Senegal - Casamance Rice Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 2056 FILE COPY PROJECT PERFORMANCE AUDIT REPORT SENEGAL - CASAMANCE RICE PROJECT (CREDIT 252-SE) May 10, 1978 Operations Evaluation Department 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 1972 US$1 - CFAF 277,71 1975 US$1 - CFAF 225.00 CFAF 1 - US$0.0036 CFAF 1 - US$0.0044 1973 US$1 - CFAF 230.20 1976 US$1 - CFAF 225.00 CFAF 1 - US$0.0043 CFAF 1 - US$0.0044 1974 US$1 - CFAF 250.00 1977 US$1 - CFAF 245.00 CFAF I - US$0.0040 CFAF I - US$0.0041 WEIGHTS AND MEASURES METRIC SYSTEM 1 millimeter (mm) - 0.0394 inches 1 meter (m) - 39.37 inches 1 kilometer (km) - 0.62 miles 1 hectare (ha) 2 2.47 acres 1 square kilometer (km ) 0.386 square miles 1 kilogram (kg) - 2.205 pounds I ton (t) - 2,205 pounds 1 liter (1) = 0.26 gallon ABBREVIATIONS AP - Agricultural Program BNDS - Banque Nationale de Developpement du Senegal CCCE - Caisse Centrale de Cooperation Economique (France) FAO - Food and Agriculture Organization FAO/CP - FAO/IBRD Cooperative Program FED - Fonds Europeen de Developpement IBRD - International Bank for Reconstruction and Development IDA - International Development Association INSTRUPA - Institut fur Strukturforschung und Planung GMBH (West Germany) ISRA - Institut Senegalais de Recherches Agricoles MDRH - Ministere du Developpement Rural et de l'Hydraulique PA - Programme Agricole PMU - Project Management Unit PRS - Projet Rizicole de Sedhiou Projet de Developpement Rural de Sedhiot RMWA - Resident Mission West Africa SATEC - Societe d'Aide Technique -et de Cooperation (France) SICCA - Societe Industrielle Commerciale de la Casamance SISCOMA. - Societe Industrielle Senegalaise de Construction Mecanique et de Materiel Agricole SODAICA - Societe de Developpement Agricole et Industrielle de la Casamance SOMIVAC - Societe de Mise en Valeur de la Casamance USAID - United States Agency for International Development (USA) FISCAL YEAR Government July 1 - June 30 FOR OFFICIAL USE ONLY Project Performance Audit Report SENEGAL - CASAMANCE RICE PROJECT (Credit 252-SE) Table of Contents Page Preface basic Data Sheet Highl-ights PROJECT PERFORMANCE AUDIT MEMORANDUM I. Project Summary 1 II. Main Issues Project Impact at Farm Level 6 Price Policies 9 Institutional Aspects 10 Farmers' Organization 11 Monitoring and Evaluation 12 III. Conclusions 13 PROJECT COMPLETION REPORT I. Introduction A.1 II. Preparation and Appraisal A.1 III. Implementation A.3 IV. Institutional Development and Performance A.11 V. Agricultural and Social Impact A.14 VI. Economic Rate of Return A.20 VII. Special Issues A.21 VIII. Changes in the Second Project A.22 IX. Government and Bank Performance A.23 Thie doument has a restricted distribution and may be used by recipients only in the performance - ii - Annex 1 - Yearly Project Progress Annex 2 - Farm Budgets Annex 3 - Economic Rate of Return Annex 4 - Difficulties Encountered According to the Project Director Annex 5 - Monthly Rainfall Map Project Performance Audit Report SENEGAL CASAMANCE RICE (Credit 252-SE) Preface This report presents the results of an audit of accomplishments under Credit 252-SE, for US$3.7 million, which was signed in June 1971, and closed, fully disbursed, in May 1977. It was followed by the Second Sedhiou project for which a credit of US$6.3 million, Credit 647-SE, was signed in June 1976. The report consists of a Project Completion Report (PCR) issued by the Western Africa Regional Office in July 1977 and a Memorandum pre- pared by the Operations Evaluation Department. The Memorandum is based on the PCR, a cursory review of project files, and discussions with Bank staff involved in the project. An OED mission also visited Senegal in July 1977. The mission had discussions with project management, the orga- nizations and ministries concerned with the project and visited the project area. The Project Manager contributed to the ex-post evaluation of the project by listing difficulties encountered. His contribution, which is included in the PCR, was part of the Borrower's request for a follow-up project and is striking for its frankness in commenting on the administrative and organizational difficulties. The audit concurs with the PCR's main findings about the project success but feels that the causes of this success deserve further analysis. The audit memorandum includes the summary and conclusions of the PCR as well as OED comments. The comments are relevant for lessons to be learned from this project as well as three other small farmers' projects in West Africa: the Sierra Leone Integrated Agricultural Development Project (Credit 323-SL); the Cameroon Semry Rice Project (Credit 302-CM); and the Benin Hinvi Agricultural Development Project (Credit 144-BEN). The Project Performance Audit Reports on these projects are being prepared simultaneously. The valuable assistance provided by the Government of Senegal proj- ect staff and the numerous farmers visited in the preparation of this report is gratefully acknowledged.  PROJECT PERORMACE A0DIT BASIC DATA SHEET SENEGAL CASAMANCE RICE PROJECT (CREDIT 252-SE) KEY PROJECT DATA Item Appraisal Actual or Expectation Current Estimate Total Project Cost (US$ million) 4.9 5.4 /1 Overrun (%) - plus 11% - Loan/Credit Amount (US$ million) 3.7 Disbursed )- Cancelled ) Aug. 31,1977 - Repaid to ) Outstanding to ) 3.7 Date Physical Components Completed Dec. 1976 Dec. 1976 1 Proportion Completed by Above Date (%) 100 65 /- Proportion of Time Overrun (%) -10 Economic Rate of Return (%) 18.5 22.7 OTHER PROJECT DATA Item Original Actual or Plan Revisions Est. Actual First Mention in Files or Timetable - - Oct./67 Government's Application - - Jul./70 Negotiations - - Nov./16/71 Board Approval - - May/11/71 Loan/Credit Agreement Date - - Jun./18/71 Effectiveness Date, Oct./31/71 Jan./06/72 Closing Date Borrower Government of Senegal Executing Agency Project RIZ SEDHIOU Fiscal Year of Borrower July 1 - June 30 Follow-on Project Name Second Project SEDHIOU Loan/Credit Number Credit 647-SE Amount (US$ millim) 6.30 Loan/Credit Agreement Date June 25, 1976 MISSION DATA Month/ No. of No. of Date of Item Year Days Persons Manweeks Report Identification 2/70 1970 Preparation 7-8/70 Appraisal 11/70 21 5 15 4/19/71 Supervision I 7/71 9 1 2 8/30/71 Supervision II 1/72 9 1 2 2/17/72 Supervision III 5/72 10 2 3 6/05/72 Supervision IV 9/72 12 2 4 11/14/72 Supervision V 1/73 9 1 2 2/15/73 Supervision VI 5/73 12 1 2 7/25/73 Supervision VII 9/73 12 1 2 11/12/73 Supervision VIII 9/74 6 1 1 4/19/74 Supervision IX 11/74 8 3 4 12/13/74 Supervision X 4/75 6 2 L 6/05/75 Total supervision 23 COUNTRY EXCHANGE RATES Name of Currency Communautd Financigre Africaine Franc (CFAF) Year: Appraisal Year Average Exchange Rate: US$1 - 277.71 (CFAF) Intervening Years Average US$1 - 232.55 (CFAF) Completion Year Average US$1 - 245.00 (CFAF) /1 Project was scaled down as several components proved unnecessary.  Project Performance Audit Report SENEGAL - CASAMANCE RICE PROJECT (Credit 252-SE) Highligh s The Casamance Rice Project supported the improvement of 2,000 ha swamp rice of which 1,000 ha through water control schemes, the clearing of 9,500 ha for agricultural production, and provided for agricultural credit and infrastructure. The project was successful due to yields substantially above appraisal estimates, prices higher than foreseen and lower input costs. Good management by local staff, effectively assisted by consultants also conttibuted to good project performance. The project deviated in several aspects from the appraisal design because suggested technical improvements were inadequately tested, and exceptional climatic conditions required changes in cropping patterns. Due to the Bank's limited knowledge of socio-economic conditions prevailing in the project area, farm labor and producer price constraints were underestimated, division of on-farm labor and family, structures were ill-defined. Resulting difficulties were overcome by the Bank's flexibility in adjusting project objectives to changing circumstances. The recalculated rate of return is 22.7% as compared to 18.5% estimated at appraisal. The following points may be of special interest: - substantially higher yields than assumed at appraisal (PPAR, paras. 4,6; PCR, paras. 5.03, 5.05, 5.08-5.10, 6.02); - underestimated unit costs combined with high inflation led to significantly reduced investments in road cons- truction and irrigation works (PCR, paras. 3.03, 3.06, 3.08, 3.10, 3.19, 3.25); - detrimental effects of Government's price policies on rice production (PPAR, paras. 19(iii), 21, 22; PCR, paras. 3.15, 7.02, 9.06); - Bank's lack of knowledge of local socio-economic conditions led to changes in project design and expected project impact (PPAR, paras. 19(i), 19(iii), 211 PCR, paras. 2.02, 3.02, 3.07, 3.12, 3.14, 3.15, 4.05, 4.06, 5.16, 5.19, 7.02, 9.04, 9.06); - return on man-days worked is a decisive factor in farmers' acceptance of recommended crops or farming systems (PPAR, para. 22); - farmers' participation in decision-making renders village committees into efficient organizations channeling credit and inputs to farmers (PPAR, paras. 24, 25, 29; PCR, paras. 4.04. 4.06).  Project Performance Audit Memorandum SENEGAL - CASAMANCE RICE PROJECT (Credit 252-SE) 1/ I. PROJECT SUMMARY- Background 1. The Casamance Rice Project (Credit 252-SE) was identified and pre- pared by the FAO/IBRD Cooperative Program in 1970. It was appraised in November 1970, the Credit Agreement was signed on June 18, 1971 and it became effective on January 6, 1972. 2. The Project was to: (a) establish a Project Management Unit (PMU); (b) improve 2,000 ha of swamp rice fields of which 1,000 ha by water control works; (c) clear 9,500-ha of land to grow rainfed rice in rotation with groundnuts and millet; (d) construct and operate 24 rice mills; (e) improve 244 km of roads; (f) provide seasonal and medium-term credit to project farmers through cooperatives and arrange for marketing of project production. 3. Although the development period was scheduled to be ten years, the bulk of the development was to be completed within the five-year disbursement period of the project (1972/76). Project costs during the first five years were estimated at CFAF 1.3 billion (US$4.8 million) of which US$3.7 million was to be financed by IDA. 1/ Adapted from the PCR.; -2- Proiect Implementation 4. The main results of the Project can be summarized as follows: PY4 1/ PYS I/ Difference Actual (A) Appraisal (A-E)/E Estimate (E) Z Number of Participants 23,000 20,000 + 18 Area Cultivated (ha) Swamp paddy 2,400 1,900 + 26 Grey soil paddy 3,000 1,400 +114 Upland paddy 300 670)0 - 96 5,700 10,000 - 43 Foodcrops 1,700 4,700 - 64 Groundnuts, 4100 8,000 - 49 11,500 22,700 - 49 Production Total Yield Total Yield Total Yield (t) (t/ha) (t) (t/ha) Paddy 18,000 3.2 15,000 1.5 + 21 +113 Foodcrops 2,600 1.5 5,800 1.2 - 55 + 25 Groundnuts 6,900 1.7 10,900 1.4 - 37 + 21 Net value of incremental 2/ production at farm gate prices (CFAF million) 757 515 + 47% Prolect Costs (CFAF million) 1,110 1,050 + 6 Economic Rate of Return 22.7% 18.5% I/ As a result of completing the project in PY4 instead of PYS, PY4 actual achievements are compared with PY5 appraisal estimates throughout this report. 