(STILL) A BURNING ISSUE: FUEL-EFFICIENT STOVE USE IN RURAL KENYA

A Thesis

Presented to

The Faculty of Graduate Studies

of

The University of Guelph

by

ALLISON PATRICIA GRATZ

In partial fulfilment of requirements

for the degree of

Master of Science

April, 2009

©Allison Patricia Gratz, 2009 ABSTRACT

(STILL) A BURNING ISSUE: FUEL-EFFICIENT STOVE USE IN RURAL KENYA

Allison Patricia Gratz Advisor: University of Guelph, 2009 Professor James P. Mahone

Could a stove hold a meaning beyond merely being a tool to cook with? This thesis investigates the knowledge and perceptions of rural residents concerning raw-biomass stoves in Trans Nzoia district, Kenya. The research aimed to identify factors that influenced the use of fuel-efficient stoves in that area. Qualitative data from rural residents, Ministry of Agriculture staff and local administrators were gathered using semi-structured interview techniques and participant observation. Research findings underscored the importance of identifying practical and strategic needs of technology users and the role of local innovation in addressing those needs. There were differences in knowledge and exposure to fuel-efficient stoves depending on the ethnic heterogeneity of an area, age and sex of participants. A conceptual framework developed in the course of this study highlights the importance of culture, exposure, personal factors, and stove characteristics to develop ones capacity to use a new stove.

ACKNOWLEDGEMENTS

This thesis, although authored by me, would not be in its present form without the help of hundreds of other people. I would like to start by thanking my family, especially my parents, Bob and Verna Duncan. They instilled in me a deep sense of joy and wonderment in discovering the stories and experiences of others quite different from myself, as well as a recognition of the responses that knowledge should bring. They have always encouraged me to follow my heart and be sensitive to God‟s guidance in every area of my life. Thank you for the examples you set in your life.

An insurmountably huge thanks to my husband Mike for his emotional and even physical support as he journeyed beside me through this process of coursework, researching and writing a thesis. I couldn‟t ask for a better partner in life.

Next I will thank my advisor, Dr. Jim Mahone, for giving me confidence to apply for all sorts of funding and for the prodigious amounts of support, often simply in the form of a listening ear that you gave these last two years. Thank you for your feedback and unwavering encouragement. Thank you also to Dr. Helen Hambly Odame for your practical tips, sage advice, abundant offers of ideas and resources and extremely useful feedback on drafts. It was an honour to work with this advisory committee.

Graduate school would have been quite boring, uneventful and much more stress-filled without the laughter, tears, food, discussion, and encouragement shared with my classmates both in and out of academic settings. Thank you for the challenges and rewards working with such a diverse group of people gave me.

In Kenya my research assistants Okumu Nyongeza Nickson and Mwanziah S.

Grace were of insurmountable value for providing cultural and linguistic interpretation, help with data analysis and friendship throughout the study period.

i

Thank you to all of the staff at SCC-Vi‟s Olaf Palme Agroforestry Research Center in

Kitale, especially Fred Marani, Robert Musikoya, Sellah Wanjekeche, Juliet, Emily,

Japhrice, Sally, Limu, Soita, Ruth and Jane the driver. Their enthusiasm in this project and willingness to assist a novice researcher really made the experience of researching extremely smooth and rewarding.

I am immensely grateful to all of the participants in my study for having the grace to receive my research team into your busy houses, yards, kitchens, and offices, often unannounced. It is my sincere hope and desire that you will gain benefit from this project.

I was extremely privileged to receive funding for this study. This project was undertaken with the financial support of the Government of Canada provided through the Canadian International Development Agency (CIDA), the Government of Ontario provided through an Ontario Graduate Scholarship, and the University of Guelph through various awards. The Swedish Cooperative Centre‟s Vi Agroforestry Project in

Kitale, Kenya gave in-kind support.

ii

TABLE OF CONTENTS

ACKNOWLEDGEMENTS i

TABLE OF CONTENTS iii

LIST OF TABLES x

LIST OF FIGURES xii

GLOSSARY xiv

PICTORAL DIRECTORY OF STOVES FOUND IN TRANS NZOIA DISTRICT xv

PROLOGUE 1

CHAPTER ONE: INTRODUCTION TO THE STUDY 3

Background 4

Rationale for the Study 4

Objectives 5

Limitations 5

Significance 6

Overview of the Thesis 7

CHAPTER TWO: REVIEW OF LITERATURE 9

Introduction 9

Part one: Biomass Fuel and Stoves 9 Biomass Fuel 9 Traditional Stove 10 Advantages 11 Tradition 12 Disadvantages 12 Cost in Fuel 12 Cost in Time 13 Cost in Health 13 Cost to the Environment 16 History of the Wood Fuel Crisis 17 Present Thoughts on Environmental Impacts 19

iii

Part Two: History of Fuel Efficient Stoves and Their Promotion 21 Successes and Failures in Promotion of Household Energy Technologies 22 Diffusion of Innovation Theory 23 Facilitating Factors for Successful Promotion and Use of FES 25

Selected FES Models in East Africa Today 27 Successful Stove Promotion Programs 30 Community Based Extension: The Case of the Kuni Mbili Stove in Kenya 30 Non-Government Organisations: The Case of GTZ and Rocket Lorena Stoves in Uganda 31 Government-Run programs: The Case of China 31 Government-Run Programs: The Case of India 32 Government-Run Programs: The Case of Kenya 33

Common Advantages of FES 33

Common Disadvantages of FES 34

Part Three: Energy Programming and Policies in East Africa 35 Women in Development 36 Women, Environment and Development 37 Ecofeminism 38 Gender and Development 39 Gender and Energy 40 Gender Planning: Incorporating Practical and Strategic Gender Needs 41 Gendered Approaches: Efficiency vs. Equity & Equality 42 Gender-Energy-Poverty Nexus 43

Traditional Fuels in Public Energy Policies 45 The Energy Ladder 45 Government organisations 45 Gender in Energy Planning and Policies 47

Gaps in Literature 48 Using a Gendered Approach to Investigate Stove Adoption 48 Qualitative research 50 Complexity of Categorisation 50

Conceptual Framework 51

Summary 53

CHAPTER THREE: STUDY CONTEXT 54

Introduction 54

iv

History of the Study Area 54

Physical and Agricultural Characteristics of Trans Nzoia District 56

Cooking Habits of Study Participants 58

Dominant Ethnic Groups in Trans Nzoia and Their Relevant Characteristics 59

Recent Ethnic Violence in Trans Nzoia District 61

Relevant Characteristics of the Study Area 62 Cherangani Division 63 Kaplamai Division 64 Kiminini Division 66 Endebess Division 68 History of Stove Promotion and Use in Kenya 71 Specific Use of Fuel in Kenya 71 Fuel Deficit Scenario 73 Immediate Response to the Fuel Deficit 74 History of Chepkube 75

Summary 76

CHAPTER FOUR: METHODOLOGY 77

Introduction 77

Research Strategy 77

Partner Organisation 78

Research Assistance 79

Area Selection 81

Individual vs. Household Unit of Analysis 82 Kenyan Culture and Visitors 85

Literature review 85

Focus groups 86

Semi-structured interviews with individuals and groups 87

Participant Observation 89

Feedback sessions 89

Analysis 95

Limitations 95

v

Summary 97

CHAPTER FIVE: DATA ANALYSIS 99

Introduction 99

Semi-Structured Interview Participant Profiles 99 Personal Data of Semi-Structured Interview Participants 99 Other participant profiles 103

Perceived Advantages of the TSS 105 Ease of Construction and Maintenance 105 Cooking Advantages 106

Perceived Disadvantages of the TSS 107 Inconvenience 107 Fuel Consumption 108 Increased Danger 109

Circumstances Where No Other Style of Stove Can Be Substituted for TSS 110 Practical Advantages 111 Cultural Identity 111 Traditional Beliefs and Customs 111 Family 111 Marriage 112 Death 112 Disagreements 113 Curses 114

Household Roles and Benefits Concerning FES 116 Decision Making for Stoves and Fuel 116

Practical Gender Relations Influencing the Use of FES 117 Buildings and Materials 117 Acquiring Fuel 118

Perceived Benefits for Specific Family Members in FES Using Households 118 Benefits For Women 118 Benefits For Men 120 Benefits for Children 122

Perceived Need for FES Within the Community 122 Perceived Advantages of FES 124 Saving Energy and Time 125 FES Physical Characteristics 126 Avoiding Disturbance 127 Perceived Disadvantages of FES 129 FES in General 129 Clay Stoves 130 Sawdust Stoves 131

vi

Perceived Fuel Shortage in the Area 132 Perceptions of fuel scarcity 134 Perceptions of Fuel Abundance 135

Perceived Factors That Influenced Households to Use FES 136

Perceived Factors That Prevented Households From Using FES 139

Perceived Differences in Thoughts About FES Between Community Members 143 Perceived Differences in Thought Between Men and Women 143 Perceived Differences in Thought Between Older and Younger People 144 Perceived Thoughts of Other Community Members 146

Existing Channels For Transfer of Information About FES 146 Informal and individual exposure 149 Community Level Exposure 150

Summary 151

CHAPTER SIX: DISCUSSION OF FINDINGS 152

Introduction 152

A New Conceptual Framework 152

Key #1: Culture 154 Family of Origin 155 Gender Roles 156 The Role of Christianity 157

Key #2: Exposure 158 Exposure by Observation 158 Formal Extension 159 Lack of Established Vocabulary 161 General Paucity of Knowledge 162 Negative Misconceptions as a Barrier 163 Influence of Male and Female Gender Roles 165

Key #3: Personal Circumstances 166 Physical Infrastructure 166 Stove Availability 167 The Need to Save Money 169 The Need to Conserve Fuel 170

Key #4: Stove Characteristics 172 Stove Characteristics and Their Corresponding Benefits 172 Food warming capability 173 Preferences of Characteristics for Different Situations 175

vii

Assessing the Comparative Importance of Key Factors 176

Summary 179

CHAPTER SEVEN: SUMMARY, CONCLUSIONS AND RECOMMENDATIONS 180

Introduction 180

Discussion of Findings 180

Summary of Findings 188 Objective one: To determine the role of the traditional three- stone stove within a rural household 188 Objective two: To identify practical and strategic gender needs that are met by the use of fuel efficient stoves 188 Objective three: To identify current perceived need for fuel efficient stoves within the community 190 Objective four: To identify gendered perceptions around the use of fuel efficient stoves 191 Objective five: To characterise existing and potential community level training or involvement 192

Summary of Sets of Key Factors 193 Culture 193 Exposure 194 Personal Circumstance 195 Stove Characteristics 195

Limitations 196

Recommendations for Extensionists and Extension Policy Makers 196 Capacity Development 197 Recommendations Regarding Cultural Factors 198 Recommendations Regarding Exposure Factors 198 Recommendations Regarding Personal Circumstance Factors 200 Recommendations Regarding Stove Characteristics Factors 201

Further Research 202

Concluding Remarks 203

References 205

Appendix 1: Matrix of Research Goal and Questions 214

Appendix 2: Links between Gender, Energy, and the Millennium Development Goals 216

Appendix 3: Possible reasons for success or failure of stove programs 218

viii

Appendix 4: Relational chart of administrative areas within Kenya, with study areas highlighted 219

Appendix 5: History of Stove Promotion in the Study Area 220

Appendix 6: Focal areas within study area and their average rates of use of fuel-efficient stoves 221

Appendix 7: Intended and actual research participants and activities 222

Appendix 8: Survey for people whose household used a fuel-efficient stove at the time of the survey 223

Appendix 9: Survey for people whose household did not use a fuel- efficient stove at the time of the survey 229

Appendix 10: Schedule for feedback sessions in villages 235

Appendix 11: Example of feedback form for presentation at SCC-Vi’s Olaf Palme Agroforestry Centre on August 5, 2008 235

Appendix 12: List of study participants who were interviewed using the semi-structured interview guide 236

Appendix 13: List of study participants who were not interviewed using the semi-structured interview guide 238

Appendix 14: Sources of FES exposure for each household 241

Appendix 15: Breadth and depth of FES knowledge in the study area 243

Appendix 16: Fuel procurement habits in the study area 245

Appendix 17: Physical Components of Selected FES 249

Appendix 18: Direct and Indirect Short-Term benefits associated with each physical component of FES 250

Appendix 19: Decision tree for determining FES that best fits a rural resident’s personal circumstances 253

ix

LIST OF TABLES

Table 2.1: Descriptions of selected fuel saving devices found in literature 29

Table 3.1: Comparison of fuel deficit scenarios in Trans Nzoia district with the country of Kenya 73

Table 4.1: Timing and order of research activities 83

Table 4.2: Total population and number of households in Trans Nzoia district 85

Table 4.3: Methods used to gather data to answer each objective and their relative importance 86

Table 4.4: Feedback session dates, locations and attendance 90

Table 5.1: Perceived advantages and disadvantages of TSS listed by participants 110

Table 5.2: Cultural beliefs and customs involving TSS mentioned by Luhya and Sabaot participants 115

Table 5.3: Male and female responses to the question “who makes decisions about stoves in your household?” 116

Table 5.4: Stoves encountered during the research period, their descriptions, construction materials and fuels used. 123

Table 5.5: Comparison of perceptions concerning advantages and disadvantages of FES listed by participants 132

Table 5.6: Summary of fuel procurement methods in the study area 134

Table 5.7: Factors perceived by rural residents that influenced households to use FES 139

Table 5.8: Factors perceived by rural residents which influenced households to not use FES 142

Table 5.9: Sources of knowledge for FES familiarity, and the frequency for each source for 32 households in the study area. 148

Table 5.10: Use and awareness of FES in the study area 149

Table 6.1: Popular misconceptions and their corresponding statements of truth 164

Table 7.1: Comparison of methods of categorisation of FES use 181

x

Table 7.2: A comparison of factors that facilitate the use of FES found in literature with those found in this study 187

xi

LIST OF FIGURES

Figure 2.1: Conceptual framework used for the study 52

Figure 3.1: Location of Trans Nzoia district within Kenya 57

Figure 3.2: Study locations within Trans Nzoia district 63

Figure 3.3: The center of Sokomoko village, Kaplamai division 65

Figure 3.4: A farmer’s field in Mtarakwa village, Endebess division 69

Figure 3.5: Road conditions and scenery in Cheptantan village, Endebess division 70

Figure 3.6: Three girls collect wood fuel in Kiminini 72

Figure 3.7: Fraction of charcoal and solid wood fuels used by rural, urban and cottage industry users 73

Figure 4.1: Method of choosing study villages 82

Figure 4.2: Method of choosing participants for participation in semi- structured interviews in rural villages 84

Figure 4.3: Local children join the research team in a kitchen during an interview 88

Figure 4.4: A research assistant and the driver discuss FES with rural residents in Mtarakwa during a feedback session 91

Figure 4.5: A research assistant explains FES to a group of men in Nabiswa A 92

Figure 4.6: The author with a spontaneous gathering of rural residents at a FES feedback session in Cherangani 93

Figure 4.7 : Parents and teachers gathered at a feedback session in Cheptantan 94

Figure 5.1: Age distribution of semi-structured interview participants 100

Figure 5.2: Number of household members of semi-structured interview participants 101

Figure 5.3: Ethnic affiliation of semi-structured interview participants 102

Figure 5.4: Size of land used by semi-structured interview participants 103

Figure 5.5: Description of heads of households of semi-structured interview participants 103

xii

Figure 5.6: A woman in Central division demonstrates her chepkube brooder 127

Figure 5.7: Notches in the corners of each burner decreased disturbance due to smoke on this chepkube hatchery and the customised shape of burners held pots securely 128

Figure 5.8: Bicycles laden with firewood travelling from Endebess 136

Figure 6.1: Conceptual framework of factors that affect the use of fuel- efficient stoves in rural Kenya 154

Figure 6.2: Factors which influenced current FES users to use FES 176

Figure 6.3: A simplified timeline of events leading to widespread use of chepkube (CK) in an ethnically diverse village 177

Figure 6.4: Key factors which influenced participants to discontinue use of a FES 178

xiii

GLOSSARY

Agroforestry: A practice of combining elements of agriculture (plant and/or animal) with woody plant species in a mutually beneficial manner.

Chepkube: The term means “a hiding place” in the Kalenjin language. It also refers to a clay and brick stove with a metal cupboard embedded in the clay to store food and keep it warm.

Extensionist: a practitioner of extension (education) activities, usually with a rural focus.

FAO: Food and Agriculture Organisation of the United Nations

FES: Fuel-efficient stove. This term includes fuel-saving devices such as fireless cookers and solar cookers.

Githeri : A mixture of maize and beans boiled until soft, which takes not less than three hours. This is a very popular dish in places where both maize and beans are plentiful. It is simple to prepare and does not require close monitoring. It is possible to decrease the cooking time by pre-soaking, but the taste changes because of fermentation, and people don‟t prefer it that way.

GO: Government organisation (for example, the Ministry of Agriculture).

GTZ: German Technical Cooperation organisation (or Deutsche Gesellschaft für Technische Zusammenarbeit), an organisation funded by the German government for international development related projects.

ITDG: Intermediate Technology Development Group. A non-governmental organisation based out of Britain with a focus on international development.

Merry-go-round: A type of money lending arrangement used by women‟s groups in Kenya. Each woman contributes a set amount of money at regular intervals, and receives all the contributed money when it is her turn.

NGO: Non-government organisation (for example, SCC-Vi).

Pombe: A locally brewed alcohol made of fermented maize flour.

SCC-Vi: The name for the Swedish Cooperative Centre‟s projects in East Africa. „Vi‟ means „we‟ in Swedish.

Ugali: The staple food of Kenya. Usually made of maize flour and water, although traditionally other ingredients such as finger millet, cassava and sorghum were used. It is prepared by stirring over a fire until a very thick porridge or paste is formed varying in stiffness according to preference. It is usually eaten with vegetables, meat, or a stew.

xiv

PICTORAL DIRECTORY OF STOVES FOUND IN TRANS NZOIA DISTRICT, KENYA

A three-stone stove fuelled by maize Three-stone stove outside a rental house stovers is used to prepare ugali in Endebess in Kiminini

Improved three-stone stove with metal Improved three stone stove built into the bracing in Kaplamai wall of a house in Cherangani

Improved three stone stove with two Clay stove with three burners and wood burners in Cherangani drying above in Endebess

xv

Chepkube with two burners in Kiminini Chepkube with additional side burner in Cherangani

Chepkube inside a kitchen with a window Decorative chepkube fuelled by maize for light in Cherangani cobs in Cherangani

Chepkube brooder in Central division Chepkube hatchery in Cherangani

xvi

Cement chepkube in Kiminini Pot sits inside burner of cement chepkube

Now defunct chepkube bakery oven with Regular charcoal stove (left) and deluxe owner in Kaplamai charcoal stove(right) that uses wood in Cherangani

Metal sawdust stove bought in a local Clay sawdust stove made by a woman in market Kaplamai

xvii

Woman in Endebess with her kuni mbili Kuni mbili liner (left), liner encased in installed with a chimney metal (right), and fireless cooker at the Ministry of Agriculture in Cherangani

Kuni mbili liner used to cook chapati (flat Woman and her child sit with two sizes of bread) at a home in Cherangani fireless cookers in Cherangani

xviii

PROLOGUE

March 2006

Captivated, I watched three middle aged Kenyan women acted out a mime in a dusty house yard in rural Kenya. They had prepared their presentation for a team of representatives from SCC-Vi, a Swedish-based agroforestry initiative working in

Eastern Africa who I was volunteering with at the time. One actress was dressed like a man, and the other two were meant to be his co-wives; one attired with neat, tidy clothing, and the other in clothes that were more tired and ragged. The husband asked his two wives to make him tea to drink. Immediately the well-attired woman built a fire using solid fuel wood she had gathered from trees she had planted. Soon flames were licking at a pot placed on three supporting stones, and tea was boiling in minutes. She added good quantities of milk and sugar to her brew, and served it proudly to her husband. Meanwhile, her co-wife struggled to bring water to a weak boil because she was using only maize stalks to fuel her three-stoned stove. She had neither money nor cow, so all she could offer her husband was tepid brown water.

The message this drama meant to convey was that people who plant trees and practice other agroforestry methods gain the benefits of fuel, healthy animals, extra income, and a happy family.

This scene was etched in my mind, along with the question why, if the maize- stalk using co-wife could see her peer succeeding so well with her alternate farming methods, did she not follow her example and emulate her success? Watching this play and observing the action and inaction of people in Kenya and Canada concerning habits that could drastically change their lives and the lives of their families and communities was the basis for my research. Specifically, I investigated the factors that led some people to use fuel-efficient cooking devices while others did not. The

1

answers I got from my questions for this project have given clues to possible answers to my initial question about agroforestry practices.

July 2008

A chance conversation with an English-speaking college teacher I met on the streets of Kitale helped me clarify my purpose for this study, in situ. At first she acted defensive when I described the project, as if she wondered what business did I as a foreigner have doing this kind of research that obviously had rich cultural nuances? We gradually found common ground when she spoke of how much she enjoyed using her gas cooker. Using the gas cooker allowed her to relax, spend more time with her husband and children, and be involved in other activities like church and a job. I explained to her that this was why I was interested in knowing why some women want or are able to achieve that level of freedom in their lifestyles by choosing to use a certain stoves and others don't or aren't. I explained that the research was dealing with how to be able to give women the options to make a choice and not be bound to one by tradition or lack of exposure and enabling people to make decisions that result in benefits for themselves, their families and their communities.

2

CHAPTER ONE

INTRODUCTION TO THE STUDY

This study investigated the factors that influenced people to use fuel-efficient stoves (FES) in rural Trans Nzoia District, Kenya. It sought to understand why certain stoves were used in preference to others, what factors influenced their popularity, and why some people failed to use FES at all. The research involved visiting and interviewing a range of rural residents who differed in culture, age, and sex as well as local administrators and employees of the Ministry of Agriculture who were actively promoting certain FES models. It was conducted in partnership with SCC-Vi

Agroforestry, a Swedish based non-governmental organisation who has been working in

Trans Nzoia District for over twenty-five years.

SCC-Vi Agroforestry requested that a qualitative study investigating FES use be undertaken as part of their objective to reduce the pressure of unsustainable wood harvesting in the project areas. The use of FES in Trans Nzoia was lower than expected despite numerous and varied attempts to promote them. Extension workers in the area suspected this was due to unidentified cultural and gender related reasons, and they wished to know what the reasons were.

The study aims to respond to that request by providing perspectives primarily from rural residents themselves. The results are meant to inform and influence future extension efforts, and inform future revisions of program and policy for government and non-government organisations that are active in technology promotion. While examples are specific to portions of one district in Kenya, the results and implications can be broadly applied to technology promotion efforts all over the world.

3

Background

Can the type of cook stove a household uses impact the environment they live in? In Kenya this is arguably the case, where biomass fuels provide 98% of all energy used in rural areas (Mugo, 1999), which are home to 60% of the country‟s population

(FAO, 2007). Such widespread use of biomass fuels, combined with high rates of deforestation from clearing land for arable agriculture, has led to a deficit in biomass production in that country (Kgathi, Hall, Hategeka, & Sekhwela, 1997; Mugo, 1999).

Both the Kenyan government and a myriad of non-governmental organisations (NGOs) actively promote more efficient production and usage of biomass energy (Kenya

Ministry of Energy, 2003; Simms & Reid, 2005)

Fuel-efficient stoves have been celebrated as a major part of the solution to decrease energy deficits in developing nations, as well as decrease health, social and local environmental problems. Affordable, fuel-efficient stoves have been promoted for decades but have met with mixed success in terms of extent of usage. Despite presumed advantages, community members are not using the fuel-efficient technology as readily as local non-governmental organisations expected (SCC-Vi, 2008).

Rationale for the Study

Numerous studies have documented a lack of gender sensitivity in energy programming. Programs related to using or producing wood fuel were typically considered women‟s projects, while most other energy programs were gender neutral, or had no reference to gender (Aravjo & Quesada-Aguilar, 2007; Simms & Reid, 2005;

Skutsch, 1997). To effectively promote the usage of energy-related technology, it is imperative that researchers and extension agents understand the cultural, social, gendered and practical implications of adopting that new technology for both sexes

4

(Kgathi et al., 1997; Skutsch, 1997). Extension agents suspected there were socio- cultural, gender-related and practical reasons why the stoves they promoted were not being used. This study is relevant because it will identify those reasons.

Goal

To understand factors that influence fuel-efficient stove (FES) use in rural households in Trans-Nzoia district of Rift Valley province, Kenya.

Objectives

1. To determine the role of the traditional three stone stove within a rural

household1

2. To identify practical and strategic gender needs that currently exist for fuel-

efficient stoves

3. To identify current perceived need for fuel-efficient stoves within the

community

4. To identify gendered perceptions around the use of fuel-efficient stoves

5. To characterise existing and potential community level training or involvement

for FES

Limitations

Due to limitations in resources and time this study involved a diverse but small selection of participants in the research process. Recent violence restricted the areas available to visit, and time was not sufficient to allow for multiple iterations of research to explore each avenue of research thoroughly. Most specific findings of this study arose from conversations in eight villages and several government offices scattered throughout the district. The significant variation between villages and their

1 Refer to Appendix 1 for a matrix outlining research questions relating to the objectives

5

occupants meant that the selection process was by no means exhaustive in terms of finding a representative of each „typical‟ household situation. The divergent perceptual experiences of participants were analysed to generate interpretations of patterns and relationships, but the very act of interpretation necessitated the introduction of the interpreter‟s bias. As such, the strength of this study lies in providing general factors affecting FES use and guidance for future areas of research and lines of questioning for extension agents rather than generating predictive or extrapolative results. The qualitative data were not numerous or representative enough to generate statistically significant conclusions even at a village level regarding FES use.

Significance

In spite of the above-mentioned limitations, this study successfully met its goal and objectives. The study was significant to undertake for many reasons. Although traditional stoves were said to be inferior to improved models, this could only be in terms of energy used for cooking. Knowing the current role of traditional stoves in a household (objective 1) helps extension agents understand under what circumstances community members would consider using a new style of stove. For “anyone wishing to understand how and why people may want to switch to stoves, must first understand how and why people cook over fires, as well as what advantages fires have over stoves” (Foley, Moss, & Timberlake, 1984, p. 10).

Is it possible that a stove holds a meaning beyond merely being a tool to cook with? Identifying the practical and strategic gender needs within a household

(objective 2) shows the significance of having a stove in a household. Knowing specifically how a stove fulfills or fails to fulfill practical and strategic gender needs is useful for tailoring dissemination, education, and stove engineering programs. This

6

objective also reveals who is responsible or has the right to make decisions, under what circumstances, and why.

Knowing the perceptions various stakeholders within a community or household hold regarding an FES (objective 3) are imperative to discern so that any invisible barriers (those not addressed by tangible things like stove design or program and policy structure) can be identified and addressed. Finally, determining what the perceived need for fuel-efficient stoves is in a community (objective 4) is crucial to know how to direct any current or potential community level training efforts

(objective 5) and avoid failure. These are key pieces of information for a project that has the potential to bring users immediate significant social, health, and environmental benefits.

Overview of the Thesis

This thesis is organised into seven distinct chapters. Chapter Two presents a review of literature on the traditional three-stone stove, the importance of conserving fuel, the history of fuel-efficient stove promotion, and the interplay of energy programming and policies in East Africa. It ends with a list of gaps in literature that this study sought to fill and the conceptual framework on which the investigation was based.

Chapter Three describes and defends the methodological and analytical approach applied to this exploratory study. The chapter outlines the range of qualitative research methods employed for data collection including semi-structured interviews, focus groups, key-informant interviews, literature review and participant observation. It also describes how and why each participant was chosen to take part in the study.

7

Chapter Four describes the context for the study. It introduces elements of local history, present ethnic makeup of the area, food preferences, geography, history of fuel-efficient stove promotion, and descriptions of the study communities that are relevant for interpreting the data and discussions that follow.

Chapter Five presents the research findings organised according to objective.

This chapter is rich in narrative and is presented with many charts and graphs to aid in comprehension. Chapter Six discusses the findings using a conceptual framework that arose as a result of data compilation. It also relates the findings to elements of the literature review.

Finally, Chapter Seven summarises the findings and delivers the main conclusions drawn from this study. Recommendations are listed and suggestions for further research are included.

8

CHAPTER TWO

REVIEW OF LITERATURE

Introduction

This review of relevant literature covers several topics establishing the usefulness of this study. Part one outlines fuel use in Kenya, including information on the traditional open fire cooking system. Following that is an overview of various arguments for why promotion of fuel-efficient stoves (FES) is important in the East

African context. Part two highlights the history of FES design, current models and the history of household energy technology promotion. Part three looks at the paradigm shift from women in development through to gender and development and the relevance of that shift to energy policies and programs. It will also briefly describe the status of gender within energy technology promotion in the global south. The chapter ends by outlining the gaps that exist in literature, and introducing a conceptual framework that guided the investigation.

Part one: Biomass Fuel and Stoves

Biomass Fuel

The history of bioenergy2 is as long as humanity itself. Wood and other bioenergy resources were the main source of energy for humans since the Stone Age, only being replaced by more „modern‟ forms of energy relatively recently (FAO, 2007).

They remained the primary source of energy for heat and cooking in Europe and North

America until the middle of the 19th century. Steadily replaced by more efficient,

2 Bioenergy can be defined as „energy obtained from biological and renewable sources (biomass), normally in the form of purpose-grown energy crops or by-products from agriculture, forestry or fisheries‟ (FAO, 2007, p. 1). Bioenergy resources include: fuelwood, charcoal, sugar cane bagasse, maize, sweet sorghum stocks, livestock manure, biogas, microbial biomass and algae (FAO, 2007).

9

convenient energy sources (coal, oil, gas, electricity), it was not until the early 1900‟s that wood lost prominence as the principal energy source in rural areas of most developed countries. The decline continued rapidly and continuously in industrialised countries, however biomass fuels remain the most significant source of energy in the rural household sector of less industrialised areas of the world to this day (Arnold,

Köhlin, Persson, & Shepherd, 2003).

Worldwide, 7% of energy used is from renewable resources, mainly in the form of hydro power and biomass energy. The developing world uses 70% of the renewable energy, mainly in the form of biomass, for cooking and heating in the household and at cottage industry levels (FAO, 2007; Karekezi & Kithyoma, 2002). In most developing countries, less than 5% of the population use modern forms of energy. In 2000, this figure in Kenya was 3% (Kenya Ministry of Energy, 2003). The wide spread nature of rural settlements filled with the region‟s poor make the connection to and building of a conventional power grid much too costly, thus Sub Saharan Africa is the least electrified region in the world. Additionally, cost and logistics of dissemination of fuel and technology are barriers in rural areas to „modern‟ energy sources like liquefied petroleum gas (LPG) and kerosene due to unreliable roads and increasing transportation costs (Clancy, Skutsch, & Batchelor, 2003; Karekezi & Kithyoma, 2002).

Traditional Stove

Cooking as a passion and a pastime is a position held by a privileged minority of people worldwide. Over half the world, primarily in the less- economically developed south (Africa, South America and Asia), still relies on biomass fuels burned on open fires or rudimentary stoves to provide the bulk of their household energy needs. For the cooks in those places (primarily women and children), activities related to cooking food represent a chore and potentially a threat to their lives (Rehfuess, 2006). The

10

most rudimentary stove, commonly referred to as the three-stone fire or three-stone stove (TSS), is made of a simple cooking fire surrounded by three or more stones, bricks, mounds of mud or other fireproof material. Stones, or sometimes metal bars, support a pot which sits partly down between them. Styles differ in each country and region; Ghana, Guatemala, India, Fiji, Senegal, Ecuador and Kenya would all have distinct differences in their construction (Foley et al., 1984).

Advantages

The traditional fire has many practical and strategic advantages, which explains its ubiquity and existence for millennia (Foley et al., 1984). Mahiri and

Howorth (2001) exhort us to keep in mind that:

“the multiple/simultaneous functions of open fires have largely been ignored. If not only cooking, but light, space heating, and insect control are included, the efficiency of open fires becomes greater. Efficiency is more social than physical, and should be defined by end-user rather than end-use” (p. ).

The heat, smoke and light produced by the fire are beneficial if the fire is built inside a small cooking hut, as is commonly done in Kenya, because dawn and dusk can be cold, smoke keeps insects out of dried foods stored in a compartment over the fire, and the light gives some respite to the darkness of the hut or illuminates the night

(Mahiri & Howorth, 2001).

A TSS is free, portable, doesn‟t require special tools or skills to build, does not require repairs, it is able to accommodate any size, shape and type of fuel, and has capacity for any size of pot. An adept user can move the stones closer together or further apart to decrease or increase the heat reaching the pot, and can juggle fuel to provide the right type and amount of flame for cooking (Foley et al., 1984). Some people like the smell of smoke and believe it adds to the flavour of foods (FAO,

2006b). Additionally, the secure base made by stones is crucial for the prolonged and

11

vigorous stirring needed to prepare thick porridges that many countries, including

Kenya, use as a staple food (Foley et al., 1984).

Tradition

Integral components of households, such as the family stove, can often be steeped in cultural tradition, making their disuse unfathomable. For instance, some tribes of Kenya have traditions regarding how disputes are settled between men and their wives. Following a large disagreement, a wife could take with her a stone from the fire from where she traditionally cooked the family‟s meals as she left the household of her husband. With one of the three stones gone from the family fire, food would not be cooked until the husband reconciled with his estranged wife and convinced her to return to reconstruct the three-stone stove (C. B. Kariuki, personal communication, February 2004).

Disadvantages

While the light, heat, and smoke these traditional fires produce has many advantages outlined above, they have many disadvantages as well.

Cost in Fuel

The three-stone fire is often accused of being inherently wasteful of energy for cooking purposes. The main reason for this is due to its construction. The loose placement of stones around the base of the fire permits heat to radiate laterally, allowing only a small amount to be transferred to a cooking vessel. When wind is introduced, cooking efficiency drops further as there is a simultaneous increase in both heat production and amount of heat that fails to be intercepted by cooking pots.

This translates to TSS typically only achieving 7-15% efficiency, compared to simple

FES designs, which can easily double that efficiency in the field (Foley et al., 1984).

12

In places where fuel is purchased, household or cottage industry expenditures are significant. Inefficient methods of fuel usage unnecessarily increase money spent on procuring fuel (Dutta, Matinga, Panjwani, & Cecelski, 2005; Rehfuess, 2006).

Cost in Time

Authors have recorded that women and children often walk long distances

(Chikoko, 2003) or spend long hours each day collecting fuel, actions which they term drudgery. They argue the work of procuring fuelwood is deleterious to a woman‟s quality of life, and that her time could be better spent doing other productive or reproductive work (Dutta et al., 2005; Rehfuess, 2006).

Cost in Health

Death due to complications from smoke inhalation is the fourth leading cause of death among women and children in sub-Saharan Africa; the majority of which is produced from tending fires in enclosed cook-houses (Rehfuess, 2006). It is estimated that 2 390 000 people die from smoke- related health complications annually in Africa

(Ezzati et al., 2002). When biomass is not completely combusted, pollutants are released into the air, visibly as smoke and soot, as well as invisibly as gases or fine particles. Typical 24-hour levels of small particles3 in a cooking hut in Asia, Africa, or

South America are 300-3000 g/m3, with peaks during cooking time up to 10 000

g/m3. For comparison, the US environmental protection agency says a standard for

3 outdoor PM10 levels is 50 g, while the European Union sets the standard at 40 g/m .

Fine particles (less than 2.5 microns in diameter, PM2.5) are even more damaging, causing inflammation of airways, lungs and impairing immune responses (Rehfuess,

2006).

3 3 The amount of small particles (less than 10microns in diameter, PM10) per m is the most widely used measure of indoor air pollution when determining health hazards.

13

When children are small, African mothers typically carry them around strapped to their backs (World Bank, Barnes, Openshaw, & Smith, 1994). Whatever pollutants the mother is exposed to, the baby is also. Inhaling this cocktail of harmful pollutants

(primarily carbon monoxide and small particles, but also including nitrogen oxides, benzene, butadiene, formaldehyde and polyaromic hydrocarbons among other damaging chemicals) is equivalent to (Rehfuess, 2006) or significantly more damaging than (World Bank et al., 1994) the damage done by smoking two packages of cigarettes a day. There is strong evidence of correlations between exposure to indoor air pollution and acute lower respiratory infections, chronic obstructive pulmonary disease, asthma, lung cancer, cataracts, and tuberculosis (Sagar & Kartha, 2007).

Epidemiological studies conducted since the 1980‟s have shown that exposure to indoor air pollution due to the incomplete combustion of solid fuels doubles the risk of children under the age of five to develop pneumonia and acute infections of the lower respiratory tract and triples the likelihood of women to suffer from chronic obstructive pulmonary disease (i.e. bronchitis or emphysema) (Rehfuess, 2006; World

Bank et al., 1994).

Women in some areas are vulnerable to sexual exploitation by men when they are out gathering fuel; either from roving attackers or real or supposed officials in government-owned woodlots. They also have an increased risk of suffering from snake bites during collection trips and are more prone to developing back problems from carrying heavy loads of firewood (Dutta et al., 2005; Komuhangi, 2006).

In some areas, lack of fuelwood has meant a reduction in the number of meals cooked in a day and nutritious options like dry-preserved protein-rich beans are overlooked in favour of foods that require less cooking time (Foley et al., 1984;

Komuhangi, 2006). Additionally, anecdotal evidence reveals that people will decrease

14

the amount of water that is boiled (leading to an increase of water-borne illness and decreased sanitation), decrease the cooking time of foods (this decreases the digestibility and ability for nutrient absorption of some foods, which affects children disproportionately), and adjust the amount of food that is cooked in a household, or the number of mealtimes, sometimes down to once a day (Drigo, 2006).

Impacts from the HIV/AIDS pandemic have been acutely felt by many families in Africa. In the western part of Kenya including Trans-Nzoia district, infection rates average at 30%, leaving no family untouched (Chianu et al., 2008). The battle against

HIV/AIDS requires effective treatment and prevention. Availability and type of household energy plays an important role in both those areas (Rehfuess, 2006).

Proper care of persons living with HIV/AIDS includes frequent heating of water for sterilization of instruments, preparation of food, hot compresses and bathing, all of which mean the amount of energy consumed in a household with a sick person is greater than one without. In order to avoid developing the effects of full-blown AIDS, one must combine drug treatment with proper nutrition and maintain energy and fitness levels (Dutta et al., 2005; Rehfuess, 2006). Adequate energy availability “is likely to be important in reducing women‟s burden of care for persons living with

HIV/AIDS and in enabling self care for women living with HIV/AIDS” (Dutta et al., 2005, p.8).

Immunocompromised persons infected with HIV are more prone to opportunistic diseases such as tuberculosis and other smoke-related infections.

Another repercussion of such a high number of people dying from secondary diseases due to HIV infection is that there are an increasing number of child-headed households. Gathering fuel is a major household task, so with no alternative of help,

15

children must leave their academic endeavours to make lengthy treks for biomass fuel

(Dutta et al., 2005).

Cost to the Environment

Harvesting woody biomass for rural household use is not inherently bad or unsustainable, but when demand surpasses supply, as often happens close to urban areas or due to charcoal production, trees will be cut down for the express purpose of providing fuel. Some people argue that burning animal dung and crop residues as alternates to wood fuel robs agricultural land of potential nutrients, making high usage of those an environmental cost as well (World Bank et al., 1994).

In times of wood scarcity, crop residues and animal dung are used as fuel, which removes the potential to use them as non-chemical soil nutrient and biomass enhancers. Leaving crop residues on fields decreases erosion that in turn protects waterways from siltation. Additionally, the breakdown of biomass by soil microorganisms allows nutrient cycling and builds the soil structure, which are both essential for good soil health.

Complex ecosystems serve as the pharmacy, water company, power company, grocery store, and supplier of building materials to subsistence level communities all over the world (Rehfuess, 2006). What would happen if these services were removed from their neighbourhood? There has been a growing concern surrounding the impacts of deforestation felt globally, and the role wood fuel collection has played in it. The actual influence of “woodfuel collection on forests has been controversial and its role in rural livelihoods and deforestation the subject of considerable debate”

(Arnold et al., 2003). The next section outlines historical and current theories on the impact of wood fuel collection on ecosystem health.

16

History of the Wood Fuel Crisis

Over 100 years ago severe fuelwood shortages were reported in central and western Kenya (Mugo, 1999), however interest in woodfuel didn‟t start to gain ground until the 1970‟s. Initial efforts were directed towards the “perceived, potentially devastating effects of escalating fuel demand on forest resources” (Arnold et al.,

2003) and the accompanying negative socio-economic consequences that were predicted for the rural poor (Arnold et al., 2003). A different idea was proposed in

1979, during the OPEC oil price rise which impacted Sub-Saharan African economies, which prompted a closer look at power sources. Upon closer examination, it seemed that forests were being cleared for agriculture, not fuelwood. Fuelwood was harvested primarily from trees outside and only secondarily from within forests (Foley et al., 1984; Mahiri & Howorth, 2001).

A fire had already been lit though. The initial panic of the 1970‟s had spawned the Woodfuel Gap theory, which raised alarms about a looming fuelwood crisis.

Founded on the projection of fuelwood consumption based on population increases, the theory anticipated serious discrepancies between supply and demand (Arnold et al., 2003; Mahiri & Howorth, 2001). For instance, in 1980 Kenya reported a national fuelwood deficit of 30%, and projected that it would increase to 74% by 1995 and 89% by 2000 (Mugo, 1999). The theory assumed that the shortfall (gap) would be met by overharvesting forest resources, leaving the country to spiral toward further forest decline (Arnold et al., 2003). Woodfuel issues went on to attract major attention from funders, who started multiple programs aimed at increasing wood supply through wood planting programs and emphasising efficient use of wood-based fuels (i.e. using fuel- efficient stoves), many of which are still in existence today (Arnold et al., 2003;

Mahiri & Howorth, 2001).

17

A related approach to viewing the deficit problem is that of Woodfuel

Orthodoxy. It is simplistic in its linear association of deforestation or lack of trees directly with woodfuel consumption because it was based on prevailing theories such as the Malthusian correlation with population dynamics or Von Thunen‟s concentric zones of fuel exploitation. These approaches “merely seek out generalised relations among various empirical objects and events themselves, and not abstractions about what produced them” (Mahiri & Howorth, 2001). Solutions under this model are frequently promoted by top-down planning institutions and call for increasing plantations of trees and farm forestry, without taking into account complex issues surrounding land tenure and who has access to new biomass resources (Mahiri &

Howorth, 2001).

In the late 1980‟s and early 90‟s interventions were assessed, and the focus shifted from availability of fuel to gender issues and integrated land use approaches

(Arnold et al., 2003; Mahiri & Howorth, 2001). A systematic survey conducted in 1982 showed that there were many sources of fuel outside the forest, including twigs and branches from hedgerows and fallen trees and wood residues from other uses in the rural economy. Analysis of aerial photos from 1967 and 1986 presented evidence for a general shift of woody vegetation from bushland to be concentrated by settlements.

In some cases, woody biomass actually increased closer to concentrations of people

(Mahiri & Howorth, 2001). This analysis helped to reinforce the view that demand for fuel is unlikely to deplete forest cover on a large scale (Arnold et al. 2003).

