brushing, bathing, making breakfast, making lunch), two kinds of water c ontainers (10 L buckets given to participants and cups), and five water sources (river water, tap stand, rainfall collection, aguatero neighbor). Artwork by Ashley Yoder, 2004.

PAGE 91

73 Figure 4-3. The second page of the water data collection tool. It depicts six water-use tasks (making dinner, making beverages, dishwashing, toilet flushing, bathroom cleaning, washing laundry, and mopping), water containers, and five water sources. We asked respondent s to color in the correct number of buckets and draw an “X” over the corr ect water source. Artwork by Ashley Yoder, 2004. of paper. This two-page chart became the basic format we used to collect the data for two water measures, a self-administered direct measure and a task-based estimate of water use. The first measure, the self-administered direct measure of daily water use, was collected for each household on one randomly-a ssigned day in each of the first four interview cycles. One day before the meas urement was to take place, we went to participants’ houses, gave them a ten-liter bu cket, and asked them to measure and record all of the water they consumed for the thirt een tasks. We showed them the chart, and

PAGE 92

74 explained that they should record the amount of water their families used and where the water came from by coloring in the appropriate drawings on each of thirteen tasks. We scheduled the household’s interview for th e day following the water measurements. During the interview, we reviewed participants ’ water charts, verifying that all responses were correctly recorded. The second measure, the task-based estima tion of weekly water use, was collected directly after we verified data for the self -administered measure. Once people’s memories of water use had been jogged, we asked them how often and with what quantities of water each member of their household perfor med each of the thirteen water tasks on the list in the last week. Hous ehold heads gave very thor ough descriptions of household water use on each of the thirteen tasks. Gi ving a typical answer, one woman reported that, in her seven-person household, four people bath e one time a week with 10 liters of water, and three people bathe two times a week with 10 liters of water. In each interview, we probed until participants reported their househol d water use on each task with this level of detail. After recording the task-based estimation, we collected data on a third measure, the overall estimate. Because most household h eads consider water quite expensive and difficult to acquire, I hoped th at they would be able to recall how many times they had acquired it over the last wee k. Also, since water containers are visible, I thought that household heads would monitor provision levels much in the same way that I watch the gauge on my truck’s gas tank dropping each day. To get the overall estimate, we asked the question “how much wate r do you believe your househol d consumed in the last week?” directly, without the use of additional prompts or probes.

PAGE 93

75 In addition to the water use measures, we also collected data on the experience of water scarcity. Based on data from participan t-observation and interviews, I created a list of 33 questions about the experience of wate r scarcity. The English and Spanish versions of these questionnaires are pr esented in the chapter on wate r scarcity. For each question, we asked if the household head had the water-re lated experience in the last week, or not. We asked household heads the same 33 questions over four interview cycles. These data form the basis for the water experience scale, which I discuss in greater depth in the chapter on water scarcity. Because we could not directly meas ure the amount of water coming into households, I devised four diffe rent measures of household wa ter provision and scarcity (self-administered direct measure of daily wate r use, task-based estimate of weekly water use, overall estimate of weekly water use, and 33 questions a bout the experience of water scarcity). Data on water storage space and access for each household were also collected. These six measures were all created from pa rticipant-observation data, and were adapted to the specific needs and circumstances in Villa Israel. In the chapter on urban water scarcity, I will discuss each of these measures in greater detail. Direct Observation In my original design for this research, I argued that observation data should be collected to test the quality of recall-based interview data. For each interviewing cycle, three hours of observations were to be done in every partic ipating household. The observations would have yielde d reliable data that could be compared with recall-based interview responses about water use, soci al interactions, and economic exchanges. During the informed consent period, however, it became clear that most participants

PAGE 94

76 would not tolerate household observations, and those who did would modify their behaviors so extensively that the data would be useless. Instead of household observations, I decide d to do direct observ ations of social interactions in public places. While these observations would not bear directly on the interview data, they would provide complement ary information about social interactions and economic exchanges. Moreover, the obs ervations could be conducted unobtrusively, so we could record observations with a reasonable degree of confidence in the authenticity of observed pers onÂ’s behavior (cf. Altman 1974, for a thorough discussion of validity and reliability in direct observation). Observations were carried out in four 42-day cycles over 8 months. This schedule worked in tandem with the two-month in terviewing cycles; for each 60 days of interviewing, we did 42 days of observati ons. Once we had the sampling frame of 60 public places (intersections and endpoints), we used a random number table to select 42 sets of 12 public places. For each of the four observation cycles, we repeated the same sets of 12 observations over the 42-day period. Each of the 42 interviewing days was assigned a randomly selected time drawn from 12 daylight hours (between from 7 am and 7 pm). Observations of the 12 randomlyselected public places, then, took place dur ing a randomly-selected hour-long block (e.g., from 7 am to 8 am). All 12 observation point s had to be observed within a 60 minute time span during the assigned observation hour. Three interviewers and the part-time observer were trai ned to conduct observations following a standardized method. Each obser ver carried a large green notebook that contained grids for entering observations, loca tions of intersections and endpoints, the

PAGE 95

77 randomized observation schedule, and a list of 40 activity codes (drawn from participantobservations). The observer was to walk to the po ints he or she was assigned in a circular direction, starting from the place where my re nted room was located. If the observers were questioned by a curious onlooker as to wh at they were doing, they were instructed to say that they were “making appointments for the interviews”—an activity that required a notebook and was often combined with observations. When observers arrived at an observation poin t, they were instructed to look in all directions to a half-block distance, or to half a river-bed, or to half a field; they were not allowed to record everything that occurred within their field of vision. Because each observation point only extended halfway to the next point, no two observation points overlapped or shared the same public space. Observers proceeded to the center of an observation point, turned 360 degrees to look at everything within view, observed for up to 5 seconds, and continued walking. When there was a large crowd at an observation point, observers were allowed to sit down n earby to record, but th e initial observation should only have lasted 5 seconds. The 5 sec ond rule was instituted to make sure that observers recorded interactions and activities that occurred over a standard snapshot of time. In spite of my efforts to make observa tion procedures uniform, there were some problems. It was very difficult for observers to record all the activities that went on in groups of six or more people. Additionally, there was some variation in the amount of detail that different observers recorded; some observers always recorded more activities than others. Finally, there was also some vari ation in our comprehension of the behaviors we observed. For instance, I was less skill ed than Bolivian observers at correctly

PAGE 96

78 identifying different activities. Because of this variation in activity coding, I have decided not to use the activity variable in this analysis. Over four cycles, from June 2004 to January 2005, we conducted 1,986 observations of randomly-selected public places during randomly selected daylight hours in Villa Israel. The same observa tions were repeated in each cycle, and so there were 496 observations repeated four times over an ei ght-month time period. Data were collected on nine variables, and will be analyzed in the chapter on soci al interactions and economic relations. Summary of Field Methods The data collection for this project bega n in January 2004 with a five-month period of participant-observation. In March, I traine d a field research team, mapped Villa Israel, and drew random samples of occupied cons tructions and public places. In April, the research team began conducting interviews w ith 102 participating household heads. The interviews produced repeated measures of so cial interactions, economic exchanges, and water use, and were conducte d over five two-month cycles. In June, we began conducting observations of 60 public places in Villa Israel. The obser vations produced data about social interactions in Vill a Israel, and were conducted over four 42-day cycles of randomized observations. The data were immedi ately coded for statis tical analyses; those analyses will be presented in th e three chapters that follow. Contextualizing the Research Design: Rethinking Community and Cooperation As I planned the research design, I put a gr eat deal of work into creating sampling frames, interviewing protocols, observati on schedules, coding schemes, and other elements needed to conduct the research in a standardized an d reproducible way. Soon

PAGE 97

79 after I began the research proj ect, I realized that I had not put enough work into adapting the research design to local traditions th at emphasize community and cooperation. With the help of my research team, I adopted new approaches to community relations and collaborative research that, ultimately, improved the projectÂ’s data collection and quality. The relationship between the research team and community members was quite bumpy at the beginning of the research pr oject. While some community members did welcome us, many more were ambivalent, fri ghtened, or hostile to the prospect of participating an American-run research proj ect. Many Bolivians believe that American involvement in Bolivian domestic politics has be en to blame for some of the most serious economic and political crises in recent Bolivia n history. In recent ye ars, Bolivians living in the Chapare-growing coca region have sa id that Americans in military uniforms burned their crops, homes, and possessions, leaving them destitute. Understandably, then, some feared that the fieldwork was part of an American intellig ence project and would lead to some new calamity for the community. As we discussed potential solutions to th is problem, Wilda argued vigorously that we needed to do more to involve the community in the research. To st art, we held a large meeting for study participants after the first round of interviews. In June, we delivered personal invitations to house hold heads, organized snacks an d skits, and planned a raffle for those who attended. The first meeti ng was a huge success: the schoolroom was packed, participants were relieved to see fr iends and neighbors seat ed nearby, and people had the opportunity to ask questions about th e interview process. Af ter the first meeting, we found that participants were more open in interviews. During the course of the year,

PAGE 98

80 we held more meetings for Villa Israel residents—and found community members were generally very happy to participate. I quickly learned that inclusiveness was just as important to the research team as it was to study participants. At the team’s insi stence, I modified the interviewing protocol to include team-based interviewing. We im plemented weekly meetings, in which we discussed problems that arose in the data co llection process. We also created “coding retreats” in the seventh week of each interview cycle, in which we took a week-long break from Villa Israel to double-check our coding. The team interviewing, weekly meetings, and coding retreats a ll helped improve the quality of the data we collected. As I mentioned in the section on the re search team, having a well-trained and unified team proved invaluable to me during this project. Team me mbers were dedicated to the project, and suggested important and appropriate modifications to the original research design. One of the most important se ts of modifications they recommended was the inclusion of cooperative a nd community-oriented elements in the research. Because these elements had not been included in pr ojects I had worked on in other ethnographic contexts, I had not anticipated the need for th em in Cochabamba. In the future, I will be sure to think more carefully about local comm unity and work traditions as I develop new research designs.

PAGE 99

81 CHAPTER 5 THE POLITICAL ECOLOGY OF URBAN WATER SCARCITY Introduction Urban water scarcity is a ra ther new topic in social science research, and few, if any, published articles have attempted to de fine or operationalize the concept. While water scarcity has been studied extensivel y in rural contexts, urban water scarcity emerges out of distinct infrastructural, political, and economic systems. Theoretical frameworks, like those posited by Sen and Homer-Dixon, help us understand how complex social-ecological systems produce ineq uities in water distribution. However, we must go further to develop a science of urban wa ter scarcity. In this chapter, I present an analytic framework for defini ng and operationalizing the urban water scarcity concept. The chapter proceeds in three broad sectio ns, and uses data from Cochabamba to illustrate how the analytic framework should be applied. In the first section, I analyze the political ecology of urban water scarcity in the Cochabamba Valley. Because socialecological systems underlie all human water distribution, an analysis of urban water scarcity should begin with a thorough discus sion of water resour ces in the region under study. Following recent case studies of ur ban water scarcity (Ennis-McMillan 2006, Swyngedouw 2004), I will also discuss how po litical and economic power shape the distribution of water in Cochabamba. In the second section, I describe water provision in Villa Israel, including the water resources, provision, and storage in the community. In the third section, I present three approaches to defining and operationalizing urban water scarcity. Using data collected in Villa Is rael, I examine urban water scarcity with

PAGE 100

82 measures of water provisi on, water security, and experi ence. Together, these three sections combine to create a detailed ethnogr aphic picture of urban water scarcity in Cochabamba in general and Villa Israel in particular. Section 1—Political Ecology of Urban Water Provision and Scarcity in Cochabamba In this first section, I examine the political ecology of water provision and scarcity in the Cochabamba Valley. First, I discuss hydrology and water resources in the area, including groundwater resources, surface water resources, and rainfall. Next, I examine the social structures that sh ape water distribution, such as Cochabamba’s municipal water service, urban residential segregation, a nd water provision systems. Although not all of the data that would ideally be used for this kind of study were avai lable, I was able to piece together an extensive analysis of natu ral, economic, political, social, and cultural aspects of water provision and scarcity in Cochabamba. Introduction to Cochabamba’s Water Situation The Valley of Cochabamba has an area of about 100 km2, and contains a number of alluvial fans, a spring zone, confined a quifers, two rivers, a la ke, and a playa zone. Despite its rich water resources, industri al and domestic pollution have seriously degraded the surface water resources, and overabstraction and pollution now threaten the aquifer system. Since local water reserves alone cannot support the valley’s population, the municipality now brings in water from outside sources to s upport 62 percent of the city’s population. However, 38 percent of the c ity’s poorest residents are not connected to the municipal system (SEMAPA in Terhor st 2003). Unconnected residents use a combination of relatively unreliable services—such as rainwater collection, private water vendors, and community water systems—to get the water they need to survive.

PAGE 101

83 Hydrology of the Cochabamba Valley The Cochabamba Valley is located on the eastern slopes of the Andes. The valley has an average elevation of 2540 meters, and ranges from a depth of 100 meters in the northern valley to 800 meters along the nor thern escarpment (UN-GEOBOL in Stimson et al. 1996). The Cordillera Tunari rises to the north of the city of Cochabamba, and forested Tunari National Park slopes down the mountain side. Beyond the foot of the mountains, at the north side of the vall ey, wealthy Cochabamba neighborhoods, dotted with parks and swimming pools, flourish in the water-rich environment. Running through central and south Cochabamba, the valley has three major surf ace water bodies: the Rocha River, the Tamborada River, and the Laguna Alay. The Rocha River bisects the city of Cochabamba, which was originally settled on the RochaÂ’s northern banks but has grown to occupy both sides equally (Ledo 2002). To the southeast of the Rocha River is Lake Alay, which is surrounded by a series of flood-prone plains. To the south of the Rocha river is the Tamborada River. Cochab ambaÂ’s squatter settle ments are located in the regions to the southeast of Lake Alay and to the south of the Tamborada River. Like many semiarid intermontane basins, th e Cochabamba Valley is the site of a large alluvial fan aquifer system. Such system s are formed when the gradient of a stream suddenly levels off, as at the foot of a m ountain, and fluvial and de bris flow sediments are deposited (National Res earch Council 1996). Over time, these fluvial deposits accumulate into a cone of sedimentary mate rial, called an alluvial fan (Lutgens and Tarbuck 1998). From the top of a fan (apex) to the bottom (base), a series of streams flow away in radial or interbraided channel be ds. Water runoff usually deposits the largest debris near the apex of the alluvial fan, a nd carries smaller sediments away toward the

PAGE 102

84 Figure 5-1. Map of the Cochabamba Valley (f rom Stimson et al. 2001, p. 1100). Alluvial fans and groundwater wells are concentr ated in the northern edge of the Cochabamba Valley. The Rocha River bisects the city of Cochabamba. Area C delineates the southern half of the city of Cochabamba. base. As a result, hydraulic conductivity tends to be highest at the apex of the fan and along channels where deposits are largest. H ydraulic conductivity declines toward the base of the fan, and away from entrenched channels (Stimson et al. 2001). Often, runoff drains from mountainsides in a number of st reams, each forming its own alluvial fan. As the fans receive more sedimentary deposits, they grow, eventually connecting into one large system of alluvial fans that abuts the mountainside. Such a landform is called a bajada. When the streams are flooded by major downpours, excess water flows down over the bajada in sheets and collects in temporary lakes called playas. The Cochabamba Valley was formed during the Pliocene era, and has since been covered with fluvial and alluvial deposits of detritus and lacustrine clay deposits. These deposits produced a valley floor layered with lenses of clay, silt, sand, and gravel (Stimson et al. 1996). To the northwest of the Cochabamba Valley, a large bajada is located along the edge of the Cordillera Tuna ri mountain range. The three major alluvial

PAGE 103

85 fans found in the bajada system are the C hocaya, Llave, and Pairumani. These three alluvial fans are located on the remote north western side of the valley, far from urban settlements and the city itself (see alluvial fans in Figure 5-1). The principal recharge areas of the alluvial fans are located at th eir apexes, in a four to five kilometer-wide area at the northwest of the valley. The most permeable portions of the bajada are located at th e apex and along the main axes of the alluvial fans. At the apex and central axis of the fans, groundwater flows more qui ckly than in other areas of the valley. In the central axis of the Cho caya fan, for instance, groundwater velocity is estimated to be between 0.3 and 0.6 m d-1. The hydraulic conductivity of the area is believed to be between 1.0 x 10-4 and 3.0 x 10-4 m s-1 (Stimson et al. 1996). A five to ten meter-thick clay cap extends from the base of the alluvial fans, down the valley, to the Rocha River. A one kilomete r-wide spring zone is located near the base of the alluvial fans, where aquifers are semi -confined In this area, the water table is approximately two to ten meters below ground surface. In and below the spring zone is an area in which artesian we lls are located (Stimson et al 1996, 2001). Artesian wells are underground water sources in which water is co nfined and rises to surface level without the use of pumps. The zone that contains artesian wells is closest to upper class neighborhoods, providing CochabambaÂ’s wealth iest residents with access to the richest water resources in the valley. The Cochabamba Valley experiences pronounced seasonal variations in precipitation. In the summer, between Decem ber and March, average monthly rainfall reaches as high as 200 mm. In the winter m onths, between May and September, average monthly rainfall falls below 10 mm. Some data indicates that average annual rainfall in

PAGE 104

86 the Cochabamba Valley is higher (470 mm) than in the south side of the valley (400 mm) (AIG 2003, Claure Periera & Asociados 2001). Figure 5-2. Seasonal variations in rainfall for the city of Cochabamba (from Vera undated). There are two additional indicators of th e difference in water distribution between the north side of the valley (where wealt hy residents live) and s outh side (where poor neighborhoods and squatter settlements are located). First, the north side of the Cochabamba Valley is forested with vegetati on year-round, while the so uth side is nearly devoid of vegetation. Second, the north side is dotted with wells, while south-side has few wells with poor output or no output. Two photos, taken just as the dry season was ending in September 2004, illustrate the stark di fference between water availability in the north and south. Since water resources in the southern zone are less studied, there is little available data on the water resources there. However, the lack of alluvial fans, wells, surface water, springs, and vegetation on the south side al l indicate that it has less water (or less accessible water) than the north si de of the Cochabamba Valley.

PAGE 105

87 Figure 5-3. Northern Cochabamba has alluvial fa ns, forests, plant life, high rises, and the residences of the cityÂ’s wealthiest people. Figure 5-4. Southern Cochabamba is arid, barren, and contains the residences of the cityÂ’s poorest people. Water Resources in the Cochabamba Valley CochabambaÂ’s water is currently supplied by well fields located in the alluvial fan system to the northwest of the valley (see well fields in Figure 5-1, Stimson et al 2001), and by reservoirs located outside the valle y. Surface water sources located outside the valley provide 40 percent of CochabambaÂ’s wa ter, while 30 wells provide the remaining 60 percent of the water supply (SEMAPA 2006). Well fields located in the alluvial fan system tend to be unconfined, less than thirty meters deep, and built in sand and gravel sequences. Those wells located closer to the city tend to be semi-confined, built in ground containing lacustrine clay depos its, and can be as deep as two hundred meters (Stimson et al 1996).

PAGE 106

88 SEMAPA is CochabambaÂ’s municipa l water company. SEMAPA currently administers 56,148 connections to the potab le water distribution system and 58,046 connections to the sewage system (SEMAP A 2006). In 2000, control briefly passed to Aguas del Tunari, a consortium led by the Bech tel Corporation. After months of protest against its sale, the water system reverted back to municipal control. Since then, SEMAPA planners have worked to improve problems in the existing water system. The SEMAPA system had three major problems in 2004: water loss, contamination, and overabstraction. One of the most pressing problems in the Cochabamba water distribution system is water loss; an estimated 55 percent of the volume produced is lost to leakage and theft (SEMAPA in Terhorst 2003). For instance, the total output of water produced from all sources in 1998 was 19,259,120 cubic meters (Cen tro de Investigacin Multidisciplinario 1999). Total water loss, then, amounted to more than 10 million cubic meters in 1998 alone. Since the reversion of the Cochabam ba water system to municipal control, SEMAPA has focused primarily on improving the water loss problem. One particularly successful program offered amnesty to hous eholds with illegal water connection (Terhorst 2003). Another major problem in Cochabamba is water contamination. SEMAPAÂ’s wastewater disposal and sewage treatment se rvices cover only 55 pe rcent of the cityÂ’s population. As a result of improper waste disposal, large septic ponds containing untreated wastewater collect in impoverished areas of the c ity (Ledo 2002). In addition to the sewage leaked by these ponds, septic tanks and the municipal sewage system have significant leakages. For instance, recent tests of groundwater in the southern and eastern

PAGE 107

89 areas of the Cochabamba Valley show that ni trate (NO-3) concentrations have almost reached the regulatory drinking limit of 45 mg l-1 near the city (Stimson et al 2001). Although there are clay lenses distributed throughout the floor of the Cochabamba Valley, the alluvial fan systems are not sepa rated hydraulically from pollution sources. Because the valley aquifers are interconnecte d, contamination from point sources in the city threatens water quality in the entire system (Stimson et al 2001). A drought in the 1990s compounded the effects of over-abstraction, causing significant drops in the potentiometric surfaces of valley aquifers. Mo re intensive use of groundwater resources on the Cochabamba valley floor increases th e likelihood that surface contaminants will infiltrate, possibly reaching aquifers thr oughout the valley (Stimson et al 1996). The threat of spreading contamination adds ur gency to Cochabambans’ search for outside water sources. A third problem is over-abstraction, or the unsustainable withdrawal of groundwater resources. Cochabamba’s populat ion is estimated be about one million people, and is growing quickly. Within th e Cochabamba Valley, two local lakes have been drained, and groundwater levels have dropped enormously. Areas that once housed wells with a depth of only 20 me ters now require wells that re ach to over five times that depth (Finnegan 2002). The city’s rate of growth, coupled wi th over-abstraction, has put enormous additional strain on a water distribution system that has never caught up to the needs of its population (L aurie and Marvin 1999). Power, Privilege, and Water Provisio n in the City of Cochabamba Beyond its hydrological and technical probl ems, Cochabamba’s municipal water system has another serious problem—inequit y. Long ago, the city of Cochabamba was located only in the central valley, while a la rge swath of land to the south was reserved

PAGE 108

90 for agricultural development. When populat ion was very low, people say, there was enough water in south Cochabamba to live, grow crops, and keep animals. As national economic crises pushed Bolivians into cities to find work, squatter settlements have filled in the south side of Cochabamba. Just as in squatter settlements around the world, Cochabamba land speculators buy up large pieces of land, and then divide them into lowpriced residential lots (c f. Gill 2000: 55-62 for a discussion of how land speculation works in urbanizing Bolivia). Rural migrants, downwardly mobile urban families, and new residents recruited through social networ ks buy up the land, and then must work to legalize their claims to occupy the land. The legalization process now takes about twenty years from settlement to authorization in Cochabamba’s Districts 7, 8, 9 and 14, the southern zones where squatter settlements are located. Squatter settlements w ith land claims that are not yet legalized are considered “rural” by the Cochabamba Prefecture—although officials recognize that many of the communities have a high human population density, no crops, and no farm animals beyond chickens (e.g., Claure Periera & Asociados 2001). In the southern zones, even legally recognized communities and land cl aims, such as those in long-established communities like Villa Pagador and Valle Herm oso, do not entitle residents to all the rights of Cochabamba residency. Municipal wate r and sewage service, for instance, is not offered to residents of Cochabamba’s south side. SEMAPA services are currently availabl e to 62 percent of Cochabamba’s one million residents (SEMAPA in Terhorst 2003, Ledo 2002). The water system covers only the north and central parts of the city, where the wealthie st residents live. About 80 percent of Cochabamba’s water supply goes to the north and centra l zones of the city

PAGE 109

91 (Ledo in Terhorst 2003). Water provision in Cochabamba’s wealthiest households runs up to 165 liters per person per day, and wa ter costs only one percent of the household income (Ledo 2004). Water service in Coch abamba’s north side typically includes multiple indoor water taps and indoor sewage Despite the relative advantage of their water situation, residents of north and central Cochabamba are also affected by the city’s water scarcity problem. Households conn ected to the municipal system commonly experience rationing, intermittent loss of servi ce, and periods in which water is delivered for only a few hours a day (SEMAPA 2006). Most households cope with shortages by keeping water reserves on hand, which they replenish when the system is functioning. Finding safe drinking water is an even greater challenge for people whose households are not connected to the municipal system. About 38 percent of residences are located in the south side of Cochabamba, where SEMAPA does not have a mandate to offer service. This area is Cochabamba’s poorest—in which 93 percent of the population is considered poor by Inter-American De velopment Bank standards (SEMAPA in Terhorst 2003). Finding affordable water is a major burden for households on the south side. For example, Cochabamba’s poorest househol ds have as little as 20 liters per person per day, and water costs more than ten per cent of the household’s income (Ledo 2004). Households that lack access to the municipal water system must find alternative sources of water for daily use. These alternatives in clude community water systems, private water vendors, rainwater, and surface water. Among the urban poor, community-based wate r systems are often considered the best alternative to municipal service. In 120 south-side neighborhoods, Water Committees are responsible for securing wate r access for local residents (Los Tiempos

PAGE 110

92 2004). While many Water Committees have succeeded in erecting some water infrastructure, not all households can rely on water supplied by community-based systems. In my observations in south Cochab amba neighborhoods ranging from relatively new ones (like Villa Israel) to very establis hed ones (like Valle Hermoso), I found that many local water sources provided an insuffi cient amount of water for daily household needs. In some systems, water sources have disappeared permanently, dry out seasonally, or lack the water needed to support grow ing populations. Even when community water systems have water, some household heads are unable to use them because fees are too high, they lack the political connections to gain access, or queuing up is too timeconsuming. As a result, many of the poorest or most disenfranchised households have to rely on private vendors, rainwater collect ion, and surface water collection for water provision. About 19 percent of CochabambaÂ’s resident s buy water from a truck operated by a private vendor (Ledo in Terhorst 2003). In 2004, private vendors charged 4 Bolivianos ($0.50 USD, 0.020 Bs/L) for one turril which holds 200 liters of water. When households buy in bulk or have a long-standing relationship with a vendor, they can negotiate to buy a turril of water for about 3.5 Bolivianos ($0.018 Bs/L). Also, most vendors give a yapa (bonus), filling one or two additional 20 liter buckets for each turril purchased. Yapas can significantly drop the perliter price of water for those who buy in bulk, to 0.015 Bolivianos per lite r. Filling a 2000-liter tank of water at full price costs 40 Bolivianos, while the price is only 30 Boliv ianos with a discount and yapa. Since 2000 liters normally provides about two weeks of water for a household, bulk discounts can provide a 20 Boliviano per month savings, or a 25 percent discount. It is significant that

PAGE 111

93 these kinds of bulk savings are often only av ailable to households with a large water storage capacity (2000 L tanks or larger); the poorest households pay full price to fill one or two 200-liter turriles at a time. In communities where there is no municipal water service, the water vendor often provides the most convenient and reliable servic e. As a result, vended water is considered to be quite precious, and even a luxury fo r some families. Even so, people wonder and worry about the quality of water being provi ded to them. One rumor, which many people I spoke to believe to be true, is that wate r vendors fill their tanks out of the swimming pools and lakes of CochabambaÂ’s wealthiest neighborhoods. This rumor is a powerful expression of the inequities in CochabambaÂ’s water distribution system; many south-side residents believe that even the best water they have access to is only deemed swimmingpool quality in the north side of the c ity. The swimming pool rumor makes clear how deep the divide in access to wealth, economic justice, and citizenship rights is in Cochabamba. For households that struggle to buy wate r from private vendors, there are alternatives. During the summer wet season, ma ny households collect and store rainfall. However, rainwater collection requires an initi al capital investment in equipment, which is prohibitively expensive for some households For households that lack the resources to obtain water from rainwater collection, there remains only one alternative. Surface water sources, such as rivers, ponds, and canal beds can be used to collect low-quality water. People generally use this water for cleaning tasks, which offsets the cost of water provision. Minimal household water needs, wh en augmented with surface water, can

PAGE 112

94 often be satisfied with one turril of vended water or rainwater a week, which is then used only for ingestion-related tasks like drinking, cooking, and dish washing. Section 2—Water Resources, Provision, and Storage in Villa Israel In this section, I examine the complex human-environment interactions that underlie water scarcity in Villa Israel. First, I discuss local water resources, such as groundwater and surface water. Next, I assess water provision options, including surface water collection, rainwater collection, th e community water system, water delivery trucks, water loans, and water sales. Fina lly, I describe two t ypes of water storage containers and purification techniques. In each subsection, I ex amine the natural, technological, social, political, economic, and cultural aspects of water availability in Villa Israel. Local Water Resources Few, if any, published studies of Villa Is rael’s groundwater resources have been conducted. What evidence does exist indicates that water is not plentiful. In 2001, the engineering firm Claure Periera & Asociados conducted a study of Villa Israel, and found that the community’s two existing wells do not provide sufficient water to meet the needs of all residents. They also concluded that su bterranean rock layers in Villa Israel could not be perforated with the dr illing equipment available in Bo livia at the time. Along with basic sanitation organizations UNASBVI and PROSABAR, they recommended that Villa Israel build a distribution sy stem to import water from a source located 3995 meters to the north of the community. In March of 1999, a well was perforated to a depth of 177 meters. The maximum daily output from th e well was 5.832 liters per second. The well was intended to provide 80 liters of water pe r person per day to Villa Israel residents (Claure Periera & Asociados 2001). Unfortunate ly, before the water distribution system

PAGE 113

95 was built—but after the community invested somewhere between $10,000 and $29,666 USD in the project—it was revealed that the Villa Israel Ne ighborhood Council had not secured the appropriate access to the underg round water resources in a neighboring residential zone called Ictocta, where the well was located. To date, the community investment funds have not been recovered, and no water was ever provided to Villa Israel residents from the Ictocta well. During the wet season, Villa Israel also ha s three surface water so urces. First, there is a large riverbed that ru ns along the eastern edge of the community. Although it is generally dry for about eight months of the year, from March to November, it does provide a steady trickle of water during the wet season. A B Figure 5-5. Villa Israel’s river A) during th e rainy season, in January 2005 and B) during the dry season, in June 2004. In Villa Israel, there are also two canal systems. Because Villa Israel and its surroundings are quite barren of vegetation, flash flooding poses a serious threat to people and land. As a result, the community ha s built the canals to channel floodwaters in heavy rains, minimize the risk of erosion, and enable people and vehicles to move safely through the community during the wet season. The canals, however, are only full when

PAGE 114

96 the entire community is inundated with water, and thus are not very useful as a household water source. Figure 5-6. Villa IsraelÂ’s runoff canals are dry year round, except during heavy rains. Finally, there is a surface water source high up in the hills behind Villa Israel. Study participants say that th e source has water year-round. Th ey estimate that the water is over an hourÂ’s walk away. While it is no t widely used by community members, some Villa Israel residents do walk that far to fi nd clean water for bathing and clothes washing. Water Provision in Villa Israel In this section, I discuss each of Villa Israel residentsÂ’ water provision options in turn. Those options include surface water colle ction, rainwater collection, the community water system, water delivery trucks, water loans, and water sales. For each water provision option, I analyze the quantity, quality, and cost of water av ailable, as well as the social, political, and cultural issues that affect access to water sources. The analyses are supported by data collected from participant-observati on, key informants, and five cycles (10 months) of in-depth interviews with a household heads from 76 randomlysampled residential constructions.

