" 9 ~22812 Local Level Institutions Working Paper No. 8 Local Level Institutions THE LOCAL LEVEL INSTITUTIONS STUDY: LOCAL INSTITUTIONS AND SERVICE DELIVERY IN BURKINA FASO Anand Swamy, Christiaan Grootaert, and Gi-Taik Oh The World Bank Social Development Family Environmentally and Socially Sustainable Development Network September 1999 LOCAL LEVEL INSTITUTIONS WORKING PAPER SERIES #1 The Local Level Institutions Study: Overview and Program Description (also available in Spanish and French) #2 The Local Level Institutions Study: Program Description and Prototype Questionnaires #3 Estudio Sobre Las Instituciones Locales En Bolivia: Panorama General Y Descripcion, Muestreo Y Recoleccion De Datos, Y Encuestas (by Godofredo Sand6val, Julio Cordova, Beatriz Ascarrunz, Alfredo Balboa, Griselda Gonzales, and Gloria Velasques) (only available in Spanish) #4 Grassroots Organizations and Local Development in Bolivia: A Study of the Municipalities of Tiahuanacu, Mizque, Villa Serrano, and Charagua (by Godofredo Sand6val, Julio Cordova, Beatriz Ascarrunz, Alfredo Balboa, Griselda Gonzales, and Gloria Velasques) (also available in Spanish) #5 Local Institutions and Service Delivery in Indonesia (by Christiaan Grootaert) #6 Social Capital, Household Welfare and Poverty in Indonesia (by Christiaan Grootaert) #7 Social Capital and Development Outcomes in Burkina Faso (by Christiaan Grootaert, Gi-Taik Oh, and Anand Swamy) #8 Local Institutions and Service Delivery in Burkina Faso (by Anand Swamy, Christiaan Grootaert, and Gi-Taik Oh) #9 Local Institutions, Poverty and Household Welfare in Bolivia (by Christiaan Grootaert and Deepa Narayan) #10 Does Social Capital Help The Poor? A Synthesis of Findings from the Local Level Institutions Studies in Bolivia, Burkina Faso and Indonesia (by Christiaan Grootaert) Working papers can be viewed at _tt14p:l.worldbagkpousoc ialdevelopment, or obtained from: The World Bank Social Development Department Social Capital Working Paper Series Attention Ms. Gracie M. Ochieng 1818 H Street, NW, Room MC 5-410 Washington, DC 20433, USA Tel: (202) 473-1123 Fax: (202) 522-3247 Email: gochieng( dorldbank.org or: The World Bank Social Development Department Publications 1818 H Street, NW, Room MC 5-232 Washington, DC 20433, USA Fax: (202) 522-3247 Email: sd p ns(vworldbank.org Papers in the Local Level Institutions Working Paper Series are not formal publications of the World Bank. They are published informally and circulated to encourage discussion and comment within the development community. The findings, interpretations, judgements, and conclusions expressed in this paper are those of the author(s) and should not be attributed to the World Bank, to its affiliated organizations, or to members of the Board of Executive Directors or the governments they represent. LOCAL INSTITUTIONS AND SERVICE DELIVERY IN BURKINA FASO TABLE OF CONTENTS ACKNOWLEDGMENT ................................................ ii 1. INTRODUCTION: LOCAL INSTITUTIONS AND SOCIAL CAPITAL . 1 2. SOCIAL CAPITAL IN BURKINA FASO: THE EXISTING LITERATURE . 3 3. THE DATA SET .5 4. THE MAP OF LOCAL INSTITUTIONS .14 5. CHARACTERISTICS OF THE MOST IMPORTANT ASSOCIATIONS .23 6. SOCIAL CAPITAL AND POVERTY .26 7. WHO PROVIDES SERVICES? .28 8. COLLECTIVE ACTION AND TRUST .31 9. SUMMARY AND CONCLUSIONS .36 ANNEX: TABLES BY PROVINCE .39 REFERENCES .41 i ACKNOWLEDGEMENT The authors would like to thank Anthony Bebbington and Paula Donnelly-Roark for helpful comments on an earlier version of this paper. The Local Level Institutions (LLI) study was conducted under the leadership of Gloria Davis, Director, Social Development Department. In the initial phases, task Managers were Anthony Bebbington (July 1995-June 1996) and Christiaan Grootaert (July 1996- December 1997). The final phase of the study was undertaken as a joint venture between the ESSD and PREM Networks and co-managed by Christiaan Grootaert and Deepa Narayan. The study received financial support from the Government of Norway. The Bolivia country-study was undertaken by a team from the consulting firm Sinergia. Coordinator was Godofredo Sand6val and the research team consisted of Julio Cordova, Beatriz Ascarrunz, Afredo Balboa, Griselda Gonzales, and Gloria Velasquez. The Burkina Faso country-study was coordinated by Paula Donnelly-Roark. The field work was undertaken under the auspices of the Commission Nationale pour la D6centralisation. The Indonesia country-study was coordinated by Scott Guggenheim and the research team consisted of Kamala Chandrakirana, Pieter Evers, Sjari Manaf and Silvia Werner. The processing and management of the data files was the responsibility of Gi-Taik Oh and Kalpana Mehra. General research assistance was provided by Susan Assaf. Staff assistants for the study were Gracie Ochieng and Anju Sachdeva. ii I. INTRODUCTION: LOCAL INSTITUTIONS AND SOCIAL CAPITAL As the world's technological possibilities continue to expand, the problem of underdevelopment becomes even more of a puzzle. Given available techniques and resources, we have the capacity to, at the minimum, eliminate hunger and other extreme manifestations of poverty. Why are opportunities being missed? The proximate causes are easy to see: low productivity, over-population, low levels of education, etc. However, though these are all valid explanations in at least some contexts, economists are increasingly seeking to go beyond them. While there are no "ultimate" causes to be found, many economists now believe the problem of underdevelopment is to a significant extent a problem of underdeveloped institutions (Olson, 1982; North, 1990). Some examples may clarify the point. Consider the problem of low productivity in agriculture. The immediate cause may be that the technology is obsolete and unproductive. But why doesn't the farner introduce a better technology? One reason could be that s/he lacks the necessary funds. Why doesn't s/he borrow them? Possibly because banks lend only against collateral, which s/he does not have. The solution to the problem of low productivity then lies, at least in part, in the development of lending institutions which do not rely on collateral for loan recovery. Alternatively, the farmer may lack the skills necessary to use new techniques. The absence of these skills may be related to the availability and quality of education provided in the (typically) government-run schools, or the absence of appropriate "extension" programs to spread knowledge of new techniques. The solution may lie in improving the performance of these government institutions. As the two examples provided above indicate, economic development requires not only the presence of appropriate institutions, but also their effective functioning. Following the work of Putnam (1993) many now use the term "social capital" to describe development- enhancing institutions as well as the norms and values that sustain them. There are many debates around the precise definitions of "institutions" and "social capital" and the extent to which these ideas overlap. We are interested in both and, at least for the purposes of this paper, will not attempt to disentangle them conceptually'. Our task is to document local institutional development/social capital in the rural areas of Burkina Faso, and examine their relationship with economic and social outcomes. We use survey data collected as part of the World Bank's Local Level Institutions (LLI) Study. We have divided our work into two parts. In this paper our goal is mainly to describe the characteristics of households, their performance or outcomes in various economic and social dimensions, and their participation in associational activity. Our second paper is more ambitious: it investigates the causal relationships between associational activity/social capital and economic and social outcomes, using multivariate regression analysis (Grootaert, Oh, and Swamy, 1999). See Grootaert (1997) for a discussion. 1 The remainder of this paper is organized as follows. We first briefly discuss what we have learned about social capital in Burkina Faso from the existing literature (Section 2). The third section describes the data sources and summarizes the current situation in Burkina Faso with regard to various social and economic outcomes that are of interest to us. In this section, the variables we will look at, such as per capita expenditure, poverty, etc., have long been of interest to economists, and pose no special difficulties. The measurement of social capital, however, is a much more difficult task. Our goal is to measure a community's ability to cooperate for mutual benefit: this is surely a difficult characteristic to quantify. Following Putnam's pioneering work (1993), however, one approach has appealed to many researchers. A community's ability to cooperate should be reflected in the formation of associations which, directly or indirectly, promote the common good of their members. These associations can cover a range of activities, from groups of people coming together to pray, to play cards, to discuss agricultural practices, to share credit, etc. The fourth section of this paper, therefore, provides a "map," a description of the number and types of organizations that exist. The fifth section discusses the characteristics of the most important associations in greater detail. In the sixth section, we provide descriptive statistics on the relationship between social capital and poverty/income. The seventh and eighth sections look at the provision of social services and collective action/trust, respectively. The final section summarizes and concludes the paper. 