Ironically, although biomass increased overall in the landscape during this time, there was a marked decrease in access to wood, especially for women. Traditionally fuel was treated as a „free good‟ and was readily available and within reach for rural communities in Kenya. In recent times registration of land and its transfer to private

18

ownership eliminated traditionally free and accessible communal sources of wood, making the prospect of free wood collection much more difficult for land-poor rural persons (Arnold et al., 2003; Mahiri & Howorth, 2001). Indeed,

“[a] complex mosaic of user rights embedded within the household structure such as access rights, title rights and agricultural rights . . . coupled with the cultural overlay of rights, which determine access to and end-use of wood resources… create[d] paradoxically huge „scarcity‟ differentials between individual households in villages with a seeming „abundance‟ of wood” (Mahiri & Howorth, 2001, p. 212).

For instance, some communities have cultural restrictions dictating how and if a woman can plant or cut from a tree on her farmstead (Mahiri & Howorth, 2001).

From the 1990s until present, government expenditures on agricultural and forestry extension has decreased, accompanied by an increasing emphasis on cross- sectoral programs and partnerships with private and voluntary NGO sectors. The emphasis is to be decentralisation, encourage local development and use participatory approaches. This approach appreciates the range of social factors that influenced economic and energy poverty4 (Mahiri & Howorth, 2001).

Present Thoughts on Environmental Impacts

Even with the evidence presented above, there exists a difference of opinions within current literature on the ecological or environmental impact of biomass fuel procurement within a Sub-Saharan African context. Literature is moving towards the opinion that harvesting of fuel for household use is not inherently bad, if done in a sustainable manner where demand is less than supply (Dutta et al., 2005; Mahiri &

Howorth, 2001; Rehfuess, 2006). Trade in biomass fuels like wood and charcoal are critical income for millions of people on the continent, and many authors argue they are not to blame for deforestation nearly as much as previously thought. It is agreed

4 Energy poverty is defined as „the absence of sufficient choice in accessing adequate, affordable, reliable, high quality, safe, environmentally benign energy services to support economic and human development‟ (Clancy et al., 2003, p. 5).

19

that a fuel scarcity exists; however one must turn to the hunger and demand for agricultural land to find the root cause for deforestation (Arnold et al., 2003; Foley et al., 1984). It seems that all countries with major fuelwood problems also have major pressure from growing populations, who view forested land simply as unused agriculture potential (Foley et al., 1984). Arnold et al. (2003) argue that fuelwood supplies are drawn from a much larger base than just forests, and that users have a range of responses to adapt to a changing availability of fuelwood.

Mahiri and Howorth (2001) agree that in rural areas, “fuelwood is usually a by- product or part product of trees which have a different primary purpose from that of fuel” (p. 212). Essentially, they say, the fuelwood problem is more of a “land use problem that can be addressed as an energy problem” (p. 212). Around urban areas and in rural areas where people do not have access to land, population pressure can cause indiscriminate felling of trees, whereas those in rural areas with access to land have full capacity to increase wood production (Mahiri & Howorth, 2001).

Notable organisations and authors who appear to attribute deforestation on fuel gathering include the FAO who stated in their 2005 Global Forest Resources

Assessment: “Africa reported a steady increase [in wood removals in recent years]. It is estimated that nearly half of the removed wood was woodfuels. Informally or illegally removed wood, especially woodfuel, is not usually recorded, so the actual amount of wood removals is undoubtedly higher” (p. 5). The International Energy

Agency, an agency founded during the oil crisis of the 1970‟s and “acts as a policy advisor to 27 countries to ensure reliable, affordable, clean energy to citizens”

(International Energy Agency, 2008) stated in a 2007 publication that biomass energy itself was not a concern, but rather it‟s unsustainable harvest and inefficient use

(Birol, 2007). The Government of Kenya (2005) maintains that “the greatest threat to

20

[their] environment is posed by poverty”, which is manifested in part by “the high and increasing demand for wood fuel [which] poses a major threat to existing forest and other terrestrial ecosystem resources” (p. 24). A Tanzanian environmental development organization cited reducing fuelwood consumption as a major way to keep their environment sustainable (TIPTZ, 2008).

While the harvest of unprocessed biomass fuel, such as that commonly used in rural households is not a driver of deforestation, the harvest of charcoal specifically, a fuel commonly used by urban dwellers, can be quite unsustainable (Simms & Reid,

2005). Others qualify their statements by saying that although ecosystems have an astounding capacity to accommodate increasing needs, it‟s foundation of existence is threatened by over-exploitation of its resources by growing populations. Where populations are dense and trees are scarce, environmental problems do sometimes arise (Drigo, 2006; Rehfuess, 2006).

Part Two: History of Fuel-efficient Stoves and Their Promotion

As discussed briefly earlier in this chapter, the traditional TSS has been criticised for being inefficient in terms of fuel use because it produces excess smoke and does not direct much heat to the pot surface (Foley et al., 1984). Fuel efficiency is also called combustion efficiency. As the combustion efficiency of a stove increases, the production of heat increases and smoke decreases from the same amount of wood (FAO, 2006a). The goal then of improved stoves is to increase the combustion efficiency and also to direct heat towards the pot instead of losing it laterally. Many cultures have developed stove styles independently from each other, which seek to increase combustion efficiency in some way. The first step in the evolution of improved stoves was avoiding lateral heat loss. By building the fire in a depression in the ground, and placing more, smaller stones closer together, the gaps

21

filled in, and the pot could be fitted more tightly above the flames. The fire was also more protected from wind (Foley et al., 1984).

The Woodburning Stove Group at Eindhoven University in the Netherlands classified traditional improved stove models in the early 1980‟s. They divided them into three types: the shielded lightweight, the shielded heavyweight, and the closed heavyweight. For them, any effort to enclose or shield an open fire broadly defines a stove in their terms. The lightweight varieties were often made of scrap metal shaped into squat cylinders or rectangular prisms by local tinsmiths and were able to burn a variety of fuels (charcoal, fuelwood, rice husks, or sawdust). Local potters cast slightly more hefty ones in ceramic moulds. They lasted slightly longer than the one- two year lifespan of the metal ones. Yet more stoves were build directly into the ground, fashioned by artisans from bricks made of mud, sand or clay stabilised with wire, metal, straw or dung (Foley et al., 1984).

Successes and Failures in Promotion of Household Energy Technologies

Development of efficient and convenient woodstoves occurred in industrialised countries with no assistance, however in non-industrialised countries, there was very little spontaneous indigenous development. A generally low income prevented people from actively investing in improved stoves, but governments were keen to promote them because they feel their growing populations were putting too much of a strain on the resource base (World Bank et al., 1994). Stove development and dissemination started in the early 1980‟s and proceeded in earnest by the individual and collaborative work of several organisations following the 1980 United Nations conference on New and Renewable Sources of Energy (Okello 2005, World Bank et al.,

1994).

22

In Kenya, the organisations involved in FES at that time were the newly created

Kenya Ministry of Energy, the Appropriate Technology Centre, the Kenya Energy and

Environment Organisation, United Nations Children‟s Fund, Maendeleo ya Wanawake,

CARE-Kenya, the Intermediate Technology Development Group (ITDG) and the German

Technical Cooperation organisation (GTZ) (Okello 2005). The Woodfuel Gap theory was popular so NGOs, donors and governments instigated interventions because it was assumed local innovations would not keep pace with the rapid change in biomass demand from increasing population pressure. Acting out of a desire to prevent or mitigate forest deforestation, and recognising that it was impossible for people to adopt more modern but also more costly energy alternatives, multiple stove programs were started around the world. Researchers came up with amazing results of stove models in the lab. Initial efforts showed that improved stoves increased efficiency by three to six times that of a TSS, which was believed to have combustion efficiency between 5-10%. The successes (or lack thereof) of these programs now serve as valuable sources of lessons, tips and tricks when promoting stove use (World Bank et al., 1994).

Interest in biomass energy has renewed in the last ten years with the crafting of the Millennium Development Goals. Although there are no goals dealing explicitly with energy, it can be seen as an element of all eight. The gender-energy-poverty nexus (Clancy et al., 2003) is the current focus, with environmental effects of wood fuel harvesting taking a back seat. Part three of this paper discusses these concepts further.

Diffusion of Innovation Theory

Everett Rogers (1983) proposed five characteristics of technologies or innovations that help explain different rates of adoption. These characteristics

23

referred to those perceived by the potential adopter, not an absolute value. The first characteristic is the relative advantage of the technology. This can be defined as the relationship between price and value; a high performance merits a high price, but the performance must be perceived to be significantly higher than the thing it replaces.

Secondly, the technology must be compatible with things the person already has. In this case, that would mean fuel and physical space. Third, it must not be too complex. As complexity of a technology increases, an average consumer is less likely to adopt it until they have had ample support from friends to help them navigate the complexity. Fourth, a person is much more likely to use a technology if they can use it first. This concept of trialability is based on the premise that if a person can not physically try a new technology, they are unsure if the value of it is good. Finally, adoption happens much more quickly when the results of a technology can be seen in practice, observed, and communicated about effectively. The easier it is to see the results of a technology, the more likely people are to adopt it.

Rogers (1983) further proposes that each person responds to these five characteristics differently and falls beneath a bell-shaped curve divided into five categories. The first group of people to adopt a technology is small (2.5% of the population). These innovators are generally younger, more liberal, more educated, have more disposable income and are highly socially mobile, enjoying talking about trends and taking risks. They are also often too excited about an innovation to give a balanced assessment of a technology. The second group of people called early adopters make up about 13.5% of the population. They are more likely to perceive risks associated with a technology and do research prior to purchase, so consequently are able to offer a balanced assessment and are more likely to act as opinion leaders for non-adopters.

24

The third and fourth groups are most of the population (34% each). They are the early majority and late majority. These groups are pragmatic; more interested in the applications of a technology, and need research and concrete reasons to purchase it. People in the early majority are ready to buy something when they find a strong match between application needs and a technological solution, whereas the late majority will wait until a technology is a well established standard, and need assistance to buy it, often from well-established companies. The final category (16% of the population) is non-adopters, or laggards. They are generally older and want nothing to do with new technology. They will buy a technology only if it is hidden within another product that they are already using. Promotion programs should not bother about pursuing this final category of people (Rogers, 1983).

A tipping point occurs when competing technologies go out of business, or in the case of stove technology, become obsolete. Rogers (1983) says that the take-off point occurs once the influential early adopters begin spreading the technology to their peers. These early adopters are who technology promotion programs should be aimed at. They are perceived to be credible, willing to innovate, self-confident with new technologies, and are often sought for information, advice or opinions by their peers.

Facilitating Factors for Successful Promotion and Use of FES

Despite Rogers‟ (1983) informative work, there are no hard and fast rules or magic formulas to unlock the mystery of what makes stove dissemination programs work or not (stove dissemination programs are considered to be successful if stoves are bought and then actively used). Available literature suggests the following

25

patterns which have been revealed through more than two decades of promotion efforts 5:

1. FES are more apt to be used in places where fuel is purchased than in places

where it is not. Savings in money seems to be the primary factor in

determining if a stove is used or not. Spending money on smaller quantities of

fuel relieves strains on household budgets (Rehfuess, 2006).

2. FES are more apt to be used in households where the primary stove user‟s time

is valued. When the opportunity cost of women‟s labour activities is high,

there is more incentive to save time, in order to spend more time in generating

income. Even if overall household income is raised but a woman‟s income

remains constant, there is no incentive to adopt a labour-saving stove (Kelkar &

Nathan, 2002; World Bank et al., 1994).

3. Long term stove programs are more successful than programs that have not

been established for long (World Bank et al., 1994).

4. The best programs incorporate significant interactions between those who

design, produce, and use the stoves. Women will not use or maintain a stove if

it does not fit their needs, so it is imperative stove designers have input from

the intended users (Rehfuess, 2006; World Bank et al., 1994).

5. Subsidies are often positive influencers of stove adoption (Kelkar & Nathan,

2002); however they only work if they are very small. It is better not to use

them at all (World Bank et al., 1994). It is preferable to offer a range of styles

and choice of interventions, allowing the user to choose based on differing

demands and abilities to pay (Rehfuess, 2006).

5 Refer to Appendix 3 for a more detailed list of possible reasons for success and failure of stove programs.

26

6. Targeting a user group most likely to benefit from (and consequently adopt) a

FES is more strategic than general promotion (World Bank et al., 1994).

7. Stoves are more likely to be successful when the stoves are easily and locally

manufactured (World Bank et al., 1994).

Stoves that are properly designed, meet the needs of their users and are adapted to their environment (Foley et al., 1984) as well as achieve a balance between perceived and social benefits (which depend both on the stove and the cooking customs of the users) are hallmarks of effective stove promotion programs (World Bank et al., 1994). Gerard Foley and others (1984) further emphasise that “the future justification of stove programmes has to be found in their capacity to deliver such direct benefits to the people who will use them. If they don‟t, there is really no point in bothering” (p. 15).

Selected FES Models in East Africa Today

The Woodfuel Gap theory postulated in the 1970‟s and 80‟s instigated a flurry of research into and dissemination of FES. Some models used biomass fuel as an energy source, while others, perhaps more aptly named cookers rather than stoves, used the stored heat from food and heavy insulation to complete the cooking process (basket cookers) or the sun‟s rays to cook food (solar ovens). Some stoves were moveable, like the popular Kenya Ceramic Jiko which is a more efficient way to cook with charcoal, while others are built into a wall, like the Rocket Lorena stove.

FES are available in a variety of styles, reliabilities, efficiencies, and prices. An efficient stove is one that has a good combination of complete burning of fuels and good heat transfer from flame to pot (Komuhangi, 2006). In general, fuel-efficient stoves have the following features in common (FAO, 2006a):

27

1. High combustive efficiency: it is possible to achieve over 90% efficiencies by

controlling the type of fuel used, the air and flame mix, and the burn intensity.

2. Reduction in wood consumption: efficient stoves use 40-60% less firewood for

the same task. Two pot stoves further cut down on fuel use compared to single

pot stoves.

3. Good transference of heat from the fire to the pot: insulation, shortening the

fire flow path and reducing the distance between the flame and the pot

increases heat transference.

4. Minimal smoke emission: high combustive efficiency and using chimneys to

draw smoke through a chimney may achieve almost 100% reduction in

emissions. The opening to the chimney needs to be designed to draw the

smoke up but at the same time to constrict the flow so that too much heat is

not lost up the chimney.

5. Ergonomics and structural stability: stove layout must be user friendly and

enable pots to be moved around. The stoves must be stable and the pots must

be secure when they are on the stove to prevent tipping and causing burns.

6. Materials: stoves are made from a wide range of materials including pottery-

lined mud, metal or ceramic-lined metal. The inner material must be strong

enough that constant heating and cooling does not cause it to disintegrate too

fast.

28

Table 2.1: Descriptions of selected fuel saving devices found in literature

Fuel-efficient Description Technology Basket cooker, These devices continue cooking foods once they have been brought fireless cooker to a boil by traditional cooking methods by allowing the heat to (SCC-Vi, 2008) or radiate slowly from the cooking vessel. They also keep foods warm insulated for hours. They are constructed from a basket or box lined with an basket/box insulating layer of cloth waste, newspaper or hay, and covered with (FAO, 2006a) a sort of pillow filled with the same materials. They are actively promoted by the Kenyan Ministry of Agriculture (AG1).

Kuni mbili These names refer to a pre-fabricated ceramic liner which is typically made by trained potters (most often women), then fired in a kiln. It was popularised following a partnership established between GTZ and the Kenyan Ministry of Agriculture in the mid- 1990s and is still heavily promoted by the Ministry of Agriculture. The liner lasts for four to ten years, and is supposed to be priced so that an ordinary woman can sell a bunch of bananas or a chicken to procure one. Fuels include agricultural residues (i.e. sugar cane waste) and wood (Karekezi, 2002; Okello, 2005; Simms & Reid, 2005).

Sawdust stove This name refers to stoves made of clay or metal that are used by firmly filling sawdust (which may include wood chips and shavings) around a central core. The central core is removed and a lit piece of kindling is inserted into a port at the bottom of the chamber where it rests in the air cavity left by the central core (Mukunda, Dasappa, Swati, & Shrinivasa, 1993; Thornton, 1984). The metal versions of this stove are portable and made by local jua kali artisans. The clay versions are permanent and usually constructed by the user.

Rocket Lorena The key component of this stove is a combustion chamber that is stove specially designed to ensure a correct fuel/air mix. The stove is properly insulated to minimise heat loss and maintain a maximum temperature within the combustion chamber. Heat transfer efficiency is maximised because the cooking vessel sits inside a shielded fire chamber (Komuhangi, 2006). A stove similar to this is now being promoted by GTZ near Kakamega, in a province neighbouring the study area (AG12).

29

Successful Stove Promotion Programs

The first two examples described below are of programs run by community or other non-governmental groups; however the literature suggests that the most successful stove promotion programs were government-run programs that limited government involvement to assistance in technical advice, dissemination and quality control. Programs with the greatest success were ones where the government is not involved with the construction or sale, but does have a long-term commitment to project continuity.

Community Based Extension: The Case of the Kuni Mbili Stove in Kenya

In order to reduce significant pressure fuelwood harvesting caused in their forest (one of Kenya‟s only rainforests, and a national park), the Kakamega Integrated

Conservation Project, in partnership with ITDG introduced acceptable and affordable fuel-efficient stoves to communities adjacent to the forest. By the end of 2001, 4000 kuni mbili („two sticks of firewood‟ in Kiswahili) liners had been made and installed into homes by local artisans as well as ten larger stoves installed into local institutions

(schools, hotels, hospitals). This kuni mbili stove, also known as upesi („fast‟ in

Kiswahili) or maendeleo („progress‟ in Kiswahili) is made of clay and fired in a kiln. It was able to accommodate agricultural residues (i.e. sugar cane waste) as well as wood. It reduced the load of fuelwood used to half, reduced smoke and carbon emissions, and the program provided employment to community members (Simms &

Reid, 2005). The Kakamega project was a spin-off of earlier projects ITDG sponsored together with GTZ and the Kenya Ministry of Agriculture home extension officers throughout the 1980‟s. Okello (2005) described it as an evolution of a subsidised project to one run primarily by local women‟s groups without outside monetary aid.

30

Non-Government Organisations: The Case of GTZ and Rocket Lorena Stoves in Uganda

When the residents of a fuelwood-strapped region of Western Uganda were approached by the EAP (Energy Advisory Project) of GTZ through local NGOs to take part in a program promoting the construction and use of the Rocket Lorena FES, they enthusiastically accepted. Rocket Lorena stoves saved 50-70% of energy compared to

TSS, and were made with free or cheap locally available materials. In a region where small farm sizes make growing their own fuel in addition to subsistence agriculture difficult and publicly available fuel as much as ten kilometres away, this program was well positioned for success. The stove dissemination strategy built on the capacity of local NGOs to technically and administratively manage the program. Sub-county coordinators and visiting trainers identified village mobilisers, and every parish was given a proven artisan to train further stove builders. Lasting benefits of this program include the positive social image stove users held in the community, the marketable skills many villagers gained comparable to house construction, and most importantly, the satisfaction of users with their new stoves (Komuhangi, 2006).

Government-Run Programs: The Case of China

China has the largest and arguably most successful fuel-efficient stove program in the world. Between 1982 and 1990, it is estimated that 90% of improved stoves installed world-wide were in China, in more than 120 million rural households (World

Bank et al., 1994). As of 2006, 200 million Chinese households had FES (Rehfuess,

2006; World Bank et al., 1994). The program identified areas of greatest need and concentrated efforts there. Direct contracts between the central government and county administration meant that self-sustaining rural energy companies that manufactured installed and serviced stoves were established to fill local markets

(World Bank et al., 1994). The local artisans would modify general designs for local

31

users, although key components like the internal combustion chamber were produced centrally to ensure quality control. National-level stove engineering challenges created healthy competition between manufacturers and designers and ensure ongoing ingenuity (Rehfuess, 2006; World Bank et al., 1994).

Support in the form of technical training and program implementation was the responsibility of local rural energy offices. Although the government subsidised stove programs by helping producers with construction training, administration and promotion support, the actual price of stoves was not subsidised at all (World Bank et al., 1994). Part of China‟s success was that women were directly involved in income- generating activities, which meant that a savings in time had monetary value to them

(Clancy et al., 2003). The overall result of this program was a shift in societal norms.

Now, there are more improved stoves than non-improved stoves in use in the country

(Rehfuess, 2006).

Government-Run Programs: The Case of India

India had the second largest fuel-efficient stove program in the world; however it did not enjoy the same successes felt by China. In the 1980‟s programs were centrally planned and relied on several layers of bureaucracy, which hindered their progress. The programs were implemented country-wide which meant resources and efforts were not strategically targeted (World Bank et al., 1994). Indian women did not value time savings, and did not appreciate the stove design, which for cultural reasons did not meet the requirements and constraints of their lives (Clancy et al.,

2003). Stoves were promoted by local officials who were simultaneously promoting many other government-instigated initiatives, and were not very effective with this new burden added to their workloads. The stoves in India were heavily subsidised- at

50%, which meant that the stove producers were constructing stoves to please the

32

government rather than tailored to local needs. Later programs were modified to overcome the downfalls of central planning (World Bank et al., 1994).

Government-Run Programs: The Case of Kenya

During the early 1980‟s, following the 1980 United Nations Conference on New and Renewable Sources of Energy, the newly created Kenya Ministry of Energy spearheaded a design process which resulted in the engineering of the Kenya Ceramic

Stove (KCJ). The KCJ is an improved charcoal burning stove which quickly gained acceptance among urban stove producers “who were initially offered free training and marketing support by [the Kenya Energy and Environment Organisation (KENGO)], working with the ministries of Energy, Agriculture and Environment and Natural resources” (Okello, 2005, p. 2). Rural stove programs proved more difficult to gain acceptance because they sought to replace a free system (the TSS) with one that cost money (FES) (Okello, 2005).

Common Advantages of FES

The major advantages of FES are that they reduce the user‟s reliance on fuelwood, which gives them more time to devote to other tasks, eases the pressure on local biomass sources; saves the user time where fuel is gathered; and saves money where fuel is purchased (also in terms of economic externalities associated with overharvesting of trees). Most FES models also reduce the amount of smoke produced, which improves the health of users and any children close by, reduces their risk of adverse health effects and contributes to a cleaner living environment, where clothing and walls are free of soot (World Bank et al., 1994). The use of FES has also been known to increase the social stature of a using household (Okello, 2005).

33

FES programs, like other income-generating activities and cottage industries can have many positive side effects for people involved in them. For artisans involved in making stoves, they gain a new skill, create secondary jobs for others, gain income to support their families, and teach their children valuable lessons in income generation and energy and environmental conservation. For women involved in stove making groups, Okello (2005) says “stove production has also provided a launching pad for realising other ambitions” (p. 5) like increasing their participation in community development activities.

Common Disadvantages of FES

Depending on the design, many FES have specific disadvantages that make them considerably less alluring to prospective users. Firstly, they are almost never free. Any monetary outlay, no matter how minimal, is more than a free TSS.

Stationary models can also, obviously, never be moved, which is opposite to a TSS

(FAO, 2006a). Since FES are pre-constructed, they are not flexible enough to accommodate the wide range of cooking methods like a TSS does (Karekezi &

Kithyoma, 2002). Also, FES can be temperamental and break if made (World Bank et al., 1994) or used improperly (for instance using water to douse a fire will crack a clay liner) (FAO, 2006a). Some designs have fuelwood holes that are so diminutive users must devote extra time to chopping fuel into small pieces (a difficult feat for many tropical hard woods). Other designs that meant to increase heat transfer efficiency by decreasing air flow actually increased smoke emissions. Some chimneys decreased smoke but also decreased efficiency (World Bank et al., 1994). Although a reduction in smoke emissions is typically considered an advantage by stove designers, many users miss the roles it played in deterring insects and animals and adding flavour to food

(FAO, 2006a).

34

Introducing a FES into a home, particularly those that require payment for materials, labour, or the stove itself may be an avenue that disrupts the intra- household balance of power. In order for a rural woman in India to garner financial support from her husband, she told him a stove cost 15 rupees rather than its actual cost of 40 rupees. She was thus able to get financial buy-in when she felt she wouldn‟t have been able to had she asked for the full price (Ghertner, 2006).

Part Three: Energy Programming and Policies in East Africa

Essential for the basic needs of cooking, boiling water, producing light and heat, and a prerequisite for good health, provision of energy is “a reality that has been largely ignored by the world community” (Rehfuess, 2006, p. 6). She proceeded to write strongly that “international development agendas still fail to recognise that missing out on clean energy equals missing out on life” (p. 8). It is imperative then to examine the situation of energy policies and programming, specifically concerning biomass based household energy, in East Africa.

During his inaugural speech in the aftermath of the Second World War

President Truman coined the terms development and underdevelopment. For the next two decades as these terms were fleshed out, the economic contributions of women were largely overlooked. The prevailing belief was that the basic economic role of women was reproductive6. Women‟s needs were not promoted among development initiatives involving the poor, the participation of women in programs was not sought, and as a result women were left out of many of the ensuing benefits (Martine &

Villarreal, 1997).

6 See „Gender and Energy‟ on page 40 for a discussion of women‟s triple role, including their reproductive role

35

Women in Development

The term Women in Development (WID) was coined in the 1970‟s in reaction to the pervasive marginalisation of women in development activities. This school of thought devised ways to incorporate women into current projects, regardless of their scope or methodology. They measured success in the number of women involved, and focused on women as a special group (Martine & Villarreal, 1997).

The WID approach coincided with the energy crisis of the 1970‟s, bringing to the forefront thoughts on energy and gender. At that time many planners regarded women as the root cause of the fuelwood problem. They believed that women were wilfully cutting down forests for their own needs (Skutsch, 1997). The programs of the time focussed on improving the welfare of women by meeting their practical needs, for instance by giving them fuel-efficient stoves (albeit without the tools or knowledge to make repairs). Two UN women‟s organisations were formed in this decade, and several international conferences were held, increasing awareness about gender, however little changed in the field. The programs continued to function with the basic needs approach. They believed people would „develop themselves‟ if given the basic requirements for development (Skutsch, 1997).

During the 1980‟s there was significant growth in women‟s movements in developing countries, coupled by a startling economic crisis that increased poverty in developing countries. The term feminisation of poverty was coined in response to the realisation that women were the most affected by escalating poverty in their roles as providers, and as the primary victims of government funding cutbacks. Women‟s movements criticised the concept of development, striving instead to create alternative visions from a feminist perspective (Martine & Villarreal, 1997).

36

WID was a good start. Female participation and benefits from development activities did increase, but criticisms of the WID approach were unavoidable. The primary criticism was that it did not question the type of activity going on, the implicit development model, nor the probability that any gender inequity in the models would inevitably reproduce themselves over the long term. Women were simply inserted as small components into larger projects or involved in small women-only projects, being left out of the main development interventions (Martine & Villarreal, 1997). WID was also called „paternalistic‟ by proponents of a new theoretical position called Women,

Environment and Development (Skutsch, 1997).

Women, Environment and Development

Supporters of the Women, Environment and Development (WED) approach argue that women are saviours of the environment, which means that improving the status of women is a first step in ameliorating environmental problems because they are closer to nature, due to the sexual division of labour (Locke, 1999). WED practitioners believe that it is not enough to look at women‟s needs in isolation. One must consider the social and environmental context, and consider questions like: „who controls the resources?‟, and „who has access to private land and common forest?‟

(Skutsch, 1997). In energy projects, this translates into “a concern for women as key actors or privileged participants whose additional or even sole participation will lead to value added in terms of project efficiency” (Locke, 1999, p. 268). Whereas men are primarily responsible for growing cash crops for profit, women look to natural biodiversity for fuel, medicine, water, and foods. Therefore, they are the “most appropriate participants in environmental conservation” (Locke, 1999, p. 268).

The WED position advocates participation and intervention. It “accepts the socially-constructed nature of women‟s roles in resource use but assumes that women

37

in their current roles are naturally more informed and interested in environmental issues” (Locke, 1999, p. 268). Environmental policies that regard gender inequalities garner ideological support from WED (Locke, 1999).

Ecofeminism

Ecofeminism is a school of thought that is loosely related to WED. While “WED focuses on the material roles of women, ecofeminism emphasizes the „natural‟ and spiritual content of women‟s „closeness‟ with nature” (Locke, 1999, p. 269). It places women as victims of environmental degradation and postulates that women are inherently superior resource managers than men because of their biology

(reproductive abilities) or spiritual linkages. Vandana Shiva speaks of “caring, nurturing, sustaining and non-violent” women and nature being in the same category, both suffering from the “mechanistic” approach and relation of dominance and oppression from “patriarchal”, “profit-driven” men (in Martine & Villarreal, 1997, p.

7).

The primary criticisms of ecofeminism are that it defies the concept of gender, and promotes inequality between the sexes. If women are innately better than men because of some spiritual or physical attribute, change is dependent on genetic evolution, rather than an evolution of social roles. This biological determinism would mean that all women everywhere have the same universal sex roles, which are unchangeable and different than men, which is not the case. Likewise, it is difficult to deny that spirituality is a cultural construct, since religions differ so much even in the involvement of women. Removing the concept of gender places women back in the place they fought (and are fighting) to be emancipated from (Locke, 1999; Martine

& Villarreal, 1997).

38

Gender and Development

Nearing the end of the 1980‟s, the Gender and Development (GAD) school of thought emerged as a reaction to the meagre results of WID stemming from it‟s patchwork approach to integrating women in development agendas. GAD seeks to empower women rather than merely incorporate them into pre-existing projects and methodologies. The goal of GAD is to transform unequal relations within a framework of economic development. It works to transform unequal gender and social relations by questioning the type of development models used (Martine & Villarreal, 1997).

Furthermore, a GAD perspective

“regards current social roles as established and maintained through power and authority, and therefore intrinsically contested and dynamic . . . Women‟s and men‟s relation to environmental resources is part of general entitlements and capabilities ascribed to individuals by social relations of gender, class and so on” (Locke, 1999, p. 269).

The role of women as keepers of the environment can and has been contested.

A gendered approach recognises that men also have considerable knowledge about environmental issues, and have reason to strive to conserve them, especially as they apply to his realm of work. As such, GAD does not presume that women will necessarily benefit from environmental conservation and questions policies “asserting a synergistic relationship between women and the environment” (Locke, 1999, p.

269). Gender analysis suggests environmental interventions will be an arena where gendered bargaining processes will be acted out (Locke, 1999).

A notion continues to exist which implies that “the world is and continues to be divided into two unequal parts” which is “inadvertently accepted and perpetuated by focussing on women in poor households” (Martine & Villarreal, 1997, p. 9). Clancy et al. (2007) further argue that gendered energy needs need to reflect gender roles, embedded in gender division of labour. The gendered division of labour puts the onus

39

primarily on women in the global south to meet their energy needs through procurement of fuel.

Gender and Energy

Why speak about gender and energy instead of women and energy? Firstly, one must understand the difference between the terms sex and gender. Simply put, sex incorporates biological and physical differences between women and men. These are differences that as a rule can not be changed, and are immovable and innate (i.e. women can carry babies and men are generally physically stronger). Gender refers to those differences between men and women that can be changed- differences that are socially constructed, taught, differ between societies and are dynamic (i.e. women wear dresses and wash dishes while men wear trousers and build houses) (Moser, 1993;

Skutsch, 1997).

Gender roles then are based on society, not biology. Another important family of terms when discussing gender are the „triple roles‟: the reproductive role is any actions related to reproducing the labour force, including preparing food, tending the sick, child bearing and rearing; the productive role includes actions related to doing work for payment or in kind contributions, i.e. the production of wealth through subsistence or market enterprises), and the community role encompasses tasks undertaken in order to increase community rather than personal wellbeing, including volunteer work, participation in religious or communal activities. The extent to which each gender participates in each of the triple roles depends on socio-economic status, and culture (Moser, 1993; Skutsch, 1997).

40

Gender Planning: Incorporating Practical and Strategic Gender Needs

An underlying rationale of gender planning is recognising that men and women occupy different roles in society and achieve different levels of control of resources therefore they often have differing needs. Gender planning is concerned with identifying gender interests, or prioritised concerns specific to one gender, and then satisfying them by way of fulfilling corresponding gender needs (Moser, 1993).

Gender needs can be divided into two categories: practical and strategic.

Practical gender needs are often more easy to address and plan for, as they do not challenge the gender division of labour or the position of women in the society.

Rather, they are “the needs women identify in their socially accepted roles in society”

(Moser, 1993, p. 39). These are needs relating to roles within the traditional domestic arena: child care, food preparation and provision, and domestic income-earning activities. Strategic needs are more difficult to plan specifically for. They are “the needs women identify because of their subordinate position to men in society” (Moser,

1993, p. 40). These needs are highly contextual; they could relate to domestic violence, legal rights and equal wages (Moser, 1993).

Margaret Skutsch (1997) exhorts planners to consider the practical and strategic needs of stakeholders when considering changes in habits and practices and the introduction of new technologies. Examples of practical needs in terms of stove choice include things like: does the stove allow me to cook all the foods I need to cook? Does it hold my pot well? Is it within my budget to purchase? Is the fuel accessible? Does it fit in my kitchen? Practical needs can be broadly categorized as those that stem from concrete, tangible problems. Conversely, strategic needs stem primarily from non-tangible issues, such as improving or maintaining relationships, status, and freeing up time or energy. In this context, strategic needs include: being

41

able to produce warm food on time, being able to maximize time spent cooking to be free to do other activities, having a stove that men feel free to use as well as women, and keeping children safe.

In energy programming, attention has been paid to women as a specific group only in terms of household energy projects (stoves, cookers, fuelwood projects) appealing to a woman‟s reproductive role. General energy projects have lacked a gendered approach, not usually considering the cultural context, and the potential impacts of the innovation for men and women in the societal roles they fill.

Gendered Approaches: Efficiency vs. Equity & Equality

Margaret Skutsch (1997) gives two possible starting points to consider gender as a factor in rural planning: efficiency and equity or equality. Supporters of the efficiency argument recognise that women and men have different roles, and that what is positive for one may not benefit the other. Therefore, if only the heads of households are questioned during the course of a project, this primarily male group will not necessarily put forth ideas that benefit the entire family. Rather, there have been multiple instances of well-intentioned programs which actually increase labour for women while decreasing it for men. Assumptions were made about women‟s wants, needs, and general lack of willingness to participate in projects without actually consulting them. Efficiency is achieved only when both genders are part of the planning process due to their differing ideas.

The equality or equity argument is about the empowerment of women and a desire to „fix‟ cultures. From this standpoint, women are systematically underprivileged by their socially and economically weaker gender roles. These roles are socially constructed; therefore they have the potential to be re-constructed and create a shift towards emancipation of women. This is imperative because

42

interventions that are aimed at benefiting everyone are unavoidably skewed in favour of men due to their relative prominence in society, commercial enterprises, and ability to own land and procure loans (Skutsch, 1997). If one considers increases in household productivity levels but does not investigate who is supplying labour, who is garnering benefits and the uneven advantages males inevitably possess, the result is gender-blindness (Clancy et al., 2003; Clancy, Ummar, Shakya, & Kelkar, 2007;

Skutsch, 1997).

Hybridising efficiency and equity arguments creates a more moderate approach to gender programming. This approach recognises that women are exploited but does not advocate for a complete sexual revolution. A goal in the middle ground would be to encourage some societal reforms in order to improve the lot of women. For energy projects, this usually starts with the efficiency approach at the planning stage and equity mindset during implementation (Skutsch, 1997).

Gender-Energy-Poverty Nexus

The gender-energy-poverty nexus has increasingly become a point of interest for researchers in the energy field for many reasons. The nexus discussion looks at the extreme interconnectedness of gender, energy use and economic well-being. Okello

(2005) remarks “women and children suffer the most from over reliance on limited biomass energy resources in rural areas” (p. 2). Women primarily gather and use biomass resources and very seldom have access to any “modern, clean and efficient energy technologies” (Okello, 2005, p. 2). Clancy et al. (2003) explain that “energy poverty interacts with other manifestations of poverty” (p. 5) and that “it is important to explore the issues that surround it, including the gender aspects” (p. 5).

Developers and governments argue that in order for economic development to happen, constituents must escape energy poverty by climbing the energy ladder.

43

Energy poverty can be defined as “the absence of sufficient choice in accessing adequate, affordable, reliable, high quality, safe, environmentally benign energy services to support economic and human development” (Clancy et al., 2003, p. 5). It is based on the idea that energy is an essential input for sustaining livelihoods at the most basic level, providing food, boiled water, light and warmth. The poorest people of the world (between 2-3 billion people) rely primarily on dwindling biomass resources for their energy source, for which women are the primary finders and users

(Clancy et al., 2003).

Energy is not addressed directly in any of the Millennium Development Goals

(MDGs) adopted at a United Nations Assembly in 2000 (Dutta et al., 2005; Rehfuess,

2006). Rather, energy poverty is seen as another manifestation of the poverty that is a prevailing feature in low-income households throughout East Africa (Rehfuess, 2006).

The UN Millennium Project report identified “energy, including electricity and safe cooking fuels, as an essential infrastructure service and part of the means to a productive life” (Dutta et al., 2005). Linkages between improved access and more efficient use of energy exist with all eight MDGs (Dutta et al., 2005; Rehfuess, 2006).

These linkages are summarised in Appendix 2.

While most major environmental worldwide problems (i.e. global climate change, accumulation of radioactive waste, over appropriation of biomass) are a result of actions in developed countries where gender gaps are smallest, most secondary problems (i.e. desertification, decrease in natural resources, increase of erosion, increase in air and water pollution) are more localised and reversible if enough money and political policy is in place, and occur in places where the gender gap is the greatest (Martine & Villarreal, 1997). Many of these secondary, reversible environmental issues are associated with energy issues and affect women

44

disproportionately more than men, increasing the gender-energy-poverty nexus

(Martine & Villarreal, 1997; Simms & Reid, 2005).

Traditional Fuels in Public Energy Policies

The Energy Ladder

In a world that is feeling the pressure of increased prices of petroleum-based fuels, the allure of renewable energies is strong. Not all types and uses of energies are considered equally desirable though. Rehfuess (2006) explains that “being poor condemns half of humanity to dependence on polluting household energy practices”

(p. 8). Clancy and others (2003) agree, saying “the impacts of poor quality fuels and energy scarcity extend beyond cooking” (p. 20) to threaten well-being and increase vulnerability.

A popular image in current energy literature is that of an energy ladder. At the bottom of the imaginary ladder would be various „polluting‟, „poor quality‟ biomass fuels: animal dung, crop residues, fuelwood, and charcoal. Biomass fuels are further divided into two categories- inefficiently used (traditional fires/stoves) and efficiently used (burned in FES). At the top end of the hypothetical ladder is coal, fossil fuels such as kerosene and LPG, with electricity at the highest rung. The general assumption, which is backed up by multiple studies, is that people will get energy from the highest economically feasible point on the ladder (Barnes & Floor, 1996).

Energy on this ladder is progressively cleaner and more modern.

Government Organisations

Governments generally consider the drudgery and efficiency arguments of fuel collection and take one of two approaches to addressing fuel shortages. The first is to promote efficient, sustainable use of traditional biomass fuels (such as introducing

45

efficient stove programs) and the second is to encourage residents to switch to modern cooking fuels. Governments generally advocate implementation of policies that promote cleaner, more efficient fuels and technologies and break down barriers to access, affordability and supply as a part of their broader development strategy

(International Energy Agency, 2008).

A recent FAO publication examining policies and programs of selected developing countries (including Kenya) revealed there was a general focus on liquid fuels and electricity with little reference to biomass fuels. The policies were also relatively limited in scope, addressing regulatory measures much more than research and development, extension, and economics. The countries set ambitious targets, perhaps because of the early stage of the programs (FAO, 2007). This meant that most energy policies did not address energy poverty because they focussed on commercial energy rather than biomass fuel which is largely produced and sold in an uncontrolled fashion in the informal sector (Murphy, 2006; Clancy et al., 2003). Wood fuel remains the dominant source of energy for most households7, small service and industrial establishments in Kenya and will continue to be for the near future, but little is spoken about it in the country‟s energy policy (Ministry of Energy, 2003).

The intentions of governments of developing countries to address the myriad of issues that interact with energy are stymied by a lack of cooperation across government ministries (Clancy et al., 2003). For instance, in Kenya the Ministry of

Energy is concerned with promoting fuel-efficient technologies (Ministry of Energy,

2006), but they carry out this task jointly with the Ministry of Agriculture who promote energy saving stoves at the household level. The Ministry of Energy is left to increase the efficiency of improved stoves (mainly charcoal and institutional stoves) and

7 In 2005/6 87.7% of rural households used wood for cooking while 30.2% of urban households used charcoal. This is a 52% decrease in charcoal use in urban areas since 2000 due to a shift to LPG and kerosene (Nyoike, 2008).

46

promote them at their Energy Centers8. Their goal is to increase adoption of efficient charcoal stoves from 47% to 100% and FES from 4% to 30% nationally by 2020 (Nyoike,

2008). By 2024, the Kenyan government aims to expand rural electrification penetration to 40% in the country (it is currently at 3%) and to eliminate the fuelwood deficit completely (Kenya Ministry of Energy, 2003).

The Ministry of Forests sets forestry objectives that must integrate meeting the demand for fuelwood and charcoal and facilitate access to supplies for biomass fuel dependent people (Arnold et al., 2003). Meanwhile, the Ministry of Planning and

National Development is concerned with meeting the eight Millennium Development

Goals, of which provision of “clean, appropriate, affordable, sustainable energy” is a key component (Kenya Ministry of Planning and National Development, 2005, p. 27).

Gender in Energy Planning and Policies

Clancy and others (2003) urge government and non-government organisations to address energy use as a poverty issue and a venue to empower rural people, especially women. If they don‟t, supposedly gender- neutral policies may backfire and discriminate against women (Karekezi & Wangeci, 2005). Huyer and Westholm (2001) sagely quipped “no data, no visibility; no visibility, no interest” (in Clancy et al., 2007, p. 243), referring to the lack of interest on the part of energy planners to create policies with a gender approach.

Clancy and others (2007) argue elsewhere that “the invisibility of women‟s needs to energy planners [stems] from a lack of appropriate gender-analysis tools to meet the particular data requirements of the energy sector” (p. 241). Karekezi and

Wangeci (2005) note that “interviews with [East and Southern African energy] policy

8 The Energy Centers aim to “effectively serve as frontline field extension, promotion and training facilities of the Ministry [of Energy]” (Ministry of Energy, 2003, p. 3)

47

makers demonstrated that they either failed to perceive a need for gender mainstreaming or were constrained by their unfamiliarity with gender issues” (p. 15).

The very fact that energy, especially biomass energy, is „a female matter‟ and most policy makers are not poor or female contributes to gender-blind policies.

Indisputably there is a growing movement in the global south fighting for emancipation and rights of women. Energy projects can support the process of empowerment of women because gender roles are culturally determined, and therefore can be changed. Gender-blind planners who look only at increasing productivity or income at a household level without scrutinising the internal workings of a family unit hinder equity based gender planning (Skutsch, 1997).

Gaps in Literature

As this review of literature has shown, there are wide reaching advantages and disadvantages for using biomass fuel, TSS and FES. Past research has shown that successful stove promotion is based on a myriad of contextual factors. Philosophical orientations underlying policies regarding household energy are usually gender-blind, which has a negative effect on FES use. This study aimed to add to that body of literature by making a qualitative study that takes an efficiency gendered approach

(Skutsch, 1997) by intentionally incorporating the views of both young and old women and men, and looking at the nuances of stove adoption, not merely using categories of adopter or non-adopter of a technology.