PAGE 115

97 Surface Water Collection The only surface water source that is widely used by Villa Israel residents is the river. As I mentioned above, the river provides only a small amount of water. During the wet season, the riverÂ’s width is normally between 30.5 and 91.5 cm and its depth is between about 2.5 and 15 cm. Figure 5-7. The Villa Israel river during the wet season. While people in Villa Israel avoid ingesti ng the river water, they do rely on it for some household cleaning activities. Households located near the rive r collect buckets of water, and bring them home to wash floors and water plants. Some people bring bathing mats and shampoo, and bathe their entir e families in the river water. Figure 5-8. People in Villa Isra el rely on river water for bathing. Note the bath liner, shampoo, and clothing.

PAGE 116

98 Women also congregate in groups to chat and wash laundry when there is enough water. Even when there is no visible wate r, a few industrious women dig out small hollows, which fill with water and can be used for washing. Figure 5-9. The riverbed is a center for social interactions where women wash and chat. Villa Israel residents that grew up in the countryside see the river as a social spot, and enjoy washing and visiting there. Parents al so allow their children to splash and play in the river. Figure 5-10. Children splash and play in the river during the wet season. People also compete for the best locations in the water-scarce river. When one familyÂ’s laundry is being wa shed upriver, it visibly pollute s the water for those washing below. As a result, arguments sometimes br eak out between the wa shing groups that dot the river. Some women wake up early to go to the river alone and wash in the fresh morning flow, before it is contaminated by mid-morning washers.

PAGE 117

99 While visible pollution causes conflicts among river-users, invi sible pollution of the river is a much bigger health problem. In Villa Israel, there is no community sanitation system, nor are there minimal standa rds for wastewater disposal. As a result, wastewater is disposed of in the street, in septic tanks, or in unsealed dumping holes. Figure 5-11. Wastewater is di sposed of in unsealed dump ing holes. Here, the hole is located inside a housing compound. Many households lack bathrooms, and peopl e use uninhabited areas such as fields and the bluffs above the riverbeds as makeshift toilets. Consequently, Vi lla Israel is filled with point contamination sites. During the wet season, rains wash away feces and other wastes directly into the riverbed. Figure 5-12. The open area between these two bushes is used as an outdoor toilet by many Villa Israel residents. Wastes drain directly into the riverbed below.

PAGE 118

100 Contaminated water then collects in river and canal beds. Although people do not use surface water for drinking or cooking, acciden tal ingestion almost certainly occurs as they wash and play in the ri ver. Further, the contaminated water, when used to clean, spreads bacteria into the hous ehold via clothing, cl eaning items, and direct contact. While I have no data on surface water quality, this pr oblem likely contributed to the 24 incidents of water-related illnesses repor ted by the 76 Villa Israel household heads during the study period. Rainwater Collection In Villa Israel, households collect rain water during the wet season with rooftop rainwater collection equipment. Such equipm ent consists of slanted roofs made of impermeable materials like tile, gutters affixe d to the roofs, and a pipe or tube that extends from the roof gutters to a turril or water tank underground. Figure 5-13. Rooftop rainwater collection equi pment includes a drain from roof gutters, a turril and a pipe that extends to a water tank underground. The quantity and quality of water collected from rooftop collection systems varies enormously, and depends on the size of the r oof, the maintenance of gutters and pipes, and the cleanliness of roofs and turriles When roofs are left dirty, rainwater collection

PAGE 119

101 yields filthy brown water that barely serves to clean patios. When the entire system is carefully maintained, however, large amount s of high-quality water can be collected. In a major downpour, households can easily amass 200 liters of rainwater, which lasts between one and three days for a nor mal family. During the wet season, arguments erupt when household members forget to main tain rainwater collection equipment, and large quantities of rainwater are lost. If a household has vigilant residents and the necessary storage capacit y, it is possible to fill a 10,000 lit er storage tank over the course of the wet season. One woman managed to sust ain her family using the rainwater supply in her 10,000 liter tank until late August. Th is woman, however, was the same one who walked hours to wash her familyÂ’s laundry in the year-round surface water source, and was unusually conscientious about water conservation. Several other households were able to make their rainwater supplies last into the first months of the dry season (through mid-June ) or began using rainwater just as the wet season began (in November). Households that engage in frugal rain water use tend to be downwardly mobile. They are families that have constructed large water storage tanks, but can no longer afford to fill the tanks w ith commercially-purchased water. Poorer households (that have small water storage areas) and wealthier ones (that can afford to fill their tanks with commercial water) generally only collect and use rainwater in the wet season months of December, January, and February. Community Water System: Tap Stands Residents of Villa Israel worked hard to build a community water system that draws water from the communityÂ’s two small hillside wells. The wa ter quality is high compared to other sources in Villa Israel, although it does contain visible silt deposits. The water system currently delivers water to about ten tap stands around the community.

PAGE 120

102 Tap stands are small brick columns built around a water pipe that, when unlocked, produces a thin stream of water. A tap stand normally serves ten to twenty households in an area. Each tap stand is funded and c onstructed by the beneficiaries themselves. Because of the cost, many community zones have not erected tap stands. Some have chosen not to do so at all, while others ar e still working to orga nize the necessary funds and labor. In 2004, one new tap stand was bu ilt, and another was under construction but remains unfinished. Figure 5-14. Tap stands are small br ick columns built around a water pipe. Local beneficiaries are also responsibl e for the maintenance of the tap stands. However, community members do little to ma intain water systems and taps stands once they are installed, probably because they lack expertise, time, and money. At least one tap stand has been abandoned for years, as water flow is blocked. Be neficiaries have not organized to repair it, and have no plans to do so. Even functioning tap stands have some problems. Water leakage is high, and pools of water form on the ground above the underground pipes that carry water from the wells to the tap stands. Water output is also quite low, only allowing for a 20 to 40 liter allotment per household per day.

PAGE 121

103 Water is distributed once a day, betwee n 4:30 am and 9:00 am, from Monday to Saturday. One community member, an aged man, is responsible for water distribution. He gets up at 4:00 am, and opens the main pump fr om the wells. Then he begins his rounds, visiting each tap stand in turn. He randomizes the sequence of his tap stand visits, so that no one has to get up early or be late for wo rk every day. Before opening the tap stand, he blows his whistle loudly to al ert the zone that the tap sta nd is about to be opened. As people stumble sleepily out of houses with one or two large buckets in hand, he unlocks the tap stand. People then line up, waiting up to 20 minutes for their turn. The tap stands are one place in Villa Israel where social interactions are quite rich—people argue, gossip, and help each other out regularly. Conflicts at the tap stand usually deve lop around one issue—access. Because home owners build and maintain the pumps, only they are eligible to use the tap stands. Each month, home owners must pay 10 Bs (bet ween 0.016 and 0.008 Bs per liter) to gain access to the tap stands. While everyone knows the rules, not everyone agrees with them. Some homeowners who originally helped cons truct their zone’s tap stand no longer have the time or money to collect water each day. Even so, they feel that th ey have the right to water, and occasionally ask that the water distributor give them 10 or 20 liters of water to make ends meet. Other people, who have paid the monthly fees, ask to be given more than 40 liters of water. In 2004, there was one man who repeatedly snuck to his tap stand when it was not supervised, opened it, and took water. Outright wate r theft, however, is quite rare at the tap stands. Problems genera lly arise when one community member tries to convince others to bend the rules.

PAGE 122

104 At each tap stand, the water distributor a nd local beneficiaries enforce the access rules at their own discretion. At some pum ps, people do make exceptions for longtime residents and renters. For instan ce, several of the beneficiaries of one tap stand worship at a nearby evangelical church. One day, a fello w congregant (who happens to be a renter) asked to be granted access to the tap stand. Over the protests of non-evangelical home owners, the woman was given permanent acce ss to the water source. This situation, however, is atypical. Only a few needy rent ers belong to powerful coalitions in the community. More often, the most vulnerable people in Villa Israel—renters, working single mothers, and child-headed households —lack the status, time, and money to gain access to the tap stand system. In 2004, only 30 percent of households used the community water system even once. There are several reasons why people do not use the tap stands. First, only 53 percent of residents are homeow ners, and are officially eligible to use the water system. Even so, some of these eligible homeowner s do not live in zones that are served by working tap stands. Of those who do have access, many say that they do not want to get involved with the politics that play out at the water pumps. Other households do not have any members that can wait four hours each day fo r the water distributor to arrive, or that can carry the heavy water buckets home. A few household heads prefer not to drink “dirty water” that contains visible silt deposits. Finall y, even though the tap stand is cheaper (per liter) than vende d water, some households simply do not have a 10 Bs lump sum to invest in water at the start of the month. Aguateros, or Water Delivery Trucks Water delivery trucks, or aguateros as th ey are called in Cochabamba, are by far the most common source of water in Villa Israel. The trucks carry enormous cisterns behind

PAGE 123

105 them and, when full, hold 10,000 liters of wa ter. During the course of the study, 100 percent of Villa Israel households reported buying wate r from an aguatero. Each day, between one and four aguateros make the rounds in Villa Israel The drivers do not follow set routes, and tend instead to circulate in Villa Israel’s main streets. Each truck is operated by two men; one man drives, and th e other controls a large water hose on the back of the cistern. Only one of the aguateros lives in Villa Israel; the others come from outside to sell water. Because their water source is not located in Villa Israel, all aguateros must return to the north side of Co chabamba to refill their cisterns with water. Figure 5-15. Water delivery trucks, or aguateros are the most common source of water in Villa Israel. Note the aguatero passing through the entr ance to Villa Israel. Aguateros are independent entrepreneurs, a nd work to earn a profit. As a result, they prefer clients that can buy large quantities of water in one transaction. The largest storage tanks in Villa Israel hold 10,000 liters of water, and an aguatero can sell his water out in one transaction if he lands such an account. As a result, aguateros often give out their cellular phone numbers to owners of large water tanks. One woman who had a 10,000 liter storage capacity said that she called the aguatero whenever she needed to fill up her tank, made an appointment, and was even allowed to buy on credit. When they cannot sell all of their water in one transaction, aguateros creep down Villa Israel’s main roads—th e ones with the most houses and best driving conditions. By

PAGE 124

106 sticking to main roads, aguateros can sell wa ter to more houses without running the risk of damaging the truck on poorly-maintained ro ads with potholes, m ud, and large rocks. People who live on main roads said that getting the water they need was much easier than did people who live far away. Fo r instance, a restaurant-owner who lived directly on the main street said that she never had trouble getting water, since the aguatero passed her house every day—even though she rarely bought mo re than a few hundred liters of water at a time. People with little water storage capacity ( 200-400 liters) and remote homes have a hard time convincing aguatero s to sell them water. Since there is no scheduled route, people in need of water must wait at home until they hear the beeping horn of the aguatero in the community. Once they hear the horn, they dash out into the street looking for the water truck. It is common to s ee three or four people running through the community when the aguatero is out, trying to anticipate which street he will turn down and how they can find him. Over four interview cycles (from June 2005 to January 2005), 52 percent of household heads said that they ha d run through the streets in search of the aguatero during the last week. People can run for up to 20 minutes just to find the aguatero. Once they reach the truck, they must still convince the aguatero to come to their homes and sell them water. When people have little money to spend or little storage space to fill, the aguateros sometimes refuse outright to make the sale. Since aguateros have no formal obligati on to supply community members with water, the poor have little recourse for getting the water they need. Those who have trouble securing service from the aguatero become angry, worried, or desperate to find water. I often heard people leveling the same criticism at aguateros, “they are so creido ”.

PAGE 125

107 It is interesting that people nearly always chose the word creido (arrogant, self-important) to describe aguateros, rather than egoista (selfish, self-interested) or flojo (lazy). On face, it would appear that aguateros’ refusal to work hard to distri bute water has much more to do with self-interest or perhaps laziness than arrogance. However, from the perspective of Villa Israel’s poor, the aguatero’s offense is not simply the refusal of a service—it is the power dynamic created by the refusal that real ly angers people. On ce this power dynamic is established by the aguatero’s refusal to se ll water, Villa Israel residents only have one option— rogar In highland Bolivia, rogar is an important social tactic, used to cajole, guilt, or persuade people in just a bout any social situation. Rogar literally means to beg, but is actually a begging performance, in which a person uses a whining tone, self-effacing statements, and guilt-inducing arguments to get something done. By begging, one person shows submission to another, and shows that he or she relies only on the goodwill of the other to achieve an important goal. The la nguage used in begging establishes a clear power dynamic between two people, even if just for the moment that the favor is being requested. Villa Israel residents who cannot induce aguatero s to sell water in any other way use rogar as a last recourse. Over four interv iew cycles (from June 2005 to January 2005), 47 percent of household head s said that they had to be g the aguatero to sell them water in the last week. The tactic, while quite effective, forces pe ople to lay bare the power relationship between vendors and client s. While people understand the need for aguateros to make a profit, they resent th e humiliation involved in having to beg to buy water. For this reason, many Villa Israel residents see aguateros as creidos rather than people understandably motivated by va lue-neutral economic incentives.

PAGE 126

108 Despite the annoyances, cost, and hum iliation involved in buying from the aguatero, it is clearly the most common way to get water in Villa Israel. There are two main reasons why this is the case. First, more water volume is distributed in Villa Israel via aguateros than any other source. Since it is nearly impossible to acquire more than 40 liters of water from the other sources year -round, people must rely on aguateros to buy water for cooking, hygiene, and house cleaning. S econd, Villa Israel residents work hard to avoid social entanglements—whether cooper ative or conflictive— in the community at large. Becoming reliant on communal water so urces, such as the tap stand system and surface water, obligates people to involve them selves in social inte ractions, cooperative exchanges, and community politics. In c ontrast, the aguatero provides a market transaction that is relatively removed from the politics of daily life in Villa Israel. Clients do not have to cooperate, negotiate, or inte ract with neighbors in any way to acquire water from the aguatero. As I explain below, the ability to get water while avoiding social obligations and interactions is an important c onsideration for many Villa Israel residents. Water Loans, Gifts, and Charity When people run out of water and money, some ask nearby family members or neighbors for help. Asking for water is delicate business, and people try to avoid doing it whenever possible. Over four interview cycles (from June 2005 to January 2005), 14 percent of respondents reported receiving water from someone else in the last week, and an average of 23 percent reported giving water to someone else in the last week. The fact that 86 percent of household heads said that they did not borrow water and 77 percent did not lend water shows just how strong the aversi on to water exchanges is in Villa Israel. When asked why they did not participate in water exchanges, house hold heads said that they wanted to avoid conflicts, that neighbors do not return borrowed wa ter, or that they

PAGE 127

109 “just don’t like lending and borro wing”. Another factor may be that lenders rarely give more than 10 or 20 liters of water; water borrowing only provides a short-term solution to household water problems. The act of giving water is generally called prestar (to lend or loan) in Villa Israel. When people make water loans to a neighbor, th ey hope that she will return the water as soon as her household supply is replenished. That the ideal water loan has a quick turnaround and would be repaid in kind demons trates clearly that most households do not maintain reciprocal relationships with their neighbors. An ideal-type exchange occurred in November 2004 when one woman, the head of a large and economically-stable household, borrowed a bucket of water from her neighbor across the street. Later that day, the woman filled up her own tank and returned the borrowed water. Her prompt return of the water enabled her to avoid the establishmen t of a long-standing social obligation to the neighbor. Some families do engage in reciprocal water exchanges without the expectation that water will be repaid immediately. For instance, two impoverished households are located at the northwestern e dge of Villa Israel, far from the main street where the aguatero passes. When one of the heads of hous ehold is urgently in need of water, she always visits the other to ask for a water loan. Often, however, the other woman is in equally dire circumstances and cannot afford to help her neighbor. If she does have water to spare, she gives it. Immediate repayment is not usually an option for these women, but they are able to maintain th eir exchange relationship desp ite frequent rejections and delayed repayments. Out of necessity—not choice—these neighbors maintain a trustbased water exchange relationship.

PAGE 128

110 Water exchanges can also be part of a larger generalized reciprocal exchange relationship. That is, one household may loan wa ter to another, and later be repaid when the borrower looks after th e loaner’s house or children, brings them a gift of fruit, or does them some other small favor. People in longstanding exchange relationships say that they “loan” each other water; they also use the terms regalar (to gift) or ayudar (to help) to describe the exchanges. Such exchange s tend to exist between family members who live close by. For instance, one young couple live s next door to the wife’s parents. The couple has a water storage capacity of only 400 liters, while the parents have a tank that holds 10,000 liters. The parents offer the coupl e unrestricted access to their water tank; in return, the wife helps her moth er run her small restaurant. Complex exchange relationships can also develop between Vill a Israel residents who are economically entangled though employme nt or rental agreements. For instance, people who own home-based businesses like bake ries or leatherworking shops give their employees free access to household water re sources. Employees drink, cook, and even shower in the workplace. The employees are not charged for the use of water, nor is it considered part of their income. Instead, em ployers consider the water to be a gift. Similarly, generalized reciprocal exchanges develop between landlor ds and renters in Villa Israel. Since renters tend to live within a family’s housing compound, along with the landlord’s family, there are many opportun ities for the exchange of goods, loans, and favors. Landlords frequently offer renters some or complete access to household water resources. In return, renters keep an eye on the house and children, help out around the house, or loan food and other goods to the landlord.

PAGE 129

111 Thus far, I have described water excha nges that work—between people who create relationships with clear exp ectations and consistent beha viors. However, when people engage in water exchanges that disappoint expectations, all ki nds of tensions and conflicts can develop. Because most people wi ll not ask for water unless they are truly desperate, there is a strong prohibition against refusing water to a neighbor. As a result, many people go to great lengths to avoid bei ng asked for water. Some people do not talk to their neighbors at all, while others ignore the door when they hear knocking. Those who cannot avoid their neighbors become reluctantly involved in water exchanges. One impoverished elderly woman ow ns a tiny corner store. Because she must attend her store, she feels forced to loan water when people come in search of a loan. Another evangelical family ha s become a water patron to a needy neighbor. Because their church encourages charity, the evangelicals grudgingly “l oan” water to the neighbor several times a week—even though they know that such loans will never be repaid. Among family members, too, resentments can develop around water loans. One woman lives with her husband, children, and the husba nd’s aunt. The aunt, who is supposed to maintain a separate household economy, take s water from the family’s tank without asking permission or repaying the loan. As a result, a nasty three-way dispute arose between the woman, her husband, and the aunt—and remains unresolved. If people who feel obligated to give wate r become resentful, those who are refused water take much greater offense. In three separate families, parents abandoned their children in Villa Israel. The eldest children, ranging in age from 16 to 11, were left to fend for their younger siblings. Desperate to ma ke ends meet, the child household heads ran out of money and water every few weeks. Eventually, neighbors began to refuse their

PAGE 130

112 requests for water loans, saying they had nothing left to give In two of the cases, the families got into bitter arguments over the refused water loans. No adult household head admitted to being refused a water loan. This is probably because most household heads would not ask fo r a water loan from a neighbor that they believe would refuse them. Several people did say they avoided asking around for water loans to prevent conflicts. Rather than ask for a water loan from someone that might refuse them, household heads said they prefer to go without water, ask for a favor from the tap stand distributor or buy water from a business or neighbor. Buying Water from Businesses and Neighbors When household heads wish to avoid giving and receiving water loans, some rely on monetized water exchanges instead. Over four interview cycles (from June 2005 to January 2005), 6 percent of the 76 household heads reported that they had bought water from someone in the community in the last week. Over the same study period, 2 percent of household heads reported selling water to some one else in Villa Israel in the last week. There were probably more buyers and sellers because there are several local businesses that sell water to multiple clients. In Villa Israel, I knew of five businesses that sold water: a public shower, three corner stores, and a restaurant. Although I sa w a sign advertising th e public shower, I was never able to locate it. Only one family ever said that they used Villa IsraelÂ’s public shower during the entire course of the study. More common was the purchase of water fr om a corner store. One popular corner store, located near the entrance of Villa Israel on the main street, sold 10 L of water for 0.50 Bs (0.050 Bs/L). While the store-bought wate r was much more expensive than water purchased from any other local source, people did seem to accept that this was the going

PAGE 131

113 market rate. I only heard one accusation of price-gouging in Villa Israel. A woman went to her corner store to buy a bucket of water, and found the store closed. She approached the proprietress of a nearby re staurant, and asked her to se ll 10 liters of water. The proprietress offered to sell the water for 1.00 Bs (0.100 Bs/L), and the woman became incensed. She could not believe that someone ha d the nerve to sell 10 liters of water at double the retail rate. Although she bought the wa ter that day—her family had not bathed or eaten cooked food in 24 hours—she swore ne ver to patronize the restaurant again. Figure 5-16. People buy water from businesses like public baths, corner stores, and restaurants. Here, a stand sells fru it juices and perhaps water from the turril in the background. People also reported buying and selling wate r with neighbors. There did not appear to be any one situation in which buying and selling water was appropriate. In some cases, people said they bought water because they did not know anyone with water to lend. In other cases, people offered m oney to neighbors who extended a “gift” of water. In both cases, it appears that househol d heads volunteered money to avoid conflicts (that could arise out of a neighbor’s refusa l to lend water) or to avoid so cial obligations (that come with accepting a true gift). The child-headed households, too, had considerably more

PAGE 132

114 water purchases than other households. Intere stingly, no one said that they resented having to buy water from stor es or neighbors—they only re sented being overcharged. Water Storage and Purification As I showed in the section on water provi sion, a household’s wa ter storage capacity determines how much water a family has, how long it lasts, and on which water sources they can rely. All households have small water storage vessels: baldes (10 liter buckets), tachos (20 liter buckets), and baadores (wash basins). Small water storage vessels facilitate surface water collec tion, use of the community wate r system, water loans, and water sales. However, for rainwater collection and aguateros—the two sources that provide the bulk of Villa Israel’s water—large-scale water storage space is vitally important. There are only two types of large-s cale water storage containers used in Villa Israel: tanks and turriles Water tanks are concrete boxes that are de signed to hold large quantities of water. The tanks are built underground, and are acce ssed through a small, square hole with a metal lid. The access hole is installed in a house’s outdoor patio, above the water tank. To access the water inside the tank, people generally lean into the access hole with a bucket, and scoop water out by hand or with a rope. Th ree households in the study had tanks with engine-run pumps, which supply in-house taps with running water. Water tanks in Villa Israel hold between 1800 and 10000 liters of wate r. In all, 37 percen t of Villa Israel’s households have water tanks. The average stor age area in Villa Israel’s water tanks is 5474 liters (SD = 3390). While many people aspire to have the larg e water store areas th at tanks provide, not all can afford to construct them. Such a proj ect, of course, requires homeowners to make a large investment in their property—and in staying in the community. Building a tank

PAGE 133

115 costs about $1000 USD, and ta kes at least two construc tion workers to complete. Construction quality is very important, as the tank should not have cracks, an inclined floor, or access problems. Tanks can be quite difficult to clean when they become dirty, as they must be drained and scrubbed by ha nd. As a result, some people allow dirt and household items (like toys and pens) to accu mulate at the bottom of water tanks. After tanks, Villa Israel residents have only one other large-scale water storage option— turriles Turriles are cylindrical containers that hold 200 liters of water. Some households use only turriles to store water, while others supplement water tanks with one or two additional turriles In all, 63 percent of Villa Israel households have turriles Households have between 200 and 1000 liters of turril storage capacity—or between one and five turriles For households that had at least one turril the average turril -based storage space is 332 liters. No househol d in Villa Israel has less than one turril or 200 liters, of water storage capacity. Unlike water tanks, turriles are kept above ground level. Most turriles are placed near the edge of housing compounds, where the aguatero can easily reach them with his hose. In some homes, turriles are also kept near the house, so that rainwater collection equipment can empty into the vessel. Turriles are relatively easy to clean because they are not deep. However, many households do not cover the vessels, and they are left open to dust, dirt, animals, and other environmental contaminants. While turriles can be made of metal or plastic, most in Villa Israel are second-hand meta l ones that cost about 80 Bs. Some of these turriles are in very bad condition—they are rusted, covered in tar, or were once used to carry flammable or poisonous mate rials (and still carry the warning signs).

PAGE 134

116 The contamination of stored water appears to be a serious problem in Villa Israel. People normally do little more than skim off vi sible dust and let dirt settle to the bottom of water containers before bringing the water to the kitchen to wash dishes, clean salad ingredients, or cook. However, since Villa Isra el residents do not consider plain water to be an appetizing beverage, most people drink mate (boiled herb tea) or moqochinchi (water boiled with dried peaches, sugar, and cinnamon) rather than unboiled water. As a result, most drinking water is consumed in a safe, boiled manner. Recently, a group from the Universidad Mayor de San Simon in Cochabamba began promoting the use of an alternative wa ter purification techniqu e, called SODIS, in Villa Israel. Using the SODIS method, people expose a small amount of drinking water to sunlight in a clear 2-liter soda bottle for 5 to 48 hours (d epending on the weather). The process uses the radiation from solar energy to inactivate microorga nisms such as fecal coliforms (EAWAG and SANDEC undated). Although the SODIS technique has been promoted through the local health center, it is only used by a few households and does not seem to be catching on am ong Villa Israel residents. Section 3—Defining and Operationaliz ing the Urban Water Scarcity Concept In the last section, I gave many examples of the ways in which the natural, technological, and social aspect s of water distribution affect people in Villa Israel. The community lacks clean and productive wate r sources. Access to water resources has become monetized and politicized. Household levels of water provision are mediated by factors like socio-economic status, housing location, and water storage capacity. People who do not have enough water feel stresse d, get sick, and even forgo washing and cooking. All of these phenomena are asp ects of one overarching problem—water scarcity.

PAGE 135

117 Water scarcity is a concept that, while it may seem intuitive, is not easy to define. I argue here that water scarcity can be understood in three ways : (1) the absolute lack of water needed to stay healthy and clean, (2) the inability to secure access to a stable source of water, and (3) the experience of feeling deprived of water. The challenge that I take up here is to create a series of measures that captures all thes e dimensions of water scarcity. In this section, I will explore three differ ent approaches to studying water scarcity. The first approach simply determines how mu ch water people have. The second approach assesses the security of peopl eÂ’s access to water sources. Th e third approach examines water scarcity as an experiential phenomenon th at should be studied within its specific cultural context. Each of thes e three approaches is incomple te without the others, and all are needed to understand urban water scarcity. Approach One: Absolute Mea sures of Water Provision Any study of water scarcity should begin with an absolute measure of water provision, such as the amount of water a commu nity, household or person has. Absolute measures enable us to compar e levels of water provision with baseline biophysical water requirements for health and hygiene. They also provide valuable data that can be used to compare levels of water provision within communities, across cultures, or across historical periods. The minimal human water requirements for drinking and household use in lowtechnology situations have been established by a number of international studies. Adults need to consume, on average, 5 liters of clean water daily to survive in tropical climates and 3 liters in temperate climates. To mainta in minimal health standards, humans need about another 45 liters of water a day: 20 lite rs for sanitation, 15 lit ers for bathing, and 10

PAGE 136

118 liters for cooking and other household chores (Gleick 1996). In total, then, adults need about 50 liters of clean water a day to stay healthy and clean. Collecting data on water pr ovision can be quite difficu lt when water sources are distant, have different distribution mechan isms, or are used by many household members. In her work in rural Uganda, Sugita (2004: 58-60) used her know ledge of local water collection to develop a sophi sticated recall-based method for measuring household water provision from local wells. In Villa Israel, however, water sources were too numerous and the recall task too complex for us to coll ect provision data. Instea d, I chose to rely on recall-based water use measures. While such water use measures are only a proxy for water provision, they are a star t. In the future, researcher s should develop better recallbased measures and, ideally, observation-base d measures of water provision that are appropriate for use in complex urba n water distribution systems. As I explained in the chapter on field met hods, I used three different approaches to estimate how much water households in Villa Israel had: a self-administered direct measure, task-based estimate, and overall es timate. Each of the approaches was designed to measure the same concept and, if they are truly accurate, should produce the same perperson per-day estimate of water provision. Here, I describe the three approaches, and discuss the results of each one. The self-administered direct measure For the first measure, the self-administered direct measure, we asked participants to record all of their householdÂ’s water use on an illustrated water use chart. Participants measured household water use on four randomly-selected days over an eight-month period, from April to November 2004. Once I calculated the amount of water consumed by household members in one day, I divided it by the number of people living in the

PAGE 137

119 household that day. This provided a rough estima te of the amount of water allocated to each household member that day. Because daily water use is quite idiosyncratic, a measure of daily water use cannot be used to compare household provision levels. When averaged over time, however, it does yiel d an estimate of average water use across households in Villa Israel. For the 72 households that participated in the self-measure, the household members used an average of 44.8 lite rs (SD=27.3) of water per person per day over the eight-month study period. The minimu m average of water that any household had over the eight-month period was 6.6 liters per person per day. The maximum amount that any household had over the eight-mont h period was 189.5 liters per person per day. The task-based water use estimate The second measure is the task-based wate r use estimate. After we asked household heads to verify their water use self-measures, we asked th em to estimate the amount of water that each household memb er spent on thirteen household tasks over the course of a week. The thirteen tasks are: (1) face wash ing, (2) hair washing, (3) tooth brushing, (4) bathing, (5) making breakfast, (6) maki ng lunch, (7) making dinner, (8) making beverages, (9) dishwashing, (10) toilet fl ushing, (11) bathroom cleaning, (12) washing laundry, and (13) mopping. For each task, we asked household heads how often each household member normally performed the task over a week-long period, and what amount of water they used. Fo r the 74 households that part icipated in the task-based estimate, the household members used an av erage of 32.9 liters (SD=18.2) of water per person per day over the eight-month study peri od. The minimum average of water that any household had over the eight-month peri od was 10.4 liters per person per day. The maximum amount that any household used ove r the eight-month period was 136.5 liters per person per day. The task-based estimate, then, produced estimates of household water

PAGE 138

120 provision that are 11.9 liters smaller, on average, than those produced by the selfmeasure. The overall estimate of weekly water use The third measure is the overall estimate of weekly water use. We asked household heads to estimate the overall amount of wate r their household consumed in the last week, one time during each of five interviews over a ten-month time period. Since the study compares households, not household heads, I averaged household head sÂ’ estimates into one household-level statistic (for households with more than one head). I then produced a per-person per-day average for each household using the five estimates. For the 76 households that participated in the task-bas ed estimate, the household members used an average of 20.1 liters (SD=12.5) of water per person per day over the ten-month study period. The minimum average of water that any household used over the eight-month period was 4.4 liters per person per day. The maximum amount that any household used over the eight-month period was 88.1 liters per person per day. Comparing three measures of water provision Of the three water provision measures, the self-administered direct measure provides the most conservative assessment of wa ter scarcity in Villa Israel. According to the data from this measure, mean water provision (44.8 liters) nearly met the minimal requirement of 50 liters per person per day, and 52 households (72 percent) did not use enough water to meet their basic needs. Howe ver, several patterns in respondentsÂ’ behavior indicate that they tended to unde restimate water scarc ity using the selfadministered measure. First, we caught seve ral respondents hastily filling in responses just before their interview appointment, which suggests that some did not actually perform the self-administered direct measur e at all. Second, when we reviewed each

PAGE 139

121 household head’s responses with him or her, respondents frequently said that their data were incorrect. Respondents repeat edly told us to correct the data by reducing the liters of water used on a household task. Finally, even after correcti ng a large number of responses, there were still respondents whose responses did not fit the data I collected during participant-observation. For instance, wh en I made the typical lunch of stew for four people, I always used under 15 liters of water to wash and cook ingredients. Yet some respondents reported using 20, 30, or 40 li ters to prepare and cook one family-size lunch. One person even reported using 110 liters (although she later retracted)! In contrast with the self-administered measure, the overall estimate yielded the smallest mean water provision—only 20.1 liters. According to the data from this measure, 73 households (96 percent) did not use enough water to meet their basic needs. However, based on my knowledge of the 76 hous eholds in the survey sample, I know that there were more than three households in th e sample with a suffici ent water supply. Why, then, did respondents under-re port water provision using the overall estimate? As I explained in the field methods chapter, we did not use any additional memory prompts for the overall estimate, such as asking pe ople to specifically remember buying water, refilling tanks and buckets, or seeing the dr opping water level in ta nks. Since respondents knew that the study was about water scarc ity, they may have underestimated water provision as part of a deferen ce effect, in which the respondent tells the interviewer what they believe she wants to hear (Bernard 2002: 232 -233). Although there was no way to remove people’s knowledge about the goals of the study, the use of cues and prompts improve the response accuracy c onsiderably (cf. Brewer 2002).

