- i--, .- ' - ~~~~~~E N V I R O N M-E N T D E P A R lT M E NT P A P E R S Paper No. 034 ESS~~~~~~~~f1A1A,iH1V RMAI"' ..t;izrf-WW KI.:,,m Ws|*3 .y.:.,.; -5'--v.PARTICI iATION SERIES Lear ning ftrm the Po'or' A Particjpat6n, Poverly 4Assss-ment inK-ya2- 'Deepa Na yan. May 1996 AMR.EFunm Envionmntaly ustain able Dbevelopment 2The World Bn ESD Environment Departmeit Papers Participation Series 001 -: Participation in Educaton' Nat J. Colletta Gi1lian Perkins . 002 Participation in Water & .Sanitation Gabrielle Watson: N. Vijay Jagannathan ': :t : -0O3 - 0 ~ . - 0 - . Pa,rtici,pation in Irrigation,;' 00 -. :; ' Ruh Meinzzen-Dick ; - . - Richard Reidinger Andrew Manzardo 004 Participation in Social Funds Mary Schmidt Alexandre Marc 006 :articipation z in ountry EconoTmic ' - an R. Aronson and Sector Work Ellen Tynan 007 Designing 'Community Based -Deepa Narayan ,Development' .:; . - 020' :: -: ' ;.ff 'Partiction in -Andrew Norton Poverty Assessments -ThomasStephens 021. Participation,and Indigenaus Peoples Shelton H. Davis Lars T. Soeftestad 031 - Participation Through Tom Carroll IntermedialryNGOs Mary Schmidt Tony Bebbington - xxx Participation in Forest and Ajit Baneqee Conservation Management Gabriel Campbell Maria C. Cruz Shelton H. Davis Augusta Molnar Copies are available frml the World Bank's Enavironment Department, social Policy & Resettlement Division. L~ Social Policy and Resettlement Division Lealrning from the Poor: A Participatory Poverty Assessment in Kenya Deepia Narayan David Nyamwaya May 1996 Papers in this series are not formal publications of the World Bank. They are circulated to encourage thought and discussion. The use and citation of this paper should take this into account. The views expressed are those of the authors and should not be attributed to the World Bank. Acronyms and Abbreviations AMREF African Medical and Research Foundation CBS Central Bureau of Statistics (Government of Kenya) NGO Nongovernmental Organization PPA Participatory Poverty Assessment WMS Welfare Monitoring Survey (Government of Kenya) SARAR Self-esteem, Associative Strength, Resourcefulness, Action Planning and Responsibility Contents Acknowledgments iii Kenya PPA Study Team iv Executive Summary 1 1. Introduction: Methodology and Processes Used 7 What is Participatory Assessment? Objectives of the Study Research Methods Used? Selection of the Study Sites Nature of Data 2. What is Poverty? Who are the Poor? 12 Perceptions of Poverty Major Problems Perceived by the Poor How the Poor Perceive their Future How Does Someone Become Poor? Livelihoods and the Poor Poverty and Female-Headed Households 3. Coping Strategies and Mechanisnns 30 Seasonality Access to Food Access to Water Access to Education Access to Health 4. The Untapped Potential of Indigenous Self-Help Networks 45 Harambee and Social Networks are Under Stress Cooperatives of Self-Help Groups NGOs and Church Groups Women's Groups i 5. Urban Nairobi 52 The Situation of Female-Headed Households in the Mathare Valley and Korogocho Slums Infrastructure Coping Strategies and Institutional Support Perceptions of Poverty by Women Residents of Mathare Valley and. Korogocho Slums Conclusion 6. Mandera District Study 59 Who are the Poor and Why are they Poor? Coping Strategies District Officials and Service Provision Gender Analysis Education in One Village 7. Afterword 62 Establish Credibility Don't Hurry the Process to Meet Deadlines Tensions and Turning Points Building Capacity While Rushing Internalization Requires Time and Follow-Up Involve Other Agencies ii Acknowledgments This study was a collaborative venture between staff of the National Council of Churches of the British Overseas Development Administria- Kenya, particularly, Violet Wainaina, and the tion (ODA), AMREF, UNICEF, the World Bank, pastors and staff of Holy Trinity and St. John and the Government of Kenya. Churches in Korogocho. We would like to thank the Office of the Vice Special thanks also go to the drivers of AMREF President and Ministry of Planning and Na- and Government of Kenya who travelled days tional Development for supporting this Partici- and nights on the remotest roads in each district, patory Poverty Assessment. and sometimes to places where there no roads at all, carrying study teams to the villages. We are We would also like to thank the Provincial indebted to Betty Atieno, Callen Juma and Sarah Commissioner, Nairobi Area, and his staff, and Chitechi of AMREF and Peninah Nyakweba of the District Commissioners and their staff in the World Bank for their secretarial and adminis- Bomet, Busia, Kisumu, Kitui, Kwale, Mandera trative support. From the World Bank, we wish and Nyamira, who gave their time and expertise to acknowledge Gurushri Swamy, Jack Maas and to assist this study. Tim Marchant. The report draws upon the village reports produced by the PPA field teams Special thanks go to the Planning Bureau, the supervised by Katua K. Munguti, Helen Odido, Central Bureau of Statistics (CBS) and their field Richard N. 0. K'Okul, Violet Kimani, Charity staff for assisting in locating the sampled Kabutha, Ismail Abdirahman, and Mary Kabuka. clusters and villages, and to Francis Munene of The report was edited by Ted Howard and CBS who helped in all stages of the project, formatted by Cristy Tumale and Kerstin Canby. including sampling, training, field work and. data processing. Most of all, we are grateful to the residents and poor people of the 35 villages and two urban In urban Nairobi, we would like to thank the slums who participated in this Participatory chiefs and elders in Mathare and Kariobangi, the Poverty Assessment. iii Kenya PPA Study Team The World Bank Deepa Narayan, Study Director AMREF (African Medical & Research Foundation) David Nyamwaya, Country Coordinato]r Robert W. Morgan Justus I. Mwanje UNICEF Alfred Okinda Ogwande & Conny Nyatta District Study Teams Bomet District Kwale District Katua K. Munguti, Supervisor Violet Kimani, Supervisor Margaret Chepngetich Zena Juma Evalyne Kalya Kairachi Marimba Mary Kandie Faith Mbai Wesley Kirui Ali Shufa Busia District Nyamira District Hellen Odido, Supervisor Charity Kabutha, Supervisor John Kennedy Alumasa Veronica Bosire Ezekiel Esipusi Janet Mang'era Robert Wamoka Kenneth Nyangena Beatrice Wandera Edna Omn.wenga Kisumu District Mandera District Richard N.O. K'Okul, Supervisor Ismail Abdirahman, Supervisor Caroline Adhiambo Adell Mohamed Nur Adan Zakary Ndege Njoga Mohamud Abdalla Hassan Fred Otieno Ojuang Mohamud Maalim Issak George Achieng Okinda Robert Ou7ino Omolo Okuthe Abdikarirn Suleman Dan Oruko Idi Abdi Suleman Kitui District Urban Nairobi Katua K. Munguti, Supervisor Mary Kabuka, Supervisor Josephine Musyoka Bathsheba Kerubo Janet Mwinzila Maria Musomi Catherine Ndinda Julius Ngundo iv Executive Summary Reducing poverty is the overriding objective of Methodology the World Bank. Data on poverty are essential for policy formulation. In the past, the primary Since the primary purpose of the PPA is to type of data gathered about poverty by the ensure that findings are relevant to policy- World Bank was quantitative, with heavy making and have credibility, the methodology reliance on pre-structured household used in Kenya evolved in consultation with key questionnaires. stakeholders both in -- country and in the Bank. To take maximum advantage of information To complement the conventional approach, the that was already available and to narrow the World Bank now supports the use of more focus of the study, key stakeholders were asked open-ended, interactive and qualitative dal:a to specify what questions they wanted an- collection methods to understand poverty, swered by the PPA. particularly from the perspective of the poor. This approach is called the Participatory Getting Consensus on the Issues Poverty Assessment (PPA). PPAs pay special attention to process. The aim is to engage a Choosing Methods, Establishing range of stakeholders in the study process to Credibility generate involvement, to maximize local ownership and to build commitment to change. The key questions that emerged were: What are Each PPA is different, reflecting the country the perspectives of the poor on poverty? What context, the time available and the information are their indicators of poverty? What are the needs of policymakers. long-term trends? What causes poverty? What is their expectation for the future? Are female- Objectives of the Study headed households poorer? If so, why? How do the poor cope with poverty? What roles do A Participatory Poverty Assessment (PPA) was formal and informal groups play? What are the AnPartiatory Povert ment o PPa) was key problems and priorities of the poor? Do the undertaken by the Government of Kenya and the World Bank during February-April 1994 to poor have access to basic services, education, complement statistical studies of poverty in health and safe drinking water? Why do they Kenya. The Kenya PPA had three primary not use services? What are the perspectives of objectives: first, to understand poverty from the district officials regarding poverty? the perspective of the poor and those providing Many community level studies are dismissed services to them; second, to start a process of dialogue between policymakers, district-level as anecdotal because the findings are not seen providers and the poor; and third, to addr;ess ngit ra nd thy e particance. the issue of the "value added" of the PPA nity and thus are of little policy significance. apprissueoah toe undluerstandi poverty. PPTo overcome this hurdle, sampling issues were discussed and communities selected in discus- Participation Series Learning from the Poor sion with senior officials from the Directorate team was trained intensively for two weeks in of Planning and the Central Bureau of statistics instrument development and field testing. (CBS) of the Government of Kenya. The PPA was linked to the National Welfare Monitoring Data Analysis Survey (WMS) which used a sample of 12,000 The primary methods for analyzing the data people from the national cluster sampling were systematic content analysis and gradual framework. Based on findings from the just aggregation of data based on themes from the completed WMS, five of the poorest districts household, group, village, district and national across Kenya were chosen. Within each district, levels. five "clusters" were chosen randomly using maps available at the CBS. Two other districts - K Fid - Kisumu and Mandera in an arid area in the ey Fidngs north -- were added at the last minute because of funds made available by UNICEF. The final What is Poverty? Who are the Poor? PPA sample was 3,500 people from 35 clusters Poverty as measured by wealth ranking was or villages from seven districts, and a small widespread in the seven districts studied. As sample of female-headed households from two defined by the study's participants, poverty had Nairobi slums. increased and become pervasive: estimates of the percentage of each district's population that To address specific issues, 11 different data was "poor" or "very poor" ranged from 54 collection methods were used. Several focused percent to 85 percent. on the same issues to cross check information and to probe issues in depth. The methods In order to minirnize risk and survive, the poor were a combination of SARAR, Participatory reported that they diversified their sources of Rural Appraisal and brief household level income. The natLre of their income-generating questionnaires. Methods included mapping, activities -- including small-scale subsistence wealth ranking, seasonal analysis, trend and farming, peddling, manual labor, illicit brew- price analysis, focus group discussions, key ing, welding and cobbling, and small-scale informant interviews; visual card methods business -- varied by region depending on the were used for problem identification, gender opportunities available. Many respondents analysis, understanding health seeking behav- spoke about the worsening situation over the ior; and incomplete sentences were used to past decade. A typical statement was that, "Ten understand children's evaluation of education. years ago we had a crop in the field, well- stocked granaries, a cow. Today, because of Each team also held a day long district level land subdivision, inflation, drought, the high workshop to understand poverty from the price of seeds and fertilizers, we have small perspective of the key decisionmakers and to and patchy yields, the granaries are empty, the brief officials about the nature of the study. The cows have been sold, and sometimes to sur- in-country work was financed by the British vive, we depend on government relief food." Overseas Development Administration (ODA) with a contribution from UNICEF. The study Comparison of Findings team was coordinated by the African Medical Between PPA and WMS and Research Foundation (AMREF), an NGO. Statistically, the findings of the PPA and the The field research team consisted of 35 re- searchers, primarily sociologists and anthro- were strikingly similar. In three of the five pologists who could speak the local language of districts in which cluster sampling was used, the district. Teams of five to seven people spent the results are nearly identical. In two districts, two or three days in each village. The urban team was comprised of three people. The study therPP thod re Muc higher le of poverty than did the WMS; however, the 2 Environment Department Papers Executive Summary two districts in question had both been se- interviewed in privacy. The last question dealt verely affected by a drought which had o,'- with their vision of the future. Overall 85% of curred since the WMS. It is possible that the the poor felt that their situation would get even level of poverty in these two districts incteased worse in the future. The percentages went up as a result. to 97% in some villages. Perceptions of the Poor How Does Someone Become Poor? Village residents and the poor had no problem The process of impoverishment is understood identifying who was poor and the reasons for by villagers as having two dimensions. First, their poverty. When people were asked to list almost everywhere poverty is seen as an inter- the major problems they faced, more than generational phenomenon: "Poverty is inher- three-quarters of the poor reported that their ited. If you are born to a poor father, he cannot most important challenge was either lack of educate you and cannot give you any land or food (46%) or lack of money (30%). very little land of poor quality; every genera- tion gets poorer." Perceptions of District Officials In many districts, sharp differences were noted But becoming poor is also understood as a between the district leaders' understanding of process that can happen to anyone over time. the intensity and spread of poverty in their Primary factors cited include land subdivision, areas and the perceptions and experiences of soil erosion, low productivity, rising prices, village people and the poor. District officials reduced government services and having many linked poverty with illiteracy, whereas village children. Sickness, physical disability, old age people said the poor were often educated, and (no energy and inability to work, no children) they did not associate poverty with illiteracy, and being orphaned were other routes through The negative stereotypes most commonly used which families or individuals plunged into by district officers were that the poor are lazy, poverty. idle, drunkards, criminals, prostitutes, poor thinkers and uncaring about the future. Village nflahon people by contrast described the poor as In all districts, people linked growing poverty religious and hardworking, deeply worried with increasing prices, and compared current about their security and future. While a few prices to those of 10 and two years ago. Ac- vilagers viewed the poor as lazy, the majority cordingly to village people, prices have risen said a poor person who is lazy is unlikely to substantially on a continuous basis over the survive and, in fact, has to combine multiple past decade, and dramatically so in the past survival activities "to barely keep afloat.' two years. During the period 1984-1992, for example, prices of sugar and cooking oil, two This gap in perception was seen by the poor as basic commodities, increased 20-25 percent per an important factor in blocking programs year on average in all districts. In the past two which might otherwise be helping them. While years, these increases have averaged 100-150 interviews with village people showed a high percent per year, indicating a significant degree of awareness about poverty and who acceleration in the trend. the poor are, district leaders tended to give textbook descriptions of poverty, and the poor Cost-Sharing were viewed as an amorphous group. Cost-sharing over the past 10 years has been credited with bringing about slow improve- ments in Kenya's educational and health In every village, households that were identi- systems. At the same time, cost-sharing is fied as poor by the community at large vvere slowly depriving the poor of education and Participation Series 3 Learning from the Poor health care, as this study shows. Under the maintenance strategies. However, the poor visit cost-sharing program, while the government government health facilities only as a last resort provides teachers, parents must maintain the when they are desperate. This is not because schools and provide blackboards and chalk, they do not believe or respect the curative desks and chairs, roofs and walls, and play- powers of health providers, but because their grounds. However, in most schools visited, experience with health facilities has been so parents have simply not been able to afford to dismal over the past few years. Problems cited pay the escalating fees to do this. According to by the poor included distance, "unofficial fees," reports by the study team, Kenya's educational lack of drugs ad rude health personnel. infrastructure -- from buildings to materials to teachers themselves -- seems to be giving out Poverty and Female-Headed Households all at once. Study participants recommended For the purposes of this study, people defined bursaries targeted to the poor to assist them in female-headed households as those in which meeting their cost-sharing payments. there was no adult man living with the woman or supporting her from another location. At Analysis of household data showed thatthcoluinofhesiampigad (depending on the district) between 32 and 63 the conclusron ointhe social mappwig and percnt o poo houehods hd on or orewealth ranking exercises which were con- percent of poor households had one orimore ducted publicly {n 35 villages, people were children who had dropped out of primary asked to mark all of the female-headed house- school. When poor parents had to make a holds on a map. Overall, while 25% of the choice about who would stay in school and hlsoamp. Ovrl hle2%o h choie aboutd w w ldrtayin, schoolers and study population was categorized as very poor, who shuld bewithdrwn, dagheswr there were over itwice as many female-headed almost always the first to go. Overall, 60% of households (44%) as male-headed the oorfelttha girs souldbe ithdawnhouseholds (44%) as male-headed households the poor felt that girls should be withdrawn (21%) in this group. While 59% of the male- headed households were categorized poor or very poor, this was true for 80% of the female- Access to Water headed households. The pattern of greater While the WMS did not identify access to poverty among female-headed households was drinking water as a significant problem, the true for every district and for all 35 villages. In PPA found that access to water was a major Kitui and Busia, more than 95% of the female- challenge facing the poor. In particular, the headed households were classified as poor or study found that: in all districts except one, very poor. poor people reported access to safe water as a problem; in all districts the water problem In every district, women identified food as the became acute during the dry season; where most immediate and pervasive problem. In demand was high and community groups had some villages, female heads of households the capacity (skills, technical and business consistently limited the number of meals eaten management), local groups were managing to cope with the ever-present lack of food. simple water systems effectively without One-third of the female-headed households in government support; and successful water user Elugulu village had one meal per day (the groups branched into income production remaining two-thirds had two). Some women activities, some of which fed back into water resorted to begging for food, others relied on systems to expand the system or for repairs. brewing changaa and ajono, despite being harassed by police. Women often traded their Access to Health labor for food. In Emuramia, community The study found that the poor are health members observed that "most of the people conscious and engage in a variety of health who worked on the farms of the well-to-do in exchange for food were women." The ingenu- 4 Environment Department Papers Executive Summary ity of women to survive on very little in the a disproportionately important role in the lives face of starvation was reflected in every of poor people who are constrained in every district. way except by their labor. In addition, the PPA study found in particular, that: Women's limited control of land was an issue in all districts. Despite the fact that legally The self-help system of "Harambee" is women have access to land, when divorce or stressed, and the informal safety nets of the death of the husband occurs, customary law poor are breaking down in the absence of takes over and women are generally deprived timely support from the outside. of land. The issue of ownership and rights, to * There is a large number of strong indig- land -- which has important implications for enous self-help groups; the majority either targeting poverty alleviation programs -- was do not reach the poor or are under capital- not captured by the WMS. ized and therefore cannot meet the needs of their potential clients. The Role of Harambee and Indigenous Networks * NGOs and church groups, with a few The poor, and especially the very poor, are exceptions, are providing welfare support difficult to reach through centralized govern- to the poor, rather than supporting the ment services, particularly when the quality of poor in building strong self-help groups. service is inadequate and when cost-sharing is * The potential of women's organizations as involved. Like all societies, Kenya has indig- income producing groups is large un- enous social networks which form the basis of tapped. resource and labor exchanges and reciprocity. In every village visited, there were at least 5, Lesson Learned About Process and sometimes as many as 17, associations. This suggests that nationwide, there are between 200,000 and 300,000 groups in thE the PP essris as ints the sub- rural areas. The social capital embodies in these ane in ensuringuseo ings hen soms informal networks binds a group of people are designed and conducted in a hurry, at some together and fuels development. Under normal negative consequences. circumstances, these informal associations play Participation Series 5 i 1. Introduction: Methodology and Processes Used This Participatory Poverty Assessment (PPA), procedure in several key ways. The distinguish- conducted in 1994, was based on a sample ojF ing characteristics of the PPA are found in the approximately 3,500 people living in 35 villages participatory, interactive and sociological in seven rural districts across Kenya and one methodologies used which give "voice" to the low-income area in Nairobi. The PPA was poor and allow the exploration of issues in designed to complement and provide better depth because of their open-ended nature. understanding of the statistical findings regard- Participatory research embodies an approach to ing poverty in Kenya obtained in the December data collection that is two-directional - both 1992 Welfare Monitoring Survey (WMS) of from the researcher to the subject and to the 12,000 households conducted by the Govern- researcher from the subject. To some extent, all ment of Kenya's Central Bureau of Statistics persons involved in the study become the data (CBS). The PPA study was coordinated in- gatherers, including the local leaders who are country by the African Medical and Research contacted, but most particularly the village Foundation (AMREF) and financed by the residents and the poor people themselves. British Overseas Development Administration (ODA), UNICEF and the World Bank. This vvas Objectives of the Study the first time that the PPA methodology was applied to assessing poverty in Kenya. The study had two primary objectives: first, to understand poverty from the experience and This report, based on the Kenya study, was perspective of the poor and those providing written to inform readers about the PPA services to the poor; and second, to begin a process and findings. It deliberately does not process of dialogue between policymakers, make recommendations based on the findings. district-level service providers and the poor. For those, see the following World Bank publication: "Kenya Poverty Assessment," Questions explored during the PPA included Report No. 13152-KE, March 15, 1995. The the following: Afterword highlights some process lessons learned. * When do people consider themselves poor? *What are the perceived causes of poverty? What is Participatory Assessment? * How do the poor cope in their difficult