2 2. SOCIAL CAPITAL IN BURKINA FASO: THE EXISTING LITERATURE There is, so far as we know, no existing literature (prior to the LLI study) that specifically addresses the issue of social capital in Burkina Faso. Several authors, however, have commented on aspects of social relations that are relevant. The following facts or claims stand out in this literature: highly developed traditions of cooperation within ethnic groups, harmonious relations among ethnic groups, and a disjunction or gap between the state and civil (especially rural) society. We briefly discuss each theme below. The most numerous and best documented ethnic group in Burkina Faso, the Mossi, are said to place a very high value on sharing and mutual support. In his book Structures of Social Life, anthropologist Alan Fiske (1991) describes extensive cooperation among the Mossi with regard to allocation of land, water, mutual insurance2, and pooling of labor3. He emphasizes that common good, rather than individual benefit, is affirmed in such activities [p. 268]: The Moose [Mossi4] preference for Communal Sharing does not seem to be driven by technological and utility-maximization constraints.. .That the Moose pool their labor collectively, and share their food, often in the face of incentives to do otherwise, demonstrates that their primary motivational orientation is toward corporate participation and belonging, and that their paramount goals concern mutual solidarity, a sense of common identity and belonging, unity and kindness. Harmonious relations between ethnic groups in Burkina Faso have also been attributed, at least in part, to Mossi norms. Englebert (1996, 125) argues that Burkina Faso's "relative uniqueness" in the African context (in that there is little conflict among ethnic groups) is in part due to the "assimilationist nature" of the Mossi, and their tendency to de- emphasize ethnic differences. He also suggests that the state has distanced itself from ethnic authority and thereby reduced ethnic competition for "ownership" of the state. Finally, the paucity of economic resources at the disposal of the state itself means there is less to compete over. The relative lack of ethnic conflict and other factors favorable to development are also discussed by Kevane and Englebert (1997). However, the distance between the state and ethnic authority, which is cited approvingly by Englebert, may be partly responsible for the complaints voiced by Maclure 2 Given the variability of incomes in Burkina Faso (Reardon, Matlon, and Delgado, 1992), much can be gained from mutual assistance. This has been demonstrated by Carter (1997), using a detailed household level data set. Carter finds that a typical household, if it is entirely dependent on self-insurance, will in a given year face a subsistence crisis with probability 0.21. However, if 10 households form a network and agree to donate (if necessary) all of their above-subsistence income to a household which is falling below subsistence, this probability falls to 0.16. 3 For a discussion of crisis-support within "compounds" or extended families, see Veirech (1986). 4 Mossi is the more common spelling; Fiske is following the recommendation of a language commnission. 3 (1994) and Spiers (1991) to the effect that rural communities are cut-off from the decision- making of governmental institutions. Maclure's argument (1994) is developed in the context of decentralization of education in Burkina Faso, which has involved an effort to include communities in the management of primary schools (World Bank, 1991). He argues that this is a very difficult task in Burkina Faso's rural communities which "all too frequently are impoverished and politically marginalized." He suggests that governmental institutions and resources tend to be dominated by elites and that "the majority of the rural populace in Burkina Faso is restricted from participating actively in decision-making." (p. 249). Our reading of the existing literature suggests that the picture with regard to social capital may be mixed. Social capital within communities may be at a high level; on the other hand, there may be a lack of social capital as embodied in relations between communities and state institutions. In this context it is important to note that a major new initiative of the government of Burkina Faso addresses precisely this issue. Burkina Faso has recently embarked on a wide-ranging program of decentralization with the specific objective of involving rural indigenous institutions, so that the majority of the rural population can be engaged in the task of economic development. The purpose is to combine the strengths of formal state agencies and indigenous institutions that are deeply rooted in communities. In one sense the decentralization can be viewed as an attempt to bridge the gap between the state and communities, which was referred to above5. The LLI Study was conceived as a broad survey of social capital, not specifically meant to address the three themes discussed above; still, we will comment below, to the extent possible, on how much the survey data are consistent with anthropologists' claims. The reader will notice that the literature's claims about co-operative behavior among the Mossi are only partly borne out. Our analysis does not reveal a clear pattern; the Mossi seem to be more co-operative than other ethnic groups in some dimensions and less so in other dimensions. We now turn to a description of the survey data. 5 The new Decentralization Law, passed in August 1998, was based on the recormnendations of the Commission Nationale de la D6centralisation (CND). It allows groups of villages to come together and request Commune Rurale status from the government. It also allows any group, such as a Parent-Teachers' Association, for exarnple, to obtain legal recognition, and enter into contracts with other legal entities. Our description is based on Donnelly-Roark et al. (1999), which provides a more detailed account of the decentralization law. 4 3. THE DATA SET The data set for this paper comes from the Local Level Institutions (LLI) Study, a comparative study of three countries (Bolivia, Burkina Faso and Indonesia), which aims to investigate the role of local institutions in providing service delivery and in affecting welfare and poverty outcomes.6 Data were collected at the level of the community, the district, and the household. At the level of the community, interviews with focus groups of households and with community leaders were held to establish a map of functioning institutions in the community. Three instruments were used: * Information on community services was obtained through interviews with key informants such as the village chief, teacher, health provider, etc. This was supplemented with information on the local economy (infrastructure and distance to markets), local society (ethnic/religious composition) and local institutions. * The community services were also discussed with groups of households, with the objective of learning the community's perspective on the quality of service, its experience with collective action, and its views on local institutions and development projects. * For the most important local institutions, interviews were held with leaders and members, as well as non-members, in order to get a balanced view of the role of the institutions in the village, their development over time, their main activities, relations with other institutions and government, and their main strengths and weaknesses. At the district level (defined as the administrative level above the village or community), data were collected about the extent of service coverage and the institutional arrangements for the provision of services. Information was also obtained about the general functioning of the district administration and its relation with civic organizations, through interviews with general and sectoral managers at the district level. The third and critical part of the data collection was a household survey which aimed to capture households' actual participation in local institutions, their use of services, and information that identifies the welfare level of households and their coping strategies. The questionnaire consisted of six sections: 6 The objectives of the Local Level Institutions study and the questionnaires are discussed further in World Bank (1998). 5 * demographic information on household members * participation in local institutions * characteristics of the most important groups * service provision profiles * perceptions of community trust and collaboration * household economy and coping strategies. The limited resources available did not make possible a sampling framework such that the studies would be representative of the countries at the national level. Instead, three or four areas were selected in each country (municipios in Bolivia, provinces in Burkina Faso and Indonesia), which represent different economic, social and institutional environments. In the case of Burkina Faso, the data collection covered the rural areas of four provinces: Yatenga, Sanmatenga, Houet and Sissili (Table 1). These four provinces were selected to represent different regions of the country and different social, economic and institutional settings (CND, 1998). The next three paragraphs provide brief descriptions of each province. 7 Yatenga and Sanmatenga provinces are situated in the North of Burkina Faso, are predominantly populated by the Mossi, and are largely Muslim. Yatenga is the traditional seat of the Mossi Kingdom. Both provinces have relatively high population density. Both provinces suffer from low rainfall, which makes for low productivity agriculture. Livestock raising however is prominent. Mossi social organization is quite hierarchical in a formal sense, but relatively flexible in terms of daily practice. Houet, which lies to the West, is culturally and ethnically diverse. The dominant ethnic groups have non-hierarchical and flexible elders councils. Houet is a high rainfall area and is often called the granary of Burkina Faso. Human development is at a relatively high level. Houet draws immigrants, both farmers and herdsmen, from the North. Farmer-herder relations can involve conflict, with disputes arising over land and water. Sissili, which lies to the south-east, is the home of the Gourounsi. Like in Houet, non-hierarchical and flexible elders' councils are active; in Sissili these councils have promoted various development initiatives. Due to high levels of rainfall, Sissili is able to grow a variety of agricultural products, and attracts Mossi from Yatenga and Sanmatenga who are in search of arable land. 7 These are based on CND, 1998 and Donnelly-Roark et al., 1999, which contain extensive descriptions of each province. 6 Table 1: Selected Socio-economic Indicators of the Four Study Areas Yatenga Houet Sissili Sanmatenga All Population 443,935 883,320 153,627 469,684 11,000,000 % Urbanized 12% 46% 12% 10% 30% Literacy rate 16% n. av. n. av. 10% 9% % of Heads of Household who are 91 76 85 94 86 Farmers Household Expenditure per Capita 58,712 69,768 71,686 60,886 65,265 " ('000 CFAfr/year) Gini-coefficient 0.40 0.33 0.27 0.26 0.32 1/ Based on the four study areas only. Sources: Donnelly-Roark et al. (1999) for first three rows of provincial figures; World Development Report (1997) for first three figures in All column; remainder computed from survey data. Within each province, one administrative department was selected for the data collection (in Houet, two departments were selected). In each province, twelve villages were chosen purposively based on four criteria: organizational level, economic situation, cultural diversity, and proximity of services. Within each village, 20 households were selected randomly to participate in the survey, leading to a total sample of 960 households (CND, 1998) (Table 2). Table 2: The Local Level Institutions Study Sample for Burkina Faso Province District Number of _Households Yatenga Oula 237 Houet Toussiana 60 Peni 179 Sissili To 240 Sanmatenga Korsimoro 243 Total 959 Table 3 shows selected characteristics of the respondents to the LLI survey. The vast majority of respondents are men and household heads. There are as many as 60 distinct ethnic and language groups in Burkina Faso (Englebert, 1996); the Mossi are by far the largest, being close to half of the population. In the LLI survey sample, the Mossi constitute 60% of respondents. Yatenga and Sanmatenga are overwhelmingly Mossi, whereas Houet and Sissili are ethnically diverse. Muslims are by far the largest religious group; Catholics and animists are also significantly represented, except in Yatenga. 