PAGE 140

122 The most accurate of the three measures, I believe, was the task-based estimate. This seems quite unintuitive, as many researchers have documented the unreliability of recall based responses (cf. Deutscher 1973, Bernard et al. 1984, Ayjan and Isiksal 2004) However, by cueing memories, keeping the r ecall period short, and prompting for the activities of each household member, we helped respondents give thorough and thoughtful responses (cf. Bernard 2002: 237-239 ). Before doing the task-based estimate, we cued respondentÂ’s memories by reviewing th eir self-administered water use data with them. Then, we asked them about their water habits over the last w eek, a relatively brief response period. For each task, we prompted th em to recall the usual behaviors of each household member. While talki ng to people about their water use habits, I never caught respondents giving sloppy or suspicious respons es, as I did with the other two measures. While I realize that forgetfulness surely did affect the accuracy of respondentsÂ’ estimates, I am convinced that respondents gave their mo st reliable water-use data during the taskbased estimate questions. According to the da ta from this measure, 68 households (92 percent) did not use enough water to meet their basic needs. Table 5-1. Descriptive statistics for three measures of water scarcity SelfAdministered Direct Measure Taskbased Water Use Estimate Overall Estimate Average of Three Measures N of cases 76 74 72 76 Minimum 4.4 10.43 6.6 11.21 Maximum 88.14 136.52 189.48 134.207 Mean 20.059 32.879 44.79 31.189 Standard Deviation 12.485 18.154 27.32 17.322

PAGE 141

123 0.00 20.00 40.00 60.00 80.00 100.00 120.00 140.00 160.00 180.00 200.00 15913172125293337414549535761656973 Household (counts)Liter s Task-Based Esimate Self-Measure Overall Estimate Figure 5-17. Line graph depicting average statistics for three measures of water provision, collected across five survey periods and 76 households. The selfmeasure yields high statistics, while th e overall estimate yields low ones. The task-based estimate was selected for use in the study. If all three methods yielded accurate measur es of water provision in Villa Israel households, the three means would be the sa me. In reality, there is a rather large difference—ranging from 11.9 to 24.7 liters—be tween means for the self-administered direct measure (44.8 liters), overall estimat e (20.1 liters), and task-based estimate (32.9 liters). Even if the measures are not accurate, however, th ey could still be reliable— meaning that they all yield a high number fo r households with a lot of water, and a low number for households with a little water. To test this, I correlat ed the three measures with each other. As it turns out, none of the measures is perfectly correlated with any other. The correlation between the task-based estimate and the self-administered direct measure is .86. The correlation between the ta sk-based estimate and the overall estimate is .76. The correlation between the two measures I identified as less accurate—the selfadministered direct measure and the overall estimate—was the lowest at .64. The low correlations indicate that at least two of the three measur es are neither accurate nor reliable.

PAGE 142

124 To deal with the possibility that all three measures are inaccurate, I considered averaging the three st atistics to yield one overall m easure of household-level water provision. Doing so creates a statistic that summarizes all three measures. The summary measure has a mean of 31.2 liters (SD = 17.3). In this case, however, I have chosen to use the task-based estimate instead of the summa ry measure for two reasons. First, when I tested the difference between the task-based estimate and mean of three measures, I found that they were not very different. The correlation between the two measures is also quite high (.96). Further, a t-te st showed that the difference between the two means, while statistically significant, was quite small —only 1.3 liters (p = .05). Second, it takes an enormous amount of time and labor to collect three measures of water provision. It is unreasonable to expect that another research er would reproduce these methods, and use the average mean in an analysis. By choos ing one method and explaining why it was the best, I hope to create a model for estimati ng water provision that scholars can reproduce in future studies of urban water scarcity. Using the data throughout the analysis helps scholars assess the utility of the measure. Discussion: What do the water provision data really mean? According to international consensus, peopl e need at least 50 liters of water a day to meet minimal human water requirement s for drinking and household use in lowtechnology situations. If people in Villa Israel have an aver age of 33 liters per day, what can we conclude about water pr ovision there? What is it li ke to live beneath minimal water standards? Between January 2004 and 2005, no resident of Villa Israel died of causes related to water scarcity or contaminat ed water. Rather than creating a dramatic death toll, the daily eff ects of water scarcity in Villa Israel were mundane, relentless, and immiserating. People use filthy water for house cleaning and washing. They are unable to

PAGE 143

125 maintain good standards of hygiene, and suffer lice infestations and other related health problems. At times, families even lack the water needed to cook or drink. The absolute measures of water provision provide an essential baseline measure of water scarcity in a community. However, vulne rability to water scarcity is about more than the amount of water one has at any give n time. The struggle to refill tanks and the constant search for water is another importa nt aspect of water scarcity, and is not captured very well by the absolute water meas ure. In the next section, I discuss water security, or what it means to have a stable and reliable water source. Approach Two: Proxy Measures of Water Security Water scarcity means more than just not having enough water. Another important aspect of water scarcity is wate r security, or a household’s ability to maintain its access to water. When people have secure access to productive water sources, they can always meet minimal water needs, such as drinki ng, cooking, and basic hygiene. However, when households lack secure access to productive wa ter sources, they may run out of water or be unable to replenish hous ehold water supplies at any time. When this occurs, households suffer very serious consequences —such as not drinking, cooking, or bathing. In this section, I present f our different approaches to measuring water security. Each approach provides a proxy measure, that is, it measures an indicator of water security rather than measuring water securi ty directly. The first approach measures monthly precipitation in Cochabamba. The second approach measures the physical location of households in Villa Israel. The th ird approach measures involvement in Villa Israel’s political and religi ous institutions. The fourth and final approach measures household water storage capacity. With each of these approaches, I examine the role that water security plays in determining how vulnerable people are to water scarcity.

PAGE 144

126 Precipitation The first measure of water security in Villa Israel is a measure of precipitation. As I explained in the section above, rainfall and the river are importan t water sources for families in Villa Israel. When it rains, many families use rainwater collection equipment to capture and store large qua ntities of water. They also use river water, which is generated from the runoff created by rainfall for washing and cleaning. Rain and river water can be considered “secure” water so urces because access is free and relatively unrestricted. Precipita tion, then, provides a straightfo rward proxy measure of the water available to all Villa Israel residents. In this study, I utilize mo nthly precipitation data that were collected from the Cochabamba weather station (located at 17.45 S and 66.09 W). The data were published in Official Monthly Bulletins by SENAMHI, Bolivia’s National Meteorological and 153.4 129.2 13.5 24.2 4.7 0 39.9 0.1 6.1 0.04 83 66.9 101.2 0 20 40 60 80 100 120 140 160 180 1234567891011121314 Months, during the Study PeriodRain (mm) Figure 5-18. Line graph depi cting monthly rainfall in th e Cochabamba Valley (17.45 S and 66.09 W), from January 2004 to January 2005. Hydrological Service. The data are collected each day from the Cochabamba station, and summed to create a monthly measure of rainfall in millimeters. The precipitation data

PAGE 145

127 analyzed here were collected in Cochabamba from January 2004 to January 2005 (SENAMHI 2006). The average monthly rainfall in Cochabam ba over the 13-month study period was 47.87 mm (see appendix, Chart 5-18). The minimum monthly ra infall was 0.00 mm (recorded in June 2004), while the maximu m monthly rainfall was 153.40 mm (recorded in January 2004). Rainfall only exceeded 100 mm during three months: January 2004, February 2004, and January 2005. Rainfall was less than 10 mm during five months: May 2004, June 2004, August 2004, September 2004, a nd October 2004. When graphed, it is clear that rainfall-based water security is highest during the wet season months (November to February) and lowest during th e dry season months (March to October). Physical location of households The most important source of water for Vill a Israel residents is the water delivery truck, or aguatero. To measure the security of householdsÂ’ access to the aguatero, I created a measure of the physical distance of households from the aguateroÂ’s main route. Since the entrance to Villa Israel is via the ma in street, aguateros must drive on that street every day. As I mentioned previously, people who live on or near the main street have little trouble finding the aguate ro and convincing him to sell them water. People who live at the far edges of the commun ity, however, have a hard time finding the aguatero, and an even more difficulty trying to convince him to dr ive to far-flung residences to sell water. The physical location measure, then, is a proxy measure of the difficulty people have finding the aguatero and convincing him to sell them water. To create this measure, I divided the map of Villa Israel into four main zones. The first zone consists of households that are located within one half-b lock from the main st reet (on the east and west sides). The second zone consists of households that are located within one half-

PAGE 146

128 block of the second streets from the main st reet (on the east and we st sides). The third zone consists of households that are located within one half-block of the third streets from the main street (on the east and west sides). The fourth z one consists only of households that are located past the third stre et and the runoff canal on the west side of the community; the river lies directly beside the third street on the east side of the community. People with the best access to the aguatero live in zone one, within one block of the main street. In the study sample, 22 percen t of households had very secure physical access to the aguatero. People with good physical access to the aguatero live in zone two, about two blocks away from the main stre et. The largest percentage of households, 41 percent, is located in this zone. People w ith poor physical access to the aguatero live in zone three, about three blocks away from th e main street. About 33 percent of households lived in this zone, and had insecure access to the aguatero. The fourth zone is quite far from the main street, and also has a majo r physical barrier to access, the runoff canal. Only 4 percent of households in the sample we re located in this zone; their access to the aguatero was extremely insecure. Political and religious connections Another factor that affects a household’s ab ility to secure access to water sources is political and religious involve ment (Ennis-McMillan 2006). As I explained in the section in the community water system, people w ith connections—via the Neighborhood Council or a local evangelical church—have more secu re access to the tap stand system than those without connections. While it does not provide enough water fo r exclusive long-term use, the tap stand system is an important resource for households that are experiencing a water crisis. The involvement of a household in Villa Israel’s Neighborhood Council and

PAGE 147

129 evangelical churches is a proxy measur e for the household’s politico-religious connections in the community—a nd their ability to get acce ss to the tap stand system during a crisis situation. Over 10 months, or five interview cy cles, we collected data on household participation in the Neighborhood Council a nd local evangelical churches. In each interview, we asked household heads to tell us the number of meetings they attended in Villa Israel during the last week. The averag e number of meetings that household heads attended per week is a measure of how conn ected they are to political and religious groups in Villa Israel. People who wield more politico-religious connections, then, should have more secure access to the tap stand sy stem than people with fewer connections. Over the 10-month study period, the av erage number of politico-religious interactions that household heads had over a week-long period was 0.3 meetings (SD = 0.41). The maximum average number of meeti ngs that any household head had over the study period was 1.5 meetings per week, and the minimum was 0.0 meetings per week. It is, perhaps, more illuminating to examine the difference between household heads that have some connections in Villa Israel and hous ehold heads that have none at all. Once the measure is dichotomized, it yields a simp le variable in which household heads are assigned either a 0 (meaning that the never a ttended a political or religious meeting in Villa Israel) or 1 (meaning that they attended at least one political or religious meeting in Villa Israel). Over ten months, 36 percent of household heads reported that they attended no meetings at all in the community; 64 per cent of household heads reported that they attended at least one poli tical or religious meeting. Those households who had no connections at all with political and relig ious groups probably had more difficulty

PAGE 148

130 maintaining secure, long-term access to the co mmunity tap stand system than households that had political or re ligious connections. Water storage capacity The last measure of water security is hous ehold water storage area. As I explained in the section on water provision above, each household in Villa Israel uses turriles an underground water tank, or some combination of the two to store water for household use. Some households also keep 20-liter bidone s or 10-liter baldes as supplemental water storage containers. Households with one turril of storage capacity can make 200 liters of water last up to a week, while households with a large storage tank can use water at a rate of 2500 liters per week. While th ere is variation in rates of consumption, households with larger water tanks generally have more secure access to water. Having a larger tank gives people more space to collect rainwater, more water to use, and more time to save and plan to refill the tank. As a result, people w ith water tanks have more ability to keep their tanks filled with a stable supply of water. Water Storage Capacity in Villa Israel0 2000 4000 6000 8000 10000 12000 147101316192225283134374043464952555861646770 Households (count)Storage Capacity in L Figure 5-19. Line graph depicting water stor age capacity among Villa Israel households. Over 8 months, or four interview cycl es, we asked household heads how much water storage area they had in their homes. Because people buy or sell turriles and build

PAGE 149

131 or damage water tanks, storage capacity doe s vary over time. Water storage capacity ranges between 200 and 10,600 liters in Villa Is rael. The average ove rall water storage capacity in Villa Israel hous eholds over the 8 month peri od was 2240 liters (SD = 3172). Eleven households (15 percent) had only one turril of storage space. On the other end of storage capacity, five houses (7 percent) had 10,000 liters or more of water storage area. Summary: Proxy measures of water security In this section, I pres ented four different proxy measures of water secu rity in Villa Israel. In the first approach, I assessed the av ailability of precipita tion, a water source that is not restricted by economic, political, or social factors. In the second approach, I examined the physical proximity of individua l households to the most important water source in Villa Israel, the agua tero. In the third approach, I examined household headsÂ’ involvement with political and religious inst itutions that mediate access to Villa IsraelÂ’s tap stand system. In the fourth and final approach, I measured householdsÂ’ water selfsufficiency in terms of their storage capacity. Each of these measures captures some as pect of a household ability to secure access to water. The approach is not designed to reflect the amount of water to which a household has access. Instead, the security m easures assess how likely a household is to get and maintain access to some amount of water, however small. This distinction is important because people in Villa Israel can and do survive with very small daily water allotments. It is only when they lose water access completely that people are unable to complete crucial household ta sks like cooking and drinking. Approach Three: Measures of the Exp erience of Urban Water Scarcity The third and final approach that I develop here is a measure of the experience of water scarcity. Because the amount of water th at people need to maintain their standard

PAGE 150

132 of living varies enormously across cultures, people in different cu ltural contexts can experience the same water provision level as ab undance or scarcity. For instance, people with less water develop cultural practices that attenuate the effects of water scarcity (cf. Vargas 2001). Conversely, people with highly water-consumptive technologies need far more than the recommended water allotment to care for themselves (cf. Gleick 1996). A measure of the experience of water scarcity indicates whether people have experienced water scarcity from the perspectiv e of people in that culture. From interpretation to measurement In creating the measure of the experien ce water scarcity, I drew heavily from Ennis-McMillan’s work on sufriendo del agua (suffering from water) in La Purificacin, a water-scarce, urbanizing community in th e Valley of Mexico (2001, 2006). Sufriendo del agua is the embodiment of worry over th e inequitable distribut ion of scarce water. Ennis-McMillan explains that medical anthr opologists examine bodily distress and social suffering to understand the “nexus of people’s negative physical, emotional, psychological, and social experiences…to ex amine how people experience and articulate their bodily distress in relation to unequal social relations” (2001: 370). Ennis-McMillan showed that community residents experience bodily distress over scarce water—a variety of psychosomatic symptoms—when they see ev idence of injustice in water allocation. One of the reasons that I find the bodily di stress concept so compelling is that it offers a way to understand the culturally-media ted experience of water scarcity that is both widely accepted by interpretive anthropol ogists and amenable to hypothesis testing. Although Ennis-McMillan describes sufriendo del agua as a local idiom, I thought that the embodiment of urban water scarcity might be a universal phenomenon. EnnisMcMillan included rich et hnographic descriptions of sufriendo del agua and provided a

PAGE 151

133 list of words that residents us ed to express their suffering: frustracin (frustration), angustia (anguish), molestia (bother), preocupacin (worry), coraje (anger), and of course sufrimiento (suffering) (2006: 117). During the participation-observation phase of the research, I asked people about how they feel when they expe rience water scarcity. As it turn ed out, Villa Israel residents used some of the same word s Ennis-McMillan recorded, molestarse (to get upset) and preocuparse (to get worried), to talk about water scarcity. Villa Is rael residents also used new words that appeared to describe the embodiment of water scarcity: renegar (to fume) and sentir miedo (to feel scared). I created a qu estionnaire to ask about these four feelings in the Cochabamba dialect. The culturally-constructed aspects of wate r scarcity, of course, are enormously complex and cannot be fully captured only by examining bodily distress. I wanted to gather data on all of the aspect s of the experience of water scarcity, as I had heard Villa Israel residents talk about it. I wrote 29 additional questi ons, and created a questionnaire with 33 questions in all. All of the questions came from wo rds and ideas that I learned about during participant-observation and the first round of survey research; no question came solely from categories or ideas that I broug ht with me to the field. This is important, because it means that the questionnaire reflec ts native experiences of water scarcity. It also means that replication of the questionn aire probably would not work very well for researchers in places other than south-side Cochabamba, or perhaps Bolivia. Rather, each researcher should create their own list of questions based on ethnographic research. Data collection and analysis Between June 2004 and January 2005, we asked household heads to respond to the 33 question four times. Each of the 33 que stions asked if the household head had

PAGE 152

134 experienced a water-related f eeling or event during the last week. The questions required only a simple yes/no answer, and were coded using 1 (yes) and 0 (no). I took the step of transforming rich ethnographic observations to spare yes/no questions because the data they yield enable me to create Guttman scales. A Guttman scale is a kind of composite measure made up of indicators that poi nt to a progressive, unidimesional concept (Guttman 1950). Usi ng Guttman scales, we can go beyond just talking about experience—we can define the steps people go through when they experience different aspects of water scarcit y. We can also calculate which step each Table 5-2. English translations of thirty-thr ee yes/no questions about the experience of water scarcity in Villa Israel. 1. In the last week, have you gotten angry with someone from your house over water? 2. In the last week, have you argued with someone from your house over water? 3. In the last week, have you gotten angry with someone from Villa Israel over water? 4. In the last week, have you argued with someone from Villa Israel over water? 5. Sometime in the last week, have you fumed about water scarcity? 6. Have you worried every day over the last week about water scarcity? 7. In the last week, have you been afraid that water would run out in your house? 8. In the last week, have you felt annoyed about having to fetch water? 9. In the last week, have you lacked the money you need to buy water? 10. In the last week, did you have to run after the aguatero in the street? 11. In the last week, have you had to beg the aguatero to come and sell you water? 12. Have you borrowed water from anyone in Villa Israel in the last week? 13. Have you loaned water to anyone in Villa Israel in the last week? 14. Have you bought water from anyone in Villa Israel in the last week? 15. Have you sold water to anyone in Villa Israel in the last week? 16. In the last week, have you wasted time because of water scarcity? 17. In the last week, have you wasted time because of the aguatero? 18. In the last week, have you lost out on an opportunity to earn money because of water scarcity? 19. In the last week, have you lost out on an opportunity to earn money because of the aguatero? 20. In the last week, have you conserved water when bathing? 21. In the last week, have you conserved water when cleaning the house (i.e., mopping)? 22. In the last week, have you conserved water when washing the laundry? 23. In the last week, have you conserved water when cooking? 24. In the last week, has anyone gotten sick in your house because of water scarcity? 25. In the last week, have you been unable to cook because you lacked water? 26. In the last week, have you been unable to bathe because you lacked water? 27. In the last week, have you been unable to clean the house because you lacked water? 28. In the last week, have you been unable to do the dishes because you lacked water? 29. In the last week, have you been unable to wash laundry because you lacked water? 30. In the last week, have you thought of leaving Villa Israel because there is no water here? 31. Have you mopped your house with clean water in the last week? 32. Have you mopped your house with reused water in the last week? 33. Have you used the same water more than once in the last week?

PAGE 153

135 Table 5-3. Thirty-three yes/no questions about th e experience of water scarcity in Villa Israel, written in the south Cochabamba Spanish dialect. The data were used to create five different Guttman scales. 1. En esta ltima semana te has enojado con alguien de tu casa por el agua? 2. En esta ltima semana has discutido con alguien de tu casa por el agua? 3. En esta ltima semana te has enojado con alguien de VI por el agua? 4. En esta ltima semana has discutido con alguien de VI por falta de agua? 5. Alguna vez en esta ltima semana has renegado por falta de agua? 6. Te has preocupado cada da en esta ltima semana por falta del agua? 7. En esta ltima semana has sentido miedo porque el agua se va a acabar en tu casa? 8. En esta ltima semana te ha molestado conseguir el agua? 9. En esta ltima semana te ha faltado plata para comprar el agua? 10. En esta ltima semana tenas que ir corriendo por la calle detrs del aguatero? 11. En esta ltima semana has tenido que rogar al aguatero para que venga a venderte agua? 12. En esta ltima semana te has prestado agua de alguien en VI? 13. En esta ltima semana has prestado agua a alguien en VI? 14. En esta ltima semana has comprado agua de alguien en VI? 15. En esta ltima semana has vendido agua a alguien en VI? 16. En esta ltima semana has perdido tiempo por falta de agua? 17. En esta ltima semana has perdido tiempo por culpa del aguatero? 18. En esta ltima semana has perdido la oportunidad de ganar dinero por falta de agua? 19. En esta ltima semana has perdido la oportunidad de ganar dinero por el aguatero? 20. En esta ltima semana has medido el agua para baarte? 21. En esta ltima semana has medido el agua para limpiar la casa? 22. En esta ltima semana has medido el agua para lavar ropa? 23. En esta ltima semana has medido el agua para cocinar? 24. En esta ltima semana alguien se ha en fermado en tu casa por falta de agua? 25. En sta ltima semana no has cocinado por falta de agua? 26. En sta semana no te has baado por falta de agua? 27. En sta semana no has limpiado la casa por falta de agua? 28. En sta semana no has lavado los platos, vasos, ollas por falta de agua? 29. En sta semana no has lavado ropa por falta de agua? 30. En esta semana has pensado en irte de VI porque no hay agua? 31. Has baldeado tu casa en sta ltima semana con agua limpia? 32. Has baldeado tu casa en sta ltima semana con agua sucia? 33. Has usado ms de una vez la misma agua en sta ltima semana? household reached, and use the number assigned to them as ordinal data in later analyses. Using Guttman scales, then, culturally-unique experiences can be defined, measured, and ranked. The data from 67 household heads inte rviewed between August and September 2004 were analyzed in Anthropac, a suite of programs that has a routine for checking the unidimensionality of a scale with the Gu ttman technique (Borgatti 1996). Using the program, I organized the data into columns, so that many respondents said yes to the question in the first column, fe wer said yes to the second, and so on. If the variables scale

PAGE 154

136 perfectly, each question is pr ogressive, and no one answers yes to one question without also saying yes to the question that preceded it. To test how well variables scale, I calculated the coefficient of re producibility (CR—the statistic that summarizes errors in the scale) for each scale. If the CR was grea ter than 0.85, I considered the questions to scale sufficiently enough to discuss further here (Bernard 2002: 302-207). I found that variables for five of the dimensions of wa ter scarcity experience examined in this chapter—bodily distress, wate r-based tasks, aguatero, wate r exchanges, and conflicts— did scale. Below, I examine each in turn. Bodily distress The first concept that I tested using the V illa Israel data was Ennis-McMillan’s idea of bodily distress experienced in response to water scarcity. As I explained above, there were questions on the yes/no questionnaire that dealt with four indicators of bodily distress: molestarse (to get upset), preocupa rse (to get worried), renegar (to fume), and sentir miedo (to feel scared). I found that one of the indicators—to feel about water running out—does not scale very well. The othe r three variables do scale, with a CR of .85. The first indicator is “worrying ever y day about water scarcity.” The second indicator is “feeling annoyed about getting water.” The thir d indicator is “fuming about water scarcity.” The series of feelings that people experience—worry, anger, fuming— represents the escalati on of negative emotions about water scarcity. People in Villa Israel do not link thei r suffering to wider political economic conditions in any way, as they did in EnnisMcMillan’s study. When talking about water scarcity, people focus on the experience of not having water rather than the socioeconomic phenomena that cause water scarcity. As a result, it is not cl ear that the feelings of worry, annoyance, and fuming that they express can really be considered bodily

PAGE 155

137 distress. What is clear, however, is that pe ople experience feelings a bout water scarcity in a measurable, predictable, and progressive way. Water-based tasks The second concept that I tested is a ve ry intuitive one—wate r-based tasks. The idea behind this scale is that people prio ritize water-use tasks when water becomes scarce. It makes sense, for instance, that people would stop watering plants before they stop drinking water. We asked people about fi ve different household water-use tasks that they said they stopped doing when water was low—cooking, bathing, cleaning the house (mainly mopping floors), washing dishes, and doing laundry. A ll five task indicators do scale, with a CR of 0.88. The first task that people eliminate when water is scarce is cleaning the house. After that, they eliminat e in succession: bathing, doing the laundry, cooking, and finally doing the dishes. The rankings determined by the scale ar e, for the most part, unsurprising. Obviously, people would stop cl eaning their homes before th ey stop bathing, and they would stop bathing before they stop cooking. What appears a b it strange, however, is the placement of two tasks—doing laundry and di shes—at the middle and far end of the scale. However, when one looks closely, it appe ars that each task is located directly after the task that logically precedes it. Fo r instance, people stop cooking, and then stop washing dishes. Similarly, people stop bathing, and then stop washing clothes. The scale, then, seems to represent both th e priority of tasks in wate r scarce conditions, as well as the order in which they are performed (and halted). Buying water from the aguatero One interesting idea I tested was that people pass through a common set of experiences when they attempt to buy water fr om the aguatero. I co mpiled a short list of

PAGE 156

138 the things people said happen when they had trouble getting water from the aguatero. The list includes running out of money, chasing after the agua teroÂ’s truck, begging the aguatero to sell water, and wasting time becau se of the aguatero. The four indicators do scale, with a CR of .89. The first experience that most respondents had was begging the aguatero to sell water. After that, people re ported running after the aguatero. Next, people said they wasted time because of the aguater o. Finally, people said that they lacked the money they needed to buy water from the aguatero. Superficially, this scale app ears to be about the experi ence of being unable to get water from the aguatero. However, what unde rlies the inability to get water from the aguatero is the amount of money that househol d heads have available to buy water. For instance, people who do not have a lot of money allocated to buy water have to beg the aguatero to come to their houses. The agua tero may avoid people who repeatedly buy water in small quantities, forcing them to run after the truck. People who run and beg without convincing the aguatero that it is wo rth selling them water have wasted their time. Finally, when people can not induce the agua tero to sell them water, it is clear that they lack the sufficient quantity of money th ey needed to buy water from the aguatero. In view of this interpretation, the scale seems to summarize the experience of lacking money to buy water from the aguatero. Water exchanges Another important water acquisition experien ce in Villa Israel is participating in water exchanges. As I mentioned above, there were not a large number of water exchanges reported during the in terviews. As a result, I was unable to identify trends regarding how people decided to sell water, buy water, loan water, and borrow water with their neighbors. Luckily, a Guttman analysis can isolate trends that are difficult to see in

PAGE 157

139 participant-observation and semi-structure d interviews. The four water exchange indicators scaled very well, and had a CR of .94. The water exchange that most people engage in is loaning water to someone else in the community. After that, people borrow water from someone else. Then, people buy wa ter from another Villa Israel resident. Finally, they sell water to anot her Villa Israel resident. When I ran the Guttman procedure on these indicators, I was surprised that they do scale. I believed that people with more water would loan and sell the resource, while people with less water would borrow and buy it. In fact, the scal e proves that the monetization of water exchanges is a tactic adopted only by people who also participate in water exchanges like loaning and borrowing. Buyers and sellers, then, are people who choose to monetize just a few of their wate r exchanges. I suggested above that people monetize water exchanges to avoid creating social obliga tions, and it would be very interesting to do a more comprehensive st udy of when and why people monetize some water exchanges and not others. Water conflicts The last set of indicators that I tested d eals with water-related conflicts. This is a category that I was quite interested in at th e onset of the study. I learned that people in Villa Israel only use two terms to describe water-related conflicts—enojarse (to get angry) and discutir (to get in an argument). Both of these terms are rather euphemistic, and actually encompass a variety of interactions Enojarse is word th at describes feeling anger, making an angry comment, making a nasty face, or storming away. When the anger is two-sided, it is said that people di scutir (get in an ar gument). Discutir can describe anything from a disagreement to a shouting match to a minor scuffle. Once an altercation involves pushing, hi tting, or punching, though, it is called pelear (to fight).

