Cost recovery for health and water projects in rural Mali : household ability to pay and organizational capacity of villages

PHN Technical Note 87-3 COST RECOVERY FOR HEALTH AND WATER PROJECTS IN RURAL MALI: HOUSEHOLD ABILITY TO PAY AND ORGANIZATIONAL CAPACITY OF VILLAGES by Martha Ainsworth Francois Orivel Punam Chuhan Originally issued in Country Policy Department Discussion Paper No. 1983-8, February 1983, "Three Studies on Cost Recovery in Social Sector Projects." Reissued January 1987 Population, Health and Nutrition Department World Bank The World Bank does not accept responsibility for the views expressed herein which are those of the author(s) and should not be attributed to the World Bank or to its affiliated organizations. The findings, interpretations, and conclusions are the results of research supported by the Bank; they do not necessarily represent official policy of the Bank. The designations employed, the presentation of --terial, and any maps used in this document are solely for the convenience of the reader and do not imply the expression of any opinion whatsoever on the part of the World Bank or its affiliates concerning -the legal status of any country, territory, city area, or of its authorities, or concerning the delimitations of its boundaries, or national affiliation. PHN Technical Note 87-3 COST RECOVERY FOR REALTE AND UATER PROJECTS IN RURAL MALI: HOUSEROLD ABILITY TO PAY AND ORGANIZATIONAL CAPACITY OF VILLAGES ABSTRACT This paper deals with several aspects of the feasibility of a plan for village-level cost recovery for health and water supply projects in southwestern Mali, specifically: 1) Can rural villagers collectively or individually afford to finance these services; and 2) does the organizational capacity exist at the village level to implement such a scheme, i.e., to collect and manage the funds thus mobilized? The analysis utilizes two sources of data: a 1981 household survey and a 1982 survey of the 38 villages from which those households were drawn. The households surveyed have significant sources of cash income and the contributions envisaged by the projects are within the means of most households. Data from the household and village surveys suggest that these communities have some experience with financing the investment costs of local projects, but little, if any, in financing recurrent costs for collective services. The ability of villages to mobilize cash for purchase of spare parts and repairs is as critical to the upkeep of a tubewell as is a .trained mechanic. Projects in other parts of Mali utilizing the "village contract" approach are experiencing conditional success, but -have been operating only a short time; the longer-term feasibility of village-level finance of recurrent costs is unknown. These projects have often been associated with other development activities in villages which are less remote than most villages in the region studied. Projects elsewhere which rely on local finance of village health workers have run into numerous problems, but it is not clear to what extent they can be attributed to inadequate finance mechanisms, rather than low willingness to pay for the services provided. Exploration of financing alternatives with village authorities, and insuring that the option chosen is viable and accepted by those who will use the service, is recommended. It cannot simply be assumed in delegating responsibility for recurrent finance to villages that appropriate and viable structures will appear where they previously never existed, particularly in a zone with low literacy and where recordkeeping and accounting are virtually unknown. Prepared by: Martha Ainsworth, Francois Orivel and Punam Chuhan Consultants to the World Bank This is one in a series of papers prepared under Research Project 672-72: Demand for and Willingness to Pay for Health and Schooling in Mali. Originally issued in Country Policy Department Discussion Paper No. 1983-8, February 1983, "Three Studies on Cost Recovery in Social Sector Projects." Reissued January 1987 Cost Recovery for Realth and Waater Projects in Rural Mali: Household Ability to Pay and Organizational Capacity of Villages Contents I n t r o d u c t i o n . . . . . . . . . . . . . . * * * * * * * * * * * * * * * * * * * * * * **. 1. Household Ability to Pay.....*.t..******************.. Background. e ... . . . ..s..... ..** ** *** ** *** ** ** *** ***3 Ca a sh Income. . . . . . . . . **.*. * * . . . . * * * * . . * * . * * 6 Cash Expenditures....... . .a**********************9 II. Organizational Capacity of Villages for Cost Recovery... . -.... .. ..*** .* * .** . ** .*.*......... 3. Project Specifications...o....0.0..