Upper Volta - Third Rural Development Fund Project

Document of F IL E C - PY The World Bank FOR OFFICIAL USE ONLY Report No. 3621-UV UPPER VOLTA THIRD RURAL DEVELOPMENT FUND PROJECT STAFF APPRAISAL REPORT February 24, 1982 Western Africa Projects Department Agriculture Division 5 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS SDR 1 US$1.18 = CFAF 319 US$1.00 = SDR 0.84 = CFAF 270 CFAF 1,000 = US$3.70 = SDR 3.11 WEIGHTS AND MEASURES Metric Systems ABBREVIATIONS AVV - Volta Valleys Development Authority BCEAO - Central Bank of West African States BOAD - West African Development Bank CERCI - Irrigated Crops Study and Research Center CNCA - National Agricultural Credit Bank DAFR - Directorate of Forestry Management and Reforestation DSA - Agriculture Department (MRD) FAER - Rural Equipment Fund HER - Rural Water Supply and Engineering Department (MRD) ICRISAT - International Crops Research Institute for the Semi-Arid Tropics IMTC - Inter-Ministerial Technical Committee IRAT - Tropical Agriculture and Foodcrops Research Institute IRCT - Cotton and Textiles Research Institute IRHO - Oils and Oilseeds Research Institute MRD - Ministry of Rural Development OFNACER - National Cereals Agency ORD - Regional Development Organization PY - Project Year RDF - Rural Development Fund RMWA - Regional Mission in West Africa SC - Standing Committee of IMTC SOFITEX - Voltaic Textile Fibre Company (Parastatal) FISCAL YEARS Government January 1 - December 31 Project October 1 - September 30 IDA : July 1 - June 30  UPPER VOLTA FOR OFFICIAL USE ONLY THIRD RURAL DEVELOPMENT FUND PROJECT STAFF APPRAISAL REPORT TABLE OF CONTENTS Page No. I* BACKGROUND . *.......*e*ee.ceee***.e*e*e.e****. .*ee........ 1 A - The Economic Setting * .................................. 1 B - The Rural Sector .... 1 C - Sector Strategy, Development Constraints and Matjor Issues ...6 II. THE PROJECT AREA .................... 8 A - Ecology and Population . ......a8 B - Earlier Projects 9 III. THE PROJECT ..........11 A - Project Origin and Preparation ..................11 B - Objectives and Summary Description... 0*0...... 11 C - Detailed Features .... 12 D - Cost Estimates .... .... 21 E - Proposed Financing.. . . . 22 F - Procurement ........ 25 G - Disbursement .............**.. . ......... .e. .. 26 H - Accounts, Auditing and Reporting .................* .... 27 IV. ORGANIZATION AND MANAGEMENT ............................. 28 A - Sub-project Selection Criteria and Project Cycle ........ 28 B - RDF Organization and Staff .............. ............. 31 V. TECHNOLOGICAL AND PRODUCTION SPECIFICATIONS .............. 33 * A - Investment Phase ceeecceeecec.....e.ce....c.0.c... c 33 B - Operating Phase ...c... ... eece.....c...*..e 33 C - Yields and Output cc...e..............e.................. 34 Vl. DEMAND, MARKETS AND PRICES; FINANCIAL RESULTS ....... ...... ... . 36 A - Demand, Markets and Prices ....c.. o....ece..e....e..e.. 36 B - Financial Results ....eecc. ......eooo.o...c... o ......c.. 37 This report is based on the findings of an appraisal mission consisting of Messrs. Osei, Cole, Meimaris, Simsolo (RWA), Thiam and Traore (BOAD), van Bolhuis and Zijderveld representing the Dutch Government and Loaec (Consultant) which visited Upper Volta in March/April 1981. This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  TABLE OF CONTENTS (Cont'd) Page No. VII. BENEFITS AND ECONOMIC ANALYSIS ..... ......................... 39 A - Major Benefits and Benefit Distribution ................. 39 B - Economic Rate of Return ..................... ............ 40 C - Risks ................................................... 41 VIII. AGREEMENTS REACHED AND RECOMMENDATION ....................... 41 ANNEXES 1. Organization and Management 2. Supporting Tables Table 1: Detailed Project Cost Table 2: Financing Plan and Estimated Schedule of Disbursements Table 3: Government Cash Flow Table 4: Economic Rate of Return Calculation 3. Selected Documents and Data in Project File A. Working Papers to Guide Implementation 1. Land Development Sub-projects 2. Agricultural Development 3. Agricultural Support Services 4. Agricultural Credit 5. Production of Animal Traction Equipment 6. Rural Forestry 7. Village Water Supply (in French) 8. Basic Data for Economic Analysis B. Reports and Studies Related to Project 1. Troisieme Fonds de Developpement Rural; RDF; November 1980. (Preparation Report in 3 vols. + summary and 12 Annexes, French) 2. Instructions Generales sur les operations etc; RDF; November 1973. (Subproject Appraisal Procedures; in French) 3. Rapport de Suivi sur les Amenagements Hydroagricoles, campagne 79/180, RDF. (First Evaluation Report on Land Development Schemes (1978/79) in French). LIST OF TABLES IN TEXT 1. Project Cost Summary 2. Financing Plan 3. Disbursement Categories and Amounts 4. Key Technical Parameters: Land Development Schemes 5. Production Specifications; Yields; Output 6. Farmgate Prices, Market Prices and Economic Values 7. Summary Farm Budgets 8. Summary Results of Economic Rate of Return Calculations IBRD Map No. 16064  UPPER VOLTA THIRD RURAL DEVELOPMENT FUND PROJECT STAFF APPRAISAL REPORT I. BACKGROUND A. The Economic Setting 1.01 Upper Volta is classified as among the 29 "least developed coun- tries" with GNP per capita estimated in 1979 at about US$180. Agricultural production activities which provide employment for the 90% of the population classified as rural, are severely constrained by low and uncertain rainfall and generally poor soils and the concentration of such activities on the densely-populated central plateau where nearly 65% of the total population occupy 33% of the land area. Domestic expenditures for both consumption and investment exceed GDP (by about 30% in recent years) and a large part of the difference is made up by emigrants remittances and pensions paid by France. Even after such transfers, Upper Volta still needs foreign aid grants and loans equivalent to as much as one fifth of its GDP in order to make ends meet. There are plans to rework alluvial gold seams and other mineral resources which are known to exist (manganese and rock phosphate) but the distance of the country, some 600 km from ports on the coast, makes it unlikely that manganese deposits would be exploited in the near future; rock phosphate is already being mined on a small scale and is used as fertilizer. 1.02 The dismal prospects for development have been, in part, responsible for a massive migration; as much as 0.9 million Voltaics, equivalent to 16% of the total population of 5.7 million and about 22% of the adult population between 15 and 59 years of age, are migrant workers, mostly in the Ivory Coast. There is also a sizeable movement from rural to urban areas. These migration patterns account for the high growth rate of the urban population (4% per annum) compared to the overall growth rate of 1.6% and the slow rate of increase in the rural population (1.4%). B. The Rural Sector 1.03 Agricultural production (including livestock, fishing and forestry) accounts for about 45% of GDP and nearly 65% of exports. Traditionally, livestock and meat products dominated exports but in recent years cotton lint and oil seeds exports have consistently earned nearly half of export revenues, reflecting better prices for these products and stepped-up promotion, espe- cially of cotton. By contrast the traditional coastal markets for Upper Volta's livestock and meat products with the notable exception of Nigeria have softened considerably under the impact either of local economic difficulties lowering effective demand (in Ghana), or increased local production activity (Ivory Coast) and access to cheaper imports from alternative sources. Produc- tion of coarse grains (millet and sorghum), which form the basic staples, has been generally sufficient (at about 1 million tons) in normal rainfall years to feed the Voltaic population. However population growth trends and the  - 2 - apparent stagnation of yields make it likely that the country could move into a position of structural deficit before the end of the present decade. The rapidly growing urban population would also increase the demand for rice which is at present imported for the most part (about 30-50,000 tons each year). Cotton is the dominant cash crop and the rapid increases in both yields and production achieved in recent years give further grounds for believing that productivity in cereal production can also be improved; in effect cotton is the only crop which benefits from the complete range of services--input supply, organized extension, assured marketing and research. 1.04 Irrigated production in Upper Volta lags behind its potential. It is estimated that only some 7,000 ha of the 120-150,000 ha of potential irrigable land is developed. The constraints which have prevented a faster rate of development include: high investment costs per ha (up to US$6,500/ha for large-scale schemes); inadequate price and marketing arrangements for irrigated crops, principally rice (typically less than 20% of paddy output Ls marketed); and lack of experience in planning, executing and maintaining major irrigation works (ONBI, the national agency responsible for this is only 5 years old). Against this must be weighed the growing demand, both locally and in nearby export markets, for rice, and a highly motivated agricultural labor force which has demonstrated its adaptability to the rigorous discipline required for irrigated crop production. Further development in the sub-sector will be based largely on the result of the pilot project at Niena-Dionkele and on a master plan for the Black Volta and Sourou basins which is nearing completion. 1.05 Provision of rural water supplies is inadequate in most regions in Upper Volta. As indicated by a recent report prepared by a Bank consultant, the basic requirement of 10 liters/capita/day (lcd) is met for only 15% of the population, with areas like the Koudougou ORD receiving even lower levels of service. A national water supply strategy, developed in 1976 and updated in 1979, states that, to keep the same standards (or higher for villages or centers of 500 inhabitants or more) some 6,700 water points remain to be drilled or deepened. At present some 900 water points are created each year in Government and NGO-supported schemes. The main issues in rural water development are (i) the danger of lowering the water table if a program of the dimensions proposed were to be realized; (ii) the high cost, fragility and multiplicity of pumps currently in use or proposed for Upper Volta; (iii) inadequate maintenance; (iv) the financial ability of Government to provide water at no charge to users; and (v) inadequate attention to parallel efforts in health education. The rural water supply component of the proposed project addresses these issues (para. 3.22). What is required, however, is a national policy for the sector as a whole which takes account of each of these factors. UNDP has already financed two studies which will, respectively, determine the optimum performance criteria for pumps suited to the Upper Volta environ- ment and examine appropriate levels of, and collection arrangements for, user  - 3 - charges. IDA staff have been closely following both studies and some prelim- inary results would be incorporated in the proposed project (para. 3.22). Further sector work, focussing especially on maintenance, is to be undertaken by IDA staff in the coming months. 1.06 Principal Institutions. The Ministry of Rural Development (MRD) has centrally based "staff" units with technical advisory functions and field- based "line" organizations responsible for execution. The central staff comprise a planning and administration unit and specialized technical units for agriculture (Direction des Services Agricoles-DSA), livestock (Direction d'Elevage-DE) and Rural Water Supply and Engineering (Direction de l'Hydrau- lique et l'Equipement Rural-HER). In the field, MRD is represented by 11 regional development organizations (ORDs) which have broad responsibilities in their areas of operation. They cover the whole country, with the exception of the valleys around the Volta and its tributaries, where river blindness is endemic and population consequently very low. In these areas an autonomous Volta Valleys Development Authority (AVV) is responsible for a resettlement program whose progress to date has been poor. ORDs depend upon the central services for specialist technical advice. 1.07 The four ORDs covering the central plateau (Center, East Central, North Central and North), where most of the activities under the proposed project would be carried out, are among the weakest in Upper Volta--apart from the funds available for investments under the second RDF project, their only other major source of financing has been Government budgetary grants which coier only the salaries of established civil servants, less than 40% of employees. The ratio of extension workers to farm families ranges from 1:800 to 1:1,100 compared to 1:330 in Bougouriba and 1:420 in the two ORDs of Bobo-Dioulasso and Dedougou where the IDA-assisted West Volta Cotton Project is currently underway. There is a general lack of logistical support for extension work, farm-level extension workers receive no training except in Yatenga (North) and the use of modern inputs and improved husbandry practices is still far below the potential. 1.08 Government has recently received the report of a commission which it set up to study the functioning and financing of ORDs. Its major recommenda- tions are: (i) limitation of the functions of ORDs to improving agricultural production and promoting farmer organizations to this end, (ii) giving greater autonomy to ORD directors especially in personnel management (at present all employees are either hired on contract in which case their salaries are uncertain from month to month, or civil servants whose transfer, conditions of service, salary and promotion are decided at the central level) and (iii) most importantly, instituting a tax on all agricultural exports and imports for the express purpose of providing ORDs with operating funds. The report also significantly recommends that externally financed projects should have a longer duration and should provide for the operating costs of ORDs which execute them. For the 1982 fiscal year, Government has taken measures which have the effect of limiting the annual deficits of ORDs to their 1981 levels; a long-term solution still needs to be found.  - 4 - 1.09 Central Services of MRD. The planning unit is essentially respon- sible for planning, programming and controlling ORD activities. In ORDs where there are no externally-funded projects--such as in Fada-Ngourma and Sahel--ORD activities are ineffective. Elsewhere, the control of the unit (and therefore of MRD) over ORDs is purely nominal, being replaced by such control and supervision as is exercised by the funding agency. Active discussions are underway to mount a technical assistance project with IDA support which iwould, inter alia, attempt to remedy the situation. 1.10 The two other services which would play a role in the execution of the proposed project are: (a) Rural Water and Engineering (HER) which has national responsibility for the design and construction of rural water points and small- scale rural works (bottomland development, small irrigation peri- meters etc...). In the past HER concentrated almost exclusively on well and borehole construction and on irrigation projects. In recent years, responsibility for large-scale irrigation works has been transferred to an autonomous agency, the National Dams and Irrigation Agency (Office National des Barrages et de lIrrigation, ONBI). HER has thus been forced to concentrate on small-scale rural works in which, especially under RDF 2, it has now built some expertise. Its role has been limited to providing design and periodic technical supervision services to ORDs for erosion-control and bottomland development works; this would be continued under the proposed project. In rural water supply, the division of its activities between the construction of new wells and boreholes and the maintenance of existing water points has been too much in favor of the former. Under the proposed project, emphasis would be placed on increasing its planning, programming and supervision capability and reorienting operations towards maintenance of existing water points (para. 3.22); (b) Agriculture Department (DSA). DSA's professed function is to provide the link between research activities and extension work by ORDs. In externally-financed projects, DSA is normally by-passed and the ORDs are provided with the means to conduct field trials and demonstrations, sometimes with the support of the research agencies. Elsewhere, DSA, because it lacks operating funds, has made little impact. The proposed project provides for the means to enable DSA to give technical support to the ORD demonstration program (para 3.12). 1.11 Forestry related activities are carried out by the Directorate of Forestry Management and Reforestation in the Ministry of Environment and Tourism. This directorate will receive material assistance to enable it to provide required technical support to the village reforestation sub- project under RDF 3 (para 3.20). 1.12 Agricultural Research. A number of research institutions are responsible to MRD or to the Ministry of Higher Education. For cotton,  the main agency is the Cotton and Textiles Research Institute (IRCT), and for cereals the Tropical Agriculture and Foodcrops Research Institute (IRAT). The Irrigation Crops Study and Research Center (CERCI) conducts work on irrigated crops, particularly rice. CERCI is financed by and receives tech- nical assistance from UNDP/FAO; the other two institutions are funded by France. Sorghum and millet breeding and agronomy are pursued by a regional team of the International Crops Research Institute for the Semi-Arid Tropics (ICRISAT) carrying out both on- and off-station trials and an associated farm systems study on the central plateau. The development of improved cereals (millet, sorghum, maize) and food legumes (cowpeas, groundnuts), and cultural practices compatible with small farms in semi-arid regions is the purpose of the Semi-Arid Food Grain Research Development (SAFGRAD) program financed by USAID under the auspices of the Scientific, Technical and Research Committee of the Organization for African Unity. The new programs are centered on the rehabilitated Kamboinse research station near Ouagadougou. Staff to support the program are provided by the International Institute of Tropical Agriculture, Nigeria; the International Crops Research Institute for the Semi-Arid Tropics, India; Purdue University, USA; and IRAT. Usable results, except for cotton and maize, are as yet meagre, and even then, not always known to ORD staff who are supposed to communicate them to farmers. 1.13 Agricultural Credit. Seasonal loans principally for cotton and fertilizer purchases are provided by 2 consortia of local banks, including a small (about 5%) participation by the recently formed Agricultural Develop- ment Bank (CNCA). These loans, which are rediscountable by the central bank (up to 65%) have been repaid reasonably promptly (about 90% repayment rate when due and most overdues are subsequently collected). CNCA, which is owned jointly by Government and public bodies (54%), and the regional Central Bank (BCEAO), its development banking subsidiary (BOAD) and the French bilateral aid agency (Caisse Centrale), has an authorized capital of CFAF 1300 million (US$4.8 million) of which half is paid in. For medium-term credit to farmers, CNCA planned to lend, in 1981, about CFAF 400 million (US$1.5 million) mostly for animal traction equipment. Most of the projected resources available for this purpose would come from central bank rediscount facilities and from earmarked funds included in externally financed projects. This latter source of funds would continue to be necessary to enable CNCA to expand lending for medium-term asset acquisition by farmers. CNCA is expected in its initial years to concentrate upon management and control of funds at a central level, establishing lending criteria and ceilings, and training ORD credit and supply staff who will continue to handle day-to-day credit operations (technical appraisal of loan requests, disbursements, and recovery). Until the advent of CNCA in 1980, ad hoc arrangements for medium-term credit were operated in the context of development projects, including the ongoing RDF II project. Repayment at about 80% nationally is expected to improve with the strengthen- ing of ORD credit and supply bureaus as in the proposed project (para. 3.11). Conventional medium-term credit would gradually become the exclusive respon- sibility of CNCA (para 3.14). 