INFORMATION TO USERS

This manuscript has been reproduced from the microfilm master. UMI films the text directly from the original or copy submitted. Thus, some thesis and dissertation copies are in typewriter face, while others may be from any type of computer printer.

The quality of this reproduction is dependent upon the quaiity of the copy submitted. Broken or indistinct print, colored or poor quality illustrations and photographs, print bleedthrough, substandard margins, and improper alignment can adversely affect reproduction.

In the unlikely event that the author did not send UMI a complete manuscript and there are missing pages, these will be noted. Also, if unauthorized copyright material had to be removed, a note will indicate the deletion.

Oversize materials (e.g., maps, drawings, charts) are reproduced by sectioning the original, beginning at the upper left-hand comer and continuing from left to right in equal sections with small overlaps.

Photographs included in the original manuscript have been reproduced xerographically in this copy. Higher quality 6” x 9* black and white photographic prints are available for any photographs or illustrations appearing in thte copy for an additional charge. Contact UMI directly to order.

ProQuest Information and Learning 300 North Zeeb Road, Ann Arbor, Ml 48106-1346 USA 800-521-0600 UMÏ

SURVIVAL TO SURPLUS: VARIATION IN LIVELIHOOD STRATEGIES AMONG WOMEN IN PORTO NOVO, BENIN

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

By Jennifer L. Mandel, M.P.A, M.A.

***

The Ohio State University 2001

Dissertation Committee Apjvoved by Professor Lawrence A. Brown, Advisor Professor Paul A. Robbins Advisor Professor Eugene McCann Department of Geography UMI Number 3022535

UMI’

UMI Microform 3022535 Copyright 2001 by Bell & Howell Information and Learning Company. All rights reserved. This microform edition Is protected against unauthorized copying under Title 17, United States Code.

Bell & Howell Information and teaming Company 300 North Zeeb Road P.O. Box 1346 Ann Arbor, Ml 48106-1346 Strategy is a technical term meaning the positioning of oneself to others in order to accomplish one’s goals. Whereas one’s reasons for action may be subjective and strictly personal, one’s strategy is shaped by social circumstances...the strategy becomes social insofar as individuals recognize the actual or possible influence of others, their values and actions, upon their own goals. (Hamilton 1985) ABSTRACT

Economic development research and practice has primarily emphasized access to financial capital and the acquisition of westem-style business skills to move people out of the informal economy, in a developing and developed country contexts, or off welfare, in developed country settings. This orientation reflects misunderstanding of how people actually earn a living and the ways in which their life circumstances shape their opportunities to do so. Although access to resources (including financial capital) is important, research on women’s economic strategies suggests that spatial mobility may play an equally critical role. Given that resources are not distributed in geographically uniform patterns, access involves not only availability and eligibility, but also ability to get to where those resources are. This study argues that geographically localized social institutions are (re)produced through people’s daily social and economic practices. The results are differential opportunities for income generation. Therefore, variation in the livelihood strategies people assemble fiom different economic activities reflects their life circumstances. The object of this research is to identify and explain basic patterns of economic strategy usage and outcome in one context, by examining social processes that give rise to different configurations. Employing primary data including 522 in-depth surveys and 15 life/economic histories of women living in Porto Novo, Benin, this study identifies the kinds of strategies women use, the enabling conditions that give rise to particular strategy configurations and examines the ways in which life-circumstances shape women’s economic opportunities. Findings indicate that women use a wide array of different strategies including mono-activity forms involving only one economic activity and

ui diverse approaches comprising multiple income sources. Specific strategy configurations emerge firom two enabling conditions - Assets and Mobility. Women's opportunities in livelihood strategy configuration, and hence, their ability to earn a living, reflect localized social institutions {Structures o f Constraint) within their households and communities. While Assets are important. Mobility is even more so. Assets impact type and scale of enterprises, but Mobiiity enables women to overcome limited resources under some conditions, and facilitates their use under others.

IV DEDICATION

In loving memory of my grandmothers, Ruth Press Jackson and Molly Mandel and my aunt, Elizabeth Velma Hacket Who were all great teachers and provided me a legacy of strong, independent, inspirational women. ACKNOWLEDGMENTS

Although writing a dissertation is often a solitary affair, this work could only have been completed with support and input from many people. Listed below are just a few of the major contributors to this endeavor. First and foremost I must thank the wonderful OSU faculty with whom I had the privilege to work. Chief among them is Larry Brown, my advisor, who had the patience to let me find my own project and do it in my own way. I am also grateful to him for showing me how to do research, providing opportunities to pursue my interests and trying to teach me two fundamentals to our profession; writing concisely, and "rule number one: everything takes longer than you think its going to." Nancy Ettlinger, also had an enormous impact on my professional development. Her guidance, insight, honesty, and friendship, not only inspired me to work to the highest possible standards, but also gave me confidence in myself and my research. Finishing this project and Ph.D. is in no small measure a function of her support. Paul Robbins stepped onto my committee upon joining the Geography Department. Although I'm sure he had no idea what he was in for, he never turned me away from his office door and provided invaluable feedback and advice. He also taught me to focus on the BIG questions. I am deeply grateful for his support, fiiendship and inspiring me to do "good science". Several other people assisted at critical junctures in this project. Without patience and insight from Steve Holloway this research would never have gotten off the ground. He helped me find the researchable question in my general interest in Afiican women's work. Claire Robertson taught me about Afiican women traders and the importance of an historical perspective. She also provided invaluable guidance on data collection in an

VI Afiican context. Jennifer Olsen also shared her vast Afiican fieldwork experience and helped me through the rough spots when I was in Benin. Alan Murray, provided the statistical means for doing what I wanted to with my data when everyone else said it was impossible. Cathy Rakowski was instrumental in providing opportunities to promote my research within Women's Studies. Last, but certainly not least, I must thank Eugene McCann for jumping onto my committee near the end and providing thoughtful feedback. I owe a great debt to many people in Benin as well. Foremost among them are the women of Porto Novo who shared their lives with me by answering so many silly "jovo"questions. Dr. John Igue provided office space at Laboratoire d’Analyse Régionale et d’Expertise Sociale (LARES) and he. Dr. Bio G. Soule, and Yvette Onibon contributed essential advice. 1 am also extremely grateful to my research assistants, Jacques Bangbola, Marguerite Dossou, and Leocadie A1 Hasa, for their fiiendship, enthusiasm and tireless efforts on my behalf. I wish to thank the organizations that funded this research. These include: The National Science Foundation for a Dissertation Enhancement Grant (No. SBR-9711900); the U.S. Department of Education and OSU African Studies Program for a Foreign Language Area Studies Fellowship; and OSU for the Women’s Studies Elizabeth E. Gee Grant, and Graduate School Alumni Research Award. I am deeply indebted to Simon Akindes and his family in Porto Novo, for "adopting" me, especially, Meme, Regine and Clotilde. And to Damien who made the work possible by managing the mundane details of daily life in Benin. Numerous firiends in Porto Novo taught me about Béninois culture and made me feel at home particularly, Leon and Christian Midognin, Madame Pauline Pedro and her daughter Raymonde, Sofia Barabota, and the Ahounou family, Wasi, Wasia, Kamal, Kaseem, Kadija, and Madupe. I am similarly blessed by wonderful fiiends in the U.S. My cheerleaders and stats gurus, Sarah Zapolslqf and Emily Cole-Bayer made the work lighter and easier. Many insightful late night conversations about Africa with Tom Bayer helped me keep the work real. Jim Fields, never lost confidence and provided whatever support was most needed. Janet and Andy Ingraham Dwyer made it possible to go to the field without guilt by

vu taking care of Gwen. Although I am grateful to all my fiiends and colleagues at OSU for their comradery, I owe a special debt to Fernando Bosco, and Tara Maddock, for their fiiendship and always being available to provide feedback. Most importantly, I am grateful to have an incredible family. My aunt, Marian Bressel, provided practical and emotional support, especially while I was in the field. She also did a tremendous job editing my dissertation. My parents, Jill Jackson and David Mandel never lost faith, even when all my (younger) siblings were actually earning a living. They were my first geography teachers, showing me the world and teaching me about cultures other than my own. I am especially appreciative of both my parents, experienced development professionals working in the field, for encouraging me to think about my research in practical development terms and my father who read every word of my dissertation.

vui VITA

1965 ...... Bom, Washington, DC.

1989 ...... B.A. English, Indiana University, Bloomington, IN.

1992 ...... M.P.A., Indiana University, School of Public and Environmental Affairs, Bloomington, IN.

1994 ...... M.A. Geography, The Ohio State University, Columbus, OH.

1997-2001 ...... Ph.D. Candidate Geography, The Ohio State University, Columbus, OH.

PUBLICATIONS

1998. Brown, Lawrence A., Jennifer L. Mandel and Victoria A. Lawson. "Occupational Composition and Economic Adjustment: Ecuador 1982-1990." International Regional Science Review, vol. 20, pp. 183-209.

1995. Brown, Lawrence A., Jennifer L. Mandel and Victoria A. Lawson. "The Uprooting of People, Migration and Labor Force Experiences: Ecuador 1982-1990." Journal for Development Policy, vol. 3, pp. 331-348.

FIELDS OF STUDY

Major Field: Geography

IX TABLE OF CONTENTS

Page

Abstract...... iii

Dedication ...... v

Acknowledgments ...... vi

Vita...... ix

List of Tables ...... xv

List of Figures ...... xvii

Chapters:

1. INTRODUCTION...... 1 1.1 Introduction ...... 1 1.2 Problem Statement ...... 3 1.3 Theoretical Considerations ...... 5 1.4 Regional Context and Themes of the Research ...... 6 1.5 Organization of the Dissertation ...... 9

2. EMPIRICAL FRAMEWORK: DO STRATEGIES VARY? ...... 12 2.1 Introduction ...... 12 2.2 Multiple Livelihood Economic Strategies in General ...... 13 2.2.1 Introduction ...... 13 2.2.2 Use of Multiple Livelihood Economic Strategies Among Households ...... 13 22.3 Use of Multiple Livelihood Economic Strategies Among Individuals...... 16 2.3 African Women and Multiple Livelihood Economic Strategies ...... 18 2.3.1 Introduction ...... 18 2.3.2 The Pre-Colonial Period ...... 20 2.3.3 The Colonial Period ...... 23 2.3.4 The Early Post-Colonial Period ...... 27 2.3.5 The Era of Structural Adjustment ...... 30 2.3.6 Conclusion ...... 32 2.4 Differences Among Afiican Women ...... 33 2.5 Conclusions ...... 35

3. CONCEPTUAL FRAMEWORK: HOW CAN STRATEGY VARIATION BE EXPLAINED?...... 36 3.1 Introduction ...... 36 3.2 Dimensions of Difference ...... 39 3.3 Structures of Constraint ...... 40 3.4 Localizing Structures of Constraint ...... 42 3.5 Localized Structures of Constraint and Labor Force Experience; An Example ...... 45 3.6 Gender and Development; Applying Localized Social Institutions ...... 46 3.7 Conclusions ...... 47

4. RESEARCH CONTEXT AND DATA COLLECTION; HOW CAN THE FRAMEWORK BE APPLIED TO BENIN?...... 49 4.1 Introduction ...... 49 4.2 The Political-Economy of Benin; An Historical Overview ...... 49 4.2.1 The Pre-Colonial Period ...... 51 4.2.2 The Colonial Period ...... 52 4.2.3 Early Post-Colonial Period 1960-1988 ...... 54 4.2.4 The Era of Structural Adjustment 1988-Present ...... 55 4.3 Data Collection Procedures ...... 58 4.3.1 The In-Depth Survey ...... 60 4.3.1.1 Assets...... 60 4.3.1.2 Formal Rules ...... 61 4.3.1.3 Cultural Norms ...... 61 4.3.1.4 Personal Preferences ...... 63 4.3.2 Life and Economic History Interviews ...... 64 4.4 Sample Representativeness ...... 64 4.5 The Geography of Social Groups ...... 67 4.6 Analysis Procedures ...... 71 4.7 Conclusion ...... 73

XI 5. CLUSTER ANALYSIS: WHAT STRATEGIES DO WOMEN IN PORT NOVO USE? ...... 74 5.1 Introduction ...... 74 5.2 Content Analysis: Developing an Economic Activity Typology ...... 75 5.2.1 Economic Activity Categorization ...... 76 5.2.1.1 Occupational Category ...... 76 5.2.1.2 Employment Classification ...... 77 5.2.1.3 Type of Good ...... 79 5.2.2 A Typology of Economic Activity ...... 81 5.3 Identifying Number of Activities ...... 84 5.4 Identifying Livelihood Strategies ...... 85 54.1. Content Analysis: Rationale for Combining and Prioritizing Activities ...... 85 5.4.1.1 Rationale for Combining Activities ...... 85 5.4.1.2 Rationale for Prioritizing Activities within Strategies . .. 87 5.4.2 Cluster Analysis: Identifying Livelihood Strategies ...... 88 5.4.2.1 Matrices of Similarity Measures ...... 89 5.4.2.2 Clustering Cases ...... 91 5.4.2.3 Results: Thirteen Strategy Types ...... 92

6. CHI-SQUARE ANALYSIS: WHAT ENABLING CONDITIONS SHAPE STRATEGY CHOICE?...... 96 6.1 Introduction ...... 96 6.2 Data ...... 97 6.2.1 Location/Mobility ...... 97 6.3 Identifying Enabling Conditions: CHAID Analysis ...... 100 6.3.1 CHAID Analysis Description ...... 100 6.3.2 CHAID Analysis: Four Models ...... 102 6.3.2.1 Assets Models ...... 102 6.3.2.2 Cultural Norms M odels ...... 105 6.3.2.3 Location/Mobility M odels ...... 105 6.4 Conclusion ...... I ll

7. WHAT IS THE RELATIONSHIP BETWEEN ENABLING CONDITIONS AND STRATEGY ACCUMULATION POTENTIAL?...... 113 7.1 Introduction ...... 113 7.2 Data and Analysis ...... 116 7.3 Enabling Conditions and Accumulation Potential ...... 117 7.4 The Meaning of Mobility for Women in Porto Novo ...... 121 7.5 Trade-Offs: Mobility vs. MLES ...... 123

8. LIMITED MOBILITY: HOW DO DIMENSIONS OF DIFFERENCE PRODUCE LOCALIZED ENABLING CONDITIONS?...... 126 8.1 Introduction ...... 126

XU 8.2 Data and Analysis ...... 127 8.2.1 D ata ...... 128 8.2.2 Analysis ...... 129 8.3 Explaining Limited Mobility ...... 130 8.3.1 Introduction: Identifying Themes ...... 130 8.3.2 Life-cycle Effects Limiting Mobility: Lack of Autonomy for Young Women ...... 134 8.3.3 Life-cycle Effects Limiting Mobility: Loss of Family Assistance for Middle-aged Women ...... 137 8.3.4 Life-cycle Effects on Limiting Mobility: Isolation among Older Women ...... 139 8.3.4.1 Household Security Issues ...... 140 8.3.4.2 Frailty due to Age ...... 141 8.3.5 Conclusions to Life-cycle Effects Limiting Mobility ...... 142 8.3.6 Class effects Limiting Mobility ...... 142 8.4 Asset Differentials among Women with Limited Mobility ...... 143 8.4.1 Introduction: Identifying Themes ...... 143 8.4.2 Class Effects on Asset Distributions: The Role of Family Background ...... 145 8.5.3. Class Effects on Asset Distributions: The Impact of Formal Rules ...... 146 8.4.4 Class and Life-cycle Effects on Asset Distributions: Husband’s Interference ...... 150 8.4.5 Conclusions to Class effects Limiting Assets ...... 153 8.6 Conclusions ...... 153

9. SUBSTANTIAL MOBILITY: HOW DO DIMENSIONS OF DIFFERENCE PRODUCED LOCALIZED ENABLING CONDITIONS? ...... 155 9.1 Introduction ...... 155 9.2 Explaining Substantial Mobility ...... 156 9.2.1 Introduction: Identifying Themes ...... 156 9.2.2 Ethnicity and Class Effects: Economic Independence among Yoruba Women ...... 163 9.2.3 Life-cycle Phase and Class Effects: Economic Independence among Goun Women ...... 169 9.2.4 Conclusions ...... 172 9.3 Class and Ethnicity Effects: Differential Assets among Women with Substantial Mobility ...... 172 9.3.1 Introduction: Identifying Themes ...... 172 9.3.2 Ethnicity Effects: Differences among Women with Limited Assets...... 173 9.3.3 "Purposeful Choice": Ignoring Social Expectation ...... 175 9.3.4 Class Effects: Differences among Women with Substantial Mobility ...... 178

x iii 9.3.5 Conclusions ...... 183 9.4 Conclusions ...... 184

10. CONCLUSION: EXPLAINING LIVELIHOOD STRATEGY VARIATION . . . 186 10.1 Introduction ...... 186 10.2 Explaining Livelihood Variation: How do people earn a living? ...... 186 10.2.1 Identifying Strategies and Enabling Conditions ...... 187 10.2.2 Strategy Accumulation Potential Relative to Enabling Conditions ...... 188 10.2.3 Explaining Strategy Configuration Opportunities ...... 189 10.3 Methodological and Theoretical Contributions ...... 191 10.4 Development Implications ...... 192 10.5 Future Research...... 194

SOURCES...... 195

APPENDIX A: FRENCH VERSION OF SURVEY...... 210

APPENDIX B: ENGLISH VERSION OF SURVEY...... 225

APPENDIX C: DATA DICTIONARY...... 240

APPENDDC D: S-PLUS CODE FOR MATRIX CREATION...... 321

XIV LIST OF TABLES

Table Page

4.1 Age Cohorts of Women in Benin, Oueme, and the Sample ...... 65

4.2 Distribution of Ethnic Groups in Benin, Oueme, and the Sample ...... 66

4.3 Level of Education among Women over 15 in Benin, Oueme, Sample...... 66

4.4 Residential Arrangement by Ethnic Group ...... 68

4.5 Residential Arrangement by Area of the City ...... 68

4.6 Ethnicity by Area of the City...... 69

5.1 Typology of Economic Activity ...... 83

5.2 Number of Activities by Percentage of Women ...... 84

5.3 Order of Importance among Economic Activities by Respondents ...... 88

5.4 Summary of Livelihood Strategy Configuration Types ...... 94

6.1 Analysis Variable Summary (Part I)...... 98

6.1 Analysis Variable Summary (Part H)...... 99

6.2 Characteristics of Substantial Asset Strategies ...... 103

6.3 Characteristics of Limited Asset Strategies ...... 106

6.4 Characteristics of Substantial Mobility Strategies ...... 107

6.5 Characteristics of Limited Mobility Strategies ...... 108

XV 6.6 Strategy Groupings based on Enabling Conditions ...... 112

7.1 Strategy Groupings based on Enabling Condition Pairings ...... 114

7.2 Interviewees and Strategy Type ...... 117

7.3 Financial Capital Growth Relative to Enabling Conditions ...... 117

7.4 Time Allocated to each Activity by Percentage of Women ...... 119

7.5 Mean and Median Number of Economic Activities by Enabling Condition Group ...... 124

7.6 Women’s Responses Regarding MLES ...... 124

8.1 Percentage of Women with Limited Mobility by Ethnicity ...... 131

8.2 Percentage of Women with Limited Mobility by Age Range and Residential Arrangement ...... 132

8.3 Percentage of Women with Limited Mobility by Age Range and Residential Location ...... 132

8.4 Household Characteristics of Women with Limited Mobility...... 133

8.5 Asset Related Characteristics of Women with Limited Mobility ...... 144

9.1 Percentage of Women with Substantial Mobility by Ethnicity ...... 159

9.2 Residential Arrangement of Women with Substantial Mobility...... 160

9.3 Residential Location of Women with Substantial Mobility ...... 160

9.4 Household Characteristics of Women with Substantial Mobility...... 162

9.5 Asset Related Characteristics of Women with Substantial Mobility...... 174

XVI LIST OF FIGURES

Figure Page

1.1 Conceptual and Empirical Framework for Dissertation ...... 8

1.2 Organization of the Dissertation ...... 10

3.1 Conceptual Framework ...... 38

4.1 Study Area: Republic of Benin ...... 50

4.2 Detail of Study Site: The City of Porto Novo ...... 59

4.3 Research Steps...... 72

5.1 Percentage of Women by Activity Two Occupation Categories ...... 77

5.2 Responses on Activity Importance by Percentage of Women ...... 80

5.3 Rationale for Combining Activities by Percentage of Women ...... 86

5.4 Cluster Solution Efficiency Scores ...... 93

9.1 Reasons for using MLES by Percentage of Women with Substantial Mobility ...... 157

xvu CHAPTER 1

INTRODUCTION

1.1 Introduction Economic development research and practice has primarily emphasized access to financial capital and the acquisition of westem-style business skills to move people out of the informal economy, in a developing and developed country contexts, or off welfare, in developed country settings. This orientation reflects misunderstanding of how people actually earn a living and the ways in which their life circumstances shape their opportunities to do so. Although access to resources (including financial capital) is important, research on women’s economic strategies suggests that spatial mobility may play an equally critical role (Pratt and Hanson 1991; England 1993a; England 1993b; Clark 1994; Hanson and Pratt 1995; Robertson 1997; Schroeder 2000). Given that resources are not distributed in geographically uniform patterns, access involves not only availability and eligibility, but also ability to get to where those resources are. Lack of understanding as to how people earn a living is evident in four critical silences within the economic development literature. First, recognition of indigenous economic strategies employed by people within diverse socioeconomic contexts has been largely absent. Second, development research fails to fully appreciate that a diverse set of economic strategies, including those comprised of informal sector activities, may provide opportunities for stable income generation and even accumulation. Third, current approaches fiequently treat people in a uniform fashion, or makes distinctions only based on class and gender. Even women in development (WID) and gender and development (GAD) approaches tend to essentialize women, assuming that all are equally disadvantaged or prioritize class based distinctions. Fourth, enabling conditions that

1 shape individual choice in economic activities have been poorly understood. One important dimension of this issue has been inadequate comprehension of the multiplicity of ways cultural norms and practices around reproductive responsibilities constrain or enable mobility and therefore, people’s economic activities, especially for women. This study argues that geographically localized social institutions are (re)produced through people’s daily social and economic practices. The results are differential opportunities for income generation. Therefore, variation in the livelihood strategies people assemble 6om different economic activities reflects their life circumstances. The object of this research is to identify and explain basic patterns of economic strategy usage and outcome in one context, by examining social processes that give rise to different configurations. While process is clearly important, the main goal of this study is to understand fundamental patterns of difference. To develop the argument this research proceeds in four steps, based on the experiences of women in Porto Novo, The Republic of Benin (hereafter referred to as Benin). First, strategy configurations employed by women in Porto Novo are identified, where strategy configuration is defined as all possible combinations of distinctive income generating activities. Second, enabling conditions fiom which various strategy configurations emerge are identified. Enabling conditions are the necessary factors for pursuing particular opportunities. Third, the accumulation potential of different configurations, relative to enabling conditions that facilitate their use is assessed. Here accumulation potential refers to the ability to maintain and/or increase one’s financial capital. Finally, strategy configuration relative to enabling conditions fiom which they emerge, are explained in terms of social processes that produce those conditions. In examining these issues, this work offers three contributions to development research. First, it extends current gender and development theory to account for multiple dimensions of differences among women, and the role of geography in access to resources. Second, it develops a methodology for examining the range of economic strategies people employ. Third, it provides empirical evidence of variation in strategy configuration and outcome in one context. 1.2 Problem Statement This research examines people’s livelihood strategies - how they earn a living. I became interested in this topic while in Nigeria during the summer of 1994. At that time, all oil workers went on strike to protest the annulled 1993 elections and subsequent imprisonment of Moshood Abiola, the presumed winner. As a result, there was no fiiel available, including kerosene for cooking, except when sold on the black market at exorbitant prices by the military who controlled the country. Because banks were closed, there also was little money in circulation. These factors added to on-going economic difficulties faced by the Nigerian population resulting from structural adjustment and military dictatorship. Yet, people adapted by changing their mix of economic activities, or adding new enterprises to their on-going economic strategy. They had the flexibility to do so because their capital was not invested in a single business venture. Most had several different sources of income, called here, a Multiple Livelihood Economic Strategy (MLES)'. A variety of studies have documented the increasing tendency for people to simultaneously maintain several economic activities as a function of changes in the global economy (Stinson Jr 1986; Stinson Jr 1987; Stinson Jr 1990; Bin 1992; Rongguang 1993; Reardon, Crawford et al. 1994; Coombes and Campbell 1996; Valdivia, Dunn et al. 1996; Amirault 1997; Murray 2000; Owusu 2000; Tienda 2000; Slater 2001). As Mingione (1991:1) noted,

recent patterns of tertiarization and industrial restructuring have rendered more and more complicated and heterogeneous even the offlcial employment structure of advanced capitalist countries

' A multiple livelihood economic strategy is defined here as any combination of two or more income generating activities, including two or more distinct own-account activities, but not including unpaid family labor, self-provisioning and/or barter activities.

3 These studies suggest it is inappropriate to think of economic strategies simply in terms of a single dominant activity. Rather, people use a wide array of livelihood strategies ranging from mono-activity forms, the typical object of economic research, to an assortment of pluri-activity patterns, comprised of two or more sources of income. Previous work on MLES is largely descriptive, however, emphasizing its use of as a means of survival. The possibility that various forms of this approach to income generation can lead to different kinds of accumulation is not considered. Additionally, how and why people choose particular strategies is not explained. Nonetheless, thinking about individual and household income generation capacity in terms of MLES facilitates moving beyond dualisms that commonly characterize livelihood activity, including; production/reproduction, formal/informal, petty trader/mamabenz, survival/surplus^. Elaborating on these themes addresses two fundamental questions. How and why do people use different livelihood strategies? What are the outcomes of various strategy configurations in terms of accumulation potential? Considering especially life circumstances, this work investigates enabling conditions emerging from diverse socioeconomic contexts that facilitate or inhibit use of different strategy configurations including mono-activity forms involving only one economic activity and diverse approaches comprising multiple income sources. Specifically, through a case study of women in Porto Novo, Benin, this research illustrates how differences among women relative to their life-cycle phase, ethnicity, residential location and class provide unique possibilities for strategy creation.

^ Mamabenz is a term used in West Africa, in particular, to describe large scale traders usually of imported goods such as fabric because they are often known to drive Mercedes. 1.3 Theoretical Considerations Within the sustainable livelihoods literature, (Blaikie, Cannon et al. 1994) have argued, with respect to natural disasters, that people’s ability to cope with adverse conditions depends on access to resources including financial, material, and social capital. They define access as the ability of an individual or group to secure resources necessary for survival. The capacity to do so always depends on social and economic relations. They recognize that resources are not spread evenly, and therefore, access includes spatial as well as social, political and economic elements. There are two limitations to this approach, however. First, is the survival orientation. Second, it does not recognize differences within groups such that some members might have better access to resources than others. For example, (Blaikie, Cannon et al. 1994), consider gender as simply a primary division among people in society, arguing that women are always disadvantaged in their access to and control over resources. While this may be true, the assumption that all women are equally disadvantaged is not. Also, the argument can be extended and applied to the resources available for creation of livelihood strategies, not only for survival, but also for accumulation. A feminist theoretical perspective provides an alternative means of considering how variation in access to resources among individuals and groups enables or inhibits the creation of livelihood strategies. Specifically, feminist theory fiom minority and developing country perspectives argues that people’s opportunities are shaped by the interaction among broad social structures, called here Dimensions o f Difference, including gender, class, race, ethnicity, life-cycle phase, sexual orientation, etc. (hooks 1981; hooks 1984; Collins 1990; Young 1990). The interplay among these factors is manifest in geographically situated social institutions in the form of assets distribution, cultural norms, political rules and personal preferences, known collectively as Structures o f Constraint, which produce enabling conditions for livelihood strategy configuration ^olbre 1994; Hanson and Pratt 1995; WGSG1997). Preferences are included here because while people make "purposeful" choices about their lives, these are conditioned by the socioeconomic and political context within which they live (Folbre 1994).

1.4 Regional Context and Themes of the Research The empirical foci of this work are women in Porto Novo, Benin. Many people across the globe increasingly use an array of livelihood strategies, women in Africa, and Benin specifically, have a long well-documented history of using MLES (Ngaza and Koda 1980; Stichter 1988; Robertson 1990; Aina 1991; Mustapha 1992; Guyer 1995; Koopman 1995; Robertson 1995; Davison 1996). Thus, Benin provides an excellent case study in which to examine the questions outlined above. Data for the study were collected during fourteen months (September 1997 to November 1998) of fieldwork in Benin. Three different data sets were gathered: 232 preliminary surveys, 535 in-depth surveys, and 15 life and economic histories. The wide variety of data facilitate use of quantitative analysis to identify broads patterns in strategy configuration and enabling conditions. Qualitative data including interviews and open- ended survey questions, permit a complementary analysis, specifically focusing on social processes behind the identified patterns. Although Benin does not have critical resources, and therefore is not a major player in the global economy, it performs a vital function within the regional economy as an entrepot state (Igue and Soule 1992; Maula 1997). Goods imported and produced throughout the West African sub-region are re-exported through Dankotpa, one of West Africa's largest markets, located in Cotonou. Porto Novo is a key transit point in this trade due to its location on the Nigerian border and close proximity to Cotonou. Women play an integral role in this lucrative trade, which is often conducted informally and illegally (Igue and Soule 1992; Flynn 1996; Maula 1997). With the imposition of structural adjustment in 1991 and subsequent currency devaluation in 1994, trade activities have increasingly comprised at least one dimension of Béninois women’s livelihood strategies, even where they have wage employment. As the overall number of traders increased, competition among them has likewise rose. As a result, women perpetually look for profitable commercial opportunities. Often this search involves mobility outside Porto Novo to Nigeria, Togo, and other regions within Benin. Further, in this form of trade access to substantial capital is not always necessary. The ability to be mobile outside Porto Novo reflects differences among women in cultural norms and practices around reproductive responsibilities. Ethnicity is a key dimension of difference. For the Yoruba, the second most prevalent group in Porto Novo, women’s reproductive responsibilities include the ability to feed themselves and their children through their trade activities. In contrast, among the Goun, the dominant group in Porto Novo, women’s reproductive responsibilities are customarily defined in terms of care-taking. Variation among the Goun is evident, however, depending on residential arrangement (female-headed, nuclear or extended families) and location (rural or urban)\ Likewise, as women’s care-taking responsibilities decrease with age, they frequently gain freedom of movement, but this again depends on residential arrangement and location. Finally, women of higher socio-economic status are also more mobile, but this is often associated with larger households, which results in increased financial responsibilities and/or residential arrangement and location (usually urban nuclear families). Geographic mobility, therefore is a critical factor in Béninois women’s abiUty to earn a living. Being mobile can make the difference between a strategy of survival and surplus (accumulation). The ability to be mobile, however, emerges from interactions among broad social structures, which are manifest in geographically localized institutions including assets distribution, formal rules, social norms, and personal preferences (Folbre 1994). In other words, a variety of social processes shape the kinds of strategies women can create and therefore, their ability to earn a living. A flowchart depicting the overall argument of the dissertation is presented in Figure 1.1.

^ Nuclear families in the Béninois setting include parents and children. Extended families on the other hand, usually include several generations, but do not usually involve a polygynist man living with several wives. CONCEPTUAL FRAMEWORK Imaiswmo

4 i-i -■ ::'-

$ EMPIRICAL FRAMEWORK

SiKa

nabmg^n

4i:^iSéÛh0^^

"v # * sg 8 B ' ' ' Mbno^Activi^ lutalfAc

Survival

Figure 1.1 Conceptual and Empirical Framework for Dissertation 1.5 Organization of the Dissertation Having provided a glimpse at the issues this study addresses, subsequent chapters focus on the key questions that underlie this work (see Figure 1.2 for a flow chart of the dissertation organization). Chapter 2 asks the question, ‘Do livelihood strategies vary and is there an increasing tendency for use of MLES?’ In answering this question a critical a review of the literature on household and individual economic strategies, as they pertain to MLES, is presented. The second key question, addressed in Chapter 3, is ‘How can we explain differences among people, even within the same group, in their access to resources?’ This chapter provides a conceptual framework for explaining differences among people grounded in a feminist theoretical approach. In particular, it synthesizes three arguments that facilitate understanding the social construction of difference. First, feminist theory from minority and developing country perspectives focusing on the social construction of difference is reviewed. Second, woric in feminist economics linking theories of difference with an institutional framework is discussed. Finally, feminist geographic theory provides an explanation for the spatiality of these social processes. Chapter 4 provides an overview of socio-economic and political history of Benin. The central question here is "How can the framework be applied to Benin?’ Special attention is given to women’s role within this political economy. Additional discussion centers on the operationalization of the theoretical framework in the Béninois context. Data collection methods are then presented. The chapter ends with a presentation of sample characteristics and discussion of its representativeness. Results from the analysis are reported in Chapters 5 through 9. In Chapter 5 the question ‘What strategies do women in Porto Novo use?’ is posed. Content analysis of the open-ended survey questions regarding the relative importance of each economic activity and the basis for combining activities is used to determine the nature of strategy configurations. Subsequently, a cluster analysis is used to group women according to their particular strategy configuration. Introduction {Chapter 1) y Empirical Framework: Do Strategies Vary? {Chapter!) y Theoretical Framework: How can strategy variation be explained? {Chapter 3) y Context and Data Collection: How can the framework be applied to Benin? {Chapter 4) y Cluster Analysis: What Strategies do women in Porto Novo use? {Chapter 5) y Chi Square Analysis: What enabling conditions shape strategy choice? {Chapter 6)

What is the relationship between enabling conditions and accumulation potential? {Chatter 7)

What social processes give rise to various enabling conditions? {Chapters 8-9)

Conclusions {Chapter 12)

Figure 1.2 Organization of the Dissertation

10 A Chi-square Automatic Interaction Detection (CHAID) analysis is used in Chapter 6 to determine the nature of the constraints and opportunities that give rise to particular strategy configurations. The purpose here is to answer the question ‘What is the nature of constraint and opportunity influencing strategy choice? This leads to the question posed in Chapter 7, which is ‘Do different opportunity structures provide variable possibihties for accumulation?’ To answer this question comparison of median starting and current capital is used in combination with interview data to assess the importance of capital and mobility for income generation. Finally, Chapters 5 and 9 answer the question ‘In what way do the opportunity structures reflect different social processes’? Here, descriptive statistics are used to identify the principle processes at work in shaping enabling conditions. Having identified the dominant processes, interview data are used to explain how they work.

11 CHAPTER 2

EMPIRICAL FRAMEWORK: DO STRATEGIES VARY?

2.1 Introduction As noted in the introduction (Chapter I), the main purpose of this chapter is to examine the question 'Do livelihood strategies vary?’; focusing on those consisting of more than one economic activity. An overview of relevant literature reveals three lines of inquiry underlie the answer to this question. First, studies firom a wide variety of settings demonstrate that strategies vary among households. Second, recent studies suggest that use of Multiple Livelihood Economic Strategies (MLES) is increasingly used by individuals in response to globalization and economic restructuring'*. Third, A6ican women have a long well-documented history of MLES for income generation, emphasizing a wide array of distinctive strategy configurations. The preponderance of research evidence notwithstanding, studies focusing on economic strategies has yet to explain in a systematic fashion why and how people use particular strategy configurations. This reflects the fact that many studies accentuate survival as the principle aspect of MLES use, largely ignoring other outcomes such as accumulation. Variation among people in goals and outcomes of their various strategies is rarely considered. Given the evidence of increasing MLES usage across the globe, it seems past time to examine the conditions that give rise to its use.

■* Multiple Livelihood Economic Strategy (MLES) refers to any economic strategy comprised of two or more distinctive economic activities. Other terms used to describe this Qfpe of economic strategy include: moonlighting, pluriactivity, and multiple modes of livelihood (MML).

12 This outline is elaborated in three steps. First, an overview of literature on MLES is presented. Second, work speciEcally documenting this approach to income generation by Afncan women is discussed. Third, reasons for the high level of MLES usage among African women is examined with specific reference to the gender division of labor within Afiican economies and households. In addition, changing patterns of strategy configuration with respect to historical changes in Afiican political-economies are considered. The chapter concludes with a discussion of gaps in these literatures, which the current study seeks to address.

2.2 Multiple Livelihood Economic Strategies in General 2.2.1 Introduction Much literature documents the use of MLES, in both developing and developed world settings. Three themes are evident in. First, some studies explain MLES in terms of economic restructuring and globalization. Second, many studies cite survival as the principle reason people take on numerous economic activities. Third, a few studies implicitly consider outcome variability, arguing those with better access to resources can more effectively utilize MLES,. In both developing and developed country contexts, work on households emphasizes MLES in rural areas by farm families and in urban locales by the under and unemployed. Work addressing use of MLES by individuals has become increasingly prevalent. 2.2.2 Use ofMultiple Livelihood Economic Strategies Among Households Research focused on rural households indicates MLES is used to diversify income sources and spread risk, especially as profits from small-scale farms decline. One important body of literature addressing MLES in developed country contexts comes from rural sociology. These studies suggest that as the viability of family farms have declined, households increasingly use non-farm and/or off-farm work by one or more family members to maintain household income ODeseran, Falk et al. 1984; Hallberg, Findeis et al. 1991; Lobao and Meyer 1995; Coombes and Campbell 1996; Ubery, Healey et al. 1996). This work acknowledges that non/off-farm income to supplement farm earnings is

13 not a new trend, but indicates the pattern of part-time farming is becoming permanent (Hallberg, Findeis et al. 1991; Coombes and Campbell 1996). The variety of strategies used include combining full or part-time farming with one or more of the following: non­ farm enterprises, such as farm shops and zoos, and off-farm activities such as wage and self-employment activities (Coughenour and Swanson 1983; Deseran, Falk et al. 1984; Lobao and Meyer 1995; Ilbery, Healey et al. 1996). Similarly, in developing country contexts, studies indicate that farm households use a myriad of strategies to supplement income as the viability of small-scale farming declines (Stark 1991a; Stark 1991b; Reardon, Crawford et al. 1994; Valdivia, Dunn et al. 1996). One important body of work focuses on sustainable rural livelihoods (Carney 1998; Francis 2000; Vogel and Smith 2000). As with farm families in developed country contexts, MLES in developing settings involves multiple sources of income such as non/off-farm wage and self-employment including trading, selling services, and/or producing goods for sale (Francis 2000). Rural households change their mix of activities seasonally, in response to various environmental and economic conditions and as household characteristics change as a function of age, educational attainment, time and resource availability. Strategy goals may likewise shift over time from survival to security and possibly even to accumulation (Reardon, Crawford et al. 1994; Evans and Pirzada 1995; Carney 1998; Francis 2000; Vogel and Smith 2000). By combining economic activities families diversify income sources and spread their risks. Ability to do so is critical because any single activity is unlikely to provide sufGcient income (Francis 2000). Moreover, households are using an increasingly diverse array of strategy configurations than in the past (Francis 2000). Another MLES noted in research on rural farm families in developing regions; involves migration, wherein one or more family members move temporarily to a major employment site and send remittances back to the household. In this instance, household member may have limited opportunities to pursue wage employment in the local area. Therefore, the family has no choice, but to look for alternative income sources elsewhere (Trager 1988; Stark 1991a; Stark 1991b; Evans and Pirzada 1995; Galli and Funk 1995).

14 Urban households, in both developing and developed regions, also choose among a variety of diversification strategies when income from the primary breadwinner is insufGcient and/or unstable. Two bodies of literature dominate here; work on the new immigrants in developed regions and the informal economy in developing countries. In developed country contexts new immigrants use a variety of strategies to cope with intense competition in their particular business niches. Apart from self-exploitation, many use MLES through vertical integration of their enterprises by moving up/down the production chain, or through expansion/ diversification by opening other shops (Aldrich and Waldinger 1990; Waidinger, Aldrich et al. 1990). For example, an entrepreneur may have both wholesaling and retailing enterprises (Werbner 1984). Alternatively, new inunigrants may combine wage and/or self-employment. More recently, Tienda and Raijman (2000) argue the variety of strategies new immigrants employ is underestimated due to lack of accounting for informal activities. They found once multiple job-holding by all members of a household, including informal activities, are considered the variety of strategies employed by immigrant households increases dramatically. Although households are the unit of analysis in all this work, the research also documents use of MLES by individual family members as well (Aldrich and Waldinger 1990; Waldinger, Aldrich et al. 1990; Tienda 2000). In developing and developed country contexts, work on the informal economy, especially in Latin America and Asia but more recently also in Italy, demonstrates use of MLES for purposes of survival in the face of economic restructuring: As Moser (1992:99) notes, "fewer households can now depend on the income of one earner." Accordingly, all household members with the capacity to contribute to household income, including children. Here strategies usually involve combining wage employment and/or various own-account activities. In this work, not only is MLES evident at the household level, but also often individual family members have multiple sources of income. (McGee 1979; Roldan 1985; Mingione 1991; de Oliveira and Roberts 1994). Many of these studies suggest labor force participation is undercounted and income generating capacity under appreciated (Roldan 1985; Mingione 1991). One type of research focuses on

15 women’s industrial outworking as a mean of supplementing family income. For example, Roldan (1985) shows that women may have as many as three or four separate sources of income, only one of which is industrial outworking. The others usually comprise various service activities. MLES use by households has been addressed in a wide variety of research. Four themes are common to all. First, households employ diverse sets of strategies. Second, MLES approaches to income generation are usually employed as a survival strategy in the face of adverse conditions, particularly recession, economic restructuring and/or limited employment opportunities. Third, supplementary activities often take place in the informal sector, and as a result, labor force participation is often underestimated. Fourth, MLES use occurs within a household context, but also by individuals. Although this work conclusively documents the use of MLES in a wide array of settings, these studies have not explicitly examined causal mechanisms and strategy outcomes. This is especially evident in emphasis on survival as the principle explanatory process for MLES use. Therefore, variation in strategy usage is not effectively explained. 2.2.3 Use o f Multiple Livelihood Economic Strategies Among Individuals Several studies document MLES employment by individuals. Three themes are evident. First, changing patterns of MLES reflect changes in the economy. Second, motivations for using this approach are primarily viewed as survival oriented. Third, research focus emphasizes the impact of MLES use on employers. Starting in the late 1950s researchers began examining ‘moonlighting’. Early studies explained MLES use in terms of shorter work weeks that allowed people to have more than one job. The main motivation given was financial need. Of particular concern was the impact on businesses. This work argued that moonlighting diminished the objective of reducing the work week and could potentially be costly to employers as accidents increased due to worker fatigue (Habbe 1957; Mott 1965). Research by the U.S. Bureau of Labor Statistics, throughout the 1970s and 1980s, documents MLES as an increasing trend in U.S. labor markets. This reflects economic fluctuations such as decreasing unemployment and reductions/flexibility in work hours.

16 Patterns identified suggest those with the greatest access to resources (e.g. social capital in the form of business networks) will more effectively mobilize MLES. Financial need, however, continued to be viewed as the dominant motivation (Perrella 1970; Hayghe and Michelotti 1971; Grossman 1975; Jamal 1986; Stinson Jr 1987; Stinson Jr 1990). Although men were identified as the primary utilizers of MLES, increasing usage among women was noted in the late 1980s (Stinson Jr 1986). Starting in the 1990s a shift in the trend of MLES use was noted. While earlier studies focused on use by ‘blue collar workers’, more recent studies

indicate that relatively well paid, highly educated persons have more than one job because their schedule allows for it, because their expertise is in demand, or because their financial reasons extend beyond meeting basic living expenses and paying off debts. (Amirault 1997)

As with earlier patterns this change reflects economic restructuring. These findings are consistent with another study that documented use of MLES among university professors (Jamal 1986). The increased importance of MLES is evident in that the 1994 and later Current Population Survey, a monthly household sample survey, was redesigned to account for multiple job-holding (Stinson Jr 1997). Studies in other contexts also note individual MLES use. As with studies in the U.S. context, MLES is explained in terms of financial need and opportunity. A series of research projects on four regions of Italy document the increasing tendency for individuals to have more than one source of income, especially in the informal sector. Specifically, this work argues that the system of job guarantees facilitates MLES use and that those with better resources will more effectively mobilize the strategy (reviewed in Mas! 1987). Similarly multiple job-holding in the European Community (EC) countries is common. The focus of this study, however, is effective accounting of job creation in member states. As a result, not much information is provided about the individuals who actually employ the strategy (Alden and Spooner 1982). Finally, two recent articles in the Beijing Review document MLES usage among Chinese professionals. The main

17 focus of these articles is the opening up of the Chinese economy, permitting multiple job- holding for the first time. The central concern is the implications of MLES usage for the Chinese economy ^ in 1992; Rongguang 1993). The orientation of these studies on individual MLES use is survival. Although variation in strategy type is suggested by findings that those with better resources will more effectively mobilize MLES, this is never examined or explained in detail. The lack of information on the way individuals use MLES reflects the foci of these studies, which is economic restructuring and appropriate accounting of job creation. Thus, while it is clear MLES is used by both households and individuals, little is known about how and why people adopt this strategy.

2.3 African Women and Multiple Livelihood Economic Strategies 2.3.1 Introduction Although no studies explicitly examine MLES use among Afiican women, research investigating other aspects of their lives and economic activities clearly document it. Indeed, the fact that they engage in multiple economic activities is often seemingly taken as given and therefore, not scrutinized. This work substantiates Afiican women’s long standing practice of using MLES and provides a basis for understanding why they use this approach to income generation. Ethnic, class and age differences among women in their economic strategies are clearly evident, but general themes are nonetheless identifiable. Two themes especially relevant for explaining MLES usage are: gender ideology vis-a-vis women’s reproductive responsibilities and the changing gender division of labor within Afiican economies. In documenting MLES applications among Afiican women, this section will historically examine these related processes. Employment of MLES is evident in most work examining Afiican women’s lives. Research detailing women’s life histories firequently include descriptions of simultaneous maintenance of numerous economic enterprises ^g aza and Koda 1980; Deniel 1985; Davison 1996). One such work includes several interviews with Béninois women, raised during the colonial period, all of whom engaged in wage employment and/or various

18 trading activities QDeniel 1985). Studies analyzing specific aspects of women’s lives and/or work also incidentally refer to the use of MLES (Robertson 1990; Aina 1991). For example, Robertson's (1990) work on the impacts of socioeconomic change on Ga women in Ghana, mentions that women engage in economic activities supplementary to their primary occupations. In rural settings women usually combine farming with various ofFnon-farm activities including wage-employment, service provision, and trading. Women employ similar strategy configurations in urban areas, but replace farming with market gardening. While these issues have not been explored in detail, it is possible to summarize women’s historical use of MLES as a fimction of the two inter-related processes identified above. Limited employment opportunities and an on-going intensification in financial and domestic responsibilities characterize women’s economic history in Afiica^ These conditions are apparent in two ways. First, men usually dominate occupational categories that offer the most income generating opportunities by virtue of their high economic status. Second, gender ideology in virtually all settings define women as the main providers of their families’ basic needs (Aspaas 1998). Moreover, the gender division of labor has become increasingly segregated such that today many women throughout Africa are the sole providers for their own and children’s needs no matter their marital status. Simultaneously, women’s opportunities for earning sufficient income (even through an array of economic activities) have diminished substantially. Thus, while African women often have a high degree of economic autonomy and independence, their constrained economic prospects renders this situation impossible for many. The persistence of women’s limited occupational opportunities relative to men is explained by access to capital, labor, education, and waged employment, and the co-optation of women’s work by men when it becomes lucrative.

* For an excellent review of Afiican women’s history that echos the argument being made here see Berger, 1999.

19 2.3.2 The Pre-Colonial Period In pre-colonial times women had three primary roles: reproduction, production for husbands, and own-account economic activities. The first two fimctions superseded women’s own account enterprises. This was problematic because in polygynous family settings, prevalent throughout Afiica, women did not expect husbands to provide for their needs or those of children (Awe 1977; Fapohunda 1983; Afonja 1986; Dennis 1991; Koopman 1995). They met these responsibilities primarily through subsistence agriculture, augmented in some settings with trade. In many pre-colonial Afiican societies, which were primarily agricultural in organization, the gender division of labor tended to be complementary. Men did jobs that required greater strength such as clearing land, whereas women and children preformed fimctions such as preparing land, planting, and weeding. Everyone participated in the harvest (Afonja 1986; Awe 1989; Ogbomo 1995). In addition,, in some instances, there was gender division in crops grown by men aqd women. Among the Esan, for example, men grew yams, considered “the king of crops”, whereas women grew all other produce (Ogbomo 1995). Prior to imposition of colonialism, many Afiican societies began to engage in capitalist modes of organization and production, stemming in large part firom growing external trade to North Afiica and Europe. While these changes created new opportunities for some groups of women, such as the Yoruba, they also increased the segregation of responsibilities between men and women. Moreover, most women continued to have fewer economic prospects than men. Men always engaged in the most lucrative form of production, with the possible exception of the Yoruba (Koopman 1995; MCCall 1995). There was strict distinction in artisanal activities between men and women in many societies. For example, among the Esan, men engaged in wood carving, metal smithing and basket weaving. Women, on the other hand, worked in pottery-making, soap-making, cloth dyeing and weaving, palm-oil and food-processing (Ogbomo 1995). Likewise among Yorubas, women worked as potters and soap-makers, in addition to processing food for sale. Men, on the other hand.