****..... *. .13 Cash Management Experience of KBK Villages... ......14 III. Conclusions..... .......... ...************.*. *o * *.. ... 18 Appendix A. Note on the Calculation of Household Income.......20 Introduction In attempting to reconcile the need for vastly expanded social services with the limited availability of public funds, African governments and aid donors are closely examining new sources of finance. In Mali, for example, fewer than a third of school age children are enrolled in primary school and perhaps ouly one quarter of the population has access to modern medical care. Levels of finance for existing public services are woefully inadequate, however, as evidenced by the pervasive lack of drugs, textbooks and recurrent funds for fuel at public facilities. Given the poor economic climate, the resulting slow growth in conventional sources of finance -- mainly through taxation - and the reluctance of aid donors to finance the recurrent costs of social services, the rate of expansion of limited systems of health and schooling will rely increasingly on the ability to mobilize additional resources from untapped sources. One option for social sector finance is direct cost recovery from beneficiaries of health and schooling to finance at least some of the cost of providing these services. For many historical, political and equity-related reasons, user fees or local subscriptions are a little used source of finance in the social sectors, despite the considerable private benefits arising from many. Aside from the currently debated equity and efficiency effects,of such a scheme L/, its success in terms of providing additional social sector finance will rely equally on households' ability and willingness to pay, on 1/ c.f. Nancy Birdsall, "Strategies for Analyzing the Effects of User Charges in the Social Sectors", CPD, July 1982; David de Ferranti, "Realth Sector Financing and Expenditure in Developing Countries: Current Issues", PHN, February 1983; Mateen Thobani, "Efficiency and Equity Implications of User Chnarges in Social Sector Services", CPD, May 1982. the one hand, and on the capacity of the administration at the relevant levels to collect, manage and disburse funds appropriately, on the other. This paper will deal with several aspects of the feasibility of village-level cost recovery for health and water supply projects in Southwestern Hali, specifically: ') can rural villagers collectively or individually afford to finance these services; and 2) does the organizational. capacity exist at the village level to implement such a scheme, i.e., to collect and manage the funds thus mobilized? Our analysis 4i1l utilize vo sources of data: a household survey conducted in 1981 as preparation for World Bank health and water supply projects; and a survey of the 38 villages from which households were drawn, conducted in 1982. The first section of the paper will evaluate household 'ability to pay" based on expenditure data, information on the sources of income and the estimated cost of the proposed health and water projects. Key issues here will be the availability of cash to rural households and the composition of 'their reported cash expenditures. The second section will evaluate the experience of these villagers in collecting and managing cash utilizing data on community organizations collected during the 1982 village survey. An additional issue cen=al to feasibility of village-level cost recovery household "willingness to pay" - is the subject of separate paper. 1/ t/ Nancv Sirdsall and Punam Chuhan, "Willingness to Pay for Health and ;ater in Rural Mali: Do WT Questions Work?", Country Policy Depar:-ent,.arch 1983. -3- I. Household Ability to Pay for Health and Water Projects Village-level cost recovery has often been dismissed as a potential source of recurrent finance based on the presumption that many -ural populations are simply too poor to contribute or lack cash. Following an overview of the 1981 household survey in Mali, the proposed health and water projects and the costs to be financed by villages, this section will examine the sources of income and cash expenditure patterns to. assess household ability to pay" for the projects. The'discussion will show that the households surveyed - located in one of the least developed and most remote areas of Mali -- do have considerable sources of cash income and that the contributions envisioned by the Bank health and water projects appear to be within the means of the average household surveyed. Background The area to be covered by the proposed health and water projects comprises three contiguous districts, or "cercles", in the region of Kayes in Southwestern Mali: