Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

The relationships between livestock and human wealth, health, and wellbeing in a rural Maasai community… Glass, Catherine Sian 2019

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

Item Metadata

Download

Media
24-ubc_2020_may_glass_catherine.pdf [ 2.53MB ]
Metadata
JSON: 24-1.0384551.json
JSON-LD: 24-1.0384551-ld.json
RDF/XML (Pretty): 24-1.0384551-rdf.xml
RDF/JSON: 24-1.0384551-rdf.json
Turtle: 24-1.0384551-turtle.txt
N-Triples: 24-1.0384551-rdf-ntriples.txt
Original Record: 24-1.0384551-source.json
Full Text
24-1.0384551-fulltext.txt
Citation
24-1.0384551.ris

Full Text

THE RELATIONSHIPS BETWEEN LIVESTOCK AND HUMAN WEALTH, HEALTH, AND WELLBEING IN A RURAL MAASAI COMMUNITY OF SOUTH-WESTERN KENYA by Catherine Sian Glass  B.Sc., The University of British Columbia, 1986 M.Sc., The University of Toronto, 1993  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF  DOCTOR OF PHILOSOPHY in THE FACULTY OF GRADUATE AND POSTDOCTORAL STUDIES (Population and Public Health)  THE UNIVERSITY OF BRITISH COLUMBIA (Vancouver)   October 2019 © Catherine Sian Glass, 2019     ii  The following individuals certify that they have read, and recommend to the Faculty of Graduate and Postdoctoral Studies for acceptance, the dissertation entitled: The relationships between livestock and human wealth, health, and wellbeing in a rural Maasai community of South-Western Kenya   submitted by Catherine S. Glass in partial fulfillment of the requirements for the degree of Doctor of Philosophy in Population and Public Health   Examining Committee: Dr. Trevor Dummer, School of Population and Public Health Supervisor  Dr Monika Naus, School of Population and Public Health Supervisory Committee Member  Dr Jerry Spiegel, School of Population and Public Health University Examiner Dr David Fraser, Land and Food Systems University Examiner Dr Guy Palmer, Washington State University External Examiner        Additional Supervisory Committee Member: Dr. Marina Von Keyserlingk, Land and Food Systems     iii  Abstract Livestock are critical to the livelihood of up to two billion global poor and thus represent an ideal focus for poverty amelioration. For traditional keepers, livestock are: culturally significant, nutritionally important, and serve as “daily currency” and household “savings”. However, they may also increase infectious disease risk, especially via zoonoses which can reduce both human and livestock health and quality of life. Although many studies exist on livestock-dependent communities, including the Maasai and other pastoralists, significant knowledge gaps persist regarding the relationships between traditional livestock-keeping and human wellbeing.  This dissertation investigated associations between pastoral livestock and owner health through a series of cross-sectional and longitudinal studies conducted in Olkoroi, a rural Maasai community. The objectives were to: 1) review the literature on connections between livestock health and productivity, and human wealth, health and wellbeing; 2) describe Olkoroi sociodemography and capital; 3) assess local human and livestock disease priorities and livelihood challenges; 4) conduct longitudinal studies of livestock growth, livestock and human infectious disease; 5) measure adult psychological wellbeing; and 6) use the collected data to build predictive models of human wellbeing, herd size, livestock growth, livestock and human infectious disease frequency.   I found livestock were the primary livelihood and predicted psychological wellbeing, but 40% of households, primarily female-headed, had insufficient animals to support themselves. Men and women identified similar factors affecting wellbeing but differed in proportional attribution: women uniquely spoke of restrictions on autonomy. Community disease prioritizations were similar to national priorities, however, disease management was inconsistent and causal understanding was low. Households self-rated husbandry practices highly, but felt financial constraints prevented adoption of best practice. Household variables were associated with herd size, but climate was the best predictor of livestock growth, and livestock and human infectious disease: livestock disease prevalence did not predict human disease. My results suggest livestock research must prioritize gender and local context to better understand livestock-human health relationships. Claims about the contribution of livestock to human disease burdens must also be clarified through more consistent research    iv  frameworks which allow inter-study comparisons, and more longitudinal studies to better identify causal relationships between exposures and disease incidence.       v  Lay Summary How livestock affect human wellbeing in rural poor communities is not fully understood. My research explored human-livestock relationships in Olkoroi, a Kenyan Maasai community. Livestock remains culturally important, and residents are confident about their herding practices. Herd size reliably indicates wealth and predicts herder happiness. Single parents are almost exclusively women, typically lack livestock, and consequently struggle to support their children. Some women feel restricted by tradition, and a lack of government infrastructure increases health risks and limits livelihood diversification. Community members believe that livestock does not contribute significantly to human disease: my research supports their belief, as I find that human and livestock disease are both primarily associated with climate. I recommend that keeper perspectives should be prioritized in future research and development initiatives in rural poor settings, because keepers have the most knowledge about their livelihoods and how they are affected by the ecosystems in which they live.     vi  Preface This dissertation is original, unpublished, independent work by the researcher and author, Catherine Glass. Data presented and analysed in chapters 3-7 were covered by UBC Ethics Certificate H07-02752 and the Kenya Medical Research Institute Non-Scientific Steering Committee Protocol Number 164.      vii  Table of Contents  Abstract ................................................................................................................................... iii Lay Summary ...........................................................................................................................v Preface ..................................................................................................................................... vi Table of Contents .................................................................................................................. vii List of Tables ........................................................................................................................ xiv List of Figures ...................................................................................................................... xvii Abbreviations ..................................................................................................................... xviii Glossary: Maasai/Swahili/Kenyan Vernacular ................................................................. xxi Acknowledgements ............................................................................................................. xxii Dedication ........................................................................................................................... xxiv  Introduction, Background, Research Justification, Goals and Objectives .....1 1.1 Introduction ............................................................................................................... 1 1.2 Justification of Research Focus ................................................................................ 3 1.3 Background (emphasis on Africa, Kenya and the Maasai) ....................................... 5  Modern Pastoralism .............................................................................................. 5  Health and Pastoralism ......................................................................................... 9  Wellbeing and Pastoralism ................................................................................. 11 1.4 The Potential of Livestock-based Development Initiatives to Improve Health and Wellbeing in Poor Populations (Africa/Pastoral focus) ...................................................... 12  Sub-Saharan Africa (SSA) in a Global Poverty Context .................................... 12  Poverty in Kenya................................................................................................. 13    viii   Knowledge Gaps and Poverty in Narok County and Narok South Sub-County 15  Livestock and Poverty......................................................................................... 17  Livestock in Kenya (focus on pastoralists) ......................................................... 18 1.5 Existing Reviews of Livestock Based Interventions .............................................. 21 1.6 Research Goals and Objectives ............................................................................... 24 1.7 Livestock-Based Interventions: Opportunities and Challenges .............................. 25 1.8 Thesis Overview ..................................................................................................... 29  Data Collection and Analytical Methodology ..................................................31 2.1 Setting, Population and Ethics Approval ................................................................ 31 2.2 Research Development, Field Work Timeline, and Field Assistance .................... 32 2.3 Qualitative and Mixed Methods Data Collection ................................................... 34  Personal Record Keeping and Informal Community Perspectives ..................... 34  Single Interviews with CMF Employees and District Officials ......................... 35  Wealth Marker Focus Groups ............................................................................. 37  Livestock duties/SWLS/Wellbeing interviews ................................................... 37  Livestock and Human Health Disease Prioritization, Rationalization and Causation......................................................................................................................... 38  Self-Assessed Livestock Husbandry and Best Husbandry Practice Interviews .. 39 2.4 Sociodemographic and Quantitative Data Collection ............................................. 39  Baseline Community Health Assessment and Sociodemographic Data ............. 39  Baseline Livestock Health Data .......................................................................... 41  Community Characteristics ................................................................................. 41  District Level Health Data .................................................................................. 41    ix   Narok District Weather Data .............................................................................. 42  Longitudinal Human and Livestock Health Data Collection .............................. 42  Young Livestock Growth Data ........................................................................... 43 2.5 Data Analysis .......................................................................................................... 44  Wealth and Wellbeing......................................................................................... 44  Livestock Duties and Relationship of Livestock to Wellbeing .......................... 46  Livestock Disease Prioritizations and Cultural Competency.............................. 46  Herding Duties: Self-Assessment and Best Practice .......................................... 47  Herd size Models ................................................................................................ 47  Young Livestock Growth Rate Model ................................................................ 48  Human and Livestock Infectious Disease Frequency Models ............................ 48  Researcher Positionality, Research and Community Context .......................50 3.1 Introduction ............................................................................................................. 50  Objectives ........................................................................................................... 50  “The Perspective and Position” of the Researcher ............................................. 50  The Importance of Single Community Studies ................................................... 52  Olkoroi as a Representative Maasai Community................................................ 54 3.2 Setting: Physical Geography, Climate, Ecosystem ................................................. 55 3.3 Geopolitical Structure ............................................................................................. 57 3.4 Community Structure .............................................................................................. 57  Origin, Ethnicity, Community and Family Structure.......................................... 57  Housing: Structure and Cultural Practice ........................................................... 60  Community and Family Decision Making .......................................................... 61    x  3.5 Household Vulnerability ......................................................................................... 63  Single Heads of Household ................................................................................. 63  Livestock Poverty ............................................................................................... 65  Violence .............................................................................................................. 66  Alcohol and Vulnerability................................................................................... 67 3.6 Children: Family size, child “ownership” and roles of children ............................. 68 3.7 Education ................................................................................................................ 69 3.8 Adult Education Attainment ................................................................................... 71  Current Attitudes and Barriers to Education ....................................................... 73 3.9 Community Resources ............................................................................................ 74  Education: Olkoroi Primary School and Secondary Opportunity....................... 74  Medical ............................................................................................................... 75 3.10 Religion ................................................................................................................... 77 3.11 Livelihoods ............................................................................................................. 77  Traditional Pastoral Livelihood ...................................................................... 77  Cropping and Land Ownership ....................................................................... 78  Small business ................................................................................................. 78  Casual labour .................................................................................................. 80  Salaried work .................................................................................................. 80  Professional Work ........................................................................................... 81 3.12 Conclusion: Constraints on Wealth, Wellbeing and Health ................................... 81  Material vs Subjective Wellbeing in a Traditional Maasai Community: A Gendered Perspective ............................................................................................................82    xi  4.1 Introduction ............................................................................................................. 82 4.2 Results ..................................................................................................................... 87  Wealth Marker Focus Groups ............................................................................. 87  Sociodemographic Survey .................................................................................. 88  Cluster Analysis of Wealth Markers ................................................................... 90  Psychological Wellbeing Interviews................................................................... 91  Contributors and Detractors to Wellbeing: Men vs Women .............................. 92  Exploratory Wellbeing Model ............................................................................ 95 4.3 Discussion ............................................................................................................... 96  Economic Differentiation of Olkoroi Families ................................................... 96  Wellbeing .......................................................................................................... 100 4.3.2.1 Male vs Female SWLS Scores .................................................................. 100 4.3.2.2 Current Contributors and Detractors to/from Wellbeing .......................... 105 4.3.2.3 Future Contributors and Detractors to/from Wellbeing: Men vs Women 106  Explanatory Wellbeing Model .......................................................................... 107 4.4 Conclusion ............................................................................................................ 110  Livestock Disease, Rearing Practices and Contribution to Wellbeing: Community Perceptions and Exploratory Models ...........................................................113 5.1 Introduction ........................................................................................................... 113 5.2 Results ................................................................................................................... 117  Livestock Disease Prioritization and Understanding ........................................ 119  Household Husbandry Self-Assessment and Best Husbandry Practices .......... 122  Self-reported Vaccination Prevalence and Timing ........................................... 124    xii  5.2.6.1 Small Ruminant Disease Prevalence ........................................................ 126 5.2.6.2 Cattle Disease Prevalence ......................................................................... 127 5.2.6.3 Cattle Trypanosomiasis Prevalence .......................................................... 128  Perceptions about the Contribution of Livestock to Wellbeing, and the Relationship between Gender and Livestock Responsibilities ..................................... 129  Self-assessed Livestock Husbandry, Best Practices, and Owner Perceived Barriers to Maximising Herd Productivity ................................................................... 130  Livestock Disease Prioritization and Understanding ........................................ 132 5.3.3.1 Local Perceptions about Zoonoses ............................................................ 136 5.3.3.2 Vaccination Practice ................................................................................. 137  Herd Size Predictive Model .............................................................................. 139  Young Livestock Growth Rates Model ............................................................ 140  Disease Prevalence Models............................................................................... 141 5.4 Conclusion ............................................................................................................ 146  Human Health in Olkoroi ................................................................................150 6.1 Introduction ........................................................................................................... 150 6.2 Results ................................................................................................................... 154  Baseline Health Indicators ................................................................................ 155  Health Promotion Behaviour ............................................................................ 156  Human Disease Prioritization, Understanding and Rationalizations ................ 159  Community Reported Mortalities ..................................................................... 161  Exploratory Models of Self-Reported Human Morbidity ................................. 162 6.3 Discussion ............................................................................................................. 165    xiii   Baseline Health ................................................................................................. 165  Health Risks/Prevention Activities ................................................................... 168  Disease Prioritization, Rationales and Understanding of Causation and Treatment ...................................................................................................................... 170  Human Disease Models .................................................................................... 174 6.3.4.1 Malaria ...................................................................................................... 174 6.3.4.2 Respiratory Infections ............................................................................... 176 6.3.4.3 GI Illness ................................................................................................... 177 6.3.4.4 Total Infectious Disease Incidence ........................................................... 178 6.4 Conclusions and Recommendations ..................................................................... 178  Summary, Relevance, Recommendations and Conclusion ...........................184 7.1 Summary of Findings ............................................................................................ 184 7.2 Strengths and Unique Contributions ..................................................................... 188 7.3 Limitations ............................................................................................................ 192 7.4 Future Research .................................................................................................... 195 7.5 Conclusion ............................................................................................................ 199 References .............................................................................................................................202 Appendices ............................................................................................................................251  Sociodemographic and Health Questionnaire ........................................... 251  Livestock duties Script, SWLS, and Wellbeing........................................ 259  Livestock Disease Prioritization and Understanding ................................ 261  Human Disease Prioritization and Understanding .................................... 267    xiv  List of Tables Table 3-1: Security of Olkoroi Single Heads of Household (2008-2010) ......................... 64 Table 3-2: Reasons for Single Head of Household Status in Olkoroi, 2008-2010 ........... 64 Table 3-3: Olkoroi Adult Educational Attainment by Sex and Level of Education (2008)................................................................................................................................................. 73 Table 3-4: Classroom sex ratios, Olkoroi Primary School, 2008 ...................................... 73 Table 3-5: Number of Boys and Girls Reaching Grade 8 at Olkoroi Primary School, 2006-2009 ............................................................................................................................... 74 Table 3-6: Kenyan Certificate of Primary Education Results for Olkoroi Primary School, 2006-2009 .................................................................................................................. 74 Table 4-1: Community Focus Groups’ Delineation of Wealth Categories by Marker .. 87 Table 4-2: Baseline Community Characteristics of Study Population: Adult Members of Olkoroi Village, Narok District South, Kenya, 2008 (n=150) ........................................... 88 Table 4-3: Household assets, and education status of community children in Olkoroi (2008) ...................................................................................................................................... 89 Table 4-4: Livestock Owned Relative to Poverty Threshold, Sex of Head of Household and Family Structure in Olkoroi ......................................................................................... 90 Table 4-5: Relative Proportions of Men vs Women by Specific Wellbeing Category in Olkoroi (2009)........................................................................................................................ 91 Table 4-6: Unadjusted and Adjusted Linear Regression Coefficients and 95% Confidence Intervals of Variables Associated with Individual Life Satisfaction (n=150)................................................................................................................................................. 96 Table 5-1: Performance of Olkoroi Livestock Duties by Sex, 2009 ................................ 118    xv  Table 5-2: Positive Contributions of Livestock to Wellbeing, 2009 ............................... 118 Table 5-3: Do Livestock Detract from Wellbeing? If yes, how? (2009) ......................... 119 Table 5-4: Averaged Rank of Self-Selected Livestock Diseases of Local Importance (2009) and Rank of Total Self-Reported Livestock Disease Prevalence, Olkoroi (2008-2010) ..................................................................................................................................... 120 Table 5-5: Heads of Households Self-Assessment of Household Husbandry Practice, Olkoroi 2009 ........................................................................................................................ 122 Table 5-6: Summary of Interviews on Best Husbandry Practices (Olkoroi 2009) ....... 123 Table 5-7: Owner Reported Vaccination Prevalence and Timing: Olkoroi, 2009 ........ 124 Table 5-8: Unadjusted and Adjusted Parameter Estimates and Confidence Intervals (95% CI) of Head of Houshold Variables Associated with Preliminary Herd Size (TLU): Olkoroi, 2008 .......................................................................................................... 125 Table 5-9: Unadjusted and Adjusted Parameter Estimates and 95% Confidence Intervals (95% CI) of Variables Associated with Young Livestock Growth Rates: Olkoroi, May 2009-November 2010 .................................................................................. 126 Table 5-10: Unadjusted and Adjusted Odds Ratios (O.R.) and 95% Confidence Intervals (95% CI) of Variables Associated with “Olodua” Prevalence in Small ruminants ............................................................................................................................. 127 Table 5-11: Unadjusted and Adjusted Odds Ratios (O.R.) and 95% Confidence Intervals (95% CI) of Variables Associated with Total Cattle Disease Prevalence ...... 128 Table 5-12: Unadjusted and Adjusted Odds Ratios (O.R.) and 95% Confidence Intervals (95% CI) of Variables Associated with Trypanosomiasis Prevalence in Cattle............................................................................................................................................... 129    xvi  Table 6-1: Adult Baseline Health Indicators: Olkoroi 2008 ........................................... 155 Table 6-2: Frequency of Health Risk/Prevention Activities ........................................... 158 Table 6-3: Community Prioritization of Human Disease, Frequency of Self-Reported Illness and Frequency of Cause for Clinic Visits in Narok District ............................... 160 Table 6-4: Community Reported Causes of Mortality, 2006-2016................................. 162 Table 6-5: Unadjusted and Adjusted Parameter Estimates and 95% CI of Climate, SES, and Livestock-Related Variables Associated with Total Self-Reported Morbidity, Malaria, Respiratory Infections, and GI Illness in Children .......................................... 164 Table 6-6:Adjusted Parameter Estimates and 95% Confidence Intervals (95% CI) of Climate, SES, and Livestock-Related Variables Associated with Total Self-Reported Morbidity, Malaria, Respiratory Infections, and GI Illness Frequency in Adults ....... 165   xvii  List of Figures Figure 2-1: Research timeline .............................................................................................. 34 Figure 3-1 Olkoroi from the Western Hills ........................................................................ 56 Figure 3-2: Reasons of Lack of/Interruption of Education in Olkoroi Adults (2008) .... 72 Figure 4-1: Most Important Contributor to Current Wellbeing in Olkoroi (2009): Women vs Men (n=150, p=0.025) ........................................................................................ 92 Figure 4-2: Most Important Detractors from Current Wellbeing in Olkoroi (2009): Women vs Men (n=150, p=0.0036) ...................................................................................... 93 Figure 4-3: Most Important Future Contributors to Wellbeing in Olkoroi (2009): Women vs Men (n=150, p=0.18) .......................................................................................... 94 Figure 4-4: Most Important Future Detractor from Wellbeing in Olkoroi (2009): Women vs Men (n=150, p=0.005) ........................................................................................ 95     xviii  Abbreviations ACCC – Association of Canadian Community Colleges (now CICan - Colleges and Institutes Canada) AMREF – African Medical and Research Foundation ASAL – arid and semi-arid lands ASF – animal source foods AU – African Union BP – blood pressure Bpm – beats per minute CE – cystic echinococcosis CMF – Christian Missionary Fellowship DEO – District Education Office DFID – [U.K.] Department for International Development DHS – Demographic Health Survey DVO – District Veterinary Office FHH – female headed households GCC – global climate change GBD – Global Burden of Disease GDP – gross domestic product GI – gastrointestinal Hb – hemoglobin HCW – healthcare workers HIV – human immunodeficiency virus    xix  HoH – head of household IGAD – International Authority on Development  ILRI – International Livestock Research Institute IMF – International Monetary Fund KSh – Kenyan shillings LID – Livestock in Poverty-Focused Development LMIC – low and middle income countries MDG – Millennium Development Goals MeSH – Medical Subject Headings MHH – male headed household mmHG – millimetres of mercury MPI – Multidimensional Poverty Index MUAC – Mean upper arm circumference NGO – non-governmental organisation OD – open defecation OIE – Office International des Epizooties (OIE aka World Organisation for Animal Health) OPHI – Oxford Poverty and Human Development Initiative PRSP – Poverty Reduction Strategy Paper PVS – Performance of Veterinary Service SAP – Structural Adjustment Program SSA – Sub-Saharan Africa STIs – sexually transmitted infections SWLS – Satisfaction with Life Scale    xx  TBD – tick-borne disease TFR – Total fertility rate TLU – tropical livestock unit UNDP – United Nations Development Programme UTIs – urinary tract infections WASH – water, sanitation and hygiene WB – World Bank WHO – World Health Organisation WVS – World Values Survey        xxi  Glossary: Maasai/Swahili/Kenyan Vernacular  Agrivet – non-veterinary vendors of agricultural pharmaceuticals with widely varying training (in rural areas, usually informal if any) Ashe oleng - thank you very much (Maa) Enkashumpai – Caucasian female (Maa) Kati-kati – in the middle, between (Swahili) Maa – the Maasai language Mayiolo – I don’t know (Maa) Moran – warrior (Maa) Ole – approximately equivalent to “Mr.” (Maa) Shamba – farm or garden plot (Swahili) Sidai- good/excellent/beautiful (Maa)   xxii  Acknowledgements It has been a very long road to completion, and the list of people and institutions to whom I owe thanks is lengthy. In the space allotted, I can only tip my metaphorical hat to both those who have provided key support at specific times, and those who never stopped keeping the faith that I would one day finish (or if they did, did a great job of hiding it)! Starting with my workplace, enormous thanks go to Langara College and my many colleagues who encouraged me, as well as department chairs, administrators and support staff, who allowed me to bend my schedule, tolerated my odd hours and section numbers, and always did their best to accommodate my needs. Special thanks to Ken Naumann, a faithful work friend through good times and bad, and Don MacDonald who started it all, when he introduced me to Kenya with his invitation to be part of the Canadian Field School in Africa. Thanks also to the Langara Faculty Association for funding made available through the Langara Research Committee. I cannot omit acknowledgment of the cheerful help of the Langara security staff, who opened the doors for me endlessly on Sundays, and checked on me regularly late at night. At ILRI, thanks are owed to Thomas Randolph, Mohammed Said, and Esther Schelling for early advice and assistance, as well as a place to call on in Kenya, connections to equipment, and perspectives on pastoralism and livestock disease in Narok. I owe extra thanks to Esther for meeting with me and providing accommodation in Switzerland. At UBC, a big thank you to Monika Naus who got me started on my doctorate, has been endlessly patient and kind in providing support and feedback, not to mention connecting me to my statistical angel, Robert Balshaw. Thanks also to Craig Stephen for early guidance on preliminary chapter drafts. I am very appreciative for the late entry of Marina von Keyserlingk and her reading and commentary on the final drafts of the thesis. Huge thanks to Trevor Dummer for taking on a supervisory role, and always making time for me, from reading and correcting drafts, to answering questions, filling in forms, and guiding me step by step to the end. I am very grateful for your steady, calm influence. Robert Balshaw, my aforementioned statistical guiding light, is, more than anyone, the person who enabled me to complete my analysis and write-up. He taught, encouraged, and helped me to make sense of a veritable data mountain. But as important as the statistical   xxiii  guidance provided, was his unfailingly goodwill and care. I could not have done this without him. That he continued to provide critical support even after leaving BDCDC and moving to Manitoba, is just one of many examples of the way he consistently went above and beyond. I can’t thank Patricia Jansen enough for getting me back on track and making it possible for me to finish. Words cannot express how transformative it was to meet Susan Cox, who has truly been balm to my academic soul. Emily Van Gulik deserves a medal for her patience and skill in helping me navigate through the endless bureaucracy an extra-length PhD required, and the sympathetic shoulder she provided for me, more times than I can count. Moving out of the university setting, as I continue expressing my appreciation, I must start with my parents, Marjorie and Anthony Glass who supported various aspects of this journey in a myriad of ways (including keeping the kids and feeding me in comps week). Not only did they keep me standing in some very dark hours as I began my studies, it was also their values that set me on a path to the School of Population and Public Health. Much gratitude for the love and care received from my husband, Adrien Raynal, and the millions of ways he demonstrated his faith in me. I promise I will do some cooking now. Thanks also to my wonderful in-laws and all the “courage” (with a French accent) they wished me. My four children have been a critical part of my support system, and have helped in countless ways, from keeping me company in the field, doing expert transcription and formatting with their superior keyboard skills, crunching numbers, not holding the bedbugs and GI illness against me, and always having words of encouragement and sympathy when I was down. All of you, represent the best of my life work thus far.  Thanks to my dear friends, Sean Graham and Catriona Gordon. I don’t have the words to describe how much your steadfast faith, encouragement and affection have helped. Lastly, but by no means least, thank you to all the residents of Olkoroi for your incredible generosity in sharing your lives, experiences and perspectives with me. The lessons and insights you gave me have transformed my worldview. Thanks most of all to Ole Alfred Leroka Koshal, a field assistant non pareil. Ashe, ashe, ashe, oleng.    xxiv  Dedication This dissertation is dedicated to my children, Sarah Glass, Brynmor Crookall, Cadan Glass and Daniel Crookall, who changed my life irrevocably for the better.  