2 0 were blacksmiths, iroa-smelters, and carpenters (Fadipe 1970). This is not to diminish women’s contributions to family maintenance, which were highly valued, but simply to point out that woman’s primary role in household food production limited their access to economic status. Among the Senufo, men engaged in the cultivation of staple crops and livestock raising, which were activities of higher economic value, whereas women engaged in commercialized agriculture of rice and groundnuts (Guyer 1995). For the Gikuyu, in East Africa, the pattern was reversed. If the crop had commercial value, men cultivated it. If it was primarily for subsistence purposes, women did the farming (Davison 1996). Long-distance, as opposed to local, trade was usually considered more profitable. Among the Igbo in West Africa, men engaged in long distance trade primarily for safety reasons, but also because it was the most lucrative (Peil 1975; Van Allen 1976; Njoku 1980). Two exceptions to this are the Yoruba of Nigeria and Benin and the Asante of Ghana. In both cases women engaged in long-distance and local trade, as did men (Fadipe 1970; Hades 1980; Falola 1995). The commodities sold by men and women differed substantially, however, with men controlling the most profitable items (Clark 1991; Dennis 1991; Clark 1994). Among the Yoruba, men controlled trade in such imported goods as gun powder, whereas women controlled locally produced foods and crafts (Afonja 1986). Moreover, men controlled cash cropping, a major source of the increasing level of urbanization among the Yoruba in the 18th century. This, in large part, stemmed from women’s lack of access to capital (Afonja 1986). The result was an increasingly separate gender division of labor in which men farmed and women traded (Spiro 1987). Women compensated for lack of access to the most profitable activities by combining trade in a number of different items, as well as selling in numerous locations (Falola 1995). Despite the necessity of generating independent incomes in many parts of Afiica, due to the wide variety of roles they occupied within the household, women’s lack of access to the means of production, labor and their right to control their own time limited their ability to do so. For example, Igbo women only had the right to work their own

2 1 plots after working in their husbands fields. A similar pattern is evident among the Gikuyu in Kenya in that women (Davison 1996). As a result, use of other people’s labor was critical, as was sharing household responsibilities among wives in polygynous marriage. In the latter case senior wives usually had control over younger wives labor and therefore had more time for their own account economic activities (Robertson 1990). One way women, in West Africa particularly, increased their own-account economic activities was through use of slaves and pawns. Due to the gendered and geroncratic hierarchical nature of social structure in most Afiican societies only elderly women had any rights to others labor. Even a woman’s children were first a source of labor for her husband. Robertson and Klien (1983) argue this is the main reason for greater use slaves among women than men. Although often only elite women had access to slaves, in some societies even “commoners” could acquire them. Whereas their husbands usually gave elite women slaves, common women had to purchase them. Thus, in many West Afiican societies where slavery was widely practiced, women had a means of divesting themselves of some of their roles, especially their domestic responsibilities, which in turn, facilitated maintaining independent income generating activities (Robertson and Klien 1983; Afonja 1986; Dennis 1991). Women still lacked access tOo other means of production, however. In rural settings women's usufiiictuary rights to land depended on a male family member. Consequently, while women traded and earned independent incomes, ability to do so was dependent on male allocation of land since women's main commodities were agricultural goods in either a raw or processed form. Among the Ga, male priests initially allocated land to lineages, in which male heads would then allocate land use rights (Robertson 1990). Yoruba women were similarly denied the right to inherit land as a fimction of the patriarchal social structure of Yoruba society (Afonja 1986). Even in societies with matrilineal lineage systems women derived their land rights from their brothers or uncles (Guyer 1995; Koopman 1995). Yoruba, Ga, Igbo, Senufo, and other West Afiican women all had independent rights to trade. They frequently obtained their merchandise from husbands, however. Ga

2 2 women fish traders acquired the fish they sold firom a male relative - usually fathers or husbands (Robertson 1995). One exception to this was Yoruba women traders, who sold their husbands agricultural products and their own crafts including pottery and fabric (Guyer 1995; Koopman 1995; Saul 1995). Another exception was sale of farm produce in urban areas. Generally, West Afiican women kept profits firom sale of surplus produce firom their own plots, which they often sold in processed form, such as millet beer sold by Senufo women and palm oil sold by Yoruba women (Guyer 1995). Moreover, Yoruba women, became increasingly exceptional in that as Yoruba society urbanized, some women were no longer expected to engage in agriculture at all, allowing them to engage exclusively in trade. They were, however, also increasingly expected to have their own independent sources of income through which to support themselves and their children (Afonja 1986; Dennis 1991; Guyer 1995; Koopman 1995; Saul 1995). Not only did women have multiple roles within their societies and households, but they also often had multiple occupations. In rural areas women worked for their husbands farming and selling the produce, keeping a portion of the proceeds. In addition, they had their own-account activities including farming, artisanal activity and trading. In urban areas, women sometimes worked for their husbands selling farm produce. They also managed multiple income generating activities of their own. 2.3.3 The Colonial Period With the introduction of colonialism, women's economic position became changed. Colonials, with help fi-om indigenous men, formalized women’s lack of control over the means of production by codifying land tenure rights such that women were denied independent ownership of land. They also limited women’s access to Western education which had implications for waged employment opportunities. The ultimate result was women were caught in a "catch-22" between their primary responsibilify for family maintenance, and any legal and "appropriate" means fulfilling that responsibility except trade. Furthermore, under this system colonial administrators fi^quently denied women men's assistance providing for family needs. They achieved this by recruiting

23 men into wage employment through necessity of paying taxes in cash. The result was that many men lived away from their families. This further intensified the gender division of labor (Berger and White 1999). Lack of education denied women access to wage employment in the urban sector as well as access to political power, both in the colonial and post-colonial periods, due to the importance of literacy. Even where some education was provided to girls, it usually emphasized training them for Western idealized roles as wives and mothers (Van Allen 1976; Wilson 1982; Seidman 1984; Robertson 1986; MacGaffey 1988; Awe 1989; Hay and Stichter 1995; 0"Barr and Firmin-Sellers 1995). The only opportunities available to women who gained full education were in occupations considered "traditionally" (according to Western Victorian ideology) female such as nursing and teaching, which are generally characterized by their low status and low pay (Van Allen 1976; Fapohunda 1983; Seidman 1984). Among women in Nigeria low levels of wage employment are clear in that in 1950 (the latter part of the colonial period) "less than 2% of all paid employees were women." (Afonja 1986). In rural areas, lack of education did not play a significant role in determining access to wage employment, such as export oriented agriculture. Victorian ideology of colonials, however, resulted in the hiring of men because they saw women as too weak for that kind of work, though Afiican women were historically primarily responsible for agricultural production (Van Allen 1976; Matsepe 1981; Koopman 1995; O'Barr and Firmin-Sellers 1995). Thus, here too, women were left with only "traditional" female occupations related to household maintenance, including food production, and in a few cases wage labor in domestic service ^atsepe 1981). Rural women also engaged in trade. For example, among the Akan, women continued to sell both their own and their husband’s produce as well as engage in other forms of trade (McCall 1959). In this way, the introduction of export economic activity intensified the customary gender division of labor and simultaneously reduced women's income generating opportunities (Guyer 1995; Wipper 1995).

24 One of the few avenues for income generation left available to West African women was trade. Initially, in the colonial period women gained access to those areas of trade that men formerly dominated (McCall 1959). Thus, one could argue women gained some ground during the colonial period with respect to trade (Dennis 1990; Ekechi 1995). Women’s trade activities grew in two specific areas. First, in the sale of food stuffs because there was no effective European competition, at least initially (Clark 1991). Second, women began retailing imported goods. In Ghana this occurred through the pass book system and women’s control of local markets (Njoku 1980; Okine 1993; Clark 1994). In some areas, however, men maintained control over the most lucrative forms of trade. In Benin women dominated all aspects of trade except wholesaling imported goods (Tardits and Tardits 1962). This pattern also pertained to Yorubas, with men maintaining control over wholesale of imported goods until the twentieth century (Afonja 1986). Similarly, in some areas men continued to dominate long-distance trade. They were able to do this because they had greater access to European currency and new modes of transportation including railways, bicycles, and trucks (McCall 1959; Ekechi 1995). We should note, however, that trade became a marginal activity relative to those occupational opportunities available to men. Moreover, increasing availability of European goods undermined women’s indigenous craft and trade activities. This resulted from the higher status of imported goods and the fact that they tended to be cheaper due to economies of scale in production and distribution available to colonials, but not African women (Dennis 1991). Colonialism also undermined women’s indigenous food processing in some areas through introduction of mechanized processing technology. Once technology was introduced, activities became male dominated (Awe 1989). For example, palm-oil processing, purview of women in many pre-colonial societies, was taken over by men as production became mechanized (Ogbomo 1995). Furthermore, gains women achieved, in terms of trade activities, failed to last. In the latter colonial period in Ghana, price controls on the cost of imported goods and the increasing preference of colonial administrators to distribute these goods through large

25 stores, once again reduced women’s trade activities (Clark 1994). Therefore, gains made by men, coupled with growing separation in the gender division of labor, in which women, alone, increasingly bore the burden of family maintenance undermined any gains women made in terms of income generating opportunities. The result of intensification in the gender division of labor severely impacted women across Afiica, especially in white settler colonies. In these areas Colonial administrators expected women's subsistence agriculture to maintain families thereby justifying paying men extraordinarily low wages (Wilson 1982; Seidman 1984; Guyer 1995; Koopman 1995; Davison 1996). Households had few opportunities for survival under these conditions as colonialists took the best land leaving Afiicans virtually landless or with only marginal land (Matsepe 1981; Guyer 1995). As in other parts of Afiica the few occupations available to women were consistent with customary avenues to independent income including petty trade and beer brewing (Matsepe 1981; Seidman 1984; Wipper 1995). Additionally, out-migration of men fiom rural areas to mines and plantations resulted in a new occupation for women, prostitution. Both colonial Victorian and indigenous ideologies considered this means of income generation illegal and immoral, however (Wilson 1982). Fear that women would engage in prostitution resulted in indigenous men working with colonial officials to limit women’s mobility in trade. Women were denied the right to trade in Nairobi, Kenya, for example, which also reduced their income generating capacity (Robertson 1997). Through these persistent patterns of marginalization women were increasingly responsible for maintaining households with fewer economic opportunities for fulfilling those obligations (Dennis 1990). This meant they had to work for wages despite limited opportunities. These conditions are evident in life-history interviews conducted with Béninois women, who came of age during the colonial period. For example, one woman combined work as a nurse with trading, allowing her to save enough money to open her own clinic when she retired fiom colonial service. Another woman worked as both a restauranteur and trader (Deniel 1985). One interviewee stated:

26 It is necessary that a woman can buy what she needs without being obliged to ask her husband all the time. But for her to be independent and at the same time furnish her part in the household (Deniel 1985:26)

Consequently, many women had multiple sources of income even where they did not have wage employment. 2.3.4 The Early Post-Colonial Period These patterns persisted in the post-colonial period and as a result women's income generating opportunities continued to erode. Women had less access to schooling despite introduction of universal education in many parts of Africa (Little 1973). This reflects the need to engage in income generating activities and the unwillingness of parents to educate girls who will marry and no longer be a source of support for them in their old age. Girls also continued to lag behind boys in access to educational facilities. This pattern was clear in that only 81 girls were in primary school for every 100 boys. The ratio drops further to 72 girls for every 100 boys for secondary school (World Bank 1995 reported in Gordon 1996a). Because of low levels of educational attainment, Nigerian women remained the most illiterate and uneducated group in society (Peil 1975; Fapohunda 1983; Awe 1989; Robertson 1995). In some countries new opportunities arose for women through windfall profits from sale of major export commodities. The result was that both educational and waged employment opportunities increased. Even where women obtained a high degree of education, however, they were frequently limited in the type of "formal" sector wage employment available to them, assuming they could find any employment at all (Little 1973; Scott 1986; Brown 1989; Robertson 1995). Men, being more skilled, were more likely to "occupy positions of greater power and authority as well as the more highly paid jobs" (Afonja 1986). Women, on the other hand, were more likely to engage in "female" oriented occupations as teachers, typists, secretaries, teachers and nurses. For example, in Benin women represented seventy percent of all primary school teachers (Little 1973). Even considering these occupational categories, in 1966 Nigerian women only occupied 2.9 percent of all professional positions (Afonja 1986). Further, these positions tended to

27 be the lowest paid and status (Peil 1975; Fapohunda 1983; Dennis 1991). Yet, women only have access to occupations left vacant by men (Little 1973). Ironically, Robertson (1995) argues that education may prove detrimental to women in terms of their income generating capacity. While in school they are not learning &om their mother’s or through some form of apprenticeship about how to trade, which may result in lesser ability once schooling is finished and no wage employment is found. Moreover, she contends that even if women obtain wage employment, they are unlikely to have access to childcare assistance fiom family in urban areas, which makes it impossible for them actually to take a job. Peil (1975) supports these findings. Women’s lack of access to education also hurts them in another way. Since independence, most Afiican governments have provided access to capital and training through development programs. Women’s lack of education, however, prevents them fiom finding information regarding these programs, which results in their simply not knowing about them. This reflects their illiteracy and lack of access to social and economic networks through which such information is passed. Moreover, women are fiequently denied access to any form of credit, even subsidized programs through the State, due to their lack of collateral. This is indicative of women’s income earning capacity, which prevents them fiom accumulating savings or those material possessions that could function as collateral, especially land (Dennis 1990; Okine 1993; Osirim 1996). Likewise in unskilled wage employment, such as agriculture, which did not require education, women are relegated to lower status and paying positions (Daddieh 1989; Guyer 1995). Men continued to dominate the most lucrative occupations, not only in terms of formal waged labor, but also within the "informal sector” (Afonja 1986; Wipper 1995). Therefore, women, for the most part, continued to be limited to subsistence farming, trading, food processing and selling, and prostitution as sources of income. This pattern pertained even in areas such as Benin and Ghana where some industrial occupational opportunities developed (Fapohunda 1983; Allen 1989).

28 Women frequently engaged in trade even if they also had some fonn of wage employment. This stems from the need to augment earnings and establish a viable businesses as a source of income in retirement (MacGaffey 1988; Dennis 1990). For example, secretaries also sell clothes and food items (Dennis 1990). In Nigeria, trade activity increase and wage employment decreases with age. hi this regard Béninois women stand out from other women in Francaphone Africa. Whereas in Mali and Senegal, women’s involvement in the labor force, immediately following independence, was negligible, 66% of women in Benin in 1970 had some form of economic activity. In part this difference is attributed to the influence of Islam in Mali and Senegal (Peil 1975). Trade remained the dominant economic activity available to women. Even secluded women in the Islamic countries of northern West Africa maintain independent commercial enterprise. Here, too, use of MLES is evident as women sell numerous varieties of goods through the help of their children and relatives (VerEecke 1995). For other groups, trade became institutionalized as women’s dominant economic activity. Among Yorubas two-thirds of all adult women are traders (Spiro 1987). Although trade remained the main occupational category available to women, it is not as profitable as in the past. This stems from the drive of many newly-independent governments to "modernize" systems of production and distribution within their countries. As a result, they veiwed women’s trade activities as "backward" and thus, hindering progress. Moreover, in periods of economic decline, they scapegoat women’s trade activities as causes of deterioration. For example, in Ghana, the government argued that market women were the source of economic problems because they did not heed price controls and hoarded goods creating the situation of high prices and shortages in order to make themselves rich (Robertson 1995). In fact, it is women’s economic marginalia that makes them vulnerable to such attack. Given that many market women cannot afford the cost of booth rental in government maintained markets or taxes, they frequently simply organize themselves into illegal markets that characterize the informal economy. In some extreme cases the ultimate result is that the government destroys

29 whole markets. Such was the case in Ghana with the destruction of Mokala Market (Robertson 1983). Finally, we should note that women’s entrance into some new occupational categories does not relieve them of their domestic responsibilities (Maula 1997). While men may have been both more willing and able to help financially with household responsibilities, during the period of relative prosperity^ immediately following independence, the pattern of women as primary provider persisted (Afonja 1986; Maula 1997). Thus, with independence the gender division of labor remained highly separate with women having limited occupational choices and opportunities to generate income. At the same time, women continue to bear the greater part of the burden of financially maintaining households (Gordon 1996a). 2.3.5 The Era o f Structural Adjustment With the introduction of structural adjustment policies (SAP) people, but especially women, again face an increasingly competitive environment. This results from decline in "formal" wage employment for both men and women in the public and private sectors. This, in turn, increased levels of unemployment and subsequent competition for alternative sources of income. Women’s persistent economic marginalization is extremely evident under these conditions of economic decline. Any slight gains made by women during the early independence period disappeared. Parents are even less likely to send girls to school. Likewise, women are among the first to experience retrenchment from the wage labor force. For example, in 1983 the unemployment rate among Nigerian women was 10.6%, almost twice that of men (Fashoyin 1989; Olukoshi and Olukoshi 1995). Consequently, most women engaged in trade, as do men, who also experienced retrenchment, salary freezes, and decline in “real” wages (Dennis 1991). Whereas in the past trade was the strict purview of women alone, under SAP men are reasserting themselves into this economic sector. This pattern is clear in that prior to SAP only 50% of the population was engaged in “informal” trading activities, whereas today the proportion has increased to at least 70% (Mustapha 1992). While sex appropriate

30 positions are redefined to suit men a similar process did not occur for women (Fashoyin 1989; Matanmi 1989; Clark 1994). Therefore, women faced increased competition &om a group that enters this sector of the economy with better access to assets, including financial, social and material capital (Aina 1991). They are, therefore, better equipped than women (Clark 1991; Dennis 1991). A concomitant dimension of these changes is a new gender differentiation within the ‘informal’ economy. Women are once again limited to trade in locally produced goods, principally food related items, household goods, and some types of fabric and clothing. Men, on the other hand, control more profitable commodities such as electronics and also engage in a wider array of artisanal activities including, carpentry, construction, auto and shoe repair. Women artisans primarily work as seamstresses and hairdressers (Osirim 1996). Competition among women in trade is also growing as they seek the most lucrative opportunities and increase their number of activities. An important source of competition is the increasing tendency for people with wage employment (both male and female) to have multiple sources of income (Mustapha 1992; Galli and Funk 1995; Zack- Williams 1995). Another factor increasing competition is reduced profitability of most activities due to rising prices and loss of clientele. With decline in ‘real wages’ people only buy what is necessary. As a result many traders are even losing money. These changes have reduced people’s flexibility to change their economic activities as they become increasingly dependent on credit (Galli and Funk 1995; Clark 1997). Under these circumstances, MLES is increasingly important for virtually everyone throughout Afiica today (Mustapha 1992; Galli and Funk 1995; Owusu 2000; Slater 2001). In rural areas women are also intensifying their use of MLES. In addition to production on their own-account and for husbands, women increased their trade activities. In some settings women also grew fiuit trees, fished and engaged in food production such as beer-brewing and home-made snacks (Palmer 1991; Galli and Funk 1995).

31 Despite diminishing economic opportunities, women's financial responsibilities within the household have not decreased. Rather they have intensified as women became increasingly independently responsible for meeting their own and children's livelihood needs. This is reflected in the growth of female-headed households in both urban and rural areas (Palmer 1991; Galli and Funk 1995). In Ghana the situation of men abdicating their financial responsibilities became so severe that government instituted laws to make men accountable (Robertson 1990). Furthermore, one critical aspect of the impact of SAP is that with inflation and currency devaluation the price of essential food items increased (Clark 1997). In the face of diminishing income generating opportunities and dramatically increasing cost of living some groups of women found it impossible to maintain their households in urban areas and began return migration to rural areas (Dennis 1991; Clark 1994). 2.3,6 Conclusion We can understand women's economic experience in Africa in terms of their limited occupational opportunities. When new opportunities arose for men in waged employment, women gained ground in terms of trade. Through this process they acquired access to such activities as long-distance trade, wholesaling and retailing imported goods. Women were only able to enter these sectors of trade however, because new opportunities existed for men. Thus, women’s economic activities remain marginal relative to men. This pattern persists today. With introduction of SAP men are losing ground economically in wage employment. Therefore, they are reasserting themselves into trade such that activities thought of as “women’s” are taken over by men. This is possible despite women’s long history in trade, because men have greater access to financial capital and social and economic networks that facilitate such activities. A disproportionate burden for family maintenance further exacerbates women’s situation. During the pre-colonial period, men were residents in the same areas as their families and more likely to bear some cost of household expenses. Today, with increasing segregation in the gender division of labor, however, even in areas where men and women co-reside, men are increasingly unlikely to help with family maintenance.

32 This pattern persists because, for many groups, women’s reproductive responsibilities are defined to include provision of household basic needs. This relates, at least in part, to the prevalence of polygyny. The result of these processes is that women have increasingly home the burden of family maintenance without assistance from male partners while simultaneously having limited economic opportunities. As a result of these two conflicting gender ideologies, MLES has always been an important means of meeting their financial responsibilities. While women have a long well-documented history of engaging in multiple economic activities, men are also increasingly turning to it. This trend, coupled with the declining profitability of women’s trade enterprises, explains increasing MLES usage. This reflects difficulties in earning sufficient income firom a single economic activity.

2.4 Differences Among African Women As noted above, there are clear differences in the economic strategies Africans configure. These vary along a number of dimensions; wage and self-employment; wholesaling and retailing; type of goods and services provided; localized and geographically extensive activities. Moreover, the relative profitability of any strategy configuration is also a function of the number and types of activities comprising it. For example, although wage employment does not necessarily pay well, provides a stable income, which may support other enterprises. Wholesaling is apparently more lucrative than retailing, as are goods people do not buy all the time such as fabric (MacGaffey 1988; Oldne 1993; Clark 1994). Likewise literature on Afiican women suggests that spatially extensive mobility is always more remunerative than localized trade relying on local wholesalers (Eades 1980; Okine 1993; Clark 1994; Falola 1995; Robertson 1997). Although this information is clearly discernable fi:om the literature on Afiican women’s lives, the ways in which women combine different economic activities to create livelihood strategies is not. Additionally, while these studies make it abundantly clear that women frequently have more than one economic activity, the reasons for and consequences of combining particular activities remains unexamined.

33 Research on Afiican women also points to difTerences among women along a variety of dimensions including ethnicity, class, life-cycle phase, residential arrangement and location (Clark 1994; Falola 1995). Moreover, these differences correspond to variation in the economic activities to which women have access. With respect to residential arrangement and location, for example, women in urban areas, living away fix>m their families may lack access to childcare, limiting the number and type, and geographic extent of enterprises with which they can configure a livelihood strategy (House-Midamba 1995; Osirim 1996). Additionally, it is evident firom numerous sources that West Afirican women, generally, and Yoruba women specifically, are extremely mobile in their economic activities, in contrast to other ethnic groups (Eades 1980; MacGaffey 1988). Among Yorubas, however, under some conditions, women’s mobility may vary over the life­ cycle. Similarly, stratification among Yoruba women is evident based on family background and wealth (Eades 1980; Dennis 1990; Falola 1995). These differences are apparent in the kinds of goods women sell as well as their degree of geographic mobility (Eades 1980; Spiro 1987; Dennis 1990; VerEecke 1995; Robertson 1997). Ethnic variation may also play a role in the types of goods women sell (Clark 1994; Musisi 1995). Finally, women are impacted differently by macro-economic changes such as SAP (Galli and Funk 1995). The ways in which these differences shape women’s choices vis-a-vis the configuration of a livelihood strategy and variation in outcome as it corresponds to particular types of strategies has yet to be explored. Although oûiec Dimensions o f Difference^ are occasionally investigated, most studies focus primarily on class or ethnic differences. In general, no more than one or two variations among women are explored in any given work. For example, in (Robertson and Berger 1986), essays focus on differences among women as a function of class. Likewise works that provide an overview of some aspect of women’s lives in Afiica tend to center on ethnic variation and occasionally life-cycle (Hay and Stichter 1995; Berger and White 1999). As a result, the

^ Important theoretical concepts are highlighted in italics.

34 interactions among these various dimensions of difference remain unexamined. One exception to this is (Clark 1994), who explains variation in commodity areas and trader hierarchies in terms of life-cycle, ethnicity, and class. In many studies, however, although differences among women are acknowledged at the outset, they are glossed over in favor of broad generalities (Zack-Williams 1995; Berger and White 1999).

2.5 Conclusions This chapter reviews literature on MLES in general and pertaining to African women specifically. Although work in a variety of settings points to the importance of multiple sources of income causal mechanisms and outcome variation is largely unexamined. With respect to Afiican women research points to a.number of issues relevant to the current study. First, it is evident that MLES use has a long history among African women owing to gender ideologies. Second, women use a wide array of different strategy configurations. Third, differences among women relative to ethnicity, class, and life-cycle shape the kinds of activities to which they have access. Nonetheless the ways in which the specific interactions among all these dimensions of difference shape the kinds of strategies women configure and outcomes thereof have yet to be explicitly examined. In part, this reflects the fact that gender differences are the dominant focus of most of these studies, which often subordinate other differences. It also reflects the fact that women’s use of MLES is essentially taken for granted. These are the silences within the literature on MLES in general, and Afiican women’s lives specifically this study seeks to address.

35 CHAPTERS

CONCEPTUAL FRAMEWORK: HOW CAN STRATEGY VARIATION BE EXPLAINED?

3.1 Introduction As discussed in Chapter I, the central issue of this study is explaining how women earn a living. Research conducted in diverse locales reveals that individuals and households deploy a wide array of livelihood strategies (Chapter 2). These include mono-activity forms and distinctive combinations of two or more economic activities (MLES). These studies imply that people use particular kinds of strategies for security, or accumulation, as well as survival. But reasons for choosing specific strategic configurations, and the outcomes thereof, require further study. Literature on Afiican women, the focus of this research, suggests that differentiating among women is important to understanding how and why they configure particular strategies (Chapter 2). Although this work points to several Dimensions o f Difference - Gender, Class, Ethnicity, Life-cycle Phas£ - shaping women’s opportunities in creating a livelihood strategy, it does not address interactions among these factors. The question remains, therefore, how can strategy variation be explained? I argue that interactions in Dimensions o fDifference manifest in geographically localized social institutions {Structures o f Constraint), which are (re)produced through people’s daily social and economic practices. These practices result in a variety of livelihood strategies reflecting life circumstances within differing geographically - localized institutional contexts. The results are differential income generating strategies

^ Important theoretical concepts are highlighted in italics.

36 as a function of the types of economic enterprise to which women have access relative to their resources and degree of mobility. Each of these concepts are defined below. To elaborate the argument, 1 draw on a conceptual fiamework synthesizing three bodies of literature: general feminist theory as well as that focusing specifically on economics and geography. These are examined fi-om an institutional perspective that seeks that elucidates the interaction between individual agency and social structures (Polanyi 1957; Weber 1968; Bourdieu 1977; Soja 1980; Thrift 1983; Giddens 1984; Pred 1984; Gregory and Urray 1985). My goal is to identify a set of concepts that will facilitate understanding differences among women in economic opportunities. Developing this framework proceeds in the following manner (Figure 3.1). First, lessons from feminist theory, specifically minority and developing country perspectives, provides a foundation for understanding how one’s opportunities differ as a function of interactions among social structures such as Ethnicity, Class, Gender, and Life-cycle Phase, known collectively as Dimensions o f Difference^. Second, feminist economist Nancy Folbre (1994), extending this firamework, shows how interactions among Dimensions o f Difference, manifest in social institutions including distribution o fAssets, Cultural Norms and Formal Rules (legal systems), which are the Structures o f Constraint shaping people's opportunities. Within sets of constraint, people express "purposeful choices" in pursuit of their interests. Third, feminist geography facilitates understanding the spatial situatedness of social institutions, which are geographically localized within communities and households. In other words, social and economic processes defined by interactions among Structures o f Constraint (social institutions) vary spatially both between and within social groups. Finally, feminist work also reveals that labor force experience is an excellent lens through which to understand how Dimensions o f Difference operate in everyday experience to shape life circumstances and, more importantly, economic

* Key concepts are highlighted in italics.

37 CONCEPTUAL FRAMEWORK

unensto erence

. ___

❖ Inferacfen^^TeTdimensio^__ co n stn iiâëâ^W S ^ u S ^^ em y . _w..^ A## ' <■

# % # 3 # .’.C ; '\ • *. v*j^!V -*.>; vT • .• r

A6elA^#^Ndirmi ules faPfe&wnces

^ Meràé^0B5’m < m v[m aistgm ^^ institutions SociàlmStitutions IhAgWmKeKwCoMslpaw&th^ - ^

Structurt^Ma mpiMuced'and^ang

# #

Figure 3.1 Conceptual Framework 38 opportunities. While work in gender and development has accounted for variation in economic opportuni^ based on gender differences, it is equally important to appreciate interactions among a wider array o( Dimensions o f Difference in these processes.

3.2 Dimensions o f Difference (Figure 3.1, Box I) Early feminist theory assumed all women faced similar conditions in their experiences of sexism (Friedan 1963). Even when writers acknowledged variation, for example as a function of Race and Class, some such as Fritz (1979) argued that "women’s suffering under sexist tyranny is a common bond among all women, transcending the particulars of the different forms of that tyranny." These researchers see gender as the most important distinction between people, whereas Marxists prioritize class. In both cases, however, a single broad social structure -- gender or class - is seen as sufficient to explain variation in experience among social groups; diversity within groups is not considered. More recently, post-structural feminists, especially working from minority and developing world perspectives, argue that interaction among broad social structures produces differences in individual experience. Not only are Class and Gender important, but also Race/Ethnicity, Nationality, Physical Ability, Sexual Preference, among others (hooks 1981; Dill 1983; hooks 1984; Jayawardena 1986; Bunch 1990; Collins 1990; Young 1990; Young and Dickerson 1994; Pettman 1996). A further element is often implied - Life-cycle Phase. This is frequently subsumed under Gender because social expectations over life are often related to gender roles. Amott and Matthaei (1996), for example, mention the importance of marital status in women’s employment (in the United States). The impact of marriage on life choices is, however, not simply a function of Gender. Both men and women face new responsibilities and opportunities when they marry. Rather, variation in Cultural Norms occur as people move through the life-cycle — childhood, adolescence, young adulthood, marriage, child-rearing, mid-life and old-age QCertzer 1989; Katz and Monk 1993; Laws 1997). That expectations of men and women differ in this respect reflects interaction

39 between Life-cycle Phase and Gender as two unique, but inter-related Dimensions of Difference. Variation in Cultural Norms associated with Life-cycle is also apparent in Racial/Ethnic and Class differences. According to feminist theorists writing from minority and developing country perspectives, interaction among Dimensions o f Difference produce a socially constructed over-arching system of domination such that different groups experience varying degrees of penalty and privilege. For example, while white women in the United States may feel discriminated relative to white men, based on Gender, they are likely to be privileged in comparison to black women as a result ofeffects (hooks 1981; hooks 1984). The relative importance of, and interactions among, broad social structures (Dimensions o f Difference) - Gender, Class, Race/Ethnicity, Nationality, Life-cycle Phase, Sexual Orientation, etc. - play out in specific ways. This stems from the fact that the meaning associated with Dimensions o f Difference varies cross-culturally and over time (Ferguson 1991). Interaction among social structures give rise to context specific social institutions, which are Structures o f Constraint shaping opportunity. Four Dimensions o f Difference, identified in the literature on African women, are of key importance to this work - Gender, Class, Ethnicity, and Life-cycle Phase.

3.3 Structures o f Constraint (Figure 3.1, Box II) Folbre (1994) offers a generalized framework for examining the historical- geographic nexus of social structures. She argues that social institutions are the material manifestation of interactions among Dimensions o fDifference called here. Structures o f Constraint. These are categorized broadly as distribution o fAssets, Formal Rules or legal systems. Cultural Norms and Personal Preferences. Assets include financial, material, human, and social capital that an individual brings to decision-making processes. Legal frameworks or Formal Rules structure business and personal interactions, hi some instances, however. Formal Rules may be systematically ignored, as in widespread acceptance of unregulated economic activities in developing countries.

40 When legitimized through social acceptance, informal rules represent the Cultural Norms of an area. Thus, norms are the implicit rules that govern behavior. The first three dimensions o( Structures o f Constraint (Assets, Formal Rules, Cultural Norms) are the social institutional fiamework within which individuals express Preferences, the fourth dimension. Personal Preferences are included as a Structure of Constraint because, while people make what Folbre (1994) calls "purposeful choices" about their lives, these are conditioned by socioeconomic context (institutional setting) within which one lives. Feminist theorists recognize that women’s self­ conceptualizations are socially constructed and therefore constrained, but at the same time, that women may step outside constraints to act in their own self-interest. Using "purposeful choice", instead of rational choice, dispenses with the strict dichotomous approach to structure and agency (Folbre 1994). Taken together these four aspects (Assets, Formal Rules, Cultural Norms, and Personal Preferences) constitute Structures o f Constraint that shape individual opportunity. People are bom into particular institutional settings. Through daily social and economic practice, including decision-making such as with whom to associate and economic strategy configuration, one or more aspects of a community’s social institutions are (re)produced and changed. Ability to effect change, however, is constrained by institutional context. Some groups have advantages over others, embedded in one or more Dimensions o fDifference (Gender, Class, Race/Ethnicity, Life-cycle Phase). Privilege in one area is associated with benefits in others. For example, those who have greater access to economic assets generally establish legal firameworks (and Cultural Norms). My argument to this point may be summarized in the following manner. One’s degree of economic opportunity is a function Structures o f Constraint — choices (Preferences) they make within the context of social institutional setting. Variation in social institutional contexts (Assets, Formal Rules, Cultural Norms, Personal Preferences) can be expressed along a continuum of social relations ranging fiiom those that are highly constrained (few choices) to those that offer greater opportunity (many

41 choices). Although, feminist theory suggests a spatiality in Structures o f Constraint, this is not elaborated. The question remains, therefore, ‘how do Structures o f Constraint vary geographically within households and communities?’

3.4 Localizing Structures of Constraint (Figure 3.1, Box HI) Feminist geographers have extended the argument that people are situated within multiple Dimensions o fDifference, Three inter-related themes emerge from this work. First, Structures o f Constraint (social institutions) are localized within communities such that they vary across space in households, neighborhoods, cities, regions, nations, etc. In other words, social and economic processes are inevitably spatial. Second, variation in social institutions are (re)produced through geographic as well as social and economic processes. Third, individuals are differentially impacted by spatial and socioeconomic processes (WGSG 1984; Rose 1993b; Kobayashi 1994; Massey 1994; Pratt and Hanson 1994; Hanson and Pratt 1995; McDowell 1997; WGSG 1997). Work by feminist geographers indicates that Dimensions o fDifference characterizing individuals and groups are localized within households and communities. As Pratt and Hanson (1994:8) note, "systems of difference are constructed in place and different systems are constructed in different places." Their study of women and work in Worcester reveals there is a geography to communities within the city. These are shaped by differences in Class, Nationality, Race, Life-cycle Phase and other Dimensions o f Difference. Diversity in community identity plays out not only within cities by neighborhood, but also within households and firms. They operate at other spatial scales, including the regional and national (Massey 1994). Arguably, communities based on ethnic and religious affiliation operate across a variety of scales, but may still be inscribed in particular places. An example of this are Nigerian home-town associations, in which members live in diverse locales, but the cotmmmity is nonetheless situated in a particular place, the home-town (Trager 1998).

42 The second point follows from this. Geographies of difference are (re)produced through spatial processes. Again, Pratt and Hanson (1994,1995) provide an excellent example. With respect to the spatial division of labor, firm location and hiring decisions reinforce patterns of segregation among groups within cities and firms. Where firms locate, in part, reflects their understanding of labor force differences within particular areas. Their awareness results in differential employment opportunity in specific locations. In an Afiican context. Porter (1995), explains how road construction in developing countries, often a development priority, changes the geography of rural markets such that village markets may die as new roadside ones emerge. This spatial process, in turn, changes income generating opportunities. Finally, spatial expression of social and economic processes differentially impacts people according to interactions among Dimensions o fDifference. Variation is most apparent in mobility patterns and spatial mis-match. Here, localization o f Structures o f Constraint within cities in the form of residential segregation and availability of transportation may limit mobili^ and, hence, access to employment. Likewise, gender ideology around women’s reproductive responsibilities are reflected in interactions among social institutions. As a result, women experience constrained mobility (Blackley 1990; McLafferty and Preston 1996). Further, drawing on time-space geography, others reveal that even where women are mobile, they may experience more constrained and fixed activity spaces then men, reflecting gender ideologies that render women primary care-takers within households ^ratt and Hanson 1991; Kwan 1999a; Kwan 1999b). Another aspect of work on women’s activity spaces is residential arrangement. The presence of other adults in the household may mitigate domestic responsibilities, enabling women to experience less fixity in activity time-spaces ^w an 1999b; Schroeder 2000). This has particular salience in Afiican settings where extended-family residential arrangements are common. It should not be assumed, however, that the mere presence of other adults will necessarily relieve women of care-taking responsibilities. Women’s mobility will still be shaped by interactions among Structures o f Constraint, reflecting gender ideologies.

43 As Laws (1997) points out, patterns in mobility may also change over the life- course in response to institutions; in this case, for example, the state. Her study of women’s mobility over the life*course demonstrates that young, unmarried women with children, dependent on welfare, are likely to have severely curtailed mobility due to a lack of money. Even if childcare facilities are provided to reduce this responsibility, lack of financial means may still limit activity spaces. Similarly, widowed-women retirees may also be spatially constrained by lack of social security benefits and/or lack of transportation. Although both instances limit women’s mobility, the reasons for this vary, reflecting unique intersections of Gender and Life-cycle Phase. Likewise, in the example of Nigerian road construction, all people were not equally impacted. Men tended to be more mobile, partly as a function of gender ideologies that affords them greater mobility, and partly as a function of greater access to resources. Likewise, some women did not experience a decline in economic opportunity because mobility was not restricted (they lived close enough to have regular access to the new road and market) or the whole village moved. But where women’s mobility was restricted, and they lived in villages removed firom the new road, possibilities for income generation declined. These differences in access reflect intersections in Dimensions o f Difference, especially Ethnicity and Gender, which manifest in variation in Structures o f Constraint (Porter 1995). The point is that the meaning o£Dimensions o fDifference are socially, economically, historically, politically, and geographically situated. These intersections manifest through geographically localized social institutions, called here Structures o f Constraint. In other words. Dimension ofDifference and Structures o f Constraint are both place and time specific. The myriad ways in which social institutions play out impact people differentially, specifically in terms of their mobility. The relative importance of mobility, in turn, reflects geographically-localized social institutions.

44 3.5 Localized Structures o f Constraint and Labor Force Experience: An Example The framework articulated ed above is general, applicable to many contexts. Below is an example of its application to American women. This illustration demonstrates that labor force experience is an excellent lens through which to examine differences among women. Labor force experience reflects how localized social institutions {Structures o f Constraint) differentially impact people. The main goal of the early-1970s feminist movement in the United States was an Equal Rights Amendment. This was grounded in white middle-class experiences as housewives who were denied through Cultural Norms, opportunity to work outside the home, defined as more fulfilling. But this view was inconsistent with the experience of many black women who often worked outside the home - as a necessity, rarely as a choice (hooks 1981; hooks 1984; Collins 1990; Ammott and Matthaei 1996). In fact,

[t]he bourgeois individualistic theme present in the contemporary women’s movement led many Black women to express the belief that the movement existed merely to satisfy needs for personal self-ftilfillment on the part of white middle-class women (Dill 1983).

The different experience of blacks is a function of social institutions. Class differences that result in disparate access to Assets are key to understanding variation in Cultural Norms across racial groups. The most disenfi-anchised groups in society (those with least choice) are assigned the lowest paying, least secure, and most difficult work. This pattern reflects labor market segmentation, which is one means by which Dimensions o f Difference are maintained and (re)produced, both within and between groups (Ammott and Matthaei 1996). Equally important is the gender division of labor in articulating Cultural Norms governing men’s and women’s reproductive responsibilities. Individuals belong to many groups. As a result, even when people share a biological heritage such as race, they will not necessarily experience, or respond to, social structures m the same manner. They choose with whom to associate and these choices may be shaped by a shared biological heritage such as race, or non-biological dimensions

45 o f difference such as class, or a combination of both race and class. Likewise, people may change afhliation over time (Collins 1990; Young 1990; Folbre 1994). Folbre (1994) articulates this idea in terms of an individuals’ "conflicting allegiances" among various groups in which they have, or choose, association. The issue of group membership is critical because it is used by theorists to indicate the degree to which one experiences discrimination. Someone who encounters penalty along various Dimensions o f Difference will have limited choices, not only about the structure of their lives, but also about with whom to associate (hooks 1981). Nonetheless, even when highly constrained, people behave in "purposeful" ways to optimize well-being as they perceive it (Folbre 1994). The idea that people make choices about whom and under what conditions to affiliate is important because through choices they act to resist and support social institutions (hooks 1981; hooks 1984; Collins 1990; Young 1990; Ferguson 1991; Folbre 1994).

3.6 Gender and Development: Applying Localized Social Institutions Work cited above is concerned with social and spatial construction of difference so as to better understand commonly accepted social categories (Gender, Class, Race/Ethnicity, etc.). What remains to be done, however, is considering how Dimensions o fDifference among individuals, manifest in localized Structures o f Constraint to provide differential opportunities and hence, strategy configuration. This is especially relevant for work on gender and development. The main emphasis of research on gender and development has been gender differences both at the macro- and micro- economic level (Parpart 1989; Elson 1990; Afshar 1991; Massiah 1993). Through this work, we know that roles of women within their society has been largely ignored by development theorists and practitioners (Boserup 1970; Parpart 1989; Elson 1990; Afshar 1991 ; Massiah 1993). Likewise, we know that women have been differentially impacted by development policies, especially structural adjustment programs (SAP) (Comia, Jolly et al. 1987; Vickers 1991; Afshar and Dennis 1992). As noted in Chapter 2, some empirical studies on the impact of socioeconomic change in developing regions, Afiica in

46 particular, acknowledge uneven impacts among women. These studies, however, tend to emphasize only one or two dimensions of difference, e^ecially class. The result is dichotomous conceptualizations of women’s economic activities; mamabenz (wealthy wholesalers)/petty traders (small-scale retailers); formal (regulated)/informal (unregulated); and survival-oriented/growth-oriented. To appreciate causes and consequences of the multiplicity of economic strategies people employ, it is necessary to move beyond dichotomies. It is essential to appreciate how numerous dimensions of difference manifest in localized social institutions {Structures o f Constraint) shape people’s opportunities.

3.7 Conclusions The focus of this chapter is development of a theoretical framework for examining differences among women. This is achieved by synthesizing three areas of feminist theory. Each body of work offers a particular set of concepts which, taken together, provide a mechanism for explaining differential economic opportunities. Dimensions o f Difference {Gender, Class, Ethnicity, Life-cycle Phase) are the broad social structures, which characterize individual identity. Interactions among these lead to social institutions, called here Structures o f Constraint {Assets, Formal Rules, Cultural Norms, Personal Preferences). These are localized within communities that operate across space - including households, neighborhoods, cities, regions, and nations. Localization of social institutions shapes the range of opportunities people have in economic activity and, hence, livelihood strategy configuration. As Gordon (1996b:27) points out, while it is important to recognize differences among women, "we should not lose sight of the fact there are widely repeated patterns and enough historical and cross-cultural continuity to make generalization and causal theorizing possible." To explain variation among women in the economic opportunity, through the lens of their livelihood strategies, this study will mobilize concepts discussed above in a disaggregation among women. While process is clearly important, to understanding differences among individuals, it is first necessary to appreciate patterns in

47 those differences. The goal of this study, therefore, is to understand patterns of difference among women.

48 CHAPTER 4

RESEARCH CONTEXT AND DATA COLLECTION: HOW CAN THE FRAMEWORK BE APPLIED TO BENIN?

4.1 Introduction Having now provided empirical and conceptual frameworks for this research (Chapters 2 and 3), attention turns to the case study, women in Porto Novo, Benin. Of particular interest is application of the conceptual framework to the study context. Accordingly, this chapter begins with an introduction to Benin, specifically an historical overview of it’s political-economy and women’s place within it. Brief discussions of Porto Novo’s history are interwoven with this. Section two covers data collection procedures including operationalizaton of the conceptual firew ork. Section three provides a general overview of the sample with reference to its representativeness both within Benin and the Oueme, the province in which Porto Novo is located. The geography of Porto Novo’s social groups relative to ethnic and residential arrangement patterns is also addressed. Finally, a brief overview of data analysis is presented in section four.

4.2 The Political-Economy of Benin: An Historical Overview Benin is a small West Afiican, Francaphone country situated between Nigeria and Togo (Figure 4.1). It extends 700 km northwards from the Gulf of Guinea toward Niger and Burkina Faso, it’s neighbors to north. While only 120 km wide at the coast, it broadens to ISO km in the south and 350 km in the north. It’s geography has had

49 Figure 4.1 Study Area: Republic of Benin

50 important implications for its economic development. Specifically, most of southern Benin is proximate to both Nigeria and Togo, which has facilitated its development as an entrepot state within the sub-region. Despite an inadequate transportation system, except along the coast, cross-border trade, in which women play a central role, is the primary vehicle for this economic activity (Allen 1989; Igue and Soule 1992). 4.2.1 The Pre-Colonial Period Three empires dominated this area of West Afiica prior to colonization. Danhomé, was founded in the early 16* century by the Adja people (Fon and Goun). These were migrants fiom present day Togo, into which the kingdom, located in the southwest and along the coast, extend. Three cities, still important in Benin, were established by the Adja including Abomey, the capital of Danhomé, Allada in the southwest and Porto Novo in the southeast (Akindele and Aguessy 1953). The latter two were seats of semi-independent kingdoms within the Danhoméan empire. Likewise, the Yoruba Empire of Oyo was comprised of several kingdoms including Ketou and Sabe, still Yoruba dominated cities Benin. This empire, located somewhat inland in the southeast, ranged into present day Nigeria. The north was controlled by the Bariba Kingdom, which was significantly smaller than either of the others (Ronen 1975; Quenum 1983; Allen 1989). All three empires were actively involved in the Atlantic slave trade, especially Danhomé and Oyo, and experienced external infiuence through these trade relations. The Bariba were also engaged in trans-Saharan trade with Arabs. The earliest European contact came with the Portugese who built a trading post in Porto Novo in the 17* century, naming it for a town in Portugal. In the southwest Brazilians settled in the Guidah in the early 18* century. In both instances, missionaries followed these groups, but had little success converting the population. Islamic jihad by Moroccans in the north had a considerably greater effect (Ronen 1975; Quenum 1983; Allen 1989). As noted in Chapter 2, West Afiican women were actively involved in international trade relations. Indeed, Yoruba women were especially known for their commercial abilities. One important difference between Yoruba and Adja women.

51 however, is that while the former engaged in long-distance trade, the latter did not (Fadipe 1970; Awe 1977; Eades 1980; Afonja 1986; Guyer 1994; Falola 1995). In the north, as well, women only conducted local trade, particularly once Islam became prevalent. Another note of interest regarding the role of women, is that there was an elite cadre of women soldiers, in the Danhomean empire, which became known as Amazons by Europeans (Ronen 1975; Quenum 1983; Allen 1989; Maula 1997). 4.2.2 The Colonial Period Apart firom the Brazilian and Portuguese traders, these early empires had little European contact until the 19* century. Having established control over the coast in 1882, the French conquered the entire area, known today as Benin, inl890. In 1900 Porto Novo was made the capital of the new French dependency of Dahomey. The French felt themselves to be liberators, having exiled the last "tyrannical" Danhomean King, Benhazin. Early writings by colonial administrators indicate that they had no intention of changing fimdamental political structures, instead imposing indirect rule through local chiefs (Ronen 1975; Allen 1989). Beginning in 1899, however, the nature of colonial administration changed. Through the imposition of taxes, which had to be paid in French money, as opposed to cowry shells or livestock, colonial administrators forced the population to engage in wage labor and export crop production. Through this process the French introduced plantation agriculture and new crops to Benin, especially palm oil, cotton, maize and peanut production (Ronen 1975; Allen 1989). In addition, they disrupted indigenous political systems by making chiefs responsible for prompt and complete tax collection, changing chiefs who did not comply. Thus, began direct rule in French West Africa. Another administrative change deepened class and ethnic divisions. Through education the French sought to create an Afiican elite with the idea that each ethnic group should be governed by their own members. Toward this end colonial administrators created political sub-divisions, called cantons, that delineated boundaries between ethnic groups. Again, their stated objective was to free each ethnic group from the subjugation by any other (Ronen 1975).

52 To facilitate economic development, enhance their control over and "civilize" the population, the French established public education in 1904. These were in addition to missionary schools instituted earlier. Despite the fact that education was free, access and attendance were uneven. More schools were constructed in urban areas, such as Cotonou, Porto Novo, and Abomey, resulting in greater access for those who lived in or near these towns. Additionally, parents in rural areas were less inclined to send their children to school because their help was needed in farming. Given that the vast majority of Benin’s population lived in rural areas, relatively few people actually obtained education. As with earlier education initiatives, this had the effect of deepening class differences within the population (Ronen 1975). Moreover, as discussed in Chapter 2, gender differences in type of education were also evident. Girls were less likely to receive education and to the extent they did, it emphasized instruction in French idealized women’s roles as wives and mothers. Women’s professional training included gender specific occupations such as teaching and nursing (Ronen 1975). This contributed to widening class differences among women. This division was also enhanced by changes in trade. New commodity areas opened up for women, especially imported goods, as men became increasingly involved in wage labor. All women, did not have equal access to these opportunities, however, as they required large amounts of capital and credit accounts with import firms. Yoruba women were particularly poised to take advantage as a function of their long involvement in lucrative long-distance trade (Ronen 1975; Afonja 1986). Although initial impetus for French colonization included empire building and accessing resources, ultimately, they impacted every aspect of life. The result was a deepening of gender, ethnic and class divisions within the indigenous population. This aspect of colonial administration is critical to understanding current differences among the Béninois.