Bafoulabe, Kenieba and Kita ("KBK'). This region is distinguished by its rugged terrain, poor communication and transport and dispersed population. Per capita income of less than $200 is among the lowest in the world. Fewer than one third of school age children are enrolled in primary school, infant mortality is in the vicinity of 255 per 1000 and life expectancy at birth is only 32 years. I/ The vast majority of the population is engaged in subsistence agriculture. To help in the design and preparation of health and water supply projects requested by the Malian government,. a household survey of K3K was 1/ Ecole Nationale de Medicine et de Pharmacie du Mali, "Evaluation Sanitaire des cercles de Kenieba, Bafoulabe, Kita", Rapport preliminaire, mai, 1981. -4- commissioned by the Bank in 1981. The sample consisted of 186 households in 38 villages; 64 households were located in Bafoulabe, 52 in Kenieba and 70 in Kita. The villages were selected on the basis of water scarcity; 4 to 5 households were randomly interviewed. in each village. I/ Data was obtained on agricultural production, cash expenditures, income, demographic character- istics of households and a number of issues related specifIcally to project design. The rural water supply project proposed for KBK consists of drilling and installation of 325 tubewells equipped with hand operated pumps which would be maintained and repaired, for the most part, by a pump 'caretaker" selected by the village and trained by the project. Each village would finance a portion of the investment costs, remuneration in cash or in kind of the pump caretaker, the cost of spare parts, and reimbursement to arrondissement and cercle authorities for repair and emergency assistance beyond the capabilities of the caretaker. These expenses are summarized in Table 1. The village contribution to the invest=ent-270,000 M, or about $500 2/ is extremely small compared to the average installation cost per productive equipped borehole of $20,000. The recurrent contribution of villages, including remuneration of the pump caretaker, would represent about 65 percent of the estimated recurrent cost per water point per year. 3/ Each tubewell will be able to serve, at a maximum, 500 persons. The mean size of 1/ The sample was not statistically representative of MK. See the discussion of how survey villages may differ in "The Mali Household, Co=muni:7 and Service Provider Surveys: Documentation', C?D,. March 1983. 2/ 1981 exchange rate, US$t.00 - 543 Mr. 3/ Not included in the Table is the cost of pump replacemen: (about 6C0,C00 MF each), which must be done every 5-6 years. -5- the villages surveyed was closer to 800 persons,-however. .Therefore, for the average" village in our sample 2 wells would have to be drilled with 400 persons supporting each. Table 1: Costs to be Borne by Villages in the Rural 'ater Suoply Project (per tubevell) Item Amount (MF) Inves tent costs 270,000 One third of the purchase price of the pump 220,000 Ten bags of cement 50,000 Recurrent costs (Per year) 62,500 Spare parts 15,000 Reimbursement of arrondissement repairman and for emergency assistance by cercle 11,500 Remuneration of pump caretaker a/ 36,000 a/ Estimated value, to be paid in cash or in kind; amount and nature of remuneration to be determined by each village. At the time that the household survey was conducted, one of the options under consideration for the health project was a village health worker scheme in which one or more persons selected by the village would be trained by the project in first aid, sanitation and preventive care and financed by the village. The annual cost per village was estimated to be 65,000 MF or the in-kind equivalent. 1/ I/ See Birdsall and Chuhan, 1983. The investment and recurrent casts per capita and per househoId o: both projects are summarized in Table -2. The investment cost per household for our average village of 800 persons ranges from 7700-10,400 MF and the combined recurrent costs of both projects amount to 2700-3700 per household per year. Table 2: Per capita and per Household Costs for Health and Water Projects (1F) b/ (Assumption: Village of 800 persons, two tubewells] Per capica Per Household a/ Bafoulabe Kenieba Kics Investment cost, well 675 7817 7736 LO,402 Recurrent cost, wall 156 1806 1788 2404 Recurrent cost, health worker 81 938 928 1248 Total recurrent costs for well 237 2744 2716 3652 and health worker a/ Mean household size: Safoulabe, 11.58; Kenieba, 11.46; Kita, 15.41. b/ 1981 exchange rate: US$1.00 - 543 4F. Cash Income Although remuneration of the pump caretaker and village health worker may be in cash or in kiad, a number of village-f inanced compocencs of