Using a Gendered Approach to Investigate Stove Adoption

The role of men in influencing decision making regarding what stove to use on a regular basis is little understood. Available examples in stove literature include a study in Sierra Leone documenting preferences of traditional fires for the taste the

48

smoke imparts to food (Karekezi & Kithyoma, 2002) and a case in Zimbabwe where men influenced their wives to shun new solar cooking technology because technology and development were traditionally a male-dominated arena (Clancy et al., 2003;

Ladipo, 1994; Simms & Reid, 2005).

The pervasiveness of the idea that women make decisions surrounding stove use is evidenced by a search of the World Bank‟s (World Bank et al., 1994) comprehensive report on factors that would make someone want to cook with an improved biomass stove. The terms „man‟, „men‟, „husband‟, „wife‟, „spouse‟ and

„gender‟ do not appear once in the 36-page document. A more recent review by

Clancy et al. (Clancy et al., 2003) recorded more extensively of the role of men in household decision making, especially decisions about energy and technology. In a household with both adult men and women, the gendered division of labour dictates that women make decisions in her realm of influence (typically the kitchen), however when energy or technology is to be purchased, men enter the decision making process

(Clancy et al., 2003; Wilson & Green, 2000).

Men also make decisions about the construction of the cooking hut- the design and choice of roof and wall materials influence the cooking and smoke patterns

(Clancy et al., 2003). Some men, upon being presented with the prospect that their wives could use a time-saving solar cooker expressed concern over how their wives would use their new time but also saw it as an opportunity for them to do more productive activities (Wilson & Green, 2000). On a similar note, women pragmatically realised that savings in one area of drudgery would equal an increased amount of time doing drudgery work elsewhere (Clancy et al., 2003). The questions that remain are: who makes decisions? and who benefits? This study sought to answer the call for a

49

shift from a focus on women to a gender approach, incorporating also the views and influences of men.

Qualitative research

The FAO (2006) includes „attitude‟ in a list of disadvantages associated with

FES stoves. They argue that certain traditional beliefs and attitudes may need to be changed in order to regularly use a FES. SCC-Vi Agroforestry, a non-governmental agroforestry organisation active in East Africa suspect that unexplained social and cultural factors are inhibiting the adoption of FES in their region (SCC-Vi, 2008).

Bearing in mind that each stove promotion program faces a distinct set of challenges and benefits depending on local conditions, it is prudent to seek the particularities of the interface of communities and programs so that programming can be tailored to increase the probability of stove adoption (World Bank et al., 1994). This research aimed to discover yet undocumented beliefs or attitudes that inhibit the use of FES in

Trans Nzoia district, and identify them so that local FES extension programs can be tailored to take those beliefs into consideration.

Complexity of Categorisation

Most studies related to stove use can be compared to snapshots, meaning data was gathered and analysed at one point in time. This more simplistic analysis lends itself to a two dimensional view of adopters or non-adopters. Kiptot et al. (2007) questioned the validity of conclusions from studies like that, and suggested looking at longitudinal data that categorised people as adopters, re-adopters, pseudo-adopters, testers/rejecters and non-adopters of technology. This study anticipated taking that line of reasoning one step further and fill gaps in stove adoption literature by looking

50

more closely at the reasons people have used a particular FES in the past, present, and the reasons they accept or reject to continue using one in the future.

Conceptual Framework

The framework for this study flows out of the literature review and gaps found in literature, particularly focussing on four major proposed factors that determine the active usage of FES within a rural Kenyan household. For the purposes of this study,

„active usage‟ is defined to be using the stove at least once a week to prepare a meal and „FES‟ is defined to be any stove that is more efficient than the TSS and still uses raw biomass fuel. The framework is based on a study looking at FES adoption in

Southern Africa (Makame, 2007).

This framework builds on four core concepts. Firstly, there must be FES available to be used. In Kenya stove dissemination happens through the market place and special programs set up by community based and non-governmental organizations.

The extent and way the government implements its plan to promote fuel-efficient stoves likely also has an impact (Kenya Ministry of Energy, 2003). FES dissemination programs which are long-lived, de-centralised, and have good coordination between stakeholders (support agencies, stove builders and sellers) have a much better chance at survival than ones that are not well established, managed with layers of bureaucracy, and not coordinated with similar efforts in the area (World Bank, 1994).

Secondly, the FES must have obvious advantages over the model currently in use, which in most cases is a TSS. If it is complex to use, complex to prepare fuel for, or repair then it will quickly fall into disuse even if it is present in a household (Foley et al., 1984; Makame, 2007).

51

Positive impacts on Negative impacts on social wellbeing, social wellbeing, human and human and environmental health environmental health

DECISION TO TOVE USED TOVE NOT S USE FUEL S ACTIVELY ACTIVELY EFFICIENT USED STOVES

PERCEPTION OF STOVE BY INDIVIDUAL PERCEPTION OTHERS OF STOVE By spouse Socially acceptable By other friends and peers Useful User friendly

STOVE ATTRIBUTES INDIVIDUAL FACTORS Relative Strategic needs advantages Practical needs Complexity Information Comparability Education Economic

MANAGEMENT SUPPORT Policies Projects Programs

Figure 2.1: Conceptual framework used for the study (based on Makame, 2007).

Thirdly, the FES must fulfill some individual perceived need, either strategic or practical, and the education and information must be available for the person to know that those needs can be met. For instance, reviews of dozens of stove dissemination programs found that people who perceive a fuel shortage are more likely to use a fuel- saving technology than those who have no awareness of a fuel deficit. Education must be focused to be relevant and address local issues, which makes larger, decentralized programs less effective than smaller, grassroots ones. Economic factors also factor

52

prominently in this category. Previous studies have found that people who need to pay for fuel already are more willing to pay for a technology which reduces the amount of fuel used than those who scavenge fuel, regardless of the time savings a reduction in fuel needs represents (Foley et al., 1984; World Bank, 1994).

Finally, according to Everett Rogers‟ (1983) work on diffusion of innovation, friends and peers are the primary influencers during the evaluation and trial stages of an innovation and the second most prominent influencers at the initial awareness stage. Other studies have shown specifically how the attitude of a husband can negatively affect the adoption of technologies into a household (Ladipo, 1994), or negatively affect their use once they have been incorporated into a household (Mahat,

2003). It is suspected that in the end, community peer pressure and intra-household pressure are the strongest influencers of whether an FES is adopted, and how actively it is used. This research is based on the assumption that there are more benefits than negative consequences on the social, environmental and human health levels in using a FES as opposed to maintaining the status quo.

Summary

This chapter examined relevant literature in the fields of traditional and fuel- efficient stove use, promotion and adoption and the role and relevance of gender and raw biomass fuel in energy programming and energy policies. It also suggested gaps in literature that this study aimed to fill, and introduced a conceptual framework on which this study is based. The following chapter discusses the socio-political, historical and geographic context of the study in more detail and summarises some literature on the history of fuel-efficient stove promotion and use in Kenya.

53

CHAPTER THREE

STUDY CONTEXT

Introduction

Large and small scale historical, physical and social contexts shape the lives of participants, influence how they see the world and how they act within it (Bamberger,

Rugh, & Mabry, 2006). This chapter outlines some of those contexts, in both small and large scale, particularly focusing on the geographic areas participants lived in. This chapter also provides a context to explain why FES have been promoted in the area by the Ministry of Agriculture and some non-governmental organizations. Unless otherwise indicated, the information in this section was gathered from administrator

(AD) and Ministry of Agriculture (AG) key informants listed in Appendix 13. The study took place in four out of seven divisions in Trans Nzoia, so this chapter focuses on those in exclusion of others9.

History of the Study Area

During colonial times the original inhabitants of Trans Nzoia were displaced by large scale European-owned farms. Following independence in the early 1960‟s, laws were passed and settlement schemes were set up to allow African Kenyans to take loans and buy the large farms that made up the land previously referred to as the

White Highlands‟ (Odingo, 1971). Each farm was sold to one group, whose members came from a single ethnic group. Each original group member had a piece of land up to 6 acres in size. During the last forty years, these group members have subdivided their land, sometimes passing it on to sons, and sometimes selling to a

9 Refer to Appendix 4 to see the organisation of divisions within Trans Nzoia District

54

willing buyer. This has resulted in a patchwork of communities, which are more homogenous in some areas than others due to different rates of selling off individual portions of land (Brown, 2003). In general, areas close to Kitale were more mixed, and comparatively remote areas were more homogenous (such as Cherangani and

Endebess), since they were settled last). Endebess and some parts of Kaplamai also had large scale (less than five acres) farms mixed with the small scale farms. This was especially true in Endebess where Agriculture Development Company and Kenya Seed farms dominate large swaths of landscape.

Following the redistribution of colonial lands, the population of the area increased at a rate similar to the rest of the country. Population density maps show that Cherangani, Kiminini and Endebess experienced higher rates of population growth from 1989-1999 than Kaplamai in the same period, likely because Kaplamai was settled first and had a higher density to begin with. In 1999 there were 100-300 people living per square kilometre in most of the study area. At the same time 45-55% of the population throughout the area lived below the poverty line (World Resources

Institute, 2007). Population growth is estimated at 3.8% per annum throughout the district (SCC-Vi, 2006).

The division and partition of this nutrient-rich land has meant that previously treed areas have been cleared for plant agriculture, which now covers over 75% of the land (World Resources Institute, 2007). The increased population pressure has put escalating pressure on fuel sources in the area to the extent that 78% of residents experience mild to extreme fuel deficiencies (Drigo, 2006). In areas where grandparents could not have dreamed wood would become scarce, their grandchildren are now struggling to burn maize stovers to replace the dwindling supply of firewood.

55

Maize cobs and stovers are important secondary (and seasonally primary) fuel sources, and sawdust and charcoal are tertiary sources of fuel.

Physical and Agricultural Characteristics of Trans Nzoia District

Trans Nzoia district is located in an agriculturally rich portion of the Rift Valley province of Kenya commonly referred to as the country‟s „granary of maize‟ (Nyangito

& Ndirangu, 1997, p. 2). Both small and large scale farmers in Trans Nzoia produce

40% of domestically marketed maize in Kenya (Nyangito & Ndirangu, 1997). It is 246

700 ha in size, which is 0.43% of Kenya‟s total land cover (Kassam, van Velthuizen,

Fischer, & Shah, 1993). Mount Elgon covers the western boundary, part of which forms a portion of the border with Uganda. The semi-arid region of West Pokot is situated to the north-east, and Uasin Gishu, a division similar to Trans Nzoia in terms of land use lies along the southern border. The major town in the area is Kitale, a modestly sized and centrally located agricultural town. The closest cities are Eldoret approximately two hours to the south and Kisumu, approximately three and a half hours to the west.

The district lays 1800-2100m above sea level and has an average temperature of 18.6 C. Most of Trans Nzoia experiences a relatively high average annual rainfall compared to the rest of the country. On average, 800-1200mm of rain falls per year in the eastern region, and 1200-1600mm of precipitation falls near Mt. Elgon (World

Resources Institute, 2007). Kitale receives a mean annual rainfall of 1168mm. Rain is distributed through all twelve months, with peaks in May and August and drier times from November through February. The study took place from May to August, during which time the rain fell five to six times a week. The villages were located within two of the five major upper watersheds for the country (Upper Eastern Watersheds of

Mount Elgon and Upper Southern Watersheds of the Cherangani Hills). The volcanic

56

soil in most of the district is a reddish sandy loam over a basement complex, but in

Endebess it transitions to dark-red and red-acid loams (Odingo, 1971).

Figure 3.1: Location of Trans Nzoia District within Kenya (source: Drigo, 2006)

The average land size in Trans Nzoia is 3.9 acres, but the mode is 0.9 acres

(SCC-Vi, 2006). Composite maps show that land is cultivated most intensely in

Kaplamai, Kiminini and Cherangani which also happen to be the areas with smallest land holdings. Despite differences in land size, the predominant livelihood strategy

57

throughout the area was mixed farming of food crops, cash crops and livestock10. In some areas (for example Cheptantan village in Endebess) casual waged labour on large scale farms is the predominant livelihood strategy (World Resources Institute, 2007).

A baseline survey conducted in 2006 by SCC-Vi confirmed that 80.5% of rural residents in Trans Nzoia were farmers. Over 95% of them grew maize, which grew particularly well on the soils in Endebess district near Mount Elgon where plants were grown without fertilizer yet were more vigorous than plants in any other area. The abundant crop residues (maize stalks and cobs) were used preferentially as cooking fuel rather than as mulch for fields. The next most popular crop was beans (grown at between 79% and 94% of farms in the study area), followed by sweet potatoes, millet, and irish potatoes (11-26% in Cherangani, Kaplamai, and Endebess). Bananas were grown by residents in Kiminini (21%) and Cherangani (25%) but not at all in the

Kaplamai and Endebess. Cherangani (81%) and Kaplamai (72%) led Endebess (57%) and

Kiminini (54%) in chicken production. Between 28% and 41% of people kept at least one cow through all areas (SCC-Vi, 2006).

Cooking Habits of Study Participants

Diets for rural residents in the study area were based around the foods they grew: maize, beans, potatoes, green vegetables, and fruits; supplemented by ingredients easily available in the markets: tomatoes, onions, oil and spices. Chicken, goat or beef was eaten on special occasions. A typical daytime meal was made of a starch and a vegetable accompanied by tea, while breakfast was often a thin fermented porridge called ugi or tea. Depending on the season and economic status of a household, people ate between one and three times a day.

10 Area residents received 25-50% of their cash income from livestock sales (World Resources Institute, 2007).

58

Ugali, a stiff porridge usually made of maize meal (though traditionally made of other grains such as sorghum and finger millet, mixed with cassava and fermented) was the staple food that no day was complete without11. This food is made by vigorously stirring flour and water in a pot over a heat source. In the last stages, the porridge is so thick it is forcefully turned over and over, necessitating a brace of some sort to prevent the pot from tipping over. Cooking ugali to feed a family of eight would take about 30-40 minutes. Luhya elders said traditionally ugali was made with cassava, finger millet and amaranth (FG4).

Githeri was the second most common food in the area. It was made of dry maize and beans boiled until they became soft (about three hours). Since maize and beans were grown on almost every farm, this dish was nearly ubiquitous. It was simple to prepare in large quantities, but required more fuel than any other food. Less fuel was required if the maize and beans were pre-soaked, but respondents complained that it tasted badly when that happened.

Dominant Ethnic Groups in Trans Nzoia and Their Relevant Characteristics

Kenyans generally identify strongly with a particular ethnic group (Bratton &

Kimenyi, 2008), in part due to the divisions and ethnic titles reinforced by colonial influences. In this study an ethnic group refers to a group of people with a common ancestry or history, similar language and culture and usually the same „homeland‟

(Hall, 2000). Culture in turn refers to the collective representation of customs and activities found in kinship, marriage, friendship, ritual and ceremony that are manifested in individual behaviour (Cohen, 1974). In other words, an ethnic group is something one is typically born into and cannot change, whereas culture encompasses

11 Many people reported that if they had not eaten ugali in a given day, they would say they had not eaten food at all. The terms „food‟ and „ugali‟ were often interchanged in this way.

59

a dynamic mix of customs that are usually associated with a particular group of people

(ethnic, national, regional or otherwise).

Although data could not be located that gave an ethnic breakdown of the area,

District Officers and Area Chiefs reported that Kalenjin (Sabaot and Nandi) and Luhya

(Bukusu) were present in the highest numbers, while Kisii, Kikuyu and Turkana were the next most prevalent groups. Every study participant identified with one of those five broad ethnic groups. Some customs of local ethnic groups were relevant to the study.

Kalenjin is a name given to a broad group of peoples (sometimes also called

Highland Nilotes) who would not necessarily identify each other as relatives. They had in common their language group and their preference to be cow herders (Mokoloo,

2005). They are also known for drinking more milk than their neighbours, and perhaps because of this are known for preferring to have clean kitchens (to keep flies from the milk). Trans Nzoia is the traditional land of Kalenjin people, particularly the Nandi and

Sabaot (sometimes called Elgon Maasai) subtribes (Collier & Sambanis, 2005).

Sabaot people had been traditionally more resistant and slow to change than other people groups in the area. Polygamy was very common in Sabaot communities.

They used three stone fires almost exclusively and had significant cultural beliefs attached to that device. They originally inhabited the slopes of Mount Elgon and are still in the majority there. Mount Elgon was one of the last places in Kenya to be reached by institutions such as the Christian church, so they remain with one of the highest proportions of people following traditional beliefs.

The Nandi subtribe was based in the Cherangani and Nandi Hills to the east and south of Mount Elgon. They had more contact with other groups of people than the

Sabaot did, so were comparatively more progressive. Nandi women were the first

60

ones to create a chepkube („hiding place‟ in their language) attached to their stove to store their food warm. They were also the innovators of the brooder and hatchery chepkube stoves.

Luhya (or Abaluyia) was a name for another broad group of peoples with slightly varying dialects and customs (Mokoloo, 2005). Most Luhya people in the study area were of the Bukusu subtribe. Polygamy was very common within the older generations. Those who still followed traditional belief systems are known for having strong customs and taboos about many things- from TSS to conversation topics allowable on the day of a funeral12. Their traditional area lay to the south west of

Trans Nzoia district.

The Turkana are also a group with several subtribes (Mokoloo, 2005) who migrated from a very arid area north of Trans Nzoia. Their traditional stove was an improved TSS that was somewhat more fuel-efficient. They did not traditionally have the luxury of abundant fuel, so this was likely the reason they closed in the three stones. Turkanas were not necessarily known for following their traditions when they lived far from their traditional homelands.

Recent Ethnic Violence in Trans Nzoia District

For two months following Kenya‟s federal election in late December 2006 the country suffered great losses and disruptions in lives and properties as minority groups in otherwise politically or ethnically homogenous areas were (often violently) displaced by their neighbours. The majority of Trans Nzoia residents supported the opposition party, but Kikuyu, Kamba and Embu residents were particularly targeted for their suspected support of the party in power as well as for historic grievances related

12 One older gentleman declined to answer questions about culture during an interview because he had returned from a funeral that day and was afraid he could be reported to the elders.

61

to historic land tenure issues (Bratton & Kimenyi, 2008). Within Trans Nzoia district,

Endebess, Saboti, Kwanza and Cherangani divisions were the most adversely affected, possibly because they had less ethnic and political diversity, so minority groups could be more easily targeted. During the study period, two camps for internally displaced persons were encountered less than ten kilometres away from study villages in

Cherangani and Endebess, and three participants were recently widowed, possibly due to the post-election violence.

A compounding factor for residents of Saboti and Endebess divisions were their geographic proximity and familial ties to the Sabaot of Mount Elgon in the neighbouring district. A local militia called the Sabaot Land Defense Force (SLDF) first emerged to resist evictions due to government land re-distribution policies in 2006.

Thousands of people were displaced and hundreds were killed because of the violent actions of both the SLDF and the Kenyan government‟s combined military and police response. This violence ended in late May 2008 when the militia leader was captured and killed (Human Rights Watch, 2008).

Relevant Characteristics of the Study Area

In early 2008, Trans Nzoia district was divided into Trans Nzoia East, Trans

Nzoia West and Kwanza districts, however for the purposes of this document they are all referred to by the old district name of Trans Nzoia. Trans Nzoia was composed of seven divisions, five of which were visited during the course of this study. Most research took place in two villages from each of four divisions: Cherangani, Kaplamai,

Kiminini and Endebess. The next few pages contain elements common to all areas, and then follow with a brief description gained by participant observation during the study period of notable characteristics of each division and village.

62

Figure 3.2: Study locations within Trans Nzoia district. (Adapted from SCC-Vi, 2006)

Cherangani Division

Cherangani was historically predominantly Kalenjin, and in the remote areas they remain that way. Closer to the centers land was divided and sold to people of all ethnic groups. SCC-Vi surveys show that residents of Cherangani division owned an average of 5.4 acres (mode of 5 acres) which was the largest in Trans Nzoia. Eighty- six percent of household heads had attended school, and 39% had studied past the primary grades. 80.7% of people kept chickens, but only 0.7% kept goats, which was low compared to the district average of 4% (SCC-Vi, 2006).

It was nearly impossible to find someone who did not use a fuel saving device in

Kiboino village. The village itself was about three kilometres from Kajibora, the main trading center for the division, which also housed the District Official‟s office. Most plots of land were much smaller than the area average because the original pieces of land had been heavily partitioned, which led to the creation of a highly populated ethnic patchwork dominated in this area by Turkana. The houses were mainly semi-

63

permanent with thatched roofs. Agroforestry was not intensive, and there were not many non-agroforestry trees either. It was rare to see animals larger than a chicken.

Most houses branched off of the main road and one side road. Residents were very friendly and curious about the study.

The name Lumboka sounds like a Bukusu word, and indeed many Bukusu people were found in that village. This area was a mix of small and large scale farms, and was situated about four kilometres from the main center in the opposite direction from Kiboino. Most residents used fuel-efficient stoves, however it was noted that those that still used three-stoned stoves had not lived in the area for very long. Maize was grown on every farm, but agroforestry was not very extensive. Everyone was very welcoming, but it seemed this area more than any other had been affected by the post-election violence. Two of the four people interviewed were recently widowed and another spoke of land tenure issues that started earlier that year.

Kaplamai Division

Kaplamai was an ethnically mixed area with Kalenjin, Luhya and Kikuyu in the majority. SCC-Vi‟s 2006 baseline survey showed that 86% of household heads were male and 94.2% had attended school, although they most commonly stopped at the primary level (53%). On average they had lived on their farms for 15 years, but the mode was 8 years. In terms of agricultural products, Kaplamai was notable for producing twice the district average of eggs (SCC-Vi, 2006).

Sokomoko felt the most congested and compact of all the areas in the study.

An SCC-Vi extension officer warned that people may give an alternate name for the village because of the negative connotations of the name Sokomoko, which translates into „where bad things happen‟ in one of the local dialects. The village was thus named because it was locally well known for the numerous illegal pombe (maize beer)

64

brew shops in the small market area. The village was located about two kilometres from the main road, and about five kilometres from Maili Nane, the main trading center for this division.

Figure 3.3: The center of Sokomoko village, Kaplamai division

Most lots were between 0.1 and two acres in size, and most had trees planted alongside a kitchen garden of maize, beans, and green vegetables. The trees were mainly for fruit production or building materials and fuel. Some chickens were visible, but few cows or other animals were on the compounds. Houses generally had iron sheets for roofing material and walls made of mud and stick. The village was ethnically diverse, though Kisii, Kalenjin, and the Luhya subtribe Bukusu dominated.

People here were generally very friendly, however when residents of one of the houses that was brewing pombe were approached for an interview they became suspicious. The research team left without conducting an interview.

Kemeloi was a village made of both small (less than five acres) and large (up to

200 acre) farms. It was about two kilometres from the main road, and about six kilometres from Maili Saba, the nearest small trading center. The most striking aspect of this village was that it was impossible to find a TSS. The ethnic makeup was primarily Kalenjin, so the most popular stove style was the basic chepkube. Although

65

some people had said that they had Bukusu neighbours that used both chepkube and

TSS, non-Kalenjins were not located despite multiple enquiries and transecting the entire village.

The area around Kemeloi had quite a few trees, especially eucalyptus planted along the roads of large scale farms. Even some small scale farmers had trees but they were not necessarily using them for agroforestry purposes. The lack of agroforestry methods could have been partially attributed to the lack of activity by

SCC-Vi due to the large percentage of large-scale landowners in this area.

Kemeloi seemed to be slightly better off than others in the study. All houses used iron sheets for roofing, but while three compounds had semi-permanent mud walled houses and chickens scratching through kitchen gardens of maize, beans, and green vegetables, the fourth house was permanent, had a store house made of timber, a cowshed, and no evidence of plant agriculture. That farm seemed to sustain itself well on cows, sheep and chickens. People were very friendly and frank, and freely shared information

Kiminini Division

Kiminini division was an ethnically mixed area with a dominant Luhya presence

(particularly the Bukusu subtribe). SCC-Vi surveys showed that 18% of household heads were employed, a higher than average number likely influenced by Kiminini‟s proximity to Kitale and positioning on a busy travel route from Kitale to Lake Victoria.

The average farm size was the lowest in the area (two acres) but the mode was comparable to Kaplamai and Endebess (one acre). Kiminini grew far more fruits than any other area in the study (banana, guava, mango, avocado) but was average or below average in frequency of animal, maize and bean production.

66

Nabiswa A had a large Luhya population and was made primarily of small scale farmers about four kilometres from the division‟s main town of Kiminini. Most fields were filled with maize intercropped with beans, chickens were more common than any other animal, and nearly every house was semi-permanent. Farms had many indigenous trees on them, but they were not grown for agroforestry purposes. Despite of the dense settlement pattern, it was rare to find someone buying wood. People usually gathered it from their own farm or from the bushes at the roadside, and indeed men, women and children were observed carrying everything from logs to twigs back to their compounds. One small kiosk and one church flanked the Nabiswa A area and provided it‟s only non- residential buildings. The lowest level of education an interview participant had reached was standard seven, which meant that people in this village had more formal education than those in other study areas.

Masaba B was the name of an area that stretched about one kilometre up from the main road starting near the large Kiminini trading center. Land parcels were smaller and more expensive than elsewhere (many were less than one acre), so there were many more permanent houses here than any other village in the study.

Electricity poles dotted the roadside; the roads were much smoother than Nabiswa A.

Agroforestry practices were very strong in this area. There was an abundance of

Sesbania13 trees planted, which usually indicated SCC-Vi had been working in this area.

There were also several instances of zero-grazing cattle. This area was unique for the diversity of fuel-efficient stoves, and the fact people generally did not consider the food-warming capabilities of the chepkube to be its greatest attribute. The area was ethnically cosmopolitan, though people of Luhya descent dominated.

13 Sesbania sp. Are popular multi-purpose agroforestry trees heavily promoted by SCC-Vi

67

Endebess Division

Endebess was the most unique of the four divisions visited. Past SCC-Vi baseline surveys showed that 71% of household heads did not study past primary school, and they were least likely of all the divisions to have their own business (3% compared to a district average of 6%). More farms produced maize, beans, millet, irish potatoes and beans than any other area and although Endebess had fewer milk cows on average, it had far more goats than anywhere else (18% compared to district average of 4%) (SCC-Vi, 2006).

Mtarakwa was the most difficult village to access during the study. Although many cow herders were seen on the road, no other vehicles were visible for the 45 minute truck ride from the junction of the main road to the village. Even armed, uniformed officers were walking on the road instead of driving. The road conditions were very poor, resembling large ruts and gulleys caused by high erosion due to rainfall. The topography of the area was hilly with rocky outcrops. The soil was rich and yielded maize that looked better than any other area yet did not require fertilizer. The village had a catholic church but no businesses with storefronts. The closest kiosk, cafe, shop and schools were about two kilometres away.

Land parcels appeared to be uniformly three acres. Everyone grew maize, but trees and shrubs were relegated to thin fence rows or a few around a house, which meant everyone we interviewed relied on wood from off their farm for fuel. A round- trip walk to the forest where people collected wood was three hours.

68

Figure 3.4: A farmer’s field in Mtarakwa village, Endebess division

Every village resident encountered was from the Sabaot tribe, but no one reported experiencing post election violence or effects from the Sabaot Land Defense

Force militia on the other side of Mount Elgon first hand14. Almost all houses in the area were built in the traditional Sabaot round-walled, thatched roof style. The remaining houses had square walls and iron-sheet roofs. Evidence that this area had been settled relatively recently were: the area was homogenous in terms of ethnic groups, the land was still in uniform packages, the soil did not need fertilizer even though maize had been growing on it continuously, and it was close in proximity to

Mount Elgon National Park and forest.

Cheptantan was composed of a strip of shops, kiosks, and cafes just off a main road, with many farms associated with it that were three to ten acres in size. It was located about three kilometres from one of many large- scale farms in the area, where it was assumed many residents found casual labour. As in Mtarakwa, the soil allowed maize to grow well without the aid of additional fertiliser, but it was of poor quality for covering mud houses or constructing mud stoves as it cracked and fell off very easily.

14 The Sabaot Land Defense Force was active primarily between August 2006 and May 2008. Their activities were focused on the other side of the mountain, in a neighbouring division.

69

Figure 3.5: Road conditions and scenery in Cheptantan village, Endebess division

The presence of school age children who were not in school despite the government‟s free primary education policy was evident in this village more than any other. The research assistants as well as a local principal said that people in this area have a reputation for sending their children to school at a much later age, not valuing education, and removing them from class to work as manual labourers. Indeed, two of the study participants (both adult females) had never attended school. Presence of locally brewed pombe was seen strongly here. The first interviewee made reference to her husband often being drunk, the co-wife of the second interviewee and the mother of the third participant were obviously intoxicated during the visits, which both happened before noon.

Cheptantan residents were primarily Sabaot, although they called themselves by the pseudonym, „Elgon Maasai‟. Perhaps they preferred the second name to distinguish themselves from the Sabaots who had recently caused unrest elsewhere.

Almost every house in the village had at least one cow, and everyone grew maize and beans. Even with large plots of land, there were very few trees on farms. Everyone spoke of collecting wood off farm, along the roadside or in a small forest.

70

History of Stove Promotion and Use in Kenya

Specific Use of Fuel in Kenya

Despite lofty goals to increase rural electrification in the coming decades to as much as 40% annually by 2024, standardise LPG products and greatly increase oil exploration and pipeline extensions in Kenya (Kenya Ministry of Energy, 2003),it is most likely that biomass energy will remain the dominant form for a long time yet

(Karekezi, 2002), as poorer areas restrict their purchase of fuels to those used for lighting (Clancy et al., 2003).

The bulk of energy consumed in rural Kenya is for household use, 97% of which is biomass fuels (Kenya Ministry of Energy, 2003; Mugo, 1999). Within the household, the lion‟s share of energy is used for cooking, with lighting, heating, and battery power for radios being secondary uses. Wood is the most common cooking fuel, while other forms of biomass such as crop residues (for instance straw or corn stalks or husks), sawdust, and animal manure are also used, in their traditional and unprocessed form. For low-income rural households, any negative effects associated with biomass fuel are outweighed by the fact that in most cases, the use and procurement of biomass fuels have no monetary cost (Karekezi, 2002). Urban areas predominantly purchase wood processed into charcoal for cooking (Kenya Ministry of Energy, 2003).

Province-wide, wood fuel use in rural and urban Rift Valley decreased in the period between 1989 and 1999 from 79.3% of fuel used to 73.7%, as did „other‟ fuel15

(7.9% down to 2.5%). Data for lighting in that same time period showed a doubling in the use of electricity to 9.5%, paraffin lamp use grew from 65% to 75% and „other‟

15 „other‟ fuel could refer in part to crop residues or animal dung, as there were not other categories detailing those in the survey.

71

fuels16 decreased from 30% to 15% (Kenya Ministry of Energy, 2002). These data show a shift from unprocessed fuels to more expensive, processed ones. This data reflects either rural to urban migration, or that income is rising resulting in an overall upward movement on the energy ladder.

Figure 3.6: Three girls collect wood fuel in Kiminini

The Kenyan Ministry of Energy compared types of fuel used in rural and urban households and industries throughout the country in 2000. Ninety percent of all wood fuel used in Kenya was in rural households, 2% in urban households and 8% in cottage industries17. While charcoal was the dominant form of energy consumed in urban households, when averaged across the country, 46% of all charcoal was used in rural households, 36% in urban houses and 17% in the cottage industry. Modern energy

(electricity, LPG and kerosene) made up only 3% of the total energy used in households, while petroleum and electricity used by the agriculture, transport and commerce industries made up only 17% of the country‟s total energy use (Kenya

16 „other‟ fuel in this instance could refer in part to lighting from biomass fuels through open fires, as there was no category detailing that in the survey. 17 „Cottage industry‟ refers to home-based income generating activities like beer brewing, bread baking and fish drying. It may also include fuel used in charcoal production and brick making.

72

Ministry of Energy, 2002). The use and production of biomass energy is very much a major concern for Kenya.

100

90

80

70

60 Charcoal 50 Solid Wood 40

30

20

Fractionusedfuelof category each by 10

0 Rural Urban Cottage Industry Categories of fuel users

Figure 3.7: Fraction of charcoal and solid wood fuels used by rural, urban and cottage industry users. Adapted from Kenya Ministry of Energy (2002).

Fuel Deficit Scenario

Maps and data generated from the Woodfuel Integrated Supply/Demand

Overview Mapping (WISDOM) methodology analysed spatial wood fuel production and consumption patterns in several African countries. In Trans Nzoia, two-thirds of the area showed a high or medium-high fuel wood deficit. When compared to the country- wide statistics, Trans Nzoia is in a much more strained position (Drigo, 2006).

Table 3.1: Comparison of Fuel Deficit Scenarios in Trans Nzoia District with the country of Kenya Deficit Surplus Medium- Medium- Medium- Medium- High Balanced High high low low high Trans % of 29% 48% 1% 1% 2% 6% 12% Nzoia population All of % of 27% 29% 5% 7% 6% 20% 7% Kenya Population Note: adapted from Drigo (2006).

73

Immediate Response to the Fuel Deficit What then are the strategies employed by people who rely on wood fuel for practically all of their energy needs, yet face an extreme deficit in its supply? Firstly, one can use alternate fuels. In most cases the alternates (agricultural residues and animal waste) are not purchased, but they are inferior in terms of combustion efficiency (Arnold et al., 2003; World Bank et al., 1994). Secondly, a household can divert money from other essentials such as food, medicine, or schooling to purchase commercial fuels, an expense that was previously avoided (Arnold et al., 2003; Drigo,

2006; Rehfuess, 2006).

Thirdly, one can decrease the amount of fuel used by modifying cooking techniques. Coping strategies of this nature include serving food cold or employing less fuel-intensive cooking methods including using fuel-efficient stoves (Arnold et al.,

2003; Clancy et al., 2003). Foley (1984) noted seeing women pouring water on a fire or smothering flames immediately with sand to conserve fuel wood. The subsection

Cost in Health on page 13 of Chapter Two details some negative health impacts of modified cooking methods listed by Drigo (2006). Fourthly, one can continue to use fuel at the same rate, thus exerting unsustainable pressure on available woody biomass sources, leading to further depletion of bushes and trees and increased adverse environmental and social effects (Drigo, 2006).

Curiously, a paradox exists when considering fuel wood deficits and fuel saving techniques. In places where fuel is plentiful, for instance where land has recently been cleared for agricultural purposes, or trees are being grown specifically for the purpose of fuel production, people consume much more of it. Where fuel is scarce, people are already employing multiple fuel-saving strategies. People would save more wood in places where no deficit is felt, however those people would have significantly less incentive to use a fuel-saving stove for its fuel saving capacities (Foley et al.,

74

1984). Per capita consumption of wood for energy in sparsely populated rural areas is consistently less than in rural settlements (FAO, 2006b), likely because people tend to settle densely in places where trees and plants grow well (a place where wood is available to be cut) and live sparsely in environments which do not favour the establishment of much life at all (plants or people).

History of Chepkube

The chepkube stove was developed in the 1970‟s in the Nandi communities of the Cherangani and Nandi Hills by women who wanted to keep their kitchens clean by containing the ash from the fire as well as keep food warm so they wouldn‟t be disturbed to cook more than once in a day. The technology quickly spread to other

Kalenjin communities as the Nandis were a Kalenjin sub-tribe and their language, customs and traditions were similar to several other neighbouring Kalenjin subtribes

(see Appendix 5 for a timeline showing the evolution and spread of the chepkube).

People said that the food warming ability of the chepkube was its most salient feature that made it popular initially. When asked why they used a chepkube, most Kalenjin participants said something like “I‟ve never seen another kind of stove” and others said of the Kalenjin: “chepkube is just like culture . . . they are used to doing it for a long time . . . their mothers and grandmothers used it too”.

Gradually non-Kalenjin people started buying land in areas that was primarily

Kalenjin (i.e. Cherangani) and learned about the stove, its benefits and its construction from their neighbours. Chepkube stove users also started to move to other areas such as Kaplamai and Kiminini, thus spreading the technology more.

Simply knowing about the chepkube was not enough to make people start to use it, because as one participant pointed out, “some [people] walk, have exposure, and

75

have some small knowledge but of the ones that walk, some think differently and some don‟t”, however it was an important first step.

The use of chepkube was so widespread in some communities (Kiboino in

Cherangani and Kemeloi in Kaplamai) that it was impossible to find someone who had lived there for longer than a year that lacked one. In those areas, people started using chepkube “just to be uniform with others in the area” (4.2.4), regardless of tribe.

Summary

This chapter briefly describes the relevant cultural, social, geographic and historical contexts for the research. Information was sourced from available literature, informal interviews and participant observation. Trans Nzoia is unique in Kenya for its cosmopolitan ethnic mosaic as a result of land settlement patterns. Maize is the primary crop throughout the region, which is the key ingredient in two of the staple foods in the area and also a valuable fuel. Of particular interest were the ethnic patchworks that made up Trans Nzoia district, the cooking habits of participants, the local fuel deficit situation and the types of stoves that were common in the area. The following chapter describes the methods used to select participants and collect data as well as the rationale behind choosing those methods.

76

CHAPTER FOUR

METHODOLOGY

Introduction

As demonstrated in the Chapter Two, for FES promotion programs to effectively translate into active FES use, barriers to and facilitating factors of FES use must be identified. Chapter Three described the factors related to the geography, agriculture, and ethnic groups that formed the context in which this research was based. Although studies have been carried out in other parts of the world, there have been none carried out in Trans Nzoia District. This research therefore was exploratory in nature as it sought to identify factors that affected the choice of rural residents in that area to use FES. The data collection methods chosen lent themselves to a qualitative analysis of data and provided triangulation of data sources. These methods included: review of literature, focus groups, semi-structured individual and group interviews and participant observation.

This chapter describes and provides rationale for the research approach taken, which engaged a wide diversity of stakeholders who held knowledge about the use of

FES in Trans Nzoia. It outlines the data acquisition methods used in the study, and how data was qualitatively analysed. Lastly, the limitations of the research approach are presented.

Research Strategy

This study follows a philosophical orientation known as social constructivism, which “emphasize[s] the actor‟s definition of the situation” and seeks “to understand how social actors reorganize, produce, and reproduce social actions” as well as share

77

understandings of life circumstances” (Schwandt, 2007, p. 39). When this approach is taken, one must be prepared to hear truth as the teller sees it, and not attempt to align them to some pre-determined framework that is created outside their reality.

Therefore, the research is rooted in the belief that subjective meanings are “formed through interactions with others . . . and through historical and cultural norms that operate in individuals‟ lives” (Creswell 2003, p. 8). This research focussed on the thoughts of average community members rather than policy makers and program managers.

An epistemological framework that arises from this orientation is called standpoint theory, which essentially reasons that if one wishes to know what drives someone to do something (in this case, use a new style of stove), it is sensible to talk to them (the people doing the action) directly, ask questions about their behaviour, why they might do that instead of something else, and what influenced them to act that way (Letherby, 2006). This theory originally developed out of feminist criticisms of the absence of women‟s voices or the marginalisation of them in social science research. The theory is now used in studies such as this one that is interested in the standpoint of the relatively disadvantaged rural poor by researchers who believe that the participants themselves are experts of their own lived experiences (Maynard,

2004).

Partner Organisation

This research was carried out in partnership with SCC-Vi Agroforestry, a

Swedish-based agroforestry organisation actively working in four East African countries since the early 1980‟s. Their mandate is to “contribute to improved livelihood and empowerment of small scale farmers in the Lake Victoria Basin through sustainable management of natural resources and farm enterprise development” (SCC-Vi, 2004, p.

78

2). They are an international NGO with several foci: training and community empowerment, monitoring and evaluation, and extension through community groups.

Field coordinators carry out extension activities with a large network of community groups within two divisions in the study area as well as several divisions outside the study area. In the broader picture, they also worked in another province in Kenya as well as five locations throughout Tanzania, Uganda and Rwanda18 (SCC-Vi, 2004).

Research Assistance

The majority of this research was carried out by a team comprised of two research assistants and myself. Before the assistants were hired, I was accompanied by a member of SCC-Vi staff on all visits. One female and one male research assistant were chosen because older people in the research area, especially older men felt much more at ease conversing with someone of their own sex. Both assistants had degrees from post-secondary institutions, grew up close to or in Kitale and were fluent in English, Kiswahili and a mother tongue. Their familiarity with culture was invaluable as they often pointed things out that would have otherwise gone unnoticed, such as accents of people denoting a certain mother tongue, markings in the yard indicating a freshly dug grave, or possible reasons why someone ate a certain type of food19. They were heavily relied upon as cultural as well as linguistic interpreters.

18 I became familiarised with the activities and philosophy of SCC-Vi in Kitale when I volunteered with them for three weeks in early 2006. I approached them about partnering with me for my masters research in late 2007, and they were favourable to the idea, giving me a list of proposed topics they wished to see research done on. The topic of social and cultural factors affecting stove choice was the most appropriate for my area of interest and field of study. Although SCC-Vi has similar projects in four East African countries, I preferred to work with the Kitale project because I knew the staff, I knew the area, and I already had an established support network in town. I also had a basic knowledge of Kiswahili, so wanted to be in a place where that could be exploited. 19 In one memorable case, we interviewed a woman who was feeding her children and herself a porridge made of the by-products from brewing local beer. From this observation as well as one or two other casual remarks we could surmise that having a cooking device that could accommodate a large beer brewing cauldron was important to her.

79

Previous employment with SCC-Vi conducting surveys meant that the assistants knew each other, were familiar with the philosophy, approach and activities of SCC-Vi, and familiar with the work ethic that was expected within the organisation. For instance, they were well aware that SCC-Vi did not remunerate survey participants, so when this question arose they were easily able to answer with authority. This helped immensely to decrease the initial time spent on orientation to the project and decrease miscommunication during the study.

The male assistant was ethnically Bukusu20. His fluency and ease at translation between Kibukusu and English was advantageous on several occasions when older

Bukusu participants were more expressive in their mother tongue. He was very personable and was generally the spokesperson for the research team upon arrival at new households. The female assistant was Kamba, an ethnic group that is minimally present in the Trans Nzoia area. Kamba were among those that were persecuted during recent post-election violence. Although almost every participant ensured that they knew what group both assistants were from before starting the interviews, they did not seem to respond differently to her.

In addition to arranging for two research assistants, assistance was received from the field coordinators at SCC-Vi whose areas the study was based in terms of coordination of meetings with local administrators and community focus groups and key informants. The four coordinators gave varying degrees of assistance and participation, often due to communication difficulties rather than lack of effort on their part. The managers at SCC-Vi also made a vehicle and driver available to be hired for a reduced rate when travelling to the farthest areas (Cherangani and

Endebess).

20 Bukusu is a subtribe of Luhya, a community that was widespread within our study area. See the previous chapter for more information on this ethnic group.

80

Area Selection

Data collected by baseline and annual monitoring surveys carried out by SCC-Vi from 2004-2007 showed vastly different usage rates of FES use across focal areas21.

After consultation with managers, field coordinators, extension and monitoring and evaluation staff of SCC-Vi, it was decided that the study would look at areas with a range of FES usage rather than focus solely on areas with a history of low or high FES use. It was decided that the reasons someone did not use FES were just as interesting as the reasons someone did use one.

The study areas were chosen from thirty-two focal areas in Trans Nzoia

District22 using two criteria. The first criteria allowed comparison of areas with different rates of FES usage. Focal areas were grouped into four categories based on use of a fuel-efficient cooking device within the household. Focal areas were then randomly selected from each category.