1.14 Marketing. The record of official cereals marketing policy in Upper Volta is reminiscent of similar efforts elsewhere in West Africa: because of the small size of the market in relation to total production,  - 6 - small variations in output cause large variations in marketed quantities and prices and the official marketing agency (OFNACER) does not have the flexibility and financial resources to enable it to react efficiently to such changes. While OFNACER has done well in distributing food aid in drought- induced emergencies, its record in regulating price fluctuations between harvests has been much less impressive. It buys only a small proportion of marketed output (about 20%); its costs (especially transport costs) are high and it operates at a loss; storage losses are high; it is inadequately financed; and ORDs acting as buying agents for OFNACER lack transport and storage facilities. Without the incentive of reliable markets and prices high enough to justify cash outlays for inputs, the modernization of cereal produc- tion will continue to be slow. One possible solution to this problem is to promote a more active role for farmers groups in crop storage, processing and marketing. Based on pilot actions begun under the Second project, the proposed third project provides for the construction and operation of village cereal banks and small cereals dehullers in furtherance of this objective. Cotton marketing is satisfactorily handled by SOFITEX, a joint state/public enterprise. C. Sector Strategy, Development Constraints, and Major Issues 1.15 Governments Objectives. Successive Governments in Upper Volta have stressed national self-sufficiency in food production as the topmost priority for the agriculture sector. In examining the ways in which it has intervened in the sector, one can infer that its strategy consists in: (a) maintaining, and if possible improving the productivity of rainfed crop production through extension advice and inputs. The ORD structure is the best evidence of this: it is not selective and it is predicated on the belief that marginal improvements in traditional cultivation practices are possible and can be acquired by farmers through the action of extension agents; (b) promoting migration to the under populated southwest and especially in the Volta Valleys where rainfall is adequate and reliable, and soils more fertile. The AVV project, now in its eleventh year, shows the persistence of Government efforts in this area in spite of the poor results; and (c) swampland and irrigated production schemes to counter the effects of insufficient and unreliable rainfall. 1.16 Bank-Group Support to the Sector. To date there have been ten IDA-supported projects in the rural sector of which four have been completed. Two of these, the Cotton project (Cr. 225-UV signed in 1970) and the Bougou- riba project (Cr. 496-WV approved in 1974) supported the ORD-managed regional development approach. Although the cotton project had a clear commodity focus, its major result was to demonstrate that where an improved technology existed which was within the means of small farmers, it could be spread widely in a fairly brief period of time with the ORD organization. The high recurrent cost effect of this approach did not create any problems during the execution of the cotton project because it increased the production of a commodity  - 7 - which had a high market value and whose marketing was and remains easily monopolized by a public agency. A recalculation of the economic rate of return upon project completion showed a slightly higher return (37%) than the appraisal estimate (32%). By contrast, the first Bougouriba project had only a small cotton component and cotton prices had meanwhile not risen as much as during the cotton project; the rate of return on completion (15%) indicated that appraisal estimates (68%) were much too optimistic. The major lesson was to demonstrate the difficulty that Government has in paying for agricultural development programs organized and operated in this manner. The other two completed projects were the drought relief project (Cr. 442-UV of 1973) and the first Rural Development Fund project (Cr. 317-UV of 1972) both of which aimed at financing small discrete and mostly physical assets using the same institution (RDF) to channel the funds and control expenditures for activities which were physically carried out by existing Government services. Overall, results were satisfactory (paras. 2.02-2.06). 1.17 The remaining IDA-assisted operations in the rural sector include follow-up projects to the cotton (West Volta, Cr.706 of 1977) and Bougouriba (Credit 1097 of 1980) projects, both of which explicitly promote regional development on a broad front rather than narrowly focussing on specific commodities. There are also projects in irrigation (Niena Dionkele, Cr.1013-UV of 1980); Livestock (Cr.557-UV of 1975); Forestry (Cr.982- V of 1980) and a Second Rural Development Fund project (Cr.640-UV of 1976). Project execution is progressing satisfactorily in all these projects. 1_. d Major Constraints and Elements of a Strategy. The summary evalua- tion of completed IDA-supported projects discussed above leads to the identi- fication of the major issues which have to be taken into account in a develop- ment strategy for the sector: (a) the physical environment (climate, hydrology and soils) is not favorable to crop and livestock production. The indirect way of tackling this problem is to develop a technology which makes the most of these given facts of nature. Research results which are adapted to actual farm conditions are a small part of the work currently being done. This leads to the second major constraint, viz; (b) lack of suitable technologies. In part this is a matter of organi- zation since some improved husbandry practices which have been developed by research can be adapted for use by small farmers at reasonable cost. For example, in parts of the central plateau, better land preparation methods using available (and affordable) animal traction equipment coupled with erosion-control measures and organic fertilization have been shown to give yield increases of 20-30% compared to traditional methods using hand-hoes. With better organization of the design, construction and maintenance of erosion-control bunds, of extension advice on suitable husbandry practices and of credit as under the second RDF project, farmers were shown to be willing to take up this technology; and  - 8 - (c) institutional weakness. This has several facets: organizational, as in the above demonstration; financial as demonstrated by the persis- tent lack of operating funds for the ORDs which are the major executants of Government policy; and human, given the paucity of skilled technicians in general and specifically those willing to work in the rural area. 1.19 Government's principal objective of meeting national food demand = from local production can be further promoted by the strategy it has pursued thus far with some important qualifications and safeguards. First, the heavy cost of maintaining the present ORD structure can only be justified where it can be demonstrated that practicable technological packages exist which can be passed on to, and be economically usable by, small farmers. By this reasoning, the present structure of ORDs is premature for some ecological zones (such as the central plateau). Even for areas where suitable packages exist, expenditures on the ORD should be carefully controlled to bring them in line with the value of the output of the farmers benefitting from the extension package. Secondly, a major effort is required both in research and in the organization of the transfer of research results into usable packages at the farm level; only thus can the ORD structure find its full justification. Side-by-side with this, there is room for promoting productivity-increasing investments which through simplicity of design and reduced scale, can be brought within the management and maintenance capacity of rural communities. Resettlement should be encour- aged in the context of standard regional development projects planned for underpopulated zones rather than being organized separately and at high costs as is presently done by AVV (para 1.06). Given the developmental character of the "operating costs" necessary for key functions such as research and exten- sion, external financing should make provision for such expenditures. II. THE PROJECT AREA A. Ecology and Population: 2.01 Agricultural production activities under the project would be concentrated on the central plateau covering the ORDs of Ouagadougou (Center), Kaya (North Central), Ouahigouya (North), and Koupela (East Central). In addition a pilot action would be mounted in Sahel in the sector of Dori. The water supply component would be limited to Koudougou (West Central) which would also benefit from RDF financed subprojects in those parts of the ORD not included in the Koudougou Pilot Project currently being prepared. Climate is primarily Sudanian with annual average rainfall declining from around 1,000 mm in the south, to less than 700 mm north of Kaya and Ouahigouya and to below 600 mm in Dori. There is a four- to five-month wet season (May-October) but the beginning of the rains is erratic; they are generally interrupted between mid-June and early July, and they may cease as early as mid-September. Temperatures range from a maximum of 34/400 C to a minimum of 16/260 C in the dry season. The dry season divides into a cool period of three months between December and February followed by a hotter period prior to the rains. Most of the arable land consists of shallow, poorly structured tropical ferruginous soils often acidic and deficient in nitrogen, phosphate and  - 9 - organic matter. Generally a distinction can be drawn between the thin, over- exploited soils around villages towards the top of the watershed and deeper less gravelly soils further down the toposequence. Bottomland soils are more fertile but sometimes difficult to work by hand. Total rural population is estimated at about 2.4 million, about 300,000 farm families. Density averages some 35 in/km2 and is over 60 in/km2 in the parts of the Ouahigouya (North) ORD. B. Earlier Projects. 2.02 Credit 317-UV in the amount of US$2.2 million for a rural develop- ment fund project was approved in June 1972. The final disbursement was made on June 23, 1977. The objective of RDF was to finance a variety of small productive investments which, because of their size and scattered location, were otherwise not suitable for conventional Bank Group lending. The project, which was conceived as the first experimental phase of a longer program was to consist of: (i) 500 wells to improve village water supply; (ii) 700 small village storehouses for farm inputs; (iii) the development of 1,500 ha of bottomlands for rice cultivation; (iv) erosion control works on 1,000 ha to improve crop yields (sorghum, cotton, groundnuts); (v) 200 ha of small irri- gation schemes below existing dams, to produce rice; (vi) the upgrading of 270 km of feeder roads; and (vii) studies and a number of unidentified sub- projects. An essential point in sub-project selection was the interest of the people concerned, expressed through their willingness to provide free labor for sub-project implementation and subsequent operations and maintenance. 2.03 Project execution was entrusted to RDF, created as an autonomous unit within the National Development Bank (BND) for this purpose. Since 1979, RDF has been operating independently of BND, with its management reporting directly to the Secretary-General of MRD (para. 4.04). RDF appraises projects identified in the first instance by potential beneficiaries themselves and largely prepared by technical and field services such as HER and ORDs. After approval of RDF's annual program by an interministerial committee chaired by senior MRD staff, the project is executed by the competent agency (HER or ORD for example) which signs a contract with RDF for this purpose. Financial accounting and control remain the responsibility of RDF which also follows up on completed projects. 2.04 Actual implementation was as follows: Component Appraisal Actual Wells (no.) 500 262 52 Village storage (no.) 700 644 92 Bottomlands (ha) 1,500 1,719 115 Erosion control (ha) 1,000 1,395 140 Small irrigation (ha) 200 96 38 Feeder Roads (km) 270 366 136 Instead of three years, actual project implementation took over four years. Actual costs were 7.5% higher than appraisal estimates.  - 10 - 2.05 Concentrating on the central plateau, Credit 640-UV for the second rural development fund project financed the same type of sub-projects as under the first project with the exception of feeder roads. In addition, it provided for training and technical assistance to overcome staff con- straints and financed studies of development potential in adjacent areas. Other components included agricultural credit and the creation of a well maintenance brigade. The project was to be executed over a five year period, but has been extended by one year while activities in the first, and partly the second, year were still financed under RDF 1. Additional financing provided by Dutch Aid (DF1 5.15 milion equivalent to about US$2.0 million) will ensure that no substantial reduction of the work program will be neces- sary in the last project year. Actual implementation to date compared to appraisal objectives is: Component Appraisal Actual _ Bottomlands (ha) 2,700 1,885 70 Improved bottomlands (ha) 300 145 48 Small irrigation (ha) 200 170 85 Erosion control (ha) 9,200 15,051 164 Open wells (no.) 520 569 109 Tube wells (no.) 180 180 100 2.06 The major lessons learned from RDF 1 were: (i) the need to strengthen staff in the technical and executing agencies (HER and ORDs); (ii) the problem of ensuring maintenance of assets created with RDF financing, especially feeder roads; and (iii) the absolute necessity for ensuring the availability of credit to enable farmers to acquire complementary inputs and equipment in the absence of which the full benefits of the land development schemes could not be realized. Overall, however, the project was described in the PPAR issued in September, 1979 as a "highly innovative project which was successful in implementing diversified and scattered subprojects". Benefi- ciaries were estimated at some 230,000 people and a recalculation of the economic rate of return on agricultural subprojects showed a result much higher (22%) than the appraisal estimate of 16%. The follow-up project incorporated the lessons of RDF 1 since it continued with the subprojects proved successful under its predecessor, included elements which had been revealed to be essential (credit and strengthening of staff of technical and executing agencies) and excluded feeder roads for which a more competent national body had meanwhile been created, thus ensuring better maintenance. Results as shown above have been very satisfactory, especially for erosion control and water supply. Credit, primarily medium-term loans for animal traction and agricultural equipment, has been a popular comlementary program. Repayment rates at farm level are not known precisely (para. 3.28) and in any case it is still too early to judge; however, there is ample evidence that results would be at around the national average for such credit (about 80%). Poorer results in bottomland development have been the result of: (a) land tenure problems, (b) the heaviness of bottomland soils which make land prepara- tion difficult, (c) poor design and quality of works, and (d) an undeveloped marketing system for paddy. The proposed third project includes bottomland development but on acreage less than 50% of the actual realizations of RDF 2  - 11 - (para 3.06), whose design would be improved as a result of their review by an independent panel of experts (para 3.07), and on which cultivation rights for participants would be given legal underpinning (para 3.07). The major successes of the earlier projects however have been in creating an efficient organizational tool for appraising small productive rural investments, providing sound financial control during their execution, and building up a capability for the monitoring of such projects. The proposed project is based on a full exploitation of these capabilities. III. THE PROJECT A. Project Origin and Preparation 3.01 The Second Rural Development Fund Project is in effect the third in a series of projects financed with IDA assistance and was substantially completed by September 1981; Dutch funds are available to continue activities at substantially the same rate as in recent years until the expected date of effectiveness of the proposed third project. A third RDF project was first mentioned in 1977 but it was not until August 1979 that Government and the Bank reached agreement in principle to begin project preparation. The work itself has been undertaken largely by RDF staff under the supervision of RMWA. Successive drafts of the preparation report were reviewed by RMWA staff during the fall of 1980 and the full report, incorporating most of the suggestions mace then, was completed before the end of 1980. An appraisal mission visited the country in March-April 1981. B. Objectives and Summary Description 3.02 Objectives. The project would aim at completing the process, pursued under the preceding two projects, of developing a reliable and cost- effective capability for managing small-scale rural investments. As demon- strated in the earlier projects, there is both a demand and a need for such investments, which have attracted high rates of participation especially in the heavily populated and poorer-than-average central plateau. The specific subproject types identified in the course of project preparation and appraisal have been based on experience acquired under the previous projects; they would increase crop and fuelwood production, improve the availability of safe drinking water for rural communities, and increase the value of agricultural produce through processing and better storage. 