situations? Participatory research is a process of collabora- * What are the long-term trends? Is the tive problem solving through the generation and situation improving or getting worse? use of knowledge. It is dynamic, demand-based situat irovinger getting w orse and change oriented, and seeks to raise people's * What are the gender differences? Regional awareness and capacity by equipping them with differences? new skills to analyze and solve problems. The * What access do the poor have to basic "participatory" method differs from the trad:i- social and agricultural services? When do tional interviewer/respondent survey research they use these services? Participation Series 7 Learning from the Poor * What is the perspective of district information and probing of issues with different decisionmakers with respect to the poor? groups of people. What contact do the two groups have? 1. Mapping. Groups were asked to draw a To maximize local ownership and to initiate map of their community on the ground, change, participatory techniques were used in marking the key features of the village and defining the research agenda (including the drawing irn individual households. If it was sampling framework) and in the data collection raining, the activity was conducted on process. The research agenda was defined in poster paper using markers. consultation with senior decisionmakers and staff in the Ministry of Planning and relevant sectoral 2. Poverty characteristics and wealth ranking. ministries. The Central Bureau of Statistics helped Once trust was established, people were in the selection of districts and defined the sam- asked to identify the characteristics of differ- pling clusters in the chosen districts. In each ent wealth groups, ranging from "very poor" district, meetings were held with local officials and to "rich." When consensus was reached, the some NGOs. At the end of the field work, immedi- groups used colored stickers to classify ate feedback was given to district officials where households as belonging to one of these possible. categories aind marked them on the map. The findings were discussed at a national work- 3. Seasonal analysis. The lives of poor people shop hosted by the Government of Kenya and dependent on agriculture change with the AMREF in the summer of 1995. The workshop rhythm of the seasons. People were asked to was attended by 80 government officials across draw a matrix on the ground, with the sectors, NGOs, university staff and donor repre- months written across the top on the sentatives. This was followed by a half-day horizontal axis and activities or resources meeting to discuss institutionalization of the PPA on the vertical axis. The discussion focused process within the CBS while depending on the on how people cope at different times of the private sector, NGOs and universities for expertise year and thte periods of greatest stress in in participatory methods. their lives. Research Methods Used 4. Trend and price analysis. Groups plotted changes in their lives over the past 10 years Eleven different data collection methods were and during the past two years. Prices of used. Several methods focused on the same commodities considered important were types of issues to allow cross-checking of collected. Trends related to school fees, Box 1: Ranking of Household Status Participants determined wealth categories of households in their community. In six of the eight study areas, people used four categories: "Very Poor, " 'Poor," "Medium," and "Rich." In Mandera, the large nomadic district in extreme northern Kenya, three categories emerged: "Very Poor," "Poor," and "Non-Poor." Residents did not classify any households in these villages as "Rich." In urban Nairobi, a fifth classification, "Sub-Human,' was suggested by Father Mario Porto to describe some 1,000 persons living under plastic sheets on a vast mountainside of garbage known as Mukuru, adjacent to the Mathare Valley and Korogocho slums. Speaking of the people who are turned out of overcrowded households in rural homes, many of whom end up in the Nairobi slums, Father Mario said: "When they are too poor to live in Mathare, they move to Korogocho. When they are too poor to live in Korogocho, they move to Mukuru. This is the end of the line in Kenya." 8 Environment Department Papers Introduction: Methodology and Processes Used Box 2: Dynamics of the Visual Exercises in the Villages In most of the villages, use of visuals as part of the research techniques was the high points of the study. They generated excitement and laughter, somretimes pain and a great deal of information. Because each team could spend only three days in each village, an unexpected problem caused by these exercises was that the shouts and laughter generated by the visuals would sometimes draw participants away from a focus group session or some other activity that was happening simultaneously. In one poignant episode reported in the study, a woman taking part in the Problem Identification exercise examined a picture of a husband beating his wife and suddenly shouted, 'That's me! That's what they do to us here!" And she began to cry. school dropouts, changes in water ava:il- then handed to the participants. The group ability and other factors were also re- was asked to place each card under the corded. picture of the person who owns the object in a marriage. Once this discussion was 5. Focus group discussions. In these discus- complete, people were then asked what sions, men, women and youth met in happens during divorce or separation, and separate groups usually numbering between finally what happens to the division of 7 and 10 persons. Topics included charac- property if the husband dies. teristics of the poor, ways of coping with poverty and programs that address poverty. 9. Understanding health-seeking behavior. Participants were shown a picture of a 6. Key informant interviews. Much of the health clinic, and then discussed the good information obtained through groups was and bad points of the clinic they visited. also asked of key informants such as Cards depicting family members were used district and village leaders, teachers, to elicit information about who received priests, village elders, and officials of priority treatment in health care; other NGOs, relief agencies, men's and woiren's cards depicting common health center groups and work associations operating in problems (such as lack of drugs and the sampled communities. crowding) formed the basis for discussion about the quality of the local clinic. 7. Problem identification. To avoid intro- ducing sectoral bias, a set of about two 10. Household questionnaires. Fifteen dozen drawings was developed depicting a households ranked poor or very poor were variety of problems and issues. Blank sampled in each village. In the seven paper was also provided so that partici- districts, a total of 514 (327 male-headed pants could draw other local problems. and 187 female-headed) households were Voting with stones, people ranked the live interviewed. In addition to demographic most important problems affecting thei.r data, the household surveys covered some lives. These were then discussed in depth. of the same topics as the focus group discussions. 8. Gender analysis. Three large pictures of a man, a woman and a couple were placed on 11. School questionnaires using "incomplete the ground. Smaller pictures of a variety of sentences." These were administered to possessions and objects - such as a house, 6th-grade children to learn their level of land, babies and young children - were satisfaction with their school experience. Participation Series 9 Learning from the Poor Selection of the Study Sites these became the 35 sampled villages for mapping and inclusion in the PPA (five villages Based on the 1992 Welfare Monitoring Survey in each of the seven districts). data, five districts shown to be among the poorest in Kenya were selected for the study. Teams visited the District Commissioner and The final selection included consideration of the district leaders (including district officers in security situation in the country. In addition, health, water, education, agriculture, livestock because funding related to UNICEF's special and social services) upon first arriving in each interests became available at the last moment, study area. District officials were informed Kisumu and Mandera districts were added to about the purpose of the study and an over- the study. A small sample study of the poor in view of the district was obtained. The District urban Nairobi was also included. The districts Statistical Officer then accompanied teams to were as follows: locate clusters, and CBS enumerators assisted 1. Kwale District, in Coast Province, extreme in locating the sample villages. Because of southKwaeDisterict,n Coyas Porovrince, exTrem strict random sampling, the villages were south-astemi-arn Kenyculralbrderinon.Tn representative of each district. In some cases, research teams were required to make ex- 2. Kitui District, in Eastern Province, some tended journeys to reach their designated distance east of Nairobi, also a semi-arid areas. On one occasion, a team had to cross agricultural region. into a neighboring district for a time because of 3. Bomet District, in Rift Valley Province, limited road availability, and then re-cross into western Kenya, a higher rainfall agricul- its chosen district to reach the correct villages. tural region. These distance and road factors are mentioned because they have special importance in the 4. Nyamira District, in South Nyanza Prov- findings of the PPA. ince, western Kenya, a higher rainfall agricultural region. Each study teajm consisted of a senior anthro- 5. Busia District, in extreme western Kenya pologist or social scientist as supervisor and bordering Uganda, a higher rainfall agri- four university graduates as research assistants. cultural region. The project schedules called for teams to spend 6. Kisumu District, also in South Nyanza three days in each of the five villages (15 days) Province in western Kenya, a higher and a total of three days before and after the rainfall agricultural region. field work meeting with the district leaders (a follow-up debriefing meeting was held with 7. Mandera District, in North-Eastern Prov- leaders in each district). Despite the numerous ince, a very dry, expansive and sparsely exercises performed in each village, all teams populated nomadic pastoral region border- were able to hold to this timetable (although a ing Ethiopia and Somalia. number of teams worked until 2 a.m. and arose 8. Urban Nairobi. A small urban sample of at 6 a.m. in order to keep to this rigorous 60 female-headed households was chosen schedule). from the adjacent slum areas of Mathare Valley and Korogocho. Special attention was given to urban Nairobi in the study. Mathare Valley and Korogocho are Following selection of the seven rural districts, among 14 identified "informal settlement" or five clusters in each district were chosen by slum areas in Nairobi that are often regarded simple random sampling from the master list as the areas in greatest poverty. The focus of the maintained by the Central Bureau of Statistics urban study was threefold: to understand the in its ongoing national surveys. Each cluster factors that triggered the arrival of women in contained one village or part of a village, and the slums and their coping strategies once 10 Enviromnent Department Papers Introduction: Methodology and Processes Used there; to identify some of the formal and field, a well-stocked granary and two cows, informal support systems found in urban whereas today they have only small plots with Nairobi; and to examine the ways in which poor soils and bad yields, no granary, no cows rural poverty contributes to urban poverty. A and must depend on relief food, then one can slightly modified set of research instrumenlts say that the situation has probably gotten much was used in Mathare and Korogocho. worse. While it is not possible to say how much worse with the precision deriving from Nature of Data quantitative data, an assessment can be made with respect to certain indicators, such as price The strength of the participatory method is that increases of common goods reported by respon- it permits exploration of topics in depth. dents over the past 10 years. Because all discussions are open-ended, one is not limited to pre-designed questions or Through a process of content analysis, some of answers. the data were converted to frequencies and percentages. In addition, a limited amount of Most data collected by participatory methods quantitative data from a pre-coded household are inherently qualitative. When a family says questionnaire was used. that 10 years ago they had a sizeable crop in the Participation Series 11 2. What is Poverty? Who Arel the Poor? "Poverty lives here. Ten years ago zve were richi, you could see it in the local markets which wereffunctioning. Today there is nothing to sell, there are no markets anymore." --A villager in Kitui "Fifteen years ago we had aninmals, a granary and land. Today the animals are gone, the granaries are enmpty, the land is subdivided and our clothes are in tatiers." - A villager in Bomet Poverty was widespread in the seven districts of animals, send their children to high cost studied. As defined by the study's participants schools, can afford hospital bills and have -- poverty had increased and was pervasive; expensive possessions such as household estimates of poverty in individual villages utensils and goods, cars and televisions. ranged from 33% to 91%. In Kitui district, the "very poor" and "poor" totalled 76%; in Busia, * Average: own about 10 hectares of land, 68%; Kwale, 62%; Bomet, 64%; Nyamera, 54%; have a good semi-permanent house, send and Kisumu, 57%. Poverty was highest in their children to school, can meet hospital Mandera, where more than 85% of the popula- bills, buy clothes from shops, are clean and tion of the villages was classified as very poor are employed by the private or public or poor. Thus the following profiles and sector. descriptions of life for the poor of the seven districts apply to 50% or more of the total * Poor: have a very small amount of land population in the study villages. It is interest- (not more than two hectares), do not have ing to compare the pattern of results obtained enough food (two meals a day), a bit clean, through the participatory wealth ranking work through contracts or odd jobs, have a methodology used in the PPA with the results large family and have difficulty paying of conventional expenditure surveys (Box 3). hospital bills and school fees. Perceptions of Poverty * Very poor: have inadequate food and shelter, no land, no livestock and tattered Respondents had no difficulty in describing in clothing; they may abandon their children, graphic detail the characteristics of different are destitute, often beg and may steal. They levels of wealth, and in making distinctions may go for d.ays with one meal a day and between the very poor and poor in their com- sometimes just starve. Sometimes "they munities. Generally, their perceptions fell along move like dogs from place to place" - a the following lines: practice known as kununa in Kikamba, when the poor go to different homes to tell * Rich: well dressed, clean, own large farms long stories while waiting for food to be (of more than 20 hectares), have large herds served. 12 Environment Department Papers What is Poverty? Who are the Poor? Box 3: Comparing the Absolute Poverty Line with People's Own Definitions of Poverty Throughi Mapping and Wealth Ranking There are many ways of measuring poverty. The most common and accepted method is through establish- ing an absolute or relative poverty line. Poverty lines are usually based on estimates of income required for the minimum caloric intakes needed for survival adjusted for gender, age and intensity of physical labor performed. The Kenya Welfare Monitoring Survey (WMS) 1992 data based on 12,000 households establishes the national absolute poverty line and disaggregates the data by region and district. Another approach is that used in this study - e Participatory Poverty Assessment (PPA) in which the wealth ranking method was used. The sampling framework for the PPA was based on the national cluster sampling used for the WMS. The Central Bureau of Statistics (CBS) assisted in the random selection of five clusters, which often coincided with village boulndaries, from five of the poorer districts. This systematic procedure was followed for five districts but nc t for Mandera and Kisumu which were added to the study at a later stage. The method for establishing the levels of poverty in the PPA villages involved a four-step process carried out in large groups which met in public. In the first stage, a map of the village was created on the ground, with sticks and stones used to mark the key features of the village and households. The second stage involved a discussion - which became quite heated - centered on the characteristics of people who are very poor, poor, average and rich. In the third stage, once consensus was reached, people were asked to categorize each household in the community as rich, averetge, poor or very poor. (This process is commonly known as wealth ranking.) In the final stage, the female-headed households in the village were identified. The entire four-step process sometimes took several hours, and created great excitement and involvement. Subse- quently, the maps were copied on poster papers for later use as a poverty monitoring tool. The comparison of the results from the conventional absolute poverty profile approach and the participatory mapping and wealth ranking approaches are reported below. It is important to note that for three of the five districts in which the cluster sampling approach was followed closely, the results from the two approaches are almost identical. The two districts where the PPA method gives much higher levels of poverty are from very arid districts, Kitui and Kwale, both severely affected by the drought in the last two years. Thus it is possible that the levels of poverty in these two districts severely affected by drought has increased. The WMS did not include the semi-arid areas like Mandera in the north. Absolute Poverty Line,* Participatory Poverty Assessment District Adult Equivalence (Social mapping, wealth ranking) 1. Bomet 65% 64% 2. Busia 68% 68% 3. Nyamira 54% 54% 4. Kisumu** 39% 57% 5. Kitui 58% 76% 6.Kwale 50% 62% 7. Mandera NA 85-90% * Kenya: Poverty Profiles, 1982-92, by John Thengusi Mukui, December 1993. ** Cluster sampling procedure not followed precisely. In general, village people ascribed "positive" * Positive characteristics: The poor are very and "negative" characteristics to the poor as hardworking, especially the women, to cope follows: with having nothing and to just survive; Participation Series 13 Learning from the Poor they are very religious, they need God and pray a lot; they are caring, help each other Box 4: The Boy Who Tried to Commit with material and moral support; they are Suicide for Lack of Shoes honest and cannot afford to be dishonest; they are receptive to new farming ideas; The research team working in the urban Nairobi they think and worry a lot about how to slums found a handsome, very serious looking make ends meet; they have pride (as one 10-year old boy being counselled by a church said, "We may be poor in material things worker. "He tried to commit suicide this said, Wremay be poor i material thngs morning," the c-hurch worker said. He wanted but we are rich in the eyes of God.") to go to church, but was too poor to own a pair of shoes. He was ashamed to go to church Negative characteristics: The poor are a barefooted, so he tried to commit suicide. burden to the community - they are lazy, "That's poverty," the church worker stated, idle, beggars, criminals; they are drunkards matter-of-factly. and alcoholics who drink to forget their problems; they are poor planners and make poor use of land; their children are mal- while women placed fewer people in these nourished; they beat their wives and engage categories. During PPA interviews and other in prostitution; they have no power to research activities, these differences were change anything. publicly negotiated until consensus was reached to the satisfaction of all. Perceptions in One Village The perceptions of the poor by people in Children's Perceptions Nyaigesa village, Nyamira, were fairly repre- Children's perceptions of the poor mirrored sentative of those found in other villages and those of adults, but with the addition of more districts: detail. As one sixth grader noted, "The poor live under the trees and have leaking houses * Poor people have large families; rich people and stand or move all night to escape the have large farms. rain." But children also turned out to be the .The poor are religious; they are reigious harshest judges of the poor (even more than because only Godcanandleare the district level service providers), problemse perhaps because they are terrified by how close they see themselves living to the edge of * The majority of the poor are alcoholics who poverty, and how they are directly affected by drink to temporarily blur reality. poverty and by the foibles of their poor * In general, poor people are not lazy, they parents. In Nyamira, Kiongongi village, are hardworking. school children described the poor as follows: "the poor are lazy, steal food from barns, steal quorre homeigs are usble,acting. by cows and goats, lack respect, use bad words, quarrels, fights and wife beating, are drunkards, stupid, use money badly, buy * Some of the very poor people in the village silly things, have large families and are those are the old. They have no energy left to who beat their wives." take care of themselves and they have no external assistance. In Mandera district, children showed a clear understanding of poverty. They viewed a poor An interesting and important gender difference person as one who has no animals, no shamba emerged in people's perceptions of the poor. (garden), no job, no source of food. They also Men generally classified larger numbers of stated that poverty is brought about by epi- people in the poor or very poor categories, demics and other tragic incidents that make 14 Environment Department Papers What is Poverty? Who are the Poor? children orphans. In Neboi village on the the poor are lazy, idle, drunkards, criminals, outskirts of Mandera town, the community prostitutes, poor thinkers and uncaring about identified the very poor as able-bodied people the future. Village people by contrast de- without adequate resources such as livestock, scribed the poor as religious and farming land, food and shelter. They were also hardworking, deeply worried about their seen as a destitute group of beggars, the security and future. While a few villagers disabled, the aged, the blind and the sick. viewed the poor as lazy, the majority said a Hawkers are also considered poor since their poor person who was lazy was unlikely to business is highly seasonal and collapses in the survive and, in fact, had to combine multiple dry season when the village is vacated due to survival activities "to barely keep afloat." lack of water. The separation between leaders and the poor How District Leaders Perceive the Poor was viewed by some respondents as having a In many districts, the study teams spent a day geographical component. District leaders were conducting key informant interviews with based in the larger towns and tended not to go local officials. Sharp differences were noted to villages on bad roads or those with danger- between the district leaders' understanding of ous roads. When they did go to these places, the intensity and spread of poverty in their said one of the village leaders, they went in areas and the perceptions and experiences of groups on special purpose tours, passed village people and the poor. This gap in through quickly and did not pause long enough perception was seen by the poor as an impor- to talk to the poor or hear about their problems. tant factor in blocking programs which might Numerous people made the point that despite otherwise be helping them. While interviews the poor roads, district leaders should be with village people showed a high degree of coming to their villages to see the problems awareness about poverty and who the poor directly and talk things over. are, district leaders tended to give textbook descriptions of poverty, and the poor were Major Problems Perceived by the Poor viewed as an amorphous group. oThe poor were asked to list the problems they Some interesting contrasts emerged between faced (Table 1). More than three-quarters district leaders' perceptions of poverty, and the reported that their most important challenge perceptions given by local people themselves. was either lack of food (46%) or lack of money District officials linked poverty with illiteracy, (30%). Problems associated with illness (7%) whereas village people said the poor were and access to education (6%) were the only other often educated. The negative stereotypes most challenges cited in each district. At the district commonly used by district officers were that level, the problem of food was most serious in Table 1: Major Challenges Facing the Poor (%) Category Food Money Education Water Sickness Land Other Kwale 53 27 1 8 11 - - Kitui 33 37 5 11 9 - 5 Mandera 55 18 4 19 1 - 3 Nyamira 57 17 1 1 11 5 8 Kisumu 61 23 4 - - - 12 Bomet 26 39 4 - 7 11 13 Busia 37 35 9 - 8 - 11 National 46 30 1 6 7 2 8 Participation Series 15 Learning from the Poor Table 2: How the Poor Perceive rich and others reportedly bewitching the their Future ("/o) poor so that they would stay poor. Wil.be Better Worse In addition to the above, several long-range factors emerged in the study, each contributing Kwale 25 75 to increased poverty broadly across all areas of Kitui 4 96 Kenya. These included inflation, cost-sharing Mandera 8 92 strategies, demographic factors and social Nyamira 8 92 breakdown. Kisumu 3 97 Bomet 29 71 Busia 16 84 Inflaton Total 13 87 In all districts, people linked growing poverty with increasing prices, and compared current prices to those of 10 and two years ago. Accord- Kisumu (61%), Nyamira (57%), Mandera (55%) ing to respondents, prices have been rising and Kwale (53%). substantially on a continuous basis over the past decade, and dramatically so in the past two How the Poor Perceive Their Future years. During the period 1984-1992, for ex- ample, prices of sugar and cooking oil, two basic In every village, households that were identi- commodities, increased 20-25% per year on fied as poor by the community were inter- average in all districts, according to the "Trend viewed in privacy. The last question dealt with Analysis" conducted in the national study. In their vision of the future. Overall, 85% of the poor felt that their situation would get even Box 5: Why Someone Becomes Poor, worse in the future. The percentages were as Ndetani, Kitui high as 97% in some villages (Table 2). The following reasons for poverty were given How Does Someone Become Poor? by community groups in Ndetani: 1. Unreliable rainfall leads to crop failure. The process of impoverishment was understood 2. Being widowed leaves a woman with a by villagers as having two dimensions. First, heavy burden. almost everywhere poverty was seen as an 3. Many children, because they use up all the intergenerational phenomenon: "Poverty is resources. inherited. If you are born to a poor father, he 4. Not belonging to a group because one is too cannot educate you and cannot give you any poor just makes you poorer, because as an i land or very little land of poor quality; every individual you can never do what the group generation gets poorer." can do. 5. Giving birth in old age, when one is too But becoming poor was also understood as a tired to work process that can happen to anyone over time. 6. People are poor because of landlessness. Primary factors cited included land subdivi- 7. People cannot find jobs. sion, soil erosion, low productivity, rising 8. People are poor because the government prices, reduced government services and does not provide agricultural tools. having many children. Sickness, physical 9. People are poor because their produce is disability, old age and being orphaned were bought at very low prices, while the price of other routes through which families or indi- things they buy is hiked. viduals plunged into poverty. Witchcraft also 10. People are poor because they have neither emerged as a factor (in Kitui and Kwale), the spouses nor children. 