53 4.2.3 Early Post-Colonial Period 1960-1988 Although France began to give colonial subjects greater autonomy starting in the mid-1950s, this was an attempt to maintain control over the territory. Ultimately, however, it led to independence for all of French West ASrica in 1960, most of which, except Guinea, became a commonwealth of sorts, known as the French Community. France still maintained strong ties to its former colonies, however. This is most evident in the fact that the Community’s currency. Communauté Financière Africaine franc (CFAF), was tied to the French franc at 50 CFAF = 1 franc (Ronen 1975; Allen 1989). Porto Novo was retained as the capital of the newly independent country, despite the prominence of Cotonou as the country’s economic capital. Some industrialization, primarily focused on food processing, occurred in the period immediately following independence. Economic stagnation and clientelism on the part of elected officials, however, led to a series of military interventions and subsequent rapid turnover in the government, at two to three year intervals in 1963,1965,1967,1969 and 1972. With the 1972 coup, Mathieu Kerekou came to power establishing an ostensibly Marxist regime, which remained fairly stable until the late 1980s when economic decline again forced change (Ronen 1975; Allen 1989). Kerekou’s regime emphasized diversification in trading partner’s and reduction in reliance on France for budget subsidies. Although industrialization was state led and controlled through nationalization of existing firms and the creation of partially private companies (parastatals), most of the agricultural and commercial sectors were never brought under state control (Allen 1989; Maula 1997). As a result women’s economic activities were not impacted. One seemingly important legal development, for women, during this period was the passage of Aritclel24 in 1972. This law gave women equal rights, especially in employment (Tomieri 1997) (and fieldnotes). Interestingly, when asked about this law, most interviewees either did not know of it or did not feel any impact from it. frideed, one interviewee, who is employed by the government as an extension worker, commented that "it is the domain of politics...because one cannot bring Article 124 into the home"

54 (interview with Kadija, October 1998 - see Chapter 7 for a complete list of interviewees). Her comment suggests that although the law may provide women with equal opportunity in employment, gender ideology within Béninois households does not. A more cynical view was expressed by Christian, an older trader, who suggest that the only reason the Kerekou government passed the law was so that they could conscript women into to the military should the need arise (interview with Christian, September 1998). Her comment reflects the timing of the legislation, which was immediately after a failed mercenary invasion sponsored by the French government (Allen 1989). The lack of impact from Article 124 is apparent in that a fairly rigid gender division of labor has persisted. Although women occupy some managerial positions within the government, as in Kadija’s case, most who have wage employment work in gender specific jobs. In fact, in 1979 only 30.1% of professionals, 23.5% of administrators, and 24.9% of those in the service sector were women. In all likelihood the employment categories occupied by women in these sectors were gender specific including secretary, teacher, nurse. This is in contrast to the commercial sector which is almost entirely comprised of women (93%) (INSAE 1992). Unlike other parts of West Afiica (e.g. Ghana), most, if not all. Béninois governments have had a fairly relaxed attitude toward this sector (Maula 1997). 4.2.4 The Era o f Structural Adjustment 1988-Present Despite some economic growth through the 1970s and 1980s, Benin’s economy, like many others in the developing world, began to decline in the late 1980s due to impacts of the world economic recession. Ultimately unable to meet debt payments, the government was forced to adopt a structural adjustment program (SAP) initiated by the World Bank and hitemational Monetary Fund (IMF) in 1988. This led to a national crisis due to growing opposition to Kerekou’s regime, in part because of the government’s inability to pay civil servants’ salaries. In 1990 a transitional Prime Minister and government were appointed with the charge of writing a new constitution and organizing fiee elections. A draft constitution was approved by popular referendum in 1990, and the

55 transition to a multi-party system was complete in April, 1991 with the election of Christopher Sogolo as President (INSAE). The successful transition to a democracy, notwithstanding, Benin has continued to experience economic difficulties under SAP. As in other African countries SAP in Benin emphasized economic liberalization. Included in the policies initiated in 1989 were privatization of parastatal industries, reorganization of the banking system, and downsizing of the government bureaucracy. Devaluation of the currency, from 50 to 100 CFAF to the French franc, was added in 1994. Despite these measures economic conditions in Benin have not improved. Apart from the litany of ills identified by the external lenders (e.g. The World Bank), including poor balance of payments, indebtedness, corruption, waste, bloated government sector, insufficient and inadvisable investment, Benin is hindered by lack of resources. It's principle exports are cotton and palm oil, neither of which are very profitable (INSAE). According to the World Bank these problems have been exacerbated by the population's response to SAP. Specifically, growth of the informal sector has resulted from downsizing the public sector and decline in "real" wages. In addition, those finishing their education are unlikely to find wage employment as private sector activities have not developed. The result is a high level of official unemployment. Interestingly, while some like Igue (1992) and Allen (1989) have commented positively on Benin’s role as an entrepôt state within the region, those promoting SAP see it negatively. This reflects the fact that much of this trade occurs informally and therefore does not benefit the government (INSAE). Porto Novo plays a pivotal role in this informal cross-border trade owing to its location on the Nigeria border (Figure 4.1). Although originally settled by the Yoruba from Nigeria, it became urbanized under the Goun (a subgroup of the Adja) in the 17* century. As of the 1992 census (the most recent) Porto Novo was a city of 179,138 people, 48% of whom are male, 52% female. These two ethnic groups still dominate the city: 67.8% Fon (of which 45.5% are Goun) and 19.8% Yoruba ONSAE 1992,1996).

56 For women, who have always dominated the commercial sector, the result is increasing competition Sroai the newly self^mployed and those using trade activities to augment their salaries (INSAE; Igue 1995). Women comprise the vast majority of new entrants into the commercial sector as they have been especially hard hit by declining public sector employment. Although Benin is still largely a rural country as is evident in that fishing and farming are the dominant occupations for both men (67%) and women (42%), to the extent that professional work is available, it is dominated by men. Only 1% of all men work as professionals, but less than one percent of women are represented in this occupational category. Men, however, represent 75% of all scientific occupations, 80% of other professionals and 67% of government workers. Men also dominate as artisans (74%). After fishing and farming the most prevalent form of work for women is within the commercial sector, in which 42% of all women work, representing 92% of the labor force in this category (INSAE 1992). As noted above the gender division of labor has also not changed substantially. Within the wage sector women work primarily as teachers and nurses, as artisans they are principally seamstresses and hairdressers and in the commercial sector their commodity areas consist essentially of food and household items (field notes). One further source of difficulty is the rising inflation, especially since the CFAF was devalued. Virtually everyone commented on the near doubling of prices since 1994 (field notes). While these conditions are not remarkable in an Afiican context, they are ubiquitous, and present considerable problems for women attempting to configure profitable economic strategies. Given these conditions, a high level of financial responsibility within the household, increasing competition, and limited employment opportunities, women work many long hours, 18-20 hours a day, including both their reproductive and productive activities. Women also work virtually their whole lives, 60% of women in the sample started working by age 20 (23% by age 15) and a considerable number were still working into their 80s.

57 4.3 Data Collection Procedures Having now provided a bit of contextual background attention turns to data collection procedures. As noted above, women in Porto Novo are the foci of this study. Although, as indicated by the conceptual framework (Chuter 3), it is important to consider women’s "purposeful" choices within socioeconomic context, especially their household, one cannot assume households are cohesive decision-making units. This is particularly important in an African context generally, and Benin specifically (Maula 1997). Two reasons for this are the instability of marriage and unequal benefits from pooling resources (Guyer and Peters 1987b; Fapohunda 1988; Geisler 1993; Horn 1995; Doss 1996; Maula 1997). Accordingly, women as individuals were the unit of analysis, although detailed information regarding their household composition was also obtained’. Respondents included women actively engaged in income generating enterprises. Thus, women working as unpaid family labor were not included. Data collection took place over fourteen months from September 1997 to November 1998. To address the issue of place-to-place variation in Structures o f Constraint (localized social institutions), data were collected throughout Porto Novo. Six areas within the city were chosen - two in the city center, two in the first tier of urban development or suburbs, and two on the urban firinge (Figure 4.2). Within each of these areas, one location was near a major market and the other was removed from it. The three market areas are - the Grande Marche in the city center, Houinme in the suburbs, and Ouando Market on the urban fringe. The largest of these three is Ouando, not the Grande March as the name might indicate. This reflects a shift in city development away from downtown. In addition, Ouando is ideally situated in proximity both to the city, its rural hinterland, and the Nigerian border.

’ Household was defined as all persons living under the same roof, who ate together.

58 CIRCONSCRIPTION DE DE PORTO-NOVO i

L.\

W##»# ^ , , , *»w## ###»##*#* , . ,

V*-

Figure 4.2 Détail of Study Site: The City of Porto Novo 4.3.1 The In-Depth Survey Two types of data were collected - an in-depth survey and life-history interviews. Within each of the six city areas, a snowball sampling method was used to contact survey respondents (French and English versions are provided in Appendices A and B, respectively). Initial contact was made with neighborhood chiefs, who then selected someone to introduce me and my two research assistants to women in the neighborhood. Appointments were then made to return to do the survey, so as not to disrupt their work any more than necessary. Women were also asked at the initial meeting if they knew anyone else who would be willing to participate. Despite using a snowball sampling method, every attempt was made to ensure that the overall sample was representative in terms of age, occupation and ethnicity. In total 535 surveys were collected, of which 522 were useable for analysis. Data gathered included demographic information, training/education history, details on current economic activities, abbreviated economic histories, marital status, family background, household configuration and responsibilities, financial responsibilities, revenue, expenses, savings, and associational life (for a complete list of variables see Appendix C). Each area of inquiry represents one or more aspects of the Structures o f Constraint. The following is a discussion of the variables (identified in bold) associated with each of the four Structures o f Constraint and the ways in which they are important to strategy configuration choices. 4.3.L I Assets As defined in Chapter 3, Assets are all the financial, material, human, and social capital upon which individuals can draw in the creation of the livelihood strategy. Thus, critical variables included sources^” and amounts of starting capital, employees — both paid and unpaid (family), education and training. Financial and material Assets are particularly important because they facilitate access to merchandise, necessary training/education and equipment (e.g. sewing machines, hair dryers, etc.). Family background (mother’s and father’s occupation) may, in part, determine this, and is

Variable names are highlighted in bold.

60 reflected in sources of starting capital and education/training. Access to all forms of capital not only reflects Cultural Norms dnd Assets" distribution within and between groups, but also an individual’s relationship with the state {Formal Rules). This occurs because the state may be a source of assistance through training programs and loans. Employees may increase the scale of an enterprise as in the case of artisans with apprentices who can draw on their labor to serve a greater number of clients. Apprentices are also a source of income because they pay for their training. Alternatively, additional labor may multiply the number of enterprises in which an individual can engage. Education and training tend to enhance economic opportuni^ by providing access to wage labor and artisanal activities requiring specialized skills such as dressmaking and hairdressing. Training may also take place informally such as learning how to trade by helping one’s mother. 4.3.1.2 Formal Rules Formal Rules play a significant role in who can be an artisan because a special certificate earned through apprenticeship is required to open a workshop and hire apprentices. For access to wage labor one’s relationship to the state may also be significant because it is the principle employer in this context. Again, formal rule structures determine who has access. Indeed, in many instances it is necessary to pass an exam before being eligible for a job with the government. Each job type has its own exam. Moreover, some education is also usually required. Although no specific variables were used to capture this dimension of the Structures o f Constraint, it is intrinsic in the type of enterprises in which women engaged. 4.3.1.3 Cultural Norms Cultural Norms impact and reflect a number of issues in women’s life circumstances that, in turn, shape their opportunities for income generation including: expectations, their level of responsibility (both care-taking and financial), the extent to which they can call on others to help, and their degree of mobility. Critical variable here

" Key theoretical concepts are highlighted in italics.

61 include: marital status, residential arrangement, number and age of children in the household, childcare (dependency burden), and financial responsihllities. Although most women over the age of 19 claim to be married, frequently this is not an official marriage. There are three types of formal marriage in Benin: customary (associated with particular ethnic rituals), civil, and religious. Where the marriage is official and Christian, the expectation is that it is monogamous and that the couple co- reside together. In all other cases it may be polygamous, which limits the expectation of co-residence. Marriage for women in Benin confers a certain social status such that even if the husband is not in the household, the ability to claim marital status of any kind backed up by evidence of children is critically important to women’s social standing. In addition, marriage reflects fulfillment of expected Cultural Norms based on life-cycle phase. An assumption often associated with Africa is that people live in extended family settings. While this is often the case, it is not universal. In fact, in Benin an important frequently expressed life objective is the ability to live a home of one’s own. If it is financially feasible people will move to rented quarters in order to live outside the extended family setting until they can buy land and build their own house. The perceived benefit of this is that it is possible to hide one’s financial resources, thereby reducing demands for assistance from family members. This facilitates savings for investment in land, house construction, and/or economic activities. The choice to live in a nuclear family setting, however, may also limit opportunity if there are young children in the household because it results in a lack of assistance in care-taking. On the other hand, living in an extended family setting is not a guarantee of child-care assistance. To some extent the availability of help with care-taking responsihllities depends on whether a woman is living with her own family or her husband's. Thus, residential arrangement, which reflects Cultural Norms as well as access to Assets, may either constrain or enhance opportunity depending on the specific circumstances.

62 The degree to which extended family provides assistance with care-taking responsibilities, and/or demands financial and other forms of aid, influences the dependency burden experienced; care-taking, specifically of young children and elderly relatives, and financial support of family members. Although reproductive responsibilities such as care-taking are generally only applicable to those in the household, financial support can extend to family members outside the household. If one is perceived to be relatively well-off by the extended family, for example if engaged in wage employment, they may then foster their young children to the household. Depending on the age of foster children and number of children in the household already, this may diminish financial resources and increase care-taking responsibilities. These patterns reflect Cultural Norms around obligation and reciprocity in the extended family. Dependency burden, particularly related to care-taking, may seriously impinge on mobility if there are a number of young children in the household and a lack of assistance from other family. Here too, age range and number of children in household make a difference. Other factors also may affect mobility, such as Cultural Norms among some groups that restrict women’s activities to the family compound or neighborhood until she has had two or three children. Lack o f Assets will also limit mobility to walkable distances. 4.3.1.4 Personal Preferences Women’s preferences in strategy configuration were obtained through open-ended questions asking why they engaged in, and combined, particular activities, the relative importance of each activity and impressions regarding the success of their strategies. Responses were then categorized into variables representing each question. Answers reflect the "purposeful" choices women make within the context of their life- circumstances. Each set of variables discussed above represents the principle mechanisms impacting range of choice in economic strategy configuration through which Folbre’s generalized institutional framework (Assets, Cultural Norms, Formal Rules and Personal

63 Preferences) plays out in the Benin context. It should be noted that many are inter­ related, and all are contingent. Thus, how they impact individual’s lives is highly variable, not only across groups but also within them. This reflects the intersections among various Dimensions o f Difference (Gender, Class, Ethnicity, Place-in-the-life- cycle). 4.3.2 Life and Economic History Interviews In addition to detailed surveys, data include 15 in-depth interviews describing women’s life and economic histories. Although somewhat more difScult, again an attempt was made to ensure that participants represented a variety of occupations, ages, and ethnicities. The purpose here was to understand processes behind decision-making vis-a-vis strategy configuration. Of particular interest was the ways in which life- circumstances shaped women’s opportunities. Additionally, given the historical nature of interviews, it was possible to explore how strategy configuration changed over the life- course.

4.4 Sample Representativeness As noted above every attempt was made to ensure sample representativeness in terms of age, ethnicity, education, and occupation. Comparing sample data with those from the 1992 census, this section examines the extent to which that goal was achieved. Similar to other developing countries, age structure in Benin tends to be fairly young. Given that the focus of this study is working women, and that most people do not initiate independent economic activities until age fifteen, discussion will begin with this age cohort, considering only women over age 15. As can be seen the Table 4.1, very young and older women were slightly under represented in the sample relative to the population of women over age 15 in the Oueme and Benin. On the other hand, middle age women were somewhat over sampled. In general, however the sample is fairly representative in terms of age of women in Benin.

64 AGE COHORT BENIN OUEME SAMPLE

>20 16.0% 15.0% 3.3%

21-30 32.0% 31.0% 30.0%

31-40 22.0% 22.0% 33.0%

41-50 12.0% 12.0% 19.7%

51-60 7.0% 8.0% 7.5%

61-70 5.0% 6.0% 4.6%

71+ 5.3% 52% 13%

Total 99.3% 99.2% 99.4%

Table 4.1 Age Cohorts of Women in Benin, Oueme, and Sample

In ethnicity, greater variation was observed between the sample and the populations of Benin and the Oueme. Nonetheless, it is still reasonably representative. Table 4.2 indicates that the sample contained a somewhat higher percentage of Adja, especially Goun, and a slightly lower proportion of Yoruba. Statistics for the Mina were not provided in the 1992 Béninois census. Unfortunately, the 1992 census data did not include a simple break down in level of education for the Oueme. Relative to the population of Béninois women over IS in general, however (Table 4.3), the sample is somewhat better educated. This probably reflects the fact that Porto Novo is an urban milieu, in which one would expect to find a somewhat higher level of education. Data on occupation in the 1992 Béninois census were grouped into rather large categories. Given that no break down was provided for specific jobs contained within each category, it is difficult to make a comparison with sample data. Nonetheless, as noted earlier, the vast majority (92%) of women in Benin are engaged in commercial activities. This is fairly consistent with the proportion of

65 ETHNICITY BENIN OUEME SAMPLE

Yoraba 6.2% 24.4% 19.9%

Goun n/a 45.5% 63.0%

Fon 30.1% 22.3% 9.2%

Mina n/a n/a 5.0%

-Other" 63.4% 7.8% 29%

Total 100.0% 100.0% 100.0%

Table 4.2 Distribution of Ethnie Groups in Benin, Oueme, and Sample

LEVEL OF EDUCATIONBENINOUEME SAMPLE

None 82.0% 83.0% 51.3%

Primary 10.0% n/a 26.1%

Secondary 7.7% n/a 20.7%

Terminal 0.3% n/a 10%

Total 100.0% 99.1%

Table Level of Education among women over 15 in Benin, Oueme, and Sample

The high proportion of other ethnic groups in the Benin column reflects the geography of ethnici^ in Âe country. The Oueme is dominated by Goun and Yoruba. Other groups are more prevalent in different parts of the country.

6 6 women in the sample, 82.5%. While only one percent of all women were categorized as having professional or service employment, a somewhat higher proportion of women with these occupations were apparent in the sample (professional=6.7%, service=2.3%). As with education, this undoubtedly reflects the fact that Porto Novo is a city, in which one would expect to And more wage employment.

4.5 The Geography of Social Groups One additional general feature of the sample bears discussion, the geography of social groups. As discussed in chapter 3, feminist geographers have shown that social and economic processes differ spatially across communities within cities (Hanson and Pratt 1995). This variation is articulated here in terms of patterns of residential arrangement and the distribution of ethnic groups across the city. As expected, a clear geography emerged when looking at the spatial patterns in these two variables by area of the city. Table 4.4 shows that there is a relationship between ethnic group and residential pattern. Different patterns in residential arrangement are observable for the various ethnic groups in Porto Novo (Table 4.4). Yoruba women are much more likely to be living without their husbands (57.7%), than either Fon or Goun women. A high proportion of Fon women are also likely be living without their husbands (45.8%), but a substantial number live in nuclear family settings as well (39.6%). Goun women, on the other hand, are more likely to live in extended family settings (45.9%) and with their husbands (51.3%), than women in either of the other two groups. Variation in these patterns reflect and (re)produce particular localized social institutions. These, in turn, will shape women’s opportunities in strategy configuration. Geographical localization in these social institutions is evident in that there is a relationship between residential arrangement and area of the city. Table 4.5 reveals the spatial distributions of residential arrangements within Porto Novo.

67 Ethnicity Residential Arrangement other Yoruba Goun Fon

Nuclear 40.0% 18.3% 26.1% 39.6%

Female-headed 40.0% 22.1% 15.5% 22.9%

Family w/ Husband 0.0% 1.9% 0.9% 2.1%

Family w/o Husband 0.0% 35.6% 23.7% 22.9%

In-laws w/ Husband 20.0% 14.4% 25.2% 12.5%

0.0% 7.7% 8.5% (L0%.

Total 100.0% 100.0% 100.0% 100.0%

Table 4.4 Residential Arrangement by Ethnie Group

Area of City Residential Arrangement Urban- Urban- City Center City Center Suburb Suburb - fringe fringe - M arket No Market Market No M arket M arket No Market

Nuclear 13% 24.5% 23.9% 18.6% 36.6% 37.0%

Female­ 12.5% 18.4% 23.9% 16.9% 22.0% 16.7% headed

Family 1.0% 2.0% 1.5% 1.7% 0.5% 0% w/husband

Family w/o 46.9% 24.5% 19.4% 42.4% 12.9% 11.1% husband

In-laws 15.6% 24.5% 23.9% 16.9% 22.0% 31.5% w/husband

In-laws w/o 16.7% 6.1% 7.5% 3.4% 5.9% 3.7% Huühand

Total 100.0% 100.0% 100.0% 100.0% 100.0% 100.0%

Table 4.5 Residential Arrangement by Area of the City

68 In the city center near the market the dominant residential arrangement, for women in the sample, is living within their own extended families without their husbands. Based on evidence in Table 4.4 people in this area are expected to be predominantly Yoruba. Indeed, 44.2% of all Yoruba women surveyed live in the city center near the market (see Table 4.6). Likewise, in the city center area away from the market, extended family, both women’s own without husbands and in-laws with husbands, are the dominant arrangements. The type of housing found in this area tends to be original housing stock, often homes belonging to particular family lineages.

Ethnicity City Area other Yoruba Goun Fon

other 6.7% 1.0% 2.4% 2.1%

City Center - M arket 6.7% 44.2% 14.3% 0.0%

City Center - No Market 0.0% 9.6% 82% 12.5%

Suburb - Market 26.7% 6.7% 13.1% 18.8%

Suburb - No M arket 13.3% 23.1% 7.3% 12.5%

Peri-urban Market 46.7% 11.5% 41.9% 41.7%

Peri'Urhnn - Nn Market 0.0% 3.8% ------118%______1 2 Æ

Total 100.0% 100.0% 100.0% 100.0%

Table 4.6 Ethnicity by Area of the City

69 As expected, the main form of residential arrangement found in the suburbs, both near and removed from the market, are nuclear and female-headed households. This area is dominated by relatively new construction. As noted above, most people in Benin expressed the desire to build their own homes. The first tier of urban development (suburbs) is the first area in Porto Novo where this has occurred. Two types of housing stock are found, large multi-story owner occupant homes and large concessions containing a number of rental housing units akin to row-houses. The former are primarily homes of nuclear families, while occupants of the latter may be either nuclear families or female-headed households. Again a concentration of Yoruba are found here (see Table 4.6). Finally, on the urban-fringe two types of residential arrangements are evident. Near the market the dominant form are nuclear family households (36.6%), but female­ headed (22.0%) and extended-family (in-laws with husband=22.0%) settings are also prevalent. Removed from the market, women are somewhat more likely to be living with their in-laws and husbands (31.5%), but nuclear family arrangements (37.0%) are also evident. The mix of housing in this area reflects the fact that it is in transition. While older mud-brick family compounds containing large extended families dominated in the past, new houses are being built. The city government is facilitating this processes by putting in roads and parceling land into saleable lots. In some instances, people’s homes and/or shops have been destroyed. The geographic patterns evident in residential arrangement by ethnic group across the city reflects localized social institutions, which produce unique interactions among Structures o fConstraint. As a result, women’s opportunities in strategy configuration will not only vary as a function ofherXssgrs, education, and training, but also as function of where she lives and with whom. This aspect of women’s lives reflects the ways in which Cultural Norms, as one aspect o f Structures o f Constraint, shape women’s opportunities.

70 4.6 Analysis Procedures The relationship between women’s strategy configurations, opportunities for income generation, and localized social institutions within Porto Novo, Benin are investigated in four stages using quantitative and qualitative methods, and data fiom both the surveys and interviews (Figure 4.3). Research starts with identification of livelihood strategies employed by women in Porto Novo, and moves toward explaining strategy configuration in terms of the enabling conditions that give rise to them, and finally, the social processes that produce those conditions. The first stage (Chapter 5) identifies the kinds of strategies women employ. Investigation begins with examination of how and why women combine particular economic activities, as well as their relative importance. This is based on a content analysis of open-ended survey questions. The purpose is to "establish a set of categories and then count the number of instances that fall into each category" (Silverman 1993:9- 10). Having identified the how and why women create their livelihood strategies, subsequent analysis draws on women's categorization of economic activity. A cluster analysis is then preformed to determine specific combinations of economic activities that form the range of livelihood strategies used by women in Porto Novo. Stage two (Chapter 6) explores ways in which particular localized social institutions produce enabling conditions that facilitate employment of particular strategy configurations. Using a Chi-square Automatic Interaction Detection (CHAID) analysis association between variables representing Structures o fConstraint and specific strategy types is identified. Of particular interest is the relative importance of specific types of Assets and mobility as enabling conditions. Through this process, strategy configurations are grouped according to enabling conditions necessary for their deployment. In stage three (Chapter 7), enabling conditions identified in stage two are linked to opportunities for income generation through the use of descriptive statistics. This enables exploration of the ways in which sets of enabling conditions produce unique opportunities for accumulation. Emphasis is placed on differences between starting and current capital. Also, important is a content analysis of open-ended survey questions and

71 idenmüig

acnvitylcoinbmanpns;

Mahoib

m m Identifying Enàbiing Conditions Id(mti^3^B&èKcna%oî\iohibii3ëplo^

Aftr/Aotf: Chi-square Automatic Meraction'Detêctioirterassess'relatibnsbip

potential consider relations%,betweenspecifîcen«)ling%M&titm#nmncoimë% generation cmuid^ M e t & W ^ ] '^blm estion&w

Jnvailgaîe^socioeèonomm^^mM r^:mea#dmmân ahdinablmecM Mmod: ^D^mçtweSgtig^ CdiSen

Figure 4 J Research Steps 72 interview data. The object is to understand the meaning of different enabling conditions for women in their strategy configuration. Finally, stage four (Chapters 8 and 9) investigates the socio-economic and spatial processes behind strategy configuration. Initially dominant processes for each group of strategy configurations are identified through descriptive statistics. Content analysis of interview data is then used in a preliminary examination of ways in which localized social institutions shape opportunities for strategy configuration as a fimction of Structures o f Constraint. Results of all these analyses are synthesized in the concluding chapter (10), which explains how opportunities for strategy configuration emerge fix)m localized social institutions, manifest in Structures o f Constraint, to produce conditions that enable or inhibit deployment of particular strategies. Additionally, the association between particular strategic arrangements in economic activity and income generation potential are discussed. Finally, the theoretical and empirical implications of this study are explored, emphasizing ways in which it is a platform for fiiture research.

4.7 Conclusion This chapter focuses on application of the conceptual firamework to the Béninois context. This began with an historical discussion of Benin’s political-economy, development of Porto Novo, and women’s place within it. Subsequently, data collection procedures were discussed, focusing especially on operationalizing the theoretical firamework. Structures o f Constraint. The chapter then turned to describing basic sample characteristics, emphasizing ways in which it is representative of Benin’s population in general, and the Oueme, particularly. Of special interest was the geography of social groups within Porto Novo, focusing on spatial variation in residential arrangement and ethnicity. Finally, research procedures, foci of the following chapters, were presented.

73 CHAPTERS

CLUSTER ANALYSIS: WHAT STRATEGIES DO WOMEN IN PORT NOVO USE?

5.1 Introduction The empirical framework (Chapter 2) indicated a variety of economic strategies are used by households and individuals. These can be categorized into two general types; mono-activity, a single income source, and plural-activity configurations, comprised of multiple economic activities (MLES). MLES is widely used by Afirican women in general, and in Benin, specifically. In most cases women engage in several activities simultaneously and/or seasonally, but activity combinations differ. The resulting outcomes, in terms of income generation potential, likewise vary. It is important to understand, then, that MLES is not a single economic strategy, but rather a category of livelihood strategy encompassing numerous possible configurations and outcomes. Therefore, strategy configurations are distinguishable based on the discrete type and number of activities comprising them. The conceptual framework (Chapter 3) indicates that women’s opportunities in strategy configuration reflects their socioeconomic setting, conceptualized as localized social institutions, which manifest in Structures o f Constraint^^. Before explanation can be offered as to why women choose different strategic combinations of economic enterprises, and how these produce varying outcomes, however, it is first necessary to identify the variety of strategy configurations among women in Porto Novo - the objective of this chapter.

" Important theoretical concepts are highlighted in italics.

74 Identification of strategy configurations utilized by women in Porto Novo proceeds in three steps. First, an empirically and conceptually meaningful typology is developed based on distinguishing among economic activities based on a content analysis of open-ended survey questions. The result is three discrete variables - occupational category, employment classification, and type of good - which are subsequently combined to produce a single variable encompassing all three aspects of economic activity. Second, the typology provides a basis for determining the total number of activities comprising women’s livelihood strategies. Third, generalized strategy configurations are identified through cluster analysis, grounded in a content analysis of open-ended survey questions which facilitates understanding women’s rationale for combining and prioritizing the activities comprising their livelihood strategies. As noted in Chapter 4,522 comprehensive surveys are used for analysis. Data gathered included information on economic activities, demographics, household composition, family background, and financial responsibilities (French and English versions of the survey are provided in Appendices A and B respectively). The data of principle interest in this chapter concerns economic activities.

5.2 Content Analysis: Developing an Economic Activity Typology Characterizing the types and numbers of economic activity in which women engage proceeds in two stages. Content analysis of women’s responses to open-ended survey questions provides a basis for distinguishing among their enterprises. This approach first constructs a coding scheme containing all possible theoretically and empirically grounded themes identified in women’s responses to open-ended survey questions (Silverman 1993). The coding scheme is subsequently used to create categorical variables capturing the Ml range of responses. This facilitates identification of three unique aspects of activity ^ e - occupational category, employment classification, and type of good for trade activities. The resulting classification of economic activity is then used to create an empirically and conceptually grounded

75 typology of economic activity (For a complete list of variables/categories the data dictionary in Appendix C). 5.2.1 Economic Activity Categorization To identify types of livelihood strategies women used by women in Porto Novo, two main questions were posed; I) What is your principle source of income? 2) Do you do other things to earn money? In women’s responses to these questions three aspects of economic activity were identified as critical to distinguishing among them: occupational category, employment classification and type of good. 5.2.L I Occupational Category Initial categorization of women’s responses to the questions above resulted in identifying of nineteen occupational categories. Wage earners distinguished themselves based on their profession, artisans their craft, and traders on the scale of their activities as well as their role in the production process. As illustrated in Figure 5.1 shows the various occupations women identified in their second economic activity*'*. Given the high degree of detail in, and smallness of some divisions, they were summarized into six general occupational categories (Table 5.1, Column 1): wholesaler, retailer, producer, artisan, service provider, high-wage employment, and low-wage employment. For example, all craft activities such as seamstressing and hairdressing were classified as artisan. Commercial activities were categorized in greater detail than other activity types because they represent the greatest proportion of all economic activities - 82.6% of all enterprises in activity one alone. Other activities are distinguished based on level of education and training required. Artisanal activities are grouped together because although specialized instruction is required, formal education is not. Service providers, are, likewise, distinguished by the type of training required. For example.

*■* AU occupation types were not identified for every economic activity. Activity two occupational categories are displayed here as an example because activity one included too many to display efiectively. In aU other instances Activity one is used as the example unless otherwise noted.

76 healaf______«wvant oUwf swvices trainer heallhworker teacher sales derk hairdresser wholesaler seamstress

producer retailer

Figure 5.1 Percentage of Women by Activity Two Occupation Categories

work as a traditional healer requires training, but not formal education. In contrast, a health-worker or nurse usually has both and usually has formal wage-employment. Therefore, traditional healers are classified as service providers, and health-workers as high-wage employment. Individuals with wage employment were classified based on the requisite level of education. Occupational categories requiring a high degree of education and training, such as teacher, are classified as high-skill occupations, whereas those involving little or none, such as servant and sales clerk, were identified as low-wage Jobs. 5.2J.2 Employment Classification A second aspect of economic activity characterization pertaining to employment classification. This reflects with whom, if anyone, a woman works and/or if she is employed by someone else. For example, traders often indicate that a particular activity

77 is carried out by one of their children. Likewise, an important distinguishing characteristic among artisans is whether or not they have apprentices. Although they are not paid, (in fact, they pay for their training) the presence of apprentices greatly enhances profitability of an enterprise because they are a source of income and increase the potential number of clients a woman can serve. To verify this information women were asked about their relationship to employees. Three types of responses emerged: no relation (e.g, apprentices); family (e.g. children, younger siblings); and friends. Additionally, women identified themselves as wage workers in specifying their occupation. Combining these two pieces of information facilitates identifying women’s employment status vis-a-vis each activity. The resulting classification is consistent with that developed by Brown et. al. (1997). While their scheme includes nine categories’’, to enhance applicability to the Béninois context, these are collapsed into seven (Table 5.1, Column 2): owners-employees, owners-family, self-employed, high-skill public, high-skill private, low-skill public, low skill private. Their approach is modified in the following ways. First, while Brown et. al. (1997) characterize self-employed as including both those who work alone and those who employ unpaid family labor; self-employed only includes people who work alone. Persons who have family assistance with activities are classified, instead, as owners-family. The distinction between people with family assistance and those who truly work alone was made because family assistance is understood as influencing strategy configuration. The original category owners is renamed owners-employees, distinguishing it from owners-family, but its definition, employers of wage labor, remained the same. In addition, to paid employees this category includes employers of apprentices. Their categorization with wage labor, as opposed to unpaid family work, reflects that apprentices are generally not related to employers.

” Their scheme, building on work by Lawson, (1990), identifies nine employment categories: owners, self-employed, managers public, managers private, high-skilled public, high-skilled private, low-skilled public, low-skilled private and family (Brown, et. al., 1997).

78 Second, managers public, managers private, high-skill public and high-skill private are collapsed into a two categories - high-skilled public and private. One reason for this reduction is that relatively few individuals in the sample engaged in high-skill wage labor, most employed in the public sector. Also, there are only three categories of high-skill wage employment - teachers, health-workers, bureaucrats. The low-skill classification is left as it was in Brown et. al. (1997), again because it represents a relatively small proportion of activity types in the sample. Jobs in these categories include sales clerks, secretaries, servants, and law/military officers. Finally, family employment is dropped from the typology because those individuals are not income earners and therefore not represented in the sample. 5.2.1.3 Type o f Good This aspect of economic activity only applies to commercial enterprises (retail, wholesale, producer). Traders differentiated themselves by type of good they sell. The initial categorization of goods sold by women in the sample included 63 discrete items. Women’s rationale for distinguishing among goods was relative profitability and frequency of sale. This was most apparent in women’s responses to the "Why is this activity important?" Figure 5.2 reveals that for activity one, the two most frequent responses were "profitability" and "quick sales"It is interesting that these two responses are mutually exclusive. For example, as one woman with five activities explained, the gum and candy she sells "feeds the kids everyday" whereas fabric and jewelry "have higher profits". Implicit in this remark is the awareness that fabric and jewelry are bought with less regularly, but have a higher value. Likewise, another woman identified food items as the most important because "she sells it regularly. People buy food first.’’"

The “only activity” response was given by women with only one activity, hence its importance as their only source of income.

” All quotes taken from surveys collected in Porto Novo, Benin January to September 1998.

79 X

Responses: Activity 1 - Why is this activity important?

Figure 5.2 Responses on Activity Importance by Percentage of Women

Economic anthropological classifications of goods provide a similar categorization. This argues that consumption is part of production and exchange and, therefore, all three are socially constructed. As a result, social perception of the use of goods influences exchange. While these classifications focus on the social meaning of goods, and implicit power relations reflected in ownership, another element is the firequency with which goods are purchased and used. For example, food as a category of goods is consumed daily, irrespective of the social meaning in particular meals, types of foods, or food items associated with rituals (Orlove and Rutz 1989; Douglas and Isherwood 1996).

80 The influence of perceived consumption patterns on traders is evident in their explanations of the relative importance of economic activities. Drawing on understanding that the frequency with which goods are sold reflects the socially constructed nature of their value, four general categories of goods are identified (Table 5.1, Column 3); (1) daily-use goods refers to items, such as food and fuel, that meet basic needs; (2) as-needed goods are bought if people have money and/or a need, but are not necessarily consumed on a daily basis (e.g. medicines, housewares, shoes); (3) occasional-use goods are those people buy for special circumstances (e.g. people save money all year to buy fabric for a new outfit for Ramadan, Christmas, and/or New Years celebrations'^); (4) intermediate goods are used as inputs to other activities such as hardware and lumber. 5.2.2 A Typology of Economic Activity The three variables resulting from this categorization - occupational category (actxocat), employment classification (actxclas), type of good (actxcom)” - are combined to produce an empirically and conceptually grounded typology of economic activity (Table 5.1). Creation of the typology proceeded in the following manner. First, occupational categories are subdivided to incorporate employment classification. Second, commercial activities (wholesaler, retailer, producer) are further differentiated to include type of good. The resulting 46 distinct types of economic activity and abbreviations used are (Table 5.1, Column 4):

WHL-OWNEMP-D AILY, WHL-OWNEMP-ASNEED, WHL-OWNEMP- OCCAS, WHL-OWNEMP-INT: wholesalers with employees, trading in daily, as-needed, occasional, and intermediate goods

'* Ramadan is a Muslim holiday celebrating the end of a forty day fasting period.

Actxcat, actxclas, and actxcom are the variable names for each discrete dimension of economic activiQr, occupational category, employment classification, and good type. Actx refers to activity number, x is used in place of a specific number because data are coded for up to six activities.

81 WHL-OWNFAM-DAILY, WHL-OWNFAM-ASNEED, WHL-OWNFAM- OCCAS, WHL-OWNFAM-INT: wholesalers using unpaid family labor, trading in daily, as needed, occasional, and intermediate goods

WHL-SELF-DAILY, WHL-SELF-ASNEED, WHL-SELF-OCCAS, WHL- SELF-INT: self-employed wholesalers, trading in daily, as-needed, occasional, and intermediate goods

RET-OWNEMP-DAILY, RET-OWNEMP-ASNEED, RET-OWNEMP- OCC AS, RET-OWNEMP-INT: retailers with employees, trading in daily, as- needed, occasional, and intermediate goods

RET-OWNFAM-DAILY, RET-OWNFAM-ASNEED, RET-OWNFAM- OCCAS, RET-OWNFAM-INT: retailers using unpaid family labor, trading in daily, as-needed, occasional, and intermediate goods

RET-SELF-DAILY, RET-SELF-ASNEED, RET-SELF-OCCAS, RET- SELF-INT: self-employed retailers, trading in daily, as-needed, occasional, and intermediate goods

PRD-OWNEMP-DAILY, PRD-OWNEMP-ASNEED, PRD-OWNEMP- OCCAS, PRD-OWNEMP-INT: producers with employees, trading in daily, as-needed, occasional, and intermediate goods

PRD-OWNFAM-DAHY, PRD-OWNFAM-ASNEED, PRD-OWNFAM- OCCAS, PRD-OWNFAM-INT: producers using unpaid family labor, trading in daily, as-needed, occasional, and intermediate goods

PRD-SELF-DAHY, PRD-SELF-ASNEED, PRD-SELF-OCCAS, PRD- SELF-INT: self-employed producers, trading in daily, as-needed, occasional, and intermediate goods

SERVICE-OWNEMP, SERVICE-OWNFAM, SERVICE-SELF: service providers employing wage labor, unpaid family labor, and self-employed

ARTISAN-OWNEMP, ARTISAN-OWNFAM, ARTISAN-SELF: artisans employing wage labor, unpaid family labor and self-employed

HIGHSKILL-PUBLIC, HIGHSKILL-PRIVATE: high-skiU wage employees in the public and private sectors;

LOWSKILL-PUBLIC, LOWSKILL-PRIVATE: low-skill employees in the public and private sector

82 l»OccuBattonil II -Employment ClaeeNlcatlon lll-TVpeofGood IV-3 Part Economic Srtw?r«ff Cateooriee (actxcom) AcMvltv Variable (actxocat) (actxclas) (actxc)

Wliolmaler Owner-employees Oaily-Use Goods whi-ownemp-daily (owns productive resources, (food, fuel) whhownemp-asneed Retailer employs wage labor) whi-ownemp-occas As-Needed Goods whi-ownemp-prd Producer Owner-hmily (Medicines, Housewares, whl-ownfism-daily (owns productive resources, Used Clothing/Shoes) whi-ownfiim-asneed Artisan employs unpaid bmily labor) whl-ownAm-occas (seamstress, Occasional Goods whl-ownfism-prd hairdresser, Self-Employed (Fabric, Alcoholic/ whi-self-daily photographer) (owns productive resources, Carbonated Beverages, whi-self-asneed works alone) Beauty Supplies) whi-self-occas Service Provider whi-self-prd (traditionai healer, High-Skill Waged Public Producer Goods religious leader, typist, (employment in the public sector, (hardware, lumber, paint, ret-ownemp-daily healtlHvorker) some autonomy In work - e.g„ car parts) ret-ownemp-asneed teacher, bureaucrat) ret-ownemp-occas High-SkillWage ret-ownemp-prd Employment High-Skill Waged Private ret-ownfam-daiiy (teacher, health- (employment in the private sector, ret-ownfsm-asneed wotlrer, bureaucrat) some autonomy in work - e.g„ ret-ownAm-occas teacher, health-worker) ret-ownfam-prd Low-SkillWage ret-self-daily Employment ImOw-Skill Waged Public ret-self-asneed (secretary, servant, (employment in the public sector, ret-self-occas sales clerk) little autonomy in work - e g., ret-self-prd secretary, law/military officer) prd-ownemp-daily Low-Skill Waged Private prd-ownemp-asneed (employment in the private sector, prd-ownemp-oceas little autonomy in work -e .g ., prd-ownemp-prd bartender, sales clerk, servant) prd-ownfsm-daily prd-ownfam-asneed prd-ownhm-occas prd-ownhm-prd prd-self-daily prd-self-asneed prd-self-occas prd-self-prd

artisan-ownemp artisan-ownfam artisan-self

service-ownemp service-ownfam service-self

highskill-public highskill-private

lowskill-public towskilFprrvate ____ Source; Wright, e t al. (t982); Lawson, (t990); Brown, e t ai. (t99S, t997) Note: Classification for each indhrklual is based on the intersection of three variables: ACTXC=ACTXOCAT( occupational category)+ ACTXCIAS (employment category - ovmer and self-employed categories applied to all but, waged categories)+ ACTXCOM ( type of good - applied only to occupational categories: wholesaler, retailer, producer)

Table 5.1 Typology of Economic Activity

83 This three dimensional variable enhances differentiation among activities utilized by women, while simultaneously facilitating comparison across cases.

5.3 Identifying Number of Activities Determining the total number of activities in which women engage was primarily based on their own assessment and characterization of distinct activity types, as indicated above. Activities were also differentiated where they took place in unique locations, merchandise was obtained &om different suppliers, and/or occurred at different times during they day. Data were recorded for up to six economic activities for each woman. Table 5.1 specifies frequencies for the number of activities engaged in by women, for which the mean is 2.06 (std. deviation=1.08). Subsequent analyses include only the first four activities, because only a small percentage of women have five(2.7%) or six (.4%) activities. It bears noting that the majority of women (62.8%) have more than one activity, most use some form of MLES.

Number of Cumulative Activities Freauencv Percent Percent 1 194 37J 37.2 2 17Î 34.1 71 J 3 92 17.! 89.1 4 41 7 i 96.9 5 14 2.7 99.6 6 3 .4 100.0 Total 522 100.C Mean 2.0( Std. dev. 1.0!

Table 5.2 Number of Activities by Percentage of Women

84 Having now distinguished among, and determined, the total number of economic activities women employ configuring their livelihood strategies, attention turns to identifying specific strategy types.

5.4 Identifying Livelihood Strategies Identification of women’s livelihood strategies proceeds in two stages. First, a content analysis of open-ended survey questions provides an understanding of the rationale for combining and prioritizing particular activities. Second, a cluster analysis is used to identify unique strategy configurations, drawing on the content analysis. 54.1. Content Analysis: Rationale for Combining and Prioritizing Activities 5.4.1.J Rationale for Combining Activities Women were asked why they combined particular activities in configuring their livelihood strategies (Figure 5.3). Although a number of responses were given, the three main ones were: to make ends meet (31.1%), increase income (30.5%), and diversification (18.4%). These answers reflect those most frequently given by traders, who dominate the sample. Responses from women in other occupational categories varied somewhat. Whereas diversification was important to traders (21.7%), "making ends meet" and "increasing income" were more important to service providers (making ends meet=25.0%, increasing income=25%) and artisans (making ends meet=38.6%, increasing income=l 8.2%). For traders the importance of diversification reflects the type of goods sold. As one woman who combines a number of daily-use goods indicated,"! find a little bit of profit from each thing”. In contrast, a combination of daily-use goods and occasional-use goods allows them to both meet financial obligations and accumulate savings. A woman who employs this strategy stated, “[t]hanks to my commerce I can do lots of things like feed my kids and have savings." Artisans, on the other hand, expressed the need for additional activities because they did not earn regular income from their craft. As one hairdresser noted “[i]t is only during the holidays that one works well, the

85 profitability

complimentary activi

salary insufficient ability other lias savings

making ends meet diversification

social expectation family heritage

personal preference

seasonality increase income flexibility

Figure 5.3 Rationale for Combining Activities by Percentage of Women

other days one can stay without anything to do firom morning till night, yet I have to pay the rent, electricity, etc." Women with high-skill public and private employment frequently cited "insufficient salary", a relatively minor category within the whole sample (4.4%), as the most important reason for having additional activities (52.4%). Yet, as one teacher commented "teaching is the most important activity because it provides a regular income and retirement."”

” All quotes taken from surveys collecting in Porto Novo, Benin January to September 1998.

86 As indicated by the diversity of responses to why women employ MLES, it is important to distinguish among specific strategy configurations. Since answers varied by occupational category this provides the basis for doing so. 5.4.1.2 Rationale for Prioritizing Activities within Strategies As indicated by the teacher quoted above women also prioritized activities based on income generation potential and regularity. As a result, when asked to order activities in terms of their importance, women had no trouble doing so. Table 5.3 lists fi-equencies for women’s prioritization. Most women stated that the first activity mentioned, in answering the questions "What is your principle source of income?" "Do you do other things to earn money?", was their most important activity. In general, women engage in activities one and two all the time, and activities three and four only occasionally. For example, a number of traders worked as maids intermittently when they need to earn money quickly. Similarly, many women raised livestock, which only sold once a year or when they had an emergency. Based on this prioritization among activities, activities one and two are considered primary activities, whereas activities three and four are seen as supplementary. The reason for considering activities one and two as principle is that women were inconsistent in identifying their principle activity^' as their most important activity^. This is reflected in Table 5.3 in that 10.3% of women identified activity two as the most important. In all likelihood, this indicates a difference between the importance one’s vocation and the main means of income generation. For example, a artisan may consider her vocation (e.g. seamstressing) the principle activity, inspite of the fact that it is not her main source of income. She might then identify the second activity as the most important because it provides the greatest earnings.

Answer to the survey question: “What is your principle economic activity?”

^ Answer to the survey questions: “Which is the most important activity to you?” “Why?”

87 O rder of Activity Activity Activity Activity Importance One Two Three Four

1 87.0 lO J 1.7 .4

2 9.8 48.5 3.1 .2

3 1.7 2.7 23.8 .8

4 .4 4 10 10A

No Response” 1.1 1.3 2 1.0

Not Annlicahle^ 00 70.3 8 7 4

Total 100.0 100.0 100.0 100.0

Table 5.3 Order of Importance among Economic Activities by Respondents

Having now differentiated among activities, determined the total number in which women engage, and their prioritization within strategy configuration, it is possible to identify the range of livelihood strategies employed by women in Porto Novo. 5.4.2 Cluster Analysis: Identifying Livelihood Strategies Identifying women’s livelihood strategies is based on grouping them according to their combinations of economic activity. Degree of homogeneity among activity combinations (activity type - Table 5.1, Column 4) and number (Table 5.2) is determined and then cases are clustered using a similarity measure. This proceeds in the following manner. Economic activities for each pairing of women are compared and the weighted value of matches calculated. Every case is compared to every other producing a matrix of similarity measures for each pair. The matrix o f similarity measures is then

“ No Response indicates that did not prioritize economic activities - all were equally important.

Not Applicable indicates that a woman did not have a second, third, and/or fourth activity.

88 used as the basis for clustering women according to the homogeneity of their livelihood strategies (activity type and number). Each step is discussed in detail below. 5A.2.1 Matrices o fSimilarity Measures To determine the degree of homogeneity in type and number of activities women engage in, a symmetrical matrix of similarity measures is produced using S-plus based on the value of activity matches between each case relative to the total possible matches. Every case is compared to all others in the calculation of the matrix (see Appendix D for the program coding)^^. This is done through a similarity measure specified as^‘;

Sii = where:

Sy = similarity between case i and case/ a, = activities for i a, = activities for/ //a ,, Uj) = fimction of value of activity matches between case i and case j c,= number of activities for/ Cy = number of activities for/

In other words: Sy = total value of activi tv matches between case / and case/' all possible combinations for / and/

Given women’s prioritization of activities and differences in rationales for combining activities, matches^^ in activity type are weighted to differentiate among them based both on positionality within the strategy configuration (primary or secondary activity) and occupational category (Table 5.1, Column 1).

“ n=522, so the resulting matrix is 522 x 522.

“ This similarity measure is consistent with (Everitt, 1993).

Matches indicates where two women have the same activity

89 Weightings are assigned as follows;

artisanal/service match in activities one and/or two = 2.00 high-skill match in activities one and/or two = 1.50 low-skill match in activities one and/or two = 1.00 any other match in activities one and/or two = 0.75 matches in activities three and/or four - 0.50 matches between activities three/four and one/two = 0.25

The value of wieghtings are not meant to indicate inherent hierarchy in activity types. The purpose is simply to differentiate among them. In this formulation cases comprising similar types and number of activities have measures close to one, whereas dissimilar ones are close to zero. As an example consider the following case:

CASE Activity 1 A ctivity 2 A ctivity 3 A ctivity 4

1(0 high-skill public retfam-asneeded not applicable not applicable

2(0 high-skill public retfam-occasional retfam-asneeded not applicable

The match between activity one for both cases has a value of 1.5, the match between activities 2 and 3 has a value of .25. Thus, the similarity measure is calculated as follows:

\5*25 1.75

In this instance the two cases are fairly similar, indicated by the closeness of the similarity measure to one. As noted above, comparison among all cases results in a symmetrical matrix (522 X 522) of similarity measures between all cases. This measure provides a basis for

90 grouping women relative to the similarity of their combinations of econontic activity (strategy configuration). S.4.2.2 Clustering Cases A non-hierarchical optimization approach was used for identifying similarity among cases based on the similarity measure. The basis for defining groups was a median clustering approach, in which representative cases serve as the key for identifying clusters (Everitt 1993; Bailey and Gatrell 1995). One advantage is that this approach is based on actual data, in this case similarity measures, in that medians are known a priori because they correspond to observed cases. In detemtining clusters, cases are divided into groups based on their similarity measure, the median similarity measure for all cases in each group is then identified, and the case closest to the median is used as the group focus. The objective of the approach is to "nunimize total weighted assignment of observations to selected medians" (Murray and Estivill-Castro 1998). This approach partitions cases based on the criterion of minimizing total within group variance. The clustering model utilized ensures all solutions are within 1% of optimality. The number of clusters are specified a priori and each new division of the data requires rerunning the analysis (Murray and Estivill-Castro 1998). Where hierarchies within the data are not sought, non-hierarchical optinuzation approaches have been shown to produce much more efficient results. As noted above, the optimality of each solution is specified beforehand (within 1% in this case). Thus, the results are considered to be optimal within 1% (Murray and Estivill-Castro 1998). This approach to clustering produces an efficiency measure for each solution by which the relative effectiveness of various data divisions can be assessed. As the efficiency measure decreases, the level of efficiency of the results improve. The curve that results firom plotting the efficiency mesarue relative to the number of clusters reveals the point at which the level of efficiency firom additional partitions of the data increases at a decreasing rate. The area around this point indicates a range of cluster solutions that are likely to be meaningful. While the ideal number of clusters must be determined by the

91 researcher, this mechanism provides one means of doing so (Murray and Estivill-Castro 1998). Determining the final cluster solution depends on use of both the efficiency measure and the researchers subjective determination of what constitutes clear, consistent clusters. The efficiency curve produced through this analysis is presented in Figure 5.4. Although the curve drops significantly at first, it then tapers off continuing to decline a steady rate rather than flattening. Cluster solutions between nine and fifteen appear to be significant and require further investigation. Examination of these solutions revealed that thirteen was the ideal number of clusters. This was apparent fix)m the fact that, in each additional division of the data starting with fourteen groups, only thirteen out of the total number had a clear logic to the organization of cases. Likewise, when only thirteen clusters are used each grouping is internally consistent in terms of the cases contained within it. Thus, thirteen was both the minimum and the maximum number of clusters that could be produced through the weighted similarity matrix that results in coherent groupings. S.4.2.2 Results: Thirteen Strategy Types Each of the thirteen clusters represent, a distinct approach to configuring one’s livelihood strategy. These vary both in number and type of economic activity. Table 5.5 lists all configurations including the name assigned to each, its abbreviation, a brief description, and the frequency with which each is employed. Strategy types are named for the dominant activity characterizing it. The tremendous importance of trade activities is evident in that many strategy types are comprised of different kinds of commercial enterprise".