PAGE 158

140 People did pelear in Villa Israel while I was there, but no one reported doing so over water. Using only these native categ ories, I asked people about four kinds of conflicts— getting angry with people at home over wa ter, getting angry with people in the community over water, getting into an argumen t with people at home over water, getting into an argument with people in the commun ity over water. The water conflict indicators scaled very well, with a CR of .96. It turn s out that household heads get angry with people at home over water first. After that they get into arguments over water with people at home. Then, people get angry with community members over water. Finally, people get into water-related argum ents with community members. As I mentioned in the section on water exch anges, people in Villa Israel are very conflict-averse. It is not surprising, then, that people always get angry before they get into interpersonal conflicts, whether in the home or in the community. However, if inequity in water distribution were the main driver of water-related anger and conflict, we could expect community-level incidents to preced e household-level ones. Instead, it appears that people turn their frustrations over the water situation toward community members only after water-related tensi ons boil over in the home. Summary: Measures of the experience of water scarcity In this section, I developed five Guttman scales that measure different aspects of the experience of water scarcity in Villa Israel. The five scal es deal with feelings, waterbased tasks, the aguatero, wate r exchanges, and water confli cts. These five aspects of water scarcity are not exhaustive. In fact, the experience of water scarcity is very multifaceted, and could be examined alone in a much more extensive study. In this section, I seek only to show that interpretive co ncepts, such as bodily suffering, can be

PAGE 159

141 operationalized and measured. By doing so, we can understand the extent to which experiences associated with water scarcity are shared within and across communities. Further, using methods like Guttman scaling, we can link in terpretive and scientific approaches to understanding experiences. In doing so, we promote the growth and development of a rich, diverse anth ropology of urban water scarcity. Discussion: Defining and operationalizin g the urban water scarcity concept In this chapter, I have argued that peopl e experience water scarcity as more than just unmet biophysical needs, and the water requirement guidelines can only provide one of the tools that are needed to examine urba n water scarcity. Because the amount of water needed to complete everyday household tasks varies across people and cultures, measures of experience can tell us a great deal about how and when water scarcity hits a society— regardless of the number of liters of water they have per day. In Villa Israel, many people survive with less than 10 liters of water per day. In this context, it is the complete loss of access to water sources that causes people to re ally experience a water crisis, in which they do not bathe, cook, or drink with water. In this section, I developed three main a pproaches to measuring water scarcity. The three approaches are designed to capture diffe rent facets of the complex water scarcity phenomenon. The first approach is a measure of water provision, or the amount of water a household has per person per day. The second approach to measuring water scarcity examines water security, or the stability of a household’s access to water supply sources. The third approach measures e xperiences of water scarcity. Each of the approaches provides importa nt tools for understanding water scarcity across cultures and time periods. Measures of water provision have already been reproduced successfully and used to compare le vels of water scarcity in many societies.

PAGE 160

142 Measures of water security and experi ence are bound to loca l systems of water distribution, and tell us about the unique ways in which wate r scarcity is structured and experienced across cultures. Pu t in anthropological terms, measures of water provision and security can tell us a great deal about the etics of urban water scarcity, while measures of water experience do a good job of telling us a bout the emics of urban water scarcity. Etic knowledge holds true across cu ltures, while emic knowledge is situated within the beliefs and language of a specifi c cultural context (L ett 1996, Harris 1979). It is important to study both the emics and etic s of urban water scarc ity to understand how the phenomenon is created and experienced across cultures.

PAGE 161

143 CHAPTER 6 SOCIAL RELATIONS AND EC ONOMIC EXCHANGES In this chapter, I discuss the dependent va riables that are examined in this study. In the first two sections, I briefly review anthropological studies on sociability and reciprocity. After that, I expl ain how the studies relate to Laughlin-Brady model and the hypotheses I tested in Villa Isra el. Finally, I examine how two factors, nested cycles and cultural institutions, complicate and contribute to the analyses I conclude that, while it is important to examine nested cycles and cultural institutions, it is also appropriate to focus on the sole theoretical question under study he re—the effects of ur ban water scarcity on social relations and economic exchanges. Anthropological Research on Reciprocity The reciprocity concept is one that emer ged in early anthropological inquiry, and continues to fuel anthropological theory-mak ing today. Reciprocity has been analyzed by anthropologists from two main perspectives. The older approa ch views reciprocity as the phenomenon underpinning human society, base d on exchanges of goods, labor, ideas, and sentiment that form the building blocks of all social systems. Another approach treats reciprocity more narrowly, as a system of economic exchange in which goods and services are given with the expectation that they will be returned later. Both of these perspectives are important, and will help e xplain hypothesized change s in sociability and reciprocity. In the sections that follow, I discuss the two anthropol ogical perspectives on reciprocity, and how they relate to this study.

PAGE 162

144 Reciprocity as a Social Foundation One of the earliest anthr opological works on reciprocit y, Malinowski’s study of the Trobriand Islanders, influenced many of the major works in economic anthropology (1922). Malinowski conceived of reciprocity as the principle that organized the entire social life of Trobriand Islanders—legal, economic, moral, social, religious, and psychological. He explained th at reciprocity, or the “princ iple of give and take”, establishes binding obligations which compel honorable citizens to accept offers and repay them so as to not be excluded from the social order. Another influential early work drew heavily on Malinowski’s analysis of reciprocity: Mauss’ The Gift (2000). Mauss also believed that reciprocity was a system of exchange in which everything from people, to items, to labor passes “to and fro”. The system, as he saw it, permeated the legal, economic, moral, re ligious, and aesthetic dimensions of social life. Followi ng Durkheim (1982), he called this a total social fact. Homans, too, believed that exchange permeated social life, and ar gued that all social interactions should be analyzed as exchanges (1958). Blau, in his work on exchange and power, followed Homan’s conceptualization of social exchange and expanded it, clarifying how systems of recipr ocal exchange can bestow be nefits, but also widen gulfs in status, power, and differences (1964). While the reciprocity concept encompasse s both social and economic exchanges, early anthropologists focused on the importance of social exchanges of ideas, power, and prestige. They showed that these reciprocal exchanges form the social foundation of all societies. Early approaches to understa nding reciprocity, then, demonstrate the importance of understanding patterns of soci al interaction and ex change as part of reciprocity.

PAGE 163

145 Reciprocity as a System of Economic Exchange In the late 1960s, scholars began to focu s more closely on explaining the economic aspects of reciprocity. Pola nyi was one of the first sc holars who conceptualized reciprocity as an economic system, which c ould be analyzed in comparison to other economic systems like redistribution and mark ets (1968). Following Polanyi, a number of scholars began to examine when, where, and why systems of reciprocal economic exchange are established. Sahlins explained that reciprocal exchange s of goods and services could be less or more self-interested, depending on the social di stance of the actors. He defined the most altruistic of exchanges as generalized reciprocity where there is a prohibition on immediate payback. The midpoint of the continuum is balanced reciprocity where goods of equal value are exchanged at the same mo ment or perhaps after a small delay. The far end of the continuum is negative reciprocity where people attempt to get a thing for less than its value (1972). Most anthropologists, when they discuss reci procity, refer to the kind of exchange that Sahlins called generalized reciprocity. Later economic theorists showed that reciprocal exchange systems are intended to maximize security, and function like informal social insurance systems. In her work on the !Kung, Wiessner showed that reciprocal exchanges are a “social method of pooling risk thorough storage of social oblig ations” (1982: 65). When people cannot independently store the goods they need to su rvive, they create long-standing exchange relationships in which obligations are st ored. That way, when one person needs food, water, or help, they can call in a favor (a “stored obligation”) from another person. In her work on Basarwa reciprocity, Cashda n (1985) used the economic theory of risk minimization to explain how reciprocal exchange systems work. Reciprocal systems,

PAGE 164

146 like any insurance system, only work when scarc ity hits individuals at different times. If everyone in the system is hit by scarcity at the same time, the system breaks down. Cashdan suggested that reciprocal insuran ce systems would exist in urban areas when people have independent economic fortunes, but would be truncated where urbanitesÂ’ fortunes are all tied to the same risks. Urban Reciprocity Anthropological theories of reciprocity were created and tested in societies that were described as primitive, pre-industrial and rural. Lomnitz was one of the first anthropologists to examine urban reciprocity as an economic system (1977). She defined reciprocity as a form of economic exchange that is social, is recurring, and maximizes security (1977: 189). Her work on reciprocity in a Mexica n shantytown showed that systems of reciprocal exchange protect th e urban poor against severe scarcity (1978, 1978). Research conducted between the 1960s and 1980s documented how the urban poor used reciprocal exchanges to combat scarc ity, particularly in Latin America and the Caribbean. Reciprocal exchange systems dealt with needs as diverse as job assistance, loans, goods, services, and f acilities (Lomnitz 1977), chil dcare (Stack 1974, Safa 1974, Scheper-Hughes 1992), help during crises like accident, illness, fire (Safa 1974, Lobo 1982), moral and emotional support (Lomnitz 1977), protection against alcoholism and spousal abuse (Lomnitz 1978), and companionship (Safa 1974, Lobo 1982). Most importantly, however, reciprocal social ne tworks were used to supplement the low incomes (Leeds 1971) and frequent joblessness (Lobo 1982, Perlman 1976) with which the urban poor had to contend.

PAGE 165

147 Sociability, Reciprocity, and the Laughlin-Brady Model From the brief survey of anthropological research that I provi ded above, a picture of reciprocity begins to emer ge. Reciprocity describes, firs t, a kind of social exchange system that permeates all societies. The ex change of values, information, and social interactions provides a base on which co mmunities and cultures are built. These reciprocal social interactions are a necessary but insufficient condition for the construction of reciprocal economic exchanges. The emergence of reciprocal exchange systems depends on the kinds of risks and protections found in an economy. When people in a society are threatened by severe scarcity, lack sufficient material means to pr otect themselves against scarcity, and are not all threatened by scarcity at the same time, reciprocal exchange systems can develop. When basic necessities are abundant, can be st ored, or become scarce for everyone at the same time, reciprocal exchange systems generally do not thrive. The conditions for reciprocal exchange systems, then, can develop in any economy, depending on the configuration of risks and protections found there. The Laughlin-Brady model (1978) predicts how people will respond when the third precondition for successful a reciprocal exchange system is violate d, and scarcity hits everyone in a society at the same time. Laughlin and Brady showed that, when scarcity initially hits, people invest more heavily in the reciprocal exchange system. They engage in more social interactions, participate in more reciprocal exchanges, and seek cooperative solutions. However, after this initial period of increased sociability and reciprocity, people lack the resources and ener gy to sustain this kind of enhanced sharing and cooperation. Instead, they begin to withdr aw from social rela tionships, reciprocal exchanges, and cooperative solutions. As scar city continues to worsen, people tend to

PAGE 166

148 become completely isolated as they pursue individual survival. The figure below depicts the hypothesized relationship between resource scarcity, sociabil ity, and reciprocity. Figure 6-1. Relationships hypothesized by th e Laughlin-Brady model. The X axis represents increasing resource scarcity, and the Y axis represents varying levels of reciprocity and sociability. A test of the Laughlin-Brady model should examine changes in three dependent variables over time: sociability, reciprocity, and community participa tion. In this study, I used five separate measures to test changes in sociability, reciprocity, and participation in Villa Israel. By explaining exactly what kinds of exchanges I investigated, detailing the coding rules I employed, and discussing how they fit into native categories, I provide clear and reproducible measures of the th ree concepts under study. Before I proceed, however, I will briefly discuss how the complexity of scarcity cycles complicates our ability to understand cha nges in reciprocity and sociability over time. Studying Nested Scarcity Response Cycles The periods of privation that the Laugh lin-Brady model describes can be short or long. Short-term scarcities can be cause d by seasonal drought, crop failure, road blockages, or damage to urban infrastruc ture. Long-term scarcities can be caused by natural disasters, economic crises, or envi ronmental degradation. A recent theory of integrated social-ecological sy stems explains that short-term and long-term cycles can be NormalModerateSevereExtremeMore Extreme Resource Scarcit y High Med Low

PAGE 167

149 nested within each other, causing responses at multiple scales (Gunderson and Holling 2002). Like most social-ecological systems, the one under study in Villa Israel has nested scarcity response cycles. For instance, short-term seasonal changes in water provision are nested within a long-term cycle of globa l capitalism. While the short-term cycle— seasonal water scarcity—is the focal point of this study, it is also important to understand how that cycle it fits into larger economic trends. In the next section, I explain how a twenty-year shift toward an informal economy, punctuated by periods of severe economic crisis, appears to have undermined the ur ban reciprocal insurance systems among Latin America’s urban poor. As I explained in the first chapter of this study, the economic crises of the 1980s and economic restructuring of the 1990s destab ilized existing economic systems in Latin America. Squatter settlements were hit ha rder than downtown areas because their residents lacked physical and social access to the diminishing job opportunities (Eckstein 1990). The economic crisis—and particularly th e decline in the poo r’s opportunities to earn regular wages, intensify and diversify la bor, and self-employ—di srupted their ability of invest in and redeem social ob ligations (Gonzlez de la Rocha 1994, 2001, 2004, Moser 1996). As a result, families intensified strategies that generate income and attempted to minimize income expenditures, consumpti on, and social investments (Halebsky 1995). Reciprocal relationships and participation in community organizations declined because people lacked the time and resources to invest in maintaining them (Moser 1996). Between 1980 and 1995, poor families in urban La tin America withdrew from the social,

PAGE 168

150 reciprocal, and cooperative relationships th at were the hallmark of shantytowns and squatter settlements between 1960 and 1980 (Gonzlez de la Rocha 2001, 2004, Moser 1996). In the section that follows, I examine data that characterize reciprocal exchange systems in Villa Israel over the 10-month survey period, from April 2004 to January 2005. Using data on average social interactions, reciprocal exchanges, and participation, I conclude that reciprocal exch ange systems, while perhaps di minished, are still active in Villa Israel. Data on Sociability, Reciprocity, and Participation in Villa Israel In this section, I discuss the data my fi eld team and I collected over 18 months in Villa Israel. As I explained in the chapter on field methods, data on social interactions and economic exchanges were collected us ing two methods: dir ect observations and semi-structured surveys. First, I explain how the data from each method were coded. Then, I analyze the data on reciprocity, sociab ility, and participation, and discuss them in view of the literature on contemporary reciprocal excha nge systems in Latin America. Public Social Interactions Data collection and coding Direct observations were used to collect data on public social interactions. The method yielded the only data on sociability that could not be a ffected by respondent accuracy problems. The direct observations were collected in four repeated 42-day cycles, and observations were made in 12 ra ndomly-selected public places at a randomlyselected time each day. Researchers obser ved and coded “social interactions”. A grouping of two or more people was considered to be one social interaction. When there were multiple groupings in one public pl ace, each grouping was considered a unique

PAGE 169

151 social interaction. One social grouping was c oded as distinct from another when there was physical space between the groups of people and there was no interaction observed across the groups. The observation data represent a gross count of social interactions that occurred at a specified public place over a 5-second period of time. For instance, if a public place receives a score of 2 on “social interaction” that means that there were two social groupings observed at the randomly-selected time. One social gr ouping could have had two people sitting on a stoop talking, while the other social grouping had six people playing soccer together. The two social groupings could also have any number of other configurations; what is important is that they are social and socially distinct. On the other hand, when a public place receives a score of 0, the code could represent a situation in which no people at all were present. It could al so mean that there we re four people there, but not one of them was inter acting with any other person. Data analysis The mean number of public social in teractions observed over 414 repeated observations of 60 public places was .82 (S D = .76). The minimum number of social interactions observed was 0 and the maximu m number of social interactions was 10. Public social interac tion was highest around the taxi-trufi stand where vendors and drivers congregate, around the Villa Israel mark et, and in front of the UCE church. Social interactions, then, tend to develop around econom ic and religious centers. Public social interaction was lowest at the edges of Villa Isr ael, and particularly near the river that runs along the east side of the settle ment and the drainage ditches that run along the west side of the settlement. It is extremely signifi cant that Villa Israel’s waterways are the settlement’s social “black holes”, sin ce “watering holes” are generally known as

PAGE 170

152 Figure 6-2. Map of Villa Israel with dots representing 9 of 60 randomly sampled public places. Red dots represent areas with th e highest number of public social interactions, and blue dots represent areas with lowe st number of public social interactions. places of high social interaction in water-scar ce settlements. Of course, the Villa Israel’s waterways are dry most of the year, and this explains why they do not act as social hubs. Household-level data on Sociabili ty, Reciprocity, and Participation Respondent selection Because the household is the basic unit of resource acquisition and distribution in poor urban communities (Gonzlez de la Roch a 1994, Gonzlez de la Rocha and Gantt 1995), I chose it as the unit of analysis for the semi-structured interviews. In the field site chapter, I described the com position of households and the di vision of labor within them. In the field methods chapter, I explained that we only intervie wed people defined as “household heads”, or those who were respons ible for the acquisiti on and distribution of MARKET TAXI-TRUFI STOP DRAINAGE DITCH RIVERBED DRAINAGE DITCH UCE CHURCH

PAGE 171

153 household goods. Of the households we interv iewed, 41 percent were households with more than one head. Of those, two households had three heads (a mother, father, and teenaged daughter) during the course of the st udy; the rest had just two household heads. We interviewed multiple household heads because it helped us understand household decision-making, power dynamics, a nd division of labor. However, each household had to be represented by one head in the quantitative analyses to preserve the independence of observations and generalizab ility of the sample. I chose the household head that was most knowledgeable about resource acquisition and distribution to represent the household. Of the household heads chosen for quantitative analysis, 86 percent are female and 14 percent are male. Of these, three are girls under the age of 18 and one is a boy under the age of 18. Data collected at the household level Semi-structured interviews with household heads were us ed to collect recall data about the household headÂ’s social and economic activities over the la st week. We used a large number of cues to prompt respondentsÂ’ memories of social interactions, economic exchanges, and participation in meetings. Social interactions and economic exchanges were only counted if they took place be tween the respondent and a resident of another household located in Villa Isra el. We counted the interactio n and exchanges whether they occurred inside or outside Villa Israel. For in stance, if two Villa Israel residents meet up in downtown Cochabamba, this counts as a soci al exchange. Similarly, if one Villa Israel resident feeds a neighborÂ’s dog while the nei ghbor is out of town, this counts as an economic exchange. Each participant in the activity received credit for the economic exchange, whether that person was on the giving or receiving end.

PAGE 172

154 The semi-structured interviews yielded da ta on five variables relevant to the discussion here: visiting, food sharing, loan ing, helping, and community participation. First, we collected data on two measure of social interaction—visi ting and food sharing. The “visit” measure counts the number of visits people made or received in the last week. According to the native construct visitar (to visit), a visit occurs when one person intentionally goes to another pe rson’s home to see them. If the people run into each other on the street, at the store, or at a social even t, the interaction is not considered to be a visit. The “food sharing” measure counts th e number times a person gave or received edible foods in the last week. Invitar comida (to offer food) is a very important native category in which one person gives another f ood for immediate consumption. Foods that are commonly offered include a banana, a few mandarin orange segments, about 10 kernels of puffed corn, a small plate of fried intestines, or a cup of soda. Food sharing is used to establish a relationship, build ra pport, show friendship, or smooth over rocky relations. The food has no signifi cant practical purpose, as th e amount offered is usually less than a handful. Next, we collected data on two meas ures of economic exchange—loaning and helping. The “loan” measure counts the number of times a person gave or received an item to or from someone else in the last week. According to the native construct prestar (to loan or borrow), a loan occurs when one person gives another any material item, like money, a bucket of water, or a hammer. F ood ingredients that are not intended for immediate consumption, like onions, a bunch of bananas, or uncooked potatoes, are counted as loans, because giving th em is not considered an act of invitar comida in Villa Israel. The “help” measure counts the number of times people gave or received help over

PAGE 173

155 the last week. There is no strong native categ ory that describes labor assistance. Instead, we asked about a number of common favors people do for each other, including keeping an eye on a neighbor’s house or children, help ing someone carry heavy bags to or from the bus, or helping someone wash the dish es. I classify these favors as economic exchanges because the labor has an economic value. Some of the favors were remunerated at the moment the favor was done with cash, a meal, or a gift, and others were not. We coded them as help whether another exchange accompanied it or not. Finally, we collected data on one meas ure of participation in community organizations—the number of meetings that pe ople attended in Villa Israel over the last week. We asked people how many meetings of the Neighborhood Council or Villa Israel churches they attended in the last week. Wh ile there were also ot her organizations that met in Villa Israel, such as the School C ouncil or PLANE, their meetings were periodic and would have skewed the meeting attendance data. Data analysis In this section, I discuss the data from five measures of sociability, reciprocity, and community meeting attendance. Data were co llected over 10 months, from April 2004 to January 2005. The mean number of social inte ractions in which household heads reported participating was 6.72 visits and 12.34 incidents of food sharing over the week-long recall period. The mean nu mber of economic exchanges in which household heads reported participating was 3.72 loans and 5.64 helping events over the week-long recall period. Finally, the mean number of mee tings in which household heads reported participating was 0.34 for the week-long recall period. There are some interesting differences in the reciprocal behaviors of female and male household heads. T-test analyses reveal that women engage in significantly more

PAGE 174

156 food sharing (women = 13.47, men = 5.68, p = .02) and loans (women = 4.14, men = 1.24, p = .00) than men. However, there is no si gnificant difference in female and male household headsÂ’ participation in visiting, he lping, or attendance at community meetings. These data indicate that material goods like food, water, and mone y, tend to flow through womenÂ’s networks in Villa Israel, while in tangible exchanges lik e visiting, helping, and participation circulate eq ually among men and women. 6.72 12.34 3.72 5.64 0.34 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 VisitFoodLoanHelpMeetings Measures of Reciprocity, Sociability, and ParticipationCoun t Figure 6-3. Means for five measure of recipr ocity, sociability, and meeting participation. Looking over the mean measures of r eciprocity, sociability, and meeting participation, it seems clear that people engage most frequently in social interactions, less frequently in economic exchanges, and rarely in meetings. To understand why, we must consider the level of investment required for each kind of interaction, and the kind of payoff that can be expected from each one as well. Social interactions, such as visiting or sharing food, ge nerally require a relatively small capital investment. While the expected payoff is also quite small, social interactions provide people with a low-risk way to build social relationships and store small social obligations. In contrast, stored obligations to give a loan or help can be quite onerous. This explains why people tend to participat e more actively in social exchanges than

PAGE 175

157 economic ones. Community participation carrie s with it an even h eavier burden of time, financial, and social obligation, and the i nvestments do not necessarily result in the storage of explicit social or economic oblig ations. For this reason, people prefer making investments in interpersonal relationships rather than in organizational ones. Two excellent studies, one longitudinal (Gonzales de la Rocha 1994, 2001) and one cross-cultural (Moser 1996), indicate that ur ban reciprocal exchange systems are in decline. While those studiesÂ’ findings cannot be generalized to Villa Israel, the authors make strong arguments that the trends th ey found in Mexico, Ecuador, Hungary, the Philippines, and Zambia are global. While reciprocal exchanges may have declined over the last 15 years in Villa Israel as well, the 2004-2005 data show that Villa IsraelÂ’s household heads do participate actively in so cial and economic exchanges. They also indicate that people invest more heavily in social and intangib le economic exchanges than in tangible economic ones. Cultural Systems and Values are Unique This study analyzes the relationship between resources and reciprocity. The theory underlying it is a materialist one; it posits that changes in the environmental base lead to changes in the way people interact with each ot her. As I mentioned in the theory chapter, materialist theories in ecological anthropol ogy have historically been vulnerable to accusations that they are cra ss and environmentally deterministic. However, I do not believe these to be valid criticisms of this study for two reasons. First, it is not crass to make the obvious obs ervation that the presence or absence of some environmental resources determines how people interact. For instance, if there is water in the Villa Israel river, people meet there to wash laundry, talk to friends, and argue over who gets the best wa ter. If there is no water in the river, no one goes there.

PAGE 176

158 The observation is simple, but, as I show ed in the map above, it has enormous implications for how, when, and where people interact in Villa Israel. The second reason that the cri ticism is not valid is that ecological anthropologists do not contend that simply being able to predic t changes in the resource base allows us to predict how, when, and where pe ople will interact in any society. Laughlin and Brady, in their discussion of the scarcity response mode l, emphasized that reactions to scarcity are culturally distinct, and depend on the social and economic norms of each society. Messer, too, explained in her review of the literature on seasonal scar city that “the environment, while it explains the periodici ty and duration of social cycl es, does not adequately explain the form of sociality which prevails in each phase or the institutional forms which express it” (1988: 133). These scholars underscore th at while environment and economy are the most important drivers of scarcity response, they are do not act alone. Local cultures, embedded in history, place, and belief, must also be examined if we are to predict changes in social and economic inte ractions in specific societies. In the section that follows, I discuss ho w social interacti ons and reciprocal exchanges are expressed through four cultur al institutions in Villa Israel: family, neighborhood, religion, and business. In the literature on Latin America’s urban poor, scholars have shown how each cultural institution shapes reciprocal exchanges. However, these institutions are not the same in every society, and it is important to acknowledge when and how they vary across and within cu ltures. During the hei ght of the dry season, in August and September, we asked household heads to whom they go when they run out of basic necessities. Drawing on household head s’ responses and data collected during

PAGE 177

159 semi-structured survey interviews, I an alyze how family, neighborhood, religion, and business shape survival-oriented recipr ocal exchanges in Villa Israel. Data on Exchanges within Four Cultural Institutions Between August and September of 2004, we collected detailed data about household headsÂ’ reciprocal exchange partners over a week-long time period. The repeated measures ANOVA depicted in Figur e 6-4 shows that household heads conduct significantly more exchanges within business and family relationships, and fewer exchanges with neighbors and churchgoers (F = 5.265, p = .003). People loaned and 0.21 0.57 1.09 1.44 0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 ChurchNeighborhoodFamilyBusiness Number of Loans by Cultural Institution, Aug. Sept. 2004 Figure 6-4. Means for loans that took place within four cultural institutions in Villa Israel (church, neighborhood, family, and busin ess) during August and September 2004. borrowed the least, an average of .21 items with members of their own church. After that, they loaned and borrowed .57 items w ith neighbors. People loaned and borrowed more, an average of 1.09 items, with fam ily members. People loan and borrowed the most, an average of 1.44 items, with people they knew through a bus iness relationship, such as client-vendor, renter-landlord, or employee-employer relationships. It is important to note that only generalized re ciprocal exchanges are counted; balanced reciprocal exchanges like paying rent or buying goods are not counted here.