Also, the people of Olkoroi for their welcome, kindness, patience and extensive collaboration in the research reported within, most especially to Ole Alfred Leroka Koshal, and to Noosokon Sironik, a young woman cruelly taken from life too soon.              1   Introduction, Background, Research Justification, Goals and Objectives 1.1 Introduction  Numerous livestock-based interventions intended to improve national and to a lesser degree, household wealth, have been implemented in the developing world since the turn of the 19th century. Initiatives have been driven by varied motives: supplying colonial and post-colonial metropoles;1 stimulation of economic growth and poverty alleviation via agriculture (a significant contributor to the economy of many developing countries);2, 3 and ameliorating imbalances in domestic supply versus demand for livestock products in many poor nations.4, 5 Livestock are an ideal focus for poverty reduction efforts because of their dual role in national and family economies, relevance in the livelihoods of a large proportion of the global poor,6 and potential for health promotion, both through the nutritional value of animal-source foods (ASF)7 and the association between economic security and health.8-10 Livestock-based interventions, past and present, have primarily focused on increasing revenues rather than health improvements of animal keepers,a with the assumption (if health was even considered) that greater wealth would directly translate into uniform improvements in health and socioeconomic status (SES).9 This can be the case, but increasing revenue from traditional livestock systems such as pastoralism, upon which the poor typically depend, has rarely been a simple endeavor: “Despite…bright spots demonstrating the possibility of alternative pathways, overall, mainstream pastoral development is a litany of failure, involving substantial sums of wasted resources”.12 To expect that increased revenue alone will lead to improved SES, much less, a commensurate increase in wellbeing and health for each member of all participating households is not a validated assumption.9  To ensure that benefits from livestock development are at least potentially available to everyone in a poor community requires extensive initial consultation with representatives from all sectors of the community. Marginalized members may need support to access the                                                  a There are some exceptions, for example, targeted efforts to reduce the prevalence of parasites such as T. solium, however, these efforts are also often hampered by infrastructural weakness in the poorest nations.11. Gabriël S, Dorny P, Mwape K, Trevisan C, Braae UC, Magnussen P, et al. Control of Taenia solium taeniasis/cysticercosis: the best way forward for sub-Saharan Africa? Acta tropica. 2017;165:252-60.    2  discussion, and external actors must have a thorough understanding of local capital from the individual, to the household, community and ecosystem.13-15 Historically, such necessary preliminary research has rarely happened and the general consensus of the research and development community is that most livestock-mediated poverty interventions have been unsuccessful.14, 16-23 This has primarily resulted from inadequate consultation with those affected. But it is also the consequence of a lack of genuine focus on the poor, in turn leading to a failure to understand the varied drivers, perpetuators, and outcomes of poverty.14, 16, 22, 24 There has also been a consistent inability to learn from failures.25 Moreover there is a scarcity of research on livestock health and welfare, and the wealth and health of poor livestock keepers simultaneously.26-28 Literature exists on each subject separately, as well as theoretical discussion of how they relate. But, without studies which collect information on both concurrently, it is difficult to fully understand the complexities,26 or test hypotheses about animal-human health and wealth relationships. Perry and Grace summed up this lack of clarity, noting: “…livestock probably matter to poverty reduction, but we are not sure exactly how or how much”.29 A further complication is that livelihood, livestock and human demographic data, and health metrics from the developing world are often insufficient in quantity and quality.30 Evidence is not always current,31 is primarily derived from cross-sectional studies,32, 33 and is frequently non-generalizable because of difficulties in collecting representative data from rural areas (where most poor livestock owners live8) and marginalized populations,8, 15, 26, 34-42 especially in insecure regions.43 Lastly, research in rural communities, for either preliminary investigations or intervention evaluation, is rarely conducted long enough to draw fully informed conclusions on the relevant descriptive epidemiology, possible outcomes and/or effectiveness of interventions.44-46  In discussion of the potential, necessary priorities, and roadblocks in pro-poor development, insufficient agricultural research funding in many countries has been highlighted 29 particularly with regards to livestock and initiatives relevant to smallholders. National funding deficits have been exacerbated by a global decline in funding for both basic research47 and development initiatives focused on agricultural animals.48-50 Perry and Sones further suggested much of the existing research was unlikely to help the poorest citizens of the poorest nations because: “…sectors of the affluent world are still basing their science    3  contributions to poverty reduction on self-interest…At the moment, only the crumbs go to the poor”.51 Adding to these challenges are barriers such as lack of resources and infrastructure,52 corruption,53, 54 social feasibility and political will at all levels including the individual29 in target nations.19 At the same time, no one prescription is likely to fit all circumstances, even within single communities.19, 45, 55, 56 The associations between livestock and the wealth, health and psychological wellbeing of their keepers are diverse and complex57 although common underlying relationships frequently exist. There is substantial evidence to support the conclusion that livestock can positively affect the SES and health of their owners. With sufficient resources and sustained market demand, 6, 19, 52, 58, 59 livestock provide income for maintenance and improvement of household social, material and human capital. ASF are an accessible supply of high value nutrients, and in many traditional societies, livestock play important cultural roles which contribute to psychological wellbeing. On the negative side, livestock may cause injury to their keepers and increase risk of zoonotic infection. Livestock-human relationships, and the potential of livestock to ameliorate poverty and improve health, are the focus of a substantial quantity of academic and development literature. Yet, for a variety of reasons, there is still insufficient evidence to support many of the widely mooted claims about these relationships,27, 60 and guide more effective development efforts. This is especially the case for pastoralism, an ancient practice still followed by millions globally, and now recognised as one of the most environmentally sustainable, traditional, livestock-dependent livelihoods.61   1.2 Justification of Research Focus Globally, the highest proportion of poor livestock keepers live in Asia and Africa.14 In Africa,  pastoralists make up a major fraction of livestock-keepers, and are important contributors to agricultural productivity.62 Nonetheless, pastoralists have not been supported or incorporated into national decision making in a manner commensurate with their contributions to the livestock economy.17, 63 In 2013, the African Union (AU) recognised and affirmed the importance of pastoralism to African economies but simultaneously acknowledged “…that pastoralists are among the most politically and economically marginalized communities”.64 There also remains a paucity of accurate sociodemographic    4  information on pastoral households,61, 65 and stereotypes about pastoral motivations and behaviour persist.66-68 Little information is available on evolving structures of pastoral households69 and possible increases in the number of female-headed households (FHH), although some information has been reported on shifting cultural values, and changing roles for pastoral women.70, 71 FHH may already make up a high proportion of community households due to a combination of traditional marriage practices and social changes. While it has been suggested that changing pastoral practice is increasing autonomy for some women, FHH may be among the most vulnerable of a marginalized population.71 Historic evidence suggests livestock research and development agendas in general, but also specifically as regards pastoralists, have not been effective because of frequent failure to identify or focus on:  production systems used by pastoralists72, 73  the self-identified needs/priorities of target populations16, 17  the most vulnerable within poor populations17, 66  women’s role in livestock keeping74, and the livestock (typically small ruminants such as sheep and goats) which are most accessible to women and the poor75-77  In addition, measurement tools frequently fail to take into consideration intra-household differences in autonomy, information access, and control over family assets, especially livestock.71, 74 Despite much discussion of the importance of livestock in the wellbeing of the poor, most human-livestock health relationships discussed in the literature are theoretical60 because few studies have collected data on livestock and human health simultaneously.  In the last three decades, there has been a strong publication emphasis on the role of zoonoses as a barrier to movement of livestock keepers, especially pastoralists, out of poverty,9, 14, 78-82 although insufficient evidence has been collected to generalize this assertion. While some zoonotic illnesses can have substantial impact on household wellbeing, distribution and relative importance vary widely.83 Some of the most recent studies and reviews of zoonoses appear to contradict the widespread and repeated claims about their importance and prevalence in humans and livestock, especially as regards brucellosis, one of the most studied of the zoonoses in the Global South.83-91     5  Livestock-focused development initiatives make enormous intuitive sense for the multiplicative benefits, including the health of livestock keepers, their success could bring. Not only could livestock-keeping populations benefit, but also consumers and related industries, and ultimately national economies through the chain of livestock production, and export markets. Changing the history of failure in livestock development initiatives, however, requires a direct focus on poor livestock keepers, accurate knowledge of their circumstances and needs, and most importantly a meaningful participatory process that allows the populations of interest to articulate their needs, priorities, and goals for themselves.   1.3 Background (emphasis on Africa, Kenya and the Maasai)  Modern Pastoralism  Pastoralism is practiced around the world from Asia, Europe, and the circumpolar North, to North, Central and South America, India and Africa73, but approximately half of all pastoralists are found in Africa.92 Although pastoralism is conceptually well known, and some practitioners, like the Maasai are “famous” (in part for their warrior tradition, but in greater part because of their “cattle complex”93 a term coined in the 1920s by Melville Herskovits, which encompassed not just a livestock dependent livelihood, but a tradition in which livestock were central to all significant aspects of life), there is a deficiency of quality data on pastoral numbers and livestock holdings.94 In Kenya, pastoralists in Northern Kenya were excluded from the National Demographic and Health Surveys (DHS) until 2003, and similar exclusions from national data collections have been common in other countries.95  Pastoral practice is not uniform, but exhibits common characteristics: a livelihood primarily dependent on livestock; occupation of arid and semi-arid lands (ASAL) unsuited to plant-based agriculture because of erratic water supply;96 cyclical movements driven by water and/or pasture availability and sometimes disease distribution; communally owned land; a legal system based on customary law;97 and a livestock-centred culture that imbues much of daily and ceremonial life.98 Low productivity of ASAL has been traditionally offset by communal ownership of extensive territory, which supports substantial herds. The combination of large land holdings, mobility, and relatively low inputs of labour, materials and money, make pastoralists, contrary to much historical representation, extremely efficient    6  producers and stewards of fragile ecosystems which may not support other livelihoods.29, 61, 99 Politically, the meaning of pastoralism has varied dependent on the defining body. In Kenya, for example, official pastoral designation is lost with a move to semi or fully urban environments.95 However, there is a long, documented history of pastoralists passing in and out of “classic” pastoralism, including extended time spent in urban centres, without relinquishing their identity.98 Pastoralism requires adaptability to patchy and unpredictable events, resources, and climates.17, 68 This necessary flexibility is another defining feature of pastoral life.94, 96 Pastoralists are also frequently marginalized.68 They may literally live on the “edges” of their countries, inhabiting or relegated to unwanted or inhospitable lands,17 but metaphorically they frequently inhabit national peripheries, disenfranchised, misrepresented, underrepresented and underserved by governments.100, 101 Although marginalization is usually viewed as a detriment, at least one pastoral researcher has suggested marginalization has benefits, at least to those with the most power within pastoral communities.102 In Africa and academia, pundits have argued about the viability of traditional pastoralism since first colonization.68 The modern consensus is that pastoralism is resilient, but pastoralists globally face similar, ongoing challenges. Governments have long endeavored to settle them, not always for pastoralist benefit.73, 103-106 In a 2008 debate about future pastoral viability, four options were identified: maintaining exclusively traditional practice, supporting traditional practice via diversification, adoption of high intensity production (rarely feasible due to cost and lack of resources in ASAL), or settling and leaving pastoralism for other occupations.13 The last option, without access to alternate livelihoods which may require capital, training and higher education,107, 108 has a high likelihood of causing (further) impoverishment and diminished health.109, 110 Settling can however, create new opportunities and allow women in particular to explore avenues that might otherwise be unavailable in a traditional livelihood system.102, 111, 112  Huge customary land areas in Africa have been lost to colonial and post-colonial governments. “Land grabs” continue, via creation of hunting concessions, “development” projects, conversion of pasture into cropland, and acquisition of “waste or underutilized territory” by foreign owners.66, 94, 113, 114 Government mandated subdivision and privatization of communal land, in part driven by unsubstantiated theories about agricultural    7  productivity,25 have occurred for decades in East Africa and other pastoral regions. Privatization results in both increased and decreased security,25 and increases settlement pressure (in part due to reduced mobility).115 Security theoretically increases because legally documented land title can be used as collateral in commercial transactions and may prevent land manipulation/theft by outsiders. Some Kenyan work, however, suggested that holding land title neither increased productivity, nor facilitated access to credit, particularly for the poor and women.25 Decreased security further occurs when individual ownership, in combination with sales, increased settling, and population expansion lead to territorial fragmentation. Fragmentation precludes rights of movement required for extensive pastoralism70, 116, 117 and may also increase susceptibility to drought and global climate change (GCC).115 Kenyan pastoralists with newly held private land have also been repeatedly manipulated by those with more commercial experience and/or corrupt intent.118-120 Misunderstanding of the requirements of individual landownership has also led to land loss when owners have not collected their deeds, or when head of households (HoH) have died and the property is not, or cannot be (often for women121) re-registered.25, 122 Customary law still holds strong in many pastoral communities but tends to be eroded by national law. In rural Maasailand, increased contact with the outside world via higher rates of formal education and technological advances, has led to more awareness of legal rights,123 especially for women and children. As with settlement, national laws may provide increased protection and avenues by which the less powerful in Maasai hierarchy can seek “justice”. However, they also weaken cohesiveness and the strong fabric of Maasai custom that have historically preserved tribal identity against outside forces.123 Politicians, development agencies, scientists, and pastoralists themselves have asserted the need for diversification and there has been substantial documentation of this process.108 Diversification is frequently posed as a survival necessity caused by reduced capacity to follow historic livelihood due to: changing agricultural practices;124 population growth;125 civil and national conflicts;66, 125 corruption;108, 126-129 loss or enclosure of land;120, 125, 130 impoverishment due to livestock loss, drought and livestock disease;131, 132 and/or breakdown of the cultural safety net.108, 112, 133-135 The safety net still constitutes approximately 10% of pastoral holdings,136 most commonly in the form of animal gifts or    8  loans to those who have suffered livestock losses. It is, however, believed by both academics and pastoralists to be significantly diminished134 due to increased poverty,107, 115 loss of cultural tradition,137, 138 and adoption of modern economic practices.139 Many diversification interventions and strategies have been promoted and described,68 but historic accounts indicate the Maasai and other pastoralists have always diversified by their own initiative as necessary. Non-pastoral revenue generation activities include: remittances; salaried work in game reserves, tourist camps, service industries, and civil service positions; small business ranging from petty trade to more formal activity such as construction and transportation; investment in new breeds and livestock technology such as artificial insemination; crop-based agriculture; mining; and even land and resource speculation.108 Rarely mentioned is the participation of both Maasai women (from colonial times)137, 140, 141 and men (recently catering to female tourists in locations such as Mombasa and Zanzibar142) in the sex trade. Another revenue generator infrequently referenced is participation in Western funded missionary work which may be one of the few educational avenues and/or opportunities for salaried employment in remote African communities. Although diversification trends are often explained by the “decline” or “unsustainable” nature of pastoralism, diversification is often pursued to support traditional practice.124 Multiple authors have documented pastoralists across East Africa directing diversification revenue back to a core family unit to support herd recovery, maintenance and/or expansion.94, 112, 143 Some, forced out of pastoralism by livestock loss, diversify for survival, but hope to return to pastoralism.13, 107, 108, 144-146 Settlement, also, does not preclude continuation of mobile pastoralism within an extended family, and may facilitate diversification. Research in Marsabit, Kenya, found settled Kargi households had more mobile livestock than nomadic families in North Horr. Livestock rather than family mobility was hypothesized to explained higher Kargi livestock productivity.107 Despite government and development sector enthusiasm for greater diversification as a solution to livelihood difficulties, Kenyan evidence suggests both diversification from13, 147 and maintenance of primarily pastoral activity can predict better health, financial stability and wealth,108 even within the same community.144 Conversely, diversification may be associated with economic decline, especially when driven by significant or complete    9  livestock loss.13, 66, 107, 148 While diversification appears to be a pastoral tradition, there is some debate about current diversification patterns. Some research suggests that the poorest and wealthiest are most likely to diversify,108, 112 the former because of need and the latter to protect assets and increase resilience. Others have suggested that all strata are diversifying due to cultural, economic, and civil influences.124, 125, 130, 143 There is further question as to whether current trends are similar to past coping mechanisms or represent a permanent livelihood shift with potential long-term consequences for pastoral identity.124, 143   Health and Pastoralism  Hypotheses about the effects of pastoralism on human health have primarily been tested via studies on health relative to size of livestock holdings and comparisons of zoonoses prevalence in livestock and their keepers. Studies include comparisons between different types of pastoralists, between pastoral and settled tribes, and between settled (with or without livestock) and mobile pastoralists of the same tribe.144, 149-151 A variety of study designs and data sources have been used although most research is cross-sectional. A longitudinal study on livestock-human health associations is ongoing in the non-pastoral Luo in Western Kenya.27 Overall, conclusions are inconsistent and may not be generalizable. Studies which use DHS, for example, are limited by the breadth and detail of data collected, and a focus on women and children.152 In addition, pastoralist health status may be unrelated to livelihood, data collection can be difficult and/or biased due to social factors and sampling challenges,152, 153 and the body of literature is variable in focus, study design, and is relatively small.26 Infrastructure limitations in many developing countries result in unreliable morbidity and mortality data, especially for rural populations,154 and data collection often omits mobile pastoralists65, 139 and insecure regions.155 For example, an Ethiopian review of maternal mortality concluded available data were inadequate to make valid conclusions about causation or calculate accurate national mortality rates, and there was no data at all on pastoralists.156 Researchers documenting morbidity and nutrition in Chadian pastoralists were unable to follow-up on individuals from season to season due to mobility effects.152 Research bias also interferes with data validity because few researchers are willing to live in remote locations and/or travel for long with mobile communities.    10  Based on a limited pool of data, it appears there are both health advantages and disadvantages to pastoralism. However, poor rural communities, whether settled or mobile, crop or livestock-dependent, tend to suffer similar problems: high rates of infectious diseases, malnutrition (particularly in droughts), and high infant, child and maternal mortality.149 In addition, specific disease prevalence in any particular community is strongly influenced by the local environs, such as in a study from Mali, where nomads settled by water suffered high rates of malaria, bilharzia and a variety of parasite infections.157 A review of Sahelian health research suggested pastoralists experienced higher infant mortality than settled communities.158 Other studies have suggested that pastoralists, particularly men, suffered higher rates of zoonotic infections such as brucellosis and tuberculosis (TB),152, 157, 159, 160, but correlations between zoonoses risk and pastoralism are variable.26, 161, 162 Evidence also indicates that pastoralists suffer increased risk of respiratory, vaccine-preventable (typically when coverage is limited by poor infrastructure),26, 152 and sexually transmitted infections (STI).153, 157, 163, 164 Zinsstag claimed that pastoralists have generally poor health and almost always suffered from higher rates of the zoonoses brucellosis, echinococcosis, and rabies, as well as, depending on locale, trypanosomiasis, leishmaniasis, and plague.100 However, Kenyan studies149 on the impact of settlement concluded children of mobile pastoralists suffered less malnutrition due to higher livestock productivity and greater milk intake, than children in settled families, even those which owned livestock. This association held true during droughts, when researchers expected active pastoralists to face more difficulty. Similar correlations were found in Mali.152 Nomadic women in Chad experienced higher rates of malnutrition during the dry season,165 but as with other research, there was little difference between nomadic and settled children.166 Settled pastoralists in Uganda experienced increased rates of epidemic diseases (hepatitis E, yellow fever, cholera, and meningitis) compared to mobile groups of the same ethnicity.167 In Mali, lower rates of parasitic infections in mobile pastoralists were observed, but infant mortality was higher. Chabasse and Schelling working in Mali and Chad, respectively, found morbidities were similar in settled and pastoral communities.152, 168 A number of studies have indicated that mobility interferes with pastoral healthcare access and subsequent health status, but economic and social factors are important too.158 An    11  investigation in Chad found social networks had both positive and negative health access effects for women, but religion created barriers for Muslim pastoral women specifically, because settled medical practitioners were predominantly male.110 Pastoral women, regardless of faith, often need permission and support from their husbands before they can obtain healthcare.169 Therefore gender may exacerbate health challenges associated with pastoralism.26 Incomplete health data, combined with limited communication during travel, may impede healthcare service and delivery even when available.26  The needs of pastoralists, still neglected in many countries, may be low on national priority lists. Mobile pastoralists, especially in remote areas, have less access to infrastructure be it education, health, communication or legal. Delivery systems suited for settled communities are often inappropriate and/or expensive to provide for livestock-keepers on the move. Neglect, land loss, and historical conflict can also contribute to pastoral avoidance of government initiatives.100, 170 Some studies have noted that pastoralism was correlated with delays in diagnosis and treatment for infections such as TB, hepatitis, and a variety of STI’s152, 171, 172 until illnesses reached advanced states,158 in part because of clinic avoidance.   Wellbeing and Pastoralism Wellbeing is variably defined and measured using a wide array of methods. Edward Diener, the “father” of modern wellbeing research (and developer of the widely used Satisfaction with Life Scale, SWLS), terms it simply happiness, or more formally, self-evaluated life satisfaction.173 Dodge et al. explained it as “a state of equilibrium or balance that can be affected by life events or challenges”.174 Some research suggests people tend to have a “set” wellbeing to which they return even after major life trauma,175 but other authors have challenged this “hedonic adaptation”.176, 177 Historically, little wellbeing research was done in the developing world, and even less in the poorest communities of low income countries.178 Studies in poorer nations have increased in the past two decades, but when the term is used in low and middle-income countries (LMIC) research, it is rarely used in the academic sense. In Staying Maasai,108 an important text on Maasai livelihoods in Kenya and Tanzania, it was often used to describe the economic status of entire households. This despite the fact that household measures may mask or omit variation in individual perspectives,3, 143    12  and general acceptance3 that “…poverty cannot be conceptualized or measured in isolation from some concept of wellbeing.”179 Multi-dimensional poverty indexes that include wellbeing are now widely used to reflect the varying repercussions of poverty, for example by the United Nations Development Programme (UNDP) in annual Human Development reports.180, 181  It has been suggested the lack of wellbeing research in LMIC may be due to a perception that wellbeing is a lower, or even inappropriate priority compared with urgent survival pressures such as hunger and extreme income poverty.182 It may also be incorrectly assumed that increased income automatically correlates to enhanced life satisfaction.183 To the contrary, the small body of research that has been conducted in LMIC generally, and in poor communities globally, indicates that just as for the better off, wellbeing is multi-dimensional. Poverty affects self-assessed wellbeing, especially under conditions of serious material deprivation, but it does not preclude happiness.184 Furthermore, some participatory research has found that community defined wealth-ranks tend to incorporate both material and psychological wellbeing rather than material or income-based comparisons alone.3 A single previous study on pastoral wellbeing as measured with the SWLS was conducted on Kenyan Maasai from the Siana Plains.185 Other “wellbeing” studies reviewed were almost exclusively material or wealth oriented, although a recent qualitative study reported on differences in conceptual wellbeing associated with age and gender in a northern Tanzania Maasai community.186 The Siana investigation was part of a larger study of three relatively isolated, materially “simple” cultures (the others were Inuit and Amish) comparing their wellbeing to industrialized societies. Researchers concluded that the Maasai were the most satisfied of the three communities studied.  1.4 The Potential of Livestock-based Development Initiatives to Improve Health and Wellbeing in Poor Populations (Africa/Pastoral focus)  Sub-Saharan Africa (SSA) in a Global Poverty Context  World poverty has more than halved since 1981. A recent review of progress on Millennium Development Goals (MDG) celebrated substantial successes in poverty reduction but highlighted rural location, lack of quality education (explicitly at the secondary    13  level), infrastructural deficiencies, and social exclusion as continuing obstacles in lifting those who remain most entrenched in poverty.187 All of these factors are highly pertinent to pastoralists. Although poverty rates have dropped in every developing country and region, including SSA, by best estimates the absolute number living in extreme poverty has almost doubled in SSA, from 205 to 389 million people.188 However, as 61% of countries in SSA lack effective national poverty measurement tools, it is difficult to assess the accuracy of cited numbers.189 It is believed, though, that roughly half of the citizens of SSA, mostly rural dwellers, live in extreme poverty190-192 and the region has consistently been a negative exception in global poverty trends because population growth overwhelms economic advancements.14, 187 Predictions that one-third of the world’s extremely poor would be living in SSA by 2015 have come true and been exceeded.192 The extremely poor in SSA also suffered from a 21% poverty gap (the percentage living below the national poverty line) in the last available data, double the poverty gap of the next poorest global region.191   Poverty in Kenya The Kenyan gross domestic product (GDP) is higher than average for the developing nations of SSA, and it is categorized as a LMIC.193 However, World Bank (WB) country categorizations can mask high proportions of citizens living in poverty and the Oxford Poverty and Human Development Initiative (OPHI), locates 72% of the multidimensionally poor in middle income countries.194 The Kenyan GDP growth rate increased from 3% to 7% between 2003-2007190 but during roughly the same period, the proportion of poor almost doubled.190 The GDP growth rate has stayed fairly stable since,193, 195 and has overall contributed to a drop in national poverty. Most recently the OPHI ranked Kenya 85th out of 120 countries (a low numerical value corresponds to high poverty) on the Oxford Multidimensional Poverty Index (MPI) using data from the 2014 Kenyan DHS (based in part on a 39.9% national poverty, 14.5% severe poverty, and 12.9% destitution rate).196 However, because of population growth and high rates of rural poverty, Kenya had the sixth highest absolute number of citizens living in extreme poverty.197, 198 Significantly, it also has the second highest inequality ranking in East Africa,199 and ranks ninth in Africa,200 a probable consequence of a prioritization of urban over rural development201 and extensive    14  corruption.202-205 Major decentralization initiatives have been implemented in recent years, but rural Kenya is still substantially excluded from political decision making.206 Corruption penetrates almost every level of government and development initiatives, exacerbating inequality and hindering inequality reduction efforts. The latest MDG report for Kenya showed that some gains had been made, but in many arenas progress was slow. Poverty, under-five mortality, and infant mortality rates were double the 2015 MDG target, while maternal mortality rate was triple the target.207  Rural poverty rates were higher than urban, 38.8 and 29.4% respectively in 2015/2016.208 Because a high proportion of citizens live rurally (68%209), most of the poor (85%) and extremely poor (91%) are found in rural Kenya,210 which was a focus region for the WB Voices of the Poor participatory poverty assessment. Voices of the Poor is 20 years old and did not include Narok district, but its findings were similar to those of the current work. Some relevant perceptions from poor Kenyans included: pessimism about the future, and a feeling that relative poverty had worsened; a belief that poverty was generational; a tendency for the wealthy to attribute negative and false stereotypes to the poor; and identification of fees-for-services as a contributor to increased difficulty in accessing education and healthcare. The research also concluded that FHH were disproportionately represented in the poor (20% higher rates than male-headed households (MHH)) and very poor categories (double the number of MHH on average) in every region surveyed.211 Pastoral poverty is predominantly associated with no/low livestock holdings.144 Insufficient holdings can result from any or a combination of: generational poverty,212 drought,94 livestock disease,94 raiding/thefts,94 land loss,94, 118 land privatization,213 misfortune,214 gender (female),186 corruption,120 sickness,215 lack of resilience,216 or decline of the social safety net.147, 217, 218 Small herd owners tend to be more vulnerable and least able to recover from loss.108, 214, 219, 220 Pastoralists are also often politically neglected and Kenya is no exception.147, 221 Kenyan data suggests some of the highest poverty rates are in pastoral districts95, 108 and northern pastoralists are among the worst off due to extra stressors of civil and cross-border conflict222 (which has become increasingly weaponized with wars in the Democratic Republic of the Congo, Northern Uganda, Sudan and Ethiopia).94, 223, 224  In contrast, in spite of many reports on increased pastoral poverty14, 94, 115 as well as more than a    15  century of forecasts of pastoralism collapse,225, 226 there is evidence to suggest that only some pastoralists are struggling.94, 95, 108 Recent evaluations have also provided well supported claims for substantially larger national pastoral holdings in Ethiopia, Uganda, Sudan and Kenya than government figures suggest,136, 227, 228 confirming results found in smaller scale research. These conclusions, however, are likely not mutually exclusive as research has also suggested increasing inequity in pastoral populations.46, 94, 95, 108, 115, 218, 229   Knowledge Gaps and Poverty in Narok County and Narok South Sub-County  Personal observations of unreported human and livestock morbidities and mortalities, and lack of adherence to national law in Olkoroi and surroundings over 14 years, suggested local representatives of the Kenyan veterinary, medical, educational and legal systems did not document rural events that might have been recorded in urban centres, in large part because they lacked monitoring resources. Specific examples related to reportable diseases (of both people and livestock), crime (including murder), education access, female inheritance, marital age and circumcision. Medical and veterinary officials in Narok town openly admitted that little rural data was gathered (the district is classified as 92.9% rural compared to a national average of 67.7%).181 Law enforcement officers posted in Ololaimutia (a vehicle-hour away from Olkoroi) stated there was effectively no police presence in Olkoroi and similar communities. Even when data were collected, personal review of reports (from the Olkoroi medical clinic) revealed frequent tabulation errors, under-representative statistics (some residents rarely attended the clinic when sick), poor data security (two years of Narok District rural clinic reports were lost due to computer malfunctions when the district was sub-divided), and research could be shelved without analysis.230 District Veterinary Office (DVO) officials stated that for the most part “the focus of the office is on major towns”. They were also frustrated because “…there is no funding for recruitment, employment and deployment…” and simultaneously dismissive of the people they served: An expectation that the poor villager might understand globalization is ludicrous. The World Bank believes that a Maasai citizen should understand what is a public versus private good. This is a person who doesn’t perceive the consequence of sharing shelter [with livestock].     