16 Environment Department Papers What is Poverty? Who are the Poor? the past two years, these increases have aver- Cost-Sharing aged 100-150% per year, indicating a significant Cost-sharing for education and health services, acceleration in the trend (Tables 3 and 4). introduced over the past 10 years, is credited with slow but visible improvements in Kenya's The poor coped with increased prices by eaLting economy. At the same time, these programs fewer meals per day, buying lower quality are having a negative effect on the poor and the foods and increasing their reliance on relief' ve poor food distribution. Another coping mechanism very poor. was greatly increased school dropout rates In most districts people stated that the cost- which people say have risen two to four tirnes sharing program in education, introduced in in the past two years. The general trend was Kenya eight years ago, is resulting in the col- exemplified in Busia district where they said, lapse of primary educational facilities -directly "khusaba lala", meaning they washed their with respect to the physical plant and indirectly hands just once a day - after their only meal. with respect to teachers -and, by extension, to Table 3: Trend Analysis as Perceived by Poor People During the PPA, community groups were asked what changes they had seen in the last 10 years and in the last two years. People spoke at length about the deterioration of the overall situation, and in which areas of life it was worse than others. People were able to list prices of commodities that were important to them, using stones to represent the proportional increase or decrease. (Price data was gathered for 1982, 1992 and 1994 to keep the data parallel to the statistical sets on poverty kept by the Kenyan government.) Following is an example of reported changes in prices and how they have affected people's lives. Each zero (o) reflects one stone; thus, villagers reported that sugar prices were three times higher in 1994 than they were in 1992. Changes Between 1992-1994 Parameters/item 1992 1994 Percent Change 1. Sugar prices o ooo 200% 2. Salt o ooooo 400% 3. Cooking fat o oooo 300% 4. Roofing sheets o ooooo 400% 5. Mattress galvanized sheets o 00 100% 6. Employment oc 0 -50% 7. School drop outs o ooooooo 600% 8. Proportion of the poor o oo 100% in the community Changes Between 1984-1994 Item 1984 1994 Percent Change 1. Livestock per family oooooo 0 -83% 2. Water availability oooooo 0 -83% 3. Cost of education oooo 000000 50% 4. Tree cover OOOO 00 -50% 5. Food availability oo o00 50% Participation Series 17 Learning from the Poor Changing Demographic Characteristics Table 4: Changes in Prices Reported Thenlink larami toapovertytwas in All Districts The link of large families to poverty was apparent to respondents in all districts who Prices of basic commodities have more than mentioned the fragmentation of land among doubled in the last two years, as has the sons, smaller plots and lower yields as causes proportion of the poor. People reported that of poverty. In urban Nairobi, many female school dropouts have increased seven fold in heads of households, when asked what they the last two years because of higher fees would change in their lives if they were able, demanded. said they would not have had children. Actual Price Changes (Ksh) A resident in Nyaigesa village, Nyamira, said, Item 1984 1992 1994 "Poor people have large families. Rich people have large farms." In Bomet, a villager said, Fertilizer (50 kg) 450 600 1300 "The poor can be identified by the many Maize Seeds 70 120 550 children they have," and added, "Many chil- Milling/debt 7 25 40 dren implies many mouths to feed, hence food Sugar (kg) 7 17 40 shortage, lack of proper clothing and even lack Bar of Soap 5 9 60 of proper health."' In Emuramia village, Busia, Fees (secondary school) 4000 12000 18000 a community member said, "A poor man looks School uniform 40 70 400 weak and has a big family [an average of seven children]; daughters from such families are education itself. Under the cost-sharing plan, prone to early marriages and pregnancies, and the government supplies teachers and the usually leave their children with the old poor parents must maintain the physical plant, grandparents." including buildings and playgrounds, black- boards and chalk, desks and chairs, school Social Breakdo,wn books and other teaching aids. In many schools For adult women, poverty was linked primarily visited, the parents had simply not been able to to widowhood and to a lesser extent to aban- do this. donment by the husband, wife beating and divorce. This was reported both by men and These comments apply mainly to primary schools. women, and was obvious even to school As noted elsewhere in this report, secondary children. On average, female-headed house- schools have to all intents and purposes priced holds (in which no husband is present) were themselves out of the market for the poor and the rated as poorer than male-headed households very poor, and are no longer an option for most of in every village a:nd district. Women had "use these severely disadvantaged families. and harvest rights but no land rights;" because they had no ownership of land, they could not With respect to health care, the cost-sharing sell it if they became widowed or divorced. program has had a different impact. The Ksh 10 (For more information on social breakdown, fee charged to register at the health facility, and poverty and female-headed households see the additional Ksh 10 fee charged to get prescrip- pages 21-28.) tions for drugs, are proving too costly for the poorer patients. These charges are in addition to Livelihoods arid the Poor other costs required for a health center visit, such as patient transport (often over long distances) Income Generating Activities and in several districts, the requirement that patients supply the paper on which their drug In order to minimize risk and survive, the poor prescriptions are written. (See Chapter 2 for diversified their income. The nature of their more on the impact on the poor of cost-sharing.) activities varied by region depending on the 18 Environment Department Papers What is Poverty? Who are the Poor? opportunities available. Chewerta village in the tions. Well-off families also provided employ- district of Bomet provided an example of the ment for the children of the poor, who were often range of income generating activities of the poor: sent by parents to serve as maids or farm hands. Monthly payments were paid directly to the * Small-scale subsistence farming: those parent and used to support the family. Brewing who owned or had access to land grew and selling illicit liquor and engaging in informal tomatoes, maize and potatoes for local trading were other methods of earning income. consumption. They used traditional nieth- The poor sometimes sold a small part of their ods which resulted in small yields. Some agricultural produce, kerosene, salt or fish. Those also may raised livestock. who owned bicycles, known locally as "boarder- * Peddling/hawking: women sold foods such boarder," ferried goods and passengers for a fee. as milk, eggs and vegetables in local mar- kets. In Elugulu, the poor also exploited the natural resources in the region to earn income. The * Manual labor: the poor picked tea, weeded men "dressed" stones for ballast and made land and fenced property owned by rich charcoal, while the women made baskets, farmers nearby. pottery and mats. During difficult times, some * Illicit brewing: the poor brewed and sold of the poor also resorted to stealing from shops busaa to augment their incomes. or farms in order to get food. * Jua kali (welding) and cobbling: in the local markets, the poor repaired shoes, farm People engaged in subsistence farmig and grew tools and household items such as radios. a patch of vegetables such as tomatoes and potatoes; a few grew cash crops such as cotton in * Small-scale business: these included Busia and tea in Bomet. Many had sold their operating a small kiosk or selling mittumba land at throw-away prices (Kwale and Nyamira) (second-hand clothing). and had become squatters on their own land. In Kisumu, it was reported that a majority of the In addition to these activities, community poor had sold their land to rich developers. members stated that the poor sometimes relied on stealing and selling farm products, begging Considering all the study districts together, the from the well-off and borrowing essentials from main reported sources of income subsistance friends and relatives. were farming (60%) and odd jobs and remit- tances from relatives (25%). The least important A similar diversity in types of livelihood activi- source was poultry farming (less than 1%). No ties among the poor is illustrated by the village rural respondent reported formal employment of Elugula in Busia district. The poor were or begging as a source of income (Table 5). most commonly engaged in subsistence farming and working on other people's land. In Elugulu Odd jobs included work in the informal sector village, about 73% of the poor practiced farming (lua kali), bicycle and radio mechanics, tailoring as a source of income. They grew a variety of and basketmaking; casual work (contracts) on food crops, including cassava, millet, sorghum, people's farms in exchange for food; prostitu- groundnuts and indigenous vegetables, and tion; hawking and petty trade of cooked foods, raised cattle, goats and poultry. fruit, paraffin, match boxes; cross-border trade (Busia); and illicit beer brewing. The poor who engaged in gariagaria (manual labor/odd jobs) did so for money or food. Some Land Ownership and Access Rights poor people vacated their homes and became Land ownership is of central importance to most squatters on land of the well-off and worked as Kenyans, especially in the rural areas. In every laborers in exchange for food and accommoda- district studied (with the exception of Bomet) Participation Series 19 Learning from the Poor Table 5: Main Sources of Income By District (%) Categorp Farmin Livestock Poulty Trad;Yg Odd Tobs Remittances Kwale 72 1 - 12 15 Kitui 76 - - 9 15 Mandera 37 10 1 25; 27 Nyamira 71 5 - 1C 14 Kisumu 40 - - 15 45 Bomet 49 3 3 4 41 Busia 79 - - 7 14 Total 60 2 1 12 25 and coffee estates, landlessness among the poor Table 6: Land Ownership by District (%) was high, with 49% reporting owning no land. This finding partially explains the puzzle of Categony Yes No Bomet, which appeared lush and rich to the eye Kwale 76 24 but still emerged high in overall poverty. In the Kitui 85 15 nomadic Mandera district, the majority do not Mandera 49 51 own land. Nyamira 91 9 Among those who owned land, 57% of the poor Kisumu 92 8 owned fewer than 2 hectares, followed by those Bomet 51 49 who owned 2-8 hectares (33%). Just 10% of the Busia 87 13 study participanlts reported owning land Total 75 25 exceeding 14 hectares. In Nyamira, 98% owned 2 or fewer hectares, while just 2% owned as the reported level of land ownership was high, many as 8 hectares; not a single person in even among the poor (Table 6). The majority in Nyamira owned more than 8 hectares. The Kwale, Kitui, Nyamira, Busia, and Kisumu relationship of poverty to the amount of land (76%-92%), reported owning some land. In owned was evident to all. "The small pieces of Bomet district, a large part of which is in the land (0.5 Ha) are unable to occupy the poor high agricultural potential zone consisting of tea people throughout the year," one said. Table 7: Access to Land By District (°/0) Categonr Cultivation/Harvesting Rights Sellinig Rights All Rights Others Kwale 13 2 83 2 Kitui 27 6 66 1 Mandera 33 8 54 5 Nyamira 93 7 - Kisumu 30 4 63 3 Bomet 76 - 22 2 Busia 39 3 58 - Total 43 3 53 1 20 Environment Department Papers What is Poverty? Who are the Poor? With regard to access and use rights to larnd, preference for land (79%), while just 10% in 53% of the people enjoyed all rights, followed Mandera saw land as the most valuable asset; by 43% who had only cultivation and harvest- 32% of the poor in Bomet regarded agricultural ing rights. There are some significant differ- products as most valuable, while not a single ences in degree of land rights between dis- person interviewed in Mandera or Kisumu tricts: in Nyamira, for example, just 7% had all identified such products as their most valuable rights to land, while 83% had such rights in possessions. Kwale (Table 7). Access to Credit Value Placed in Assets The poor had very limited access to credit from Land was regarded by 44% of the people a1s the formal institutions (Table 9). The majority of the single most valuable asset. This was followed poor (70%) observed that they had no access to by children (25%) and agricultural products credit at all. Small loans were obtained from (12%). Fewer than 2% regarded a spouse as family members, self-help groups, women's their most valuable asset (Table 8). groups and occasionally from church groups. Many school Parent-Teacher Associations Again, inter-district variations were consider- allowed the poor to delay payment of school fees able: the highest preference for children was or pay their fees in installment. They coped with recorded in Mandera (80%), the lowest in Kwale payment of hospital bills in a variety of ways, (7%); residents of Kisumu exhibited a high including leaving their national identity card in Table 8: Perceptions of the Most Valuable Asset in the Household by District (%) Agricultural Wifel Livestockl Household Category Land Products Children Husband Poultri Goods Others Kwale 59 16 7 - 8 4 7 Kitui 44 7 39 3 4 3 1 Mandera 10 - 80 5 1 1 3 Nyamira 17 14 8 - 49 11 2 Kisumu 79 - 12 4 - 5 1 Bomet 23 32 23 - 12 3 8 Busia 76 12 8 - 1 - 3 Total 44 12 25 2 10 4 4 Table 9: Access to Credit Facilities by District (%) Class Banks & Financial Inst. Cooperatives Friends/Relatives Others No Access Kwale 7 1 5 5 81 Kitui 3 1 23 35 39 Mandera - - 4 8 88 Nyamira - 5 - 12 83 Kisumu 3 4 7 16 71 Bomet 7 3 21 20 49 Busia 7 7 1 3 83 Total 4 3 9 14 70 Participation Series 21 Learning from the Poor hospitals together with the patient while they subdivide land if they had a limited amount to looked for work to raise the money to pay for pass on to their children. Overall, 61% said that treatment and to retrieve their card. they would give all of their land to their sons and another 8% said they would give the Overall, only 4% of the poor had access to credit majority land to their sons. There were strong through banks and another 3% through coopera- district variatio.ns: 95% in Nyamira said they tives. It is clear that most of the poor are not being would give all their land to their sons while in reached through various micro-credit schemes, Bomet only 12% said they would do so. Only either individually or in groups. 28% said they would divide the land equally While credit is available in theory from the Agricul- among their sonts and daughters. Only 1% tural Finance Corporation (AFC), its policies in fact (primarily in Kwale) would favor girls (Table exclude the poor. For example, while AFC clients 11). are required to own 5 hectares, the majority of the poor own less than 2 hectares. Second, when a man abandoned his wife or a woman walked out on her husband with or Poverty and Female-Headed Households without a divorce, the land stayed with the man. If the husband died, the woman had a stronger Definitions and forms of female-headed house- claim to the lancl, but invariably lost her rights holds are ever changing. For the purposes of either to the parents-in-law or brothers-in-law this study, people defined female-headed who seized the land. The woman might be given households as those in which there was no limited cultivation and harvesting rights. adult man living with the woman or support- Women's limited control of land was an issue ing her from another location. in all districts. In Elugulu village in Busia district, men stated that "when a husband dies At the conclusion of the social mapping and and the woman has children with him, she may wealth ranking exercises which were con- keep all the household assets." The women, ducted publicly in 35 villages, people were asked to mark all of the female-headed house- holds on a map. Table 10 highlights a clear Box 6: Coping Mechanisms Used by pattern. Overall, while 25% of the study Women in Nairobi Slums population was categorized as very poor, there "You just go to bed. You pray to God to give you were over twice as many female-headed somethg to be You day.' households (44%) as male-headed households (21%) in this group. While 59 % of the male- "If I'm left in this house, I guess I'll continue headed households were categorized poor or staying. I can't kill myself. What will happen to very poor, this was true for 80% of the female the children?" headed households. The pattern of greater poverty among female-headed households was "I have nothing at all. When I gave birth to true for every district and for all 35 villages. In twins, I almost died in the house. I was taken to Kitui and Busia, more than 95% of the female- Redeemed Church to be fed. I didn't have headed households were classified as poor or money for kerosene, water, food. I stay hungry." very poor. By contrast, the findings from the "There was a time when I had nothing, com- WMS did not reveal these differences. pletely. So I stopped some white men in a car and they gave m,e 200 shillings." Ownership and Rights to Land and Other Property "The house has one room which allows only one bed. I have no furniture. The utensils are made Women lost out on land twice. First, girls were from collected tins and plastics (cooking oil generally discriminated against in land inherit- containers)." ance. Poor people were asked how they would 22 Environment Department Papers What is Poverty? Who are the Poor? Table 10: Male and Female-Headed Households, by Poverty Ranking Male Head Female Head Total Entire Sample Very Poor 594 21% 293 44% 887 25% Poor 1,083 38% 235 35% 1,318 37% Medium 899 31% 118 18% 1,017 29% Rich 283 10% 18 3% 301 9% Total 2,859 100% 664 100% 3,523 100% Bomet Very Poor 125 21% 48 36% 173 23% Poor 231 38% 60 45% 291 40% Medium 184 30% 23 17% 207 28% Rich 66 11% 2 2% 68 9% Total 606 100% 133 100% 739 100% Busia Very Poor 22 6% 6 17% 28 7% Poor 222 59% 27 77% 249 61% Medium 100 27% 2 6% 102 25% Rich 29 8% - - 29 7% Total 373 100% 35 100% 408 100% Nyamira Very Poor 19 21% 17 36% 36 24% Poor 100 38% 27 45% 127 39% Medium 126 30% 16 19% 142 28% Rich 26 11% - - 26 9% Total 271 100% 60 100% 331 100% Kisumu Very Poor 92 15% 87 33% 179 21% Poor 218 36% 100 38% 318 36% Medium 245 40% 67 26% 312 36% Rich 52 9% 7 3% 59 7% Total 607 100% 261 100% 868 100% Kitui Very Poor 123 29% 61 78% 184 37% Poor 183 44% 14 18% 197 40% Medium 82 20% 3 4% 85 17% Rich 30 7% - - 30 6% Total 418 100% 78 100% 496 100% Kwale Very Poor 213 36% 74 77% 287 42% Poor 129 22% 7 7% 136 20% Medium 162 28% 7 7% 169 25% Rich 80 14% 9 9% 89 13% Total 584 100% 97 100% 681 100% Participation Series 23 Learning from the Poor Table 11: Gender Discrimination in Land Inheritance (%) Son Equally Son iDaughter Category Inherits Divided Gets More Inherits All Others Kwale 59 30 5 5 1 Kitui 28 48 23 - 1 Nyamira 95 2 2 - 1 Kisumu 91 1 1 3 4 Bomet 12 75 12 - 1 Busia 81 11 5 - 3 Total 61 28 8 1 2 however, told a different story. They noted Even where women actually "own" the land, that when a husband dies, the "brothers-in-law complete control may be absent. In Ikapolok .... take all the valuable assets, leaving the village in Busia, for example, women had cultiva- widow with barely enough to give her a new tion rights to the land but could not use or deal start." In Ombo, in Kisumu district, men and with the land in any other way. Selling the land youth both stated that land, cattle, the fly- for income would therefore be impossible. In whisk and the children belonged to the man Kitui district, people pointed out that a woman and his family in life and in death. This, lacked control over her land if her father-in-law despite the fact that all agreed that women remained alive. In this case, they said, "she has should receive everything after a husband's no right to make decisions about the property." death except fly-whisks, traditional stools and Similar trends were seen throughout other a few other possessions most associated with districts. In Chewerta village in Bomet, for men. example, women had "access" to all household Box 7: Gender Analysis Tool Gender differences in ownership, control and access to resources is a culturally sensitive and emotionally laden topic. Hence a gender analysis visual tool - consisting of three large pictures of a man, a woman and a couple, along with 15 smaller cards depicting assets commonly held by households - was used to approach the subject. The activity is conducted by first placing the three large pictures on the ground. The smaller cards are then placed before the participants, who are asked to assign each asset to the person most likely to own it - does it belong to the man, the woman or the couple? Once this task is completed, partici- pants are asked a set of questions, such as "Which assets can the woman use even though she may not own them? What happens to the assets in case of death of the husband and if there is a separation or divorce?" The pattern of response that emerged was very consistent across districts: most assets are jointly owned while a couple remains married and the husband is alive; with divorce, most assets move to the man's column; the same is true if the man dies, although the woman may retain more assets depending on the kindness of the parents-in-law. More likely than not, with some time lag, assets move out of the wife's column and into the man's column and thus revert to his family. Research teams reported that this data collection tool - along with other visual research tools - was the most effective available means for generating involvement and eliciting discussion on sensitive issues. So much so, in fact, that in some districts they detracted from research techniques using key informant and household interviews. 