7 Table 3: Selected Characteristics of Respondents (in %) Yaten a Houet Sissili Sanmatenga All Gender of Respondent 95.4 97.1 99.2 73.7 91.2 - Male 4.6 2.9 0.8 26.3 8.8 - Female Position in Household 98.3 100.0 100.0 95.9 98.5 -Head 1.7 - - 4.1 1.5 - Spouse . Language -Moore 93.7 1.7 45.9 97.5 59.7 -Dioula 0.8 28.0 0.4 - 7.3 - Fulfunde 5.5 9.6 0.8 2.1 4.5 - Gourounsi - 52.5 0.4 13.2 -Bobo 5.0 0.4 - 1.3 -Other - 55.7 - 14.0 Religion -Muslim 96.2 51.0 58.8 48.1 63.4 -Catholic 3.4 11.7 10.8 13.2 9.8 -Protestant 1.3 2.1 2.1 1.4 -Aniniist 0.4 36.0 28.3 36.6 25.4 Average Years Lived 53.9 44.3 31.9 49.3 44.8 in CoMmunity_. Table 4 provides a broad breakdown of some of the key economic characteristics of the households in the sample8. Approximately 12% of the respondents describe themselves as "inactive." The figure is exceptionally high for Houet (22.3%). Among those who report occupations, the vast majority are farmers/fisherman/cattle raisers. Almost all the respondents describe themselves as self-employed or unpaid family workers, with virtually no wage labor. The vast majority have no education at all, and differences among provinces along this dimension are small. It should be noted that Table 4 is calculated over all adult individuals in the sample while Table 3 pertained only to the respondents. 8 Table 4: Employment and Education Characteristics of the Adults (aged 15 and above) in the Sample (percentage) Yatenga Houet Sissili Sanmatenga All Main Occupation - Farmer/Fisherman/Cattle 88.0 73.8 92.9 92.9 86.3 Raiser 0.2 1.7 0.2 - 0.6 - Trade 0.7 2.0 0.2 0.2 0.8 - Other Private Sector 0.0 0.2 - - 0.1 -Public Sector 11.1 22.3 6.7 6.9 12.2 - Inactive 100.0 100.0 100.0 100.0 100.0 All Employment Status of Economically Active People -Wage Earner 0.7 1.9 0.1 0.2 0.7 - Self-employed/Unpaid Family 99.3 98.1 99.9 99.8 99.3 Worker All 100.0 100.0 100.0 100.0 Education -None 83.5 82.4 86.0 83.5 83.9 -Primary Incomplete 4.0 7.1 4.1 2.5 4.6 -Primary Complete 0.6 2.0 1.5 0.7 1.3 - Secondary Incomplete 0.9 4.8 1.1 1.9 2.3 - Secondary Complete 0.1 0.2 - 0.3 0.1 - Vocational School 0.2 0.2 0.1 0.1 -University 0.1 0.2 - - 0.0 -Other 10.8 3.1 7.1 11.0 7.7 All 100.0 100.0 100.0 100.0 100.0 Table 5 provides summary statistics on the level and composition of per capita expenditures. These are significantly lower in Yatenga and Sanmatenga than in the other two provinces. In addition, the level of inequality is relatively high in Yatenga (a gini-coefficient of 0.4). Consistent with this, the proportion of the population below the poverty line, and the expenditure gap, are relatively high in Yatenga9. For the sample as a whole the Gini- coefficient is 0.32. On average 67% of household consumption expenditure is on food, and the food shares are relatively high in the provinces with lower per capita income; they are 69.5% and 77.8% in Yatenga and Sanmatenga, respectively, compared to 57.2% and 65.5% for Houet and Sissili, respectively. The proportion of household expenditure on housing is very low in all provinces, varying between 1.9% and 3.1%. This is as expected, since housing is very modest and is self-constructed. The proportion of expenditure on education is relatively low in Yatenga and Sanmatenga, as compared to Houet and Sissili. Consistent with this, the percentages of children aged 5-14 attending school in Yatenga and Sanmatenga are 22.4% and 20.6% 9 The poverty line is set at two-thirds of median per capita consumption expenditure. 9 respectively, compared to 38.3% and 35.6% for Houet and Sissili. Overall, the proportion of household expenditure on education is quite high in Burkina Faso, 11.1%. Table 5: The Level and Composition of Household Expenditure Yatenga Houet Sissili Sanmatenga All Composition of Household Expenditure (%) Purchased Food 41.3 25.7 28.0 34.5 31.9 Home Produced Food 28.2 31.5 37.5 43.3 35.2 Education 7.7 15.2 14.2 6.2 11.1 Health 2.7 4.4 3.9 1.5 3.2 Housing 2.5 3.1 1.9 2.2 2.5 Private Transfers 0.6 1.5 1.4 0.4 1.0 Other 17.0 18.6 13.1 11.9 15.1 All 100.0 100.0 100.0 100.0 100.0 Household Expenditure per 58,712 69,768 71,686 60,886 65,265 Capita (CFAfr per year) Coefficient of Variation 76.42 61.79 54.14 49.98 61.25 Gini-coefficient 0.40 0.33 0.27 0.26 0.32 Ratio of Richest Over 4.6 2.5 1.7 1.8 5.2 Poorest Village Poverty Incidence - Poverty Head Count* 40.0 26.7 13.9 23.4 26.0 - Expenditure Gap** 42.2 31.2 22.7 22.4 32.7 * Percentage of population which is poor. This is based on a relative poverty line of two-thirds of the median of household expenditure per capita in the four provinces. ** The expenditure gap equals the shortfall, relative to the poverty line, of expenditure per capita of the average poor person, expressed as a percentage of the poverty line. Table 6a presents several non-monetary indicators of household welfare. Twenty-five percent of households report that they often go hungry. It is striking to note that the percentage is as high as 13.5 even in the richest quintile. As many as 55% of the households report that they have had to sell assets in order to buy necessities. The percentage is lowest among the poorest, presumably because they do not have assets to sell. In 89.8% of households, at least one child in the age-group 5-14 does not go to school, which highlights the importance of expanding education in Burkina Fasol°. This issue is discussed at greater length in the companion piece (Grootaert, Oh, and Swamy, 1999). Access to electricity and tap water is also very limited. Finally, only a small fraction of households reported that they had pulled a child out of school to help in the home or the farm; this could partly be because so few children are in school to start with. 'O Prinay enrollment rates (gross) in Burkina Faso in 1990 were 37%, compared to 73% for Sub-Saharan Africa as a whole (World Development Report, 1997). 10 Table 6a: Selected Indicators of Household Well-being, by quintile Quintile I Percent of Households Poorest 2 3 4 Richest All Which often go hungry 38.7 27.1 26.0 18.2 13.5 24.7 Which sold land, livestock or 41.4 58.3 59.4 57.3 59.4 55.2 equipment in order to buy necessities Where not all children 5-14 attend 96.6 93.1 88.6 86.9 82.1 89.8 school Which pulled children out of school 1.1 2.9 1.8 1.9 2.1 2.0 to help on farm or at home With access to electricity 1.6 0.5 0.5 0.0 1.6 0.8 With access to tap water 5.8 1.6 2.1 0.0 3.1 2.5 When we break down the data by language (Table 6b) we see an interesting pattern: the lowest incidence of hunger is among the Fulfunde-speakers, who are most likely to have sold an asset in order to buy a necessity. This may reflect the fact that the Fulfunde, who specialize in raising and selling cattle, may find it relatively easy to smooth consumption by selling them in consumption crisis: the sample average for number of animals (small and large) owned per household is 32.4; for the Fulfunde it is 97.7. The lower incidence of hunger among the Fulfunde could also partly reflect the fact that mean per capita household expenditure is slightly higher among them than in the sample as a whole (83,633 CFAF as compared to 65,265 CFAF). Another interesting finding is that not a single Fulfunde household sends all its children aged 5-14 to school. This could be because child labor is especially useful for minding cattle. Our regression analysis (see companion paper) confirms a lower propensity among the Fulfunde to send their children to school, controlling for other (observable) relevant characteristics. 11 Table 6b: Selected Indicators of Household Well-being, by language/ethnicity Language | Percent of Households Moore Dioula Fulfunde Gour- Bobo All ounsi and l_________ _________ others Which often go hungry 29.3 8.6 4.7 26.8 18.5 24.7 Which sold land, livestock or 59.5 37.1 69.8 52 45.2 55.2 equipment in order to buy necessities Where not all children 5-14 attend 93.2 83.3 100 80 85.6 89.8 school Which pulled children out of school 1.9 0 0 2.6 3.4 2.0 to help on farm or at home With access to electricity 0.5 1.4 0 0.8 2.1 0.8 With access to tap water 2.3 0 2.3 7.1 0.7 2.5 Table 7 provides data on household asset ownership. Unsurprisingly, there is virtually no car-ownership. A significant fraction of households in all three provinces owns a motorcycle/moped, a bicycle, or a radio. We have seen above that Sissili and Houet have higher average per capita incomes; consistent with this, we see that motorcycle/mopeds, radios, bicycles, ownership of "major durables," and ownership of large animals are higher in Sissili and Houet. Land ownership per household in Yatenga is higher than in Houet and is the same as in Sissili. This suggests that land scarcity may not be responsible for lower per capita incomes in Yatenga, but rather the low productivity of agriculture due to the low rainfall. When we break down the figures by ethnicity, we see that only 84% of the Fulfunde own or rent land, compared to 98% for the sample as a whole. The mean ownership is 3.11 hectares among the Fulfunde, in contrast with Mossi (4.58), Dioula (4.25), Gourounsi (5.39), and Bobo and others (4.53). Again, these figures are consistent with the specialization in livestock by the Fulfunde. The reader should note that in this instance, and in several others in the remainder of this paper, we do not provide a detailed breakdown (tables) of the figures by ethnicity, in the interest of brevity. 12 Table 7: Ownership of Household Durables, Land and Animals Yatenga Houet Sissili Sanmatenga All Percent of Households Owning Car/Truck 0.0 1.7 0.8 0.0 0.6 Boat 0.0 0.0 0.0 0.0 0.0 Outboard Motor 0.0 1.3 0.0 0.0 0.3 Motorcycle/Moped 25.3 41.8 27.5 9.5 26.0 Bicycle 75.1 85.8 94.6 67.9 80.8 Color TV 0.4 1.7 0.0 0.8 0.7 Black/White TV 0.0 1.3 0.0 0.0 0.3 Audio System 2.5 20.9 0.8 0.0 6.1 Telephone 0.4 1.3 1.3 0.4 0.8 Radio 35.4 72.0 63.3 40.3 52.8 Refrigerator/Freezer 0.0 0.8 0.4 0.0 0.3 Sewing Machine 4.6 3.8 4.6 1.2 3.6 Washing Machine/Dryer 0.0 0.4 0.0 0.0 0.1 Pressure Lamp 0.0 15.5 0.0 0.0 3.9 Other Major Durable 13.5 30.1 36.7 17.3 24.4 Average Asset Scorelr 1.6 2.8 2.3 1.4 2.0 Average Number of Large Animals 12.8 19.1 21.8 10.8 16.1 Owned Average Hectares of Land Owned 5.2 4.4 5.2 3.7 4.6 or Rented I I I 1/ Average number of durable goods owned per household. Our findings in this section are consistent with what we know about Burkina Faso from other sources. The standard of living is very low, whether measured by household consumption expenditure, asset-ownership, or by the attainment of "functionings"1' such as education, access to electricity, tap water, and the like. Among the provinces, Sissili and Houet are better off, in terms of both consumption and asset-ownership measures. We also noted some language/ethnicity-based differences, especially in educational outcomes. This preliminary finding suggests that it may be important to control for language-group in more detailed statistical analyses. With this overview of the economic status of households in these four provinces, we now turn to a discussion of membership in associations. I This term was coined by Sen (1988). 