PAGE 178

160 Religion and church In the section on the field se tting, I explained that the ev angelical-Catholic divide has caused major tensions among Villa Israel residents. I noted that one reason for the rivalry is a struggle for political power. In this section, I explore another major reason that tensions developed between evange licals and Catholics— conflicts over how reciprocal exchange systems s hould be built and maintained. Historically, Latin American Catholicism produced hierarchical systems of reciprocal exchanges through the rotation of cargos fiestas and godparenting. Cargos are ritual offices that are held on a ro tating basis in Catholic communities. Cargo office holders redistribute their wea lth in exchange for prestige and political power (Cancian 1965, Dow 1997, in press). In Bolivia, fiestas that play homage to local virgins are a particularly important part of local cargo system (Buechler and Buechler 1971, Goldstein 2003). Parents also name godparents to establ ish formal relationships that connect children to powerful community members. The fiesta and godparenting traditions ensure that reciprocal relationships link together highand low-status families in reciprocal exchange relationships. Over the last 20 years, a wave of evange lical conversions has swept Latin America (Stoll 1993, Cleary 1997). In he r superb ethnography of Pentecostals in Bogot, Bomann argues that the believers’ relationship with the divine, quest for redemption, and spiritual passion are crucial to understanding this movement (1999). However, Bomann and other scholars agree that converts are initially attracted to the material benefits that evangelical churches offer. For instance, believers use the institutional resources and informal networks provided by evangelical churches to cope with poverty (cf. Mariz 1994). In Bolivia, the poorest shantytown dwellers are often drawn to evangelical churches because

PAGE 179

161 of the material benefits and social support they provide (Gill 1990, 2000). Among evangelicals, there is a strong prohibition ag ainst drinking, dancing, music, saints, and parties. Instead of spending on celebrations, evangelicals advocate saving and reinvesting family earnings in capitalistic en terprises (Dow 1997, in press). In Villa Israel, the prohib ition and stigmatization of parties, dancing, music, and drinking has thwarted the form ation of a Catholic reciproc al exchange system. Local Catholics are extremely bitter about this, and so me profess a deep hatred for Villa Israel’s evangelicals. One well-off Catholic retiree expl ained, “I like living in Villa Israel because it is tranquil, but I dislike that the majority of the people here ar e Protestants. Those people are shameless hypocrites.” He went on to discuss the evangelicals’ social shortcomings, “one Protestant here…is one of the cheapest people I have ever met in my life. Although he has quite a bit of money, he refuses to socialize with his equals… the Protestants always separate themselves in an antisocial manner whenever there is a community event” (Interview P334, 4/3/2004). Although most evangelicals refuse to interact socially with Cat holics, they do not believe that Catholics should be excluded from reciprocal exchange systems. For instance, Villa Israel’s UCE Church runs an excellent daycare program, and allocates treasured spots in the program to both eva ngelical and Catholic children. Evangelicals universally express the idea that recipro cal exchanges should be charitable. One evangelical woman explained how she implements this ideology in her own life: “I know the people here, and I know who most needs help. Usually it is women whose husbands have left them with small children, or re nters, or people who have serious marital problems. I would always choose to help the people who are most needy first” (Interview

PAGE 180

162 M116, 9/6/2004). Indeed, evangelicals did pa rticipate in significantly more loans (evangelicals = 4.31, Catholics = 2.86, p = .06) helping (evangelicals = 6.79, Catholics = 3.53, p = .02), and food sharing (evangelical s = 14.51, Catholics = 8.65, p = .02) than 4.31 6.79 14.51 2.86 3.53 8.65 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 16.00 LoanHelpFood Differences between Evangelicals and Catholics on Sociability and Reciprocity Evangelicals Catholics Figure 6-5. Means for the number of loans, helping, and food shari ng that evangelicals and Catholics engaged in betw een April 2004 and January 2005. Catholics over the 10-month st udy period. It is quite significa nt that, while evangelicals participate in more reciprocal exchanges than Catholics, there are fewer exchanges between churchgoers than any other group examined here. This indicates that evangelicals do exchange with people in the co mmunity at large, rather than just people who attend their own church. Evangelical reciprocal va lues are built around bottom-up charity, while Catholic reciprocal values are built around top-down hi erarchy. Catholics understand this, and feel their exclusion from access to political power, the community water system, and reciprocal exchange systems proves that evange licals are hypocrites. It is indeed difficult to reconcile evangelical charity with th e exclusion of Catholics from important community resources. Although I do not yet fully understand the dynamics at play, I believe that evangelicalsÂ’ pr ofessed belief in indiscrimina te charity is genuine and evidenced by their loan activity. However, I also suspect evangelical social exclusion of

PAGE 181

163 Catholics prevents them from building trust, familiarity, and social relationships with Catholics, and precludes th e establishment of meaningf ul reciprocal exchange relationships across religions. The evangelical prohibition on socializing with Catholics, then, effectively excludes Catholics from info rmal reciprocal exchange systems, even though evangelicals believe that Catholics s hould be beneficiaries of charity and reciprocal exchanges. Neighborhood In Latin America’s urban migran t neighborhoods, much ethnography has documented the pivotal role that neighbors pl ay in urban survival networks (Lomnitz 1977, Safa 1974, McFarren 1992, Halebsky 1995). In Villa Israel, only a few families maintain close reciprocal exchange relationshi ps with neighbors. For instance, one single mother works 12 hours a day in the cancha leaving her three child ren at home alone. She relies on her next-door neighbor to keep an eye on her children and home. When the neighbor was pregnant, the single mother provid ed her with a special bread, bananas, and medical advice. She also bri ngs the woman’s family meat regularly. The single mother said that, if she needed something, she would go to her neighbor before going to anyone else (Interview M8, 8/13/2004). Most Villa Israel household heads have a much lower level of trust with neighbors. Although some occasionally watch a neighbor’s house, loan a laundry brush, or give a bucket of water, many avoid getting involved in close, long-term reciprocal exchange relationships with neighbors. One woman e xplained why, “I would not want to ask for help from anyone. Here, people don’t help you. Th ey talk nicely to your face, and then they knife you in the back. If you ask them for help, they are just as likely to spread the news that you need help or be happy that you are in trouble. For that reason, I don’t tell

PAGE 182

164 anything personal to anyone and I don’ t trust anyone” (Inte rview M116, 9/6/2004). Another woman concluded, “If I am in need, I just have to endure it. Why would I go and ask for help? Everyone would just say no. They would just gossip about it. You have to keep it to yourself” (I nterview M190, 9/12/2004). Over the 18 months of field research, people constantly brought up the same themes when talking about reciprocal exchanges with neighbors: mistrust, humiliation, and gossip. In Villa Israel, people are unable to build trusting, reciprocal relationships with their neighbors because they fear that showing neediness will make them targets of gossip and humiliation. Gossiping has long been recognized as a tool that is used to suppress undesirable behaviors across cultu res (cf. Stack 1974, Hill and Gurven 2004). In Villa Israel, gossiping and humiliation suppress the appearance of neediness and participation in reciprocal exchanges. It is possible that a new value system—one that valorizes economic independence and deni grates economic interdependence—has emerged to support the widespread withdrawal from reciprocal exchange systems that was documented by Gonzlez de la Rocha and Moser among the urban poor. The stigmatization of neediness, then, may be a cu ltural belief that develops in systems that are transitioning from a reciprocal exchange system to a non-reciprocal one. To draw definitive conclusions about this, however, data would have to be collected using natural experiments or crosscultural research. Envy, gossip, and their efficacy in s uppressing community cooperation are common themes in anthropological studies of small-scale communities (e.g., Foster 1965). Shakow (personal communication) is currently engaged in fieldwork on envy in Choro, a rural town just outside Cochabamba Bolivia. Shakow argues that envy, anxiety,

PAGE 183

165 and community conflict are incr easing in response to the eco nomic inequities associated with neoliberal expansion (2005). Her fieldw ork also indicates that Cochabambans see envy as a growing moral ill that prevents them from creating successful economic development programs. Both Shakow and I have observed the presence of inequity, envy, gossip, and community conflict in Cochabamba. While we employ different theoretical approaches, we both believe that the trends are relate d to the expansion of neoliberalism and associated economic changes. An important cont ribution for future researchers will be to unravel the mechanisms—economic, po litical, and cultural—through which neoliberalism creates new configurations of cooperation and conflict. Such research will require fieldwork in a variety of contex ts and explicit cross-cultural comparison. Family In the Andes, survival networks have been constructed around extended kin groups for centuries, if not millennia (Moseley 1992, Kluck 1989, Isbell 1978, Lobo 1982). Among urban migrants, reciprocal networks frequently draw on nuclear family, extended family, and fictive kin (Lobo 1982). In Villa Israel, too, household h eads often rely on local family members for emergency loans. Many Villa Israel household heads said that they had a high le vel of trust with immediate family members. One woman who liv es across the street from her parents in Villa Israel explained that she “would never he sitate to borrow or take what I need from my mother, because no matter what happens, she is my mom and she has to help me” (Interview M253, 8/22/2004). Another woman w ith a large extended family in Villa Israel told us that if she needed help, she would go directly to her brother because “he is

PAGE 184

166 the only person who would never reject me. If he couldn’t help me, I would go to my other family members to ask fo r help” (Interview M194, 8/22/2004). Despite the high level of trust many pe ople develop with family members, reciprocal relationships do not flourish in all Villa Israel families. As one elderly woman said when asked what she would do if she ran out of essentials, “I couldn’t ask anyone for help, not even my daughter. I would di e, what else can I do?” (Interview M182, 8/20/2004). Another women explaine d, “I wouldn’t ask anyone for help. I’m afraid to ask and be rejected. Even my sister-in-law c ould say no. I don’t trust anyone enough to count on them or ask them for favor s” (Interview M342, 8/18/2004). In Villa Israel, fictive kin relationshi ps are neither common nor particularly important. Despite this, some Catholic house hold heads say that th eir godparents would help them out in an emergency. One woma n, who has no immediate relatives in Villa Israel, said that she “would try to go to th e store to buy what I need but, if I couldn’t, I would go to my godmother. I wouldn’t feel afraid to borrow anything, because I know that I would give it back” (Interv iew M106, 8/22/2004). Although people do see godparents as possible sources of help in an emergency, few maintain long-standing reciprocal exchange relationships with them or express a high degree of confidence that the godparents would help them out during a crisis. Businesses One surprising finding from the field research is the extent to which Villa Israel household heads use business relationships to bu ild reciprocal exchange partnerships. In the literature on urban reciprocity, there is little theo retical discussion of reciprocity among vendors, clients, renters, and em ployees. The only exception is Peattie’s discussion (1970) of monetizati on of reciprocal exchange re lationships. Peattie explained

PAGE 185

167 that, in the shantytown where she conducted re search, people paid for favors as a way to avoid the creation and storage of reciprocal obligations. In Villa Israel, people turn the tactic on its head, and use business relati onships to build reciprocal ones. When reciprocal exchanges take place under the cam ouflage of business relationships, people can avoid the stigmatization of neediness and enjoy the benefits of reciprocity. In Villa Israel, many household heads main tain reciprocal exchange relationships with owners of local corner stores. In excha nge for the credit that vendors extend, clients reciprocate with loyalty, information, and pa yments. One woman, who has several family members in Villa Israel and is a member of the UCE church explained, “I am afraid to borrow and ask for favors from people here. In an emergency, I would only go to the corner store to borrow the things I really need” (Inte rview M339, 8/22/2004). Another woman said she relies on credit from the corner store because “I know that I can pay them back somehow, but I do not go to othe r people because I know they would deny me help” (M338, 8/16/2004). Many others said th at they only trust the owner of their preferred corner store to help them during a crisis. Interestingly, household heads’ reliance on corner stores creates its own set of problems for store owners. One store owner sa id that she avoids giving customers items on credit because, “when you charge them the only thing you earn is enemies, and then they end up going to other people’s st ores” (Interview M181, 5/18/2004). Another storeowner said that he fre quently gives credit, but only to customers he knows and considers trustworthy (Interview P366, 9/15/200 4). The storeowners’ comments illustrate that clients expect to r eceive loans and gifts from storeowners, and punish the storeowners when they try to force bala nced or negative reci procal exchanges.

PAGE 186

168 Storeowners know that they must be very ca reful to establish reci procal relationships only when they have trust with a client. Among co-workers, employers, employees, land lords, and renters, close reciprocal relationships are also built in Villa Israel One restaurant owner maintains a close friendship with the owner of a nearby corner st ore. In a crisis, the restaurant owner said, she “would go to the one and only person I always go to for help (the store owner). She is the only person with whom I have trust” (Interview M99, 9/7/2004). Similarly, household heads that had renters or employees reported that they maintained extensive reciprocal exchange relationships that included food sh aring, numerous loans, and help with childcare. Based on observations and interviews, it ap pears that household heads prefer to create reciprocal relationships out of business ones for three main reasons. First, business partners have an economic in terest in the welf are of reciprocal partners. Second, reciprocal partners do not a ppear to be needy when exchanges occur under the cover of business relationships. Such relationships are rarely the objects of gossip in the community. Third, people can repay favors w ith money alone if they choose, and no long-standing social obligation will be create d. In this way, a recipr ocal relationship can easily be re-converted into a busines s relationship if things go sour. Conclusion In the first half of this chapter, I disc ussed the three dependent variables that are examined in this study. Drawing on classi c anthropological studies of social and reciprocal exchange, I explained how the Laughlin-Brady model predicts that resource scarcity will affect sociabili ty, reciprocity, and community participation. In the second half of the chapter, I examined how tw o important phenomena—nested cycles and

PAGE 187

169 cultural institutions—affect the Laughlin-Brady hypotheses. I showed that, while the long-term cycle of global capitalism has eroded urban reciprocal exchange systems, Villa Israel household heads do still engage in num erous reciprocal exchanges. I also showed that cultural institutions like religion, neighborhood, family, and business shape reciprocal exchanges in unique ways. In Villa Israel, conflicting value systems, mistrust, and gossip prevent people from creating exte nsive reciprocal exch ange relationships through religion and with neighbors. Instead, pe ople have developed reciprocal exchange relationships with business partners and fam ily members. Now that I have established how and why urban water scarcity is expect ed to affect sociab ility, reciprocity, and community participation, I w ill test the hypothesized relationships in the next chapter.

PAGE 188

170 CHAPTER 7 DATA ANALYSIS In this section, I test the four hypotheses presented in chapter one. In section one, I use repeated measures anal ysis of variance (ANOVA) to determine whether household heads engage in more social interact ions, economic exchanges, and community participation during the wet season than th e dry season. In section two, I develop a multiple regression model to determine wh ether household heads with more water participate in more social and economic intera ctions than those with less water. Finally, in the third section, I fit the regression model I developed in section two to 10 datasets, and show how the effect of water availab ility on sociability changes seasonally. I conclude that water availabil ity did affect sociability, but did not affect reciprocity or participation, in Vill a Israel during 2004-2005. Section 1—Changes in Water, Sociability, Reciprocity, and Participation over Time A Note on Methods of Analysis The statistical analyses in this section were performed using repeated measures ANOVA to determine if there are significant di fferences in mean measures, such as water scarcity and food sharing, over five study periods from April 2004 to January 2005. Since the repeated measures ANOVA procedur e analyzes only data with no missing observations, data from 57 of the 76 households surveyed were used in these analyses. For variables that do show a significant di fference over the five time periods, I use qualitative data from semi-structured survey interviews, key informant interviews, and participant-observations to explain change s over time. For some measures, the changes

PAGE 189

171 were clearly related to seasonal changes in water availability; for other measures, the qualitative data point to other causal mechanisms. Before I begin the discussion on temporal chan ge in Villa Israel, it is important to briefly revisit the issue of nested cycles. As I explained in chapter six, multiple nested cycles—ecological, economic, political, and social—act together to shape social and economic interactions in every society. In th e data presented here not all cycles act equally on all measures. My ability to disent angle the influences of rainfall, economic, and political cycles on social and economic in teractions depends, to a large extent, on knowledge I gleaned from in-depth interv iews and participan t-observation. Where possible, I let Villa Israel re sidents speak for themselves about how these cycles work. Often, however, people are unaware of how nest ed cycles shape their micro-interactions; in such cases, I rely on my own observation and analyses. Changes in Water Availability over Time In this section, I determine the extent to which water availability changes over time in Villa Israel. To do so, I examine three meas ures of water scarcity: rainfall, the waterbased task scale, and normative water use. I chose these three measures because they represent each of the three approaches I used to characterize water scarcity—security to water access, the experience of water scar city, and household wate r provision. I begin with a discussion of seasonal rainfall in Villa Israel. Secure access to water: Rainfall measure To understand the role that Cochabamba’s seasonal rainfall patterns play in creating urban water scarcity, it was essent ial that 2004-2005 follow a typical wet-to-dry season cycle. The graph below depicts actual seasonal rainfall patter ns across the study’s six data collection periods. The rainy s eason lasted from January to March 2004, and

PAGE 190

172 began again in late November 2004. The dry season lasted from March to November 2004. 296.1 28.9 39.9 6.2 83.4 168.1 0 50 100 150 200 250 300 350 P-O: JanFeb-Mar Time 1: AprMay Time 2: JunJul Time 3: AugSep Time 4: OctNov Time 5: DecJan Total Cochabamba Rainfall in mm over Six Study Periods, Jan. 2004 Jan. 2005 Rainfall Figure 7-1. Total rainfall in mm across six da ta collection periods, from January 2004 to January 2005, in Cochabamba, Bolivia. (R ainfall data from SENAMHI 2005). As I explained in chapter five, rainfall is a proxy measure for changes in water security over time. When it ra ins, Villa residents have acce ss to two free and unrestricted sources of water—rooftop rain water collection and the Vill a Israel River. In 2004, the last heavy rains fell between March 17 a nd March 20 (SENAMHI 2005). Soon after, the Villa Israel River went dry. Some house holds had access to stored rainwater through April and May. The first rains of the ne w wet season fell between November 16 and November 17, 2004 (SENAMHI 2005). This indi cates that Villa Israel residents had access to free and unrestricted water sources before May 2004 and after November 2004; they lacked access to free and unrestricted water sources from June 2004 to October 2004. While access to two secure water sources disappeared during the dry season, this does not necessarily mean that people in Vill a Israel experienced water scarcity once it stopped raining. In Villa Israel, people re ly on multiple water sources—market-based water trucks, the community tap stand syst ems, rainfall collection, and surface water

PAGE 191

173 sources—to obtain the water they need. Wh ile rainwater and surface water sources depend entirely on rainfall, the tap stand syst em and water deliver trucks do not. Using rainfall data alone, then, we cannot determin e whether seasonal rain fall patterns actually result in seasonal water scarcity in Villa Israel. To do so, we must examine a direct measure of the experience of water scarcity. Experience of water scarcity: Water-based task scale As I explained in chapter five, the experience of water scarcity is multifaceted, and involves feelings, conflicts, and other dime nsions. However, the measure that best represents the amount water people have in their homes is the Guttman scale of waterbased tasks. The Guttman measure uses fi ve indicators—house cleaning, bathing, doing laundry, cooking, and washing dishes—to dete rmine if a household had enough water to complete common water-based tasks in the la st week. A household with a score of 0 had enough water to do all five tasks, while water scarcity prevented a household with a score of 5 from doing all five tasks in the last week. The water-based task scale provides a sensitive measure of the amount of water s carcity households experienced in the last week. 1.12 1.33 1.04 0.77 0 0.2 0.4 0.6 0.8 1 1.2 1.4 P-O: JanFeb-Mar Time 1: Apr-May Time 2: Jun-Jul Time 3: Aug-Sep Time 4: Oct-Nov Time 5: Dec-Jan Mean Composite Scores of Water Scarcity, Jun. 2004 Jan. 2005 Water scarcity Figure 7-2. Mean composite household scores for Guttman task-based measure of water scarcity over four study periods, from J une 2004 to January 2005, in Villa Israel.

PAGE 192

174 Figure 7-2 depicts the mean scores of households over four study periods, from June 2004 to January 2005. A repeated meas ures ANOVA for the Guttman water-based task scale shows that there are small but fairly significant differences in the means over the four study periods (F = 2.40, p = .08). Th e means for June to July 2004 (m = 1.12), August to September 2004 (m= 1.33), and October to November 2004 (m = 1.04) are all between 1 and 2, which indicates that, on aver age, households struggled to get the water they needed to bathe and do laundry. In December 2004 and January 2005, the mean decreased to 0.77, which indica tes that households, on aver age, had enough water or struggled only to get the wate r they needed for house cleani ng. The data show that the experience of household water scarcity inte nsified in August and September 2005, at the height of the dry season. There are only four data points available for the Guttman task-based measure of water scarcity because it took over three mont hs to research, develop, and test the 33 indicators used for the Guttman scales. Ev en so, the four data points are enough to illustrate that temporal variations in the experience of water scarcity do correspond to seasonal variations in rainfall. In Figure 7-3, mean composite household scores for the task-based measure of water scarcity are pl otted against data for total rainfall over the four study periods. Because the scales for th e two measures are so different, the line graphs are shown without scales. The graphs clearly demonstrate that, as rainfall decreases between June and October 2004, the household experience of water scarcity increases. Conversely, as rainfall increas es between November 2004 and January 2005, the household experience of water scarcity decreases.

PAGE 193

175 Time 2: Jun-JulTime 3: Aug-SepTime 4: Oct-NovTime 5: Dec-Jan Plot of Rainfall against Water Scarcity Rainfall Water scarcity Figure 7-3. Total Cochabamba rainfall is plo tted against the task-based water scarcity measure for data collected between June 2004 and January 2005. The line graphs indicate that household perc eptions of water scarcity are linked to seasonal rainfall variations in Villa Israel. While the water delivery trucks and tap stand system clearly do mitigate householdsÂ’ dependence on seasonal water resources, they do not de-link them entirely (as they do in urba n centers with municipal water delivery). Instead, rainfall and surface water collection bi nd households to natural rainfall cycles, causing the experience of water scarcity to fluctuate seas onally in Villa Israel. Household water provision: Task-based water use estimate To assess the extent to which households su ffered from absolute water scarcity, we obtained estimates of household water use. Here, I examine data for the task-based water use estimate, which was introduced in chapter five. The task-based water use estimate reflects the amount of water household heads perceive that household members normally allocate to 13 tasks. Since the data reflect water use norms, the task-based water use estimate is not a sensitive measure of short-te rm changes in water av ailability. Instead, it provides a broad view of household-level e fforts to budget and plan for the use of existing water resources.

PAGE 194

176 Figure 7-4 depicts mean estimates of per-p erson per-day water use in Villa Israel households between April and November 2004. Looking at the mean per-person per-day water use over four study peri ods, it is clear that household members never came close to using the 50 L of water that are required to meet daily household needs in a lowtechnology setting. In fact, average househol d water deficiency ranged from 14.02 L per day (during April and May 2004) to 22.37 L (per day during October and November 2004). Although Villa Israel households su ffered from water scarcity throughout 2004, levels of household water use did vary acr oss the four data collection periods. 35.98 33.60 33.37 27.63 0.00 5.00 10.00 15.00 20.00 25.00 30.00 35.00 40.00 Time 1: Apr-MayTime 2: Jun-JulTime 3: Aug-SepTime 4: Oct-Nov Mean Water Use Estimates, Apr. 2004 Nov. 2004 Water Use Norm Figure 7-4. Mean estimates of household water use (per person, per day) over four study periods, from April 2004 to November 2004, in Villa Israel. A repeated measures ANOVA for the wate r use estimates shows that there are significant differences in the means over th e four study periods (F = 5.38, p = .003). In April and May of 2004, mean household water use was 35.98 L per person per day. As the dry season progressed, mean per-person per-day household water use decreased to 33.60 L in June and July 2004 and 33. 37 L in August and September 2004. Finally, in October and November 2004, mean household water use was at its lowest, 27.63 L per person per day.

PAGE 195

177 Because the research design originally incl uded only four survey cycles, we do not have data for the fifth survey period (in December 2004 and January 2005). When it still had not yet rained in October 2004, I asked the research team and respondents to participate in one extra round of interviews. Everyone agreed, with the stipulation that they would no longer do the despised water self-measure and the task-based estimate exercises in round five. The last cycle of interviews was less onerous for everyone involved, but, as a result, the extent to wh ich household water use changed after the rainy season began in late November remains a mystery. The data collected for April to Septembe r 2004 show that water use dropped as the dry season progressed. It is unsurprising that as water becomes scarcer, people report using less water. However, there are two asp ects of the graph that are unexpected. First, water scarcity worsened considerably from June and July 2004 to August and September 2004, but mean household water use only dropped by 0.23 L during that study period. Second, household heads reported that they decreased household water use by 5.74 L per person per day from August to November 2004. However, the water scarcity measure indicates that the water problem began to improve in October and November 2004. Why would people decrease water use just as the water scarcity problem begins to improve? A plot of water use against the rainfall data, shown in Figure 7-5, illustrates the two problems.

PAGE 196

178 Time 1: Apr-MayTime 2: Jun-JulTime 3: Aug-SepTime 4: Oct-Nov Plot of Rainfall against Water Use Norm, Apr. 2004 Nov. 2004 Rainfall Water Use Norm Figure 7-5. Mean Cochabamba rainfall is plotted against the household water use measure for data collected be tween April 2004 and November 2004. The two anomalies make sense if changes in normative water use lag behind real changes in water availability. This would explain why normative behaviors at time three seem more appropriate for time two rainfall levels, and normative behaviors at time four seem more appropriate for time three rainfall le vels. The data indicate then, that peopleÂ’s water budgeting behaviors are shaped by past experiences, rather than current or anticipated ones. Conclusion: Changes in water availability over time In Villa Israel, the experience of water scarcity is complex, and is shaped by a number of factors including economy, po litics, technology, and hydrology. Despite the complexity of water scarcity, it is possible to identify two simple water patterns in Villa Israel. First, the experience of water scarcity is closely related to rainfall patterns. While Villa Israel households use the water delive ry truck and tap stand systems all year, rainfall and surface water provide crucial s upplements to market-based water purchases. As a result, people have more water (from four sources) when it rains, and less water (from only two sources) when it does not ra in. Second, people budget water in response to past experiences of abunda nce or scarcity. For this r eason, people maintain the same

PAGE 197

179 level of water use after they ha ve experienced relatively stab le access to water, and they conserve water after they have experi enced decreasing access to water. Piecing together data for rainfall, water scarcity, and water use, we can see how water availability changes over time for residents of Villa Israel. During the wet season, from December to March, water availability is highest. In April a nd May, the rains end but people can still subsist on stor ed rainwater. In June and Ju ly, rainwater stores begin to run out, and household members begin to cons erve water by allocating less water to common household tasks. In August and Sept ember, experiences of water scarcity intensify, as rainwater stor es have long run out and no rain falls. In October and November, people plan for severe water s carcity and allocate le ss water to household tasks. In November, the first rains of the we t season begin to relieve water-related stress. In December and January, heavy rains fall again and fewer households experience extreme water scarcity. In the next section, I examine changes in sociability and reciprocity over time. I also relate those changes to seasonal water availability in Villa Israel. Changes in Sociability, Reciprocity, a nd Community Participation over Time Over ten months of survey research, th e research team collected data on seven measures of sociability, recipr ocity, and community participati on in Villa Israel. In this section, I discuss three measures of sociabil ity (visits, food shari ng, and public social interactions), two measures of reciprocity (loans and helping), and two measures of community participation (church attenda nce and Neighborhood Council attendance). I begin with an analysis of changes in sociability.

PAGE 198

180 Sociability: Changes in food sharing over time The first measure of sociability discussed here is the number of times Villa Israel residents shared food in the last week. Fi gure 7-6 depicts the mean number of food sharing events in which household heads par ticipated over five study periods. A repeated measures ANOVA shows that there are sign ificant changes in f ood sharing across the five study periods (F = 4.09, p = .01). The chan ges in food sharing shown here appear 7.97 19.84 9.83 6.84 12.97 0.00 5.00 10.00 15.00 20.00 25.00 Time 1: AprMay Time 2: JunJul Time 3: AugSep Time 4: OctNov Time 5: DecJan Mean Food Sharing, Apr. 2004 Jan. 2005 Food Sharing Figure 7-6. Mean numbers of food sharing over five study periods, from April 2004 to January 2005, in Villa Israel. exactly as predicted by the Laughlin-Brady model. As water scarcity intensifies, between April and July 2004, there is a sharp incr ease in food sharing. Then, in August and September, as the dry season progresses, food sharing drops off. In August, an evangelical woman recounted what happened when she offered food to one of her neighbors at Villa Israel’s Sunday market, “I gave her some of my lunch, on the top of the pot I was eating out of. Sh e did not give me anything in return because she doesn’t know how to invitar comida She hides her food as she eats it. You have to teach people how to share food” (Interview M8, 8/13/2004). In July and August, several other women also brought up the topic in interviews, and e xplained to me that pe ople in Villa Israel

PAGE 199

181 just do not know how to share food properl y. Clearly, the dry season change in food sharing habits does not escape people ’s attention in Villa Israel. However, the idea that people “just don’t know how to share food” is not borne out in the data. Food sharing reaches a low point during the crisis m onths of October and November, when water, income, and market activity are all scarce. In November, one Villa Israel resident explained how the three factors interrelate, “The market is empty, and no one has any money. There isn’t any rain either, so that things can be produced again” (Interview M32, 11/10/2004). In respons e to this overwhelming scarcity, people avoid maintaining social relationships th at might be a draw on scarce economic resources. In December and January, when resources are abundant again, people resume food sharing activities. The data indicate, the n, that people do know how to share food, but they only choose to do so when they are in a stable economic position. In the next section, I explore how and why social intera ctions increase in December and January. Sociability: Changes in visiting over time The second measure of sociability discussed here is the number of visits made and received by household heads in Villa Israel. Below, Figure 76 depicts the mean number of visits in which household heads particip ated over five study periods. A repeated measures ANOVA shows that there are signif icant changes across the five study periods (F = 3.23, p = .02).

PAGE 200

182 4.30 7.00 7.75 7.05 11.53 0.00 2.00 4.00 6.00 8.00 10.00 12.00 14.00 Time 1: AprMay Time 2: Jun-JulTime 3: AugSep Time 4: OctNov Time 5: DecJan Mean Visiting, Apr. 2004 Jan. 2005 Visiting Figure 7-7. Mean number of visits over fi ve study periods, from April 2004 to January 2005, in Villa Israel. In particular, there are two sizeable in creases in visiting—one between April and July 2004 and another between October 2004 and January 2005—across the ten months of survey research. Both increases are hypothesized by the Laughlin-Brady model. As water scarcity begins to worsen between Ap ril and July of 2004, people intensify social interactions. Then, as water becomes avai lable during the wet season in December 2004 and January 2005, social inter actions increase again. During the wet season in December and Ja nuary, however, the increase in visiting is higher than we would predict as a response to changes in wa ter availability alone. That increase in visiting is actually related to ra infall, holiday, and economic cycles that recur each year. As one Villa Israel entrepreneur explained in October 2004, “The economy in Cochabamba is bad now, as it always is at this time of year. But it will pick up again in December, as always.” (Interview M101, 10/19/2004). In December, the onset of the rainy season frees up time and capital that had been spent on the aguatero Two important holidays, Christmas and the New Year, also help stimulate the Cochabamba economy. The sharp increase in visits in December and January, then, reflects the

PAGE 201

183 combined effects of the start of the ra iny season, holiday soci al activity, and the revitalization of the Cochabamba economy. One unexpected finding here is the small increase in visiting between July and August of 2004. Following the Laughlin-Brady m odel, we would expect people to visit less as the dry season progresses, from A ugust to November. However, in August and September, the number of visits in which pe ople participate actually increases; visiting only decreased during October and November of 2004. Although the drop in visiting was delayed, people’s reasons for withdrawing from visiting relationships did coincide with the hypotheses under study here. For instance, one mother explained that she did not make any social visits in early November b ecause she “did not have any money left over” and visiting always involves some sma ll expenditure (Interview M32, 11/10/2004). Many other respondents said that they did not have time to go and visit people, as they were occupied with numerous economic activities as they tried to make ends meet. As predicted, then, as water scarcity worsened, people intensified income-generating activities and withdrew from social relationships. Sociability: Changes in public social interactions over time The third measure of sociability examined here is public social interaction. Figure 7-8 depicts changes in public social interactio ns over four study periods in Villa Israel. A repeated measures ANOVA shows that ther e are small but moderately significant changes in public social interactions ove r the four study periods (F = 2.72, p = .04).

PAGE 202

184 0.91 0.77 0.86 0.75 0.00 0.20 0.40 0.60 0.80 1.00 Time 2: JunJul Time 3: AugSep Time 4: OctNov Time 5: DecJan Mean Public Social Interactions, Jun. 2004 Jan. 2005 Public Social Interactions Figure 7-8. Mean number of public social intera ctions over four study periods, from June 2004 to January 2005, in Villa Israel. In June and July, public social interacti ons average .91 interactions per site, and they decrease to .77 in August and September. The spike and drop in interactions do, to this point, follow the changes predicted by the Laughlin-Brady model. However, after October, trends in public social interactio ns are the opposite of what we expect. In October and November, public social interacti ons increase, and then they fall again in December and January. After a review of the data from participant-observation and semistructured survey interviews, I could find no reason for the observed changes in public social interaction. Unlike the patterns in food sharing and vi siting, public social interactions do not appear to follow the trend predicted by th e Laughlin-Brady model. Public social interactions involve ve ry little expenditure of time and re sources, and so there is probably no reason to reduce them during the dry season. Instead, some variable other than water scarcity must drive change in public social interactions, and that variable was outside of the scope of this study.

PAGE 203

185 Reciprocity: Changes in loans over time A repeated measures ANOVA showed no significant difference in loans (F = 1.89, p = .13) over the five study periods. Reciprocity: Changes in helping over time A repeated measures ANOVA showed no significant difference in helping (F = 1.82, p = .14) over the five study periods. Participation: Changes in church attendance over time A repeated measures ANOVA showed no significant difference in church attendance (F = 1.79, p = .13) over the five study periods. Participation: Changes in neighborhood council meeting attendance over time The last measure examined here is a ttendance at Neighborhood Council Meetings. Figure 7-9 depicts changes in Neighborhood Council meeting attendance over five study periods in Villa Israel. A repeated me asures ANOVA shows that there are very significant changes in public so cial interactions over the four study periods (F = 6.59, p = .006). 0.35 0.26 0.09 0.050.05 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 Time 1: AprMay Time 2: JunJul Time 3: Aug-Sep Time 4: OctNov Time 5: Dec-Jan Mean Neighborhood Council Attendance, Apr. 2004 Jan. 2005 Neighborhood Council Attendance Figure 7-9. Mean number of public social intera ctions over four study periods, from June 2004 to January 2005, in Villa Israel.