The second criteria sought to minimise the chance that rural participants had experienced post-election violence within their area. SCC-Vi Field coordinators were consulted regarding the extent of disturbance that occurred within each of the chosen focal areas during the previous six months. In some cases, notably in the category with the lowest use, several potential study areas were disqualified this way.23

Subsequently, two villages were chosen randomly from each focal area based on lists obtained from SCC-Vi field staff, sometimes with the aid of area chiefs.

It is important to note that while SCC-Vi preferentially collaborated with groups in communities with high percentages of small scale farmers (farming less than five acres of land), that bias did not affect the choice of villages for this study. This

21 „Focal area‟ is the term used by SCC-Vi, whereas „sub location‟ is the analogous government term. Since all SCC-Vi data is partitioned according to focal area, I opted to use this term 22 See Appendix 4 to understand the organisation of various administrative areas in Kenya 23 See Appendix 6 for a list of average rates of FES use and notes regarding qualification in the study

81

meant that in some areas SCC-Vi was well known, while in others the staffs were not familiar with the residents, and neither were the participants familiar with SCC-Vi activities. In all cases, being associated with SCC-Vi brought about increased rapport, especially with small scale farmers.

Figure 4.1: Method of choosing study villages

Individual vs. Household Unit of Analysis

Individuals rather than households were chosen as the unit of analysis based on several assumptions. Firstly, it was assumed that the multifaceted arrangement of relationships between members of the household and members of the community could not be collapsed into a simple function resembling that of an individual decision maker, as is the case when employing a theoretical approach such as the New

Household Economics model (Moser, 1993). Choosing individuals assumed that responsibilities were divided primarily along gender lines in rural Kenyan households so

82

that the role of a decision maker was shared by different people under different circumstances. Secondly, it was assumed that labour was divided inefficiently, i.e. the gendered division between women‟s unwaged and men‟s waged work was not

“legitimized as both functional and efficient for household welfare maximisation”

(Moser, 1993, p. 21). Finally, it was assumed that each member of the household would habitually make decisions that were not utility-maximising for all members of the household. For instance, men may not always fully understand the issues the female members of their households faced in terms of childcare, cooking, and fuel and water procurement because they have never (or rarely) done it.

Table 4.1: Timing and order of research activities

2008 Task May June July August Preliminary information gathering from SCC Vi staff, women‟s stove X X making groups Determined study area, wrote first X draft of interview guide Hired and oriented research X assistants to the project Practiced semi-structured interviews using the interview guide, and X X revised it. Conducted focus groups, and interviewed key informants. Held in-depth semi-structured X X interviews with 32 households Conducted feedback sessions for the X X study participants and SCC-Vi staff Compiled report of preliminary X X findings

83

Upon arrival in a village, a random starting spot was chosen and a random number of houses (3-6) were counted. The research team attempted to fill a strategic quota sample of two FES users and two non-users in each village by resorting to snowball sampling techniques once a quota had been filled by strategic sampling, however we were unable to fill the quotas in a few villages due to lack of candidates.

Figure 4.2: Method of choosing participants for participation in semi-structured interviews in rural villages (continued from Figure 4.1)

When possible, an adult male and female from each household were interviewed. This resulted in 40 interviews from 32 households. This sample was extremely small compared to the population of the area (see the table below) so it did not allow for quantitative analysis, and the results of the study are not statistically valid.

84

Table 4.2: Total population and number of households in Trans Nzoia district Total Projected total Selected population Total population at growth District Divisions (1999 households rate of 3.8% for 2008 census) from 1999 census Cherangani 58 194 11 377 81 386 Trans Endebess 61 481 12 398 85 983 Nzoia Kiminini 64 685 12 601 90 463 Kaplamai 89 858 16 901 125 668 Trans Nzoia Total (all divisions) 274 218 126 994 861 677 Source: SCC-Vi (2006), 2008 projection revised by author based on guidelines used by SCC-Vi

Kenyan Culture and Visitors

Kenyans are generally very welcoming to visitors. Visitors are treated with respect, and made to feel comfortable to stay as long as they like. When the research team approached houses for interviews, the host or hostess stopped his or her business, greeted the team warmly and offered chairs to sit on. Only once we were all settled did we state our names, affiliations, and intentions, and ask to interview them, a question to which they always responded in the affirmative24.

Literature Review

Questions posed of study participants were based on findings of previous studies that proposed hypotheses of factors that affected stove use, and uncovered lessons learned about technology adoption. Literature pertaining to gender and policy making were also examined. The results of the literature review were used to make the semi-structured interview questionnaires and establish context for the study.

24 I learned by the end of the research period why people never seemed to be in a hurry for us to leave even though they had work to do: it was because it would have been a sign of rudeness on their part to indicate that they wanted a visitor to leave.

85

Table 4.3: Methods used to gather data to answer each objective and their relative importance Method

Objective Focus Semi-Structured Participant Literature Group Interviews Observation Review 1. Role of X XXX XX XX traditional stove 2. Gender needs XXX X X 3. Perceived need X XXX XX for FES 4. Gendered XX XXX X perceptions 5. Community level X XX XX training Note: XXX= most of the data was gathered using this method, XX= some data was gathered using this method, X= a comparatively small amount of data was gathered using this method

Focus Groups

Focus group participants were selected by field coordinators of SCC-Vi. There was a bias in their choice of participants, but for the purposes of the study, it was of no consequence25. Although groups of a certain sex and user category (user or non- user of FES), were requested, one group was made of both men and women, and all groups were heterogeneous in terms of FES usage26.

Two focus groups on FES were conducted. One was with a women‟s group made of a mixture of ethnicities, ages, and experience with FES. It took place outside, in the yard of a house with a nursery school. The other was with select members from an established men‟s group made mainly of members of the Kalenjin community, but with a mix of ages and experience with FES. This took place inside a church. The focus groups were conducted in Kiswahili. One research assistant acted

25 Bias because the field coordinator needed to deliver fertiliser, or because the group was already meeting with people from SCC-Vi that day 26 Refer to Appendix 7 for a list of what was asked for and what was given in terms of study participants

86

as a translator, I facilitated and wrote notes in English, and the other assistant wrote in Kiswahili.

Semi-Structured Interviews With Individuals and Groups

Interviewing is arguably the most basic mode of enquiry that helps one to understand the experience of people and the meaning they make of that experience.

It “allows us to put behaviour in context and provides access to understanding… action” (Seidman, 1991, p. 4). For the inexperienced interviewer and researcher, maintaining some structure was advantageous to introduce some form of comparability to the data, and it also required fewer interview skills as compared to a free-flowing model (Kumar, 2005).

The research team conducted thirty-five individual and group key informant interviews to gather two main kinds of information27. Key informant interviews followed a loose structure to elicit particular information. Questions about the history and current socio-political situation of their area were asked of local administrators and government officers as well as personal thoughts about the use of FES in their communities. Trips to Eldoret and Kakamega yielded interviews with officers from the

Ministry of Agriculture, Ministry of Energy, and a representative from GTZ about the history of the chepkube, kuni mbili and other FES models. Other key informants were asked questions about cultural beliefs associated with the TSS or chepkube stoves and about their thoughts about why people do or do not use FES. Some key informant interviews were in the original study design, some were arranged upon recommendation by a previous participant, and some were completely spontaneous.

Individual semi-structured interviews were carried out with 40 rural residents from 32 strategically sampled households and lasted between twenty minutes and one

27 Refer to Appendix 13 to see a list of interview participants and their conversation topic

87

hour, depending on the willingness and ability of the person to give information.

These interviews were conducted using the interview guides in Appendices 8 and 9.

They took place both outside and inside the homes and kitchens of participants.

The social nature of life in rural Kenya was such that curious neighbours, children, and other household members would inevitably come and leave the conversation at different points, although the conversation usually stayed with only one person. It was noted when someone other than the primary participant contributed information, and that dialogue was allowed to proceed unimpeded. Only if the person‟s spouse interrupted would they be asked to save their responses for their own interview.

Figure 4.3: Local children join the research team in a kitchen during an interview

Interviews with rural residents were usually conducted in Kiswahili or Kibukusu and translated to and from English (by the end of the study period some were in

Kiswahili only) however interviews with administrators or Ministry of Agriculture staff were conducted exclusively in English. Questions were not always presented in the

88

same order and all questions were not asked to all people. If the participant

mentioned something unusual, a new line of questioning was pursued to fleece out

new information. As the principal researcher I directed the lines of questioning in the

interviews, but the research assistants also asked additional questions periodically to

clarify points.

Participant Observation

Using both eyes and ears to observe purposefully, systematically, and selectively was a subtle and completely unobtrusive way to gather data (Schwandt,

1997). Throughout the research period the research team observed the community, the market place, the roadsides, woodlands, gardens, households and gatherings of people that we interacted with. The notes gathered formed a base for and augmented additional information that was expanded on further through interviews. Often the research assistants noticed many intricacies of traditional life that escaped me as someone new to the area and culture.

Feedback sessions

After all interviews and focus groups were completed, a series of eight feedback sessions were held. The purpose of the first seven sessions was to give information back to the communities that had contributed information to the study, as well as verify some of the preliminary analysis. The field coordinators for each area were designated the task of finding suitable meeting places in each village.

Participants were invited to these pre-arranged places at the conclusion of each semi- structured and key-informant interview and were encouraged to bring with them friends and family who may be interested to learn about FES.

89

Table 4.4: Feedback session dates, locations and attendance Date and Time Location Number of Participants28 July 26, 10am EN/Endebess- Mtarakwa – Catholic Church 8 July 26, 2pm EN/Endebess- Cheptantan – Chorlim Primary School 40 July 28, 9am KA/Sitatunga- Sokomoko – Sara Maridadi‟s house 20 July 28, 2pm KA/Sitatunga- Kemeloi – Kemeloi Primary School 8 July 29, 9am KI/Nabiswa- Masaba B – ACK Church No one July 29, 2pm KI/Nabiswa- Nabiswa A – SCC-Vi Office 4 July 30, 10am CH/Cherangani- Kachibora – Ministry of Agriculture 25 Office August 5, Kitale- SCC-Vi Olaf Palme Agroforestry Center 25 10am Conference Room Note: all feedback groups were male and female except Nabiswa A, which was entirely male.

The content of the rural feedback sessions were fairly uniform (Appendix 10 shows a sample schedule). Each FES was introduced, and their benefits and disadvantages were described. Each member of the research team took part in facilitating a section of the session. Questions were also asked of the group, and some time for dialogue took place at the end. Fireless cookers, two examples of portable kuni mbili liners as well as a metal sawdust stove were the only stoves physically present. Other stove models were shown using large laminated photographs as well as poster-sized sketched drawings.

Although most semi-structured interview participants confirmed they would attend the feedback times, there were a surprisingly low number of participants that returned (four in total over three sessions). This low number surprised the entire research team, because we were convinced the intentions of the rural participants were sincere.

28 Most feedback sessions were attended by people who arrived and departed at different times. No official attendance or count was taken, so the numbers listed here refer to our best approximation of the total number of attendees.

90

Figure 4.4: A research assistant and the driver discuss FES with rural residents in Mtarakwa during a feedback session.

Most people who attended feedback sessions were either gathered randomly (in

Mtarakwa, Sokomoko, Kachibora and Nabiswa A) or were recruited by messages sent home with students or were teachers recruited by principals (Cheptantan and

Kemeloi). The result was that the flow of information did get reciprocated within the communities, although it did not go directly to the people who were involved with the study. The formality of the sessions varied depending on the number of participants.

In Sokomoko people gathered on benches in a woman‟s yard, in Nabiswa A men ate rice out of cupped hands and in Kemeloi the feedback session started with one person and ended with eight.

91

Figure 4.5: A research assistant explains FES to a group of men in Nabiswa A.

The most interesting feedback session happened in the grounds of the Ministry of Agriculture office in Kachibora, the main administrative center for Cherangani district. The research team waited for almost two hours for former participants to arrive. When none did, I walked back to our truck to return the demonstration stoves.

At that point, some curious women approached the truck. After being shown what the

FES were and how they worked, these excited women in turn began teaching others.

Soon, up to 25 people had spontaneously gathered in several small groups examining the stoves and laminated photographs of permanent FES, one explaining to the next what they had learned.

92

Figure 4.6: The author with a spontaneous gathering of rural residents at a FES feedback session in Cherangani

There were three reasons speculated for low attendance of study participants to the feedback sessions. First, although the physical locations were as neutral as possible (two churches, two schools, two offices and one home) and the sessions were mostly in the same village the interviews took place in, the commuting distance still ranged from 100 metres to five kilometres. In two cases, due to logistics, the session was in a different village. The market center Kachibora was chosen as a central meeting place for the two villages in Cherangani and feedback for Cheptantan village in

Endebess was held at a neighbouring village (four kilometres away) because of a more suitable meeting place.

93

Figure 4.7 Parents and teachers gathered at a feedback session in Cheptantan

Secondly, the timing was not optimal. The end of July was the time for harvesting beans, which nearly everyone planted. Also, since there was a span of up to one month between interview and feedback session, the date may have simply been forgotten. Third, the style of invitation was not ideal. Invitations were given verbally rather than on a piece of paper, but if the process was repeated, it would be recommended to distribute small paper invitations so that it would appear more important, and the paper would be a guide to record the details of the gathering.

The final feedback session was held at the SCC-Vi Agroforestry offices in Kitale.

Gender and Home Economics staff from the Ministry of Agriculture were invited as well as all staff from SCC-Vi. A powerpoint presentation of the preliminary results of the study was presented, followed by a discussion with participants about the findings.

Each attendee was also asked to fill out a short questionnaire to indicate which findings they agreed or disagreed with, as well as what areas they felt should be further researched29.

29 Refer to Appendix 11 to see the questionnaire

94

Analysis

Interviews and focus groups were recorded as accurately as possible in both

Kiswahili and English during the interview itself. As soon as possible (usually after two field days) the research team compiled notes to reconstruct the interviews or focus groups. At that time significant details of the context (i.e. participant observation) were discussed and extra explanatory notes supplemented the corresponding interview notes.

At the end of the research period, data from focus groups, semi-structured interviews and participant observation were analyzed using open coding. Quotes were extracted from reconstructed transcripts and observation notes and organised first according to specific research questions, then according to sub-themes within each broader category. The theme names arose from the text and notes themselves by using the in vivo coding technique, which uses phrases or words repeated by the participants (Bernard, 2000). Special attention was shown to gender, age and house composition variables to elicit differences between males and females as well as among males or among females in different geographies of different ethnic groups.

Limitations

The methods and style of research had several limitations. Firstly, the flexibility of semi-structured interviewing meant not all questions were asked to every participant, so most responses could not be reported in a quantitative manner.

Additionally, nearly every question asked gained null or neutral responses at some point. Either respondents could not think of or did not know an answer, or their answer was “none”, or “no difference”. Except in instances where a lack of answer or

95

neutral stance was deemed a significant finding, these responses were not listed in the findings.

Semi-structured interviews were often conducted out of doors in the presence of curious bystanders. It is expected this had effects on participant‟s responses, however it is unknown if the effects were positive or negative. Also, the general chaos of children and animals running around, food that needed to be cooked or shops that needed to be tended were among general distractions that condensed several interviews.

The simple presence of a foreigner may have also influenced the responses of participants. In one instance, members of a women‟s group suggested (in Kiswahili) to one another that they should focus their responses on a particular technology in order to get more empathy and funding from the foreigner in their midst. In other cases the presence of a foreigner likely enhanced the willingness of participants to take part in the study. Some scholars also believe that the presence of multiple people in focus groups and group interviews has unforeseen consequences (both negative and positive) on the responses of participants.

Finding common language and meanings was at times difficult. Although

Kiswahili was a common language for people in Trans Nzoia, it was not their mother tongue, so their vocabulary lacked richness. The English word stove had multiple words in Kiswahili however the word fuel did not have a commonly used equivalent.

The research team quickly identified patterns of misunderstanding and worked to correct them by adjusting the phrasing of questions.

Communication with local staff (of SCC-Vi and the Ministry of Agriculture) was often difficult. Cultural differences gave way to mutual misinterpretations of meanings. Communication was also hampered by the use of cell phones that had poor

96

coverage and loud background noise in public transit (which users were often using at the time of the phone call). In-person meetings did not happen often, so it did not provide an effective space to pass on detailed messages.

There were also limitations related to maps and directions. Maps were difficult to find or have drawn. The differences between secondary roads, paths and trails were often difficult to discern. There were absolutely no road signs off of major arteries, and no delineations between villages. A series of three adjacent homesteads could give three different names for what they considered to be their village. When searching for a certain place, bystanders would be questioned at every fork about which direction to go, for there was inevitably a village center at each fork. This process was repeated until the village was found.

Sometimes, the villages that were selected from lists provided were not villages at all30. Cheptantan village in Endebess was selected after an unsuccessful search for a place called Chorlim. Chorlim was the name of a large scale farm owned by the Agricultural Development Corporation and also a primary school, but no one knew of a village with that name. Cheptantan was the next village after Chorlim primary school. Mtarakwa was chosen after several failed attempts to locate Kostoi.

For the sake of the project the name of the village was inconsequential as long as it was randomly chosen, which both Cheptantan and Mtarakwa were.

Summary

Despite these limitations, data was successfully gathered from focus groups, semi-structured individual and group interviews encompassing nearly 100 participants by a research team of three people.

30 A village was defined by the person who made the list. The Ministry of Agriculture, SCC-Vi and area chiefs all had different lists of villages within exactly the same area. For the purposes of this research, the definition of a village did not matter, since it was essentially used to

97

This study used a constructivist approach and standpoint epistemology to identify a myriad of factors that affected the use fuel-efficient stoves within rural households. To do this, three primary data collection techniques were employed

(focus groups, semi-structured interviews and participant observation). Information was also gathered from available literature. While an initial list of participants was compiled, the research approach was flexible and pragmatic, so data collected from people outside of the initial plans were also analysed.

The study areas were chosen from data collected using stratified sampling techniques. The study areas were chosen for their relative assumed variation in usage of fuel-efficient cooking devices. The study participants were chosen using strategic and snowball sampling. Study participants included rural residents, local administrators and government officers. The following chapter analyses the data collected from these participants.

98

CHAPTER FIVE

DATA ANALYSIS

Introduction

This chapter summarises the research findings and presents them in relation to the five study objectives. The findings are drawn primarily from semi-structured interviews with forty rural residents, but also include data gathered from sixty more participants through involvement in focus groups and additional semi-structured interviews. The narratives of research participants are presented throughout the chapter to give voice to different stakeholder perspectives encountered during the data collection process.

The first section gives a brief profile of participants who took part in the study.

The second presents the significance of a TSS within a rural household and gives reasons for its continued presence. Gendered differences in FES construction, use and benefits are presented in the third section. The fourth section outlines the perceptions about saving fuel and FES use within the communities. It also includes accounts of different FES models encountered during the study and their individual characteristics.

The fifth section presents differences in how different segments of the community view and interacts with FES. The final section provides an understanding of how people learned about fuel saving technologies in Trans Nzoia.

Semi-Structured Interview Participant Profiles

Personal Data of Semi-Structured Interview Participants

Profiles of all rural residents interviewed using the semi-structured question guides are provided in Appendix 12. Statistics used to make the figures in the next few pages refer only to the interviews that took place in the eight randomly selected

99

villages, not the semi-structured interviews conducted as practice in other areas.

Forty rural residents were interviewed using the semi-structured interview guide during the main part of this study. They are referenced throughout this chapter using a code that refers to their district, village, and household31 with a or b added to differentiate between multiple interviews within a house.

Twelve participants were male and 28 were female. They ranged in age from

16 to 68, with mean ages of 36 (female) and 39 (male). Figure 5.1 shows the uneven distribution of ages according to gender. Most women interviewed were between the ages of 21 and 60, but only one man was between the ages of 31 and 50. Data from the practice interviews were similar to this, with the exception of one elderly couple who were ages 78 and 97. There was not an equal representation of men who were of the same age cohort of most women interviewed. As a result, the responses are biased toward representing perspectives of older and younger men and middle-aged women.

10

9

8

7

6 male 5 female 4

3 number of participants 2

1

0 <20 21-30 31-40 41-50 51-60 >60 age categories

Figure 5.1 Age distribution of semi-structured interview participants (n=40)

31 For instance, 2.1.1 is Kaplamai division (2), Sokomoko Village (1), first household.

100

Female participants had between 0 and 12 years of formal schooling (average

7.0 years), whereas men had between 4 and 16 years32 (average 8.9 years).

Households were comprised of an average of number of 6.4 people, with a range from three to 13 (see figure 5.2).

12

10

8

6

4 number number of households

2

0 3-4 5-6 7-8 9-10 11-12 13-14 unknown family size (number of people)

Figure 5.2 Number of household members33 of semi-structured interview participants (n=34)

Four major tribes were represented in our sample of semi-structured interview participants. Kalenjin were technically the most numerous (18) but since the Sabaot subtribe was so different in terms of culture and stove usage they were routinely treated as a separate category (see figure 5.3). Nandi was the next most numerous

Kalenjin subtribe, so they are often referred to separately in the text. The Luhya population was primarily made of the Bukusu subtribe (11 out of 15) so they are also

32 One male participant was a retired veterinarian, so I calculated he had at least 16 years of schooling. 33 Determining household size is difficult when polygamous families and non-resident household members are considered. Therefore, the data presented in this table refer to the question “how many people do you cook for?”

101

often referred to separately in the text. Other ethnic groups encountered in practice interviews were Kamba (1) and Kikuyu (4).

16

14

12

10

8

6

Number Number of participants 4

2

0 Luhya (including Sabaot Kalenjin (including Kisii Turkana Bukusu) Nandi) Ethnic group (tribe)

Figure 5.3 Ethnic affiliation of semi-structured interview participants (n=40)

Twenty-three households had access to between 0.1 and seven acres of land

(see figure 5.4), with an average use of 1.7 acres. Four households had larger pieces of land somewhere in the 35-50 acre range, and data was not collected for five households. Twenty-six households had a male head (see figure 5.5), and at least four of those were polygamous households. Seven households were headed by women who were most often widowed and one was comprised of orphans, the oldest of which was over 20 years old.

102

8

7

6

5

4

3

number of households of number 2

1

0 0-0.4 0.5-0.9 1-2.9 3-5.9 6-8.9 9-19.9 20+ unknown land size (acre)

Figure 5.4 Size of land used by semi-structured interview participants (n=33)

30

25

20

15

10 number ofnumber households 5

0 Male Female Adult children household head

Figure 5.5 Description of heads of households of semi-structured interview participants (n=32)

Other participant profiles

Profiles of participants from whom information was gathered without the use of the semi-structured interview guide, including individual and group key informant interviews, focus groups, feedback sessions and participant observation are shown in

Appendix 13.

103

Seven group interviews were conducted with rural residents, two of which were highly structured focus groups with four specific questions, three were unstructured interviews with stove producing groups, one was a semi-structured interview with a group of elders, and one was an impromptu group interview with people gathered in a chief‟s office34. These group interviews are denoted in the text with the prefix FG.

Fifteen rural key informants were interviewed in a semi-structured manner. Five were specifically sought for their knowledge of culture, three for knowledge of stoves, four were as a result of spontaneous conversations with residents in town. One spontaneous meeting was with an urban, educated woman who approached me while walking home from town, and two were the other members of the research team who gave great cultural insights and anecdotes and discussed the cultural implications of comments as the interviews were transcribed.

Three pre-planned and one spontaneous conversation with local area administrators were carried out for the purpose of gaining contextual knowledge of the area (denoted by the prefix AD). Seven conversations with Ministry of Agricultural officers who worked in the study area gave insight into the Ministry of Agriculture‟s programming in the area and their personal thoughts on why some FES are more popular than others. Additionally, interviews were held with three Ministry of

Agriculture and one Ministry of Energy officers in the city of Eldoret to talk about a locally developed stove that brooded and hatched chicks. Later, a Ministry of

Agriculture officer seconded to GTZ gave insight on the history of stove promotion programs through the larger geographic area. These are all referred to with the prefix

AG.

34 In this situation, I had arranged an interview with the chief, who turned out to be a difficult individual. When my research team arrived, he informed us that we were there to speak to the people assembled in his office (who, incidentally, were there to address the chief, and were not interested in us). Instead of lecturing the people, I asked them questions about the stoves they knew about and used.

104

Perceived Advantages of the TSS

In communities where chepkube was the primary stove, the thought that TSS could out-perform a clay stove, or have any positive characteristics was almost laughable. Despite this, most people though could list at least one advantage of TSS.

Ease of Construction and Maintenance

Above all, participants noted that TSS was easy to construct, both in terms of expertise and materials needed. It was also portable, the materials were readily available and it did not break easily or need repairs. A woman in Cherangani mentioned “older people use the TSS because there is no expense of purchasing materials. They can just collect stones for free” (4.1.2). They were also good in situations where one needed to cook outside, needed a back-up stove or while travelling. One respondent said “if you rent a house, you have to build a TSS outside.

You can‟t cook in the house because the landlord doesn‟t want the roof disturbed by smoke” (4.1.1). A male respondent recounted, seemingly from personal experience,

“you can build TSS outside, for instance if you have gone somewhere for a job, you can take stones, arrange them and start cooking without struggling” (FG2).

Several respondents indicated that fuel, especially maize stalks, were free and plentiful and that the TSS was the best stove to burn them in. A female focus group participant said “the TSS and chepkube are both good because you do not spend much time moving to get kerosene and charcoal because you only use firewood” (FG1). In addition, people said TSS is easier to light and cooks food faster than other options:

“the person who wants to cook faster can use three stones” (4.2.1) because “when you place the firewood from both ends, it makes the cooking happen faster” (FG2).

Some participants even reported that the TSS used less fuel as compared to other clay

105

stove models, especially if using small firewood (4.2.4) or if you are only cooking something small like tea (FG7).

Cooking Advantages

The second most common category of responses was about the usefulness of a

TSS when one is actively cooking. Adding supporting stones or repositioning one or more stones makes it adjustable for small and large pots. It is especially good for stirring a large pot of something stiff and heavy, like ugali. An older woman intimated

TSS were necessary when cooking for a crowd when she nodded towards the 22-year old research assistant and said:

“when [she has] offspring, it would be too difficult to use the other stoves. . . when she is now starting to cook, she is using one hand [to stir], using the stove. When many children are there, your hand gets heavy, therefore you must make a TSS” (3.1.1).

Many women said that the three stones were able to grip the pot, thus stabilising it when it is being stirred.

During a group interview with Bukusu elders, there was general agreement that traditional foods, such as ugali made with millet, sorghum and cassava were prepared better on TSS. One man added TSS was good “because the large firewood made a big fire which cooked githeri very well, so that it was softer and better in taste” (FG3).

Others agreed with him and also spoke about boiling a cow leg on a TSS “so that it was cooked so soft that an old lady with no teeth could eat it” (FG3). The TSS, paired with traditional pottery pots and a period of fermentation resulted in especially tasty preparations of the traditional vegetables sujaa and saga (FG3). Finally, TSS was perceived to be better than other fuel-efficient stoves (FES) for roasting cobs of maize, a popular and wide-spread snack.

106

Most users of the three-stoned stove (TSS) learned about it from their parents, and some participants listed this specifically as an advantage. Knowledge of the TSS had been passed “. . . just like a chain. Great grandparents learned from great-great grandparents, grandmothers learned from great-grandparents, and mothers learned from grandmothers, and daughters from their mothers” (3.1.1). Stories of experiences with TSS are woven into the familiar fabric of their culture.

Perceived Disadvantages of the TSS

Participants listed a broad range of disadvantages of TSS. People who had significant experience with FES had responses related to fuel consumption, inconvenience and safety ready however a few participants who had not seen other models had a difficult time answering the question directly. A middle-aged woman summarised those thoughts this way: “you know, we have been using this stove since we were born so when you are used to something, you can‟t see its disadvantages”

(4.2.3).

Inconvenience

The most commonly mentioned category of perceived disadvantages was the theme of disturbance and inconvenience. Many users of FES thought that TSS was slower to cook, especially when the wind blew or the fire spreads because of lack of walls. One couple explained “a clay stove saves fire, but TSS spreads fire outside and the food cannot be cooked fast. . .because the fire is moving away” (3.1.4a,b). Other people expressed displeasure with smoke when the stove was used inside, with wind and rain when the stove was used outside, and with ash or smoke in food as a result of blowing air. A woman in Cherangani said “when you blow on the fire it causes dust and ash to come into the food” (4.1.1). Two other women spoke of taste being

107

compromised because of “the smell of smoke in food” (2.4.4a, FG7). Chepkube users also were quick to cite that a major disadvantage of FES was that “if ugali is cooked early, it may get cold because a TSS lacks a place to keep it warm” (4.1.1). Also, “if you want to go on a journey, you can‟t keep food warm” (FG1). Which means “. . . you cook twice and use twice the wood” (2.2.2).

Other aspects of TSS use that bothered people were: “when you put a pot on

[the TSS] it will not fit; you need small stones to support the pots” (1.2.3), “when you are washing a pot, it makes your hands to be dirty by soot” (1.1.1) and “the TSS may deform pots easily because of the force of stirring ugali” (FG2). Also, “there is accumulation of ash which makes the firewood not to ignite and burn fast. There is no space for the firewood to breathe” (2.1.1b).

Fuel Consumption

The second category of perceived disadvantages that emerged was the amount of fuel that was consumed. A man from a TSS-using household explained “the TSS, because of it having three openings, wastes a lot of firewood, therefore wastes a lot of heat. . . this causes a lot of firewood to be used” (2.1.1a). Women in a focus group pointed out that wind increased the amount of burning because the fire was not protected (FG2), and women in Endebess recounted “you buy a lot of firewood and you have a long distance to walk searching for wood” (1.2.6) and “the place where we collect the wood, people don‟t want us to go because it is illegal or dangerous because of animals or people that live there” (1.1.1).

108

Increased Danger

The final group of responses pertained to the increased danger associated with

TSS. One woman acknowledged that the TSS “produces a lot of smoke that can cause an eye disease” (1.1.3). Almost everyone else focussed on the fact that the TSS is at crawling level for small children, which often led to injury. A father recounted his fear that “it is so easy for [kids] to touch the pot, and when it turns it may pour something and the kid will be burned” (3.1.4a). Some focus group participants said the

TSS “may cause crawling babies to be burned by the fire” (FG1) because “the flames are not enclosed, and also because food is not reinforced for stability” (FG2). One of my research assistants could personally attest to the results of a child‟s interest with a

TSS, because her hand was completely scarred from her encounter with a pot of boiling water as a toddler.

More irritation and danger came from the instability of pots because rocks broke or moved. A woman in Kiminini said “when using TSS, if you slide or come in contact with the stones, you can make the food to fall” (3.1.4b), whereas a woman in

Endebess remarked “when you get soft stones, they explode. When they explode, they can burn kids” (1.2.2). A man who only got exposed to TSS as an adult summed up his thoughts by saying “when I‟ll be sitting there around a TSS, I‟ll sit with a lot of keen because I‟ll have known that I am sitting next to something that is dangerous”

(3.1.4a).

109

Table 5.1: Perceived advantages and disadvantages of TSS listed by participants

Perceived advantages of TSS Perceived disadvantages of TSS

Easy to construct Lacks a place to keep food warm Portable Stones can explode Fuel is free and plentiful (i.e. User must spend money and time maize stalks) procuring fuel (i.e. wood, maize cobs) Easy to light Difficult to light when ash Cooks food quickly accumulates Uses less fuel than clay stoves Slower to cook than clay stoves Adjustable to stabilise cooking More fuel is used because energy is vessels lost Improves the taste of traditional Cooking vessels are unstable food Food tastes worse due to ash and Better for roasting maize smoke An important part of culture Causes hands to get dirty from soot Better than any other stove when Unsafe for children because of cooking with large pots spreading fire, height and instability Smoke is good to dry wood and of pots keep the kitchen free of insects Fire gets disturbed by wind and rain Produces more heat than other Produces smoke which is stoves uncomfortable and causes eye disease Deforms pots because of heat and pressure from stones

Circumstances Where No Other Style of Stove Can Be Substituted for TSS

Although at least half the people said there was nothing that a TSS could do better than any other style of stove, there were three main themes about instances where this was felt to be the case. First, there were some practical advantages the

TSS had over other stoves, most of which were listed in the previous section on perceived advantages. Secondly people spoke about the significance TSS had to cultural identity, and finally there was significant conversation around the role the TSS played in traditional beliefs and ceremonies.

110

Practical Advantages

Several respondents, some of whom used some form of FES thought that “during ceremonies big pots may be used in a TSS because they fit . . . you can place firewood from three ends to boil [the contents] quickly” (4.1.1). Others thought “with a TSS a lot of people can surround it and get a lot of heat” (2.1.1a) and “smoke [from TSS] is good to keep the kitchen free of insects and dry the [fire] wood” (FG7).

Cultural Identity

A chief in Kaplamai compared the cultural differences in stoves with cultural differences in alcoholic beverages consumed at a teenager‟s circumcision ceremony

(AD4). An officer from the Ministry of Agriculture in Endebess said each ethnic group

“has its own way of using stoves each according to its own” (AG4). For some people having a TSS physically present in their kitchen is a way of identifying them with a particular group. This is especially true for Sabaot and the various Luhya subtribes.

For instance, a woman in Kiminini told us “if you are Bukusu, it is essential that [TSS] be there” (3.1.1). In Endebess, upon being asked if she used a TSS, a Sabaot woman replied by laughing, saying of course, it was the normal thing to use there, as if any other option was absurd (1.1.3).

Traditional Beliefs and Customs

Family

The deepest meaning of the TSS comes from the cultural beliefs surrounding them. In the study area, only Sabaot and Luhya participants reported strong (and similar) beliefs related to the TSS. They explained that “the TSS is a symbol for a man that allows him to keep leadership of the home and to look after children, wife, cattle and everything else” (FG3) and “the three stones represent the husband, wife and

111

children. You need them all to make a meal” (KI9). For some who had embraced the

Christian faith the three stones also symbolised “life and trinity- Father, Son and Holy

Spirit” (FG3). An older Bukusu woman said that when a woman rests one foot on the large stone (the man‟s stone), she felt that could get closer to her husband (KI4).

Marriage

When a couple is married, it was tradition for some mother-in-laws to “make the three stones for her son . . . which were made of two stones and a mound of clay”

(FG3). An agriculture officer who had worked in Trans Nzoia more than 12 years explained that “a stove is given by the mother-in-law, therefore you can‟t throw it out, or else you will be thrown out of the marriage” (AG12).

After getting comfortable in a group interview, some Luhya elders shared a

Bukusu tradition that tested a perspective bride‟s cooking skills in a sort of pre- wedding ceremony. The bride- to-be had to carry food, utensils, and even a broom to clean up from her parent‟s house. She then cooked a meal of traditional (millet) ugali and chicken for her new family using the new TSS. If her new family approved of her cooking skills, she was accepted. If not, she was sent back to her parents and the wedding didn‟t take place (FG3). On a sourer note, a Sabaot woman told us that if a woman married a widower, she must destroy the hearth and make a new stove, out of fear that the previous wife had put a potion on the hearth before she died, which would have an adverse effect on the husband (KI3).

Death

A woman in Kiminini believed “my TSS is not quite traditional because none of the stones are permanent. If my husband were to die, I should make a permanent one just so that I can remove a man‟s stone” (3.1.1), because a traditional Luhya and

Sabaot belief was that if the husband‟s stone is not removed when he dies, “then a

112

curse will come to the household” (FG3). The widow must remove the stone with her foot, not her hands (3.1.1), and then hide it in the forest (FG3). Other Luhya elders said in their home areas “when the husband dies, a whole new stove was constructed.

The new one has two large and one small stone” and “when the husband dies, you must remove the ash from the fire and mix with dung to smear the stones and the hearth. This symbolises that the man is not there, and only the mama and the children remain” (FG3). A Sabaot woman had a similar story. She said that instead of moving his stone, they just removed the ash (FG1). The practices are the same if a wife dies (3.2.4), but most people only mentioned husbands unless pressed further. If children die there was nothing done to the stones. The only exception was in the case of a miscarriage, in which case the Bukusu “dig near the stove and bury the baby near the hearth” (FG4).

Disagreements

The TSS played a major role in marital disagreements or other instances where quarrels had taken place. Firstly, if a woman had a disagreement with her husband and wanted to leave the household, she destroyed the stove or simply removed her stone, and returned to her parent‟s place. If the woman returned (sometimes after coaxing from the husband), they had to hold a ceremony where a sheep (FG3, KI5) or a cow (3.1.1) was slaughtered and a new stove was constructed. Disrupting the hearth was considered serious. Even if stones “meet each other accidentally” by uprooting from their holes during the heating and cooling process, “you must come with a goat and then slaughter it in a ceremony” (1.2.2). A Sabaot man added that at their ceremonies, “when a husband and wife reconcile . . . there is an old person who acts as a counsellor for them” (KI9). The Bukusu research assistant also remembered seeing an old man at a reconciliation ceremony for his uncle and aunt that had a

113

conversation with them about seeking forgiveness, after which his uncle and aunt washed their hands the stomach lining of the freshly killed sheep (KI5).

The same research assistant also remembered seeing a quarrel happen once between two of his cousins, the first-born sons of each of his uncle‟s two wives. Their father took a hen, dislocated the neck, pulled off the head, built a TSS and swung the hen around the fire. Then, he roasted the chicken and each boy had to eat half in order to rebuild the relationship within the family (KI5).

Curses

Although Luhya and Sabaot men did not typically enter the kitchen to use the

TSS, they would touch the stones if they wanted to disown their wife by removing her stone (AG11, FG3) or curse a member of the family. The following stories were told to my male research assistant by an elder from his community he met after a day in the field. In very serious tones she told him that when a wife moves out, especially if she has had an affair with someone, and the man suspects that this is what the wife was doing, he directs her to jump across the TSS. If she was truly unfaithful, then the act of jumping across the hearth would make her cursed, and she would then be free to move out. If she was in fact faithful, nothing bad would happen. Also, if a man wanted to curse his family, he would enter the kitchen naked and sit on the stones, saying things like “move and fall in the river. . . die on the way” or “if you are not truly my children, may this bad thing happen to you” or, to a woman “may you be barren” (KI4). This practice was also traditional in the Sabaot community (KI9).

114

Table 5.2: Cultural beliefs and customs involving TSS mentioned by Luhya and Sabaot participants

Family Marriage Death Disagreements Curses Three stones A man‟s When a man or A man or A man sits represent family and mother must woman dies, woman may naked on the life (father, mother make the TSS their stone move the TSS to utter and children) for his new must be woman‟s stone curses for his bride removed and which means family

replaced with a the woman

moveable one must leave

To Christians, the The new wife Sometimes ash If the woman A suspected three stones could must prove is removed returns, a adulteress is also represent Father, herself by from the fire ceremony must cursed if she Son and Holy Spirit cooking a meal and mixed with be done to jumps across on the new dung to smear make the the hearth

hearth the remaining hearth ready to when directed stones and use again to hearth

The presence of TSS If a woman The stone TSS also a key allows a man to keep marries a should be put part of other leadership over the widower, she in the forest reconciliation household must make an ceremonies entirely new

stove

A woman can rest her If a child is foot on the man‟s miscarried, it is stone to get closer to buried near the her husband hearth

Of all the statements related to cultural beliefs, the ideas of the stones representing the family, removing stones after the head of the household died, and women disturbing the hearth after marital disagreements were by far the most common responses. Most of the other responses were reported less than five times, and some only once. The general feeling among Bukusu was that “even if there is

[another style of stove], there must be a TSS also” (FG1).

115

Household Roles and Benefits Concerning FES

Decisions Making For Stoves and Fuel

There was a distinct difference between the way men and women responded to questions regarding decision making for stove procurement. Eight men spoke about being involved in stove-themed discussions and five said they were never involved at all. This contrasted to two female respondents saying their husbands were consulted and thirteen respondents who said it was only women who thought about such things.

A man in Kaplamai said that he and his wife “would like to buy a kuni mbili stove, and have discussed it as a couple” (2.4.1a), and a woman from the same area “had discussed getting kerosene [with her husband]. . . [and] had also discussed getting a sawdust stove” (2.4.4a). Two women in Endebess “both communicate[d] with their husbands” (1.2.4) and a young man in Endebess said “everyone in the family will do the decision making together” (1.2.3).

Table 5.3: Male and female responses to the question “who makes decisions about stoves in your household?” (n=27)

Male respondents Female respondents “women “women “both” “both” only” only” 8 4 2 13

When men were involved in conversations, they were about stoves that used kerosene (2.1.3, 2.4.4a), charcoal (2.1.3, 4.1.3), sawdust (2.4.4a), were for raising chickens (3.2.3, 3.2.4, FG4) or were stoves that they had seen somewhere, like in a cafe (4.1.3), at a group meeting (2.1.1a, 2.4.1a) or growing up at home (2.1.3, 3.1.4a,

1.2.3). Only one pair of women (who currently use TSS) reported discussing clay stoves with their husbands (1.2.4).

116

There were two cases of men who knew quite a bit about certain FES models

(2.1.1a, 3.1.3), but despite engaging their wives in discussion, their wives continued to use the older models. One man explained that he thought it would be a good thing to be working together with his wife, but in the end he thought “the wife may think he is trying to interfere in her program” (2.1.1a). Both men said that if their wives failed to agree to use another stove model, “then it would be impossible for them to have it”

(2.1.1a).

It was difficult to determine if all respondents meant that the final decision was always the woman‟s but the trend of responses certainly went that way. The majority of women felt that “the duty of the mama is to talk and deal and plan about the kitchen” (4.2.4) and that the husband‟s role “was only to eat” (4.1.1). When queried further about the influence of men if money was involved, one woman replied “there are some things as a mama you have to save and buy, not ask the man” (3.1.1), thus inferring that she would not involve her husband in stove purchases. In terms of fuel procurement, this was largely a woman‟s task, but several people spoke of men supplying the money to purchase it.

Practical Gender Relations Influencing the Use of FES

Subtle practical gender relations came into play when materials were needed for stove construction, money was needed for purchasing materials or labour was needed to build or repair a kitchen or house. There were also some instances of men participating in fuel and kitchen related tasks.

Buildings and Materials

Although one elderly gentleman initially said he would never be involved in anything to do with stoves, later he explained that “. . . the man‟s realm is to make

117

bricks and other materials” that could be used in making a stove (2.3.1b). Other men spoke of preparing clay for his wife to build a chepkube (3.1.4a) and building a chimney because there was too much smoke for his wife (3.2.4). Some women spoke about needing help from male family members to build or repair kitchen buildings and houses before they could start to use a FES (1.1.2, 3.1.2). It seemed that many women lacked money to improve their stoves, but they were not always successful getting money from their husbands. At the end of one long conversation about different FES that a woman knew about, she was asked if her husband could assist her to buy a new stove. The husband, who was sitting in the same room partially listening to the conversation broke in saying he could indeed “assist to give money to buy a new stove”, which resulted in a glare from his wife, who looked like she wanted to quote something quite different (4.2.4).

Acquiring Fuel

Splitting wood seemed to be a task shared across genders, at least in some households. Men were observed splitting wood on two occasions (2.1.3, FB3). One man spoke of his workload decreasing in terms of searching for fuel since his wife started using an FES (KI7). Conversely, other men “have fear and feel ashamed to go collect firewood” so would preferentially use a charcoal stove (4.2.1).