" It is important to note that the employment classification of the strategy defining activity does not necessarily apply to all enterprises included in a livelihood strategy. A woman may be self-employed in one activity and have assistance firom unpaid family members in another. Likewise, women may trade in a variety of good types, but it is usually the case that the good type of the strategy defining activity also applies to other activities.

92 VO

Clusters

Figure 5.4 Cluster Solution Efficiency Scores Strategy Name (abbreviatton) Description Frequency Percent

Producef'Family (prodfam) 1-2 activities, producing and/or retailing daily-use goods, 26 4.98 usually food related, with family assistance

Producer-Selfemployed-1 -2 Daily- 1-2 activities, self employed producers, wholesaler, 61 11.69 uw Good Activities (prodself-daily- and/or retailers of daily-use, as-needed, and occasional l •2-daily) goods, at least one food related

Produccr-Self-employed-2-t- Daily- 2+ activities self-employed producers, wholesalers, and/or 50 9J8 use Good Activities (prodself-2+- retailers of daily-use, as-needed, and occasional goods, at daily) least one food related

Retailer-Family {retfam) 2-3 activities retailing daily-use, as-needed, and 42 8.05 occasional goods with family assistance

Retailer-Self-employed-l Daily-use 1 activity only - self-employed retailer of daily-use goods 50 9.58 Activity {reuelf-l-daily)

Retailer-Self-employed-2-3 Daily- 2-3 activities, low-skill wage work, self-employed 77 14.75 use Good Activities (/etseif-2-3- production and/or retailing of daily-use and as-needed daily) goods

Retailer-Self-employed-3+ Daily- 3+ activities, self-employed retailer of daily-use, 36 6.90 use Good Activities (,retself-3+- occasional, and intermediate goods daily)

Retailer-Self-employed-As-needed 1-2 activities, low-skill wage work, self-employed 27 5.17 Good Activities (retself-asneed) wholesaling and/or retailing of as-needed goods

Retailer-Self-employed-Occasional 2-3 activities, self-employed retailers of as-needed and/or 36 6.90 Good Activities (tetself-occasion) occasional goods

Artisan-Self-employed (artself) 2-3 activities, self-employed unlicensed artisans (mostly 28 5J6 hairdressers) and retailers of daily-use, as-needed and occasional goods

Artisan-Employees (artemp) 2-3 activities, licensed artisans, wholesalers, and/or 23 4.41 retailers of as-needed and/or occasional goods with employee assistance

Mixed-I Activity Only (mixed-l I activity only, service providers and wholesalers of as- 50 9.58 act) needed, occasional, and intermediate goods, with employee and/or family assistance

High-skill-Public(highslcill) 1-3 activities, high-skill public sector wage work, self- 16 3.07 employed wholesaling and/or retailing of daily-use, as

Total 522 100.00

Table 5.4 Summary of Livelihood Strategy Configuration Types

94 5.5 Conclusions The purpose of this chapter is to identify a parsimonious set of livelihood strategies employed by women in Porto Novo. This is achieved first by exploring ways in which women characterized, and distinguished among various activities. First, distinct aspects of women's economic activities, occupational category, employment classification, and good type, were identified through content analysis of open-ended survey questions. Combining these leads to 46 discrete activity types. Second the rationale for combining and prioritizing activities was determined also using content analysis. Third, women were compared in terms of their livelihood strategies, specifically based on type and number of activities using a similarity measure. Fourth, cases were classified into a set of thirteen distinct strategy types using cluster analysis.

95 CHAPTER 6

Cm-SQUARE ANALYSIS: WHAT ENABLING CONDITIONS SHAPE STRATEGY CHOICE?

6.1 Introduction Discussion to this point has considered the importance of various types of livelihood strategy for African women in general (Chapter 2); the socioeconomic experience of women in Benin (Chapters 2 and 4); and the specific livelihood strategies employed by women in Porto Novo (Chapter S). Attention now turns to identifying the Enabling Condition^^ that shape various forms of economic strategy utilized in Porto Novo by women. The conceptual firamework (Chapter 3) specified that intersections among various Dimensions ofDifference manifest in Structures o f Constraint, categorized broadly as Assets, Formal Rules, Cultural Norms, Preferences. Interactions among Structures of Constraint and spatial processes, in turn, generate localized social institutions, which produce localized Enabling Conditions shaping economic activity choice. This analysis focuses specifically on three sets of factors representing diverse localized social institutions: Assets, Cultural Norms, Location/Mobility. The objective of this chapter is to examine the ways in which these three sets of factors —Assets, Cultural Norms, Location/Mobility ~ produce Enabling Conditions shaping women’s strategy configuration choices (combined economic activities). This is

^ Important theoretical concepts are highlighted in italics.

96 accomplished through the use of Chi-square Automatic Interaction Detection (CHAID) analysis. Following a discussion of data used and description of the analytic technique, four different models are presented. The first three models focus on either: Assets, Cultural Norms, sad Location/Mobility. The purpose is to explore to what extent, and in what way, each of these set of factors produce conditions for employment of the thirteen strategy types identified in Chapter 5. The fourth model incorporates all three sets of factors to examine which among them are most important in strategy configuration.

6.2 Data Data used here are from 522 in-depth surveys of women in Porto Novo. In Chapter 5, this data was used to identify specific kinds of strategies. Here the life circumstances of respondents is the issue. Table 6.1 summarizes the set of variables characterizing Assets, Cultural Norms, Location/Mobility used in the analysis. Those representing Assets and Cultural Norms are discussed in chapter 4. Variables (highlighted in bold) characterizing Location and Mobility are discussed below. In addition, four control variables^" are used in all four models: ethnicity, religion, age, and number of activities ' 6.2.1 Location/Mobility Location and Mobility are characterized by several nominal variables. Residential location refers to the six areas of Porto Novo designated in the data collection process (city center, suburb, and urban fringe — each near to or removed from an area of intense economic activity or market). Activity and supplier locations for each enterprise are identified as home, Porto Novo market, another area in Porto Novo, Cotonou, the Oueme, another region in Benin, or outside Benin (usually Togo and/or

In the event that none of the variables associated with Assets, Cultural Norms, and Location/Mobility are significant, control variables will be the ones most strongly associated with strategy configurations.

Variable names are highlighted in bold.

97 Faetons Variable Cateanriea

Assets Source of Capital parents, husband, relative, friend, userie loan, bank loan, credit

Starting Capital <$50, $51-100, $101+

Current Capital <$50, $51-100, $101+

Education none, primary, secondary, terminal, university

Training 98hairdresser, seamstress, photographer, pastry chef

Employee Relationship children, younger siblings, other family, Mends, no relation

Cultural Norms Children’s Age-Range 0-10,11-20,0-20,5-15,20+^*

Principle Care-taker mother, relative, friend, maid, husband, themselves

Care-taking Burden low, medium, high”

Contribution to Household <25%, 26-50%, 51-75%, <75% Budget

Residential Arrangement female-headed, nuclear, in-laws with/without husband, family with/without husband

Household Head self, husband, parents, in-laws

Table 6.1 Analysis Variable Summary (part 1)

Although categories are over-lapping they are mutually exclusive in that each respondents household was only classified in one group. Categories are over-lapping to reflect the range of possible house-hold configurations, hence producing a nominal rather than ordinal variable.

“ Low=children >10, at least one >20; Medium=some supervision required by mother - children <10 and mother is not primary care-taker or children are 5-15; High=children <10 and mother is primary care-taker.

98 Enablluv CnnâMims V nriahle

Cultural Norms (contd.) Marital Status unmarried, married (polygamous or not)

Type or Marriage unofficial, civil, religious, customary, civil and religious

Years of Marriage number of years

Number of Children number of children

Number of Dependent Adults number of children over 20 and/or older (non*working) relatives

Percentage Own Children proportion o f women's own children in the household

Location and Mobility Residential Location city center, suburb, urban fringe - near to or removed from market area

Activity Location home, Porto Novo market, other area o f Porto Novo, Cotonou, Oueme, other Benin region, outside Benin (Nigeria/Togo)

Supplier Location home, Porto Novo market, other area o f Porto Novo, Cotonou, Oueme, other Benin region, outside Benin (Nigeria/Togo)

Mode of Transportation walking, moto-taxi, taxi, rented truck, personal/family moto or car

Control Variables Ethnicity Goun, Fon, Yoruba, other

Religion Catholic, Christian Celeste, Protestant, Muslim, Voudoun

Age Range 15-20,21-25,26-30,31-35,36-40,41- 45,46-50,51-55,56-60,61-65,66-70, 70+

Number of Activities number o f activities

Table 6.1 (contd.) Analysis Variable Summary

99 Nigeria). Mode of transportation characterizes Spatial Mobility in terms of distance in that taxis (taxi) are used exclusively for travel outside the city whereas walking and motorcycle taxi (moto-taxi) are the main modes of transport within it. Access to a personal or family vehicle implies gteatec Mobility.

6.3 IdentifyingEnabling Conditions'. CHAID Anaiysis The three sets of variables control. Assets, norms, and Location/Mobility are linked to the Structures o f Constraint reflecting localized social institutions. They are thought to produce Enabling Conditions, which shape women’s choice in economic activity and hence strategy configuration. To determine, which factors best characterize Enabling Conditions experienced by women in Porto Novo, as well as to identify the ways in which they condition women’s "purposeful" choices, a CHAID analysis is conducted. Four separate models are presented, one for each set of factors - Assets, Cultural Norms, Location/Mobility - independently, and one including all factors together. Each of the thirteen strategy configurations, identified in Chapter 5, is analyzed separately. By identifying variables most strongly associated with each strategy configuration, the three individual models facilitate understanding how the factors produce Enabling Conditions for each strategy type. The aggregated model allows identification of the factor(s) that best characterize Enabling Conditions for women in Porto Novo” . Prior to discussing each model, CHAID analysis is briefly explained. 6.3.1 CHAID Anaiysis Description CHAID analysis (SPSS 1998) differentiates cases within a given strategy type based on specific characteristics of their yiwe/s. Cultural Norms, Location/Mobilit^^.

” These analyses provide a check on the cluster analysis through which the thirteen strategy configurations were identified in that they assesses the degree to which cases within each strategy type are homogeneous.

This is based on strength of association measured through chi-square statistics in multi­ dimensional cross-tabulations. CHAID chooses among all the possible independents those which have the strongest association with the dependent and indicates in what way

100 For example, of persons using a Producer-Family strategy, one group are characterized by nuclear family residential arrangements and another extended-family settings, even though both use the same strategy. The object of the analysis is to find characteristics that are most strong associated with a particular strategy. CHAID analysis proceeds through a series of steps. First, for each variable the best partition among its categories is determined by merging categories in terms of their association with a given strategy type^‘. Second, CHAID continues to merge categories of each independent variable until each compound category is statistically significant in terms of chi-square association with the strategy. Third, the program tests the significance of each variable relative to the strategy through a chi-square statistic. Fourth, having assessed the strength of association between each variable and the strategy, CHAID orders the variables relative to their degree of association (chi-square) with the strategy. Fifth, CHAID grows a decision tree splitting the strategy cases into sub-groups based on association with various categories of each variable, starting with the variable that is most strongly associated with the strategy. For example, for the Retailer-Self-employed-2-2 Daily-use Good Activities, the most important variable was number of activities. Cases within this strategy were split into those that had two activities and those that had three. At each level a new analysis of association strength between each variable and the strategy is conducted. This form of decision tree analysis enables assessment of multiple interactions between a given dependent (strategy type) and a set of independents (Assets, Cultural Norms, Location/Mobility variables), identifying the most statistically significant. Each successive sub-group is characterized in terms of the independent variable categories with which it is most strongly associated. As a result it is possible not only to determine which variables are most strongly associated with strategies, but also

the independent is associated with the dependent variable through a test of significance (Kass 1980; Hawkins and Kass 1982; Loh and Shih 1997).

“ This is determined through a chi-square cross-tabulation of the each independent with the dependent.

101 the nature of that relationship in terms of variable categories QCass 1980; Hawkins and Kass 1982; Loh and Shih 1997). One draw back to this procedure is that it merges variable categories based on statistical significance without consideration of theoretical or empirical coherence. Once CHAID identifies variables most strongly associated with each strategy types, therefore, it is sometimes necessary to regroup categories in a more substantively meaningful fashion and then reassess the level of association of that variable with the dependent. In some cases regrouping results in the variable no longer being statistically significant and the next strongest must be assessed. In this way, growth of the decision tree can be directed such that both statistical significance and empirical and theoretical coherence are maintained. 6.3.2 CHAID Analysis: Four Models As noted above, four separate models are analyzed using CHAID, one for each of the three factors thought to produce Enabling Conditions for women, and one including all factors. The first three models facilitate examining specific characteristics ofXf.rers, Cultural Norms, and Location/Mobility associated with strategy. The last, aggregate model, focuses on the factors themselves, identifying which among them is most important in producing Enabling Conditions for a given strategy. Thirteen separate analyses, one for each strategy types, are conducted for each of the four models, producing a total of fifty-two decision tress. The results of each model are discussed below. 6.3.2.1 Assets Models Tables 6.2 and 6.3 summarize results of the Assets models for each of the thirteen strategy types (see Table 5.5 for a complete list of strategy types and their abbreviations). Table 6.2 includes strategy types that are identified as being Asset dependent {Substantial Asset Strategies), while Table 6.3 includes those strategies that do not require Substantial Assets {Limited Asset Strategies). Substantial Asset strategies (Table 6.2) are characterized by strong association between strategy type and significant levels of human and/or financial capital. This

102 ASSET VXBSABLES

Strategy Starting Current Source of Education Training Employee Capital Capital Capital Relation

RctSeir- >5100 >5100 N/E Primary/ not not applicable Occasion level 1 level 4 Secondary applicable level 2 x*»20.79*** x ^ 2 J 3 level 5 level 3 x^lO J9*** x*=2.72 x W .3 7 '

Mlxcd-1 Act ' >5100 >5100 N/E primary/ not not applicable level 4 level 1 secondary applicable level 3 x*=2J4 level 7 level 4 x^8.52*** x*=3.ll x’=9.87***

HighSkill >5100 N/E personal secondary/ N/E N/E level 5 savings terminal x^3.70 level 3 level 1 x^l3.37**‘ x*-59.70***

ArtEmp >5100 N/E N/E N/E seamstress/ no relation level 4 hairdresser level 1 x*»ll.S6*** level 2 x*»113.00‘ ** x*-59.61***

<5100 <5100 family primary/ seamstress/ not applicable ArtScIf level S level 6 level 7 secondary hairdresser level 2 x*= 1.88 x*=2.l9 x^2.40 level 3 level 1 x*=7J0*** x ^ l.7 0 x‘-71.05***

ProdFam <5100 N/E N/E none N/E chlldren/otbe level 2 level 3 r family x*-10.88*** x W .0 5 " level 1 x*-78.55***

RctFam <5100 <5100 family/perso N/E N/E chlldren/otbe level 2 level 4 nal savings r family x*»9.05** x^5.63 levelS level 1 x*=2.l5 x*-118.07*** Note: Level indicates the decision tree branch at which the variable entered the analysis. No Entry (N/E) indicates that the variable did not enter the analysis. •*•= significant at p value=0.00, ••=significant at p value=O.IO, *=significant at p value=0.50

Table 6.2 Characteristics ot Substantial Asset Strategies

103 linkage suggests Assets are critical to strategy usage and are, therefore. Enabling Conditions for the strategy. Two distinct patterns are evident. Some strategies are only distinguished by the importance of human capital, women's own, in terms of education and training, and/or in the form of other’s labor. A second set of strategies are characterized by a high level of both financial and human capital. Human capital is critical to employment of three strategies: producer-family {prodfam), retailer-family (retfam), and artisan-self-employed {artself). For the two family oriented strategies {retfam, prodfam) relationship to employees is crucial, specifically use of children’s unpaid labor. Personal human capital, in the form of training, is critical to users of the artisan self-employed {artself) strategy. As noted earlier, the importance of unpaid family labor is likely to manifest itself in the number and scale of enterprises in which women engage. Artisans, on the other hand, cannot ply their craft without some form of training. Strategy types requiring both financial and human capital include: artisan- employee {artemp), high-skill public {highskilt), mixed-l activity {mixed-l act), and retailer-self-employed-occasional good activities {retself-occasion). For trade-oriented strategies {retself-occasion, mixed-l act) substantial (> $100) financial capital at the start (starting capital), and/or through regular turn-over in merchandise (current capital), is critical. Also important, but secondarily so, is human capital in the form of education. This suggests that the type of goods and/or scale of enterprises require large amounts of money (both for start-up and maintenance). In contrast, human capital was more important for users of the artemp and highskill strategies. For example, for the artemp strategy starting capital was the fourth most important variable, but still characterized the majority of women in this group (19 out of 23). Women using either of these strategies caimot do so without training, in the case of the former, or education, in the case of the latter. Importance of financial capital for artemp refiects the fact that opening a workshop and employing apprentices require a substantial investment in equipment. As with the two trade-oriented strategies

104 characterized as Asset dependent, the importance of financial capital for the highskill strategy reflects the type of goods sold and/or the scale of the enterprise. Limited asset strategies (Table 6.3) are characterized either by smaller amounts (< $50) of financial capital or do not have any Asset variables figure prominently in the models. Even where starting and/or current capital entered the analysis, in many instances they are not statistically significant. Moreover, to the extent that human capital variables entered these analyses, there is a negative relationship between strategy type and the independent variables. For example, a few strategies were characterized as not having any employees (employee relationships not applicable). In sum, human and/or financial capital is critical to the use of prodfam, retfam, artself, artemp, highskill, retself-occasion, and mixed-l strategy configurations. These seven strategy types are, therefore, classified as Asset dependent (Substantial Asset Strategies). Assets in the form of human and/or a large amount of financial capital is seen as a necessary Enabling Condition for the formation of these strategy types. Other strategies do not require significant human and/or financial capital and, therefore. Assets do not produce critical Enabling Conditions for their employment. These are considered Limited Asset Strategies. 6.5.2.2 Cultural Norms Models CHAID models using just Cultural Norms and control variables do not produce clear, coherent results. Although each model for each of the thirteen strategies identifies Cultural Norm variables as significantly associated with the strategies, groups are not homogeneous. This suggests that different Cultural Norms can produce Enabling Conditions firom which various strategy configurations emanate, but they are not uniquely relevant. 6.3.2.3 Location/Mobility Models For all, but one strategy, supplier location is the dominant variable. Some strategies are characterized by extensive Spatial Mobility in that they require accessing suppliers outside Porto Novo. These are identified as Mobility dependent (Substantial Mobility Strategies) (Table 6.4). Others emphasize local suppliers (within Porto Novo),

105 .tSS’f r VARIABLES

Strategy Starting Capital Current Capital Source o f Capital E ducation T raining Em ployee Relation

ProdScIM -2-Daliy <$100 <$100 husband/family N/E N/E not applicable level 2 level 3 level 5 level 4 **=2.82 **“ 5.01 **=4.61 **=6.77*

ProdSeif-2-f-Daily N/E <$100 loan/credit N/E none not applicable level 2 level 3 level 4 level 5 **=14.16*** %*= 10.03* x*=4.32* **“4.43

RctSelM-Daily <$100 <$100 personal savings none/primary N/E not applicable level 4 level 5 level 3 level 6 level 2 *’=9.20»* **=2.89 **=10.11* **=1.79 **=12.00***

RctSeir-Asnccd <$100 N/E husband/family N/E N/E not applicable level 2 level 8 level 3 %'“ 5.94* **=1.63 **“ 2.65

N/E <$100 N/E none/primary none not applicable RetScir-2-3-Dafly level 4 level 3 level 5 level 3 **=3.50 **=4.13* **=2.42 **=4I.87***

RetScir-3-KDaily <$100 <$100 personal savings N/E none none level 4 level 5 level 6 level 3 level 2 **=3.86 **“3,90 **=2.73 **=4.77 **=8.14*** Note; Level indicates the decision tree branch at which the variable entered the analysis. No Entry (N/E) indicates that the variable did not enter the analysis. ••*= significant at p value=0.00, ••“ significant at p valuc“O.OI, •“ significant at p valuc“0.05

Table 6.3 Characteristics ofLimited Asset Strategies LOCATION/MOBILITV VARIABLES

Strategy Activity 1 Supplier Activity 2 Suppler Residential Location Activity 2 Location M ode of Transportation

RetScIf Occasion outside Porto Novo outside Porto Novo suburb Cotonou/Porto Novo N/E level 2 level I level 3 level 4 *•=8.03* k*=4I.08*** *'=7.47 x'=5.24

FrodSelf-l -2-Dally N/E outside Porto Novo all-market neighborhood m oto-ta*l level 4 level 2 level 4 level 1 *'=10.34* x '= ll.96** x'=6.04 *'=9.09*

RetSclf-Asneed outside Porto Novo N/E city center market Porto Novo N/E level 1 level 3 level 2 **=5.75* x'=6.89* x'=7.33

A rtE m p N/E outside Porto Novo all-market Porto Novo taxl/moto-taxi level 4 level 2 level 1 level 3 *2=8,95* % '= II2I*** x'=30.00*** *'=11.28**

N/EN/E center city/suburbs Porto Novo personal vehicle IllghSkili no m arket level 2 level 1 level 4 x'=3.69 *'=139.76*** *'=6.43*

Mixed 1 Act outside Porto Novo N/E city center Cotonou/Nigeria taxi/moto-taxi level 2 level 5 level 4 level 3 *‘=3I.58*‘ * *'=4.85* x'=4.92 x'=5.23 Noie; Level indicates the decision tree branch at which the variable entered the analysis. No Entry (N/E) indicates that the variable did not enter the analysis. *••= significant at p valuc=0 00, ♦♦=significant at p value=O.OI, ^^significant at p value =0.05

Table 6.4 Substantial Mobility Strategies and are, therefore, considered strategies that do not require Substantial Mobility {Limited Mobility Strategies) (Table 6.5).

LOCATtOmiOBlLlTYyAXaABLES

Mode of Strategy Activity 1 Activity 2 Residential Location Activity 2 Transporta Supplier Suppler Location tion

Prodfam N/E Porto Novo all-market Porto Novo foot/moto- level 3 level 2 level 4 taxi x*“3.64 x'=ll.96** x*“6.04 level 1 x*=8.09*

ProdSelf-2-t-- Porto Novo Porto Novo suburb/urban fringe Porto Novo foot/moto- daily level 2 level 4 market level I taxi x*=20J4*** x^3.6l level 3 x*=21.85^*^ level 5 xM .58 xW .60**

Rctfam Porto Novo Porto Novo suburb/urban fringe N/E foot/moto- level 2 level 1 no market laxi x M .I4 x*=37.0*** level 3 level 4 x*“9.23** xM.OO

RctScif*l> Porto Novo N/E all-market N/E N/E daily level 2 level 4 x2“ l4.83*** x*=2.45

Porto Novo Porto Novo N/E N/E foot/moto- RetSelM*3“ level 3 level 2 taxi daily x*=8J3* x*“ 20J3*** levels x*=4.64*

RetSclf-3+- Porto Novo Porto Novo urban-fringe market Porto Novo N/E daily level S level 3 level 4 level 2 x*=^92 x*“ 18.52*** xM 7.73* x^l8.37^^^

ArtScIf Non-trader Porto Novo suburb/urban &inge N/E Ibot/moto- level 3 level 3 market taxi x^25.59^*^ j ^ . 6 3 level S level 1 ,^ 1 .2 9 x*=27.72** *

Note: that the variable did not enter the analysis. ••*= significant at p value=0.00, ••“ significant at p value4).01, •“ significant at p value=0.05

Table 6.5Limited Mobility Strategies

108 Four strategy types - retailer-selfemployed-occasional good activities {retself- occasion), tetailer-selfemployed-2-3-daily-use activities {retself-2~2-daily), producer- selfempIoyed-l-2-daily-use activities (prodself-l-2'daily), artisan-empioyees (artemp) - are characterized by supplier location(s) of outside Porto Novo: the Oueme, Cotonou, another region in Benin and/or outside the country, usually Togo or Nigeria. These are characterized as Mobility dependent (Substantial Mobility Strategies). The importance of Mobility in using these strategies is also evident the modes of transportation, associated with each strategy. Specifically, women using these strategy types indicated that taxis and personal vehicles are their principle means of transportation, forms used primarily for long-distance travel. The highskill strategy is included in this group despite the fact that supplier location does not enter its analysis. Examination of supplier location firequency for this strategy reveals that 80% of women using it obtain supplies outside Porto Novo. Given the importance of accessing these non-local suppliers for all these strategy types. Spatial Mobility is considered a critical Enabling Condition for engaging them. These strategies are there. Mobility dependent or Substantial Mobility Strategies. The remaining seven strategies - producer-family (prodfam), producer- selfemployed-2+-daily-use good-activities (prodself-1-2-daily), retailer-family (retfam), retailer-selfemployed-l-daily-use good activity (retself-l-daily), retailer-selfemployed-2- 3-daily-use good-activities (retself-2-3-daily), retailer-selfemployed-3+-daily-use good- activities (retself-3+-daily), artisan-selfemployed (artself) - were also strongly characterized by supplier location. In these instances, however, women obtained supplies locally, within Porto Novo. The modes of transportation they use are also associated with local travel (e.g. foot, moto-taxi). These findings suggest that women using these strategies are more circumscribed in their Mobility. Therefore, strategy configurations in this group are characterized as Limited Mobility Strategies. For women using these strategies Spatial Mobility is not a critical Enabling Condition. In sum, CHAID analysis o fLocation/Mobility variables relative to each of the thirteen strategy types, reveals that supplier location, specifically within or outside Porto Novo,

109 is a critical factor producing Enabling Conditions for engagement in particular kinds of strategies. Moreover, findings relative to supplier location are substantiated by the mode of transportation associated with each strategy type. The reasons for the importance o îMobility are explored in the next chapter. Interestingly, the Location/Mobility models also indicate there is a strong linkage between particular kinds of strategy configuration and residential location. This supports the argument that localized social institutions produce differential opportunities for strategy configuration. 6.3.3.4 Aggregate Models To determine which among the three sets of factors best characterize Enabling Conditions for each of the thirteen strategies, a final CHAID analysis includes all variables together, i.e. the sum of the three smaller models discussed above^^. In examining findings firom the aggregate models, the importance of geography becomes evident. While human capital variables figure prominently in for a few strategies, spatial patterns, specifically as they relate to supplier location, most strongly characterize all strategies. This suggest that while various forms o f Assets are important to strategy configuration. Mobility is even more so. Supplier location, factored into the analyses in virtually all cases no later than the third level, and was statistically significant at the 0.05 level or better. For all but the two artisan-oriented strategies (artemp, artself), mode of transportation is also strongly associated. These findings not only validate earlier ones concerning the importance o f Mobility for strategy configuration, but also suggest that among all factors (Assets, Cultural Norms, Location/Mobility) potentially shaping Enabling Conditions, Mobility is the most important.

No table is included for this analysis because too many variables are included.

110 6.4 Conclusion This chapter identifies Enabling Conditions shaping women’s choice strategy configuration. Three sets of factors - Assets, Cultural Norms, and Location/Mobility - characterizing Structures o f Constraint and geographic processes are thought to produce Enabling Conditions. To assess the extent to which, and in what way, these factors are influential in strategy configuration three independent CHAID analyses are conducted. The results indicate that both Assets, in the form of human and/or financial capital, and Mobility, characterized by supplier location and mode of transportation, produce Enabling Conditions critical to strategy configuration. Virtually all strategy types are strongly associated with six variables representing these two factors: starting capital, current capital, education, training, supplier location and mode of transportation. The nature of that association differs, however, among strategy configurations. As a result, strategy types are grouped according to whether they require Substantial/Limited Mobility and Assets (Table 6.6). A final aggregate model, including all variables, reveals that among these two Enabling Conditions, Mobility is the most important. In addition to identifying the underlying Enabling Conditions shaping strategy configuration, the CHAID analysis validates identification of the thirteen strategy types. While strategies configuration is defined by similarity in type and number of economic activities, this analysis reveals that cases included in each strategy are also homogeneous in terms Assets and Location/Mobility. Two issues remain unclear. First, why and how Mobility is important to strategy configuration must be explored. This is taken up in the next chapter (Chapter 7). Second, although models for Cultural Norms variables are inconclusive in terms of their relationship to strategy configurations, they are not unimportant. Rather, heterogeneity among cases within strategy types vis-a-vis Cultural Norms suggests a number of processes related to this factor may produce Enabling Conditions. This is investigated in chapters 8 and 9.

Ill Cumulative

Iraup I —No As$e!s/NoMobility 213 40.8 40.8

Producer-Self>employed-2-*' Daily-usc Good Activities SO 9.6

Retailer-Self-employcd-1 Daily-useGood Activity so 9.6

RetaiIer-Self-employed-2/3 Daily-use Good Activities 77 14.8

Relailer.Selftemnlnvmt.3-(- Hnilv-iise Good Activities Id 6 9

Iroup 2 —Asuts/No Mobility 96 18.4 59.2

Artisan-Se[f>employed 28 S.4

Producer-Family 26 S.O

Retflfler.Familv A2 s o tmup 3 —No Assets/Mobiiity 88 16.9 762

Producer-Self-empioyed-l/2 Daily-use Good Activities 61 11.7

Retailer-Relf-emnloved As-Needed Good Activities ______2L ------5 J _

Iroup 4 — Assels/Mobility 125 23.8 100.0

Artisan-Employees 23 4.4

High-skill Public 16 3.1

Mixed 1 Activity SO 9.6

Retaiier-Self-employed-Gccasionai Good Activities 36 6 J

rotal 522 m o 100.0

Table 6.6 Strategy Groupings based on Enabling Condition Pairings

112 CHAPTER?

WHAT IS THE RELATIONSHIP BETWEEN ENABLING CONDITIONS AND STRATEGY ACCUMULATION POTENTIAL?

7.1 Introduction Thirteen specific strategy types used by women in Porto Novo were identified Chapter 5, the two main Enabling Condition^* underlying strategy choice. Assets and Mobilit/*, were identified in Chapter 6. Cohesiveness among strategy configurations relative to the amount and type o î Assets and degree of Mobility suggests these two factors facilitate particular kinds of strategies. Accordingly, strategies are classified by the amount and type ofidssers necessary for their utilization (Substantial/Limited Assets) and the degree oî Mobility required {Substantial/Limited Mobility). Then, four Enabling Condition groupings are identified based on interactions between these two factors (Table 7.1). The question remains however - what are the consequences for (financial) capital accumulation, of using different strategy types emerging from these Enabling Conditions! The conceptual firamework (Chapter 3) specified that intersections among various Dimensions o fDifference {Gender, Ethnicity, Class, Life

Theoretical concepts are highlighted in italics.

Enabling Conditions are highlighted with bold italics.

113 Cnmubitive S lra te w Crniininp« iroup I —No AsseaWo Mobility 213 40.8 40.8

Producer>Self'employed-2-)' Daily-usc Good Activities 50 9.6

Retailer-Self-employed-l Daily-use Good Activity 50 9.6

Retailer-Seif-einployed-2/3 Daily-use Good Activities 77 14.8

R«tailer.RglC#mnlnvfH.1+ Dailv-iiseGood Activities ------3fi_ 6 9 iroup 2 — Assets/No Mobility 96 18.4 59.2

Artisan-Self-employed 28 5.4

Producer-Family 26 5.0

Retailer.Pnmilv 42

Iroup 3-No Assets/Mobility 88 16.9 762

Producer-Self-employed-1/2 Daily-use Good Activities 61 11.7

Retailer.SelfLemnloved A (.Needed Good Activities 27 ------i2 _

"iroup 4 — Assets/Mobility 125 23.8 100.0

Artisan-Employees 23 4.4

High-skill Public 16 3.1

Mixed 1 Activity 50 9.6

Retailer-Self-employed-Occasional Good Activities 36 6.9

rotal 522 100.0 m o

Table 7.1 Strategy Groupings based on Enabling Condition Pairings

114 Cultural Norms, Preferences, Interactions among Structures o f Constraint and spatial processes, in turn, reflect localized social institutions manifest in loca/ized Enabling Conditions, which shape women's economic activity choices. And women's differential opportunities for strategy configuration result in disparate possibilities for accumulation (Chapter 2). Both Assets and Mobility were identified as critical Enabling Conditions, but Mobility was determined to be most important (Chapter 6). Specifically, Mobility outside Porto Novo to access suppliers is crucial to configuring some types of strategies. While Assets impact the type and scale of activities. Mobility enables women to over-come a lack ofXsseA under some conditions, and facilitate their use under others. This suggests there is a relationship between degree o(Mobility in strategy operationalization and accumulation potential. This argument is developed first by examining outcomes of different Enabling Condition groupings in terms of growth in financial capital, based on comparison of median starting and current capital^. Analysis demonstrates that growth of financial capital cannot be understood in dichotomous terms ~ survival/surplus, petty trader/mamabenz. Rather, outcomes reveal a continuum of growth potential ranging from survival (virtually no growth), to stability (some growth), to surplus (significant growth). Secondly, content analysis of open-ended survey questions and interview data is used to explore the importance of Mobility for accessing suppliers fi’om women's perspective. This analysis reveals that women compensate for Limited Mobility by associating numerous activities. The resulting strategy configurations reflect that in combination low profits from each activity to generate sufGcient income.

Variable names and concepts associated with accumulation potential are highlighted in bold.

115 7.2 Data and Analysis To examine the importance and consequences of mobility as an Enabling Condition two inter-related analyses are pursued. Data are taken firom the 522 in- depth surveys and 15 life/economic history interviews. Of specific interest firom the surveys are starting and current capital data. To examine the relationship between Enabling Conditions and accumulation potential the median difference between median starting and current capital is calculated for each of the four strategy groupings. This facilitates placing strategy groupings along a continuum relative to growth in financial capital. Having identified the relationship between Enabling Conditions and accumulation potential, women's perspectives regarding the importance of mobility is explored using content analysis of open-ended survey questions and interview data'". Unlike the content analysis used in Chapter 5, where identification of categories was of key importance, the approach is used here to explore the meaning of mobility (Silverman 1993). Of specific interest is women's explanations for why accessing suppliers outside Porto Novo can facilitate accumulation. For a complete list of interviewees, see Table 72*^.

All interviews were conducted in French or a local language, Goun or Yoruba, through the help of a translator. Transcriptions were, likewise, written in French. Interview text used in this analysis was translated by me. In so doing, I have tried to remain true to how women actually communicated including colloquialisms, repetitions of speech, and included punctuation sufficient only to facilitate understanding.

Interviewee's names have been changed to protect their anonymity.

1 1 6 Interviewees StrateevTvne Date of Interview Blandine RetailerSelf-employed-Occasional Good Activities October 1998 Cheti RetailerSelf-employed-As-needed Good Activities October 1998 Christian RetailerSe!f-employed~3+-Daily-use Good Activities September 1998 Claire RetailerSelf-employed-As-needed Good Activities October 1998 Delphine RetailerSelf-employed-As-needed Good Activities October 1998 Gladys Artisan-Self-employed November 1998 Kadija High-skill-Public October 1998 Marguerite Retailer-Self-employed-l Daily-use Good Activity September 1998 Pauline Retailer-Self-employed-2-3-Daily-use Good Activities October 1998 Regine Artisan-Employees November 1998 Wasia Retailer-Self-emoloved-Occasional Good Activities October 1999

Table 7.2 Interviewees and Strategy Type

7.3 Enabling Conditions and Accumulation Potential To examine the relationship between enabling conditions and growth in financial capital, the difference between median starting and current capital is computed for the four strategy groupings (Table 7.3). This reveals that strategies for which Substantial Mobility is of key importance have higher growth rates in financial capital. This substantiates the CHAID analysis, discussed in chapter 6, which identified Mobility as the most important Enabling Condition.

Median Median Starting C urrent Growth in Accumulation Enabling Conditions Capital Capital ra n ita l rnnrinnnm

Limited Mobility/Limited Assets S16.00 S20.00 S4.00 Survival

Low Stability Limited Mobility/Substantial Assets 2100 30.00 8.00

Substantial Mobility/Limited Assets 10.00 24.00 14.00 High Stability

100.00 140.00 40.00 Sumlus

Table 7.3 Financial Capital Growth Relative to Enabling Conditions

117 The importance of Mobility is further evident by comparing the Limited Mobility/ Substantial Assets m i Substantial Mobility/Limited Assets groupings (Table 7.3). Despite having a significantly lower median starting capital, the Substantial Mobility/Limited Assets strategy grouping has almost twice the growth in financial capital. This suggests that a lack of .<4sse/s can be compensated for by accessing suppliers outside Porto Novo. It appears much more difficult, however, to counteract limited Mobility through a high level of .^ss«rs. This finding is corroborated by the high level of growth for the Substantial Mobility/Substantial Assets grouping. Moreover, the increasing levels of growth in financial capital from one strategy grouping to the next clearly indicates that livelihood strategies do not fall neatly in two classes (e.g petty traders/mamabenz or survival/surplus). Rather there is a range of possible outcomes are better expressed as a continuum. In looking at livelihood strategies included in each grouping (Table 7.1) a pattern is evident in economic activities comprising each configuration^^ Strategies in the Limited Mobility/Limited Assets group are comprised primarily of one or more daily-use good activities. These economic enterprises have relatively low profits, but offer stable income generation potential (Chapter 5). The main reason for considering strategies associated with Limited Mobility/Limited Assets survival oriented is that limited growth in financial capital suggests restricted opportunities to acquire savings. Indeed the median savings among this group is only S1.20 a week. In the event of a crisis such as a medical emergency, women using strategies in this group will not have means to cope, and might well draw on what little capital they have accumulated. Strategies included in the Limited Mobility/Limited Assets, also tend to comprise multiple daily-use good activities, txceçi Artisan-Self-employed, which combines artisan activity with trade in daily-use goods. The higher level of assets among this group, especially use of unpaid family labor, enables women using strategies emerging finm conditions of Limited Mobility/Limited Assets to manage numerous activities simultaneously. While women concentrate on one activity, their children focus on others.

See Table 5.5 for a description of activities included in each strategy type.

118 Table 7.4 indicates the time allocation between activities by percentage of women using strategies in the Limited Mobility/Limited Assets and Limited Mobility/Substantial Assets groups. It is apparent that women 'NiOtt Limited Assets are more likely to engage in second and third enterprises for only a portion of the day (mornings or evenings). On the other hand, those who have children's unpaid labor as zn. Assets are able to manage multiple endeavors all day long. This in part explains the somewhat higher level of growth in financial capital among women with Limited Mobility and Substantial Assets. Slightly higher earnings capacity enables women within this group to likewise have moderately higher savings and therefore greater insurance against loss of capital in the event of a financial crisis. This is the reason for considering strategies in this group low stability in terms of accumulation potential.

Limited Mobility/ Limited Mobility/ Limited Assets Limited Assets

Activity Activity Activity Activity Activity Activity Time Spent 1 2 3 1 2 3

All Day 92.0 673 45.8 92.6 91.0 88.3

Mornings Only 4.2 16.7 25.3 4.5 0.0 0.0

Evenings Only ■ 3.8 12.6 25.3 2.9 5.4 8.8

Evenings/Weekends Only 0.0 3.4 3.6 0.0 1.8 2.9

T.ate Niohts Onlv _ O iL ____ QSL _____&JL 0 0 18 ------12 .

Total 100.0 100.0 100.0 100.0 100.0 100.0

Table 7.4 Time Allocated to each Activity by Percentage of Women

119 Whereas strategies va Limited Mobility groupings tend to emphasize daily-use goods, those with Substantial Mobility involve trade in as-needed and occasional-use goods. Women using Substantial Mobility/Limited Assets strategies combine trade in daily-use goods, which provide a regular income, and occasional/as-needed-use goods, which provide regular capital infusion. Given the possibility of consistent capital growth, these strategies are classified as high stability since women using them are unlikely to experience capital loss when faced with financial crisis. Strategies in the Substantial Mobility/Substantial Assets group associate wage or artisanal activities with trade in as-needed/occasional goods. Alternatively, they involve trade in one or more occasional/as-needed-use good activities. In the case of the former, a fairly high stable income is generated &om wage or artisanal activity, whereas commercial activities provide regular growth in capital, but also augment income. Where strategies in this group only comprise trade activities, enterprises are generally fairly large (e.g. wholesaling). Here scale of enterprises facilitates high level of capital growth therefore warranting identification as strategies of surplus on the accumulation continuum. By comparing median starting and current capital, it is possible to more fully appreciate the importance of Mobility. Strategies that involve accessing suppliers outside Porto Novo are clearly more profitable than those that do not, irrespective of the level of Assets. Moreover, varying levels of growth in financial capital clearly demonstrate that dichotomous categorizations of livelihood strategies (e.g. survival/surplus, petty trader/mamabenz) are less appropriate than a continuum for expressing the flill range of accumulation potential possibilities. Finally, examining activities that comprise each strategy within Enabling Condition groupings not only facilitates understanding level of accumulation, but also suggests a relationship between trade in particular kinds of goods and Mobility. This is explored further in the next section, where women's perceptions on the importance ofMobility is considered.

120 7.4 The Meaning of Mobility for Women in Porto Novo Women whose strategies are characterized by Substantial Mobility access suppliers outside Porto Novo (Chapter 6). These strategies tend to be more profitable because of types of goods sold. This is also evident in women's reasons for engaging in these activities. Although a majority (33%) of women stated it was a matter of "heritage", something they learned fi-om their mothers, 17% gave "profitability as the reason. Further, a small, but significant group (3.4%) indicated they followed someone else in the activity, which women usually only do if they perceive the other person's enterprise as profitable. For example one women using a Retailer-As-needed Goods strategy said "my sister-in-law sold used clothes and that earned her a lot, so that's why I also started." The question remains why is it important to have Mobility for profitable trade in as-needed and occasional goods. Interview data reveal that the high profit margin, associated with trade in these commodities can only be maintained by going to the source of those goods. This stems, in part, fi-om the fact that wholesalers of these items are only located in particular places. Where people wholesale such items locally there is mark-up on the goods which, in turn, reduces profit earned finm their sale. Goods which comprise the production and retail activities of women using strategies in the Substantial Mobility/Limited Assets and Substantial Mobility/Substantial Assets groups require that women be mobile. Daily-use goods, such as food and fuel, are readily available through wholesalers in Porto Novo. As- needed goods, such as used clothing and medicines are not as easily available locally. For example, used (western style) clothing, which constitutes many daily wardrobes, is only available wholesale at one market in Cotonou (Misebo or the Dead Jovo Market**). Second-hand clothing and household linens (e.g., curtains, sheets) arrive firom the United States and Europe on a daily basis in large bails. These are bought by wholesalers, who

** Jovo was the slang term for white person. The Market name reflects the belief that no one other than a dead person would have no further use for their clothes and the fact that the clothes all come fi-om Western (White) countries.

121 then take the bails apart, sort clothing by type, and then resell it to other wholesalers or retailers. If one does not know the wholesalers, and where to buy the clothing for retail, it is difficult to engage in used clothing trade. As-needed and occasional-use goods also have a higher profit margin than daily- use goods, but one can only take advantage of this by going to the source of supplies. An interviewee, Delphine, explained why this was the case. She said, one can not engage in particular types of trade without knowing the "secret" for it. When asked what she meant by "secret" of the trade, she explained it has to do with where to purchase particular goods.

Imported food one buys in Cotonou, so the medicines firom Lome one buys in Cotonou, sometimes the Grand Marche, beauty products, those are firom Nigeria, that is the secret... Because the products from Nigeria there, if you go to Cotonou to buy them it is expensive. It is less expensive to buy them here (Porto Novo)... because the people who bring them firom Nigeria to Cotonou pass by here before going to Cotonou. (Interview with Delphine, November 1998).

One would expect the cheapest place to buy goods from Nigeria is Nigeria. If one does not know the best place to buy goods, it is possible to lose a lot of money. Another interviewee, Claire explained why this was the case. Although she sold medicines and beauty supplies firom Nigeria, she bought her supplies firom wholesalers at the Grand Marche in Porto Novo. She recognized, however, that in so doing she was losing money.

The wholesalers, sell to us the retailers at a somewhat cheaper price because we buy in quantity. Still there is the maximum price that people will pay. So, I can oiüy charge the maximum, if I charge more people will just go to the Grand Marche themselves to buy. Wholesalers also sell retail. That is why I watch the prices. As soon as prices start to go up a little on something, I buy as much as I can. Then when the maximum price also goes up, I wUl make more profit. (Interview with Claire, October 1998).

122 Because these as-needed-use goods do not sell all the time, their marginal profitability was somewhat higher than for daily-use goods. As a result women try to get as much profit as possible on every article sold. One way to achieve this is to go to the source of goods to buy supplies. Potential for loss increases the importance of being able to get to the source of supplies, and of having good relationships with wholesalers in that place. These comments reveal that for trade in certain kinds of goods (as-needed and occasional-use). M obility outside Porto Novo is critical. There are two reasons for this. First, large-scale wholesalers for these items are only found in certain markets. Second, purchasing these items fiom local wholesalers results in a loss of profits because there is a limit to the mark-up on these items that will ensure sales. Mobility, therefore, means knowing the "secret" of trade - where particular items are sold - how much they can be marked-up, and having good relationships with wholesalers in those locations. Without Mobility, then, the potential profitability of trade in these goods is reduced.

7.5 Trade-Offs: Mobility vs. MLES Having multiple income generating activities is a means of compensating for Limited Mobility. This is evident in comparing the mean and median number of activities by Enabling Condition group (Table 7.5). Women using livelihood strategies grouped as Limited Mobility engage in a greater number of activities than those with Substantial Mobility. For women with Limited Mobility marginal profit fiom each activity produces sufficient income to enable making ends meet. This is evident in women's answer to the question regarding why they had more than one source of income - the most common of which was "my strategy works because I get a little profit fiom each activity". Despite the fact that women consistently gave similar answers for why they use MLES, variation in the importance of particular answers is consistent with their degree o f Mobility (see Table 7.6). For women with Limited Mobility, "making ends meet" and "increasing income" were the most important, while "diversification" and "insufficient salary" are also critical to women with Substantial Mobility. "Diversification" is especially

123 S tandard Enabling Condition G rouoines N Mean Deviation M edian

Limited Mabilitv/Limiled Assets 213 2 J 5 1.12 2.00

Umileti Mobilitv/Substantial Assets 96 235 0.97 2.00

Substantial Mobility/Limited Assets 88 1.68 1.12 1.00

Substantial Mobility/Substantial Assets 125 1.60 0.79 1.00

Total 522 2.06 1.08 2.00

Table 7.5 Mean and Median Number of Economic Activities by Enabling Condition Group

Percentage of Women by Enabling Condition G ro u p Litnited Limited Substantial Substantial Mobility/ Mobility/ Mobility/ Mobility/ Why use MLES? Limited Substantial Limited Substantial T otal Assets Assets Assets Assets Sam ple

m ak in g ends m eet 32.9 3 3 J 31.0 222 18.8

increase income 30.9 28.2 27.6 33.9 18.4

divereification 15.8 19.2 27.6 19.6 l l . l

salnrv ineiiffiripiit 7.6 '3 3.4 2.7

Table 7.6 Women's Responses Regarding use of MLES

124 significant to artisans because they frequently earn less than a regular income from their craft. As a result, they engage in additional activities. 7.6 Conclusions This chapter examines the accumulation potential of various livelihood strategies relative to the Enabling Conditions by which they are characterized. Calculation of growth in capital verified the importance o f Mobility as an Enabling Condition. Strategies that require Mobility are more profitable, offering greater opportunity for accumulation. In contrast, women using strategies reflecting Limited Mobility compensate for relatively low profit activities by engaging in a great number of enterprises. Further, outcomes in capital growth reveal that strategies cannot be dichotomously grouped,;rather, they are better expressed along a continuum ranging from survival (limited growth), to stability (some growth), to surplus (significant growth). Women's perceptions regarding the importance of mobility provides an explanation for differences in accumulation potential for different strategy configurations. Knowing the "secret" of trade means knowing where to obtain particular goods. This is especially critical for as-needed and occasional-use goods. Whereas daily-use goods are readily available within Porto Novo, more profitable items are not. Moreover, the high cost of obtaining such items coupled with the limits to mark-up indicates that high potential profits can only be assured by obtaining them from the source. Given that as- needed and occasional-use goods are only available through wholesalers in particular locations, mobility is required. Having now explained the importance of mobility to the operationalization of particular kinds of livelihood strategies, the next two chapters will explore how intersections in Dimensions ofDifference manifest in Structures o f Constraint produce Localized Enabling Conditions. In other words the reasons for women's mobility or lack thereof will be examined relative to their life-circumstances. Chapter 8 addresses explores the lives of women using strategies identified as having limited Mobility and chapter 9 those that require substantial Mobility.

125 CHAPTERS

LIMITED MOBILITY: HOW DO DIMENSIONS OF DIFFERENCE PRODUCE LOCALIZED ENABLING CONDITIONS^

8.1 Introduction To this point, this research has focused on identifying patterns of women’s strategy configurations (Chapter 5) and Enabling Conditions'^ from which they arise (Chapter 6). Having differentiated thirteen strategy types and two Enabling Conditions, Assets and Mobility"^, Chapter 7 explored the accumulation potential of strategies relative to Enabling Conditions. Evidence from Chapters 6 and 7 indicate that Mobility is the critical factor shaping women’s opportunities to earn a living and increase capital. Therefore, the importance of Mobility was explored in Chapter 7. The question remains, however, what determines women’s degree of Mobility and access to Assets’! The conceptual framework (Chapter 3) specified that intersections among Dimensions o f Difference (Ethnicity, Gender, Class, Life-cycle Phase) manifest in Structures o f Constraint, categorized broadly as Assets, Formal Rules, Cultural Norms, and Preferences. Interactions among Structures o f Constraint (localized social institutions), in turn, produce Localized Enabling Conditions, which sh ^e strategy configuration. Attention now turns to understanding how variation in life circumstances, as represented by Dimensions ofDifference and Structures o f Constraint, limit MobUity for

Important theoretical concepts are highlighted in italics. ^ Specific Enabling Conditions are highlighted in bold italics.