3.03 Summary Description. The project would be implemented in the four- year period 1982/83-1985/86 and would comprise the following subprojects: (a) land development comprising engineering design services, equipment and supervision of construction of (i) anti-erosion bunds to protect about 20,000 ha of dry land farms, (ii) simple bunding on 640 ha bottomlands to provide partial control of flooding, and construction of earth dams, intake structures, spillways and flood protection dykes to provide for complete control of flooding and supplementary  - 12 - irrigation on 240 ha of existing bottomland developments, and (iii) small-scale irrigation works to provide full water control irrigation on 160 ha downstream of existing reservoirs or from ponds and shallow wells; (b) support to ORDs consisting of training and initial funds for the departments responsible for training of extension staff and farmers; credit administration and input supply; and research and demonstra- tion; (c) credit to farmers for the purchase of ox-drawn implements and oxen to be administered by CNCA; to farmers' groups to finance small rice dehullers, small grain mills and initial stocks of cereals and general merchandise for group-owned and operated village stores, to be administered by RDF; and to the Rural Equipment Support Fund (Fonds d-Assistance a lEquipement Rural-FAER) to finance purchase of inputs required for the manufacturing of animal drawn implements; (d) village storage sheds which would be used either to store inputs and implements in which case they would be operated by ORD staff, or as community-owned and operated stores for traded cereals (cereals bank) or general merchandise; (e) tree plantations consisting of investments in 6 nurseries and about 400 ha of trees to provide wood, fruits, fodder and the equivalent of a further 160 ha in trees to fix bunds on soil erosion sites; (f) village water supplies consisting of 60 new wells; deepening of 200 existing wells; 150 boreholes; provision of equipment and operating funds for wells/boreholes maintenance brigades; and a stock of spare parts for pumps which would be sold for cash to villagers; and (g) project management overheads consisting of a headquarters building for RDF and other costs associated with managing the investment program outlined above. The project also provides funds for the financing of other small-scale rural projects which would be identified in the course of project execution. Such projects would be selected using substantially the sane criteria of tecbnical and economic feasibility as for identified projects (para. 4.02). C. Detailed Features 3.04 Erosion Control Works. High population densities leading to progressively shorter fallow periods have made large areas of the central plateau practically unusable for cultivation, with topsoils completely washed off by successive rains. The construction of simple earth bunds, by allowing more time for infiltration and by braking the force of the flow of rainwater, has been shown to have a marked effect on crop production; evaluation surveys carried out in 1980 on bunds constructed in 1978 under RDF II indicate yield increases of 20%-65% above yields on similar plots with no erosion bunds.  - 13 - 3.05 About 20,000 ha of farmland would be protected against soil erosion under the proposed project. Except for a small pilot project to check the effects of wind erosion in the northern Sahel ORD (about 1,600 ha), soil erosion works would be concentrated on the three most populous ORDs of Kaya, Ouagadougou and Ouahigouya (about 5,800 ha each), with a much reduced program (1,000 ha) in Koupela ORD where soil erosion is not a major preoccupation. Typically each site covers about 25 ha and is farmed by some 15 farm families (1.5 ha each) who generally have an equivalent area of land off-site. Con- struction methods are simple: after ORD survey teams stake an area, a tractor- drawn disc plow traces the line. Bund construction is then completed by beneficiary farmers under the supervision of ORD staff. The project provides for the necessary design, extension and supervision services; equipment; operating costs of ORD land survey teams; and spare parts and small tools. The labor contributed by beneficiaries on such sites during initial construc- tion is substantial: about 20 days/ha (or 500 working days per site) and a further 3-5 working days/ha per year for maintenance. Increased extension activity would be required after land development works are completed in order to train farmers in the use of the full range of improved cultivation practices necessary to enable maximum benefits to be derived from such works. Thus, for erosion control and simple bottomlands average area per extension worker increases from 50 ha in the year of construction to 200 in the sixth year after construction; comparable areas for improved bottomlands and irrigated lands are 5 ha rising to 20 ha in the sixth year. The costs of such additional extension work are included in the project. It is understood that this would not require recruitment of new staff; rather existing ORD staff would be redeployed from lower priority functions. This was confirmed by Government during negotiations. No problems of land allocation on developed sites have been encountered so far and are unlikely to occur under the proposed project since ownership of such land is well established before the construction of erosion control works. A major change in procedure would be initiated under the proposed project: whole watersheds would be studied instead of small scattered sites at present. If conditions are judged suitable (willingness of rural population to participate, appropriate physical and hydrological conditions) an erosion control and bottomland development program covering several seasons would be planned for the locality. Other refinements in design initiated under the current project which would be continued under the proposed project include planting perennial grasses or bushes on bunds after construction and suitable vegetation cover to protect waterways against gully erosion. About 13,500 farm families would benefit from such works. Further details are in Working Paper 1 and project costs are detailed in Annex 2 of this report. 3.06 Bottomland development works. Under the first and second projects some 3,000 ha of relatively fertile bottomland soils have been equipped with simple earth bunds to provide some control of flooding and improved distribu- tion of water. In bottomlands where flooding is more severe, water control is further increased by the construction of a larger earth dike (1.5 m high) above the system of bunds with a drainage canal and an intake structure so that the introduction of extra water into cropped-areas does not depend solely on the physical configuration but can be increased at will by local farmers (improved bottomland developments). The command areas of such improved works  - 14 - constructed under the second project amounted to about 120 ha. Construction methods are identical to those used in soil erosion works. Although problems have been encountered in the operation of bottomland developments (para 2.06), sufficient farmer interest and suitable areas exist to justify a modest subproject under the proposed project in carefully selected areas of the Kaya, Koupela and Ouagadougou ORDs. 3.07 The proposed project would provide for engineering design services, skilled craftsmen's services including an initial ploughing by tractor, equipment, building supplies, and construction supervision to develop 640 ha using simple bunds and a further 240 ha of improved bottomland areas; improved designs would need the approval of a panel of experts to be set up under the project (para 3.18) and HER would directly supervise the construction of improved bottomlands. A typical bottomland site would be between 10 and 20 ha with individual parcels of 0.2-0.6 ha; in all some 4,400 families would benefit from such works. A major improvement in design apart from the incorporation of such sites in watershed planning schemes (para 3.05) would consist in adding a perpendicular "wing" to the end of each bund, which slopes down and away from the main structure, so that excess water flows at a slower pace from one bund to the next, lower one. It has been demornstrated that this would greatly reduce breaches. In addition, the pre-appraisal socio-economic survey (para 4.03(c) would be used to examine the land-tenure situation rigorously and eliminate those sites where cultivation rights for farmers contributing iat;.- t ck.riEtruction are not assured. In particular, the preappraisal would exa:2,ne and evaluate to the cohesiveness of informal groups organized for the purpose of cultivating improved lands by evaluating the source of authority that the leadership of such groups lay claim to (para. 4.03). There is a clause in existing laws governing land ownership which vests lands improved with public funds in the Government. During negotiations Government gave a formal undertaking to enforce such laws more rigourously and, in particular, that it would formally transfer ownership of such developed land to farmers particiDating in construction. 3.08 Small Irrigation Perimeters. Under the two preceding projects some 260 ha of small irrigated perimeters (gravity and pumped) were developed and contrary to experience with similar projects financed by voluntary aid agen- cies, for example, the percentage of the area which was still under cultivation 3 years after construction and the cropping intensity are both very high, 75% and 130%, respectively. An initial survey conducted jointly by HER and ORD staff in the Kaya and Ouagadougou ORDs has identified several very promising sites out of which it is proposed to developed 160 ha spread over about 10 sites and cultivated by individual farmers each with a plot of about 0.2 ha. The project would provide for the necessary technical advice, skilled craftsmen and equipment, including pumps for irrigation sites requiring them. Spare parts for such pumps and construction material and tools and equipment required for maintenance would be sold for cash (para 4.03). Construction methods and responsibilities would be slightly different from those which were used in the previous projects. Design, survey and construction supervision would be undertaken by HER, not, as previously, by ONBI. Further, a panel consisting  - 15 - of 2 engineers and one economist chosen from qualified professionals resident in Upper Volta--the University of Ouagadougou, the International Center for Hydraulic Studies (CIEH) and the Pan-African Institute of Development all have suitable staff for this purpose--would approve all designs for irrigation works and improved bottomlands proposed by HER before they are included in RDF-s annual work program. Agreement was reached with Government during negotiations that it would nominate people with qualifications and experience satisfactory to IDA to serve on this panel. Government also gave undertakings with regard to cultivation rights in irrigated perimeters similar to those described in para 3.07. 3.09 Support to ORD services. Full benefits from land development schemes (above) are obtained only to the degree that such investments are accompanied by improved cultivation practices at the farm level. To ensure that ORDs can effectively provide advice on such improved practices, funds would be made available to strengthen the operations of the four central plateau ORDs in three key areas: (i) training; (ii) credit administration and input supply; and (iii) trials and demonstration. 3.10 Training. The European Development Fund (EDF) is currently financing a training program--buildings, equipment, staff salaries and operating costs-- in the Ouahigouya ORD. A similar program with funding from EDF is expected to commence soon in the Koupela ORD. In the Ouagadougou and Kaya ORDS, however, although training units exist in organization charts, there is no training activity aside from those which have been financed and organized by the present RDF project in connection with its investment programs in these two ORDs. The proposed project would finance vehicles, equipment and initial operating costs to enable the training units in the 2 ORDs to function. The priority action of such units would be the organization of in-service training for existing extension staff. The project provides for one-day training sessions each month for the 200-odd extension agents and about 20 sector chiefs who would be involved in training farmers on land development schemes to be financed under the project (para 3.05). Other training is provided under the project for: tractor drivers (22 drivers in 10-day sessions during the first two years of project execution); farmers through visits to demon- stration sites (in groups of 30 and on 10 occasions a year for each of the 4 ORDs); agricultural technicians at the intermediate level (3 students would be financed for the 4-year cycle intermediate technicians course at Matourkou). A senior training specialist is included in the technical assistance team which is to be provided by UNDP to RDF for the duration of the proposed project (para 4.08). Further details of the training program are in Working Paper 3; costs are in Annex 2. 3.11 Credit and Supply Offices. ORD staff are still responsible for day-to-day credit operations (para 1.13). In the four ORDs of the central plateau where project actions are concentrated, the credit and supply bureaus are undermanned and lack operating funds; this has not been a major obstacle in the past when credit operations were modest and RDF staff were available to lend a hand (para 1.13). The proposed project provides for medium-term credit for animal traction equipment to be administered by CNCA using these bureaus; they would consequenty need to be strengthened. In the third project  - 16 - year, the volume of such credit sales is high enough that an 8% mark-up would be adequate to cover their estimated operating costs (Working Paper 4). The project provides funds for the financing of operating costs during the irlitial two years. Agreement was reached with Government during negotiations that it would require the credit and input supply offices of the ORDs in the Project Area to charge a mark-up of not less than 7% on credit sales of animal traction equipment handled by such offices and that the level of this mark-up would be reviewed and adjusted as necessary in order to ensure that revenues thus procured would be sufficient to cover the operating costs of such offices by the third year of project operations. Further details are in Working Paper 4. 3.12 Applied Research and Demonstration. The links between research and extension are weak even though there is a central service (DSA) in MDR which is responsible, among other things, for ensuring that information on more productive technologies reach farmers through ORDs. Under the current project the foundations of an agricultural demonstration system have been laid in the form of the "zones pilotes". These are village communities in which farmLers are encouraged to try new techniques with incentives being offered in the form of credit for ox-cultivation equipment, project-financed village storage sheds and water points. The results have been encouraging and farmer groups, aLfter visiting such villages, have been convinced enough to start erosion control schemes and construct commonly owned storage facilities. The proposed project would help install a systematic trial and demonstration program based on the "zones pilotes". Specifically the project would finance the equipment and operating costs of 16 trial and demonstration points (15 on the central plateau and 1 in Dori) as well as providing the means for extension staff to run such demonstration points. Responsibility for trial design and supervision would remain with RDF with technical support from ICRISAT (para 1.12) but DSA would nominate an agricultural officer to collaborate with RDF who would provide the officer thus nominated field allowances and the transport necessary to participate in field supervisions. ORDs would nominate a field trials officer to take responsibility for day-to-day operations and the project would provide each with a vehicle and field allowances. Further details are in Working Paper 3. 3.13 Credit. Animal traction has been shown to improve soil permeability and reduce labor bottlenecks. Demand for such implements is high where credit is available. In the four ORDs of the central plateau, however, it is reliably estimated that only about 30% of the area with soil erosion schemes was ploughed with animal traction equipment in 1979/80. As the sample represents more progressive farmers, the overall use of animal traction equipment is probably no more than 10%. According to appraisal mission calculations up to 50% of farmers participating in erosion-control works can use and can pay for a set of such implements (Working Paper 2). The project would provide funds to be administered by CNCA and made available for the purchase on credit of animal traction implements and oxen. Funds would be available each year for up to 1,800 sets of ox equipment and half the cost of a pair of oxen for each set and up to 400 sets of equipment suitable for donkeys. These figures are the equivalent of less than 25% of the total number of farmers expected to participate in soil erosion control and bottomland development projects under the proposed project.  - 17 - 3.14 A full set of ox-drawn implements costs about CFAF 53,150 (US$197) and a set of equipment suitable for donkeys costs about CFAF 35,500 (US$130). Credit terms would be those currently applied by CNCA. Interest would be at the rate of 11% per annum on outstanding principal, the preferential rate accorded to loans to the rural sector under existing Government policy. Although until recently this was about equal to the annual inflation rate in Upper Volta (the average for the 5-year period ending in 1979 was just about 11%), reliable projections of local inflation through 1985 are somewhat higher (12% falling to about 9% by 1985). Consequently, Government agreed during negotiations that the whole question of interest rates for rural sector loans with a view to adjusting them to take account of inflation as well as the operating costs of CNCA, required further study. Repayment is in four equal instalments, after one year's grace on principal. Thus, after paying interest and the mark-up of 7% (para 3.11) in the first year, a farmer who has borrowed money for ox-drawn implements would have a debt burden of about CFAF 17,200 (US$64) annually; farmers acquiring donkey equipment would need to pay CFAF 11,430 (US$42). About 25% of project area farmers with landholdings up to 7.0 ha would be the potential buyers of such equipment. Their gross revenues on the 4.0 ha of dry land protected by erosion control works would be about CFAF 220,000 (US$815) which would be sufficient to enable them to service such debt. Credit administration would be under the day-to-day control of ORDs who will be supervised by CNCA. Repayments of outstanding loans made under the ongoing project for animal traction equipment would continue to be made to RDF which would account to CNCA each year and pay to it sums received in this connection. Such amounts would then be available for a second round of loans under the proposed project. A protocol has been signed by both CNCA and RDF which sets out the terms of their cooperation including the progressive transfer from RDF to CNCA of credit funds repaid under the ongoing project. Further details of credit arrangements for animal traction are in Working Paper 4. 3.15 RDF would continue to manage credit made to informal farmers' groups for the acquisition of small rice dehullers, small grain mills and for initial stocks of cereals and general merchandise for cereal banks and village merchandise stores. These loans represent only a part (about 45% of the total cost of buildings plus equipment or merchandise) of a larger program which includes construction of village stores (para 3.19) and which requires more technical and managerial supervision than CNCA, with its limited mission (para 1.13) can adequately handle. Rice dehullers, grain mills and cereal banks would also be closely linked with the promotion of model villages in the demonstration program (para. 3.12); the supervision of the complete program by a single agency (RDF) is indispensable at this stage. Interest and repayment terms are identical with those applicable to loans for animal traction equip- ment. In addition a down payment of 20% would be required for these credits. RDF supervision would include both distribution of such items sold on credit and the collection of repayment and would be undertaken in the course of its supervision of the "zones pilotes" program. 