Perceived Benefits for Specific Family Members in FES Using Households

Benefits For Women

Many people argued that women had the most to benefit when an FES was used in a household, since they were the primary users. Indeed, her list of perceived benefits was much longer than any other person. Most benefits related to decreasing stress and increasing flexibility in some way. For instance, one pair of women said

118

“since we have a small business in the market, we would cook food, place it in the cupboard and the kids would come and eat lunch while we were gone” (3.2.1). Food cooked more quickly on a stove with multiple burners, so that eased the stress when there were visitors (FG7) or when they were in a rush to get somewhere like church

(3.1.2). Women also appreciated not staying close to the stove to continue adding fuel, as was the case with TSS (3.2.1). Chepkube brooders gave women the specific benefit of keeping hens as a business (2.2.4b) and one woman reported using a chepkube‟s cupboard to quickly dry maize when it was too wet outside (4.1.1).

A newlywed man smiled when asked what benefits his wife enjoyed from her chepkube. He said “she can be happy because she can‟t spend a lot of time there cooking. . . if I am late outside I can delay, so [she] can just put ugali in that place and I can get it in that place when it is warm, and eat” (3.1.4a). Others acknowledged that if their wives used a FES she would spend less time searching for firewood (3.2.2) and that wood would last longer once it was purchased (2.2.4a). One man remarked that the benefits for himself and his wife were the same, because “they are one family. When he loses, she loses . . . the money he would have spent to do another thing or buy another thing now has to buy the firewood” (2.1.1a). Another young man observed that sawdust stoves and clay stoves gave women safety because “the flame can‟t burn the hands while cooking” and when the wind is blowing, ash does not enter the pot (2.1.3).

Women also identified intangible benefits for themselves like having happier husbands and gaining respect and happiness themselves. Examples of decreasing the incidence of quarrels is discussed below together with the man‟s benefits but would easily apply here too. One woman simply said “[her husband] would be so happy because a change happens” (3.1.2). A woman in Kaplamai explained how her husband

119

would have greater respect for her when she started using a FES and how her sense of well-being would increase as well. She said,

“with the new stove, my husband [would see] me cooking and says „my wife is going on developing‟. . . A woman may go to another home and admire [their stove] and wish she could be the mama there, and despise her own home [because she didn‟t have that stove]” (2.4.1b).

Benefits For Men

If food is the way to a man‟s heart, then respondents seemed to say clean, warm food available quickly was surely the way to keep peace in the marital relationship. A woman who used TSS said “if a TSS is used to cook food when the man is home late, there are quarrels” (2.4.1b) because the TSS was comparatively slower both because of energy loss as well as only being able to cook one pot of food at a time (3.2.1). In contrast, a former user of chepkube said “the husband would be so happy [if you used chepkube] because he will be thinking that you are just coming from the kitchen, and that you had just cooked, but it was from before” (3.1.1). Some enthusiastic young women explained that “[when] the wife is working in the center, the husband may come home early from a job, and he can eat warm food”. They also said their husbands enjoyed ugali which had sat in the chepkube‟s cupboard and “form[ed] a layer just like bread” (4.1.1). In fact, two respondents spoke of discomfort similar to heartburn when eating cold ugali (1.2.3, 2.2.4). Food also looked more appetising cooked on an FES because there was no ash blowing from the hearth into the pot

(FG7). Food was not the only thing the stoves were used to heat. A woman who used

TSS knew that the multiple burners on a clay stove meant that “[the husband] can see a lady cooking food at the same time as warming water for bathing, instead of putting food aside to warm water when he comes home” (3.1.2).

A group of women who had recently started using chepkube stoves believed their husbands benefited because they could “be in the kitchen at the same time that [their

120

wives were] cooking” because there was no smoke or dirt to disturb them (FG7).

Other women reported that with FES in the kitchen their husbands were “able to cook, especially when the wife is not there” (4.2.4). A woman who had never used TSS before said,

“if a lady goes on a journey and has left behind a few sticks of wood, she can be assured that her husband will not have a difficulty cooking. He will not look for firewood and will be able to prepare food quickly in the morning so he won‟t be late for work. He won‟t just leave having boiled water and leave it on the stove, after eating nothing” (4.1.4).

He also wouldn‟t be disturbed because “cooking with a chepkube [means] dust and smoke will not dirty his trousers” (4.1.1), and clay stoves were reported to be easier to light and cooked food more quickly (3.2.1).

The final class of perceived benefits for men had to do with possible ways the family as a whole would benefit rather than him in particular. The most common response in this theme was about saving of money35 (2.1.1a, 2.2.4a, 2.4.2, FG6, KI7a).

One man said he used to buy wood three or four times a year, but since installing the new kuni mbili stove, he purchased wood only once a year (KI7a). The group of women who were pioneers with using chepkube brooders said that their husbands encouraged them to use that stove because of the income it would generate, which would help the family (FG4). Another response for the man in particular was that his children were less likely to be burned (1.2.6, 2.1.2). The father of a respondent sagely told us that the chepkube gave him and his wife benefit in their old age because “it assist[ed] students to eat on time and in the future they will have to remember their parents that cooked for them on time” (2.2.2).

35 Saving money may or may not have been a perceived benefit for the family as a unit. See Chapter Six for more discussion about this subject.

121

Benefits for Children Participants identified two areas where children would benefit in particular from a FES being used. Firstly, FES was seen as safer for young children. For instance, one mother said “in a chepkube the fire does not spread. It just goes up, but with TSS a kid could step on a firewood, cause the pot to turn and pour on him/her” (4.2.2). A father remarked “because of the kuni mbili style, it prevents danger more than TSS whereby the children when playing around the TSS may push another child and may get burned in the fire” (2.1.1a). Other mothers spoke about how her children would get burned if they lifted a pot from the TSS but not if they fetched a pot from a chepkube because the clay stove was permanent and stabilised the pot better (3.2.2,

4.2.1).

Respondents generally agreed that the food warming capabilities of the chepkube style of FES was beneficial for children, especially those who came home for lunch in the middle of the day. A group of women with very young children may have been speaking from personal experience when they said “when the children are back from school, they will find and eat food on time, so they won‟t be late for school”

(4.1.1). Likewise, another woman expressed interest in using a stove with multiple burners so that she could cook two foods at once and feed her kids in the evening before “they tire easily and fall asleep” (2.4.1b). A man said in his house “after cooking the evening‟s food, we can place water on the [kuni mbili] liner and leave, then use the water for washing in the morning while it is still warm” (KI7a).

Perceived Need for FES Within the Community

During the course of the research period nineteen stove models were encountered, fourteen of which used wood and wood substitutes. The models and their fuels are listed in Table 5.4 and the history of use of each FES is recorded in

122

Appendix 15. The first sixteen stoves in the table 5.4 were seen at rural households, whereas the final three stoves were encountered solely through contact with officers from the Ministry of Agriculture and GTZ. FES users were present in every village.

Although not everyone had personally experienced using FES, every participant could recount some information about some FES style.

Table 5.4: Stoves encountered during the research period, their descriptions, construction materials and fuels used. Stove name Description Materials Fuel

Three stone Three stones, bricks or mounds - Bricks or Wood, cobs, stove (TSS) of clay with fuel inserted from stones stalks three directions - Clay (optional) Improved three Like TSS, with two openings - Bricks or Wood, cobs, stone stove closed in by mud so that fuel is stones stalks (ITSS)* inserted only in the front. - Clay - Metal rods (optional) Clay stove (CS)* Like ITSS, but with two or more - Bricks Wood, cobs, burners. - Clay stalks - Metal rods Chepkube (CK)* Like CS, but includes a - Bricks Wood, cobs, „cupboard‟ built into the clay on - Clay stalks one side of the burners. - Metal rods - Metal pail or iron sheet Chepkube CKB: Like CK, but raised about - Bricks Wood, cobs, brooder (CKB)* 30cm to create a brooding area - Clay stalks & for chicks. A series of layers is - Metal rods Chepkube constructed between the fire - Metal pail or brooder/ bed and the chick area. iron sheet hatchery (CKBH)* CKBH: Like CKB, but with an Additional additional specially lined materials for „cupboard‟ built on the top layer CKB: for hatching eggs. - Off cuts from timber - Old gunny sack - Dry soil - Another iron sheet - Chicken wire Kuni mbili fixed KMP: The simple ceramic liner - Purchased Wood, cobs, (KMF)* or KMF: Ceramic liner inserted into ceramic liner stalks

123

portable (KMP)* a base of bricks and clay. - Bricks (KMF) - Clay (KMF) Sawdust stove – Resembles ITSS, but the front - Bricks Sawdust clay* opening is closed except for a - Clay small metal tube inserted in the - Small metal clay cup Sawdust stove – Metal stove with legs and a - Purchased Sawdust metal (SDM)* small front opening for inserting from informal kindling and an open top with artisans metal pot rests Cement Resembles a much larger and - Cement Wood, cobs, chepkube* taller version of CK. - Metal rods stalks - Unknown Chepkube A much larger and taller version - Bricks Wood, cobs, bakery* of CK with a large cavity in the - Clay stalks center for inserting trays of - Metal rods scones. - Unknown Deluxe charcoal Resembles SDM but larger and - Purchased Wood, cobs, stove* lacking a front opening. from informal stalks, artisans charcoal Charcoal stove Resembles SDM but lacking a - Purchased Charcoal front opening and has an from informal internal chamber to separate artisans charcoal from ash Kerosene stove Metal stove fitted with liquid - Purchased Kerosene fuel canister Gas cooker Metal stove fitted with - Purchased Gas pressurised fuel canister Rocket Stove* Highly efficient single-burner - Special bricks Wood, cobs, brick stove with one opening for - Soil stalks fuel - Unknown Fireless cooker* Insulated basket with an - Basket Residual heat insulated pillow-like lid - Black cotton from cooked material food - Insulating material - Needle & thread - Or, purchased Solar cooker* Wooden box with glass lid - Purchased Sun energy *Fuel-efficient stoves

Perceived Advantages of FES

The perceived advantages of FES mirrored many of the disadvantages of TSS as well as the benefits accrued to family members presented earlier in this chapter. The

124

perceived benefits were related to time and energy savings, ease of operation and construction and safety and warmth.

Saving Energy and Time

Savings of energy and time spent cooking was the most common category of perceived benefits. Time was saved in four ways. Firstly, clay stoves were seen to cook faster than charcoal (3.2.4) or TSS (3.1.4b) because the heat was more intense or concentrated. Secondly, stoves with multiple burners saved cooking time because the user could prepare three different foods, like “vegetables, ugali and eggs at the same time” (3.1.2) instead singly. When using a chepkube a person is not “disturbed to light up the fire” throughout the day because “when I am cooking chai, (usually consumed in the morning) I‟m also cooking supper” (2.2.4b). Third, time was saved because the user did not need to monitor the fire as closely to add fuel because clay stoves „ate less wood‟ compared to the TSS (3.2.1) and sawdust stoves did not need to be monitored at all once they were ignited. A young woman said “when you put sawdust in a stove, you can leave it and go for a walk (or herd cattle (4.2.2)) and when you return, the githeri is done” (2.1.2). Another woman experienced a similar phenomenon when she packed her clay stove full of corn cobs and left them to burn to cook her githeri (3.1.3). The chairman of a male community group explained the fourth way to save time succinctly: “the main benefit of a chepkube is that you only need to prepare food once, in the morning, and it keeps warm [in the cupboard all day]” (FG2).

Most participants believed FES used less fuel compared to TSS, therefore saving those who bought fuel money. For instance, two people in Endebess said that they went looking for wood once a week when they were using a clay stove, but thrice a week when using TSS (1.2.3, 1.2.6). A man who used to buy charcoal eight or nine

125

times in a year now buys it once yearly because a secondary benefit of using his kuni mbili stove was that other stoves were used less (KI7b). A group of women who recently switched from TSS to chepkube said they no longer had to purchase wood fuel even though their properties were less than an acre in size (FG7). For most households that purchased fuel36, money was saved within one or two months of buying or constructing a FES. A woman reflected that:

“TSS is nice because it saves money in the beginning. . . [because it is free to build]. . . but if you want to save money, you have to spend money. After building [a clay stove], you save more, but TSS is expensive forever” (3.2.1).

She also mentioned that sawdust stoves in particular had the advantage of using a fuel that was free and easy to procure.

FES Physical Characteristics

The second category of benefits referred to physical characteristics of different

FES styles. Metal stoves that used sawdust or in one case which used wood and cobs

(4.1.3) were durable and portable which meant they could be carried to and used in rental houses (AG3) and moved inside in case of rain (4.1.3). Several users of clay stoves spoke of destroying them and then easily rebuilding them if they broke (2.2.1), if they procured sawdust (2.1.1b), if their kitchen was destroyed (4.2.3) or if they just felt like it (3.2.1). For many women the bulk of the materials needed to make the simpler stoves (bricks, clay) were free or easily available (FG7).

In terms of operating stoves, respondents pointed out that they could leave their child to cook at a chepkube because unlike TSS, the pot wouldn‟t fall over (2.4.3b).

The sawdust stove was seen as easy to cook with once it was lit because “with proper filling, it cooks faster and won‟t go off easily” (2.1.4). Other women said that a clay stove “is so easy for a man to use, unlike TSS . . . because the clay stove is very easy

36 We gathered fuel data from 30 households and found 20 of them bought some or all of their fuel. Refer to Appendix 16 for more details.

126

to light the fire, and the food cooks quickly. Also the clay stove is clean, not dirty.

Even when cooking, the ash cannot come to you” (3.2.1). Several participants said they added mud around the stones of their TSS to close off two sides (thus creating an improved TSS) or used larger clay stoves to decrease the spread of ash (1.1.1, 2.1.3,

FG4). A benefit specific to chepkube brooder in addition to the ability to raise chicks was that you could contain the birds and sweep “the bird dirt” (2.2.4), thus keep a clean kitchen. Finally, the physical characteristic unique to chepkube stoves was the ability to keep food warm which allowed users to leave and come back to find their food still warm (3.2.3, 4.2.3, FG2) or even better cooked (4.2.1, KI2).

Figure 5.6: A woman in Central division demonstrates her chepkube brooder

Avoiding Disturbance

The third broad category of perceived benefits related to gaining warmth and avoiding disturbance due to smoke and instability of cooking vessels. Using a clay stove to produce warmth was an idea only a few respondents had. One man mentioned that while he used the charcoal stove for warmth, his wife got warmth from her clay stove (3.1.4a). When asked directly about how much heat their stoves produced as compared to TSS, users of clay stoves (1.2.6), improved TSS (2.1.4) and

127

chepkube (FG4) all said that they were always comfortable and that there was just enough warmth in their kitchen because you could “adjust the fuel to give more light or heat to the user” (FG7). A woman who used an improved TSS with a chimney fitted to it said that while the chimney drew away warmth, which was a disadvantage, it also drew away smoke, which was a benefit (2.3.2). Other respondents also said smoke was less of an issue because less was produced by chepkube (2.2.4) or sawdust stoves

(4.1.1) or it was directed to one side (FG2).

Figure 5.7: Notches in the corners of each burner decreased disturbance due to smoke on this chepkube hatchery and the customised shape of burners held pots securely (FG4)

Increased stability of cooking vessels (traditional round-bottomed clay pots or newer, more widespread flat-bottomed metal pots) was a benefit that was listed in response to several types of questions37, as was the issue of increased safety for children. One FES user said “a clay stove is safer because if I want to cook, [the stove] is stable and strong. It won‟t move. A pot fits well. You can‟t be burned because the pot won‟t move or tip” (3.1.3). Another said “when a child plays around the clay stove, the food can‟t pour on them because the stove is permanent, but with

37 For example, “what are the benefits of FES for women/men/ children?” and “what are the disadvantages of TSS?”

128

TSS they can move the stone and food might pour on them” (3.1.4b). A key safety feature of clay based stoves was that they “have been moulded so that the fire concentrates in the middle” (4.2.3) instead of “coming outside like TSS” (1.2.6), thus preventing children or the users from getting burned. Clay stoves could also be raised

“so it is difficult for the kid to reach the food” (3.2.1).

Decreasing time spent cooking, saving money on fuel, decreasing disturbance from smoke, flames and unstable pots, portability or ease of construction, food warming capacity, ability to raise chicks and keep a clean kitchen are the broad categories of benefits listed by rural residents during the study. Participants spoke of other benefits indirectly such as decreased time spent collecting wood, overcoming vulnerability to fuel shortages by having diversity in stoves and fuels, and also the aesthetic appeal of decorated stoves in the kitchen.

Perceived Disadvantages of FES

The perceived disadvantages of FES were generally in three themes: disadvantages because the stove lacked characteristics of a TSS, disadvantages referring to clay stoves in particular, and disadvantages specific to sawdust stoves.

FES in General

Several users said that food tasted different on FES compared to that prepared on a TSS. In particular, people mentioned that ugali could not be cooked well (4.2.1,

FG3). The solid walls of a FES were what made it efficient, but they were also a source of contention for people who were used to TSS because “you cannot place a drum38 or a large pot on [an FES], unlike a TSS” (1.1.4). Another perceived disadvantages was

38 A drum in this case referred to the cooking vessel commonly used to make pombe, the locally brewed but illegal alcohol made from maize flour. This participant was eating mkarango for lunch while we were speaking with her, which is a porridge made of roasted maize flour milled from expired or rotten kernels used to make pombe (KI5).

129

that FES are “expensive to build- you need bricks, metal rods etc. . . . unlike TSS, which is free” (FG1). In some places the soil was of such poor moulding quality that even that had to be purchased (1.2.4, FG4, FB1, FB2). Stoves that were purchased ready-made, such as the metal sawdust stoves were also perceived by some to be too expensive (4.1.1). Lastly, respondents said that “when you move to a new place, you can‟t carry [a stove moulded from clay] with you” (FG2), in contrast with a charcoal or kerosene stove, which could be moved.

Clay Stoves

The most common stoves found in the 32 semi-structured interview households were: TSS (19 households), charcoal (17 households), variations made of bricks and clay (16 households) and stoves that used sawdust (eight households). There were also one operational kuni mbili (4.2.2) and three variations on chepkubes which were inoperational: one made of cement (3.2.1), one made of clay that worked as an oven to bake scones and cakes as a business (2.2.4), and one that used to cook and brood chicks but was used as a chair at the time of the interview (2.2.4). Consequently, the

FES people were most familiar with were in the family of stationary brick and clay models (improved TSS, clay stove, chepkube and chepkube brooder, fixed kuni mbili stoves and clay sawdust stoves)39. The major disadvantages perceived by participants concerned maintenance, fuel consumption, smoke, lack of warmth and some issues particular to the brooder.

Several people believed stoves built of clay consumed too much fuel because the stoves were large (4.1.2), because they needed “lots of big logs of wood for the food to be cooked well” (4.2.1), because once food was placed in the cupboard you must

39 Refer to Appendix 15 to see more details on which households currently or previously used each FES.

130

have a small fire (4.1.1) or large embers (4.2.1) to keep it warm, and because enough heat was generated to cook three pots, even if you were just cooking tea (FG7).

Secondly, people spoke about the stove breaking “when you don‟t take very good care of it” (2.1.4) or when you use the wrong type of clay to construct it (3.1.4a). If a person did not know how to use it well, or how to monitor it and be vigilant to repair and smear it regularly40, it could break (FG1, FG2). The multi-burner clay stoves needed some level of expertise to build (FG1), and often required builders to purchase material (3.1.4b, FG1) which were both seen as a disadvantages. Families who cooked with a fixed kuni mbili (KI7b) and with a chepkube both said that more smoke was produced with their FES. Finally, a woman pointed out that since the chepkube brooder‟s fire base was raised to hip height, less heat was produced, especially for children when compared to the low clay stoves (2.2.2).

Sawdust Stoves

Metal sawdust stoves had to be purchased (4.1.1), sawdust itself was sometimes

“difficult to be found” (2.1.2) and one had to travel to a trading center to find both those things (4.2.2). The stove took some skill to use because “it doesn‟t start up quickly or easily, especially [with] damp sawdust. When you are in a hurry, you can‟t use the sawdust stove . . . if there is no wood, men prefer to use charcoal because sawdust is difficult to light” (3.2.1). Also, the concentrated intense heat of the sawdust “destroys pots easily” (3.2.1) and made it difficult to cook ugali “because the ugali gets burned before it is cooked” (2.1.4). The metal sawdust stove is also not adjustable. A young woman said her sister build a sawdust stove out of clay because the one in the market did not fit her pot (2.1.3).

40 Smearing a stove means to cover it with a mixture of clay, dung and sometimes ash. This is done approximately once a month (FG4).

131

Table 5.5: Comparison of perceptions concerning advantages and disadvantages of FES listed by participants

Perceived Factors A B C D E F G Saves fuel X X X X Saves time X X X Food stays warm all day, is ready on time X X X X X Saves money X X Business opportunity X X Improved marriage relations X Improved self-worth X Bath water is warmed X X Ash is contained (food is clean) X X X X Easy to build X Durable X Portable X Easy to use X X Less smoke produced X X More stable (safer) X X X X Fire bed is raised X X Warmth is produced X Ugali tastes good when left in chepkube cupboard X Expensive to build or buy X X Not portable X X Consumes more fuel than TSS X X Less heat is produced X X More smoke is produced X (Sawdust) Intense heat burns ugali and damages pots X (Sawdust) Fuel is difficult to find X X (Sawdust) Need skill to light the fire X

Key: A: Perceived advantages of FES B: Factor that influenced someone to use FES C: Perceived advantage for women D: Perceived advantage for men E: Perceived advantage for children F: Perceived disadvantage of FES G: Factor that influences someone not to use FES

132

Perceived Fuel Shortage in the Area

Every household in the study used wood fuel for cooking. Additionally, 27-2941 out of 32 households used maize cobs and stovers, 17 used charcoal, eight houses used sawdust, three houses used kerosene and one of those also used gas as cooking fuel.

Only two households used one type of fuel (1.2.1, 1.2.6) and the most a household used was six types (4.2.4). Interview questions focussed on raw biomass fuels (wood, cobs, stalks, and sawdust) as those were the fuels used in FES. Table 5.7 shows the types of fuels and stoves used as well as how the fuels were procured.

About two thirds of participants purchased some or all of their fuel and there was no significant difference between FES users (62%) and non-users (65%). The most commonly purchased fuels were wood (16 out of 30 homes), cobs (11 out of 27 homes) and sawdust (six out of eight homes). Out of thirty homes, seven purchased wood all the time, nine purchased it sometimes, six gathered it sometimes or all the time from the forest, eight gathered it sometimes or all the time from their farms, three gathered it from the road side, three did not specify a place, and one gathered it from a neighbour‟s property.

Of twenty seven households who reported using maize cobs, five purchased cobs all the time, six collected it from their farm and purchased extra once theirs was finished, fourteen only used cobs from their farm, and two did not specify where they got theirs. Maize stalks were never purchased. Twenty-one out of twenty seven homes that burned maize stalks relied entirely on their own production, while two sometimes or always gathered from other people‟s properties, and three houses did not specify where their stalks came from. Six sawdust users purchased their fuel while two gathered for free. Out of seventeen charcoal users, sixteen purchased it all of the

41 Specific data about maize cob and stover usage was not gathered from two households, but based on trends in the community, it is likely that one or both houses used one or both of those fuels (3.2.4, 4.1.2)

133

time, while one supplied all of their own from trees on their property (3.1.4). For additional information about amounts and frequency of procurement organised by area, refer to Appendix 16.

Table 5.6: Summary of fuel procurement methods in the study area (n=33) Charcoal Wood Sawdust Maize cobs Maize stalks Gathered only 2 15 2 9 17 Purchased only 15 7 6 4 0 Purchased and 0 9 0 7 0 gathered Unknown method of 0 2 0 13 3 procurement

Maize cobs and stalks were used immediately following maize harvest and stored until they ran out, after which time either additional cobs were purchased or the houses relied on the other fuels. People commonly said they used stovers “when they are found” (3.2.1), as in after harvest. Cobs were seen as a way to extend the life of the wood fuel they had (4.1.4) because “by mixing wood and cobs, you use less than just using wood or just cobs” (1.2.3). One person specified that she used cobs in the wet season and stovers supplemented by wood in the dry season (2.4.4a). Sawdust was usually used specifically to cook githeri (3.1.1) or as a back up when other fuels

(like cobs or charcoal) ran out (2.1.2, 3.1.4b).

Perceptions of fuel scarcity

People‟s perceptions of fuel scarcity varied quite a bit depending on personal circumstances and their outlook, but most people spoke about difficulties they faced throughout the year to procure fuel. One woman said, “I would like assistance in learning [about FES] because in this area we have a problem with firewood” (2.2.3). A young man and an older man in neighbouring villages explained their need to conserve fuel: “this place is changing from rural to urban and the population is increasing. They

134

are not planting trees, and many are being cut down” (3.2.2) and “in the old times there were lots of trees and lots of wood. Now, we must move with the way the world is moving [and change our fuel usage]” (FG5). A young woman in the same area said “I see that firewood is scarce, because now, there is a problem. There is no forest. It has been cleared” (3.1.2).

Apart from the forest disappearing, problems people faced varied from goats eating trees they planted intended for fuel (1.1.1), father-in-laws not allowing their daughter-in-laws who were staying on their land to cut trees (they could prune branches though) (2.2.1), to difficulty procuring dry wood during the rainy season

(2.1.4).

Perceptions of Fuel Abundance

Some people who produced their own fuel or gathered it from the surrounding area felt there was no shortage because they could always find enough that they avoided the need to purchase anything. When one was asked if she would like to use less fuel, she (a chepkube user) replied “I am comfortable with this stove” and said it was never a problem to find wood, cobs and stovers on her farm, so she had no reason to use less of it (2.2.2). A second response equated fuel with money. A woman who normally purchased all of her wood explained that “finding fuel is not a problem. The problem is just if you don‟t have money because there are some people who normally move with bicycles selling the firewood” (3.2.1).

135

Figure 5.8: Bicycles laden with firewood travelling from Endebess

Perceived Factors That Influenced Households to Use FES

Factors participants listed that influenced their decision and the decisions of their peers to either use or not use FES concerned the person‟s cultural background, extent of exposure to FES, perceived advantages and disadvantages of FES, triggers in their personal environments, and youthfulness. People spoke most often about clay stoves and particularly about chepkube rather than sawdust fuelled versions since they were the most common form of FES in the area.

The word „chepkube‟ is a Kalenjin word meaning „hidden‟ or „secret‟ place

(FG4). Respondents felt that simply growing up in a Kalenjin household was a strong factor that influenced the use of that stove. One woman said, "I know Nandis use chepkube . . . but I am not sure why it is just them and not others" (2.4.4a). A

Kalenjin woman showed us a play stove at her house for her children to learn how to mould the stove and how to cook with it (KI14). Another woman recounted learning how to use the stove by playing with a mock one behind their house and said she used

136

it as an adult because "it is like culture . . . my parents used one . . . when you have a house, you must make one yourself, because this is the way the culture is” (4.1.1).

Respondents also spoke about the amount and type of exposure someone got to the technology being a major factor. For one woman in Cherangani this seemed to be in the form of peer pressure because although she preferred to cook with TSS, she said

"anyone that stays here for some time uses chepkube” and said she started using that stove specifically because she wanted “to be uniform with the [neighbours]” (4.2.4).

Ideas spread because “all the neighbours come from mixed tribes, therefore it is easier for them to use new things” (FG7) and the dissipation of TSS related cultural beliefs of Luhya and Sabaot due to the spread of Christianity. As one older man pointed out, “now people are going to school, becoming Christians, Moslems, and are comparing to see what kind of affect does removing a stone or a stick42 really have on somebody? Is it a belief or a truth?” (3.2.4). Other types of exposure came from children learning about FES at school and then influencing their parents (2.4.1b,

2.4.3a), people sharing ideas which leads to a change (2.1.1a), and intermarriage

(4.2.1).

Some people said they or their peers began to use FES because of specific benefits like keeping food warm, saving fuel, ability to fit it in a small kitchen, and the ability to keep water warm. They used examples of a woman who owned a cafe that needed a clay stove so she could cook three foods at once (1.2.1), women who had other businesses so needed to cook in the morning before going to work (4.2.1,

2.2.4b) as well as women who didn‟t necessarily have businesses outside of the home but just appreciated the ability to “cook in the morning and eat it later" (KI13) or cook quickly (4.2.2).

42 Another tradition for Luhya and Sabaot people was to break the stick that protruded from the roof of a traditionally thatched kitchen when the husband died.

137

Other people pointed at circumstantial factors that enabled (or would enable) someone to use an FES. The first was the need for knowledge and training. A young man in Endebess was convinced that his grandmother just needed time and the right information to “get the idea and use [chepkube]” (1.2.3). A woman said, "it is imperative that I change stoves if I find [a stove that uses less fuel]" (2.2.1). The second factor was money. Several members of women‟s groups said a major facilitating factor for them was that "they organise[d] themselves in a merry-go- round43 to build the chepkubes" (FG1). Their groups helped both in terms of money and building expertise (FG1, FG4, KI9, 3.2.4, 4.2.4). Other people said that they would build a chepkube once a house or kitchen was complete (3.1.2, 3.2.2), build a sawdust stove when sawdust was available (2.1.1b), or build chepkube brooders if they got support from SCC-Vi (3.2.4). One person pointed out that “the reason a person can accept a change is because of a problem” (2.1.1a), for instance “maybe because of lack of wood, it can make a person to start using solar or sawdust stove” (2.1.4), like a family who formerly used a sawdust stove because “in the past [they] didn't have a lot of trees” (3.2.3).

A theme people spoke of that influenced all factors described in tables 5.7 and

5.8 was the difference between older and younger generations. An older lady said

"chepkube is for educated people because they come with new type . . . You young people should use the new kinds" (3.1.1), like a woman who spoke of destroying and rebuilding clay stoves on a whim, according to the latest styles (2.2.1). Others said

“this generation does not stick on traditions, therefore it is not a barrier for they to switch from TSS” (2.4.1a), “the [younger generation] have learned from the Kalenjins,

43 Members of a Merry-Go-Round each contribute an equal sum of money on a regular basis, and each receive a much larger sum of money when it is their turn to receive it. For instance, if twelve women each contributed 100 KSh each month, then each of them would received a sum of 1200 KSh one time in the year, as their „turn‟ came up.

138

whereas the older ones have less good socialisation and friendship with them” (2.3.2), and “[the younger generation] wants to please their husbands (by serving them hot food), whereas the older generation has given up” (2.3.2).

Table 5.7: Factors perceived by rural residents that influenced households to use FES Culture (19%) Exposure (21%) Personal Physical Circumstance (30%) characteristics of FES (30%)

Less traditions Neighbour‟s Knowledge (4) Cooks fast/ once because of mixed influence (4) per day/ warm (7) Member of a tribes (3) Children learn money lending Has been trained Less traditions about FES and tell and/or stove on how to build the because of parents (2) building group (4) FES (3) Christianity (3) Youth are more New house or Warms water (2) Chepkube is for likely to make kitchen is built (2) Can use alternate Kalenjins (3) multi-ethnic More money fuel (2) friends (1) Youth do not hold available (2) Cleaner kitchens as tightly to People sharing Chosen for a MoA (2) traditions (2) ideas (1) demonstration day Saves fuel (1) Youth are more Talk leads to (2) willing to change innovation (1) Fits in a small Personal (1) kitchen (1) Intermarriage (1) preference (2) Unspecified Only thing used Working away from problem (1) here, grew up with the home (1) it (1) Sawdust is Thinking more available (1) about cooking (1) Peace (1) Younger generation is more educated (1) Note: The numbers of responses are in parentheses

Perceived Factors That Prevented Households From Using FES

The factors identified by participants that negatively influenced FES use or inhibited rural residents from using FES are listed in table 5.8. The most common responses concerned physical characteristics of the FES which were undesirable or not

139

needed, specifically concerning the food warmer on the chepkube, and environmental triggers, particularly a lack of funds and a lack of adequate training. Further responses reflected people‟s lack of knowledge of FES benefits, their resistance to change and beliefs about stove preferences of certain cultural groups.

The cost of stoves and their materials was the most frequent negative factor mentioned by rural residents. The most costly stove was constructed from cement and cost the users 5000 KSh to build. The women who owned it were prohibited from using it at the time of the interview because they couldn't afford 2500 KSh to repair the cracking cement (3.2.1). Their stove was the most extreme case of inability to use an FES due to cost of materials.

A metal sawdust stove cost 300-350 KSh, a kuni mbili liner cost 200-300 KSh, and materials to build a chepkube also cost 200-300 KSh. Materials for a chepkube brooder cost 400-600 KSh plus the cost to hire a builder (500-1000 Ksh) and capital to buy chicks, making it potentially a very high start-up cost. Two people said that

“money became an issue" (3.2.4) so they couldn't buy chicks for their brooder (2.2.4,

3.2.3, 3.2.4). Several other people who otherwise would have built a chepkube

(2.4.1b, 2.4.3a, 3.1.4b, 3.2.4, 4.1.3, 4.2.4) or kuni mbili (KI7a) said that they or others lacked money to buy materials. One woman thought that compared to a chepkube, “the kuni mbili doesn't consume a lot of fuel, but it is costly, so it is better for one to built [a chepkube]" (2.2.1).

A lack of knowledge about the existence of FES or FES benefits were the primary exposure factors which were perceived to absent in rural residents who did not use FES. At least seven residents recognised that since clay stoves are handmade,

“if you don't know how to make [a stove] it is impossible [to get one]” (2.4.1b). One

140

participant said specifically "older people can't use chepkube because you need an expert to build it" (1.1.4).

Others hypothesised that non-FES users "may have seen [an FES] but they don't know the benefits, and they don't have anyone that explains the benefits, therefore they don't use [it]" (3.1.2). More people thought TSS users “think that the one that she is using is easy, but it's not" (2.2.2), because "they haven't discovered the secrets of chepkube yet" (FG4). Some stove styles simply weren‟t widespread, such as the deluxe charcoal stove bought in Kisumu, a city four hours away by bus (4.1.3).

A few people knew that FES could save them both cooking time and fuel, but they were not interested in saving either of those things. A woman who had previously used chepkube said “people with chepkube are busy people” and explained that she doesn't need to have one now because she is “always just in the door [with] no hurry of cooking and moving out” but when she gets a job, she will build it “whereby I will just cook and leave for the kids to come, take and eat”, it's just that now she has the

“free time to cook for them with a TSS” so she is not interested in changing (4.2.1).

Some women pointed out the additional labour needed to keep the stoves working well, and preferred not to use them for that reason (2.1.1a, 1.2.1).

A man in Kaplamai said “in other areas people still use TSS, but not here.

Because of traditions they stick with TSS, but also because there is more wood there"

(2.4.3b), and a man from Kiminini who knew how to build chepkube brooders, but who‟s wife used TSS said that "fuel is not a big enough issue to make one change stoves" even though his wife had previously used a sawdust stove when they lacked fuel several years ago (3.2.3). Two participants said that people who relied primarily on maize stalks to meet their energy needs would not find using a clay stove easy but, the exact reason was unclear.

141

Table 5.8: Factors perceived by rural residents which influenced households to not use FES Culture (14%) Exposure (15%) Personal Physical characteristics Circumstance (36%) of FES (33%) Resistant to change Do not know FES Cost is too high (11) Do not need to cook (5) benefits (9) once per day or keep Lacks training in food warm (6) Kisii don‟t use Do not know FES stove building (9) chepkube (3) exists (4) Clay stoves are Needs a house or labourious to maintain Bukusu use TSS (3) FES are for kitchen built (5) or build (5) educated people Chepkube is for Lacks proper soil (3) (1) Do not need a brooder Kalenjins (1) Lacks suitable place (4) Only women talk Unspecified tribes to build a stove (3) about stoves (1) Doesn't need to save use TSS (1) Post-election time (3)

Kikuyu use violence (1) Don‟t need/ want to charcoal (1) Chicks were not conserve fuel (3) profitable (1) Unexplained lack of Stove group is too interest (2) young (1) FES consumes more Sawdust is not fuel(2) available (1) Less heat is produced Unexplained (2) procrastination (1) Maize stalks do not work well in clay stoves (2) Don't need a fireless cooker (1) FES cooks food slower than TSS (1) Chepkube brooder lacks a place to rest wood (1) Builds FES according to whims (1) Note: The numbers of responses are in parentheses

Other barriers to FES use were the corollary of the facilitating factors in the previous pages. There were beliefs that certain stoves were for certain tribes (i.e.

Bukusu use TSS, Kisii would not use chepkube because they prefer to eat ugali right away, Kikuyu prefer charcoal because it is easier, and the chepkube is only for the

142

Kalenjins). Five people spoke about a person‟s resistance to change, often referring to the older generations, and also of a disparity in terms of knowledge. A former chepkube brooder user said “grandmothers only know one use of a stove, but us we can use it to keep chicks” (2.2.1).

Perceived Differences in Thoughts About FES Between Community Members

Participants were asked to comment if they thought men thought differently about FES than women, or if older men and women thought differently than their younger counterparts, as well as general impressions of community reaction to a household starting to use a FES.

Perceived Differences in Thought Between Men and Women

About two thirds of respondents believed men thought differently about FES than women. Generally, most women and some men believed that only women thought of issues to do with stoves. While this may have been true for some of the population, many male participants did have definite thoughts and opinions about stoves. A man who was formerly a cook at a primary school described an innovative stove he had made of clay that could be build outside and covered with a tarp overhanging a trench in event of rain. A minute or two after he recounted his invention, his female neighbour, who was being interviewed, said “men don‟t know much or care to know much, as long as the food was hot and on time” (2.3.2).

Another woman said that “men don‟t have any opinion about stoves” (2.1.1b) directly after her husband described with great interest and detail the kuni mbili stove he had seen demonstrated and his unsuccessful attempts to convince his wife to use it

(2.1.1a).

143

When interviewing men and women about the advantages and disadvantages of different stoves, men repeatedly gave similar descriptions about the physics of heat loss in TSS and conservation in stoves with enclosed fires, while women did not go into such detail. The predisposition for men to talk about physics was reminiscent of

Canada where men and boys are traditionally considered to have more interest and talent for engineering and physics. Several people confirmed this was the same belief in Kenya, but added the caveat that men in their area would never build stoves

“because it was not his department” (2.2.3a).

Men did not generally talk about anything to do with stove construction or use as the women did, but most men did have some knowledge of the approximate cost of fuel and could say if their stove had the ability to keep food warm. After giving the physics explanation, most men would proceed by saying “this is now the limit [of knowledge] that a man can give” (2.1.1a) and that if we wanted to know more, we would have to speak to their wives (2.2.3a, KI7a).

Several men were eager to find out more information about FES during the study. At one point during a group interview with Luhya elders, the older men started saying they wanted education about clay stoves and how to prepare them (FG5).

When told that fireless cookers would keep bath water warm overnight, several other men specifically pursued information about that technology (3.2.2, FB4, FB6, FB7).

Perceived Differences in Thought Between Older and Younger People

About three-quarters of participants believed older and younger people thought differently about stoves. A young man in Kiminini said “we think differently because we are moving with the times” (3.2.4). Participants of all ages believed age influenced how willing someone was to change, how much information and training they had access to, and the preferences they had in terms of stove characteristics.

144

Youth wanted to use stoves that cooked fast (1.2.3, 4.1.4), did not produce much smoke (4.2.3), were stylish and modern (2.2.1, 4.2.2), and were seen as progressive

(2.1.3, 3.2.2). Interestingly, the „FES‟ they referred to in these cases included charcoal, kerosene and gas. On the other hand, older people generally preferred to stay with the stove they had (most often the TSS) because they were used to it (1.2.4,

4.2.3), they thought food tasted better (2.1.2), they liked to sit by a big fire (2.4.1a,

4.1.4), or they preferred to use firewood rather than sawdust or charcoal (2.1.2,

3.2.1). One woman explained that when people got older, they “became like fools, therefore [could] not reason that it would be impossible to collect so much wood [for their TSS]” (2.4.2). Another woman told us she could leave her grandmother in a house with food and a sawdust or charcoal stove to cook it on, but return to find her still hungry because she did not find fuelwood (3.2.1).

Participants said that younger people tended to have more knowledge about new stoves like sawdust, kerosene and charcoal (2.4.1a) but lacked expertise to build good TSS (3.1.1) or in the Kalenjin community, good chepkubes (4.1.1). While many women who knew how to build FES reported building them for their mothers or other older relatives, one women expressed frustration that she, as a member of the “older generation” had interest, but was “not able to get the information from the younger ones” (2.3.1a). Curiously, when asked if she knew of a neighbour that used an FES, she accompanied us to the house of a neighbour who used chepkube and said she was willing to teach people how to build them (2.3.2).

As a whole, men did not know very much about stoves of any kind, or about the amount and kind of fuels used in them but there was a general trend that younger men knew more about these issues and used FES more than their older counterparts (1.2.3,

2.1.1a, 2.1.3, 3.2.2, 4.1.3).

145

Perceived Thoughts of Other Community Members

One interview question attempted to ascertain if there was any reaction

(positive or negative) from families or the community at large when a household starting using a FES. Of twelve people that answered this question, one young Bukusu woman said that first their neighbours thought badly when her mother started using chepkube, but then they loved the stove, and have requested that her mother build it for them (4.2.2), one respondent said that her friends reacted with disbelief when she showed them her new sawdust stove, but now they like it and want to have one too

(2.1.2), two people did not know what people thought, and the remainder said people were generally happy to see the new stove and often requested that the participant help them build one for them as well. Luhya and Sabaot people who held cultural beliefs about the TSS, confirmed their beliefs did not prohibit using another stove as long as the TSS was physically present (FG3, FG5).

Existing Channels For Transfer of Information About FES

This section has quite a few tables, however when examining these data, the reader must be aware it was inconsistently gathered due to the semi-structured nature of the data collection technique. Direct queries were only routinely made about the clay stove, chepkube, chepkube brooder, kuni mbili and fireless cooker, but not every household was even deliberately asked about these. Households who were not directly asked were: 1.2.4, 1.2.5/6, 3.2.4, 4.1.1, 4.1.3, 4.1.4 and 4.2.2, so responses for their villages (1.2, 3.2, 4.1, 4.2) were lower than they could have been. In those cases, only stoves that were actively in use or mentioned in passing were included in analysis. Data on the other FES styles (i.e. dung, solar, improved TSS etc.) was usually unsolicited, but again, not always. The intention for including these tables that could not be subjected to quantitative analysis of any kind was to introduce the

146

notion that there is a strong possibility that statistically sound trends could have been established if enough data had been gathered.

In some cases the venue through which a participant learned about the FES was not clear, in which case „unknown‟ or „another town‟ was recorded. For instance, due to interviewer bias, sawdust stoves were not a focus, which resulted in an uncommonly high number of „unknown‟ entries for that FES. Since neither the Ministry of Agriculture nor SCC-Vi actively promote that FES, it can be assumed that people learned about that stove primarily through friends, neighbours, and at the market.

Participants learned about each FES through a variety of sources. They reported being exposed to FES in private situations by friends and neighbours, family members, as well as in public situations such as group meetings, visiting markets, and taking part in formal extension activities done by SCC-Vi or „agriculturalists‟. Data showing how people learned about selected FES styles are presented in Table 5.9. Of those that were exposed to the improved TSS, clay stove, chepkube and sawdust stoves, roughly half later used it at some time, but the conversion ratio of exposure to active usage for chepkube brooder, fireless cooker, kuni mbili and other FES technologies was significantly lower (see Table 5.10).