24 Environment Department Papers What is Poverty? Who are the Poor? Box 8: A Familiar Story, Urban Nairobi A fuller perspective on the situation of women without husbands in Kenya today is provided by the life histories of women in urban Nairobi. Many of these women shifted from tlhe village to the big city by one means or another after their marriages had broken up. Their stories often follow a familiar pattern: The husband and wife are living in poverty in the village. The husband starts to become impatient with his wife. He may be annoyed about the food he does not get. More likely he misses the position of authority he used to have in the traditional homestead, when he had his own house and there was plenty of room for everybody. With no job, the husband starts drinking, sormetimes in the early morning. He starts looking at younger women. When he is drunk, wife-beating begins. In time, he may drive the woman from the house along with her children and start getting children by a younger wife. The woman and her children thus driven away are homeless and property-less. She may find another husband and start having more children. If this, too, is a household in poverty, this marriage may also break up. And so may other marriages. The end of the road is a female-headed household, almost always a household in poverty. items except bicycles and weapons. Women am her maid, and I also sell the water for and men, however, perceived ownership in her. She pays me 30 shillings a day. But this slightly different terms. Women said that men lady is not good. She does not get satisfied owned everything including women and their with what I do for her. For myself, I have labor. Men argued that because they shared nothing of my own except my children. My the work in the fields it was owned by both. parents separated and we lived with our However, once again, in divorce or separation mother. She had high blood pressure, and the woman loses all except her personal oriia- one day it was raining and when there was ments. The older children remain with the loud thunder and lightning she just died in man, while the baby goes with the woman to bed. From there, we stayed with our return when it is older. mother's sister, who frustrated us so much. Since we all had left school, we went our In Busia, men stated that they owned the farm different ways, and that is when I got (land), women, livestock, furniture, fly-whisk married. I can't even trace any of my and so forth. Women owned only the kitchen siblings, because we all went different utensils and personal trinkets and had access to ways." a variety of the man's items. As elsewhere, while groups agreed that women should receive Another woman stated: all goods in the case of a husband's death, everyone knew that relatives often confiscated "I don't have any house or any land or the goods leaving the widow with next to anything, because I parted company with nothing. my husband and he does not want us. The oldest boy whom I got before marrying the One female head of a household reported: first husband tried to go home the other day, but the men chased him away. There "When my husband died, my in-laws told is the problem of bus fare, otherwise I me to get out. So I came to town and slept could go and visit my Mum. Before, I used on the pavement. This lady who gave me to be much better, but as the children grow this house found me and brought me here. I my condition has become worse. I never Participation Series 25 Learning from the Poor Box 9: Women's Ownership and Loss of Assets in Divorce and Widowhood, Kisumu The visual Gender Analysis tool was used at least three times in each village with separate groups of men, women and youth. This example from Kisumu district is typical. While married and with the husband alive, women had access to every item illustrated on the asset cards except the fly-whisk. According to the women, men owned items such as furniture, livestock, donkeys, agricultural implements, fly-whisks, shambas, clearing lands and ox-carts. The men however claimed ownership to fewer items, listing only the fly-whisk and various livestock as being owned exclusively by them; personal ornaments were owned exclusively by women while the rest belonged to both. Table 12 shows the dramatic change which occurs when a couple is divorced or separated. Here, the true ownership of items becomes apparent. Those items which were "jointly owned" under normal circum- stances suddenly become the sole property of the husband. Except for the house (local culture dictates that a house belongs to the woman), ornaments and young children (until age six), the woman is left with nothing. In addition, the babies go back to the father when they are no longer deperidant on the mother's care for survival. Women state that if they own livestock they can take them should a divorce occur. However, neither men nor youth mentioned this. When a husband dies, everything except the fly-whisk officially belongs to the woman. However, there are once again a number of qualifiers to this rule of ownership. The household furniture and radio belong to the widow only as long as she remains in the home of the man; if she leaves, they revert to the husband's family. The livestock belong to the woman, but she cannot sell them without consulting her deceased husband's parents. intended to have these children, but they approve, so they got him a younger girl. came by bad luck. I wish I had not become From there, I decided not to marry again pregnant before, ever. My first husband because I felt I had bad luck." also gave me bad luck, when he abandoned me. So I married another man within Many of these women find their way to the Korogocho and we stayed for two years. I growing numbers of Nairobi slums, where got one child with him, and he left me dilapidated housing is at least low-cost and a because the family told him that I had chil- number of church groups and NGOs are work- dren from another marriage and they did not ing to assist those in greatest need. This situation, which appears from the study to be worsening, is placing still another form of poverty-related Table 12: Who Owns What in a House- bude ongvrmn neifaece. "W hol in Caeo.iocadSprto burden on goviarrunent and relief agencies. 'We hold inCsfDvren eaaare operating a vast housing relief agency here," Possession Man Woman said the Korogocho chief. Household furniture/radio In Kitui, all village groups pointed out that this Animals (sheep, oxen, situation was related to the customary "bride- donkey, etc.) # price" paid by the man to the woman's family at Farm implements/tools marriage. Because of this payment, all money or Kitchen utensils property acquired by the woman during the Foodstuffs # marriage belongs to the husband. With respect Ornaments # to children, they revert to the father's homestead Fly-whisk in case of separation or divorce; babies stay with Babies and children in caeo eartordioc;bbe tywt Land cl the mother until they have passed the weaning House age, and then they too revert to the husband. With respect to children of the woman by a 26 Environment Department Papers What is Poverty? Who are the Poor? previous marriage or by another man, they are tion is somewhat better, but only somewhat. In driven from the house along with the woman (in Bomet, the wife inherits everything owned by the case of separation or divorce) thereby making the household, but the family of the deceased her position all the more precarious. husband keeps an eye on her to see that she does not squander the wealth. Roughly the Previously, a divorced woman would normally same situation was reported in other districts return to the home of her birth. Today, with including Mandera, the nomadic area, the poverty and overcrowding in the already too rationale being that the sons of the deceased small houses, she may be turned away, espe- husband are the ones who should eventually cially if she is bringing children with her. She inherit the property. may wander around, try to remarry and acquire more children. In many cases, she becomes the The basic fact of life for women, as reported in female head of a new household, usually a all the districts studied, is that sons inherit and household in poverty. daughters don't. Women may have access to land and use land, but they do not own land. A further factor that came out in the life histories This means they have no collateral and cannot obtained in the urban Nairobi interviews was that get credit. This fact weighs particularly heavily the divorced woman's ancestral home in the village on female-headed households. may also have broken up, so that for one reason or another she can't return. If she has children by Coping Strategies various men, the father may not want her. And the Female-headed households utilzed a variety of mother may also have become the impoverished strategies to survive. Although many of the female head of a household and may not want her strategies to s e.altho any ofsthe > . ~~~strate ies used by female-headed households daughter and various children in her alreacdy are agso common to male-headed households, overcrowded dwelling,ascomntmaehddhueol, there were over twice as many female-headed In the csowdwo,households as there were male-headed house- In the case of widowhood, the woman's situa-hodmtevryprcagr. holds in the very poor category. Box 10: Ownership of Goods in Syomunyu, Kitui In the village of Syomunyu in Kitui district, most household items are shared by a couple. In the event of separation or divorce, the woman loses all rights of ownership, except to her jewelry. If she has a small baby at the time of divorce, she is allowed to care for it until it stops breastfeeding when she must return the child to the man. When asked whether they would compensate the woman for having taken care of the baby, the men said they would not. The women stated that in cases of separatlon or divorce they would take with them all items they had bought with their own money. In addition, sometimes a woman may decide to take her children. In these cases, they are often not challenged because children are seen as a woman's only asset after a divorce. The women stated that, "A woman makes a home. When she leaves, it is no longer a home and, if left, the children will suffer." Some women confessed that when preparing to leave, they would take all the money they could lay their hands on, and deny taking it if asked. After all, they said, unlike taking a cow or goat, there would be no evidence that they had taken the money. In the case of death, the women said they might inherit everything or lose everything to their in-laws, depending on their kindness. In addition, if a woman's father-in-law is still living, she has no right to make any decisions about property (such as land and cattle) she may have inherited. Participation Series 27 Leaming from the Poor Box 11: Coping Strategies of Female-Headed Households, Kitui In Kitui, 78% of female-headed households fell into the "very poor" category. How do these women and their children survive? Food is the number one priority. Because it is usually in short supply, women eat one or two meals a day under the best of circumstances. While they may grow small amounts, food is3 often bought. Women engage in a range of casual jobs - Kuveekea - to eam money to buy food. They also work on other people's farms in exchange for food received at the end of the day. During feimine, they rely on relief food provided by the government. Children's labor is often an important asset in a female-headed household's coping strategy. Mothers may give their children up to others to perform work, with payment for the child's labor coming directly to the mother. Because most female-headed households find it difficult to pay the fees for primary education, children are fortunate if they can remain in school for even a few years. Often, they are removed from school at least seasonally so that they can work and provide income for the family. For the most part, however, income is generated by the women themselves. A variety of income-generat- ing strategies are used: making handicrafts, kiondos, at night; engaging in 'petty trade" in bananas, mangoes, cassava, cow peas and millet; selling firewood, charcoal, water, wild fruits, honey and pots; carrying out casual jobs; and, in some cases, resort to prostitution. Gaining access to health care is another challenge facing female-headed households. In some instances, mothers are able to use credit to get treatment for their sick, leaving behind their identity cards as a guarantee for payment. In addition, they may borrow money from relatives, friends, and neighbors, especially to meet emergency transport needs. They are often forced to carry their young children on their backs or in wheelbarrows and stretchers made from sacks and tree branches to the health care facilities. For most illness, women rely on traditional herbs. A woman said "God has provided trees free of charge and one has only to go to the right herb, boil it and take it..." Herbs are seen as the "savior" of the commu- nity. If witchcraft is suspected, women tum to traditional healers. The basic necessities are extremely difficult to obtain. Women rely on second-hand clothes, hides and skins. Shoes are rare, tyre-shoes are tied to the feet; racks made of reeds (itaa) serve as beds; some families sleep under trees or use tree branches to block the doors of their housing structures. These households also face significant problems in fetching water from great distances (5-20 km). Women try and borrow donkeys or carry water on their backs. Some are too poor to afford jerricans end must use gourds to fetch water. Broken gourds are also used for wash basins, and clothes often washed without the cheapest detergent (panga soap). Food was the most immediate and pervasive police. Women often traded their labor for problem in every district. In villages in Busia, food. In Emuramia, community members female heads of households consistently observed that "most of the people who limited the number of meals eaten to cope worked on the farms of the well-to-do in with the ever-present lack of food. One-third exchange for food were women." The of the female-headed households in Elugulu struggles and ingenuity of women to survive village had one meal per day (the remaining on very little in the face of starvation were two-thirds had two). Some women resorted reflected in every district. to begging for food, others relied on brewing When women were forced to remove their changaa and ajono, despite being harassed by children from school to save school fees and 28 Environment Department Papers What is Poverty? Who are the Poor? gain income from the child's labor, they are Investment in girls' health care was also less most apt to remove their girls. In Nyamira, valued than for boys. "The girl will get married, preference for boys' education at the expense of and investment on her health care will be lost," that for girls came out strongly. Of those said one respondent. One young man said interviewed, 83% indicated that they would parents should not spend money on their discontinue the girl's schooling in order to make daughters' health since "her boyfriend is likely it possible for the boy to continue. In Kwale, it to have her treated out of love." was reported that girls would be sent to Mombasa to work as housemaids so that their The Last Word earnings could be used to keep boys in school. Two comments by women interviewed in the In Busia, 61% of the households stated that they Nairobi slums complete this section on would drop the girl first if there was a shortage female-headed households. One woman of school fees. Boys were seen in many cases as said, "If I could be given just a plot of land, the "homeguards" of the family, while girls one acre, I would move there very fast and would eventually marry and move away. In try to start life again. I hate staying in addition, many women were not educated Nairobi." Another woman, when asked what themselves and saw no point in educating girls. event she would change in her past if she had Thus, in female-headed households, girls may the chance said, "I would be born a man." once again bear an additional burden. Participation Series 29 3. Coping Strategies and Mechanisms In order to survive, the poor have adopted a season. Using leaves, stones and symbols to range of strategies to reduce risk of total disaster identify each item, participants then used a and maximize their very limited resources. stick to mark the seasonal differences on the Protecting their children, keeping them alive and ground (Tables 13 and 14). The greatest stress finding money to pay their school fees is the was found to be from December to May, a preoccupying concern of poor parents. Despite period when food stocks, employment opportu- their best efforts, some poor starve and some die. nities and income are at the lowest. People cope In one of the districts, researchers were greeted by begging for food and by eating "lighter by a funeral in every village they visited. meals." During this period men, and, to a much lesser extent, women engage in seasonal migra- To cope with poverty, people buy cheaper goods tion to bigger farms, tea estates or wherever and second hand clothes; they wear patched they can find work. The highest incidence of clothes and "tyre shoes" which last 10 years, and disease, especially malaria and diarrhoea, use laundry soap for bathing. Increasingly, they coincides with the long rainy season, from April are forced to violate social norms: sons now sleep to July. in the same over-crowded shelter (nyamera) and sometimes in the same bed; family members Access to Food rotate among different households and even sleep in the bush. The issue of food security was probed in different ways. Participants were asked about Seasonality food sufficiency under normal circumstances and what happened during periods of crop There were sharp seasonal variations in stress. failure and drought. In addition, information In one research activity, groups identified was obtained through focus group discussions, activities or resources that changed with the key informant interviews and a range of other Box 12: Strategies for Reducing Risk, Urban Nairobi Father Mario, a local Catholic priest, introduced a research team to a group of seven youth (all boys except for one young girl) who are making batik paintings of street scenes in the slums. They work in a rundown room adjacent to the Mukuru garbage dump. All of their work is stark and dramatic. Several scenes are of young boys and girls sniffing "gum" bags - a gasoline-soaked substance that makes one drunk. One shows a determined young boy with a bag of items collected from a mountain of garbage. Another shows a young Maasai girl with flaming red wig, ready for nighttime action, which they call "Maasai Warrior." The group is led by a serious young man named Moses Kabiru, who says they are starting to sell the batiks in Europe and America. Unlike breakable pottery and other items made in developing countries for sale in specialty shops abroad, the batiks are easy to ship overseas. 30 Environment Department Papers Coping Strategies and Mechanisms Table 13: Seasonal Analysis of Poverty, Nyamira* Item Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Light Meals ooo 000 0 0 - - - - - - - 00 Begging ooo ooo o - - - - - - - - °° 000 000 Migration ooo 000 0 - -00 000 000 000 Unemploy- ooo 000 00 0 0 00 00 - - - - - ment 000 000 0 Income - - 0 - - - 0 000 000 000 000 0 000 000 000 Disease - - 0 oo 00 00 000 000 00 0 0 0 oc0 00 00 000 000 00 Rainfall - - 0000 oooo 0000 000 00 0 0 000 000 00 OC0OO * Zeroes (o) in table represent stones used by participants to indicate the degree of change by month. Thus, three zeroes in the January colurm for "Light Meals" means that light meals are three times more likely that month than they are in Marchi or April. Table 14: Seasonal Analysis of Poverty, Geta B Village* Item Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Out 0000 0000 0000 oooo 000 00 - - - 0 - - migration oooo oooo oooo o 0000 0000 Food 0000 0000 0000 0000 0000 0000 - - 0 00 0000 0000 shortage oooo 0000 0000 0000 0000 000 00 00 0000 0000 00 0000 0000 0000 Employ- o - - - - - 0000 0000- ment 00 * Zeroes (o) in table represent stones used by participants to indicate the degree of change by month. Participation Series 31 Learning from the Poor participatory data collection techniques, all of In Kitui, in Syomunyu village, for example, a which uncovered the lack of food and of food person was paid 50 cents to dig a hole 3 feet security. deep; Ksh 1 per week to watch over people's shambas at night and scare away animals; As reported earlier, when asked to name their cutting and transporting local building poles single most important problem, 45% of the poor earned 20 cents; people were paid for uproot- said they lacked food while another 30% said ing tough grass only after the rain, when there they lacked money. There were no communal was evidence that the grass did not regrow. food reserves reported in any of the villages under Perhaps not surprisingly, researchers found normal circumstances or during any periods of that youth dreaded manual labor and piece food stress, crop failure or drought Even under jobs more than anyihing else because it sig- normal circumstances, about 6% of poor house- naled to them the slide into permanent poverty. holds reported experiencing severe hunger. While over 70% classified themselves as farmers, only When food became scarce, the poor adopted 5% were food sufficient. two directly food-related strategies to limit consumption. First, they reduced the number Malnutrition among children was evident in all of meals to one or two per day. This pattern the districts and its incidence had increased was so common that one characterization of a over the years in the perception of local people. poor person was a person who cannot have three The study could not establish the extent of meals a day. Everywhere parents said that they malnutrition among children. While women reduced their own intake first and protected the received nutrition education while waiting at children as long as 1possible. health centers, in the absence of adequate food, they found it difficult to practice suggestions Second, parents switched to "lighter" foods offered by nurses. In Kwale, women laughed such as porridge, a:nd diluted meals with while reciting the names of the essential food water. They drank their tea without sugar, ate groups and said "First we must have enough porridge with magadi soda (munyek), and fed food to eat." their children misikik, a by-product of local brew (busaa). Other strategies used by the Food Quality poor included: dropping into people's homes People were aware of the issue of food quality. during meal times; begging; stealing and The majority reported that the quality of the teaching their children to steal food; and food they ate was poor or very poor (80%). gathering wild foods. Men hunted if possible Just 1% of the poor felt that their food was rich or sold their maize seed, which they then in quality. In Kisumu District, however, none needed to buy back later at higher prices, of the poor interviewed thought that the slipping further into poverty. quality of their food was very poor. This may have to do with their diet, which includes some Hunger and School Drop-Out Rates fish. The links between lack of food, hunger and school dropout rates were made by adults and The Most Common Coping Strategy children everywhere. In Kwale women said The most common strategy adopted by the "When we do not have enough to eat, how can poor - both men and women - to gain access we send our children to school?" While to food during times of acute shortages was parents went to great lengths to keep their performing odd jobs or "piece work" for cash children in school, when food was scarce, or food. They did so for richer neighbors and mothers pulled children out of school to relatives or at nearby plantations. Though such conserve their energy and save on school- jobs paid little, they were still difficult to find. related expenses. Most children walked or ran to their school, which was located as far from 32 Environment Department Papers Coping Strategies and Mechanisms Box 13: In the Words of a School Child, Kwale 'I come to school because I have decided to be a lawyer and when I grow up I want to be a leader. Our school has many problems. There are too few teachers, we have no books, blackboards, no toilet, no water and no desks. Poor children in this school need assistance so that when they pass their examination, they can buy books, clothes and pay school fees. In my family two children dropped out of primary school in the last two years because of no money to buy school uniform or school fees. "I know that my chances of going to secondary school are very low; my father wants me to go, but he has nothing. I come to school on foot, and I do not have any lunch because there is no food. I cannot read very nicely because when I am very hungry and the rnumbers swim in front of my eyes.' Boy, 16 years old, 6th grade their homes as 7 km away. In Chewerta some of the districts do benefit (Table 17). village, Bomet, consisting of 69 households, 39 However, government provided relief food primary school children had dropped out of becomes an added category in the coping school in the last two years. During times of strategies of the poor, rather than the only food scarcity, mothers placed their children strategy. This is because the government with better off family, friends or neighbors. provided relief is both irregular and insufficient (For more on hunger and school drop out in quantity by the the time it reaches house- rates, see Education section below.) holds to meet food needs. Remittances In the districts where food relief was mentioned, The poor are often thought to benefit from the two issues came out consistently -leakage and remittances of relatives. When asked bt political interference. In one village in Kwale, for describe their means of livelihood under example, village people said, 'We saw the lorry normal circumstances, people did not inention of food relief arrive and the chief told us two such