3.16 The proposed project also makes provision for a line of credit to the Rural Equipment Support Fund (Fonds d'Assistance a l'Equipenent Rural-FAER) which is responsible for operating a network of animal traction implement manufacturing and assembly workshops in Upper Volta. The fund was set up to  - 18 - promote animal traction use. Grants from bilateral aid sources (Federal Republic of Germany, U.K., the Netherlands, USAID and Switzerland) have been used to establish a manufacturing plant to produe animal traction equipment in kit form (2 plants) and an assembly unit in each of nine ORDs, and to train users of such equipment both at ORD level and at an experimental farm. Under the RDF 2 project currently under execution some 75-100 sets of such equipment have been sold each year on credit to farmers essentially in model villages but demand has consistently outstripped supply, an experience shared by two IDA-supported agricultural projects in the West Volta and Bougouriba areas. Since its inception, FAER's operations have not been subjected to strict financial accounting; the nature of its sources of financing and the loose control exercised by an overburdened Secretary-General in MDR are largely responsible for this. The result is that both technical and financial manage- ment have proved deficient and production and distribution have not kept pace with demand. Production costs are only rough estimates and a reconstructed balance sheet as at the end of September 1980 showed accumulated losses of CFAF 110 million (US$407,000) and accounts receivable (mostly from ORDs and therefore of doubtful quality) of CFAF 185 million (US$685,000). These have been largely financed by accumulating arrears with suppliers and other creditors--CFAF 170 million (US$630,000) and bank overdrafts of CFAF 120 million (US$445,000). Its goods-in-process inventory meanwhile stood at CFAF 485 million (US$1.85 million), about equal in value to total sales in the preceding 36 months. Details on FAER operations are in Working Paper 5. 3.17 During negotiations agreement was reached with Government on the measures which need to be taken to place the production of animal traction equipment on a viable basis. These include: (a) a financial restructuring of the enterprise; (b) the transfer of overall management of FAER to a suitable agency with statutes and operating regulations acceptable to IDA; and (c) the appointment of management staff acceptable to IDA and with demonstrated skills in inventory control, production planning and cost accounting. UNDP, which has provided technical assistance to FAER under a program financed by Denmark, is willing to finance such technical assistance personnel. 3.18 With effective action along the lines discussed above (para 3.17), FAER can be restructured both financially and organizationally in order to ensure the availability of animal traction equipment, the use of which is a critical first step in the modernization of Voltaic agricultural production. If such measures were taken, FAER would need financial assistance of at least CFAF 230 million (US$0.9 million) in the two-year period 1982-1983 to accom- modate the expected increase in its sales volume with only a slight improve- ment in shortening the average collection period for sales. The project includes provision for such financing providing that the conditions stated above can be met by Government before credit disbursements. Depending on the results of the financial restructuring exercise, Government, may use the IDA/Dutch funds either to increase its equity holding in FAER or have such  - 19 - funds loaned to FAER on its behalf and for its account by CNCA. CNCA would be required to examine any application for loan financing from FAER using strict commercial criteria. Eventual repayment to CNCA would be earmarked by Govern- ment for expanding credit available through CNCA for animal traction equipment. The timing and scope of these measur^s and the conditions governing the management of this line of zredit by CNCA on behalf of Government were a. 1 agreed during negotiations. 3.19 Village Stores: Under the present project, RDF has financed a model village program (zones pilotes). Villages have been selected on the basis of agricultural potential, an existing and functioning farmers' group and the possibility of establishing a variety of actions in the same community including demonstration plots (para 3.12), stables for manure production, and village cereal or merchandise stores. In each case, the principle has been to support the farmers' group which undertakes to provide free labor for the construction of such structures and to borrow money to purchase the initial stock of cereals or general merchandise as the case may be. RDF finances (in the form of a grant) the cost of building materials (roofing sheets, cement and building poles) as well as the necessary skilled labor and also the credit funds required for purchasing the initial stock of merchandise to be traded. The program has proved to be very popular. It has increased the availability of basic goods and staple foods to rural communities who have demonstrated a willingness to pay the necessary price and thus provide the means to repay borrowed funds. Its incentive effect on the zones pilotes program has been an important factor of success. Under the proposed project the HER would continue to provide building materials and skilled labor with grant funds channelled through RDF for the construction of about 75 storage units in about 40 selected villages in the project area. In addition some 30 input and equipment storage sheds would be constructed under the project. Their operation would be the responsibility of the ORD sector chief. As in the case of community-owned merchandise stores, free construction labor would be provided by local farmers. 3.20 Village Reforestation: Swiss and Dutch Aid have financed a village reforestation program since 1978 in 5 ORDs (Volta Noire, Kaya, Ouagadougou, Ouahigouya and Fada) with the dual objectives of increasing wood production both for fuel and construction and as a soil conservation measure. Under this program, 20-30 villages are selected each year on the basis of priority according to the degree of reduction of forest cover. In these villages a one-hectare community woodlot (on a plot chosen in agreement with an informal group of villagers) and an equivalent area in family compounds are planted with suitable tree species. The program is administered by the Forest Manage- ment and Replanting Department (Direction d'Amenagement Forestier et du Reboisement - DAFR) of the Ministry of Tourism and the Environment. DAFR manages nurseries, distributes planting material through ORD staff, and supervises plantation establishment and maintenance. Training and monitoring services are provided under contract by the West African Center for Economic and Social Studies based in Bobo-Dioulasso. Under the present RDF project, funds have been provided for the establishment and maintenance of seven nurseries in areas (2 sectors in each of Kaya and Dedougou ORDs and 3 in Kaya ORD) not covered by the Swiss or Dutch-financed programs but which are on the  - 20 - list of priority zones requiring reforestation. Under the proposed project, RDF would provide grant funds for the maintenance of six out of the nurseries established (the seventh is to be taken over by the Forestry Department); seedlings and technical advice for planting up to 400 ha, divided about equally between block plantings (both individual and communal) and on family compounds; and seedlings and technical advice for tree planting on erosion control sites. The project would furnish logistical support to DAFR to enable it to give the necessary extension, nursery maintenance and technical advisory services required for the execution of the program. Responsibility for the maintenance on block plantings would lie with the head of the informal group. Some 80 villages within a reasonable radius of the six nursery sites would be affected. Seedlings would be sold at the same prices and on the same terms as in the IDA-assisted Forestry project (para 6.04). Technical aspects of the village reforestation program are discussed in Working Paper 6 and cost details are in Annex 2. 3.21 Village Water Supply: The availability of clean and permanent water supply is cited as the first priority by a large majority of rural inhabitants in almost every survey conducted in the Upper Volta. The number of artisanal wells constructed by villagers themselves outnumber, by an estimated 3:1, wells and boreholes created by Government or NGO-sponsored programs. As part of a larger national rural water supply program financed in different locations by bilateral and multilateral aid sources (para 1.05), the proposed project would finance the construction of 60 new wells, the deepening of about 200 wells, the construction and equipment of about 150 boreholes all in the West Central ORD, and two rural water supply maintenance brigades, one based in Koudougou (West Central ORD) and the other in Kaya (North Central ORD) where wells and boreholes were financed under the preceding projects. 3.22 In order to improve further the performance of the water supply unit within HER, several important changes would be undertaken. First, all bore- holes would be contracted out to private operators of which several are active in Upper Volta. HER would concentrate on planning, programming and providing technical control in the execution of such contracts, and on maintenance, including training its own staff and village-designated volunteer mechanics and supplying needed spare parts at cost to village communities. Provision has been made for three-months training courses for three mechanics in neigh- bouring countries with adequate facilities and for the training of a total of 200 village mechanics nominated by traditional chiefs and paid by voluntary contributions by villagers and 200 women volunteers for health education. All of the initial project year would be devoted to planning the program including the carrying out of geohydrological studies, test drillings and the preparation of tender documents for the boreholes program. An experienced rural water supply engineer would be recruited to direct the program and to help HER's water supply division to make the transition from being involved mostly in direct construction activities to taking on a major role in planning and maintenance; consultant services would be provided for initial studies and for yearly evaluations of the performance of the program. Specifications for pumps to be procured under the project would be based on the preliminary results of a UNDP study currently underway (para 1.05). Government has given an undertaking that it would submit to IDA for comment, proposals for a national system of user charges for village water supply which would, at a  - 21 - minimum, provide sufficient revenue to cover the maintenance costs of such water supply facilities; Government has also agreed to put such a system into effect by December 31, 1982. Further details of the 'village water supply program are in Working Paper 7. D. Cost Estimate 3.23 Cost Estumares. Total project costs for the four-year investment period are estimated at CFAF 7,800 million (US$28.9 million) with a foreign exchange component of 59.0%. These costs exclude import duties and excise taxes from which Bank Group financed projects are customarily exempt (as confirmed by Government during negotiations). Costs include physical con- tingencies of 15% on civil works, and nursery and forest gallery establish- ment and 10% on operating and maintenance costs. An allowance of 20%-30% of the base costs of vehicles and equipment has been included to finance spare parts. Costs have been based on prices prevailing in March 1981 and price contingencies have been calculated by compounding the following expected rates of local and foreign inflation: Local Costs % Foreign Costs % March 1981-March 1982 1/ 12 8 1982-83 12 (18.7) 8 (12.3) 1983-84 12 (33.0) 8 (21.3) 1984-85 11.5 (48.6) 7.5 (30.7) 1985-86 11 (65.3) 7 (40.2) 1/ Expected date of Credit approval. Effective compound rates in brackets. Such price contingencies amount to some 22.5% of base cost plus physical contingencies. A summary of project costs is in Table 1 and detailed cost estimates are in Annex 2. Table 1: PROJECT COST SUNMARY -----CFAF Million---- -----US$ Million----- % of Local Foreign Total Local Foreign Total Base Cost 1. Land Development 1,191 1,268 2,459 4.4 4.7 9.1 41 2. ORD Support 53 71 124 0.2 0.3 0.5 2 3. Credit 357 843 1,200 1.3 3.1 4.4 19 4. Village Stores 39 71 110 0.1 0.3 0.4 2 5. Village Reforestation 51 57 108 0.2 0.2 0.4 2 6. Village Water Supply 241 609 850 0.9 2.3 3.2 14 7. Project Management 282 270 552 1.0 1.0 2.0 9 8. Unidentified 266 380 646 1.0 1.4 2.4 11 9. Total Base Cost 2,480 3,569 6,049 9.1 13.3 22.4 100 Physical Contingencies 131 187 318 0.5 0.7 1.2 Price Contingencies 588 845 1,433 2.2 3.1 5.3 Total Project Costs 3,200 4,600 7,800 11.8 17.1 28.9 % 41.0 59.0 100  - 22 - E. Proposed Financing 3.24. In addition to the proposed IDA credit of US $16.0 million equiva- lent, financing for the project would be completed by a Dutch Government grant of US$7.8 million equivalent; a loan from the West African Development Bank (BOAD) equivalent to about US$3.7 million; and Government's own funds amounting to US$1.4 million equivalent. The BOAD loan would be at an effec- tive interest rate of 8%, since it would benefit from an interest subsidy grant available for rural operations; repayment of interest and principal would be in ten equal annuities commencing after an initial grace period of 3 years. The Dutch grant will be administered by IDA under the terms of an agreement between IDA and the Dutch development aid agency. The proposed financing plan, which is set out in Table 2, provides for: (a) 100% financing by the BOAD loan of the cost of vehicles and equip- ment, technical assistance personel and civil works required for the village water supply program (US$3.2 million), complemented by Government funds of US$0.6 million equivalent for local salaries, studies, training and the cost of the maintenance brigades. (b) Pari-passu financing by the IDA credit and Dutch grant in the proportion of 67:33 of the total cost of: (i) civil works: including a headquarters building for RDF; construction materials and supplies for land development works, village storage sheds, and buildings required for ORD training programs; and materials required for establish- ment and maintenance of tree nurseries and forest planta- tions (US$5.3 million); (ii) vehicles and equipment required for: land development works, ORD support services, rural forestry and overall project management by RDF (US$2.2 million); (iii) operating costs of ORD support services, HER technical services and RDF project management services (US$1.6 million); (iv) salaries and allowances other than for extension staff (US$2.4 million); and (v) unidentified projects (US$2.4 million). These funds would be non-reimbursable grants to be administered by RDF. (c) The IDA and Dutch grant to finance pari-passu in the same pro- portions as under (B) above, credit programs under three main headings: (i) oxen and animal traction equipment loans to individual farmers to be administered by CNCA (US$3.0 million);  - 23 - Table 2: PROPOSED FINANCING PLAN (US$ million) A. RURAL WATER SUPPLY IDA/DUTCH BOAD GOVT TOTAL 1. Vehicles and Equipment 0.4 0.4 2. Technical Assistance 0.4 0.4 3. Civil Works 2.4 2.4 4. Salaries, Maintenance, Studies and Training 0.6 0.6 B. OTHER COMPONENTS 1. Civil Works: (a) Office Building - RDF 0.9 0.9 (b) Construction Materials and Supplies (Land Development Works) 3.6 3.6 (c) Construction Materials and Supplies (Village Storage Sheds) 0.5 0.5 (d) Materials and Supplies for Buildings (ORD Support Services) and for Nursery and Forest Establishment and Maintenance 0.3 0.3 2. Vehicles and Equipment for: (a) Land Development Works (HER and ORDs) 1.8 1.8 (b) RDF and ORD Support Services 0.4 0.4 3. Operating Costs of: (a) Land Development Works (HER and ORDs) 0.8 0.8 (b) RDF and ORD Support Services 0.8 0.8 4. Salaries and Allowances: (a) Extension Staff 1.2 0.6 1.8 (b) Land Development Works (HER and ORDs) 1.6 1.6 (c) RDF and ORD Support Services 0.8 0.8 5. Credit for: (a) Oxen and Animal Traction Equipment 3.0 3.0 (b) Cereal Dehullers; Initial Stocks for Village Stores 0.5 0.5 (c) FAER 0.9 0-9 6. Unidentified Projects 2.4 2.4 7. Unallocated 4.3 0.5 0.2 5.0 TOTAL 23.8 3.7 1.4 28.9 % 82 13 5 100  - 24 - (ii) loans to farmers groups for the purchase of cereal dehulling equipment, and for initial stocks of cereals and general merchandise foe village stores, to be administered by RDF (US$0.5 million) and; (iii) term financing for FAER to cover working-capital required for the manufacture of animal traction equipent with such funds to be managed on behalf of Government by CNCA (US$0.9 million). Satisfactory terms and conditions of onlending to be applied to these credit: programs have been agreed with Government during negotiations and are in- corporated in the legal documents. (d) Extension staff salaries would be financed by the IDA credit (US$770,000 or 37%) the Dutch grant (US$380,000 or 18%) and from Government-s own contribution to project financing (US$0.6 million or 45%) with the latter to be provided at an increasing annual rate over the four-year life of the project. To ensure this, disburse- ments out of the combined IDA credit/Dutch grant against this category of expenditures would have an absolute monetary limit for each year in addition to specified percentages of total expenditures. The rest of project financing represents unallocated funds from the IDA credit and Dutch grant (US$4.3 million), the BOAD loan (US$0.5 million), and Govern- ments contribution (US$0.2 million). 3.25 Under the ongoing project, Government has encountered difficulties in prefinancing project expenditures which are subsequently reimbursed by the IDA credit. Government has therefore requested that an amount of US$750,000 (CFAF 203 million) representing three months expenditures on salaries and operating costs and purchases of materials and supplies required for civil works done on force account (all eligible for financing from the combined IDA credit/Dutch grant) be advanced out of the porposed credit and grant to establish a working fund. To this end, RDF would open an account (Special Account) with a commercial bank; this would be a condition of effectiveness for the proposed credit. During negotiations, Government undertook to ensure that the Special Account would be operated according to procedures set out in a supplementary letter and which provide inter alia that: (a) reimbursements after certification of the eligibility of the expenditures concerned, would be made directly to this account; (b) reimbursement requests would be accompanied by a statement of the movements in the account since the preceding applica- tion, certified by the bank in which the account is maintained; and (c) any ineligible expenditures would promptly be made good by Government. 3.26 The IDA credit would thus finance 55% of the total project costs, slghtly less than foreign exchange percentage of 59%. Combined external financing (IDA credit, Dutch grant and BOAD loan) would cover all of the foreign exchange costs plus 88% of local costs. Government's contribution to project financing would be 5% of total project costs and local farmers would provide labor for construction whose value, at the opportunity cost of labor used in the economic analysis of the project, is estimated at US$300,000 equivalent.  - 25 - F. Procurement 3.27 The construction contract for RDF's headquarters building (worth about US$0.9 million) would be awarded after competitive bidding in accordance with local procedures acceptable to IDA. Such tenders invariably attract bids from enterprises in the sub-region as a whole, including representatives of several firms with worldwide operations. Orders for tractors, vehicles and equipment (except animal traction equipment) and for construction materials and supplies (with an estimated total value of US$ 6.6 million) would be grouped whenever possible and any such grouped purchase order with a value in excess of CFAF 25 million (US$93,000) would be let after international competitive bidding following IDA guidelines. Other such tractors, vehicles, equipment, materials and supplies would be procured after competitive bidding in accordance with local procedures acceptable to IDA for orders valued between CFAF 5.0 million (US$18,000) and CFAF 25.0 million and through local competitive shopping for orders with a total value under CFAF 5.0 miliion. Animal traction equipment worth about US$3.0 million would be supplied by FAER which produces competitively priced equipment suited to local conditions, oxen will be purchased by local farmers from traditional traders and credit would be accorded only after a satisfactory health inspection by local veterinary officers. In principle, goods and services (including those required for unidentified projects) with a total value of US$9.9 million are liable to be procured under ICB or local procedures with virtually the same effect; this amount represents some 42% of the combined IDA credit/Dutch grant. 