the water project will require cash: the investment costs; spare parts; reimbursement to arrondissement and cercle-level authorities for emergency assistance. The project area is extremely remote, capital markets are very poor, and most of the population is engaged in production of subsistence crops, such as millet, sorghum and maize. In assessing ability to pay, -7- therefore, it is important to know, first, whether households have any major source of cash income and second, how reliable the cash income from these sources is likely to be. In Table 3 income is broken down by source for households in the three carcles. Our estimates of per capita income, ranging from a low of 50,000 MF in Kita to about 60,000 MF in Bafoulabe and a high of 81,000 MF in Kenieba, compare well with national per capita income of 80,000 MF (US $150). The high proportion of income from nonagricultural activities in Bafoulabe and Kenieba households is unexpected and points to substantial sources of cash income. Cash remittances account for 43 percent of total household income in Bafoulabe, while income from gold prospecting, remittances and other cash-earning activities amounts to over half of total income in Kenieba. Most income from agricultural production in Bafoulabe and Kenieba is in kind; only about one fifth of crop income is from sale of crops. I/ The nature of income from livestock sales is ambiguous. In contrast to households in the other two cercles, those in Kita derived 75 percent of income from crop production; only 10 percent of income can be attributed to remittances and. other cash-earning activities. A much higher proportion of crop production in Kita is comprised of groundnuts -- a cash crop -- than in the other 2 cercles, however. Roughly half of crop income is earned from sale of groundnuts for cash, while 6he remainder is in- kind subsistence crop income. The sources of cash income in KBK may be subject to seasonal fluctuations or great variation from year to year, however. Income from 1/ Agricultural income was calculated from production figures; see Appendix A. Table 3: Household Income by Source- Bafoulabe Keniaba Kita Income source 000 MF % 000 MF % 000 MF Net value of crops a/ 228.Z 33 277.4 30 580.6 b/ 75 Annualized value of livestock C/ 119.0 17 122.6 13 105.4 14 Gold 0.7 .. 342.4 37 0.6 Remittances 291.5 43 55.3 6 28.2 4 Other d/ 45.2 7 131.8 14 60.2 8 Total income 684.7 100 929.8 100 775.1 100 Persons per household 11.58 11.46 15.41 Per capita income 59.1 81.1 50.3 Note: 1981 exchange rate, US$1.00-543 MF. a! Gross value of crops less value of inputs. Agricultural income was 7stimated based on total production, estimates of household self-consumption and local buying and selling prices (see Appendix A). b/ More than half is income from groundnuts, which most likely represents Tash income; value of groundnuts produced in Bafoulaba and Keniaba households is only 21 percent of gross value of crops. c/ See Appendix A for explanation. d/ Cash income from retail and petty trade, hunting and gathering, orchards, spinning and weaving, tailoring, sale of milk, pensions, gardening, traditional medicine, other. f9 groundnut cultivation is perhaps the most unreliable cash source, as it is subject to the vagaries of rainfall, temperature and pests. Further, income from sale of groundnuts is likely to occur immediately after harvest, meaning that in the absence of credit markets cash availability throughout the rest of the year may be quite low. Income from remittances is also likely to be subject to similar fluctuations to the extent that money is remitted from agricultural labor in neighboring African countries. I Gold prospecting is a year round source of cash income in Kenieba although it likely falls off somewhat in periods of high demand for agricultural labor. Since gold prospecting is also reliant on chance, however, the stability of .this source of cash is in doubt. Thus, the households surveyed appear to have important sources of cash income, but because the sources are tied to agriculture and credit markets are poorly developed, cash availability in Kita and possibly in Bafoulabe may peak at harvest and be subject to considerable annual variation. Cash Expenditures Given that households have substantial cash income, how much of this, if any, is available for meeting project costs? Below we will examine the composition of household cash expenditures for evidence of ability to pay for health and water projects. 2/ 1/ The proport-4ii of remittances deriving from agriculture is not known. In Bafoulabe, however, 30 percent of absent household members were living in other African countries, 28 percent were residing in France, and 39 percent were elsewhere in Mali. Only about one third of those absent were engaged in commerce or the civil service. 