transfers at all (Table 15). The domeinant weeks later that one-and-a-half bags had been food-related copalg strategy was normally received for distribution to 116 households." purchase of food and reliance on odd jobs. Similar incidents were repeated in Kitui where During periods of food scarcity, remittances people reported, "Now we just borrow donkeys from children were mentioned by 9% of poor and go to the relief center ourselves to collect households (Table 16). Remittance transfers food, otherwise we would never get anything." may be higher for better-off households. In Busia, people in one area said "We finally received one bag of maize for the entire sub- Government Relief Efforts location, over 200 households. So we decided to give it to the school, so at least our children During periods of food shortage, there was would get a few hot meals." In Bomet, people some reliance on government food relief. This reported receiving 10 kg per household. Target- was particularly true for Mandera (239%) and ing, however, may be a problem, since only 1% Kituti (20%) (Table 16). Although church of the poor reported getting food aid. groups and occasionally NGOs mentioned food assistance as a coping strategy, the Access to Water number of poor people who were reac]hed through such programs was found to be Gathering information on whether the poor minimal. have access to and consistent use of a safe Once famine has beenoficialwater supply for drinking, cooking and bathing Once famine has been officially recognized and is difficult because definitional problems and the government relief efforts begin, the poor in Participation Series 33 Learning from the Poor Table 15: How Poor People Cope Under Normal Circumstances of Food Availability (%) Rely on Beg From Seek Alter- Have Suf- Other Category Purchases Neighbors native Foods ficient Food Starve Odd lobs Kwale 65 5 1 20 - 9 Kitui 75 5 1 1 1 17 Mandera 60 11 7 - 13 9 Nyamira 25 2 - 2 - 71 Kisumu 39 15 15 - 20 11 Bomet 53 7 4 12 4 20 Busia 51 8 20 1 - 20 Total 52 8 7 5 5 23 Table 16: How Poor People Cope During Periods of Food Scarcity (%) Depend Asst. from Remittance Do Use Prev. Pur- Category on Govt. Rel. Org. fr Children Odd Tobs Gifts Stocks chases Others Kwale 1 - 12 60 1 4 5 17 Kitui 20 7 11 33 - - 14 15 Mandera 23 8 3 37 1 5 16 7 Nyamira - 2 11 71 - 3 4 9 Kisumu 1 - 15 44 7 3 19 11 Bomet 1 - 1 8 - 6 73 11 Busia - - 9 59 - 7 14 11 Total 7 2 9 45 1 4 21 11 Table 17: How Poor People Cope Following Famine Conditions (%) Depend Asst. from Remittance Do Use Prey. Pur- Category on Govt Rel. Org. fr Children Odd lobs Gifts Stocks chases Others Kwale 7 - 7 55 4 4 - 23 Kitui 64 11 12 5 2 - 5 1 Mandera 45 8 12 27 3 1 - 4 Nyamira - 2 10 70 2 3 2 11 Kisumu 20 12 12 39 11 1 3 3 Bomet 1 1 - 8 - 5 72 11 Busia 52 7 28 1 11 - - 1 Total 27 6 12 30 4 2 11 8 34 En-vironment Department Papers Coping Strategies and Mechanisms nuances abound. However both from the The Water Situation in Kwale WMS data and the PPA study it appears that When interviewed, district authorities in Kwale overal only between 33% to 40% of the poor said that the Swedish International Develop- have "access" to a protected water source. This ment Authority (SIDA) provided water. of course does not address the issues of contin- However, despite the work of SIDA and ued functioning of installed systems, seasonal- KWAHO (an indigenous NGO) all the study ity, reliability of water source or use of water villages except one reported access to safe from these sources. water as a major problem. Water-related diseases were common. In Kinyungu village, Because of the importance of water in the lives Mswambweni division, people reported that of the poor, and the difficulty in capturing the "water is life and because we have no water, issue through questionnaires, the PPA study life is miserable." The village's main sources of addressed the issue through several open- water are water pans which have green water ended methods including mini-case studies. and dry up seasonally, the river Ramisi which There~ ~ ~ ~ ~ ~ ~ ~~~~ ~sasnly thee river Rkmis whichgs There were four key findings: is 10 km away and boreholes located 5 km * In all districts - except Bomet - poor away. Because of lack of water and distances people reported access to safe water as a involved, children accompany parents to the problem; water holes to fetch water rather than go to * Everywhere - including Bomet - the schools. water problem becomes acute during the dry season; In Gora village in Samburu division, people use * Where demand is high and community seasonal pans or walk to the railway station 5 km groups have the capacity (skills, technical away where there is a tap. A few of the richer and business management), local groups household have piped water tapped from the are managing simple water systems effec- main pipe that supplies Mombasa. In Mwapala tively without government support; and village, women rise while it is still dark to walk • Sucessfl watr usr grups banchintoto unprotected springs which have very little e Siuccessful water user groups branch into wrater. Because of the waiting time at the spring, income production activities, some of which fedacitowaersyta woman must work five to six hours per day to whicndhfee backem ior wr semsit. bring home one bucket of water. During the dry season, women walk 8 km to Majimboni to fetch water. The situation is almost identical in equate accessocition baetwe undertaned inad Gulanze village. Women spend seven hours a eqatpeople ac toe waer wastrits u ernedalby day fetching one bucket of water from seasonal way people in the drier districts generallyspigorivs10kawySmeepluea defined the poor. In Kitui, for example, people dpamn ownred by a individua living 4kmpee uase a cited "lack of access to water" as a characteristic in ecne fo scooingil out o tm to of th por Eve in Boet whcmsrihii exchange for scooping sofl out of the dam to of the poor. Even in Bomet, which IS rich m eet tfo sli surface water, rivers and streams, untreated g u water from rivers was the primary source in Bumbani village was the only one sampled three of the five villages surveyed. In these which had benefited from the Government of villages, people reported higher levels of Kenya/SIDA/KWAHO partnership. Two pollutants in water during the rainy season. In boreholes were drilled and are managed by one village, women walked 15 km for water in water committees. The community is now the dry season; in another, they traveled 5 km raising funds to Tehabilitate a piped water up hilly terrain. The situation was far worse in system which was abandoned three years ago the less favored climatic zones, as was cleavr because of lack of fuel for the pump. from an examination of the water situation in the districts of Kwale, Nyamira and Busia. Participation Series 35 Learning from the Poor Box 14: Water Problems in Nyamira Nyamira is not defined as an arid district, yet every village reported serious water problems. In Sengera II village, water is obtained from two unprotected springs which dry up in the summer. Youth listed water as the most serious problem and women said "we would like to have a water source where we do not have to walk long distances uphill to fetch water." In Mosibano village, water comes primarily from a heavily polluted dam or one protected and two unpro- tected springs which are within 1 km, but dry up in the summer. People said that almost everyone in the village had amoebic parasites. Some people fetch water from a piped water system at a nearby tea factory, while some better off household have built rain water tanks. In Kiongongi village, there are two main water sources: the Nansore protected spring and the polluted Endiba River. In the dry season, when the spring is reduced to a trickle, it can take as long as six hours to get water because of long queues. Women bring many containers to fetch water to reduce the frequency of trips, tie babies to their backs to win sympathy to jump the long queues or take water from the polluted river. In Nyaigesa village, the main sources are two springs (one of which was protected with assistance from a Catholic mission) which dry up in the summer. When the springs run dry, people use polluted rivers. In Geta B village, there are no developed water sources. People use two polluted nearby rivers, Nyantuago and Agesagane. When these dry up in the summer, women trek 15 km to Kuija River to fetch water. Some people reported boiling the water before drinking, but said they were puzzled because "boiling the water changed the water color to dark red." The Water Situation in Nyamira "participation" consisted of a thin overlay of Nyamira district is not defined as an arid involvement on essentially an engineering and district yet every village sampled had serious supply driven approach in which communities water access, seasonality and quality prob- were not required to undertake any major lems. As in other villages, water borne responsibilities until construction was complete. diseases were rampant during the rainy Community groups were not required to season as the water sources became even complete substantial organizatonal tasks or more polluted than usual. make financial contributions prior to construc- tion. Consequently, communities felt no Water Committees in Busia ownership nor responsibility for the boreholes. Despite formation and training of water com- Busia district, like Kwale, has received much mittees for operation and maintenance, the donor assistance in drilling boreholes and majority of the water committees did not fitting them with handpumps "through com- mobilize themselves to purchase spare parts in munity participation." And yet, like Nyamira, case of breakdown, especially if the boreholes every village reported water problems. While drilled were nolt centrally located and alterna- Elugulu village had two drilled boreholes tive water sources are nearby. installed by the Finish International Develop- ment Agency, both had been broken down for Water committees were functioning well, on months and not been repaired. the other hand, where agency supply of bore- holes matched community demand. In some Despite instituting some elements of a commu- communities, women's groups had taken over nity participation strategy, the approach was the operations and maintenance responsibilities not based on demand or "felt need." In Busia, and with small loans started microenterprise 36 Environment Department Papers Coping Strategies and Mechanisms activities, proceeds from which were used for Importance of Primary Education pump repairs as well. In Ikapolok village, Because a primary school education is a where the borehole was being managed well, prerequisite for entering secondary school, the members were required to pay Ksh 10 per PPA focused on understanding the dynamics month. However, the poor had dropped out of surrounding primary school attendance. In all the group and were using the two rivers near districts, poor parents placed a very high the village to draw water. premium on keeping their children in school. In order to do so, they would sell their posses- Knowledge of District Leaders sions, beg, steal, brew and sell beer, pray, go to Of all services provided in the districts sur- church, hawk produce, join self-help groups, veyed, leaders seemed most knowledgeable cajole teachers into letting the children stay in about the adequacy or inadequacy of water schools, pay in installments, put their children supplies. In Kitui, where droughts and famines to work and sometimes become destitute trying persist, leaders were able to assess the position to keep their children in school. In some areas, in some detail. "Boreholes don't work well," women's groups were formed to assist poor they said, "because the deep-level water is parents in coping with school-related expenses. saline. Dams would be good, but donors are As a last resort, many parents sold their maize shying away from them. The Athi River, a seed in the hope of putting together enough source used by some villages, is badly polluted. money. Some poor parents reported marrying The new pipeline from the Tana River to their daughters off early to receive a bride Nairobi bypasses the district." price, so as to enable the rest of the children to continue school. In Bomet District, leaders accurately appraised the water supply as good in the fertile upper Despite the best efforts of parents, many children levels comprising most of the district, but a were forced to drop out of school seasonally to problem in the lower levels with less rainfall earn money for their families during the difficult bordering Narok. In Nyamira, the Districi: periods. This phenomenon seemed more Water Office was aware that the few protected widespread in the arid districts. When enough springs in the area were not enough to meet money had been made, the children then re- community needs. The larger sources included turned to school. As a result, there were many unprotected springs and rivers which were older children in the lower grades. polluted. "In this district," the water office said, "the issue of water is quality, not quantity." Decline in Quality of Primary Schools Almost everywhere there was widespread Access to Education decline in the physical quality and basic assets of schools, from classrooms to chalk and In the hope of making their tomorrow better, textbooks. Physical deterioration was ob- the poor are obsessed with getting their chil- served in school buildings, a large number of dren through primary school. For them, despite which were incomplete or collapsing. Govern- evidence to the contrary, education is the way ment of Kenya policy requires communities to out of poverty. The PPA touched upon primary build and manage their school while the education through five different data collection government provides and pays for teachers. methods: problem identification activity; focus While some communities continue to manage group discussion; household interviews; mini- well through Harambee and church efforts case studies; and questionnaires and sentence (see Chapter 3), in the face of escalating costs, completion tasks completed by standard six most schools in poorer communities were in school children. decline, despite the continuing struggle of Participation Series 37 Learning from the Poor people to keep the schools functional. Some community, headmasters and teachers trying to villages have abandoned their local school keep the rural schools running was quite altogether and sent their children to schools in different. The situation was the worst in the other villages. arid districts - Kitui, Kwale and Mandera - where schools were sometimes more than 10 km All primary schools are expected to teach a round-trip away, involving crossing rivers wide range of technical subjects requiring which flooded in. the rainy season. In Mandera, specialized equipment, however few schools once the rains start the schools empty out, as have such equipment. Required subjects for the nomads move their animals to greener which equipment is needed include home pastures - "camnels before education." science, art and design, wood work, metal work, building construction, power/ motor Everywhere school children in the sixth stan- vehicle mechanics, electricity, aviation, and dard reported the many things they found typewriting and office practice. lacking in their schools. A shortage of books was the most frequently cited problem (65%). Additionally, schools were characterized by This was followed by lack of desks and a long absence of books, desks, water, toilets and list of other shortages, including chalk, black- shortage of teachers. In many cases, indifferent boards, classrooms and laboratory equipment. teachers, or teachers demoralized by their When asked what they would change (in circumstances, were also problems. This makes addition to supplying books and desks), stu- the poor's faith in education as the way out of dents often cited the teachers and headmasters poverty a shaky proposition at best. Many who they said caned them, sent them on headmasters reported a decline in academic errands during class time or were so lazy that performance as well. they either came late, did not teach, or simply The perceptions about education among district did not show up for work. While children said decisionmakers were in stark contrast to those they would like better trained teachers, they of the poor themselves. District officials be- also expressed deep appreciation for those lieved that "primary education is freely pro- teachers who they felt cared about them and vided, so there are no problems in attendance their teaching. even among the poor." The perspective of the Box 15: The Struggles of Harambee Primary Schools The Almodi Primary school in Busia district was built in 1975 based comple tely on Harambee. The first community Harambee raised Ksh 10,000 which was used to build one perminent classroom; two more windowless classrooms were added in 1986 and 1987. Standards 1-3 are housed in the permanent block. Conditions are crowded and children sit on the floor. Older children are in thatched classrooms with mud walls that have gaping holes. The headmaster reported that "we put the older children in these classes so that when the walls or roof collapse they can run to save themselves". The school is run by a Parent-Teacher Association and an education board. Parents are required to contribute toward the building fund in installments. While the school tries to be kind to the poor, children are not allowed to take their examinations until contributions are paid up. In addition to lack of desks and chairs, the school lacks textbooks, stationery and laboratory facilities. During the dry period when food is short, the school is half empty as parents keep children at home to conserve their energy or send them out to work. Not surprisingly, the academic performance of the school has suffered: last year it was second to last in the 17 schools in the zone. Both the headmaster and parents felt that with increasingly frequent drought, land erosion and subdivi- sion, poverty in the area had increased. The headmaster ended the interview saying "with a little bit of help from NGOs or others there is still a chance for Almodi to come out and shine one day." 38 Environment Department Papers Coping Strategies and Mechanisms have increased. Increasing poverty and esca- Box 16: Problems Commonly Cited By( lating school costs have resulted in quadru- Primary School Children pling of school drop outs in some areas in Nyamira. The single most frequent reason for . sitting on the floor dropping out of school was lack of money for . being caned by teachers school fees. . cutting grass . crowding in the classrooms School children were asked if anyone from their l lack of water and dirty compound famihles had dropped out of school in the last two . mud-thatched classrooms . being chased from school due to lack of school years. Between 50% and 68% of children in fees various districts reported that at least one child, and often more, had been forced to withdraw from school in the last two years because of Cost-Sharing and Dropout Rates inability to meet school expenses and fees. In As noted in Chapter 1, children of the pooI' Bomet, for example, 70% of these households had drop out of primary schools in large numbers more than one child - sometimes as many as because of their inability to pay school-related seven - who had dropped out of school. For expenses, even though officially there are rLo secondary school, the district drop-out rates for school fees. Although primary schools are the last two years ranged from 18% (Nyamira) to supposedly "free," parents must pay for school 81% (Bomet). In Nyamira, to cut costs, parents uniforms and constantly increasing contribu- sent their children to vocational schools. In Katolo tions. Trying to cope with declining physical sub-location in Kisumu, a 1993 study demon- facilities and increasing costs, Parent-Teacher strated the declining enrollment from Standard 1 Associations and schools have increased levies to Standard 8 across seven primary schools in the on parents. These come in various forms and area. While the enrollment in Standard 1 was 427 include building funds, activity fees, exercise pupils, the numbers declned steadily to 128 notebook fees, extra tutoring fees, and contri- pupils in Standard 8 (K'Okul, November 1993). butions of pens, pencils, chalk, erasers, labora- tory equipment, sewing machines and agricul- Despite the challenges posed by poverty, for the tural equipment. Many schools now levy moment the school children also hang on to their examination fees before exams can be taken dreams - including their hopes of becoming even in lower grades (Ksh 200/child), another professionals, doctors, nurses, engineers, pilots, new cost that parents must bear. teachers and even the country's president. Many children hold on to these aspirations even as they Analysis of household data showed that (de- acknowledged that the chances of their going on pending on the district) 32% to 63% of poor to secondary school were less than 50%. households had one or more children who had dropped out of primary school (Table 18). Of Girls are Withdrawn from Primary 187 households with children who had dropped Schools First out of school, 155 reported inability to pay When poor parents had to make a choice about school fees as the reason. Marriage, pregnancy who would stay in school and who should be and sickness were mentioned by a small minor- pulled out, the daughters were the first to go ity as the other reasons resulting in school drop (Table 19). Overall, more than twice as many outs. respondents felt that girls should be withdrawn from school before boys. The primary reason Increased Dropout Rates in the given was that girls would marry and move Last Two Years away (37%), and that they would not make In the past two years, because of high inflation money or were a waste of money (4%). Boys and drought, primary school drop-out rates were left in school because they were the Participation Series 39 Learning from the Poor "homeguards" (12%) and would also bring Access to Health income to the family (20%). In Mandera, the gender preference was reversed, with parents The poor were health conscious, knew names of reporting pulling boys out first. However this common diseases and their causes and engaged is done in an overall cultural context where in a variety of health maintaining strategies, most parents in this district are nomadic including use of govermnent health clinics. pastoralists and do not value formal education Numerous studies have shown that the poor as much as people in other areas. use public health facilities less often than the well-to-do. Hence, the policy question posed in Access to Bursaries by the Poor the study was: what factors determine the use Poor parents who reported high costs of of health facilities among the poor? primary education as a major issue were asked how they coped with these costs. Only The picture that emerged was clear, yet com- three households in the entire sample men- plex: the poor visited government health tioned bursaries from the government, NGOs facilities only wlhen they were desperate and or church-related groups. Many participants often as a last resort. This is not because they said they had never heard of a student actu- do not believe or respect the curative powers of ally getting a bursary, or if they knew about the health providers, but because their experi- such an award, they said it went to a student ence with health facilities has been so dismal in who wasn't actually poor. Because many the last few years. bursaries are designated for secondary school students - and most poor children do not While in theory the fees of the poor are sup- attend secondary school - they automatically posed to be waived, except in Kitui and Kwale fall to children in better-off homes. Another none of the poor mentioned waiver of fees by observation made by poor people in the hospitals or clinics. Generally, people reported villages was that bursaries were awarded by that they had to pay Ksh 10 to register at a district leaders to students identified by health facility and an additional Ksh 10 to get village chiefs and elders. A better plan, many prescriptions for drugs. In some areas, patients villagers said, would be to have winning were required to buy children's exercise note- students selected by school headmasters, who books to supply the paper on which prescrip- were in a better position to know who were tions could be written. In other districts, the the brightest and the neediest students. poor were forced to tip somebody on arrival at the health center in order to be seen at all. "If Table 18: Number of Households Table 119: Preference of School in Which Children Dropped Out Dropouts by Gender by District (%I/o) of School by District (%) Either Category Yes No Category Boy Girl Boy or Girl Kwale 43 57 Kwale 33 52 15 Kitui 59 41 Kitui 25 60 15 Mandera 32 68 Mandera 61 37 2 Nyamira 32 68 Nyamira 17 83 - Kisumu 39 61 Kisumu 19 57 24 Bomet 27 73 Bomet 16 60 24 Busia 63 37 Busia 16 61 23 Total 42 58 Total 27 58 15 40 Environment Department Papers Coping Strategies and Mechanisms Box 17: Health Problems in Two Districts, Kwale and Kisumu Kwale District In Gora village, the government health center serving the population charges a low fee (Ksh 5), but is 7 km from the village and very poorly equipped. It has no drugs most of the time, no transport for serious cases, no dressing materials and no electricity or electrical generator. "When a woman comes in the night to deliver, they