16  In addition, officials were rueful, “…the truth is bitter…” and resigned “For us Africans, if it does not kill it is not a big problem.” The DVO put the blame directly on the: …West…industrialization…SAP [Structural Adjustment Programs]...The private sector has been promoted, but implementation of this strategy was premature and better sources of funding have not been developed. SAP have been destructive. With structural changes, the economy cannot bear the weight of all the sectors.   A 2011 Kenyan inequality assessment231, 232 ranked Kenyan countiesb on a variety of health, education and infrastructure indicators. Narok ranked well on markers such as proportion of the population living below the national poverty line, and Narok South sub-county (location of Olkoroi) had similar rankings to the county as a whole. Narok was also the second most equal county nationally, with a Gini Coefficient of 0.315,232 but ranked less well on education and health-related indicators. Malaria and TB frequencies for Narok placed it in the upper 1/3 of county rankings, but 25th of 47 for human immunodeficiency virus (HIV) prevalence, 22/36 for rural counties. County level infrastructure outside town centres was relatively poor,231, 233 and average figures masked major within-county variations. Naikarra ward (where Olkoroi is situated), had the lowest proportion of citizens with primary and secondary education, and the highest with no education: at 81%, more than double than the best ward (43%).231 Many small communities of Narok South Sub-County, like Olkoroi, had significant deficiencies in educational, judicial, medical, veterinary and sanitation services. Although it is one of the richest counties because of game reserve revenue, there were enormous rich-poor divides in Narok (raising some questions about inequality rankings), and it was widely believed that systemic corruption and revenue diversion were the primary reason for the lack of infrastructure.234-237 Despite ranking forth of Kenyan counties for revenue generation, Narok County was also one of 14 counties which qualified to receive equalization payments from the Commission on Revenue Allocation (CRA) in 2015:                                                  b After 2010 constitutional amendments, the highest level of geopolitical organisation in Kenya became the county, of which there are 47. The next level down are sub-counties, within which are wards.     17  ‘Only a few tycoons control the wealth of Narok,’ CRA research and policy director Linet Oyugi told the Business Daily on Tuesday, adding that a small number of Narok residents benefit from the billions of shillings generated by the Maasai Mara.238  Corruption affects income inequality and economic growth, and is a negative predictor of wellbeing.239, 240 While it is difficult to find county specific information on corruption in Kenya, there has been documentation of corruption associated with land subdivision in Narok and Kajiado.70, 118, 119, 241 Staying Maasai, which looked at Maasai household economies in Kenya and Tanzania, with a specific focus on livelihood, income diversification and the impact of conservation policy found, like others,242 that corruption associated with game reserves and consequent lack of local benefit were the norm.108    Livestock and Poverty  Numbers are uncertain,136 but it is estimated that more than a billion global poor depend on livestock for part or all of their food security and livelihoods.6, 243, 244 Up to two-thirds of rural populations may be livestock-dependent, twice urban rates,29, 79, 245 and pastoralists are the most numerous livestock-dependent peoples.244 Just as it is difficult to accurately determine the number of livestock-dependent poor, it is also challenging to enumerate their livestock: estimates range from 1-20 billion.246, 247 Livestock keepers typically earn one-fifth to one-half of family income from their animals, though pastoralists often derive more, and “pure” pastoralists may survive on livestock alone. Up to 95% of  keepers in some developing countries live below national poverty lines,246 and in SSA, home to more than 400 million poor livestock-keepers, roughly 85% live in extreme poverty.243  Unfortunately, as recently as 2003, most national Poverty Reduction Strategy Papers (PRSP’s) ignored the role of livestock in the lives of the poor, the potential of livestock for poverty alleviation, and the contributions of livestock to national economies. Only four of 49 available PRSP’s in a 2003 review contained detailed strategy and budgetary consideration of livestock, an omission that was passively encouraged by the International Monetary Fund (IMF) and the WB which required no substantive consideration of livestock or other resource sectors typically important to the poor.248 In addition, the self-identified needs of poor livestock keepers have not been effectively documented, and neither are the positive nor negatives impacts of livestock on the health of the world’s poor well understood. Both needs    18  and health effects are likely variable both between and within countries. Lastly, what is prioritized in academic literature about human health-livestock relationships of poor livestock keepers (zoonoses, for example) does not always accurately reflect the real-world priorities of academics and/or poor livestock keepers.249, 250  According to Heffernan:22 ...poor livestock keepers are those who are economically and/or socially at risk and whose animals, at most, provide subsistence or the minimum augmentation of daily nutritional requirements...a poor livestock keeper does not own enough livestock to meet basic subsistence needs, yet depends upon his or her livestock.22  The word “risk” highlights the vulnerability of those who own insufficient livestock to support family needs, to diversify, to improve their SES,66, 130 and/or to facilitate escape from poverty for their children (as per Voices of the Poor). In addition, they are more vulnerable to complete loss of holdings and descent into extreme/persistent poverty if disaster strikes.130, 251, 252 Pastoral research has shown that significant livestock loss is associated with major risk of severe and chronic poverty and reduced health indices for adults and children.215 Nonetheless, livestock are also an important potential pathway out of poverty. A study of 1706 households in two Western Kenya communities, found 42% of families who escaped poverty in the preceding 25 years did so via livestock-based diversification.253    Livestock in Kenya (focus on pastoralists)  As is the case across Africa, agriculture is critically important to the Kenyan economy employing approximately 60% and 80% of the total and rural population respectively. Agriculture generates 51% of the GDP, 24% directly and 27% indirectly,254 however, the economic contribution of livestock has been significantly underestimated since colonial times.95, 136, 255 In 2009, the  International Authority on Development (IGAD), an 8-member East African trading bloc, concluded the value of livestock to IGAD nations may be 150-350% higher than official government estimates, depending on the country of focus.136 For Kenya, the revaluation was 2.5 times official estimates, and, at approximately 350 billion Kenyan Shillings (KSh), roughly on par with crop-based contributions to the agricultural GDP, estimated to be worth 410 billion KSh.256    19  It has been estimated that pastoralists are two-ten times more productive per land unit than any other system advocated for ASAL.14, 99 The lack of current, accurate, pastoral demographic data,95 however, makes even rough calculations of productivity difficult. In Chad, it was suggested pastoralists produced up to 15% of the GDP despite constituting less than six percent of the population.100 In Kenya, pastoral population size is between eight-ten million.257 Population estimates combined with production data136 imply pastoralists could be generating as much as 20% of the Kenyan agricultural GDP, roughly proportionate to their population size. Since pastoralists make other economic contributions, most prominently in tourism, their contribution may exceed their proportional demographic. Kenyan livestock productivity has been perceived as lower than its potential since the colonial era, and initiatives to increase output have targeted different producers and stages of the production process, from disease control1 to fattening pens and slaughterhouses.258 As already noted, these initiatives have rarely been successful, especially in pastoral regions.24 Some of the most commonly cited reasons for failures include: underfunding of the livestock sector in favour of plant-based agriculture;259, 260 a focus on medium and high intensity108 and/or large-scale production systems or services/models for such systems (for example disease free zones),260 which excludes the majority of Kenyan livestock producers; inability of most producers to meet Western standards of product quality, thus preventing Kenya from accessing lucrative export markets;255, 260 tariff systems that prevent nations like Kenya from even potentially penetrating Western markets;260 widespread endemic and periodic epidemic livestock diseases,29, 94, 260 insufficient veterinary resources to serve most rural populations, and since the SAPs of the 1980’s and 1990’s, difficulty in  accessing privatized veterinary services;108, 114, 261, 262 promotion of breeds and technologies ill-suited or unavailable to most of the nation’s smallholder or rural producers;8, 255, 263 and, lastly, not only lack of support for, but a long-standing prejudice against pastoral producers.108, 260  Livestock productivity barriers in Kenya, as listed above, are many and diverse. From a human health perspective, one of the most important is veterinary resources, because of their potential impact on animal health which in turn has multiple effects on human health. Recognition of the potential of veterinary services to improve global public health led the Office International des Epizooties (OIE aka World Organisation for Animal Health), to    20  include One Healthc  in all OIE Performance of Veterinary Service (PVS) evaluations, as of 2013.265 U.K. Department for International Development (DFID)-funded research in Kenya found veterinary costs were a barrier for the poor, but accessibility was the most significant impediment. It was also noted that knowledge about effective use of veterinary drugs was lacking in both the non-veterinary vendors and keepers. Inappropriate use of veterinary medicines has the potential to cause negative health outcomes to both livestock and their owners, and evidence suggests such problem already exist.266-268 The DFID research further observed wealthier livestock owners were more able to access and benefit from veterinary services, and study participants felt veterinary professionals favoured the wealthy.245 Since poor keepers have fewer resources for treatment, and are likely to experience proportionally higher livestock morbidity and mortality, better livestock disease control could increase national productivity and simultaneously contribute to poverty alleviation.  Poor veterinary infrastructure in Kenya additionally precludes accurate assessment of livestock disease frequencies, including zoonoses, and official records significantly underrepresent true rates. Both the World Health Organisation (WHO) Global Burden of Disease Report (GBD) and OIE collect some information on zoonoses but fail to report on a number of zoonoses important to the developing world. A 2013 publication reported that of more than 600 known zoonoses, at least 100 are of medical significance world-wide, but the GBD reports on only 11 and OIE, 13. Similarly, many livestock diseases are found in both richer and poorer regions, but those endemic and unique to developing nations are often neglected in research.269 In 2012 it was estimated that 99.99% of all livestock mortality in Africa was not being captured in official OIE reports and records.79 In addition, only three African nations have surveillance programs to monitor antibiotic use in domesticated animals.268 For this thesis, when livestock morbidity and mortality reports were requested from the Narok DVO in 2008, most months contained little or no data. Although the WB was in great part responsible for deterioration of veterinary services in many African countries, it                                                  c One Health is a health model that seeks to combine human and veterinary health knowledge and resources in research and intervention initiatives for greater effectiveness, especially in resource-poor settings. 264. Bardosh K. One Health: science, politics and zoonotic disease in Africa: Routledge; 2016.      21  is now exerting pressure along with other agencies, on these same nations to improve veterinary capacity.269 Reinvestment in veterinary services, especially through a One Health model, could improve human health both directly through dual service delivery but also indirectly due to a reduction in zoonotic disease transmissions, increased productivity and consequent improved SES. In writing on the lack of consensus in defining poverty, Akindola noted poverty creates deprivation in many realms that in combination reduces human capital, further elaborating that to allow “experts” to define poverty based on economic measures alone must result in an inevitable failure to capture the full experience of the poor. He also suggested that it may be necessary to define poverty according to local experience because “What constitutes poverty for one individual, for example, is not necessarily the same for another”.270 In the context of pastoralism wherein livestock play a multidimensional role that goes far beyond a source of livelihood, veterinary deficiencies affect more than household revenue. The predominant focus on productivity improvements and livestock as a pathway out of poverty without consideration of the centrality of livestock to pastoral life is another form of pastoral undervaluation.   1.5 Existing Reviews of Livestock Based Interventions There are numerous widely cited papers and narrative reviews on the potential impact of livestock-based interventions on human health and wealth.9, 16, 29, 79, 271-275 Although some include consideration of internal and external barriers to livestock-based poverty alleviation, there are very few systematic reviews on the topic. The well-known, non-systematic but extensive Livestock in Poverty-Focused Development (LID) review,16 assessed trends and causes of failures in livestock-based development, and is still highly relevant. Most of the projects reviewed were not specifically poverty focused, but those that were generally had a positive, albeit modest impact. Technological and service oriented projects were the most common but frequently, technology either did not reach target communities, or, when it did, was inappropriate for poor recipients and more likely to benefit already privileged community members. The LID team also reviewed organisational and institutional projects. The former also had a high    22  failure rate, primarily because they were not economically viable without external funding, or new initiatives were not ultimately integrated into existing frameworks. Institutional projects, however, though relatively small in number, showed promise (the approach was relatively new at the time of the review). An example of such a project is the Oxfam Wajir Pastoral Development Project which ran from 1994-2003 in Northern Kenya and worked with communities affected by drought and ethnic violence. The initiative successfully increased average pastoral incomes through support for the formation of pastoralist associations which became models for similar organisations across East Africa. The project plan explicitly aimed to and succeeded in incorporating multiple levels of government in a dialogue and process of peace, community development, service improvement, and poverty alleviation.276 Another non-systematic review, by Wanyoike et al. (2011), assessed a random sample of livestock development projects. Performance indicators were based on development agency criteria. Cluster analysis was used to differentiate successful versus unsuccessful projects, and identify determinants of success. It was concluded that project size, participant diversity, incorporation of institutional development, and effective monitoring and evaluation were important predictors of success. Inconsistency on the part of government collaborators and inclusion of non-livestock focused initiatives were associated with project difficulties. The authors concluded that 60% of the reviewed projects had failed.277 Three systematic reviews of interventions related to livestock and the poor were located. Although the first, Community animal health services for improving household wealth and health status of low-income farmers278 concluded that eight of 14 projects reviewed had produced identifiable improvements, only two measured outcomes that directly affected farmers. None of the projects considered gender or initial SES, and it was noted that methodological quality, study design and sampling strategies were variable and unclear, hindering study comparisons. Only five of the studies had clearly described outcomes. A Cochrane Review protocol (2002) and updates (2006, 2011) by the same group with the same title, sought to include only individual, cluster and quasi-randomized controlled trials and controlled before-and-after studies on the topic, but remained a protocol only (given the paucity of literature in the field). The second review A review of the effectiveness of agriculture interventions in improving nutrition outcomes279 only included two livestock    23  studies. Authors again found it difficult to compare studies across different study designs and interventions, and reported that study designs were often inappropriate for determining cause-effect relationships. The last review Can Interventions to Promote Animal Production Ameliorate Undernutrition? 280 also found major limitations in study design, evaluation and analysis, and a lack of statistical testing and analysis in many studies. The authors noted that little attention was paid to intermediate outcomes which can be important in explaining how interventions have their effect. They concluded that better designed studies with appropriate evaluation of intermediate outcomes were required to properly understand the potential and mechanisms of livestock-based productivity interventions. Publication bias undoubtedly influences published reports (reports of failure may exist only in the grey literature281), especially given that interventions are often funded by non-governmental organisations (NGOs) and development arms of national governments.282 Nonetheless, there is evidence to suggest that research, projects and intervention designs do not adequately incorporate concerns of targeted populations, success rates are not optimal, and successful interventions do not always reach those in greatest need.283 Specifically, within poor communities there are often differences in relative effects between more and less advantaged populations such as women,284, 285 culturally excluded groups, or those in the lowest socioeconomic categories.286 In 2015, given the scarcity of available reviews, I conducted my own literature search on the effectiveness of livestock-based interventions to improve wellbeing in the poor. No time limits were set on publication dates and outcome terms were selected to include potential health, economic, knowledge and social benefits. More than 1500 academic publications were extracted from Medline using four basic conceptual search terms: poverty, livestock, interventions and outcomes, and no time limits on publication dates. Boolean logic searching capacity, keywords and medical subject heading (MeSH) were used to optimise sensitivity and specificity. Only 16 papers attempted to measure health outcomes for human participants as well as livestock-related productivity and health outcomes. Many of the reviewed papers were theoretical only. Of the 16 relevant papers identified, three did not explicitly measure human outcomes reporting only descriptive conclusions, and four measured only financial outcomes. Furthermore, only: four had a control arm, four had considered SES, three followed up longer than a year, and one included    24  gender in study design and outcomes. Methodology, communities and livestock types investigated were highly variable making comparisons extremely difficult. Although the initial papers examined were from Medline only, and therefore potentially excluded more agriculturally oriented articles, searches of other sources such as AGRIS, Ag-Econ, and Agricola, yielded few additional articles.  The most recent relevant review, Evaluating one health: Are we demonstrating effectiveness? (2017), specifically examined One Health oriented literature published between 2003 and 2015 using abstract search terms which captured “One Health research, action (e.g. collaboration, surveillance, zoonotic disease control program integrated across animal-human-ecosystem interface) or case studies”. As in my own search, although a large number of potentially relevant articles met the initial search criteria, only seven of 1839 papers included any quantitative measurements of outcomes. The authors concluded there was a significant lack of evidence to support the claims of One Health proponents.60  1.6 Research Goals and Objectives The overarching goals of this thesis were three-fold. Firstly, to develop a data-informed understanding of the relationship between pastoral livestock and owner health via a longitudinal health study on both keepers and their livestock simultaneously. Secondly, to both prioritize the perceptions and values of the community and illuminate the quantitative data through a series of cross-sectional surveys on psychological well-being, disease priorities, and livestock rearing practice, with an emphasis on possible gender differences in perception and practice. Thirdly, to develop a series of exploratory statistical models to describe key livestock and human wealth and health outcomes. A literature review was performed and data was collected via an in-depth, participant-informed, mixed-methods investigation of a single, rural, traditional Maasai community, Olkoroi, in South-West Kenya according to the six primary research objectives listed below:  1. To review the literature on livestock health, welfare, and productivity, livestock-mediated development projects, and livestock-human wealth, health and wellbeing relationships in poor and/or traditional livestock-keeping communities, with a primary focus on Sub-Saharan Africa (SSA), pastoralists and the Maasai.     25  2. To develop a detailed demographic, socioeconomic and cultural understanding of individuals, households, community structure, wealth perceptions, and livelihoods of Olkoroi in order to most effectively understand and assess how livestock-keeper relationships varied along social gradients in small, rural communities such as Olkoroi.  3. To identify Olkoroi human and livestock disease priorities, livelihood challenges, and explore individual selection and evaluation rationales.  4. To compare longitudinal, self-reported household livestock and human morbidity to Olkoroi clinic reports, district medical reports, community disease priorities, and the pastoral literature. 5. To assess individual psychological wellbeing within the community and compare wellbeing between the sexes. 6. To use the data collected to build models to identify the household and livestock variables that explained the most variation in individual psychological wellbeing, herd size, young livestock growth, herd and human household health status.  1.7 Livestock-Based Interventions: Opportunities and Challenges Livelihood is inextricably associated with health in a bidirectional manner. Capacity to sustain livelihood is strongly affected by health status. Globally, but particularly in poor countries with weak social services, ill-health is one of the most common routes into poverty.287 Conversely, livelihood affects health through multiple pathways from remuneration to occupational exposures and corresponding risks for chronic and infectious diseases, as well as physical injuries. Agriculture is particularly important as a determinant of global health for both agricultural workers and consumers of the sector’s products. In richer nations, less than five percent of the workforce is now found in agriculture.288 In SSA, agriculture employs 60% of the population, and, according to the FAO, drove 50% of job growth between 1999 and 2009.289 Despite these strong links, health and agricultural domains do not always engage cooperatively or effectively to optimize health.290 Although One Health and Ecohealth research paradigms have gained prominence beginning with avian flu pandemic concerns in the early 2000s, resistance to veterinary and medical collaboration    26  continues to be a problematic obstacle to productive integration of health concerns of livestock keepers and their animals.264 Interdisciplinary engagement emphasized in Ecohealth, and necessary for such integration, particularly the inclusion of social scientists who might help to bring the voices and priorities of keepers to the fore, is also limited.264 Barriers to attainment of livelihood security for the livestock–keeping poor include concerns such as acquiring, maintaining and retaining livestock, optimising productivity, and accessing effective marketing opportunities.16 For most pastoralists there are additional challenges of coping with marginalization and climatic variability which appears to be worsening due to GCC. Given the current diversity of problems faced by the poorest pastoralists, recovery from periods of drought and disease may be becoming more difficult. FHH may also face gender restrictions that affect accessibility of livestock-keeping resources (including the first step of livestock acquisition) and/or participation in capacity/productivity improvement initiatives. Research indicates, for example, that women are much more likely to be excluded from producer organisations291 which help to strengthen smallholders through increased group production and market impact.292 Although there is little specific research on the relationship between livestock health and/or productivity and the health of pastoral keepers,27 a wide variety of studies on pastoral productivity have identified a number of variables associated with herd size and resilience (resilience referring both to continuation of pastoral tradition, as well as ability to maintain herds through, and rebuild after drought and/or disease epidemics). Critical correlates include: family size, household composition, shifting gender roles, education, diversification, land ownership, and geographical location.108, 219, 293-297  Livestock associated human health hazards can increase risk and vulnerability to poverty in a potentially downward spiral.290 It has been claimed that reduction or, if possible, zoonoses elimination would have a significant impact on the most vulnerable.14, 29 If true, simply tackling endemic zoonoses could theoretically result in a dual investment benefit of reduced human and livestock morbidity and mortality. A significant improvement in the health and productive capacity of poor livestock keepers could additionally improve earning capacity in part through enhancement of the health, productivity and consequently value of their livestock assets. Unfortunately, with some exceptions, (for example current efforts to    27  control rabies, and the successful eradication of rinderpest) global priorities and research investment tend to focus on outbreak diseases that pose potential threats to rich nations such as Ebola and Avian flu.264 A further complication is that in practice, zoonoses rarely appear in self-reported lists of the challenges faced by poor livestock keepers. A number of studies on the pastoral perspectives on zoonoses have concluded pastoralists do not prioritize or even recognize zoonotic illnesses within their health landscape.29, 298-303  Development projects have frequently been predicated on developed world agriculture, leading to the repeated introduction of large scale, high intensity production systems,24, 248, 291, many of which have failed.24, 291 There is also an ongoing tendency to prescribe intensification of smallholder and extensive production systems as the best route out of poverty255 despite evidence to suggest smallholders can be highly efficient even using traditional practices.291 Suggestions of intensification are often unrealistic, particularly when applied to livestock-based systems like pastoralism where: livestock play a multiplicity of roles, some of which could be lost with a shift in production mode; intensification is initially expensive and requires speciality breeds, feed and resources which are unaffordable for most poor livestock keepers; national veterinary systems already fail to meet needs of poor livestock producers and high intensity systems usually need greater veterinary inputs; poor producers commonly live in relatively isolated regions that are impractical as locales of intensification due to potential environmental harm to fragile ecosystems, expense and lack of infrastructure; and many rural producers face obstacles due to illiteracy, which already causes harm due to misuse and overuse of various pharmaceutical products and tools.304-307 In addition, given that pastoralists appear to already be much more productive and contribute significantly more to national economies than they have been given credit for, it might be more effective to find culturally and location specific mechanisms to address problems faced by existing production systems rather than try to impose western production methods.   Fundamentally, even though livestock-based poverty alleviation interventions have been promoted for over a century now, significant challenges and knowledge gaps remain: 1. Many governments still appear to lack interest and/or awareness of the economic importance and contribution of their low-input, traditional livestock producers, and    28  continue to ignore their perspective even when making decision which may have major impacts on their livelihoods.  2. There are still enormous knowledge gaps, and a serious lack of comprehensive, quality data on relevant populations, from basic human and livestock demographics, to productivity, morbidity and mortality statistics in livestock and their keepers.79, 80, 308 There is insufficient data on many neglected tropical livestock diseases and the complexity and diversity of livestock diseases-human wellbeing relationships make it very difficult to draw conclusions on how to prioritize and effectively address the problems they create.27, 264, 275  3. While there are many intervention studies, most are so highly variable in quality, measurement tools, and study design that they cannot be used for systematic reviews. Even the higher quality studies are so diverse it is basically impossible to draw any clear conclusions on intervention efficacy or to make comparisons between studies. Furthermore, most intervention studies use designs that are highly prone to bias, lack long term follow-up and consideration of socioeconomic variability within communities or consideration of gender factors in intervention outcomes. 4. There are few models to assist in predicting economic, health and wellbeing impacts of different interventions. Proposed interventions, whether small or large, rarely have goals beyond economic improvement. Measurement of benefit distribution or concrete health outcomes is often omitted, and the impact of increased revenue on cultural and social aspects of participating communities is rarely considered. 5. There is a lack of consensus on which diseases, whether livestock specific or zoonotic, should be the focus of interventions. Without consensus, it is difficult to prioritize interventions appropriately. To complicate matters further, controlling high impact diseases does not always produce high impact outcomes due to issues of feasibility and costs. Lastly, there is often priority disagreement between livestock keepers and “experts”. Effectively it is not really known which diseases matter the most, nor is it agreed as to who should make priority decisions.29 6. There is no clear agreement on how best to support poor livestock producers, in great part because of lack of comprehensive data, and lack of agreement on priorities.264     29   1.8 Thesis Overview This thesis is structured with seven chapters. The first chapter provided background on the current status of traditional livestock keepers with a predominantly African perspective and a specific emphasis on Maasai pastoralists. The contribution of livestock to Maasai livelihood, SES, wellbeing and health is described as well as the position of the Maasai within the Kenyan political system. Chapter one also included an overview of rural poverty in Kenya, the significance of livestock to the global poor, and the contributions of livestock to the Kenyan GDP. The chapter concluded with a discussion of the reasons that livestock-based development initiatives have been mostly unsuccessful for over a century and highlights the knowledge gaps which continue to be a barrier to more effective use of livestock to improve the wealth, health and wellbeing of poor livestock keepers.   Chapter two details the research methods used to collect data for each of the following chapters three-six, and the analytical approaches used to interpret the data.  Chapters three-six present the main results for each component of the research with a discussion, illustrated by quotations extracted from the qualitative semi-structured interviews. Chapter three reports on the descriptive epidemiology of Olkoroi including community resources, household structure and SES, culture, and livelihoods. The data described were also used in model building for chapters on livestock and human health and wellbeing. Olkoroi characteristics are linked to the larger body of research on the Maasai.   Chapter four presents the conclusions of community focus groups on wealth markers, cluster analysis of household wealth markers, the results of exploratory modeling of psychological wellbeing, and gender differences in self-reported perspectives on wellbeing.    Chapter five describes self-reported perspectives on livestock-based livelihood including livestock duties, the contribution of livestock to wellbeing, livelihood constraints, and livestock disease prioritization and understanding. The results of three exploratory analyses to describe HoH and household-level variables most strongly associated with the size of household livestock holdings, young livestock growth rates, and self-reported livestock disease prevalence are discussed.    30    Chapter six reports on the results of a baseline cross-sectional community health survey, describes health promotion behaviours, and human disease prioritization and understanding. Comparisons are made between district level, and self-reported disease frequencies. A final set of human health exploratory models identifies variable most strongly associated with child and adult total infectious disease burdens, as well as incidence of the most commonly reported individual disease categories: malaria, respiratory infection and gastrointestinal (GI) illness.   Chapter seven is a summary and integrated discussion of the most important finding of this research, unique contributions, and recommendations for future research and intervention development.    31   Data Collection and Analytical Methodology 2.1 Setting, Population and Ethics Approval Olkoroi is a small, traditional Maasai community situated in Naikara Ward, Narok Constituency South, Narok County, Kenya (latitude and longitude, 35E 1.7S). Ecologically, Olkoroi is located in typical savannah, but has access to a year-round water source. As a result, while dry and drought periods severely impact crops and livestock, the human population is always able to easily obtain water for basic needs. The community strongly identified as traditional pastoralists, but most households were in permanent residence, particularly families with younger children who attended the local primary school.  Approximately 75 households lived in Olkoroi, comprised of 459 individuals, 150 adults and 309 children. Roughly forty children attended boarding school but returned for holidays. The ratio of women to men was 1.5:1 (on average, 91 women and 59 men), due primarily to polygamy, but also because of a traditional age gap at marriage which frequently lead to early widowhood. Three quarters of livestock-owners moved their animals due to drought or seasonal pressures, but the community size remained relatively stable because when animals were moved, they were often left with relatives or hired herders. Therefore, even if HoH moved their animals, they did not usually stay with them for a prolonged period. Consent of Olkoroi adults was 100% for participation in the research, with two exceptions in 2008 (the first year of data collection): only 63% (95/150) participated in the cross-sectional health survey, primarily because it was carried out by a Canadian volunteer nurse who was new to the community, and only 45% (67/150) provided blood samples for hematological tests, because of widespread suspicion of unauthorized and surreptitious HIV testing by medical professionals in Kenya. Full participation occurred for all other studies. When residents were away during longitudinal data collection, their information was excluded until their return. Those absent for cross-sectional studies completed interviews when available, however total participation varied because of death (3.3%), marital conflict (1.3%), community exclusion (2%), and residence in more than one geographical location (2.7%). All data on children was obtained through parental interviews, primarily maternal. Ethics approval for the research was given by the UBC Clinical Research Ethics Board (H07-02752) and the Kenya Medical Research Institute (Non-Scientific Steering    32  Committee Protocol Number 164). I was an official affiliate of the International Livestock Research Institute (ILRI) for the duration of my research, through which I was granted permission to engage in livestock-related research in Kenya (ILRI has blanket research clearance under a host country agreement with Kenya). In addition, I completed both the UBC Experimental Animal User Training, and the Human Ethics Training programs. Household consent to research participation was obtained in February 2008. HoH were contacted, and the consent form reviewed in Swahili or Maa. Agreement was documented by signature or thumbprint (the norm) on the form, and a Swahili copy was left with the family.   