147

Table 5.9: Sources of knowledge for FES familiarity, and the frequency for each source for 32 households in the study area. FES Improved Clay Chepkube Chepkube Fireless Kuni Sawdust Other Source of TSS (n=6)* stove (n=26) brooder cooker Mbili (n=22) (n=8)* Level of exposure (n=15) (n=15) (n=6) (n=7) exposure Informal, Friend/ 3 3 13 4 2 7 1 individual neighbour (chepkube focussed (n=34) hatchery) Parents (n=18) 3 6 9 Sister or sister 2 1 in law (n=3) Daughter (n=3) 1 1 1 Formal, Market (n=6) 2 1 3 community Group (n=7) 1 2 1 1 1 (dung) focussed 1 (cement) SCC-Vi (n=5) 1 4 Agriculturalists 2 1 1 1 (solar) (n=6) 1 (dung) Radio (n=1) 1 Circle of Light 1 (n=1) (propane) Uncategorised Unknown 3 3 2 1 2 8 1 (solar) (n=18) Another town 3 1 (deluxe (n=4) charcoal) Note: The information for the columns „Improved TSS‟ and „Other‟ was unsolicited. Twenty-seven participants were directly asked about the other FES styles. Only unsolicited information was obtained from 1.2.4, 1.2.5/6, 3.2.4, 4.1.1, 4.1.3, 4.1.4, 4.2.2.

148

Table 5.10: Use and awareness of FES in the study area Current Past user Aware but user never used Improved TSS* 3 1 2 Clay stove 2 5 8 Chepkube 9 5 12 Chepkube brooder 1 1 12 Fireless cooker 0 0 6 Kuni Mbili 1 0 5 Sawdust 6 4 10 Deluxe charcoal* 1 0 0 Cement chepkube* 0 1 0 Chepkube bakery* 0 1 0 Chepkube 0 0 1 hatchery* Dung* 0 0 2 Solar* 0 0 2 Propane* 0 0 1 *Unsolicited responses only

Informal and individual exposure

Participants were exposed to most FES most commonly in informal situations, usually as individuals. Improved TSS, clay stove and chepkube were the only FES participants learned about through their parents but information about other FES was commonly obtained from friends, neighbours, group members and female family members. Most people spoke of asking women they knew casually to come help or teach them when they were ready to build a clay based FES (1.1.3, 2.1.1b, 3.1.1,

3.1.2, 4.2.3). Women also spoke about being willing to share their skills with their friends and neighbours (2.1.2, 2.4.2, 2.4.3b, 3.1.4b, 4.1.1, 4.2.2), sometimes for a fee

(2.2.3b, 3.2.3, FG5, KI11).

Only four people mentioned hiring experts to build regular chepkube (4.2.4), cement chepkube (3.2.1) or chepkube brooder (3.2.3, 3.2.4). The responses recorded for sawdust stoves suggested it was likely people got the most exposure to them at an informal level but only one woman reported learning about fireless cooker informally

(through her daughter who was employed as a teacher in a large town about two hours

149

away and also told her about kuni mbili) (4.2.4). Sources of exposure for each FES a household was aware of or was currently or previously using are presented in Appendix

14.

Community Level Exposure

Formal extension exposed people to FES that were not common in the area.

There were 13-19 instances of formal extension activities instigated by NGOs or GOs that taught about and promoted the use of FES models. Four stoves (solar, dung, propane and cement chepkube) were learned about in formal settings exclusively, however participants were not asked directly about their knowledge of them, so their level of awareness may have been higher than the data show. The most common FES people learned about in formal settings were chepkube brooder, followed by sawdust stoves and the fireless cooker.

Five people reported having direct contact with SCC-Vi and one with the NGO

„Circle of Light‟ which was locally affiliated with International Christian Ministries, a pastor training institute based out of Kitale. Circle of Light promoted a propane stove and battery powered light system that could be rented by rural residents, but only one person mentioned them (1.1.1) and no one said they used the system, although one of their outlet stores was observed about ten kilometres from Lumboka village in

Cherangani. Two people used the term „agriculturalists‟. One mentioned learning from an agriculturalist about kuni mbili (2.2.3), which likely could have meant either

Ministry of Agriculture or SCC-Vi staff. The other mentioned „agriculturalists‟ teaching about a stove that used sawdust or dung as well as a solar cooker (1.1.1).

The Ministry of Agriculture promoted solar cookers, but it is unclear who taught about the dung stove. The only other woman that mentioned sawdust and dung said she learned about it through a group (4.2.1), and one person mentioned they knew of solar

150

stoves (2.1.4) but did not elaborate further, so it is possible there was an additional

NGO working in the area that was not identified by participants, partners, research assistants or myself. One man heard a radio broadcast talking about fireless cooker from which he gained only a very rudimentary knowledge of the stove44 (3.2.3). SCC-

Vi was the only organisation identified in the area that has produced radio broadcasts about FES, so it was likely theirs was the show the man referred to.

Summary

This chapter summarised the data gathered from semi-structured interviews, informal conversations, focus groups and participant observation. Validity and reliability of responses were shown through triangulation of data from related questions answered by multiple participants. The data were organised according to objective and presented in narrative form. The next chapter analyses the data according to a theoretical framework that emerged during the organisation of data in this chapter.

44 He expressed confusion to us about how a basket could possibly cook food

151

CHAPTER SIX

DISCUSSION OF FINDINGS

“Trying to convince women that improved, more efficient stoves are in their interest is doomed to failure unless the factors which women themselves consider important are taken into account” (Clancy, 2003, p. 12).

Introduction

This chapter is divided in two parts. The first contains analysis of data according to a conceptual framework that emerged while writing the previous chapter. The second part of the chapter summarises the answers to the research questions using these theoretical frameworks.

A New Conceptual Framework

The conceptual framework described in Chapter Two was adapted from a study that investigated the factors affecting the use of FES being promoted elsewhere in

Africa (Makame, 2007). After compiling and organizing data in Chapter Five, it became apparent that the concepts used to frame this research did not fully reflect the data gathered. The difference between Makame‟s (2007) study and this one is that most of the stoves in this study were never formally promoted by an organisation.

The study design was conceived on the premise that many more people would be encountered that used FES promoted by SCC-Vi and the Ministry of Agriculture (kuni mbili, fireless cooker, chepkube brooder) then was. The framework in Chapter Two does not emphasise the importance of culture and tradition or the role informal extension plays, and it places too much emphasis on the opinions of community and extended family members and the effectiveness and extent of management support.

Therefore, this discussion of the results has been carried out using a completely new conceptual framework shown in Figure 6.1.

152

Figure 6.1: Conceptual framework of factors that affect the use of fuel-efficient stoves in rural Kenya

This conceptual framework suggests that there are four keys that must be gathered before a person can consider using a FES, and recognizes that the doors that these keys open (particularly those of „personal circumstances‟ and „stove characteristics‟) may be closed and re-opened multiple times, thus leading the user to cease and restart using a certain technology. For example, one woman said that she built a clay sawdust stove when she had sawdust available, but destroyed it when she lacked the fuel. Another woman said that keeping food warm was important when she had a job in town, but was no longer important when she stayed at home all day.

Each key is linked to a multitude of factors that influence it. Some people gather certain keys seemingly effortlessly. Youth, people living in ethnically diverse areas, and people who have no cultural beliefs about stoves get the „culture‟ key easily.

Other keys, like „personal circumstance‟ are harder won, especially if the sticking point is money, space or ability to build a stove.

153

Since each FES has different characteristics, a „new set of keys‟ must be collected for each style. For instance, a person that is exposed to fireless cookers is not automatically enabled to use a clay stove. There are interrelationships between previous and current enabling factors. If one‟s personal circumstances make it amenable to use a clay stove, then likely they would be able to also use a chepkube without much change.

Key #1: Culture

As mentioned in Chapter Three, the term culture in this study refers to the collective representations of symbolic formations and activities (i.e. customs) found in kinship, marriage, friendship, ritual and ceremony that are made evident in individual behaviour (Cohen, 1974). The customs that make up culture are dynamic and often related to ethnicity, which is generally static. In this study an ethnic group refers to a group of people with a common ancestry or history, similar language and culture and usually the same „homeland‟ (Hall, 2000; Makoloo, 2005).

At the outset of this research staff of SCC-Vi Agroforestry staff and the Ministry of Agriculture strongly suspected that culture (a term they tied closely to ethnicity) was the biggest single factor preventing people from using a FES. In fact, family of origin determined what stove would be present at a household, but not what stoves would not be there. Other aspects of culture (the impact of gender roles and religion) encountered were outside the realm of ethnicity.

Family of Origin

Ethnicity strongly influenced which stove would be present within a house because traditionally people learned of stoves from their mothers. Participants spoke of learning about those stoves (and variations), their meanings and histories from their

154

parents. As knowledge about how to build and operate a stove was passed through generations, the use of a particular stove became entwined with cultural beliefs and a sense of belonging. Kalenjins were especially well known for chepkube; it was called

„the stove of the Nandis‟ or „the Kalenjin stove‟ by others, and it‟s creation was generally associated with positive things like clean kitchens, time savings and warm food.

People said that you could usually “tell the tribe of a person by what stove they have in their kitchen”. The TSS, chepkube and improved TSS were all examples of stoves that could be associated with certain ethnic groups (Luhya/Kisii/Sabaot,

Kalenjin, and Turkana respectively). There were instances of women who married into a family of another ethnic group who adopted her in-law‟s stove (in this case from clay stove to TSS) or brought her stove technology with her and them to change their stove (from using TSS to chepkube).

If there were space for only one stove in a house, traditions and culture in this sense would play a major role in determining the stove that is present. However, numerous participants said that as long as the TSS (which for many held strong beliefs not related to cooking) was physically present, culture as such would not be a barrier to using another stove in addition to the first. This statement was backed up by the fact most people (20 out of 32 households interviewed) used multiple stoves, including those from Bukusu and Sabaot households.

For a growing number of youth, especially those who had grown up in ethnically diverse neighbourhoods like those found in parts of Kaplamai, Kiminini and

Cherangani, traditions of parents and grandparents were less important and they reported that their actions were less dictated by culture than by other factors. Youth were seen to be more likely to make friends with people of other tribes, learn their

155

stoves and forget their traditions as communities become more homogenous. In short, youth seemed to be more pragmatic. They were more likely to take the most advantageous route rather than sticking to something uncomfortable for the sake of custom.

Gender Roles

Gender roles were determined by deeply entrenched cultural expectations that male and female priorities were completely different when food preparation was considered. As a result of deeply entrenched gender roles, several FES had the ability to improve marriage relations. Clay based stoves contained ash and decreased smoke, which were both major disturbances that dirtied the kitchen. At least one woman said that since she switched stoves her husband felt comfortable enough with the new kitchen atmosphere to spend time talking with her as she prepared a meal. Other women spoke of gaining respect from their husbands by using a progressive stove.

When men spoke of their wives they often implied that not only was it her role and responsibility to carry out activities related to the kitchen, but also a privilege that they would not want to compromise by being involved. It seemed that if men took over decision-making in this area then a woman‟s comparatively small scope of influence would be contracted even more. Traditional beliefs seemed to strengthen these thoughts: some men said it was difficult for them to enter a kitchen, and one said he could be reported to elders in the community if he was cooking. The stoves were “the department of the women”, and many men interpreted that to mean they could not be involved. In this way, the kitchen, often referred to as a woman‟s office was regarded with a kind of respect.

The community pressure for men not to be involved with cooking did not mean that men had no experience with cooking. There were two instances when men

156

perhaps less bound by tradition were reported to cook. First, men would cook if they were living away from their female relatives. They were more likely to cook with charcoal or gas to avoid the inconvenience of cooking with smoke and ash producing raw biomass fuels and the embarrassment of searching for wood. Women spoke of the ease with which their husbands could cook with enclosed clay stoves because it took less time and was cleaner, thus meeting some of his needs. Second, men of the younger generation were more likely to cook than older men that still held very traditional beliefs.

Administrators and agricultural officers spoke much more candidly about the influence gender roles had on FES use in their communities. One said “the lady‟s priority is cooking, the men‟s priority is not, but he still wants good food on time”.

They were also much more likely to mention that men controlled finances within the household to the extent that “when a man is away, chairs, beds, and posts are split without his knowledge.” When he returned, he would say “yes, I‟d rather spend money on a basket or plant trees than [have this happen]”.

The Role of Christianity

Many participants spoke of the influence of traditional beliefs waning amongst converts to Christianity. This subject particularly came up when asked for traditional beliefs about the TSS. Although the removal of a TSS stone after the death of a parent was a symbol that the man was no longer in the house, it was not a symbol of mourning like wearing a black arm band was in European tradition. The removal of the stone was a practice actively discouraged when a person embraced a new faith, which people would explain by saying, “Jesus died on the cross, and he said the old is over . . . this is part of the old that is over” because “if you make a sign, it is believed that you are doing it because you are afraid something bad will happen”. Now, most

157

people “see that this is just [superstition so] they just got rid of it”. Stones were removed because of fear that a curse would come to the house if traditions were not carefully followed. Faith in Jesus and the Christian God meant that people did not live in fear of those curses anymore, so they did not need to have the stones in their house for that reason.

Key #2: Exposure

FES were learned about through multiple information channels but their reach was not exhaustive. Ambiguous and incomplete knowledge led to misconceptions which affected the use of FES. Also, most formal and informal extension was focussed at women since they were the primary users and men did not generally show interest in the subject since it was not their department or priority. This generalisation was to the detriment of men who actually were interested in these technologies.

Exposure by Observation

The most effective information channels were friends, family and neighbours.

Ideally participants gained accurate knowledge of benefits and disadvantages of stoves and only waited for the other „keys‟ to fall into place. Families were the most ideal way to learn of FES because when a child learned to operate and construct a stove growing up they had that skill for life. Habits of neighbours were also very influential.

There was a noticeable trend of increased tendency to use FES in ethnically heterogeneous areas. For instance, when families moved into the villages of Kiboino in

Cherangani or Kemeloi in Kaplamai, residents used chepkube to the extent that it was difficult to find someone who did not have one secured to the walls of their kitchen, regardless of family of origin. Residents in Kaplamai, Kiminini and Cherangani acknowledged that if they had stayed in their home places (which were ethnically

158

more homogenous) they would not have been exposed to these different technologies to the extent that they have in the new community. Intermarriage of youth who grew up in FES using families also impacted the spread of this technology. Several participants listed learning about stoves from their sister-in-laws or mother-in-laws.

Formal Extension

Sixteen semi-structured interview participants indicated they held current or previous membership in a community group45. This meaningful interaction was a conduit for exchange of ideas, information and innovations as well as a way to make a cooperative business (like baking, brick making or chick raising) and provide financing for projects. The strength of groups lay in their potential to spread ideas quickly through membership, a characteristic NGOs and GOs have capitalised on by working through them almost to exclusion of individuals. Within groups devoted to stoves46 there were usually one or more expert builders who then acted as a resource for the rest of the community, usually at a fee.

SCC-Vi worked in partnership with the Ministry of Agriculture who also had a mandate to promote FES (specifically kuni mbili and fireless cookers). On the surface the government programs did not appear to be as extensive or effective as those of

SCC-Vi, even though there were up to ten times the number of staff working in the same area47. Many factors influenced the difference in program reach. Two types of internal conflict within the Ministry of Agriculture were most salient to this investigation.

45 Five out of 16 semi-structured interview participants were previously or currently in groups specifically related to stoves (three for chickens and two for baking) and key informant interviews were obtained from four more stove groups. 46 Stove making groups encountered in the study had only female members, although groups did exist in Trans Nzoia that had both male and female members. 47 SCC-Vi had one field officer working in each of Cherangani, Kaplamai and Kiminini and four in Endebess while the Ministry of Agriculture had an office with up to ten or twelve staff working in each of those areas.

159

First, Home Economics officers indicated that their male colleagues were far less likely to speak about FES simply because they were men48. The Ministry of

Agriculture carried out extension activities principally through public meetings and farmer field days as well as some meetings with community groups. Primarily men attended the public meetings and farmer field days and most of the agricultural officers were also men. Since most men found it uncomfortable to talk about stoves, this subject was not given premium billing or air time, and no rural residents that were interviewed remembered hearing anything about stoves at field days and farm tours they had attended. This information contrasted with the study‟s findings that many men were very interested to talk about stoves, particularly the fireless cooker which the Ministry of Agriculture promoted. The irony was that men were not interested in the fireless cooker for food. Rather, the fact it could keep bath water warm overnight appealed to their sensibilities.

Secondly, Ministry of Agriculture Livestock officers promoted the chepkube brooder stoves to rural residents as a way to increase their poultry production, but

Gender and Home Economics officers said the chepkube was inefficient and discouraged people from using any kind of chepkube stove. Instead residents were encouraged people to use kuni mbili and fireless cookers. On the other hand, SCC-Vi staff promoted stoves of both branches of the Ministry of Agriculture.

The instances of FES exposure in a group setting reported in Chapter Five may or may not have meant that a representative from the Ministry of Agriculture or SCC-Vi was present, although both organisations worked almost exclusively with community groups rather than individuals. During the study period, two people were found that

48 This sentiment was echoed by a male Ministry of Energy manager. When asked about the connection to energy and other issues, he indicated a female Home Economics officer and said “the cross-cutting issues is why we are partnering with this lady here . . . Do you see how all the Home Economics Officers are ladies? And I am a man? [Promoting FES] is a better fit for them”.

160

used a kuni mbili stove did so as a result of the Ministry of Agriculture. Those stoves were installed because a field day demonstration49 took place at their farms, not because they actively pursued acquiring it. One person mentioned hearing about fireless cookers on the radio but he was confused about the information presented so he did not make any further enquiries. Formal extension efforts like these were a first step in gathering information to acquire the exposure key, but were not effective on their own.

Evaluating the effectiveness of educating parents through their children and entrepreneurs in markets was beyond the scope of this study, however several people mentioned first hearing about fireless cookers, kuni mbili liners and sawdust stoves at a market place as well as being positively influenced by their children‟s knowledge of

FES at school.

Lack of Established Vocabulary

Participants were asked during semi-structured interviews if they were aware of any stove that used wood that was not their stove, or if they knew of any stove that used less wood than theirs. Their responses revealed that people lacked a common vocabulary to describe stove styles in the area. In particular, the term chepkube meant to some people only the cupboard itself (i.e. a stove can contain a chepkube), to others the term meant the entire stove including burners and cupboard, and to yet others it meant any clay stove whether it had a cupboard or not. Many agricultural officers referred to the chepkube brooder by the simplified term chepkube. There were also a large number of participants that could not identify a stove (even their stove) with a name, instead calling it the usual one or the clay one or making up a name like the three-holed one. Likewise, the term fuel-efficient stove was not in

49 A field day is a day of demonstrations set up at one residents property

161

common usage, so it was not used during interviews. When asked for a stove that used less wood, people would often list charcoal and gas instead of clay based stoves.

This reflects their propensity to use multiple fuels to make each one last longer. Using charcoal or gas did decrease the amount of wood used, but not in the way anticipated by the research team.

General Paucity of Knowledge

Appendix 15 is a graphic description of people‟s lack of awareness concerning

FES styles. Only one of 40 semi-structured interview participants knew about all five stoves asked about (kuni mbili, fireless cookers, chepkube, chepkube brooder, and sawdust). The paucity of knowledge concerning the existence of FES as well as the potential benefits of FES were major hindrances to the use of FES stoves. At the end of interviews participants were asked if they had any further questions. At that point several women, (usually women who already had extensive experience and knowledge of FES) asked for additional information about FES.

There were unmet needs in terms of stove extension activities for these women who were eager to know more. Even when people were trained and had knowledge to construct or use an FES, their knowledge was not uniform. One man described a brooder he had built that had as it‟s top layer not a chepkube stove, but three TSS to heat the clay of the brooder. He said he was trained by SCC-Vi, but this was not the type of stove they promoted, so information was lost or misunderstood, which in turn propagated multiple misconceptions about the technology.

Particularly in Endebess residents noted that there was a need for local experts in the community. In other areas it seemed easier to locate multiple people who knew about stoves, but even then, women did not always know who in the community had knowledge to make stoves. It did not appear that the potential for

162

knowledge sharing of elements of stove construction between genders was exploited.

Men in the community were experienced making bricks and constructing kilns to bake them in, which are skills that would be partially transferable to stove building, but this type of knowledge sharing was not observed.

Many participants also seemed to lack knowledge of elements of simple physics such as efficiency, energy conservation, and energy production. One person spoke of friends reacting with disbelief that her sawdust stove would work so well. Others were unclear about how the fireless cooker could keep food warm or if food cooked better if it was poured directly in or kept in the original pot. Still more were unsure if fuel was saved when fire was directed towards the base of a pot instead of being spread in the TSS.

Negative Misconceptions as a Barrier

Despite multiple possible exposure experiences, people did not find the exposure „key‟ if their exposure was not of good quality. The quality of information available to a learner was limited by the depth of knowledge of the teacher and the capacity for comprehension and retention of information of the student. If, for instance a woman learned that the proper way to use a chepkube was to leave a fire burning to keep food warm in the cupboard, she was left with the misconception that chepkubes consumed more fuel than TSS. People recounted the extent of their knowledge about FES, which in many cases contained large misconceptions, as well as a complete lack of knowledge about key characteristics like cooking time and fuel consumption. Most misconceptions were about clay stoves (particularly chepkube) and the fireless cooker.

Popular misconceptions and their responses are summarised in the table on the next page. Rural residents as well as agricultural officers made these statements.

163

The statements labelled „misconceptions‟ were usually made by non-users of a particular technology and those labelled „truth‟ were corollary statements made by users of that same technology.

Table 6.1: Popular misconceptions and their corresponding statements of truth.

Misconception Truth

Large pieces of firewood are It is not only the embers that give heat to the food in needed for embers to heat the cupboard. It is mainly the heat radiating from the the cupboard or brooder, or clay. You can use small or large pieces of wood and you must always have a fire maize cobs equally well. going to maintain heat Fuel efficiency of chepkube is Since the fire is enclosed, you can cook multiple foods comparable to TSS at once and then keep them warm without reheating them, the chepkube is much more efficient compared to TSS. Chepkube is only useful to You can keep any food warm, and it is good for people keep (soft) ugali warm who enjoy hard ugali too. If someone does not want to keep food warm, you do not need to construct it with a cupboard. Chepkube is only useful for Even if you spend your day around the house the people who do not spend chepkube is still very convenient because the user time at home does not need to use wood to cook multiple times in a day. Saving time in cooking would also suit someone at home. Clay based stoves do not You can custom-make clay stoves to accommodate any accommodate large pots size of pots. You can also place the pot over multiple holes to access more heat. Clay based stoves do not grip When building the burners in an clay stove, make a (large) pot well when them in such a way that the pots you use the most fit stirring ugali well and build a depression for them to sit in. Eating food stored in a It is more likely that the increased amount of smoke Chepkube cupboard gives you from a TSS will give you asthma. You cannot get asthma asthma from eating food. Clay based stoves cook slower This is true only if you are cooking one item. If you than TSS cook more than one thing, they can both be cooked at once, thus significantly reducing overall cooking time. A clay based stove can only Clay based stoves have been used successfully be used in a kitchen with a regardless of roof type. There are more sparks and iron roof embers that fly from TSS that could ignite a thatch roof

164

Influence of Male and Female Gender Roles

Having an education about a subject was not always enough to feel enabled to make a decision without support from your husband, family or community (Makame,

2007). This study found that people found it easier to try new stove technologies when living in communities that were mixed in terms of cultural traditions concerning stoves.

Detailed and accurate implications of cultural gender differences were beyond the scope of this study. There were contradictions at every turn concerning this topic.

One woman would say “you can‟t capture the husband‟s [attention], the ladies are the ones that are ready [to learn about FES]”, but then a man would thank us for speaking to him about stoves, because usually men were left out of those discussions. An elderly Bukusu woman, the first of eight wives, said that her husband, who was evidently very traditional in other ways, was interested in progress, and so would likely be interested in the stoves we were talking about, but other women would swear that their husbands only had one duty: “to come and get the food and eat when it is ready”. In the end, customs dictated that women learned and exchanged ideas about stoves and fuel more easily because “it is only the mama who can walk and feel comfortable in another household‟s kitchen”. An agricultural officer expressed genuine surprise when she saw a photograph of a feedback session that was attended only by men (Figure 4.5).

In fact, participants said that most people did not normally talk about stoves as a part of daily conversation. When stoves were talked about, men were excluded from conversations. Information „passes through the mamas‟ because they are the ones that mostly interact with stoves, unlike the men who have less to say because

165

they have less experience. This cycle of exclusive conversation is compounded by the customary stance that issues concerning the kitchen were not the department of men.

Key #3: Personal Circumstances

The term „personal circumstances‟ refers to the capacity a person has to consider using a new stove. Ten out of fourteen reasons cited by participants who had previously used FES concerned a change in personal circumstances. They included things like having adequate physical infrastructure and access to ready-made stoves, fuels or building materials, as well as access to training and construction trends. They also included being in a position where saving time and money was perceived to be beneficial.

Physical Infrastructure

Of 29 houses physically visited, only one was considered „permanent‟. All the kitchens were „semi-permanent‟, being made of wood frames and mud walls. In reality houses, kitchens, and stoves were in constant flux, gliding along the gradient between „constructed‟ and „deconstructed‟ as time passed, walls and roofs broke, fires happened, money came, and physical energy was available. Citing the need to finish a house or kitchen (or even build a new, bigger one that would accommodate a larger stove) were common factors, and often one that needed the help of male family members. Although women mixed clay, patched and smeared walls, men were generally called on to make the frame and fix roofs.

Stove Availability

Locating a ready-made FES or a person that had skill and knowledge sufficient to build one was difficult for four FES in particular. The kuni mbili stove heavily promoted by the Ministry of Agriculture was based on a ceramic liner that had to be

166

purchased by the user. When the kuni mbili first came into existence they were promoted in partnership with GTZ who provided training to women‟s groups, built them kilns for firing the product, and gave funding to the Ministry of Agriculture to advertise and distribute the liners. The project was funded heavily in the early

1990‟s, but the GTZ project ended in 1995, which had a great impact on the ability of the Ministry of Agriculture to distribute the liners (Okello, 2005).

In some areas of Western Kenya the kuni mbili proved to be popular and eight women‟s groups near Kakamega and Kisumu have continued successfully producing liners. This was in part due to the influence of projects funded by ITDG, which aimed to develop marketing and quality control capacity amongst potters. Partially because of the high quality of clay needed to construct usable liners, only one women‟s group was established and produced liners in Trans Nzoia. That group was established in

1989 and had about twenty female potters who were mainly older widows at the time of this study, as well as two kilns to fire the liners. They have been inconsistent in producing pottery because of an increase in the price of wood fuel, which they purchased by the truckload as well as financial issues within the group. Additionally, demand was not high because apart from Ministry of Agriculture extensionists, they did not have many ways of actively marketing their product. This in turn meant that after hearing about the benefits of kuni mbili, rural residents either had to pay to travel to visit this group and then carry the heavy liner back with them by public transport (in some cases doubling the cost of the stove), or rely on the under-funded

Ministry of Agriculture staff to collect and carry them.

The Ministry of Agriculture Gender and Home Economics staff were passionate about this stove and its potential benefits, however they acknowledged that after a woman placed an order with them for a liner, it could take up to a year before she

167

received it. This was because they waited until they had enough orders to make it economically feasible to request SCC-Vi or another partner to pick them up and transport them to the divisional offices. Some staff used their own money to purchase a supply of liners and kept them on hand at the office for interested farmers, but this was not common50.

After a liner was purchased the user had the option of encasing it in a portable metal casing, keeping it unencumbered, or building it into a clay base to be permanent in the kitchen. Finding qualified kuni mbili liner installers was at times difficult, as the GH&E staff did not have enough time to install every liner and teach every group that needed assistance. The same held true for people wanting to use chepkube and chepkube brooder. SCC-Vi trained chepkube and brooder builders were scarce in some areas and numerous in others.

Fireless cookers were also promoted by the Ministry of Agriculture, produced locally and were plagued with multiple issues. First, its high cost (800-1200 KSh) was prohibitive for most people. The cost of a ready-made basket was roughly the same as buying materials and possibly labour to make a chepkube brooder, or purchasing three metal sawdust stoves. A second reason was that not very many women knew how to make the baskets, and the groups that did it as a business used the highest quality and most expensive materials, which increased the final price. Material prices were about

80% of the price of the finished basket, and they were high because baskets had to be bought from Uganda, the insulating material (blanket waste) was from one source, and the heavy gauge plastic and black cloth used had increased in prices in the last few

50 At the end of the research period a women‟s group in Kiminini was preparing to open an outlet store in the main center which would have the potential to increase the reach of this technology substantially in that area.

168

years. A group in Mount Elgon had made baskets for 200 KSh each using alternate materials in 200651, but none were available close to this price locally.

The Need to Save Money

Every rural resident interviewed mentioned money in some way. This was factored into everyone‟s personal circumstances. People were inhibited from purchasing ready-made FES, building clay stoves because of the cost of bricks, metal rods and other materials, and from using brooders because of the start up money needed to buy chicks. For these reasons, some people stayed with TSS or used an inferior FES instead of switching to a more desirable one.

Getting micro-loans from merry-go-round groups to purchase stove materials or stoves themselves was a major asset, but for those who needed merry-go-round money for other projects or who were not part of a money lending group, opportunities to have that much money at one time were scarce. Husbands generally controlled other money in a house, but ultimately women made decisions about stoves because cooking was their priority.

Some people argued that it seemed foolish to pay money to build a stove when

TSS is free, but many more argued that that thinking was flawed. Twenty-four of 32 households interviewed purchased some or all of their fuel but many residents pointed out that their neighbours did not always understand how much money could be saved by changing stoves (this statement was confirmed with interview data). Not everyone was aware of the money spent on weekly or monthly fuel purchases. While there may have been available money in a household, men and women generally had separate expenditures, so while women spoke of their husbands purchasing fuel, many women

51 This activity was organised by me on a previous trip to Kenya using locally produced baskets, cloth scraps from local tailors, thick plastic bags, and lighter lining material. Most of the baskets were subsequently lost during the violence in that division, but several women said they had found them useful and hoped to remake them now that peace has returned.

169

had to „struggle to find their own money and not ask their husbands‟ when it came to stoves. Some men understood that spending less money on fuel meant they could benefit in other ways, but it appeared there was some factor which stopped the specific conversation about money savings with regard to decreased fuel consumption from taking place between husband and wife.

The only interview that specifically indicated saving money benefitted the family was with a women‟s stove making group (FG6), who in other ways gave a highly biased interview52. There was no triangulation to their statement, so this study can not support a claim that less fuel purchases will always financially benefit the family.

In the worst case scenario, a diversion of money from fuel purchases could result in more money spent on alcohol or other purchases that do not benefit the family as a whole53. In positive scenarios, savings of fuel could result in increased ability to buy food, pay for school fees and less conflict over money shortages in the house.

The Kenyan saying “the fruits of education are sweet, but the roots are bitter” echoes the sentiment of the Canadian saying “no pain, no gain”. In other words, a distasteful beginning leads to a desirable end. A comparatively small cash input at one time would save people money in the long term, but making that financial commitment was beyond the capacity of many people.

The Need to Conserve Fuel

Conserving fuel was only desirable to a person who felt they would benefit by using less of it. A few participants and many acquaintances of participants did not believe they needed to use less fuel, or did not consider how they might benefit from

52 These women purposely gave answers that they felt would benefit their group, so they exaggerated on several points. A member of SCC-Vi staff who was present later translated what they had been saying to eachother. 53 This is the kind of scenario theorised with the Home Economics Model (Moser, 1993) mentioned in Chapter Four.

170

using less of it. Three people said since they had ample fuel on their properties they wouldn‟t want to use a technology that conserved more of it, even though one of them purchased cobs to supplement their supply. Another two women said they did not lack fuel but would appreciate using less of it because “even if I have some, it can run out”. Surprisingly, only one person mentioned the time savings in terms of time spent gathering wood as a result of using less fuel.

For the majority of participants fuel was scarce at some point in the year.

Sawdust was scarce because the carpenters and small scale lumber mills did not produce enough to meet demand, cobs and stalks were scarce when the supply ran out until harvest, and wood was especially difficult to get from April to August when the constant rain made procuring dry wood complicated or dangerous (see Appendix 16 to examine people‟s fuel procurement habits). Several participants identified deforestation and low tree cover as the cause of their fuel shortage. Local administrators and agricultural officers spoke of people in their jurisdictions eating semi-raw food or burning plastic margarine containers because they lacked other fuels.

The final component of the personal circumstances key was being in a position to gain new knowledge. People that lived in areas infrequently visited by extensionists from NGOs or GOs had limited opportunity to be formally exposed to new technologies and people living in ethnically homogenous areas had limited opportunity to be informally exposed to new technologies. Also, the post-election violence that occurred earlier in the year of this study negatively impacted the ability of several participants to gain information about or even use FES. One could not attend group meetings because she was recently widowed, another had her kitchen (and chepkube) burned down in that time, and the groups of two others folded as a result of the

171

chaos. Another participant was an internally displaced person who was staying for a short time at the compound of her absentee brother. She used primarily maize stalks for cooking, and said she just learned of chepkube when she arrived in the area and cooked with TSS because she wasn‟t ready to change yet.

Key #4: Stove Characteristics

The final set of factors influencing people to actively use a FES lay in the stove itself. Knowledge and ability to use a stove were not sufficient to translate into active usage if the technology did not meet a need or solve a user‟s problem in some way.

For instance, if a woman routinely cooked ugali for her large family in large pots that required both hands on a stirring stick, a kuni mbili stove would not meet her need to have a pot gripped well. Conversely, solid clay stoves were advantageous for most other instances when the solid base could be custom-made to fit each pot and drastically reduced the danger of a pot tipping or flames escaping to burn a family member.

Stove Characteristics and Their Corresponding Benefits

Ten physical characteristics of FES were identified and displayed in a table along with the FES they are associated with in Appendix 17. The direct, indirect, and long-term benefits listed by participants that stemmed from each physical characteristic are listed in Appendix 18. The most important findings related to those analyses were that chepkube brooder, and secondly chepkube had the greatest number of associated direct benefits, and that of the FES listed, kuni mbili, chepkube, clay stove, and metal sawdust stove all had the same long-term benefits.

Despite widespread perception of fuel scarcity, only one participant said that they started to use their stove specifically because of it‟s fuel conservation qualities.

172

Other people listed it as a benefit, and something they appreciated, but not the characteristic that served as the tipping point in their decision making process.

When people listed the reasons they began using FES, they specifically listed only five direct or indirect benefits of physical characteristics (ability to cook early, to spend less time cooking, to use less fuel, to have a cleaner kitchen, and to keep bath water warm). That being said, about 90% of the direct and indirect benefits in

Appendix 18 and about half of the long term benefits were based on responses from rural residents to other questions. The remainder of benefits were based on responses from agricultural officers or me. All long-term benefits ultimately contributed to the improving a user‟s self worth.

Food warming capability

Keeping food warm was a major advantage recognised by both men and women used chepkubes and fireless cookers; however it was not a salient point for others. An agriculture officer said that when men came home late, they were used to their wife

“waking up and welcoming them”, so often they “could not accept to just take food from a basket”. A particularly poignant interview was with a 26 year old woman who had no formal education, had been married for ten years, and knew about chepkube because her mother used it, but said she chose to use TSS “because when you have a husband that is a drunkard you need to make food quickly”. She believed the speed at which she could make a fire and cook food on a TSS when her husband came home late was faster than cooking on a clay stove. It seemed taking food from the food warmer was not sufficient to meet her husband‟s expectations, so she remained with the less fuel-efficient technology.

Ugali was prepared slightly differently by each ethnic group, and this fact played a role in what kinds of stove characteristics were useful to them. Kalenjins

173

preferred „soft‟ ugali (prepared with more water), while Kisiis preferred „hard‟ ugali

(prepared with less water, which also made it more difficult to „turn‟ as it finished cooking). Other ethnic groups fell somewhere in between. When soft ugali was put in a chepkube cupboard to keep warm, it essentially continued to cook as water evaporated, creating a layer „like cake‟ or „bread‟, which people enjoyed eating. A

Kisii woman who previously used chepkube said that while her Kisii friends would like to save fuel, they would not like the chepkube because they would have no use for the cupboard as they would have to cook their ugali differently to take advantage of the food warmer, and they also preferred to eat food as soon as it was cooked. No one mentioned specifically listed „food wastage‟ as a disadvantage, but this was implied when people spoke of foods (like ugali) being unpalatable once cool.

Several women had no use for the food warmer because they said they could cook whenever they wished as they did not work away from the house. For these women, the extra savings of wood each time they lit a fire was either not recognised or not a big enough issue to switch stoves. Ironically, a few people said they knew of people who could not or did not use FES because they lacked wood all together. For those people whose primary fuel was maize stalks, the TSS seemed to work better than any other stove for commonly accepted but unexplained reasons.

The increase in a woman‟s flexibility in time is the most striking impact of using the chepkube stove, however Sen argues:

“given a highly traditional division of labour . . . women‟s perceptions of their

own welfare are likely to be far less distinct from their perceptions of the

interests of their families as a whole than are men‟s, with the result that even

less value will be attached to their own personal interests, within the already

unequal bargaining situation” (in Okin, 1995, p. 288).

174

Therefore, it is important that new technologies can be seen to have benefits for the family as a whole for optimal use. For the chepkube, the characteristic that was repeatedly reported to benefit every family member was the ability to keep food warm.

Preferences of Characteristics for Different Situations

Different characteristics of stoves appealed to different people. Using a fireless cooker to warm water, producing cleaner food in improved TSS, and making money from a chepkube brooder were the stove characteristics that stood out to men.

Women in particular were swayed by the ability to cook faster and keep the kitchen clean and both genders appreciated saving fuel and keeping food warm. Other benefits which were mentioned as general benefits but as factors that made people switch included increased safety for children, pot stability, improved marriage relations and an increased feeling of self-worth.

One pair of brothers recounted how they encouraged their wives to build improved TSS specifically because it would decrease the amount of ash in their food.

This characteristic was important to them, but others are either not bothered by ash or do not find it to be a reason to change stoves for. Other stoves were used specifically because they shielded the fire from wind and rain, produced less smoke, or in the case of the metal sawdust stove, were portable.

Chepkube brooders and chepkube bakers were typically used not because of their cooking ability, but rather for their ability to keep chicks warm or produce baked goods, respectively. The Chepkube Women‟s group in the Cherangani Hills said that their husbands especially encouraged and supported them to start making chepkube brooders because of the added income it would bring. Men in other areas were also

175

interested in seeing chepkube brooders made because of the financial windfall that could come to the house as a result of a successful chicken selling business.

For some women, using a new technology was inconvenient. Finding money to buy materials, and then building and upkeeping the stove weekly or monthly was too bothersome to consider switching. Women in Endebess who did not know of many sources of suitable soil, were especially discouraged at the prospect of having to purchase soil or face constant repairs (several times a week) if they used local soil.

Younger people generally found it easier to change with styles, but older (and some younger women) women found change difficult because they had spent their life used to a style of cooking, sitting low to the ground, feeling a special kind of heat radiation, using a certain type of fuel, and tasting food with a certain flavour.

Assessing the Comparative Importance of Key Factors

Agricultural officers expressed frustration that despite their best extension efforts, rural residents were “being difficult to change”. They asked “how do you help people who don‟t want to help themselves?” If money was an issue, one said, they could sell a chicken and buy an FES right away.

Figure 6.2: Factors which influenced current FES users to use FES

176

SCC-Vi and Ministry of Agriculture extension staff were certain that culture and exposure to information were the primary factors influencing one‟s decision to use an

FES. Figure 6.2 shows that cultural expectations and exposure were the two main keys that led to the use of chepkube, but predictably culture did not affect the use of any other stove. The data show that in most instances the most important factor someone listed (perhaps the final key to be gained) was linked directly to the characteristics of the stove. For instance, several people who used the improved TSS said they learned about it from their families, but the reason they started to use it was for fuel conservation or to keep food clean (i.e. stove characteristics). Data were not included for the factors that led previous FES users to originally start using FES.

The figure below is compiled using data from this study. It shows a chain of events that led to the use of chepkube in an ethnically diverse village such as Kiboino in Cherangani. The shaded boxes show that different conceptual keys were at play at different stages of the spread of the chepkube.

Figure 6.3: A simplified timeline of events leading to widespread use of chepkube (CK) in an ethnically diverse village

177

People discontinued using FES because personal circumstances changed, weakening their capacity to use the technology, or eliminating the need for it. For instance, the woman in Kaplamai stopped using her chepkube bakery because fuel increased too much in cost, and she stopped using her brooder because she couldn‟t find a market for the chickens. A few women stopped using their chepkubes because their kitchen or stove was destroyed (personal circumstances), and one stopped using it because she decided she didn‟t need the food warmer any longer as she was no longer working away from home (stove characteristics).

Figure 6.4: Key factors which influenced participants to discontinue use of a FES

Figure 6.5: Key factors that influenced a person to not use a FES of which they were aware

178

Participants reported not using an FES of which they were aware for reasons that fell in all key categories except culture.

Summary

This chapter interpreted and analysed the study findings according to a new conceptual framework, and then outlined the main findings to show how each objective was met. Four keys must be gathered to unlock each hypothetical door to a particular FES. The keys relate to cultural and traditional beliefs and practices, quality and amount of exposure to the technology, personal circumstances of the potential user, and the characteristics of the technology itself. The final chapter will summarise the findings in context of technology adoption and fuel-efficient stove literature, present the main conclusions of the study, and finally offer recommendations for future action by extension workers, the policies that govern their actions, and for future research in this topic.

179

CHAPTER SEVEN

SUMMARY, CONCLUSIONS AND RECOMMENDATIONS

Introduction

This study examined the beliefs, perceptions, and type of exposure rural residents of Trans Nzoia District, Kenya had for the traditional three-stone stove (TSS) and several fuel-saving stoves and devices. The study was originally organised around a conceptual framework that represented formal extension efforts for one type of FES.

Research was carried out using qualitative methods and a standpoint epistemology, which privileged the voices of the average citizen. Data was then analysed using open coding and organised according to themes. During the data analysis process a new conceptual framework emerged, which more accurately incorporated factors reported by participants that affected the use of formally and informally promoted FES.

This chapter discusses and summarises the study findings, provides the main conclusions of the study organised by objective and conceptual key, gives recommendations regarding policy for extensionists and for future research. The paper is concluded by some final remarks.

Discussion of Findings

This study divided people into two or four categories of FES use but recognised that people fell along a more stratified gradient between the simple terms „user‟ and

„non-user‟. The nine stratified categories proposed below are a result of the interplay between three variables: when a person started using FES, if they used one at the time of the study (and if so, for how long), and if they intended to use one in the

180

future. Each person in the stratified categories had collected (or failed to collect) different „keys‟ in their life that impacted their current position.