use candles, torches, or paraffin lamps." Lack of drugs at the hospital has led to people resorting to traditional treatment. Bumbani has its own health center built through Harambee efforts. While cost is not an issue (fee of Ksh 5), the center suffers from a perennial lack of drugs, and has neither water nor electricity. During the ranking of health problems in the village, hostile health personnel ranked as the number one problem followed by lack of drugs, "dirty health care center" and poor 1ransport. The village of Mwapala is served by a dispensary and health care center both about 7 km from the village. While the center suffers from water and drug shortages and has insufficient staff, 60% of respondents said they would still take sick family members the:re. The maternity ward, however, is hardly used. The commu- nity has formed a health committee to improve the facilities, make repairs and build a kitchen for the maternity ward to encourage use. Gulanze village is served by a dispensary and hospital both 10-15 km away. Transport is a major problem. While there is a public bus which makes one trip per day through the village, many people cannot afford the fare and the bus is usually over-crowded by the time it reaches the village. The sick are thus forced to walk if they need medical attention. Hostile staff and a shortage of drugs were also cited as problems and some villagers turned to the traditional healer as the only viable alternative. The nearest public health facility to Kinyungu village is 23 km away. While there is a private clinic 7 km from the village, the prices are too high for most of the villagers to afford. Again, transport is a severe problem. There is no public transport and the sick are normally carried on someone's backs, in a wheelbar- row, or on bicycle to the health facility. In adclition, the roads are completely impassable in the rainy season. District decisionmakers admitted that the health facilities were inaccessible to many, and that due to lack of transport and lack of money, the very sick often die in their rural homes before getting treatment. Kisumu District Villagers from Ombo rely on a local dispensary (4 km away) and the hospital in Kisumu town (20 km away). All groups of community members cited the shortage of drugs as the number one problem. This was followed by lack of money and transport, with crowded facilities and hostile staff a distant third and fourth. For the village of Kabour, the nearest health facility is the Kisumu District Hospital 13 km away. Shortage of drugs at the facility, lack of money, and poor transport were ranked as the number one problem by all. In Sare village, community members, assisted by UNICEF, initiated the building of the Kinasia dispensary about 1.5 km from the village. Despite the efforts of the community, health problems remain severe, many related directly to the abject poverty in the area. Malnutrition is an endemic problem in children and expectant mothers and outbreaks of malaria a:nd diarrhoea are rampant. Supplies of drugs were at best inconsistent and have been completely unavailable for six months. Transport to better facilities is almost non-existent and the area is completely inaccessible by road during the rains. In Kosano village, men felt that hostile staff were a major deterrent in using the health care facilities located about 7 km away from the village. They stated that patients are abused and turned away. The women and youth of the community, however, felt that the shortage of drugs and lack of money were more to blame for the problems associated with using the facilities. In Okana village all three groups (men, women, and youth) agreed that the major problem with the Rabuor Health Center (8 km away) was the perennial shortage of drugs. Rude nurses, lack of money, over-crowd- ing, and poor transport were also cited as problems. Participation Series 41 Learning from the Poor you do not pay someone, you can wait all day our prescriptions?" one patient said. in the sun and the doctor will leave at the end of the day and you will still be waiting." In Kitui, people atlso said that they could not avail themselves of the free services, because District leaders did not seem to be aware of the the distance to the health clinics and absence of extent of this problem. The Medical Officer of roads meant that the sick had to be carried in Health in one district (the chief medical person stretchers and wheelbarrows for 25 km. In in a district), said all people were treated, no addition to the difficulties in reaching clinics, matter how poor they were. If a poor person once having arrived the patient may find that could not pay, he said, then a waiver was the facilities are in such disrepair they can granted and the patient was treated without barely be used. For example, the Samburu charge. In the villages, by contrast, the poor health center which serves villages in the Kwale were firm in insisting that waivers were in district had no generator to provide electricity, general never granted. Instead, quite the no drugs, no transport for the sick and no opposite was true. Patients had to pay an bandages or dressings. In other health centers additional fee even to reach the registration in the district, clean water was not available. table. "If you don't have 20 shillings, no one The perennial lack of drugs was most com- will ever touch you," said one respondent. In monly ranked as the number one problem by Kitui, poor patients complained that they had the respondents. In Nyamira district, commu- to provide the paper on which their prescrip- nity members were concerned that health care tions were written, and when there was no staff were "diverting" the drugs, thus making paper they were told to go to the shop and buy access impossible for the general public. In a school child's exercise book. "When we don't other districts, people stated that the drugs for have enough money to pay for the exercise government hospitals were sold to private books for our children in school, why do they hospitals where high prices meant the poor think we have enough money to buy them for were excluded. Box 18: Rudeness of Health Personnel Affects Clinic Use While much is already known about the general lack of availability of even common drugs in health clinics, widespread rudeness of health staff has not previously emerged as a major issue. People everywhere reported that they were abused at health clinics and would only continue to go "because we have no choice and need the services." Men, women and youth over and over again stated that they were treated like animals "worse than dogs." They reported that even before they could explain their symptoms, they would be shouted at, told they smelled bad, and were lazy and good for nothing. Some people stated that they chose to treat themselves with over-the-counter medications ratlher than go to the clinic. As some of the youth from Kitui stated, "We would rather treat ourselves than go to the hospital where an angry nurse might inject us with the wrong drug." An older man in desperate need of spectacles braved the abuse of a nurse for two days until he got his glasses. But, he said, he would never again go back and be so humiliated. In Kwale, women said that they would rather die on the roadside than go to the maternity center to give birth because of the harsh and uncaring behavior of nurses. Even baby growth monitoring has been per- verted in some areas. Women reported that if the nurses found that their babies had not gained weight, they would scream at the mothers and accuse them of eating the food meant for the babies. In Nyamira, when a mother took an injured child to the Kijauri dispensary, the mother was reprimanded, refused treatment for being careless with the child and told to go to a private hospital. One woman was refused treatment because she didn't speak Kiswahili, even though the nurse spoke the woman's local language. In the one hospital, a women in labor was left unattended by a nurse as she gave birth on the floor. While others watched, the nurse hurled insults at her and prevented them from helping the woman. Finally, after giving birth, the mother was made to clean the mess on the floor. 42 Environment Department Papers Coping Strategies and Mechanisms Choosing Who Gets Health Care vote to the man as the income earner, followed When the poor are short of money and have to by the baby and the boy. For example, in Geta make tough choices about who to invest in, B village in Nyamira women always gave men father, mother, boy, girl, or baby, some gender the preference for health care, even in examples and age differences emerge. When asked to of extreme poverty. The youth usually gave the indicate which family member they would give baby first priority, the mother the second priority to when investing scarce resources, priority and then the boy. With an exception, all men often chose the mother, the boy and baby. the groups of men, women and youth in all the A typical response went something like this: "If districts left out the female child. I die, I will die happy knowing that the mother who is the light of the house will take care of Coping Strategies the family. The son is the lusiola (a tree which When a family member is sick, health facilities produces hard wood) of the home and will far away and money scarce, people used a continue the lineage. The baby is innocent and variety of strategies for coping with health- fragile and cannot speak for itself and needs to related problems (Table 20). Traditional birth be deeply protected (Yeye ni kama malaika). attendants were used except when they re- quired that the mother bring her own materi- Women, on the other hand, often gave their als, such as an exercise book, blades and soap. Box 19: Understanding Health Seeking Behavior of the Poor Using a Visual Participatory Method Health clinic use is determined by at least two sets of factors: the attributes of the person seeking health care and the attributes of the health center. To understand the health-seeking behavior of the poor, a research activity was conducted outdoors with separate groups of men, women and youth. The technique involves a set of cards with simple line drawings developed by a local artist. First, a picture showing a health clinic is set on the ground. Referring to the picture, participants discuss the health clinic they use, and its good and bad points. Five cards (depicting a man, a woman, a girl, a boy and a baby) are then put in a vertical row and the discussion centers around when these different members of a family are likely to be taken to the health clinic. Once this topic is exhausted, seven new pictures are introduced and spread under the picture of the clinic. These focus on the attributes of the health center - lack of drugs, crowding, no doctors, closed clinic, distance, rude nurses and payment of fees. Once the clinic's attributes are clearly recognized, people are given 10 stones (which in the second round are reduced to three to encourage prioritization), and are asked to distribute them to indicate the qualities of the health center which are most important to them as users. In the last stage of the activity, the group's attention is turned back to members of the family. People are once again given 10 stones and are asked to distribute the stones to reflect their priorities in taking family members to the clinic. To force prioritization, the number of stones is reduced to three. People are asked what they would do if all family members were sick and their resources were limited. Everywhere the task generated great involvement and much heated discussion among the participants. Several key findings emerged from this activity. In terms of the attributes of the health center (with some variation), the three most important factors that have discouraged use were the distance and associated transport costs, lack of drugs and being informed of their unavailability only after the consultation fee had already been paid (thus, the patient had a presc:ription for which they now had no money to fill) and the general rudeness of the health staff. In terms of who would get priority medical attention within a family, one finding was almost universal: girls were at the bottom of the heap. In almost all cases, they were the last to be chosen for medical attention by all groups - the men, women and youth. Participation Series 43 Learning from the Poor Women also helped each other give birth with well developed in Kitui/Mwingi District where no attendant present. Community members in the female household heads said, "God has all districts consistently turned to traditional provided trees free of charge and one has only herbs and prayer because of the high cost of to go to the right herb, boil it and take it." In medical services, transportation and drugs. In this case, herbs were seen as the savior of the two villages in Nyamira district, for example, community ("niii'i kuut ni mningi na niyo ya over half the respondents (53%) said they tuokoie"). would stay home and pray for help rather than go to the clinics. Mission hospitals received mention and high praise in many districts. Mention was also made A number of widows in Busia District observed of community-based distribution systems that they normally preferred to go to hospitals, modeled on the B3amako initiative which began but when they had no money they then resorted in the last few years and appeared to be working to traditional herbs like the Khalulu and Kasanga satisfactorily in some areas. Through such Liuba, bitter roots well known for alleviating systems, the poor said they were able to buy stomach ailments. Such a coping strategy was drugs at cheaper costs and closer to their homes. Table 20: Action Taken in Case of Sickness When There is No Money (%) No Consult. Use Trad. Taken to Purchase Borrow Category Action Trad. Healer Herbs Hospital pDgs Pray Monev Other Kwale 3 8 5 53 8 3 5 15 Kitui - 1 41 20 1 5 12 20 Mandera 4 - 14 16 - 55 4 7 Nyamira 3 11 28 3 2 42 3 8 Kisumu 9 1 24 51 7 7 - 1 Bomet 3 - 1 61 3 3 23 6 Busia 12 1 27 29 9 11 3 8 Total 5 3 20 34 4 17 7 10 44 Enviromnent Departmnent Papers 4. The Untapped Potential of Indigenous Self-Help Networks The poor, and especially the very poor, are poor in building strong self-help groups. difficult to reach through centralized govern- The potential of women's organizations as ment services, particularly when the quality of income producing groups is largely uin- service is inadequate and when cost sharing is tapped. involved. How then do the poor survive and where do they turn for help? Each of these findings is discussed below. All societies have indigenous social networks Harambee and Social Networks which form the basis of resource and labor exchanges and reciprocity. Kenya is no excep- are Under Stress tion: in every village, there were at least 5, and sometimes as many as 17, associations. This "Harambee," originally a strategy promoted by suggests that nationwide, there are about the government across all sectors, involves 300,000 groups in the rural areas. The social people getting together to help themselves in capital embodied in these informal networks tackling a common or specific problem in the binds a group of people together and fuels community. While Harambee once represented development. Under normal circumstances, an important coping mechanism for the poor, in these informal associations play a dispropor- the present climate of high inflation it seems to tionately important role in the lives of poor be over-extended. In the absence of timely people who are constrained in every way support, technical guidance, business manage- except by their labor. ment training and financial support, Harambee may have reached its limits across sectors, However, during periods of stress, the poor fall especially in the arid rural areas. outside most informal and formal self-help systems. The PPA study found in particular that: Hundreds of cases were recorded everywhere of the poor investing their resources in misguided * Harambee and other social support net- Harambee efforts related to water, farming, works are overextended and are breaking livestock, education, health and a range of down in the absence of timely support income generating activities. School buildings from the outside. without books, health clinics without drugs, chickens that die before they can be sold and o g cotton that does not grow are of little use to enous self-help groups, but they either do anyone. As people said during study inter- not reach the poor or are trapped by views, "Now there is too much Harambee; we undercapitalized and therefore cannot meet do not want any more. the needs of their potential clients. * NGOs and church groups, with a few While the underlying system of people organiz- exceptions are providing welfare support ing themselves to solve problems is still strong, it to the poor, rather than supporting the is extremely stressed. The study also found that Participation Series 45 Learning from the Poor the poor receive very little support from their grow, they often form associations to leverage kin in solving food or school fee problems. their strength inI order to gain access to re- Lacking support, the poor, particularly single, sources or for marketing. widowed, divorced and abandoned women and There are several types of self-help groups in the their children migrate into the slums of city rural areas in addition to the informal extended center. kinship networks described above. Most villages have multiple self-help groups, some of Cooperatives and Self-Help Groups which exclude the poor either because of high fee requiremenis or because they require mem- Experience from around the world establishes bers to own certain assets. Others of the groups two factors as key to successful group formation make specific provisions for the poor, and some and functioning. First, the group must coalesce are specifically targeted to the poor. around a common interest, be embedded in local culture and be self-governing (have its Cooperatives own clear rules regarding entry, exit and In Western Kenya, several villages had dairy, governance). Second, the group should have coffee and agricultural cooperatives. Ranching the skills and resources, both material and cooperatives were more common in Kwale. The financial, for the task at hand. cooperatives tended to be of two types: village- based cooperatives such as the Butula Dairy Since successful groups are based on local Cooperative Society in Busia, or externally- interest, culture and invisible sociological based cooperatives, such as the Coffee Coopera- factors such as kinship clan and friendship ties, tive Society in Nyamira, which lie outside the it is usually easier to train existing groups in village. Both types exclude the poor either technical and business management skills than because of fee or asset requirements. it is to induce group formation. Hence, pro- grams that require group formation but do not As noted in an earlier section of this report, the allow time for the evolution of socially viable Agricultural Finance Corporation (AFC) lends to groups fail. As the number of successful groups the agricultural sector, but does not reach the Box 20: Village-Based and Externally-Based Cooperatives The Butula Dairy Cooperative in Busia evolved from a village women's group which prospered when it received high milk yielding cattle from the Norwegian govemment. To expand and increase profits, the group requested government permission to form a cooperative to market rmilk nearby rather than travel 25 km to the capital of Busia. Membership is open to all who own cattle after a registration fee of Ksh 50 is paid. There are currently 270 members. The cooperative is managed by an elected committee which sells shares at Ksh 100. If a member owns 10 shares, he or she become eligible to apply for loans from the cooperative. Milk is collected twice a day at a fixed time by milk men on bicycles. They check the milk for quality and enter the amount bought in a register. Payments are made promptly at the end of the month. A livestock veterinary officer regularly inspects the cattle and advises on appropriate breeding methods. Future cooperative goals include buying a vehicle for transport and construction of a food processing plant. By contrast, the Coffee Cooperative Society in Nyansankio, Nyamira, is based in the sub-location headquar- ters and is part of the coffee cooperative organization in Kenya. Members can obtain loans for fertilizers and pesticides and also sell their coffee to the cooperative. Similarly, agricultural cooperative societies such as the Pyrethum Society provide animal feed, fertilizers and farm implements to their members on loan. While both village-based and externally-based cooperatives provide their rmembers with many benefits, the poor - who do not qualify as members - do not benefit directly from either type of organization. 46 Environment Department Papers The Untapped Potential of Indigenous Self-Help Networks poor because it requires a minimum of 5 acres low cost items, such as household utensils and of land as collateral. In western Kenya, the small tools which are shared or owned indi- inability to access AFC loans was frequently vidually. These groups were more widespread mentioned. in poor communities and in the dry zones. Self-help Groups NGOs and Church Groups There are a great number and range of small nmd large self-help groups, some of which exclude the Church-related groups and activities were poor and others of which target the poor. How- mentioned almost everywhere and were usually ever, even in villages where there were many localized. In Bomet, health and livestock groups - one village in Nayamira, for example, extension services attached to the Tenwek had eight well-functioning groups - they barely Mission Hospital played an important role in had an impact on the lives of the poor. This was some of the communities. The Kaplong Hospi- due both to the large number of poor families tal community workers, in addition to health and the multiplicity of their problems which education, introduced fuel saving stoves and keeps them going from crisis to crisis. Box 21 assisted women to organize harambees to raise summarizes some examples of groups which money for school fees. exclude the poor from membership, but whose mandate includes helping the poor. NGOs were mentioned particularly in Kitui and parts of Busia, Kisumu and Kwale. Externally Self-help Groups for the Poor financed programs were mentioned in Kitui, In contrast to the cooperatives and self-help Kisumu, Busia and Kwale. NGO activities were groups which exclude the poor, most groups generally welfare-oriented. For example in for the poor focus on shared labor, pooling Kitui, the activities focused on forming groups cash resources for emergencies and purchasing Box 21: Self-help Membership Groups Which Exclude The Poor, Nyamira Groups Not Directly Benefiting the Poor Obwanchi Self-help Group, Nyaigesa village. The group's major goal is to raise money to buy commercial plots and assist each other in tasks which are too difficult for an individual to accomplish alone. The group has 45 members, including 25 women. The monthly contribution is Ksh 150. Kiongoni Self-help Group. The group raises funds to buy dairy cows for members on a rotational basis. The group has 35 members (6 women) who contribute Ksh 100 each month. Members who are able to consis- tently meet monthly payments benefit the most. Groups Helping the Poor but Excluding theni from Membership Omorembe Umoja Ni Nguvu Group, Sengera .I village. The group's 35 members (15 women) raise money to purchase goats for one another through a rnonthly fee of Ksh 70. By pooling resources, members receive goats on a rotational basis which they then own individually. The group 'ilso raises money for emergen- cies which can be used by members and the poor. It has successfully helped the poor with school fees, hospital bills and small business start-up loars which must be repaid with interest. Boyamondo Self-help Group. This group developed out of the Omorembe group mentioned above, specifi- cally for those who were too poor to consistently meet the Ksh 70 monthly contribution. The group's 70 members are required to pay registration fees of only Ksh 25; in emergencies, fees of only Ksh 5 are collected. The group has assisted the poor with food, clothes, shelter, hospital bills and school fees. Participation Series 47 Learning from the Poor Box 22: Self-help Groups of the Poor Kyamwui Health Self-help Group, Kitui. Created by Action Aid, an NGO active in the area, the group focuses on health-related activities. These include building pit latrines and rubbish pits, cultivating and terracing shambas, digging wells and raising funds for member emergencies (such as paying hospital bills). Anyone who assists in group activities is a member. Ndetani Men's Group, Kitui. This group has no membership fees. Members help each other with farming by pooling their labor. The group's effectiveness is limited due to a lack of farming implements and finan- cial resources. Kiabakungu Shelter Group, Nyamiiira. The Kiabakungu group was started in 1989 to improve houses and to help members purchase dairy cows. The initial monthly contribution was Ksh 100, but this was lowered to Ksh 20 so that the poor could participate. The group has 32 members 24 of whom are women. Thus far, the group has built 5 houses and bought 8 cows for its members. Michira Dairy and Horticultural Group, Bomet. This group, formed in 1990, has a membership open to all with a registration fee of only Ksh 10. Additional monthly contributions are decided by members. The group's goal is to market vegetables, attract government extension services, start zero grazing units and build a stall for selling farm produce. Group capital currently stands at Ksh 1000. The group is stagnating because of the inability to raise more money. Mary Goretti and Immaculate Groups, Kisoko, Busia. These health groups were formed by the Catholic Church and focus on household sanitation, hygiene and primary health care. The groups also pool labor to assist each other with gathering thatch grass and building roofs. Members contribute Ksh 20 in membership fees. Box 23: Issues Associated with NGOs Several study teams reported that NGOs, while doing good work in their target: areas, seemed to cluster their efforts in a few locations at the expense of the rest of the district. A few district leaders also voiced this concern. In Kwale, district decisionmakers reported that a number of programs were on the ground to address the needs of the poor. These were mainly NGO programs dealing with water, education, health and general development. These NGOs were coordinated tliough monthly meetings of the District Development Committee. It was noted that of some dozen NGOs operating in the district, only two were actually located in the district, with the rest located in Mombasa. The study team also found that NGO activity was concen- trated in the coastal zone which includes the principal Mombasa-Tanzania highway and a number of beach resort hotels, while the hinterland was less well served. This larger interior area is characterized by drought, lack of water year-round, tsetse fly attacks on livestock and various forms of wildlife which devastate crops. "We have many NGOs working here," said a chief in another district. "They all work in one small circle of communities" (he made a circle with his finger on the table). "The largest number of people are ignored" (he spread his arms to indicate the rest of the table). "I don't know why." District leaders agreed that some of the most effective services aiding the poor were provided by NGOs. At the same time, despite mechanisms such as the Kwale District Development Committee, there were frequent comments to the effect that the work of NGOs must be better coordinated in order to distribute assistance to all areas of the district, target services and prevent overlapping services. 