13 4. TuE MAP OF LOCAL INSTITUTIONS The LLI survey identified 53 associations spread over 8 broad categories (Table 8). The respondents reported a total of 6048 memberships, or 6.3 per household, but there were only 1446 active memberships, i.e. slightly more than one per household. This relatively small fraction of active memberships can in part be attributed to the fact that membership in some traditional groups and religious groups is not really a matter of choice by the household, but is sirnply a feature of its social identify, without necessarily requiring any active participation on its part. For example, one can be "born into" a church, without being active in it. More than half the organizations (29 out of 53) fall in the categories of "farmers service group" and "farmers group without specialization". The latter category comprises the so-called "Village Groups" (groupements villageois) which dominate agricultural life and resource mobilization in most of Burkina Faso. There were only 12 such organizations listed in the associational inventory, but they account for 62% of active memberships. The Village Groups were originally set up by the government during the 1970s and were intended to promote decentralization and improve farners' access to agricultural credit and extension. Although many of these groups have become well integrated in the community and are now in fact frequently seen as community-initiated, they still have strong ties to the central state and are dependent upon it for resources. In Yatenga, there is a parallel organization called "Naam Group" (groupement Naam) which was initiated in 1967 by a local leader who wanted to preserve cultural traditions and the performance of village service. Initially oriented towards young people, the Naam movement has developed into a general institution for village development, encompassing people of all ages (Pradervand, 1989; Smale and Ruttan, 1994). The Naam is now organized as a federation of groups in 18 of the 45 provinces (Donnelly-Roark, et al., 1999). While the Village Groups and the Naam Group deal with all aspects of agricultural life, the "farmers service groups" are usually specialized in a single sector such as water supply, health, or credit. Box 1 discusses further the different types of local associations in Burkina Faso. 14 Box 1: Types of Local Level Institutions in Burkina Faso 1. Religious Groups. This category consists of Muslim, Catholic or Protestant associations which coordinate local religious activities and provide the link with national-level religious organizations. 2. Traditional / Customarv Groups. Only two such groups were recorded in the survey, namely a traditional animist group operating in Houet, and the traditional Village Council in Sissili. 3. Social Groups. The main organization in this category is the Association of Parents and Mothers (Association des parents d'6l1ves et des meres educatives, APE/AME) which is responsible for many aspects of schooling, including maintenance of buildings and purchase of furniture. The various local associations are connected to a national parents' association. The category of social groups also consists of groups assisting in health care, as well as the Association of the Mossi which assists Mossi migrants in Sissili. 4. Soil Rehabilitation Groups. This small category consists of a few organizations that undertake reforestation and anti-erosion activities. Such organizations operate only in Yatenga. 5. Farmers Production Groups. The organizations in this category provide specialized crop assistance to farmers, mainly to promote the production of cotton and fruits and vegetables. There are also organizations that assist in livestock raising. 6. Farmers Service Groups. This is a very diverse category consisting of 17 types of associations which assist farmers in matters of water supply, savings and credit (including traditional "tontine" groups), health, and artisan activities. 7. Farmers Groups without Specialization. As discussed in the main text, this category consists of the "Village Groups" and the "Naam Group" which assist farmers with a wide range of activities. Village Groups can be organized separately for men and women or in an integrated way. The characteristic features of the Naam Group are its focus on cultural and social aspects and its organization in a federated structure. 8. Rural Youth Groups. This category consists of various associations aimed at organizing recreation for young people, introducing them into agricultural activities, and promoting religious activities. Source: CND, 1998 15 Table 8: Number of Local Organizations and Number of Memberships (Functional Classification) All Member hips Active Memb erships Type of Group Number of Number of % Number of % Organizations Memberships Memberships Religious Group 4 726 12.0 92 6.4 Traditional / Customary Group 2 219 3.6 5 0.4 Social Group 4 603 10.0 109 7.5 Soil Rehabilitation Group 5 103 1.7 8 0.6 Farmers Production Group 4 656 10.8 133 9.2 Farmers Service Group 17 1235 20.4 181 12.5 Farmers Group without 12 2230 36.9 890 61.6 Specialization Rural Youth Group 5 276 4.6 28 1.9 Total 53 6048 100.0 1446 100.0 Table 9 breaks down the active memberships by province and type. In all four provinces active membership is concentrated in farmers groups that are not specialized. As discussed before, this reflects the predominant role of the "Village Groups" in rural resource mobilization in Burkina Faso. FarTners service groups are mainly operational in Sissili and Sanmatenga. In Sanmatenga, this reflects primarily the strength of groups dealing with the conservation of natural resources, while in Sissili it reflects the strength of traditional and modem credit groups. In the other two provinces, service groups are fairly weak, but production-oriented groups have a stronger presence, especially those dealing with livestock and cotton. Yatenga suffers from chronic food deficits and livestock is essential for families' survival. In contrast, Houet has a thriving rainfed agriculture, supported by an array of organizations that support production and marketing. Among the other categories of associations, social groups (mainly parents' associations) are most active in Yatenga and Sissili. Active participation in religious groups is fairly even across the provinces (except in Sanmatenga). Active membership in traditional groups, soil rehabilitation groups and rural youth groups is quite low overall and entirely absent in some provinces. 16 Table 9: Active Memberships in Local Organizations, by province Yatenga Houet Sissili Sanmatenga All % of ______ ~~~~~~~~~~~~~Total Refigious Group 20 5.6% 28 7.5% 44 9.9% - - 92 6.4 Traditional / Customary - - 1 0.3% 4 0.9% - - 5 0.3 Group Social Group 34 9.5% 14 3.8% 46 10.4% 15 5.5% 109 7.5 Soil Rehabiltation 8 2.2% - - - - - - 8 0.6 Group Farmers Production 47 13.1% 59 15.8% 23 5.2% 4 1.5% 133 9.2 Group Farmers Service Group 2 0.5% 16 4.3% 87 19.7% 76 28.0% 181 12.5 Farners Group without 235 65.5% 253 67.8% 226 51.0% 176 65.0% 890 61.6 Specialization Rural Youth Group 13 3.6% 2 0.5% 13 2.9% - - 28 1.9 Total 359 24.8% 373 25.8% 443 30.6% 271 18.7% 1446 100 Table 10 displays the information on active memberships in a different way, namely as percentages of households actively participating in each type of association. This confirms that the most prominent differences are the higher percentages of households in Sissili and Sanmatenga that belong to farmers service groups and the relatively high participation in farmers production groups in Yatenga and Houet. It is also clear that the percentage of households participating in social groups is relatively high in Sissili and Yatenga. We also checked to see (detailed results not reported) whether there were significant ethnic differentials in active participation. The percentages of households that were active in at least one association were as follows: Moore (79.2%), Dioula (80%), Fulfunde (90.7%), Gourounsi (92.9%), Bobo and other (84.3%). Table 10: Percentage of Households Participating in Local Associations, by province I_______ ________ Province I Yatenga Houet Sissili Sanmatenga All Religious Group 8.4 11.7 17.5 9.4 Traditional / Customary Group - 0.4 1.7 - 0.5 Social Group 13.9 5.9 19.2 5.4 11.1 Soil Rehabilitation Group 3.4 - - - 0.8 Farmers Production Group 19.8 23.0 8.3 1.7 13.1 Farmers Service Group 0.8 5.4 31.3 25.1 15.8 Farmers Group without 82.3 71.6 70.0 56.8 70.1 Specialization Rural Youth Group 5.5 0.8 5.4 - 2.9 In the top panel of Table 11 we look at the incidence of active membership in groups limited by gender. The pattern that emerges is quite striking: in Yatenga and Sanmatenga 17 the overwhelming majority of memberships are in associations which have members from both sexes whereas in Houet and Sissili more than 50% of memberships are in groups with men only. When we broke down the data by ethnicity we were able to see that ethnic differentials seem to underlie these regional differences. The percentages of active memberships which are in mixed-gender groups are: Mossi (75.9%), Dioula (34.6%), Fulfunde (61.7%), Gourounsi (20.8%), Bobo and others (28.3%). The number of Fulfunde in our sample is quite small, therefore, regional differences are driven by the presence of Mossi. In the two regions where they are numerically dominant (Yatenga and Sanmatenga) mixed- gender association memberships are the norm12. The same pattern exists with regard to restrictions on membership by age: ethnic differences appear to underlie the provincial differences. We see a stark contrast in the incidence of mixed age groups between Yatenga and Sanmatenga on the one hand and Houet and Sissili on the other. The incidence of mixed-age groups in various ethnic groups is as follows: Mossi (65.6%), Dioula (27.1%), Fulfunde (58.3%), Gourounsi (9.5%), Bobo and other languages (20.4%). The provinces where the Mossi are numerically dominant, Yatenga and Sanmatenga are the ones in which mixed-age groups are most prevalent. Table 11: Distribution of Associational Memberships, by gender and age Yatenga Houet Sissili Sanmatenga All Gender -Female Only 0.8 17.2 18.1 2.2 10.6 - Male Only 5.3 50.4 62.3 6.6 34.6 -Mixed 93.9 32.4 19.6 91.2 54.8 All 100.0 100.0 100.0 100.0 100.0 Age -Age 25 Under 2.5 7.2 4.1 0.4 3.8 -Age 25 and Over 18.1 67.3 86.7 14.0 51.0 - Mixed 79.4 25.5 9.2 85.6 45.2 All 100.0 100.0 100.0 100.0 100.0 In the LLI survey, households were also asked whether they considered their memberships "important." As Table 12 indicates, the farmers groups are the ones most likely to be considered "very important." This is not surprising given that they contribute over 80% of all active memberships. Farmer service groups are considered important in all provinces, except Yatenga (where their presence is very limited). Farmers production groups, on the other hand, are less prevalent and are less likely to be considered important in Sanmatenga, compared to the other provinces. 12 When we break down active memberships by religion and gender we see that the percentage of memberships in mixed gender groups was 59.9% for Muslims, 39.3% for Catholics, and 44.3% for Animists. 