2/ Estimated prices of products adjusted for inflation as foreseen at appraisal (a cumulative 3% per annum). - 3 - 5. Staffing and Training. Government set up the Project Management Unit (PMU) in 1971. Suitable Senegalese staff was appointed and, with the technical assistance of four expatriates provided by INSTRUPA, extension staff was selected and well trained. 11ousing and office accommodation was constructed but not made available to the project until 1975. In the mean- time staff had been reasonably well accommodated in existing Government premises in Sedhiou. 6. Agricultural Development (see Table at para. 4). As a whole the project was successful. While upland rice production was disappointing mainly because of drought, the Project successfully developed rice production on new lands though partly for reasons not foreseen at appraisal. The great 1972-73 drought induced PMU.and farmers to shift rice cultivation from the inadequately flooded and increasingly saline valleys,and from the rainfed fields on the plateau to the moist grey soils on the fringes of the valleys. These soils proved espe- cially suitable for rice cultivation and permitted very good yields (3t/ha) particularly since farmers applied fertilizers correctly. 7. Demand for farm inputs on credit, especially fertilizers, draft oxen and implements, has been strong, and the related repayment performance has been satisfactory (85-90%). 8. Water Control. Drainage and irrigation works of the type foreseen at appraisal were not carried out because the civil works unit set up under the project lacked experience and development costs were underestimated at appraisal. Nevertheless PMU has successfully developed four small pumping schemes (35 ha) to grow rice during the dry season. 9. Rice Processing. Instead of 24 rice mills as anticipated at appraisal, PMU acquired only two mills in 1975 of which only one is oper- ating. The need for rice processing equipment was less than foreseen (a) due to the price structure it was more profitable to farmers to sell hand pounded rice to private traders; (b) farmers had sufficient family labor available to handpound paddy themselves; and (c) home consumption of rice was larger than expected. 10.. Feeder Roads. PMU improved only 52 km of feeder roads instead of the 224 km anticipated because the cost of the road itprovement program was higher than expected. Ill. Marketing. Despite excellent performance in growing paddy, espe- cially grey soil paddy, little of it has been marketed through official channels because family consumption was higher than expected, the official marketing network'was inadequate, and the official price for paddy was not attractive compared to the price of groundnuts. Most of the marketed paddy was sold through private channels. 12. Financial. The credit was totally disbursed in June 1976, six months earlier than anticipated because of an underestimation of costs at appraisal, and the fall in the value of the US$ dollar against the CFA franc. Government and IDA agreed, however, to finance a second-phase project; this follow-up project called the Second Sedhiou Project became effective in November 1976. 13. Although overall actual costs exceeded appraisal estimates by only 62, equipment costs exceeded appraisal estimates by about 100%, staffing costs by about 20% and operating costs by about 250%, _while infrastructure costs were only about 30% of appraisal estimates due to the deletion of almost all the rice mills and water control schemes as well as the rescheduling of the road program. 14. Administrative. In 1972 Government and IDA funds were inefficiently chanAeled to the Project because of poor administrative procedures. The sit- uation improved in 1973 when Government made working capital available in the form of a revolving fund. Despite this improvement the project continued to have some difficulty in replenishing the revolving fund in time to meet current expenditures. 15. Credit and Input Supply. The appraisal mission anticipated that cr6dit for inputs and implements would be channeled to farmers through arrangements to be established between BNDS, ONCAD, farmers' cooperatives and PHU. It soon became apparent that cooperatives were not a convenient channel because: (a) paddy growers were mostly women (50-60% of participants) who could not be cooperative members and whose husbands were not keen to accept credit on their behalf; and (b) credit channeled through cooperatives is primarily used for groundnuts and not for paddy production so another channel had to be found for rice, millet and maize. 16. To cope with these difficulties PKU entered into an agreement with OCAD by which ONCAD was made responsible for supplying inputs to the project and the project was put in charge of input delivery to farmers and of credit recovery in the framework of village committees. This arrangement has worked satisfactorily. Moreover, when PHU started to deliver rice inputs to women, men became keen to get high-yielding seed varieties and fertilizers and began to grow grey soil rice, and to help the women. 17. Although the credit and input supply arrangements have been success- ful the overall performance of ONCAD has been poor throughout the project period and substantial delays have occurred in the delivery of inputs and transportaffionof groundnuts. Under the follow-up project, attempts are made to eliminate these obstacles: the project executing agency will integrate special project credit and cooperative credit, and take over the responsibility for supervision of cooperatives from ONCAD. Conclusions 18. The Casamance Rice Project succeeded in overcoming a number of problems and, largely because of a major reorientation in design, it sur- passed its objectives thereby achieving a higher than anticipated return. -5- It made a significant contribution to institution building, made or promoted technological innovations and had a beneficial social impact. Finally, it paved the way for a more integrated and better balanced development of the Casamance region and its experiences are being replicated not only in the follow-up Sedhiou II project but in other development activities in the Casamance. - 6 - II. MAIN ISSUES Project Impact at Farm Level 19. The audit mission concurs with the PCR (para. 4.11) and the Project Manager (Annex 4, pages 1 and 2) that the project has suffered from the lack of socio-economic data on the beneficiaries. The authors of the PCR succeeded to some extent in overcoming this lack of information and corrected several appraisal misconceptions. The area of rice cultivated per active person was appropriately reduced from 0.4 ha to 0.24 ha, which is closer to actual con- ditions. They also stressed the role of women who as it turned out not only help produce rice in Casamance but in fact also largely determine whether the crop will be used for subsistence or for sale, and they identified the low Government-fixed producer price as a major constraint to the development of rice as a cash crop. They were, however, less successful when comparing.the number of actual participants registered by PMU with the appraisal target of active persons instead of the number of farm families (see below). Certain other aspects of project impact remain vague or are incorrect: (i) actual participants. PMU registered about 24,000 participants, who are considered in Bank documents as active persons. The figure includes, however, extended family chiefs, household heads, and indi- vidual active persons. Also, registration methods vary from village to village and from year to year. The mission estimates that the actual number may be as high 0s 50,000 to 60,000 active persons, or close to 50% 1 of the active population in the project area.2/ (ii) area developed. The PCR estimates that the project increased the rice area to about 5,700 ha, of which about 3,200 ha were incremental. These figures are derived from discussions with extension agents who may have based their estimates on the quantities of fertilizer distributed by PMU, assuming 'that farmers apply fertilizer at the level recommended. They are also based on the assumption that, without the project, grey soils would not have been cultivated to any 1/ Comparison with the appraisal target remains difficult. At appraisal of this project, 5,000 families each with four active persons were to benefit from the project. The appraisal report of the second-phase project assumes seven active family members. Thus, it remains unclear how many active persons the project was actually intended to reach. 2/ In villages visited by the audit mission, more than 80% of the families participated in the project. In some, only widows, veterans and minority ethnic groups did not participate. Since the project covers 70% of the villages in the project area, the figure of 50% does not appear excessive. - 7 - significant extent. However, farmers actually use less-fertilizer than recommended initially 1/ thereby fertilizing about 50% more land. Some farmers also use ox-drawn equipment supplied by PMU on land which is not fertilized at all. Furthermore, the PCR does not account for inputs distributed by ONCAD,2/, , which has supplied farmers -many belonging to families that were registered with PMU- about 30% of total rice input requirements. On the basis of these findings, the mission estimates that the total rice area improved under the project was about 9,400 ha. However, the incremental area was probably not more than 1,600 ha because farmers intermittently cultivate grey soils and such culti- vation would have increased also without the project because of the drought which drove farmers from the uplands and the swamps.(PCR summary, para. 6) to the grey soil areas. The rice area, according to the audit's findings, increased mainly as a result of the large number of draft oxen distributed by the project. This raised labor productivity at a time when shortage usually prevented farmers from expanding their rice area. (iii) rice marketing. Bank documents assumed that farmers sell most of their rice production (100% of incremental production in the appraisal report, and 97% of total production in the PCR). However, the project area had always been a grain deficit area (PCR, para. 7.02) and changed to have a surplus only in 1975. 