2/ Questions on debt and savings as well as a number of expenditure items (e.g., food, transport, schooling costs) were not included in the household survey. There will be no attempt, therefore, to reconcile or compare income and expenditure data. The KBK household questionnaire collected data on annual cash expenditures for clothing, taxes, medicines and childbirth, agricultural inputs, community projects, feasts, celebrations, marriages, baptisms, gifts, cattle, durable goods, and loans to others. Itemized food expenditures for the year were not collected, but we were able to obtain a rough estimation of annual cash expenditure on subsistence grain by comparing annual household Table 4: Annual Household Expenditures, 1980-81 (p proportion of households reporting an expenditure item) Bafoulabe Kenisba Kita Exvenditure 000 MF (p) 000 IM (p) 000 Is (p) Agricultural inputs 17.2 (.34) 21.7 (.48) .16.3 (.40) Clothing 10.3 (.37) 18.9 (.69) 12.8 (.46) Taxes 19.2 (.90) 15.8 (.89) 32.2 (.97) Drugs 13.7 (.58) 16.7 (.41) 23.2 (.66) Childbirth and baptism 4.7 (.26) 3.3 (.31) 2.9 (.43) Cattle purchase 2.9 (.03) 5.7 (.06) 0.9 (.04) Durable goods a/ 36.6 (.44) 46.4 (.69) 30.9 (.57) CoMIMnity projects b/ 10.5 (.50) 5.9 (.67) 12.1 (.23) Feasts 11.8 (.98) 4.6 (.85) 10.6 (.89) Celebrations 36.5 (.06) 143.2 (.09) 21.6 (.14) Marriage 28.9 (.18) 69.7 (.13) 136.4 (.23) Gifts 1.2 (.15) 1.9 (.06) 3.9 (.37) Loans to others 23.6 (.26) 3.7 (.13) 37.6 (.26) Value of grain purchased c/ 309.2 254.4 352.8 Total Expenditure 326.3 611.9 694.2 Persons per household 11.58 11.46 15.41 Per Capita 'xpenditure 45.4 53.4 45.0 NOte: Grain as a percent of total: 58.5% 41.6% 50.8% a/ Hoped, radio, bicycle, gun, lancern, flashlight, house with tin roof. T/ School, mosque, dispensary construction. c/ Food deficit (production - consu=ptlon) t retail price of subsistence irains. consumption and production figures, and multiplying the production deficit (i.e., the quantity presumed purchased) by the retail price of subsistence grain. Thus, it was possible to reconstruct the approximate amount and composition of cash expenditures by households in KBK, shown in Table 4. The households surveyed had cash expenditures in the range of 45,000-53,000 MF per capita, or roughly US $80-100 at 1981 exchange rates. Purchase of grains accounted for 42-59 percent of total expenditures. This proportion may seem low, since up to 80 percent of total household expenditures may be spent on food by the poorest income groups in some countries. It must be kept in mind, however, that only expenditure on grain has been estimated here; 6xpenditures on other foods may be considerable and therefore raise this proportion (see Table 5). A second important consideration is that households produce and consume a substantial amount of their requirements for food grains; the proportion of total expenditures spent on food grain reflects only that amount which must be purchased, a fraction of the total consumed. This also implies, however, that in a poor agricultural year we can expect expenditures on food.grain to jump considerably as a proportion of the total, as families must buy additional grain to compensate for their poor yields. Note that households in Bafoulabe and Kita, with the lowest per capita expenditures (45,000 MF), spent a higher proportion of total expenditure on grain (59 and 51 percent, respectively) than did those in Kenieba (42 percent) where per capita expenditures are higher (53,000 MF). These findings generally agree with the observation that as incomes rise, the proportion of expenditure spent on food declines. - 12 - Table 5: Exoenditure on Other Foods- Proportion of Households Which Spend... Less than 5,000 MF - More than Food Itam Nothing 5000 IMF 10,000 aF 10,000 MF Meat 0.00 0.51 0.24 0.24 Fish 0.18 0.69 0.09 0.03 Sugar 0.06 0.63 0.12 0.19 Salt 0.03 0.13 0.45 0.39 Somewhat surprising and indicative of an ability to pay for the health and water projects is the high expenditure absolutely and proportionately on social activities (community projects, feasts, celebrations, marriages, gifts), durables and loans to others in Table 4. These expenditures collectively amounted to 28 percent of total household expenditures in Bafoulabe, 45 percent in Kenieba and 36 percent in Kita, or from 13,000-24,000 MF per capita. Closer examination of the data reveals that some of these 'average" household expenditures are in fact large outlays by only a. few of the households in the sample, particularly for marriages and celebrations (see figures in parentheses in