2.2 Research Development, Field Work Timeline, and Field Assistance The studies described in this thesis were carried out from January 2008-December 2010. Because I was employed full time for the duration of my research, I personally carried out data collection over three field seasons from January-April 2008, May-August 2009, and October-December 2010. I was present for all cross-sectional studies, and longitudinal data collection in the three field seasons, with the exception of physical assessments and related questions that were part of the 2008 baseline health study (which was carried out by a Canadian, volunteer, qualified nurse). When I was unable to be in the community, my field assistant, Alfred Koshal, was responsible for maintaining the longitudinal studies.  The initial research plan was to conduct five studies: focus groups to identify locally relevant wealth markers; a sociodemographic survey of household history, structure, and capital; a cross-sectional survey of household health and health-related behaviours with an emphasis on zoonotic risk factors; a baseline cross-sectional livestock health survey, and a longitudinal human and livestock health study. The cross-sectional household studies were to be carried out in Olkoroi and three surrounding communities: Laleta (to the east of Olkoroi), Ilkisaruni, and Oltulele (to the south), and completed in the 2008 field season. The longitudinal study was to begin in 2008 and run for two years. However, in 2008, a number of significant research issues became almost immediately obvious. The first was that it would not be feasible to collect longitudinal data from all four local communities with the time and resources available. Therefore, when the longitudinal study was initiated in May 2008, it focused only on Olkoroi households.    33  Secondly, household wealth and individual physical health would not likely fully describe individual wellbeing, particularly as regarded possible gender effects. Additionally, as we conducted the cross-sectional livestock health survey, and also interacted socially with the community in 2008, we received numerous owner reports suggesting young livestock morbidity and mortality rates were high, potentially having an important effect on herd productivity. The demographic survey also indicated significant variation in size of livestock holdings and a high proportion of households with no livestock.  To address these concerns and better capture the diverse influences on human and livestock health, six cross-sectional studies were added to the 2009 field season, and a second longitudinal study, of young livestock growth, was initiated from May 2009-December 2010. These studies were also carried out in Olkoroi only. The additional cross-sectional studies investigated: past and/or current responsibility for livestock-related duties in household adults (HoH and any affiliated adults, for example unmarried sons, sons and daughters-in-law who lived in an extended family unit, or widowed parents), adult life satisfaction via the SWLS (Satisfaction with Life Scale), prioritization and understanding of livestock and human disease, livestock vaccination status, self-assessment of household husbandry, and opinions on best husbandry practices for optimizing livestock productivity. My primary field assistant throughout the research was Ole Alfred Leroka Koshal, a long-time local resident whose father was one of the first settlers in Olkoroi. Ole Koshal was fluent in Ma, Swahili and English, and collaborated on all field work except the cross-sectional human health assessments, maintained the longitudinal data collection while I was in Canada, and did the majority of interview translations. A volunteer Canadian nurse accompanied me in the 2008 field season and performed the human health assessments. A number of individuals provided brief periods of assistance with tasks ranging from translations, filling in for Ole Koshal on rare occasions when he was unavailable during my field seasons, organization of the preliminary focus groups, community meetings to discuss the research, and coordinating the end-of-research livestock vaccinations.       34  Figure 2-1: Research timeline   2.3 Qualitative and Mixed Methods Data Collection  Personal Record Keeping and Informal Community Perspectives The semi-structured nature of much of the data collection, and community interest in the research, meant that respondents often added commentary or elaborated on their perspectives, with or without prompting. Open-ended interviews except the wealth focus groups, were voice recorded with participant permission (SWLS/livestock duties, disease prioritization, and best practice interviews). Other data collection was translated/transcribed at the time of the interview. Informal conversations on market days, during data collections, and home visits outside research activity provided further perspectives on individuals, families, local culture and livelihood practice. I was frequently invited to community meetings about the school, or celebratory events such as weddings, church activities, and fundraisers, and also hosted community meetings about my research in each field season. Detailed notes were taken at or following all meetings/events, for research and/or personal record-keeping and reflection, and I kept daily field journals in which I recorded events and interactions of interest. Data collection entailed visiting every community home every week while in the field, and I made field notes throughout each research day. We were often asked to have tea during data collection and discussed various aspects of research as per home-2008-2010: Longitudinal human and livestock health surveys2008: February: wealth marker focus groupsFebruary-April: cross-sectional livestock, human health, and sociodemographic surveys2009-2010: Longitudinal young livestock growth survey2009:May: SWLS study+ livestock duty interviewsJune: Disease prioritization interviewsJuly: Husbandry self-assessment and livestock vaccination statusAugust: Best husbandry practice survey   35  owner’s interests, at such times. Other forms of community engagement occurred via interactions with students and teachers (my own children attended Olkoroi Primary for two months, and my oldest daughter and I taught occasionally in the school), and discussion with over 250 women from Olkoroi and adjacent villages during interviews associated with a livestock-based women’s microcredit organisation (since 2004). I continue to correspond regularly with residents into the present and am usually notified about significant events such as the progression of land subdivision, criminal activity, serious livestock morbidities and mortalities, human mortalities, and major weather events. Throughout the following chapters quotes from research interviews (predominantly the SWLS/livestock duties interviews) and a small number of informal conversations are used where they provide community perspectives which relate to data interpretations. The specific source is identified with each quote. No thematic analysis was done of qualitative/mixed method responses, though answers were grouped into categories for the purpose of comparisons and analysis.   Single Interviews with CMF Employees and District Officials Perspectives and practices of relevant government employees, officials, and representatives of CMF were obtained via open-ended interviews in 2008, but also through observation, and interactions over the research period. As these interview were conducted in English, they were transcribed at the time of the interview. The individuals interviewed and the specific interview focus were as follows:   The Olkoroi head teacher (principal) was asked in 2008 about the structure of the Kenyan education system, school funding sources, sex ratios at the school, exam performance, government staffing provision, and challenges associated with education delivery in the community.   The Olkoroi clinic nurse was asked in 2008 about: his background, training and clinic responsibilities; the most common health problems in the region; services provided by the clinic; and solutions or approaches that could help to improve community health. In addition, the Canadian volunteer nurse was given permission to document the clinic pharmaceutical and equipment resources.    36   The directors of CMF Narok (one of whom was from, and had practiced as a nurse in Olkoroi after CMF training) were asked in 2008 about: CMF history, mandate, operational locations, and specific activities in the Maasai Mara; organizational and funding structure, and relationship to the Kenya Ministry of Health; the most serious problems seen at both the Olkoroi and Ewaso Gniro clinic in Narok town (the biggest CMF clinic); and human health concerns related to livestock.  Officers from the District Public Health (DPH) and Medical Office (MO) were asked in 2008 about: the position of their office within the Ministry of Health infrastructure; office funding; and the biggest regional challenges for the office.d   Officers from the District Veterinary Office (DVO) were asked the same questions as the DPH and DM officers, but also about: services provided to rural areas; liaisons with other district offices and agencies; and zoonoses focused program delivery. In addition, there were numerous posted flip-chart sheets from a recent brainstorming exercise (on office performance, potentials, and challenges) which I was allowed to copy. The veterinary officers were much more engaged and interested than the DM and DPH officers and extemporized widely.   Additional perspectives were obtained when visits of government officials to the community overlapped with my time in residence (for example the assistant chief for the Naikarra subdivision provided information about his official responsibilities and local law on brewing alcohol, and informal conversations were had with District Education Office officials about responsibilities and challenges of education delivery). Similarly, conversations were held with the community missionary whenever he and I were present in Olkoroi simultaneously, and he also provided his written perspectives on community health and wellbeing, by email in October 2008.                                                   d Access to these officers was very limited and the interview, of necessity, brief.    37   Wealth Marker Focus Groups At the beginning of the 2008 field season, separate groups of eight men and eight women, were simultaneously interviewed to obtain community identified wealth indicators associated with an a priori list of socioeconomic characteristics from the literature combined with community observation (some standard literature markers were not relevant). Participants were selected based on consultation with an informed community member, a Maasai primary school teacher who had lived in Olkoroi for two years. Residents from all SES levels were invited. Discussions were concurrently translated and transcribed. The list of SES characteristics was sequentially reviewed with each group. Participants were asked to identify possessions and practices associated with each characteristic which distinguished the wealthy, the coping (those who could meet basic needs but did not have extra resources), and the poor (those who struggled to maintain their households). The indicators discussed were: diet composition and sufficiency, health, income, employment, modes of transportation, educational of HoH and children, veterinary and medical access, livestock ownership, marriage/family composition, housing, clothing/footwear, household furnishings and energy sources (for light, cooking, and devices such as cellphones).    Livestock duties/SWLS/Wellbeing interviews The livestock duties script, the SWLS, and wellbeing questions can be found in Appendix B  Household adults (n=150) were first asked whether or not they participated/had participated (for “retired” adults) in a list of nine livestock duties compiled from observed, routine livestock tasks performed by community members, and potential routes of zoonotic exposures: herding, disease diagnosis, disease treatment, livestock buying, livestock selling, milking, slaughtering, butchering, and assistance with livestock births.  Following the duties interview, respondents were asked to self-assess their life satisfaction via the SWLS.309 The SWLS poses a series of five statements to which respondents are asked to rate their agreement on a Likert scale of one-seven. An overall measure of wellbeing is obtained by summing the total numerical responses. After the initial numerical rating, participants were asked to consider their life and identify the most important positive and negative contributors to their current wellbeing, then asked to think    38  about their future and identify what they thought was most likely to positively and negatively influence their future wellbeing. Participants did not have difficulty with the SWLS, however, when asked about contributors to and detractors from wellbeing, the first participant interviewed had some difficult grasping the scope of the question. In collaboration, Ole Koshal and I developed a list of life factors which was recited to participants to give them a sense of life components they could reflect upon. The list included: family (birth and current), marriage, children, health, educational attainment, tribal identification, location, dwelling, possessions, livestock holdings, shamba, land owned, daily workload, opportunities for employment, religious affiliation and/or church attendance. To conclude the interview, interviewees were asked first if livestock were a positive contributor to their wellbeing and to explain their answer, and secondly if livestock were a detractor from their wellbeing, and again to explain.    Livestock and Human Health Disease Prioritization, Rationalization and Causation Disease prioritization interviews were initiated by asking the interviewee to freelist the livestock diseases that had the greatest impact on the community. Participants (n=124) were encouraged to list as many diseases as occurred to them, and after the initial listing, were prompted for any additional diseases that came to mind. When complete, the list was read back to the subject who was then asked to consider the complete list and choose the disease which affected the community the most. Upon selection, the respondent was asked why they had identified that disease (rationale for choice). The process was repeated with the remaining diseases until the full list was ordered. Each disease was then reviewed in turn and the interviewee asked: what caused it, what was the most effective treatment, and in what season it most commonly occurred. When livestock disease had been fully discussed the process was repeated for human diseases. I transcribed the initial disease list and order at the interview. Discussions of rationale, causation, treatment and seasonal associations were translated and transcribed after completion of all the interviews.      39   Self-Assessed Livestock Husbandry and Best Husbandry Practice Interviews In order to better understand barriers to livestock productivity from the practitioner perspective, HoH and affiliated adults were asked to self-assess household performance of common livestock husbandry activities, herd characteristics and infrastructure (pens). The livestock duties list was used as the basis of the survey, with the omission of disease diagnosis, slaughtering and butchering, and assisting with births (because of the difficult of assessing these tasks), and the addition of pen quality (critical for livestock protection), vaccine delivery (because of the conversationally reported low prevalence), herd composition (balance of different types of livestock, because of the traditional Maasai preference for cattle) and cross-breed ownership (based on the frequently expressed desirability of such animals during the cross-sectional livestock health survey). Participants were asked to consider each activity/characteristic and self-rate the household on a scale of 1-3, where 1 was poor, 2 was average (kati-kati) and 3 was good (sidai).  Livestock husbandry was also explored via separate open-ended interviews in which HoH and affiliated adults were queried about the best approach for optimizing livestock productivity with regards to each of the tasks from the self-assessment. Each task or herd feature was introduced, and the respondent asked what they thought was the best method for maximizing productivity via that task or characteristic. After discussing each item on the list, participants were asked which activity was the most important overall for maximizing productivity, the best and worst performed by their specific household, and the rationale for their choices.  2.4 Sociodemographic and Quantitative Data Collection  Baseline Community Health Assessment and Sociodemographic Data Cross-sectional baseline health and demographic data was collected from February-April, 2008 via a modified version of the Nomad Health in Chad questionnaire previously used for human-livestock health research on mobile pastoralists,310 (Appendix A  A Canadian nurse fluent in Swahili, assisted verbally by a Kenyan assistant fluent in Swahili and Maa, conducted physical and physiological measurements including height, middle upper arm circumference (MUAC), blood pressure, and heart rate, followed by a series of questions    40  about recent illness, treatment-seeking, preventative health practices, and potential zoonotic exposure routes (“General Health” onwards in the questionnaire). Ole Koshal and I, collected demographic and household information. 63% of community adults participated.  I carried out blood tests taking samples via a finger prick in a school classroom. A QBC Autoread Plus Hemotology Analyser (Drucker Diagnostics, USA) was used to measure hematocrit, hemoglobin, white blood cell counts, platelets, granulocytes (%), lymphocytes (%) and mean corpuscular hemoglobin concentration (MCHC). Due to widespread rumors and suspicions of unauthorized and surreptitious HIV testing by medical professionals in Kenya, only 45% of adult community residents (n=67) participated in blood tests. All residents agreed to provide household demographic information. Village boundaries were set at the most distant homes on the north side of the community, and the geographic boundaries of the hills on the west, and the Sand River on the east and south. For  the purpose of household data collection, adult status included: males or females of any age who were in or had been in a marital relationship; unmarried females of any age with children; or single males of any age, who lived separately from the main family home but within the family compound (there were no female equivalents of this category). HoH were designated as any adult with or without children who lived in their own compound, or adults who lived and ate independently within an extended family compound. Sociodemographic information collected from participants at baseline in 2008 included: age (often estimated), tribal affiliation, time of arrival in Olkoroi, marital status (single, married, separated, abandoned, divorced, widowed), type of marriage (monogamous or polygamous), male age set (also applied to married women), male and female HoHs education attainment, number of children, education level of children, reasons for lack of education or prematurely terminated education (in adults and children), church attendance, sources of income, amount of family land, approximate proportion of family land under active cultivation, and preventative health practices (use of mosquito nets, treatment of milk, and source and treatment of household water supply). Additional household information was visually gathered for each household during home visits and daily encounters (type of house and roof, presence of a latrine, and vehicle, solar panel and/or cellphone ownership).     41  Sensitive information, such as physical abuse, alcohol consumption and brewing practice, was obtained indirectly throughout all field seasons (2008-2010) using a combination of longitudinal health self-reports, responses to the SWLS survey, direct observations, and informed residents.    Baseline Livestock Health Data Baseline livestock health data and initial size of household livestock holdings were collected in the 2008 field season from Olkoroi and nearby communities of Laleta (east of Olkoroi), Ilkisaruni and Oltulele (south). Each household with livestock was visited, and data collected on numbers of cattle, sheep and goats, current livestock health, and any additional livestock health observations made by the owner when prompted at interview conclusion for perspectives on animal health.    Community Characteristics Community-level attributes, such as the school, community stores, other businesses (such as a mill and small hair salon), hotels (small café/restaurants rather than lodging places), and churches, were visually identified. Class sizes and male: female ratios in each grade in 2008 were obtained from the head teacher at Olkoroi Primary. Kenya Certificate of Primary Education results from 2006-2009 were obtained from the Kenya Open Data government portal (http://icta.go.ke/open-data/). Informal observations were also made of teaching practice, presence/absence of government teachers and disciplinary interactions between teachers and students (which were sometimes public).    District Level Health Data In April 2008, computer files of the 2007 monthly reports from all Narok District South clinics (including Olkoroi) were obtained from the Narok District Hospital. In 2009 I attempted to obtain 2008-2009 reports, but during division of the district into two administrative units (North and South), Narok South records had been lost.     42   Narok District Weather Data Official weather records of daily highs, lows, hours of sunshine, and mm of rainfall for 2008-2010 were provided by the Narok weather office. Narok is approximately 50 km from Olkoroi in linear measurement, 3 hours travel time on rough roads, but weather patterns are similar enough to provide a general picture of climate variability.   Longitudinal Human and Livestock Health Data Collection Self-reports on family and livestock health were collected biweekly from May 2008-November 2010. Interview were usually carried out at family homes between 7:00-11:00 am before household duties necessitated HoH departures. Livestock interviews were occasionally conducted in pasture. Residences were visited consecutively, over 3-4 days at the start of each two-week period. If HoHs were absent, return visits were made in the evening or the following day. Family health data was obtained preferentially from female HoH, and livestock data from male HoH, because of traditional responsibilities for children and animal holdings respectively. However, family data was occasionally reported by male HoH, and livestock interviewees included male and female HoH, hired shepherds, and older sons, daughters and in-laws. In most months (24 of 31), 2 visits were made to each family. Interviews were conducted in Maa, translated and transcribed into English concurrently during the visits. No diagnostic confirmations of disease self-reports were made. Participation in human health interviews was universal but the community research assistant excluded families without livestock for the first year of data collection, therefore predictive human infectious disease models were built using May 2009-November 2010 data only. Data was collected from 75 distinct households between 2008 and 2010, but incomplete data sets resulted in exclusion of 23 households. Human health was tracked by individual household member. A total of 2820 household visits were used for model-building. Participation in livestock health collection was also universal but more variable than household health because some households moved their livestock to and from Olkoroi, most typically in dry/drought seasons. The highest proportion of animals were moved in the May-November 2009 drought. When animals were moved, data collection continued only on any remaining livestock but resumed if/when livestock returned. Forty-seven livestock-owning    43  households participated in the study, and 2407 visits were made over the duration of the study. 1787 small ruminant and 1749 cattle morbidity/mortality reports were tabulated.  Most livestock could be directly attributed to one family unit, however extended families sometimes herded cattle or their small animals together (no families shared all livestock herding). A few single women with small holdings left their animals with extended family or neighbours, and some households kept part of their holdings in Olkoroi and part with relatives outside the community. Data was only collected on livestock present in Olkoroi, and for analytical purposes, animals were attributed to the primary herder. In families where adult sons lived with their parents, all livestock were attributed to the paternal head of the extended household except in one circumstance where an elderly father had disbursed his holdings.  Due to capacity constraints, only biweekly livestock disease prevalence was monitored.    Young Livestock Growth Data From May 2009 to November 2010, growth rates of up to three each of young goats, sheep and calves per livestock-owning household were followed using biweekly heart girth measurements. The study was started with selection of the youngest available animals of each type within the defined age parameters. Measurements began at birth when possible, and continued to six months for goats and sheep, and a year for calves. As animals aged out, they were replaced if new subjects were available. Loss to follow up of the measured animals was frequent via disease mortality, literal loss during grazing, predation, or transfer to other herds. Mortality rates were highest in the 2009 drought. As a result, measurement duration per animal varied substantially. Animals for which fewer than three measurements were obtained were not included in analysis.  All livestock owners were willing to participate in the growth study, but 14% were excluded because they had no young livestock (literally or not present in Olkoroi) over the study, or too few measurements were taken. Lack or insufficiency of measurements occurred due to a combination of: loss to follow up as detailed above; birth timing; a high rate of livestock abortions exacerbated by the drought in the first year of data collection; and movement of livestock (mostly in dry season/drought, but also according to family needs and    44  activities). Most measurements were taken on goats, followed by sheep and cattle (283, 151 and 103 distinct animals respectively) because poorer families usually had no or few cattle, goats were the cheapest and most common small ruminant, and the pregnancy rate of goats and sheep was much higher than cattle. In addition, the combination of a longer gestation period, and the high abortion and mortality rate of adult cattle during the drought meant almost no calves were born through 2009 and early 2010. Lastly, wealthier families (more likely to own cattle) moved their animals much more frequently during dry season/drought.  Calves were measured in 29 families. The number measured ranged from 1-7 per family, mean 3.5, standard deviation (SD) 4.9. Goats and/or sheep were measured in 37 families. Total small stock measured per family ranged from 3-28, mean 11.6, SD 5.8. Data was collected biweekly, alternating with livestock health, using fabric tape measures calibrated to weight via heart girth measurements. To measure heart girth, the tape measure is wrapped firmly just behind an animal’s front legs and read between the shoulders. Each animal was measured twice per visit and the measurements averaged. At the first measurement of each animal, the age, sex and a detailed physical description was recorded to enable accurate follow up. Heart girth correlation with weight has been extensively validated in a wide variety of species, and a diversity of locations. It is typically used when scale availability is limited or is a low priority due to resource constraints.311-313 Analysis was carried out on the weights corresponding to girth measurements.  2.5 Data Analysis  Wealth and Wellbeing Frequency of household wealth markers with the addition of land ownership and HoH demographic characteristics, were tabulated and compared by sex using Pearson’s chi-square tests. Markers with little community variation (90% + uniformity), such as clothing, home furnishing, utilities, medical, and veterinary practices were excluded from further analysis. Centroid-based cluster analysis of household wealth markers was used to economically differentiate households. Final variables included, based on focus group discussion and standards from the literature were: marriage type,108 number of children,108 initial herd size, type of livestock owned,130, 229, 314 percentage change in herd size over the duration of    45  research, diversification130, 314 categorized into 5 livelihood groups (none, traditional with no livestock, traditional including livestock, some non-traditional, and primarily non-traditional income), approximate size of landholdings and proportion under cultivation,108 house type (traditional vs traditional and metal roof),315 vehicle314, 316 and cellphone ownership.317, 318   Tallies of SWLS response were grouped according to Diener interpretations.319 Because of the small number of dissatisfied (4 men and 5 women) and very satisfied respondents (2 men and 4 women), for analytical purposes “dissatisfied” subjects were added to the “slightly dissatisfied”, and the “very satisfied” to the “satisfied”. Qualitative contributors and detractors from wellbeing were grouped into related categories (for example answers related to livestock ownership, or business activity). Proportional differences between men and women in SWLS categories and choice of qualitative contributors and detractors, were assessed using Pearson’s chi-square tests. Variables used to build an explanatory model for total SWLS were selected based on wealth and wellbeing literature (pastoral/Maasai focused), Olkoroi focus groups, self-identified contributors and detractors to wellbeing, community observations and/or longitudinal health reports (alcohol consumption and physical abuse). Variables used were: sex,185, 239 age sete ranging from 1-7 (women are traditionally affiliated with their husband’s age group),185, 239 diversification (non-traditional livelihood practices vs traditional),108, 314 marriage type (monogamous or polygamous vs unmarried), marital status (married vs any form of single HoH status, including never married, separated, abandoned, divorced, or widowed),239 number of children,108, 314 alcohol consumption (none vs any), domestic abuse (none vs observed or documented in data collection), church attendance (any vs none),239 and log of tropical livestock units (TLU),108, 229 land ownership (any vs none),108, 314 and status as decision making HoH (yes or no).320  Correlation of the following variable pairs was anticipated and confirmed by chi-square analysis: sex and current marital status, sex and domestic abuse, sex and HoH, and HoH and control of livestock. In addition, church attendance was 100% correlated (100%)                                                  e Most adults did not know their age accurately but identified with a known age sets. Age-set 1 was the oldest.    46  with teetotalism (abstention was required in local Christian churches). Correlated variables were included in the full models because of the exploratory nature of the investigation. Analysis was carried out with SAS for Windows, 9.4. The SAS procedure GLM (generalized linear model) was used to explore the capacity of the variables to predict total wellbeing. All variables were evaluated first as univariate predictors then combined in a multivariate model without exclusion due to the exploratory nature of the investigation. Backwards elimination was used to identify the most parsimonious combination of variables to predict wellbeing, and results expressed as coefficients with 95% CI.    Livestock Duties and Relationship of Livestock to Wellbeing The proportions of men and women who currently or had previously engaged in each livestock task were summarized by task. Tests for differences in livestock task responsibilities by sex were carried out using Pearson’s chi-square analysis. First answers to queries about the positive and negative influence of livestock on wellbeing were summarized and overall differences between men and women for positive influences were tested using Pearson’s chi-square analysis. Sex differences for the proportion of men versus women who agreed with the statement, “Do livestock detract from wellbeing” were also tested using Pearson’s chi-square but due to the small number of respondents who agreed, sex differences in specific negative attributions were not tested.   Livestock Disease Prioritizations and Cultural Competency Prioritization data was interpreted with both quantitative and qualitative methodology. Quantitative ranking was determined by assigning each of the diseases a value corresponding to total frequency of identification. The most commonly mentioned disease was assigned a value of 20 (the total number of distinct diseases identified), the next most frequent, 19, down to the least mentioned which was valued as 1. The average of the respondent’s unused ranks was assigned to the remaining diseases of the 20 that were not selected by each interviewee. The sum of the values derived from each respondent for each disease (prioritization score) was used to order the overall ranking. As the number of    47  respondents who selected diseases ranked 11-20 were low (seven or fewer, and six diseases identified by only one person), only the top ten ranked diseases were reported on. Qualitative ranking was determined by weighted frequency which was subsequently used to calculate a livestock disease cultural competency (familiarity with a subject relative to other members of a group) rating for each respondent. Each disease was weighted by the proportion of respondents who had selected it. The total of weighted values for each individual was added to yield a domain familiarity or cultural competency.321 Male vs female cultural competency for livestock disease was tested using a 2 sided t-test.    Herding Duties: Self-Assessment and Best Practice Self-assessments of household herding tasks were tabulated by rating (1, 2, and 3, corresponding to excellent, average, and poor) and by sex. Chi-square analysis was used to compare the proportion of men vs women who selected each possible ranking for each task.   Herd size Models Explanatory herd size (log TLU) models used only the characteristics of the primary decision maker for each household, the male or defacto head in partnered households, or the de juref head of solo households. Variables included were sex (male vs female), age-set of male HoH (present or absent), diversification (non-traditional vs traditional), current relationship status (partnered vs non-partnered), number of children, alcohol consumption (none vs any), and church attendance (any vs no Christian church affiliation). Analysis was carried out using SAS for Windows, 9.4. A generalized linear model (proc GLM) was used to predict log herd size. All variables were initially evaluated by univariate analysis then combined in a multivariate model without exclusion. Backwards elimination was used to identify the most parsimonious combination of explanatory variables (p<0.05), the parameters and 95% CI associated with each significant variable. Parameters                                                  f   Defacto FHH are households in which the male partner is effectively absent although the marriage is intact, while de jure FHH are ones in which a MHH is absent due to divorce or death.    48  and confidence intervals were back converted into TLU units via an inverse log transformation.   