PAGE 204

186 As the graph shows, changes in Neighborhood Council meeting attendance do not follow the pattern predicted by the Laughlin-B rady model. Instead, meeting attendance decreased consistently throughout the study period. A crisis in the leadership in the Neighborhood Council, which unfolded during 2004, caused the drop-off in meeting attendan ce that year. In April and May of 2004, meeting attendance and participation was rela tively high. Then, news of several serious embezzlement scandals involving the Neighbor hood Council hit the community. In June, there was also a public fistfight between the Council leader and a Council board member. All of these events shook community membersÂ’ trust in the leadership of the Neighborhood Council. In June, several boar d members resigned from the Neighborhood Council in protest of the curre nt leader. However, community members voted to give the leader one last chance to right the wrongs of the past. In la te August, the leader finally resigned, and was replaced in an emergency el ection. After August, pe ople said that the change in leadership was an improvement, but few showed any interest in attending meetings. As a result, attendance at Neighborhood Council meetings dropped off considerably during the last four months of the study period. This drop appears to have had little or nothing to do w ith increasing water scarcity. Conclusion: Changes in sociability, reci procity, and participation over time In this section, I presented the results of repeated measures ANOVAs for data on water availability, social in teractions, economic exchanges, and community participation. The data showed that there were significan t changes in water availability and social interactions over the 10-m onth study period. There were no significant changes in economic exchanges and community partic ipation over the 10-month study period. Clearly, it is important to ask why the hypot hesized effects app ear only in social

PAGE 205

187 interactions, and not for economic or part icipatory behaviors. To understand how and why people disengage in patterned ways, we must first understand why people engage in each kind of interaction. Soci al interactions, economic exchanges, and community participation fulfill different needs in th e life of a Villa Israel resident. People engage in social interactions for tw o reasons. First, visiting and food sharing are fun. People participate in social intera ctions to pass the time, gossip, and joke. Second, social interact ions provide an important founda tion for the initiation of later economic exchanges. While social interactio ns also occur between people who already have economic relationships, they are initi ated more frequently, occur between less trustworthy partners, and can broken off more easily than economic relationships. In contrast, reciprocal economic relationships allow each partner to draw on the otherÂ’s goods, labor, and social capital. As a result economic relationships are normally only established between partners wi th a long social history, high tr ust, and a strong interest in maintaining the relationship. To preserve economic relationships, people avoid refusing requests and making requests that will be re fused. Like economic exchanges, church participation occurs as part of a very st rong commitment to religion and to a religious community. People who stop attending church ar e targets of gossip and criticism in the community. Participation in the Neighborhood C ouncil is required and enforced with a small fine, but is also widely shirked by Villa Israel household heads. Severe short-term scarcity, it seems, impacts low-trust, low-commitment relationships first. Since soci al relationships are not instru mental to peopleÂ’s survival, they break them most easily. Additionally, people also appear to cut off social interactions that act as gateways to ec onomic interaction. People who are engaged in

PAGE 206

188 reciprocal relationships avoi d social interactions with existing and potential economic partners, so they will not be put in a pos ition to refuse requests when times are tough. People who are actively engaged in reci procal economic exchanges and church attendance, however, continue these inter actions. This is because disengaging from economic and religious commitments would i nvite censure and damage one’s long-term economic and social security. Severe short-term scarcity, then, does not appear to prompt a withdrawal from interactions within high-trust, high-commitment relationships. While ecological and economic scarcity did cause household heads to withdraw from social interactions, the 2004 dry s eason was not severe enough to prompt a withdrawal from economic exchanges and chur ch participation. Had it not begun to rain, however, it seems likely that the social and economic impacts of water scarcity would have increased during November and December. Section 2—Explaining Social In teractions across Households In chapter one, I hypothesized that h ousehold heads with more water will participate in more social interactions and economic exchanges than those with less water. Here, I test those relationships by bu ilding a multiple regression model for each of the dependent variables. Identifying Independent Variables To analyze trends over the entire 10 months of field research, I took the mean of the data collected for each variable over the five data periods. I then identified six potential independent variables: (1) self-administered direct measure of water use, (2) task-based water use estimate, (3) overall es timate of weekly water use, (4) task-based Guttman scale of water scar city, (5) water storage capac ity, and (6) household economic well-being. As I explained in ch apter five, all of th ese variables were designed to capture

PAGE 207

189 some aspect of the water scarcity concept. Variables 1, 2, and 3 were designed to measure water availability. Variable 4 was designed to measure experiences of water scarcity. Variable 5 was designed to measure water secu rity. Variable 6 was designed to measure access to water markets. While other measures (such as distance from the cistern route and Guttman scaling of water conflicts) were also designed to captur e unique elements of the water scarcity concept, those measures we re exploratory, and are too noisy for use in a regression analyses. In Table 7-1 and Figure 7-10, I show the re sults of a principal components analysis of the six variables under study here (with no rotation). I al so performed the principal components analysis with varimax and equama x rotations, and the factor loadings were the same. In factor one, three variables lo ad high: Water Norm (.95), Water Use (.86), and Water Estimate (.86). Factor one clearly re presents water provisi on. In factor two, the Guttman measure of water scarcity load s high (.76), while the economic well-being measure loads low (-.77). Factor two, then, represents water scarcity loaded against economic well-being. Factor three represents water security, and only the water storage capacity variable loads high (.98). The factor loadings support the argument I made in chapter five that water scarci ty has three aspects: water pr ovision, the experience of water scarcity, and water security. These three aspe cts, in addition to the measure of economic well-being that opposes water scarcity on compon ent three, will be used to predict social interactions in the multiple regression models developed here. While I could have used the individual factor loadings as independent variables in the regr ession model, I have chosen to use the raw data for three reasons. Firs t, as I explained earlier in this chapter,

PAGE 208

190 Table 7-1. Factor loadings for six potentia l predictors of social interactions, for summaries of data collected over five interview periods. Component loadings 1 2 3 Water Norm 0.95 0.18 -0.08 Water Use 0.86 0.36 0.00 Water Estimate 0.86 0.13 0.05 Econ. Well-being 0.37 -0.77 0.23 Guttman Water Scar. -0.40 0.76 0.12 Water Storage Cap. 0.00 0.09 0.98 Percent of Total Variance Explained 44.51 % 22.35 % 17.35 % Factor Loadings Plot 1 0 0 5 0 0 0 5 1 0F A C T O R ( 1) 1 0 0 5 0 0 0 5 1 0F A C T O R ( 2 ) 0 5 0. 0 0 .5 1 0F A C T O R ( 3 ) WATERNORMWATERUSEWATERESTIMSTORAGEECONLEVGUTTASKAVE Figure 7-10. A plot depicting f actor loadings for three factors. Factor one represents water provision, factor two represen ts economic well-being against water scarcity, and factor three represents water security. I do not have data on water norms and water use for time period five Without period five data, it would be impossible to fit a regressi on model using factor loadings to the dataset for time period five. Second, as I explained in chapter five, the three water provision measures are highly correlated, so it is unlikely that using a summary measure would

PAGE 209

191 significantly improve predictive power over using one of the thr ee variables. As a result, I have chosen to use the water estimate vari able, for which I have complete data, to represent the water provision factor in the model here. Third, I have worked throughout this study to create clear and simple analys es that have high expl anatory power. As I argued in chapter five, doing so helps resear chers who lack the time and funding I had in Villa Israel to reproduce the methods in othe r contexts. By testing the effects of four simple variables—water estimate, storage capacity, water scarcity, and economic wellbeing—I facilitate cultural anthropologists’ ability to reproduce the methods of data collection and analysis in ot her ethnographic contexts. Building the Regression Model Using the four independent variables identi fied here, I fit a regression model to the summary data to predict two measures of so ciability (visiting and food sharing). It is appropriate to question the va lue of using a regression m odel to predict two closely related dependent variables, like visiting and food sharing (Pearson correlation = .67, p = .000). Both measures were actually desi gned to operationalize the same concept— sociability—so it is unsurprising that they ar e so highly correlated. However, as similar as the two measures are, ther e are two important differences between them. Visiting is a high-commitment social activity with no re source exchange required, while food sharing is a low-commitment social activity in wh ich resource exchange must occur. The implications of these two conceptual differe nces will be explored in the section that follows this one. The independent variables had no significan t effect on the other four dependent variables (loans, helping, church a ttendance, and Neighborhood Council meeting

PAGE 210

192 attendance). The results of the first two models are presented in Table 7-2; the results of the remaining four models ar e contained in the appendix. Table 7-2. Regression models predict two ki nds of social intera ctions, using four independent variables. Data are means for five data periods. Dep. Variables N Data source R sq P Indep. Variables Coeff. P Visiting 72 Means (Apr 2004-Jan 2005) 0.13 0.05Water storage 0.21 0.07 Water estimate 0.24 0.05 Gutt. W. Scarcity 0.10 0.46 Econ. Well-being 0.12 0.39 Food Sharing 72 Means (Apr 2004-Jan 2005) 0.09 0.16Water storage 0.26 0.03 Water estimate 0.11 0.38 Gutt. W. Scarcity -0.12 0.36 Econ. Well-being -0.02 0.91 The model accounts for a modest and si gnificant amount of the variation in visiting. It is not significantly associated with food sharing. Looking at the independent variables, it is clear that there are some pr oblems. Water storage is the only independent variable that is significantly associated w ith both dependent variables. Water estimate is significantly associated with visiting, but not with food sharing. The Guttman water scarcity and economic well-being measures are not significantly associat ed with either of the dependent variables. I argue that two of the inde pendent variables, Guttman water scarcity and economic well-being, should be removed from the model. While it is important to avoid data-driven model building, there are legitimate met hodological reasons to exclude these two variables. As I explained in chapters three and five, Guttman water scarcity and economic well-being are both ordinal-level measures that were designed duri ng the field research.

PAGE 211

193 The assignment of scores on the economic well-being measure was done cooperatively by the research team after the research was conducted, so th ere are no inter-rater reliability scores to test the reliability of the scoring. The Guttman water scarcity scale is an experimental measure that has never b een tested before. The removal of the two measures from the model is warranted because the measures are not reliably coded, welldefined, or extensively tested. As a result, they are probably t oo noisy to contribute meaningfully to the analyses here. I stress, however, that fu ture studies should attempt to collect better data on economic well-being and further de velop Guttman scales of water scarcity so that these variables can be included in regression models. Understanding the Regression Model Once the Guttman water scarcity and economic well-being measures are eliminated, the resulting multiple regression m odels account for a modest but significant amount of variation in both vi siting and food sharing. Since th ere is neither a statistical nor a conceptual association between the two independent va riables (Pearson correlation = .03, p = .77), the inclusion of an interaction variable would be inappropriate for this model. The parameters for models predicting visiting and food sharing appear in Table 73 below. The independent variables still had no signi ficant effect on the other four dependent variables (loans, helping, church a ttendance, and Neighborhood Council meeting attendance). The results of the remaining f our models are contai ned in the appendix. The model accounts for 12 percent of the va riance in visiting behaviors (p = .01). Both water storage and water estimate variab les have a weak positive association with visiting. The food sharing model accounts for 8 pe rcent of variation in food sharing (p =

PAGE 212

194 .06). While there is a signifi cant association between storag e capacity and food sharing, the water estimate is not significan tly associated with food sharing. Table 7-3. Regression models predict two ki nds of social inte ractions, using two measures of water availability. Data are means for five data periods. Dependent Variable N Data source R sq P Independent Variables Coeff.P Visiting 72 Means (Apr 2004-Jan 2005) 0.120.01Water storage 0.23 0.05 Water estim. 0.25 0.03 Food Sharing 72 Means (Apr 2004-Jan 2005) 0.080.06Water storage 0.25 0.04 Water estim. 0.12 0.29 The water storage variable is a better pred ictor of both social in teraction variables. When household heads can better budget water use, they engage in more social interactions. People who constantly live with the threat that water will run out, however, avoid and are excluded from social interactio ns. This is probably because people with little water security (1) are too stressed out to really enjoy social inte ractions and (2) want to avoid getting entangled in economic obligations that th ey can not fulfill. People who are better able to budget their water use are probably more able to enjoy themselves and less wary of incurring economic obligations. The water estimate variable is only signi ficantly associated with visiting, not food sharing. Household heads who believe that they have more water engage in more visiting than household heads who believe they have less water. As with the water storage variable, people who have more water are more able to have fun and more willing to incur economic obligations. This explains why th ey are less avoidant of social situations. The question remains, however, why is the wa ter estimate associated with only visiting, and not food sharing?

PAGE 213

195 The explanation is that visi ting behaviors are more sensitive to changes in water scarcity than food shar ing. There are two reasons for this First, visits occur inside people’s houses, so it gives people the opportunity to assess one another’s water provision and storage situation. In contrast, much food sharing o ccurs in the street, and so it does not make anyone vulnerable to predat ory water borrowing or gossip about water poverty. Second, visiting is a much more intimate behavior that food sharing. Once someone has entered another person’s home, it is easy for one person to obligate the other to do a favor. Food sharing, however, is much mo re casual, and can easily be cut short or disengaged from without giving offense. In the next section, I explore the use of regression models to examine how these two variables predict social interaction when water is available and when it is not available. Section 3—Explaining Household Social Interactions over Time The Laughlin-Brady model predicts that so ciability and reciprocity will only be affected when resources are scarce. If the re source supply is abundant, then, there should be no correlation between resour ce provision and sociability. To test this using the water data, I tested the regression models using 10 datasets (from five study periods) to predict visiting and food sharing. The model only fits th ree of the datasets—for the data collected during the height of the 2004 dray season. Ta ble 7-4 shows the re sults for the three regression models. The independent variables are not associated with variance on the dependent variables in the other 27 mode ls, with one exception. The regression model accounts for 11 percent of the variance in “helping” beha vior in the period two dataset (p = .03). However, since this one finding is not expl ained by theory and does not fit the overall

PAGE 214

196 pattern in the data, I do not cons ider it to be significant. Th e appendix contains the results of the other seven regression models for vi siting and food sharing, and the 20 regression models fitting the data to the remaining four dependent variables. Table 7-4. Regression models predict visi ting and food sharing for two study periods: June—July 2004 and August—September 2004. Dependent Variable N Data source R sq P Independent Variables Coeff. P Visiting 65 Jun 2004 -Jul 2004 0.110.02 Water storage 0.33 0.01 Water estim. 0.07 0.54 Visiting 60 Aug 2004 -Sep 2004 0.230.001Water storage 0.27 0.03 Water estim. 0.38 0.002 Food Sharing 60 Aug 2004 -Sep 2004 0.140.01 Water storage 0.33 0.01 Water estim. 0.18 0.16 Together, water storage capacity and the wa ter use estimate account for a moderate amount of the variation in vi siting and food sharing during th e dry season. In June and July, the model explains 11 percent of the va riance in visiting (p = .02). In August and September, the model explains 23 percent of the variance in visiti ng (p = .001). In August and September, the model explains 14 percen t of the variation in food sharing (p = .01). In each of the models, water storage yielded a positive, significant correlation to social interactions. The water use estimate yielde d a positive, significant correlation with visiting only during August and September. In that model, water estimate had the strongest and most significant correlation of any predictor used in the three models. During the other study periods and for food sharing, the water estimate did not account for any variation in the dependent variable. Using the models, we can identify several im portant trends in the data. First, water availability is only significantly associated w ith social interaction dur ing the height of the

PAGE 215

197 dry season. Second, household heads began to cu t back on visiting early in the dry season (in June and July), but they did not reduce food sharing until later in the dry season (in August and September). Third, household heads with less water storage capacity were affected earlier and reduced both kinds of so cial interactions, while people who believed they had less water were affected later in the dry season and reduced only visiting behaviors, and not food shari ng. Fourth, when the water estimate emerges as a significant predictor of social interacti on, it exerts a stronger and more significant effect than the storage capacity variable. Looking at the three regression models, the mechanisms that mediate social responses to water scarcity become clear. Du ring the rainy season, water availability has no effect on social interactions. As the dry season sets in, household heads cut back on social visits. As the dry season advances, th ey also cut back on food sharing. Household heads with less storage capacity are initially more responsive to the onset of the dry season than are household heads with less wate r, because they are more conscientious about budgeting social and economic expenditu res. However, when water scarcity becomes severe, people who believe that th ey have less water cut back on social interactions more drastically than do people with less storage space. Based on these findings, I would predict th at, had the dry season been worse: (1) people with less water would cut back on food sharing; (2) then, people with less storage capacity would reduce economic exchanges; a nd (3) finally, people with less water would reduce economic exchanges. Data Analysis Conclusions In this chapter, I examined the effects of urban water scarcity on social interactions, economic exchanges, and community particip ation. In the first section, I examined

PAGE 216

198 changes in water provision and scarcity ove r time in Villa Israel. Using repeated measures ANOVAs, I identified the two depende nt variables that changed in response to water scarcity in Villa Israel—visiting and food sharing. In the sec ond section, I created a multiple regression model that tests the effects of water availability on social interactions for the Villa Israel data. I explained how a nd why two measures of water availability (water storage and water estimate) predict tw o measures of sociability (visiting and food sharing). Finally, in the third section, I tested the fit of the regression models for the data from each of the five study periods. I showed that water scarcity affects sociability only during the height of the dry season. During th e wet season and trans itory periods, water availability variables are not correlated w ith measures of sociability, reciprocity, and community participation. I also showed that water storage capacity affects sociability earlier in the dry season, but that the water estimate has a stronger effect once dry season water scarcity intensifies. I conclude that the 2004 dry season in Villa Israel was not severe enough to cause effect s on reciprocity or community participation. I hypothesize, however, that under more severe water scarci ty conditions, sociability, reciprocity, and community participation would continue to change as predicted by the Laughlin-Brady model.

PAGE 217

199 CHAPTER 8 CONCLUSIONS In this last chapter, I discuss the study’ s results and their implications for theory, practice, and future research. In the firs t section, I summarize th e research agenda, methods, and findings. In the second section, I discuss the study’s contributions to theory and practice. In the thir d section, I explore the implications for future research in cultural anthropology. Section 1—Summary of Findings This research examines the effects of urban water scarcity on reciprocity and sociability. Following Laughlin and Brady’s 197 8 model of human adaptation to resource scarcity, I hypothesized that f our effects could be observed in Villa Israel: (1) household heads would participate in more social inte ractions during the wet season than during the dry season, (2) household heads would par ticipate in more reciprocal economic exchanges during the wet season than duri ng the dry season, (3) household heads with more water would participate in more social interactions than those with less water, (4) household heads with more water would pa rticipate in more reciprocal economic exchanges than those with less water. To test these hypotheses, I developed meas ures of water scarcity, sociability, and reciprocity that were appropr iate to the ethnographic cont ext of Cochabamba, Bolivia. Measures of water scarcity were designed to function in an urban context where water comes from multiple sources, including a ma rket-based system, a politicized community system, and seasonal rainfall-based sources. I designed three approaches to measuring

PAGE 218

200 water scarcity: water provision, water securit y, and the experience of water scarcity. Each of the approaches contributes unique inform ation about the emics and etics of water scarcity. Measures of reciprocity and sociability were designed to capture the complex social landscape of a Bolivian shantytown, including the heterogeneity of origins, religion, ethnicity, household composition, and economic well-being in the research population. I designed three measures of soci ability, two measures of reci procity, and two measures of community participation. Each of the measures represents emic categ ories of sociability, reciprocity, and participati on, but they also operationaliz e the concepts in clear, reproducible ways that contribute to our etic understanding of the topic. After four months of part icipant-observation, data were collected over 10 months, in five rounds of survey research and direct observation. The results of repeated measures ANOVAs show that participation in two kinds of social interacti ons (visiting and food sharing) changed over time. Participation in public social inter actions, reciprocal exchanges, and community meetings did not change over time. A multiple regression model showed that two measures of water ava ilability (water storag e capacity and water use estimate) had a moderate, positive associ ation with social in teractions. The other measures were not significantly correlated w ith the dependent variables. However, by fitting the regression model to datasets from 5 study periods, it became clear that water scarcity only affected vis iting and food sharing during th e dry season; there were no effects during the rainy and transition seasons. The findings indicate that the Laughlin-Brady model does accurately predict the effects of urban water scarcity, to a point. As the dry season progresses, household heads

PAGE 219

201 engage in fewer social interactions. Peopl e with less water and less ability to budget water use engage in fewer social interac tions. Changes in reciprocity and community participation did not occur as predicted. Ec onomic and religious relationships are much more valuable than social ones, and water scarcity was not severe enough to motivate people to break valuable ties. Participa tion in the Neighborhood Council changed in response to cycles other than seasonal wate r scarcity. I c onclude that seasonal water scarcity in Villa Israel was not severe enough to cause all the effects hypothesized by the Laughlin-Brady model during 2004-2005. This research described changes in wate r scarcity, sociability, reciprocity, and community participation during a typical seasona l cycle in Villa Israel. As a result, the findings should be helpful in drawing conclu sions about survival ta ctics in water-scarce urban areas during a typical dry season. I do believe, however that during an atypical year with more severe or prolonged water scarcity, we could observe the effects on reciprocity and community participation predicted by the Laughlin-Brady model. Section 2—Contributions to Theory and Practice Theoretical Contributions to Anthropology In this section, I examine the study’s contri butions to anthropological theory. First, I discuss urban water scarcity and the surv ival strategies of the urban poor. Then, I discuss human adaptations to resource scarcit y. Finally, I discuss the findings in view of the larger project of theorybuilding in cultural anthropology. Urbanism, Poverty, and Water Scarcity In this study, I argue that the emergen ce of urban water scarcity can best be explained using a political ecological appro ach. In chapter one, I explained that a center/periphery urban structure, character ized by the inequitable distribution of

PAGE 220

202 municipal services to outlying residents, de veloped in many Latin American cities. In chapter five, I showed that such an urban st ructure exists in Cochabamba, Bolivia and has caused a serious water scarcity problem to develop among residents of the cityÂ’s south side. I explained that histori cal factors, including BoliviaÂ’s national economic crises, the hydrological characteristics of the Cochabamba Valley, political decisions of Cochabamba policymakers, and the settlement patterns of highland migrants, have all contributed to the exclusion of the urba n poor from CochabambaÂ’s municipal water system. I concluded that, to understand the em ergence of urban water scarcity in local contexts, it is necessary to examine hydrological social, political, and economic factors. I also argued that water scar city is a complex concept th at should be studied using etic and emic approaches. First, measures of water provision help us assess the severity of unmet biophysical needs for water. Second, measures of water security help us understand how ecological, technological, a nd infrastructural f actors shape water scarcity. Third, measures of the experience of water scarcity help us understand how culture mediates the ways in which water scar city is felt and unders tood in local contexts. Using all of these approaches helps us crea te a more complete picture of the causes, distribution, and effects of water s carcity within an d across cultures. Urban Reciprocity Another important phenomenon under study here is decline of reciprocal exchange systems. As I showed in chapter six, economic conditions in the 1960s and 1970s supported the creation of reciprocal netw orks. After the 1980s, economic crises and neoliberal restructuring underm ined reciprocal exchange relationships. My research on urban reciprocity in Villa Isra el indicates that new systems of belief and behavior have begun to develop in response to new economic conditions.

PAGE 221

203 First, I found that there is a strong prohibition on ec onomic interdependence among neighbors in Villa Israel, which is enfor ced with gossiping and public embarrassment. Second, I found that, while household heads avoid creating reciprocal exchange relationships with neighbors, they maintain extensive reciprocal exchange relationships with family members and business partners (like landlords, renters, storeowners, and clients). Third, I found that reciprocal exch anges are not distributed equally over all community members. Catholic exchange s systems, which are built around an unsustainable and stigmatized fiesta system, are in decline. In contrast, evangelicals are actively involved in reciprocal exchange s, which are built around an accumulationoriented value system rather than a redistri butive one. Together, thes e findings indicate that new values and reciprocal exchange stru ctures have emerged to replace the old ones, and they are more compatible with the new economic system than the old values and reciprocal exchange structures. Human Adaptations to Resource Scarcity This studyÂ’s main focus is on testing the applicability of th e Laughlin-Brady model for urban water scarcity. The results show that the Laughlin-Brady model does accurately predict the effects of water scar city on sociability in Villa Israel. This is particularly significant because is it indicates that, how ever complex the societies and scarcity phenomena we study, we can still disentangl e the effects of ecological and economic cycles on social and cultural phenomena. It al so provides further evidence that material changes in the ecology and economy are the pr imary drivers of changes in human belief and behavior. In this study, I used both ideographic and nomothetic theories to explain the effects of urban water scarcity on reciprocity and soci ability in Villa Israel Ideographic theories

PAGE 222

204 advance case-specific explanations, while nomothetic theories advance universal explanations that apply to many cases (Ber nard 2002: 77-83). By using both kinds of theory, I showed that each one has much to contribute to our unde rstanding of how and why people respond to resource scarcity. First, I used ideographic (local and regional) theories to examine urban water scarcity, soci ability, and reciprocity in Villa Israel. Then, I used a nomothetic theory of culture cha nge to explain why water scarcity affects sociability in Villa Israel. As I explained in chapter one, ecologi cal anthropologists have spent decades embroiled in debates to determine the extent to which changes in the resource base cause changes in human belief and behavior. Following the work of political ecologists, I have argued here that the relations hip between resource scarcity and human behavior is strong, but not mechanistic. Changes in the availabi lity of ecological and economic resources are the primary drivers of culture change. Howeve r, the cultural, politic al, and psychological dynamics that create and mediate human percep tion of resource scarcity also shape the ways—collective and indivi dual—that people respond. Interdisciplinary Contributions The research was designed to contribute to two interdisciplinary research areas: environmental sustainability and human wellbeing. In this section, I examine how the research findings advance interdisciplinary research in each of these study areas. The research was designed to examine hum an-environment interactions in a waterscarce urban environment. This study made several contributi ons to this research area. First, the study contains an in-depth ethnographic descript ion of life in a water-scarce urban area. The ethnographic description s hows how extractive technologies and distributive systems shape human water use. Second, it details the political ecology of

PAGE 223

205 urban water scarcity in Cochabamba, Bo livia. The political ecological analysis demonstrates how power, politics, and econom ics shape access to na tural resources in developing cities. Third, it pr esents a three-pronged appro ach to understanding the urban water scarcity concept. The analysis of wate r provision, water security, and water scarcity experience illustrate that “u rban water scarcity” is a co mplex concept that contains elements of biophysical need, infrastruc tural adequacy, and culturally-mediated perception. Fourth, the research presents twelve reproducible ways to measure the urban water scarcity concept. By presenting the definition, operationaliz ation, and descriptive statistics for each measure, the study facilita tes scientific research on the human aspects of urban water scarcity. Fifth, the study demonstrates the utili ty of nested cycles (as in nested seasonal and economic cycles) and tr ansformation (as in the transformation of reciprocal relations through new social cons tructions like Latin American evangelism) in explaining the human responses to changes in the resource base. In doing so, the study shows how theories of social-ecological systems can help explain cycles of human behavior. Together, these findi ngs reveal new information a bout the workings of human systems of water distribution and consumption, and point to future avenues for research on human-environment interactions around water, as well as to programs for ameliorating problems of urban water scarcity. In addition to its focus on environment, the research was designed to examine how water scarcity affects human well-being, pa rticularly at the family, community, and cultural levels. The study contri buted to this research area in several ways. First, the study identified the people responsible for th e acquisition and distri bution of important household resources, including water. The research revealed that child-headed

PAGE 224

206 households are most vulnerable to water scarc ity, and that dual-headed households were most successful in acquiring and distributing the resources they need. Second, the study examined water allocation across thirteen household tasks. It showed that people suffering from water scarcity cut back firs t on hygienic activities, and then on water consumption. Third, the study showed th at water provision and budgeting affect household headsÂ’ ability to par ticipate in social interactions which act as a gateway to reciprocal economic exchanges. Over the l ong-term, this may affect household headsÂ’ ability to participate in reciprocal favors a nd exchanges. Fourth, it showed that the local water distribution system provides unequal access to water resources for community residents. This explains why many Villa Israel residents do not benefit from the water the system provides. Fifth, the politicization of community-level water systems also causes people to avoid participating in the maintenan ce and extension of local water systems. As a result, the infrastructure has fallen into disr epair with extensive water leaks and at least one non-functional tap stand. Sixth, over time serious tensions have developed around the water issue at the community level. As a result, the Water Committee is defunct, community members refuse to fund soluti ons that the Neighb orhood Council pursues, and the creation of community-oriented soluti ons to water problems has been inhibited. Together, these findings reve al new information regardi ng how urban water scarcity affects family and communities, and point to future avenues for research about the cultural aspects of water scarcity. Drawing on ethnographic, observational, and survey data, the study makes contributions to two areas of interdiscipl inary research. First, it examines hydrologic aspects of human-environmen t interactions. It shows how human systems of water

PAGE 225

207 extraction and distribution are created and main tained in developing cities. It also shows that the survival of the most vulnerable popul ations is closely tie d to natural hydrologic systems, because vulnerable populations are ma rginalized from municipal water systems. Second, the study examined the effects of urban water scarcity on one vulnerable population, residents of Villa Is rael. It shows that, although h ouseholds vary in terms if their ability to access water resources, the enti re community is affected by the stresses, politics, and conflicts that deve lop around urban water scarcity. Ending Urban Water Scarcity: Six Recommendations As I mentioned in chapter three, this re search project was conceived and executed with the help of Susana Southerwood a nd Abraham Aruquipa, tw o hydrologic engineers from the water-delivery NGO Water for People. While in the field, I was conscious of the need to generate applied findings that would help real-world practitioners like Susana and Abraham with their work to end urban water scarcity. While practitioners are certainly capable of drawing their own conclusions fr om the findings presented here, I hope to facilitate the process by pointing to six s uggestions for improving water delivery and water relief projects. First, while water scarcity is a seriou s problem for people all year, water relief projects are urgently needed for residents of water-scarce communities during the dry season. Between the months of June and Oct ober, many families are in need of better access to water, cheaper water, and more wate r. Any kind of relief project that could distribute or facilitate th e distribution of water duri ng the dry season would be enormously helpful to residents of communities like Villa Israel (Recommendation 1). Water relief projects would probably prevent hunger, disease, and hygienic problems during the dry season. If water relief projects are implemen ted, they should target the

PAGE 226

208 neediest households first. Re latively affluent household heads can be incredibly savvy about appropriating food, mate rials, jobs, and money that are meant for the poorest households. To avoid this, relief projects shou ld distribute relief to households in the following order: child-headed households households headed by elderly women, households headed by single mothers, hous eholds headed by out-of-work men, and renters (Recommendation 2). The qualifications of project particip ants can be easily verified by checking with the owne rs of nearby corner stores. The most serious obstacles to the implementation of community-based water projects appear to be (1) political corrup tion that targets development projects and (2) power struggles over access to and control of resources. It is extremely important for practitioners to do everything they can to prevent local corrupti on and power struggles (Recommendation 3), because such incidents live on in communities members’ memories and undermine their willingness to participate in community projects for years. However, corruption and power struggles are complex problems, and the solutions lie far beyond the scope of this study. Even when nothing can be done about local corruption and power struggles, practitioners can improve water distribution by working with water delivery truck operators. Currently, water delivery trucks provide the most reliable, accessible, and affordable water access for most community members—and everyone uses them. However, water delivery service could be im proved in two ways. First, people who live far away from main roads cannot get reliab le access to the trucks. By negotiating for a fixed route that includes back roads, even just one day a week, NGOs could significantly improve access to water sources for familie s in outlying areas (Recommendation 4).