Table 4). Expenditures on durable goods, community projects and feasts, on the other hand, were reported by a much larger proportion of households. Expenditures per capita on these items are compared to per capita cost of the water and health projects in Table 6. Households spent 3500-5000 'IF per capita on durable goods, community projects and feasts in 1980-81, while per capita inves=ent and recurrent costs for :he projects amount to only 675 MF and 237 MF, respectively. These costs appear to be within the means of the households surveyed. Table 6: Per Capita Expenditures on Durable Goods and Social Activities Compared to Estimated Project Costs (000 HF) Bafoulabe Kenieba Kita Per capita expenditure on: Durable goods 3.16 4.05 2.01 Community projects 0.91 0.51 0.79 Feasts 1.02 0.40 0.69 To tal 5.09 4.96 3.49 Project costs per capita: Per capita investment cost, well project 0.675 Per capita annual recurrent cost, well and health project 0.237 II. Organizational Capacity of Villages for Cost Recovery Successful mobilization of community resources for financing the investment and recurrent costs of the water and health projects will rely not only on the ability and willingness of households to pay, but also on the capacity at the village level to collect and manage these funds. Data collected from the KBK household survey and a subsequent survey of village organizations-discussed below-suggests. that these villages had some experience in financing investment costs for community projects but little or no experience in collecting, managing and disbursing cash for ongoing community services. Project Specifications The water supply and health projects specified the types of costs to be recovered at the village level, but not the finance mechanisms to be utilized. The water supply project, for example, will enter into a contract with a village committee that specifies which services will be financed by the village and which will be provided by the project. It is left to the discetion of the committee, chief, or other authority to work out the more complicated finance issues, such,as: 14 - whether the various components (e.g. salar±es) will be financed collectively or through other methods, such as fee for service or drug sales (for the health worker); whether payments will be assessed per household, per person or per use, in cash or in kind; whether certain persons or households will be exempted; M whether collections should occur annually, monthly or periodically; - who will manage, collect and disburse payments; and how compliance will be enforced. Cash management experience of KBK villages Data from. the household survey suggest that the villages studied had had some experience in mobilizing cash, labor or materials for cons truction of local health or schooling facilities. Recall, for example, that half of the households surveyed in Bafoulabe, two thirds of those in Keniaba and about one fourth in Kita reported cash expenditures for construction of schools, dispensaries or mosques in 1980-81. Further, of the 21 primary schools serving these households, 19 had been built with comunity assistance and for 15 the assistance was partly in cash. Information on local experience in financing recurring activities was collected 9 months after the K3K household survey when fieldworkers returned to the 38 villages from which households had been drawn with a village-level questionnaire administered to "respondent groups", usually consisting of a village chief, members of a council and other villagers. I/ One section of the questionnaire focussed on existing village organizations, particularly those which rely on cash contributions and which maintain some type of treasury. Information was collected on: the type of payment (cash, .kind, labor); who contributes; how the resources mobilized are used; who makes decisions on the allocation and management of funds; major operational problems; and the ethn4.c composition of the organization. A total of 58 organizations were reported by respondent groups in 30 villages; 8 villages reported no organizations with a treasury. There were 5 major types of organization which can be classified as either traditional or modern. The two major types of traditional organizations were tons and classes d"age. Briefly, tons are composed of the young people within a village, roughly between 15 and 35 years of age, and headed by a member of the chief's household. These groups hire out their labor to individuals in the community for which they typically receive payment in kind-food during the work period and perhaps a sheep or head of cattle. This, in addition to any cash payments, comprises their "treasury". The ton may also donate its labor to community projects, such as school construction. Classes d'age are groups of young people who were initiated (circumsized) at roughly the same date. 1/ The'purpose of this survey was to collect data on. communities and the health and schooling services utilized by them to link with the household data as the basis of an analysis of demand for health and schooling. See "The Mali 'dousehold, Community and Service Provider Surveys: Documentation", Country Policy Department, 1983. - 16 - They are subgroups within the ton, but may also offer their services independently and maintain their own accounts. 