1 2 6 women in Porto Novo and shape their access to Assets. Through examination of the relationship between these sets of factors, this chapter demonstrates that Enabling Conditions, specifically Limited Mobility, arise from particular intersections among Dimensions o f Difference and their localized manifestations, Structures o f Constraint. Analysis proceeds in three steps. First, particular themes representing reasons for women’s immobility relative are identified through descriptive statistics (frequencies and contingency tables). Second, women’s life-histories illustrate the processes through which intersections among Dimensions o fDifference produce conditions o î Limited MobUity that restrict women’s economic activity choices and, hence, their opportunities in strategy configuration. The final section explores the sources and meaning of disparate levels of/fssers experienced by women using strategies within the Limited Mobility/ Limited Assets and Limited Mobility/Substantial Assets groups'’^ In keeping with the principal objectives of this project, which are to understand why women in Porto Novo use different livelihood strategies, this chapter focuses on the basic patterns in intersections among Dimensions o fDifference. While social, economic and geographic processes are clearly important, the primary goal here is to understand fundamental patterns of difference among women. Thus, only dominant processes are examined.

8.2 Data and Analysis As in Chapter 7, both quantitative and qualitative data are used for this portion of the study. Descriptive statistics (frequencies and contingency tables) based on survey data indicate general patterns of association between women’s life circumstances and Mobility. Content analysis of both open-ended survey questions and life/economic histories facilitate understanding how various intersections among Dimensions o f Difference, manifest in Structures o f Constraint, shape women’s choice of strategy configuration with respect to Enabling Conditions.

For a complete list of all strategies within each group see Table 7.1.

127 8.2.1 Data As earlier, data are from 522 in-depth surveys. In part, analysis centers on the same set of Cultural Norms and Assets variables identified in Chapter 6^*. These variables are used because CHAID analysis revealed that Cultural Norm variables do not directly produce Enabling Conditions (Chapter 6). This was interpreted to indicate that strategy choice is shaped by a range of Cultural Norm processes, resulting from interactions among Dimensions o f Difference and Structures o f Constraint. Therefore, the same set of Cultural Norm variables used in the CHAID analysis are considered in assessing how Dimensions o fDifference impact Enabling Conditions. Additional variables regarding respondents’ upbringing are also considered because they influence both women’s choice of economic activities and ability of parents to provide education, training and starting capital'*’. These include mothers’ occupation, raised the respondent (raised by), and who paid for education/training. These variables focus explicitly on why women may or may not have access to Substantial Assets. Open-ended survey questions provide insight into why women make particular choices in creating livelihood strategies. This includes explanation for why they choose economic activities and enterprise locations. Fifteen life/economic histories (in-depth interviews) provide further information as to how specific intersections among Dimensions o fDifference, manifesting in Structures o f Constraint, influence women’s livelihood strategy choices relative to Enabling Conditions. Women’s life/economic histories provide a means of understanding interactions between familial responsibilities relative to residential arrangement/location and choices in economic activities, enterprise and supplier location. Details concerning family background and upbringing

■** See Table 6.1 for a complete list of variables and chapter 4 for discussion of how each variable is thought to shape strategy configuration. Variable names are highlighted in bold.

128 contribute to understanding women’s access to Assets, including financial and human capitaP. 8.2.2 Analysis Given that both quantitative and qualitative data are used in this chapter, two separate, but interrelated analyses are required. Frequencies are run for relevant Cultural Norm variables to identify basic characteristics o(Enabling Condition groupings {Limited Mobility/Limited Assets, Limited Mobility/Substantial Assets), Frequencies for both groups together are then evaluated relative to the whole sample. This facilitates identifying the most significant variable categories and ways in which groups diverge substantially from the sample as a whole. Where variables are strongly interrelated (e.g., care-taking burden and residential arrangement) contingency tables are also used to determine patterns of association among variables. Through analyses described above, themes representing particular interactions among Dimensions ofDifference are identified for further investigation using content analysis of open-ended questions and interview texts. As in chapter 7, content analysis provides insight into women’s perspectives on the ways in which their life circumstances shape economic activity choices (Silverman 1993). The whole analysis aids understanding how various patterns in Cultural Norm variables are associated with Enabling Conditions, specifically. Limited Mobiiity^K

“ See Table 7.2 for a complete list of interviewees. " It should be noted that while this analysis centers on women’s current situation, historical orientation of interviews provide insight into the dynamism of women’s livelihood strategies. Thus, conditions of constraint and opportunity {Enabling Conditions) should not be seen as fixed or immutable; rather, they are a snapshot of women’s lives in Porto Novo at one particular moment in time.

129 8.3 Explaining Limited Mobility 8.3.1 Introduction: Identifying Themes Women who have Limited Mobility tend to use livelihood strategies consisting of self-employed trade activities in daily-use goods” . These activities primarily involve manufacture and/or distribution of food items that constitute the staple diet of Béninois in this region including canned goods (e.g., tomato paste used in sauce preparation), fruit, and various types of grains (com, rice, and gari” ). All strategies, but one {Retailer-Self-employed-I-Daily-use Activity), comprise a combination of artisanal, production and/or retail oriented activities with women making and/or buying goods in local markets, then reselling them from their homes, the market, or ambulant sales throughout the city. The marginal profit from each enterprise produces sufficient income in combination to permit making ends meet, resulting in Limited Mobility strategies oi survival and low stability*. The main question addressed here is; why do these women \àc)c MobUity outside Porto Novo? A second question concerns differences among women with Limited Mobility in access to Assets. Most women using strategies within the two Enabling Condition groupings - Limited Mobility/Limited Ass^s and Limited Mobility/Substantial Assets - are Goun (Table 8.1). Ethnicity, then, is constant for women in these groups. As shown in sections two and three of this chapter. Class effects apply primarily to Assets. Therefore, Life-cycle Phase, which is shown to be most important, is discussed first.

” See Table 7.1 for a complete list of strategy configurations by Enabling Condition group. ” Gari is a grain made from manioc (sometimes called cassava), which is used in many staple foods. It is arguably the most important staple eaten in that part of Benin. ^ Accumulation potential continuum placement is highlighted in italics.

130 Women with Total Ethnicity Limited Mobility Sample

Other 1.9 2.9

Yoruba 14.9 199

Goun 67.6 63.0

Fon 10.4 9.2

M ina 5.2 5.0

Total 100.0 100.0

Table 8.1 Percentage of with Limited Mobility Women by Ethnicity

Life-cycle Phase, one Dimension o f Difference, is reflected in operant Cultural Norms, a Structure o f Constraint. Three patterns reflecting changes in Cultural Norms as women move through the Life-cycle explain Limited Mobility. Young, newly married, Goun, women live in extended-family settings with in-laws on the urban-fringe. This indicates a customary value system, in which young women lack autonom/^, especially in reference to economic activities. Middle-aged women may gain autonomy by moving to their own homes (female-headed and nuclear households) in the suburbs. In so doing they lose family assistance, which still limits their Mobility. Finally, older women are often isolated because they continue to live on their own or in their family’s homes in the city center. Thus, they are still left without family assistance, and, hence, still do not have Mobility outside Porto Novo. This corresponds to residential arrangements, which are associated with residential locations within Porto Novo (Chapter 4). Residential arrangement and location, represent the localization of Cultural Norms associated with Life-cycle Phase (Tables 8.2 and 8.3).

” Themes are identified in bold.

131 Percent]ige of Women with Limited Mfobility by Age Range Residential Total Arrangement <30 31-50 51+ Sample

Nuclear w/ Husband 22.7 36.5 4.9 26.4

Female-Headed 9.1 17.9 36.6 18.8

Family w/ Husband 1.8 0.0 0.0 1.1

Family w/ o Husband 22.7 17.9 36.6 24.5

In-laws w/ Husband 39.1 17.9 73 21.6

in-laws w/o Husband 4.5 9.6 14.6 7.5

Total 100.0 100.0 100.0 100.0

Table 8.2 Percentage of Women with Limited Mobility by Age Range and Residential Arrangement

Percentage of Worn* en with Limited MobiiUy by A ge Range Residential Total Location <30 31-50 51+ Sample

City Center - Market 10.1 12.3 48.7 18.4

City Center - No Market 6.4 11.0 5.1 9.4

Suburb - M arket 9.2 16.2 lO J 12.8

Suburb - No Market 5.5 13.6 7.7 113

Urban-fringe - Market 51.4 35.7 25.6 35.6

Urban-fringe-No Market 17.4 11.0 2.6 10.6

Total 100.0 99.8 100.0 973

Table 8 J Percentage of Women with Limited Mobility by Age Range and Residential Location

132 Patterns in Cultural Norms associated with Life-cycle Phase are evident in women’s household characteristics (Table 8.4). The mechanisms by which these conditions emerge are discussed below.

Percentage of Women with Limited Mobiiity by Age Group

Characteristic <30 31-50 51+ Total Sample (35.6) (50.5) (13.3)

Childcare - mother 39.1 32.1 14.6 27.4 father 3.6 2.6 0.0 2.3 relative 36.4 293 4.9 27.1 maid 2.7 3.2 0.0 5.4 children 4.5 25.0 36.6 20.7

Household Head - herself 20.9 35.3 75.6 37.2 husband 62.7 59.0 19.5 52.5 shared 0.9 1.3 0.0 0.8 parent/in law 15.4 0.6 2.4 9.0

Husband in Household - yes 63.6 54.5 12.2 49.2 no 29.1 44.2 87.8 46.0

Years Married - < 10 60.0 12.1 0.0 29.8 11-20 23.6 37.8 4.9 28.2 20+ 0.9 3 7 J 78.0 26.6

Marital Status - not married 13 3.2 4.9 7.1 only wife 263 26.9 14.7 26.7 1“ wife 51.8 50.7 63.4 473 2™* wife 14.5 19.3 17.0 18.8

Children's Age Range'*- < 10 69.1 22.4 0.0 33.0 5-15 10.0 25.6 7.3 16.3 0-20 9.1 44.9 56.1 36.0 20+ 0.0 2.6 4.9 2.1

Table 8.4 Household Characteristics of Women with Limited Mobility

Although categories overlap, women were only coded once depending on the actual age-range of children in their households.

133 8.3.2 Life-cycle Effects Limiting Mobiiity: Lack ofAutonomy for Young Women One group who have Limited Mobility are young women (under 30) (Table 8.4). The majori^ live in extended family settings, primarily with in-laws on the urban-fringe. Cultural Norms among Goun living in more rural-like settings, such as this, dictate that women stay home until they have several children, who are school-age or older. The majority of these women have been married less than ten years and their children are likewise less than ten years old. While some women have relatives care for their children while they work, most are primary care-givers. Even where someone else takes care of their children, however, women do not necessarily have the freedom to be mobile outside Porto Novo. Lack of autonomy is most clearly reflected in their identification of their husbands or parents/in-laws as household heads. Women’s lack of autonomy is also evident in responses to survey questions regarding reasons for locating enterprises in a particular place, usually at home or nearby market. Answers indicate they are told to do so by their husbands and/or in-laws. For example, one 30-year-old woman {Retailer-Self employed-l-Daily-use Activity) stated "it was my husband who decided I should stay home." Another woman {Retailer-Self employed-3+-Daily-use Activities) said, "I used to sell at my parent’s house, but my husband forbade me to walk around to sell (ambulant sales) and that’s why I sell only in front of the house." This prohibition against having economic activities away from home, not only limits where women can trade, but also how far they can travel even on an occasional basis, such as to purchase supplies. An interviewee whose life circumstances fit this pattern explained why women living in this type of setting lack autonomy. Marguerite {Retailer-Self-employed-l-Daily- use Activity), a 28-year-old, was raised in Cote d’Ivoire by her mother, a French citizen and nurse, and her father, a teacher. She went to school, and at the time her parents divorced, she followed her father back to his home, Porto Novo, where she continued her studies through high school. Because her father was a teacher, her education was free. She decided to leave school at 19, before passing her exams, to marry her husband, Cedric, against the wishes of her parents. According to Marguerite, her father’s

134 objections stemmed from his belief that Cedric’s family was too "rural" and that they would not allow her to pursue a career as a journalist as he hoped. In fact, his concerns were bom out. At the time of the interview, she had been married eight years and had four children; three girls aged 7,5, and four months and one son, aged three. After six years of not working she decided she wanted to begin trading. Although her husband was always generous with her, she wanted to buy her own things and also found it "shameful" that she could not give her children money when they asked for it, having to get it &om her husband first. Both her father-in-law (the old man) and her husband, however, were very opposed to the idea of her working. With respect to her father-in-law she recounted the following exchange on the subject:

The fact is that people will think that I am not at ease, that they don’t give me money here, and I start selling. That is what the old man thought and then refused. He refused when I started selling alcohoP and kerosene. He opposed energetically. He asked me what I lack and want to sell, to tell him what I need and he’ll give me money to satisfy my wants. When he said that, I said no, I want to learn how to earn my own life. That is all that I told him and then he said it was not good what I was thinking about doing, that it wasn’t yet 10 years or IS years since I came here and I wanted to begin commercial activities. It is not good, wait. (Interview with Marguerite, September 1998)

She had a similar exchange with her husband, but was ultimately able to convince him:

He said no. He opposed. He asked the reasons why I want to trade and I told him that it is to have my own money. And he asked if the money he gives me is not enough, I said no it is not, that I am not satisfied. The children ask you, and I ask you also, I prefer to go about other activities. When I told him that, he agreed and then I started with alcohol. (Interview with Marguerite, September 1998)

” Alcohol refers to a locally produced variety called Sodabi, which is distilled from palm wine.

135 She goes on to explain that her father-in-law’s concern about waiting 10 or IS years after marriage has to do with Cultural Norms among the Goun about when it is appropriate for married woman to leave the family compound.

Normally after, ....before five years a woman mustn’t do anything. Before three children, she does nothing. After the third child that you can begin going about your business. You can even go to your parents. Otherwise, if you don’t have three children, you can’t even go out or visit your parents, friends, no . . . (Interview witli Marguerite September 1998)

So, for the first five years of marriage women are not allowed to leave the family compound at all even to visit their own families. After that, for the next five to ten years, they may be somewhat more mobile, but only with a chaperone from their husband’s family. Cedric’s aunt went with Marguerite the first time she visited her father at the end of the five-year period. Women are not supposed to start independent economic activities until a ten to fifteen waiting period is over. When asked if this custom is particular to the family or more wide spread she said that it is common throughout Benin, even in Abomey where her father’s family is from^*. Two Cultural Norms are at odds in Marguerite’s account of how she began to trade. On the one hand, there is the interdiction against women even leaving their homes until they’ve had at least three children or engaging in trade for ten to fifteen years after marriage. The other norm, expressed by Marguerite in her desire to earn her own money, is the expectation that women should not rely on their husbands for everything. This sentiment was also voiced by many other women who lack Mobility when asked why they began their own independent economic activities. For example, one woman (Retailer-Self-employed-I-Daily-use Activity) said that "the woman without a craft, should sell something, if not it’s bad." Another vfomaii(RetailerSelf-employed-3-^~

The dominant ethnic group in Abomey is the Fon, which is closely related to the Goun. In fact, Goun may be thought of as a sub-group of the Fon. Thus, her experience among her husband’s Goun family would not be expected to be all that different fixim what would be common among her father’s people in Abomey.

136 Daily-use Activities) stated that "when one marries one must do something." Yet another (Retailer-Self-employed-Daily-use Activity) said that "a woman must not stay without doing a trade." For women in Benin, earing an income is part of their reproductive responsibilities. At a minimum they should be able to provide for their own needs. Thus, Marguerite’s pride in earning her own income is at odds with her husband’s and father-in- law’s desire to take care of her. Marguerite explained her ability to overcome her husband’s objections as a function of her own strength of conviction that it was the right thing to do and because their marriage was not an official one. He had not paid a dowery yet and this restricted his ability to control her.

...if he pays the dowery it means that she mustn’t move, she mustn’t say no or yes for what he [her husband] does. It means that the man has right of way in the household if he gives the dowery. (Interview with Marguerite, September 1998)

By locating her economic activities at home and purchasing her merchandise 6:0m local suppliers. Marguerite is able to mediate her own desires with those of her husband’s and his family. Her husband even helped her establish her business by giving her a small amoimt of starting capital. He also helps her negotiate these conflicting Cultural Norm by obtaining supplies for her. In this way, she can have her own source of income, satisfying her needs, without overly embarrassing the family by working or traveling outside the family compound. Thus, while Marguerite clearly has agency in her economic activities, her choices are conditioned by Cultural Norm related to herljj^- cyc/e Phase, which are socially and geographically situated within a particular residential arrangement and location. 8.3.3 Life-cycle Effects Limiting Mobility: Loss o f Family Assistance for Middie-aged Women

A second group who experience Limited Mobility are middle-aged women (31- 50). As with the young women, the majority live on the urban-fringe (Table 8.3). The proportion, however, is somewhat less than that for the total sample, hi contrast women

137 in this age group live in suburban and city locations at a higher percentage than women in the total sample. Suburbs are an area within Porto Novo, where nuclear and female­ headed households are prevalent (Chapter 4). While some of these women still live with in-laws, most reside in nuclear family settings or on their own with their children (Table 8.2). This setting is often associated with greater autonomy in decision-making. Lack of assistance &om extended family, coupled with the presence of school-age children in the household, however, results in zoriiymzà. Limited Mobility. Relevant characteristics for these women are detailed in Table 8.4. One significant difference between middle-aged women and their younger counterparts, is the age of their children. Children of these women are school-age or older, enabling women to use their unpaid labor in the management of economic activities. This allows women to simultaneously maintain several enterprises (Chapter 7). Nonetheless, while children are old enough to help with economic activities and housework, they are still young enough to require supervision. This limits women’s Mobility outside of Porto Novo. Another important difference among young and middle-aged women is their marital status. Wliile newly married young women tend to be the only wife of their husbands, who are usually living in the household, middle-aged women are more likely to have polygynous marriages. Often, as the first wife they are left to manage their households alone, when their husbands marry a second wife with whom they live in another location. This pattern is evident in that fewer middle-aged women live with their husbands. In the past a man would live with all his wives in the same compound, allowing women to share household and care-taking responsibilities. Today, however, men are much more likely to essentially abandon their first wives, leaving them to raise their children alone. This accounts for the high proportion of female-headed households among this group. These conditions are also evident in the share of household budgets for which women are responsible. Among women with Limited Mobility/Limited Assets 39.8% contribute 75% or more to the household budget. For women with Limited Mobility/Substantial Assets the percentage is 47.9%.

138 That these conditions reduce mobility is apparent in women’s answers regarding why they locate their activities at home. Equally evident in their answers are the ways in which they use older children to compensate for their own immobility. One young producer, who lives with her husband in a nuclear family setting, stated, "I just had a baby, so I don’t go out, but I give a quantity to the children to sell in the neighborhood." These circumstances impact women’s Mobility in obtaining supplies as well as in the exercise of their economic activities. One retailer stated that her activities are located at home because she "can’t leave the house because [she has] small children." This woman also does not go to the market for supplies. Instead, she uses the unpaid labor of her younger sister in this capacity. "My sister goes every day to the market to buy supplies. I am in front of the house all the time and my sister sells ambulant for me." As with young women, lack of mobility for the middle-aged reflects localized Cultural Norms reflecting Life-cycle Phase. Whereas young women lack autonomy because they live with husbands and in-laws, these somewhat older women lack assistance. Though use of their children’s unpaid labor enables them to manage a greater number of activities simultaneously, it does not permit them to travel outside Porto Novo. This reflects that women in these circumstances live in nuclear or female-headed households. While Life-cycle Phase is still the dominant Dimension o fDifference at work here, it impacts women differently at this life-stage. This is evident women residential arrangement and location reflecting the social and geographic situatedness of Cultural Norms. 8.3.4 Life-cycle Effects on Limiting Mobility: Isolation among Older Women Although most women in this group fit the patterns described above, a few (13.3%) are somewhat older (50-70+). The majority live in the city center in their family’s home. This is not necessarily an extended-family setting, however. Family homes in the city center are often the original houses of particular kinship groups, which are frequently left to the poorest family members who do not have sufflcient means to construct their own homes (Chapter 4). Older women living in this context, therefore, generally live alone. Even where children reside with their mothers, the majority are over

139 twenty and living their own lives. Moreover, most of these women do not live with their husbands. Lack of household responsibility suggests that older women would have considerably more autonomy and therefore Mobility. Yet, they also, do not travel outside Porto Novo. So, the question remains, why was this the case? Two answers emerge when examining at open-ended survey questions and interviews, household security and frailty due to age. 8.3.4.1 Household Seatrity Issues People in Porto Novo often expressed fear of leaving their homes unattended due to the possibility of robbery if no one is home. This is a particular concern for women living on their own with grown children who have their own activities. This anxiety was most clearly expressed by a retailer of daily-use goods, "there are too many robbers in the neighborhood. I cannot go out." Similarly, one 55-year-old {Producer-Self-employed- 2+-Daily-use Activities), when asked why she chose to sell in front of her house said: "I don’t want to leave the house, because I am frequently alone." Women in these circumstances may still have children at home, but they are engaged in their own activities. As one woman {Producer-Self-employed-2+-Daily-use Activities) said "the girls used to go [sell] in the city for me, but now they do their own activities." As result these women are on their own at home, particularly if they live are heads of their own households. As another woman {Retaiier-Self-employed-3+-Daily-use Activities) whose children were all over 20 with their own activities said, "I can’t leave whenever I want. I am obliged to find a trade that does not make me go out all the time." This experience was best expressed by Christian (Retailer-Self-employed-3+- Daily-use Activities), a 61-year-old interviewee, who, like Marguerite, lives on the urban- fringe. Unlike Marguerite, however, she is the head of her household consisting of her five grown children and two young adopted daughters. Although at one time she worked as a seamstress, her eyesight is too poor to continue her craft. Her livelihood strategy, therefore, consists of three retail activities: selling finit, firewood, and school supplies. When asked why she chose these activities and locates them at home she said:

140 I sell the mangos because they were already in my concession (courtyard). I sell the school supplies because I am across fiom the school and the firewood feeds me every day. I work here because I live here and I can’t have an activity that makes me go out. Since the children are grown and have their own activities, someone must be at home to look after the house. That is me. (Interview with Christian September 1998)

The fruit is readily available during the season and firewood is brought to the house by wholesalers. She, therefore, does not go out to get these goods. She relies on her children to purchase school supplies for her as their activities frequently take them to Cotonou. Mobility is hence, not required for her activities. Here again, Life-cycle Phase plays a significant role in shaping women’s opportunities. Older women are more likely to live alone. Given the fear of leaving a house unattended, when women’s children have their own activities, no one else is left to stay at home except the woman herself. Thus, just at the moment in her life when she might have more Mobility because she has fewer familial responsibilities, in fact her situation does not change. 83.4.2 Frailty due to Age Another explanation older women gave for their lack o f Mobility is that they are too old. Although some older women engage in ambulant sales and even travel to Cotonou to buy supplies, older women within this group, gave age as the reason for circumscribing their activities. For example, one 70-year-old (Retailer-Self-employed-l- Daily-use Activity) stated that "before I used to travel to sell, but now I have problems with my sight." Two others {Retailer-Self-employed-l-Daily-use Activity and Retailer- Self-employed-3+-Daily-use Activities), a 60-year-old and a 72-year-old expressed similar sentiments, saying "[bjecause of my age I am obliged to sell at the house. 1 cannot go far." This lack o f Mobility in their economic activities applies equally to where they purchase merchandise. All these women buy supplies in local markets, often ones right in their neighborhood even if it was not a very large market. Consequently, they are

141 probably paying higher prices for goods relative to women who travel at least to one of the two large markets in Porto Novo, Ouando Market or the Grande Marche. Life-cycle Phase not only limits Mobility for these women, but also ability to earn a profit. 8.3.5 Conclusions to Life-cycle Effects Limiting Mobility In all three situations discussed above women’s Mobility is restricted by Cultural Harms associated with Life-cycle Phase. Variation is localized according to residential arrangement and location. For young women living with in-laws in the urban-fringe Cultural norms require they do not travel far from home, indicating a lack of autonomy. Middle-aged women gain greater independence by living in female-headed and nuclear households. In so doing, however, they make a trade-off in family assistance. Difficulties associated with this are compounded by Cultural norms that make polygyny acceptable, which often results in abandonment by their husbands. Lack of assistance coupled with a high degree of care-taking responsibilities, results in relative immobility. Older women frequently become isolated living on their own or with grown children. Although they no longer have a care-taking responsibilities, they still lack Mobility either for fear of their households security or because they feel too frail for travel. In all three instances women’s life circumstances shape their economic activity choices and hence, their strategy configurations. Their decision-making reflects "purposeful choices" in which their agency is conditioned by socially and geographically situated Cultural norms associated mHa. Life-cycle Phase, which result xa. Limited Mobility. 8.3.6 Class effects Limiting Mobility Another explanation for lack o f Mobility, irrespective o f Life-cycle Phase, is lack of merchandise. This reflects socioeconomic Class, another Dimension o f Difference. For example, one 42-year-old {Retailer-Self-employed-3-^-Daily-use Activities), stated that "it is because I don’t yet have a lot of money that I stay in place." Another 46-year- old (Producer-Self-employed-2-i~-Daily-use Activities), said that "I don’t sell anything important (large) so I stay at the house." Likewise, a 26-year-old, {Retailer-Self­ employed-1-Daily-use Activity) when asked why she located her activities at the house said that "I don’t have enough means. What I do is just to find something to eat.”

142 All women who expressed reasons such as these for home-based activities, also purchased supplies in local markets, usually Ouando Market or the Grande Marche. For these women it does not make sense to travel outside of Porto Novo to buy merchandise. They all sell daily-use goods which are available &om wholesalers in these local markets. Given their restricted capital, whatever small savings in price they might receive by going to Cotonou or Nigeria would be offset by the cost of transportation to those locations. This reason for lack oiM obility reflects socioeconomic Class, manifest in^^^ser distributions, a Structure o f Constraint, The question therefore remains, why do these women lackXssers?

8.4 Asset Differentials among Women with Limited Mobility 8.4.1 Introduction: Identifying Themes While all women grouped as Limited MobUity/Limited Assets and Limited Mobility/Substantial Assets lack Mobility, they experience disparate Asset Distributions. Human capital is the distinguishing characteristic among these women vis-a-vis access to Assets. This primarily reflects Class distinctions, which are evident in family background. Women with Substantial Assets have some form of human capital, such as education, upon which to draw in the creation of livelihood strategies. For users of family-oriented strategies {Producer-Family, Retailer-Family) human capital consists of children’s labor. In contrast. Self-employed Artisans have training in some craft. There is no distinction among women mth Limited and Substantial Assets, however, in access to financial capital. ¥ot Self-employed Artisans this is especially critical as it prevents them from opening a workshop, for which certification is required. This reflects the impact of Formal Rules, not only reducing earning capacity, but also results in their working outside the legal fiamework. Limited Assets in the form of financial capital in some instances also reflects husband’s interference and/or abandonment. Asset related characteristics of women in these two groups are depicted in Table 8.5.

143 Percentage of Women with by EiiabUHg ConditloH Group

Characteristic Limited MobiBty/ Limited Mobility^ Total Sample Limited Assets Substantial Assets

Education - none 6 4 J 49.0 51.3 primary 25.8 26.0 26.1 secondary 9.4 25.0 20.7 university 0.5 0.0 1.0

T rade Learned - seamstress 10.8 18.8 12.6 hairdresser 0.5 6 J 3.1 none 83.1 70.8 76.2

Training Paid by - parents I2J 18.8 16.1 guardian 1.9 1.0 1.5 husband 0.9 5.2 1.9 self 0.5 4.2 1.5 not applicable 84.4 70.8 78.2

Employee Relation - no relation 9.0 10.4 15.7 children 1.9 36.5 8.6 family 2.8 21.9 8.5 not applicable 8 6 J 31.3 66.7

Median Starting Capital S16.00 520.00 520.00

Median Current Capital S22.00 530.00 530.00

Raised by - mother only 48.1 54.7 47.7 father only 2.9 2.1 2.3 both parents 12 J 16.8 17.8 other family member 35.1 25.0 20.8 family friend 1.4 1.1 1.4

Mother’s Occupation - petty trader 7 2 J 75.0 68.4 trader 9.4 12.5 15.1 farm er 1.4 3.1 1.9

Starting W ork Age -10-15 28.1 23.9 24.9 16-20 46.7 43.2 41.0 21-25 20.6 25.0 26.0 26+ 4.5 7.9 8.1

Source of Capital - personal savings 34.7 313 32.9 husband 23.5 10.5 19.4 family 22J. 41.0 5 0 J loan/credit 19J 10.6 16.6

Table 8.5 Asset Related Characteristics of Women with Limited Mobiiity

144 8.4.2 Class Effects on Asset Distributions: The Role o f Family Background Women with Limited Assets have virtually no education in relation to those with Substantial Assets. The latter group have either some form of artisanal training or unpaid family labor upon which to draw in configuring livelihood strategies. Having a craft expands the range of economic activities in which women can engage. Likewise, as discussed earlier, having unpaid family labor enables women to increase the number of activities they can maintain simultaneously. In both cases, there is potential for somewhat better income generation as compared to women with Limited Assets. This is evident in a comparison of median current capital. One explanation for this variation Is family background in that parents of women with Limited Assets are unable to provide education/training or starting capital. The relative poverty of these women's families is apparent in that many were raised by someone in their extended family. In many African contexts people foster their children out to wealthier relatives when they cannot support them. While aimts and uncles raise their sibling's children, this does not mean they will provide education, training or starting capital. Further evidence lies the fact that the majority of these women did not obtain starting capital from their families, rather their principle sources were personal savings, their husbands, loans or credit. The relative poverty of these women is also evident in the early age at which they started working. These findings are substantiated in women's reasons for starting their first activity. One woman {Retailer-Self-employed-2-3-Daily-use Activities), who started at age 12, said she did so because "I needed to earn my own money because the person who raised me didn’t take very good care of me." A woman {Retailer-Self-employed-2-3- Daily-use Activities), who was raised by her parents stated that "[they] did not have the means to take care of me. My mother was sick, so I had to manage on my own in order to eat." For these women. Limited Assets reflect the low socioeconomic Class of their families. This is evident in their low levels of starting capital. Lack of financial capital not only reduces women’s choice of economic activity, but impacts their Mobility.

145 8.5.3. Class Effects on Asset Distributions: The Impact o f Formal Rules Women using a Self-employed-Artisan strategy have their own human capital, in the form of artisanal training, upon which to draw in creating livelihood strategies. Lack of financial capital, however, hinders their ability to make effective use of their skills. In this instance Class impacts women’s choice of economic activity, but manifest Asset distributions and Formal Rules. Béninois law requires artisans to be certified in order to open a workshop. This dictates they must be formally trained, have a diploma, and pay taxes regularly all of which require significant financial capital. Many women who use this strategy indicated they worked from home because they could not open a workshop. Three explanations for working at home are evident in responses to open-ended questions regarding activity location and reasons for having multiple enterprises. First, women were not formally trained and therefore could not legally open a workshop. Second, women were in training and/or just finished training, but did not yet have their diplomas. Third, certified women, who had the right to open a workshop, could not afford to do so. Women still in training and/or who had not yet paid for their diplomas indicated they could not open a workshop because they were not certified. For example, one young artisan said that she "is not liberated yet (does not have her diploma), so [she] can’t work anywhere, but the house." In this instance, she only works part-time on evenings and weekends, as she was in training the rest of the time. She uses what she has already learned about braiding as a means of earning the money to continue her training. Another woman indicated that she had two sources of income (hairdressing and selling candy) because she "needed to earn the money to get her diploma." Until she obtained that, she would not be able to practice her craft legally. Other women, who had not been formally trained, also did not have the right to open a workshop. Most women to whom this applies work as hairdressers rather than seamstresses. Many women in Benin cannot afford to go to a salon to have their hair done. As a result women learn to braid hair. This form of hairdressing is practiced by those without a workshop, as it does not require any tools other than a comb. Women

146 who simply learned how to braid hair without formal training have an effective means of earning a living, requiring little or no financial capital. This is evident in one woman’s explanation for why hairdressing was her most important activity. "Because braiding does not need capital, everything I earn for this work is mine." These artisans provided the service to women who do not have time or cannot afford to go to a salon. In this instance, women work &om home or wander through markets offering their services to traders. As one woman who worked from home indicated, "As I have not learned hairdressing, I cannot open a workshop." The research assistant to who did this survey added the following note: "It is against the law [to open a workshop]. It is necessary to have a diploma first." In a few instances, women, who are certified as hairdressers or seamstresses work from home because they cannot afford the equipment and/or rent to open a workshop. For example, one seamstress said, "I rented a place at Catchi” , but because of the cost of rent, I left it." Similarly, a certified hairdresser said she worked form home because "I don’t have the money to open a workshop." Inability to pay for formal training, the certificate and/or the equipment necessary for opening a workshop reflects the relative poverty of these women as indicated by lack of financial capital. This reflects their family background, in that while their families were able to provide education and training they lack the resources necessary for establishing a workshop. In this sense these self-employed artisans are the same as other women with Limited Mobility. This is the situation for interviewee, Gladys {Artisan-Self-employed), a 35-year- old hairdresser, who had just opened her own salon at the time of the study. She is single, did not have any children and lives with her parents in the city center near the Grand Marche. Her salon, which had been open about a year, is also in the city center, on a major road. From the time Gladys left school she has engaged in trading off and on to earn money. She did this initially to pay for her training as a hairdresser. Although she

” Catchi is a major shopping area in suburbs (first tier of urban development).

147 received money from friends and family when she graduated, she invested this in land on which to construct a house someday rather than using it to open a salon. After receiving her diploma, she went to Cote d’Ivoire to work with another hairdresser for more advanced training. While there, she was also able to earn some money working for him. The money paid by every client she worked on was split three ways. Gladys explained it as follows.

If I did the hairstyling it was $3.00, we divided it in three. He (the patron), took $1.00 for the salon, $1.00 for him and then $1.00 for me. (Interview with Gladys, November 1998)

Although she was living with her uncle, Gladys had to use this money to feed herself. Thus, when she returned to Benin she did not have enough money to open a salon. Nonetheless, she was able to get the necessary equipment from various people. For example, her mother bought her a hair dryer and her old patron gave her a broken chair, which she fixed. She also indicated that another male Mend gave her things like an electric curling iron. With all this she was able to open her own salon, but this only lasted about seven months as she had problems with the landlady. While she was looking for another location, on a main road, she engaged in trade again. This time she bought yams and cheese in Parakou and sold them to producers of prepared food on credit®®. Although she earned a significant amount from this, she again decided to invest in land, when the plot next to her first one became available. She continued this trade for about 18 months, but eventually, she stopped this because it was too much trouble. About that time, she found a workshop for rent and once again opened her salon. Although she still had all her equipment from before, she indicated that it was not enough even for her working alone.

“ Parakou is a city about the same size as Porto Novo located in Northern Benin. It takes five hours by taxi and twelve by train to get there from Cotonou.

148 I need many more things, but there isn’t the money. Now the hair dryer there, I would like two. If two clients come I have to do one after [the other]. If I had two...(Interview with Gladys, November 1998)

She went on to explain how hard it is to earn money for these things due to competition.

That girl that came earlier, normally if one does the hairstyle she wanted, it is $1.50. She said [she would pay] 75 cents . . . Here in Porto Novo everyone pays S1.00, she took off another 25 cents. (Interview with Gladys, November 1998)

In addition she recounts how expensive it is to live and maintain the salon.

I take a motorcycle taxi [to get back and forth], I do everything, I do the tontine®'. I do the tontine, one for electricity here, tontine for the electricity at the house, everyone uses the electricity. . . , one for the salon, which is $12.00 a month. (Interview with Gladys, November 1998)

She is, therefore, unable to save very much money with which she could buy more equipment. Her dearth of equipment, in turn, means that she cannot employ any apprentices, who might help increase her income both because they pay for training and by enabling her to work on more clients at a one time. Also, although she has the salon, keeping it open is very difficult under these conditions. So, she continues to trade in food from time to time to supplement her earnings from hairstyling. Gladys’ interview reflects two important points. First, it indicates ways in which restricted financial capital reduces women’s opportunities to make effective use of their skills. Second, it reveals the importance of MLES. While her parents did not have the resources to pay for her training, she was able to herself through commercial activities.

*' Tontines take two forms. One is a rotating savings group, in which all participants contribute at pre-determined regular intervals and then take turns using the money. The second is a form of savings and loan, in which a man collects money weekly &om clients and then returns it to them at a pre-determined interval (e.g. monthly). He usually keeps a portion as payment for the service and does not pay interest. He also loans money if the relationship is well established. In this instance, he charges interest, which is usually quite high (e.g. 50%)

149 Another issue raised in Gladys life history is the increasing level of competition women face. This reflects two macro level social and economic processes at work. First, the imposition of structural adjustment means that fewer people are continuing in school because they are unlikely to find wage employment afterward. Rather many more are obtaining training as artisans. Given the rigid gender division of labor in which only two artisanal activities, hairdressing and seamstressing, are available to women, competition among women with these skills has increased dramatically. In addition to these macro level processes. Class also limits women’s opportunities in strategy configuration. Here Class effects manifest in differential Xsser Distributions and impacts oiFormal Rules. Lack of financial capital makes it difficult for women to pay for training, obtain certification and equipment necessary to open workshops. Moreover, without formal training and the diploma one earns fi’om it, women do not have the right to open a workshop. 8,4.4 Class and Life-cycle Effects on Asset Distributions: Husband's Interference Another subset of women Limited Assets experienced both Class and Life-cycle effects on their access to Assets. This applies to women who cited their husbands as the source of starting capital, because it is unlikely that he provided very much. There are two possible reasons for this. One is that he simply did not have it. Alternatively, providing a wife with only a small amount of capital might be seen as a means of controlling the scale of her activities such that she would not have a reason for venturing far fix)m home. The latter explanation is consistent with the case of young women whose autonomy is hampered by expectations of husbands and in-laws and with all women who felt their merchandise is too low to warrant having activities outside the home. Under these conditions women’s lack o îAssets reflects the intersection oiLife-cycle Phase and Class. The ability of husbands to severely hinder their wife’s activities is evident in the case of Pauline {Retailer-Self-employed-2-3-Daily-use Activities). She was raised by an aunt because her mother died when she was six-years-old. Her aunt was a teacher and as a result Pauline was able to get an education, but only through the fifth year of primary

150 school because she did not pass the exams to continue. When she left school, her aunt paid for her to attend secretarial school, but she did not fim'sb the course. When Pauline quit secretarial college at about age 18, she left Cotonou, where she had lived with her aunt, and returned to her father’s house in Porto Novo. Throughout her childhood, her aunt paid for all her expenses, from which Pauline saved some money to use as starting capital.

My aunt gave me breakfast money, I hid a little and that is how I collected to start selling small things. (Interview with Pauline, October 1998).

With her return to Porto Novo, she had to support herself, which she did through trade and managing her cousin’s hardware store for him. She primarily sold snack foods such as peanuts and fruit. When asked why she chose these activities she said that it was because she did not have a lot of capital. When she married, her husband, a teacher, moved the family to Cote d’Ivoire. Once there he gave her $ 140 to start trading again. She began by selling toiletries, later adding snack foods, such as peanuts which she sold wholesale. Eventually, she increased her capital to $800 at which point she switched to wholesaling jewelry. She bought merchandise in Benin and sold it in Cote d’Ivoire. During this time of prosperity she and her husband split all household expenses including children’s school fees. In 1992, when the CFAf was devalued by 50%, she lost a good portion of her capital because retailers to whom she sold jewelry on credit were unable to pay her.

The people didn’t pay. You run, you take the taxi, the taxi counter that quickly counts the money, I saw finally that I was wasting my money. . . the people didn’t pay. (Interview with Pauline, October 1998).

151 About this time, her husband, with whom she had problems throughout their marriage, decided he wanted to take a second wife. As a result, he not only stopped contributing to the household, but also started to take her money. Finally, in frustration she lefr him and returned with the children to her father’s house in Benin. When she came back, she had only about $300 left. With that she began to engage in cross-border trade between Nigeria and Benin.

I did trade in com, beans, eggs, even drinks (sodas) with the rest of the money, I bought for [S300] here and I ate it too until it was finished. All I have left is [S20]. (Interview with Pauline, October 1998).

Because she had to cover all household expenses for herself and the children including their school fees, Pauline was unable to hold on to her capital. At that point her aimt turned over a small boutique in front of the family house from which she sold religious articles. At the time of the interview, this was Pauline’s principle economic activity along with the sale of snack foods, which provided a very small, but regular income. Religious articles do not sell very often or offer much of a profit. For example, she only earns about thirty or forty cents on the sale of a Bible. With these few enterprises and occasional help from family members she supports herself and her children even managing to keep them in school, but at the expense of having something to eat regularly. Pauline’s experience is consistent with many middle-aged women from poor families. Her family did not have much in the way of resources with which to help her start in trade. Although, she had an education, it did not cost her aimt anything because she was a teacher. Her husband provided her with considerable starting capital once he began earning a lot himself, but then re-appropriated it when he chose to marry again. Her capital was further eroded with the economic decline in the 1990s. Ultimately, she was left to raise her seven children on her own with very few resources. The low levels of financial capital among women with Limited Mobiiity^ is, therefore, explained by socioeconomic Class of families and under some conditions by Cultural Norms associated with Life-cycle Phase. These circumstances severely reduce women’s opportunities in economic activities and therefore, livelihood strategy

152 configuration. The restricted financial capital they amass is used primarily for retail trade in daily-use goods wltich are readily available without leaving Porto Novo. These goods, however, provide few opportunities to increase capital as they have only marginal profitability, resulting only in strategies o(survival or low stability. 8.4.5 Conclusions to Class effects Limiting Assets Although women with Limited Mobility differed in their access to Assets, this is only apparent in human capital. Women vnûi Substantial Assets have education and/or training and ,in the case of women using family-oriented strategies (Producer-Family, Retailer Family), unpaid family labor. Lack of financial capital among all women with Limited Mobility results from socioeconomic Class, reflecting family background. While families of women with Substantial Assets are somewhat better off, evident in that they provided education and/or training, they lack sufficient resources to provide financial capital. This is especially limiting for artisans who cannot make effective use of their skills without formal training, proper certification and money to open a workshop. For many women, especially those whose strategies are comprised primarily of trade activities. Limited Mobility and lack of financial capital are inter-related. Women choose to purchase supplies locally because there is no benefit to traveling outside Porto Novo. Finally, for many women. Class and Life-cycle intersect reducing both Mobility and Assets. This is particularly the case for young women, who often lack autonomy and middle-aged women, whose husbands abandon them.

8.6 Conclusions Despite having disparate access to Assets women in Enabling Condition groups discussed above. Limited Mobility/Limited Assets and Limited Mobility/Substantial Assets are not really all that different. In both cases, women choose to center their activities at home and do not travel outside Porto Novo. They are also similar in their lack of financial capital. Nonetheless, examination of these women’s life-circumstances reveals that processes from which these conditions emerge vary significantly. While Ethnicity is not a critical Dimension o fDifference among these women, Life-cycle Phase

153 and Class interact in a myriad of ways manifesting in localized social institutions (Structures o f Constraint) through Distributions, Cultural Norms, and Formal Rules. These intersections are socially and geographically localized in women’s residential arrangements and locations producing Ioca//ze(/ Enabling Conditions.

154 CHAPTER9

SUBSTANTIAL MOBILITY: HOW DO DIMENSIONS OF DIFFERENCE PRODUCE LOCALIZED ENABLING CONDITIONS?

9.1 Introduction Chapter 8 examined how specific intersections among Dimensions o f Difference (Gender, Class, Ethnicity, Life-cycle Phase‘S) give rise to Localized Enabling Conditions (Asseis, Mobility^\ specifically focusing on strategies emerging from conditions of Limited Mobility. Following the same analysis procedures^ this chapter explores the relationship between Dimensions o fDifference and conditions Substantial Mobility. Issues of Mobility are the main concern because this was identified as the dominant Enabling Condition from which specific strategy types emerge (Chapters 6 and 7). Moreover, strategies associated with Substantial Mobility^^ provide women greater opportunities for income generation by enabling them to access suppliers outside Porto Novo, thereby affording a wider range of choice in economic activity and, hence, strategy configuration (Chapter 7).

Theoretical concepts are italicized. Enabling Conditions and accumulation potential are bold italicized. ^ For a discussion of analysis procedures see Chapter 8, section 8.2. A list of strategy types grouped by the Enabling Conditions firom which they emerge are presented in Table 7.1.

155 Drawing on data Grom 522 surveys and 15 life/economic histories, analysis centers on interactions among variables associated with Cultural Norms and Assets^, two aspects of Structures of Constraint, localized within particular residential arrangements^^ and locations. This proceeds in three steps. First, themes reflecting reasons for women’s Substantial Mobility are identified through descriptive statistics (Grequencies and contingency tables). Second, women’s life/economic histories** illuminate processes through which intersections among Dimensions of Difference produce conditions of Substantial Mobility. The second section examines differences among women using strategies associated witli this Enabling Condition in their access to Assets.

9.2 ExplainingSubstantial Mobility 9.2.1 Introduction: Identifying Themes Strategies associated with Limited Mobility almost exclusively involve retail trade in daily-use goods. In contrast, those attributed to Substantial Mobility include a wide array of different configurations involving high-skill wage employment, artisanal activity, service provision, retail trade and wholesaling®. The common point among all strategies in this Enabling Condition group, however, is at least one retail oriented commercial activity in as-needed (e.g. medicines) and/or occasional-use (e.g. fabric) goods. As discussed in Chapter 7, these goods are often only obtainable at wholesale prices in particular locations outside Porto Novo. Even where some may be purchased locally, doing so reduces profit potential. Women combine trade in these goods with other activities not only to increase their income, but also because their other enterprises often provide only irregular (e g.

“ A list of relevant variables is provided in Chapter 6, Table 6.1 and are reviewed in Chapters 4 and 6. ® Variable names are highlighted in bold. ** Table 7.2 lists all interviews and their strategy types. ® See Table 5.5 for a description of strategy types.

156 seamstressing) or insufficient (e.g. high-sldll wage employment) incomes. This is evident in women’s reasons for having more than one activity (see Figure 9.1). Diversification is the dominant reason given by artisans as they only have regular sales during holidays and therefore, need supplementary activities to ensure steady earnings. As one hairdresser put it, "there isn't work all the time, I sell things to be able to live." For retailers of occasional and as-need goods, diversification can be equally important because often these types of merchandise do not sell all the time. In these circumstances women might combine the sale of several of higher profit items and/or daily-use goods. As one retailer of occasional goods explained, "beauty products don't work, that's why I decided to add fabric to assure a revenue." By combining retail trade in as-needed and/or occasional-use goods with other trade and non-commercial activities, women create strategies of/iig/i stability and surplus.

complimentary activi

salary Insufficient making ends meet

diversification

family tieritage

personal preference increiM Income seasonality

Figure 9.1 Reasons for using MLES by Percentage of Women

157 Strategies among women in this group tend to comprise a small number of activities, however. This primarily reflects inadequate time, due to having a full-time job, travel, or lack of desire as income earned firom the few enterprises is seen as sufGcient. For example, one seller of beauty products {Retailer-Self-employed-Occasional Goods) who traveled regularly to Nigeria for supplies, stated "[w]ith the one activity alone, I am already bothered too much. I cannot do other things anymore." Another wholesaler of gari (Mixed-1-Activity), who buys her merchandise from producers in the Oueme, said "I don’t find the extra time to do other things." Time required for travel to purchase supplies relative to that available for trade is evident in women's descriptions of how they organize their activities. For example, a woman using a Retailer-Self-employed-Occasional Goods strategy portrayed her typical week as follows:

I go to Nigeria to buy merchandise and the next day I go sell a little. The third day I go to Cotonou to buy clothes and I sell the fourth day until the end of the week. (From survey with Retailer-Self-employed-Occasional Goods)

Essentially, this woman is only trading every other day. Lack of time for additional activities reflects the time she spends traveling to get supplies. Similarly, an artisan described her travel as somewhat less frequent, but further afield.

When the clients ask for something I travel to Lome. At most two times a month to do the shopping. (From survey with Artisan-Employees)

When this woman travels her apprentices manage the salon. In this instance, economic activities are less impacted by women's Mobility. The main question answered here is why are women, using these strategies mobile outside Porto Novo? Table 9.1 depicts the ethnic distribution among women with Substantial Mobility. As discussed in Chapters 2 and 4, Yoruba women have engaged in long-distance trade since pre-colonial times. Thus, it is not surprising to find them heavily represented. Nonetheless, Goun remains the dominant ethnicity.

158 Women with Total Ethnicity Substantial Mobility Sample

Other 4.2 2.9

Yoruba 27.2 19.9

Goun 563 63.0

Eon 7.5 9.2

Mina 47 ------5Æ

Total 100.0 100.0

Table 9.1 Percentage of Women with Substantial Mobility by Ethnicity

These findings suggest that Ethnicity, as one Dimension of Difference, is important to understanding why women have Substantial Mobility. As in Chapter 8, among the Goun, Cultural Norms associated with Life-cycle Phase, localized in particular settings (residential arrangement and location), are also critical to explaining ^uAs/a/i/ia/ Mobility. Finally, Class effects are evident in women’s mobility for both ethnic groups and explain differential access to assets among women with Substantial Mobility. Tables 9.2 and 9.3 reveal variation in residential arrangement and location by Ethnicity and Life-cycle Phase. Yoruba women are much more likely to live with their own families at every age, either in the city center or suburbs. The latter pattern is particularly evident among young women who are probably still living at home with their parents, not in multi-generational extended family settings. Middle-aged Yoruba women also live in the suburbs, but are heads of households. Unlike Goun women, they infiequently live with in-laws or in nuclear family settings until older.