3/ Traditionally, 1/ The research organization has since found that, because of the high natural fertility of grey soils, smaller fertilizer doses might be justified. 2/ Office National de Coopgration et d'Assistance au Dgveloppement, the existing institution in charge of input distribution and credit handling, mainly for groundnut growers. ONCAD also sells fertilizers for cereal cultivation and ox-drawn equipment. These items are estimated to correspond to half of ONCAD's sales which grew during the project period from CFAF 13 million to CFAF 91 million. Corresponding figures for the PMU are CFAF 8 million and CFAF 113 million. 3/ Grain surplus (deficit) in the project area: 1972/73, (23,420tons); 1973/74, (7,240 tons),; 1974/75, (6,400 tons); 1975/76, 8,595 tons. - 8 - farmers in Casamance grow rice for subsistence, to serve visitors, and to eat it during religious ceremonies; 1 it is also a matter of prestige to have guanaries full (reserves for up to two years). Finally, women control the crop (partic- ularly of the swamp and grey soil areas, which are reserved for subsistence crops V and sell only small quantities at a time. Hence, appraisal ex- pectations and PCR assumptions about large quantities of rice being marketed are unrealistic. Exactly how much rice is actually marketed (in addition to the 4% sold through official channels for debt repayments) remains, however, unknown. Thus, while efforts were made by the PCR to adapt the appraisal farm model to prevailing conditions, data are still insufficient to permit a final conclusion on the contribution the project!! has actually made to the average farmer by the end of the investment period. 20. Time available to the audit mission was too short to adequately supplement available but insufficient data, and the figures advanced are tentative. The picture that emerges, however, is different enough from that presented in Bank reports to cast doubts on a number of assumptions, objectives and explanations given in these reports. Assumptions that can be challenged 1/ One socio-economist familiar with the area estimates that in certain tribes, only one-third of the rice produced actually reaches the granary. Two-thirds are to compensate occasional hired labor, are eaten during harvesting and serve social obligations. 2/ The appraisal report of the first-phase project indicated that tra- ditionally women cultivated rice but the matter was not considered sufficiently important to warrant more attention. In fact women control rice grown in the swamps and on grey soils, and determine production intensity (area inputs) as well as the further disposal of the crop after subsistence needs have been met. The PCR suggests that men have become more involved in grey soil cultivation. The only indication the mission could find to support this suggestion, is that men are cultivating the women's fields with their oxen and control upland rice production. It is interesting to note that in Sierra.Leone (preface) this division of labor is reversed, upland rice is cultivated by women for subsistence while swamp rice is grown by men mainly for cash. 3/ The mission recalculated the rate of return on the basis of its findings, and found an incremental production of 15,000 tons of paddy which is in line with the PCR estimates. The other discrepancies by and large offset each other and therefore the audit supports the PCR rate of return. - 9 - include the zero opportunity cost of farm labor and the decrease of groundnut and millet production as a result of drought, both used in preparing farm budgets and in calculating economic rates of return. The finding that about 50% of the active population is likely to have been reached during the first- phase project, suggests that the objective of the second-phase project--to increase project beneficiaries from 30% to 40% of the area's active population-- may have been accomplished even before the second project was started (para. 19 (i)). A similar question could be raised with regards to the potential for further grey soil rice development. 21. The audit's findings that farm labor is not overabundant and that rice was grown essentially for subsistence suggests that the explanations given by the PCR for the lack of farmers' interest in the experimental project rice mills (para. 3.15) are invalid. Rice milling in the traditional way is done by women by pounding paddy in a strenuous and time consuming manner. The success of the Chinese-financed rice mill where women are queuing-up demonstrates that there is a demand for processing facilities which would ease a burdensome task and at the same time permit women to spend more time in the fields. This demand is so strong that it is encouraging private operators to install new rice mills. The main reason for the underutilization of project mills is to be found in the inflexible system which compels farmers to sell their paddy to the mills, at unattractive Government prices, as compared to the "contract" work done by the Chinese mill. There, farmers deliver their paddy which is husked for a fixed charge (CFAF 7-10/kg) and the rice, which is destined mainly for subsistence, is returned to the producers. If the audit's analysis is correct, a change in this procedure would have improved farmers'utilization of project mills. Such a change could be envisaged under the second-phase project which also provides for small rice mills.i/ Price Policies 22. The PCR correctly emphasizes the detrimental effects of the Govern- ment's price policies on rice production. The mission agrees with this con- clusion but feels that the criticism calls for qualification. First, a higher price would probably not have induced farmers to sell significantly more rice, because under the first-phase project this crop was primarily grown for sub- sistence and stored against potential future crop failures. However, it can be speculated that more attractive prices might have prevented rice production on grey soils from decreasing after 1975. Second, at the time of appraisal of the follow-up project (June 1976), the Bank endorsed a farm gate price that was even below actual Government-fixed levels based on the assumption that additional funds would thereby become available for strengthening the financial basis of cooperatives. The audit concurs with the PCR's reservations on 1/ Two additional conditions should be met: (a) as indicated in the PCR, mills with about one-third less capacity than envisaged at appraisal should be selected; and (b) Government officials with a bias towards large mills should allow construction of the smaller, more efficient mills. - 10 - Government price controls which are a disincentive for expanded rice production (PCR, para. 7.02). It is, however, unfortunate that the Bank's contradicting views on the price policy conveyed to the Government within a year, may lead to a possible and unnecessary confusion of Senegalese officials. A possible loss of faith in Bank judgement could have been avoided, if the price policy issue and its consequences had been analysed more in depth in the appraisal of the follow-on project. Third, the Bank also recommended that fertilizer prices should be brought in line with actual costs. The Government accepted this recommendation, but due to the freeze on rice prices, the cost/benefit ratio of fertilizer application on rice narrowed and this explains why fertilizer purchases were reduced so drastically after the 1976 price increase. Fourth, the price of rice per se is not a constraint, but compared with the groundnut price, which reached CFAF 41.5/kg, the same price as paddy V/ in 1975, upland rice is no longer competitive when judged an returns per man-day. Upland rice is grown mainly by the men as a cash crop in excess of subsistence needs and is therefore more strongly influenced by price changes than grey soil rice. Expansion of upland rice was a major component of the two Casamance projects providing an additional cash crop to reduce--seen on a national basis--the risks of groundnut monoculture. The less favorable returns per man-day for rice due to the price equilibrium with groundnuts will give little chance to sizeable production increases of this crop and thus the Government diversifi- cation policy seems to face defeat. Institutional Aspects 23. PMU. Several factors contributed to PMU's successful performance, two of which deserve special attention: (a) the continuity and the experience of its staff, and (b) the good cooperation between the Project Manager and the four expatriate assistants from INSTRUPA, a German consulting firm. Many PMU field staff had participated in previous pilot schemes in the project area (map) and acquired experience in rice cultivation under prevailing con- ditions in Casamance. They were particularly acquainted with the agricultural potential of grey soils, which had been developed on an experimental basis in two of the schemes, and thus were instrumental in bringing PMU to focus on the development of rice on grey soils. The Government's favorable assessment of expatriate assistance is largely due to the Project Manager's capacity to use this assistance properly and to the expatriates' ability to use their 1/ Price changes over project years for main crops in CFAF/kg 1971 1972 1973 1974 1975 Groundnut 17,5 23,1 29,5 35 41,5 Millet 14 25 25 30 30 Maize 14 25 25 n.a. n.a. Paddy 21 25 25 41,5 41,5 At these prices, upland rice had a lower return per man-day (CFAF 394/man-day) than groundnuts (CFAF 453/man-day) and millet (CFAF 483/man-day). Grey soil rice had a return of CFAF 702/man-day. - 11 - authority diplomatically. Although they were in executive positions, the expatriates made a special effort to involve their counterparts in decision- making and to keep the Project Manager well advised. The unobtrusive role played by the expatriates in this project contrasts with the more obvious executive powers of those in the Sierra Leone project (preface). As a result, physical achievements in Sierra Leone were in line with appraisal goals while in Casamance they fell below appraisal objectives (Basic Data Sheet). However, the institutional accomplishments are quite different: the project executing agency for the Casamance project is showing no signs of deteriorating efficiency as in the ease of the Sierra Leone project after the departure of most of the expatriates at the end of the first-phase project. Admittedly, PMU requires continued external assistance under the second-phase project to maintain efficiency, but this is not in contradiction to the appraisal findings of the first-phase project which anticipated a 10-year development period for PMU until it would carry on completely on its own. The good working relationship with the expatriates has led to an extension of their contracts so that continuity will be maintained during implementation of the follow-up project. Farmers' Organization 24. In view of the inefficiency of the existing ONCAD cooperatives, PMU created village committees (Comitis Villageois de Diveloppement, CVDs) (PCR, paras. 