3.28 Construction of land development works and village-storage sheds, and establishment and maintenance of tree nurseries and village forest plant^ations would be done cn force account, with skilled labor and supervisory personnel from technical departments (HER, DAFR, and ORDs) and unskilled labor provided by beneficiary families. Total costs of such force account works in the form of salaries and operating costs of the concerned technical departments, is estimated at US$2.4 million. The rest of the combined IDA credit/Dutch grant would finance other salaries and operating costs US$2.8 million) and credit for village groups (US$0.5 million) and FAER (US$0.9 million); a further US$4.3 milion would be unallocated. Procurement of goods and services to be financed under the proposed BOAD loan would follow the procedures described in para 3.27 applicable for similar goods to be financed by the IDA credit/Dutch grant; and the contract for the construction of fully equipped boreholes (US$1.3 million) would be let after international competi- tive bidding. Contracts for technical assistance personnel required for the rural water supply component, worth an estimated, US$0.4 million, would be awarded in accordance with the procedures of BOAD which has however agreed to give IDA staff the opportunity to comment on the experience, qualifications and terms of reference of eligible candidates before the contract is finalized. Using the procurement arrangements described above, prior review of contracts for goods and services eligible for financing out of the combined IDA Credit/ Dutch grant would cover goods and services with an estimated value of US$13.8 million (71% of the total IDA credit/Dutch grant when unallocated funds are excluded).  - 26 - G. Disbursements 3.29 The IDA credit and Dutch grant would be disbursed pari-passu, and would cover expenditure items in the amounts and proportion of total cost set out below. Table 3: DISBURSEMENT CATEGORIES AND AMOUNTS US$ SDR % of Total Category and Description Equivalent Equivalent Expenditures 1. Construction of Office Building - RDF 900,000 770,000 100% 2. Construction materials and supplies for land development works, village storage sheds, buildings for ORD support services and for establish- ment and maintenance of tree nurseries and forest plantations: 4,100,000 3,480,000 100% 3. Salaries and Allowances: (a) Extension Staff 1/ 1,100,000 940,000 55% (b) Other 2,200,000 1,870,000 100% 4. Financing for FAER 900,000 770,000 100% 5. Animal Traction Equipment and Credit for Purchase of Oxen 3,000,000 2,550,000 100,% 6. Cereal dehulling equipment and initial supplies for village group stores 500,000 430,000 100% 7. Vehicles and Other Equipment 2,200,000 1,870,000 100% 8. Operating Costs 1,500,000 1,280,000 100% 9. Unidentified Projects 2,400,000 2,040,000 100%, 10. Working Fund (Special Account) 750,000 640,000 11. Unallocated 4,250,000 3,560,000 Total 23,800,000 20,200,000 In light of experience under the ongoing project, it is expected that the credit would be fully disbursed over a five and half year period (annexe 2, table 2), or one year earlier than the average experience. The average reflects delays experienced in start-up of new as distinct from follow-up projects. I/ These would be disbursed as follows: 90 percent of total expenditures in PYI up to a maximum of CFAF 130 million; 80 percent in PY2 up to a maximum of CFAF 120 million, 60 percent in PY3 to a maximum of CFAF 110 million, and 30 percent thereafter.  - 27 - Applications for withdrawals from the credit account would be fully docu- mented, except for local staff salaries and allowances, operating costs and minor civil works, which would be submitted with certified detailed statements of expenditures, whose supporting invoices would be retained for verification by the project auditors and for inspection in the course of project supervi- sion. H. Accounts, Auditing and Reporting 3.30 At present RDF keeps its accounts in conformity with the system recommended for use by non-revenue earning enterprises in the six countries of the West African Monetary Union. The system is satisfactory and RDF accounts are, with a slight exception, very well kept. The one exception relates to the way it accounts for credit operations: because RDF reimburses ORDs globally for all operations carried out by them on RDFs behalf, out- standing credit amounts due to FDR are deducted from amounts due to ORDs on account of other operations. Thus the impression is created that RDF- administered credit has a perfect repayment record. Under the proposed project, RDF would be required to keep supplementary accounts showing the real state of repayment by farmers to ORDs. RDF's accounts would be audited by independent auditors with experience and qualifications and under terms of reference acceptable to IDA. Certified copies of the audit and the audit report, including an evaluation of the appropriateness of the accounting system and internal controls, the use of the IDA credit for the purposes intended, and the accuracy of summary statements of expenditures, would be sent to IDA within six months of the end of each fiscal year. For reporting purposes, RDF would be required to submit periodic reports to IDA, modeled on present practice, as follows: (i) quarterly reports based on the progress reports submitted to RDF's director by the projects supervision division which focus on specific implementation problems and suggested improvements (para 4.03 g); (ii) annual reports based on the continuous monitoring and evaluation of a sample of subprojects chosen from each year's annual work program (para 4.03 h); and (iii) the annual work program including selection criteria upon approval by the IMTC (para 4.05). In addition, at the conclusion of the fourth year-s investment program RDF would be required to produce a project completion report. During negotiations agreement was reached with Government on the accounting, auditing and other reporting procedures described above.  - 28 - IV. ORGANIZATION AND MANAGEMENT A. Subproject Selection Criteria and Project Cycle 4.01 Subproject Selection Criteria. The success of RDF as a channel. of funds for small-scale rural development projects depends on its ability to select such projects according to principles which ensure their technical. and organizational feasibility and their economic efficiency. To this end, it has developed training manuals for use by ORD staff responsible for initial project identification in which clear directions are given for the conduct of a preappraisal. These good intentions have not always been realized partly because the information demanded of ORD staff is frequently unavailable but also because the work involved has been both materially and conceptually beyond the capabilities of such ORD staff. Frequently, also, not enough attention has been given to social factors such as land tenure arrangements. The result has been that some subprojects which seemed faultless on paper have turned out to be failures, the abandonment of developed bottomlands being the most striking example. Under the proposed project, identification and apprai- sal procedures for subprojects would be simplified and the use of selection criteria would be the main mechanism of control during the preappraisal phase. 4.02 The selection criteria would remain as under the two earlier projects, viz: (a) need as evidenced by a request from potential beneficiaries; (b) effective demand as evidenced by a willingness on the part of such beneficiaries to contribute in kind or in cash to the investment; (c) technical feasibility which implies a design simple enough to be realized by local farmers under the technical supervision of staff of executing agencies (ORD, HER, DSA, DAFR); (d) economic efficiency, which is deemed to be satisfied by a minimum economic rate of return of 12% on productive projects and, for infrastructure investments by a design judged to produce the lowest cost per beneficiary; and (e) social acceptability, which requires RDF to examine proposed subprojects taking account of the larger social context to ensure their acceptance by target populations. To ensure that these conditions are met, RDF would redesign its identification and preappraisal checklists to put more emphasis on social factors and to obtain summary, verifiable information. Sample checklists have been developed for land development works (Working Paper 1). RDF would also organize, at least two times per year, training seminars for staff of executing agencies on appraisal and supervision procedures. IDA supervision missions would, as far as possible, be programmed to coincide with such training sessions.  - 29 - 4.03 Project Cycle. For identified projects, the responsibilities of RDF and executing agencies would remain unchanged except for the introduction of a preappraisal phase: (a) Identification. It is expected that there would be an increase in the trend towards identification by potential beneficiaries. They would seek the help of the executing agency to put the request for RDF financing in a standard form which is currently used by the RDF Project Coordinator in each ORD. For land development schemes and village stores, ORD staff would be required to identify by name the members and leadership of the group making the request and the availability of affected land for cultivation by the members so named. Similarly, for reforestation schemes, availability of suitable land and the right of access of individuals to village block plantings should be ascertained by DAFR and ORD staff. For water points, it would be important that women are associated with the request since they are the most frequent users. It is not expected that this would create any social discord especially in villages with demonstration activities where women have actively participated in the choice and location of cereal dehullers. (b) Preparation, design, engineering. On receipt of requests and after putting such requests in standard form, ORDs would submit the requests to technical services--DAFR for reforestation, HER for water points, small irrigation, improved bottomlands and village stores--who would proceed to detail the tasks required, bills of quantity, construction methods, suitable species, etc. The detailed reports would be sent back to ORDs with copies to RDF. For small erosion works and simple bottomlands, the preparation task would be the responsibility of ORDs themselves who have the staff with the necessary experience. (c) Preappraisal. This consists in executing agency staff (principally the RDF Project Coordinator in the ORD and the sector extension agent) rediscussing the technical design, construction methods and social feasibility with potential beneficiaries and checking to see that, e.g., topographical and hydrological conditions are as assumed in the detailed design blueprints or that there are no impediments to free access to developed lands or village woodlots for any group of villagers. The object would be to obtain confirmation of the commitment of potential beneficiaries and of local leadership struc- tures both traditional and modern to the original request after it has been given concrete form. ORD staff would be responsible for this phase, but would call for the participation of technical serv- ices staff (HER, DAFR) where there are particular technical difficul- ties indicated. ORDs would send the complete documentation consisting of the original demand, the technical preparation report and a report of the preappraisal to RDF. For irrigation perimeters and improved bottomlands, HER would also at this time, submit copies of its design for approval by the Panel set up for this purpose (para 3.08).  - 30 - (d) Appraisal. RDF staff would be responsible for appraisal which would be on the basis of samples drawn from the mass of requests for (each kind of subproject and per ORD. Appraisal would involve visits to sites of the selected sample, discussions with potential benefici- aries, collection of data and detailed economic and financial analy- sis of the sample of subproject types on the basis of the data col- lected. For infrastructural projects (village input stores and wells) a simple measure of the cost per beneficiary would be enough. A work program per ORD and per subproject type, indicating execution tasks both geographically and over time would be prepared from the subprojects meeting the criteria discussed in para 4.02 along with execution contracts. RDF would prepare work programs in conjunction with executing agencies. (e) Financing Approval. Annual work programs are submitted to the 'Inter- ministerial Technical Committee (para 4.05) which gives its approval with modifications as necessary. Work programs thus approved carry with them the authority for RDF's director to disburse funds in accordance with the terms of the execution contract. (f) Project Execution would be supervised by ORDs for soil erosion work, simple bottomlands, village stores and reforestation and by HER and ORD for improved bottomlands and small-scale irrigation. Technical support in the form of periodic visits in the course of project execution would be provided to ORDs by DAFR for reforestation and by HER for land development works and village stores. For village water supply points, all boreholes would be executed by contractors under HER's technical supervision. Simple wells would be dug using voluntary labor by villagers under direct HER supervision and with the assistance of skilled masons under contract with HER. (g) Supervision. Works would be commissioned by RDF with the assistance of HER (boreholes, wells, village stores and land development schemes) and DAFR (village reforestation blocks). Project super- vision would be the joint responsibility of RDF and ORDs. While ORD staff would be responsible for seeing to it that assets created with RDF financing are used in the best conditions, RDF itself would confine its supervision activity to detailed examination of a major problem (or a set of problems) encountered in the operation of subprojects. This would be done selectively and the object wouLd be to suggest practical improvements to ORDs and to ensure the carrying out of such improvements. A priority area for study in the first year of project execution would be land tenure security on irrigated perimeters and improved bottomlands. RDF staff responsible for supervision would also ensure that spare parts and small tools mnd equipment required for some subproject operations are available for purchase by village communities.  - 31 - (h) Evaluation. RDF staff would be responsible for evaluating the impact of project actions on the basis of continuous monitoring of the sample of subprojects chosen at the appraisal stage (4.03 d). For this purpose, it would design single-paged questionnaires for each subproject which would seek to measure the evolution of the key parameters underlying its appraisal. Such questionnaires would be administered by ORD staff who would receive a briefing from staff of RDF's Studies and Programming Division (para 4.04) both before and after they are supposed to fill in the questionnaires. B. RDF Organization and Staff 4.04 Under the proposed project, RDF would be operating under revised legal statutes which provide it with the legal authority to act as an autono- mous institution under MRD supervision. Its major function would continue to be the management, on behalf of Government, of funds destined for investments in small-scale rural projects. The new statutes would also allow RDF to provide appraisal, supervision and monitoring services for other such projects sponsored by independent voluntary organizations, especially foreign based non-government institutions which have a large volume of activities in this domain in Upper Volta. RDF would be entitled to charge fees for these services. Government would also empower RDF to ensure that such small-scale rural investments are carried out under conditions which promote Government's development objectives or at any rate do not contradict such objectives. To this end, Government gave undertakings during negotiations that it would use its influence to encourage independent agencies planning small-scale rural projects to submit their plans to RDF for its comments. To provide back- ground on the institutional evolution of RDF, Annex 1 analyzes alternative institutional forms and the desirable characteristics which justify the form ultimately chosen. 4.05 As at present, RDF's effective Board of Directors would be the Interministerial Technical Committee (IMTC) chaired by the Minister of Rural Development (or his representative) and including eleven other members repre- senting key agencies. This committee is responsible for giving broad policy guidelines to RDF and would continue to do so under the proposed Third Project. In addition the new statutes also provide for a Standing Committee (SC) of IMTC, consisting of representatives of the departments of DSA, HER, DAFR, CNCA and RDF which would meet more regularly than IMTC--at least four times a year--and which would be responsible for ensuring the most practical way of carrying out IMTC's broad policy guidelines and also act as a tender board in the awarding of contracts following guidelines required by external financing agencies. 4.06 There are, at present, four divisions in RDF which report directly to the Project Manager: (a) Studies and Programming which is responsible for appraising sub- project proposals, preparing annual work programs and monitoring the impact of RDF subprojects. It is currently staffed by three expatriates: a rural engineerfinanced under a FAO project, a Dutch Aid financed economist and a sociologist also provided under Dutch Aid funds.  - 32 - (b) Project Supervision headed by an expatriate rural engineer follows up on all RDF financed projects. The division is made up of 2 teams, each composed of an agro-economist and a technical specialist (agronomist or rural engineer). Project management has divided ils geographical zone of operation into two halves for purposes of project supervision. (c) Training and Special Projects. In addition to organizing farmers' and extension agents' training programs, this division also has special responsibility for supervising development programs in demonstration zones (zones pilotes). It is staffed by an agronomist aided by a senior agricultural technician. (d) Finance, Credit and Accounting which does RDF's budgets, keeps accounts, prepares disbursement requests to external financing sources and processes disburseamnt requests from executing agencie.s. In addition, it keeps credit accounts for the pilot credit opera- tions undertaken under the Second Project. This division is fully staffed by Upper Volta nationals. 4.07 Under the proposed project, these four divisions would be streamlined into three by the dissolution of the training and special projects unit. Its training function, described in Working Paper 3, would be taken over by ORD training units (para 3.10) aided by a training specialist in RDF project super- vision division (para 4.08) and the supervision of demonstration and trials under the proposed project would become part of the normal work of the same division but with the technical support of DSA which would be given essential logistical support to accomplish its task (para 3.12). 