Table 7.1 Comparison of methods of categorisation of FES use

This study‟s Users Previous users Aware Unaware perspective

Adopters, Kiptot et testers or No Testers or No al.'s (2007) Adopters pseudo- category pseudo category Non-adopters perspective adopters in model adopters in model (unknown)

Rogers‟ (1983) All All EA, EM, LM All All L L none perspective but L *

Simplified Users Non- users perspective Stratified „zones‟ Zone 1------2------3------4------5------6------7------8------Zone 9 perspective Awareness Parents Used as Used as an Parents Used Used as Parents Aware Unaware of FES used it an adult adult used it as an an adult used it but has of FES adult not used FES Are they Yes, has Yes, for Yes, for No No No No No No currently always >7 years <7 years using FES? used it

Do they Yes Yes Yes Yes Yes No No No No intend to use FES in the future? *Key for Rogers‟ (1983) categories: All= innovators (I), early adopters (EA), early majority (EM), late majority (LM), and laggards (L)

Kiptot et al. (2007) suggested four discrete categories of technology users: adopters, who are in „for the long haul‟; testers, who are not sure if they will stay with the technology long term; pseudo-adopters who plan only to use the technology for a short time to get some external benefit (usually from the extension program) and then discard it; and non-adopters, who have rejected the technology outright. This study does not support Kiptot et al. (2007)‟s model on several counts. First, only two people encountered in the study could have potentially fit the pseudo-adopter

181

category (they had both hosted farmer field days at their houses and got a kuni mbili stove installed) but as many participants pointed out, “it is too difficult . . . to know another person‟s thoughts”, so their intentions to continue using the stove could not be deduced. Second, Kiptot et al. (2007)‟s categories assumed first generation use of a technology and did not consider people who had used the stove in their original home, but were prevented from using it or chose not to use it in their adult home for unknown reasons. Third, Kiptot et al. (2007) assumed that a person‟s current state of using or not using a stove was based on a concrete decision rather than the unpredictability of an individual‟s circumstances (such as the ability to access the proper fuel, living in a permanent or rented house, or the present state of their kitchen‟s structure).

This study supports the theory that people tend to occupy the highest possible rung(s) that they can afford on a hypothetical energy ladder, but also includes the caveat that fuels used in stoves that are purchased (such as a metal sawdust stove) are higher on the ladder than fuels (such as maize stovers) that are most commonly used in stoves that are free to build. Also, fuels that are simpler to collect should be situated lower on the energy ladder than those that require more effort. Depending on the season and thus the availability and price of different fuels, the rungs must be rearranged slightly.

There were slight differences between men and women‟s position on the energy ladder. Several participants said that if men needed to cook for themselves they would preferentially use a charcoal stove over one that used raw biomass fuel.

Several women stated that they would also prefer to just use charcoal stoves, but circumstances prevented them from doing that (namely the difficulty of preparing

182

large amounts of food on a small stove and the cost of using it for everyday cooking).

This set a double standard that fuelled the gender-energy-poverty nexus.

Several authors have argued that when discussing fuel efficiencies of traditional cooking fires, one must be mindful of their multiple and simultaneous functions. This study found that while the TSS gave off light, it was not used to illuminate anything apart from cooking, and that other improved biomass stoves gave off sufficient light for that task. Some women also made windows in their kitchens to help with illumination. Similarly, while the TSS gave off heat that was appreciated especially by the very old and young in the cool evenings and mornings, people who switched to FES built of clay did not complain about a lack of heat. In fact, charcoal stoves, which were used by about half the households, seemed to be the preferred space heaters in the evening. Finally smoke, although reduced, was still produced by

FES so its insect ousting benefits were still obtained.

The TSS was more convenient for supporting pots containing food for large groups of people at a celebration such as a wedding, but the nature of TSS meant that one could be specially assembled for these infrequent occasions. Apart from that one practical benefit, the cultural significance of TSS for some ethnic groups was the only barrier preventing its eventual disappearance from permanent status in households in

Trans Nzoia.

If the TSS is becoming redundant, what factors then are facilitating the takeover of the traditional rural hearth by a FES? People appeared to be on an adoption gradient, a notion supported by Everett Rogers (1983). He suggests that the first people in a community to use a technology typically have more disposable income, youthfulness and more education. The young English-speaking man who purchased a deluxe charcoal stove just because he thought it was innovative and

183

interesting was an example of this innovator type. The stove cost more than most FES and no one else in his home area used it but these facts did not deter him. Innovators are typically an easy sell for extension workers, but they typically make up only 2.5% of a population but do not easily influence non-innovators. They contribute to the informal extension process because they fulfill the gatekeeping role that lets in a flow of ideas into their social system.

Willingness to innovate and the ability to teach and influence others made the leader of a women‟s group who recently started using chepkube stoves a true Rogers- style early adopter. She was young, but was very interested in innovation and was successful when she tried it. On 0.1 acre she raised enough goats, cows, chickens, fruits, vegetables, fodder and fuel for her family and had extra income to build a permanent house and purchase another piece of land. She was the first in her neighbourhood to build a chepkube stove, and then was instrumental in teaching others about it. Early adopters also represent an audience that is relatively easy to reach with extension because early adopters gain the „culture‟ key quickly, and are eager to innovate and teach others once they have sufficient information themselves.

The spread of chepkube technology within this woman‟s group is an example of extensionists making themselves redundant when extension efforts are successful with early adopters (Rogers, 1983).

The group members who built chepkubes after observing the early-adopting woman represent the early majority. They built their stoves a year after the first woman as they waited to see if she would achieve success. The early majority category represents people who are more interested in the applications of a technology, and are more likely to use it if they see a match between their needs and the technology‟s solution. Many of the participants that were in zones 2, 3, 5 and 6 in

184

Table 7.1 fell into this category. It is hardest for them to find the „stove characteristics‟ key, so informal extension, which allows potential users to interact directly with people similar to them who have used the technology in question, is the best way to reach this group. Formal extension efforts with this type of person are less useful, and must focus on how the stove will benefit them, and facilitate linkages with early adopting experts and opinion leaders.

Rogers‟ (1983) late majority classification represents the portion of people in zone 8 of table 7.1 who found it hardest to gather the „stove characteristics‟ key.

They are similar in many ways to the early majority, and likely represent most of the group members who have not yet adopted a technology. Members of a chepkube group reported that some members had not started using the technology yet because the group was still too young, which I interpreted to mean they were still ruminating on the idea. Many older participants who were interested in FES were part of this category. Extension workers must take more of a coaching and supportive role for this type of person. Their apprehensive tendencies contributed to a frustrated extension worker‟s comment that “in a three year time you will see what you expected them to do at first is now being implemented”.

Laggards is Rogers‟ (1983) name for the remaining people in zone 8 of table 7.1 who are aware of the technology but will not use it unless it is incorporated in some other thing they are using. According to Rogers‟ (1983) model, 16% of members of zone 1 were also likely in this category. Since the chepkube technology has spread so ubiquitously through Nandi Kalenjin culture it was no longer a new technology that needed to be adopted. People used it as unquestioningly as TSS users used TSS in communities where its use was widespread.

185

Rogers‟ (1983) diffusion of innovation model is useful in many respects, but similarly to Kiptot et al. (2007), Rogers (1983) only addressed the issue of initial (first generation) adoption, and he lacked an explanation for people who stopped using an innovation due to circumstances outside of their control. In other words, he addressed the issues behind the conceptual exposure and stove characteristics keys satisfactorily, but fell short on the cultural factors, and did not directly address factors related to personal circumstances at all.

Studies on FES use are ongoing because there are no hard and fast rules or magic formulas which unlock the mystery of what made people use new FES in the literature. Seven general themes are presented on the next page, along with the findings of this study which support or fail to support them.

This study uniquely contributes to technology dissemination literature by developing a framework with four sets of key factors that lead to active technology use. This model differs from others (for instance, Kiptot et al., 2007 and Rogers,

1983) because it analyses the roles of locally developed technology and informal extension. The term use is used in preference to adoption because the four keys framework also addresses the issue of first and second generation users and recognises that even for people who have adopted an FES, the use of certain stoves may be in a state of flux due to their whims and changing circumstances.

Creating a record of the chepkube stove is a valuable contribution to FES literature. Although chepkube stoves have been in use in Trans Nzoia and the neighbouring Uasin-Gishu districts for over thirty years, no record of it or the more recent iterations of chepkube brooder and hatchery can be found in published academic literature. This study represents the first time these innovations are presented in print.

186

Most FES literature refers to the benefit of time savings primarily in terms of time spent collecting fuel. In contrast, this study shows that women value savings of time spent cooking much more readily than time spent collecting fuel.

Table 7.2: A comparison of factors that facilitate the use of FES found in literature with those found in this study Facilitating factors found in FES Corresponding findings from this study promotion literature FES are more likely to be used in There was no correlation between FES usage and places where fuel is purchased fuel purchasing. In fact, people were more likely than in places where it is not. to use a stove based on factors unrelated to fuel consumption.

FES are more apt to be used in People valued time savings over any other type households where the primary of benefit accrued to an FES, but there was no user‟s time is valued. valid way to assign a value to participant‟s time in this study.

Stoves are more likely to be used The locally developed clay-based and sawdust when they are easily and locally fuelled stoves were significantly more popular manufactured. than the stoves manufactured elsewhere.

The best programs incorporate Locally developed clay-based stoves were very significant interactions between popular because they could be customised to those who design, produce, and each user. The stoves promoted by the Ministry use the stoves. of Agriculture were not sufficiently customisable to meet the needs of users.

External subsidies only work if The funding from GTZ to the Ministry of they are very small. Agriculture subsidised the costs of promotion and distribution of kuni mbili and fireless cookers so the initial results were positive. However, when funding disappeared the reach of the extension efforts and the number of people exposed to the FES decreased.

Targeting a user group most likely Working through women‟s groups who had an to benefit from a FES is more interest in a specific stove was more effective strategic than general promotion. than general promotion at farm field days.

Stoves from long term programs In a sense, the Nandi community is the longest are used more readily than those established stove „program‟ in the area. They from programs that have not been have undergone numerous iterations of their established for long. enclosed stove to respond to the needs of the users, which has resulted in its widespread use.

187

Summary of Findings

The key findings from this research are summarised here according to the five objectives of the study.

Objective one: To determine the role of the traditional three-stone stove within a rural household

The main benefits of TSS perceived by participants were: it‟s simple and no- cost construction, its portability, ease of use, adjustability, and ability. The TSS held significant cultural meaning for Luhya and Sabaot people who believed the three stones represented the family (father, mother, and children), and had traditions relating the TSS to marriage, death, disagreements and curses, but they recognized it did not need to be used actively for cooking to satisfy those cultural traditions.

Practically speaking, few people knew of a way to cook with large pots apart from the three stones, so for them it was indispensable in that capacity. TSS was also used when people were renting or in a temporary location, when it was needed for ambiance outside (for instance for people to gather around and listen to stories in the evening), or when people used maize stovers as their primary fuel.

Objective two: To identify practical and strategic gender needs that are met by the use of fuel-efficient stoves

This study determined that FES had the potential to address divergent practical needs of men, women and children in relation to their gendered roles and position within the culture as well as some strategic needs specific to women. It also determined that male and female gender roles affected the use of FES in Trans Nzoia.

Unlike practical needs, strategic needs are difficult to identify and plan for because they are highly context specific. In this study, practical and strategic needs are

188

entwined in issues to do with food preparation and fuel use. The stove became more than a tool for food preparation. It emerged as a way to keep harmony in the household, as a tool to make money, as a tangible way to gain freedom from time constraints, as a symbol of progress and as a venue for abundant innovation.

The perceived benefits of using a FES were different for men, women and children. Men‟s benefits centered around practical things like good tasting food and physical comfort as well as a family‟s well being and an increased ability to use the stove himself. Children were perceived to benefit primarily from increased safety and having access to warm food and bath water. Women‟s benefits addressed practical needs like producing timely, good tasting food for her family, keeping a clean kitchen, saving and earning money and having more options in how she arranges her daily schedule. Women identified their strategic benefits (like increased sense of confidence and self-respect, improved marital relations, and ability to make money from constructing stoves) much more than men did.

Women gained self-confidence and reported feeling more valued by their husbands because they used fuel-efficient stoves. This was both because they could produce food faster and cleaner and because they could earn an income by rearing chicks or as a stove builder. FES were cleaner, easier to use, and more efficient in terms of time spent cooking, so men were much more likely to cook on one of those styles than a TSS. When men are comfortable cooking, women are freer to be absent from their homes over a mealtime. Women were also able for the first time to prepare food according to their schedule and serve it when asked (by keeping it warm in the chepkube) rather than needing to prepare food on demand. Women who built the original chepkube stoves as well as women who followed making dozens of innovations were directly meeting their strategic need to exercise one‟s own knowledge. The

189

opportunities presented by free time and time flexibility as well as the self-confidence related to FES use were the most obvious strategic needs met by FES.

Avoiding domestic violence is a strategic need for women. Marital disagreements were avoided on four counts by using FES: food was clean, food was warm, it was ready quickly and money was saved by not buying extra fuel. While some couples54 discussed what kind of FES to use, it was almost always the women who made the final decision about what stove to use in „her office‟. Despite this, because of their role in society, women were usually less financially able to purchase stoves or materials than their husbands, and they relied on male family members to build kitchens to house most FES. Not all women felt free to ask their husbands for money for a stove, although some men said they usually purchased firewood for the house. These results echo Moser‟s (1993) sentiments that gender planning should identify the practical and strategic gender issues and prioritised concerns specific to each gender.

Objective three: To identify current perceived need for fuel-efficient stoves within the community

This study determined that most people in the study area perceived a fuel shortage. It also determined that the majority of people not currently using a FES desired to use one in the future and that the majority of people currently using an FES desired to know about other fuel-efficient stoves. Curiously, these three facts were correlated but there lacked sufficient evidence to say that the perceived fuel shortage caused people to seek out fuel saving devices. Although many people listed fuel savings as a positive characteristic of FES, only one rural resident directly said they

54 This term refers to one man and one woman joined in a polygamous or monogamous relationship

190

started to use an FES because of its fuel saving ability. Instead of looking for fuel- energy efficiency, they were interested in the efficient use of human energy.

The majority of rural residents used different styles of stoves for different cooking tasks or situations and used multiple fuels to decrease their vulnerability to shortages in any one type. About two thirds of participants purchased some or all of their fuel so participants who did not perceive a fuel scarcity were both self-sufficient in terms of fuel production or equated fuel with money, and said there was a scarcity of money, not fuel.

It appeared people were ultimately more often influenced by characteristics related to a person‟s personal circumstances and stove characteristics than those related to their culture or exposure to the technology. This was different than what agricultural officers, who most often supposed low FES use was due to culture and lack exposure, predicted.

Objective four: To identify gendered perceptions around the use of fuel-efficient stoves

Most people perceived a difference between the thoughts of older and younger people and men and women concerning FES. Older people found it more difficult to change habits because they were used to a stove they had used for decades.

Generally they also preferred the TSS because they appreciated the warmth that came from it and the different way it cooked food. They also preferred using solid wood fuel. No one gave or met disapproval to or from any member of the community who used FES.

Women knew more about stoves than men did, but many men (especially younger ones) were interested to learn more about them. Men were more likely than women to discuss the physical properties of stoves that contributed to fuel efficiency,

191

but women were more likely to discuss properties that contributed to time efficiency.

When men chose to use a stove, they would preferentially use one that kept them clean, saved them time, and did not use wood fuel. When women chose to use a stove, they would choose one based on practical needs, for instance the availability of fuel, the speed with which they needed to prepare things, and the type of food they were cooking. All people expressed the opinion that the kitchen and tasks associated with it were included in the rights and responsibilities of women.

The study findings related to this objective support the gender and development discussion (Locke, 1999; Martine & Villarreal, 1997, Clancy et al., 2007) by supporting the view that women can be empowered by the stove(s) they choose to construct because they have chosen them based on the needs they want met as opposed to the needs formalised institutions (such as the Ministry of Agriculture) think should be met. Women use and know more about stoves and fuel because their socially constructed gender roles predispose them to that but men, if given the opportunity, would be able to achieve the same level of interest and skill. The shift in gender roles seen by men who use FES with improved characteristics is transforming unequal relations within the household. The socially constructed roles of men and women have the potential to be reconstructed, and create a shift towards the emancipation of women (Skutsch, 1997).

Objective five: To characterise existing and potential community level training or involvement

Stoves developed and promoted by members of the community or sold in the market were much more popular than stoves promoted by outside agencies. People learned about novel FES from formal extension, most often through involvement with

192

community groups. Peers were considered the most common source of additional information about stoves.

Few participants had attended large scale formal extension activities, and those that had did not recall learning about FES there. Participants had a general paucity of knowledge about the existence and characteristics of FES styles, likely because their primary sources of knowledge were informal channels that were regionally located. Misconceptions were common, especially about clay based stoves, and neither rural residents nor extension agents had a consistent vocabulary to identify types of FES.

Summary of Sets of Key Factors

In addition to satisfying its objectives, this study was able to respond to the main research question concerning factors that affected the use of fuel-efficient stoves in rural communities in Trans Nzoia, Kenya. The factors are most easily conceptualised in relation to the four keys that make up the conceptual framework developed through the study.

Culture

Culture determined which stove would be in a house, but it did not influence what stove would not be there. The findings of this study link FES use to societal change: the importance of traditions is lessening as communities become more heterogeneous, young people are seeking progressive technological solutions more than their elders, and men feel more able to use stoves that are cleaner and easier to operate.

The use of chepkube has become entrenched in the culture of one ethnic community, and several geographic communities. Chepkube was the most well-known

193

locally developed FES and had several physical characteristics which the kuni mbili and fireless cooker lacked, which added to its popular appeal. It was custom made for each user, had multiple burners, and kept food warm.

The traditionally divergent gendered priorities and roles served as a hindrance to the use and innovation of FES. The role of learning about and bringing new stove technology into the home was the priority of women, but without the support of their husbands, their efforts were often limited. Also, since women are concerned with thinking about these tasks, the problem solving and engineering skills of men are not accessed adequately. Male involvement in stove issues was most often limited by cultural norms that dictated men were not to be involved in issues concerning the kitchen.

Exposure

People living in more ethnically diverse communities were more likely to be exposed to and use FES that were informally promoted, such as the chepkube and sawdust stoves. Nearly all fuel-efficient cooking devices used by participants were locally developed and informally promoted.

A general disparity in terms of information flow existed between and among different areas of society. This applied to age, sex, and in some cases geography.

Information about stoves generally passed through women and young people. Men and older people were less likely to have a high level of exposure to FES, even if they were interested in them. People in more remote areas were not visited as often by agricultural extensionists so had less exposure to formally promoted FES.

People that were part of groups were more likely to be exposed to FES that were formally promoted by the Ministry of Agriculture, SCC-Vi Agroforestry and other

NGOs than people with no group affiliation.

194

Personal Circumstance

A lack of adequate funds and building expertise were two major hindrances to using clay-based FES. Being part of a women‟s group enabled women to gain practical knowledge about FES and to borrow money from each other to acquire one. As people‟s personal circumstances changed they changed which stove they relied on.

Some people stopped using FES when their circumstances made it undesirable or impossible to use one.

Most people used multiple cooking devices and accessed multiple fuels for daily cooking. This meant that most people were able to consider using another stove.

These practices decreased their dependency on any one fuel and increased their flexibility to use different methods of cooking for different tasks.

Stove Characteristics

Few people chose a stove based on its ability to conserve fuel, therefore although saving fuel is a perceived benefit of FES (just as increased safety for children is a benefit), it is not a characteristic that widely influenced participants to choose a stove for. For FES to be widely used in this area, it must significantly reduce cooking time and/or be adjustable for pot sizes, be able to cook ugali well and keep kitchens clean. These characteristics increased the likelihood of a man using the stove as well.

Participants rarely mentioned a savings of time spent collecting fuel as a specific benefit.

Locally innovated fuel-efficient stoves were significantly more popular than models promoted by formal extension. These models met more practical and strategic gender needs, and people were able to see them in practice more.

195

Limitations

The findings of this study were limited by the fact some specific information

(particularly relating to strategic needs) was difficult to obtain verbally from participants without asking leading or closed-ended questions, but these types of questions were not asked because then the researcher would be speaking for the participant in some way. Also, due to constraints in time and energy, there was not sufficient representation of participants in general or any stratified group (i.e. different ethnic groups, age or sex) to make quantifiable results. The findings show trends rather than indisputable conclusions. The results would have been stronger if there had been triangulation within households, but only five husband-wife pairs and one father-daughter pairs were questioned out of 32 households.

Despite efforts to stratify the sample to select for areas that had minimal impacts from post-election violence, there were at least three instances where participant households were directly disrupted by the violence. Those households added to the richness of the study rather than detracting from it because there is always some element of unpredictability in people‟s lives, and this study was able to reflect that.

Recommendations for Extensionists and Extension Policy Makers

This study was undertaken to respond to the needs of extensionists of SCC-Vi in

Trans Nzoia (needs that are shared by Gender and Home Economics officers of the

Ministry of Agriculture in that region). With that in mind, these action-oriented recommendations are intended for the extension field workers of Ministry of

Agriculture (MoA), SCC-Vi and any other NGO working on similar issues as well as those who make policy by which the extensionists are guided. The recommendations are

196

organised according to the sets of factors from the conceptual framework developed by this study.

Capacity Development

The following recommendations should be considered within framework of capacity development55 of the people of Trans Nzoia, the organisations and institutions at work there, and ultimately the system in which they are enmeshed. Peter Morgan

(2006) argues that capacity is about:

Empowerment and identity: “properties that allow an organization or

system to survive, grow, diversify and become more complex” (p. 6);

Collective ability: “that combination of attributes that enables a system to

perform, deliver value, establish relationships and to renew itself” (p. 7);

Recognising a systems phenomenon at play: “it usually deals with complex

human activities which cannot be addressed from an exclusively technical

perspective” (p. 7);

Seeing a potential state: “it is elusive and transient . . . [it] is dependent

to a large degree on intangibles [thus it is] hard to induce, manage and

measure” (p. 7) and finally;

The creation of public value: “the ability of a group or system to make a

positive contribution to public life” (p. 7).

The factors influencing the use of locally innovated or locally improved stoves in Trans Nzoia are also factors that influence the capacity of individuals to create value in a system. These recommendations are intended to develop capacity in both rural residents and extension workers.

55 Capacity can be defined as “that emergent combination of attributes that enables a human system to create developmental value” (Morgan, 2006, p. 8)

197

Recommendations Regarding Cultural Factors

The study results do not support the conclusion that customs of ethnic groups in Trans Nzoia prevent the use of FES. Therefore, Ministry of Agriculture (MoA) and

NGO field staff should not be pre-occupied with ethnicity as a barrier, and instead look at other hindering factors. They should support and encourage the cultural aspects that have resulted in the innovation and spread of various chepkube technologies rather than discourage people from using them. The Nandi people developed a stove that meets more needs and has more popular appeal than kuni mbili or fireless cookers. Instead of spending energy trying to promote fuel saving devices that people are less interested in, support their innovations and work with them to identify and promote modifications that increase fuel efficiency.

Recommendations Regarding Exposure Factors

MoA and NGO field staff should refine and be flexible with marketing and promotion messages, especially concerning the fireless cooker. Men generally have more money at their disposal than women do, and more men come out to farmer field days and public meetings than do women. Since most men do not pay attention to cooking devices in public situations, promote the basket cooker as a bath water warmer. Focus on the benefits for the buyer, and the family will benefit in all the other ways.

People‟s circumstances have a significant impact on what stove appeals to them. Therefore MoA and NGO field staff should take advantage of the variety of FES in the area, and ask several simple questions before recommending one stove over another. Appendix 19 holds a decision tree that can be used as an aid to recommend appropriate FES.

198

MoA and NGO field staff should start a pilot project for women to make and promote chepkube stoves (not just chepkube brooders) as a business in their local area. In other areas (Uasin Gishu, Nyanza provinces) it has been shown that the technology spreads faster when someone other than a governmental or non- governmental organisation is promoting it. If someone is relying on the business of building or installing a stove for their daily sustenance, then they will have more impact then an extension officer with many other duties in their dossier.

MoA, Ministry of Energy and NGOs should produce and distribute attractive, illustrated print media showing stoves and benefits. SCC-Vi produces calendars each year, so they should consider designing a year‟s calendar for issues about wood fuel and new technologies. Or, alternatively produce small brochures with simple line drawings illustrating the benefits, drawbacks, materials needed, and where to access additional information about different stove styles. People do not throw away calendars as much as small brochures. Distribute these calendars to area administrators, other government offices and places of high visibility.

All organisations dealing with stoves should share the successes of projects and technologies in Trans Nzoia with other projects and Ministry of Agriculture offices.

Experts on the technology benefit by doing paid training workshops, and users of the technology benefit by the advantages the technology brings. When one section of society holds information that would benefit other members of the society but does not share it, people remain ignorant or uninspired to change.

SCC-Vi should hire a contractor to document in detail and make training brochures about how to build the family of chepkube stoves.. Documenting the most efficient designs and spreading this information with different NGOs would be a

199

valuable way to honour the women that have made the technology and also benefit others.

Recommendations Regarding Personal Circumstance Factors

MoA and SCC-Vi field staff should be sensitive to unusual barriers within communities that prevent them from using FES. In Endebess, a major practical barrier was the lack of suitable building soil. Some women have solved the problem by buying the proper soil. Work with groups to procure suitable soil in bulk and sell it to make stoves. If someone has soil well-suited to moulding on their land, facilitate linkages between them and interested stove builders.

SCC-Vi should design and run (or facilitate women‟s groups to run) formalized trainers and formal training workshops, where people pay to learn. In some areas, notably Endebess there is a shortage of skilled artisans to make stoves or install kuni mbili liners. If people pay to learn the program would be sustainable, and people would also have some investment in their education. For instance, a women‟s group could send one person to be thoroughly trained, who could then build a stove for each member, merry-go-round style.

SCC-Vi and the Ministry of Energy should increase the reach of skilled stove builders. This can be done both by improving the business and marketing skills of stove builders, increasing the number of trained builders, and facilitating connections between skilled builders and rural residents interested in time or fuel savings. There is a wealth of knowledge within communities that can be capitalised on.

SCC-Vi should offer a course at the new SCC-Vi Agroforestry Academy on stove building to train „professional‟ builders who understand the logistics of fuel conservation. Understanding the physics of combustion within a chamber and the

200

engineering of efficient stoves would aid in a person‟s innovation of new models of stoves.

SCC-Vi needs to follow up with groups that were in the LIFE project with chepkube brooders. Farmers in Kiminini and Kaplamai expressed displeasure that SCC-

Vi had failed to keep a promise they made them regarding marketing of chicks. Their groups had raised chicks in chepkube brooders but did not have good ways to market them at the end so sold them at less than market value.

Recommendations Regarding Stove Characteristics Factors

SCC-Vi, working through environmental clubs and community groups should facilitate a series of participatory stove innovation workshops with the goal to refine efficiency and to improve on something already in existence. The top five entries could showcase their stoves at the Kitale and Kisumu Agricultural fairs. Men may have transferable skills from building kilns, just as women may have knowledge that could help men in their kiln construction. Perhaps make a contest between environmental clubs from different schools in a district or between women‟s and youth groups in one area at division and district levels. Provide contestants with guidelines to use locally available materials within a certain budget, and specify the characteristics the stove must have. At the end, test them and award the winners. Encourage people to try designing their own stove, or modifying a current model to make it their own rather than the stove of the Nandis or the stove from Vi. Also, encourage people to network with each other and share ideas, because when you move together, you move far.

SCC-Vi and MoA field workers could work together with rural schools to involve efficient stove design into their physics units and environmental club activities.

Innovations are coming from the people who use them. Youth are most willing to try out new things, and schools are natural meeting grounds for people of other

201

backgrounds. The diversity of schools can be a real advantage for generating new ideas, and in the end the school will benefit by gaining an efficient stove or two.

Every field worker involved in stoves should encourage innovators to incorporate chimneys into their designs. Although very few people described problems with human health connected to smoke inhalation, literature is full of references to the deleterious effects of inhaling smoke from cook stoves.

Since most people adopted stoves because of time benefits, all extension workers should discuss benefits of saving time and gaining flexibility in work schedules into discussions with rural residents. Encourage sceptical people with stories and examples of what people have done with their adjustable work day, and the secondary and long term benefits.

Further Research

SCC-Vi, and the Ministry of Agriculture often have student researchers with them for a two or three month attachment. Some of these projects could easily be accomplished in that period.

SCC-Vi should work in partnership with the Ministry of Energy through the MoA to determine the most efficient style of chepkube. Every chepkube stove observed was different than the next. Some people said that in their experience, chepkube style stoves use the same amount of firewood as three stone fires. The design of the internal chamber, the size and shape of openings for firewood and burners, the depth of the brooder section, and the materials used in construction all significantly affect efficiency of the stove. Research into optimal dimensions should be done not to make them entirely uniform (because one of the most appealing aspects of the chepkube type stoves is that they can be modified according to user preference), but so that guidelines can be established to show the best way to make the design.

202

The Ministry of Energy working through the Ministry of Agriculture should engineer an efficient and locally acceptable combination of a chepkube oven with kuni mbili liners. Currently, the Ministry of Agriculture is at odds with the community.

While the community seems to prefer and spread the technology of their indigenous innovation of home-made clay stoves with multiple burners (often with a cupboard attached), the Ministry of Agriculture is promoting only single burner stoves without cupboards built in. Research should be carried out to see the best way to combine the engineering and uniformity of the kuni mbili liner with the local preferences of cooking more than one food at a time and being given the opportunity to keep food warm and even keep chicks underneath.

SCC-Vi in partnership with the Ministry of Agriculture should document and design the most efficient model of brooder and hatchery. There is great variation in styles and models of chicken brooders and hatcheries. There have been problems reported about uneven heat distribution and of problems with rates of hatching.

Giving standardizing guidelines will help with curb these problems.

SCC-Vi should do more research into indigenous innovations and promotion in terms of fuel saving techniques among their project areas, and facilitate learning exchanges between women‟s groups in those areas. This could include other models of stoves, or alternative fuels people are using, for instance compressed bricks or dung pellets.

Concluding Remarks

The use of fuel-efficient stoves in Trans Nzoia is increasing. SCC-Vi requested that a qualitative study be carried out to determine why people were not using two particular models of FES. The primary factor influencing the low use of those models was not a lack of willingness or ability to use a new technology, but rather that rural

203

residents had developed their own technologies, which better suited their needs. The innovations in these communities represent progress because the communities are transitioning to using technologies that make more efficient use of fuel and time at times without the outside influence of formal extension.

At this time, we are moving into a world pre-occupied with energy. A study on energy efficiency as opposed to energy creation, and a study of devices powered by raw biomass fuels as opposed to something more socially acceptable, like sun, wind, or water flow? Quite simply, it was because this study sought to understand rather than prescribe factors that affected stove type and use. If acted upon, the recommendations of this study will develop capacity within individuals, organisations and systems to see and use everyday objects like stoves and actions like food preparation to address strategic needs to positively transform society.

Trans Nzoia holds a uniquely favourable position compared to other regions in

Kenya because of its cosmopolitan composition. Mixing of thoughts and exchanges of ideas happen much faster when people of different backgrounds are neighbours. In the case of Trans Nzoia, this is illustrated by the widespread use of the chepkube stove outside of the originating Nandi community. The pressure to find cooking fuel is a unifying force within communities. Despite other differences, fuel scarcity generally sees no cultural lines in a village. Thus, the drive to find a solution to it must also come from all facets of a community. The irony in Trans Nzoia is that the most fuel- efficient models have been adopted because of their efficient use of human energy rather than raw biomass fuels.

204

References

Aravjo, A., & Quesada-Aguilar, A. (2007). Gender and bioenergy. Factsheet 32, IUCN.

Arnold, M., Köhlin, G., Persson, R., & Shepherd, G. (2003). Fuelwood revisited: What

has changed over the last decade? (Occasional Paper No. 39). Bogor, Indonesia:

Center for International Forestry Research.

Bamberger, M., Rugh, J., & Mabry, L. (2006). RealWorld evaluation : Working under

budget, time, data, and political constraints. Thousand Oaks: Sage Publications.

Barnes, D. F., & Floor, W. M. (1996). Rural energy in developing countries: A challenge

for economic development. Annual Review of Energy and the Environment, 21,

497-530.

Bernard, H. R. (2000). Social research methods: Qualitative and quantitative

approaches. Thousand Oaks, CA: Sage Pub.

Birol, F. (2007). World energy outlook 2006. Paris: International Energy Agency.

Bratton, M., & Kimenyi, M. S. (2008). Voting in Kenya: Putting ethnicity in perspective

No. 2008-09). Storrs, CT: University of Connecticut.

Brown, S. (2003). Quiet diplomacy and recurring "ethnic clashes" in Kenya. In C. L.

Sriram, & K. Wermester (Eds.), From promise to practice : Strengthening UN

capacities for the prevention of violent conflict (pp. xii, 429 p.). Boulder, Colo. ;

London: Lynne Rienner Publishers.

205

Chikoko, M. G. (2003). A comparative analysis of household-owned woodlots and

fuelwood sufficiency between female and male headed households: A pilot study in

rural Malawi, Africa. (PhD, Oregon State University).

Clancy, J. S., Skutsch, S., & Batchelor, S. (2003). The gender-energy-poverty nexus:

Finding the energy to address gender concerns in development DFID Project

CNTR998521. London: DFID.

Clancy, J. S., Ummar, F., Shakya, I., & Kelkar, G. (2007). Appropriate gender-analysis

tools for unpacking the gender-energy-poverty nexus. Gender and Development,

15(2), 241-257.

Cohen, A. (1974). Urban ethnicity. London: Tavistock Publications.

Collier, P., & Sambanis, N. (2005). Understanding civil war: Evidence and analysis.

Washington, D.C.: World Bank.

Creswell, J. W. (2003). Research design : Qualitative, quantitative, and mixed

methods approaches (2nd ed.). Thousand Oaks, Calif.: Sage Publications.

Drigo, R. (2006). Woodfuel integrated Supply/Demand overview mapping (WISDOM)

methodology: Spatial woodfuel production and consumption analysis of selected

African countries. Woodfuel integrated supply/demand overview mapping

(WISDOM) methodology. Rome: FAO.

Dutta, S., Matinga, M., Panjwani, A., & Cecelski, E. (2005). Empirical evidence for

linkage: Energy, gender and the MDGs. ENERGIA News, 8(2), 6-8.

206

FAO. (2006a). Fuel-efficient stoves. Retrieved April 12, 2008, from

http://www.fao.org/sd/teca/tools/lst/lstp26_en.html.

FAO. (2006b). Global forest resource assessment 2005: 15 key findings. Rome: FAO.

FAO. (2007). Forests and energy in developing countries. Forests and energy working

paper 2. Rome: FAO.

Foley, G., Moss, P., & Timberlake, L. (1984). Stoves and trees : How much wood would

a woodstove save if a woodstove could save wood?. London: International Institute

for Environment and Development.

Ghertner, D. A. (2006). Technology and tricks: Intra-household technology

implementation and gender struggles. Gender, technology and development.

10(3), 281-311.

Hall, C. J. (2000). Racial and ethnic antagonism in Kenya. In M. A. Watson (Ed.),

Modern Kenya : Social issues and perspectives (pp. xxii, 380 p.). Lanham, Md.:

University Press of America.

Human Rights Watch. (2008). “All the men have gone” war crimes in Kenya‟s Mt.

Elgon conflict. USA: Human Rights Watch.

International Energy Agency. (2008). International energy agency. Retrieved 10/16,

2008, from http://www.iea.org/about/index.asp

207

Karekezi, S. (2002). Poverty and energy in Africa- A brief review. Energy Policy, 30,

915-4 p.

Karekezi, S., & Kithyoma, W. (2002). Renewable energy strategies for rural Africa: Is a

PV-led renewable energy strategy the right approach for providing modern energy

to the rural poor of Sub-Saharan Africa? Energy Policy, 30, 1071-7 p.

Karekezi, S., & Wangeci, J. (2005). The role of gender research in influencing

power sector policy in eastern and southern Africa. ENERGIA News, 8(2), 15-17.

Kassam, A. H., van Velthuizen, H. T., Fischer, G. W., & Shah, M. M. (1993). Agro-

ecological land resources assessment for agricultural development planning - A

case study of Kenya. World Soil Resources Reports, No. 71. Rome: FAO.

Kelkar, G., & Nathan, D. (2002). Gender relations and the energy transition in rural

Asia. Collaborative Research Group on Gender and Energy, ENERGIA International

Network on Gender and Sustainable Energy Use, DFID.

Kenya Ministry of Energy. (2002). Study on Kenya's energy demand, supply and policy

strategy for households, small scale industries and service establishments: Final

report. Nairobi, Kenya. Retrieved 9/26/2008, from http://www.knbs.go.ke/#

Kenya Ministry of Energy. (2003). Ministry of Energy sessional paper for the year 2003.

Nairobi, Kenya. Retrieved 9/26/2008 from

http://www.energy.go.ke/publicationsnew.php

208

Kenya Ministry of Planning and National Development. (2005). Millennium

development goals status report for Kenya 2005. Nairobi, Kenya: 40p.

Kgathi, D. L., Hall, D. O., Hategeka, A., & Sekhwela, M. B. M. (1997). Biomass energy

policy in Africa: Selected case studies. London ; Atlantic Highlands, N.J.: Zed

Books in association with African Energy Policy Research Network (Afrepren) and

Stockholm Environment Institute (SEI).

Komuhangi, R. (2006). Mass dissemination of Rocket Lorena stoves in Uganda. Boiling

Point, 52, 21-22.

Kumar, R. (2005). Research methodology: A step-by-step guide for beginners (2nd ed).

London; Thousand Oaks, Calif.: SAGE Publications.

Ladipo, P. (1994). The value of participant observation and experimentation. In H. S.

Feldstein, & J. Jiggins (Eds.), Tools for the field: Methodologies handbook for

gender analysis in agriculture (pp. 136-2 p.). London: IT Publications.

Letherby, G. (2006). Standpoint epistemology. In V. Jupp (Ed.), Sage dictionary of

social research methods (pp. 288-289). London; Thousand Oaks, Calif.: SAGE

Publications.

Locke, C. (1999). Constructing a gender policy for joint forest management in India.

Development and Change, 30(2), 265-285.

Mahat, I. (2003). Gender dimensions in household energy. Boiling Point, 49, 27-30.

209

Mahiri, I., & Howorth, C. (2001). Twenty years of resolving the irresolvable:

Approaches to the fuelwood problem in Kenya. Land Degradation & Development,

12, 205-10.

Makame, O. M. (2007). Adoption of improved stoves and deforestation in Zanzibar.

Management of Environmental Quality, 18(3), 353-366.

Makoloo, M. O., & Ghai, Y. P. (2005). Kenya: Minorities, indigenous peoples and ethnic

diversity. London: Minority Rights Group International.

Martine, G., & Villarreal, M. (1997). Gender and sustainability: Re-assessing linkages

and issues. Santiago, Chile: FAO.

Maynard, M. (2004). Standpoint epistemology. In M. S. Lewis-Beck, A. Bryman & T. F.

Liao (Eds.), Sage encyclopedia of social science research methods (pp. 1072-1073).

Thousand Oaks, Calif.: Sage.

Morgan, P. (2006). The concept of capacity. ECDPM Synthesis Paper. Maastricht, The

Netherlands: European Centre for Development Policy Management.

Moser, C. O. N. (1993). Gender planning and development: Theory, practice, and

training. London; New York: Routledge.

Mugo, F. W. (1999). The effects of fuelwood demand and supply characteristics, land

factors, and gender roles on tree planting and fuelwood availability in highly

populated rural areas of Kenya. (PhD, Cornell University).

210

Mukunda, H. S., Dasappa, S., Swati, B., & Shrinivasa, U. (1993). Studies on a stove for

powdery biomass. International Journal of Energy Research, 17(4), 281-291.

Nyangito, H., & Ndirangu, L. (1997). Farmers' response to reforms in the marketing of

maize in Kenya: A case study of Trans Nzoia district. Institute of Policy Analysis

and Research Discussion Paper, DP 003/97, 1-23.

Nyoike, P. M. (2008). Opening speech by the Permanent Secretary, Ministry of Energy

during the workshop to present the national biodiesel strategy to stakeholders at

the Kenya School of Monetary Studies, 13th March 2008. Retrieved April/15, 2008,

from www.energy.go.ke/publicationsnew.php

Odingo, R. S. (1971). Kenya highlands; land use and agricultural development.

(Nairobi) East Africa Pub. House (C1971):

Okello, V. (2005). The upesi rural stoves project. Boiling Point, 51, 2-5.

Rehfuess, E. (2006). Fuel for life: Household energy and health. Geneva: World Health

Organization.

Rogers, E. M. (1983). Diffusion of innovations (3rd ed.). New York: Free Press.

Sagar, A. D., & Kartha, S. (2007). Bioenergy and sustainable development? Annual

Review of Environment and Resources, 32, 131-167.

SCC-Vi. (2004). The Vi AgroForestry programme in Kenya, Tanzania and Uganda.

Upplands Vasby, Sweden.

211

SCC-Vi. (2006). 2006 baseline survey results. Kitale, Kenya: SCC-Vi Agroforestry

Project.

SCC-Vi. (2008). Personal communication. Kitale, Kenya.

Schwandt, T. A. (1997). Qualitative inquiry: A dictionary of terms. Thousand Oaks:

Sage Publications.

Seidman, I. E. (1991). Interviewing as qualitative research: A guide for researchers in

education and the social sciences. New York: Teachers College Press.

Simms, A., & Reid, H. (2005). Africa: Up in smoke? London, UK: New Economics

Foundation: International Institute for Environment and Development.

Skutsch, M. M. (1997). Gender in energy training pack. Enschede, the Netherlands:

University of Twente.

Thornton, J. (1984). A successful sawdust-burning stove. Appropriate Technology,

10(4), 30.

TIPTZ. (2008). Energy saving stoves. Retrieved 10/16/2008, from www.tiptz.org.

Wilson, M., & Green, J. M. (2000). The feasibility of introducing solar ovens to rural

women in Maphephethe. Journal of Family Ecology and Consumer Sciences, 28, 54-

61.

212

World Bank, Barnes, D. F., Openshaw, K., & Smith, K. R. (1994). What makes people

cook with improved biomass stoves? - A comparative international review of stove

programs. Washington, D.C.: World Bank.

World Resources Institute. (2007). Nature's benefits in Kenya: An atlas of ecosystems

and human well-being. Washington, DC and Nairobi: World Resources Institute.

213

Appendix 1: Matrix of Research Goal and Questions

GOAL OBJECTIVES QUESTIONS FG SSI PO LR What are the perceived advantages of a TSS? 1. To determine the What are the perceived disadvantages of a

role of the TSS? traditional What are alternate (non-cooking) uses for a three-stone TSS? stove (TSS) within a rural Under what circumstances can no other , Kenya. , household style of stove be substituted for a TSS?

(FES) use (FES) in rural

Who makes decisions in the house regarding stove purchases? Fuel procurement? What practical gender relations influence 2. To identify the use of FES? (construction, maintenance, practical and design, info transfer) strategic What can an FES do better than a TSS?

gender needs

efficient stove

Rift Valley Rift Province

- which How does a woman benefit from using an currently exist FES? for FES How does a man benefit from his wife using an FES? How does a child benefit from a mother using FES? What are the perceived disadvantages of a FES? What are the perceived advantages of a FES? Is there a perceived fuel shortage in the 3. To identify area? Why/ Why not? (i.e., do they grow current their own fuel?) perceived need What circumstances/ factors influenced for FES within using households to start using an FES? the community What circumstances/ factors would influence a non-using household to start

households in TransDistrict Nzoia of using an FES? What circumstances/ factors would

influence using households to stop using an To understand To factors that influence fuel FES?

214

GOAL OBJECTIVES QUESTIONS FG SSI PO LR Do men think differently about FES than women?

Do older men think differently about FES than younger men?

efficient 4. To identify - gendered Do older women think differently about

TransNzoia perceptions

FES than younger women?