4.03 and 4.04) to channel credit and inputs to farmers. The audit mission found that CVDs were less efficient than ONCAD cooperatives in credit recovery, but better in getting inputs to farmers on time because of the active involvement of the extension service in organizing input deliveries. In contrast to ONCAD cooperatives, which are noted for their over-centralized management and inadequate farmer participation in decision-making, the CVDs have become a forum in which farmers can voice their opinion and, these views, as the mission found, are effectively relayed to PMU's management through the extension agents. PMU's flexibility in regard to the appraisal proposal, . rightly considered by the PCR as one of the cornerstones of project success, is closely related to this interaction between CVDs and the PMU. On the whole, CVDs appeared to be more efficient tools for mobilizing farmer participation than ONCAD cooperatives, which is an advantage not mentioned in the PCR. 25. The future of the CVDs is, however, uncertain. The existence of two channels of credit and input supplies creates conflicts between political factions in villages and between the two development agencies. Farmers are also starting to play one agency against the other and switch from one credit channel to the other to avoid debt repayment. Finally, PMU and the CVDs have no incomes (rice is not marketed), while ONCAD generates revenue through its monopoly of the groundnut trade. - 12 - 26. The second-phase project proposes more clearly defined respon- sibilities: cooperatives would be reorganized to follow more closely the CVD system and the activities of ONCAD would be taken over by a regional develop- ment agency-the Socift6 de Mise en Valeur de la Casamance (SOMIVAC)--created in 1976 (PCR, para. 4.01) and headed by the former PMU manager../ Tangible results, however, are so far limited and ONCAD shows no inclination to forego its authority over the cooperatives. Considering that farmers' eventual support to SOMIVAC will depend on a thorough reorganization of ONCAD cooperatives which is in no way assured and remembering that previous attempts to reorganize ONCAD have met with little success V, the outcome of the project institution building efforts remains uncertain. Failure by ONCAD to adopt the CVD innova- tions of farmer participation, credit to women and the multicrop development objectives could impair the future success of the project. 27. Monitoring and Evaluation. PMU extension staff spend considerable time (more than 30% of total time) collecting and compiling information on yields. (2,229 samples taken in 1976, covering 10% of the participants), and these data can be considered as relatively reliable. Yield data, however, are collected without a clear view of what they will be used for. For instance, data on rice yields in fields cultivated without fertilizer or ox-drawn equipment are not collected and it is therefore impossible to judge the impact of these inputs. Also no data are collected on family composition, farm sizes, rice consumption or labor requirements. There are five main tribes in the area, each with a different family structure, concept of the women's role, and attitude toward selling rice. No attention has been paid to these differences and Bank guidance was particularly poor in this respect. 28., The PMU case does not bear out one finding of OED's First Review of Built-in Project Monitoring and Evaluation and confirms another. The above review found that in general yields were not measured regularly; this is clearly not the case here. The review also showed that cost effectiveness of data collection could be improved by paying more attention to the intended end-use of the gathered information. Monitoring and evaluation in the Casamance project could have been more cost-effective by reducing yield sampling on project fields and including samples of non-project fields, and by paying more attention to the important socio-economic variables for a small number (10 or 15) of families representing different categories of farmers (for instance: in each tribe, CVD members, ONCAD farmers, farmers using no inputs). Hopefully, the creation of a separate unit for monitoring and evaluation in the second- phase project will lead to a better approach. I/ PMU estimates would be coordinated by SOMIVAC. 2/ See the Project Performance Audit Report on the Senegal Agricultural Credit Project (Credit 140-SE), Report No. 1319 of October 1976. - 13 - III. CONCLUSIONS 29. In retrospect, project success is due largely to (a) the farmers starting to till the fertile grey soils as a result of an exceptional drought that impaired cultivation on other soils, and (b) good project management by a competent Senegalese Project Director, effectively assisted by expatriates and experienced field staff. PMU's main contribution to the success of the project is the establishment of a credit distribution and input supply system that was more flexible and efficient than the previous one, making inputs readily available to all farmers, including women, to support their efforts in cultivating the grey soils. Hence, PMU's ability to understand farmers' intentions and support their needs was more important in project implementation than merely carrying out project appraisal recommendations. 30. The project deviated from appraisal proposals for several reasons: assumptions concerning technical improvements were inadequately tested and proved over-optimistic on the whole, exceptional climatic conditions required major changes in cropping patterns; and because, as indicated in the PCR (para. 9.02), the project was inadequately prepared, farm size and family composition were poorly defined, land tenure constraints were not identified and subsistence requirements were underestimated. The audit supports the PCR in its favorable assessment of PMU and the Bank's flexibility in adjusting project objectives according to changing circumstances in the early years of project implementation. 31. Although the second-phase project incorporates several improvements (strengthening monitoring and evaluation, coordinating project activities with existing institutions, emphasizing crops other than rice), it does not fully take into account the lessons that could have been learned from the first project. Rice is still expected to provide more than half the incremental benefits, although several constraints on swamp and upland rice development encountered in the first project have not been resolved. Construction of rice mills is again included, albeit on the condition that their need can be confirmed by reviews but without appropriate changes having taken place that could enhance the farmers' response (para. 21). Also, farm labor and producer price constraints appear to have been underestimated and farm and family structures remain ill-defined. Because socio-economic factors which determined the results of the first project to such a large extent have still not been adequately studied, the second-phase project must remain insufficiently defined; flexibility by the Bank and project authorities will again be needed during implementation to make it also a success.  SENEGAL CASAMANCE RICE PROJECT (CREDIT 252 SE) COMPLETION REPORT .Table of Contents Page No. I. INTRODUCTION ............................... A.1 II. PREPARATION AND APPRAISAL ......................... A.1 A. Origin ...................................... A.1 B. Project Formulation .......................... A.2 C. Targets .. .. ..... .......... . A.2 D. Other donor agencies' roles................... A.3 III. IMPLEM ENT ATION ............................ A.3 A. Effectiveness and Start-up.*.**o****..*.. A.3 B. Mid-course Revision .......................... A.4 C. Physical Progress ... ............... .... .. A.4 General infrastructure .................. 4 Land Development and farm investments ... A.6 Processing and Marketing................. A.6 Special studies ......................... A.7 D. Procurement and Construction ..........o..... A.7 E. Project Costs and Disbursements .............. A.8 Total Project Costs .....................A.8 Cost Structure ...... .k.................. .8 Financing ....................... A.9 Subsidies ........... ... . ... ..... A.9 F. Covenants ...... ..... A.10 IV. INSTITUTIONAL DEVELOPMENT AND PERFORMANCE . ********, A.11 A. Institutional design and growth .............. A.11 B . Extension .......4......... ................ A.12 C. Input supply and Farm Credit ................. A.12 D. Staff and training issues .................... A.13 E. Foreign technicians .......................... A.13 F. Accounting .................. .... ....... .... A.14 G. Reporting .........*.... .................o.....o . A.1-4 (Table of Contents continued) Page No. V. AGRICULTURAL AND SOCIAL IMPACT ..................... A.14 A. Incremental Output .......................... A.14 Overall output ........................... A.14 Farm income .....................*** A.15 Reliability of output data ............... A.15 B. Technological impact .......................... A.16 'C. Prices and Inflation .......................... A.18 D. Number and Selection of Participants .......... A.19 E. Secondary Goals and Unexpected Effects ........ A.19 VI. ECONOMIC RATE OF RETURN .................*.. ****** A.20 VII. SPECIAL ISSUES ................ ................. A.21 A. Weather and risk .......................... A.21 Bo Sector influence ............................. A.21 C. Integrated development ...................... A.21 D. Replication ******.. ** ***********....... A.22 E. Settlement ........***************** * A.22 VIII. CHANGES IN THE SECOND PROJECT .................