1/ Modern organizations in the survey villages included: parent associations (associations des parents d'eleves), all linked to schools within the villages; branches of the UDPH, Mali's political party, to which all households must contribute; and FGRs, rural consumer cooperatives. The financial experience of village organizations is presented in Table 7. The modern organizations were more likely to collec: cash, but with the exception of the FGRs, it was collected and managed by outside authorities. Respondent groups cited problems of shortfalls in funds, mismanagement and corruption in these organizations. Traditional organizations were controlled from within the village, but generally dealt with collection of payments in kind for labor provided to individual households-not cash transactions for community services. There appeared to be few problems with the traditional organizations, although respondents no doubt understated problems with those organizations for which they were responsible. Thus, with the exception of the 3 ?GRs, villages had virtually no experience in collecting and managing funds. And the record for FGZs was spotty: all 3 had management problems and 2 were short of cash. 2/ 1/ Traditional organizations in K3K are discussed in greater detail in SNED, "Project d'alimentatlon en eau et developpement sanitaire: Rapport sociologique", Samako, fevrier 1982. This report gives Gxample of prices charged by tons, both in cash and in kind. Our survey data indica:a that =ost are paid in kind, however. 2/ In a fourth village the FGR had complemly disbanded. ... . .. - 17 - Table 7: Village Organizations a/ (Figures are the number of organizations.) Characteristics Traditional Organizations Modern Organizations Type Ton 23 Parent association (APE) 14 Classes d'ages 5 Political party (UDPM) 6 Cooperatives (FGR) 3 Total 28' Total 23 Use of Contributions Fastivities 27 School finance 14 Community dev't 6 Political 6 Other 3 Cooperative 3 Type of Contributions Cash 8 Cash 22 Kind 9 Kind 0 Labor 21 Labor 8b/ APE 271.2 Amount of cash UDPM 100.0 payment (M,F) c/ FGR 550.0 Decision maker Council 19 Outside authority 17 Membership 8 Council 3 Chief 1 Membership 1 Problems None 18 Shortage of funds 12 Decline in 5 Mismanagement 4 membership Corruption 4 Other (including 4 None 3 corruption) a/ A total of 58 organizations were reported; this table summarizes 5 major types of which there were 51. Others included religious groups (4), other traditional young peopleos groups (2) and an organization of village youth who had spent time in France (1). b/ All 8 were parent assocations. c/ Insufficient response -18 - III. Conclusions In assessing the feasibility of financing part of the invescent ard a significant portion of recurrent costs of health and water projects in K3, we have examined household ability to pay and the experience of villages in collecting and managing funds for collective activities. Based on analysis of sources of income and household expenditures, it appears that the households surveyed do have significant sources of cash income and that the contributions envisaged by the projects are within the means of most households. A third crucial issue--the willingness of households to pay for the proposed services--is discussed in a separate paper. / Data from the KBX household and village surveys suggest that these communities had some experience with financing :he investment costs of local projects, but little, if any, in financing recurrent costs for collective services. The ability of villages to mobilize cash for purchase of spare parts and repairs is as critical to the upkeep of a tubewell as is a trained mechanic. Projects in other parts of Mali utilizing the "village contact" approach are experiencing conditional success, but have been operating only a short time; the longer-tar= feasibility of village-level finance is unknown. 