, Kenya. , around the For a woman: what does she think her use of FES. husband, mother, mother-in law, think about using a FES? For a man: what does he think community or family members would think if his wife used an FES? How did FES users become familiarised with FES? Where does/would a current/potential FES 5. To user get more information about FES?

characterise RiftValley Province

What has been the participant's existing and experience with FES at the community potential level? (I.e. demonstrations, construction, community- etc.) level training Does or has the participant ever or

FES) use FES) in rural households in participated in community level training ( involvement Districtof specifically about FES? Does or has the participant ever

participated in community level training stove stove

To understand To factors that influence fuel where FES was mentioned?

215

Appendix 2: Links between Gender, Energy, and the Millennium Development Goals (adapted from Dutta et al., 2005 and Rehfuess, 2006)

Millennium development A gendered perspective of the contribution of fuel- goals efficient stove use in household energy practices Saving time spent being ill or having to care for sick children will cut health care expenses and increase earning capacity Where fuels are purchased, money spent on fuels will decrease Goal 1: Eradicate extreme Improved household energy technologies and poverty and hunger practices will open up opportunities for income generation Time and effort spend collecting fuel will be saved Increased opportunities for income generation with savings of time and a more efficient technology With less time lost in collecting fuel and due to ill Goal 2: Achieve universal health, school attendance, time spent with primary education homework and overall school performance will increase Alleviating the drudgery of fuel collection and reducing cooking time will free women‟s time for productive endeavours, education and child care. Reducing the time and distance that women and girls Goal 3: Promote gender need to travel to collect fuel will reduce the risk of equality and empower assault and injury, particularly in conflict situations. women Transforms household gender roles Increases voice and participation of women Stove programs provide employment for women in the energy sector Reducing indoor air pollution will prevent child morbidity and mortality from pneumonia and alleviate chronic respiratory problems among women

Goal 4: Reduce child Protecting the unborn baby from indoor pollution can mortality help avert stillbirth, perinatal morbidity and low & birth weight Goal 5: Improve maternal A less polluted home can improve the health of new health mothers who spend time close to the fire after having given birth Getting rid of open fires and kerosene wick lamps in the home can prevent infants and toddlers being

216

burned, scalded, or poisoned Monetary savings mean quality health care can be afforded Improved health and nutrition because not as much fuel required to cook foods or boil water Reduced exposure to disease vectors for persons suffering from HIV/AIDS. Lowering levels of indoor air pollution levels can help prevent 1.6 million deaths from tuberculosis annually. Reduced burden to care for HIV/AIDS patients Goal 6: Combat HIV/AIDS, malaria and other diseases Reduced drudgery for a woman suffering from HIV/AIDS Improved health and nutrition because not as much fuel required to cook foods or boil water Reduced risk of infection during fuel collection Where biomass is scarce, easing the reliance on wood for fuel through more efficient cooking Goal 7: Ensure practices will lessen pressures on forests environmental Moving up the energy ladder and using improved sustainability stoves can increase energy efficiency and decrease GHG emissions Recognition in development agendas and by Goal 8: Develop a global partnerships of the fundamental role that household partnership for energy plays in economic and social development development will help achieve the MDGs by 2015.

217

Appendix 3: Possible reasons for success or failure of stove programs (adapted from World Bank et al., 1994, p.14)

Reasons for success Reasons for failure

Program targets region where traditional fuel Program targets region where traditional fuel or and stove are purchased or fuel is hard to stove are not purchased or fuel is easy to collect. collect. People cook in environments where smoke People cook in the open, and smoke is not really causes health problems and is annoying. a problem.

Market surveys are undertaken to assess Outside "experts" determine that improved potential market for improved stoves. stoves are required.

Stoves are designed according to consumer Stove is designed as a technical package in the preferences, including testing under actual use. laboratory, ignoring customers' preferences.

Stoves are designed with assistance from local Local artisans are told or even contracted to artisans. build stoves according to specifications.

Local or scrap materials are used in production Imported materials are used in the production of the stove, making it relatively inexpensive. the stove, making it expensive.

The production of the stove by artisans or The production of the stove by artisans or manufacturers is not subsidized. manufacturers is subsidized.

Stove or critical components are mass produced. Critical stove components are custom built.

Similar to traditional stove. Dissimilar to traditional stove.

The stove is easy to light and accepts different- The stove is difficult to light and requires the use sized wood. of small pieces of wood. Power output of stove can be adjusted. Power output cannot be easily controlled.

The government assists only in dissemination, The government is involved in production. technical advice, and quality control.

The stove saves fuel, time, and effort. The stove does not live up to promised economy or convenience under real cooking conditions.

Donor or government support extended over at Major achievements expected in less than 3 least 5 years and designed to build local years, all analysis, planning, and management institutions and develop local expertise. done by outsiders. Monitoring and evaluation criteria and Monitoring and evaluation needs are not planned responsibilities chosen during planning stages and budgeted, or criteria are taken uncritically according to specific goals of project. from other projects or not explicitly addressed. Consumer payback of 1 to 3 months. Consumer payback of more than 1 year.

218

Appendix 4: Relational chart of administrative areas within Kenya, with study areas highlighted

**note: After the 2007 federal election Trans Nzoia district was split into Trans Nzoia East, Trans Nzoia West and Kwanza districts. Fourteen other districts were also created in Rift Valley Province (Musikoyo, 2009).

219

Appendix 5: History of Stove Promotion in the Study Area

1970‟s: Some members of the Kalenjin/ Nandi community begin building raised clay jikos with a „cupboard‟ to keep their food warm and multiple burners. The „Chepkube‟ is born. Early 1980‟s: GTZ forms a partnership with the Kenyan Ministry of Agriculture (MOA) Early 1980‟s: SCC-Vi starts an agroforestry project in West Pokot to help arrest desertification Mid 1980‟s: SCC-Vi moves their project area to be centered in Kitale. Early 1990‟s: the MOA begins promoting the Kuni Mbili liner and many women‟s groups get funding to build kilns and produce the liners Mid 1990‟s: the MOA begins promoting the Fireless Cooker Early 2000‟s: Women around the Nandi Hills begin constructing Chepkube stoves with brooders underneath. Early 2000‟s: GTZ loosens partnership with MOA and women‟s groups, encouraging them to continue their businesses without external sponsorship. MOA continues to act as brokers and distributors of KM and FC. Mid 2000‟s: Women around Nandi Hills begin constructing hatching compartments in their Chepkube Brooders.

Unknown: the Sawdust jiko gains in popularity as fuel becomes scarce.

220

Appendix 6: Focal areas within study area and their average rates of use of fuel- efficient stoves (SCC-Vi 2006, 2007) Abbreviations CH= Cherangani Division KA= Kaplamai Division CE= Central Division KW= Kwanza Division KI= Kiminini Division SA= Saboti Division EN= Endebess Division PEV= Rejected due to Post-Election Violence

2006 Baseline 2007 Average % survey Progressive of Selection Division/ Focal results survey households Category status Area (%) results (%) using FES Selected CH/Cherangani 85 85 >70% CH/Chepsiro 74 91 83 KA/Motosiet 63 48 55 rejected- PEV CH/Milimani 63 48 55 Selected KA/Sitatunga 54 56 55 40-69% CH/Suwerwa 54 54 CE/Kibomet 44 50 47 KW/Kwanza 43 43 KA/Kaplamai 40 40 KW/Kapomboi 31 39 35 CE/Waitaluk 33 34 34 KW/Kaisagat 29 29 KI/Kiminini 26 26 20-39% Selected KI/Nabiswa 11 38 25 KI/Sikhendu 29 16 22 KA/Sinyereri 21 21 CE/Namanjalala 21 21 KW/Keiyo 17 24 20 KI/Matunda 18 18 SA/Kinyoro 17 18 18 CE/Matisi 15 PEV EN/Mabere 14 14 Selected EN/Endebess 11 11 <20% SA/Saboti 6 12 9 SA/Machewa 9 9 SA/Kisawai 9 9 PEV EN/Matumbei 3 14 8 PEV EN/Chepchoina 6 10 8

221

Appendix 7: Intended and actual research participants and activities

Intended List of Research Participants and Activities Area Focus group KII- culture KII- MinAg KII- area admin CH/Cherangani FES- Male users 1-2 people Home Ec officer DO, chief KA/Sitatunga FES- Female users 1-2 people Home Ec officer DO, chief KI/Nabiswa FES- Female non- 1-2 people Home Ec officer DO, chief users EN/Endebess FES- Male non- 1-2 people Home Ec officer DO, chief users Actual List of Research Participants and Activities Area Focus group KII- culture KII- MinAg KII- area admin CH/Cherangani 1. FES- Male none 1. Crop specialist* None users 2. Home Ec officer* 2. CK women‟s group KA/Sitatunga – 1. FES- 1. A man who 1. DAEO (Dist ag 1. DO* villagers chosen Female users didn‟t know much ext officer)* 2. chiefs/ b/c had to give 2. Impromtu 2. A man and 2. Home Ec officer* councillors- fertilizer to conversation woman who knew 3. environment random* their groups in chief‟s some things officer* 3. chief, office Sitatunga* KI/Nabiswa 1. group KII- male and females- 1. DAEO & Home Ec 1. chief* culture and FES officer* EN/Endebess none 1. didn‟t say 1. Home Ec officer- 1. DO, chief – anything plus tour to KM told us to write 2. KM home w/ home* a letter but ag officer never replied.* Other Key 1. Kuni Mbili group, Central Division Informant 2. Chepkube group, Central Division Interviews 3. the old lady Okumu knew- Bukusu culture (interviewed by a research assistant only) 4. Juliet - Kalenjin culture* 5. Ann- general culture* 6. Crispin - Sabaot culture* 7. Ann - Chepkube stove* 8. Sarah - Chepkube stove* 9. Wood seller, Kiminini 10. Okoa CBO- Kiminini* 11. Eldoret Min of Ag- Livestock* 12. Eldoret Min of Ag- Home Ec* 13. Eldoret Min of Energy* 14. GTZ- Kakamega Notes: KII= Key Informant Interview *= interview was conducted in English

222

Appendix 8: Survey for people whose household used a fuel-efficient stove at the time of the survey

Fuel-saving stoves survey (Users) 1 June-July 2008

SCC-Vi Agroforestry University of Guelph Wood-Saving Technology Survey (Users)

District Division Focal Area Village Date

Section 1: Identification particulars Household Name Age Gender Tribal affiliation

Relationship to HHH Polygamous Land size # people in Education level HH? household

Do you belong to any groups?

Section 2: Mawe Tatu What kinds of jiko do you use? Do you also use mawe tatu?

Do you use kuni? How many months Where do you get it? How much is How Msogoro? Masagari? of the year are you (purchase, collect wild, consumed? much Makaa? SD? Other? self-sufficient i.t.o. grow) Who does this cost? fuel? work? How long?

(fem) Who are the people that help with cooking? (male) have you ever used an FES?

223

Fuel-saving stoves survey (Users) 2 June-July 2008

What challenges do you face while using Mawe Tatu? Tell me about them…

Section 3: Wood Saving Devices What kind is it?

How often do you use it? (why?)

Apart from cooking, what other purposes does the jiko serve?

What are the advantages? Any more?

What are the benefits for different members of the family? Apart from practical benefits, are there any kinds of social benefits? Husband‟s benefits Wife‟s benefits Children‟s benefits

224

Fuel-saving stoves survey (Users) 3 June-July 2008 How did you come to have this jiko in your house? Where did you first learn about your jiko? Any other place? How long ago? Did you pay anybody?

Who made the decision about having one in your house? Did you discuss?

What did your husband think then? What does he think now? Did he have an opinion about you changing jikos?

What about other people- did anyone disapprove when you started using this jiko?

Have you noticed a difference in the opinions of men vs. women?

Have you noticed a difference in older people vs. younger people?

Have you noticed any other differences that affect what someone thinks of the jiko? (i.e. grow own wood, don‟t cook much, tradition…)

225

Fuel-saving stoves survey (Users) 4 June-July 2008

Section 4: Jiko knowledge Is there anybody else that you know that uses some sort of fuel-saving jiko? Who?

What other types of fuel- saving jikos do you know about?

Name From where? Seen Advantages Disadvantages only or used?

Have you ever gone on any farm tours?

During these tours, did you see a jiko being demonstrated?

Where would you go if you wanted more info, or had questions about a/ your jiko?

226

Fuel-saving stoves survey (Users) 5 June-July 2008

Do you know people who do not use any kind of efficient jiko/meko?

Why do you think they don‟t use one? Anything else? Why? What makes them that way?

Have you ever taught people how to use or make this jiko?

Are there disadvantages to your jiko? What are they?

Section 5: comparing TSF and FES Is there anything that Mawe Tatu is better for than an FES?

Is there anything a Mawe Tatu can do that a FES cannot do?

Is there anything that a FES can do that a Mawe Tatu cannot do?

Is there any reason why you would stop using this jiko? (i.e. if you moved to a place with plentiful fuel)?

Is there anything you would like to change (or have changed) about your current jiko?

227

Fuel-saving stoves survey (Users) 6 June-July 2008

What values do you attach to Mawe Tatu? Do you have any cultural traditions to do with the three stones?

Is there anything else you would like to say?

Key HHH= Head of household HH= Household Mawe tatu= TSF (three stone fire) or TSS (three stone stove) Jiko, meko= stove Kuni= firewood Msogoro= maize cobs Masagari= maize stovers Makaa= charcoal SD= sawdust

228

Appendix 9: Survey for people whose household did not use a fuel-efficient stove at the time of the survey

Fuel-saving stoves survey (Non-users) 1 June-July 2008

SCC-Vi Agroforestry University of Guelph Wood-Saving Technology Survey (Users) District Division Focal Area Village Date

Section 1: Identification particulars Household Name Age Gender Tribal affiliation

Relationship to HHH Polygamous Land size # people in Education level HH? household

Do you belong to any groups?

Section 2: Mawe Tatu What kinds of jikos do you use? Do you also use mawe tatu?

Do you use kuni? How many months Where do you get it? How much is How Msogoro? Masagari? of the year are you (purchase, collect wild, consumed? much Makaa? SD? Other? self-sufficient i.t.o. grow) Who does this cost? fuel? work? How long?

(fem) Who are the people that help with cooking? (male) have you ever used an FES?

229

Fuel-saving stoves survey (Non-users) 2 June-July 2008

What challenges do you face while using Mawe Tatu? Tell me about them…

Section 3: Other jikos & fuel What have you heard about Fuel saving jikos? Advantages Disadvantages

Are there any benefits for different members of the family by using an FES? Apart from practical benefits, are there any kinds of social benefits? Husband‟s benefits Wife‟s benefits Children‟s benefits

230

Fuel-saving stoves survey (Non-users) 3 June-July 2008

Who would make a decision about having one in your house? Do you/ have you discussed this issue? Why/ why not?

Have you noticed a difference in the opinions of men vs. women?

What does your spouse think of using MT vs. CK or other jiko?

Have you noticed a difference in older people vs. younger people?

Have you noticed any other differences that affect what someone thinks of a new jiko? (i.e. grow own wood, don‟t cook much, tradition…)

What about other people- would anyone disapprove if you started using this jiko?

231

Fuel-saving stoves survey (Non-users) 4 June-July 2008

Section 4: Jiko knowledge Is there anybody else that you know that uses some sort of fuel-saving jiko? Who?

What other types of fuel- saving jikos do you know about?

Name From where? Seen Advantages Disadvantages only or used?

Have you ever gone on any organized farm tours?

During these tours, did you see a jiko being demonstrated?

Where would you go if you wanted more info, or had questions about a/ your jiko?

232

Fuel-saving stoves survey (Non-users) 5 June-July 2008

Why is it that you don‟t use a FES?

Section 5: comparing TSF and FES Is there anything that Mawe Tatu can do better than an FES?

Is there anything a Mawe Tatu can do that a FES cannot do?

Is there anything that a FES can do that a Mawe Tatu cannot do?

Is there anything that you would change (or have changed) about your current jiko?

Have you heard of Kuni Mbili or Maendeleo Jiko? Sawdust? Fireless cooker? CK? Brooder?

Apart from cooking, what other activities happen in the kitchen?

233

Fuel-saving stoves survey (Non-users) 6 June-July 2008

What values do you attach to Mawe Tatu? Do you have any cultural traditions to do with the three stones?

Is there anything else you would like to say?

Key HHH= Head of household HH= Household Mawe tatu= TSF (three stone fire) or TSS (three stone stove) Jiko, meko= stove Kuni= firewood Msogoro= maize cobs Masagari= maize stovers Makaa= charcoal SD= sawdust

234

Appendix 10: Schedule for feedback sessions in villages

5 min- Welcome, prayer and introductions Okumu

5 min- Introduction to study- what we did, where, and why Allison

10 min- Discuss three-stone fire and improved three-stone fire with one and

multiple burners (include cultural issues) Okumu

10 min- Discuss Chepkube, Chepkube brooder and Chepkube hatchery Allison

10 min- Discuss metal and clay sawdust stoves, kuni mbili and fireless cooker

Grace

10 min- final discussion, wrap up, thanks and encouragement Okumu, Grace and

Allison

Appendix 11: Example of feedback form for presentation at SCC-Vi’s Olaf Palme Agroforestry Centre on August 5, 2008

Presentation of Preliminary Results August 5, 2008 SCC-Vi Agroforestry, Kitale Kenya

(Still) a burning question: understanding factors that influence fuel-efficient stove use in rural Kenya

MSc. thesis by Allison Gratz University of Guelph, Canada

Who are you? (circle one) Vi Agroforestry Officer Agriculture Officer Student

What did you agree with from the presentation?

What did you not agree with?

What would you correct?

What suggestions do you have for further research?

What suggestions do you have for further analysis?

What suggestions do you have for further feedback to SCC-Vi or Ministry of Agriculture? Please turn over to write further comments.

235

Appendix 12: List of study participants who were interviewed using the semi- structured interview guide # peopl years Perso e in House- Land of n Se Ag the hold size educa primar code x e house head (acres) -tion tribe y stove Topic Other notes daughter-in- law was present but did 1.1.1 f 45 13 m 3.00 ? Sa I Official SSI not talk 1.1.2 m 17 6 f? ? Sa T Official SSI 1.1.3 f 38 8 m 3.00 11 Sa T Official SSI 1.1.4 f 42 6 p 3.00 5 Sa T Official SSI father' 1.2.1 f 26 8 p s 0 Sa T Official SSI 1.2.2 f 67 9 m 7.00 0 Sa T Official SSI 1.2.3 m 16 9 m 6.00 8 Sa T Official SSI spontaneous 1.2.4 f 39 10 m ? Sa T Official SSI conversation snowball ? (cafe sample from 1.2.5 f 32 ? m owner) ? Sa L Official SSI 1.2.4 snowball sample from 1.2.6 f 32 ? m ? ? Sa L Official SSI 1.2.5 2.1.1a m 37 6 m 0.25 8 Ki T Official SSI 2.1.1b f 30 6 m 0.25 8 Ki T Official SSI 2.1.2 f 19 10 c large 8 Na T Official SSI Brother (27) was present but did not talk 2.1.3 m 25 13 m 0.50 8 Bu I Official SSI much 2.1.4 f 28 5 m 0.10 6 Tu I Official SSI 2.2.1 f 32 6 m in-laws 8 Ka C Official SSI 2.2.2a f 21 11 m 1.00 6 Ka C Official SSI 2.2.2b m 60 11 m 1.00 ? Ka C Official SSI 2.2.3a m 56 3 m 2.00 4 Na C Official SSI 2.2.3b f 47 3 m 2.00 9 Na C Official SSI 2.2.4 f 33 7 p big 4 Na C Official SSI 3.1.1 f 49 6 p 4.00 9 Lu T Official SSI 3.1.2 f 28 5 m 0.30 7 Bu T Official SSI snowball sample from 3.1.3 f 43 7 f 0.75 7 Ki L Official SSI neighbour 3.1.4a m 22 4 m 50.0 8 Bu C Official SSI 3.1.4b f 20 4 m 50.0 11 Bu C Official SSI 3.2.1 f 26 6 f 0.5 8 Lu T Official SSI 3.2.2 m 24 8 m 0.75 8 Bu T Official SSI 3.2.3 m 52 11 m 3.00 11 Lu T Official SSI snowball sample from 3.2.4a f 60 4? m ? ? Bu B Official SSI 3.2.3. snowball sample from 3.2.3. Former 3.2.4b m 65 4? m ? 12+ Bu B Official SSI veterinarian. 4.1.1 f 22 3 m 0.10 8 Tu L Official SSI two friends

236

were also present and spoke 4.1.2 f 32 8 f 0.50 9 Na C Official SSI 4.1.3 m 30 4? m 0.50 ? Lu D Official SSI 1-5 neighbours were also present and 4.1.4 f 45 5 f 0.50 6 Ki T Official SSI spoke 4.2.1 f 38 8 f 0.10 11 Bu T Official SSI 4.2.2 f 22 8 m 0.30 12 Bu C Official SSI 4.2.3 f 42 7 f 1.00 6 Ki T Official SSI More material goods present here than at 4.2.4a f 63 5 m 35.00 8 Bu C Official SSI other houses 4.2.4b m 68 5 m 35.00 ? Bu C Official SSI

Practice 2.3.1a f 78 4? m 30 ? Bu T SSI Practice 2.3.1b m 97 4? m 30 ? Bu T SSI Practice 2.3.2 f 45 ? m ? ? Lu T SSI Practice 2.3.3 f 32 ? f ? ? Km T SSI Practice 2.4.1a m 39 ? ? ? ? ? ? SSI Practice 2.4.1b f 43 ? ? ? ? ? ? SSI Practice 2.4.2 f 65 ? m ? ? Ku I SSI spoke Englis Practice 2.4.3a m 38 8 p 2 h well Bu C SSI Practice 2.4.3b f 31 8 p 2 ? Ku C SSI Practice 2.4.4a f 24 5 m 2 ? Ku I SSI Practice 2.4.4b m 28 5 m 2 ? Ku I SSI Key e.g. 2.1.1a Division.Village.Household.Person Sa Sabaot 1.1 Endebess division, Mtarakwa village Ka Kalenjin 1.2 Endebess division, Cheptantan village Na Nandi 2.1 Kaplamai division, Sokomoko village Bu Bukusu Tribe Person 2.2 Kaplamai division, Kemeloi village Lu Luhya code 2.4 Kaplamai division, unknown village #1 Ki Kisii 2.4 Kaplamai division, unknown village #2 Tu Turkana 3.1 Kiminini division, Masaba B village Km Kamba 3.2 Kiminini division, Nabiswa A village Ku Kikuyu 4.1 Cherangani division, Kiboino village I Improved three-stone stove 4.2 Cherangani division, Lumboka village T Three-stone stove m male Primary L Clay stove House- f female Stove C Chepkube stove hold p male (polygamous) B Chepkube brooder head c child D Deluxe sawdust stove Topic SSI Semi-structured interview

237

Appendix 13: List of study participants who were not interviewed using the semi- structured interview guide Person Area code Person Sex Age Topic Other perceived advantages and This was a women's group Unknown 20's- disadvantages of that dealt partly with 2.0 FG1 Women's group f 60's TSS & FES building chepkube. perceived advantages and Unknown Men's 20's- disadvantages of This was a man's 4.0 FG2 group m 60's TSS & FES agroforestry group. These were all contemporaries of each Luhya culture other. Some men were 3.0 FG3 focus group f/m 50+ Luhya culture chiefs or village elders. We visited four homes Chepkube 20's- Chepkube and interviewed the four 4.0 FG4 women's group f 60's history, logistics women as a group. three men were there purposefully (employed by chief) and others chief's officers 20's- FES knowledge were just there to speak 2.0 FG5 and visitors f/m 60's in the area to the chief This resembled more of a Cen- Webolela 20's- KM liner group conversation than tral FG6 Women's group f 60's production a focus group. We toured to four homes Chepkube stoves in this group before Cen- Central Woman's 20's- and the role of conducting a focus tral FG7 group f 40's groups group. didn't know any cultural stuff b/c had lived away from the area for a long 2.0 KI1 m 60 Luhya culture time wife used a CK - she showed me warm ugali 2.0 KI2 m 45 Bukusu culture for the first time Either this woman did not know much or she did not want to talk in TSS & Sabaot front of the neighbour 1.0 KI3 f 40's culture girl. This was an acquaintance of Okumu's family… who TSS & Bukusu he just met and struck 0.0 KI4 f 70's culture up a conversation with Research Bukusu and Kitale KI5 assistant m 26 Kenyan culture 3.0 KI6 Okoa CBO m 20's fireless cookers spoke with husband first- was hesitant but obliged us. Did not want to talk 1.0 KI7 farmer f/m 30's KM jiko once wife was there.

238

Kamba, Kikuyu Research and general Kitale KI8 assistant f 22 Kenyan culture 1.0 KI9 Crispin Juma m 35 Sabaot culture Sabaot culture We met randomly on the street and started a conversation. She was a culture, barriers well educated education Kitale KI10 Ann Keya f 50 to FES use professional. Serah was a friend of mine. I visited her and 28, her mother twice to f 60 learn about their Serah and Ann Chepkube chepkube brooder that Kitale KI11 Biwott brooders Ann had made. I noticed her selling wood next to a vegetable f 50's Wood usage in stand, so I struck up a 3.2 KI12 wood seller the area conversation 4.3 KI13 rural resident f ~60 Chepkube stoves informal interview 4.3 KI14 rural resident f ~60 Chepkube stoves informal interview 4.3 KI15 rural resident f 35 TSS & FES informal interview usefulness of 3.0 AD1 Chief of Kiminini m 40's study District officer history of TN 2.0 AD2 of Kaplamai m 40's E/W Chief of Motosiet, Councillors of Motosiet and 50's- need for fuel 2.0 AD3 Sitatunga m 60's saving This chief did not really want to talk with us. He Historical and would rather talk at us or Chief of current issues in get us to talk to his 2.0 AD4 Sitatunga m 50's Sitatunga gathered people. area info, ideas 2.0 AG1 Min Ag DAEO m 50's for stove use Ministry of extension Agriculture Crop methods and 4.0 AG2 Specialist f 40's challenges extension DAEO did not talk much, methods and and left once she saw 3.0 AG3 Min Ag G&HE f 40's challenges she wasn't needed extension methods and we went from her office 1.0 AG4 Min Ag G&HE f 40's challenges to a farmer's place extension methods and 2.0 AG5 Min Ag G&HE f 40's challenges interviewed at SCC-Vi Min Ag Environment FES use in the interviewed at the side 2.0 AG6 Officer m 40's area of the road Eldor- AG7 Min Ag Livestock f 50's History of I spoke with this woman

239

et Officer chepkube while waiting for my real interview person to arrive. She was very helpful. I got this contact through a contact from some conference proceedings History and that I found on the extension of internet re: indigenous Eldor- chepkube style innovation and the et AG8 Min Ag G&HE f 30's stoves. chepkube brooder Extension and Eldor- Min Ag G&HE chepkube style et AG9 district head f 40's stoves Ministry of Energy improved After the interview, he cookstove center policies re: walked us to the jua kali Eldor- manager for promotion of stations where liners et AG10 North Rift Valley m 40's chepkube were being assembled. Crop specialist we had spoken with earlier, but extension this time she was very methods and quiet with the G&HE 4.0 AG11 Min Ag G&HE f 40's challenges there. GTZ context, Kaka- new stove mega AG12 GTZ coordinator f 40's models

Key to Abbreviations 1.0 Endebess Division 2.0 Kaplamai Division 3.0 Kiminini Division 3.2 Nabiswa B Village, Kiminini Division 4.0 Cherangani Division Area 4.3 Unknown village, Cherangani Division Kitale Town where research was based, center for Trans Nzoia District Central Central Division (nearest to Kitale) Kenya's 4th largest city, about two hours from Kitale, center for Eldoret Uasin Gishu District Kakamega Town about two hours away from Kitale in Western Province FG Focus Group or Group Interview Person KI Key Informant Code AD Administrator AG Agricultural officer Min Ag Ministry of Agriculture Min Ag G&HE Ministry of Agriculture Gender and Home Economics Officer Ministry of Agriculture District Agriculture and Environment Person Min Ag DAEO Officer Officer for Deutsche Gesellschaft für Technishe Zusammenarbeit GTZ (German Technical Cooperation)

240

Appendix 14a: Sources of FES exposure for each household

KEY FES participant currently used FES participant previously used FES participant was aware of but did not use

B- Chepkube brooder J- Chepkube Bakery C- Chepkube K- Kuni Mbili D- Deluxe charcoal L- Clay F- Fireless cooker M- Cement chepkube H- Chepkube hatchery N- Dung I- Improved TSS O- Solar S- Sawdust FES house learned code Source of exposure about another town (man)- group 211 meeting K 211 friend S FES 211 neighbour C house learned 212 neighbour C code Source of exposure about 212 sister/friend another village S 111 Agriculturalists O 213 ? S 111 Agriculturalists S 213 husband's mother I 111 Agriculturalists N 213 Husband's mother C 111 friend I 214 ? O 111 friend C 214 other town S 112 neighbour I 214 parent I group- from a woman from 214 parent C 113 Uganda L 221 neighbours L 113 ICM- Circle of Light G 221 parent C 114 neighbour C 221 SCC-Vi, group B 121 neighbour - another town C 222 ? B 121 parent - another town L 222 market S 122 the 'Nandis'- neighbour? C 222 parent C 123 ? S 223 ? S 123 market F 223 Agriculturalists F 123 parent C 223 group J 124 neighbour L 223 neighbour H 124 neighbour, church C 223 parent/ community C 125 sister in law- another town L 223 SCC-Vi B 126 another town B 224 neighbour B 126 from friend L 224 parent C

241

Appendix 14b: Sources of FES exposure for each household (continued)

KEY FES participant currently used FES participant previously used FES participant was aware of but did not use

B- Chepkube brooder J- Chepkube Bakery C- Chepkube K- Kuni Mbili D- Deluxe charcoal L- Clay F- Fireless cooker M- Cement chepkube H- Chepkube hatchery N- Dung I- Improved TSS O- Solar S- Sawdust FES house learned code Source of exposure about FES 311 ? L house learned 311 market F code Source of exposure about 311 neighbour C 411 ? S 311 neighbour S 411 parent C 312 friend I 411 parent L 312 neighbour S 411 parent/ community I 312 sister in law- another town L 411 parents B 313 ? S 412 neighbour C 313 parent L 413 ? C 313 parent C 413 another town D 314 ? S 414 neighbour C 314 SCC-Vi B 421 group N SCC-Vi (woman) parents 421 group B 314 (man) C 421 neighbour K 314 unknown (a only) F 421 neighbour & group S 321 ? S 421 parent C 321 ? C 422 ? K 321 ? K 422 from friends S 321 ? L 422 neighbour (Nandi) C 321 neighbour/ group M another town- people 322 ? L 423 teaching B 322 ? S 423 neighbour C 322 market & neighbour K 423 neighbour L 322 neighbour B 424 another town B 323 market S 424 daughter- another town F 323 radio F 424 daughter- another town K 323 SCC-Vi B 424 daughter- another town S 324 SCC-Vi B 424 from neighbour C

242

Appendix 15a: Breadth and depth of FES knowledge in the study area

Focal area, Division Endebess, Endebess Sitatunga, Kaplamai Village Mtarakwa Cheptantan Sokomoko Kemeloi Participant 111 112 113 114 121 122 123 124* 125,6 211 212 213 214 221 222 223 224 * Ethnic Group Sa Sa Sa Sa Bu/ Bu/ Sa Sa Sa Sa Ki Na Bu Tu Ka Na Na Na Sa sex/age f/45 m/17 f/38 f/42 f/26 f/67 m/16 f/39 f/30 f/30, f/19 m/25 f/28 f/32 f/21, f/35 f/47, FES asked about m/37 m/60 m/56 (solicited data) CHEPKUBE BROODER A A P A A A CHEPKUBE P A A A A A P A A A C C C C CLAY STOVE A A A C A FIRELESS COOKER A A KUNI MBILI A OtherSAWDUST stoves mentioned A A P C C A A A (unsolicited data) GAS A KEROSENE A C A CHARCOAL C C C C C C C CHEPKUBE BAKERY P DELUXE CHARCOAL SOLAR A A CEMENT CHEPKUBE IMPROVED TSS C A C C A THREE STONE STOVE P C C C C C C C P C C P P A A A A DUNG A

Notes Ethnic groups * only unsolicited information received Sa= Sabaot Bu= Luhya (Bukusu) C= Currently in use Lu= Luhya (general) P= Previously in use Na= Kalenjin (Nandi) A= Aware but never used Ka= Kalenjin (general) Ki= Kisii m= male Tu= Turkana f= female

243

Appendix 15b: Breadth and depth of FES knowledge in the study area (continued)

Focal area, Division Nabiswa, Kiminini Cherangani, Cherangani Village Nabiswa A Masaba B Kiboino Lumboka Participant 311 312 313 314 321 322 323 324* 411* 412* 413* 414* 421 422* 423 424 Ethnic Group Bu Bu Ki Bu Lu Bu Lu Bu Tu Na Lu Ki Bu Bu Ki Bu sex/age f/49 f/28 f/40 f/20, f/26 m/24 m/52 f/55, f/23 f/32 m/23 f/45 f/38 f/22 f/42 f/63, FES asked about m/22 m/60 m/68 (solicited data) CHEPKUBE BROODER A A A C A A A A CHEPKUBE A A C A A P C C A P C P C CLAY STOVE A A C A P P P FIRELESS COOKER A A A A KUNI MBILI A A A C A SAWDUST A A C C C A P C C A C A

Other stoves mentioned (unsolicited data) GAS C C KEROSENE C C C C CHARCOAL C C C C C C C C C C CHEPKUBE BAKERY DELUXE CHARCOAL C SOLAR CEMENT CHEPKUBE P IMPROVED TSS A C THREE STONE STOVE C C P C C C C P P P P C C C C P DUNG A

Notes Ethnic groups * only unsolicited information received Sa= Sabaot Bu= Luhya (Bukusu) C= Currently in use Lu= Luhya (general) P= Previously in use Na= Kalenjin (Nandi) A= Aware but never used Ka= Kalenjin (general) Ki= Kisii m= male Tu= Turkana f= female

244

Appendix 16a: Fuel procurement habits in the study area

Key:

G- gathered from unspecified locations gorogoro, a.k.a. debe, basin- small bucket Sa= Sabaot m= male GS- gathered from farm gunia- gunny sack Bu= Luhya f= female (Bukusu) GF- gathered from forest Lu= Luhya (general) GR- gathered from road * only unsolicited information received Na= Kalenjin (Nandi) GO- gathered from other places (see notes) Ka= Kalenjin (general) $- purchase (100KSh= $1.50CDN in May 2008) Ki= Kisii U- unknown Tu= Turkana

Focal area, District Endebess, Endebess Village Mtarakwa Cheptantan house code 111 112 113 114 121 122 123 124* 125,6* ethnic group Sa Sa Sa Sa Bu/ Sa Bu/ Sa Sa Sa Sa sex/age f/45 m/17 f/38 f/42 f/26 f/67 m/16 f/39 f/30 Gas/ propane Kerosene Charcoal $ Wood GF G, GR GS, $ G, $ GR $ GF GF, U U Sawdust Maize cobs U U U U U Maize stalks U U U U U U Purchase notes: charcoal Purchase notes: rarely wood- 50KSh/ 50KSh/day (githeri) wood purchase 3 days from or 100KSh/3 days forest Purchase notes: 40KSh/bag <2 days cobs if cooking githeri Purchase notes: sawdust Purchase notes: other Gather notes- 30cmtree/3 one one headload- 3-4 1 trip/ day- 4 hours wood days, 3hrs armload/ 3 hours- x3days headloads/ 2 hrs- one round trip to forest day none week armload - on farm 1km away Gather notes- cobs and stalks

245

Appendix 16b: Fuel procurement habits in the study area (continued)

Focal area, District Sitatunga, Kaplamai Village Sokomoko Kemeloi house code 211 212 213 214 221 222 223 224 ethnic group Ki Na Bu Tu Ka Na Na Na sex/age f/30, m/37 f/19 m/25 f/28 f/32 f/21, m/60 f/35 f/47, m/56 Gas/ propane Kerosene $ Charcoal $ $ $ $ $ $ Wood $ GS $ GF, $ GS($), GR GS GS GS Sawdust $ $ Maize cobs GS, $ GS, $ $ GS, $ GS GS GS GS Maize stalks GS GS GS GS GS GS GS GS Purchase notes: 20KSh/ 2days 400KSh/ 20KSh/ gorogoro charcoal gorogoro 2months = lunch only

Purchase notes: 20KSh/day 200KSh/ 1 week 20KSh/ day in 50KSh/day wood rainy season <10sticks from other villages

Purchase notes: 20KSh/day (1 buys from April- 80KSh/1week 20KSh/day 70KSh/gunia cobs basin) September purchases from villagers w/ large farms, during maize shelling time

Purchase notes: 40KSh/2weeks sawdust gunia Purchase notes: takes 1- 1 1/2 other hrs to find wood and stalks Gather notes- cuts from own Removes cuts branches 6 sticks/ day. wood trees branches from cut whole own trees. trees from the Pays someone swampy area to climb and or collect <1hr. prune them 1- 2x/yr.

Gather notes- cobs lacks from April in harvest time only used when only used uses cobs Dec- and stalks to September only available when March & stalks available during harvest

246

Appendix 16c: Fuel procurement habits in the study area (continued)

Focal area, Division Nabiswa, Kiminini Village Nabiswa A Masaba B house code 311 312 313 314 321 322 323 324 ethnic group Bu Bu Ki Bu Lu Bu Lu Bu sex/age f/49 f/28 f/40 f/20, m/22 f/26 m/24 m/52 f/55, m/60 Gas/ propane $ Kerosene $ $ Charcoal $ $ GS $ $ $ Wood GO GR GF, $ GS $ GS, $ GS, $ $ Sawdust GO GO $ $ Maize cobs U GS, $ GS GS GS $ GS, $ ? Maize stalks U GS GS GS GS ? Purchase not used much- 140KSh/ only a little bit notes: charcoal just for warmth week 2 debe

Purchase one indigenous 150KSh/ 500KSh August- 150- uses 3 notes: wood tree 1600KSh week November (rain 200KSh/ stick/day April-Nov. season) week

Purchase 500KSh/ 250KSh/ notes: cobs year (10 week sacks) Purchase 20KSh/ notes: sawdust gunia 2 weeks Gather notes- cuts branches 2 hrs, gather more in 2 days/ month prune from trees plants wood from 3x/week dry season (Nov- gathering kuni (vs. December- July trees, cuts neighbour's Mar) 7days for TSS), also branches farm branches from charcoal Gather notes- only during only during only during 80KSh/gunia, 1 per stalks only cobs and stalks harvest harvest harvest month from during neighbour harvest Gather notes- sawdust free sawdust from wood other from splitter woodsplitter 0- 10 bags/ year

247

Appendix 16d: Fuel procurement habits in the study area (continued)

Focal area, Division Cherangani, Cherangani Village Kiboino Lumboka house code 411 412 413 414 421 422 423 424 ethnic group Tu Na Lu Ki Bu Bu Ki Bu sex/age f/23 f/32 m/23 f/45 f/38 f/22 f/42 f/63, m/68 Gas/ propane $ Kerosene $ $ Charcoal $ $ $ $ Wood $, G G, $ U G GF GF, $ $ $ Sawdust $ $ Maize cobs $, GS ? U GS $ GS $, GS $ Maize stalks GS U GS GO GS GS, GO U Purchase 450KSh/gunia/ notes: charcoal month

Purchase 50-100KSh/ 500KSh (5x per rainy season: 2000KSh/ tractor trailer 1000KSh/tractor notes: wood week year, 100KSh each 500KSh/tree/ April-aug or Dec (w/ trailer Jan-July, need time) month cobs) to buy more Purchase 50KSh/4 days 50KSh/bag 1per 50KSh/gunia 20bags/ 50KSh/bag 4-5 days notes: cobs month mar-aug season (25KSh/gunia during harvest. May before PEV/>renters, also purchase stalks. >pop) Purchase 20KSh/gunia - notes: sawdust doesn't last long Gather notes- uses 10 sticks of uses one heeadload dry season: 2 she sold her trees wood wood/day brother's every weekend headloads/ and her land is taken timber scraps from forest - week over, so now must sometimes branches only purchase

Gather notes- only during harvest usually uses stalks free from only during stalks free from uses @ harvest cobs and stalks (did not mention stalks- 1m tall people's fields harvest people's fields unclear if she cobs) bundle/day purchases stalks

Gather notes- other

248

Appendix 17: Physical Components of Selected FES

multiple pre- food burners solid custom- Sawdust made warming per fuel walled enclosed built fuel brooding ceramic capacity source burners firebed burners chamber compartment liner durability portability Improved TSS X X X Clay stove X X X X Chepkube X X X X X (x) Chepkube brooder X X X X X (x) X sawdust (clay) X X X X sawdust (metal) X X X X X kuni mbili X X X fireless cooker X X

X characteristic is always associated with FES (x) characteristic is sometimes associated with FES

249

Appendix 18: Direct and Indirect Short-Term benefits associated with each physical component of FES

MULTIPLE FOOD WARMING CAPACITY BURNERS RAISED FIRE BED bath food ugali fuel, food and can cook early or water can cook up to 3 ash is off the Direct benefits stays tastes fire is out of once per day stays things at once floor palatable better child's reach warm spend Can less consume less food less fuel is kitchen is indirect benefits less time cook fuel is less fuel is wasted used cleaner cooking faster used food is food is kitchen is Secondary indirect ready on ready more benefits time on time comfortable Long-term benefits 3, 6 2, 4 4, 5, 6 2, 4 6 3,6 2, 4 5 6

Key to long–term benefits: 1- Money is earned 2- User is less vulnerable to fuel shortage 3- User has increased flexibility in time 4- Money is saved 5- Increase in health and safety of user and family 6- Marriage relations improve Benefits 1-6 also contribute to: 7- Increase in self-worth of user

250

ENCLOSED FIRE BED heat is Direct user is not burned by more don't need to monitor ash is contained fire burns more efficiently benefits flames concentra as closely to add fuel ted

can less less time food is stove is indirect kitchen is cook fuel user is can cook easier to spent cleane less smoke is produced easier to benefits cleaner faste is healthier faster operate attending r operate r used stove

food kitchen is food kitchen Secondary food is man can food is man can more tastes is more user is indirect tastes ready use the ready on use the comfortabl bette comfor healthier benefits better on stove time stove e r table time Long-term 6 6 3, 6 6 6 5 2, 4 5 6 3, 6 6 3 benefits

251

SAWDUST FUEL CUSTOM-BUILT NON-ADJUSTABLE LINERS CHAMBER BROODING COMPARTMENT Direct can use alternate chicks are kept guano is increased pot stability benefits fuels warm contained chicks are raised indirect easier to kitchen is less chance of pot tipping more efficiently benefits operate cleaner than outdoors Secondary men can decreased risk of kitchen is food is not indirect use the injury to user and more wasted benefits stove children comfortable Long-term 6 3, 4 5 2 1 6 benefits

PRE-MADE CERAMIC LINER DURABILITY PORTABLILTY can use it in Direct fire burns more efficiently do not need to repair it several benefits locations indirect can cook less fuel is less smoke is produced save time save money save money benefits faster used Secondary kitchen is food is ready food tastes user is indirect more on time better healthier benefits comfortable Long-term 3, 6 6 6 5 2, 4 3 4 4 benefits

252

Appendix 19: Decision tree for determining FES that best fits a rural resident’s personal circumstances

Key: C- chepkube F- fireless booker L- clay stove S- sawdust stove (fixed or portable) K- kuni mbili OR improved TSS (K*= portable; K^= with a cupboard built in, similar to chepkube)

253