48 Environment Department Papers The Untapped Potential of Indigenous Self-Help Networks Box 24: Women's Groups Using Social Capital to Generate Income Ombo Women's Group, Kolunje Sub-location, ]Kisumu. The group formed in 1983 to start income produc- tion activities. All the members belonged to the same clan and started with rope making and weeding rich people's farms. Wanting to diversify, they rented two fish ponds from a neighbor, restocked the pond, purchased fish food and marketed the fish at the local market. Despite increased fish production, the activity was abandoned because open access to the fish ponds resulted in high levels of stealing of fish and a declining fish harvest. The group now focuses on basket making, hiring of vehicles for the very sick for transport to hospitals and, when income is available, extending small loans to members for petty trading and hawking activities. Sare Women's Group. The group also began in -1983, emerging from a community group which built the local primary school. Members focus on the needs of school children and through harambee have bought a cupboard, textbooks, desks and a blackboard. They diversified into cotton growing, which proved to be a disaster, and are now growing some vegetables, ihough they are limited by a lack of water. The group has saved Ksh 27,000 in a bank account maintained by the Finland-Kenya Friendship Association. Kitui Women's Groups. Most of Kitui/Mwungi district was rated poor or very poor by community members. Because of the poverty of their groups. they have little working capital and can therefore provide only small amounts of money to members in dire need. The groups also help women buy small household utensils for cooking and occasionally help with school fees. One group does weave baskets, however they have difficulty selling them in the local market, which has declined in recent years due to increasing poverty and recurrent drought. Nyamira Women's Groups. Both Muchenwa (80 members, of which 6 are men) and Omoteme Women's Groups (47 members, of which 5 are men) started with Kshs 20 as entry fees. Because of high interest in becoming a member, fees have been increased to Ksh 500 and 200 respectively. The goal of the groups is to buy commercial plots, rental houses, posho mills and household items. The Omoteme Group has assisted in building 20 houses for its members. Menyenya W/Vomen's Group and several like it have a membership fee of Ksh 20 and rent land to grow vegetables which are then sold to the secondary school. With the meager profits from the vegetables, the groups buy household utensils. Other groups invested in poultry raising, but this was abandoned when poultry disease killed all the chicken. A private hospital-based livestock extension service is now advising the group on rabbit rearing. Bomet. This district, which has a high agricultuwral potential zone and a low potential area in the south also has women's groups that engage in petty trading activities. Membership in the groups is open to the poor, who are given free maize. Some groups support the poor by assisting them during the planting season, buying clothes and paying hospital bills. for health and sanitation. In Busia, the activi-. helpful to the individual poor in their struggle ties focused around provision of water through to survive, they have not assisted in strength- drilling of boreholes (most of which were not ening the capacity of existing groups to become functioning). In all areas, NGOs also received autonomous. some mention in being helpful with payment of school fees and purchase of books. However, Women's Groups there was little or no evidence that any groups formed or assisted by NGOs had achieved any There are more than 23,000 registered women's level of autonomy. Nor had they evolved into groups in Kenya. Grounded in Harambee and larger groups or diversified their activities. evolving from a social welfare orientation, Thus while NGO and church inputs are directly women's groups today are struggling primarily Participation Series 49 Learning from the Poor national survey, less than 2% of the registered During discussions of coping strategies at the women's groups reported social welfare as their individual, group and community levels, primary activity (Women's Bureau, Ministry of women's self-help groups received frequent Culture and Social Services, 1993). The pri- mention in every district. Mini-case studies on mary activity of the groups was related to the functioning of these groups provide addi- increasing the income of its members through tional insight: agriculture, livestock, finance, sales and ser- vice, construction, real estate, manufacturing * Every village had more than one active or and provision of transport. defunct women's group. Often, they emerged from existing collective action While some groups receive some financial community groups. assistance, and there are some success stories, * Membership in women's groups was an by and large the efforts of women's groups are important part of coping strategies particu- misdirected and invested in low return or f l inappropriate activities. Overall, they are also constrained by the low levels of funds available * In addition to income production, two to the group and their own limited business welfare activities were frequent objectives of acumen and technical skills. groups: first, raising cash to pay school fees and to meet hospital expenses; and second, It appears that Kenya's large number of well- assisting with transport costs for burial. organized women's groups are a major un- Women's groups were financially more tapped resource. While there are large success- prosperous in the medium and high ful cooperative movements - for example in potential aricultural areas. In the oorer coffee and more recently in dairy - women's areal engroups ared resoures - ~~~~~~~areas, women's groups pooled resources groups have not generally been viewed as for basic survival and petty trading. The important social capital (associations and majority of the groups felt constrained by networks) that can fuel production growth in their limited financial resources. the rural areas. There has been no systematic effort to develop the capacity of women's * Most groups levied membership fees and groups as important production groups. monthly contributions which ranged from Box 25: Even in Mandera, Women's Groups Persist Even in the harsh circumstances of Mandera - a dry semi-desert district which has reported the harshest environment of any of the areas studied, and which is isolated from markets and city centers - women's groups persist in their efforts. In Arda Kalacha, a village in which everyone was categorized poor or very poor, a women's group came into being seven years ago to assist needy people, initiate harambees (public fund-raising events) and help the poor pay for school fees. The group has 30 members. Despite inadequate and short rains, women undertake small-scale farming during the two brief rainy spells and walk 18 kilometers to the nearest trading center to sell their produce. A few years ago, the group invested in poultry raising; for several years they sold eggs successfully at the trading center. Recently, a disease wiped out a majority of the hens, but the project is continuing. After a while, the women collected resources to start goat herding, but these were also killed, this time in the drought of 1992. The few that survived were sold. Rather than giving up totally, the group then resorted to buying stock for petty trading; they now sell sweets, cigarettes and biscuits in the community. Despite their very limited resources, they continue to extend help to the most destitute in the community who are unable to participate in or contrib- ute to the group's efforts. 50 Environment Department Papers The Untapped Potential of Indigenous Self-Help Networks Ksh 5 to Ksh 1,000 per month. Often these trying to make their lives better through fees were used to establish excludability of collective effort. benefits. * Women's groups often had men as mem- • While the poor were excluded when bers, were supported by men and the membership fees were greater than community at large and were organized Ksh 100 per month, many women's groups along clan or kinship lines. targeted their activities specifically to the poor. If this was the primary purpose of Money alone will not help the women's the group, the poor were exempt from groups to become economically viable pro- payment of fees. Such groups helped the duction groups that reach the poorer seg- poor with food, school fees and housing ments of rural society. In order to tap the construction. potential of Kenya's existing women's groups, * Even the women's groups with fairly large major obstacles must be overcome. These membership fees (Ksh 500 to Ksh 1,000) include legal registration of the groups so that often invested in failed activities. Most they are eligible for credit and provision of groups, however, worked hard and despite management training. experiencing many failures persisted in Participation Series 51 5. Urban Nairobi The Situation of Female-Headed Households The majority of women living in Mathare in ftie Mathare Valley and Korogocho Slums Valley and Korogocho left their rural homes because of significant stress. Of the 60 women In addition to the seven rural districts studied for interviewed in the study, 17% migrated to the PPA, research was also conducted in two of Nairobi becauase of the break-up of their Nairobi's worst slums. The Mathare Valley and marriage; 13% said they had come to the city Korogocho slums, an area covering about 15 because of abandonment and another 13% square kilometers, are some of the 78 informal because they were widowed. Other factors settlements in the city and are situated in included having a child out of wedlock (13%) Kasarani Division. The division has the highest and fleeing a violent husband (12%) or alco- number of informal settlements within Nairobi. holic parents (5%). The rate of population growth in the slums is very rapid due both to high fertility rates and in- Most of these women arrived in Nairobi virtu- migration. In 1975, for example, Mathare Valley ally without assets of any kind. Many are from had an estimated population of 70,000; by 1987, the same districts studied in the rural poverty the population of Mathare Valley was estimated assessment; the cycle of poverty which often at between 150,000 and 180,000, with a popula- starts there with female-headed households tion density of about 1,250 persons per hectare. continues in the urban setting. The majority of women in the urban study were heads of The women and children of these urban slums households through divorce (34%), through are among the poorest people in Kenya. When being widowed (23%), or because they were interviewed more than 90% of the residents "never-married, single mothers" (23%). In categorized themselves as poor or very poor. addition, while many of the women came from Box 26: Women's Voices from Nairobi's Slums "My brothers chased me from home because I had two children. You know, living where people hate you is difficult. So I decided to leave. I got my room in Korogocho. The death of my Dad led to my not going to school when I was in the village. In the process I got my first child, who led to the second one. I wish my father had not died." "When my husband died, my in-laws told me to get out. So I came to town and slept on the pavement. This lady who gave me this house found me and brought me here. I anm her maid, and I also sell the water for her. She pays me 30 shillings a day." "Both my parents used to drink, and therefore neglected the children. They could not do anything worth- while to assist us. I got married in 1982 and divorced in 1987. We divorced because my husband was an alcoholic. He started selling property ... to get money for alcohol. We had no shamba. When I stopped him from selling things, he beat me. He chased me, and I came to Korogocho." 52 Environment Department Papers Urban Nairobi rural areas, few can go back, as illustrated by in; an overwhelming majority (80%) rented the following statement from one of the women: their homes. The inhabitants were, for the most part, squatters on land which is predomi- "I don't have any house or any land or anything nantly owned by the government or city because I parted company with my husband ancl council. he does not want us. The oldest boy whom I got before marrying the first husband tried to go horne The houses which most of the women (82%) the other day, but the men chased him away." lived in were made of mud and wattle; card- board, plywood or even milk cartons were used Infrastructure to patch walls. Roofs were made of corrugated iron and tins, and in most cases were old and Housing leaking. Ventilation is poor. Room size averaged around 8' x 8'; very few women had two rooms None of the women in the study owned the for their own use (33% of the women lived in 8' x land on which they lived. Very few of the 8' rooms and 25% in rooms of 16' x 16'). Rents in women (12%) owned the structures they lived both slums ranged between Ksh 200 - 750; the majority of inhabitants paid between Ksh 200 _ and Ksh 350. More than two-thirds of the Box 27: Focus Group Discussions: women described their housing as "inadequate." Responses by Vumilia Women's Group Yet even this inadequate housing can be taken from them. Because people are squatters, As part of the Participatory Poverty AssessmenI, owning only their house (at best), the local a focus group of women living in Nairobi's owning o r house (ath the loca n urban slums was interviewed. When asked to government or city council (with the permission characterize who a poor person is, their basic of the central government) can, at any time, definition was: "When people are poor, it means remove the structures without compensating the they do not have food and clothing for them- owners. selves and their children." The distance from city center to these settle- Other characteristics of poor people identified ments ranges from 5 km to 17 km, the nearest by the women's group included: being the Mathare Valley slum. All roads in * no money to buy anything Mathare Valley and Korogocho are temporary, with numerous footpaths. Access by vehicles in * no land, house or property Mathare Valley is limited due to the rough * they steal scavenge and beg terrain and the high density of the built-up * they have large families structures. While the roads in Korogocho are * they are orphans planned, they have been swallowed up or * they cannot afford education, medicine, blocked by inadequate waste disposal. hospital charges, nor start a business * they are squatters Water and Sanitation * they have physical and mental disabilities While water in the urban centers of Kenya is * have ill mannered children commonly connected to individual households, * have no relations or support group this was not the case in the slums of Mathare * job insecurity Valley and Korogocho. Water is obtained from * have no husband standby taps owned by private individuals, * share room with children who draw it from the main City Council pipes. * inflation battered Some people get water from the nearby rivers, * dependent on God's Grace to survive especially those who cannot afford to buy * no means of livelihood even with prayer supplies on a daily basis. Vendors who own and deep thought. water kiosks control and sell water to buyers. Participation Series 53 Learning from the Poor Water costs between Ksh 0.50 and Ksh 1 for a paying for treatment, most did not have any 20-litre container (debe). In a few cases, land- money left with which to buy drugs. In addi- lords had water connected to one room. tion, the cost of food and rent were so high that the majority of slum residents simply could not The sewage disposal system in the area is poor. afford to treat their sick. Inhabitants use pit latrines, damping grounds and the river to deposit human waste. While there are Education Facilities some public toilets, they are generally overcrowded Mathare Valley has three primary schools, each and dirty with poor drainage. For example, one pit of which has a nursery. Korogocho slum has latrine in Korogocho was used by sixteen house- three primary schools, two nursery schools and holds. The overflowing latrines force the inhabit- two informal primary schools. Total enrolment ants to use the damping ground instead. In in Korogocho is 4,720 pupils in formal primary general, because of the pressure on space, the pit schools and 1,300 in the informal schools. It is latrines are built very close to the households, estimated that only 50% of the eligible children making the area extremely unhealthy. are enrolled in the few educational facilities. HealthFacilities Coping Strategies and Institutional While health facilities are located not far from Support the slums, the majority of women in the study said they would only go to the hospital as a last Food resort. Prayer, over-the-counter medications and traditional herbalists were usually relied Food is abundant in the slums, but the resi- upon first. Hospitals were expensive and after dents do not have enough money to pay for it. Box 28: Life Histories of Focus Group Respondents Woman A: displaced from Molo in 1992 due to the tribal clashes; has eight children who helped her with her casual labor in Molo on people's farms; her house was burned in Molo and all her belongings lost. Woman B: also a victim of tribal clashes in Molo where her husband was killed; has seven children with whom she stays in Mathare Valley; they collect paper for sale at Ksh 2.50 per kg. at Eastleigh. Woman C: born in Mathare Valley slum; has eight children and is jobless; was given a place to sleep by an elderly man and shares the room with his children; her mother suffers from high blood-pressure. Woman D: her father died and then her mother and children moved to Nairobi; there are now eleven in their family - seven brothers and four sisters; she has eight children of her own. Woman E: chased from Narok during the tribal clashes; came with her mother who is now very sick; a friend gave them accommodations; she does casual labor in order to feed her mother. Woman F: chased by husband when she had her fourth child; now has five children and buys and sells vegetables to raise money for food. Woman G: came from Muranga; her mother died in 1961 when the three children were very young; the father moved to Kayole (Nairobi) where he was involved in casual labor as a farm hand; she did not go to school and labored with her sister to put the brother through school, but lhe dropped out in Standard 3; in 1990, her father died and soon afterwards her sister also died; in most instances she is not able to feed her only daughter so she sends her to the Kiboro feeding program; she is not able to pay rent and is supported by church members; she is HIV positive. 54 Environment Department Papers Urban Nairobi Most of the women relied on "rejects from the Education butcher," such as animal hoofs, and heads, legs Most of the women from both slums had two or and intestines of chickens. About 85% of the three children in school and on average three women said they did not have enough food for school-age children in the household not yet their entire family. Seventy-three percent of enroled. In the Mathare Valley slum, 32% of the children eat only one or two meals per day. lI families had no drop outs from school; 18% had a food crisis, 50% of the women said they one child drop out; and 50% had two or more would starve; 34% would borrow from children who have dropped out of school. In the neighbors. case of Korogocho, 56% of the families had no drop outs from school; 14% had one child drop out; and Most families prepared their food at home though 31% had two or more children who have dropped there were numerous food vendors who sold ot cooked food right at their doorsteps. Food com- out. monly sold included chapati, mandazi and parts of The majority of women (56%) cited lack of chicken ( mainly the heads, legs and intestines school fees as the reason children had to drop which the sellers obtained from large-scale chicken- out. These families faced a number of other meat processors). Also on sale for consumption constraints as well. The women spoke of not were the hoofs of animals like pigs, goats and cows. having enough money to buy school uniforms, The most commonly used fuel sources were pay for the "building fund" or purchase statio- firewood, charcoal and paraffin stove (Table 21). nery. In some instances, when the household firewood, charcoal and paraffin stove (Table 21). ddnthv nuhfod hlrnrfsdt Paraffin was considered the least expensive got h ool, hg itsel beame a fuel, although most respondents seemed to go to school in hung itself school prefer charcoal. Firewood was a constraint to significant factor in "dropping out" of school. some respondents because their landlords did The average cost of maintaining a child in not alow its use. nursery school ranges between Ksh 100 and Ksh 500; in primary school, Ksh 501 -1,000; and in Some respondents described their cooking secondary school, over Ksh 1,000 per year (Table facilities as barely adequate. Those with large 22).n(at theoim ofe study te ean rate houseolds ould ot us parafin soves22). (At the time of the study the exchainge rate houseoldscoul notuse arafin soveswas Ksh 63 to US1$.) The women indicated that because of the amount of food they had to cook. they ri to vus1 worg niz ati f s t, Those households resorted to using more they relied on various organzations for support, expensive charcoal. Overall, most respondents such as bursaries from the District Development Committee, the Redeemed Gospel Church of e fw ( a Kenya, the National Council of Churches of (5%)°J Kenya (NCCK), Action Aid Kenya (AAK) and World Vision International. Table 21: Sources of Cooking Energy in Mathare Health Valley and Korogocho Slums as Common childhood diseases include diarrhoea, Prioritized by Residents (%) measles, vomiting, coughing, kwashiokor, marasmus, tuberculosis, worms and dysentery. _ oe Sr2 reThe youth suffer from stress, worms, typhoid, None 2 0 O scabies, nervous breakdown, drug abuse, gun- Firewood 59 5 17 shot wounds, infections and bleeding from Charcoal 25 55 83 abortions, non-pill related bleeding, contracep- Somebody Else 0 3 0 tive pill complications and sexually transmitted Paraffin 5 24 0 diseases. Parents had stress, blood pressure, Other 9 13 0 sexually transmitted diseases (including HIV/ Participation Series 55 Learning from the Poor Table 22: Cost of Educating Children Institutiona! Support in Mathare and Korogocho Slums (%) The area is served by both local and external agencies. However, local groups have been Cost Cost of Transport plagued by problems including theft of assets, Category Fees Books Cost often by the group's treasurer. Other women's Less than Khsh 100 1l2 7 12 groups have raised fees, effectively barring 101-500 Ksh 18 15 5 very poor women from joining. The National 501-1000 Ksh 8 8 5 Council of Churches of Kenya, an NGO, was 1001+Ksh 15 18 5 cited by many respondents as "being of great Ext. Asst. 18 22 14 benefit." The group counsels women, pays Do Not Know 29 30 59 school fees, buys