PAGE 227

209 Second, while most families can afford to buy the water they need from the water delivery truck, some families simply cannot By implementing a system of vouchers, subsidies, or graded pricing, water provision levels could be improved significantly for the most vulnerable members of wate r-scarce neighborhoods (Recommendation 5). While it is true that water de livery trucks are independent profit-oriented operators, they are also accustomed to negotiating over th e price and distribution of water. NGOs command a considerable amount of power a nd capital across the developing world, and could do much to make water markets physic ally and economically accessible for all community members. Projects that focus on water delivery truc ks could probably do more to improve water distribution than any other kind of project in water-scarce urban communities like Villa Israel. Finally, this study shows that, over the c ourse of the dry season, household water storage capacity does more to improve househol d well-being than any other kind of water infrastructure. Having more water storage cap acity allows people to negotiate for bulk water discounts and to budget their water use more effectively. This indicates that wateroriented NGOs should consider building water storage capacity instead of or in addition to making improvements to water distributi on infrastructure (R ecommendation 6). The importance of water storage cap acity, of course, is relate d to household dependence on water markets, the reliability of those water markets, and the availability of a functioning local water distribution system. In Villa Israel, however, conditions would be ideal for a pilot program and further study of the propositio n that water storage capacity is the best infrastructure for protecting families from water scarcity.

PAGE 228

210 Section 3—Looking Ahead in Cultural Anthropology Moving Forward: Anthropological Re search on Urban Water Scarcity As I explained in chapter one, social scie ntists have only rece ntly begun to focus their attention on urban water sc arcity. Because the area of inquiry is so new, it presents many opportunities for anthropologists to cont ribute. In this section, I briefly examine three potential areas for future research on urban water scarcity. First, anthropologists should do much more to understand urban water scarcity. To begin, we must determine where and wh en urban water scarcity emerges around the world. To build a cultural an thropology of urban water scar city, we must examine both material and cultural mechanisms that shape urban water scarcity, including water distribution systems, water ex traction technologies, and soci o-economic systems. We also need to test the efficacy of the three appro aches to understanding water scarcity presented here, and develop better measures as our understanding of the phenomenon deepens. Second, I asserted that dry season water s carcity in Villa Isra el was not severe enough to all cause all the eff ects predicted by the Laughlin-B rady model. In the future, we should determine whether very severe wate r scarcity affects sociability, reciprocity, and community participation as predicted by the Laughlin-Brady model, or not. It is, of course, very risky to plan a field resear ch study around seasonal urban water scarcity, since the seasonal rainfall cycles are ge nerally unpredictable. However, by doing comparative research in communities with differe nt levels of water scarcity, we can also determine the effects of severe water scarcity on the dependent variables. In such studies, it will also be important to determine how th e relative importance of different aspects of water scarcity (like water security, water prov ision, and market efficiency) change as dry season unfolds.

PAGE 229

211 Third, we should use cross-cultural resear ch to determine how the socio-economic effects of urban water scarcity are expressed in different cultural systems. In particular, the study should be reproduced in societies with different econo mic bases, social structures (such as family and religion), and kinds of reciprocal exchange systems. This will help us understand how culture mediates the effect of resource changes on human belief and behavior. It will also help show that materialist explanations of culture change (Steward 1959, Carneiro 1970, Rappaport 1974, Harris 1979) are not mechanistic, but rather make full use of culture in developing explanations for social phenomena. Cultural AnthropologyÂ’s Future: Collaboration, Method, and Theory Collaboration As I explained in chapter one, natural scientists have become increasingly interested in collaborating w ith social scientists to understand socialecological systems. Such collaborations have enormous potential to yield influential theory and practical applications. Generally, natural scientists are interested in social scientistsÂ’ ability to explain how human beliefs, perception, and be havior affect (and are affected by) the natural resource base. This is th e kind of contribution that any social scientist can make to an interdisciplinary collaboration. However, anthropologists also have unique knowledge that no other discipline can offer, including knowledge of human biophysical and cultural adaptations, human manipulation of the na tural and constructed environments, the historical depth of human experience, and a comparative, evolutionary perspective. Anthropologists are in a unique position to contribute to studies of humanenvironment interactions, then, because they offer more sophisticated analyses of the human-environment nexus than any other social science discipline. Ecological anthropologists laid the th eoretical foundations of ecological anthropology (Steward

PAGE 230

212 1959, Carneiro 1970, Rappaport 1974), created in fluential integrative theories (Wolf 1969, Vayda and McCay 1975, Harris 1979), embr aced political ecology (Durham 1979, Schmink 1982, Stonich 1989, 1993), and continue to innovate in their use of socialecological and new ecological theory (van der Leeuw and Redman 2002, Abel and Stepp 2003). To remain at the forefront of innovativ e interdisciplinary co llaborations, however, cultural anthropologists maintain the attention to methodol ogical rigor that our colleagues in biological anthropology and archeology have. Method To remain in dialogue with other discip lines, cultural anthropologists must use scientific methods. At a minimum, we shoul d continue to produ ce rigorous ethnographic work, make plain our sampling and data co llection methods, use reproducible methods, explicate causal links, and test hypotheses relati onships with statistical analyses. We must also be clear about whether our argument s are based on reliable data, exploratory evidence, or conjecture. Over the last two decades, cultural anthropol ogists have continued the disciplineÂ’s tradition of creating in novative research methods. The an thropological tool kit contains many new items, including consensus analysis (Romney et al. 1986), personal network analysis (Bernard et al. 1990, McCarty and Wutich 2005), economic experimentation (Henrich et al. 2004), and eco logical experimentation (Cas agrande et al. in press). Methodological innovation is particularly important in cultural anthropology because anthropologists are often positioned to observe the emergence of new social phenomena and structures before anyone else. Ethnography gives us knowledge of these new cultural constructions, but methods give us the abil ity to report on those findings in concise, scientific ways.

PAGE 231

213 Theory For the past 25 years, there has been something of a paradigmatic shift in anthropological research, from scientific m odel-building to humanistic interpretivism. Scientific model-building us es reproducible methods to test hypotheses and make falsifiable contributions to knowledge. In contrast, human istic interpre tive research identifies new discourses, challenges domina nt narratives, encourages reflexivity, and promotes engagement with the power and politics. Both are focu sed on understanding cross-cultural changes in human belief a nd behavior, and both make important contributions to the development of anthropology as a discipline. However, when humanistic interpretivism goes unchecked by scientific contributions, these strengths can become weaknesses. In cultural anthropology, for instance, scant evidence has occasionally been used to posit grandiose theories of crosscultural change. Conversely, ethnog raphic research can become so particularistic that it is disengaged from the effort to create crosscultural theories. Both trends are undesirable because they undermine the quality of res earch produced, prevent anthropologists from influencing interdisciplinary collaborations and social po licy, and detract from the strength and growth of the discipline. By balancing critical, discourse-oriented research with scientific, data-based research, anthropology has the potential to gr ow and improve. Anthropological theory is at it most powerful when humanistic and scie ntific approaches are used to complement each other. For instance, cultural anthropologi stsÂ’ ability to put themselves in unusual ethnographic contexts, identify new discourses, and analyze their idio matic implications is one of the disciplineÂ’s gr eat contributions to social science. But we must also

PAGE 232

214 determine the extent to which new discourses an d behaviors are global, regional, or local. Once we do, we can explain how and why they develop as they do. In my work here, I opera tionalized Ennis-McMillanÂ’s bodi ly distress concept, and determined the extent to which it can be a pplied to water scarcity in Villa Israel. I determined that, while Villa Israel resident s do not express concern about water scarcity using the idiom of bodily distress, emotion a nd conflict are important aspects of the water scarcity experience. Like Ennis-McMilla n, I analyzed the political, ecological, and economic aspects that shape urban water scarcity. Combining our work, a cultural anthropology of urban water scarcity that is interpretive, scientific, cross-cultural, and explanatory begins to emerge. For interpretivists and scientists to build an integrated cultural anthropology, we must read each otherÂ’s work and use it in the field. While combining approaches is not as common as it ought to be, it is very important for the future development of cultural anthropology. To my way of thinking, the biggest challe nge for scholars of my generation is to bridge the gap between humanism and sc ience that develope d during the 1980s and 1990s. For scientists, the challenge is to fi nd novel ways of operationalizing and studying the questions humanists raise. For humani sts, the challenge is to understand how discourses can be operationalized and explained by mid-range and nomothetic theories of cross-cultural change. If young scholars can find ways to cross sub-disciplinary boundaries, we will facilitate future collabora tions, preserve the strengths of cultural anthropology, and create resear ch that influences the way pol icymakers build our world.

PAGE 233

215 APPENDIX REGRESSION MODELS Regression Models with Four Independent Variables, with Data for Means from Five Study Periods Table A-1. Regression model predicts loans us ing four independent variables. Data are means for five data periods. Dep Var: ITEM N: 72 Multiple R: 0.239 Squared multiple R: 0.057 Adjusted squared multiple R: 0.001 Standard error of estimate: 3.444 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 6.204 1.863 0 3.329 0.001 WATERESTIM 0.034 0.033 0.126 0.928 1.02 0.311 STORAGE 0 0 0.047 0.975 0.391 0.697 ECONLEV -1.096 0.591 -0.257 0.733 1.855 0.068 GUTTASKAVE -0.188 0.458 -0.056 0.761 0.411 0.683 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 48.233 4 12.058 1.017 0.405 Residual 794.546 67 11.859

PAGE 234

216 Table A-2. Regression model pred icts helping using four inde pendent variables. Data are means for five data periods. Dep Var: HELP N: 72 Multiple R: 0.169 Squared multiple R: 0.029 Adjusted squared multiple R: 0.000 Standard error of estimate: 6.500 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 6.42 3.517 0 1.825 0.072 WATERESTIM 0.016 0.063 0.033 0.928 0.262 0.794 STORAGE 0 0 0.16 0.975 1.311 0.194 ECONLEV -0.427 1.116 -0.054 0.733 0.382 0.703 GUTTASKAVE -0.496 0.865 -0.079 0.761 0.573 0.568 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 83.413 4 20.853 0.494 0.74 Residual 2830.867 67 42.252 Table A-3. Regression model predicts participation in Neighborhood Council meetings using four independent variables. Da ta are means for five data periods. Dep Var: JV N: 71 Multiple R: 0.286 Squared multiple R: 0.082 Adjusted squared multiple R: 0.026 Standard error of estimate: 0.205 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 0.159 0.113 0 1.416 0.162 WATERESTIM -0.004 0.002 -0.236 0.932 -1.93 0.058 STORAGE 0 0 0.034 0.977 0.285 0.777 ECONLEV 0.031 0.035 0.122 0.736 0.886 0.379 GUTTASKAVE -0.029 0.027 -0.144 0.759 -1.06 0.293 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 0.248 4 0.062 1.471 0.221 Residual 2.785 66 0.042

PAGE 235

217 Table A-4. Regression model pr edicts participation in c hurch meetings using four independent variables. Data are means for five data periods. Dep Var: IG N: 72 Multiple R: 0.102 Squared multiple R: 0.010 Adjusted squared multiple R: 0.000 Standard error of estimate: 0.795 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 0.452 0.43 0 1.05 0.298 WATERESTIM 0.005 0.008 0.082 0.928 0.653 0.516 STORAGE 0 0 0.062 0.975 0.505 0.615 ECONLEV -0.028 0.136 -0.029 0.733 0.206 0.837 GUTTASKAVE Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 0.442 4 0.11 0.175 0.951 Residual 42.374 67 0.632 Regression Models with Two Independent Models, with Data for Means from Five Study Periods Table A-5. Regression model predicts loans us ing two independent variables. Data are means for five data periods. Dep Var: ITEM N: 72 Multiple R: 0.074 Squared multiple R: 0.005 Adjusted squared multiple R: 0.000 Standard error of estimate: 3.485 Effect Coefficient Std Error Std Coef ToleranceT P(2 Tail) CONSTANT 3.47 0.815 0 4.258 0 WATERESTIM 0.018 0.032 0.068 0.999 0.57 0.571 STORAGE 0 0 0.025 0.999 0.207 0.837 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 4.566 2 2.283 0.188 0.829 Residual 838.213 69 12.148

PAGE 236

218 Table A-5. Regression model pred icts helping using two independent variables. Data are means for five data periods. Dep Var: HELP N: 72 Multiple R: 0.154 Squared multiple R: 0.024 Adjusted squared multiple R: 0.000 Standard error of estimate: 6.422 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 4.841 1.502 0 3.224 0.002 WATERESTIM 0.016 0.06 0.033 0.999 0.273 0.786 STORAGE 0 0 0.149 0.999 1.252 0.215 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 68.751 2 34.376 0.834 0.439 Residual 2845.529 69 41.24 Table A-6. Regression model predicts participation in Neighborhood Council meetings using two independent variables. Data are means for five data periods. Dep Var: JV N: 71 Multiple R: 0.182 Squared multiple R: 0.033 Adjusted squared multiple R: 0.005 Standard error of estimate: 0.208 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 0.196 0.05 0 3.952 0 WATERESTIM -0.003 0.002 -0.18 0.999 1.509 0.136 STORAGE 0 0 0.028 0.999 0.239 0.812 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 0.1 2 0.05 1.16 0.32 Residual 2.933 68 0.043

PAGE 237

219 Table A-7. Regression model pr edicts participation in c hurch meetings using two independent variables. Data are means for five data periods. Dep Var: IG N: 72 Multiple R: 0.098 Squared multiple R: 0.010 Adjusted squared multiple R: 0.000 Standard error of estimate: 0.784 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 0.387 0.183 0 2.11 0.038 WATERESTIM 0.005 0.007 0.075 0.999 0.625 0.534 STORAGE 0 0 0.06 0.999 0.502 0.617 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 0.408 2 0.204 0.332 0.718 Residual 42.408 69 0.615 Regression Models with Two Independent Models, with Data from Five Study Periods Table A-8. Regression model predicts food sh aring using two inde pendent variables. Data were coded from interviews co nducted during study period 1 (April-May 2004). Dep Var: FOOD1 N: 68 Multiple R: 0.189 Squared multiple R: 0.036 Adjusted squared multiple R: 0.006 Standard error of estimate: 11.121 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 6.311 2.1 0 3.006 0.004 STORAGE 0.001 0 0.191 0.981 1.55 0.126 ESTIM1 -0.006 0.064 -0.012 0.981 0.101 0.92 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 298.508 2 149.254 1.207 0.306 Residual 8038.713 65 123.673

PAGE 238

220 Table A-9. Regression model pr edicts visiting using two in dependent variables. Data were coded from interviews conduc ted during study period 1 (April-May 2004). Dep Var: VISIT1 N: 68 Multiple R: 0.122 Squared multiple R: 0.015 Adjusted squared multiple R: 0.000 Standard error of estimate: 5.258 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 3.585 0.993 0 3.611 0.001 STORAGE 0 0 0.018 0.981 0.147 0.884 ESTIM1 0.029 0.03 0.118 0.981 0.952 0.345 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 27.223 2 13.611 0.492 0.613 Residual 1796.895 65 27.645 Table A-10. Regression model predicts loans us ing two independent variables. Data were coded from interviews conducted dur ing study period 1 (April-May 2004). Dep Var: LOAN1 N: 68 Multiple R: 0.134 Squared multiple R: 0.018 Adjusted squared multiple R: 0.000 Standard error of estimate: 3.520 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 2.937 0.665 0 4.419 0 STORAGE 0 0 0.125 0.981 1.009 0.317 ESTIM1 -0.011 0.02 -0.067 0.981 0.541 0.59 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 14.658 2 7.329 0.592 0.556 Residual 805.342 65 12.39

PAGE 239

221 Table A-11. Regression model predicts helpi ng using two independent variables. Data were coded from interviews conduc ted during study period 1 (April-May 2004). Dep Var: HELP1 N: 68 Multiple R: 0.141 Squared multiple R: 0.020 Adjusted squared multiple R: 0.000 Standard error of estimate: 5.442 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 3.5 1.028 0 3.406 0.001 STORAGE 0 0 0.142 0.981 1.144 0.257 ESTIM1 -0.002 0.031 -0.007 0.981 0.057 0.955 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 39.058 2 19.529 0.659 0.521 Residual 1924.942 65 29.614 Table A-12. Regression model predicts participation in Neighborhood Council meetings using two independent variables. Data were coded from interviews conducted during study period 1 (April-May 2004). Dep Var: JV1 N: 66 Multiple R: 0.176 Squared multiple R: 0.031 Adjusted squared multiple R: 0.000 Standard error of estimate: 0.655 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 0.194 0.126 0 1.539 0.129 STORAGE 0 0 0.164 0.983 1.308 0.196 ESTIM1 0.001 0.004 0.048 0.983 0.383 0.703 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 0.87 2 0.435 1.012 0.369 Residual 27.069 63 0.43

PAGE 240

222 Table A-13. Regression model pr edicts participation in c hurch meetings using two independent variables. Data were coded from intervie ws conducted during study period 1 (April-May 2004). Dep Var: IG1 N: 67 Multiple R: 0.212 Squared multiple R: 0.045 Adjusted squared multiple R: 0.015 Standard error of estimate: 1.307 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 0.36 0.248 0 1.448 0.153 STORAGE 0 0 0.129 0.981 1.043 0.301 ESTIM1 0.009 0.008 0.152 0.981 1.23 0.223 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 5.134 2 2.567 1.503 0.23 Residual 109.284 64 1.708 Table A-14. Regression model predicts food sh aring using two inde pendent variables. Data were coded from interviews conducted during study period 2 (June-July 2004). Dep Var: FOOD2 N: 65 Multiple R: 0.069 Squared multiple R: 0.005 Adjusted squared multiple R: 0.000 Standard error of estimate: 28.820 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 18.127 5.967 0 3.038 0.003 STORAGE 0.001 0.001 0.066 0.997 0.518 0.606 ESTIM2 -0.037 0.189 -0.025 0.997 0.194 0.847 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 246.205 2 123.102 0.148 0.863 Residual 51496.349 62 830.586

PAGE 241

223 Table A-15. Regression model predicts loans us ing two independent variables. Data were coded from interviews conducted during study period 2 (June-July 2004). Dep Var: LOAN2 N: 65 Multiple R: 0.024 Squared multiple R: 0.001 Adjusted squared multiple R: 0.000 Standard error of estimate: 5.682 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 4.923 1.176 0 4.185 0 STORAGE 0 0 0.004 0.997 0.032 0.975 ESTIM2 -0.007 0.037 -0.024 0.997 0.189 0.851 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 1.17 2 0.585 0.018 0.982 Residual 2001.815 62 32.287 Table A-16. Regression model predicts helpi ng using two independent variables. Data were coded from interviews conduc ted during study period 2 (June-July 2004). Dep Var: HELP2 N: 65 Multiple R: 0.324 Squared multiple R: 0.105 Adjusted squared multiple R: 0.076 Standard error of estimate: 9.174 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 2.89 1.899 0 1.521 0.133 STORAGE 0.001 0 0.304 0.997 2.529 0.014 ESTIM2 0.048 0.06 0.096 0.997 0.799 0.427 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 611.583 2 305.791 3.633 0.032 Residual 5218.356 62 84.167

PAGE 242

224 Table A-17. Regression model predicts participation in Neighborhood Council meetings using two independent variables. Data were coded from interviews conducted during study period 2 (June-July 2004). Dep Var: JV2 N: 64 Multiple R: 0.207 Squared multiple R: 0.043 Adjusted squared multiple R: 0.012 Standard error of estimate: 0.440 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 0.152 0.092 0 1.653 0.103 STORAGE 0 0 0.207 0.997 1.647 0.105 ESTIM2 -0.001 0.003 -0.028 0.997 0.223 0.824 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 0.53 2 0.265 1.367 0.263 Residual 11.829 61 0.194 Table A-18. Regression model pr edicts participation in c hurch meetings using two independent variables. Data were coded from intervie ws conducted during study period 2 (June-July 2004). Dep Var: IG2 N: 64 Multiple R: 0.134 Squared multiple R: 0.018 Adjusted squared multiple R: 0.000 Standard error of estimate: 0.755 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 0.295 0.158 0 1.87 0.066 STORAGE 0 0 0.12 0.997 0.942 0.35 ESTIM2 0.002 0.005 0.054 0.997 0.425 0.673 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 0.634 2 0.317 0.556 0.577 Residual 34.803 61 0.571

PAGE 243

225 Table A-19. Regression model predicts loans us ing two independent variables. Data were coded from interviews conducted dur ing study period 3 (August-September 2004). Dep Var: LOAN3 N: 60 Multiple R: 0.146 Squared multiple R: 0.021 Adjusted squared multiple R: 0.000 Standard error of estimate: 4.980 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 2.477 1.123 0 2.206 0.031 STORAGE 0 0 0.11 0.998 0.835 0.407 ESTIM3 0.028 0.039 0.092 0.998 0.7 0.487 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 30.947 2 15.474 0.624 0.539 Residual 1413.786 57 24.803 Table A-20. Regression model predicts helpi ng using two independent variables. Data were coded from interviews c onducted during study period 3 (AugustSeptember 2004). Dep Var: HELP3 N: 60 Multiple R: 0.266 Squared multiple R: 0.071 Adjusted squared multiple R: 0.038 Standard error of estimate: 10.133 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 2.846 2.284 0 1.246 0.218 STORAGE 0.001 0 0.204 0.998 1.597 0.116 ESTIM3 0.101 0.08 0.161 0.998 1.259 0.213 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 446.01 2 223.005 2.172 0.123 Residual 5852.723 57 102.679

PAGE 244

226 Table A-21. Regression model predicts participation in Neighborhood Council meetings using two independent variables. Data were coded from interviews conducted during study period 3 (August-September 2004). Dep Var: JV3 N: 58 Multiple R: 0.118 Squared multiple R: 0.014 Adjusted squared multiple R: 0.000 Standard error of estimate: 0.311 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 0.066 0.072 0 0.91 0.367 STORAGE 0 0 0.111 0.999 0.831 0.409 ESTIM3 0.001 0.002 0.034 0.999 0.255 0.8 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 0.074 2 0.037 0.386 0.682 Residual 5.305 55 0.096 Table A-22. Regression model pr edicts participation in c hurch meetings using two independent variables. Data were coded from intervie ws conducted during study period 3 (August-September 2004). Dep Var: IG3 N: 58 Multiple R: 0.225 Squared multiple R: 0.051 Adjusted squared multiple R: 0.016 Standard error of estimate: 0.966 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 0.727 0.224 0 3.245 0.002 STORAGE 0 0 -0.131 0.999 0.999 0.322 ESTIM3 -0.01 0.008 -0.178 0.999 1.355 0.181 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 2.738 2 1.369 1.467 0.239 Residual 51.33 55 0.933

PAGE 245

227 Table A-23. Regression model predicts food sh aring using two inde pendent variables. Data were coded from interviews conducted during study period 4 (OctoberNovember 2004). Dep Var: FOOD4 N: 62 Multiple R: 0.047 Squared multiple R: 0.002 Adjusted squared multiple R: 0.000 Standard error of estimate: 14.353 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 8.12 3.316 0 2.449 0.017 STORAGE 0 0.001 0.029 0.968 0.217 0.829 ESTIM4 -0.027 0.113 -0.032 0.968 0.243 0.809 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 26.594 2 13.297 0.065 0.938 Residual 12154.373 59 206.006 Table A-24. Regression model pr edicts visiting using two in dependent variables. Data were coded from interviews c onducted during study period 4 (OctoberNovember 2004). Dep Var: VISIT4 N: 62 Multiple R: 0.128 Squared multiple R: 0.016 Adjusted squared multiple R: 0.000 Standard error of estimate: 9.606 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 5.418 2.219 0 2.442 0.018 STORAGE 0 0 0.1 0.968 0.761 0.449 ESTIM4 0.057 0.076 0.099 0.968 0.756 0.453 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 90.025 2 45.012 0.488 0.616 Residual 5444.249 59 92.275

PAGE 246

228 Table A-25. Regression model predicts loans us ing two independent variables. Data were coded from interviews conducted dur ing study period 4 (October-November 2004). Dep Var: LOAN4 N: 62 Multiple R: 0.102 Squared multiple R: 0.010 Adjusted squared multiple R: 0.000 Standard error of estimate: 5.817 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 3.88 1.344 0 2.887 0.005 STORAGE 0 0 -0.062 0.968 0.475 0.637 ESTIM4 0.024 0.046 0.07 0.968 0.529 0.599 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 20.807 2 10.404 0.307 0.737 Residual 1996.612 59 33.841 Table A-26. Regression model predicts helpi ng using two independent variables. Data were coded from interviews c onducted during study period 4 (OctoberNovember 2004). Dep Var: HELP4 N: 62 Multiple R: 0.093 Squared multiple R: 0.009 Adjusted squared multiple R: 0.000 Standard error of estimate: 8.278 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 6.778 1.912 0 3.544 0.001 STORAGE 0 0 0.001 0.968 0.006 0.995 ESTIM4 -0.046 0.065 -0.093 0.968 0.704 0.484 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 35.22 2 17.61 0.257 0.774 Residual 4042.989 59 68.525

PAGE 247

229 Table A-27. Regression model predicts participation in Neighborhood Council meetings using two independent variables. Data were coded from interviews conducted during study period 4 (October-November 2004). Dep Var: JV4 N: 58 Multiple R: 0.177 Squared multiple R: 0.031 Adjusted squared multiple R: 0.000 Standard error of estimate: 0.256 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 0.114 0.06 0 1.879 0.066 STORAGE 0 0 -0.179 0.968 -1.33 0.189 ESTIM4 -0.001 0.002 -0.041 0.968 0.302 0.763 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 0.116 2 0.058 0.887 0.418 Residual 3.608 55 0.066 Table A-28. Regression model pr edicts participation in c hurch meetings using two independent variables. Data were coded from intervie ws conducted during study period 4 (October-November 2004). Dep Var: IG4 N: 58 Multiple R: 0.131 Squared multiple R: 0.017 Adjusted squared multiple R: 0.000 Standard error of estimate: 0.950 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 0.321 0.224 0 1.431 0.158 STORAGE 0 0 0.104 0.968 0.764 0.448 ESTIM4 0.006 0.008 0.101 0.968 0.747 0.458 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 0.873 2 0.437 0.484 0.619 Residual 49.627 55 0.902

PAGE 248

230 Table A-29. Regression model predicts food sh aring using two inde pendent variables. Data were coded from interviews conducted during study period 5 (December 2004-January 2005). Dep Var: FOOD5 N: 59 Multiple R: 0.178 Squared multiple R: 0.032 Adjusted squared multiple R: 0.000 Standard error of estimate: 17.176 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 11.55 3.876 0 2.98 0.004 STORAGE 0.001 0.001 0.16 0.99 1.208 0.232 ESTIM5 -0.069 0.141 -0.065 0.99 0.491 0.625 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 542.182 2 271.091 0.919 0.405 Residual 16521.377 56 295.025 Table A-30. Regression model pr edicts visiting using two in dependent variables. Data were coded from interviews conduc ted during study period 5 (December 2004-January 2005). Dep Var: VISIT5 N: 59 Multiple R: 0.230 Squared multiple R: 0.053 Adjusted squared multiple R: 0.019 Standard error of estimate: 16.383 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 9.256 3.697 0 2.503 0.015 STORAGE 0.001 0.001 0.225 0.99 1.72 0.091 ESTIM5 -0.032 0.134 -0.032 0.99 0.242 0.81 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 840.353 2 420.176 1.565 0.218 Residual 15031.308 56 268.416

PAGE 249

231 Table A-31. Regression model predicts loans us ing two independent variables. Data were coded from interviews conducted during study period 5 (December 2004January 2005). Dep Var: LOAN5 N: 59 Multiple R: 0.049 Squared multiple R: 0.002 Adjusted squared multiple R: 0.000 Standard error of estimate: 7.760 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 4.91 1.751 0 2.803 0.007 STORAGE 0 0 0.024 0.99 0.181 0.857 ESTIM5 -0.019 0.063 -0.04 0.99 0.295 0.769 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 7.954 2 3.977 0.066 0.936 Residual 3372.555 56 60.224 Table A-32. Regression model predicts helpi ng using two independent variables. Data were coded from interviews conduc ted during study period 5 (December 2004-January 2005). Dep Var: HELP5 N: 59 Multiple R: 0.070 Squared multiple R: 0.005 Adjusted squared multiple R: 0.000 Standard error of estimate: 8.604 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 7.052 1.942 0 3.632 0.001 STORAGE 0 0 0.037 0.99 0.274 0.785 ESTIM5 -0.029 0.07 -0.056 0.99 0.419 0.677 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 20.438 2 10.219 0.138 0.871 Residual 4145.223 56 74.022

PAGE 250

232 Table A-33. Regression model predicts participation in Neighborhood Council meetings using two independent variables. Data were coded from interviews conducted during study period 5 (December 2004-January 2005). Dep Var: JV5 N: 57 Multiple R: 0.103 Squared multiple R: 0.011 Adjusted squared multiple R: 0.000 Standard error of estimate: 0.261 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 0.047 0.06 0 0.781 0.438 STORAGE 0 0 0.103 0.989 0.757 0.452 ESTIM5 0 0.002 0.016 0.989 0.121 0.904 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 0.039 2 0.02 0.288 0.751 Residual 3.68 54 0.068 Table A-34. Regression model pr edicts participation in c hurch meetings using two independent variables. Data were coded from intervie ws conducted during study period 5 (December 2004-January 2005). Dep Var: IG5 N: 57 Multiple R: 0.197 Squared multiple R: 0.039 Adjusted squared multiple R: 0.003 Standard error of estimate: 0.994 Effect Coefficient Std Error Std Coef Tolerancet P(2 Tail) CONSTANT 0.821 0.228 0 3.609 0.001 STORAGE 0 0 -0.193 0.989 1.437 0.156 ESTIM5 -0.004 0.008 -0.064 0.989 0.479 0.634 Analysis of Variance Source Sum-ofSquares df MeanSquare F-ratio P Regression 2.149 2 1.074 1.087 0.344 Residual 53.36 54 0.988

PAGE 251

233 LIST OF REFERENCES Abel, T., and J. R. Stepp. 2003 A new ecosystems ecology for an thropology. Conservation Ecology 7(3): 12. Electronic document, http://www.consecol.org/vol7/iss3/art12/ accessed October 20, 2005. Altman, Jeanne 1974 Observational Study of Behavior: Sampling Methods. Behaviors 49 (3): 227-267. American Interna tional Group (AIG) 2003 Bolivia Country profile. Electronic document, http://home.aigonline.com/country_view, accessed October 25, 2005. Ayjan, H. –ztas and Semih Isiksal 2004 Memory Recall Errors in Retrospective Surveys: A Reverse Record Check Study. Quality & Quantity: Intern ational Journa l of Methodology 38(5):475-493. Bernard, H. Russell 2002 Research Methods in Anthropology: Qualitative and Quantitative Approaches. Walnut Creek, CA: Altamira Press. Bernard, H. Russell, P. D. Killworth, L. Sailer, and D. Kronenfeld. 1984 The Problem of Informant Accuracy : The Validity of Retrospective Data. Annual Review of Anthropology 13:495-17. Bernard, H. Russell, Peter D. Killworth, Christopher McCarty, and Gene A. Shelley 1990 Estimating the Size of Personal Networks. Social Networks 23:289–312. Blau, Peter 1967 [1964] Exchange and Power in Social Life. New York: John Wiley & Sons, Inc. Bojanic, Alan 2001 Integrating Biodiversity into the Forestry Sector: Bolivia. In Integration of Biodiversity in National Forestry Planning Program Workshop. Bogor, Indonesia: CIFOR Headquarters.