18 Table 12: Percentage of Memberships in Associations Listed as "Very Important" Yaten a Houet Sissili Sanmatenga All Religious Group 80.0 53.6 52.3 58.7 Traditional / Customary Group - 100.0 25.0 - 40.0 Social Group 76.5 57.1 58.7 53.3 63.3 Soil Rehabilitation Group 87.5 - - - 87.5 Farners Production Group 85.1 91.5 87.0 25.0 86.5 Farmers Service Group 0.0 75.0 77.0 85.5 79.6 Farmers Group without 60.4 74.3 76.1 74.4 71.1 Specialization Rural Youth Group 53.8 0.0 69.2 - 57.1 Average 66.3 74.5 72.0 75.6 71.9 Households were asked to identify the three most important groups to which they belonged. Table 13 identifies the ten groups most frequently cited among the three most important groups. Not surprisingly, the Village Groups and the Naam Group occupy the top four spots. We have already commented on the importance of these groups in Burkinabe rural life. Three of the other six groups on the list are also related to farming. One religious group (Muslim association) and two social groups (parents' association and health and literacy association) round out the list. (Annex Tables Al through A4 provide information on the most important groups separately for each province.) 19 Table 13: The Ten Most Important Local Associations Initiated by Organized in Group Name Category Govern- Comm Formal Informal Number of ment unity Times Cited (% of Total Citations) 1. Village Group-rmale Farmers w/o 15.8% 84.2% 72.1% 27.9% 262 (19.8%) (Groupement villageois hommes) Specialization 2. Village Group - mixed gender Farmers w/o 23.7% 76.3% 82.4% 17.6% 249 (18.8%) (Groupement villageois mixte) Specialization l 3. Village Group - female Farmers w/o 13.0% 87.0% 51.0% 49.0% 192 (14.5%) (Groupement villageois feminin) Specialization 4. Naam Group (Groupement Farmers w/o 8.6% 91.4% 100.0% 0.0% 145 (10.9%) Naam) Specialization I 5. Water Management Committee Farmers 1.1% 98.9% 3.4% 96.6% 81 (6.1%) (Comite de gestion des points Services d'eau (CPE)) 6. Association of Livestock Farmers 11.0% 89.0% 84.6% 15.4% 77 (5.8%) Farmers (Groupement villageois Production eleveurs (GVE) ou Association des eleveurs) I 7. MuslimAssociation Religious 5.1% 94.9% 34.6% 65.4% 74 (5.6%) (Communaute ou association musulmane) . l 8. Association of Parents and Social 10.8% 89.2% 12.3% 87.7% 53 (4.0%) Mothers (Association des parents d'elves et des meres educatives (APE/AME)) . l 9. Association of Cotton Producers Farmers 51.5% 48.5% 69.7% 30.3% 32 (2.4%) (Groupement de producteurs de Production coton - GPC) lIO.Health & Literacy Associations Social 69.2% 30.8% 82.1% 17.9% 25 (1.9%) (Association Bantaare; Cellule PSP; Association Wendpouyre; Association Croix-Rouge) In Table 14 we see that over a three month period households attend 5.8 meetings, on average. There are substantial regional differences, with meeting attendance being relatively low in Yatenga and Sanmatenga, compared to Houet and Sissili. If we break down meeting attendance by ethnicity, we see that the Mossi, who are numerically dominant in Yatenga and Sanmatenga, attend on average 4.4 meetings every three months. The attendance rates for the other ethnic groups are: Dioula (6.88), Fulfunde (5.23), Gourounsi (11.38), Bobo and others (5.56). Thus, the regional differences could well be driven by differences in the behavior of ethnic groups. It is interesting to note that the Mossi, who, according to the anthropological literature, have a high-level of community cooperation, are less prone to attend meetings than others. However, while meeting attendance is perhaps a good indication of time devoted to associations, it need not be a good measure of cooperation or even active participation in cooperative affairs. The multivariate analysis of the LLI study results suggest indeed that 20 meeting attendance is not a significant indicator of the role of social capital in affecting household welfare or access to services (Grootaert, Oh and Swamy 1999; Grootaert, 1998). Table 14: Average Number of Attendances to Groups Meetings (Per 3-Month Period) LLI Categories Yatenga Houet Sissii Sanmatenga All Religious Group 3.5 3.0 3.1 3.2 Traditional / Customary Group - 3.0 1.8 - 2.0 Social Group 2.7 2.1 2.1 2.2 2.3 Soil Rehabilitation Group 3.0 - - - 3.0 Farmers Production Group 2.5 3.7 8.5 3.0 4.1 Farmers Service Group 2.5 2.9 3.3 1.3 2.4 Farmers Group without 2.8 3.2 4.6 2.5 3.3 Specialization Rural Youth Group 3.0 4.4 4.0 - 4.0 All 4.4 6.1 8.8 3.7 5.8 Note: The averages are computed over the sample of households which do belong to the relevant type of association. Table 15 completes our description of the map of local associations in Burkina Faso. It indicates that local groups are overwhelmingly community-initiated. However, this classification is based on respondent perceptions, and this need not always correspond to historical reality. As mentioned earlier, the Village Groups were initiated by the government but over the years they have become assimilated by the community and are now largely seen as locally originated. This is obviously a positive process which suggests that who initiates a local association may become less important over time, provided the structure of the group is sound and flexible. The process of assimilation can be a long one, as this case indicates, since the Village Groups were created more than two decades ago. There is an interesting parallel between the experience of the Village Groups in Burkina Faso and that of the Agrarian Syndicates (Sindicatos Campesinos) in Bolivia, one of the other countries in the LLI study. The Agrarian Syndicates were also created by government (in the 1950s) and over time they gradually took on the role of the central community organization to manage resources and to resolve conflict. In this process, they also became perceived as genuinely local groups and 96% of respondents classified them as community-initiated (Grootaert and Narayan, 1999). As far as the other groups are concerned, only social groups and farmers production groups are initiated by government to a significant degree, 31 % and 20%, respectively. In spite of the high degree of community-initiation, almost three-fourths of groups are seen as formally organized. There is an interesting contrast among the different types of farmers groups. The Village Groups, the Naam Group, and the production-oriented groups are predominantly formally organized. The service groups, which are more specialized in 21 one sector or agricultural input, and which also tend to be younger organizations, are more often informally organized. Donnelly-Roark et al. (1999) argue that this reflects a less hierarchical and more consensus-oriented approach to organizational functioning. Table 15: Type of Initiation and Organization of Groups State Community Formally Informally Initiated Initiated Organized Organized Religious Group 4.4% 95.6% 31.5% 68.5% Traditional / Customary Group 0.0% 100.0% 0.0% 100.0% Social Group 31.2% 68.8% 36.7% 63.3% Soil Rehabilitation Group 12.5% 87.5% 87.5% 12.5% Farmers Production Group 20.3% 79.7% 78.2% 21.8% Farmers Service Group 5.0% 95.0% 29.8% 70.2% Farmers Group without 16.5% 83.5% 75.1% 24.9% Specialization Rural Youth Group 3.6% 96.4% 71.4% 28.6% Having completed the map of local institutions and the overview of overall group membership, the following section looks more closely at the characteristics of the three associations the household identifies as "most important." 22 5. CHARACTERISTICS OF THE MOST IMPORTANT ASSOCIATIONS Table 16 focuses on the three most important associations to which the household belongs, and provides information on several dimensions: the reasons for joining, the costs of joining, and the extent of participation. With regard to the reasons for joining, the two most common reasons are the impact on household livelihood and the impact on the community. One province, Sanmatenga, deviates significantly from this pattern: mandatory membership in associations is much higher in this province, while the impact on the community is considered much less important as a reason for joining. This is probably related to the strong organizational structure of the Village Groups in Sanmatenga, with close links to higher-level federated organizations. The traditional chieftancy is also still stronger in Sanmatenga than in Yatenga (Donnelly-Roark et al., 1999). Further, the traditional cooperative culture of the Mossi is such that there is no need to join associations in order to take care of community matters. By and large, no membership fee is required to join and to remain a member of local associations in Burkina Faso. In 67% of cases no monetary contribution is paid, and in almost half the cases, there is also no requirement of labor contribution. However, two-thirds of households report that they are "very active" participants in the decision-making process of the association of which they are a member. Some interesting contrasts are evident in the extent of use of groups to access services (Table 17). With regard to education we see that the proportions of households which say that group membership helps them access education is relatively high in Yatenga and Sanmatenga. Although the level of school attendance is especially low in these provinces, the two departments selected for the LLI study have above average education facilities. The extent to which there is an underlying ethnic dimension is not clear: the proportion of Mossi who say that association membership helps them access services is 50.4%, compared to 23.3 % for the other ethnic groups but Mossi enrollment rates (23.5%) are lower than most other groups, e.g. 48.2% for Dioula and 45.4% for Gourounsi. In the case of health, the role of group memberships is most important in Yatenga. Group membership also plays an important role in ensuring access to water supply, credit, savings, and agricultural technology. The importance of local associations for water supply is to be expected in a country where rainfall is variable and inadequate in many areas. In Sanmatenga, for example, 11 of the 12 villages in the study had a Management Committee for water wells. These function quite effectively as the province has few conflicts over water issues, which is not a trivial achievement in an area where farmers and herders continuously must compete over scarce water (Donnelly-Roark et al., 1999). The low importance of local associations in water supply in Houet is a result of the existence of a government-supported water supply program. 