school uniforms, gives loans to promote businesses for women who are heads of households, provides a feeding program for malnourished children and AIDS), joint pains, blockage of urinary tract, supports vocational training for youth. A few tuberculosis, syphilis, gonorrhea, stress, orts povided similar buthfewer ' ............... . other NGOs provided similar, but fewer, headaches, ulcers, pregnancy, delivery compli- services. cations, breast cancer and fungal infections. Despite the presence of NGOs, the respondents Among the ways that residents of Mathare copaie thtteNO seeob Valley and Korogocho attended to their illnesses complained that the NGOs seemed to be were hers (%); rug (12); uinga trdi-helping the same groups or households and -were herbs (2%); drugs (12%); usdin a were not getting to those who were most in eional healer (5%); prayer (5%); and visitig a need. The NGOs had little coordination with each other which created problems in providing Livelihood services. Of the women in the study, 85% engage in Whereas there is a Community Hall in Mathare some sort of small income generating activity Valley, Korogocho is without one. However, or business. Several of the women are involved there are associations for youth, women and in more than one income generating activity. men. The principal aim of these associations is Many women engage in hawking goods at the to combat poverty and deprivation. One markets or along the roads. They often sell example is the Mathare Youth Sports Associa- fresh foods or rejects from the larger markets tion. Groups of business women perform (leftovers - usually the poorest in quality - "merry- go-round" contributions on a weekly are fed to the children). Casual labor is another or monthly basis to promote their businesses or income-generating activity. These women buy household utensils. There is also an usually work in quarries, construction, water association of inen who sold goats. fetching, trench digging in city council works, factories and Kasarani farms. Other women do Perceptions of Poverty by Women laundry for the high-income communities. Residents of Mathare Valley and From these activities, women can raise between K Ksh 20 and Ksh 50 per day. Korogocho vlums Women's perception of the causes of poverty The remaining women were unemployed, covered a range of issues including: housing, begged, scavenged or received minimal help employment, food, health and clothing, cost of from relatives or the Catholic church. One education, orphans, waste disposal systems, woman was a prostitute. prevalence of illicit brewing and alcoholism, 56 Environment Department Papers Urban Nairobi disability, large families (many children), lack of (57%). The number of medium and rich opportunities for slum dwellers to participate in residents in the slums was negligible. The decisionmaking and prioritization of develop- majority of women (63%) felt that they were ment projects, lack of entrepreneurial skills, poor; 29% felt that they were medium; just 2% marital problems and lack of land ownership felt that they were rich. rights. The very poor women household heads had no Some felt that the people of Mathare Valley and source of income, no husband nor any property. Korogocho slums had self-imposed poverty and Their families were large and they usually had that there was only a small number of genu- very young children. In order to eat, they inely poor people. However, most were of the squatted and begged food and clothing. The impression that there were indeed pockets of parents had tattered clothes; the children were poverty in the valley, especially among single- malnourished. Those with some source of headed households with large families. The income could afford one meal a day, clothing women felt, however, that a large number of and rent. However, they were not able to take households were poor due to alcoholism and sick family members to a hospital or pay school laziness. fees. Both the poor and very poor said they rely on prayer to solve their immediate and long- In Mathare Valley, 74 % of the women under- term problems. stood poverty as meaning families had no money to meet basic needs such as food, Conclusion education expenses, adequate housing and health services. In Korogocho, the same The study has shown that the majority of definition was held by 56% of the respondents. women who have come to Nairobi's worst Length of stay in the valley was not a factor slums are either single (never married), di- cited by people as a cause of poverty. vorced or widowed. Most have several chil- dren. As a result of their marital status, the According to the women interviewed, 97% of women look for different ways and means of the residents were poor (40%) or very poor livelihood. Box 29: Role of NCCK in Mathare Valley and Korogocho Slums The goal of National Council of Churches of Kenya (NCCK) is based on the Biblical book of John 10:10: "That they may have life and have it in abundarce." NCCK's Urban Community Improvement Programme (UCIP), provides emergency relief, organizes groups of women for self help and income generating activi- ties, conducts training on using low-cost materials to improve housing and sponsors a center for nutrition and family education to fight malnutrition, The Council's Women and Children in Stress Project was started in 1991, following various government and NGO studies which identified women and children as the most vulnerable to the social, economic and psychological problems associated with urban life. The project aims to alleviate stress by providing: coun- selling to women, children and youth; educational sponsorship and uniforms to children from stressed families; assistance and skills training for youth and school drop outs; and credit facilities for small business as a way of giving women a source of income and enhancing their standard of living. The Maternal Child Health and Family Planning project was begun in 1987. It provides credit facilities to women, skills training to youth and community health and population education, including delivery of contraceptives. These activities are spread throughout Mathare, Huruma, Kariobangi and Korogocho slums. Participation Series 57 Learning from the Poor The study revealed that the women of Mathare children mostly eat at the feeding centers. Valley and Korogocho believe that people are poor if they cannot earn enough money to The poor, on the other hand, were seen to be enable them to feed their families, provide those female-headed households who con- decent housing and keep their children in ducted small business or casual work at the school. Women used two categories to classify homes of well-to-do families mainly in those living in poverty: "poor" and "very poor." Eastleigh, or worked in people's plots in The very poor are those who are handicapped Kasarani, or collected tins, paper and bottle- physically or mentally, including alcoholics. tops which could be resold to earn just enough According to the women, the very poor make money to buy one meal each day. no effort to feed their children, hence the 58 Environment Department Papers 6. Mandera Districit Study Mandera is poorer than any other district in Who are the Poor and Why are they Kenya. Nearly 92% of the people have no Poor? education, contrasted with 52% in all of Kenya. The PPA found that 100% of women heads of When asked who were the poor in the area, households rank themselves as "very poor" and most village residents replied, "We are the 55% of the total population similarly classify poor." A total of 77% of people classified themselves as very poor. In Kenya as a whole, themselves as "poor" or "very poor," compared the percentages are 21 % and 37%, respectively. with an average of 63% in the other districts. Several features distinguish Mandera from the Apart from the emphasis on drought and agricultural districts studied in the PPA. First, it migration, reasons given for poverty closely is a semi-arid desert area; during the long dry paralleled those in the other districts. The cost seasons, the nomads take their livestock and of living was said to be on the rise, with the move to where the grass is, often in neighboring price of 1 kilogram of sugar given as Ksh 10 in Ethiopia. Villages in Mandera become com- 1990, Ksh 25 in 1992, and Ksh 50 in 1994 (an pletely deserted during these dry period increase of 100% per year and 400% in four migrations. This creates acute problems for years). The cost of a camel had also risen to Ksh school headmasters who are trying to maintain 10,000. schedules. The status of women was said to have changed. Another striing feature iS that villages in In the past, the woman's place was in the house Mandera are far removed from each other. One lookin after the children. Due to increasing of the villages studied was 300 kilometers from goverter the now Due to increasing Distict eaduartrs.Accodingto Droght poverty, women now have to fend for themselves District Headquarters. According to a Droughl: alongside the men, including doing odd jobs, Baseline Study by the Central Bureau of Statis- working as servants for the well-to-do, collecting tics and UNICEF, after the 1991-92 drought, 15/0 firewood and gum arabica, hawking and begging. of the population had to walk five days or more to reach a health facility. The prevalences of The children showed a clear understanding of stunting and malnutrition among children were poverty. They viewed a poor person as one 51% and 62%, respectively. who has no animals, no shamba, no job and no Minimal agriculture is practiced along one river. source of food. They also understood that The extreme poverty in the area was attributed poverty is brought about by epidemics and by people to heavy losses of livestock (camels, children orphans. cattle, sheep, goats, and chickens) and crops during the 1984 and 1992 droughts. During the Coping Strategies two brief rainy seasons, malaria becomes a problem. The "long" rains normally fall in April- The principal coping mechanism mentioned in May, the "short" rains in October-November. Mandera, apart from seasonal migrations, was Participation Series 59 Learning from the Poor food relief provided by government and A mobile immunization program was initiated NGOs. Donors and assistance from the in 1987 by the Ministry of Health. Communi- Government of Kenya were mentioned more ties are quite satisfied with the mobile group frequently in Mandera than in the other six which is responsible for distribution of drugs districts studied and immunizations free of charge. Field officers educate the community on health- Other coping mechanisms mentioned were related proble:ms. Since 1992, agencies like doing odd jobs, hawking in the larger centers, UNICEF and MdDM (Spain) have been helping working for the well-to-do, begging and child in the programs. labor. The Arda-Kalacha Women's Group was started In meeting health and education needs, the seven years ago. Membership is open to all and Koran was mentioned, including attending there are 30 members. The group has had Koranic schools and reading passages from the projects in small-scale cereal farming, poultry- Koran over the sick. Other strategies included raising and goat-rearing. While farming efforts marrying off daughters in school to save on fees have had some success, poultry disease wiped and make room for boys. In some cases, child out a majority of the hens (but the project is labor kept some children out of school while continuing) and the 1992 drought killed most of they earned extra money to support their the goats (the r emaining few were sold). De- siblings who were attending school. spite hardships, the group continues to assist its members and other beneficiaries, - including The principal strategy for accessing water was destitutes, the sick, aged, blind, disabled and trekking long distances, often 20 to 30 kilome- students from poor families. ters, to find a water source. Assistance was rendered by the use of camels or donkeys to Gender Analysis fetch the water. In Lambur village, the following picture was District Officials & Service Provision given. Within a family, the man owns camels, sheep, goats and cows but not the chickens. He The study found that district decisionmakers also owns the :radio, land and farm implements. seemed to have a very limited view of the The woman owns kitchen utensils, her orna- extent and nature of poverty in the district. ments and the baby. Food and the house fall Some leaders, for instance, viewed the poor as under joint ownership of the couple. concentrated mainly around town centers like Mandera and felt that this would allow the poor If the man dies and the widow is childless, she to get food from distribution centers. retains her ornaments, kitchen utensils and food. If she has children, she splits land and money The situation in Arda-Kalacha village with respect with her parents-in-law and brothers-in-law. If to services provision is fairly typical. The nearest she has sons, she retains almost all the property, school and health center are both 18 kilometers which eventually is taken over by the sons. away; the closes watering point is 9 kilometers away; district headquarters is 180 kilometers away. Education in One Village Although some basic infrastructure exists, it is poorly maintained and during the rainy season the Kiliweheri village is in extreme western roads are flooded. The headmaster said a few Mandera district, 300 kilometers from District problems facing his school included lack of Headquarters. The only formal education is a teachers, desks and even walls for the school. primary school which has never gone beyond 60 Enviromnent Department Papers Mandera District Study Standard 5 due to lack of water. Because the greener pastures. Just before Ramadan, a large village is deserted during the dry season, number of weddings take place and hence an students and teachers are forced to shift to increase in the number of girls who drop out of other locations before the end of the term. school to get married. Nevertheless, the school According to the headmaster, in addition to has made some advances. Students have been lack of teachers and lack of desks, the school able to improve their performance in the needs a workshop, more books, a latrine, a national Kenya Certificate of Primary Education dining hall, protection from ants which have (KCPE) exam, and the 4K Club has been able to destroyed the bookstore and protection from attend a few seminars on soil conservation. villagers who use the school fence for firewood. There has been an increase in the number of students over the past 10 years, Most teachers In the dry season "camels before education" is hate to be posted to Mandera district, and the motto for nomads who withdraw their therefore the few who accept are considered children from school and migrate in search of "angels" by the community as a whole. Participation Series 61 7. Afterword "We didn't trust the PPA process. Now we understand it, accept it and it has become ours. " - Mr. Osundwa, Director Ministry of Planning Government of Kenya The purpose of the Kenya PPA was three-fold: Initially, the credibility of methods is insepa- first, to understand poverty from the perspec- rable from the credibiLity of the researcher. tive of the poor and those who provide services Hence, the fact that the World Bank is now to the poor; second, to bring this knowledge to promoting the use of these methods creates an policymakers to influence their attitudes and opportunity to test these methods to address decisions; and third, to begin a process of national policy issues. dialogue between policymakers, district-level service providers and the poor. The many Despite a new openness to try a participatory detailed PPA findings based on the knowledge approach to understanding poverty, there is of the poor have been used a]Lready by different still skepticisrn about the utiLity of the approach. stakeholders to make po Licy. This creates its own pressures. In this environ- ment, it is critical to Listen, learn, be patient and The PPA idea and process is new and still in the spend time with the Poverty Assessment task stage of invention and learning by doing. manager, other decisionmakers and to remain However, the methods and prmciples of in a consultation "stance" until the reports are participatory research are at least two decades completed and cleared. old. It is particularly important to document At the end of this process, in the Kenya case, the lessons learned about process so that the the task manager said "I do my statistical learning with each PPA conducted is cumula- anysis mandage back to wa stthostays tive. analysis and go back to what the poor say as a check. In many cases, the PPA provided impor- This section of the report presents the six main tant information which I am convinced cannot lessons that emerged. be obtained from surveys - for example, the information about the poor's perceptions of Establish Credibility poverty, distrilbution of property on divorce, the detailed information on school dropouts and how people cope. In other cases, the PPA ther wrk twitheonomiss ahihned stasc refuted or confirmed the numbers we had from there aeatoobstacesuwhin toe the statistical survey, on which I did not have addressed head-on to ense That e findng full confidence because of the many problems the credibiLity of the methods and researcher, experienced in doing the survey. For example, andsecond,the redibilityhiof the mhsnresearsch, the PPA pointed out the greater poverty among and second, the degJree to which the results can fml-eddhueod n lopoie be genealized female-headed households and also provided be generalized. 62 Environmnent Department Papers Afterword very convincing information about the extent of some staff from the Bureau of Statistics in field the water problem, neither of which were work, and visits by the Director of the Bureau captured by the survey." to the PPA training and with a research team during field work. A proposed film of the To address the issue of generalizability, much process was cancelled. time was spent with statisticians. The sample selection of villages was done jointly by a group Tensions and Turning Points of government and Bank statisticians and Kenyan anthropologists using the sample One of the difficulties encountered in the study framework set up for the census. The Welfare was the challenge of ensuring communication Monitoring Survey was used to anchor the between people whose research and working content and sampling of the PPA study. Based paradigms are quite different. Initially, dia- on the results of the WMS, six of the pooresi: logue with the planners would get stuck on the districts were chosen to represent different issue of comparability of subjective poverty agroecological zones. Within each district, the definitions across villages. (Since a PPA had village clusters or "enumeration" areas were never been done in Kenya, there were no chosen randomly. Adoption of this process precedents.) There were also tensions between meant that when the results became available anthropologists and PRA trainers within the statisticians and planners were interested and PPA team. Lastly, there were tensions between took the findings seriously. government officials and AMREF, the regional NGO which had been contracted to manage the Don't Hurry the Process to Meet PPA. Deadlines The turning point with senior planners and Hurrying process beyond a certain point kills statisticians came when they were shown some the process and creates a false sense of partici- of the visual materials which had been devel- pation. Most PPAs until recently have been oped to gather information of priority problems done in a hurry to gather data in time to feed and gender analysis. Seeing the materials led to into the Bank's poverty assessment documents, appreciation and understanding of the process which in turn are meant to determine the and its possibilities. Country Assistance Strategies. The Kenya F'PA was conducted under the same time pressures: Tensions between anthropologists and partici- the study was conceived, funds raised, people patory trainers were diffused by creating a identified, team trained, data collected and training team involving all the senior supervi- analyzed and draft reports produced in just sors including the statisticians, and then having four months. each person take the lead in the area of their expertise, rather that simply putting the partici- Obviously, shortcuts had to be taken. While the patory trainers in charge. participatory processes at the community level were given adequate attention, other areas The turning point for the anthropologists who suffered. Workshops to involve district offic:ials, felt they knew how to work with village people for instance, were cut back. Most neglected was and hence resisted trying the visual SARAR the continuous interaction and feedback to methods was field testing of these methods as senior government officials in the Bureau of part of the training process. Once again, seeing Planning and Central Bureau of Statistics. The was believing, and as a result the team reached price of neglect was high: The Government of consensus on a vision and strategy for combin- Kenya initially distanced itself from the PPA ing different methods with the PPA. report and findings despite involvement of Participation Series 63 Leaming from the Poor Building Capacity While Rushing officials and AMREF. Fortunately, a subsequent PPA in Tanzania provided the opportunity for Sustainable change needs to be rooted in local Bank staff to stop over in Kenya to continue a expertise and institutions. However, when dialogue with government officials. Through PPAs are put together at short notice, it can be these discussions agreement was reached to difficult for in-country participatory trainers bring the research team back together and hold and researchers to drop everything they are a national PPA seminar hosted by the govern- doing to participate. In the case of Kenya, while ment to discuss and disseminate the PPA the country has at least two networks of partici- findings. patory trainers, only one person could make herself available for the PPA - and this only As a result of this follow-up work, the Govern- after taking the drastic action of quitting her ment of Kenya eventually embraced the PPA job. approach. With government support, an in- country Kenyan research team financed by As the demand for participatory skills in- ODA is presently in the field doing the second creases, international organizations need to take round of a national PPA in conjunction with the a long-term view and invest in capacity build- National Welfare Monitoring Service. The ing in-country. PPAs, for example, require Bureau of Statistics are developing a core PPA professionals with technical skills in participa- module to monitor poverty. Once again, tory methods as well as institutions which can AMREF is coordinating the PPA for the Gov- manage large research projects involving ernment of Kenya. extensive travel. Involve Other Agencies Internalization Requires Time and Follow-up In addition to the Government of Kenya and the World Bank, there were a number of other key The lowest point in the Kenya PPA process stakeholders in the PPA. came when the Ministry of Planning distanced itself from the study in September 1994. This The process of consultation not only raised resulted from breakdowns in communication interest in the study but also quickly generated which could have been avoided. While frequent funding commitments. Both ODA and UNICEF contact between the study team and the Minis- cut through their bureaucracies to commit try was maintained in the early stages, there funds, although the actual flow of funds took was almost no contact once field work started. much longer. After field work was completed, presentations were made to ODA and UNICEF Worse, government representatives received at their request. ODA used the findings to start the green cover poverty assessment report more concentraited work to address urban which incorporated the PPA findings before poverty issues and to think through its strate- they had seen the PPA report itself. Confronted gies for working with NGOs. UNICEF with PPA findings based on processes and strengthened it work in Kisumu and evaluated research methods they did not fully understand its approach in Mandera. and could not explain to their superiors, they initially chose to distance themselves officially AMREF already had a participatory trainer from the study. working on health issues who unfortunately could not be involved in the study because of From the Bank perspective, once the PPA task prior commitments. However, AMREF's was complete, no time was made available to involvement in managing the PPA resulted in follow up or sort out issues with government support by their Director for adopting the 64 Environment Department Papers Afterword participatory approach into its other work. pologist at the university has adapted the The PPA research teams have been used since SARAR tools to AIDS and primary health care by UNICEF and UNDP. The World Bank programs; the only economist in the team education sector staff are conducting particilpa- returned to Kisumu and trained his NGO staff tory assessments to assess early childhood in the PPA methodology. education needs. Some of the senior research supervisors have adapted many of the SARAR In Kenya, the work continuous at many levels.... tools to their other work. The senior anthro- Participation Series 65 20 ZU473 36A1 202 A77 - 056 FAX0 0; X 3Printecl on 100E Postconsurner rqcycled.paper.;0(gX-0 :;