4.08 Existing technical assistance positions in FDR would continue to be financed by France, the Netherlands and UNDP/FAO. The positions involved are Senior Rural Engineer, Economist and Sociologist in the Studies and Programming Division; and Agro-Economist, Agronomist, Rural Engineer and Training Specialist in the Project Supervision and Monitoring Division. Government has received agreement in principle from bilateral and multilateral aid sources to provide such technical assistance and it is expected that such personnel would be on duty by October 1, 1982. Should such technical assis- tance not materialize, funds in the financing plan proposed in this report under the categories of unidentified and unallocated would be sufficient to enable suitable staff to be financed. For all these positions, Voltaic nationals are either actively understudying the incumbent with the objective of taking full responsibilities in the course of the execution of the proposed project or have been identified as suitable potential replacements and are being trained with funds provided under the present project which would be supplemented under the proposed project. Only two expatriates are expected to be operating in RDF in the fourth project year. Overall, there will be only two more professional staff required for the proposed project compared to RDF 2 in spite of the increased volume of activity.  - 33 - V. TECHNOLOGICAL AND PRODUCTION SPECIFICATIONS A. Investment Phase 5.01 The simple design and construction methods used for soil erosion bunds and development of simple bottomlands have evolved from the experience of the two earlier projects. The length of contour bunds does not exceed 200 m for erosion control or 300 m for simple bottomlands. At erosion control sites, spacing of bunds varies between 10 and 50 m, depending on slope, with the objective of achieving a 0.5 m difference of level between consecutive bunds. Waterways between bunds are between 10 and 15 m. The design modifica- tions (para 3.07) introduced for bunds at bottomland sites would be systema- tically applied. Bunds and waterways would be planted with appropriate perennial grasses and bushes to fix the structures. There is very little purchased construction material used in such bunds. For improved bottomlands, reinforced spillways and better designed intake structures would complete the bunding system of a simple bottomland site for better water control. The design of irrigated perimeters would depend on the topography and would also be vetted by an independent panel of engineers and economists (para 3.08) to ensure low cost and technical feasibility. Village storage sheds would be built to standard specifications successfully used in the past. For village water supplies borehole construction, requiring the use of sophisticated drilling equipment, would be let on contract to private firms operating in the country. For simple wells, concrete lining would be fabricated on site using standard molds. Details are in Working Paper 1. B. Operating Phase 5.02 Yield increases predicted for erosion control schemes are not predicated on the generalized use of purchased inputs (fertilizer, especially) since this is justified only under favorable conditions, expected to be no more than 50% of total area developed. A lower rate of fertilizer use would have only a marginal impact on predicted economic results since without- fertilizer outcome is still a high proportion of with-fertilizer results (Working Paper 2). Fertilizer recommendations under such conditions would be 100 kg/ha of NPK (14.23.14). The demonstration program would test an alter- native fertilizer dressing consisting of 450 kg/ha of rock phosphate every fourth year supplemented by a 50 kg/ha application of urea as top dressing. The basic increases in production, however, would derive from improved cultural practices after bunds are constructed: seed dressing, timely sowing in line, weeding and ridging. Two cereal varieties have shown promising results in research station tests and would be tested at farm level under the trials and demonstration program, one for the southern parts of the central plateau and another for the drier northern zones. For maize composites, short-duration varieties suitable for the intermediate zone between lowlands and dry slopes would similarly be tested at farm level. For rice, suitable varieties exist for flooded, rainfed and irrigated cultivation conditions and would be recom- mended under the proposed projects. The total amount of seeds to be distri- buted is modest (less than 25 tons per year on average) and would be exchanged for farmers' harvested produce. These varieties respond very well to ferti- lizer (100 kg/ha of NPK and 50-100 kg/ha of urea) which would be available  - 34 - through the improved ORD credit and input supply bureaus to be financed under the proposed project. Nursery and transplanting techniques for rice would be tested in the trial and demonstration program. Labor availability would not be a constraint: a recent study of a neighboring, similar ORD, shows peak labor demand using improved practices to be less than 60% of available labor. A summary of key technical parameters in land development schemes is in Table 4 and input recommendations, expected yields and output are summarized in Table 5. C. Yields and Output 5.03 Yield predictions have been based on recorded experience in similar schemes in the central plateau. The data currently available show very large ranges for yield increases (e.g., in cereals after erosion control the range is 25%-65%). Although the yield predictions used in this report are based on the lower end, it is likely that improved availability of inputs and a better organization of extension services would have the effect of increasing average yields above the predicted figures. Increases in yield are expected to be progressively attained as shown in Table 5, although 100% of areas developed would be cultivated in the season following the completion of the development works, as has been the experience. Total incremental output from land development schemes is also shown on Table 5. Further details on inputs, yields and other agricultural aspects are in Working Paper 2. Table 4: KEY TECHNICAL PARAMETERS--LAND DEVELOPMENT SCHEMES Erosion Simple Improved Irrigation Control Bottomland Bottomland Gravity Purap Area (ha) 25 10 20 20 5 Area per Family (ha) 1.5 0.2 0.2 0.2 0.2 Length of Contour Bunds (m) 200 300 300 Length of Canals (m/site) - - - 200 50 Drainage Canals (m/site) - - 50 70 20 Concrete/Masonry Structures (m3) 0.1 0.2 120 950 15 Earthworks (m3) 3,750 2,250 3,000 2,200 800 Voluntary Labor (man-days/site) 550 700 4,000 6,000 1,500 Hours of Tractor/Other Heavy Machinery (total) 25 70 2,500 1,200 60 Direct Investment Cost (CFAF/ha) 2/ 47,000 161,000 840,000 2.20M 1.96M 1/ Cost per Farm Family (CFAF) 2/ 70,500 32,200 168,000 440,000 392,000 Cropping Intensity 90% 85% 120% 145% 145% Total No. of Beneficiary Families 13,400 3,200 1,200 400 400 1/ Includes pumping station equipment and protection structures. 2/ Civil works, topographic survey, design and supervision of construction.  - 35 - Table 5: PRODUCTION SPECIFICATIONS; YIELDS; OUTPUT ------Rice (Paddy)-- Coarse Bottomlands Bottomlands Ground- Vege- Cereals (simple) (improved) Irrigated nuts tables 6/ Seeds: kg/ha 16 80 80 80 100 6/ variety 1/ IR20/C74/ C74 & IR1529- 1/ IRAT 10 VIJAYA 680-3 Fertilizer 2/ 100 kg 100 kg 100 kg 100 kg 100 kg 50-300 u/ha complex complex + complex + complex + complex ea.of NPK 60 kg urea 100 kg urea 100 kg urea depending on crop ------------------------ (kg/ha) ------------------- -Mt/ha- Yield Ref yr 450 300 800 -- 450 -- Y1 610 980 1,620 1,750 625 5.2 Y2 650 1,200 2,000 2,630 675 7.8 Y3 650 3/ 1,500 2,500 3,500 675 10.4 --_________________------(mt)----------- ------- Incremental 4/ Output PYl 520 -- -- -- 80 -- PY2 1,190 110 70 70 190 -- PY5 2,720 600 380 430 430 810 PY8 2,720 5/ 700 450 530 430 1,450 1/ Initially local varieties, improved varieties under test. 2/ Compound NPK is 14.23.14. Project will test use of local rock phosphate and urea in place of compounds for cereals and rock phosphate for ground- nuts. Fertilizer applied to 50% of cereal and groundnuts area and 100% of paddy and vegetables. 3/ Assumes fertilizer need on up to 50% of area by third year of development in yields of a site. 4/ Incremental output is after allowing for post harvest losses of 15% coarse cereals, 10% paddy and 5% groundnuts. 5/ In addition erosiion control wold arrest present decline in yields estimated at 5% per year on unprotected sites; effect of erosion control measures would be to check losses equivalent to about 2,300 tons cereals by year 8. 6/ Vegetables would be a mix of potatoes, green beans, onions, tomatoes and cabbage according to area and market; seed rates vary from 0.34 mt/ha tomatoes to 1.7 mt/ha for potatoes.  - 36 - VI. DEMAND, MARKETS AND PRICES: FINANCIAL RESULTS A. Demand, Markets and Prices 6.01 Demand and Markets. With the major exception of rice and vegetables, most of the output from the agricultural production programs included in the proposed project is expected to be consumed by producing families themselves. The Central plateau region is, in general, a net importer of basic food stapLes (millet and sorghum) in a normal year. In addition to rainfall conditions, the low cereal output per capita is also caused, in part, by the migration of the active male population leading to a high dependency ratio. The promotion of village cereal banks would contribute to the development of alternative market opportunities for such staples since village groups would have both the financial means and the storage facilities to compete with private buyers soon after the harvest. Rice and vegetables produced under the project would be destined for population centers where demand is very high for both. Marketing of local rice has been more difficult largely because it is available to private purchasers in the form of bulky paddy which has to be transported and milled in urban centers where imports (sometimes in the form of food aid sold at lower than their value in international trade) offer severe competition. Paddy farmers under the proposed project would have the possibility of offering milled rice for sale thanks to the provision of rice dehullers on credit to village groups; such milled rice also stores better, enabling sales to be spread over a reasonable period. Total rice production under the project is, in any case, a very small proportion of annual imports, less than 5%. In all, project agricultural output would be disposed of without any major problems. 6.02 Prices. Farmgate prices used in the financial analysis of project results have been based on official prices for cereals and for others on observed prices in Ouagadougou and regional markets with allowances made for intra-seasonal fluctuations and marketing costs. These prices are compared in Table 6 with official prices, recorded prices in Ouagadougou for February 1981, and the economic values calculated for tradeable outputs. As the table makes clear, marketing is competitive enough that actual prices are usually above official prices. These prices are judged to give adequate incentives. Table 6: FARMGATE PRICES, MARKET PRICES, AND ECONOMIC VALUES (F/kg) Economic Crop Farmgate Market 1/ Official Value Paddy 63 97.5 63 106 Millet/Sorghum 50 77.5 50 90 Groundnuts (shelled) 54 133 54 65 Onions 40 100 183 2/ 40 Tomatoes 40 156 79 2/ 40 Green Beans 90 150 120 2/ 90 Potatoes 75 125 n/a 75 Cabbages 45 n/a n/a 45 Fuelwood 6 13 n/a 6 1/ Recorded data for Ouagadougou market, February 1981. 2/ UVOCAM prices in 1980 for vegetables; Ouagadougou.  - 37 - B. Financial Results 6.03 Farmer Benefits. At the farmgate prices quoted in Table 6 and on the basis of the yield and output projections in Table 5, farm and crop budgets have been worked out for the agricultural production subprojects. Results of the detailed calculations in Working Paper 2 are summarized in Table 7. Table 7: SUMMARY FARM BUDGETS (CFAF) Erosion Control Simple Improved Full A I/ B 2/ Bottomlands Bottomlands Irrigation Area (ha) 3/ 1.5 1.5 0.2 0.2 0.2 Gross Return Ref. Year 26,400 26,400 700 4,100 n/a PY 5 32,300 44,400 14,500 51,400 84,900 Net Return Ref. Year 24,400 24,000 450 3,500 n/a PY 5 30,300 35,400 12,000 35,600 52,300 Increase over Ref. Year 5,900 11,000 17,550 32,100 n/a Labor Input (man-days) Ref. Year 105 105 5 12 PY 5 122 139 25 58 82 Incremental Income per man-day 347 324 578 698 638 4/ 1/ Erosion works and no purchased inputs. 2/ Improved: erosion works plus fertilizer. 3/ Areas harvested are smaller. 4/ Average return/man-day. These results indicate increases in family incomes from project activities ranging from 25% over without-project income for erosion control works with no fertilizer use to 25 times without-project incomes on simple bottomland development works. The absolute increase in incomes, while small in relation to average family incomes in the central plateau, would still be considerable. It is estimated that an average family in this zone earns the equivalent of CFAF 30,000 (US$110) per year of which some CFAF 10,000 (US$37) is from  - 38 - non-agricultural employment. This is about 65% of the national average. Such a family, if it benefits from erosion control on half of its total dryland holdings of about 3.0 ha and farms another 0.2 ha in simple bottomland devel- opments, would increase its income by about CFAF 17,450 (US$65) or about 30% at full development for an increase in labor input of about 35 days. Families with larger holdings (about 7 ha) and a cash crop such as cotton on at least 1 ha (about 5% of expected project participants) would be able Lo do better with increases in income of up to 60%. This is because they tend to have larger holdings in bottomland areas. Per capita incomes would not, however, vary much between families because of the positive correlation between family size and size of holding. 6.04 Cost Recovery. The financing of recurrent costs has been a major problem especially in the rural sector (paras 1.16 and 1.18). The proposed project addresses this issue by incorporating arrangements for user charges in those instances where Government agreement has been obtained for such charges in comparable circumstances in earlier Bank Group projects. In one instance, rural water supply, the project would initiate a national policy of user charges which would be amended in the light of the results of the study now being undertaken with UNDP financing (para 1.05). Cost recovery arrange-- ments involve: (i) Irrigated perimeters: Government is required, under the terms of Credit Agreement governing the Niena Dionkele Project (Cr. 1013-UV of 1980, Sec. 4.04) to establish water charges on water control development projects nationally by December 31, 1981; to set such charges at a level which fully covers maintenance costs and such proportion of investment costs as is financially feasible for farmers; and to review such charges annually. In the case of irrigated perimeters in the proposed project a charge of CFAF 20,000/ha/year covers maintenance costs fully. (ii) Village reforestation: National policy, confirmed during the negotiations of the Forestry Project (Cr. 982-UV) calls for the sale of seedlings at prices ranging from CFAF 100 for local varie- ties of fruit trees to CFAF 200 for improved varieties; seedlings for plants not producing fruit are to be sold for CFAF 10 each. A stumpage fee of CFAF 200/m3 is also levied on transporters of fuelwood into urban markets. Under the proposed project seedlings would be sold and stumpage fees charged in the same manner and at the same rate as for the forestry project; and (iii) Village water supply: The study referred to earlier is expected to lead to the elaboration of a national policy on user charges. Without prejudging the results of this study, an estimate was made during appraisal of the maintenance costs of a village water supply system on force account. Under reasonable assumptions, a per capita charge of CFAF 175/year would be sufficient to cover mainte- nance costs fully.  - 39 - During negotiations, assurances were given by Government that it would, in the case of village perimeters and village woodlots, enforce the national policy on cost recovery as agreed in previous Bank Group supported projects and as outlined above, and, for village water supply, institute a system of user charges in consultation with IDA and which would recover at least the maintenance costs of such water points, at the latest by December 31, 1982. The impact of the project on Government finances (para 6.05) takes into account revenues derived from water charges on irrigated perimeters and from seedling sales and stumpage charges. 6.05 Financial Impact on Government. Project supported agricultural activities do not, in general, give rise to crops which can be directly taxed by Government. Furthermore, the expected use of fertilizer on part of the land to be developed under the project would result in subsidy obligations for Government. On the other hand, most expenditures financed under the project are for creation of physical assets whose subsequent maintenance would be borne by beneficiaries. The only expenses beyond the four-year development period at the charge of Government which can be attributed to the project are expenditures to keep ORD support services for training, demonstration and nursery maintenance and on incremental extension service costs to complete the training of farmers on land development schemes (para 3.05). The financial burden (including debt service) for Government on the basis of these assump- tions is presented in Annex 2, Table 3 which shows an annual deficit of about CFAF 65 million (US$0.2 million) in PY 1 which increases to about CFAF 362 million (US$1.3 million) in PY 5 and stays around CFAF 265 million (US$1.0 million) for another eight years as Government repays the BOAD loan. There- after, the annual deficit is about CFAF 125 million (US$0.5 million). VII. BENEFITS AND ECONOMIC ANALYSIS A. Major Benefits and Benefit Distribution 7.01 The central benfit of the project would be to complete the develop- ment of an institutional capability within Upper Volta for making sound investment decisions for the promotion of small-scale rural projects and for managing investment funds intended for this sector. The possibilities for such investments have been revealed to be very large in the course of the two preceding projects and their small size and simple design have ensured a very high level of participation by beneficiaries. It is very likely that, given the country's relative poverty, the existence of such a mechanism would accelerate the flow of official development assistance. 7.02 The subprojects which have been identified in the course of prepara- tion and appraisal would also have substantial material, if not always quanti- fiable, results. Thus, village stores and associated credit for working capital (para 3.15) would reduce the dependence of poor rural populations on private traders and the parapublic marketing agencies. The village reforesta- tion program would have a real, although non-quantifiable effect on the  - 40 - preservation of the natural environment (para 3.20). The 410 water points to be provided under the project (para 3.21) would respond to a basic need of rural populations in that part of the country. Some 60,000 villagers would have access to clean drinking water with consequent improvements in health and about 17,000 families would benefit from land development works. 