23 Table 16: Characteristics of the Memberships in the Most Important Groups Yatenga Houet Sissili Sanmatenga All Reasonsfor Joining (%) Impact on Household Livelihood 27.4 28.1 24.5 41.0 29.3 Impact on Community 57.6 46.4 55.2 18.3 46.5 Everybody Belongs/Mandatory 5.9 7.2 5.7 25.1 9.8 Emergency/Future Benefit 6.9 13.8 13.9 10.0 11.4 Other 2.2 4.5 0.7 5.6 3.0 Total 100.0 100.0 100.0 100.0 100.0 Membership Fees (%X) No Membership Fee 78.5 84.0 51.1 62.5 68.6 1-100 CFAfr/Month 17.1 9.2 33.7 23.9 21.4 101-300 CFAfr/Month 3.1 3.1 4.0 10.8 4.8 301-500 CFAfr/Month 1.3 2.3 8.2 2.8 3.9 501 and up CFAfrlMonth - 1.4 3.0 - 1.3 Total 100.0 100.0 100.0 100.0 100.0 Average Amounts of Fee (CFAfr) 25.8 39.7 109.1 47.2 58.9 Labor Contribution (%) No Working Days 54.2 52.1 46.8 31.9 47.1 1 - 4 Days / Year 8.1 6.3 4.0 31.0 10.7 5 - 12 Days / Year 19.9 18.1 5.9 29.9 17.1 13 - 24 Days / Year 8.7 8.0 15.8 4.4 9.9 25 - 52 Days / Year 8.1 14.6 9.2 2.4 9.1 53 andup Days / Year 1.0 0.9 18.3 0.4 6.1 Total 100.0 100.0 100.0 100.0 100.0 Average Days of Labor 7.1 10.3 24.1 4.7 12.7 Contribution l Extent ofActive Participation in Decision Making (X) Very Active 56.7 65.9 72.4 75.7 67.5 Somewhat Active 32.4 33.0 27.1 17.5 28.1 Not Active 10.9 1.1 0.5 6.8 4.4 Total 100.0 100.0 100.0 100.0 100.0 Table 17: Involvement in Service Provision of the Most Important Groups % of Memberships in Yatenga Houet Sissii Sanmatenga All Groups Active in. Education 54.2 14.3 33.0 61.0 38.5 | Health 44.2 10.1 21.9 3.2 20.7 Water Supply and Sanitation 38.9 6.0 32.8 45.8 29.7 l Electricity 0.6 - 0.5 0.4 0.4 | Credit 16.8 14.7 28.6 25.1 21.4 Savings 10.6 7.5 4.0 11.6 7.9 Agricultural 38.6 15.8 34.8 29.9 29.8 Inputs/Technology Irrigation 8.4 0.9 0.2 0.8 2.5 Land, Forestry, Water Rights 24.9 2.0 26.9 1.2 15.0 24 The extent of heterogeneity of an association's membership can have an impact on its effectiveness. There are two possibilities. On the one hand, cooperation among members may be easier if the group is homogenous. On the other hand, if the community is split into multiple homogenous groups this can lay the ground for factionalism. However, if groups are heterogeneous, this may open up channels of communication and create an environment in which different sections of the community can interact and share each other's knowledge and infornation. Our discussion in section 2 had highlighted the fact that social tensions, especially among ethnic groups, are at a relatively low level in Burkina Faso. Based on this we should expect groups to have diverse memberships. This is precisely what we see (Table 18). The only dimension along which we have homogeneity is occupation. This is not a surprise, since the vast majority of households in Burkina Faso are farmers.13 Homogeneity with respect to gender is much lower in Yatenga and Sanmatenga. This is consistent with what we observed earlier regarding the presence of gender-specific associations (see Table 1 1), and reflects mainly an ethnic factor. When we break down the data by ethnicity, we see that 23% of memberships among the Mossi are in organizations whose members are "mostly from the same gender"; the corresponding figure for the other ethnic groups is 68%. Table 18: Internal Homogeneity of the Most Important Local Associations % of Memberships in Yatenga Houet Sissii Sanmatenga All Associations whose Members are "Mostly from the Same" Neighborhood 34.4 19.8 18.9 21.5 23.4 Kin Group 28.4 18.3 10.0 3.6 15.5 Occupation 82.5 78.5 74.5 68.1 76.3 Economic Status 7.8 1.1 29.3 0.0 11.1 Religion 35.3 19.8 14.3 1.6 18.4 Gender 10.9 83.4 49.5 21.1 43.7 Age 5.0 1.4 0.5 0.0 1.7 Political Party 24.7 13.5 8.0 3.6 12.7 Education 4.7 8.6 0.5 0.0 3.6 13 A word of caution is needed regarding the heterogeneity criteria. People's perception of what constitutes a "difference" along any given criteria can vary. For example, while 76% of association members perceive the other members to be mostly of the same occupation (mainly farmers) only 11% perceive them to be of the same economic status. Likewise, it is difficult to interpret how only 4% of members are seen to be of the same education in a country where 80% of the population has no education. 25 6. SOCIAL CAPITAL AND POVERTY A central focus of the Local Level Institutions Study is the relationship between social capital and poverty. In particular, we are interested in the question of whether social capital allows access to services and, hence, to income-earming opportunities. It is also possible, of course, that wealthier people are more able to acquire social capital. Disentangling these possibilities requires multivariate regression analysis, which we do in a companion paper (Grootaert, Oh, and Swamy, 1999). In the current paper we tabulate membership in local organizations according to the income level of the household, and the income level of the village to which it belongs (Tables 19 and 20). Our main finding in this regard is that group membership does not seem to vary systematically with income. Only one pattern is noticeable: the poorest quintile of households is less likely to belong to a social group, while households in the top two quintiles of villages are more likely to belong to a social group. Since social groups are mainly active in the areas of health and education, this suggests that the poorest households who already have less access to these services, may also be less helped by associations. A much more detailed analysis of the relationship between income-earning opportunities and social capital is provided in our companion paper. Table 19: Percentage of Households who are Members of Local Associations, by category of association and household quintile Quintile of PCH IHEXP, Household Level LLI Categories 1 2 3 4 5 All Religious Group 3.7 9.9 12.0 11.5 9.9 9.4 Traditional / Customary Group 0.0 0.5 1.6 0.5 0.0 0.5 Social Group 6.3 10.4 14.1 11.5 13.0 11.1 Soil Rehabilitation Group 0.0 1.6 0.0 0.5 2.1 0.8 Farmers Production Group 12.0 9.9 12.5 11.5 19.8 13.1 Farmers Service Group 10.0 18.2 14.1 21.9 14.6 15.8 Farmers Group without 69.6 69.8 70.8 70.3 69.8 70.1 Specialization Rural Youth Group 2.1 1.0 4.2 4.2 3.1 2.9 26 Table 20: Percentage of Households who are Members of Local Associations, by category of association and village quintile Quintile of PCHHEXP, Village Level LLI Categories 1 2 3 4 5 All Religious Group 7.2 13.8 9.7 9.0 6.6 9.4 Traditional / Customary Group 0.0 0.5 0.0 0.0 2.2 0.5 Social Group 6.1 7.9 6.5 14.8 19.9 11.1 Soil Rehabilitation Group 0.0 0.0 0.0 3.8 0.0 0.8 Farmers Production Group 15.6 1.5 10.8 9.5 30.4 13.1 Farmers Service Group 10.0 24.6 7.0 20.5 14.9 15.8 Farmers Group without 69.4 66.5 68.7 79.1 65.8 70.1 Specialization Rural Youth Group 3.9 0.5 3.2 5.7 1.1 2.9 27 7. WHO PROVIDES SERVICES? In this section we evaluate the relative role of community associations, government, and NGOs, in providing four services: education, health, access to credit, and assistance with titling of land or houses. Community organizations play a relatively small role in providing education (Table 21): only 10.6% of children who are in school go to community-owned schools. Given that less than 30% of children go to school in the first place, the role of community-run schools is indeed very small. There is some ethnic variation. The proportions of school-going children whose school is owned by the community are: Moore (5.2%), Dioula (4.4%), Fulfunde14 (18.2%), Gourounsi (15.2%), Bobo and others (10.6%). The majority of school-going children are in schools run by the central government, with another 10% going to schools run by religious institutions. However, as we pointed out earlier, local associations, especially the parents' associations, play a key role in maintenance of school buildings (even those owned and managed by the government) and in the purchase of furniture. Table 21: Who Owns and Manages Schools? Owners ip M_ Management Yatenga Houet Sissili San- All Yatenga Houet Sissili San- All l_______ matenga I matenga _ Community 3.4 14.2 16.0 0.0 10.6 6.8 13.4 5.0 0.0 7.5 District 3.4 7.3 0.8 3.7 4.0 8.1 12.6 4.2 0.9 7.4 Govemment Central 62.2 66.3 70.2 77.8 68.3 57.4 62.4 81.5 79.6 69.9 Government Religious 20.9 10.3 5.5 5.6 10.2 15.5 10.0 6.7 10.2 10.1 Organization s NGOs 0.0 0.8 6.7 0.0 2.4 0.0 0.4 0.4 0.0 0.3 Other 10.1 1.1 0.8 12.9 4.5 12.2 1.2 2.2 9.3 4.8 All 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Note: Figures show the distribution of children aged 5-14 in school, according to entity who owns/manages the school. Health services are also mainly provided by the central government (Table 22), with only 5.4% of health consultations being at community-run institutions. Community institutions have a significant presence only in Sanmatenga, where they account for 12.5% of all health consultations. District governments play a negligible role, except in Houet, where they provide 11.3% of health consultations. As was the case for education, the main role of community associations is in the maintenance of the facility - regardless of who fornally owns and operates it. What emerges therefore is a complementary role between the entity that builds an education or health facility - mainly the central government - and the local associations. The latter are needed to ensure the sustainability of the facility by maintaining 14 This figure is not very meaningful, because it is based on only 11 school-going children. 28 and repairing it, and, occasionally, by buying needed inputs. Another striking result from Tables 21 and 22 is the near absence of NGOs in the provision of education and health care. Table 22: Who Organizes Health Care? Yatenga Houet Sissili Sanmatenga All Community 1.1 4.6 3.7 12.5 5.4 Central Government 96.0 82.7 92.8 81.2 88.3 District Government 1.1 11.3 0.9 0.0 3.3 Religious Organizations 0.0 0.1 0.0 0.0 0.0 NGO 0.0 0.0 0.2 0.4 0.2 Employer/Private Business 0.5 0.2 0.4 0.2 0.3 Other 1.3 1.1 2.0 5.7 2.5 All 100.0 100.0 100.0 100.0 100.0 Note: Figures show the distribution of health consultations, according to ownership of the health care facility. Only 16.1% of households in our sample report having obtained credit in the past year (Table 23). This is a surprisingly low figure, given that that incomes are extremely low and variable in Burkina Faso, and is almost surely a reflection of an existing supply constraint. Of the limited credit that is available, friends and relatives, community credit groups, and NGOs all make a significant contribution. The role of formal banks, either government- owned or commercial is very low, at 3.1%, and 1.9%. A striking finding is that in Yatenga community credit groups are the source of credit in 52.2% of cases.'5 In our companion paper, we examine the role of community credit institutions in reducing the incidence of distress sales. We find that an individual who lives in a village where most residents belong to a community credit institution is less likely to engage in a distress sale (Grootaert, Oh and Swamy, 1999). 15 This is mainly due to the presence of large credit and saving organization, the Union Regionale des Caisses Populaires. (Donnelly-Roark et al., 1999). 29 Table 23: Who Provides Credit? Yaten a Houet Sissili Sanmatenga All % of Households who 9.7% 13.0% 26.7% 14.8% 16.1% obtained credit in the past year l Distribution of sources Household member 8.7 6.3 9.2 0.0 6.3 Friends 13.0 18.8 20.0 30.8 21.4 Relatives 0.0 21.9 10.8 51.3 21.4 Money lender - - - - Trader/store 4.4 0.0 3.1 0.0 1.9 Community credit group 52.2 15.6 15.4 5.1 18.2 Govenunent bank 13.0 6.3 0.0 0.0 3.1 Commercial bank 0.0 0.0 4.6 0.0 1.9 NGO 4.4 6.3 30.8 5.1 15.7 Other 4.4 25.0 6.2 7.7 10.1 All 100.0 100.0 100.0 100.0 100.0 l In many African countries efforts have been undertake to ensure that households obtain forrnal titles to their land in their possession. In Burkina Faso a national program for land management (Programme National de Gestion des Territoires) was initiated in the mid- 1980s. However, among the households in our study sample, only 1.1% of owners have title documents. 