159 Percentage of Women with Substantial Mobility by Age-Range and Ethnicity Total Residential Sample Arrangement Yoruba Goun

<30 31-50 51+ <30 31-50 51+

Nuclear w/ Husband 20.0 8.8 3 3 J 15.4 33.8 13.3 26.4

Female*Headed 6.7 20.6 0.0 5.1 21.5 46.7 18.8

Family w/ Husband 6.7 0.0 11.1 2.6 1.5 0.0 l.l

Family w/ o Husband 60.0 4 1 J 22.2 38.5 15.4 40.0 24.5

In-laws w/ Husband 6.7 23.5 l l .l 3 3 J 16.9 0.0 21.6

In-laws w/o Husband 0.0 5.9 22.2 5.1 10.8 0.0 7.5

Total 100.0 100.0 100.0 100.0 100.0 100.0 100.0

Table 9.2 Residential Arrangement of Women with Substantial Mobility

Percentage of Women with Substantial Mobiiify by Age-Range and Ethnicity Residential Total Location Yoruba Goun Sample

<30 31-50 51+ < 30 31-50 51+

City Center - M arket 26.7 47.1 66.7 18.9 12.7 33.3 18.4

City Center-No Market 6.7 il.8 11.1 8.1 15.9 0.0 9.4

Suburb - Market 0.0 5.9 0.0 13.5 19.0 26.7 12.8

Suburb-No Market 46.7 23.5 Il.t 133 7 3 0.0 11.3

Urban-fringe - Market I3 J 5.9 II.I 32.4 3 6 3 40.0 35.6

Urban-fringe - No Market 6.7 5.9 0.0 133 7 3 0.0 10.6

Total 100.0 100.0 100.0 100.0 100.0 100.0 97.9

Table 9.3 Residential Location of Women with Substantial Mobility

160 Goun women with Substantial Mobility display similar patterns to their counterparts with Limited Mobility in middle-age and when older; living in nuclear and female-headed households In middle-age, and heading their own households or living with their own families when olden Young, Goun women, however, have a somewhat different pattern in that they are more likely to live with their own families. Moreover, although a significant portion of young Goun women live in the urban-fiinge, some are also likely to live in the suburbs. Even where they live in the urban-fringe, the fact that they are more likely to be living with their own families suggests a different pattern in Cultural Norms associated with Life-cycle Phase than that experienced by their counterparts with Limited Mobility. In all cases, women with Substantial Mobility have greater autonomy and independence in economic activities than women with Limited Mobility. Among the Yoruba this reflects Cultural Norms, associated mih. Ethnicity, that encourage women to be economically independent. Often Yoruba women are solely responsible for the financial needs of their children. To facilitate this, they are given a sum of money at marriage as part of their dowry, which is to be invest in commercial activity. If required to ensure profitability of their enterprises, Yoruba women are also expected to travel outside Porto Novo. Even where women do not have official marriages, this practice persists™. One further difference of interest is that Yoruba women are more likely than Goun to be in ploygynous marriages as the second wife. For the Goun, Life-cycle and Class intersect and are localized in nuclear and female-headed households in the suburbs and urban-fringe, enabling these women to also be economically independent, and hence mobile outside Porto Novo. These patterns are evident in women’s household characteristics, depicted in Table 9.4. The processes by which these conditions emerge are discussed below.

™ Official marriage refers to one sanctioned through religious, civil, and/or customary ceremony. Most women in Benin do not have official marriages as it is very expensive.

161 Percentage of Women with Substaniial Mobiliiy by Age Group

Characteristic Yoruba Goun Total Sample <30 31-50 51+ <30 31-50 51+ (25.9) (58.6) (15.5) (32.8) (54.6) (12.6)

Children’s Age Range” - <10 57.1 11.8 0.0 56.4 17.2 6.7 33.5 5-15 7.1 17.6 l l .l 7.7 20.3 13.3 16.5 0-20 14.3 58.9 66.6 12.8 56.3 33.3 36.6 20+ 0.0 2.9 0.0 0.0 1.6 13J 2.1 not applicable 21.4 8.8 22.2 23.1 4.7 3 3 J 11.3

Childcare - mother 6.7 19.4 0.0 22.2 22.2 28.6 28.7 father 6.7 0.0 0.0 0.0 3.2 0.0 2.4 relative 40.0 19.4 11.1 41.7 30.2 7.1 27.3 maid 20.0 9.7 0.0 2.8 11.1 0.0 5.6 children 6.7 38.7 66.7 2.8 30.2 28.6 21.7 not applicable 20.0 12.9 22.2 30.6 3.2 35.7 13.3

Household Head - herself 20.0 47.1 44.4 20.5 36.9 80.0 37.2 husband 33.3 38.2 55.6 51.3 61.5 20.0 52.8 shared 0.0 0.0 0.0 0.0 1.5 0.0 0.8 parent 46.7 11.8 0.0 28.2 0.0 0.0 8.3 in-law 0.0 2.9 0.0 0.0 0.0 0.0 0.8

Husband in Household - yes 45.5 34.4 55.6 64.5 5 2 J 13.3 51.7 no 54.5 65.6 44.4 35.5 47.7 86.7 48.3

Years Married - <10 53.4 32.2 0.0 69.4 21.4 9.1 33.1 11-20 6.7 45.1 0.0 2.8 50.8 0.0 31.2 20+ 0.0 16.1 100.0 0.0 27.9 90.9 29.4 not applicable 40.0 6.5 0.0 27.8 0.0 0.0 6.3

Marital Status - not married 40.0 8.8 0.0 28.9 0.0 6.7 7.1 only wife 20.0 20.5 22.2 36.9 30.8 26.7 26.7 1“ wife 26.7 17.6 22.2 31.6 53.9 3 3 J 473 2*'wife 13.3 52.9 55.5 2.6 15.4 3 3 J 18.8

Table 9.4 Household Characteristics of Women with Substantial Mobility

Although categories overlap, women were only coded once depending on the actual age-range of children in their households.

1 6 2 9.2.2 Ethnicity and Class Effects: Economic Independence among Yoruba Women Yoruba women, in general, have considerable autonomy at every stage in life. This is evident in designation of themselves or parents as household heads. Moreover, at no Life-Cycle Phase do Yoruba women identify themselves as primary care-takers for their children. A further indication is the fact that only older women have a greater propensity, than women in the whole sample, to have husbands in their households. This reflects residential arrangement, in that most Yoruba women live with their own extended families. Middle-aged women also refer to having maids as a means of managing childcare responsibilities. This is suggestive of their socioeconomic Class. Thus, even where women live in nuclear or female-headed households they are likely to have someone else take of their children. These conditions reflect Cultural Norms among the Yoruba. The life circumstances of interviewee, Delphine {Retailer-Self-employed-As- needed-Goods^), 32, who sells medicines and imported food are reflective of Ethnicity effects on women’s mobility. She started selling snack foods while still in school, with a little bit of money her uncle’s wife gave her. When she married, as is custom among Yoruba, her husband provided a more substantial capital of $200, which allowed her to start selling medicines and beauty supplies. Although she has a child of three, he lives with her aunt in Cotonou. After Delphine had the baby she became sick, and so her aunt took the son. When asked if the boy was going to come back home, she said,

I don’t know, I don’t know, but since the aunt, she doesn’t want [him to leave] because she does not have children, she has two sons who are elsewhere, so she is alone. (Interview with Delphine, October 1998)

When asked if she provides money for the child’s upkeep she stated,

It’s his aunt, if I find something I buy it for the child, it’s his aunt, it’s the older sister of my husband. (Interview with Delphine, October 1998)

^ Livelihood strategies are italicized.

163 Thus, Delphine’s financial and care-taking responsibilities are both reduced. Unlike many Yoruba women, she lives in the suburbs, in a nuclear family setting with her husband and two children firom his first marriage. When asked who takes care of them while she works, she just laughed and said "they are big". Delphine has limited care-taking responsibilities, her youth and residential arrangement notwithstanding. This affords her a great deal of mobility. As a result she has the fireedom to go to Nigeria regularly where she buys medicines, which she then retails in her neighborhood. Her husband does not limit movement because of expectations of Yoruba women as traders. Blandine’s {Retailer-Self-employed-Occasional Goods) children were old enough at the time of the study to take care themselves. Even when they were younger, however, she frequently left them on their own when she traveled. She continued trade through delivery of each child, and then only waited the customary forty-one days after the each baby's birth to start again. After her first child was bom, when she traveled she said

The house was big, there was even renters to whom I could say, well, watch the child I am going and coming back. I left the child with women in the house.. JDuring that time, during that time when my co-wives came over I took the child to an Aunt who was at Dodji. I left my child there, when I came back, I went and got my child (Interview with Blandine, October 1998)

Blandine, did not want to leave her children in the house when the co-wives were there, because she feared they might hurt her children since all their children were rivals for her husband’s wealth when he died. When her oldest child was ten she left them on their own.

The children already a little big, the first was ten years old and so she took care of the second and I left the two at the house to go [travel]...When I was gone overnight my Aunt came in the evening to stay with them. (Interview with Blandine, October 1998)

Leaving her children on their own enabled Blandine to be mobile in economic activities. Her willingness to do so reflects the importance of being a successful trader to Yoruba

164 women, and being able to financially support the household. When asked if her husband objected in anyway she stated that

As he, himself, as he, himself is a trader, he didn't say that. He didn’t even want that a wife stayed without doing anything. (Interview with Blandine, October 1998)

A somewhat older woman, Kadija {High-skill-Public), 52, a rural extension worker for the government, also has never had to manage childcare responsibilities on her own. Her life-circumstances reflect both Ethnicity and Class effects on women’s mobility. At the time of the study, she lived with her husband and ten children in the suburbs. In addition to three of her own children, she also supports the child of a friend, three of her siblings children and three cousins. This nuclear family residential arrangement, somewhat unusual among the Yoruba, reflects the fact that both Kadija and her husband work for the government and, as a result, have been moved often in the course of their careers. Because they are officially married (civil and religious wedding) they are always moved together, however. Although they plan to retire in Porto Novo, they are originally from north Benin and, therefore, do not live near their extended families. Kadija’s household composition illustrates another important aspect of many African cultures, which is the tendency for poorer family members to foster children with wealthier relatives. This is especially true for teachers because primary education is free to dependents, including children who are not their own. Another interviewee, Regine {Artisan-Self-employed), who was raised by her sister, a teacher, explained why this is the case.

She had six children, but at the time, relatives, well, when they see people who are evolved, or someone is already a teacher, then they are seen as very important already, already in the family, so they entrust their child to them...we were nineteen raised by the couple, interview with Regine, October 1998)

165 When asked ifher sister could have refused she stated,

She could not refuse, they were relatives, cousins and often people who were friends. People want only that their child go to school also. (Interview with Regine, October 1998)

As a result, women who are perceived to be wealthy, especially those with wage employment, tend to have large households, but many of the children are not their own. Throughout Kadija’s adult life, she and her husband always supported many people in addition to their own children. During times of crisis this included adults as well as children. Such was the case in the mid-70s during a famine in north Benin, where she lived and worked at the time.

We were in a strategic location because all the sick people from our village were brought there and the foreigners came. It was at Nattitingou, so everyone came to the house and in one day, when there weren’t very many, we were not fewer than 25 to eat at the house...we ate morning, noon and night. (From interview with Kadija, October 1998)

When asked if she fed all tliese people herself, she emphatically responded in the affirmative. While foster children help with care-taking responsibilities and housework, their presence in the household adds considerably to financial responsibilities. As discussed above, with respect to Delphine {Retailer-Self-employed~As-needed-Goods),y/hosc son lives with an aunt, when people foster their children with relatives they do not expect to pay living expenses for them. This explains, in part, why women with wage employment have economic activities to supplement their salaries. In addition to wage employment, Kadija also always maintains at least one, and at times several, commercial activities. She explains her trade activities as a fimction of three conditions. First, she was "bom to it because my parents were both traders". In part she related this to her being Yoruba and their expectations of women as traders. Second, the number of dependants coupled with the insufficiency of her salary (even in combination with her husbands), requires that she earn additional money.

166 It cannot suffice, we are numerous, we are numerous and there are our children, there are our brothers’ children, all that there and well, it is necessary to send money to our parents. (Interview with Kadija, October 1998)

Third, economic change make commercial activities a necessity.

As the economy changed and became more influenced by the world economy the need for bureaucrats decreased. So the salaries stabilized while prices and needs increased. As a result people are obliged to do something in addition to their public service, especially if they were raised in trade. (From survey with Kadija, June 1998)

This increasing tendency for government employees to also engage in trade, helps explain the growing number of traders, upon which many other women (surveyed and interviewed) commented. This is one important source of the rising competition discussed earlier. The cousins living with Kadija, work as maids, but since all children in the household are fourteen or older, they do not require care-taking. The maids, therefore, primarily help with housework, as do other children who are all either in school or apprenticeship. When the children were young, Kadija always had a maid from her home village to help with child-care. In addition to childcare and housework, the maids helped Kadija with her supplementary economic activities. Since the start of her career, she always engaged in two types of trade - making and retailing prepared snacks, and wholesaling imported food. While living in north Benin, her sister who lived in Cotonou bought imported goods and sent them north. Kadija described a typical day as follows;

The morning already, can I say morning, at midnight already I get up to make the little cakes, well say at 10:00 PM or II :00 PM. I put the flour, I beat the flour, I left and at midnight already it had risen, so I, at midnight I start to put them in the oven. Well, as I make a lot there, many, until 3:00 in the morning, 4:001 finish that there and I put the rice on the stove, the rice cooks and from 5:30,6:001 start to peal, I start to peal my yams. I cut [them] and

167 the girls there (maids) who were already beating the cake that I must cook with the yams, they beat it. At 6:451 finish everything, I finish everything and I give it. You, you go there and you, you got fiiere by the schools, by the market and a least, at most, at 10 minutes before, I leave the house and at 7:00, it was at 7:00 that we went to work....Since I was nursing at that time, well I came home at 11:30 since we left at 2:00, if I came home at 11:301 benefit by beating more flour for the evening. (Interview with Kadija, October 1998)

Kadija went on to say she was always on time for work, but never slept more than two or three hours in the evening. She was able to manage because she had maids. Usually one maid stayed home to care for the children and sell firom the house, while the other did ambulant sales. Since moving to Porto Novo in 1990, Kadija stopped selling prepared foods. She said too many people were already selling them in her neighborhood to make it worthwhile. Instead, she makes ice and frozen yogurt, the money firom which pays for the refingerator’s electricity usage. In addition to maids selling fi’om home, she bought a bike and hired an elderly man who circulates through town selling firom a cooler. She also continues to wholesale imported foods, which are bought in Cotonou. Kadija attributes her use of MLES to necessity and family heritage. Her ability, to manage multiple activities, even while working full-time, stems primarily firom socioeconomic Class. This is evident in Kadija's ability to hire others, and that she has maids. It is also apparent in her high level of financial responsibilities as a major contributor to the household budget. If she were not considerably wealthier than her relatives, they would not foster children with her. In this sense Class can be seen as a double-edged sword. Kadija's desire to engage in trade is also reflective of hec Ethnicity. This is what she means by heritage; Cultural Norms among Yorubas that women have commercial activities. In her life-story, as with others who use livelihood strategies associated with Substantial Mobility, it is the intersection of Class and Ethnicity that explain their relative fireedom of movement. These circumstances are reflective of economic independence.

168 9.2.3 Life-cycle Phase and Class Effects: Economic Independence among Goun Women The life-style described by Kadija is also prevalent among Goim women in public service. Indeed, the majority of women using a High-skill-Public strategy are Goun (62.5%) or Fon (18.8%). For them, the ability to travel outside Porto Novo is reflective 0ÎLife-cycle Phase and Class. Virtually all (92.3%) Goun or Fon women with public sector employment are between the ages of thirty-one and fifty. Goun women in tlris age group live in nuclear and female-headed households, in the suburbs or urban-fringe, in newly built or rented houses. Despite not living with family who could assist in childcare, these women are not hindered in mobility by household responsibilities. In part, this reflects socioeconomic Class, which enables them to either hire maids or foster extended family members who work in that capacity. As described with Kadija, families perceived to be wealthy frequently foster children of relatives and friends. Under these conditions, women’s contributions to the household budget are critical. As a result, they gain autonomy and independence in economic activities, including the freedom to travel outside Porto Novo. Therefore, Substantial Mobility among this group can be explained by economic independence, which is associated with Class. Moreover, the majority of children among women in this age-group are older and probably in school. It is at this stage in life when Goun women gain the freedom to engage in economic activities and be mobile outside Proto Novo. Unlike counterparts with Limited Mobility, children of these women probably attend school, especially if one of their parents is a teacher. This also reduces care-taking responsibilities. Where these women have a maid, security against robbery is not an issue. Thus, Life-cycle Phase also works to women’s benefit in the sense of providing Substantial Mobility in this type of setting. Young Goun women are similarly mobile. As with middle-aged women, this is reflective of both Class and Life-cycle. A significant proportion of young Goun women are not married, even if they have a child. Some who are married live with their in-laws, but a prominent sub-group live with their own families in the suburbs. Under these

169 circumstances, they have considerable autonomy because parents do not interfere with their daughter's economic activities. This living situation, coupled with ability to call upon relatives for help with childcare, affords Substantial Mobility. Even where young women have small children and live in nuclear or female- headed households, but do not have maids, they may have another source of assistance. If the woman is an artisan with apprentices, she can ask them to care for her children when traveling. This is apparent in artisans' responses to questions regarding employee responsibilities. A hairdresser said they "sweep, do housework, get water, braid hair, sell", essentially whatever she wants them to do, including taking care of children. An older Goun woman, Cheri {Retailer-Self-employed-As-needed goods), has life circumstances similar to those described above. At 44, she has two grown children, the younger of whom is 19. Her daughter is in training as a computer technician. Her son is more problematic.

So I went there (son’s school), and made the child come (home) since school was not working. I asked the child what he wanted to do. He said that he wanted to learn to be an auto mechanic, so I put the child with a mechanic (apprenticeship). That did not work. He didn’t follow it at all. He ruined the patron’s things. It was his father who paid each time. So, afterward we asked the question again. He said now he wanted to be a chauffeur, that would suit him. So, we put him in that craft. He has already finished, he got his license and he drives today. (Interview with Cheri, October 1998)

Thus, although her children still live with her, they are grown and busy living their own lives. This quote also indicates her children attended school. Moreover, while Cheri lived with her husband at the time of the interview, she did not always. When first married, she was the second wife and rented a room apart firom her husband and his first wife. Eventually, she bought land and built her own home. That is where she lived at the time of the study, with her husband. His other wives (first and third) left him. He is retired, however, and does not impinge on her autonomy in anyway; Cheri is completely economically independent. He Uves in her house, but she supports the household.

170 Cheri changed activities many times over the years. The little capital to start an activity was 6om her father. Her husband neither paid a dowery nor gave her capital at marriage. She started selling a daily-use good, com, which was bought directly &om producers in the Oueme and wholesaled in Porto Novo. When she married Cheri switched to selling scarves in Nigeria because her husband, a chauffeur, could drive back and forth for free. When children were young she took them with her, leaving them in the car with their father while she purchased supplies. As they grew older, they attended school and then stayed with relatives while she traveled. Eventually, Cheri was able to save $1000, which she invested in the sale of fabric with her sister-in-law. After four years, she realized that other women were starting to sell fabric on credit, and that her profits were decreasing due to increased competition. She and her sister-in-law split their capital and remaining fabric. She invested most of the money in land, on which she eventually built her house, and then began selling candy in Nigeria, which was bought on credit in Benin. Once competition for this type of good became too great, she changed again to the sale of medical supplies. She decided to do this after talking with some nurses and asking what they needed. With this trade, instead of buying in Benin for sale in Nigeria, Cheri does the reverse. When asked how she thought to ask nurses what they needed, she explained it as follows:

When you are bora intelligent tliat’s what you are going to do, things that are very very intelligent, I thought that. I went myself to ask because I did not know where the syringes came from... So, I went and asked them if I bring you this, will you buy it. The people said yes and I started. (Interview with Cheri, October 1998)

Since she only buys on command, Cheri has had a fairly steady trade for fourteen years. She noticed recently, however, that some nurses were getting family members involved in the trade. Between this increased competition and increasing prices on everything, she wasn’t earning as much as previously. So, Cheri developed a new market — wholesaling to resellers in Cotonou.

171 She finds trade in this as-needed good, medical supplies, coupled with retail sale of sewing notions (also an as-needed good), an effective strategy. While medical supplies do not sell all the time, there is a large market and they offer high profits. Cheri’s Life-cycle Phase, the fact that her children are grown, coupled with her financial independence enable her to be mobile. 9.2.4 Conclusions In all cases described above, women are economically independent. Among the Yoruba, this simply reflects Cultural Norms that require women to be so. In this sense their mobility is best explained by Ethnicity, which is localized in city center and extended family residential arrangements. Among the Goun, Substantial Mobility also results from their financial independence, reflected in contributions to the household budget (median = 57% of total budget). These circumstances reflect Life-cycle Phase and Class, localized in residential arrangement and location. Most Goun women in this Enabling Condition group are either very young and therefore, living with their own parents in extended family settings in which they can rely on relatives for assistance with childcare or are older, in which case they tend live in female-headed and nuclear household settings. In the case of tlie latter, the presence of maids offsets care-taking responsibilities, enabling them to be mobile. Among the Goun therefore, Life-cycle Phase and Class effects explain S«6sm/i«a/Mobility.

9.3 Class and Ethnicity Effects: Differential among Women with Substantial Mobility 9.3.1 Introduction: Identifying Themes Despite having Substantial Mobility women within Enabling Condition groups of Substantial Mobility/Limited Assets and Substantial Mobility/Substantial Assets have differential access to This is best explained by Ethnicity and Class effects. Although both groups of women are considerably better off then their counterparts with Limited Mobility, there are significant differences among them. Variation is evident in their family background. While only Yoruba women with Limited Assets have human

172 capital, women of both Yoruba and Goun ethnicities with Substantial Assets also have significant financial resources. This is especially important for artisans with employees because it enables them to legally open a workshop and engage apprentices. Human capital in the form of secondary and university education enables those with Substantial Assets to access public sector jobs that would otherwise be unavailable. These differences are a direct result of family support available to artisans and high-skill wage employees. In some instances, however, women with Limited Assets are able to overcome lack of financial resources by ignoring social expectation. Asset related characteristics of these women are detailed in Table 9.5. 9.3.2 Ethnicity Effects: Differences among Women with Limited Assets Differences among women with Limited Assets are evident in their level of education, starting, and current capital. While most Goun have no education, the majority of Yoruba have at least attended primary school. The impact of education is also evident in the later starting work age of Yoruba women. Moreover, Yoruba women have significantly higher levels of both starting and current capital. Difference among these two ethnic groups is also apparent in sources of starting capital. While personal savings is important to both groups, Yoruba women are more likely to receive money fix)m their husbands at marriage. Bank loans are another significant source for Yoruba women, whereas Goun women rely on credit. Variation in women's access to /dssets is reflective of disparate family backgrounds. Yoruba women are more likely to be raised by both parents and have mother’s who were traders^. These differences are evident in the cases of Delphine and Cheri, discussed above. Cheri was raised by her mother alone because her father had five wives and forty-two children. Although her mother was a producer and retailer of

” Mother’s occupation was coded as “petty trader” for respondents who identified their mother's as primarily housewives who did a bit of petty trading on the side and “trader” for those whose mother's had full-time trading activities.

173 Substantial Mobility/ Substantial Mobility/ Lmited Assets Substantial Assets Characteristic Total Sample

Yoruba Goun Yoruba Goun

Education - none 36.8 70.2 35.9 20.6 513 primary 42.1 22.8 28.2 23.8 26.1 secondary 21.1 7.0 35.9 50.8 20.7 univenity 0.0 0.0 0.0 4.8 2.0

Trade Learned - seamstress 0.0 8.8 7.9 16.7 12.9 hairdresser 0.0 0.0 S3 11.7 3.1 typing 5.9 5 J 0.0 1.7 4.1 health worker 5.9 2.6 2.6 3-3 1.6 not applicable 88.2 82.5 84.2 66.7 78.0

Employee Relation - no relation 15.8 15.8 15.4 39.7 16.1 children 15.8 1.8 2.6 1.6 8.6 family 5.3 3.5 12.8 9.5 8.5 not applicable 63.2 78.9 69.2 49.2 66.8

Median Starting Capital 540.00 510.00 5120.00 100.00 520.00

Median Current Capital SIIO.OO 520.00 5200.00 90.00 530.00

Training Paid by - parents 10.5 10.5 12.8 25.8 16.2 guardian 0.0 1.8 0.0 1.6 1.5 husband 5.3 1.8 0.0 1.6 1.9 self 0.0 1.8 2.6 1.6 1.5 not applicable 84.2 84.2 84.6 69.4 78.5

Raised by - mother only 36.8 51.8 46.2 46.8 47.7 father only 0.0 1.8 2.6 3.2 2.3 both parents 21.1 8.9 25.6 27.4 17.8 other family 42.1 35.7 23.1 21.0 30.8 family friend 0.0 1.8 2.6 1.6 1.4

Mother’s Occupation - petty trader 57.9 82.1 35.9 64.5 693 trader 31.6 5.4 5 U 19.4 15.4 bureaucrat 5J 0.0 0.0 1.6 0.8 teacher 0.0 0.0 0.0 3.2 0.8 seamstress 0.0 1.8 2.6 1.6 0.6 farm er 5 J 0.0 2.6 0.0 1.9

Starting Work Age - IMS 11.1 25.0 17.9 24.6 243 16-20 66.7 40.9 20.5 32.8 41.0 21-25 22.2 2 7 J 38.5 37.7 26.0 26+ 0.0 6.8 23.1 4.9 7.7

Source of Capitai - savings 33J 30.4 17.1 333 319 husband 38.9 19.6 17.1 13.0 19.4 family 16.7 28.6 48.6 40.7 31.1 loan 5.6 0.0 5.7 1.9 3.2 credit 5.6 21.4 11.4 11.1 13.4

Table 9.5 Asset Characteristics of Women with Substantial Mobility

174 prepared foods, this was a relatively small scale activity that allowed her to supplement money received from her husband. Delphine's mother, on the other band, worked full­ time as a trader. As a result she was raised by an uncle, who paid for her education. Both women started their activities with relatively little capital. Delphine, a Yoruba, started with $5.00 received from her uncle's wife when she turned 18. Although still in school at the time, she started selling snack foods to classmates. When she left school, after failing exams three times, she used savings from her trade to add other daily- use items to the enterprise. Eventually, Delphine accumulated $40 in capital at which point she switched to selling beauty products, an occasional good. A couple of years later, she realized medicines, an as-needed good, would sell better because people were asking for them. At that point Delphine had $120, to which her husband added $200 when they married. Cheri started witli somewhat more money, $15, received from her father at age 19. With this she started trading com and onions, which were bought from producers in the region. This capital was insufficient for Cheri’s enterprise however, so she took an additional $35 worth of merchandise on credit. When she married, her husband did not give her any capital.

He did not give capital as such, but sometimes he gave me a little money which I kept. (Interview with Cheri, October 1998)

Despite these differences associated with Ethnic Cultural Norms, both women were eventually able to create highly profitable livelihood strategies. Cheri's case also illustrates the important role of agency in the form of "purposeful choices" that overcome a relatively poor family background. 9.3.3 "Purposeful Choice*’: Ignoring Social Expectation Despite starting with limited capital, Cheri was able to accumulate a substantial capital over the course of her life by reftising to give money to friends and family for ceremonies. This is a significant breech of social expectation. A very important custom among the Yoruba and Goun is attendance at various ceremonies for family and friends

175 including; weddings, funerals, graduations. At the ceremony guests are expected to give a gift of money. In large extended families it is possible to have ceremonies every weekend, which can be expensive, not only in gifts, but also in the necessary purchase of new fabric and clothing. These ceremonies serve an important reciprocal financial function in that if one attends regularly and later has some financial need their family and ftiends will help them. People who do not participate find it difficult to get any support. When she married, Cheri began to trade scarves in Nigeria because her husband provided free transportation as a chauffeur. This activity was very successful, especially because she did not pay for transport. For every bunch of scarves sold, Cheri earned $100. Eventually, between savings and her refusal to give money to friends and family for ceremonies, she was able to save SI000 with which to start selling fabric. Cheri explained her unwillingness to participate in important functions as follows:

Someone lost a relative, 1 should go give some money, 1 refused all that because 1 wanted to increase my capital and all my profit that 1 found 1 saved. (Interview with Cheri, October 1998)

She continued this practice for one year even in the face of people talking about her. When asked about people’s response to these actions she said that,

people there treated me like a miser, that 1 am a miser, 1 am stingy, 1 don’t want to give out money. (Interview with Cheri, October 1998).

When asked if this bothered her, she said that because she wanted her commerce to grow she ignored it. Eventually, however, this caught up with her, and explained her lack of financial capital at the time of the study. In 1998, a few months before being interviewed, her father passed away. Because he had been a well known traditional healer and village chief, the family was expected to put on a big ceremony for him. The ceremony, which lasted at least a week, was extremely expensive. The family imported a cofGn fixim Cote d’Ivoire and built a mausoleum in which to put it. They also bought two cows and many goats to feed the guests. In total this cost the family more than $4000. As was expected.

176 due to her wealth, accumulated over the years, and because it was her own father, Cheri contributed $500. Since she had not contributed to her friend’s ceremonies, however, they did not assist her with her father’s. As a result she said she had virtually no capital left. She was also, at that time, the sole support of her family because her husband was not working.

I don’t have any more money. My husband gave me $600. It is with that I will restart my commerce. The children will eat from it. The children go to private school. There is one who must pay $320 this year. Only I pay...It is with that sum I must do everything. (Interview with Cheri, October 1998).

Cheri's solution, once again, was to stop participating in ceremonies. Through this mechanism she will probably be able to accumulated significant savings. If she has another financial crisis, however, Cheri will not find support and will have to use own her capital again. Another example of a woman who controverts social expectation is Blandine, who changed religion and, in so doing was disowned by her family. This severely impacted Blandine's financial situation when her husband passed away.

It is especially because of religion that I was abandoned [by my family}. If today I say okay, now I don’t want to be Christian Celeste anymore, I want to follow you [family], 1 will find people to help me. (Interview with Blandine, October)

Blandine made a choice, however, that her beliefs were more important than getting financial assistance.

I cannot leave the church, I cannot leave because if I go to church, it is not for people, it is to pray to God, and I know that I have more benefice with God than with people. Even if I leave [the church] today, I go and follow my family, they will help me, they will give me money, but will I bring the money? I will go and leave everything and leave, then if I die, after my death, I know that I have more recompense with God. Money, well, it is something one uses here and leaves only, that does not have importance for me. (Interview with Blandine, October 1998)

177 Like Cheri, she made a "purpose choice" about her life, but doing so cost her family assistance. As a result at the time of the study, she, too had virtually no capital left. Blandine and Cheri's economic histories reflect ways in which women have agency in the exercise of livelihood strategies. They may choose to behave in accordance with Cultural Norms or ignore them. There is, however, a cost - either in lost autonomy or financial assistance. 9.3.4 Class Effects: Differences among Women with Substantial Mobility Differences in the availability of Assets among women mût Substantial Mobility reflects their socioeconomic Class. Women with Substantial Assets, are from relatively wealthy families. This is evident in a variety of factors. Women in this sub-group have three forms of Assets: personal human capital (education and training), the labor of others, and significant financial resources. Most women completed secondary school and a notable portion of Goun also have university education. Likewise, many received training in a craft or skill. Time invested in acquiring these forms of human capital are evident in the late starting work-age among women in this sub-group. Starting work-age for women with Substantial Assets differs significantly from those with Limited Assets, many of whom indicated they started working out of financial need. One seller of used clothing stated she started working "to help my husband with the household expenses.” Another woman selling used clothing said "my husband doesn’t stay at home. I am obliged to do everything to take care of my children." A woman who started work as a producer of prepared foods, and whose husband gave her $10 to start this activity, also said she did it "to take care of [her] children." Such responses suggest that these women’s current households lack resources. Another woman, who made and sold SodabP, stated that she started working because she had to take care of herself - "My mother could not do everything that she wanted to for me." This reply supports the finding that the families of women with Limited Assets are also relatively poor.

” Sodabi is a local alcoholic beverage made from distilled palm wine.

178 In contrast, parents' of women with Substantial Assets not only provided daughters with education and training, but are also the primary source of starting capital. This reflects that women with Substantial Assets are more likely than women in the whole sample to be raised by both parents. Mother’s occupations also differ markedly. The majority of these women's mothers work either as traders or in some profession. Personal human capital in the form of education and training provides women with Substantial Assets a wider range of choices in economic activities. Without education, specifically completion of secondary school and the ability to pass a state exam, women cannot work for the government. University education is also required for a number of managerial positions. Likewise, without proper certification, artisans do not have the legal right to open workshops and employ apprentices. The ability to hire apprentices can significantly increase the earning capacity of an artisanal activity. Not only do apprentices pay for training and are themselves a source of income, but they also increase the number of customers that can be served. Husbands of women in this group are not major financial contributors to their wives economic activities, but their assistance is apparent in other ways. Women’s economic activities are critical to family maintenance. Women's ability to provide substantial support, however, is grounded in parents’s assistance in obtaining assets, both human and financial capital. Life-histories of two women, Wasia and Blandine, both of whom use & Retailer- Self-employed-Occasional-Goods strategy, exemplify these processes at work. These interviews also reflect the ways in which Substantial Mobility and are deeply intertwined. Both women have been very mobile because their economic activities consisted of cross-border trade of wholesaling daily-use and, retailing occasional-use, goods. While the number of activities in which each woman engaged fluctuates over time, mobility is always a critical aspect of their enterprises. Wasia, began trading at 20. Her mother, a teacher, gave $100 to get started. With that she began buying food in Togo and Nigeria for sale in Cotonou.

179 I bought food in Lome, I came back to sell in Cotonou and I left for Nigeria to buy sugar in powder, peanut oil, a little of everything, candles, motor oil, I came back and sold in Cotonou, so I did the two like that. (Interview with Wasia, October 1998).

In addition to her mother’s help in starting capital, Wasia’s father’s assistance was also critical to her activities. He worked as a customs officer for the Béninois government. As such, he was able to help avoid paying customs by taking her to Nigeria once a week to buy goods. She noted that when she tried this trade without his help she did not earn nearly as much.

I went with my, my friend, we, we bought and came back to sell in the big market of Cotonou. But there, it did not last, I paid the customs, I paid the cost of transport...! didn’t earn much, you know. (Interview with Wasia, October 1998)

This commerce was very successful and with every trip she was able to put money in her savings account. "I put a least SIO, $20 up to $15, $20 every week" (Interview with Wasia, October 1998). After three years of this trade she accumulated $600 in capital. At that point Wasia decided to begin trading in Cote d’Ivoire instead, because she felt she would be able to earn more money. Once a month she flew to Abidjan with occasional goods such as fabric, jewelry, and beauty products for sale to bureaucrats. Although there were limits to how much Wasia could take without paying for overweight baggage Wasia had two ways of working around this.

One did the business, sometimes one takes the package, one puts it on the back, attaching it like the children, arriving...or there is the business, you know and so there are people at the airport, you go quickly when you have a plane. You should take a plane at 6:00 PM, you should be there at 1:00 PM or 2:00 PM you are there. There are young nmners, when you see them, you tell them, well, I must travel to that country, well I have an extra fifty kilos. They say, well, wait, I will go see the passengers, the people who don’t have enough kilos, well I can negociate and after you give them something. (Interview with Wasia, October 1998)

1 8 0 Young men in the airport would get other passengers to carry some packages for her. Wasia then collected them after customs in Abidjan. So, she rarely paid for extra baggage or customs. Again, this was a very successful activity for her because Wasia earned between $160 and $200 for her $600 in capital, excluding all her expenses such as the plane fare, taxis, food, etc. Wasia did not pay rent in Abidjan as she stayed with a family fnend. Throughout the seven years that Wasia engaged in these two activities, she lived with parents. Although she did not pay for rent or food, Wasia contributed to household expenses. She also paid for personal needs such as clothing. This is explains Wasia's ability to save so much and increase capital. After marrying at age 26, Wasia moved to Saudi Arabia because that was where her husband worked. Even after moving she continued to trade. According to Wasia, in Saudi Arabia women are not allowed to trade in the markets or on the streets, as they are in West Africa. So, only once a year during the Hadj, she went to hotels where African pilgrims stayed, selling souvenirs and fabric to them. She also worked as a maid. Wasia explained the need to earn a living and contribute to the household budget in terms of obligations to her husband. She contrasted expectations of Béninois women as wives with those of Arab women, who she said are treated

as a jewel by their husbands, something precious, them, they are jealous to let their wives work someplace, in case men come and see, chat with men, all that it isn’t good in Islam. So, they prefer that the wife stays at home, closed up and he does everything for her...Well, here, you can’t do that because the mister will be tired, there are the children there, you want to buy salt, give me ten cents, you want to buy pepper, give me 25 cents, oh the love will leave you like that...even if you ask a little, he will tell you, if you are a woman and you don’t do anything, you don’t work, you don’t move, you are not a courageous woman, love will leave. It is like that, it is necessary, it is necessary to have a little on your side.. Jf you take 100 cases, you will find 98 like that. You see some, the women who are ugly, they are loved by their husbands because she sells that. You see women who are pretty but she doesn’t have work, she doesn’t have a trade, she doesn’t have anything, but two days after, the husband will go cra^, he will go see the ugly one because she when he says, well I need that, I want to buy that, she will bring it out. (Interview with Wasia, October 1998)

181 Although this quote referred in general to women in Benin, she went on to explain that expectations of women are greater among the Yoruba.

The Fon, the Goun, the Mina, the wife gives a little, the husband gives a little, but among the Yoruba often the majority, when you get married they give you the dowery in a little plate, there is money for doing trade, when he gives you, it is finished, you never ask for anything again...if you are a good wife, it is necessary to know how to find money. (Interview with Wasia, October 1998)

These explanations are consistent with what many other women in the study expressed, but particularly evident among Yoruba women who are more likely to use strategies that require mobility. Due to Yoruba expectations of women as income earners, husbands do not hinder wive’s economic activities, especially by restricting her mobility. In some instances they even facilitate it. Upon their return to Benin, Wasia began trading in both imported food and seasonal occasional goods. In September, for example, with the start of school she sold school supplies. In December and January, with the holiday season, she switched to fabric and liquor. She goes several times a week to Cotonou to purchase supplies. If her husband also has business there, he drives her. In addition, when he travels to Saudi Arabia or France, if she gives him money he buys things for her to sell. Blandine, another trader in occasional goods had a similar experience. She began trading in fabric at age 12 because that was what her mother did. Although, Blandine did not have Wasia’s education, her mother helped get started by giving her $200 worth of fabric from her own supply. Blandine started accumulating her own capital with profits firom this trade. Although she said she would never stop selling fabric because that was what she learned from her mother, upon marriage she began adding other types of trade to her economic strategy.

When I married, then, I did, I began to go to my country, there where my father was bom in Nigeria there to buy timber, com, then I came to sell here

182 with my husband because my husband also was a trader. (Interview with Blandine, October 1998)

In addition

At that time, there was a moment occasionally when I went to buy tomatoes which came &om Lome, I, I rented a truck, I bought, I came to sell here, wholesale and it was like that I sold that com there, it was by season. (Interview with Blandine, October 1998)

Throughout Blandine's marriage she engaged in four activities, fabric and timber regularly, com and tomatoes seasonally. Her husband, who also sold timber, helped in this activity by providing transportation. In part, this assistance can be understood as reciprocation for Blandine's providing contacts in her father’s home village that enabled them to engage in trade. 9.3.5 Conclusions Differences in access to Assets is evident both among women with Limited Assets and ones with Substantial Mobility. Among women with Limited Assets, Ethnicity effects reveal that Yoruba women are in general better educated and have more family support than the Goun. These differences reflect variation in family backgrounds in that Yomba women are more likely to have been raised by both parents and have mothers who work full-time as traders. Moreover, upon marriage Yoruba women receive money for investment in commercial activities. As a result these women have human capital, and also significant financial capital relative to their Goun counterparts. These pattems reflect Cultural Norms among the Yoruba that require women to be economically independent of their husbands. In some instances Goun women may overcome lack of financial resources by ignoring social expectation. This has costs if they face a budgetary crisis, such that all capital may be lost. Yoruba women also may lose family support by ignoring social expectations. In the case discussed above, this involved a change of religion that resulted in disownment from the family. Hence, women's "purposeful choices" in the face of Cultural Norms shapes their opportunities.

183 Although Yoruba women with Limited Assets are better oÉf than Goun counterparts, women of both ethnicities are not as well off as counterparts with Substantial Assets. These differences are primarily attributable to Class effects evident in family background. Virtually all women in with Substantial Assets were raised by both parents, and their mother's occupations are more likely to include full-time trade or professional work. As a result, these women's families have greater resources with which to provide daughters with both human and financial capital. While only Yoruba women with Limited Assets have access to human and some financial assets, women of all ethnicities with Substantial Assets have access to all three forms of Assets - their own human capital, others' labor, and significant financial resources. These conditions afford them a wide range of choice in economic activity including wage employment and artisanal work within the formal rule structure. The latter enterprise is considerably more lucrative than craft activity outside the legal framework because it enables women to hire apprentices, themselves a source of income, and whose labor increases the scale of activities. Women's greater choice in economic activity also affords them a high level of economic independence. In this sense. Mobility and/fsMA are deeply intertwined.

9.4 Conclusions Life circumstances of women with Substantial Mobility reveal that specific interactions among Dimensions o fDifference, manifest in Structures o f Constraint and localized in residential arrangement and location, produce distinctive opportunity structures. Social processes that shape differences in Assets among women with Substantial Mobility reflect geographically localized intersections in Ethnicity, Class, and Life-cycle Phase. For Yoruba women social norms reflecting gender ideology include women’s productive activities as part of their reproductive responsibilities. These do not change over the life-course. As a result, parents and husbands will do everything they can to facilitate women's economic independence.

184 For Goun women, on the other hand, social norms related to gender ideology are variable over the life-course, and geographically in residential arrangement and location. Women living under customary values, more prevalent among extended family settings, on the urban fringe are able to be mobile when young and unmarried or older after ten to fifteen years of marriage. Goun women living in more urban environments, usually in nuclear family or female-headed household settings have considerably more autonomy in economic activities irrespective o îLife-cycle Phase. For all women within tliis group Class also plays a significant role in facilitating mobility, but is a double-edged sword. Earnings of women with Substantial Assets are often critical to household sur\'ival; hence, husbands are less likely to impinge on mobility. Again, the ways in which social processes shape women’s opportunities are socially, geographically and temporally situated manifesting in distinctive interactions among all four Structures o f Constraint, Asset distributions. Formal Rules, Cultural Norms and Personal Preferences. In all cases, women express "purposeful choices" in configuring their livelihood strategies, but they are conditioned by the social context within which women live.

185 CHAPTER 10

CONCLUSION: EXPLAINING LIVELIHOOD STRATEGY VARUTION

10.1 Introduction This chapter serves four purposes. First, findings of this study are summed up to provide a comprehensive answer to the two questions asked at the beginning - how and why do people earn a living?; what are the consequences of different approaches to livelihood strategy configuration? Second, methodological and theoretical contributions are summarized. Third, development implications of the investigation are considered. Fourth, future research avenues are discussed. 10.2 Explaining Livelihood Variation: How do people earn a living? The central goal of this dissertation is explaining variation in livelihood strategy usage and outcome by elaborating processes giving rise to different strategy configurations (Conceptual Framework -- Chapter 3). To accomplish this, the study (a) identifies thirteen discrete strategy types used by women in Porto Novo, (b) differentiates two critical Enabling Conditions - Mobility and Assets, (c) examines strategy outcomes in terms of accumulation relative to Enabling Conditions, (d) explores how distinct intersections among Dimensions ofDijference, manifest in Structures o f Constraint, produce Localized Enabling Conditions shaping strategy configuration. Women's opportunities in livelihood strategy configuration, and hence, their ability to earn a living, emerge firom localized social institutions (Structures o f Constraint) within their households and communities. While Asset’s are important. Mobility is even more so. Assets impact type and scale of enterprises, but Mobility

186 enables women to overcome limited resources under some conditions, and facilitates their use under others. Analysis results are synthesized in the next three sub-sections to explain differential strategy use and outcome among women in Porto Novo. This reveals the critical importance of Mobility and ways in which life-circumstances shape economic opportunity. Before outlining women's differential access to Assets and degree of Mobility, it is first necessary to examine pattems in livelihood strategy configuration, whether there are distinguishable variations, and Enabling Conditions from which they emerge. The purpose is to provide a basis for exploring disparate strategy outcomes and ways in which life circumstance shape opportunity. 10.2.1 Identifying Strategies and Enabling Conditions Analysis began by identifying types of livelihood strategy used by women in Porto Novo (Chapter S). Identification of strategy configuration is grounded in women's differentiation among economic enterprise and strategy types in terms of type (Table 5.1, Column 4), number and importance of activity. Non-hierarchical cluster analysis, based on a similarity measure, then compared and grouped women relative to type and number of activities. This revealed thirteen distinct strategy types. Subsequently, CHAID analysis identified the particular Enabling Conditions from which strategy configurations emerge. A separate model was used for each of the three factors ~ Assets, Cultural Norms, Location/Mobility - thought to facilitate or inhibit strategy use. Variables representingvlsse/x and Mobility were identified as most strongly associated with strategy type. This is interpreted to mean that these two conditions are critical for strategy creation. Then, strategies were grouped based on amount of and degree of Mobility required for strategy usage - Limited Mobility/Limited Assets, Limited Mobility/Substantial Assets, Substantial Mobility/Limited Assets, Substantial Mobility/Substantial Assets. A fourth aggregate model including all variables revealed that while Assets are important. Mobility is even more so.

187 10.2.2 Strategy Accumulation Potential Relative to Enabling Conditions Having identified basic pattems in strategy type and Enabling Conditions finm which they emerge, accumulation potential of strategy groupings was addressed (Chapter 7). Comparison of median starting and current capital for the four groups validated the importance of Mobility. Although all groups displayed some capital growth, levels were considerably higher for the two with Substantial Mobility. This indicates that access to Assets notwithstanding, ability to generate significant revenues is primarily a fimction of ability to access suppliers outside Porto Novo. Responses to open-ended survey questions and interview data revealed why this is the case. Strategies requiring Substantial Mobility involve trade in as-needed and occasional goods. While these items are not sold regularly, they offer considerably higher profits than daily-use goods. In combination with other activities that provide regular income, they provide opportunities for capital infusion. Benefitting firom this revenue potential, however, requires obtaining supplies at their source. Given that these goods are usually available at wholesale prices outside Porto Novo, Mobility is required to engage in strategies including of this type of trade. In addition, this analysis reveals that accumulation potential cannot be understood in dichotomous terms - mamabenz/petty trader, survival/surplus. Rather, strategy outcomes are best expressed as a continuum ranging firom survival (limited capital growth), to stability (slight capital growth), to surplus (significant capital growth). Differences at each level reflect women's ability to cope with financial crisis. Limited increases in capital provide few opportunities to save. Under these conditions, in the event of an emergency (e.g. medical crisis) women will be forced to use capital, thereby diminishing opportunities in economic activity. Women who experience this often work as maids to rebuild capital. In contrast, stable and significant capital growth provide greater opportunities for savings. Faced with a crisis women will have more resources and, therefore, are less likely to lose capital. While strategies m ût Limited Mobility only provide opportunities for survival or low stability, those with Substantial Mobility enable

18 8 high stability and surplus. Accordingly, strategy outcomes can be directly attributable to women's degree of Mobility. 10.2.3 Explaining Strategy Configuration Opportunities Having identified women's strategy configurations, Enabling Conditions firom which they emerge and accumulation potential thereof, the question remains - how do women's life circumstances shape strategy configuration opportunities. The conceptual firamework (Chapter 3) argues that interactions among Dimensions o f Difference (Gender, Class, Ethnicity, Life-cycle Phase) manifest in localized social institutions (Structures o f Constraint - Assets, Cultural Norms, Formal Rules, Personal Preferences), which in turn produce the Enabling Conditions from which strategy configuration emerges. Descriptive statistics and content analysis of open-ended survey questions and interview data reveals this is indeed the case (Chapters 8 and 9). Descriptive statistics facilitate identification of themes reflecting women's life circumstances. Interview data then provided illustrations of these localized social processes at work. Women with Limited Mobility are predominantly Goun, therefore Ethnicity is not an issue. Rather, variation among these women reflects interactions between Life-cycle Phase and Class. At every life-stage women in this group are not mobile outside Porto Novo. Reasons for this vary considerably as a function of household configurations and location. Class also explains Limited Mobility differences among women in their access to Assets. Irrespective of life-cycle stage many women in this group limit Mobility because their circumscribed resources did not warrant accessing suppliers outside Porto Novo. Further, some women, especially employers of the Artisan-Self-employed strategy, have access to personal human capital in the form of craft skills. Limited financial capital, which characterizes all women in this group limits ability to make effective use of one's craft. Class and Life-cycle effects reveal differences among women in access to Assets as well. Middle-aged women have access to the unpaid labor of their children. This enables them to manage multiple activities simultaneously, thereby providing somewhat better earnings opportunities.

189 Differences among women Substantial Mobility y in contrast, reflect interactions among all operant Dimensions o f Difference - Ethnicity, Class, and Life­ cycle Phase. Yoruba women are prevalent in this group, reflecting Cultural Norms embedded in their history of long-distance trade. Here, Ethnicity is the dominant explanation for women's Mobility. Goun women within this group are mobile either because they are young and unmarried, their residential arrangement offers assistance with familial responsibilities, or because they are wealthy enough to have maids. In their case, interactions among Ethnicity, Class and Life-Cycle Phase, situated in residential arrangement and location, provide opportunities (or Mobility outside Proto Novo. Although these women are better off than counterparts with Limited Mobility, clear differences in access to Assets reveal fundamental class differences reflected in upbringing. Those with Limited Assets have some human capital upon which to draw in configuring livelihood strategies. In contrast, those with Substantial Assets have both human and financial capital, which affords a much wider range of choice in economic activity. In each of these instances, women's opportunities in strategy configuration, and hence their ability to earn a living reflect distinct interactions among Dimensions o f Difference, manifest in localized social institutions (Structures o f Constraint), which produce Enabling Conditions from which strategies emerge. Moreover, the importance of geographic processes to understanding women's opportunities are revealed. Women's residential arrangements and locations are the spatial expression of Cultural Norms and Asset distributions. Also, these social, economic, and geographic processes impact women's Mobility differentially and, hence, their ability to earn a living. Each stage of analysis summarized above builds on preceding ones, contributing to an explanation of diverse livelihood strategies and their consequences. Women's livelihood strategies consist of a wide array of distinct configurations including mono­ activity forms, comprising only one economic activity, to those encompassing numerous distinct enterprises or Multiple Livelihood Economic Strategies ^4LES). Resulting outcomes are best expressed along a continuum ranging for survival (virtually no capital

190 growth), to stability (some growth in capital) to surplus (significant growth). These economic practices and their repercussions (re)produce socially and geographically situated social institutions, called here Structures o f Constraint, manifesting in two critical Enabling Conditions - Assets and Mobility. Women have differentially opportunities in strategy creation and outcome as a function of their access to Assets, and more importantly, their degree olMobility.