2/ Further, these projects have often been associated with other development activities in villages which are less remote than most villages in K0K. Projects elsewhere which rely on local finance of village health workers have run into numerous problems, but It is not clear to what extent they can 1/ Sirdsall and Chuhan, 1983. 2/ iaLVETAS and 'M? both have village - financed water supply projects in Mali, the forner located in the Mali-sud project zone. Village face is also under consideration in Upper Volta. For problems encountertd in a similar scheme in Lesotho, see Richard Feachem ct al., 7ater, Health and Development. London: Tri-Had Books, L978. - 19 be attributed to inadequate finance mechanisms, rather than low willingness to pay for the services provided. 1/ All of this points up the importance to the project of thoroughly exploring finance alternatives with village authorities and insuring that the option chosen is viable and accepted by those who will use the service. IL cannot simply be assumed in delegating responsibility for recurrent finance to villages that appropriate and viable structures will appear where they previously never exis ted, particularly in a zone with such low li teracy and where record-keeping and accounting are virtually unknown. 1/ A recent review of AID-assisted health projec-ts provides numerous examples of vi/llage health worker finance, both successful and unsuccessful. Villagefinance of salaries seems to encounter greater difficulty than schemes which finance health workers through fee for service or drug sales. See Hichael Favin, AID-Assisted Primary Health Care Projects: Summary Reviews, 1981. - 20 Accendix A Note on the Calculation of Household Income 1. Crop Income Income from crops was calculated from production data. Income from groundnuts-a cash crop-was obtained by aultipl-,ing groundnut production by the farmga.ce price of groundnuts, as reported by respondents. The portion of subsistence crop production which is self-consumed should be priced at its opportunity cost, which in most circumstances is the farmgate price. In this area, however, the retail price of subsistence crops. was twice the farmgate price. By using the farmgate price for valuing self- consumed grain, we would be undervaluing agriculture against alternative activities, as such a substitution would not provide the family vith an equivalent quantity of food. Thus, the portion of subsistence crop production consumed by the household was valued at the retail price and any surplus production was valued at the farmgata price. I/ The method used for calculating income from agricultural production produces a higher figure for total household and per capita income than would have been the case had we used the fargate,price as the oppotunity cost of all production as shown in the :able below. The latter method would not change the ranking of income by cercle, but would result in a per capi:a income estimate roughly 10,000 MF less for all three cercles. Note that the 1/ The quantity of grain consumed-annually by each household was esti=a:ed by multiplying the average daily consumion per capita by the number of household members, and this figu.re times 365 days. The average daily consumption of subsistence grains per capita reported by households was 0.513 kg, which provides nearly 2000 kcal. Th4s figure compares favorably With more elaborate nutrition surveys carried out in similar areas. - 21 proportion of income derived from agriculture in*Bafoulabe.and Kenieba would decline from roughly one third to less than one fourth of total income, while in Kita crop production would still comprise 69 percent of the total. Income estimates (000 HF) Bafoulabe Kenieba Kita (1) (2) (1) (2) (1) (2) Net value of crops 228.2 133.5 277.4 171.4 580.6 427.1 Household income 684.7 589.9 929.8 823.8 775.1 622.0 Per capita income 59.1 50.9 81.1 71.9 50.3 40.4 No te: Assumptions on opportunity cost of consumed subsistence production: (1) faragate price for production sold; retail price for that consumed; (2) farmgate price for all production. 2. Income from livestock The households surveyed had considerable livestock holdings, primarily cows, sheep and goats-(see Table below). Livestock is a major asset in KBK, since land has no selling value and there appears to be no housing market. I/ Annual income dervied from livestock holdings was estimatedby valuing the stock at its current market price and dividing the value of each animal by the estimated number of years of market value. 1/ There may be a market for houses with steel roofs (which are not common), but transactions remain rare. - 22 Lives tock Ownership- (number of head owned per 7household) Animal Bafoulabe Keniaba Ki ca Cows 10.67 7.79 7.41 Sheep 4.56 8.31 2.86 Goats 3.81 4.13 1.74 Total 19.04 20.23 12.01 . . . . . . ..