30 8. COLLECTIVE ACTION AND TRUST The final section in this paper addresses issues of collective action and trust. The survey data fully reflect the traditions of cooperation mentioned earlier in this paper. When asked whether they felt other households in the village made a fair contribution to collective action at the village level, 89% of households answered in the affirmative. Eighty percent of households participated at least once in collective action at the village level, and two-thirds participated at the level of the neighborhood (Table 25). For the most part, however, participation is limited to less than once every two months. Provincial variation in collective action is slight, although participation is somewhat lower in Houet. Table 25: Participation in Collective Action l _____________________ _________ Neighborho d Frequency of Yatenga Houet Sissili Sanmatenga All Participation (Per Year) 0 17.7 43.9 31.3 38.7 33.0 1-6 times 60.3 54.0 53.3 46.5 53.5 7-12 times 16.9 1.7 13.3 9.1 10.2 13-24 times 3.4 0.0 1.7 4.5 2.4 25+ times 1.7 0.4 0.4 1.2 0.9 Total 100.0 100.0 100.0 100.0 100.0 Village Frequency of Yatenga Houet Sissili Sanmatenga All Participation (Per Year) 0 15.6 26.4 4.2 32.1 19.6 1-6 times 61.2 66.1 64.6 53.5 61.3 7-12 times 18.6 4.2 25.8 10.7 14.8 13-24 times 3.8 2.5 5.4 2.1 3.5 25+ times 0.8 0.8 0.0 1.6 0.8 Total 100.0 100.0 100.0 100.0 100.0 The frequency of participation in collective action does not seem to vary systematically by the wealth of the household or the wealth of the village (Table 26a). Significant differentials emerge, however, if we break down the figures by ethnicity (Table 26b). Clearly, the Mossi and Gourounsi participate in collective action with much greater frequency than the other groups, especially the Fulfunde. This can be interpreted in at least two different ways. On the one hand the high rates of participation among the Mossi can be seen as confirmation of anthropologists' (Fiske, 1991, Kevane and Englebert, 1997) claims regarding cooperative norms among the Mossi. Another interpretation could be that households are most likely to cooperate with members of their own ethnic group; since the Mossi are the most numerous group a Mossi household is more likely to be in the company of 31 others of the same ethnic group, and hence may be more likely to participate in collective action. Finally, it is worth remembering that the extent of collective action may have both "demand" and "supply" dimensions, i.e., one region may have more collective action either because residents are more willing to cooperate, or because there are more tasks that require cooperation. For example, the Gourounsi live primarily in Sissili, which, although it is a relatively well-off province, still has poor service facilities in many areas. This may well increase the need for collective action. Consistent with this, Sissili also has more active memberships in associations per households and higher levels of meeting attendance and active participation in associations than any other province (Grootaert, Oh and Swamy, 1999). Table 26a: Average Frequency (times per year) of Participation in Collective Action, by quintile | Neighborhood Village Household Quintiles Poorest 3.3 4.3 2 3.0 3.9 3 3.1 3.9 4 3.6 4.1 Richest 3.2 4.9 All 3.3 4.2 Village Quintiles Poorest 4.1 5.2 2 2.8 3.5 3 2.1 2.7 4 4.6 4.8 Richest 2.6 4.9 All 3.3 4.2 Table 26b: Average Frequency (times per year) of Participation in Collective Action, by ethnicity Neighborhood Village Moore 3.95 4.51 Dioula 1.90 3.36 Fulfunde 1.07 1.56 Gourounsi 3.72 6.06 Bobo and Others 1.47 2.58 All 3.26 4.21 32 Regarding the motivation for collective action, it is clear that in all the provinces the vast majority of households believe that others contribute fairly to collective action (Table 27a). Fear of social criticism appears to be lowest in Houet; this is consistent with the finding above that collective action is at a relatively low level in this province. On the other hand, fear of criticism or fines is extremely high in Sissili, which of course may in part explain the high levels of collective action and participation in associations observed in that province. Table 27a: Motivation for Collective Action, by province .________ _______ Neighborhood Yatenga Houet Sissili Sanmatenga All % of Households fearing criticism or fine 73.4 56.1 78.3 72.0 70.0 % of Households encouraged to participate 38.8 20.5 81.7 11.5 38.1 because of criticism or fine' % of Households perceiving others to make fair 87.3 85.8 78.3 92.2 85.9 contribution l Village Yatenga Houet Sissili Sanmatenga All % of Households fearing criticism or fine 72.2 49.0 95.8 71.6 72.2 % of Households encouraged to participate - - - - - because of criticism or fine' % of Households perceiving others to make fair 87.3 87.8 86.3 93.8 88.8 contribution Note: ' The question regarding whether fear of criticism motivated participation in collective action was asked only once, with reference to the "community," without differentiating between neighborhood and village. The ethnic dimension confirms this. Rates of participation in collective action are high among the Mossi and the Gourounsi. Fear of censure for non-participation is also much higher among the Mossi and the Gourounsi than in the other ethnic groups (Table 27b). This lends credence to arguments like those of Olson (1982) to the effect that collective action can occur only when fear of social punishments offsets the incentive to free-ride. 33 Table 27b: Motivation for Collective Action, by Ethnicity Neighborhood Mossi Dioula Fulfunde Gourounsi Bobo and others % of Households fearing criticism or fine 73.5 52.9 48.8 85.0 57.5 % of Households encouraged to participate 34.6 25.7 18.6 86.6 21.2 because of criticism or fine' % of Households perceiving others to make fair 87.2 88.6 76.7 83.5 84.3 contribution . l Village Mossi Dioula Fulfunde Gourounsi Bobo and others % of Households fearing criticism or fine 77.1 52.9 46.5 97.6 47.3 % of Households encouraged to participate - - - - - because of criticism or fnel % of Households perceiving others to make fair 90.9 91.4 88.4 81.0 86.3 contribution . Note: The question regarding whether fear of criticism motivated participation in collective action was asked only once, with reference to the "community," without differentiatin between neighborhood and village. Given that income and consumption are highly variable in Burkina Faso, and given the anthropologists' claims of extensive cooperation in rural communities in Burkina Faso, we expected to see high levels of mutual assistance. In fact (Table 28) only 32% of households say that in an emergency they can definitely expect help from beyond the household and relatives. The results regarding ethnicity are puzzling: the percentages of various ethnic groups that report they can definitely obtain crisis assistance beyond family members are as follows: Mossi (27.2%), Dioula (22.9%), Fulfunde (44.2%), Gourounsi (26%), Bobo and others (56.9%). We see that the Fulfunde and Bobo and others, who have relatively low levels of participation in collective action, are more likely to receive crisis assistance. The opposite is true for the Mossi and the Gourounsi. We are especially puzzled by the relatively low figures for mutual assistance among the Mossi, given the anthropologists' claims of extensive cooperation among them. Perhaps the answer may lie in the concept of "household and relatives". Since the Mossi often live in entirely Mossi villages, a fairly large fraction of their community can be broadly construed as "household and relatives". It would then not be a surprise if it is hard for them to get assistance from beyond this group. However, this is merely a speculation, and this issue clearly needs further investigation. 34 Table 28: Community Assistance Yatenga Houet Sissili Sanmatenga All % of Households who think they can obtain assistance beyond immediate household and relatives -Definitely 19.8 49.0 32.1 27.2 32.0 -Probably 35.4 23.0 22.5 35.0 29.0 -Probably Not 16.5 7.9 23.3 5.8 13.3 -Definitely Not 24.9 10.9 18.8 23.0 19.4 -Difficult to Answer 3.4 9.2 3.3 9.0 6.3 Total 100.0 100.0 100.0 100.0 100.0 35 9. SUMMARY AND CONCLUSIONS This paper provides a description of the extent of household participation in local associations in Burkina Faso, and the role of these associations in service delivery. The data come from the Local Level Institutions Survey, which canvassed 960 households in four provinces of Burkina Faso: Yatenga, Sanmatenga, Sissili and Houet. These four provinces represent very different social, cultural and economic settings. Yatenga and Sanmatenga are situated in the north of Burkina Faso, are predominantly populated by the Mossi, and are largely Muslim. Both provinces suffer from low rainfall, which makes for low productivity agriculture, co-existent with livestock raising. They are the two poorest provinces in the study zone, but while Sanmatenga has a very equal distribution of income, Yatenga has be most unequal distribution among the four provinces. Houet, which lies to the west, is culturally and ethnically diverse. It is a high rainfall area and is the granary of Burkina Faso. It has the second highest income level of the four provinces. Sissili, in the southeast, is the home of the Gourounsi. Due to high rainfall, agriculture is diversified and the province has the highest income level among the four areas in the study. Across the four provinces, 53 different local associations were recorded. The number of associations to which an average household belongs is quite large (6.3), but only a small fraction of these memberships is active: each household is active in approximately one association. Not surprisingly, more than 80% of these memberships are in farming-oriented associations. Among those, the most important category is the "farmers groups without specialization", which includes the Village Groups (Groupements Villageois) and the Naam Group (Groupement Naam). The Village Groups were set up by the government in the 1970s to promote decentralization and to improve farmers access to agricultural inputs. The Naam Group was set up in Yatenga by a local leader to preserve cultural traditions and village service. We found distinct provincial patterns in the extent of age and gender integration in group membership. The Mossi-dominated regions, Yatenga and Sanmatenga, show a predominance of mixed gender and age groups, whereas in Houet and Sissili gender- and age-segregated groups are dominant. We looked more closely at the three most important associations to which the household belongs. The Village Groups and the Naam Group were most frequently cited by households as the most important groups to