Young Livestock Growth Rate Model Young animal growth rates per day were calculated for each young animal (n=501) using a linear regression model (SAS proc reg) and multiplied by 30 to generate a per month rate. Monthly rates were used as the dependent variable in a further regression to explore potential associations between growth rates and livestock, climate and owner-related variables. Before assembling the final model including climate, owner and livestock-related variables, a preliminary exploration of non-owner related variables was conducted using only climate (normal, drought or transition), animal type (sheep, goat or cow), animal sex, and age (in days). Animal sex was non-significant and was dropped from further modelling. Differences in sheep versus cattle growth rates were not significant, but all animal types were retained for the final model. The final model incorporated owner variables of sex, age, marital status, diversification, number of children, and church attendance, livestock variables of species, age, herd type, herd size (TLU), and the season variable. Preliminary growth rate calculations and final model building was conducted using SAS for Windows, 9.4. The proc genmod statement with a normal distribution and link=identity statement (traditional linear regression) was used to identify variables associated with growth rates. The genmod procedure was chosen to account for within herd correlations using GEE (Generalized Estimating Equations) methodology. All variables were first evaluated by univariate analysis then combined in a multivariate model. Backwards elimination was used to identify the most parsimonious combination of explanatory variables, and to determine the parameters and 95% CI associated with each.   Human and Livestock Infectious Disease Frequency Models Self-reported health data on both livestock and residents was collected twice a month, but for model-building, morbidities were averaged for the month. Models were constructed to explore possible associations of seasonal (normal/drought/transition), temporal (bimonthly periods), herd and household variables with total small ruminant and cattle disease burdens,    49  and the most commonly reported disease for each category of livestock (the locally identified small ruminant disease “olodua” and trypanosomiasis in cattle). A variable corresponding to the specific month (1-31 for the entire study duration) in which the data was collected was also included to ensure that family to family data were compared in appropriate time order but was not used as an explanatory variable. Livestock variables included herd type (goats and/or sheep, or cattle only vs both) and herd size (log TLU). Owner variables included: sex, age set of male HoH (or former male partner), diversification, current marital status, number of children, church attendance, herd size, and livestock movement (owner relocation of livestock during dry/drought periods). Associations with self-reported human infectious disease incidence were explored using the same temporal and household variables as for livestock disease, but with the addition of total small ruminant and cattle disease prevalence for each time interval. Because the first year of data had to be discarded (encompassing half of the “normal” climate data), the remaining data was grouped into drought and post-drought seasonal categories only.  Analysis was carried out using SAS for Windows, 9.4. A generalized linear model was used to explain the average monthly prevalence/incidence of disease (for livestock, the total number of diseased livestock for the month or “events”, divided by the product of herd size and number of reports i.e. “trials”, for human diseases, the total number of newly reported diseases, divided by the product of total adults/children and number of reports collected), using a binomial distribution with link=logit. The genmod procedure was used because of the need to use GEE methods to adjust for multiple levels of correlation in the data (which proc logistic cannot do). The “repeated subject” statement was used to account for family level correlations within herds/family members, and the “within” statement was used to account for correlation between the months (1-31). The estimate statement was used to convert coefficients and confidence intervals into odds ratios. All variables were initially evaluated by univariate analysis then combined in a multivariate model. Backwards elimination was used to identify the most parsimonious combination of explanatory variables, and the parameters and 95% CI associated with each variable.      50   Researcher Positionality, Research and Community Context 3.1 Introduction  Objectives The intent of this chapter is fourfold:  to establish my positionality as regards the community and its members  to discuss the relevance of a single community study to human and livestock health research in poor and under-served livestock dependent communities  to provide context for data and analytical models presented in subsequent chapters on livestock health, and human health and wellbeing, using descriptive epidemiology to characterize the environment, demographics, culture, customs, and capital that both supported and constrained the health, livelihoods, wealth and wellbeing of Olkoroi residents.   to demonstrate commonalities between Olkoroi and other rural Maasai communities, which have been reported upon in the human and livestock health and development literature focused on the Maasai.   “The Perspective and Position” of the Researcher  Qualitative research acknowledges “…analysis is inherently subjective because the researcher is the instrument for analysis…”322 This perspective also applies to quantitative work, as suggested by philosophers such as Hume, Schmacher, Duhem and Quine, who contributed to the development of scientific realism,323 underdetermination (uncertainty in epidemiology) and causal inference.324-327 Nonetheless, tools such as bracketing and reflexive thinking,328 used to facilitate honesty and transparency in decision making, are more commonly used in the qualitative arena. Consideration of positionality is increasingly viewed as necessary for epidemiologists whether quantitative or qualitative. It is particularly important given the problematic history of poverty oriented research into livestock,16, 44 livestock keepers,12, 74 and the Maasai specifically.329-331  My introduction to East Africa occurred via participation as a faculty member in the Langara College Canadian Field Studies in Africa (CFSIA) program for post-secondary students, over a cumulative four months, between 2001 and 2002.  While with the program, I    51  visited schools, development and women’s groups primarily in rural Kenya, but also Uganda. From 2002-2006, I was also a member of a CIDA-Association of Canadian Community Colleges (ACCC) development project in Mwanza, Tanzania. Although the work had no relation to Maasai concerns, I gained knowledge on Tanzanian history, government and health policies, which later facilitated comparison with Kenya. As part of the ACCC team I visited and interviewed community medical, legal, NGO and political representatives. In both projects I became simultaneously more interested in East Africa, and engaged with Maasai communities in particular, but increasingly uncomfortable with the dynamics of engagement between Canadian coordinators and students, East African partners and employees, and especially the disregard shown in the Tanzania project for CIDA protocols.  When I began my own research, I read qualitative texts (for example the Schensul and LeCompte series) during fieldwork preparation, kept research diaries, and wrote memos almost continuously while in Kenya, and during data analysis and writing in Canada. While motivated in part by recommended protocol, I also wrote almost instinctively to explore, alleviate, and manage tensions associated with individual, linguistic and cultural isolation in the field, as well as significant pressure I felt as a Canadian, educated, “enkashumpai” (female Caucasian) researcher working in a predominantly illiterate and poor community. Other aspects of my identity that affected my perspectives included single parenthood with full financial responsibility for my four children (who accompanied me for two of my three field seasons), and employment as a post-secondary educator. The latter were significant for several reasons. A high proportion of Olkoroi women were actual or de facto single parents who faced difficulty in providing for their children, and community access to education had been and continues to be limited. Village gossip described me as a millionaire. I regularly received requests for assistance and was sought after for answers to questions I could not always answer. Conversely, I was sometimes viewed as ignorant about livestock, generally treated patiently but sometimes laughed at when I made errors in custom and language, and resented on occasion when I could not meet the requests made of me. I was personally motivated to do this research for several reasons. My family history features marginalization, illiteracy and poverty. My experience with the failure of the CIDA project to follow mandated gender protocols and authentic participatory activities, made me    52  want to engage in a different kind of research. Most importantly, after more than a decade of working on small-scale community development in Maasailand (primarily Olkoroi), and seeing the importance of livestock first-hand, I wanted to better understand the relationships between livestock and human health/wellbeing. My preliminary readings identified many claims of positive association, but little concrete evidence.   The Importance of Single Community Studies Community studies are particularly relevant to the study of rural health and wellbeing because they identify the context of health behaviours and decisions, and extend the study of health to consider class, identity and other social determinants. Community studies expand the focus of health and wellbeing research to include the daily life of health consumers and the interacting issues facing many rural residents.332  …the priorities, capacities and needs of local [livestock-dependent] populations and health systems in resource-poor contexts are neglected and sidelined…questioning some of the normative One Health assumptions and rhetoric…demands contextualized knowledge only possible by detailed analysis of particular cases….264  Single community studies333, 334 can help to bridge knowledge gaps and identify key differences within populations that may be obscured in larger studies.332 They may be particularly useful for investigating complex health questions333, 335 such as those posed in One Health or Ecohealth research.g, 264 Single community investigations may also be useful in exploratory research where data is insufficient to confirm or reject claims, as is currently the case for many questions about the impact of livestock on the health and wellbeing of their keepers.29 Although a single community study may not provide definitive answers, nor be generalizable across other communities,337 it can help to guide future research.264, 334, 338 Local perspective obtained directly from residents, is also necessary to ensure that                                                  g One Health and Ecohealth are overlapping models that attempt to situate human health within larger, interacting systems. Ecohealth differs from One Health in that it attempts to understand the relationships between human health and ecosystems, albeit using a much wider definition of ecosystem than is in the ecological traditional, and explicitly includes the use participatory methodologies, interdisciplinary cooperation, and consideration of socioeconomic influences. 336. Bunch MJ, Waltner-Toews D. Grappling with Complexity: the Context for One Health and the Ecohealth Approach.  One Health: The Theory and Practice of Integrated Health Approaches2015. p. 415.    53  interventions by outside agencies meet community priorities.339 Lastly, single “case” studies (wherein a case can range from an individual to a nation), in the form of process tracing, have been increasingly recognised as an effective way to understand the results of development interventions. Specifically, process tracing can support external validation by identifying case characteristics associated with varied intervention outcomes.340, 341 To ensure accuracy and validity in participatory development research, it is vital to identify and include the marginalized and vulnerable, obtain perspectives from all genders, and respect and incorporate local culture, knowledge and understanding.342-345 Full community inclusion and ownership of interventions developed and implemented based on research findings are also critical for sustainability.344, 346, 347  However, such necessary inclusions may be hindered by biases associated with rural poverty/global health investigations. Common biases include gender,344 location (studies are likely to be conducted in more accessible communities, and regions),26, 44 and sampling bias (absence of “hidden” or marginalized members of communities).264, 348-354 In addition, it may be difficult to collect enough data or follow up long enough to fully capture the dynamics of values, priorities, relationships, activities and consequent livelihood choices made by residents, or long term intervention outcomes.16, 44, 342, 355, 356 In pastoral research specifically, there tends to be less data or follow up from more isolated rural communities, drought climate cycles,343 or highly mobile communities.26, 342 The challenges of rural poverty research may in part explain why the failure rate of livestock-focused development has been unacceptably high.16  Worldwide, the number of livestock-dependent households is large and their histories and political position are diverse.357 A further complication in understanding the relationship between livestock and human wellbeing/health is that customs, priorities, practices and outcomes differ between and within livestock-keeping cultures, and also in different circumstances.342, 358 For example, morbidity, mortality and productivity of livestock may vary depending on household SES, local infrastructure, geography, and climate cycles. No single research or intervention approach can possibly be effective for every group. Nonetheless, types of vulnerabilities, causal pathways, confounders and barriers to the use of livestock for wealth, health and wellbeing enhancement can be identified through representative community studies.  Although the complexity of livestock-human dynamics    54  likely preclude a single approach, cumulative knowledge from multiple community studies334 could provide a set of  templates with which to start research and evaluation, and ultimately interventions to support livestock-based livelihoods.264  Pastoral livestock keeping is somewhat unpredictable, and practice has evolved to manage uncertainty. 359 Ongoing, possibly more frequent droughts in ASAL regions, political shifts and processes such as land privatization127 and industrial development in pastoral territory, have only increased uncertainty.126, 343 It has been suggested that generalization may not be possible, in “….high uncertainty undertaking[s]”.341 Academic recognition of this difficulty has contributed to an increased interest in the application of realist study and evaluation in the health arena, and most recently in the development context. 341, 360, 361 Realistic evaluation evolved from scientific realism and acknowledges that interventions are implemented in social contexts. 362 Although the realist health literature focuses primarily on complex interventions, the principles also apply to research361 and the enormous potential variability in magnitude and range of health-related variables. These elements can include abiotic and biotic factors, as well as the responses, interactions, and behaviours of both those under observation and the observers.57 The realist approach accepts that any particular outcome is highly dependent on a host of interactions between variables that often change over time.363 Realist investigations also recognize that research, interventions and  evaluations often explicitly fail to follow up long enough to determine if collected information is representative of “real life” in the absence of outsider professionals.337, 341 Ultimately the realistic approach seeks to determine “…what works for whom, when, where and why,”332, 264 and community studies are an important part of answering this question.341   Olkoroi as a Representative Maasai Community Past and current literature108, 117, 123, 137, 139, 163, 229, 258, 330, 331, 364-374, the data obtained for this thesis, and the strong links between Olkoroi and settlements in Kenya and Tanzania (e.g. routine movement, exchange of material goods and livestock, and arranged marriages), suggest substantial commonalities between Olkoroi and other rural Maasai communities. While Kenya and Tanzania have very different geopolitical histories, original Maasai territory extended from Kenya deep into Tanzania.1 In both nations, the Maasai lost and    55  continue to lose extensive territory, and were marginalized by colonial and post-colonial regimes.1, 108, 127, 186, 330, 375, 376 Hodgson and others convincingly claim marginalization continues and may even be accelerating.126, 127, 377 The Maasai migrated throughout both countries, but their roots remain in rural areas where culture, lifestyle, and livelihood, share more similarities across borders than with urban centres.108, 377 In addition, despite repeated government interventions that have weakened traditional mechanisms of tribal cohesion, customary practices, for example male age sets (groups of age-affiliated males), and male and female circumcision, widely persist.186, 378 The Maasai have long diversified as necessary, but most continue to identify with pastoralism, and share a strong pride in tribal history, custom, and language.186, 377, 379 Many communities struggle with the same challenges: preserving heritage, livelihood, and values, while incorporating new practices, and balancing family, age and gender roles, against individual needs.380 Numerous studies by Maasai-focused academics such as Grandin, Coast, Hodgson, Galaty, Rigby, Holland, Waller, Talle, Spear, Spencer, and more recently Hughes, Homewood, Archambault, Mwangi, and Wangui, note extensive similarities between Maasai communities across East Africa. Pastoralism, by tradition a community-based livelihood, even in settled communities, values and promotes connections and tradition.    3.2 Setting: Physical Geography, Climate, Ecosystem  Olkoroi is bordered by the Sand River on the south and east, and forested hills on the west. The river ran dry during droughts, but impeded movement in the wet season. Livestock drowned in the river every rainy season and human deaths were occasionally reported upstream from of Olkoroi. The combination of savannah, forested hills, and natural water sources created prime habitat for disease vectors. Tsetse fly and ticks contributed significantly to livestock disease, and mosquitos caused a high incidence of malaria.     56  Figure 3-1 Olkoroi from the Western Hills    Olkoroi falls within the Greater Mara Ecosystem, which has a mean annual temperature of 16-18 degrees Celsius, highs around 25, and lows 8. Annual precipitation is normally between 840 and 1000 mm.381 In Kenya, including Olkoroi, long rains usually fall from March-June and short rains, November-December. Residents perceived that rainfall variability had increased causing more frequent drought, as has been reported widely in the academic literature.343, 382-385 Major droughts occurred in 2000, 2004, 2009, 2013, and 2016, and as an individual precipitator and amplifier of other routes of herd loss and resultant poverty, were a significant concern: ...if the serious drought occurs it may destroy our life. (SWLS/livestock duties interview: 37-year-old male, 2009)   Drought…ever caused great loss to my family. (SWLS/livestock duties interview: 65-year-old male, 2009)    57   There was a high prevalence of wild animals in the area, due to game reserve proximity and the nearby forested hills. Most residents did not hunt and pretended to an aversion, as per cultural narrative, for wild game, but ate it given opportunity. Wild herbivores such as savannah buffalo or waterbucks sometimes damaged crops, and occasionally destroyed entire plantings. They were also a potential source of livestock disease, for example, malignant catarrhal fever, or the zoonoses anthrax and brucellosis.386, 387 Large, aggressive herbivores such as buffalo and elephants could be a physical threat, as were carnivores, which regularly attacked livestock and, although legally protected, were sometimes killed to stop repeated predation. To a lesser extent, animals such as monkeys, snakes, and foxes, also caused problems, including crop damage, livestock injury, and transmission of livestock pathogens.  3.3 Geopolitical Structure  The politician immediately responsible to Olkoroi was the Naikarra subdivision assistant chief who visited monthly to address matters such as land demarcation, conflicts, or illegal activities such as charcoal making or home brewing. Like most Maasai, Olkoroi residents supported the Orange Democratic Movement (ODM) nationally.  3.4 Community Structure  Origin, Ethnicity, Community and Family Structure Olkoroi was a young community, first settled in 1979. Most families had arrived between 1980 and 1990, and all but three were Maasai. Two related families were Kalenjin, one Ogiek, and three wives were from the Kamba, Kikuyu and Samburu tribes. Section (independent, geographically delimited tribal sub-groups)364 lineage was Ilpurko except for a Tanzanian and Kenyan from Ilarusa and Ildamat sections respectively. Below the section level, the strongest affiliations were traditional clans (descendants of common ancestors) and age sets.365 Just as tribalism has been historically, nationally and detrimentally exploited,388 clan connections were divisive locally during elections, and in ongoing land adjudication. Traditionally, Maasai families lived in extended groupings centred on a male HoH and his wives. Women were married off in arranged marriages after circumcision to much    58  older (at least two age sets) men who had completed warrior service and graduated to junior elder status. Many rural Maasai still follow this pattern, although love and age peer marriages have become more common.389, 390 A recent study carried out in neighbouring Kajiado county, observed, counter to East and SSA trends, a decrease in average Maasai marital age, possibly due to increasing insecurity and poverty. Importantly, age of female circumcision had declined by three years over the past four decades,391 a trend which has potentially negative implications for age of marriage and female access to education.  The Maasai are customarily polygamous, but in Olkoroi, the practice may have been in decline. Of 150 adults, 60, from 28 families, had been or were in a polygamous marriage, similar to proportions reported by Coast in 2006,392 but only 20 intact families were polygamous. A possible decrease in polygamy could have been due to a strong local Christian influence and cultural shifts, but community focus groups on wealth identified polygamy as a sign of prosperity, as also reported by McCabe et. al.143 Only three of the poorest families were polygamous, and in all three either the husband or one or more of the wives had left, rendering it de facto single parent or monogamous. Some polygamous fathers had monogamous sons, but the converse was also seen, and it was not uncommon for men to be pressured by poor families to take young, additional wives from inside and/or outside the community. Despite traditional polygamy and a possible movement to increased monogamy, I was told that extra-marital relationships were historically and currently common, as reported by other researchers.390 Such liaisons frequently produced children (to whom mothers sometimes disclosed the identity of their biological fathers).  Traditionally, Maasai women marry into their husband’s family while sons stay, gradually take over family assets, and provide for aging parents. However, a traditional Maasai expression is “No son circumcised before his father” implying a son cannot make decisions for the extended family until his father dies. In the absence or death of a male HoH, wives maintained control of the family livestock until sons (or daughters, if there were no sons), married. Women who had left a marriage, were abandoned, or widowed, often but not always lived with extended family including parents, sons, or sons-in-law. Half of the Olkoroi households (38 of 75) present during this research lived in an extended arrangement (related but independent households who shared some herding duties, or independent and    59  dependent households living in close proximity) while 36 households were nuclear. Only one permanent resident lived alone: an older childless widower. Extended families took many forms. Several families hosted young, uneducated, unmarried sisters who helped adult sisters with household duties. Other forms included: single and/or married sons who shared family livestock duties but deferred decisions to their father; adult brothers and sometimes brothers-in-law (often including a widowed mother as these forms tended to evolve from sons living with independent parents); and widowed and dependent women, living with a married son and/or daughter(s)-in-law. There were also combinations of these patterns. The most common forms of extended settlements were independent or dependent widowed women living with adult sons or daughters-in-law (with children of their own), and adult sons, married or unmarried, living with their independent parents sharing herding responsibilities.  Community size varied minimally but regularly because of movement driven by two opposing reasons, as also described by Homewood and others.108, 124, 368, 393-395 Customary pastoralism requires livestock movement for trading or grazing, while “modern” traditions require movement of children to boarding schools, and adults for business, jobs in tourist camps, or government work in education, healthcare, reserves, law enforcement or the military. The largest movements observed in Olkoroi were in the drought period of 2009. Land subdivision beginning in 2010 and completed in early 2017, was ongoing throughout my research and write-up. The process had created conflict and excluded some long term residents. Communally held land is a key characteristic of traditional pastoralism, and within Maasai territory, borders historically existed primarily between sections. Negotiated inter-sectional resource use, especially during dry seasons and drought, however, was the norm,118 but in the 1960’s, the Kenyan government began mandated land privatization.117 The process is still ongoing and has been very delayed in pastoral regions. The Olkoroi community recognised that private land ownership would have a substantial impact on grazing practice, necessitate movement for most families, and potentially cause loss of community cohesion. There was also a strong awareness that privatization would increase difficulties in accessing sufficient seasonal and drought pasture, upon which traditional pastoralism depends. If land is divided then it really limit pastoralism. (23 year old male, 2014)    60  When the land is fully divided and every member issued with a title deed then nobody is allowed to graze on somebody else land/shamba. The land demarcation will affect the usual animal’s migration during drought season. The people would have to look for relatives and friends land in case of natural calamities occurred. We don't usually borrow permission to live where you want to move your animals because the land is communal. Tanzania is still communal therefore people can move there but you have to ask permission from village elders and relatives if you have any. When the land is demarcated no areas should be left for animal’s movement. (30 year old male, 2014)  However, as in other Maasai communities in both Kenya and Tanzania, Olkoroi residents had a long memory of historic land losses into the present. Although land privatization would not restore Maasai sovereignty over remaining territory,1, 330, 396 residents perceived that land title would reduce “land grabbing”8, 119, and was thus seen as a necessary evil. Any land which is not yet demarcated or registered as an adjudicated section may be grab by some corrupt leaders as the country experienced frequently in past as well to nowadays. (30 year old male, 2014)   Housing: Structure and Cultural Practice Almost all households lived in traditional Maasai houses built by the female HoH: low, wood-framed homes plastered with a mixture of mud and cow dung. Traditional homes required little or no cash outlay but deteriorated quickly and often leaked in the rainy season. Therefore, most women aspired to “modern” houses with stronger frames, plastered in more durable clay composites from the riverbanks, and roofed with iron sheets. Such houses typically had three-four rooms, each with a lockable door, did not leak, and were seen as a mark of achievement and wealth. Most households, 63/75, owned one house (or one per wife). 47/63 had a traditional home, while 16 had a “modern” house. Of the 12 families with two homes, most had a modern iron-sheet and a traditional house though two families had two modern houses and one had two traditional homes.    61  In traditional houses, families slept under the same roof until children were circumcised, when it became culturally inappropriate for them to sleep in the same house as opposite sex parents.163 Women stated this custom created female vulnerability, as girls could not sleep at home, but were not always provided secure alternate space. Mothers reported that girls were consequently more likely to be involved in coercive sexual activity or to begin sexual relations with peers or older men in whose homes they found shelter. In turn, this increased risk of early pregnancy. Along with other advantages, iron-sheet houses with separate rooms, each with their own lockable door, allowed older girls to remain at home. Men who built iron-sheet houses sometimes reserved them for their own use, but when women built such homes, typically with their personal, rather than family resources, they were shared with the entire household. Ironically, women who owned modern houses often continued to spend most of their time in a traditional home because they were easier to heat and repel mosquitoes. All new houses of any type were rapidly and heavily infested with biting insects including fleas, cockroaches and bedbugs. However, the low height, small windows and ventilation holes of traditional homes trapped heat and smoke from the fire more effectively, which deterred mosquitos, though not other pests.  A few wealthy families had solar panels and/or water-collection systems attached to their modern homes. Of the four families with solar panels, two HoH were employed by the local missionary and two had salaried jobs outside Olkoroi.    Community and Family Decision Making Major family and community decisions in Olkoroi were usually made by men. For example, women were excluded from land adjudication in contravention of Kenyan law, a commonly reported East African phenomenon.71, 397-402 A never-married woman ran the main village shop, served on the school committee, attended church regularly, had five children by local men and had lived in Olkoroi for more than 30 years. But, without a male HoH, she received no land, ostensibly because she was Kalenjin, although her Kalenjin uncle, also an Olkoroi resident, was allotted property, as were the male, Ogiek HoHs. When discussions were held about my research and potential livestock oriented interventions, a few female HoH attended but did not speak. The major exception was the school committee in which    62  both men and women participated. At the family level fathers tended to make education decisions for children although some families appeared to make joint schooling choices, also noted by Bachar in her exploration of Maasai attitudes towards female education in Kajiado, Kenya. Several women were also observed to circumvent male decisions by giving tacit or active support to daughters resistant to paternal discontinuation of their education. Much academic and popular literature depicts Maasai society as patriarchal, but some argue that female avenues of power were first eroded by colonization, and further via breakdown of tradition.393 Hodgson, a leader of academic opposition to the patriarchal narrative, asserts that historically each sex had major spheres of influence and women remain significant reservoirs and enactors of cultural and religious knowledge 330, 371, 376, 403. This appeared true in Olkoroi as women played major roles in celebrations such as marriage and circumcision and regularly led prayers at public events. A few families were supported by educated daughters, which seemed to be contributing to a perceptual shift in possible gender roles, and a move away from viewing education for boys as a better investment (since by tradition, women become part of their husband’s family but men remain with their birth family).404, 405 Archambault, who has carried out extensive ethnographic research on education and gender in Elangata Wuas and Enkop, Kenyan Maasai communities northeast of Olkoroi,70, 391, 393 has also documented increased support for female education.391 Nonetheless, there was little support for women who acted against male authority or community norms, or women subjected to violence from their husbands, other males, or, in a case of community exclusion under physical threat, women. In addition, there was no censure of male violence against women; a young, pregnant woman was beaten until she hemorrhaged to death, but there was no consequence for her husband from any realm of authority. Residents of both sexes believed male HoHs were entitled to physically discipline children and female family members, even their mothers, a widely held belief across Kenya406 and among the Maasai.407 When one husband badly beat his wife, in public, another routinely abusive man took him aside at market and said, “You must not hit your wife with a stick, only your hands” (32-year-old male, 2012). Although the Kenyan government has legislated equality and protection for women and children,391, 406 it was not enforced in Olkoroi. By tribal law, the only formal avenue for involvement with spousal conflict is    63  dispute resolution by parents of the couple. Since parents give and receive bride price, and technically have to be compensated in the case of marital breakdown, they are considered to have a stake in their children’s marriages.408  3.5 Household Vulnerability In Olkoroi 31/75 households were socioeconomically vulnerable, some in multiple dimensions. The three primary vulnerabilities were: single HoHs (24/75), livestock poverty (22/75 households had no livestock), routine neglect of female HoH and children (9/75) or violence against the female HoH (10/75). It was suggested by an informed witness, that there had been/were a few cases of violence or neglect by women towards men, but there were no self-reports of such. In combination, there were 34 distinct households with one or more significant vulnerabilities.    Single Heads of Household Solo HoH led 24 of 75 households, all but four female, and 15 (63%) with dependent children. Half of the single HoH families were relatively secureh (Table 3-1). Seven owned their own livestock (independent security), and six had free access to the livestock of extended family (dependent security). The four independently secure households with children all had livestock but were anomalous in form: an unmarried son who supported his mother and her two grandchildren; a single mother by choice who ran a successful business; a widowed woman who looked after two grandchildren and had retained her husband’s livestock; and a divorced woman with her own livestock (from earnings and gifts).                                                           h As per the Alkire (2003) definition of human security: protection of “survival, livelihood, and dignity…[and]…a minimal subset of human development and human rights.”    64  Table 3-1: Security of Olkoroi Single Heads of Household (2008-2010)  Single  HoH  Secure Insecure Total (n=24)  Independent Dependent   Dependent children 4  (16.7%) 5  (20.8%) 6  (25%) 15  (62.5%) No dependent children 3  (12.5%) 1  (4.2%) 5  (20.8%) 9  (37.5%)  7 (29.2%) 6 (25%) 11 (45.8%) 24  Abandonment (all but one household by the male HoH) and male death were the most common causes of solo households (Table 3-2). A few abandoned wives received some support from their husbands, but most abandoned families were neglected, or even abused (one wife was forbidden to educate her daughters, under threat of death). Five women had chosen to leave abusive (physically and/or psychologically) spouses. One man had been left by three wives because of his consistent neglect, and one was a widower. Three lone HoH were unmarried, one woman and two men.  Table 3-2: Reasons for Single Head of Household Status in Olkoroi, 2008-2010  HoH Sex Abandoned Widowed Left/Abuse Unmarried Total Female + dependent children) 5 4 3 1 13 (65.4%) Female (no dependent children) 2 3 2 0 7  (19.2%) Male (+ dependent children) 1 0 0 1 2  (7.7%) Male (no dependent children) 0 1 0 1 2  (7.7%) Total 8 (33%) 8 (33%) 5 (21%) 3 (13%) 24  A further eleven families were de facto single HoH because of frequent and/or prolonged absences/neglect by the male HoH. In these households, the husband controlled family assets, but the women had primary, if not complete responsibility, for the children.    65  My husband really hates me. He has abandoned me. (SWLS/livestock duties interview: 38-year-old woman, 2009)  You know that our husbands are dead [sic]. They don’t even look for food. (Conversation during health data collection: 30-year-old woman, 2009)  Seven of the eleven de facto single HoH families (eight women, as two were from a polygamous family unit who cooperated for survival), struggled to provide for their children due to lack of family livestock, or inability to access family resources.  