***** A.22 IX. GOVERNMENT AND BANK PERFORMANCE .................... A.23 Annex 1 - Yearly project progress para 1 Number of participants para 2 Crop development para 3 Farm outputs para .4 Activities per participants para 5 Farm input and field equipment para 6 Price of farm outputs para 7 Price of farm inputs and field equipment para 8 Total cost.of farm inputs para 9 Farm credit para 10 Project cost breakdown para 11 Disbursements para 12 Project staff para 13 Activities per extension worker para 14 Rice mills and storage para 15 Road program para 16 Irrigation and drainage works para 17 Cereal marketing para 18 Impact of the Project in the Sedhiou District (Table of Contents continued) Annex 2 - Farm Budgets Table 1 Y4 no oxen Table 2 Y4 with oxen Annex 3 - Economic Rate of Return Table I Economic Rate of Return Table 2 Price structure for paddy Table 3 Price structure for groundnuts Table 4 Price structure for maize Table 5 Price structure for sorghum Annex 4 - Difficulties encountered according to the Project Director Annex 5 - Monthly rainfall  SENEGAL' CASAMANCE RICE PROJECT (CREDIT 252 SE) COMPLETION REPORT I. INTRODUCTION 1.01 The Casamance Rice Project designed to be implemented over 5 years was financed by a US$3.7 million IDA Credit Agreement - Credit 252-SE - which was signed on June 18, 1971 and became effective on January 6, 1972. It covered the -Sedhiou District - "Department" of Sedhiou - in the Middle Casamance region which has high agricultural potential due to favorable rainfall. The project was the third Bank Group operation in the agricultural sector in Senegal, and the first in the Casamance area. 1.02 The main goal of the Casamance Rice Project was the development of rainfed rice production in rotation with other rainfed crops thus reducing dependence on rice imports. 1.03 Due to underestimation of costs, sharp price increases resulting from the oil crisis, and a decrease in the value of the dollar against the CFA Franc, the Credit was fully disbursed in June 1976, six months earlier than anticipated. In October 1975, IDA appraised a follow-up project, called the Second Sedhiou Project which became effective in November 1976. 1.04 This report is based on the findings of Bank supervision missions and of a completion mission in mid-1976, on PMU's comprehensive half yearly progress reports, and on evaluation reports prepared by.FPU with its effective monitoring and evaluation system. 1.05 Annex 1 to this report gives a compilation of this data, checked, updated and adjusted where necessary. 1.06 The report takes due account of the findings and comments of the project authority on the difficulties encountered in implementing the Project, which are summarized in Annex 4. II. PREPARATION AND APPRAISAL. A. Origin 2.01 The impact of rice imports on Senegal's balance of payments promoted Government in the late 1960's to prepare a program of rice development, and to explore the country's potential, both in the Senegal River Valley and in Casamance. In 1967, Government commissioned a pre-feasibility study by SATEC with FAC financing of possible rice development project. The SATEC report, prepared in January 1968, suggested two projects for the Valley to give an - A.2 - incremtntal production of 70,000 t per year, and two projects in Casamance to produce an additional 63,000 t. In view of the low investment costs compared to the Senegal Valley, the Bank preferred to finance a rice project in Casamance. The project was identified by missions from RMWA in January 1969 and FAO/CP in November 1969. The two missions spent 54 man-days in the country and the project area. B. Project Formulation 2.02 The project was prepared by a FAO/CP mission (16 man-days in the field) in July 1970 and a brief report was submitted. A five member appraisal mission (one of whom participated in the preparation mission) visited Senegal from November 1 to November 20, 1970. As experience was too brief to provide a firm data base for farm models and projections. Little information on upland crops was available at the farm level, the only data on production potential being from the local research station. As a result, the team had to rely mainly on preliminary data supplied by the SATEC report, together with experience gained by agricultural projects started in Casamance between 1967 and 1970; which had already developed 3,100 ha of rice, 50% of it rain- fed. 2.03 Discussions with Government during the formulation stage dealt mostly with the location of the project area in relation to other projects, and on organization and management. During negotiations, Government under- lined the need for flexibility in project implementation which proved to be the main factor contributing to the success of the project. C. Targets 2.04 The project aimed at bringing about the cultivation of 9,500 ha of rainfed or upland rice, which would be grown mostly in rotation with groundnuts and food crops on about 32,000 ha of cultivated land by some 5,000 farm families. It was also to improve about 2,000 ha of existing swamp rice land. In all, the project area covered about one-fourth of the Casamance region and about half its population. The project would be finished-within five years. It was to consist of: (a) establishment of a Project Management Unit (PMU) within MRDH reinforced by qualified and experienced specialists and provided with vehicles, housing, and offices. PMU would be responsible for carrying out the project and providing extension services to project farmers; (b) construction of irrigation and drainageworks for 2,000 ha of swamps; (c) construction and operation of 24 rice mills and associated storage; (d) improvement of project area roads; and -A.3 - (e) provision of seasonal and medium-term credit to project farmers through cooperatives controlled by the Project organization and arranging for marketing of project production. 2.05 Increased crop production was to be achieved by: (a) using improved varieties, fertilizers, pesticides, cleariAg equipment and ox traction; (b) introducing rice in rotation with groundnuts and foodcrops (millet and maize) in upland fields to be cleared; (c) improving agricultural techniques and implementing better water control for swamp rice. 2.06 The use of improved techniques was to be promoted through the provision of extension, cooperative, input supply and credit services. The weak data base did not permit a reliable quantification of production targets nor of the expected effect of individual factors on production. Spin-off goals were the building up of extension services, and the development of secondary economic activities in the project area such as marketing and processing. D. Other Donor Agencies' Roles 2.07 At the formulation stage, several agencies were already involved in the region. The selection of the project area resulted from the alloca- .tion of various zones of the region to various donors. The French Caisse Centrale de Cooperation Economique (CCCE) financed an interim project carried out by SATEC prior to IDA involvement in the project area; USAID also financed rice development in the area prior to project implementation. Other donors involved in Casamance included the European Development Fund (Banana project), the Dutch Government (Lower Casamance project) and the Government of the Republic of China. III. .IMPLEMENTATION A. Effectiveness and Start-Up 3.01 Effectiveness of the Credit was postponed.by 2-1/2 months due to delays in.selecting a consultant.firm acceptable to TDA to provide technical assistance and in concluding the contract. A contract with INSTRUPA was signed on January 3, 1972 and the Credit was declared effective on January 6, 1972. The other conditions of effectiveness, appointment of a (Senegalese) Project Director, the issue of a decree establishing the project authority, and a agreement between the development bank, BNDS, and the Government, were promptly met. The relatively short delay in effectiveness did not affect - A.4 - project performance, thanks to prompt start-up operations. Trained Sene- galese-personnel were available from the USAID and SATEC projects previously operating in the area. Project start-up was also facilitated by the assign- ment to PMU of vehicles, buildings and equipment available from the USAID and SATEC projects. However, some difficulties were encountered during the start-up period because of Government's failure to channel funds to PMU in advance. B.' Mid-Course Revision Project Design 3.02 The most important project revision was the official shift in em- phasis from upland rice cultivation to that of the grey soil regions in 1972. The moist grey soils -proved to be more productive, preferred by farmers, resilient to drought conditions and much cheaper to develop (see para 5.05). 3.03 Reasons for the curtailment of the water control works, the construction of rice mills and the road program are detailed in paras 3.06, 3.07, 3.08 respectively. 3.04 Cost overruns resulting from underestimation at appraisal, sharp price increases and variation in value of the US dollar against the CFA Franc prompted Government and IDA to advance the project closing date by one year from June 30, 1977 to June 30, 1976 - while the completiondate was advanced by six months from December 31, 1976 to June 30, 1976 - and to proceed with a follow-up project, preparation of which was financed under the project after reallocation of funds. C. Physical Progress General Infrastructure (Annex 1, paras 14, 15 and 16) 3.05 Actual achievements (see table below) were well below target levels, with. the notable exception of village stores, but agricultural production and development were not adversely affected, mainly due to the accelerated culti- vation of the grey soils. PY4 PYS Actual /1 Appraisal Estimate /1 Irrigated & drainage works (ha) 60 1,000 Rice mill units 1 24. Feeder road improvement (km) 52 244 Village Stores (construction/ rehabilitation) 67 0 /1 As a result of completing of the Project in PY4 instead of PYS, PY4 actual achievements are compared with PY5 apprai- sal estimated throughout this report. - A.5 - 3.06 Irrigation and drainage works. A combination of the following factors was responsible for the slow pace of irrigation development: (a) the realization that grey soils could yield good crops of rice at minimal development cost induced PMU and farmers to shift emphasis from irrigation of these soils; (b) the cost of irrigation development proved to be much higher than estimated at appraisal due to: (i) works being more s6phisticated involving dry season pumping rather than simple drainage works; (ii) price increases due to the oil crisis; and (iii) negligible use of unpaid farm labor as envisaged at appraisal. As a result, actual costs range from $520/ha to $1,520/ha as against $80/ha estimated at appraisal. (c) engineering personnel (expatriates and counterparts) had limited experience in irrigation development; (d) the data base was inadequate, especially for hydrology; The project, however, made a detailed survey of irriga- tion possibilities in the area, making it possible to speed up irrigation development under the follow-up project. 