10.3 Methodological and Theoretical Contributions This study adds to a substantial body of work within feminist geography that demonstrates how geographically localized social institutions are (re)produced through people's daily social and economic practice resulting in differential economic opportunities. It simultaneously provides empirical validation to theorized claims about how interactions among Dimensions ofDifference, manifest in Structures o f Constraint produce material disparities in people's opportunities. In so doing it makes a significant methodological contribution regarding the operationalization of MLES in economic activity research. Most studies regarding economic activity emphasize a single, dominant source of income, even where multiple sources are recognized. In part this reflects the difficulty of mobilizing multiple economic activities methodologically. Cluster and CHAID analysis used here demonstrates that this is not impossible. While the definition of economic activity used in this study is specific to the Béninois context and therefore contingent, the general method used is not. Given the importance of MLES across the globe, an effective approach for incorporating multiple income sources into the study of economic activity is critical. Theoretically, this research extends that of feminist geographers (e.g. Hanson and Pratt 1995) to a developing country context, revealing that, as in developed world settings, geographic processes are at work shaping women's opportunities. These manifest in two ways, one of which is somewhat different firom findings of U.S. based work. First, as found elsewhere, geographic localization of social institutions are the spatial expression of interactions among Dimensions o f Difference. Second, U.S. based

191 research has demonstrates that women experience spatial fixity in activity spaces, but are not necessarily hindered in income generating opportunities due to firm responses to this. In other words, while women may not be mobile, they still have economic opportunity because firms desiring their labor located in close proximity. In contrast, in this developing country setting, economic opportunity can only be obtained through spatial mobility. This reflects the importance of self-employment for women in these settings. This work also provides empirical validation for claims by feminist theorists. Women's opportunities are, indeed, shaped by distinct interactions among Dimensions of Difference, manifesting in Structures o f Constraint, which are localized social institutions. This is evident in the variety of strategies employed by women in Porto Novo, outcomes thereof and the Enabling Conditions from which they emerge. Women do not all have access to the same economic opportunities. Choice in economic activity, and hence strategy configuration, reflects women's "purposeful choices" about how best to meet their productive and reproductive obligations as conditioned by their socioeconomic and geographic setting. Within this context Mobility outside Porto Novo is of critical importance to configuring strategies of stability and surplus. Degree of mobility, however is conditioned by distinct interactions among Class, Ethnicity, and Life-cycle Phase, relevant Dimensions o f Difference in this setting.

10.4 Development Implications Major findings of this work are as follows. First, people combine numerous economic activities to configure livelihood strategies. Second, ability to create profitable strategies reflects geographic processes in two ways, (a) localization of social institutions in households and communities, and (b) under some conditions the importance of Mobility for accessing particular kinds of social and economic networks. The first, is universal having already been demonstrated as relevant in developed country contexts. The second, while not necessarily applicable to the U.S. is likely to be important in other developing country contexts where self-employment is the main means of income generation. These findings address the four critical silences within development theory and practice identified in Chapter I.

192 First, they demonstrate the importance of appreciating indigenous strategies people within diverse socioeconomic settings create, specifically MLES. Many development programs (e.g. Grameen Bank) reduce people to a single economic activity. This reflects a failure to appreciate the ways in which MLES enables people to spread risk and ensure strategy adaptability by engaging multiple economic activities. Second, development theory and practice dichotomizes economic activity types based on the sector within which they occur (formal/informal). Findings here indicate that people do not just work in one or the other and moreover, that activities themselves may cross this arbitrary boundary. Use an MLES approach as delineated in this dissertation dispenses with such dichotomies, emphasizing instead the work people actually do and outcomes thereof. Attention to dichotomies reflects concern for where and for whom money is being generated. When activities occur outside government purview they are not a source of income for the state. Nonetheless, they may offer individuals significant opportunities for income generation. More importantly, however, these finding reveal that individual daily social and economic practice does make this distinction, rendering it irrelevant. Third, current development approaches frequently treat people in a uniform fashion, or only makes distinctions based on class and gender. Even women in development (WID) and gender and development (GAD) approaches tend to essentialize women assuming all are equally disadvantaged. While it is clearly true that women have fewer economic opportunities than men, this does equate with similar conditions of constraint. This study clearly demonstrates that women's livelihood strategy generate a range of earnings opportunities. Again, a strict dichotomy does not capture the full extent of possible outcomes. Moreover, that livelihood strategies emerge from particular Enabling Conditions, reflecting interaction among Dimensions o f Difference manifest in Structures o f Constraint suggest that development approaches need to consider the multiple intersections of difference shaping women's opportunities, not just Gender and Class, In so doing the full range of accumulation potential from various strategies will be revealed.

193 Fourth, Enabling Conditions from which economic opportunity emerges, are poorly understood. This is evident in the emphasis on Assets prevalent in most development programs. For example, Grameen Bank style approaches emphasize acquisition of Western style business skill and access to financial capital. This not only demeans women's abilities to initiate and maintain their own enterprises, but also fails to recognize the importance of geographic processes. As demonstrated in this work. Mobility is often the critical factor shaping women's opportunities in strategy configuration and yet, it is rarely considered. This finding suggests that development practitioners need to consider all possible Enabling Conditions, especially the geographic. They should consider the kinds of opportunities to which women have access. In so doing they may find that providing women opportunities is not as much a matter of cash as it is access to particular kinds of social and economic networks.

10.5 Future Research Although this work has made considerable contributions methodologically and theoretically, much work remains to be done. This reflects two different, but complementary avenues of investigation. First, more research is needed on the ways in which women's life-circumstances shape their economic opportunities. Toward this end, I plan to investigate further the processes behind strategy formulation. Three issues are of particular interest: change over time in strategy use; the spatial and temporal dimensions of various strategies relative to social institutional setting and outcome; and the interplay between social and economic networks and the specific spatialities of different types of strategies. Second, the methodology developed here offers a the possibility for comparative analysis, which will enable continued examination of how social structures shape strategy composition and outcome. This, in turn, will be used to develop a generalizable fiamework for studies of the interrelationship between social institutional setting, gender ideology and economic strategy.

194 SOURCES

Afonja, S. 1986." Changing Modes of Production and the Sexual Division of Labor among the Yoruba". in E. Leacok and H. I. Safa, Eds. Women's Work: Development and the Divsion o fLabor by Gender. South Hadley, MA: Bergin & Garvey Publishers, Inc.

Afshar, H., Ed. 1991. Women, Development and Survival in the Third World. Essex, England: Longman.

Afshar, H. and C. Dennis, Eds. 1992. Women and Adjustment in the Third World. New York: St. Martin's Press.

Aina, 0 . 1.1991. Women in Business: A Study o fFemale Entrepreneurs in South Western Nigeria. Ibadan, Nigeria, The Social Science Research Council of Nigeria.

Akindele, A. and C. Aguessy. 1953. Contribution a I'Etude de THistoire de l'Ancien Royaume de Porto Novo. Dakar, Senegal: Institut Français d'Afrique Noire.

Alden, J. and R. Spooner. 1982. Multiple Job Holders: An analysis o f second jobs in the European Community. Luxembourg, Office for Official Publications of the European Community.

Aldrich, H., E. and R. Waldinger. 1990." Ethnicity and Entrepreneurship." yinnW Review o fSociology 16:111-135.

Allen, C. 1989." Benin", in C. Allen, et. al., Eds. Benin, The Congo and Burkina Faso: Politics, Economics and Society. London and New York: Printer Publishers: 1- 144.

Amirault, T. 1997." Characteristics of Multiple Jobholders, 1995.” Monthly Labor Review (120): 9-15.

Ammott, T. and J. Matthaei. 1996. Race Gender and Work: A Multi-cultural Economic History o f Women in the United States. Boston: Southend Press.

195 Aspaas, H. R. 1998." Heading Households and Heading Businesses: Rural Kenyan Women in the Informal Sector.” Professional Geographer 50(2): 192-204.

Awe, B. 1977." The lyalode in the Traditional Yoruba Political System", in A. Schelgel, Ed. Sexual Stratification: A Cross Cultural View. New York: Columbia University Press.

— 1989," Nigerian women and Development in Retrospect", in J. Parpart, Ed. Women and Development in Africa. Lanham, MD: University Press of America, Inc.

Bailey, T. C. and A. C. Gatrell. 1995. Interactive Spatial Data Analysis. Essex, Englad: Addison Wesley Longman, Ltd.

Berger, I. and E. F. White. 1999. Women in Sub-Saharan Africa: Restoring Women to History. Bloomington, IN: Indiana University Press.

Bin, L. 1992." Second Job Fever Spreading All Over.” Beijing Review: 6-7.

Blackley, P. R. 1990." Spatial Mismatch in Urban Labor Markets: Evidence 6om Large US Metropolitan Areas.” Social Science Quarterly 71(1): 39-52.

Blaikie, P., T. Cannon, et. al. 1994." Access to Resources and Coping in Adversity", in P. Blaikie, et. al., Eds. At Risk: Natural Hazards, People's Vulnerability, and Disasters. London and New York: Routledge: 46-71.

Boserup, E. 1970. Woman's Role in Economic Development. New York: St. Martin's Press.

Bourdieu, P. 1977. Outline ofa Theory o fPractice. Cambridge: Cambridge University Press.

Brown, B. B. 1989." Women in Botswana", in J. L. Parpart, Ed. Women and Development in Africa: A Comparative Perspective. New York and London: University Press of America, Inc.

Brown, L. A., J. L. Mandel, et. al. 1997." Development Models, Economic Adjustment, and Occupational Composition: Ecuador 1982-1990.” International Regional Science Review 20(3): 183-209.

Bunch, C. 1990." Making Common Cause: Diversity and Coalitions", in L. Albrecht and R. Brewer, Eds. Bridges o fPower: Women's Multi-cultural Alliances. Philadelphia: New Society Publishers: 49-56.

196 Camey, D., Ed. 1998. Sustainable Rural Livelihoods: What contribution can we make? London: Department of International Development.

Clark, G. 1991." Colleagues and Customers in Unstable Market Conditions: Kumasi, Ghana.” Ethnology 30(1): 31-48.

—. 1994. Onions are My Husband. Chicago: University of Chicago Press.

— 1997." Local-Global Interactions in Ghana's Structural Adjustment", in R. E. Blanton, et. al., Eds. Economic Analysis Beyond the Local System. New York and London: University Press of America, Inc.: 209-234.

Collins, P. H. 1990. Black Feminist Thought: Knowledge, Consciousness, and the Politics o fEmpowerment. New York and London: Routledge.

Coombes, B. and H. Campbell. 1996." Pluriactivity in (and beyond?) a Regulationist Ctisis." New Zealand Geographer 52(2): 11-17.

Comia, G. A., R. Jolly, et. al., Eds. 1987. Adjustment with a Human Face. Oxford: Clarendon Press.

Coughenour, C. M. and L. Swanson. 1983." Work Statuses and Occupations of Men and Women in Farm Families and the Structure of Farms.” Rural Sociology 48(1): 23-43.

Daddieh, C. K. 1989." Production and Reproduction: Women and Agricultural Resurgence in Sub-Saharan Africa", in J. L. Parpart, Ed. Women and Development in Africa: Comparative Perspectives. New York and London: University Press of America, Inc.: 165-193.

Davison, J. 1996. Voices from Mutira. Boulder, CO: Lynne Rienner Publishing, Inc. de Oliveira, 0. and B. Roberts 1994." The Many Roles of the Informal Sector in Development: Evidence from Urban Labor Market Research, 1940-1989". in C. A. Rakowski, Ed. Contrapunto. Albany, N.Y.: State University of New York Press.

Deniel, R. 1985. Femmes des Villes Africaines. Abidjan, Cote d'Ivoire: Inades Edtion.

Dennis, C. 1990." The limits to women's independent careers: gender in the formal and informal sectors in Nigeria", in D. Elson, Ed. Male Bias in the Development Process. Manchester and New York: Manchester University Press: 83-104.

197 — 1991." Constructing a Career under Conditions of Economic Crisis and Structural Adjustment: The Survival Strategies of Nigerian Women", in H. Afshar, Ed. Women, Development and Survival in the Third World. London and New York: Longman.

Deseran, F. A., W. W. Falk, et. ai. 1984." Determinants of Earnings of Farm Families in the U.S.” Rural Sociology 49(2): 210-229.

Dill, B. T. 1983." Race, Class and Gender: Prospects for an All-Inclusive Sisterhood.” Feminist Studies 9(1): 131-150.

Doss, C. R. 1996." Intra-household Resource Allocation in an Uncertain Environment.” American Journal o fAgricidtural Economics. 78: 1335-1339.

Douglas, M. and B. Isherwood. 1996. The World o fGoods: Towards and Anthropology o f Comumption. New York and London: Routledge.

Eades, J. S. 1980. The Yoruba Today. Cambridge: Cambridge University Press.

Ekechi, F. K. 1995." Gender, Economic Power: The Case of Igbo Market Women of Eastern Nigeria", in B. House-Midamba and F. K. Ekechi, Eds. African Market Women and Economic Power: The Role o f Women in African Economic Development: The Role o f Women in African Economic Development. London: Greenwood Press: 41-57.

Elson, D., Ed. 1990. Male Bias in the Development Process. Manchester and New York: Manchester University Press.

England, K. V. L. 1993a." Suburban Pink Collar Ghettos: The Spatial Entrapment of Women.” The Annals ofthe American Association o fGeographers 83(2): 225- 242'.

—. 1993b." Changing Suburbs, Changing Women: Geographic Perspectives on Suburban Women and Suburbanization.” Frontiers 14(1): 24-43.

Evans, H. E. and G. Pirzada 1995." Rural Households as Producers: hicome Diversification and the Allocation of Resources", in J. Baker and T. A. Aina, Eds. The Migration Experience in Africa. Uppsala: Nordiska AMkainstitutet: 65-108.

Everitt, B. 1993. Cluster analysis. London and New York: E. Arnold Halsted Press.

Fadipe, N. A. 1970. The Sociology o fthe Yoruba. Ibadan, Nigeria: Ibadan University Press.

198 Falola, T. 1995," Gender, Business, and Space Control: Yoruba Market Women and Power", in B. House-Midamba and F. K. Ekechi, Eds. African Market Women and Economic Power: The Role o f Women in African Economic Development. London: Greenwood Press: 23-40.

Fapohunda, E. R. 1983." Female and Male Work Profiles", in C. Oppong, Ed. Male and Female in West Africa. London: George Allen and Unwin: 32-53.

— 1988." The Non-pooling Household: A Challenge to Theory", in D. Dwyer and J. Bruce, Eds. A Home Divided: Women and Income in the Third World. Stanford, CA: Stanford University Press: 143-154. Fashoyin, T. 1989. Occupational Segregation and Women in a Recessed Labor Market: Nigeria.

Ferguson, A. 1991. Sexual Democracy. Boulder: Westview Press.

Flynn, D. K. 1996. Containing Cross-border Traders: National Difference and Gendered Mobility in a Benin-Nigeria Borderland. Afiican Studies Association Meetings, San Francisco.

Folbre, N. 1994. Who Pays for the Kids? Gender and the Structures o f Constraint. New York: Routledge.

Francis, E. 2000. Making a Living: Changing livelihoods in rural Africa. New York: Routledge.

Friedan, B. 1963. The Femine Mystique. New York: W.W. Norton Company.

Fritz, L. 1979. Dreamers and Dealers: An Intimate Appraisal o f the Woman's Movement. Boston: Beacon Press.

Galli, R. E. and U. Funk 1995." Structural Adjustment and Gender in Guinea-Bissau", in G. T. Emeagwali, Ed. Women Pay the Price: Structural Adjustment in Africa and the Caribbean. Trenton, NJ: Afirica World Press: 13-30.

Geisler, G. 1993." Silences Speak Louder than Claims: Gender, Household, and Agricultural Development in Southern Afiica.” World Development 21(12): 1965-1980.

Giddens, A. 1984. The Constitution o fSociety. Berkeley; University of California Press.

199 Gordon, A. A. 1996a." Women and Development", in A. A. Gordon and D. L. Gordon, Eds. Understanding Contemporary Africa. Boulder and London: Lynne Rienner Publishers.

—. 1996b. Transforming Capitalism and Patriarchy: Gender and Development in Africa. London: Lynne Rienner Publishers.

Gregory, D. and J. Urray, Eds. 1985. Social Relations and Spatial Structures. London: MacMillan.

Grossman, A. 1975." Special Labor Force Report - Multiple Job-holding in May 1974.” Monthly Labor Review 98(2): 60-69.

Guyer, J. 1.1994." The Spatial Dimensions of Civil Society in Afiica: An Anthropologist Looks at Nigeria", in D. R. John W. Harbeson, and Naomi Chazan, Ed. Civil Society and the State in Africa. London: Lynne Rienner Publishers: 215-229.

— 1995." Women in the Rural Economy: Contemporary Variations", in M. J. Hay and S. Stichter, Eds. African Women South o f the Sahara. Esses, U.K.: Longman.

Guyer, J. I. and P. E. Peters. 1987b." Introduction.” Development and Change 18:197- 214.

Habbe, S. 1957." Moonlighting and Its Controls.” Management Record: 234-237.

Hallberg, M. C., J. L. Findeis, et. al., Eds. \99\. Multiple Job-Holding among Farm Families. Ames, lA: Iowa State University Press.

Hamilton, G. 1985." Temporary migration and institutionalization of strategy.” International Journal oflntercultural Relations 9:405-425.

Hanson, S. and G. Pratt. 1995. Gender, Work, and Space. London and New York: Routledge.

Hawkins, D. M. and G. V. Kass 1982." Automatica Interaction Detection", in D. M. Hawkins, Ed. Topics in Applied Multivariate Analysis. Cambridge: Cambridge University Press: 269-303.

Hay, M. J. and S. Stichter, Eds. 1995. African Women South o fthe Sahara. New York: Longman Publishing.

Hayghe, H. V. and K. Michelotti. 1971." Multiple job-holding in 1970 and 1971.” Monthly Labor Review 94(10): 38-45.

200 hooks, b. \9% \.A in'tIA Woman: black women and feminism. Boston: Southend Press.

—. 1984. Feminist Theory from Margin to Center. Boston: Southend Press.

Horn, N. E. 1995." Women's Fresh Produce Marketing in Harare, Zimbabwe: Motivations for Women's Participation and Implications for Development", in B. House-Midamba and P. K. Ekechi, Eds. African Market Women and Economic Power: The Role o f Women in African Economic Development. Westport, CO and London: Greenwood Press: 141-155.

House-Midamba, B. 1995." Kikuyu Market Women Traders and the Struggle for Economic Empowerment in Kenya", in B. House-Midamba and F. K. Ekechi, Eds. African Market Women and Economic Power: The Role o f Women in African Economic Development. London: Greenwood Press: 100-119.

Igue, J. 0., Ed. 1995. L'Economie Béninoise sous Ajustment Structurel. Cotonou, Benin: LARES.

Igue, J. 0. and B. Soule, G. 1992. L'Etat entrepot au Benin: Commerce informel ou solution a la crise? Paris: Karthaia.

Ilbery, B., M. Healey, et. al. 1996." Agricultural Adjustment and Business Diversification by Farm Households.” Geography 81(4): 301-310.

Institut National de la Statitique et de l'Analyse Economic (INSAE). 1992. Deuxième Rencensement General de la Population et de l'Habitation. Cotonou, Benin, Ministère du Plan et de la Restructuration Economique.

—. 1996. Profil de Pauvreté et Caractéristiques Socio-Economiques des Ménagés de Porto Novo. Cotonou, Benin, Institut National de la Statistique et de l'Analyse Economique.

—. Round Table Conference o fPartners in the Economie and Social Development o f the Republic o fBenin. Cotonou, Benin, Ministry of Planning and Statistics.

Jamal, M. 1986." Moonlighting: Personal, Social, and Organizational Consequences.” Human Relations 39(11): 977-990.

Jayawardena, K. 1986. Feminism and Nationalism in the Third World. London: Zed Books, Ltd.

201 Katz, C. and J. Monk, Eds. 1993. Full Circles: Geographies o f Women over the Life Course. International Studies of Women and Place. London and New York: Routledge.

Kass, G. V. 1980." An Exploratory Technique for Investigating Large Quantities of Categorical Data.” Applied Statistics 29(2): 119-127.

Kertzer, D. I., Ed. 1989. Age Structuring in Comparative Perspective. Hove and London: Lawrence Erlbaum Associates, Publishers.

Kobayashi, A., Ed. 1994. Women, Work and Place. Montreal and Kingston: McGill- Queen's University Press.

Koopman, J. 1995." Women in the Rural Economy; Past, Present, and Future." inM. J. Hay and S. Stichter., Eds. African Women South o fthe Sahara. Esses, U.K.: Longman.

Kwan, M.-P. 1999a." Gender and Individual Access to Urban Opportunities: A Study Using Space-Time Measures.” Professional Geographer 51(2): 210-227.

—. 1999b." Gender, the Home-Work Link, and Space-Time Pattems of Non-employment Activities.” Economic Geography 75:370-394.

Laws, G. 1997." Women's Life Courses, Spatial Mobility, and State Policies", in J. P. Jones 111, et. al., Eds. Thresholds in Feminist Geography. New York and Oxford: Rowman & Littlefield Publishers, Inc.: 47-64.

Lawson, V. A. 1990." Work Force Fragmentation in Latin America and its Empirical Manifestation in Ecuador.” World Development 18(5): 641-657.

Little, K. 1973. African Women in Towns. Cambridge: Cambridge University Press.

Lobao, L. and K. Meyer. 1995." Economic Decline, Gender, and Labor Flexibility in Family-Based Enterprises: Midwestern Farming in the 1980s.” Social Forces 74(2): 575-608.

Loh, W.-Y. and Y.-S. Shih. 1997." Split Selection Methods for Classification Trees.” Statistica Sinica 7:815-840.

MacGaffey, J. 1988." Evading Male Control: Women in the Second Economy in Zaire", in S. Stichter and J. L. Parpart, Eds. Patriarchy and Class: African Women in the Home and Workforce. Boulder and London: Westview Press: 161-176.

202 Masi, A. C. 1987." Moonlighting Under the Italian Sun: A Series on Multiple-Job Holding.” Contemporary Sociology — A Journal o fReviews 16(4): 485-491.

Massey, D. 1994. Space, Place and Gender. Minneapolis, MN: University of Minnesota Press.

Massiah, J., Ed. 1993. Women in Developing Countries: Making Visible the Invisible. Oxford: Berg Publishers, Inc.

Matanmi, S. 1989. Demographic Factors and Labour Market Adjustment in Recession.

Matsepe, 1.1981." Women in the Struggle for Liberation in South Afiica". in D. Wiley and Isaacman, Eds. Southern African Society, Economy, and Liberation: 71-85.

Maula, J. 1997. Small-Scale Production o fFood and Traditional Alcoholic Beverages in Benin and Tanzania: Implications for the Promotion o fFemale Entrepreneurship. Piscataway, N.J.: Rutgers University Center for Alcohol Studies.

McCall, D. 1959." Trade and the Role of Wife in a Modem West Afiican Town", in A. Southhall and D. Forde, Eds. Social Change in Modern Africa. London and New York: Oxford University Press: 286-304.

MCCall, J. C. 1995." Social Organization in Africa", in P. M. Martin and P. O'Meara, Eds. Africa. Bloomington, IN: Indiana University Press: 175-189.

McDowell, L. 1997." Women/Gender/Feminisms: doing feminist geography.” JoMrna/o/ Geography in Higher Education 21(3): 381-400.

McGee, T. G. 1979." The Poverty Syndrome: Making out in the Southeast Asian City", in R. Bromely and C. Gerry, Eds. Casual Work and Poverty in Third World Cities. New York: John Wiley & Sons: 45-68.

McLafferty, S. and V. Preston. 1996." Spatial Mismatch and Employment in a Decade of Restructuring.” The Professional Geographer 48(4): 420-431.

Mingione, E. 1991. Fragmented Societies: A Sociology o fEconomic Life beyond the Market Paradigm. Oxford: Basil Blackwell, Ltd.

Moser, C. 0. N. 1992." Adjustment from Below: Low-Income Women, Time and the Triple Role in Guayaquil, Ecuador", in H. Afshar and C. Dennis, Eds. Women and Adjustment Policies in the Third World. New York: St. Martin's Press: 87-116.

203 Mott, P. E. 1965." Hours of Work and Moonlighting", in C. E. Dankert, et. al., Eds. Hours o f Work. New York: Harper & Row, Publishers: 76-94.

Murray, A. T. and V. Estivill-Castro. 1998." Cluster discovery techniques for exploratory spatial data analysis.” International Journal o f Geographical Information Science 12(5): 431-443.

Murray, C. 2000." Changing Livelihoods: The Free State, 1990s.” African Studies 59(1): 115-142.

Musisi, N. B. 1995." Baganda Women's Night Market Activities", in B. House-Midamba and F. K. Ekechi, Eds. African Market Women and Economic Power: The Role o f Women in African Development. London: Greenwood Press: 121-139.

Mustapha, A. R. 1992." Structural Adjustment and Multiple Modes of Livelihood in Nigeria. Authoritarianism, Democracy, and Adjustment. Uppsala: Nordiska Afrikainstitutet: 188-216.

Ngaza, M. K. and E. E. Koda. 1980. The Unsung Heroines. Dar es Salam, Tanzania: WRDP Publications.

Njoku, J. E. E. 1980. The World o f African Women. London: The Scarecrow Press, inc.

O'Barr, J. and K. Firmin-Sellers 1995." African Women in Politics", in M. J. Hay and S. Stichter, Eds. African Women South o fthe Sahara. Harlow, England: Longman Scientific & Technical: 189-212.

Ogbomo, 0 . W. 1995." Esan Women Traders and Pre-Colonial Economic Power", in B. House-Midamba and F. K. Ekechi, Eds. African Market Women and Economic Power: The Role o f Women in African Economic Development. London: Greenwood Press: 1-21.

Okine, V. 1993." The Survival Strategies of Poor Families in Ghana and the Role of Women Therein", in J. Massiah, Ed. Women in Developing Economies: Making Visible the Invisible. Oxford: Berg Publishers, Inc.

Olukoshi, A. 0. and H. Olukoshi 1995." Structural Adjustment and Female Wage Labor in the Nigerian Textile Industry", in G. T. Emeagwali, Ed. Women Pay the Price: Structural Adjustment in Africa and the Caribbean. Trenton, NJ: Afiica World Press, Inc.: 39-51.

204 Orlove, B. S. and H. J. Rutz, Eds. 1989. The Social Economy o fConsumption. The Social Economy of Consumption. Lanham, MD and London; University of Press of America, Inc.

Osirim, M. J. 1996." Women Micro-entrepreneurs in Harare and Bulawayo", in K. Sheldon, Ed. Courtyards, Markets, and City Streets: Urban Women in Africa. Boulder, CO: Westview Press.

Owusu, F. 2000. Livelihood Strategies, Economic Reform and Public Institutions: The Tale o f Two Medium-sized Towns in Ghana. Geography. Minneapolis, MN, University of Minnesota.

Palmer, 1.1991. Gender and population in the adjustment o fAfrican economies: Planning fo r change. Geneva: International Labour Office.

Parpart, J. L., Ed. 1989. Women and Development in Africa: Comparative Perspectives. Dalhousie African Studies Series. Lanham, MD: University Press of America, Inc.

Peil, M. 1975." Female Roles in West African Towns", in J. Goody, Ed. Changing Social Stnicture in Ghana. London: International African Institute.

Perrella, V. C. 1970." Moonlighters: their motivations and characteristics.” Monthly Labor Review 93(8): 57-64.

Pettman, J. J. 1996. Worlding Women: a feminist international politics. London: Routledge.

Polanyi, K. 1957." The Economy as Instituted Process", in K. Polanyi and H. Person, Eds. Trade and Market in the Early Empires. New York: The Free Press.

Porter, G. 1995." The impact of road construction on women's trade in rural Nigeria.” Journal o f Transport Geography 3(1): 3-14.

Pratt, 0 . and S. Hanson. 1991." On the Links between Home and Work: Family- Household Strategies in a Buoyant Labour Market.” International Journal o f Urban and Regional Research 15(1): 55-74.

—. 1991." Time, Space, and the Occupational Segregation of Women: a Critique of Human Capital Theory.” Geoforum 22(2): 149-157.

—. 1994." Geography and the Construction of Difference.” Gender, Place and Culture 1(1): 5-29.

205 Pred, A. 1984." Place as Historically Contingent Process.” Annals o f the American Association o f Geographers 74:279-297.

Quenum, M. 1983. Au Pays du Fons. Paris: Editions Maisonneuve et Larose.

Reardon, T., E. Crawford, et. al. 1994." Links Between Non-farm Income and Farm Investment in African Households: Adding the Capital Market Perspective.” American Journal o f Agricidtural Economics. 76: II72-1176.

Robertson, C. 1983." The Death of Mokola and Other Tragedies.” Canadian Journal o f African Studies.

— 1986." Women's Education and Class Formation in Africa, 1950-1980". in C. Robertson and I. Berger, Eds. Women and Class in Africa. New York and London: Africana Publishing Company: 92-116.

—. 1990. Sharing the Same Bowl. Ann Arbor, MI: University of Michigan Press.

— 1995." Women in the Urban Economy", in M. J. Hay and S. Stichter, Eds. African Women South o f the Sahara. Esses, U.K.: Longman Group, Ltd.: 44-68.

—. 1997. Trouble Showed the Way. Bloomington: Indiana University Press.

Robertson, C. and 1. Berger. 1986. Women and Class in Africa. New York and London: Africana Publishing Company.

Robertson, C. and M. A. Klien 1983." Women's Importance in African Slave Systems", in C. Robertson and M. A. Klien, Eds. Women and Slavery in Africa. Bloomington, IN: Indiana University Press.

Roldan, M. 1985." Industrial Outworking, Struggles for Reproduction of Working-class Families and Gender Subordination", in N. Redclifr and E. Mingione, Eds. Beyond Employment: Household, Gender and Subsistence. Oxford: Basil Blackwell, Ltd.: 248-285.

Ronen, D. 1975. Dahomey: Between Tradition and Modernity. Ithaca and London: Cornell University Press.

Rongguang, S. 1993." Intellectuals Facing the Business Sea.” Beijing Review: 27-28.

Rose, G. 1993b. Feminism and Geography: The Limits o fGeographical Knowledge. Minneapolis, MN: University of Minnesota.

2 0 6 Saul, M. 1995." Economic Life in African Villages and Towns", in P. M. Martin and P. O'Meara, Eds. Africa, Bloomington, IN: Indiana University Press: 190-210.

Schroeder, K. 2000." Spatial Constraints on Women's Work in Tarija, Bolivia.” The Geographical Review. 90(2): 191-205.

Scott, A. M. 1986." Industrialization, Gender Segregation and Stratification Theory", in R. Crompton and M. Mann, Eds. Gender and Stratification. Cambridge: Polity Press.

Seidman, G. W. 1984." Women in Zimbabwe: Postindependence Struggles.” Feminist Studies 10(3): 419-440.

Silverman, D. 1993. Interpreting Qualitative Data. London: Sage Publications.

Slater, R. 2001." De-industrialization, Multiple Livelihoods and Identity: Tracking Social Change in Qwaqwa, South Africa.” Journal o f Contemporary African Studies. 19(1): 81-92.

Soja, E. 1980." The Socio-Spatial Dialectic.” Annals o fthe American Association o f Geographers 70:207-225.

Spiro, H. 1987." Women Farmers and Traders in Oyo State, Nigeria", in J. Momsen and J. G. Townsend, Eds. Geography and Gender in the Third World. Chicago: Northwestern University Press.

SPSS. 1998. Answer Tree 2.0: Users Guide. Chicago: SPSS, Inc. Stark, 0 . 1991a. Migration in Developing Countries: Risk, Remittances, and the Family. Tel Aviv: Tel Aviv University, The David Horowitz Institute for the Research of Developing Countries.

—. 1991b. The Migration o fLabor. Great Britain: T.J. Press, Ltd.

Stichter, S. 1988." The Middle Class Family in Kenya: Changes in Gender Relations." in S. Stichter and J. Parpart, Eds. Patriarchy and Class. Boulder and London: Westview Press.

Stinson Jr, J. F. 1986." Moonlighting by women jumped to record highs.” Monthly Labor Review 109(11): 22-25.

—. 1987." Moonlighting: A key to differences in measuring employment growth.” Monthly Labor Review 112(2): 30-31.

207 —. 1990." Multiple job-holding up sharply in the 1980s.” Monthly Labor Review 113(7): 3-10.

—. 1997." New data on multiple job-holding available from the CPS.” Monthly Labor Review: 3-8.

Tardits, C. and C. Tardits 1962." Traditional Market Economy in South Dahomey", in P. Bohannen and G. Dalton, Eds. Markets in Africa. Chicago: Northwestern University Press.

Thrift, N. 1983." On the Determination of Social Action in Space and Time.” Environment and Planning D: Society and Space 1:23-57.

Tienda, M. 2000." Immigrants' Income Packaging and Invisible Labor Force Activity.” Social Science Quarterly 81(1): 291-310.

Tomieri, F. 1997. Etude Analytique sur le Statut de la Femme et la Loi au Benin. Cotonou, Benin, World Bank.

Trager, L. 1988. The City Connection: Migration and Family Interdependence in the Philippines. Ann Arbor, Michigan: The University of Michigan Press.

—. 1998." Home-town linkages and local development in south-western Nigeria - Whose agenda? What impact?” Africa 68(3): 360-382.

Valdivia, C., E. G. Dunn, et. al. 1996." Diversification as a Risk Management Strategy in an Andean Agro-pastoral Community.” American Journal o f Agricultural Economics. 78:1329-1334.

Van Allen, J. 1976." 'Aba Riots' or Igbo 'Women's War'? Ideology, Stratification, and the Invisibility of Women", in Hafkin and E. Bay, Eds. Women in Africa: Studies in Social and Economic Change. Stanford, CA: Stanford University Press: 58-.

VerEecke, C. 1995." Muslim Women Traders of Northern Nigeria: Perspectives from the City of Yola". in B. House-Midamba and F. K. Ekechi, Eds. African Market Women and Economic Power: The Role o f Women in African Economic Development. London: Greenwood Press: 59-79.

Vickers, J., Ed. 1991. Women and the World Economic Crisis. London: Zed Books, Ltd.

Vogel, C. and J. Smith. 2000. Evidence for and Determinants o f Social Resiliance in Arid Ecosystems. Berlin, Dahlem Workshop.

208 Waldinger, R., H. Aldrich, E., e t al. 1990. Ethnie Entrepreneurs: immigrant business in industrial societies. Newbury Park, CA: Sage Publications.

Weber, M. 1968. Economy and Society. New York: Bedminster Press.

Werbner, P. 1984." Business on Trust: Pakistani entrepreneurship in the Manchester garment trade", in R. Ward and R. Jenkins, Eds. Ethnic communities in business: Strategies for economic survival. Cambridge: Cambridge University Press: 166- 188.

WGSG. 1984. Geography and Gender: An introduction to feminist geography. London: Hutchinson in association with The Explorations in Feminism Collective.

—. 1997. Feminist Geographies: Explorations o fDiversity and Difference. Essex, England: Longman.

Wilson, F. R. 1982." Reinventing the Past and Circumscribing the Future: Authenticité and the Negative Image of Women's Work in Zaire", in E. Bay, Ed. Women and Work in Africa. Boulder, CO: Westview Press: 153-170.

Wipper, A. 1995." Women's Voluntary Associations", in M. J. Hay and S. Stichter, Eds. African Women South o fthe Sahara. Harlow, England: Longman Press: 164-188.

Young, G. and B. J. Dickerson, Eds. 1994. Color, Class, and Country: Experiences o f Gender. London: Zed Books, Ltd.

Young, I. M. 1990. Justice and the Politics o fDifference. Princeton: Princeton University Press.

Zack-Williams, A. 1995." Crisis and Structural Adjustment in Sierra Leone: Implications for Women", in G. T. Emeagwali, Ed. Women Pay the Price: Structural Adjustment in African and the Caribbean. Trenton, NJ: Africa World Press, Inc.: 53-61.

209 APPENDIX A

FRENCH VERSION OF SURVEY

210 QUESTIONNAIRE DE LONG SONDAGE

ATTRIBUTS PERSONNELS

Nom:

Prénom:

Groupe Ethnique: GOUN FON YORUBA MINA AUTRE,

Année de Naissance:

Quel âge avez-vous?

Quel est l’âge de votre plus grand enfant?

Stratégie Economique de Pluri-Activités

Avez-vous fréquenté? OUI NON *SI OUI - Jusqu'à quel niveau? PRIMAIRE SECONDAIRE SUPERIEUR

Avez-vous appris autre chose? OUI NON ♦SI OUI Quoi? COUTURE COIFFURE PATISSERIE SOUDURE PEINTURE AUTRE

*Qui vous a payé la formation?

♦Quelle est la durée de votre formation?

♦Pourquoi avez-vous choisi ce métier?

Quelle est votre activité économique principale?

♦Où r exercez-vous?

211 * Depuis quand exercez-vous cette activité?

Faites-vous d’autres choses qui vous rapportent de Tareent? OUI NON *SI NON - Pourquoi?

*SI OUI - Quelles sont-elles? POUR CHACUNE PRÉCISER LE TYPE D'ACTIVITE, LE LIEU,DEPUIS QUAND TYPE QUARTIER DEPUIS QUAND

SI LE COMMERCE EST INCLUS Pourquoi vendez-vous ces choses? HERITAGE TRADITION RAISON CULTURELLE REVENUS POTENTIELS AUTRE,

Où est-ce que vous achetez vos marchandises?

Comment arrivez-vous à faire venir les clients?

Oui vous a appris comment faire le commerce? PARENTS AMI CENTRE DE FORMATION, AUTRE

*Quelle était la durée de la formation?

* Avez-vous payé la formation? OUI NON

212 *Avez-vous travaillé ensemble? OUI NON *SI OUI -Avez-vous partagié les bénéfices? OUI NON

POUR LES COMMERÇANTS ET ARTISANS Ouelle était votre source de financement pour chaque activité? MARI PARENTS TONTINE PRET USURE PRET BANCAIRE___ ASSOCIATION EPARGNE PERSONNELLE.

CREDIT DON AUTRE

*Avec combien avez-vous commencé?

^Combien tournez-vous?

Travaillez-vous pour quelqu'un d’autre? OUI NON ♦SI OUI- Qui? PARENTS AMI GOUVERNEMENT PRIVE. AUTRE___

♦Que faites-vous?

♦Où?

♦Etes-vous rémunéré? OUI NON ♦SI OUI - Comment?

Faites vous aussi l’agriculture? OUI NON ♦SI OUI - ♦Quel genre?

♦Où?

♦Vendez-vous les produits? OUI NON. *SI OUI • Où les vendez-vous?

213 *Avec quelle fréquence?

Faites vous rélevage?QUI NON_ *SIOUI- *DeQuoi?

♦Où?

♦Vendez-vous les produits? OUI NON_ *SI OUI - Où les vendez-vous?

♦Avec quelle fréquence?

Etes-vous aussi membre d’un groupement? OUI NON ♦SI OUI - ♦Quelles sont les activités du groupement?

♦Avec quelle fréquence travaillez-vous dans le groupement?

♦Est-ce que vous gagnez individuellement de l’argent? OUI NON____

♦Quel autre intérêt avez-vous dans le groupement?

Avez-vous des emplovés? OUI NON ♦SI OUI-Combien?

♦Qui sont-ils?

♦Avec quelle fréquence travaillent-ils pour vous?

214 *Que font-ils pour vous?

* Est-ce que vous les payez? OUI NON

*Leur apportez-vous d'autres formes de soutiens? OUI NON____ ♦SI OUI - Lesquels? NOURRITURE, HABITS, ETC.

Pourquoi avez-vous choisi ces endroits pour vos activités? FAITES ATTENTION IL S’AGIT DE TOUTES LES ACTIVITÉS

Ouelle est votre fréquence à chaque endroit?

Comment organisez-vous vos activités?

Comment vous déplacez-vous pour chaque activité?

Combien de temps passez-vous pour chaque activité?

215 SI IL Y A PLUS D’UNE SOURCE DE REVENU Pourquoi avez-vous plus d'une activité économique?

Précisez l’ordre d’importance de chacune de vos activités?

* Pourquoi?

Laquelle aimez-vous le plus?

♦Pourquoi?

Pourquoi avez-vous choisi ces activités?

Laquelle est la plus rentable?

Pensez-vous que cette stratégie vous avantage? OUI NON_ ♦SI OUI - Comment?

♦SI NON - Pourquoi?

216 Avez-vous quelque chose d’autre qtiî vous rapporte de l'arpent? RIEN TERRES BATIMENTS BOUTIQUES VOITURES MOTOS USTENCILES CHAISES MOULINS AUTRE

Ouels sont vos revenues?

•Combien utilisez-vous pourvus besoins? PAR JOUK/SEMAINE/MOIS

•Combien économisez-vous? PAR JOUR/SEMAINE/MOIS

•Comment? TONTINE ASSOC. MUTUELLE CAISSE VILLAGEOISE CAISSE DE CREDIT. CAISSE D’EPARGNE BANQUE___AUTRE_

•Avez-vous déjà prêter? OUI NON. •SI OUI - Pourquoi?

*SI TONTINE OU ASSOC. MUTUELLE SONT MENTIONNEES •Quels sont les liens entre les membres?

•Combien cotisez-vous? PAR JOUR/SEMAINE/MOIS

•Combien de fois ramassez-vous?

•Pourquoi?

Ouelles sont vos objectives pour: •Vous-même:

•Vos enfants:

•Vos activités économiques:

217 Quelles difficultés rencontrez-vous dans l'exercice de vos activités économiques?

Histoire Economique A Quel âge avez-vous commencé par gagner votre propre argent?

*En quoi faisant?

*0ù?

* Quelle était votre source de financement au début?

*Avec combien aviez-commencé?

* Continuez-vous d’exercer cela? OUI NON *SI NON - Pendant combien de temps l'aviez-vous exercé?

* Pourquoi aviez-vous débuté cette activité?

* Pourquoi l’aviez-vous abandonnée?

*Aux bons moments quels étaient vos revenus?

*0ù est-ce que vous achetiez vos marchandises?

Quelles étaient vos activités économiques il v a dix ans?

218 *0ù les aviez-vous exercées?

"'Quelle était votre source de financement au début?

*Continuez-vous de les exercer? OUI NON *SI NON - Pendant combien de temps les aviez-vous exercé?

*Pourquoi aviez-vous débuté ces activités?

"■Pourquoi les aviez-vous abandonnées?

*Aux bons moments quels étaient vos revenus?

*0ù est-ce que vous achetiez vos marchandises?

Ouelles étaient vos activités économiques il v a cinq ans?

*0ù les aviez-vous exercées?

"■Quelle était votre source de financement au début?

"■Continuez-vous de les exercer? OUI NON *SI NON - Pendant combien de temps les aviez-vous exercées?

"■Pourquoi aviez-vous débuté ces activités?

219 * Pourquoi les aviez-vous abandonnées?

*Aux bons moments quels étaient vos revenus?

*0ù est-ce que vous achetiez vos marchandises?

Attributs Sociaux/Familiaux - Les attributs contextuels Où êtes-vous née?

Ouelle est l’origine de vos parents?

Où vivent-ils maintenant?

Oui vous a élevé?

Etes-vous mariée? OUI NON ___ *SI OUI - Depuis quand?

*Quel genre de marriage? UNION LIBRE CIVIL RELIGIEUSE

* Votre mari est-il polygame? OUI NON *SI OUI - Quel est votre rang?

Ouel est votre quartier de résidence?

* Depuis quand résidez-vous là?

*0ù est-ce que vous habitiez avant?

^Pourquoi avez-vous déménagé?

’•‘Etes-vous ici en permanence? OUI NON

Ouelle est votre religion? CATHOLIOUE PROTESTANTE MUSULMANE ANIMISTE AUTRE

220 *Avez-vous changé de religion? OUI_ NON *SI OUI - Pourquoi?

"Quand?

Nom Prénom Sexe Age Ethnie Religion Niveau Occupa Lien d’etudes tion paren ta!

Oui est le chef du ménage?

Oui est le propriétaire?

SI LES PARENTS NE FONT PAS PARTIE DU MÉNAGE Ouelle est la religion des parents/tuteurs? CATHOLIQUE PROTESTANTE MUSULMANE ANIMISTE AUTRE

Ouelle est/était l’occuoation des parents/tuteurs?

221 SI LE MARI NE FAIT PAS PARTIE DU MÉNAGE Pourquoi ne vïve7-vous pas avec votre mari?

Où habite-t-il?

Duel est son groupe ethnique? GOUN YORUBA FON AUTRE

Ouelle est son occupation?

Ouelle est sa religion? CATHOLIQUE PROTESTANTE MUSULMANE ANIMISTE AUTRE____

POUR LES ENFANTS Parmi les enfants du ménage, combien sont vos propres enfants?

Tous vos enfants vivent-ils avec vous? OUI NON ♦SI NON-Où vivent-ils?

♦Avec qui? PARENTS AMIS AUTRE,

♦Quels sont leurs âges?

A qui appartiennent les autres enfants du ménage? MARI PARENTS AMIS AUTRE

Où sont les enfants pendant que vous travaillez? A L’ECOLE A LA MAISON AVEC MOI AILLEURS

Oui s’occupe des enfants pendant que vous travaillez? MOI-MEME PARENTS AMIS DOMESTIOUE AUTRE

Oui vous aide à faire le ménage? CUISINE/NETTOYAGE/LESSIVE ENFANTS DOMESTIQUE PARENTS PERSONNE AUTRE

♦S’IL Y A DES DOMESTIQUES - Combien?

♦Est-ce qu’elles sont rémunérées? OUI NON ____

2 22 ♦SI OUI - comment?

♦Avec quelle fréquence vous aident-ils?

Ouelle est votre contribution dans le budget du ménage pour:

♦La Nouriture —

♦L’Habillement ~

♦Les Soins Médicaux --

♦Qui paye le reste?

Combien d’enfants fréquentent?

♦Ecole privée ou publique?

♦Qui paye la contribution scolaire?

♦Quel est le pourcentage de contribution de chacun?

Oui pave les fournitures et les uniformes?

♦Quel est le pourcentage de contribution de chacun?

Oui a construit la maison? OU Oui pave le lover?

Parmi toutes les choses du ménage. (LES MEUBLES. LES APPAREILS. LES USTENCILES. LES MOYENS DE DEPLACEMENTS, lesquelles avez-vous acheté vous même?

Supportez-vous d’autres personnes en dehors du ménage? OUI NON ♦SI OUI - Quel est votre lien parental?

223 *Qu’est-ce que vous leur donnez comme soutien?

NOÜRRITURE/HABITS/CONTRIBÜTION SCOLAIRE/SOINS MÉDICAUX/LOYER/ ARGENT

* Partagez-vous ces dépenses avec d’autres personnes? OUI NON____ *SI OUI - Quel est le pourcentage de chacune?

De quelle organisation êtes-vous membre?

Sociale:club sportif/association de développement/tontine/autre

Gouvernementale: parti politique

Communautaire: association de quartier/village

Professionelle: association professionelle/syndicat/groupement/ autre

Religieuse: groupe de prière/chorale/assalatou/autre

*Quel intérêt vous apporte chacune de ces associations? SOUTIEN MORALE SOUTIEN FINANCIER SOUTIEN PROFESSIONEL SOUTIEN MATÉRIEL____

*Cotisez-vous de l’argent pour chacune? OUI NON *SI OUI - combien?

MERCI BEAUCOUP POUR L’ASSISTANCE

224 APPENDIX B

ENGLISH VERSION OF SURVEY

225 LONG SUR VEY QUESTIONNAIRE

PERSONAL ATTRIBUTES

Last Name:

First Name:

Ethnic Group: GOUN FON YORUBA MINA OTHER_

Year of Birth:

How old are vou?

How old is vour oldest child?

ECONOMIC STRATEGY

Have vou gone to school? Yes NO ___ *IF YES - Until what level? PRIMARY SECONDARY COLLEGE

Have vou learned other trades? Yes No ___ ♦IF YES What? SEWING HAIRDRESSING PASTRY CHEF SOLDERING PAINTING OTHER___

♦Who paid for your training?

♦How long did the training last?

♦Why did you chose this trade?

What is vour principle source of income?

♦Where do you do that?

226 * Since when have you done that?

Do vou do other things to earn monev? YES NO____ *IFNO-Why?

♦IF YES - What are they? FOR EACH ACTIVITY SPECIFY THE TYPE, LOCATION AND DURATION TYPE NEIGHBORHOOD DURATION

IF COMMERCE IS INCLUDED Whv do vou sell these things? HERITAGE TRADITION CULTURAL REASONS REVENUE POTENTIAL OTHER

Where do vou buv vour merchandise?

How do vou attract clients?

Who taught vou how to trade? PARENTS FRIEND TRAINING CENTER OTHER

227 *How long was your training?

*Did you pay for the training? YES NO

*Did you work together? YES NO *IF YES -Did you share the profits? YES NO

FOR TRADERS AND ARTISANS What was vour source of financing for each activity? HUSBAND PARENTS TONTINE USARY LOAN BANK LOAN ASSOCIATION PERSONAL SAVINGS CREDIT GIFT OTHER

* With how much did you start?

*How much earn today?

Do vou work for someone else? YES NO ____ *IF YES-Who? PARENTS _____FRIEND GOVERNMENT PRIVATE SECTOR OTHER_____

*What do you do?

* Where do you work?

*Are you paid? YES NO *IFYES-How?

Do vou farm? YES NO *IFYES- * What type?

*Where?

228 *Do you sell the produce? YES N0_ *IF YES - Where?

*How often?

Do vou raise livestock? YES NO *IF YES - * What do you raise?

* Where?

*Do you sell the animals? YES N0_ *IF YES - Where?

*How often?

Are vou a member of a groupement? YES NO *IF YES - * What are the groups activities?

*How often do you work with the group?

*Do you earn individually from group activities? YES NO____

*What other interests do you have in the group?

Do vou have anv emplovees? YES NO *IF YES - How many?

229 *Who are they?

*How often do tliey work for you?

* What do they do for you?

*Do you pay them? YES N0_

*Do you give them any other kind of support? YES NO____ *IF YES - What kind? FOOD, CLOTHING, HOUSING, ETC.

Whv have vou chosen these locations for vour activities? GET INFORMATION ON ALL ACTIVITIES

How often are vou at each location?

How do vou organize vour work?

230 What is vour means of transport to each activity?

How much time do vou spend on each activity?

IF THERE IS MORE THAN ONE SOURCE OF INCOME Why do vou have more than one source of income?

Specify the order of importance of each of vour activities:

*Why?

Which do vou eniov the most?

*Why?

Why have chosen these activities?

231 Which is the most profitable?

Do vou think this strategy works well for vou? YES NO *IFYES-Why?

*IF NOT-Why not?

Do vou have anything else that earns vou money? NOTHING LAND ' BUILDINGS SHOPS CARS___ MOTORCYCLES UTENSILS CHAIRS GRINDER_OTHER_

How much do vou earn?

*Hovv much do you earn to meet your needs? BY DAYAVEEK/MONTH

*How much do you save? BY DAYAVEEK/MONTH

*How? TONTINE_____ MUTUAL ASSOC.___ VILLAGE BANK CREDIT AGENCY SAVINGS BANK COMMERCIAL BANK OTHER

*Have you borrowed money? YES NO *IFYES-Why?