Livestock Poverty Of 75 households, 22 had no livestock and a further 34 had less than four TLU per household member (insufficient to support the household by livestock alone).i Half of the households with no livestock had dependent children, most were single HoHs (17/22) of whom 14 were female. As tending livestock was the customary and most accessible livelihood, these households were amongst the poorest. None had secure alternate livelihoods. In combination with erosion of the cultural safety net,108 limited job opportunities, and high employment insecurity, they were extremely vulnerable. Those without extended family support were most at-risk. Households with less than 4 TLU per household member were generally better off than households without livestock, as livestock were marketable assets, but families with small numbers of animals were still vulnerable. Such households typically lose proportionally more animals in disease outbreaks, droughts, and/or when unexpected costs arise.17, 395, 410, 411 During and after data collection, more than 20% of Olkoroi households experienced major or full loss of animal holdings, due to one or more of: medical costs, drought and disease losses, failed marriages (requiring dowry return), abandonment, alcoholism, job loss, legal difficulties, and veterinary pharmaceutical errors. A few families were livestock poor because of past losses to raiding from which they had not                                                  i It has been suggested that 4 tropical livestock units (TLU) per household member is a minimum number required for a livelihood dependent on livestock alone. 409. Toth R. Traps and thresholds in pastoralist mobility. American Journal of Agricultural Economics. 2015;97(1):315.      66  recovered. Raiding between communities/tribes was a Maasai norm in the past, though long illegal, and still occurred albeit at a reduced rate. Low frequencies of livestock theft, mostly small stock, was also prevalent within the community. Occasionally households experienced substantial losses to predators. Since livestock ownership customarily started with gifts from family,376 family livestock poverty made it more difficult for new households to establish themselves and some families were livestock poor because of generational livestock poverty.  Women were more likely to be livestock-poor because of tribal rights of ownership. Traditionally, women received animals at life stages such as betrothal, marriage and childbirth.412 These customs were in decline,399, 412 and even so, women were expected to disperse their holdings to their sons at adulthood.412, 413 Single, female HoH with young children were also constrained in their capacity to tend livestock or recover and build up livestock assets if lost. Consequently, if such women had animals, they were often left with extended family. Other reasons for no/low livestock holdings included events out of the owner’s control, such as drought, disease, accidents, predators, and thefts/raiding. Four households suffered generational poverty, inheriting no livestock from their fathers.   Violence Spousal violence was common, and obvious when it occurred outside the home (not unusual, as houses were small). The size of Olkoroi, proximity of homes, and cultural acceptance of corporal “punishment” for women, meant community awareness of abusive HoH was high. My field assistant, Ole Koshal, was trusted by local women who frequently came to him for assistance with conflict resolution, and we received information on violence through multiple additional routes: incidentally; as a direct witness (Ole Koshal) or via direct observation of injuries (both of us); from reports by victims, witnesses and infrequently from the clinic nurse; and victims regularly disclosed current and past experiences in the longitudinal health and SWLS studies. Consequently, while data on experience of domestic violence was not sought out via specific questions in any one study, I was able to obtain a comprehensive perspective on frequency, degree and repercussions of abuse.  Between 2008 and 2010, one woman died, two suffered miscarriages, and another required extended clinic care for head trauma due to spousal violence. No specific inquiries    67  were made as to the triggers for violence, but the following “infractions” were reported by victims as causes for some of the worst events: funds used to feed or obtain medical care for children without permission (4); alleged infidelity (1); maternal support for daughters to continue schooling (2); insufficient attention to livestock (1); and drinking (1). Because of the frequency and normalization of violence, households were only categorized as having violence vulnerability if it was regular or extreme. By this standard, ten households, with 11 women, were at risk. In five households, violence was the only vulnerability. The woman who died came from a household that was not materially poor. Marriage. A terrible caning every time. (SWLS/livestock duties interview: 24-year-old wife, 2009)  There is no man I can say has never abused his wife. (Conversation: 28-year-old male, 2012)  In major conflict situations, women often left or were driven away to their birth family, but were frequently pressured to return to their spouses, especially if birth families were poor.    Alcohol and Vulnerability Alcohol and consumption were not hidden. Alcohol was made (although brewing was illegal) and purchased in the village, and I regularly witnessed public drunkenness and outcomes thereof. Again, there was widespread community knowledge of drinking habits.  Alcohol frequently appeared to be part of vulnerability. In seven of 22 no-livestock households, livestock poverty was the result of alcoholism in the current male HoH, or in previous generations. We observed mothers from four households (two overlapping with the aforementioned seven, and two with lone female HoH) who periodically drank to the point of self and/or family harm (for example, drinking-related physical conflicts, childcare consigned to other women while mothers drank, or food insecurity due to alcohol expenditures). In two of the intact households with violence, abuse often occurred after drinking. Two abandoned households were visited by estranged husbands roughly once a year and regularly subjected    68  to drunken violence prior to departure. In addition, in the three, secure, de facto single headed households, alcohol had contributed to male HoH absence.   The interaction of alcohol with vulnerability was complex. Alcoholism had created generational poverty in one case (the current household was teetotal but had no assets), in a few had exacerbated poverty, but in others alcohol dependency seemed to have followed livestock loss and consequent poverty, compounding the challenge of recovery. In addition, some vulnerabilities were intertwined; in the four households where violence was accompanied by other vulnerabilities, three were neglected and asset-poor, and in one household with two wives and large livestock holdings, everyone but the husband was routinely hungry.   3.6 Children: Family size, child “ownership” and roles of children The Maasai value children highly, and large households were rarely a concern except for anticipated education costs. Some polygamous families were very big (eight men had more than ten children), but the average number of children per HoH in 2010 was 4.8, a little higher than the national total fertility rate (TFR) which was 4.6 in 2008,414 and is currently just under 4.415 Because of polygamy, the average number of children per woman was lower than per man (4.4 vs 5.5). The range of family size for monogamous and polygamous women was between zero and eleven, the median for both was four, and means were very close (4.8 and 4.6 respectively). As monogamous men were the husbands of the aforementioned monogamous women, they too had a mean of 4.8 children. Family size for polygamous men, however, ranged from two to twenty, and the median and mean were ten and 10.3, respectively. Historically, reported fertility was lower in the Maasai compared to other tribes (about half peak historical national TFR), however there is little evidence to explain why.163, 331, 416 Most residents viewed large families as desirable, expressed pride about family size, or regret about low fertility, but a few HoHj sought to limit family size, especially younger or educated HoH as also found in work with urban Maasai in Ngong.316                                                  j A young wife asked me about birth control, as “Maasai women are like cows, we have a child each year whether we want to or not.” Another spoke spontaneously of using the withdrawal method to avoid pregnancy.    69  Pastoralism and rural lack of infrastructure necessitate substantial physical work. High fertility can be a rational response to labour demand, poverty and/or high child mortality rates.417 Although Olkoroi households were larger than the Kenyan average, higher rates of education had begun to create labour shortfalls in some homes, as referenced in other research.108, 418, 419 In Olkoroi, this shortage was managed by: delayed school entry; denial of schooling for some children; and approximately half of livestock-owning families used uneducated children from extended family or from poorer families in Tanzania, which alleviated cost for the birth family and provided free/low cost labour for the recipients. In a few households, parents and older siblings took full responsibility and/or hired adult herders.  Maasai children taken into or borne within marriage, belong to their mother’s husband.379 Most Olkoroi FHH who were not widowed had custody of their children but if a husband decided to keep his wife’s offspring after marital breakdown, he was supported by tribal law. An Olkoroi woman who lost her children to an abusive husband was denied help from the Narok Children’s Office because she could not pay for staff travel costs to her husband’s home: as was the norm for women who left a marriage, she had no assets or community support. Conversely, unmarried women had full responsibility for their children, even when the father was publicly known, also noted in Kajiado research.391 While a few unmarried mothers lived with their parents, in Olkoroi most were quickly married off (based on observations of numerous schoolgirl pregnancies), sometimes to men who were not the biological fathers: both outcomes have also been reported by Archambault.391 In the past, a pregnancy automatically ended female education, but in recent years, in Olkoroi, as elsewhere,393, 404 grandmothers sometimes provided childcare to allow daughters to resume education. Regardless, few young Olkoroi women completed schooling after childbearing.   3.7 Education  Education was a double-edged sword for many in Olkoroi. As reported in the pastoral literature, educated children had greater earning and diversification potential, but sometimes at the cost of traditional skills and cultural literacy.373, 393, 420 Time in school, for boarders especially, inevitably resulted in missed natural learning.421 Negative perceptions about the effects of education included, as in other research: rejection or loss of culture,422 immoral    70  behaviour,373 and potential absence of adult children from the family.70, 422 Education was viewed more positively than in the past and most families aspired to educating some children (as per SWLS/livestock duties interviews), but poverty and other barriers continued to limit access,391, 423 many families had denied education to older children (male and female), and a few households still denied education to all female children. A recent report noted that 46% of Kenyan children out of school lived in ASAL424 and in 2016, only 20% of Narok County secondary aged children were in school.425 Education has also created new cultural pitfalls for young Maasai. Locally, male secondary graduates had trouble finding jobs, but sometimes rejected a return to herding, becoming “lost” between old and new culture, as observed both by Archambault in Kenya,393 and Yao in Tanzania.422 Olkoroi children often started school late, also noted in Tanzanian work,426 and Olkoroi primary graduates could be in their late teens or older. Early marriage for girls remained common, so even if allowed school entry, they could be withdrawn before primary completion. Poor families could rarely afford high school, even if they supported education, and a daughter at home who could bring dowry, represented both economic burden and delayed financial opportunity:  …the main reason for sending my daughter for marriage is that I have no animals or resources to send her to high school…I love my daughter but unfortunately, I am poor and only the children are in my family, no cows. If she can get help [for school] ...I will not force her to get married. But if not then she should go to a husband's home. (50-year-old father whose daughter has requested assistance to attend secondary school, 2010)  Education could protect girls from early marriage, through entry into an “ageless” state, 372, 373 as per Switzer, but most parents followed traditional marital age trajectories. Another cultural conundrum stemmed from observed occasional resistance of educated girls to traditionally arranged marriages, and some community members perceived schooled women as potentially difficult spouses, also referenced by Bachar.373 Strong cultural aversion to birth control, reported by Coast in Tanzania,163 and mentioned in conversation by several HoHs, resulted in regular schoolgirl pregnancies, in both primary and secondary students. One author has suggested pre-marital pregnancies have become more common, and some Maasai parents perceive an association increased pregnancy risks in unmarried girls and female education, but no data exists to confirm or disprove these claims.391     71  Historical Maasai opposition to formal education was driven in part by fear of cultural losses but also associated with betrayal by colonial “allies”.1, 393, 427 Past Maasai “position” vis a vis education was further exacerbated by lack of rural education resources421 under colonial and post-colonial regimes, and may have been exaggerated to rationalize inadequate government support. Academic literature and personal discussions with older Maasai men in Olkoroi and other regions of Narok, iterated the theme of tribal educational intransigence, and referred to cultural marginalisation of early generations of educated boys (few girls attended in the past).393 Conversely, some research suggests the Maasai have long supported education.369, 391, 428, 429 A few Olkoroi fathers claimed that in the past, the least loved child was schooled to meet colonial demands of at least one educated child per family, while now they sent their "stupidest" children: the worst herders. The latter was a common narrative, and has been reported by others.393, 419, 430 Rural Maasai remain educationally under-served for many reasons, and stereotype persists as regards their “opposition” to education. In encounters with educators in Olkoroi, the District Education Office, and in high schools attended by Olkoroi students, education officials, even those who were Maasai, regularly disparaged the “ignorance” of community parents and the Maasai in general.   3.8 Adult Education Attainment Only 21.5% of adults had any education, mainly due to parental resistance for women, and traditional culture for men (Figure 3-2). In the past, schools were fewer, parental opposition stronger, and thus opportunity very limited.  ...parents took education as a lost way of life. Girls were not allowed at all to access in any way education. (SWLS/livestock duties interview: 58-year-old woman, 2009)           72  Figure 3-2: Reasons of Lack of/Interruption of Education in Olkoroi Adults (2008)    Of adults with education, roughly equal proportions of men and women had some schooling, but no adult women had completed secondary school (Table 3-3). A higher proportion of women than men had a full primary education, partly because the men who attended secondary had by default, completed primary. No Olkoroi HoH had attended post-secondary except two sponsored to theological college, the only option provided by the missionary, and only for men. The relative parity of adult education attainment was mostly due to the traditional two+ age set differences common between husband and wife. Married women, usually significantly younger than their spouses, were more likely to have had an opportunity to attend at least some school due to changing values and school availability, despite the ongoing lower likelihood of girls completing school and advancing to secondary.         0510152025303540Percentage of total adultsReasonsFemale Male   73  Table 3-3: Olkoroi Adult Educational Attainment by Sex and Level of Education (2008)  Education Level Sex None Some Prim. Full Prim. Some Sec. Full. Sec. Some Post-Sec.  Male 56 (78.9%) 8 (11.3%) 1 (1.4%) 1 (1.4%) 3 (4.3%) 2 (2.8%) 71 Female 85 (79.4%) 14 (13.1%) 5 (4.7%) 3 (2.8%) 0 (0.0%) 0 (0.0%) 107 Total 141 (79.2% 22 (12.4%) 6 (3.4% 4 (2.2%) 3 (1.7%) 2 (1.1%) 178   Current Attitudes and Barriers to Education  Although educationally better off than their parents, Olkoroi children still faced major obstacles due to poverty, custom, and gender. While lower grades sometimes had sex parity (Table 3-4), from 2006-2009 (there was no 2010 graduating class), the highest proportion of girls in grade eight was one third (Table 3-5). Most recently, in 2018, girls made up 15% of the grade eight class, indicating gender disparities continue. Family labour demands, gender roles, and poverty impeded both sexes, but gender disproportionately affected girls. They rarely spoke in class, or sought help from primarily male teachers (also noted by Bachar373), and did poorly on exams (Table 3-6). Archambault documented similar gendered behaviours and expectations, and suggested that formal education had been structured to reinforce traditional gender roles both nationally (based on textbooks) and in a Maasai specific sense in schools she observed, despite a shift away from pastoral livelihoods by educated Maasai.393  Table 3-4: Classroom sex ratios, Olkoroi Primary School, 2008   Kind. Gr. 1 Gr. 2 Gr. 3 Grade 4 Gr. 5 Gr. 6 Gr. 7 Gr. 8 Girls 17 10 3 8 7 6 6 3 4 Boys 28 13 12 15 12 7 10 8 8 Total 45 23 15 23 19 13 16 11 10  Some decline in progression through grade intervals was due to student transfer to better schools, but most missing children of both sexes, simply failed to complete. Furthermore, numbers alone do not fully capture the gendered nuances of educational    74  outcome. Of the four girls in 2008, the “best year”, 2 became pregnant and dropped out before finishing, one married after finishing, and one progressed to secondary school but also became a pregnancy drop out. Conversely, boys able to access secondary school were much more likely to finish.   Table 3-5: Number of Boys and Girls Reaching Grade 8 at Olkoroi Primary School, 2006-2009   2006 2007 2008 2009 Girls 3  0  4  1 Boys 9  7  8 7 Total 12 7 12 8  Competition in the Kenya Certificate of Primary Education (KCPE) national exam was very difficult for Olkoroi students. KCPE results determined secondary opportunities, but lack of resources, including teachers, resulted in high KCPE failure rates.   Table 3-6: Kenyan Certificate of Primary Education Resultsk for Olkoroi Primary School, 2006-2009    2006 2007 2008 2009  301-350 251-300 < 250 351-400 301-350 251-300 < 250 301-350 251-300 < 250 301-350 251-300 < 250 Girls 0 2 1 0 0 0 0 0 0 4 0 1 3 Boys 2 6 1 1 1 3 2 2 3 3 1 2 5  3.9 Community Resources  Education: Olkoroi Primary School and Secondary Opportunity  A 7-room primary school served Olkoroi and surrounding villages. There were no local secondary schools and the least expensive boarding high schools cost $750-$1000/year.                                                  k A pass (250/500) was viewed as a good result, marks over 300 excellent, and marks over 400, outstanding.    75  Better provincial or national schools, for which a few students qualified, were even more expensive, making secondary education beyond the reach of most families. Like many rural schools, Olkoroi Primary was understaffed. Government teachers were assigned yearly but were frequently absent and transferred quickly to positions closer to their families and urban centres. There was also a lack of books, equipment, classrooms, teacher offices and accommodation. Although free, Olkoroi Primary, like many schools, imposed fees to pay for extra teachers, non-teaching staff, equipment and/or field trips. Legally, children could not be denied schooling for financial reasons, but were routinely shamed if they did not have requested resources. In response, some children were kept home or resisted attendance.   Medical Olkoroi had a community-built (1994) two-room clinic supported by the Christian Missionary Fellowship (CMF)l. CMF had provided seed money for supplies and salaries, built living quarters for on-site Maasai nurses, and coordinated operations. Ongoing costs were covered by fees-for-service. The clinic was one of eight run in Maasailand by CMF and served Olkoroi and surroundings. Basic health care was provided, but resources were limited, there were no dental nor laboratory services, and inpatient care was only available in extreme circumstances. Because staff had limited training, they sometimes misdiagnosed or underestimated the severity of illnesses and injuries and could not treat serious or complex problems. Treatment for HIV and tuberculosis, and mosquito nets for pregnant women and mothers of children under five, were free, funded by the government. Clinic fees were not expensive compared to urban, private or government clinics, but poorer families still often choose to go untreated or used traditional remedies. In addition, most women had to seek permission and/or funds from their husbands or extended family to attend the clinic.   The clinic stocked contraceptives, but women did not generally access birth control. Privacy was not always maintained, and some residents believed the clinic secretly tested for HIV. Nonetheless, most sought clinic treatment for serious or prolonged illnesses. The                                                  l CMF is an evangelical, non-denominational missionary organisation that uses a variety of methodologies, including CHE, Community Health and Evangelism, to proselytize and create "dynamic Christ-centred communities that transform the world". (http://cmfi.org/ Accessed February 15th, 2014).    76  Maasai staff understood and respected certain cultural traditions, but they were literate, medically trained, mostly from larger communities, and CMF employees. Like teachers employed from outside, they occasionally exhibited condescending and/or judgemental attitudes. It was claimed, for example, “...alcoholism [in Olkoroi] leads to immoral behaviour. One cannot reason [with] or control this…. there is prostitution in Olkoroi- it is labelled friendship,”431  although there was no evidence to support these suggestions. The clinic had some drawbacks and CMF exaggerated available service, but was important because of the reasonable fees, service in Ma, and proximity.  Residents also patronized the CMF Ewaso clinic in Narok, and private clinics in Ololaimutia and Narok. Ewaso functioned similarly to the Olkoroi clinic but had more services. Private clinics cost more than public or missionary clinics, but had shorter waits, and often offered more sophisticated services (for example ultrasounds). They were used as an intermediate between CMF clinics and hospital services, most commonly for unusual, more serious, or persistent illnesses unresolved by the Olkoroi clinic.  The most commonly used hospitals were the closest: Narok District, Waso, Tenwek (in Bomet), and AIC Kijabe. Narok District is a secular public hospital. Service quality was relatively poor and wait times long, but fees were lower, care more advanced than in clinics, and vaccinations and some maternity services were free. Waso Hospital, just over the Tanzanian border, was commonly attended for two main reasons, lower cost and better service. Tenwek a missionary hospital, significantly more distant than Narok, was believed to provide better service and charge less. AIC Kijabe, another religious hospital, was perceived to have the best service, had regular visits from international clinicians, provided some free treatment, and had staff who spoke Ma. However, only an hour from Nairobi, it was too difficult and expensive to access for most Olkoroi families.  Kenyatta Hospital, in Nairobi, is the oldest, largest and premier Kenyan hospital, providing a wide variety of advanced services. However it sometimes has up to 300% bed occupancy.432 Despite, and in part because of its stature, Kenyatta was rarely used by Olkoroi residents. It was far away, costly, and staff did not speak Ma. The few residents who travelled to Kenyatta stood out in their tribal dress among the almost universal adoption of western attire in Nairobi, and the    77  combination of economic, social, language, and knowledge barriers made Kenyatta almost impossible to access without help.  Traditional healers practiced in Olkoroi and surrounding communities. Residents used them due to easier access, lower costs, or if they preferred customary treatment.   3.10 Religion  Christians have proselytized in Maasailand for over 100 years433 and a Ma speaking missionary has ministered in Olkoroi for more than 25 years. Historically, as with education, the Maasai significantly resisted Christianity,433 and in Olkoroi there was some antipathy to the missionary based on perceptions of favouritism. The national education curriculum included religious education with primary emphasis on Christianity, and Christian affiliation appeared to be increasing. There were 3 Christian churches despite the small size of the community. Only 20% of residents attended church regularly but there was widespread belief in the Maasai God, Ngai, and all meetings, celebrations, and important life events began with a prayer. All the local churches followed strict gender roles, and opportunities for education, travel, and positions of authority went almost exclusively to men. Such opportunities sometimes brought financial remuneration, and a few affiliates had benefited significantly.   3.11 Livelihoods  Generally, all family members, with the exception of the very young and very old, contributed to household livelihood, even if disabled. Even very young children helped: to collect wood and water, cook, tend to infants, herd, or work the shamba. Elderly women performed similar tasks and also milked livestock, mended utensils and assisted in home building or repair. By tradition, men reduced their responsibilities as their sons reached adulthood, but in poor families continued to work even when elderly.   Traditional Pastoral Livelihood  Most households owned livestock, even when they had other revenue sources. If TLUs were sufficient, small ruminants were occasionally slaughtered for consumption, but cattle were typically only killed for ceremonial occasions. Women used cattle dung for house    78  mortar and hides to make ceremonial garments and bed coverings. Other needs such as food, clothing, school, cropping and veterinary expenses, were paid via livestock sales, usually goats, unless a large sum of money was needed. Animals were also used socially as gifts, and to support the community and extended family.    Cropping and Land Ownership  Only men held land, and like livestock, it was controlled by fathers until old age or death when it was distributed to sons, but not always equally. Only one woman controlled land, as a widow with no biological children. The first families to settle Olkoroi (from the 1970s to 1982) generally owned the largest land areas (approximately 35% of households). Those without land usually fell into one of two major categories, women or recent incomers. Women without land had typically returned to Olkoroi after leaving their husband/being widowed/abandoned elsewhere but could sometimes use the land of extended family. Similarly, widows were usually permitted by sons to use family land. Some single female HoH were unable to grow crops despite having land access, because of financial constraints.  Most families (77%), grew food for home consumption, and just over a third of these also sold produce, mostly to supplement income a little. Six families sold the majority of their crops. Seventeen households grew no produce: seven of these families were vulnerable due to lack of assets, six were dependent on extended family for security with no assets of their own, and three did not crop because they were engaged in non-traditional livelihoods.    Small business  Many HoH of both sexes, had small businesses. Businesses varied by gender and rarely produced significant profit, especially women’s endeavors, but could supplement family income. A third of the men made traditional weapons for sale. Women did not make weapons, but a few assembled quivers. Six adults ran stores in the village centre, selling staples such as soap, oil, and vegetables. These stores mostly purchased inventory locally or sold family produce, and therefore had very small profit margins, would come and go depending on family circumstance, and were vulnerable to periodic thefts. Three women ran small restaurants. The woman who ran the most successful store in the village had a primary    79  education and excelled because she made an effort to buy a greater diversity of merchandise, in larger quantities, at lower prices. When she had the opportunity, she also sold lucrative items such as petrol. There was usually one agroveterinary store, owned and staffed by men, in the village centre, which sold common veterinary medicines. This store was important as it sold small quantities of drugs to those who could not afford to buy entire containers.   A few men had recently begun to provide transportation services, either by truck (only one man, on salary for the missionary, who sold most of his livestock to purchase the vehicle) or by motorbike. Six men owned motorbikes by the end of the research period, but none had driver training and rarely used a helmet. A number of residents, drivers and passengers alike, had motorbike accidents due to any or all of lack of training, poor weather, and road quality. One family lost most of their livestock assets due to the combined expense of first a motorbike purchase followed by the medical costs of an accident.  The most common small business for women was production of beaded jewelry. Few were able to make much money due to its almost universal practice, but there was some demand as all residents wore traditional beaded adornment, and it was frequently given as gifts. Women sold their jewelry at weekly markets but Olkoroi itself was not a major thoroughfare, so they had little opportunity to profit from tourists. Many women also ran tiny home stores, particularly those with young children, who were most restricted in time and flexibility. As with jewelry making, there was little profit to be made.   Nineteen Olkoroi women brewed alcohol. Brewing was a profitable and relatively easy business for women with few other options and seemed more common in FHH (12 of the 19). Of the women who brewed, ten were drinkers. Brewing was of interest, due to the characteristics of the families where it occurred, and because it appeared to be a female-only activity. Of the 19 brewers, 14 came from vulnerable family units or were dependent on extended family. Twelve were impoverished and vulnerable due to poverty or death of their husbands/abandonment, while two were vulnerable to violence. However, there was a group of six related women (by marriage or blood) who brewed, in which only one was poor.  Because of wide-spread illiteracy, diversification was predominantly restricted to small business ventures. However, lack of capital, particularly for women, restricted potential substantively. Women also had to have permission from their husbands to pursue any kind of    80  business initiative. Even men rarely had major assets so initiating a more ambitious effort (such as the aforementioned transport activities) required liquidation of the only asset many had, livestock, or relying on local lenders who charged anywhere from 30-200% interest. Olkoroi, like many rural communities, was small, so unless a business was widely patronized (such as the corn mill or market transport) most commercial activities had limited growth capacity. Even the mill, because of the high initial outlay, was started by the missionary, and employed only a few church supporters. The poorest families were the least able to run a business as they lacked the capital for even the smallest purchase or investment.   Casual labour  Approximately 35 of 150 male and female adults were observed performing casual labour over the course of the investigation. Such work could be seasonal (for example shamba work), one-off tasks for other residents, or work for the missionary family when they were in residence. Most short-term work was crop or construction-related. Casual work was typically done by the poorest HoH, very low paid, and short duration. Sometimes heavy drinkers were paid in home-brew. Housework for the missionary’s wife went only to women who attended church, although non-Christian men were occasionally hired for labour.   Salaried work A small proportion of villagers worked on salary, primarily men. Four men herded for other families but herding pay was very low and there was frequent turnover. Three men were on salary from the missionary, and two at tourist camps. Camp jobs were insecure and often terminated in low season or if tourism was depressed (for example during 2008 post-election violence). Six men had worked for periods (i.e. not simultaneously) as community teachers, three as the school guard, and three women as school cook. School employees often quit in exasperation, as salaries were erratically paid via fees levied on school parents. Most salaried work was paid approximately $100 per month which was a good salary by local standards, and comparable to the national average.434 Temporary salaried work was occasionally available through government initiatives such as voting, infrastructure installation, or public health interventions.     81   Professional Work  Low rates of literacy and opportunity in Olkoroi meant that few people worked in professional jobs. There were no fully qualified teachers, although a few residents had certificates in early childhood education. One individual had received nursing training through CMF, and eventually became a CMF administrator (but left the community to do so). Two male community members joined the army, and a third high school graduate was in the police force. The latter, however, was consistently posted away from home. Because of the size and resource limitations of Olkoroi, anyone who acquired professional training almost inevitably left the community in significant part because once hired, many public sector workers had to work wherever the government assigned them.  3.12 Conclusion: Constraints on Wealth, Wellbeing and Health  Residents of Olkoroi had the advantage of an atypical savannah location which permitted higher land productivity, but a variety of factors combined to make it difficult for community members to improve their SES, security, wellbeing and health. Some of the most critical barriers included: ongoing historical marginalization of the Maasai exacerbated by location; low literacy and limited access to education (any education in the past, and higher education currently), which not only restricted immediate opportunities for adults, but continued to constrain children and young adults, particularly those from the poorest families; failure of government infrastructure and laws to penetrate rural regions; and customary livelihoods and practice which, in combination with droughts, gender exclusion and lack of veterinary resources, provided limited capacity for advancement. Most of the challenges were magnified for women due to traditional gender roles and expectations. Some families were comfortably off, but seven of the 11 wealthiest and most secure families had achieved their positions through benefits received directly and indirectly from the local missionary, rather than traditional livelihood.    82   Material vs Subjective Wellbeing in a Traditional Maasai Community: A Gendered Perspective 4.1 Introduction The objectives of this chapter were threefold:   to describe community identified wealth markers, compare the sociodemographic characteristics and material assets ownership of Olkoroi HoH by sex, and to determine which HoH characteristics, including TLU owned, were most useful in socioeconomic categorization of households   to compare self-rated individual wellbeing (SWLS), as well as self-identified contributors to current and future wellbeing, by sex, and   to construct an explanatory model of psychological wellbeing that could help better tailor and target interventions to support livestock keepers.  