which they belong, followed by water management committees and associations of livestock fanners. Significant fractions of households reported that membership promoted the livelihood of the household and the interests of the community. Costs of membership, financial and in terms of labor-time, were found to be very low. The most important groups are significantly involved in provision of various services: education, health, water supply and sanitation, credit, savings, and agricultural technology. However, the central government plays the dominant role in the provision of education and health facilities. The role of local associations is complementary to the government's role, by 36 ensuring maintenance and the purchasing of needed inputs. NGOs are virtually absent in the provision of these services. In the all-important area of water supply, many villages have set up management committees to oversee the wells. In some provinces, such as Sanmatenga, they operate effectively and succeed in avoiding conflicts over water between farmers and herders. Elsewhere, in Houet, the committees try to compensate for the ineffectiveness of the government's water supply program. Less than one in five households in the sample reported having received credit in the past year. Given the severe fluctuations in income which are typical for Burkina Faso, and the need for funds to smoothen consumption over time, this reflects almost certainly a binding supply constraint. Friends and relatives are relied upon for almost half of all credit. Community groups are a source for only 5% of households in Sanmatenga and 15% in Houet and Sissili. Only in Yatenga, where a well established credit and savings organization operates, is a majority of households able to rely on a source of funds outside their immediate circle of friends and relatives. We also looked at the internal heterogeneity of the membership of associations, along economic and demographic criteria. Anthropologists' claims of harmonious relations among various social groups in Burkina Faso had led us to expect diversity within associations. This is precisely what we found. Association members in Burkina Faso vary by religion, economic status, and neighborhood. They vary least by occupation, since most people are farmers. Among the demographic criteria, they vary least by gender, but there are wide differences by province. Our preliminary look at the relation between household and village income levels and membership in local associations did not yield any significant patterns. The only striking finding was that the poorest households are less likely to belong to education and health groups, and households which live in the top two quintiles of villages are more likely to belong to such groups. This suggests that the poor, who already have less access to health and education, are also least helped by associations active in these areas. A much more detailed analysis of the relationship between social capital as reflected in association membership, and welfare, is presented in our companion paper. The Mossi and Gourounsi seem to participate in collective action to a greater extent than the other ethnic groups. In the case of the Mossi, this is consistent with their cooperation orientation. In the case of the Gourounsi in Sissili, the motivation seems to come in part from fear of criticism for non-participation. We could not find any evidence of income-related differentials in participation in collective action. However, we expect that demand factors are also an important reason for observed differences in collective action: collective action is organized more frequently where there are more tasks that require cooperation. For example, even though Sissili is a fairly well-off province, it still lacks many facilities. This may well explain the Gourounsi's greater participation in both collective action and in local associational life. 37 In conclusion, the evidence suggests that there is significant social capital embodied in local associations in Burkina Faso. Associations operate in many different areas that are essential to the well-being of households, and people of different religions, neighborhoods, and economic status join in associations. However, a particular dimension of social capital may be lacking, namely that embodied in relations between communities and governmental institutions. In many areas where the government can play a large role, e.g. education, health, water supply, and credit, outcomes are poor in a number of areas, and local associations need to fill the gap. In light of these findings, the efforts to draw on within-community social capital that are being made under the aegis of the Decentralization Law hold great promise. 38 Annex: Tables by Province Table Al: The Five Most Important Local Associations in Yatenga Initiated by Organized in _ Group Name Category Government Community Formal Informal Number of Times Cited (% of Total _____________ _____________ Citations) 1. Groupement Naam Farmers w/o 8.6% 91.4% 100.0% 0.0% 145 (45.2%) Specialization 2. Groupement Farmers w/o 20.5% 79.5% 100.0% 0.0% 76 (23.7%) villageois mixte Specialization 3. Groupement Farmers 4.5% 95.5% 95.5% 4.5% 36 (11.2%) villageois eleveurs Production (GVE) ou Association des eleveurs . 4. Association Social 89.7% 10.3% 100.0% 0.0% 19 (5.9%) alphabetisation/sant6 . . . 5. Communaute ou Religious 0.0% 100.0% 68.4% 31.6% 18 (5.7%) association musulmane Table A2: The Five Most Important Local Associations in Houet Initiatd by Organized in Group Name Category Government Community Formal Informal Number of Times Cited (%/e of Total Citations) 1. Groupement Farmers w/o 22.1% 78.0% 61.4% 38.6% 125 (35.8%) villageois hommes Specialization . 2. Groupement Farmers w/o 22.7% 77.3% 55.7% 44.3% 85 (24.4%) villageois feminin Specialization 3. Groupement Farmers w/o 26.3% 73.7% 50.0% 50.0% 36 (10.3%) villageois mnixte Specialization 4. Groupement ou Farmers 56.7% 43.3% 70.0% 30.0% 29 (8.3%) Association de Production producteurs de coton. 5. Communaute ou Religious 17.4% 82.6% 34.8% 65.2% 21 (6.0%) association musulmane 39 Table A3: The Five Most Important Local Associations in Sissili Initiated by Organized in Group Name Category Government Community Formal Informal Number of Times Cited (% of Total Citations) 1. Groupement villageois Farmers w/o 11.4% 88.6% 81.3% 18.7% 122 (30.2%) hommes Specialization . 2. Groupement villageois Farmers w/o 7.1% 92.9% 60.0% 40.0% 79 (19.6%) feminin Specialization 3. Communaut6 ou Religious 0.0% 100.0% 16.7% 83.3% 35 (8.7%) association musulmane 4. Association des Social 2.4% 97.6% 0.0% 100.0% 32 (7.9%) parents d'e1lves et des meres 6ducatives (APE/AME) 5. Comite de gestion des Farmers 0.0% 100.0% 0.0% 100.0% 26 (6.4%) points d'eau (CPE) Services I_I_I Table A4: The Five Most Important Local Associations in Sanmatenga I _________ IInitiated by Organized in | Group Name Category Government Community Formal Informal Number of Times Cited (% of Total Citations) I . Groupement Fanners w/o 28.5% 71.5% 78.0% 22.0% 119 (47.4%) villageois mixte Specialization _ 2. Comite de gestion Farrners 1.8% 98.2% 0.0% 100.0% 54 (21.5%) des points d'eau Services (CPE) 3. Groupement Farmers w/o 2.9% 97.1% 17.1% 82.9% 28 (11.2%) villageois feminin Specialization 4. Groupe feminin de Farmers 0.0% 100.0% 82.4% 17.6% 16 (6.4%) gestion des Services ressources naturelles __ _ _ 5. Groupement Farmers w/o 0.0% 100.0% 86.7% 13.3% 15 (6.0%) villageois hornmes Specialization 40 REFERENCES Carter, M. 1997. "Environment, Technology, and the Social Articulation of Risk in West African Agriculture. " Economic Development and Cultural Change, 46: 557-590. CND (Commission Nationale de la Decentralisation). 1998. "Decentralisation rurale et institutions locales au Burkina Faso," Ouagadougou. (mimeo) Donnelly-Roark, P., Ye, X. and Ouedraogo, K. 1999. "Burkina Faso Local Level Institutions: Preliminary Research Analysis: Local Level Institutions and Rural Decentralization in Burkina Faso." Mimeo, Institutional and Social Policy Unit, Africa Region, World Bank, Washington, DC. Englebert, Pierre. 1996. Burkina Faso: Unsteady Statehood in West Africa. Boulder, Colorado: Westview Press. Fiske, A.P. 1991. Structures of Social Life: the Four Elementary Forms of Human Relations. New York: The Free Press. Grootaert, C. 1997. "Social Capital: The Missing Link?" Chapter 6 in Expanding the Measure of Wealth -- Indicators of Environmentally Sustainable Development. Washington, DC: World Bank. Grootaert, C. 1998. "Social Capital, Household Welfare and Poverty in Indonesia." Local Level Institutions Study Working Paper, Social Development Department, Washington, DC: World Bank. Grootaert, C. and D. Narayan. 1999. "Local Institutions, Poverty, and Household Welfare in Bolivia." Local Level Institutions Study Working Paper, Social Development Department, Washington, DC: World Bank. Grootaert, C., Oh, G., and Swamy, A. 1999. "Social Capital and Development Outcomes in Burkina Faso." Local Level Institutions Study Working Paper, Social Development Department, Washington, DC: World Bank. Kevane, M. and Englebert, P. 1997. "A Developmental State Without Growth? Explaining the Paradox of Burkina Faso in a Comparative Perspective." Mimeo, Santa Clara University. Maclure, R. 1994. "Misplaced Assumptions of Decentralization and Participation in Rural Communities: Primary School Reform in Burkina Faso." Comparative Education, 30, 3: 239-254. North, D. 1990. Institutions, Institutional Change, and Economic Performance. New York: Cambridge University Press. 41 Olson, M. 1982. The Rise and Decline of Nations. New Haven: Yale University Press. Pradervand, P. 1989. Listening to Africa: Developing Africa from the Grassroots. New York: Praeger. Putnam, R.D. 1993. Making Democracy Work: Civic Traditions in Modern Italy. Princeton, New Jersey: Princeton University Press. Reardon, T., Delgado, C. and Matlon, P. 1992. "Determinants of Income Diversification Amongst Farm Households in Burkina Faso." Journal of Development Studies, 28, 2: 264-296. Sen, A. 1988 "The Concept of Development." In The Handbook of Development Economics, Vol. I., p. 10-25. Amsterdam: North Holland. Smale, M. and V. Ruttan. 1994. "Cultural Endowments, Institutional Renovation and Technical Innovation: the 'Groupements Naam' of Yatenga, Burkina Faso." Economic Development Center Bulletin No. 942. Minneapolis/St. Paul: University of Minnesota. Spiers, M. 1991. "Agrarian Change and the Revolution in Burkina Faso." African Affairs, 90: 89-110. Veirech, H. 1986. "Agricultural Production, Social Status, and Intra-Compound Relationships." In J.W. Moock eds. Understanding Rural Households and Farming Systems. Boulder and London: Westview Press. World Bank. 1991. Burkina Faso: Fourth Education Project. Population and Human Resources, Operations Division, Sahelian Department, Africa Region. Washington, DC: World Bank. World Bank. 1998. "The Local Level Institutions Study: Program Description and Prototype Questionnaires." Local Level Institutions Working Paper No. 2, Social Development Department, Washington, DC: World Bank. 42