3.07 Rice Mill Units. This program failed to materialize mainly because the projected capacity (12,000 t for milling) was not needed. In 1975/76, PBRU .collected only 772 tons of paddy and the rest of the production, some 18,000 t, was hand-milled for family consumption or sale through traditional channels. The project did acquire a rice mill for testing purposes but with a capacity of only 200 kg/hour rather than 500 kg/hour as envisaged at appraisal as it was unable to purchase a mill with the latter capacity which would maintain -a satisfactory milling yield.. Under the second project, PMU will procure and test smaller mills (1 t per day) and will adjust its construction program to marketing needs. 3.08 Feeder Road Improvement. The curtailment of the road program resulted from an underestimation of costs at appraisal, combined with the high rate of inflation. At the opening of bids in 1972, the lowest bidder offered a price of CFAF 900,000/km against,CFAF 600,000/km anticipated at appraisal. Subsequently, it was decided to set up and equip a road *unit within the engineering division of PMU. Delays in obtaining the equipment resulted in delays in implementing the program. In two campaigns - 1974/75 and 1975/76 - the project improved only 52 km at an average cost of CFAF 1.6 million/km, the considerable excess over the bid price being due to the need to use higher specifications (90% gravelling, instead of 30%). However, the slow progress of the road program had no adverse effect on project output since the bulk of the incremental production was used mostly for family consumption. - A. A - 3.09 Village Stores. The construction or rehabilitation of 67 village stor s by the project was not anticipated at appraisal. These stores (50- 60 m ) belong to village committees and are used for storing inputs and equipment. Some were built under previous projects. The action of the project consisted of providing supplies (cement, roofing) and technical advice. Cost to the project totalled CFAF 6 million. 3.10 Project Buildings. The building program was completed in.1975, two years behind schedule, because of slow local bidding procedures, and failure by the contractor to meet the deadline (13 instead of 7 months). The program was reduced, as P14U decided to forego the construction of expatriate housing, so that although costs were well above those estimated at appraisal, the total cost of about CFAF 68.9 million remained within the appraisal estimate (CFAF 74.1 million). Land Development and Farm Investments (Annex 1, para 2) 3.11 The following table compares actual area cropped in PY4 against appraisal targets for PY5: PY4 PY5 A x 100 Actual (A) Appraisal E Estimate (E) ---------ha ---------- Rice Food Crops (Maize and Millet) 5,700 10,000 57 Groundnuts 1,700 4,700 36 Total 4,100 8,000 51 11,500 22,700 51 3.12 The shortfall in area cropped of rice, food crops and groundnuts was due to the switch from upland to grey soil cultivation but, in addition, the appraisal estimates of cropped area per family were too optimistic. As shown in Annex 1, para 4, the appraisal mission estimated that an economic- ally active person would develop 1.10 ha of land in PY4 which would have required a total of 220 man-days of labor per year or more than 35 working days per cropping month. In fact, each participant developed 0.49 ha of land which is a very reasonable achievement. In any event, production reached levels well above those estimated at appraisal and more growers were reached. 3.13 Demand for draft oxen and implements, particularly ox and donkey carts was strong.(Annex 1, para 5). Metal drums for seed storage and pedal paddy threshers were added to the list of field-equipment purchases, but sprayers and winches were not needed. Processing and Marketing (Annex 1, paras 14 and 17) 3.14 As stated in paragraph 3.07, the additional rice milling capacity provided under the project was not required. The main reason appears to be - A.7 - that farmers were able to hand-pound a larger quantity than expected by in- creased use of family labor. In addition, family consumption of rice in- creased as the result of the reduced output of upland crops caused by the droughts. 3.15 The quantity of paddy officially marketed through the-project was very low (only 772 t or 4% of the total paddy production in 1975/76) partly because of the higher family consumption, partly because the official mar- keting network was inadequate, and partly because the official price for rice was not attractive relative to the price for groundnuts. 3.16 The 152 cooperatives in the project area market, under ONCAD's supervision, about 35,000 t of groundnuts in normal years, including the additional groundnut production resulting from the project. ONCAD's poor financial situation has not adversely affected marketing in the project area although it had difficulty in moving the 1975 bumper crop out of local storage points and had not completed doing so before the beginning of the 1976 rains. Special Studies 3.17 Irrigation - Drainage Survey and Design. Of the 2,000 ha of swamp land to be improved for rice production, the project proposed to develop 1,000 ha by introducing better cropping methods and 1,000 ha by implementing simple drainage and/or irrigation works. The topographic survey and design work was to be carried out by the project Engineering staff and, while the work was in fact limited to 60 ha, the studies laid the ground for a more ambitious water control scheme covering 1,650 ha of swampland and for a survey of 15,000 ha of potentially irrigable land. 3.18 Preparation of Follow-up Project. Supervision missions recommended in 1974 that a follow-up project should be prepared, with financing from the remaining funds allocated in the 1975 budget. Preparation was carried out by PMU with assistance from the Bank's Regional Mission in Western Africa (RMWA) and from irrigation consultants. The follow-up project was appraised in October 1975 and IDA financing was approved in June 1976. This was the first bank group project prepared with a significant Senegalese contribution. D. Procurement and Construction 3.19 Procurement of vehicles, clearing and spraying equipment, and rice milling equipment were to be let through international competitive bidding but only vehicles were actually purchased under the project and were procured under ICB. -For road construction and improvement, contracts totalling US$0.6 million were to be let through local bidding. Prices offered by contractors, however, proved to be excessive and it was decided to have the work done on force account, with equipment financed under the project. Delays in obtain- ing this equipment resulted in higher prices as the result of inflation. The total cost of additional equipment not foreseen at appraisal was CFAF 90.7 million (US$400,000). 3.20 Procurement of consultant services from the German firm INSTRUPA followed IDA guidelines. - A.8 - 3.21 Animal drawn implements were purchased through ONCAD, which pro- cures such equipment from SISCOMA, a local manufacturer. There were some cases of poor quality, and of a lack of service and supply of spare parts. 3.22 Building contracts were awarded through local procedures to the firm SYLLA, which completed its work six months behind schedule. E. Project Costs and Disbursements (Annex 1, paras 8, 10 and 11) Total Project Costs 3.23 Total project costs, which amounted to about CFAF 1.110 million through PY4, exceeded appraisal estimates including contingencies through PY5 by about 6%. The comparison is made with PY5 estimates because the project, as a whole, met the PY% targets and, moreover, had provided most of 1976/77 input supplies by the end of June 1976. However, the closeness between appraisal and actual overall costs conceals wide divergencies in the cost structure and between individual components. Cost Structure 3.24 Cost structures can be roughly summarized as follows in rounded figures (including contingencies): Total Total thru PYS thru PY4 Appraisal Difference Actual (A) Estimate (E) (A-E)/E CFAF million ------ % Equipment 150 13 70 7 +114 Infrastructure 120 11 395 38 - 70 Staffing 610 55 520 49 + 70 Operating Costs 230 21 65 6 -254 1,110 100 _1_05 100 + 6 3.25 The differences between.actual and estimated costs as shown above were due to: (a) an underestimation of costs at appraisal and delays in procurement (see para 3.19); (b) a deletion of 23 rice mills, most of the irrigation works and a rescheduling of the road program; (c) an increase in local wages and salaries decreed by Government in late 1976 which was partly compensated for by delayed replacement of expatriates; and (d) sharp increases in prices of fuel, lubricants, spare parts, etc. - A.9 - Financing 3.26 Financing can be summarized as follows: PY4 PY5 A-x 100 Actual (A) Appraisal E Estimate (E) CFAF CFAF Million % Million % % Prolect Costs IDA Disbursement 839 64 1,028 76 82 Government 273 21 23 1 1,186 Total Project Costs 1,112 85 1,051 1 106 Farm Inputs Incremental Farm Input Costs at non-subsidized prices 170 13 282 21 62 Farmers' participation /1 (draft oxen supplied from local herd) 24 2 26 2 92 1,306 100 1,359 100 96 ] Estimates (Annex 1, paras 5 and 7). 3.27 Due partly to the initial delays in project implementation but more to the cumbersome procedures in Senegal for processing disbursement applica- tions, and failure to claim for some allowable expenses such as civil servant salaries, disbursement started six months later than envisaged at appraisal and ran about a year behind until the closing stages of the project. A graph of estimated and actual disbursements is in Annex 1, para 11. Subsidies 3.28 Government has been subsidizing fertilizer at about 50 percent nationwide, over the project period, thus contributing about CFAF 150 million to the.project costs. With regard to subsidy policies, the measures proposed under the project were realistic. While it was realized that groundnut opera- tions envisaged were too small to be used as a lever to change the fertilizer subsidy structure on this crop -- a policy later introduced in the second Agricultural Credit and Sine Saloumu project - the project called for rice fertilizer subsidies to be reviewed annually with IDA with the objective of - A. 10 - their progressive reduction (Section 4.08 of the Credit Agreement). Although formal reviews did not take place within the context of the Casamance project, general discussions on the nationwide level of fertilizer subsidies were held starting late 1975 under the Second Agricultural Credit Project and Govern- ment has reduced the subsidy for common fertilizer and urea from 74% and 80% respectively in 1974 to 56% and 30% in 1976. F. Covenants 3.29 The covenants to the Credit Agreement other than standard claims and conditions of effectiveness were that: (a) PMU enter into an agreement with BNDS concerning