Effective poverty alleviation and enhancement of health and wellbeing in pastoralists requires an understanding of gender roles. However, gender is still inadequately incorporated into livestock-related research, despite recognition of its importance and routine integration into crop-based investigations.74 Gender is important in the agricultural sector, and livestock-keeping specifically, not only because agriculture can be very effective in reducing the most extreme poverty,435 most global poor live in rural communities,291, 436 and women likely bear the higher poverty burden in most societies,437 but also because: women perform 43% of agricultural labour in poor countries, 50% in SSA,438 and may make up as much as 2/3 of poor livestock-keepers globally.439, 440 In addition, estimates have suggested amelioration of gender-inequitable resource access could increase female agricultural output by 20-30%.438 Maasai men and women, like other pastoralists,441-443 traditionally had overlapping roles in livestock keeping.371, 376 Evidence suggests greater  diversification has increased female workloads,70, 350 and sometimes women’s autonomy.399 While both MHH and FHH may experience poverty, Maasai tribal custom makes livestock acquisition and retention especially challenging for women. By tribal law they do not inherit livestock, if widowed they hold family animals only until sons reach adulthood, and they do not receive a share of household assets if a marriage breaks down.444 Furthermore, women often have difficulty    83  accessing resources and knowledge required to improve livestock productivity.350, 440, 445, 446 Consequently, if FHH with limited livelihood options do not receive support from adult children or extended family, they fall easily into persistent poverty. Research indicates that such vulnerabilities are common in pastoral societies and pastoral women have been described as doubly marginalized by both livelihood and gender.400, 445, 447  In Kenya, a proportion of pastoralists are wealthy, and have successfully adjusted to modern financial, social and political frameworks, but many are poor with few prospects for advancement. Pastoral inequality is growing internationally95, 108, 115, 221, 448-452, and even some pastoralists feel that traditional herding has become practically and economically less viable108, 115, 314, 316, 449, 452-454. Local and global socioeconomic changes, privatization and loss of land followed by shifts in herding practice, and greater external engagement via increased access to political, educational, and legal resources, are creating new opportunities for some pastoralists, but increased marginalization and exclusion for others.17, 66, 115, 314, 449 Women may be particularly vulnerable to loss of traditional social and cultural safeguards,455, 456 and yet are also using education212 and the powers of national law to assert their rights in contradiction of customary law.399 However, there have also been instances of the use of national law to deny pastoralists and especially women, customary shared resources.350  Many factors can contribute to vulnerability in the traditional Maasai, but based on observations in Olkoroi and a small amount of relevant research, women,371, 457 the elderly,350, 458 and single HoH121 tend to be least secure. The latter two groups are also most likely to be female due to polygamy, marital age gaps and conventions of marriage dissolution. The most vulnerable pastoralists typically have the least access to the few available services and supports and in crisis, may be the first to lose access to these resources.453 However, demographic information is sparse, and even DHS may not accurately represent Maasai households.458 There is a substantial amount of literature on pastoral and Maasai material wealth,108, 229 but there is little information about psychological wellbeing (one older, 2005, SWLS survey and a more recent, 2018 Tanzanian conceptual wellbeing study)185, 186. Most research focuses on household wealth (typically livestock and land) and families with male and female HoH present. Rural populations are generally poorly documented,315, 459 especially    84  pastoralists,61, 95, 152, 156, 168 and pastoral women least of all.61, 448, 460 There is little information, for example on pastoral FHH although they appear to be relatively common. The absence of necessary data to address pastoral challenges73, 461, 462 and support efforts to escape marginalization is ironic: “Pastoralists are one of the most researched, yet least understood groups in the world.”461 The Maasai are probably the most researched of the pastoralists73, 331, 463 yet major knowledge gaps remain, and key “facts” are often based on small bodies of data which may not be sound.331, 342, 458 Until recently, agricultural data from the developing world, particularly economic analyses, were often outdated, and like poverty data usually documented at the household level.438 Despite some improvement, it remains common for current publications to cite figures from pastoral research conducted a decade or more earlier.440 In addition, gendered aspects of agriculture were rarely adequately captured or even considered, and sometimes the contributions of women were excluded altogether.464, 465  …the gender and rural development literature demonstrates the existence of a broad, pervasive (if not universal), and enduring lack of women’s inclusion in agricultural decision-making in households at scales and settings from the household to agricultural development programs and projects.466   Gender mainstreaming has been an official development expectation since the late 1990’s,467 but is not always incorporated effectively, and frequently added as a project overlay rather than from inception.438 Some agricultural development projects have even exacerbated gender inequities.376, 468 A study of 35 years of gender mainstreaming in  agricultural aid projects targeting vulnerable populations, determined that only 5.1% of 5834 projects included women or gender in project titles and descriptions, and furthermore that proportional funding allocated to women had declined since initiation of mainstreaming.465 A 2019 gap analysis of pastoral sustainability found only 1.3% of 2658 publications on pastoralism included gender as a keyword61 and concluded: …little is known about pastoralist societies and the interlinkages between their practices and the rangelands on which these depend…  Pastoral women may bring very different knowledge, skills and priorities to agriculture, for example a focus on security of food over yield.445, 469 In addition, despite    85  unequal access to productive assets, pastoral women can still be found leading agricultural innovations.470 Gender is particularly important in livestock-keeping and understanding both the roots of, and routes away from poverty, because animals may be accessible to women when it is difficult for them to purchase other kinds of material investments, and/or assert claim to family resources.74 Although pastoral women usually own fewer livestock than men, their livestock frequently make up a greater proportion of their assets.445 Another challenge in development has been the visualization of wellbeing. Material poverty is an important influence on overall wellbeing, but other factors are also critical especially once basic needs are met.471, 472 The necessity of including psychological measures of wellbeing in development has been relatively recently recognized,332, 473, 474 but project incorporation has lagged behind,475 especially in very poor communities, where basic survival concerns may seem most immediate. Furthermore, while there are clear connections between material wealth and base life satisfaction, when household wealth is controlled primarily by male HoH, family income may not adequately capture or predict the individual wellbeing of members unable to access resources due to position, gender, or both.74, 314, 452, 465, 476, 477 Those who have little access to family assets, are also less able to obtain collateral-dependent services such as credit and insurance (which are often already difficult for pastoralists to obtain314, 453) to facilitate increased productivity and wealth.466, 476, 478-480 Women typically allocate spending differently than men, investing more in the household and children. As a consequence, it has been proposed that increased female income benefits children more than increased family income.455, 476, 481, 482 Although it is widely held that single-parent are poorer than dual-parent households, and FHH are both the poorest and most common form of low income family,437, 483-485 the empirical evidence supporting these claims is neither strong, nor consistent inter-nation, though there are countries where FHH are poorer on average.484, 486 Furthermore, common negative outcomes for FHH, are frequently driven by non-monetary factors such as: gender roles, constraints, and discrimination;437, 466, 484 family structure,483, 487 though family processes (interactions) are thought to be more important;488 specific structure of FHH;483, 487 and/or are common in all wealth strata rather than concentrated at the lowest levels.437    86   There is little East African research on FHH, or on monogamous versus polygamous families.476 Available information indicated major variation in socioeconomic status (SES) and household structure associations across Africa. South African (SA) and SSA data suggested that FHH could be better off, even with reduced financial security, due to conflict reduction, and greater investment in children.484, 489-491 In West Africa, the poorest families were polygamous, but in East Africa and SA, they were FHH,492 though a Ugandan study concluded the opposite.487 A Zimbabwean study of the Shona found differences in FHH versus MHH, but noted that conclusions varied depending on whether comparisons were made of income versus assets, and types of FHH, de facto or de jure.493 A review of data from 10 countries in the Global South, found that FHH were disproportionately asset poor in Ghana and Bangladesh, slightly more likely to be poor in general, but in real numbers most poor women lived in MHH.486 Income allocation research and a small body of related findings on risk behavior and credit access, suggest that agricultural development targeted at the family level which does not consider gendered household resource investment may be inefficient.466, 476 Lending support to these conclusions are studies which suggest families are more productive with the same resources if women have better access to available assets.476  No literature describing Maasai single HoH was found but a study on the related Samburu noted the frequency of FHH with children was increasing, such households usually lived with parents or brothers, and were denied asset inheritance.452 Globally, the poorest FHH often live with extended family and may therefore be “hidden” from poverty assessment.458, 485 In Kenya, despite national land laws protecting gender equity, women tend to be excluded from land ownership.494 From 2013 and 2017 women received only 10% of registered land titles, and just 1.6% of actual land.495 Most traditionally married Maasai women cannot use family assets independently nor participate in decision making on land and livestock.314, 452, 496  In the absence of data on FHH in traditional Maasai communities, I suggest that impoverished FHH’s may be increasing because of two opposing forces: the custom of large age gaps at marriage and polygamy, and rejection of tradition by both younger men and women. The lack and/or inaccessibility of government services which might protect vulnerable FHH 453 likely exacerbates these drivers. Other contributing factors could include:    87  strongly gendered livelihoods, exclusion of women from land and livestock ownership, ongoing changes  in asset usage resulting from monetarized economies,371, 376 limited avenues of opportunity (such as access to credit and language barriers476), and the forces which cause women to become solo heads (spousal abuse, conflict, neglect and abandonment) which are also likely to magnify the other factors. As in the Samburu example, even when rural women are supported by extended family and avoid impoverishment, their opportunity to establish financial independence or develop an asset base is limited.   4.2 Results  Wealth Marker Focus Groups Male and female characterizations for each wealth marker were very similar, therefore information from both groups was combined and summarized. Only characteristics which exhibited distinguishable variation were included in data summaries, and cluster analysis. Health was not included as it is generally considered to be an outcome of sociodemographic factors. Variable wealth markers as described and categorized by the focus groups are summarized in Table 4-1.   Table 4-1: Community Focus Groups’ Delineation of Wealth Categories by Marker Marker Poor Coping/Moderate Wealthy Livestock holdings 0-5 total, may not have cattle 6-20 livestock total 100’s of livestockm Source of income Traditional livelihood Similar to poor but training may expand potential. Livestock and/or education. Live-stock are capital. Food security May go hungry. Little or no milk and/or meat.  More milk and food. Do not go hungry. As much meat and milk as is desired.  House type Traditional house  Metal roof, mud floor. Cement walls and metal roof.                                                  m Focus group descriptions left a large gap between the high end for moderate wealth and low end for wealthy (also noted by Yanda et al., 2010). One absentee owner owned ~1000 animals, however, a man in a nearby community had 6 wives and 1000’s of animals and may have been the model used by interviewees.     88  Marker Poor Coping/Moderate Wealthy Vehicle ownership None Bicycle More paid rides (no cars/trucks). Polygamy/Number of Wives Poverty prevents polygamy (0-1) 1-3 wives (monogamous by choice) 6+ wives  Education of children None, or primary school only. Can choose to send children to high school. Can send children to post-secondary.   Sociodemographic Survey Table 4-2 contains the results of the sociodemographic survey, tested by Chi-square analysis for sex differences. Men and women differed significantly for: marriage type, current marital status, domestic abuse experience, livelihood diversification, and cellphone ownership. Differences approached significance for alcohol consumption. Half of married adults were in/had been in a monogamous marriage, but polygamy resulted in a higher proportion of monogamous males relative to females. Almost two-thirds of women suffered/had suffered some marital abuse. Due to traditional age gaps at marriage, older community members were almost twice as likely to be male as female. Just over a fifth of the community received some type of support (employment, subsidized training, or direct assistance) from the local missionary.   Table 4-2: Baseline Community Characteristics of Study Population: Adult Members of Olkoroi Village, Narok District South, Kenya, 2008 (n=150)   Individual Socio-demographic Characteristics Male n=59 (39.33%) Female n=91 (61.66%) Total n=150 p value (Pearson’s Chi- square test) for Sex Difference Any Education 20  (34%) 20  (22%) 40 (27%) 0.107 Marriage type, ever (monogamous) 34  (66%) 47  (52%) 77 (51%) 0.002 Current status (married) 48  (81%) 60  (66%) 128 (85%) 0.040 Age (older than 40) 31  (53%) 36  (24%) 67 (45%) 0.118 Alcohol consumption (none) 36  (61%) 68  (75%) 104 (69%) 0.075    89  Individual Socio-demographic Characteristics Male n=59 (39.33%) Female n=91 (61.66%) Total n=150 p value (Pearson’s Chi- square test) for Sex Difference Domestic abuse experienced (none) 59  (100%) 33  (36%) 92 (61%) <0.0001 Family size (>4 children) 31  (53%) 44  (48%) 75 (50%) 0.616 High-input Livelihood Diversification 23  (39%) 8  (9%) 31 (21%) <0.0001 Cellphone Ownership 20  (34%) 14  (15%) 34 (23%) 0.0082 Church Attendance 9  (15%) 19  (21%) 28 (19%) 0.388 Missionary Assistance 11  (19%) 22  (24%) 33 (22%) 0.424  Table 4-3 summarizes household assets, and education status of community children. Chi-square tests evaluated potential differences associated with sex of the primary HoH.  Table 4-3: Household assets, and education status of community children in Olkoroi (2008)   Male-headed households (MHH) versus FHH differed significantly in major asset ownership, with the exception of house type. Although the proportion of MHH who owed HoH Material Assets Male HoH n=47 Female HoH n=28 Total n=75  p value  Livestock owned at start of research (yes) 42  (56%) 11  (14.7%) 53  (70.7%) <0.0001 TLU mean (standard deviation, s.d.) 19.5 (18.4) 16.7 (10.9) 18.9 (17.2)  Vehicle ownership (yes) 14  (30%) 1  (4%) 15  (20%) 0.006 Land owned (yes) 43  (91%) 13  (46%) 56  (75%) <0.0001 House type (Traditional and metal roof) 15  (32%) 8  (29%) 23  (31%) 0.7613 Mean proportion of children with some schooling 78.3% 77.4% 77.9%  Children with household funded secondary education (of 326 potential students) 8 (2.5%) 9 (2.8%) 17 (5.2%)     90  livestock was more than twice that of FHH, average TLU was only slightly higher for MHH. Overall community variation in livestock holdings was large, with a range from 3.1-78.4 units. Few vehicles were owned, mostly bicycles (12), and all but one owned by men as women were culturally restricted from vehicle use except as passengers. Motorized vehicles were rare (at the beginning of the research, only three men who worked for the missionary had motorbikes). Customarily, only men owned land and female ownership was primarily de facto (albeit still beneficial as it allowed supplementation of family diet and also increased income potential). Very few families independently sent their children for secondary education. Of the children who had/were attending secondary school, seven (of 17) were from the three households where the male HoH held a salaried positions with the missionary.  Four TLU per household member is thought to be the minimum required to avoid persistent poverty.409 A breakdown of livestock ownership by household structure, Table 4-4, shows only 25.3% of Olkoroi families, 16 MHH and three FHH, had enough livestock to survive by traditional livestock rearing alone, and almost a third of the households had none.  Table 4-4: Livestock Owned Relative to Poverty Threshold, Sex of Head of Household and Family Structure in Olkoroi  Livestock ownership Monogamous Polygamous Single Heads of Household Totals   Female Headed Households Male Headed Households  No livestock 3 1 15 3 22 (29.3%) Under threshold  21 11 2 0 34 (45.3%) Above threshold 10 6 3 0 19 (25.3%) Total 34 18 20 3 75   Cluster Analysis of Wealth Markers Cluster analysis of wealth markers produced two major household groups based on livestock ownership (not quantity) only. One cluster consisted of 19 families (25%) without livestock, another of 53 (71%) households which owned some livestock, and there were three unique, single household clusters (4%). Each of the latter had a HoH with an unusual    91  combinations of wealth characteristics but no livestock: a separated, salaried, female teacher with children and no assets; an unmarried, male, secondary graduate with a salaried job at the local corn mill; and a FHH who lived independently of her polygamous husband but had not completely severed ties (with a metal-roof house, a small shamba, and educated children, she lived a comfortable, relatively autonomous life).    Psychological Wellbeing Interviews The range of total wellbeing was 10-31, with a combined mean of 22.7 (the maximum score on the SWLW is 35) and an SD of 4.4. Calculated on a per question basis (dividing the total by 5), the mean was 4.5 per question (maximum self-rating for each question is seven), halfway between neutral and slightly agree, and a SD. of 0.88 per question. Mean wellbeing was 22.3 for women and 22.8 for men, and a similar proportion of men and women self-scored as dissatisfied/slightly dissatisfied. The differences in the proportions of men and women who felt an average level of life satisfaction versus satisfied/very satisfied, however, resulted in a statistically significant chi square test (p= 0.019) for sex differences (Table 4-5). When couples of discordant satisfactions were compared, women were the less satisfied in 26 of the 33 discordant marriage (27 marriages were satisfaction concordant).  Table 4-5: Relative Proportions of Men vs Women by Specific Wellbeing Category in Olkoroi (2009)n  Sex Dissatisfied/Slightly Dissatisfied Average Satisfied/V.  Satisfied Total Men 12 (20.3%) 15 (25.4%) 32 (54.2%) 59 Women 19 (20.9%) 42 (46.2%) 30 (33.0%) 91 Total 31 (20.7%) 57 (38.0%) 62 (41.3%) 150                                                  n In Diener’s Understanding Scores on the Satisfaction with Life Scale, a total score of: 30-35 indicates high life satisfaction whereby life is not perfect but as good as can realistically be expected; scores of 25-29 represent people who are satisfied even if some spheres of their lives are problematic; 20-24 suggests average satisfaction but may indicates a desire for improvement; scores of 15-19 represent slightly below average satisfaction which can represent either several areas of small but significant dissatisfaction, or one area of major dissatisfaction; and people who score below 15 are substantial dissatisfied, with multiple dimensions of life functioning poorly, or a few going very badly. As only a small number of respondents scored in the dissatisfied (9) and very satisfied range (6), the former was included with the “slightly dissatisfied”, and the latter with the “satisfied”.    92   Contributors and Detractors to Wellbeing: Men vs Women Men and women were statistically different (p < 0.05) in their attribution of most important current contributor to life satisfaction (Figure 4-1). The biggest differences were children (the choice of 38.5% of women but only 13.6% of men) and marriage/family (13.6% of women, and 35.6% of men). Men gave “other” responses (fewer than five respondents) at a higher rate (24%) than women (13.2%). The importance of this difference was unclear due to the variety and small overall contribution of each of the responses within “other”, which included business, education, livestock, myself, nothing, and respect. “Myself” was the most common “other” response (9 respondents). One woman chose respect as the most important contributor to her wellbeing. Her choice became a latterly important representation of the most extreme gender effects on wellbeing when she died after a beating from her husband.   Figure 4-1: Most Important Contributor to Current Wellbeing in Olkoroi (2009): Women vs Men (n=150, p=0.025)    Men and women also chose different detractors from current wellbeing (Figure 4-2). They felt similarly about lack of education (19.8% of women, 20.3% of men), but more men identified lack of opportunity or resources (13.2% of women and 30.5% of men) and lack/loss of livestock (12.1 % of women versus 20.3% of men). The most striking difference was in the proportions who felt negatively about marriage. One man said marriage reduced 051015202530354045Children Church/God Life/Health Marriage/Family OtherPercentages (by Sex)Self-Identified CategoriesWomen Men   93  his wellbeing, in contrast to 25.3% of women. A quarter of both sexes felt there were no detractors in their life. Responses were more definite about wellbeing detractors than contributors as “other” answers made up only 4.7% of the total, and were mainly women. “Other” responses included health, thieves, children, don’t know, and partner’s death.   Figure 4-2: Most Important Detractors from Current Wellbeing in Olkoroi (2009): Women vs Men (n=150, p=0.0036)     Male and female perspectives on future contributors to wellbeing did not differ significantly (Figure 4-3), although more men had hope for livelihood contributors such as business, salaried jobs or livestock (47.5% of men versus 29.7% of women). The most frequent response, children, was similar in in men (35.6%) and women (37.4%), as were feelings about education of children. Many respondents spontaneously elaborated that the potential of children to contribute to future parental wellbeing would depend on education attainment. Again, more female responses fell into the “other” category, 22.0%, versus 13.6% of men. “Other” response included myself/hard work, church, independence, opportunity, and health/long life.  010203040Lack ofEducationLack ofOpportunityLack/Loss ofLivestockMarriage Nothing OtherPercentages (by sex)Self-Identified CategoriesWomen Men   94  Figure 4-3: Most Important Future Contributors to Wellbeing in Olkoroi (2009): Women vs Men (n=150, p=0.18)     Men and women were again different in choice of future wellbeing detractors (Figure 4-4). Men were much more concerned about loss of family assets/poverty (55.9% versus 24.2%) while women were worried about marital or family conflict harming future satisfaction (20.9% of women but no men). Women were less certain about what might disrupt their future, with 14.3% saying it was impossible to know, but only 6.8% of men. “Other” responses (14 women and seven men) included 9 responses of “nothing” (the most common), leaving church, unemployment, addiction, children’s costs, and vulnerability.               0510152025303540Business Children,EducatingChildrenJob Livestock I don'tknow/OnlyGod knowsOtherPercentages (by sex)Self-identified CategoriesWomen Men   95  Figure 4-4: Most Important Future Detractor from Wellbeing in Olkoroi (2009): Women vs Men (n=150, p=0.005)     Exploratory Wellbeing Model In univariate analysis (Table 4-6) higher life satisfaction was associated with being: a non-drinker, a churchgoer, a landowner, married, and having more children and livestock. Position as primary household decision maker was close to significant (p=0.0599). In the full multivariate model, only increased livestock holdings and having more children predicted higher life satisfaction. Variables associated with higher life satisfaction in the final adjusted model were being a non-drinker, the household decision maker, and having more children and livestock.        0102030405060Failure ofChildrenPersonalFailureCannotknow, OnlyGod knowsMaritalconflictNothing,OtherSickness,Aging,DeathPercentages (by sex)Self-Identified CategoriesWomen Men   96  Table 4-6: Unadjusted and Adjusted Linear Regression Coefficients and 95% Confidence Intervals of Variables Associated with Individual Life Satisfaction (n=150)  Characteristic   Unadjusted Model [95% C.I.] Adjusted Full Model [95% CI] Adjusted  Final Model [95% CI] Age set [1=oldest] 0.049 [-0.38, 0.47] 0.12 [-0.46, 0.70]  Alcohol consumption [No] 1.60 [0.078, 3.12] 1.03 [-0.57, 2.62] 1.50 [0.14, 2.86] Church Attendance [Yes] 3.10 [1.35, 4.86] 0.45 [-1.49, 2.40]  Diversification  [Non-trad.] 2.02 [0.29, 3.75] 0.39 [-1.43, 2.21]  Current status [Married] 2.67 [1.22, 5.87] 0.25 [-1.85, 2.35]  HoH Decision maker [Yes] 1.26 [-0.14, 2.67] 1.47 [-.062, 3.00] 1.61 [0.39, 2.83] Education [Any] -0.039 [-1.65, 1.57] -0.38 [-2.05, 1.28]  Land ownership  [Yes] 2.95 [0.91, 5.00] 0.38 [-1.65, 2.41]  Marriage type    [Monogamous] [Polygamous] 2.47 [-0.64, 5.46] 1.99 [-1.11, 5.10] 0.83 [-2.98, 4.61] -0.25 [-4.29, 3.79]  Number of children 0.37 [0.18, 0.57] 0.32 [0.084, 0.55] 0.28 [0.10, 0.47] Log10 TLU owned 3.23 [2.29, 4.18] 2.91 [1.71, 4.11] 2.95 [2.00, 3.91] Physical abuse  [None] 1.30 [-0.13, 2.74] 1.01 [-0.48, 2.68]  Sex [Male] 0.88 [-0.57, 2.33] -1.25 [-3.12, 0.63]   4.3 Discussion  Economic Differentiation of Olkoroi Families In a Maasai society, if you have a lot of livestock, they will consider you as a rich man (SWLS/livestock duties interview: 45 year old, monogamous man).  Two distinct household groups could be characterized by livestock ownership alone (any versus none). Inclusion of a variety of other wealth measures had little effect on cluster formation. This outcome was likely due to a number of factors. Most important was the    97  traditional livelihood and culture of Olkoroi residents, in which livestock remained central. Although some customs were changing (most children at least started primary school, for example) traditional Maasai values were the norm regardless of parental education, livelihood diversification or livestock holdings. Relatively uniform lifestyle and livelihood meant non-livestock wealth markers were either too uncommon or too widespread to be useful in differentiating wealth. Nonetheless, non-livestock markers identified by focus groups and later, in wellbeing discussions, provided insight into community aspirations.  Also relevant was Olkoroi’s rural location which made resources, both human and material, difficult to access, and also, through relative isolation, supported maintenance of tradition and tribal law. Lastly, there was little local opportunity for non-traditional work except for a small number of salaried positions with the missionary (three men), other religious organisations, or at Olkoroi Primary. On occasion there was also temporary tourism, research, or government work. Most of these opportunities were exclusively for men except for low paid housework, available when the missionary was in residence. Households with no livestock included some of the poorest Olkoroi residents. Only two HoH from this group had salaried employment, a FHH who worked as a kindergarten teacher (paid by the district council) and an unmarried man who ran the corn mill (hired by the missionary). The most accessible longer-term work for this group, albeit only for males, was herding, which, though poorly paid, was traditionally part of the pastoral safety net. Herders are occasionally rewarded with gifted livestock potentially providing an opportunity to build animal holdings and move up in socioeconomic standing. 314 Other types of work opportunities for those without livestock were also poorly paid, typically short-term and intermittent, but did not have the potential benefit of “earning” animals. Since families without livestock often lived hand to mouth, any earnings tended to be used for immediate survival needs such as food. In one FHH, two young sons were denied (free) elementary education and sent to work as salaried herders outside Olkoroi. Many families delayed schooling so children of both sexes could herd for their family, or educationally excluded at least one child completely. Child income was unlikely to shift individual or family status as remuneration was lower than that paid to adults, passed directly to the parent, and resulted in    98  no asset accrual for the young workers. A small observed subset of temporary employees were alcoholics who sometimes worked to earn money for, or were even paid in alcohol.   The poorest households frequently depended on extended family and/or sometimes other community members for assistance. This group though, did include a number of widowed women and daughters (7 of 22) who lead their own households, did not own any livestock, but were free to take milk as needed, from the herds of sons, sons-in-law, or parents. While these women and any dependent children were not destitute or chronically hungry, simultaneously they had few assets to facilitate independence or improve their circumstances, and frequently faced difficulty in obtaining funds for anything beyond basic needs of food and shelter (for example school fees). Loss of access to the resources of their family, would have left them in difficult circumstances. Other types of family or community support included space upon which to build a home, occasional food assistance, child care, or monetary aid (for example to pay school fees). Of families with livestock, only 35.8% of owners, or 25.3% of all households were above the threshold required to maintain herds through drought and disease outbreaks, and support the household via livestock alone.409 As a consequence, most livestock owners needed to earn additional income.  Differentiated wealth markers in Olkoroi, were very similar to those recorded in participatory wealth ranking conducted in three mixed pastoral/agricultural and Maasai communities in Tanzania,314 and Kenyan research done in predominantly Maasai Kisaju, in Kajiado County, albeit much more urban and diversified than Olkoroi in  gender roles and livelihood options.316 Specific commonalities included the Kisaju choice of livestock ownership, number of wives, food security, business activity, and vehicle ownership, as important wealth markers, although the specific quantities which corresponded to differing levels of wealth (three categories, as used in Olkoroi) varied between the communities. Similar to Olkoroi perceptions, two of the Kisaju respondents stated: A rich man is one with money, land and livestock, family and many children. He has built plots of land and owns a vehicle (RM 8, senior moran).  A rich man is one with large herds of livestock and a large family, for example one with five wives and a car (RM 18, a 20 year old cattle trader and with no education).     99  Younger, more educated Kisaju respondents espoused similar criteria, but, unlike older participants and most Olkoroi respondents, were more likely to prioritize small families and monogamous relationships.316 In Olkoroi, monogamy was promoted by Christian churches but linked to poverty by the focus groups. Big families, the tradition in Maasai culture, were viewed almost universally positively, as remains generally true in rural Africa.458 Nonetheless, when discussing negative contributors to future wellbeing, a number of Olkoroi HoH, both male and female, commented on the difficulty of providing for large families, particularly if education was a family aspiration. Work by Switzer, Archambault and others suggest education is now widely prioritized in the Maasai, including opportunities for girls.212, 393 In casual conversations, younger and/or educated Olkoroi residents occasionally articulated values similar to those of the younger Kisaju respondents, for example, referencing the use of birth control to control family size. While the focus groups were asked to describe characteristics/behaviours/options associated with each wealth level for the list of wealth markers, participants frequently followed up by noting that most or all community members followed traditional practice. For a number of markers, education was as, if not more likely than wealth, to be the factor determining non-traditional practice.  All members of the community live in similar houses, the traditional Maasai housing made of sticks/logs and mud/dung…97% of parents of all categories have no education… All families use 4-legged stools no matter what level of wealth they possess… Most families also use traditional beds… almost everybody wears shukas [traditional fabric] though a few educated people may dress in a Western style… (Responses from the men’s wealth marker focus group, 2008)   Some poverty research on pastoral communities has suggested that cash revenue, and proximity, access and integration into market economies, often held up as key indicators of wealth or opportunity, are not good measurement tools or predictors of pastoral security and wealth.66, 108 Educational attainment or possessions such as radios, used for SES demarcation in other SSA research, may likewise provide little differentiation in communities like Olkoroi, with low literacy rates. Focus group participants explicitly articulated that radios were not wealth indicators because they held little value for uneducated adults who could not speak English or Swahili (the official languages typically used in broadcasts), and   100  furthermore could not be used while herding. Women were almost universally illiterate in both national languages, but most men spoke at least some Swahili. Livestock holdings have long been used as a wealth marker in pastoral research,229 and remain a simple, effective way to quickly categorize families. However, regardless of marker(s) used, most studies have been able to differentiate more than two wealth strata.108 It is possible that cluster analysis only identified two groups in Olkoroi as a consequence of investigating a single community, the relatively small number of households (75) and the high proportion of households without livestock. However, despite using substantially more complex models in their large, multi-team investigation of livelihoods in a range of Maasai communities in Kenya and Tanzania, the Homewood research groups found enormous variation in correlations between livelihoods and family revenue and concluded that livestock not only remained central to Maasai wellbeing, but additionally were the most common resource used to fund major diversification in all income strata.108 The wealthiest household associated with Olkoroi (an order of magnitude more livestock than any other family) was not included in analyses because the HoH was absent (but had left his livestock in Olkoroi under the care of extended family), and furthermore liquidated his holdings shortly after data collection began. Like a small number of included HoH who attempted significant diversification efforts over the period of the research, the richest man purposefully used his animals to fund a major new business initiative. Most Olkoroi families, however, owned too few, if any, animals to support major diversification. As widely reported in pastoral research, smaller-scale diversification was generally used for herd maintenance and development.   