*IF TONTINE OR MUTUAL ASSOC. ARE MENTIONED - *What is the relationship between members?

*How much to do you pay? BY DAY/WEEK/MONTH

♦How many times have you collected the tontine/mutual assoc.?

♦Why?

232 What are vour goals for: *Yourself:

*Your children:

*Your economic activities:

What difficulties have vou experienced in the operation of vour activities?

Economic History At what aee did vou start earning vour own monev?

* Doing what?

* Where?

♦What was your source of financing at the beginning?

♦With how much did you start?

♦Do you still do that activity? Yes No ♦IF NOT - For how long did you do that activity?

♦Why did you start that activity?

233 * Why did you stop that activity?

*At the high point how much did you earn?

* Where did you buy your merchandise?

What were vour economic activities ten years aeo?

* Where were they located?

* What was your source of financing for each?

*Do you still engage in that activity? YES N0_ *IF NOT - For how long did you engage in it?

* Why did you start those activities?

''Why did you stop those activities?

*At the high point how much did you earn?

* Where did you buy your merchandise?

234 What were vour economic activities five years aeo?

* Where were they located?

* What was your source of financing for each?

*Do you still engage in that activity? YES NO *IF NOT - For how long did you engage in it?

*Why did you start those activities?

* * Why did you stop those activities?

*At the high point how much did you earn?

* Where did you buy your merchandise?

Social/Familial Attributes Where were vou bom?

Where are vour parents from?

Where do thev live now?

Who raised vou?

Are vou married? YES N0_ *IF YES-Since when?

What kind of marriage? OPEN UNION CIVIL RELIGIOUS

235 *Is your husband polygamous? YES NO *IF YES - What is your rank? h what neighborhood do vou live?

* Since when?

* Where did you live before?

*Why did you move?

*Do you live here permanently? YES_ NO

What is vour religion?CATHOLIC PROTESTANT ISLAM ANIMIST OTHER

*Have you changed religion? YES NO ♦IF YES-Why?

'When?

Suecifv a I the members o f your household: Name First Sex Age Ethni Relig Level of Occupa Kinship Name city ion study tion relationship

236 Who is the head of vour household?

Who is the owner of vour house?

IF THE PARENTS ARE NOT PART OF THE HOUSEHOLD What is the religion of vour parents/guardians? CATHOLIC PROTESTANT ISLAM ANIMIST OTHER

What is the occupation of vour parents/eiiardians?

IF THE HUSBAND IS NOT PART OF THE HOUSEHOLD Why don’t vou live with vour husband?

Where does he live?

What is his ethnic groun? GOUN YORUBA FON OTHER_

What is his occupation?

What is his religion? CATHOLIC PROTESTANT ISLAM ANIMIST OTHER

FOR THE CHILDREN Amonp the children in the household how many are vour own?

Do vou all vour children live with vou? YES NO *IF NOT - Where do they live?

♦With who? FAMILY MEMBER FRIEND OTHER,

♦What are their ages?

To whom do the other children in the household belong? HUSBAND FAMILY MEMBER___ FRIEND OTHER

237 Where are the children while vou work? AT SCHOOL AT HOME WITH ME _ELSEWHERE___

Who takes care of them while vou work? MYSELF FAMILY MEMBER FRIEND MAID OTHER

Who helps vou with the housework? COOKING/CLEANING/LAUNDRY CHILDREN MAE) FAMILY MEMBER___ NO ONE OTHER

♦IF THERE ARE MAIDS - How many?

♦Are they remunerated? YES NO___ ♦IF YES-how?

♦How often do they help you?

What is vour contribution to the household budget for:

♦Food “

♦Clothing -

♦Health Care -

♦Who pays the rest?

How many children so to school?

♦Public or Private?

♦Who pays their fees?

♦What is the percentage of the contribution of each one?

Who pavs for their uniforms and school supplies?

♦What is the percentage of the contribution of each one?

238 Who built the house? OR Who pavs the rent?

Among all the things in the house. (FURNITURE. APPLIANCES. UTENSILS. VEHICLES! which did vou buv vourself?

Do vou support anyone outside vour household? YES N0_ *IF YES - What is your kinship relationship?

* What do you give them in the way of support? FOOD/CLOTHING/EDUCATIONAL EXPENSES/HEALTH CARE/RENT/MONEY

*Do you share these expenses with anyone? YES NO____ *IF YES - What is the percentage of the contribution of each?

Of what organizations are vou a member?

Social: sports club/development association/tontine/other

Governmental: political party

Community: neighborhood association/village

Professional: professional association/union/groupement/other

Religious: prayer group/chorus/assalatou/other

*What benefits do you get from each association? MORAL SUPPORT FINANCIAL SUPPORT PROFESSIONAL SUPPORT MATERIAL SUPPORT____

*Do you pay dues? Yes No *IF YES - How much?

THANK YOU FOR YOU HELP

239 APPENDIX C

DATA DICTIONARY

240 DATA DICTIONARY

Name Position

ID ID for Sorting Only Measurement Level; Nominal Column Width: 12 Alignment: Center Print Format: A 15 Write Format: A 15

CASENUM Case Number Measurement Level: Nominal Column Width: 6 Alignment: Center Print Format: A 12 Write Format: A 12

CHECKED : Measurement Level: Nominal Column Width: 6 Alignment: Center Print Format: A8 Write Format: A8

GROUPID Groupement ID Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: A8 Write Format: A8

ETHNI Ethnicity Measurement Level: Nominal Column Width: 6 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 Yoruba 2 Goun 3 Fon 4 Mina 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response

241 AGE Age 8 Measurement Level; Ordinal Column Width: 4 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

CHLDAGE Oldest Child's Age 9 Measurement Level: Ordinal Column Width: 4 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

AGERNGE Age Range 10 Measurement Level: Ordinal Column Width: Unknown Alignment: Center Print Format: F7 Write Format: F7 Missing Values: 995 thru 999

Value Label 1 15-20 2 21-25 3 26-30 4 31-35 5 36-40 6 41-45 7 46-50 8 51-55 9 56-60 10 61-65 11 66-70 12 71-75 13 76-80 14 81-85 15 86+ 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response

242 CHDAGRNG Child's Age Range 11 Measurement Level: Ordinal Column Width: 7 Alignment: Center Print Format: FT Write Format: FT Missing Values: 995 thru 999

Value Label 1 0-5 2 6-10 3 11-15 4 16-20 5 21-25 6 26-30 7 31-35 8 36-40 9 41-45 10 46-50 11 51-55 12 56-60 13 61-65 14 66-70 15 70+ 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

EDUC Education 12 Measurement Level: Ordinal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 none 2 primary 3 secondary 4 terminal 5 university 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response

243 TRADE Trade Learned 13 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 seamstress 2 hairdresser 3 crocheting 4 typing 5 computers 6 photographer 7 health-worker 8 traditional health-worker 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

TRNGPAID Trade Training Paid 14 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 parents 2 husband 3 herself 4 guardian 5 friend 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

244 TRNLNGTH Length of Training for Trade 15 Measurement Level: Scale Column Width: 4 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

TRADEWHY Why this trade 16 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value1 Label 0 other 1 parent’s choice 2 unsuccessful at school 3 vocation 4 preference 5 followed others 6 pays well 7 low investment 8 personal use 10 tradition II heritage 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

245 STRATEGY Strategy Configuration 17 Measurement Level: Ordinal Column Width: LI Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 ProdFamily 2 RetSelf-Daily I Act 3 ProdSelf-Daily 1-2 Acts 4 RetFamily-Daily 5 ProdSelf-Daily 2+ Acts 6 RetSelf-Daily 3+ Acts 7 RetSelf-Daily 2-3 Acts 8 RetSelf-As Needed 9 Artisan-Self-Employed 10 High Wage Public 11 Mixed-1 Act 12 Artisan-Employees 13 RetSelf-Luxury 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

246 1 Variables will be used as a proxy for all economic activity variables since these simply repeat for the other three activities*************

ACTIOCC Activity I ~ Occupation 18 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: FIO Write Format: FIO Missing Values: 995 thru 999

Value Label 0 other 1 wholesaler 2 retailer 3 producer 5 seamstress 6 hairdresser 7 photographer 8 sales clerk 9 money changer 10 teacher 11 bureaucrat 12 secretary 13 health-worker 14 trainer 15 other services 16 servant 17 visionary 18 law/military 19 healer 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

247 ACTIOCAT Activity 1 ~ Occupational Category 19 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 wholesaler 2 retailer 3 producer 4 artisan 5 service provider 6 hi-skill waged 7 low-skill waged 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable ACTIB Activity 1 - EconAct2 - typexclass 20 Measurement Level: Nominal Column Width: 10 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 whl-ownemp 2 whi-ownfam 3 whi-self 4 ret-ownemp 5 ret-ownfam 6 ret-self 7 prd-ownemp 8 prd'Ownfam 9 prd-self 10 artisan-ownemp 11 artisan-ownfam 12 artisan-self 13 service-ownemp 14 service-ownfam 15 service-self 16 waged-hiprv 17 waged-hipub 18 waged-loprv 19 waged-lopub 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

248 ACTIC Activity I ~ EconActS -typexclassxcom 21 Measurement Level: Nominal Column Width: Unknown Alignment: Right Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 whiemp-luxury 2 whiemp-asneeded 3 whiemp-daily 4 whiemp-producer 5 whlfam-luxury 6 whifam-asneeded 7 whifam-daily 8 whifam-producer 9 whiself-luxury 10 whiself-asneeded 11 whiself-daily 12 whiself-producer 13 retemp-luxury 14 retemp-asneeded 15 retemp-daily 16 retemp-producer 17 retfam-luxury 18 retfam-asneeded 19 retfam-daily 20 retfam-producer 21 retself-luxury 22 retself-asneeded 23 retself-daily 24 retself-producer 25 prdemp-luxury 26 prdemp-asneeded 27 prdemp-daily 28 prdemp-producer 29 prdfam-luxury 30 prdfam-asneeded 31 prdfam-daily 32 prdfam-producer 33 prdself-luxury 34 prdself-asneeded 35 prdself-daily 36 prdself-producer 37 artisan-emp 38 artisan-fam 39 artisan-self 40 service-emp 41 service-fam 42 service-self 43 hiwaged-prv 44 hiwaged-pub

249 45 lowwaged-prv 46 lowwaged-pub 47 lowwaged-com 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable ACTID Activity I-EconAct4-typexcom 22 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

'^alue Label I whl-luxury 2 whl>asneeded 3 whl-daily 4 whl-producer 5 ret-luxury 6 ret>asneeded 7 ret-daily 8 ret-producer 9 prd-luxury 10 prd-asneeded 11 prd-daily 12 prd-producer 13 artisan 14 service 15 hiskill-waged 16 lowskill-waged 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

250 ACTICOM Activity I - Commerce Category 23 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 non-trader 2 luxury goods 3 as-needed goods 4 daily-use goods 5 producer goods 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

ACTICLAS Principle Activity - Classification 24 Measurement Level: Nominal Column Width: 10 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 owner - employees 2 owner — family 3 self-employed 4 high-skilled private 5 high-skilled public 6 low-skilled private 7 low-skilled public 8 unpaid family 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

251 PACTYPE Principle Activity - Commerce Type 25 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 non-trader 2 school supplies 3 imported food 4 com 5 akassa 6 palm oil 7 gari 8 firewood 9 salt 10 prepared food 11 smoked fish 12 enamelware 13 fruit 14 bread 15 sewing notions 16 fabric 17 jewelry 18 petrol, products 19 snacks 20 Cigarettes 21 tomatoes 22 condiments 23 legumes 24 Sodabi 25 rice 26 beans 27 grains 28 plastic bags 29 coal 30 baby clothes 31 car parts 32 medicines 33 beauty supplies 34 soap 35 plastic products 36 paint supplies 37 onions 38 eggs 39 animals 40 other vegies 41 used clothes 42 yams

252 43 yogurt 44 mats 45 palm nuts 46 cross-border trade 47 drinks 48 liquor 49 ready-to-wear clothes 50 peanut oil 51 ice/water 52 empty bottles 53 coconut oil 54 under clothes 55 spices 56 religious articles 57 kid's games 58 hardware 59 shoes (tapettes) 60 fresh fish 61 used shoes 62 peanuts 63 manioc/maize 993 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response PAPTYPE Principle Activity - Production Type 26 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 non-producer 2 palm oil 3 gari 4 smoked fish 5 prepared food 6 akassa 7 cakes & snacks 8 gardening 9 animals 10 ready-to-wear clothes 11 soap 12 mats 13 yogurt 14 manioc 15 coconut oil 16 beauty supplies 17 ice/water

253 18 medicines 19 Sodabi 20 wigs 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

PALOCAL Principle Activity - Location 27 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 front gate 2 nbrhd stall 3 nbrhd ambulant 4 Adjarra Market 5 Ouando Market 6 Grand Marche 7 other market 8 PN-boutique 9 PN-ambulant 10 PN stall 11 PN-Office 12 Cot-Office 13 Cot-boutique 14 Dantokpa 15 Adjarra-Docodji 16 Catchi 17 Cotonou 18 Nigeria/Togo 19 Nigeria 20 road side 21 work-shop 22 school 23 bar 24 Ekpe 25 Congo/Benin/Nigeria 26 Ifangni 27 Gbokou Market 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

254 PALNGTH Principle Activity-Duration 28 Measurement Level: Scale Column Width: 4 Alignment: Center Print Format: F3 Write Format: F3 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

PALOCSUP Principle Activity - Location of Supplies 29 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 non-trader 2 in the nghbhd 3 ambulant sellers 4 personal contact at the house 5 middle person at the house 6 CARDER-PN 7 boutique -PN 8 gardens - PN 9 Zounkpa 10 Grand Marche 11 Ouando market 12 Adjarra Market 13 Cotonou 14 Takon 15 Adja 16 Oueme 17 Avrankou 18 Ketoukpe 19 Kitigbo 20 Ahouangbo 21 Dangbo 22 Ifangnin 23 producer in the region 24 Sakete 25 Ikpinie 26 Pobe 27 Ketou 28 other regions in Benin

255 29 Nigeria 30 Togo 31 Nigeria/Cotonou 32 Cotonou/PN 33 Nigeria/Oueme 34 Cotonou/Togo 35 Houime 36 Cotonou/Oueme 37 Gbokou Market 38 Nigeria/Benin border 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

PASORCAP Principal Activity - Source of Capital 30 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other I husband 2 parents 3 tontine 4 usurie loan 5 bank loan 6 personal savings 7 credit 8 gift 9 children 10 loans from friends/relatives 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

256 PSTRCAPR Principle Activity - Start-Up Capital - Range 31 Measurement Level: Ordinal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 0 -1000 2 1001-2000 3 2001-5000 4 5001 - 10000 5 10001 - 15000 6 15001 -20000 7 20001-30000 8 30001 -40000 9 40001 -50000 10 50001 - 100000 11 100001 - 150000 12 150001 -200000 13 200001+ 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

PSTRCAPA Principle Activity - Starting Capital - Actual 32 Measurement Level: Scale Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

257 PCURCAPR Principal Activity -Current Capital Range 33 Measurement Level; Ordinal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 0 • 1000 2 1001 -2000 3 2001 -5000 4 5001-10000 5 10001 - 15000 6 15001 -20000 7 20001 -30000 8 30001 -40000 9 40001-50000 10 50001 -100000 11 100001-150000 12 150001-200000 13 200001+ 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

PCURCAPA Principle Activity — Current Capital Actual 34 Measurement Level: Scale Column Width; Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

258 PALOCWHY Principle Activity-Location-Why 35 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format; F8 Missing Values: 995 thru 999

Value Label

0 other 1 lives near the market 2 the largest market close by 3 mother sold here 4 found a place in the market 5 family responsibilities 6 limits of age 7 health reasons 8 no place else works better 9 sells well 10 market niche 11 too many problems at the market 12 tool ittle to sell elsewhere 13 its an international market 14 on a major road 15 was moved there by the state 16 hasn’t found anyplace else 17 holding the place for someone else 18 pays no rent 19 restricted to house by husband family 20 husband will 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

PALOCFRQ Principle Activity - Location - Frequence 36 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 all the time 2 one time a week 3 every other day 4 market days 5 occasionally 6 5/6 days a week 7 one time a month 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

259 PATIME Principle Activity — Time Spent 37 Measurement Level; Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 all day 2 mornings 3 evenings 4 evenings and weekends 5 late nights 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

PAIMPORT Principle Activity-Order of Importance 38 Measurement Level: Ordinal Column Width: 3 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response

PAIMPWHY Principle Activity - Importance - Why 39 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 more profitable 2 sells quickly 3 trade learned 4 personal use 5 pays quickly 6 good market 7 less work/hassle 8 transportable 9 low investment 10 vocation 11 makes ends meet

260 12 only activity 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

PA WHY Principle Activity Why 40 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 available capital 2 heritage 3 habit 4 making ends meet 5 profitability 6 market niche 7 home use 8 parents' choice 9 education/training 10 quick sales 11 low investment 12 family responsibility/convenience 13 followed others 14 no waged labor 15 available credit 16 quick payment • no credit 17 preference 18 available equipment 19 ability 20 God’s will 21 hide income 22 unemployed 23 diversification 24 tradition 25 has savings 26 meets basic needs 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

261 OTHRACTS Other Activities 41 Measurement Level: Nominal Column Width: S Alignment: Center Print Format: F3 Write Format: F3 Missing Values: 995 thru 999

Value Label 1 yes 2 no 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

NUMACTS Number of Activities 42 Measurement Level: Scale Column Width: 3 Alignment: Center Print Format: F8 Write Fonnat: F8 Missing Values: 995 thru 998

Value Label 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 not applicable

PERIOD Period 135 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 seasonal 2 simultaneous 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

262 MRKTING Marketing 136 Measurement Level; Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 ambulant sales 2 welcoming client 3 samples 4 regular clientele 5 also sells compliment 6 God's will 7 offers credit 8 calls client 9 displays merchandise 10 word of mouth 11 reduces prices 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

SOURCAP Major Source of All Capital 137 Measurement Level: Nominal Column Width: 8 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 husband 2 family 3 tontine 4 usurie loan 5 bank loan 6 personal savings 7 credit 8 gift 9 children 10 loans from friends/relatives 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

263 TSTRCAPR Total Starting Capital - Range 138 Measurement Level: Ordinal Column Width: 8 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 0-1000 2 1001-2000 3 2001 -5000 4 5001 -10000 5 10001 - 15000 6 15001 -20000 7 20001-30000 8 30001 -40000 9 40001 -50000 10 50001 - 100000 11 100001 -150000 12 150001-200000 13 200001+ 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable TSTRCAPA Total Starting Capital - Actual 139 Measurement Level: Scale Column Width: 8 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

264 TCURCAPR Total Current Capital - Range 140 Measurement Level; Ordinal Column Width: 8 Alignment; Center Print Format; F8 Write Format; F8 Missing Values; 995 thru 999

Value Label 1 0-1000 2 1001-2000 3 2001 -5000 4 5001- 10000 5 10001 - 15000 6 15001 -20000 7 20001 -30000 8 30001 -40000 9 40001 -50000 10 50001-100000 11 100001 -150000 12 150001 -200000 13 200001+ 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable TCURCAPA Total Current Capital - Actual 141 Measurement Level; Scale Column Width; 8 Alignment; Center Print Format; F8 Write Format; F8 Missing Values; 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

265 COMTRNG Training for Commerce 142 Measurement Level; Nominal Column Width: Unknown Alignment; Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 parents 2 friend 3 training center 4 no one 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

CMTRGLNG 143 Measurement Level: Scale Column Width: 3 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable CMTRNGPD Commerce Training ~ Paid 144 Measurement Level: Nominal Column Width: 5 Alignment: Center Print Format: F8 Write Format; F8 Missing Values: 995 thru 999

Value Label 1 yes 2 no 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

266 CMSHDWRK Commerce Training — Shared Work 145 Measurement Level: Nominal Column Width: 5 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 yes 2 no 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

CMSHDPRF Commerce Training - Shared Profit 146 Measurement Level: Nominal Column Width: 5 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 yes 2 no 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

WAGEDLBR Waged labor 147 Measurement Level: Nominal Column Width: 5 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 yes 2 no 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

267 WORKTYPE Type of Work 148 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 maid 2 water bearer for masons 3 sales clerk 4 secretary 5 teacher 6 administrator 7 hairdresser 8 health aid 9 other service 10 servant 11 prays for people(visionary) 12 healf worker 13 caterer 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

PAID Paid labor 149 Measurement Level: Nominal Column Width: 5 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 yes 2 no 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

268 GRPMEMBR. Member of Groupement ISO Measurement Level; Nominal Column Width: 5 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 yes 2 no 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

GRP ACTS Groupement Activities 151 Measurement Level: Nominal Column Width; Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 gari 2 palm oil 3 gardening 4 livestock 5 crocheting 6 farming 7 syrups 8 animal husbandry 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

269 GRPWKFRQ Groupement Work Frequency 152 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 one time a week 2 2 times a week 3 3 times a week 4 2 times a month 5 I time a month 6 as needed 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

GRPBNFTS Benefits of Groupement 153 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 none 2 training 3 credit possibility 4 meals provided 5 personal contacts 6 grants 7 Govemment/NGO aid 8 personal use 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

270 EMPLYE Employees 154 Measurement Level; Nominal Column Width: 5 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 yes 2 no 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

NUMB EM P Number of employees 155 Measurement Level: Scale Column Width: 4 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

EMPRELT Relationship to Employees 156 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 no relation 2 children 3 younger siblings 4 other family 5 friends 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

271 WORKFREQ Work frequency of Employee 157 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 all the time 2 when not in school 3 as needed 4 5/7 days a week 5 momings&evenings 6 6/7days a week 7 occasionally 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

TASKS Employee Tasks 158 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 housework 2 help with sales 3 gardening 4 heavy labor 5 helps with sales&housework 6 apprentices 7 production 8 housework&production 9 health aid 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

272 PAYMENT Employees — Payment 159 Measurement Level; Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 owner pays 2 owner receives 3 no payment 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

OTHRREMU Form of Remuneration 160 Measurement Level; Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 none 2 paid/money 3 upkeep (food, clothing) 4 gifts 5 clothes occasionally 6 medical care 7 only food 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

273 TRANSPRT Mode of Transportation 161 Measurement Level; Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 on foot 2 moto 3 taxi 4 taxi and moto 5 personal moto 6 personal car 7 rented truck 8 husband's car/moto 9 on foot/moto/taxi 995 M refused to say 996 M incomputable 997 M doesn’t know • 998 M no response 999 M not applicable

NOMLESI Why no MLES - Response I 162 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 lack of capital 2 health problems 3 lack of desire 4 lack of time 5 not profitable 6 family heritage 7 sufficient income from primary act. 8 habit 9 family responsibility 10 market niche 11 available capital 12 husband’s will 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

274 N0MLES2 Why no MLES 2 163 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 lack of capital 2 health problems 3 lack of desire 4 lack of time 5 not profitable 6 family heritage 7 sufficient income from primary act. 8 habit 9 family responsibility 10 market niche 11 available capital 12 husband's will 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

MLESWHYl Why MLES? 164 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 making ends meet 2 increase income 3 flexibility 4 seasonality 5 personal preference 6 family heritage 7 social expectation 8 diversification 9 has savings 10 ability 11 salary insufficient 12 complimentary activities 13 profitability 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

275 MLESWHY2 Why MLES 165 Measurement Level; Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 making ends meet 2 increase income 3 flexibility 4 seasonality 5 personal preferences 6 family heritage 7 social expectation 8 diversification 9 has savings 10 ability 11 salary insufficient 12 complementary activities 13 profitability 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

ACTPRFT Most Protltable Activity 166 Measurement Level: Ordinal Column Width: 3 Alignment: Center Print Format: F3 Write Format: F3 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

SUCCESS Does the strategy work? 167 Measurement Level: Nominal Column Width: 5 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label J yes 2 no 3 more or less 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

276 WHYWORKS Yes-strategy works-why? 168 Measurement Level: Nominal Column Width: Unknown Alignment; Center Print Format: PS Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 meets basic needs 2 capital grows 3 good sales 4 has savings 5 increasing profits 6 regular clientele 7 meets wants 8 sells all merchandise 9 wide range of products 10 can meet problems 11 would have left 12 profitability 13 increase income 14 making end meet 15 have retirement 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

WHYDWORK whydwork 169 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 lack of profit 2 lack of capital 3 can't make ends meet 4 unmet desire 5 barely surviving 6 decreasing capital 7 competition 8 irregular profit 9 low sales 10 stagnant profit 11 increase in prices 12 lack of market 13 irregular sales

277 14 unmet goals 15 losses 16 lack of equipment 17 no produce 18 salary insufficient 995 M refiised to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

SATIS FAC Satisfaction with Strategy 170 Measurement Level; Scale Column Width: Unknown Alignment: Right Print Format: F8.2 Write Format: F8.2

Value Label 1.00 barely surviving 2.00 lack of capital 3.00 decreasing capital 4.00 lack of profit 5.00 stagnant profit 6.00 irregular profit 7.00 unmet goals/desires 8.00 making ends meet 9.00 meets wants 10.00 can resolve problems 11.00 have savings 12.00 increasing income 995.00 refused to say 996.00 incomputable 997.00 doesn’t know 998.00 no response 999.00 not applicable

OTHERINC Other income 171 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 none 2 land 3 building 4 cars 5 motorcycles 6 other 7 furniture

278 8 utensils 9 chairs 10 mills 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

WKLYREVR Weekly Revenue - Range 172 Measurement Level: Ordinal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label I 0 - 1000 2 1001 -2000 3 2001 -5000 4 5001 - 10000 5 10001 - 15000 6 15001 -20000 7 20001 - 30000 8 30001 -40000 9 40001 - 50000 10 50001 -100000 II 100001 -150000 12 150001 -200000 13 200001+ 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable WKLYREVA Weekly Revenue ~ Actual 173 Measurement Level: Scale Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

279 WKLYEXPR Weekly Expenses - Range 174 Measurement Level: Ordinal Column Width: 8 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 0•1000 2 1001-2000 3 2001-5000 4 5001 - 10000 5 10001-15000 6 15001 -20000 7 20001-30000 8 30001-40000 9 40001 -50000 10 50001 - 100000 11 100001 - 150000 12 150001 -200000 13 200001+ 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable WKLYEXPA Weekly Expenses ~ Actual 175 Measurement Level: Scale Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

280 WKLYSAVR. Weekly Savings ~ Range 176 Measurement Level; Ordinal Column Width: Unknown Alignment: Center Print Format: PS Write Format: F8 Missing Values: 995 thru 999

Value Label 1 0-1000 2 1001 -2000 3 2001 -5000 4 5001 - 10000 5 10001 - 15000 6 15001-20000 7 20001-30000 8 30001 -40000 9 40001-50000 10 50001 - 100000 11 100001- 150000 12 150001 - 200000 13 200001+ 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

WKLYSAVA Weekly Savings - Actual 177 Measurement Level: Scale Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

INCOME 178 Measurement Level: Scale Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

281 INCRANGE Income Range 179 Measurement Level: Scale Column Width: Unknown Alignment: Right Print Format: F82 Write Forniat: FS.2

Value Label 1.00 0-1000 2.00 1001-2000 3.00 2001 -5000 4.00 5001- 10000 5.00 10001 -15000 6.00 15001 -20000 7.00 20001 - 30000 8.00 30001 -40000 9.00 40001 -50000 10.00 50001 - 100000 11.00 100001 - 150000 12.00 150001 -200000 13.00 200001+ 995.00 refused to say 996.00 incomputable 997.00 doesn’t know 998.00 no response 999.00 not applicable

SAVEMODE Savings Mode 180 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 tontine 2 mutual association 3 village bank 4 credit bank 5 savings society 6 bank 7 personal bank 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

282 LOANS Loans 181 Measurement Level: Nominal Column Width: 5 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 yes 2 no 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

WHYBRRWD Why Borrowed 182 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 increase capital 2 ceremony 3 medical needs 4 school fees 5 meet needs 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable TONTRELT Relationship ofTontine Members 183 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Fonnat: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 none (tontinier) 2 colleagues 3 family 4 neighbors 5 friends 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

283 TONTAMTR Weekly Tontine Amount ~ Range 184 Measurement Level: Ordinal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 993 thru 999

Value Label 1 0 -1000 2 1001-2000 3 2001-5000 4 5001-10000 5 10001- 15000 6 15001 -20000 7 20001-30000 8 30001-40000 9 40001-50000 10 50001 - 100000 11 100001 - 150000 12 150001 -200000 13 200001+ 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

TONTAMTA Tontine Amount ~ Actual 185 Measurement Level: Scale Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

TONTPRD Period ofTontine 186 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 every month 2 one time a year

284 3 every three months 4 by turn 5 not yet 6 two time a year 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

TONTPURP Purpose ofTontine 187 Measurement Level: Nominal Column Width: 8 Alignment: Center Print Format: F8 Write Format: F8 Missing Values; 995 thru 999

Value Label 0 other 1 capital 2 ceremony 3 medical needs 4 school fees 5 buy clothes 6 celebration 7 buy someth ing for self 8 buy for the children 9 build house 10 pay the rent/utlilities 11 family contribution 12 buy a moto or car 13 cover overhead/resupply 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

ACTGOALS Activity Goals 188 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 none 2 add other activities 3 enlarge current commerce 4 change to less difficult work 5 retire

285 6 change to other activity 7 buy car 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

DIFFCLTS Difficulties 189 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 none 2 financial 3 low sales/slump 4 bad debt 5 hard work 6 lack of transport 7 lack of merchandise 8 customs 9 family financial demands 10 expensive merchandise 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

STRWKAGE Start Work Age 190 Measurement Level: Ordinal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 10-15 2 16-20 3 21-25 4 26-30 5 31-35 6 36-40 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

286 FRSTWRK First Work Activity 191 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: FIO Write Format: FIO Missing Values: 995 thru 999

Value Label 0 other 1 wholesaler 2 retailer 3 producer/wholesaler 4 producer/retailer 5 seamstress 6 hairdresser 7 photographer 8 sales clerk 9 money changer 10 teacher 11 bureaucrat 12 secretary 13 health-worker 14 trainer 15 other services 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

CHNGDACT Changed Work Activities 192 Measurement Level: Nominal Column Width: 5 Alignment: Center Print Format: F8 Write Format: FS Missing Values: 995 thru 999

Value Label 1 yes 2 no 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

287 ACTIOYRS Work Activity — 10 years ago 193 Measurement Level; Nominal Column Width: Unknown Alignment: Center Print Format: FIO Write Format: FIO Missing Values: 995 thru 999

Value Label 0 other 1 wholesaler 2 retailer 3 producer/wholesaler 4 producer/retailer 5 seamstress 6 hairdresser 7 photographer 8 sales clerk 9 money changer 10 teacher 11 bureaucrat 12 secretary 13 health-worker 14 trainer 15 other services 16 maid 17 healer 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

ACT5YRS Work Activity - 5 years ago 194 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: FIO Write Format: FIO Missing Values: 995 thru 999

Value Label 0 other 1 wholesaler 2 retailer 3 producer/wholesaler 4 producer/retailer 5 seamstress 6 hairdresser 7 photographer 8 sales clerk 9 money changer 10 teacher 11 bureaucrat

288 12 secretary 13 health-worker 14 trainer 15 other services 16 visionary 17 maid 18 healer 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

BRTHPLC Birth Place 195 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 Porto Novo 2 Avrankou 3 Dangbo 4 Bopa 5 Akpro-Misrete 6 Adjarra 7 Adjohoun 8 Seme-Kpodji 9 Abomey 10 Ketou 11 Cove 12 Cotonou 13 Bassila 14 Pobe 15 Allada 16 Ouidah 17 Sakete 18 Zangnannado 19 Abomey-Calavi 20 Nigeria 21 Togo 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

289 FAMORIG Family Origins 196 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other ■ 1 Nigeria 2 Porto Novo 3 Oueme 4 Zou 5 Attacora 6 Borgou 7 Atlantique 8 Togo 9 Ghana 10 Mono 995 M refused to say 996 M Incomputable 997 M doesn’t know 998 M no response FMCURRES Family’s Current Residence 197 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 dead 2 Porto Novo 3 Oueme 4 Zou 5 Attacora 6 Borgou 7 Mono 8 Atlantique 9 Togo 10 Ghana 11 Nigeria 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

290 RAISDBY Raised By 198 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 parents 2 mother only 3 father only 4 aunt/uncle 5 grandparent 6 family friend 7 siblings 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

MRTLTYPE Type of Marriage 199 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 never married 2 unofficial marriage 3 civil marriage 4 religious marriage 5 customary marriage 6 divorced 7 widowed 8 engaged 9 separated 10 civil/religious marriage 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

291 MARIDYRS Years of Marriage 200 Measurement Level: Scale Column Width: 4 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

POLYMRRG Polygamous Marriage 201 Measurement Level: Nominal Column Width: 5 Alignment: Center Print Format: FS Write Format: F8 Missing Values: 995 thru 999

Value Label 1 yes 2 no 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

POLYRANK Polygamous Marriage Rank 202 Measurement Level: Ordinal Column Width: 3 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

CURNBRES cumbres 203 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 Gbodje 2 Aizandjokome

292 3 Houezounme 4 Diobou 5 Zebou-aga 6 Agbokome 7 Aissoume 8 Gbezoukpa 9 Koutongbe 10 Houegbokome 11 Ita-Baglo 12 Lokossa 13 Hlinkome 14 Fiekome 15 Tokpota 16 Houeyogbe 17 Sokome 18 Kandevie 19 Gbokou 20 Sedjeko 21 Tokpanoussou 22 Dota 23 Davie 24 Ouando 25 Catchi 26 Carrefour Y 27 Maleseme 28 Ouenlinda 29 Houinme 30 Foun-Foun 31 Sadognon 32 llefie 33 Attake 34 Hassoukome 35 Akonaboue 36 Djegan -Kpevi 37 Guevie 38 Canipfoul 39 Tchivie 40 Degue -Gare 41 Togo 42 Danto 43 Dowa 44 Dodji 45 Zounkpa 46 Djegan-Daho 47 Gbecon 48 Hlogou 49 Hounsouko 50 Adjarra-Docodji 51 Djrado 52 Dangbeklounon 53 Ghana 54 Cotonou

293 55 Avakpa 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

QTRESLTH Resident Current Nieghbothood - Length 204 Measurement Level: Scale Column Width: 4 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

PRVNBRES Previous Neighborhood of Residence 205 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 Gbodje 2 Aizandjokome 3 Houezounme 4 Diobou 5 Zebou-Aga 6 Agbokome 7 Aissoume 8 Gbezounkpa 9 Koutongbe 10 Houegbokome 11 Ita-Baglo 12 Lokossa 13 Hlinkome 14 Fiekome 15 Tokpota 16 Houeyobge 17 Sokome 18 Kandevie 19 Gbokou 20 Sedjeko 21 Tokpanoussou 22 Dota 23 Davie

294 24 Ouando 25 Catchi 26 Carrefour Y 27 Maleseme 28 Ouenlinda 29 Houinme 30 Foun-Foun 31 Sadognon 32 llefie 33 Attake 34 Hassoukome 35 Akonaboe 36 DJegan-Kpevi 37 Guevie 38 Campfoul 39 Tchinvie 40 Degue-Gare 41 Togo 42 Danto 43 Dowa 44 Dodji 45 Zounkpa 46 Degan-Daho 47 Gbecon 48 Hlogou 49 Hounssouko 50 Adjarra-Docodji 51 Djrado 52 Dangbeklounon 53 Ghana 54 Cotonou 55 Avakpa 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response WHYCGRE Why Changed Residence 206 Measurement Level; Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 Joined husband 2 left husband 3 built house 4 changed job 5 bigger apartment 6 increased rent

295 7 in-law problems 8 family problems 9 other wives 10 change house 11 proximity to work 12 health reason 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

PERM RES Permanent Residence 207 Measurement Level: Nominal Column Width: 5 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 yes 2 no 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

RELIGON RELIGON 208 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 none 2 Catholic 3 Protestant 4 Celeste 5 Muslim 6 Vodoun 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

296 CHNGDREL Changed RELIGON 209 Measurement Level: Nominal Column Width: 5 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 yes 2 no 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

WHYCGREL Why Changed Religion 210 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 married 2 children 3 health reasons 4 preference 5 god’s will 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

CHGRELYR Changed Religion - Number of Years 211 Measurement Level: Scale Column Width: 3 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

297 NBCLDHH Number of Children in Household 212 Measurement Level: Scale Column Width: 7 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable PEROWNCH Percent Own Children in Household 213 Measurement Level: Scale Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

NBOWNCHD Number of Own Children in Household 214 Measurement Level: Scale Column Width: Unknown Alignment: Right Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

298 NBDEPADT Number of Dependant Adults in Household 215 Measurement Level; Scale Column Width: 4 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

NBOTHADT Number of Other Adults in Household 216 Measurement Level: Scale Column Width: 4 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

HHAGERNG Age Range of Children in Household 217 Measurement Level: Ordinal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 0-10 2 11-20 3 0-20 4 5-15 5 20+ 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

299 HHEAD Head of Household 218 Measurement Level; Nominal Column Width: Unknown Alignment: Center Print Format: FS Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 parent 2 husband 3 in-law 4 herself 5 together 6 siblings 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

RESOWNER Owner of Residence 219 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 renting 2 husband 3 herself 4 parent 5 in-law 6 children 7 family 8 husband/herself 9 siblings 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

300 MTHRRELG Mother’s Religion 220 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 none 2 Catholic 3 Protestant 4 Muslim 5 Vodoun 6 Celeste 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

FTHRRELG Father’s Religion 221 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 none 2 Catholic 3 Protestant 4 Muslim 5 Vodoun 6 Celeste 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

301 MTHROCC Mother's Occupation 222 Measurement Level; Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value1 Label 0 other 1 dead 2 retired 3 none 4 trader 5 petty trader 6 bureaucrat 7 teacher 8 health worker 9 seamstress 10 hairdresser II farmer 12 other artisan 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

FTHROCC Father’s Occupation 223 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 dead 2 retired 3 none 4 farmer 5 tailor 6 trader 7 barber 8 mechanic 9 chauffeur 10 law/military officer II bureaucrat 12 construction 13 petty trade 14 other artisan 15 teacher 16 traditional healer 17 fisherman

302 18 health-worker 19 tontinier 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response HUSBNDHH Husband-Household Member 224 Measurement Level; Nominal Column Width: 5 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 yes 2 no 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

HSBHHWHY Why Husband is not a Member of the Household 225 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 works elsewhere 2 lives with his family 3 lives with other wife 4 no place there 5 fighting 6 left him 7 traveling 8 hasn’t paid dowery 9 dead 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

303 HUSBRES Husband’s Residence (ifnot in household) 226 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 same concession 2 same neighborhood 3 other PN neighborhood 4 Cotonou 5 other country 6 in prison 7 other Benin department 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

HUSBETHN Husband’s Ethnic Group 227 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 Goun 2 Yoruba 3 Fon 4 Mina 5 Nagot 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

304 HUSBOCC Husband's Occupation 228 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Fonnat: FS Missing Values: 995 thru 999

Value Label 0 other - 1 retired 2 none 3 farmer 4 tailor 5 trader 6 barber 7 mechanic 8 chauffeur 9 law/military officer 10 bureaucrat 11 construction 12 petty trade 13 other artisan 14 traditional healer 15 cook 16 teacher 17 salesman 18 computer technician 19 fisherman 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

HUSBRELG Husband's Religion 229 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label

0 other 1 none 3 Catholic 4 Protestant 5 Muslim 6 Vodoun 7 Celeste 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

305 999 M not applicable

OTHRCHLD Other Children - Not in Household 230 Measurement Level: Nominal Column Width: 5 Alignment: Center Print Format: F8 Write Fonnat: F8 Missing Values: 995 thru 999

Value Label 1 yes 2 no 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

NBOTRCHD Number of Children not in Household 231 Measurement Level: Scale Column Width: 4 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

RSOTRCHD Residence of Other Children not in Household 232 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 other relative 2 child’s father 3 friend 4 adult/married 5 boarding school 6 apprenticeship 7 maid 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

306 OTRCH AGE Other Children’s Age Range 233 Measurement Level: Ordinal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 0-5 2 6-10 3 11-15 4 16-20 5 21-25 6 26-30 7 30+ 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

HHOTRCHD Relationship to Other Children in the Household 234 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 husband 2 relative 3 friend 4 no relation 5 adopted 6 children 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

WORKCHLD Location of Children while Working 235 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 at school 2 at the house

307 3 with me 4 elsewhere 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

CHLDCARE Childcare while Working 236 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 herself 2 relative 3 friends 4 maid 5 husband 6 themselves 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

HSEWRK Housework Help 237 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 children 2 maid 3 relatives 4 no one 5 husband 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

308 NMBMAIDS Number of Maids 238 Measurement Level; Scale Column Width: 4 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

MAIDREMU Maid's Remuneration 239 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Fonnat: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 paid money 2 none 3 upkeep 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

MAIDFREQ Maid’s Work Frequency 240 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 all the time 2 5/6 days a week 3 as needed 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

309 CHHFOOD Percentage of Contribution to Household Food Budget 241 Measurement Level: Scale Column Width: 4 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

HHFOOD Percent Contribution to Household Food 242 Measurement Level: Ordinal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value1 Label 1 10% or less 2 11% to 20% 3 21% to 30% 4 31% to 40% 5 41% to 50% 6 51% to 60% 7 61% to 70% 8 71% to 80% 9 81% to 90% 10 91% to 100% 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

CHHCLTHS Percentage of Contribution to Household Clothes Budget 243 Measurement Level: Scale Column Width: 6 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

310 HHCLTHS Percent Contribution to Household Clothes 244 Measurement Level: Ordinal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 10% or less 2 11% to 20% 3 21% to 30% 4 31% to 40% 5 41% to 50% 6 51% to 60% 7 61% to 70% 8 71% to 80% 9 81% to 90% 10 91% to 100% 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

CHHMEDS Percentage of Contribution to Household Medical Budget 245 Measurement Level: Scale Column Width: 7 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

HHMEDS Percent Contribution to Household Medicines 246 Measurement Level: Ordinal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label I 10% or less 2 11% to 20% 3 21% to 30% 4 31% to 40% 5 41% to 50% 6 51% to 60%

311 7 61% to 70% 8 71% to 80% 9 81% to 90% 10 91% to 100% 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

OTRCNTRB Other Primary Contributor to Household Budget 247 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 husband 2 parent 3 other relative 4 children 5 friend 6 siblings 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

NBCHLPUB Number of Children in Public School 248 Measurement Level: Scale Column Width: 4 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

312 NBCHLPRV Number of Children in Private School 249 Measurement Level: Scale Column Width: 4 Alignment: Center Print Fomiat: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

CCHLDEDU Percentage of Contribution to Children's Education Costs 250 Measurement Level: Scale Column Width: 4 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label I its free 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

CCSCHLSP Percentage of Contribution to Children's School Supplies 251 Measurement Level: Scale Column Width: 4 Alignment: Center Print Format: F8 Write Fonnat: F8 MissingValues: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable

313 OTRCTRED Other Primary Contributor to Children’s Education 252 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 husband 2 parent 3 other relative 4 children's parents 5 friend 6 siblings 7 children's father 8 oldest children 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

OTHRDPDT Other Dependants outside Household 253 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 parent 2 children 3 siblings 4 in-laws 5 other relatives 6 friend 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

314 SUPMODE Mode ofSupport for Dependants 254 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 upkeep 2 money 3 food 4 clothes 5 education 6 medical care 7 rent 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable CONTRSUP Percentage of Contribution to Support of Dependants 255 Measurement Level: Scale Column Width: 4 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 995 M refused to say 996 M incomputable 997 M doesn't know 998 M no response 999 M not applicable OTRCTRSU Other Primary Contributors to Support of Dependants 256 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 other relatives 2 siblings 3 parents 4 friends 5 husband 6 children 995 M refused to say

315 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

SOCORGMB Membership in Social Organizations 257 Measurement Level; Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 none 2 sports clubs 3 development association 4 tontine 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response BNFTSSOC Benefits of Social Organization Membership 258 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 none 2 moral support 3 financial support 4 professional support 5 material support 6 moral and financial support 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

316 GOVORGMB Membership in Government Organization 259 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 none 2 political party 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

BNFTGOV Benefits of Governmental Organization Membership 260 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 none 2 moral support 3 financial support 4 professional support 5 material support 6 moral and Hnancial support 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

COMORGMB Membership in Community Organizations 261 Measurement Level: Nominal Column Width: 12 Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 none 2 neighborhood association 3 village association 4 association of colleagues 5 family association 995 M refused to say

317 996 M incomputable 997 M doesn’t know 998 M no response

BNFTCOMT Benefits of Community Organization Membership 262 Measurement Level; Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 none 2 moral support 3 financial support 4 professional support 5 material support 6 moral and financial support 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

PRFORGMB Membership in Professional Organizations 263 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 none 2 union 3 professional association 4 groupment 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response

BNFTPROF Benefits of Professional Organization Membership 264 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other

318 1 none 2 moral support 3 financial support 4 professional support 5 material support 6 moral and Hnancial support 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

RELORGMB Membership in Religious Organizations 265 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: PS Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 none 2 prayer group 3 chorus 4 Assalatou 5 Permanence 6 prayer group and chorus 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response BNFTRELG Benefits of Religious Organization Membership 266 Measurement Level: Nominal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 0 other 1 none 2 moral support 3 financial support 4 professional support 5 material support 6 moral and financial support 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

319 CONTRORG Annual Contribution to Organizations 267 Measurement Level: Ordinal Column Width: Unknown Alignment: Center Print Format: F8 Write Format: F8 Missing Values: 995 thru 999

Value Label 1 none 2 occasionally 3 0-5000 4 5001-10000 5 10001-15000 6 15001-20000 7 20001-25000 8 25001-30000 9 30001+ 995 M refused to say 996 M incomputable 997 M doesn’t know 998 M no response 999 M not applicable

320 APPENDIX D

S-PLUS CODE FOR MATRIX CREATION

321 S-PLUS CODE FOR MATRIX CREATION

#Basic program to construct matrix of dissimilarity measures for all cases set up the #input data data < as.matrix (cbind (lngsrvy.econact3.rawdata$ACTlC, lngsrvy.econact3.rawdata$ACT2C, Ingsrvy.econactS .rawdata$ACT3C, lngsrvy.econact3.rawdata$ACT4C, deparse.level=l))

#this will remove rows for testing #data < gui.remove.row(target = data, target.row.spec = list(21:522))

#mdicates how much data to use in the program nrowsDATA < nrow(data)

#create output matrix scdl<- matrix (ncol=nrowsDATA, nrow=nrowsDATA)

#sets initial count for i where i=number of cases (rows) then starts the loop for comparing #each row of the data to every other row i<-l while (i <= nrowsDATA)

#starts i case loop {

#sets initial count for j where j=the comparison case then starts the loop for comparing #each case to every other case j<-l whiled <= i)

#starts j case loop {

# Get the comparison rows and remove NAs rowi < data[i,] rowi < rowi [rowi != "NA"] rowj < dataQ,] rowj < rowj[ro\yj 1= "NA"]

#sets matches equal to 0

322 matches <- 0

#sets the counter for each activty to one then starts the loop for comparing #each activity in the initial row until all the activities in the inital row have #been compared which then #ends the loop

counti < 1 while (counti <= length (rowi))

#starts the i activity loop { #sets the counter for each activity in the comparison case to one #then starts the loop for comparing each activity in the comparison #row until all the activities have been compared with the first ^activity in the intial row which then ends the loop

count] < 1 while (count] <= length (rowj))

#starts the] activity loop {

#starts to look for matches by comparing each activity in the initial #row with each activiQr in the comparison row, if a match is found #1 is added to the matches counter - then adds one to the activity ^counter for the comparison row and sends back to the top of that #loop

if (rowi [counti] = rowj [count]]) #matches < RetumMatcfaNoW(matches, counti, count]) #matches < RetumMatchWl(matches, counti, countj) matches < RetumMatchW2(matches, counti, count], rowi[counti], rowj[countj]) count] <- count] +1

} #end count] loop where count] is the activities in the #comparisoncase

323 #once all the activities in the comparison row have been compared to the #first activity in the initial row the counter for the initial row is reset by #adding one starting that loop again

counti <- counti + 1

} #end counti loop where counti is the activities in the inital case

#once all the activities in the comparison row have been compared to all the #activities in the initial row the number of matches is calculated and measure of #similarity is determined

scdl[i.j] < I - (matches/(length (rowi) * length (rowj)))

#one is then added to the comparison case counter so that it goes back to the top #ofthat loop

j<-j+l

} #end j loop where j is the comparison case

#one is then added to the initial case counter so that it goes back to the top of that loop i <- i+l

} #end i loop where i is the initial case

#sets the diagonal of the matrix to zero so that all comparisons between the same cases #are set as perfectly similar - then sets all the NAs within the matrix to 0 - then sets them #equal to the same case on the opposite side of the matrix - this produces a perfectly #symmetrical matrix diag(scdl) <- 0 scdl < ifelse(scdl = "NA", 0, scdl) scdl < scdl + t(scdl)

#makes the matrices data sheets so that they can be saved data < as.data.sheet(data) scdl < as.data.sheet(scdl)

324 S-PLUS CODE FOR ECONOMIC ACTIVITY WEIGHTING

RetumMatchNoW < function (matches, counti, countj) { matches < matches+l retum(matches) }

RetumMatchWl < function (matches, counti, countj) { if ((counti < 3) & (countj < 3)) matches < matches +1 else if ((counti >= 3) & (countj >= 3)) matches < matches + 0.5 else matches <- matches + 0.25 retum(matches) }

RetumMatchW2 < function (matches, counti, countj, valuei, valuej) { if (((counti < 3) & (countj < 3)) & ((valuei >= 13) & (valuei <= 16)) & ((valuej >= 13) & (valuej <= 16))) matches < matches + 1 else if ((counti < 3) & (countj < 3)) matches < matches + 0.75 else if ((counti >= 3) & (countj >= 3)) matches < matches + 0.5 else matches < matches + 0.25 retum(matches) }

325 RetumMatchW2 < function (matches, counti, countj, valuei, valuej) { if (((counti < 3) & (countj < 3)) & ((valuei >= 37) & (valuei <= 42)) & ((valuej >= 37) & (valuej <= 42))) matches < matches + 2.0 else if (((counti < 3) & (countj < 3)) & ((valuei >= 43) & (valuei <= 44)) & ((valuej >= 43) & (valuej <= 44))) matches < matches +1.5 else if (((counti < 3) & (countj < 3)) & ((valuei >= 45) & (valuei <= 47)) & ((valuej >= 45) & (valuej <= 47))) matches <- matches + 1.0 else if ((counti < 3) & (countj < 3)) matches < matches + .75 else if ((counti >= 3) & (countj >= 3)) matches < matches + 0.5 else matches < matches + 0.25 retum(matches) }

326