provision of funds for seasonal and medium-term credits for on lending to farmers through cooperatives; (b) PMUmaintain separate accounts and have them audited by a State Inspector and furnish a certified copy of the financial statement as so audited and the related audit report no later than six months after the end of each fiscal year; (c) the veterinary service of the Sedhiou District make available its services and facilities as needed; (d) ONCAD facilities be expanded to meet the demand for farm inputs, and appropriate arrangements be estab- lished between BNDS, ONCAD, farmers' cooperatives and PHU; (e) PMU operate project rice mills and impose and maintain an adequate per kg levy on milled rice; (f) PMU ensure purchase of paddy from farmers and resell paddy to ONCAD through farmers' cooperatives at prices no less than the wholesale price fixed for imported rice by the "office de commercialisation agricole" (OCAS); (g) the fertilizer subsidy be reviewed annually with IEA with a view to a progressive reduction; (h) Government maintain feeder roads; and (i) PMU carry out a survey This report shows that, although there were some delays, Government and PMU met all the covenants with only two exceptions: (a) PMU was unable to collect enough paddy to operate rice mill units (paras 3.07, 3.14, and 3.15); and (b) Government has not maintained the feeder roads and the maintenance of the most important of them has - A.11 - IV. INSTITUTIONAL DEVELOPHMNT AND PERFORMANCE A. Institutional Design and Growth 4.01 The Presence in Casamance of a number of projects sponsored by outside agencies, and the weakness of SODAICA, as Government agency operating in the project area, led the Bank and Government to agree on the creation of a separate project management unit reporting directly to the Minister of Rural Development, and incorporating Senegalese staff previously assigned to proj- ects in the area. Although the project still needs expatriate assistance in engineering and financial management, it can be credited with the build-up of an effective extension organization, a fledgling engineering section, and a strong and capable project manager. This organization will serve as a nu- cleus for a regional development agency designed to coordinate the several development projects in the area. The efforts of Government's organization (SOMNIVAC) will be furthered by a regional development unit financed by the follow-up project. 4.02 The PMU was given sufficient authority and the necessary means (buildings, vehicles, staff) to carry out its task effectively. A major difficulty was the slow channeling of Government and IDA funds to the project because of poor administrative procedures. The situation improved in 1973, when Government made a revolving fund available to the project. Despite these improvements, the project continued to have some difficulty in replenishing the revolving fund ("Caisse d' Avance") in time to meet current expenditures. Under the follow-up project, Government's contribution was -required to be fully paid into the project account at the beginning of each fiscal year. 4.03 ONCAD participated in the project by assisting and supervising cooperatives, supplying farm inputs on credit, and marketing groundnuts and cereals. The overall performance of ONCAD has been poor. Substantial delays have occurred in the delivery of inputs and transportation of groundnuts because of ONCAD's inadequate supervision. It operates at high cost and its financial situation is precarious. Detailed information on the perfor- mance of ONCAD is presented in the completion report of the First Agricul- tural Credit Project (Credit 140-SE). 4.04 Cooperatives and Village Committees. The Sedhiou district has 152 cooperatives, supervised by ONCAD, made up of heads of households only, and devoted primarily to input supply and marketing for groundnuts. Parallel to this structure, PMU organized the 25,000 project participants into 445 village committees. Their purpose was to: (a) provide a channel for more productive extension work; (b) streamline the supply of inputs through a complementary project credit ("Credit special"), especially for rice, which is not wholly covered by ONCAD; - A.12 - (c) provide credit not only to heads of households 1/ but also to female growers who are not eligible for ONCAD credit but who are in fact the main growers of swamp rice; and (d) promote cooperative education among farmers. Under the follow-up project, PMU is to be given responsibility for the supervision of all cooperatives in the area; the aim will be to improve cooperatives' management and consolidate the coopertive movement through the creation of a cooperative union and the organization of a regional federation. Thus, the project and its follow-up are playing an important role in the promotion of a sound cooperative system. B. Extension (Annex 1, para 13) 4.05 Under the project, an extension structure was built up, with 12 senior extension agents and 147 field extension workers (as against 100 anticipated at appraisal) operating in 1975. This increase over estimates stemmed from a parallel increase in project participants, and the agreed ratio of extension workers to farmers was maintained at the appraisal esti- mate of 1:160. With the reduction in cropped area (para 3.12) the area per extension worker fell from 74 ha to 30 ha but the increased production meant that output expressed as a ratio per extension worker increased from 102 to 124 tons of paddy. The extension workers supervised individual partici- pants rather than farm families because swamp rice growing is traditionally the responsibility of women and therefore rice extension work among heads of families would not have been productive. Until recently, extension workers have deliberately focused their activities on input supplies - especially fertilizers - for lowland rice. In order to maintain the unique balance between lowlands and upland farming in the project area the follow-up proj- ect provides that extension workers be responsible for *all aspects of farm development in both the valleys and the plateau, by inter alia, transferring input supplies to cooperatives. C. Input Supply and Farm Credit (Annex 1, paras 5 and 9) 4.06 The appraisal mission anticipated that most of the farm inputs would be supplied under farm credits channelled to farmers through cooperatives. However, it became clear early on in the project that cooperatives were not a suitable channel, firstly because most of the rice farmers are women, who are debarred from being cooperative members (membership in ONCAD coop- eratives being restricted to family heads), and whose husbands were not keen to accept credit on their behalf; secondly because credit channelled through cooperatives is primarily used for groundnuts and not for rice production. To overcome these difficulties, the project entered into an agreement with ONCAD, in which ONCAD was made responsible for supplying rice inputs to the project and the project was responsible for their delivery to farmers and for credit recovery ("Credit special"). This arrangement has worked satisfactorily. 1/ The general structure of the "carre" or family compound, is very complex in Casamance, with usually more than decision maker in the same "carre". - A.13 - When the project started to deliver high yielding seed varieties and fertil- izers for paddy to women (about 50% of project participants), men became interested in rice cultivation and began both to grow rice themselves and to help women in their fields. In addition, credit recovery was satisfac- tory despite no control by the project over paddy or rice marketing. On the other hand, although this approach has been successful, the separation of groundnut credit from rice credit caused higher costs and administrative difficulties and therefore, the second project integrates both kinds of credit. 4.07 Short-term credit was financed by BNDS using funds borrowed mainly from the Central Bank of the West African Monetary Union at 5.5 percent in- terest. Medium-term credit for draft oxen and ox-drawn implements was at 5 percent interest with repayment over 5 years. New short-term credits are made available only if repayments are repossessed in case of default. While delinquency was higher among project participants who were not members of a cooperative (85 percent repayment as against 97 percent for cooperatives), the overall rate of repayment averaged 90 percent in 1975. D. Staff and Training Issues (Annex 1, para 12) 4.08 A competent staff has been assembled at all levels and has gained experience during project implementation in many aspects including project management, agronomy seed production, cooperatives, and field extensions. Weaknesses have been evident in the areas of engineering and accounting. There are few rural engineers in the country, and good accountants are attracted by the city and by higher salaries in the private sector. The accountant had to be replaced three times in four years, and one of them disappeared with the project cash. However, the performance of the expa- triate financial analyst was good. The project provided training to the field staff and under the second project provision is made for further training of staff at all levels, including management staff. E. Foreign Technicians 4.09 Consultant services were contracted to INSTRUPA, a German firm. The consultants did a creditable job on the whole, in spite of high turn- over, resulting in part from the difficult living conditions (lack of schools, medical services, stores, limited utilities, and isolation). An engineer was terminated by the project manager for lack of competence and the contract of an agronomist who lacked initiative and leadership qualities was not renewed. Relations between expatriates and nationals have been good., With strong leadership provided by the Senegalese Project Manager, the expatriates have been fully integrated as project personnel. Before the end of the project, the expatriate agronomist was replaced by his counterpart who has performed satisfactorily but expatriate support is still needed in engineering and financial management. Under the second project, expatriate support will be provided for these posts for the dura- tion of the disbursement period, and an expatriate Planning and Evaluation Manager would be appointed with executive responsibility for two years. - A.14 - F. Accounting 4.10 Accounts were kept satisfactorily on a cash basis and were reviewed by a state inspector who visited the project but whose reports were found in- adequate. Under the second project provision has been made for full auditing of the accounts by an independent firm. G. Reporting 4.11 PMU submitted detailed, comprehensive progress reports twice a year. While there was no formal monitoring system built into the project, close monitoring was assured by PMU as part of project activities. This monitoring of project activities, trends and problems, made it possible to detect weak- nesses (upland developments, water control works and road building) and grasp the potential of grey soil rice, shifting priorities accordingly. The agron- omist section made a special effort to measure yields and production. PMU kept accurate records on the use of season inputs and implements. A notable deficiency has been the lack of knowledge of farm structures (sizes, cropping patterns, budgets) which made it impossible to measure the social impact of the project. A full scale monitoring and evaluation program is provided for in the follow-up project, which involves diversification into upland crops and animal production, and will evaluate whole farms rather than individual growers and fields. V. AGRICULTURAL AND SOCIAL IMPACT A. Incremental Output (Annex 1) Overall Output 5.01 While the upland crop outputs are much lower than envisaged, the overall output in value resulting from the project was well above estimates, as the result of the success of the development of paddy production. This is summarized by the incremental figures below: Total thru PY4 Total thru PYS Actual (A) Appraisal Estimate (E) Differencf t CFAF million./1 t CFAF million Ll 2 (A-E)/E Paddy .43,470 1,804 28,870 702 Foodcrops (13,408) (425) (1,150) (19) .Groundnuts .