7.03 The distribution of benefits among villagers under similar programs has generally been even; the major exceptions have been developed bottomlands and village woodlots. Under the proposed project, the preappraisal stage will particularly focus on the possibility of access to both village woodlots and developed bottomlands for all persons providing voluntary labor in their creation (para 4.03). Most project participants have family incomes below the national average which is itself very low (para 1.01). B. Economic Rate of Return Calculation 7.04 Economic rates of return have been calculated for all land develop- ment subprojects on the basis of prototype models (Table 4) and for the village reforestation program as a whole. Direct investment costs based on standard specifications have been compared with incremental output from a typical land development site. Basic assumptions are: (a) 10-20 year economic life depending on type of development and no residual values; (b) villagers' voluntary labor for construction and maintenance costed at CFAF 145, the estimated opportunity cost of labor; (c) full costs of required extension advice as discussed in para 2.11; and (d) a 10-20% charge for design. On these bases, economic rates of return for individual types of subprojects are satisfactory as summarized in Table 8 and detailed in Working Paper 8. The weighted average ERR for the land development subprojects is about 14 percent. Table 8: SUMMARY RESULTS OF ECONOMIC RATE OF RETURN CALCULATIONS Subproject Rate of Return Base Case Costs + 20% Economic Life-33% 1. Erosion Control 13 7 4 2. Simple Bottomlands 60 49 55 3. Improved Bottomlands 12 8 6 4. Irrigation 12 9 11 5. Village Reforestation 1/ 15 - - 1/ The whole program at an assumed price of 6 F/kg for firewood compared to Ouagadougou market prices of around 13 F/kg equivalent. Land development programs and village reforestation subprojects account for 70% of all project costs. No rate of return was calculated for village water supplies, for which there is no accepted methodology; and unidentified projects, whose economic benefits are expected to be similar to those for identified subprojects.  - 41 - C. Risks 7.05. The technical risks associated with the proposed investment program under a third phase project are minimal largely because they involve well- known and successfully tried techniques in both the investment and operating stages. The major risk derives from the possibility that assets created during the development period would be inadequately maintained either because funds for such maintenance are not provided by Government (the village water supply program) or because beneficiaries lose interest (as has happened in the past with bottomlands). Several measures are envisaged which would reduce this risk under the proposed project. First, works would be designed in such a way that maintenance would not require skills, materials and tools not readily available at the village level. For water points, financing is provided under the project for the training of village artisans in the simple repair and routine maintenance of pumps (para 3.22) and spare parts and tools would be readily available for sale to villagers (para 3.03 d). Secondly, instances of intended beneficiaries losing interest in a subproject because they are denied their fair share of benefits deriving therefrom (village woodlots and developed bottomlands) would be reduced by ensuring that no investments would be made unless all participants in the creation of such assets have fair and equal access to the ensuing benefits (para 4.03 c). In general, the location of investments would be guided by the past record of local populations in maintaining similar assets created during the first and second projects. Tests on the rates of return show that if one reduces the assumed economic life of project-created assets by one third, the rate of return for the most sensitive subproject (erosion control) falls to about 4%; for all others rates of return exceed 6%. VIII. AGREEMENTS REACHED AND RECOMMENDATION 8.01 During negotiations, Government gave formal undertakings on the following major points: (a) credit for animal traction equipment for the purchase of cereal dehullers, and of initial stocks of cereal and general merchandise for village-group owned stores would be on the terms and conditions sanctioned by the Central Bank, including interest at not less than 11% (paras 3.14 and 3.15); (b) water charges on irrigated perimeters and for village water supplies would be levied in accordance with national policy. For village water supply, Government would submit its proposals on user charges to IDA with a view to ensuring that such charges are adequate to cover at least the maintenance costs of such installations and that an acceptable system would be put into effect by December 31, 1982 at the latest (para 6.04);  - 42 - (c) title to lands benefitting from water control works financed under the project would be transferred to farmers (or groups of farmers) who have contributed in kind to the construction of such work (para 3.07); (d) a panel of three persons with experience, qualifications and terms of reference acceptable to IDA would be nominated at the latest by October 1, 1980 who would serve as a referee panel to approve designs submitted by HER for improved bottomland works and irrigated perimeters before such works are included in each year's work program (para 3.08); (e) the credit and supply bureau of the ORDs in the project area would charge a mark-up of not less than 7% of the value of animal traction equipment sold on credit and that these markups would be reviewed as necessary to ensure that they procure enough revenues to cover the operating costs of such bureaus by the third year of project opera- tions (para 3.11); (f) steps and measures judged satisfactory by IDA would be taken by Government which, inter alia, ensure the financial solvency of FAER; provide FAER with legal statutes which would enable it to function as an independent profit-making enterprise and ensure the employment of suitable management staff (para 3.17); (g) RDF would employ auditors with qualifications, experience and under terms of contract acceptable to IDA and certified copies of the reports of such auditors, including their evaluation of the appro- priateness of the accounting system and internal control mechanisms, of the use of the proceeds of the IDA credit for the purposes intended, and of the accuracy of summary statements of expenditures, would be submitted to IDA at the latest within six months of the end of RDF's fiscal year (para 3.30); (h) RDF would submit to IDA: quarterly and annual reports at the latest 3 months after the end of the period in question, and annual work programs at the latest by November 15, of each year (para 3.30); (i) RDF would prepare a completion report to be submitted to IDA at the latest six months after the closing date of the proposed credit (para 3.30). 8.02 It would be a condition of effectiveness for the proposed credit that Government has opened a bank account for a working fund which would be operated in accordance with procedures acceptable to IDA (para 3.25). 8.03 It would be a condition for disbursements out of the IDA credit for the working capital loan to FAER to be channelled through CNCA, that Government has taken appropriate steps to improve the operations and manage- ment of FAER along the lines discussed in para 3.17.  - 43 - 8.04 Providing that such assurances and conditions are obtained, the project would be suitable for a credit of US$16.0 million on standard IDA terms. l  - 44 - ANNEX I UPPER VOLTA THIRD RURAL DEVELOPMENT FUND PROJECT STAFF APPRAISAL REPORT PROJECT MANAGEMENT A. BACKGROUND 1. In Upper Volta, small rural projects encompassing the physical creation and efficient operation of both productive assets and rural infra- structure, are sponsored (identified, planned for, financed and executed) by a wide variety of agencies. In 1978/79, it was estimated that at least 16 non-governmental organizations (NGOs such as Caritas, Save the Children Fund and Oxfam) and bilateral aid sources were planning to spend in that year alone some CFAF 2.2 billion (US$8.1 million) 1/ on small rural projects ranging from irrigated market gardening to village dispensaries. The conditions governing the financing and operations of such projects vary considerably, as can be imagined. Wells and pumps on some irrigated perimeters are given as outright grants while on some perimeters they are given on credit, again under differing terms and almost always with poor repayment records. In the construction phase, some projects require beneficiaries to provide free labor while others offer payments in cash and in kind. Most of such projects are not subjected to a systematic appraisal of their technological, financial, economic or social feasibility. Above all, with the exception of small projects financed under the Second RDF Project, there is no attempt made to follow-up and monitor projects beyond the period of disbursement of project financing. 2. An implied objective of both RDF I 2/ and RDF II was to improve the planning system applicable to such small rural projects, and to provide better financial control during the investment and subsequent operations phases. Indeed, the achievement of the physical objectives of the two pro- jects, i.e. increases in agricultural production and the rural standard of living in general, depended on the evolution of a rational planning system and an efficient financing mechanism; this is recognised explicitly in both appraisal reports (paras 5.08-5.12 for the First Project, which discusses subproject selection procedures and criteria, and Annex 9 for the Second, which deals with the same subject). Already under the First Project, there was a recognition that such improvements should apply to all small rural projects and the Bank had obtained vague assurances from Government to this effect (para 5.18 of Appraisal Report). Finally, tha completion report on the First project in its last paragraph (47) summarized the main issue posed by the project: 1/ This is the equivalent of 48% of the Government's capital budget in 1977/78. Total NGO aid was CFAF 3.2 billion (US$11.9 million). 2/ When used with a Roman numeral prefix, refers to project, otherwise the institution is intended.  - 45 - ANNEX I (a) "should the organizational structure of the Rural Development Fund be strengthened and should it be given autonomy ....... (b) "or should the functions temporarily assumed by RDF be transferred to already existing institutions which would then be reorganized and reinforced". 3. A major objective of the proposed Third project would be to accom- plish the first of the two alternatives posed above and to expand RDF's responsibilities to include performing all or any of the following functions: appraisal, financial control and monitoring for small rural projects spon- sored by NGOs and bilateral aid. The rest of this Annex will consider, in turn, alternative institutional forms for RDF; its present organization,staff- ing and operating procedures; proposed changes and overall management proce- dures for RDF 3; and the cost and financial implications of these changes. B. RDF: "AUTONOMOUS INSTITUTION OR PROJECT" 4. The opportunity for development through investments in small scale projects in Upper Volta is adequately testified to, not only through the success of the first and second RDF projects but also by the volume of such activities undertaken with financing provided by NGOs and bilateral aid sources (para 1 above). The level of economic development justifies it: with Upper Volta's limited natural resource base, it is essential that the contribution of its human resources should be maximized. Since the bulk of this population is rural and possesses a generally low level of education, small, simply-designed rural investments have the best chance of ensuring high levels of participation by this population. In the specific case of the central plateau, with its high population density (44 inh./km2 compared to the national average of 20 inh./km2) and its rapidly deterior- ating soils, it is particularly urgent to pursue a program of small-scale rural investments since no reliable technical solutions are available which can be applied under a mass action program. The question then is not whether, but how, such investments should be planned and managed. 5. At present, the three alternative ways of organizing the planning, financial management and monitoring of small rural projects are: (a) a special service in the Ministry of Rural Development: If the problem of the persistent lack of operating funds for Government departments could be overcome, and if the functions were to be confined strictly to the formulation of policy, and project appraisal and monitoring this would be the ideal. Policy formula- tion rightly belongs in the Ministry. The skills required for project appraisal and monitoring are needed not only for small- scale rural projects but for all agricultural and rural sector projects and should be centralized as far as is possible. But by far the most important function of the kind of organization proposed is the management of investment funds. And this, Government depart- ments are notoriously inefficient at doing. Autonomy, flexibility  - 46 - ANNEX I and an ability to respond rapidly to changing conditions are neces- sary conditions for the successful management of such projects. The appraisal mission is fully persuaded that no conceivable reorganiza- tion of the Ministry of Rural Development will ensure that any of its centralized services would have these characteristics. (b) a "Project": It was presumably to ensure that these same characteristics were maintained that the Government and the Bank Group agreed in the first instance to create an "ad-hoc" entity, the Rural Development Fund, to manage the first project. In the beginning, because it was attached to another body which had a legal existence, namely the BND (The National Development Bank), there was no ambiguity in the status of RDF: in effect a special department was created within aLn existing bank to manage a series of investment activities. It was clear even then that this could only be a temporary solution. The situation now is that RDF is not part of BND, contrary to the requirements of the Credit Agreement of the Second project (Section ). The project director is responsible to a committee nominally chaired by the Minister of Rural Development. Since it has as yet no statutes, it would need to be "created again" theoretically, on each occasion that it is called upon to manage investment funds either local or foreign. This is unnecessarily cumbersome. Already, the implied threat of the imminent disappearance of the institution diminishes its effectiveness in dealing with other public agencies. It is also possible that potential external donors, in search of efficient and reliable local institutions which can manage their aid funds, have been deferred by the notion that RDF is only a project whose continued existence is not guaranteed. (c) An autonomous Institution The case for an autonomous body rests on the assumption discussed above (para 1), that small rural projects constitute an essential part of any development strategy for Upper Volta, and that such specific actions, limited in time and geographically scattered, will continue to be needed. The regional development agencies (ORDs) are executing agencies which do not have much capacity for carrying out their assigned duties; it is even less likely that they can assume more responsibilities for planning and project monitoring. It is also inconceivable that they themselves could have administered the lines of credit that they have used so effectively under the first and second RDF projects. Only an autonomous body, as RDF in its present form of an autonomous albeit ephemeral instituiton has amply demonstrated, can have the attributes - flexibility, rapid response capability, financial independence - which are necessary for the successful management of a program of investments whose exact content, size and location cannot be predetermined with precision. The ultimate responsibility of such an institution to MRD can be maintained side-by-side with a large degree of freedom of manoeuvre in its day-to-day operations within a clearly defined framework established in advance.  - 47 - ANNEX I C. PRESENT ORGANIZATION 6. RDF's present internal organization and staffing are dictated by the requirements of the Credit Agreement governing the second project and such modifications as have subsequently been agreed between the Bank Group and project management. There are at present four divisions which report directly to the Project Manager: (a) Studies and Programming which is responsible for appraising sub- project proposals submitted by ORD and HER (the Department of Rural Engineering) and preparing annual work programs in agree- ment with ORD directors for the geographical zones concerned. It also has responsibility for monitoring the impact of RDF sub- projects and preparing annual reports. It is currently staffed by three expatriates: a rural engineer financed under a FAO project, a Dutch-aid financed economist and a sociologist, also provided under Dutch-aid funds. This group was largely responsible for putting together the preparation report for the proposed third project; (b) Project Supervision headed by an expatriate Rural Engineer follows up on all RDF financed projects, including the commissioning of works, logistical support to ORDs and the preparation of progress reports. The division is made up of 2 teams, each composed of an agro-economist and a technical specialist (agronomist or rural engineer). Project management has divided its geographical zone of operation into two halves for purposes of project supervision; (c) Training and Special Projects. In addition to organizing farmers' and extension agents training programs, this divison also has special responsibility for supervising development programs in demonstration zones, (zones pilotes) which are in effect selected villages where new techniques are tested and which because of the combined effect of the variety of progress undertaken, serve as the point of training for other farmers in the general area. It is staffed by an agronomist aided by a senior agricultural technician; and (d) Finance, Credit and Accounting which does RDF's budgets, keeps accounts, prepares disbursement requests to external financing sources and processes disbursement requests from executing agen- cies. In addition, it keeps credit accounts for the pilot credit operations undertaken under the Second project. This division is fully staffed by Upper Volta nationals.  -48 - ANNEX I D. EVOLUTION OF MANAGEMENT ARRANGEMENTS 7. Institutional changes: If RDF were to have its role expanded along the lines described below, it would be desirable to establish it legally as an independent public body under the supervision of the Ministry of Rural Development. The formalization of RDF's legal basis for existence in this manner would essentially remedy an anomaly whereby RDF, unlike numerous and diverse organizations under the "tutelle" of MRD (e.g. CNCA, A.V.V., UVOCAM) operates with outdated legal statutes. The legal instrument embodying the establishment and operations of RDF should deal, inter alia, with: (a) responsibilities: Apart from confirming the legal authority of RDF to accept, on behalf of Government, and manage funds intended for small-scale rural projects' RDF would also be given (non exclu- sive) legal authority to ensure that such projects planned for execution by other agencies meet Government's broad development objectives. To fulfil this objective, Government would use its influence to encourage sponsors of such projects to use the services of RDF for some or all of the functions of appraisal, financial control, and monitoring and supervision that their projects might require. In particular all non-governmental organizations with operations in Upper Volta would be required to solicit RDF's comments on their planned operations before they commence. Such plans wou:Ld be presumed to have the approval of RDF eight weeks after their submission unless RDF responds otherwise in writing before the expiry of this period. RDF's suggestions for modifications in any such small-scale project would not be binding on the sponsoring body. RDF would, however, report any gross departures from national development priorities and objectives in a quarterly report to MRI). (b) finances: Apart from investment funds from whatever source which it would manage, RDF would receive Government budgetary allocations to cover (i) the salaries of confirmed civil servants on its staff (as is the case at present), (ii) normal operating costs of RDF inclucing the services of an independent auditor (see 7e below), and (iii) Government's contribution to projects managed by RDF and co-financed from other sources. It would also be empowered to receive fees from agencies which request and use its services for project appraisal, supervison, monitoring and evaluation or for management of investment funds. (c) relationship with executing agencies: RDF would rely on existing Government agencies and voluntary associations (e.g. cooperatives) for the execution of all investment programs under its management. For this purpose it would sign contracts of execution with the body concerned which would set out the obligations of both parties and specify the amounts and payment arrangements applicable to the funds required for carrying out the contract. Some contracts, such as with ORDs for the promotion of anti-erosion works development, would cover several years, with annual amounts to be agreed upon at the beginning of each year.  - 49 - ANNEX I (d) governing body: As at present, the effective Board of Directors would be the Interministerial Technical Committee (IMTC) chaired by the Minister of Rural Development (or his representative) and including eleven other members representing key agencies. 1/ This committee is responsible for giving broad policy guidelines to RDF and would continue to do so under the proposed Third project. In addition a Standing Committee of IMTC would be created, consisting of representatives of the departments of Agriculture, Rural Engineer- ing and Forestry, CNCA and RDF which would meet more regularly than- IMTC - at least four times a year - and which would be responsible for ensuring the most practical way of carrying out IMTC's broad policy guidelines and act as a tender board in the awarding of con- tracts following guidelines required by external financing agencies; and (e) audit: RDF's accounts would continue to be audited by independent auditors. Draft statutes are under preparation; they would be discussed during negotia- tions and it is expected that satisfactory legal instruments providing for the establishment and functioning of RDF along the lines described would have been enacted before Board presentation. 8. For the execution of the proposed Third project, FDR's existing four divisions would be streamlined into three, by the dissolution of the training and special projects unit. Its training function, described in Working Paper 3, would be taken over by the project supervision and monitoring division. The demonstration and trials begun under RDF II would be continued under RDF III but would become the direct responsibility of the Agriculture department (Direction des Services Agricoles) in MRD which would be given essential logistical support to accomplish its task (Working Paper 3). 9. Existing technical assistance positions in FDR would continue to be financed by France, the Netherlands and UNDP/FAO. The positions involved are Senior Rural Engineer, Economist and Sociologist in the Studies and Programming division; and Agro-Economist, Agronomist and Rural Engineer in the Project Supervision and Monitoring division. For all these positions, Voltaic nationals are either actively understudying the incumbent with the objective of taking full responsibilities in the course of RDF 3, or have been identified as suitable potential replacements and are being trained with funds provided under RDF II which would be supplemented under RDF III. Only two expatriates are expected to be operating in RDF III in the fourth project year. Overall there will be only two more professional staff required for RDF III compared to RDF II in spite of the increased volume of activity. 1/ Plan, Finance (2), Public Health BND, CNCA, the University, the Secre- tariat General of MDR, the Agriculture and Rural Engineering departments in MDR, and the Forestry Department in the Ministry of Tourism and Environment.  -50 - ANNEX I E. EXECUTING AGENCIES 10. Physical implementation of identified sub-projects would be the responsibility of existing Government departments which would be given logistical support where it is required for carrying out the assigned task. 11. HER - the Rural Engineering Department of MDR would have a major role both in the rural water supply and land development sub-projects. For water points HER would be responsible for planning, programming and technical control. In addition, it would directly carry out well maintenance activities and deepen some 200 wells using force account. A greater emphasis would be placed on advisory functions than on executive ones. Thus HER would be responsible for contracting for hydrogeological studies in advance of the establishment of each year's construction program, preparing tender documents and evaluating bids for the construction of boreholes by private contractors, training its own and village mechanics, and providing technical guidance to contractors. 12. For the land development sub-project, HER would be responsible for engineering and design work as well as for technical supervision of irrigated perimeters. It would continue to provide technical support to ORDs and help RDF to commission all finished structures. In the Dori area, where an experimental wind erosion control program is planned, HER staff would supervise closely both the construction and operational phases. The role of HER in implementation of land development works under RDF III would be much more active than before largely because of the introduction of watershed planning methods (see Working Paper 1). As a consequence, the topographic brigades (2 teams) in HER financed under RDF II would be maintained at least through PY3. Otherwise, it is not intended that HER should have an active role beyond the engineering and design stage and in the course of project supervision special attention would be given to the division of labor in land development works between HER andthe individual ORDs. 13. ORDs: As under RDF II, ORDs would be responsible for supervising land development works, construction of village stores and cereal banks, credit distribution and recovery, woodlot planting and maintenance, farmer training, and demonstration of new technology. In each of these areas, they would be supervising work done by farmers based on technical "blueprints" produced by the central services of MDR. These centralized services would also provide periodic technical supervision as appropriate: HER for land development; DSA for trials and demonstration; the Forestry Department for nursery establishment and woodlot plantation; and CNCA for credit operations. All this would be based on the practices and procedures fully tested under RDF I and RDF II. In the area of initial sub-project identification, for which ORDs have had and will continue to have, the major responsibility, stricter criteria would be introduced through the use of pre-appraisal check- lists which would require ORDs to list and evaluate the major factors which need to be considered before the investment decision is made. A model check- list has been developed for land development sub-projects and is at Working Paper 1. Similar check-lists would be developed gradually during the project investment period for all sub-projects. This should greatly simplify the  - 51 - ANNEX I appraisal task which RDF would undertake each year prior to establishing its annual investment program; it would also greatly increase the chances of picking successful projects, especially bottomland development works. 14. CNCA: would keep credit accounts for animal-traction equipment loans and supervise the work of the ORD credit units which would be reinforced under RDF III (Working Paper 4). Costs 15. Annex 1 presents the details of project management costs which are summarized below: CFAF Million Construction of RDF Headquarters 209.7 Vehicles and Equipment 59.5 Salaries and Allowances 171.2 Other Operating Costs 111.0 551.4 16. At price levels expected to prevail in PY4, the recurrent costs associated with project management would be about CFAF 110 million or about 6% of total funds disbursed in an average year.  - 52 - UPPER VO0'A ANNEX 1 Table I THIRD RIURA DEVELLORHE FUND PROJECT STAFF APPRAISAL REPORT Management Costs (CFAF 'OOO) Monthly Unit Cost/ PY 1 PY, 2 PY 3 PY 4 Total CoOt unitaire AP I AP 2 AP 3 A? 4 mensuel A. Salaries/Allowances A.1 Director General Director FDR (1) 380 4,656 4,656 4,656 4,656 18,624 Personnel Officer (1) 110 1,320 1,320 1,320 1,320 5,280 Executive Secretary (1) 150 1 800 1,800 1.800 1,800 7,200 Telephone Operator (1) 50 600 600 600 6W0 2,400 Messenger (1) 37 444 444 444 444 1,776 Watchmen (2) 37 888 888 888 888 3,552 Laborers (3) 37 1,332 1l332 1,332 1,332 5,328 Driver (1) 50 600 600 600 600 2,400 Subtotal 11,640 11,640 11,640 11,640 46,56( A.2 Studies and Programming Division iEconomist (1) 270 - 3,240 3,240 3,240 9,720 Rural Engineer (1) PM - - - - - Sociologist (1) 270 3,240 3,240 3,240 3,240 12,960 Secretary (1) 85 1,020 1,020 1,020 1,020 4,t8CI Driver (1) 45 540 540 540 540 2.16Q Subtotal 8,040 8,040 8,040 8,040 28. 92C A.3 Project Supervision Division Rural Engineer (1) PM - - - - Agriculturalist (1) 270 3,240 3,240 3 240 3,240 12,960 Agricultural Economist (1) 225 2,700 2 700 2,700 2,700 10,800 Rural Engineer (1) 225 2,700 2.700 2,700 2,700 10,800 Secretary (1) 85 1.020 1,020 1.020 1,020 4,080 Drivers (2) 45 1,080 1,080 1.080 1,080 4 320 Subtotal 10.740 10.740 10,740 10,740 42,960 A.4 Flhance Credit Accounting Division Chief of Division (1) 220 2.640 2.640 2,640 2,640 10,560 Accountant (BEP) (1) 135 1.620 1,620 1,620 1,620 6,480 Accountant Clerk (2) 130 3,120 3,120 3,120 3,120 12,480 Credit Supervision (1) 120 1.440 1,440 1,440 1,440 5,760 Supply Officer (1) 110 1,320 1,320 1,320 1,320 5,280 Secretary (1) 85 1,020 1,020 1,020 1,020 4,080 Storekeeper (1) 30 360 360 360 360 1,4440 Drivers (2) 45 1,080 1,080 1,080 1,080 4,320 Watchman (2) 25 600 600 600 600 2.400 Subtotal 13,200 13,200 13'200 13,200 52.800 Total Salaries 40 380 43-6p0 43 620 43,620 171,240 B. Vehicles and Equipment Directors Car 3,500 (1) 3,500 (1) 3,500 7,000 Station Wagon 2,500 (2) 5,000 (1) 2,500 (1) 2,500 10,000 Light Sedan 1,500 (1) 1,500 (1) 1,500 3,000 Pick-up 1,850 (2) 3,700 (1) 1,850 (2) 3,700 (2) 3,700 12,950 4 Wheel Drive 3,000 (1) 3,000 3,000 Airconditioning Units pm 1,580 1,580 Office Equipment pm 6.000 3,000 9,000 Spares pm 5,500 700 2,500 4,300 13,000 Total Vehicles and Equipment 29,780 2,550 11,700 15.500 59,530 C. Operating Cost Vehicles 45-75F/km 14 250 15,375 14,000 12,625 56,250 Utilities pm 6.190 6,190 6,190 6,190 24,760 Materials/Supplies/Maintenance pm 7,500 7,500 7,500 7,500 30,000 Total Operating Costs 27,940 29.065 27,690 26,315 111,010 D. Civil Works Headquarters Building 250F/m2 187,500 187,500 Store 30F/m2 15,000 154,POO Shed 18F/m2 7.200 7,200 Total Civil Works 209,700 209700 Grand Total 98,100 284.93'5 83,010 85,435 551,480  - 53 - ANNEX 2 UIpPR VOLTA Table I THllD RURAL DEVII0 FUNP PROJECT STAFF APPRAISAL REPORT Project Costs (CFAF '000) PY I PY2 PY3 PY4 TOTAL P.E. (N) A. Land Development i a. Civil Works 208 210 218 216 852 682 (80) b. Vehicles end Equipoent 302 103 35 58 498 423 (85) c. Salaries and Allowances 222 237 225 221 905 - d. Oper-ting Costs 49 51 51 53 204 163 (50) TOTAL A 781 601 529 548 2.459 1.268 B. ORCR Support Services Bi Trainine a. Civil Works 1.0 I.0 - 2.0 1.0 (50) b. Vehicles and Equipeent 4.7 - 0.5 3.7 8.9 7.6 (85) c. Salaries and Allowances 5.0 5.0 4.7 4.6 19.3 - - d. Operating Costs 6.1 6.3 5.4 5.6 23.4 15.2 (65) Subtotal 16.8 12.3 10.6 15.9 55.6 25.8 B2 Credit Adoinistrstion a. Vehicles and Equip=eet 7.4 1.4 - - 8.8 7.5 (85) b. Operating Costs S.2 3.2 - - 6.4 4.2 (65) Subtotal 1o.6 4.6 15.2 11.7 :3 Trisis and D-sonntrations a. Vehicles and Eqnip,eent 9.6 o.6 7.4 17.6 14.9 (85) b. Salaries and Allowances 1.4 1.4 1.4 1.4 5.6 - - c. Operating Costs 7.9 8.1 8.0 8.0 31.9 20.7 (65) Subtotal 15.9 10.0 9.4 16.8 55.1 35.6 TOTAL B 46.5 26.9 20.0 50.7 125.9 71.1 C. Credit a. FAER Credit Line 236.1 - - - 238.1 190 (80) b. CNCA - Adoinistered 147.1 186.3 265.4 222.4 821.2 533 (65) c. RDF - Adnisistered 79.1 47.5 18.3 1.5 140.7 120 (85) TOTAL C 458.5 233.6 284.2 225.9 1,200 845 D. Village Stores a. Civil Works 27.5 27.5 27.5 27.5 11D.5 71 (65) TOTAL D 27.5 27.5 27.5 27.5 110.0 71 --- E. Rural Forestry . Nursery Establishment 8.0 8.0 6.4 8.7 33.1 21.5 (65) b. Plantation Establishoent 5.8 9.0 12.4 13.3 40.E 26.3 (65) c. Vehicles and Equipsent 3.4 - - 1.8 5.2 4.4 (85) d. Salaries and Allowances 5.3 5.3 5.3 5.3 21.2 - - e. operating Costs 1.9 1.9 1.9 1.9 7.6 4.9 (65) TOTAL E 24.4 24.2 28.0 51.0 107.6 57.1 P. Rural Water SUDPly a. Civil Works 12 212.5 212.5 120 557.0 390 (70) b. Vehicles and Equipoent 23.1 35.6 35.6 - 94.3 8o (85) c. Salaries and Allowances s.6 7.5 7.5 7.5 28.1 d. Technical Assistance 16 30 30 4 80.0 hO (106) e. Maintenance Brigades 6 18 18 12 54.0 35 (6E) E. Other Operating Costs 15.1 7.5 7.5 6.9 57.0 24 (65) TOTAL F 77.8 311.1 511.1 150.4 850.4 609 G. Proiect Manacemest a. Civil Works - 210.0 - _ 210.0 147 (70) b. Vehicles and Equipnent 30.0 2.5 12.0 15.5 60.0 51 (8E) c. Salaries and Allowances 40.4 43.6 43.6 43.6 171.1 - d. Operating Costs 27.9 29.1 27.7 26.3 111.0 72 (65) TOTAL C 198.5 285.2 93.5 85.4 552.2 270 H. Unidentified Pra)ect 1958.3 1'2.6 162.6 162.2 646.i 391.2 Total Base Cost 1,671.9 1,672.1 1,445.7 1,259.7 6,049.2 3,i69.0 (59) Physical Contingencies 1) 51.0 114.2 83.8 69.1 318.1 157.7 Price Contingencies 2) 164.3 363.7 415.3 489.8 1,433.1 845.5 TOTAL PROJECT COSTS 1,887.2 2,150,0 1,944.8 1,818.4 7,800.0 4,602.2 (59) 1) 15 % on Civil Works. N-rsery and Plantation Establish-ent Costs; 10% or. Operating Costs, including cost of Well Maintenance Brigades. An allowance of 1O0%- 25% of initial u-it costs have been included in bane cost of vehicles and eqciplent to account for spares. 2) For local Costs cbopounded inflation rates are: pv 1: 18.7%; PY 2: 33%; PY 3: 45.t and PY4:63.3%. For foreign costs the c-rresponding rates are 12.3%; 21.3%; 30.79 and 40.2X.  Annex 2 9'ble 2 UP! VOLTA TRE1D ENDL DEVLOPBWM FUND PRWT:.- STAFF APPRAISAL EEPOET Financing Plan and Diebursment Schedule (CPUY lEAio.) Pfl PY2 PY3 PY4 TOTAL I. FIIUCII PLS tCFAF illio.) A. Baa wter SaWly 1. Vehicles nd uiTpment 40.9 47.6 47.6 12.0 148.1 2. kaheicel Aseiatance ad Consultants 22.8 30.0 30.0 4.0 86.8 3. CAvil WUrk 13.8 244.4 244.4 138.0 640.6 Fdmin.d by SOAD (100%) 77.5 322.0 322.0 154.0 875.5 4. Tom, Salaries, *erating Cnts Staff Training N0ovt.100%) 19.2 23.9 23.9 23.3 90.3 B. All Otker C ents 1. Civil Works (a) FnF Office Building - 241.5 - - 241.5 (b) Construction Materials and Suppliea 287.8 293.9 306.2 305.3 1,193.2 2, !ielS S d Equip-t 357.1 107.5 47.5 86.4 598.5 3. qpevtiDg Cost, 105.6 109.5 103.4 104.3 422.8 4. Credit (a) Anixal Tractin Equipment and Oxen 147.1 186.3 265.4 222.4 821.2 (b) Cereal procesBing equip. and initial trading atncks-villsge groups 73.1 47.3 18.8 1.5 140.7 (c) FAER 238.1 - - - 238.1 5. Salaries and Alnwancee (other than extension staff) 153.2 171.0 161.1 164.8 650.1 6. Unidentified Proejcts 158.3 162.6 162.6 162.6 646.i SUB-TOTAL 1,520.3 1,319.6 1,065.0 1,047.3 4,952.2 Flnomed by 100% by IDA (67%) 1,018.6 884.1 713.5 701.7 3,317.9 and buteb (33%) 501.7 435.5 351.5 345.6 1,634.3 7. rteacnim Salaries 120.9 121.3 116.9 111.1 472.2 Financed by 1DA 72.9 65.o 47.8 22.3 208.0 Dutch 35.9 32.0 23.5 10.9 102.3 Govt. 12.1 24.3 47.6 77.9 161.9 Total Bae Cwts 1,737.9 1,786.8 1,529.8 1,335.7 6,390.2 Plus. Unallcated (Representing Price Contingencies) 164.3 363.7 415.3 489.8 1,433.1 Of Wbich financed by IDA - - - - 794.1 r Ditch - 369.4 BOAD - - - - 124.5 Govt. - - - - 145.1 TOTAL I. TJEl COST - - - - 7,823.3 Bounded to 7,800 TI. DTIhAI SCREWL OF DIUSRShSES (IDA Credit and Dutch Grant - US$Millicc) Calendar Tr. 9 1985 1986 1'987 Qsrter 3 4 1 2 3 4 1 2 3 4 1 2 3 4 1 2 3 4 2 3 4 A,sout 0.5 0.5 1.0 1.0 1.0 1.0 1.25 1.25 1.25 1.25 1.5 1.5 1.5 1.5 1.5 1.5 1.0 1.0 1.0 1,0 0.S 0.3 Cuyla1tive Dish. 1.0 5.0 10.0 16.0 21.0 23.8  UPPER VOLTA THIRD RURAL DEVELOPMENT FUND PROJECT STAFF APPRAISAL REPORT GOVERNMENT CASH FLOW (CFAF MILLION) P11 P12 PY3 P14 P15 PY6 P17 PIN P19 P110O PYLL FY12 P113 A. OUTFLOWS... 1. Project Costs 1/ 1443.9 1916.9) 1660.9 1601.6 ------ 2. P081-project Admin. 2/-- -- 45.0 47.3 49.7 52.1 54.7 57.4 60.3 63.3 66.5 3. Capital Renewal 3/---- 28.7 28.7 28.7 28.7 28.7 28.7 28.7 28.7 28.7 4. Additional Extension Costs 4/ - -97.0 66.3 42.8 23.5 -- 5. Fertiliser Subsidies 5/ 19.4 40.8 53.5 63.4 43.0 23.1 --- - Total Outflows 1463.3 1957.7 1716.4 1665.0 213.7 165.4 121.2 104.3 83.4 86.1 89.0 92.0 95.2 B. INFLOWS...- 1. External Financing .1/ 1396.2 1831.9 1547.4 1440.9 (IDA) . (875.0) (994.1) (801.0) (845.4)a (Dutch) (429.4) (484.2) (388.3) (398.9).... (BOAD) (91.8) (353.6) (358.1) (196.6).. .-- 2. Water charges 6/. -0.8 1.6 2.4 3.2 3.2 3.7 3.2 3.2 3.2 3.2 3.2 3.2 3. Expenditure Taxes 7/. 1.3 3.5 7.1 11.1 15.2 17.9 19.9 21.4 22.5 23.6 24.8 26.0 27.3 Total Inflows 1397.5 1836.2 1556.1 1454.4. 18.4 21.1 23.1 ; 24.6 25.7 26.8 28.0 29.2 30.5 OPERATING DEFICIT 8/ 65.8 121.5 160.3 210.6 195.3 144.3 98.1 79.7 57.7 ; 59.3 61.0 62.8 64.7 (Before Debt Service........ C. -DEBT SERVICE .-...... 1. IDA Credit 1/ 6.6 14.0 20.0 26.4: 26.4: 26.4 26.4: 76.4' 26.4 61.3 61.0 60.8 2. BOAD Loan 9/---157 140.; 106 146 . 143.: 94.f 149.6~ 140.6 140.6 140.6 Total DbSevc-66 140 135.7 167.0. 167.0. 167.0 - 167.0: 167.0: 167.0 : 201.v : 201.6 201.4 D. NET CASH DEFICIT 65.8 128.1 174.3 346.3 362.3 311.3. 265.1 - 246.7. 224.7: 226.3 262.9 : 264.4 266.1 1/Excludes funds for credit operations but includes all contingencies. 6/ At CFAP 20 000/ha/year on irrigated perimeters 2/ For ORD Training and Research/Demonstration activities as at P14 level -71 15 4 on consumption expenditures equivalent to 25 7. of with price increases compounded at S %< p.a. value of increased production. 3/ 25 7. of cumulative purchases in PYI-PY4 for aOR activities as above. 8/ For PYl-PY4, equivalent to Government contribution to -4/ Numbers as per 14P3 Table 2 at CFAF 600 000 p.a. per extension worker project financing less expected revenues from irrigation charges plus CFAF 12 000 for supervision 1. and general expenditure taxes as per Pn 7 above. 5/ Subsidy decreasing successively from PY1 level of 697. to, respectively 9/ At 8 %/ per annum repayable in 10 annuities of principal and 60. 50, 40, 30. 20 and 0 in PY7. accumulated interest after 3 year-grace period.  UPPER VOLTA RID RURAL DEVELOI3FENT FUND PROJECT STAFF APPRAISAL REPORT Ecoomeic Rate of Return Celculation- ((FA2 '000) ---------------- A. ERlOSION CONTROL ----________ ------------- B. SIMPLE BOTLADS ---------------- ------------- C. IMOVED BOTTORANDS __________-_ --------- D. TREIGATED PRODUCTION --------------- (0T25 -he) (1-0 -has ('20 h.) (20 h.) ye-r Construction Extension Maintenance Net Constrcti-n Entens-ion Msintnnceo llnt Costruc-tion Eot-ntion Msintenance Net Coltruotton Extension Mainten e Net Costs Costs Costs Be Costs Costs -nitt -t C ost s C-t B fenefit Cn-tn Costs Cots 3enefits Costt Co.ts Cots B-enefits I 1,1i4.7 169.3 8.-j 170.6 1,013.5 67.7 10.5 16,004.0 507.9 110 - 37.423.2 1 015. 3 136.4 0 2 ' 84.6 8.7 210.5 ' 33.l3 10.5 513.3 ' 338.6 110 2,240.3 - 6,77.2 136.4 2,684.6 3 - 56.4 8.7 244.6 - 22.6 10.5 688.9 253.9 110 2,985.9 _ 707.8 136.4 14,597.2 4 4 42.3 118.7 281.5 i 16.9 127.9 928.4. _ 169.3 636 3,977.3 - 338.6 2,017.3 6,510.0 5 _ 42.3 8.7 316.0 16.9 10.5 928.4 _ 169.3 110 3,977.3 _ 330.6 136.4 6,510.0 6 -- 0.7 346.08 - 10.5 920.4 - 110 9,977.3 --136.4 6,510.0 7 8_ _ 0.7 377.8 - - 10.5 9280 _ 110 3,977.3 i _ 136.4 6,510o0 8 _ _ 118,7 40o.4 _ - 127.9 928.4 _ _ 636 3,977.3 _ 2,017.3 6,510.0 9 _ _ 0,7 405.4 - - 10.5 928.4 1 _ 110 3,977.3 - 136.4 6,510.0 10 8 _ 0.7 465.4 _ - 10.5 928.4 _ _ 110 3,977.3 - _ 136.4 6,510.0 - - ~~~~ ~~~~~110 3,977.3 --136.4 6,5io.o 636 3,977.3 ,017.3 6,510.0 _ _ 110 3,977.3 - - 136.4 6,510.0 - - ~~~~ ~~~~~110 3,977.3 --136.4 6,510.0 - 112 3,977.3 1 36.4 6,910.0 - - ~~~~~~~2,017.3 6,510.0 - - 136.4 6,51o.o - - 1~~~~~~~36.4 6,510.0 - - 1~~~~~~~36.4 6,510.0 - . - 0~~~~~~36.4 6,51o.o _EM Y RESULTS ERR (t) 12.8 59.7 11.6 12.2 SEtie.hlg Valoe- (sO L - Vst5t rnte C-n,truetinn C., - - + 9.7 -F 245.8 + 8.9 + 14.9 Be.efit. - 8.8 - 71.1 - 8,2 - 12.9 Net Present VPaie (at 12% discount rats) CFAF '000 159.7 3,088.3 1,534.3 5,808.7 N _ - % ef Pest Tests 9.7 247.2 0 9 4 9  W _ l ~ ~ ~~~~~~~~ ~~~~~~~~~~~~~~0. N. _GER_ UPPER VOLTA M A I THIRD RURAL DEVELOPMENT - d ( N,.o ~~iK UPPER LT IGERI FUND PROJECT - L ,RY moS HELNCeYOAST K t L ~~~~sso Bobo~~~~~~~~~~~~~~~~~_ib Banfoo b H A N A Third ural DevelopmenFudProectAreaDor I f v 14