Wellbeing 4.3.2.1 Male vs Female SWLS Scores Similar overall proportions of Olkoroi adults self-rated as average/neutral in life satisfaction, to those who were satisfied/very satisfied, but women were significantly more likely to report average life satisfaction than men, who were conversely more likely to be satisfied/very satisfied. Sex differences in life satisfaction are not consistent globally, and neither are the explanatory conclusions of wellbeing researchers. Overall, unlike the findings in Olkoroi, international research suggests a trend to higher wellbeing in women compared to   101  men with some country and regional exceptions.239, 497-499 The female positive wellbeing gap appears to be strongest in wealthy countries471, 499 though some work concludes an almost universal no difference or female positive sex difference.500 In contrast, some publications have noted a decline in female wellbeing in industrialized nations since the 1970’s501, 502 to the point that at least one study concluded men were now globally happier than women.501 The 2015 World Happiness Report found women were slightly happier than men globally, but sex-based differences were small compared to age and regional differences. Although Kenya is ranked as a lower-middle income country, and is no longer on the CIDA priority list of countries in greatest need, Tanzania remains poor.503 Global happiness assessments carried out by Gallup reported very low Tanzanian scores, on average, compared to Kenya,504 but regardless of national statistics, there is little to distinguish the status of rural Maasai in the two countries. As noted in chapter three, there was significant reciprocal movement between the two countries by residents of Olkoroi and surrounding communities. Even if relative East African national economic positions were relevant to gender differences in wellbeing, national trends may not be relevant to communities like Olkoroi. Gallup has also specifically acknowledged a “paucity of data on African happiness”.504 Vieira-Lima497 used World Values Survey (WVS) data to conclude that the biggest female positive gaps were in less industrialized nations with relatively low gender equality, though also remarked the reverse was true in a number of Latin American and European nations (without identifying any clear explanation for, or commonalities between “exception” countries). Of note in the Vieira-Lima publication was the finding that of the 5 countries with the highest wellbeing gap in favour of women, two were in East Africa and a third in southeast Africa (Uganda, Tanzania and Zimbabwe respectively). In contrast, some literature has claimed that a female-positive wellbeing gap did not extend to SSA498, 499 and the 2017 World Happiness report called Africa “the unhappiest continent”.504 Graham et al.499 found that countries with greater gender inequality exhibited lower female wellbeing, and available data from SSA showed higher male than female wellbeing in most years. The Graham research also noted an association between increased education and age with higher female wellbeing and concluded the generally global positive “marriage effect” did not hold true in rural areas or in the poorest nations, particularly the SSA and LAC regions. Although these   102  findings seem initially relevant to Olkoroi, further work by Graham suggested  the negative marriage effect was driven by lower levels of satisfaction in married men in all national income strata, which contradicts the Olkoroi findings.499 Nonetheless, the general conclusions of Graham may be relevant to Olkoroi and other pastoral communities where education rates are low, men control family resources and women cannot inherit critical survival assets, nor receive a share of household resources when marriages break down. More recent work by Meisenberg,502 also using WVS data from 95 countries found that women had greater life satisfaction in approximately half of the surveyed countries, and men in the other half, although differences were not great (statistically significant differences were only found in 33 countries overall). In the ten African countries included, women were more satisfied than men.502 Tanzania, though not Kenya, was included in the WVS. The only other known SWLS study in a Maasai community recorded very high life satisfaction (98% of the sample, n=127, self-rated above average, with a mean score of 5.4), and concluded that women were significantly more satisfied than men.185 However, there were some possible biases in this investigation. The researchers appeared to have spent a short time in the community and in our experience the wellbeing concept required extensive discussion to put into a Maasai context. The Maasai frequently present a proud, albeit clichéd vision of traditional life to non-Maasai and the Biswas-Diener research described Maasai customs in a superficial and outdated manner, typical of the generic narrative presented to outsiders at first interaction.505 Furthermore the Biswas-Diener work referred to interviewing participants in villages, suggesting in doing so, they had avoided selection bias. However, when the term “village” is used by the Maasai, it is not usually used in the Western sense. Instead it refers to a group of connected families, at minimum bound by friendship and allied agendas, and frequently by blood or marriage, for example a group of brothers/brothers-in-law and their families. Although it is not clear where the Diener research took place, a related publication implied it may have been in the same geographical area as my work.506 There are a number of possible explanations for my findings of proportionally different life satisfaction in Olkoroi women versus men as well as lower life satisfaction than previously reported in the Maasai. The first is the well documented global association of income with subjective wellbeing which is strongest in poorest countries.239, 471, 499, 507   103  Income is also more important when evaluating life holistically, as in my study, as opposed to contemplating specific life compartments.499 In Olkoroi, only 3/28 adults (11%) from households without livestock self-rated above average life satisfaction, in contrast with 59/122 (48%) adults affiliated with livestock-owning families. Many of the no livestock families were structurally vulnerable as well. Twenty-one of the 28 adults in the no-livestock households had no partner: 7 were widowed, 12 had been abandoned or had left an abusive spouse, and 2 were unmarried, younger men. Furthermore, 17/21 of the unpartnered adults were women, and 13/22 of the no livestock households were FHH. As well as the negative female gender-wellbeing associations in poor countries identified in Gallup poll data499, analysis of Gallup data by others concluded that social support was an important component of wellbeing.508 Olkoroi families without livestock were also much more likely to be missing social safety nets, particularly women who had been abandoned or had left abusive spouses. Stress [is the most negative contributor to my wellbeing] because I stay for many years without livestock and also my family because there are many things I cannot be able to do. Because sometimes I am not able to find food for the children and when all these things happen I get stress. Also I don’t have a husband so since my husband died I have stress at this point (SWLS/livestock duties interview: 38 year old, monogamous, widowed woman).  Although women who left high conflict marriages frequently returned to their birth family, they often faced significant pressure to rejoin their husbands because of traditions associated with bride price and ownership of children.331 In addition some women came from families that were too poor to support adult daughters (one woman who was beaten badly, told us she wanted to, but could not return to her widowed mother because of her mother’s poverty).  Growing awareness of non-traditional lifestyles through increased interaction with the outside world via cellphones, the Internet and higher rates of education in younger generations, may also have increased dissatisfaction. Many adults stated in wellbeing discussions that herding was demanding physically and time wise, but saw limited options for a different life, especially older residents, without education, and women, with little autonomy. Pastoral research has reported widely on increased inequity17, 108 and wellbeing research has repeatedly found that comparisons to richer peers weaken the correlation between life satisfaction and income especially if basic needs have been met.471, 509 In the   104  same vein, some investigations have concluded improvement in gender equity could lead to temporary or permanent declines in women’s wellbeing, possibly because increased workplace and societal equity did not often result in proportional reductions in home responsibilities501. Although this conclusion was drawn from US and other wealthy nations, some research in India has suggested challenging gender gaps may be complicated in very traditional societies as an increase in female autonomy, for example through participation in self-help organisations, if in direct conflict with traditional mores, may trigger a decline in female wellbeing.510 Furthermore, there is evidence that, as in many industrialized countries, Maasai women are taking more responsibility for income earning while still carrying the majority of household and child-rearing responsibilities.70, 316, 350, 470. However, despite the predominantly traditional lifestyle in the community, some Olkoroi women were aware of their national legal rights and had resisted gender restrictive customs and tribal law. For example: a woman who had left an abusive husband, journeyed to the Narok County Children’s Office to seek support for retrieval of her children; another encouraged her daughter to refuse circumcision; several women interfered with paternal plans to marry off young daughters by sending the girls to family outside Olkoroi to continue their schooling, or in a few cases assisted their daughters to enroll in secondary school (at risk of significant repercussion in the form of violent threats, actual violence, and temporary banishment); and a number of educated young women refused marriage to older men (no examples of marriage refusals were observed in uneducated women).  Regardless of awareness, most Olkoroi women lacked capacity and/or material capital to access legal rights, independence, or opportunity, although some received support from male extended family, or, in a few cases, adult daughters, to build livestock assets, or to enroll younger children in secondary school. A small number of studies on female autonomy (in, for example, financial or fertility realms), including some Kenyan work, suggest that women in households with discordant marital values may exercise secret autonomy, but if unable to protect themselves or their assets, are less likely to pursue potentially beneficial initiatives.511 This phenomenon was observed in Olkoroi in the form of secret female-owned livestock held by family living outside the community.   105  Many Olkoroi women lamented their lack of education or prematurely terminated education, and in a variety of contexts- during data collection, interviews, and casual conversation - addressed frustration with their lack of autonomy and independent resources: [The most negative current factor in my wellbeing is…]… marriage, because you will be under control, and it is very different when you compare the time when you were with your mother (SWLS/livestock duties interview, 20 year old, polygamous second wife).  [The most negative current factor in my welling is…] My husband’s control (SWLS/livestock duties interview, 24 year old, monogamous wife)  An interesting intersection of these issues was emphasized in a recent qualitative research on Maasai wellbeing in the context of conservation initiatives in Northern Tanzania.186 Some older Maasai men continued to view education for girls as a wasted asset, but younger men saw it as valuable for both sexes. Women perceived education as important for two general reasons: increased personal and community security, but also improving land security and understanding. For their daughters, however, women saw education as a route to independence. In Olkoroi, at a women’s community meeting, an older woman said that the primary constraint for women was “We have no power”. She elaborated that lack of female autonomy and opportunity, and high risk of poverty were direct consequences of polygamy, arranged marriages, and traditional age gaps which frequently left women widowed early with little capacity to provide care for potentially still young children. She also stated that all of these issues had particularly negative consequences for daughter’s educations.   4.3.2.2 Current Contributors and Detractors to/from Wellbeing Lending support to the hypothesis that gender roles may in part explain proportional differences in satisfaction, the two major differences between men and women in attribution of most important contribution to current wellbeing were related to marriage and children. Although there may be some overlap of the concept of family with children and marriage, most female respondents explicitly identified children as a source of happiness as opposed to   106  men who were much more likely to use the word family. In further support of this interpretation, about half (12/23) of the women who explicitly identified marriage as a primary detractor from life satisfaction chose children as the most important positive contributor to their wellbeing. A woman, who had left an abusive husband stated:  …since I left him….I feel very comfortable to live…without him. Nowadays, I don`t have a lot of stress (SWLS/livestock duties interview, 36 year old, polygamous second wife).   Another said her husband was the most negative influence on her current life satisfaction but:  There is only one thing [that would contribute to her future happiness] if god gives us to live for long life with my children (38 year old, monogamous wife).  Responses to the question about current detractors to wellbeing further emphasized gender differences in the perception of marriage. For women, marriage tied with “nothing” as the most common detractor to life satisfaction (25%), a choice made by only one man. Men were much more concerned about livelihood, possibly because, as in many societies, they were primarily responsible for household economic support. Although many women in Olkoroi contributed financially as they could, responsibility for the majority of household duties left them little time when children were young, a difficulty which was compounded by opportunities limited by tribal history, culture and location. Lack of educational opportunities also meant there were few avenues for diversification once children were grown.  4.3.2.3 Future Contributors and Detractors to/from Wellbeing: Men vs Women Responses to the question about potential future positive contributors to life satisfaction were the only ones for which there was no significant proportional difference between men and women. A possible explanation is the fatalist tendency observed in followers of traditional Maasai religion. The Christian God is often presented as actively involved with believer’s lives, and/or an entity who will bring rewards to followers, but the pluralistic Maasai God/s has sometimes been represented as a somewhat capricious being who didn’t engage directly in human activity, and was therefore an “unknown” force.403, 512   107  Respondents who didn’t respond definitively to the wellbeing questions were as likely to respond with “Only God knows”, as “I don’t know”, and often said both. In addition, a high proportion of both sexes viewed children as a key to a good future. In the Maasai tradition, the oldest son has responsibility for his father, and the youngest for his mother. As long as children do not fall into poverty as adults, both parents can thus expect care in their later years. Factors commonly perceived to contribute positively to life satisfaction in the longer-term may also have been less gendered than possible detractors, and/or, because of future uncertainties, were less clearly identifiable than current positive and negative effectors. In exploring possible future detractors from wellbeing, poverty or failure, was the top concern for both men and women, albeit almost twice as much for men as for women. However, women were almost equally concerned about marital conflict which appeared to be more likely to lead to poverty for women. In an echo of the high proportion of women who identified marriage as the most negative current contributor to their wellbeing, again only one man foresaw marital conflict as a possible future problem. The most common fears of poverty and marital conflict for women, were, based on existing demography in the community, strongly connected. The women who were most likely to be poor in the community were those who were widowed, abandoned, or who chose to leave a marriage. Poverty affected older men, but even when poor, they were rarely alone, whereas for women, being alone, regardless of cause, appeared to be frequently related to vulnerability.    Explanatory Wellbeing Model Significant variables associated with increased likelihood of life satisfaction further supported the conclusions drawn from the cluster analysis, life satisfaction, and gender differences in self-reported wellbeing contributors and detractors. In particular the significance of decision-making power but not sex, perhaps explained the differences in male and female life satisfaction. Several women commented on their inability to make decisions regarding household, or livestock brought into marriage. Livestock holdings were both the most important wealth marker, and a statistically significant positive associate for wellbeing in Olkoroi. However, as the herd size variable was measure in log10 units, the strongest effects on wellbeing in Olkoroi would have been experienced at the lower end of livestock   108  holdings (as ownership increased from none or few to 10 units). Since the largest TLU owned was 78, there was a local limit to wellbeing benefit arising from livestock. From a wellbeing perspective, each additional 10 fold increase in TLUs owned was associated with almost a full category (categories differ by 4 points on the SWLS scale)513 improvement in wellbeing (for example from neutral to satisfied to very satisfied).  Livestock can provide complete household support if holdings are large enough and are also assets for diversification and expansion of family opportunity and security. However, control and access to family livestock resources still lay almost exclusively in the hands of the male HoH which left women with little opportunity to make life changes.  I don’t have any livestock, the husband owns everything (SWLS/livestock duties interview: 35 year old, monogamous woman).  I don’t own any livestock, my husband doesn’t allow me (SWLS/livestock duties interview: 32 year old, monogamous woman).  It is difficult…I have many cows I brought from relatives but I don’t own them, they are still being controlled by my husband (37 year old, polygamous 3rd wife).  Although Maasai women usually have jurisdiction over household matters pertaining to food and children, male HoH could withhold household funds resulting in family hunger. Any activities outside the traditional female domain usually required male permission. The Global Happiness Report has consistently found that autonomy is a significant component to national life satisfaction,320 and several women explicitly commented on this issue during the wellbeing interviews. The positive association between wellbeing and capacity to make household decisions was likely associated, in part to the basic importance of wealth in wellbeing. In Olkoroi, position as primary HoH allowed free use of family assets, livestock or otherwise, which was customarily denied to women. Women who spoke about autonomy repeatedly referenced it as contingent on asset control. Since livestock were the predominant asset, control of livestock was a key component of autonomy. Women who owned their own animals made a clear link between this power and personal freedom:   109  Yes, I have [livestock] because I have a small business and when I buy any animals for my business then I must own them, the husband can’t control me at all. (SWLS/livestock duties interview: 50 year old, polygamous second wife).  I own my own livestock because I don’t have a husband (SWLS/livestock duties interview: 48 year old widow)  I have my own [livestock] because I don’t have a husband who can prevent me so I can do all because what I have is mine. (SWLS/livestock duties interview: 50 year old, divorced, polygamous woman)  Only 4/22 (18%) HoH with no livestock self-rated above neutral life satisfaction, in comparison with 35/53 (66%) of livestock owning HoH, and larger livestock holdings were associated with increased life satisfaction. At the individual level, however, possibly because of lack of autonomy, some woman living in households with livestock were still frustrated and unhappy. Specifically, in livestock-owning households, 32/53 men (60%) self-rated above neutral life satisfaction, as opposed to 27/69 women (39%).  The only thing that I see it will improve my life, I wish I will be able to have the authority to [take] charge [of] my life, the management of my livestock, and agriculture (SWLS/livestock duties interview: 49 year old, widowed, polygamous fifth wife).   Similarly, although a number of female HoH identified lack of a MHH as the primary negative contributor to their wellbeing, some, reflecting on positive contributors, spoke of relief from of stress and persecution after leaving their husbands, despite having few or no assets and frequent dependency on extended family. The FHH of an abandoned family identified as one of the most deprived in Olkoroi, was still periodically harassed by a very abusive male HoH, despite his routine absence, and stated: [The most negative contributor to my wellbeing] is my husband, nothing else (SWLS/livestock duties interview: 48 year old, polygamous wife)   110  Some of the oldest women, typically widowed and with no personal assets, if living with extended family, were highly content. Conversely, four with little family support, who lived in extreme poverty and experienced routine hunger, were among the least happy.  As noted in chapter three, children and family are highly valued in Maasai culture, so the association between number of children and life satisfaction was not unexpected, though small. Children were repeatedly referenced in the context of a good future, in the contributors/detractors discussion. However, expressed concern by some HoH about educating children, and the predominant focus of the SWLS on past and present perspectives (rather than future), may have reduced the strength of the association in the SWLS predictive model. As also discussed in chapter three, the complex associations between alcohol consumption and household vulnerabilities may explain the increase in wellbeing associated with being teetotal. Churchgoers were universally tee-total, and some (though not all) of the largest herds were owned by the men most strongly affiliated with the local missionary. As alcohol consumption was not directly questioned, more specific conclusions cannot be drawn, however, an abandoned female alcoholic with dependent children stated unequivocally that the most important possible future detractor to her wellbeing would be: If I don’t stop taking alcohol. (SWLS/livestock duties interview: 40 year old, monogamous wife)  4.4 Conclusion Livestock remained vitally important to livelihood in Olkoroi, as has been determined in other pastoral and livestock dependent community research.108 In addition, they contributed, likely through their role as primary asset, significantly to psychological wellbeing. However, the results of this study suggested gender is a critical determinant of the degree of benefit associated with livestock ownership. It is therefore imperative to continue the push to incorporate gender perspectives into livestock-related research and development in a more meaningful manner than has been achieved to date. Although material wealth in Olkoroi was still primarily determined by livestock holdings, many families had no/insufficient livestock to support themselves by traditional pastoralism alone. As increased livestock holdings were also positively associated with life   111  satisfaction, the high proportion of households with no/low livestock may have been responsible for the fact that almost 60% of community members self-rated their life satisfaction as average or less than average as opposed to satisfied/very satisfied (41%). Assessing the contribution of livestock to wellbeing was further complicated by the fact that most women lacked decision-making authority regarding livestock, with the exception of female HoH holding assets for unmarried sons. Although some women had their own animals, such holdings were typically obtained from extended family, earned independently, and/or held secretly. Thus, although the most parsimonious wellbeing model suggested larger household livestock holdings were associated with higher wellbeing, livestock benefits did not automatically translate into improvement in individual wellbeing for all. Not just Maasai women, but Kenyan women in general, are often culturally restricted from accessing major family assets of livestock and land.398, 514 Community discussion on privatization and subsequent allocation of communally owned land, which occurred over the entire period of this research, excluded women completely. Women also faced major challenges in acquiring and controlling their own assets, which in turn appeared to be critical in determining feelings of autonomy and wellbeing. Position as primary decision maker, but not sex, was positively correlated with life satisfaction of community members. This supports the effect of constrained autonomy rather than gender per se as a source of dissatisfaction. Of the varied self-identified contributors and detractors from life satisfaction, marriage appeared to be one of the strongest negative influences on women’s lives, identified by 25% of women as a current detractor, and 21% as the most likely future detractor to life satisfaction. Lack of autonomy, inability to access productive assets both at the family level and in the greater community (for example credit), and restrictive cultural norms, appeared to constrain the ability of women to both contribute effectively to family wealth, and achieve life satisfaction. Older, widowed, Olkoroi women rarely remarried in large part because cultural narratives reduced the appeal of older women as marriage candidates. 331 Without support from extended family, single female HoH of any age appear to be more likely to become poor/er, and likely less satisfied. Unhappily married women were therefore trapped in a catch 22; an unhappy marriage reduced life satisfaction, but a broken marriage could result in poverty which also reduced life satisfaction.   112  Although extensive research in many different parts of the world almost consistently demonstrates that increasing opportunity for women increases family wellbeing, equity and productivity, when women are very restricted in capacity, interventions may need to be multi-level and multi-imensional.511 Nonetheless, more nuanced perspectives on gender in project and research design may allow identification and incorporation of differences in perspectives and needs. In turn, this could facilitate female access to opportunity, more engaged participation, and consequently increased productivity, material and psychological wellbeing of all members of pastoral communities, while still respecting cultural values.    113   Livestock Disease, Rearing Practices and Contribution to Wellbeing: Community Perceptions and Exploratory Models  5.1 Introduction Research on rural Maasai, to date, has not always fully captured the diversity of within household priorities, perspectives and experience with livestock, including keeper views on productivity constraints. In addition, many livestock disease studies are cross-sectional only. In combination, this may limit capacity to support the movement of livestock holders out of poverty, and understand critical, temporal cause-effect relationships in disease incidence. This chapter therefore aims to provide perspective on these issues through investigation of the following questions:  How does gender affect responsibilities for livestock tasks and HoH perceptions about the contribution of livestock to wellbeing?  How do HoH perceive general family performance of livestock related duties, optimal, best, and worst household practices, and most important barriers to productivity?  How do HoH understand and prioritize livestock diseases and treatments?  Which variables, including seasonal, household and herd characteristics, are most strongly associated with size of livestock holdings, growth rates of young livestock, and livestock disease prevalence?  Development literature has repeatedly asserted that livestock are a potential route out of poverty for the global poor, especially the 1-2 billion smallholders who are already at least partially livestock dependent.29, 243, 515-518 However, perspectives on how to most effectively use livestock to alleviate poverty vary substantially among livestock keepers, researchers, development agencies, politicians, and commercially interested parties.264 The opinions of keepers may be affected by personal priorities, socioeconomic status (SES), geographic location, culture, and gender. The priorities of other actors may be influenced by personal, national, international, organizational and/or economic agendas.45, 83, 264 Impediments to increased productivity identified by non-keepers, include: ability to intensify production;243,   114  244 number of, and efficiency of processing facilities;10, 243, 248, 258, 519, 520 diverse aspects of technology,8 from market communication521 to artificial insemination522, 523; participation in local and global markets;147, 524, 525 livestock disease29 especially zoonoses9, 79; access to specialty breeds;18, 302, 518, 526 and recently, the contribution of livestock to GCC.527 Surmounting these barriers is sometimes presented as critical to all livestock keepers,29 but it has also been acknowledged that commercial concerns may be less relevant for the most poor.10 For the very poor, building livestock holdings,16 and for women, access to and control over both family and personal livestock assets are frequently more relevant issues.528  In surveys, pastoralists frequently identify livestock disease and the demands of dry seasons/droughts as livelihood challenges. A related issue commonly raised, is infrastructure deficiency, from medical and veterinary to education.318, 339, 342, 454, 529-532 Fewer studies report pastoral concern about market access.529, 531 Some academics perceive a disproportionate emphasis on market solutions for poverty reduction, and have pointed out that market concerns of traditional pastoralists are significantly different from those of non-extensive livestock farmers. 66 Intensified commercialization of livestock-rearing may also reduce traditional female control over assets such as milk and milk byproducts.147, 533, 534 In a major review of livestock development projects, the difficulties of the poor were succinctly summarized as livestock acquisition, maintenance, retention, and marketing. The respective constraints for each were identified as: lack of funds and/or credit; disease, inability to access veterinary resources, and fodder access; and location, poor infrastructure, and trade barriers.16 Although 20 years old, the review remains valid and highly cited, but lacked a gendered perspective (women were mentioned only twice) albeit stating, “Livestock are particularly important for women, for whom they represent one of the most widely held and important assets, and one of the most rewarding income-generating activities available”.16  Location constrains rural agriculture regardless of SES.46, 295, 535-537 Mobility and political marginalization63, 68, 538, 539 may add to pastoralist difficulties. Frequent drought and limited veterinary services in ASAL territories540 exacerbate the impact of livestock diseases and pasture/fodder deficits. The poorest keepers, and especially women,541 however, face additional challenges that reduce the return on their efforts. Women are often less able to access resources to improve productivity.542 Heffernan found farmers below the poverty line   115  earned 50% less from their livestock than those above.515 Research in poor, non-pastoral communities in Nyanza, concluded livestock improved household security, but substantively increased workload for women and children, 543 as has been observed in pastoralists generally,544 and the Maasai specifically.371 An older Maasai study concluded that wealthy pastoralists managed five times more livestock per person than the poor, with repercussions for productivity, time and labour management.229 Collaborative work conducted across heterogeneous Maasai communities in East Africa reported the poor diversified from necessity, but the better-off for risk management, and growth of livestock wealth.108 A 2016 wealth comparison study concluded diversification benefited rich Maasai households much more than poor, and diversification intensity did not effectively predict wealth.545 Poor households  may also have proportionally higher offtake because of subsistence needs219, 546 and research by Bekure found poorest Maasai families had the highest gross output (in KSh) per animal (almost 300% more than the richest households), primarily because they sold more milk products.547 High offtake, however, can come at the expense of household548  consumption, and may slow young livestock growth, stress productive animals, and contribute to high pastoral pre-weaning mortalities.549  Livestock continue to play a central role in Maasai identity, culture, and livelihood. Even when Maasai diversify and/or ostensibly leave pastoralism, strong ties and support often remain for those following traditional life, and urbanites frequently secure assets in rural livestock.108, 550 Thus, effective livestock-based interventions to improve health, wealth and wellbeing of the Maasai, and potentially other pastoralists, and rural livestock-dependent communities, require informed consideration of livestock-keeper relationships, especially individual perceptions, interpretations, and rationales for pastoral practice. Notwithstanding a large body of research on the Maasai, their views and needs have not always been fully acknowledged.126, 139, 226, 371, 377, 551, 552 From colonial time until now, Maasai portrayals have veered from impoverished to wealthy, ecosystem destroyers to ecological wardens, irrational to wise, and, powerful to marginalized.1, 376, 553-555 Galaty (2013) described pastoralists as: …desperately in need of radical change and ‘development’ but…when it arrives they tend to suffer from it; they are traditionalists who would benefit from the innovations of modernity but they are already very much part of the modern world and must be selective in which of its elements they embrace.556    116  It is not however, enough to gather community or household-level information and generalize.515, 557, 558 As in most societies, within-household variation affects livelihood benefits and disadvantages.108, 557-559 Several factors in particular appear to influence pastoral perceptions and experience. Household structure is important, including marriage status and presence of an active, MHH.108, 218, 560, 561 SES is very influential,218, 515, 557 especially herd size.144, 528, 545, 558 Gender is also a critical determinant of livestock responsibilities, decision making power and opportunity.371, 525, 528, 533, 545, 562 Historically, pastoral productivity and poverty alleviation initiatives mainly focused on: drought recovery, 276, 563-565 productivity improvement,376, 566-568 amelioration of service gaps,310, 569-571 and, more recently, support for women.572 Many projects have not achieved their intent,16, 573 and some evidence suggests pastoralists can be reluctant to maintain interventions which require personal investment.559, 574-576 Reasons for failures are numerous, but the more complex may be rooted in any or a combination of: traditional practice,559 culture-specific cost-benefit analysis,577 negative past experience,376 and expectations of outside agencies,576, 578 all dynamics which may be difficult to explicate or resolve. There may also be conflict between orthodox science and traditional knowledge and expertise.45, 579 Widespread rural information poverty creates barriers to access and effective use of scientific information and tools,515 in particular pharmaceuticals, and may create new problems, such as drug resistance.521, 540 Some academics and professionals believe traditional keepers lack requisite knowledge.515, 580 In worst case scenario, pastoralists are “ignorant”,230 sometimes they are experts,330, 376, 556 and occasionally they are both.581 Conversely, researchers have made, and continue to make errors because of failure to consider, understand582 or respect community dynamics and priorities,45, 579, 583 and include all SES strata, and gender.16, 573, 584 Knowledge translation from research to practice may fail, or not even be attempted at research completion.45, 83 Investigators may repeat the similar studies in different locations without synthesis (for example, the many individual prevalence assessments of brucellosis),85 focus disproportionately on a small number of diseases,83 and development professionals may make premature conclusions because of failure or inability to monitor interventions for sufficient time, or commit to changes in development paradigms.16, 45, 83, 376, 576, 585    117  One of the most basic hindrances to livestock productivity is lack of veterinary services. Unfortunately this barrier has become almost a given in most rural regions of SSA.264, 586 Insufficient veterinary resources exacerbate two other fundamental limiters: livestock disease incidence and prevalence, and information poverty regarding disease transmission and management.515 Furthermore there is a paucity of information on the barriers which prevent adoption of best practice. Substantial literature exists on livestock productivity and approaches for increasing productivity, but little changes in rural regions.45 In Narok, district veterinary officers stated the DVO had enough trained personnel, but insufficient funding for service delivery. They further claimed the Maasai resisted new ideas and practices but were simultaneously dismissive of traditional beliefs and livelihood.230 Olkoroi residents, on the other hand, stated veterinary presence was rare except in outbreaks.   5.2 Results The data summaries and analyses for this chapter are based on the following as described in Chapter Two: the livestock duties and contribution of livestock to wellbeing portion of the livestock duties/SWLS/wellbeing interv