PAGE 252

234 Booth, Alan 1984 Responses to Scarcity. The Soci ological Quarterly 25 (Winter 1984):113124. Borgatti, Stephen P. 1996 ANTHROPAC 4.0. Natick, MA: Analytic Technologies. Brewer, Devon D. 2002 Supplementary Interviewing Techniques to Maximize Output in Free Listing Tasks. Field Methods 14 (1): 108-118. Buechler, Hans, and Judith-Maria Buechler 1971 The Bolivian Aymara. New York: Holt, Rinehart, and Winston, Inc. Bureau of Inter-American Affairs 1998 Background Notes: Bolivia. Washingt on, DC: US Department of State. March 1998. Bureau for Democracy, Human Rights, and Labor 1999 Annual Report on International Re ligious Freedom for 1999: Bolivia. Washington, DC: U.S. Departme nt of State. September 9, 1999. Carneiro, Robert L. 1970 A Theory of the Origin of the State. Science. 469: 733-738. Casagrande, David, Diane Hope, Elizabeth Farley-Metzger, William Cook, Scott Yabiku, Charles Redman Under review Problem and opportunity : Integrating anthropology, ecology, and policy through adaptive expe rimentation in the urban American Southwest. Submitted to Human Organization, September 13, 2004. Collins, James P., Ann Kinzing, Nancy B. Grimm, William F. Fagan, Diane Hope, Jianguo Wu, and Elizabeth T. Borer. 2000 A New Urban Ecology. American Scientist 88.5 (Sept 2000): 416. Cancian, Frank 1965 Economics and prestige in a Ma yan Community: The religious cargo system in Zinacantan. Stanford: Stanford University Press. Cashdan, Elizabeth 1985 Coping with Risk: Reciprocity Among the Basarwa of Northern Botswana. Man 20(3):454-474. Castells, Manuel 1983 The City and the Grassroots: A Cross-Cultural Theory of Urban Social Movements. Berkeley: Univer sity of California Press.

PAGE 253

235 Central Intelligence Agency (CIA) 2005 The World Factbook: Bolivia. Electronic document, http://www.cia.gov/cia/publications /factbook/geos/bl.html, accessed October 25, 2005. Centro de Investigacin Multidisciplinario 1999 Realidad Numrica de Cochabamba. Cochabamba, Bolivia: Federacin de Entidades Empresariales Privadas de Cochabamba. Claure Periera & Asociados 2001 Paquete de Preinversin: Sistema de Agua Potable, Villa Israel. Cochabamba, Bolivia: Prefectura del Departamento de Cochabamba Direccin Departamental de Infraest ructura, Unidad de Saneamiento Bsico y Vivienda “UNASBVI”, y Programa de Saneamiento Bsico Rural “PROSABAR”. Cleary, Edward L. 1997 Introduction: Pentacostals, Promience, and Politics. In Power, Politics, and Pentacotals in Latin America. Cleary, Edward L. and Hannah W. StewartGambino, Eds. Boulder: Westview Press. Colson, Elizabeth 1979 In Good Years and Bad: Food Strategies of Self-Reliant Societies. Journal of Anthropological Research 35:18-29. Corbett, Jane 1988 Famine and Household Coping St rategies. World Development 16(9):10991112. Cornelius, Wayne 1975 Politics and the Migrant Poor in Mexico City. Stanford: Stanford University Press. Daily, G. 1997 Nature’s Services: Societal Dependence on Natural Ecosystems. Washington, DC: Island Press. Deutscher, I 1973 What we say, What we do. Glenview, IL: Scott Foresman. Diener, Paul, Donald Nonini, and Eugene E. Robkin 1980 Ecology and Evolution in Cultu ral Anthropology. Man 15(1):1-31. Dietz, Henry, and Richard J. Moore 1979 Political Participation in a Non-Electoral Setting: The Urban Poor in Lima, Peru. Athens, Ohio: Ohio Univ ersity Center for International Studies.

PAGE 254

236 Dirks, Robert 1978 Rethinking the Ik: A F unctional Non-Social Syst em. In Extinction and Survival in Human Populations. C. Laughlin, Jr. and I. Brady, eds. New York: Columbia University Press. 1980 Social Responses During Severe Food Shortages and Famine. Current Anthropology 21(1):21-44. Dow, James W. 1997 The Theology of Change: Evangelical Protestantism and the Collapse of Native Religion in a Peasant Area of Mexico. In Explorations in Anthropology and Theology, Salamone, Frank and Walter Adams, eds. (in press) The Expansion of Protes tantism in Mexico: An Anthropological View. Anthropological Quarterly (a ccepted for publication September 20, 2005). Durham, William H. 1979 Scarcity and Survival in Central Am erica: ecological or igins of the Soccer War. Stanford: Stanford University Press. 1995 Political Ecology and Environmenta l Destruction in Latin America. In The Social Causes of Environmental Destru ction in Latin America. M. Painter and W.H. Durham, eds. Ann Arbor: Th e University of Michigan Press. Durkheim, Emile 1982 [1895]. Rules of the Sociological Me thod. New York: The Free Press. Eames, Edwin, and Judith Granich Goode 1973 Urban Poverty in a Cross-Cultural Context. New York: The Free Press. EAWAG (Instituto Federal Suizo en Ciencias del Ambiente y Tecnologa) y SANDEC (Saneamiento para Pases en Desarrollo) Undated SODIS (Desinfeccin del Agua Me diante la Energa Solar): Proceso de tratamiento de agua para el uso domestico. Electronic document, http://www.sodis.ch, accessed January 20, 2006. Eckstein, Susan 1990 Urbanization Revisited: Inner-City Slum of Hope and Squatter Settlement of Despair. World Development 18(2):165-181. Elhance, Arun P. 1999 Hydropolitics in the Third World. Washington, D.C.: United States Institute of Peace Press. Ember, Carol R., and Melvin Ember 1992 Resource Unpredictability, Mistru st, and War: A Cross-Cultural Study. Journal of Conflict Resolution 36(2):242-262.

PAGE 255

237 Energy Information Administration (EIA) 2005 Bolivia Country Analysis Brief. Electronic document, http://www.eia.doe.gov/emeu/cabs/contents.html accessed November 3, 2005. Ennis-McMillan, Michael 2001 Suffering from Water: Social or igins of body distress in a Mexican community. Medical Anthropol ogy Quarterly 15(3):368-390. 2006 A Precious Liquid: Drinking Water and Culture in the Valley of Mexico. Belmont, CA: Thomson Wadsworth. Falkenmark, Marlin, and Jan Lundqvist 1995 Looming Water Crisis: New Approaches Are Inevitable. In Hydropolitics: Conflicts over Water as a Developm ent Constraint. L. Ohlsson, ed. Pp. 178-213. London: Zed Books. Finnegan, William 2002 Leasing the Rain. The New Yorker (April 8, 2002). Food and Agriculture Organizati on of the United Nations (FAO) 2005 Aquastat: Bolivia Database. La nd and Water Development Division. Electronic document, http://www.fao.org/ag/agl/aglw/aqua stat/countries/bol ivia/indexesp.stm accessed December 26, 2005. Foster, George 1965 Peasant Society and the Image of Li mited Good. American Anthropologist 67 (2): 293-315. Friedman, Jonathan 1974 Marxism, Structuralism, and Vu lgar Materialism. Man 9(3):444-469. Gilbert, Alan 1998 The Latin American City. London: The Latin American Bureau. Gill, Lesley 1990 ‘Like a Veil to Cover Them’: Women in the Pentecostal Movement in La Paz, Bolivia. American Ethnologist 17: 708-721. 1994 Precarious Dependencies: Gender, Cla ss, and Domestic Service in Bolivia. New York: Columbia University. 2000 Teetering on the Rim: Global Rest ructuring, Daily Life, and the Armed Retreat of the Bolivian State. New York: Columbia University Press.

PAGE 256

238 Gleick, Peter 1996 Basic Water Requirements for Human Activities: Meeting Basic Needs. Water International 21:83-92. 2004 The World's Water 2004-2005: The Biennial Report on Freshwater Resources: Island Press. Gleick, Peter, Singh, and Shi 2001 2001 Threats to the World’s Freshwat er Resources. Oakland, CA: Pacific Institute for Studies in Development, Environment, and Security. Graves, Nancy B. and Theodore D. 1974 Adaptive Strategies in Urban Migr ation. Annual Review of Anthropology 3:117-151. Grimm, NB; Grove, JM; Pickett, STA 2000 Integrated approaches to long-term studies of urban ecological systems. Bioscience, 50 (7): 571-584. Goldstein, Daniel 2003 The Customs of the Faithful: Evange licals and the Politics of Catholic Fiesta in Bolivia. Journal of La tin American Lore 21(2): 179-200. Gonzalez de la Rocha, Mercedes 1994 The Resources of Poverty: Women and Survival in a Mexican City. Oxford: Blackwell. 2001 From the Resources of Poverty to the Poverty of Resources? The Erosion of a Survival Model. Latin Amer ican Perspectives 28(4):72-100. 2004 From the Marginality of the 1960s to the “New Poverty" of Today: A LARR Research Forum. Latin Ameri can Research Review. 39(1):183203. Gonzalez de la Rocha, Mercedes and Barbara Gantt 1995 The Urban Family and Poverty in Latin America. Latin American Perspectives 22(2):12-31. Gunderson, Lance H and C. S. Holling 2002 Panarchy: Understanding Transformati ons in Human and Natural Systems. Washington, DC: Island Press. p. 3–62. Guttman, L 1950 The basis for scalogram analysis. In Studies in Social Psychology in World War II. A. Stouffer, ed. Prince ton: Princeton University Press.

PAGE 257

239 Halebsky, Sandor 1995 Urban Transformation and Survival Strategies. In Capital, Power, and Inequality in Latin America. S. Hale bsky and R.L. Harris, eds. Boulder: Westview Press. Hallpike, C.R. 1973 Functionalist Interpretations of Primitive Warfare. Man 8(3):451-470. Harris, Marvin 1979 Cultural materialism: the struggle for a science of culture. New York: Random House. Henrich, J., R. Boyd, S. Bowles, C. Came rer, E. Fehr, and H. Gintis, eds. 2004 Foundations of Human Sociality : Economic Experiments and Ethnographic Evidence from Fifteen Small-Scale Societies. Oxford: Oxford University Press. Hill, Kim and Michael Gurven 2004 Economic Experiments to Examine Fairness and Cooperation Among the Ache Indians of Paraguay. In Foundations of Human Sociality: Economic Experiments and Ethnographic Evid ence from Fifteen Small-Scale Societies. Henrich,J., R. Boyd, S. Bowl es, C. Camerer, E. Fehr, and H. Gintis ,Eds. Oxford: Oxford University Press. Homans, George C. 1958 Social Behavior as Exchange. The American Journal of Sociology 63(6):597-606. Homer-Dixon, Thomas 1999 Environment, Scarcity, and Violence. Princeton, NJ: Prin ceton University Press. Ibisch, P.L. 1998 Bolivia is a megadiversity country and a developing country. In Biodiversity a challenge for development research and policy. Barthlott, W. & M. Winiger, Eds. Be rlin: Springer-Verlag, 213-241. Instituto Nacional de Esta dstica de Bolivia (INE) 2002 Nota de Prensa No. 153: Ur banismo. Electronic document, http://www.ine.gov.bo, accessed October 20, 2005. 2003a Nota de Prensa No. 48: Religin. Electronic document, http://www.ine.gov.bo, accessed October 20, 2005. 2003b Bolivia: Caractersticas Sociodemogr ficas de la Poblacin. Electronic document, http://www.ine.gov.bo, accessed October 20, 2005.

PAGE 258

240 2004 Bolivia: Mapa de Pobreza 2001, Censo Nacional de Poblacin y Vivienda 2001. Electronic document, http:// www.ine.gov.bo, accessed October 20, 2005. Isbell, Billie Jean 1978 To Defend Ourselves: Ec ology and ritual in an A ndean village. Prospect Heights, Il: Waveland Press, Inc. Jackson, Jean E. and Kay B. Warren 2005 Indigenous Movements in Latin America, 1992-2004: Controversies, Ironies, New Directions. Annual Re view of Anthropology. 34: 549-573. Johnston, Barbara Rose, and John M. Donahue 1998 Introduction/Conclusion. In Water, Culture, Power: Local Struggles in a Global Context. B.R. Johnston and J.M. Donahue, eds. Washington, D.C.: Island Press. Kirk, Jerome, and Marc C. Miller 1986 Reliability and Validity in Qualitativ e Research. Beverly Hills, CA: SAGE Publications. Kluck, Patricia 1989 The Society and Its Environment. In Bolivia: A Country Study. Rex A. Hudson and Dennis M. Hanratty, Eds. Washington, DC: Federal Research Division. Library of Congress. Kohl, Benjamin 2002 Stabilizing neoliberalism in Bolivia: popular pa rticipation and privatization. Political Geography 21: 449-472. 2003 Restructuring Citizenship in Bolivia : El Plan de Todos. International Journal of Urban and Regional Research 27(2):337-351. Laughlin, Charles, Jr. 1974 Deprivation and Recipr ocity. Man 9(3):380-396. Laughlin, Charles, Jr., and Ivan Brady 1978 Introduction: Diaphasis and Change. In Extinction and Survival in Human Populations. C. Laughlin, Jr. and I. Brady, eds. New York: Columbia University Press. Laurie, N, and S Marvin 1999 Globalization, neoliberalism, and ne gotiated development in the Andes: water projects and regional identity in Cochabamba, Bolivia. Environment and Planning 31:1404-1415.

PAGE 259

241 Ledo Garcia, Maria del Carmen 2002 Urbanization and Poverty in the Citi es of the National Economic Corridor in Bolivia, Case Study: Cochabamba. Delft: DUP Science. 2004 Inequality and Access to water in the cities of Cochabamba and La Paz-El Alto. International Water History As sociation Conference Papers, Theme D. Leeds, Anthony 1971 The Concept of the "Culture of Poverty": Conceptual, Logical, and Empirical Problems, with Persp ectives from Brazil and Peru. In The Culture of Poverty. E. Leacock, ed. New York: Simon and Schuster. Lett, James 1996 Emic/Etic Distinctions. In Encyclopedia of Cultural Anthropology. David Levinson and Melvin Ember, eds. New York: Henry Hold Company. Library of Congress 2005 Country Profile: Bolivia. Washington DC: Library of Congress Federal Research Division. Little, Paul 1999 Environments and Environmentalis ms in Anthropological Research: Facing a New Millennium. Annual Review of Anthropology. Vol. 28: 253-284. Lloyd, Peter 1979 Slums of Hope? Shanty Towns of the Third World. New York: St. Martin's Press New York. Lobo, Susan 1982 A house of my own: Soci al Organization in the Squatter Settlements of Lima, Peru. Tucson: The Univ ersity of Arizona Press. Lomnitz, Larissa 1977 Networks and Marginality: Life in a Mexican Shantytown. New York: Academic Press. 1978 Survival and Reciprocity: The Case of Urban Marginality in Mexico. In Extinction and Survival in Human P opulations. C. Laughlin, Jr. and I. Brady, eds. New York: Columbia University Press. Los Tiempos 2004 Inauguraron proyecto para dotar de a gua a 120 familias de la zona Barrios Unidos. Cochabamba (December 20, 2004). Lutgens, Frederick and Edward Tarbuck 1998 Essentials of Geology. New Jersey: Prentice-Hall.

PAGE 260

242 Mangin, William 1967 Latin American Squatter Settlements: A Problem and a Solution. Latin American Research Review 2(3):65-98. Malinowski, Bronislaw 1970 [1922] Argonauts of the Wester n Pacific. London: Routledge. Mariz, Cecelia Loreto 1994 Coping with Poverty: Pentecostals and Christian Base Communities in Brazil. Philadelphia: Temple University Press. Marvin, Simon, and Nina Laurie 1999 An Emerging Logic of Urban Water Management, Cochabamba, Bolivia. Urban Studies 36(2). Mattos, Roger R. and Alberto Crespo 2000 Informe Nacional Sobre la Gesti n del Agua en Bolivia. Electronic document, http://www.aguabolivia.com accessed December 20, 2005. Mauss, Marcel 2000 [1923] The Gift. The Form and Reason for Exchange in Archaic Societies. London: Routledge. McCarty, Chris, and Amber Wutich 2005 Conceptual and empirical arguments for including or excluding ego from structural analysis of personal netw orks. Connections, the Journal of the Social Network Association 26(2): 80-86. McFarren, Wendy 1992 The Politics of Bolivia's Economic Cris is: Survival Strategies of Displaced Tin-Mining Households. In Unequal Burden: Economic Crises, Persistent Poverty, and Women's Work. S. Feldman, ed. Pp. 131-158. Boulder: Westview Press. Medeiros, Carmen 1990 Civilizing the Popular? The Law of Popular Participation and the Design of a New Civil Society in 1990s Bolivia. Critique of Anthropology 21(4): 402-425. Messer, Ellen 1988 Seasonal Hunger and Coping Strategi es: An Anthropological Discussion. In Coping with Seasonal Constraints. M.R.P.i.S.a. Archaeology, ed, Vol. 5. Philadelphia: The Mu seum Applied Science Center of Archaeology. Moseley, Michael E. 1992 The Inca and Their Ancestors. NY: Thames and Hudson Publishing.

PAGE 261

243 Moser, Caroline O. 1996 Confronting Crisis: A Comparativ e Study of Household Responses to Poverty and Vulnerability in Four Poor Urban Communities. Volume No. 8. Washington, D.C.: The World Bank. Nash, June 1992 Interpreting Social Movements: Bolivian Resistance to Economic Conditions Imposed by the Internat ional Monetary Fund. American Ethnologist. 19(2):275-293. National Research Council / Committee on Alluvial Fan Flooding, Water Science and Technology Board, Commission on Geoscience s, Environment, and Resources 1996 Alluvial fan flooding. Wash ington, DC: National Academy Press. Orlove, Benjamin 1980 Ecological Anthropology. Annual Review of Anthropology 9:235-273. Paredes C., Carlos and Carlos Aguirre B. 2000 Report on the Situation of Sustai nable Development in Bolivia: Assessment of the Industrial Manuf acturing Sector. United Nations Industrial Development Organizatio n (UNIDO). Electronic document, http://www.unido.org/userfiles/ti mminsk/RIO10-IND-bolivia-eng.pdf, accessed December 20, 2005. Pavao-Zuckerman, Mitchell A. 2000 The Conceptual Unity of Mode ls in Human Ecology. Journal of Ecological Anthropology 4: 31. Peattie, Lisa Redfield 1970 The View from the Barrio. Ann Arbor: Ann Arbor Paperback and University of Michigan Press. Perlman, Janice 1976 The Myth of Marginality: Urban P overty and Politics in Rio de Janeiro. Berkeley: University of California Press. Pickett, S. T. A., M. L. Cadenasso, J. M. Grove, C. H. Nilon, R. V. Pouyat, W. C. Zipperer, and R. Costanza 2001 Urban ecological systems: linking terrestrial ecological, physical, and socioeconomic components of metr opolitan areas. Annual Review of Ecology and Systematics 32: 127–157. Polanyi, Karl 1968 Primitive, Archaic and Modern Economies: Essays of Karl Polanyi. United States: Beacon Paperback.

PAGE 262

244 Portes, Alejandro, and John Walton 1976 Urban Latin America: The Politic al Condition for Above and Below. Austin: University of Texas Press. Rappaport, Roy A. 1974 Pigs for the Ancestors: Ritual in the Ecology of a New Guinea People. New Haven: Yale University Press. Redman, Charles L. 1999 Human Dimensions of Ecosyste m Studies, Ecosystems, 2: 296-298. 2005 Resilience Theory in Archaeology. American Anthropologist, 107(1): 70– 77. Redman, CL, JM Grove, and LH Kuby 2004 Integrating social science into th e long-term ecologica l research (LTER) network: Social dimensions of ecological change and ecological dimensions of social change. Ecosystems, 7 (2): 161-171. Romney, A. Kimball, Susan C. We ller, and William H. Batchelder 1986 Culture as consensus: A theory of cultural and in formant accuracy. American Anthropologist, 88:313-338. Safa, Helen 2004 From the Marginality of the 1960s to the “New Poverty" of Today: A LARR Research Forum. Latin Ameri can Research Review. 39(1):183203. 1974 The Urban Poor of Puerto Rico: A Study in Development and Inequality. New York: Holt, Rinehart and Winston, Inc. Sahlins, Marshall 1978 Stone Age Economics. London: Tavistock. Satterthwaite, David 1998 Environmental Problems in the South. In Environmental Strategies for Sustainable Development in Urban Areas: Lessons from Africa and Latin America. E. Fernandes, ed. Brookfie ld, VT: Ashgate Publishing Company. Scheper-Hughes, Nancy 1992 Death Without Weeping: The Viol ence of Everyday Life in Brazil. Berkeley: University of California Press. Schmink, Marianne 1982 Land Conflicts in Amazonia. American Ethnologist 9(2):341-357.

PAGE 263

245 Schultz, Jim 2005 The Democracy Center Online: Coverage of the Cochabamba Water War and Gas Wars. Electronic documen t, http://www.democracyctr.org, accessed March 10, 2005. Scoones. I. 1999 New Ecology and the Social Sciences: What Prospects for a Fruitful Engagement? Annual Review of Anthropology 28:479-507. SEMAPA 2006 SEMAPA: Servicio Mu nicipal de Agua Potable y Alcantarillado de Cochabamba, Bolivia. http://www.semapa.com.bo accessed February 3, 2006. Sen, Amartya 1999 Poverty and Famines. In The Amartya Sen and Jean Drze Omnibus, comprising Poverty and Famines, H unger and Public Action, and India: Economic Development and Social Opportunity. Amartya Sen and Jean Drze, eds. Oxford: Oxford University Press. SENAMHI 2006 SENAMHI: Servicio Nacional de Mete orologa e Hidrologa de Bolivia. Boletines Oficiales Mensuales. http://www.senamhi.gov.bo, accessed February 6, 2006. Shakow, Miriam 2005 The Political Life of Envy: Inequal ity and Decentraliza tion in Neoliberal Bolivia. American Ethnological So ciety Annual Meeting, April 10, 2005. Stack, Carol 1997 [1974] All Our Kin. New York: BasicBooks. Stepp, J. R., E. C. Jones, M. Pavao-Zuckerman, D. Casagrande, and R. K. Zarger 2003 Remarkable properties of human ecosystems. Conservation Ecology 7(3): 11. Electronic document, http:// www.consecol.org/vol7/iss3/art11/, accessed October 1, 2005. Steward, Julian H. 1959 The Concept and Method of Cultural Ecology. In Readings in Anthropology, Vol. 2. New York, Crowell. Stimson, Jesse, Shawn Frape, Robert Drimmie, David Rudolph 2001 Isotopic and geochemical evidence of regional-scale anisotropy and interconnectivity of an alluvial fa n system, Cochabamba Valley, Bolivia. Applied Geochemistry 16:1097-1114.

PAGE 264

246 Stimson, Jesse, David Rudolph, Shawn Frape, Robert Drimmie 1996 Interpretation of groundwater fl ow patterns through a reconstr uction of the tritium precipitation record in the Co chabamba Valley, Bolivia. Journal of Hydrology 180: 155-172. Stoll, David 1993 Introduction: Rethinking Protestantis m in Latin America. In Rethinking Protestantism in Latin America. Ga rrard-Burnett, Virginia and David Stoll, Eds. Philadelphia: Temple University Press. Stonich, Susan 1993 I am Destroying the Land: The Political Ecology of Poverty and Environmental Destruction in Honduras. Westview Press. 1989 The Dynamics of Social Proce sses and Environmental Destruction: A Central American Case Study. Populat ion and Development Review 15: 269-96. Sugita, Eri Woods 2004 Domestic Water Use, hygiene behavior and childrenÂ’s diarrhea in rural Uganda. University of Florida: Online Dissertations. Swaney, Deanna 2001 Bolivia. Melbourne, Australia: Lonely Planet. Swyngedouw, Eric 1997 Power, Nature, and the City. The co nquest of water and the political ecology of urbanization in Guaya quil, Ecuador: 1880-1990. Environment and Planning A 29:311-332. 2004 Social Power and the Urbanization of Water: Flows of Power. New York: Oxford University Press. Terhorst, Philipp 2003 Public-Popular Organizations: The Case of Cochabamba, Bolivia. Master of Science Thesis. Leicesters hire, UK: Loughborough University. Trujillo, Robert 2002 Unpublished data. Cochabamba: Unin Cristiana Evanglica. Turnbull, Colin 1972 The Mountain People. New York: Simon & Schuster, Inc. 1978 Rethinking the Ik: A Functional Non-Social System. In Extinction and Survival in Human Populations. C. Laughlin, Jr. and I. Brady, eds. New York: Columbia University Press.

PAGE 265

247 UN-Habitat 2003 Report on Water and Sanitation in the WorldÂ’s Cities. UN-Habitat. United Nations Undated United Nations Millennium De velopment Goals. Electronic document, http://www.un.org/millenniumgoals/, accessed March 1, 2006. United Nations Development Program (UNDP) 2002 Bolivia: Analysis of Environmenta l Problems in the Country. Electronic document, http://www.pnud.bo/ppd/PPDEV/Strategy/Analisys.htm accessed December 19, 2005. Van Cott, Donna Lee 2000 Party System Development and Indige nous Populations in Latin America: The Bolivian Case. Party Politics. 6(2): 155-174. Van Damme, Paul 2002 Disponibilidad, Uso y Calidad de los Recursos Hdricos en Bolivia. La Paz: Comisin para la Gestin Integr al del Agua en Bolivia (CGIAB). van der Leeuw, S and CL Redman 2002 Placing archaeology at the center of socio-natural studies. American Antiquity, 67 (4): 597-605. Vargas, Luis Alberto 2001 Thirst and Drinking as a Biocultural Process. In Drinking: Anthropological Approaches. I. de Ga rine and V. de Garine, eds. New York: Berghahn Books. Vayda, Andrew P. 1969 The Study of the Causes of War, w ith Special Reference to Head-Hunting Raids in Borneo. Ethnohistory 3:211-224. Vayda, Andrew P., and Bonnie McCay 1975 New Directions in Ecology and Ecological Anthropology. Annual Review of Anthropology 4:293-306. Vera, Raul R. Undated Bolivia: Country Resource Profiles (Pasture/Forage). Electronic document, http://www.fao.org. Wagner, Maria Luise 1989 Historical Setting. In Bolivia: A Country Study. Rex A. Hudson and Dennis M. Hanratty, Eds. Washingt on, DC: Federal Research Division. Library of Congress.

PAGE 266

248 Wali, A., G. Darlow, C. Fialkowski, M. Tudor, H. del Campo, and D. Stotz. 2003 New methodologies for interdisciplin ary research and action in an urban ecosystem in Chicago. Conserva tion Ecology 7(3): 2. Electronic document, http://www.consecol.org/vol7/iss3/art2/ accessed October 7, 2005. Walker, Peter 1989 Famine Early Warning Systems. Lo ndon: Earthscan Publications Ltd. Weeks, John 1995 The Contemporary Latin American Economies: Neoliberal Reconstruction. In Capital, Power, and Inequality in Latin America. S. Halebsky and R.L. Harris, eds. Boulder: Westview Press. Weismantel, Mary 2001 Cholas and Pishtacos: Stories of R ace and Sex in the Andes. Chicago: The University of Chicago Press. Weissner, Polly 1982 Risk, reciprocity, and social influences on !Kung San economics. In Politics and history in band societie s. E. Leacock and R. Lee, eds. Cambridge: Cambridge University Press. Wolf, Eric R. 1969 Peasant Wars of the Twentieth Century. New York: Harper Torchbooks. Whitehead, Laurence 2001 Bolivia and the Viability of Demo cracy. Journal of Democracy 12(2): 616. World Bank 2005 Bolivia at a Glance. Electronic document, http://www.worldbank.org accessed November 10, 2005.

PAGE 267

249 BIOGRAPHICAL SKETCH Amber Yoder Wutich was born in Miami, Florida. She received her bachelorÂ’s degree in anthropology, with highest honors, an d Chinese language and literature from the University of Florida in 2000. Between 1996 and 2006, Amber Wutich conducted field work in China, Mexico, and Bolivia. She has also lived in Para guay and Argentina.