( 9,385) (389) 6,800 138 Gross Value 990 821 +21 Input Costs 233 306 -24 Net Value 757 515 +47 /1 At farmgate prices. /2 Estimated prices of products adjusted for inflation as foreseen at appraisal. - A.15 - The overall incremental increase in net value was 47% above PY5 estimates and resulted from higher paddy production and farmgate prices and lower input costs. The incremental rice tonnage -- 51% - is substantially accounted for by the high productivity of grey soil rice. Yields of this crop reached over 3 t/ha, in contrast to the appraisal estimate of 1.6 t/ha from in the main upland rice Annex 1, para 4 shows that the participants produced in PY4 almost as many calories (7,300) as foreseen at appraisal on half the area. Groundnut production is expected to substantially exceed appraisal estimates by the end of the next campaign. Farm Income 5.02 Typical farm budgets have been prepared and compared with appraisal estimates, using 1971 and 1975 prices (Annex 2). Appraisal Actual Estimate 1971 prices 1975 prices ------------- CFAF '000 ------------- 1. Project Year 0 (1971) Gross revenue 52 Cost /1 24 Net revenue 28 - - 2. Project Year 4 (1975) (a) without ox-traction gross revenue 148 139 292 cost /1 38 38 73 net revenue 110 101 219 (b) with oxen gross revenue 217 245 530 cost /1 65 62 121 net revenue 152 183 409 /1 Including family consumption. The budgets show that actual farmers' revenue and costs are substantially in line with appraisal estimates when measured at constant 1971 prices and much higher using 1975'prices. Reliability of Output Data 5.03 In the course of the development period, the question was raised by various missions (supervision and preparation of second project) as to the reliability of the high yield and production figures reported, especially in PY1 and PY3. As the project marketed only a small part of the total output - A.16 - production had to be estimated by yield sampling and estimation of the areas cropped by project farmers. To do this the expatriate agronomist designed an elaborate system of yield sampling, involving large numbers of samples and comparison with yield checks made in the same fields by ISRA, the research institute. Field extension workers had responsibility for measuring and staking the areas devoted to each crop by sample project farmers and area measurements were checked by surveyors of the Rural Engi- neering Division. Although rice yields reported by the project are in line with those reported by other projects in Casamance (e.g. the SODEFITEX project in Velingara in Upper Casamance) yields of irrigated rice appear to be extremely high at 4.4 t/ha in 1972, 3.5 t/ha in 1973, 4.7 t/ha in 1974 but 6.4 t/ha with full water control. While this yield is not impos- sible, especially on small plots and with high fertilization, it is above the range obtained in similar situations in West Africa. In 1975 groundnut yields too were a good deal higher (2 t/ha) than the average yield in all of Senegal (1 t/ha) and in Sine Saloum (1.2 t/ha) in 1975. To check whether there was any bias in the method of measuring yields, PMU compared for 1975 yields from routine sampling with yields measured over whole fields, and found that they agreed within 5 to 10%. Thus, project output seems to have been assessed with acceptable reliability but the rate of return on the proj- ect has been tested nonetheless for a 10% reduction in benefits (Annex 3, para 1). B. Technological Impact (Annex 1, paras 2, 3; 4 and 5) 5.04 The project made a significant impact on agricultural technology in the area, and resulted in a marked improvement of productivity both of land and labor. 5.05 Development of grey soil rice. This is by far the most significant technological improvement. The excellent potential of grey soils for rice production was first suspected by IRAT the agricultural research institute, in 1966 and they reported in 1972 that the soils would produce high yields of rice. Grey soils are sandy colluvial, occupying the lower slopes just above the swampy valley bottoms. During the rainy season, the water table lies 20 to 100 cm below the surface, thus ensuring a steady water supply and an ideal environment for rice. With a simple package consisting of the high yielding 110 day variety (IKP-I King Pao) and generous fertilization (200 kg 8.18.27 and 150/200 kg area per ha) yields approximate those of irrigated rice with practically no development cost, as the land is clear of brush except for a light stand of volunteer oil palm and does not need any water control works. 5.06 During project preparation it was observed that farmers were growing rice on grey soils with good results, and 3 to 4.tons were observed during appraisal. The project therefore aimed at developing 1,500 ha of grey soil rice by PY11, as against 8,000 ha of upland rice. By PY4 however, farmers grew close to 3,000 ha of grey soil rice, with an average yield of 3 t/ha. Because of droughts, upland rice failed in 1970, 1972 and 1973, while grey soil rice made it possible to increase paddy production well above appraisal targets. - A.17 - 5.07 Had the project focused its activities on upland rice, it would have resulted in failure. Fortunately, PMU and supervision missions clearly per- ceived the potential of rice cultivation on grey soils and farmers were encouraged to switch rice production from the plateau to grey soils. Thus, the project can be credited for introducing a highly successful technique which combines minimal investment with high productivity. 5.08 Swamp rice improvement. Traditional swamp rice cultivation, with yield levels averaging 1.2 t/ha, was improved through the introduction of higher yielding varieties (made possible in part by the lowering of the water levels subsequent to the drought), early transplanting and fertilization with- out water control works. Yields reached 2.5 to 3 t/ha during the last two seasons. 5.09 Introduction of irrigated rice. Pumping schemes were introduced for growing irrigated rice during the dry season. This crop achieved average yields of 3.7 t in 1973 and 6.4 t/ha in 1974. 5.10 Introduction of upland rice. It was anticipated that upland rice would be introduced in the project area and would be grown on 9,000 ha at full development. This projection was based on the SATEC report, which fore- cast a yield of 2 t/ha, while IRAT, the research institute, estimated the yield potential to be 1.5 t/ha. Upland rice development failed to materialize because of: (a) the droughts of 1970 and 1972, and (b) the discovery of grey soils. At the time of project preparation, no short-cycle, drought-resistant variety was available, and the variety 1KP (110 days) did not perform well on the plateau. New drought resistant varieties are now available in the project area and were successfully tested at the farm level in 1975. This experience will enable the second project to develop a sound program of upland rice production. 5.11 Development of upland crops. Because of the deliberate stress on lowland rice by P11U, groundnuts did not develop as anticipated. However, due to the introduction of the new cultivar 69-101, together with increased fertilization and animal traction, yields and production exceeded estimates. Maize was introduced by the project and production made modest progress, through the introduction of hybrids (JDS and SMIO-BDS) and fertil- ization, which assured good yields except in 1975 due to unfavorable weather. On the other hand, millet production technology did not progress, mainly because the improved variety, "Sanio de Sefa", performed no better than the traditional varieties. 5.12 Animal production. With the exception of ox drawn development, there was no dramatic improvement in cattle production. Nevertheless, given good grazing resources, the low density of population , good animal health - A.18 - good grazing resources, the low density of population , good animal health conditions and the growing acceptance of animal traction among farmers, there is significant potential for livestock development in the project area. The second Sedhiou project is going to promote better herd management through its extension network. 5.13 Mechanization. Project farmers acquired a total of 1,700 pairs of oxen, 1,500 plows, 1,400 seeders and 500 hoes. The project also purchased on a trial basis 100 pedal threshers that will be rented to farmers. Animal traction is largely limited to upland rainfed crops and has enabled farmers to improve their labor productivity, and cropped areas as illustrated by the farm budgets (para 5.02). The impact of mechanization on yields, while likely to be positive by allowing for more timely operations, has not been measured. 5.14 Fertilizer Use. Fertilizer use has increased markedly under the project. Project farmers have accurately followed fertilizer recommendations, to the point that the proportion of fields receiving fertilizers ranged from about 60% for millet to 70%,for groundnuts and 100% for maize and rice. How- ever, fertilizer use on rice may have been excessive at up to 100 kg of nitrogen per ha and only profitable to farmers because of the high level of subsidy. Fertilizer use can be summarized as follows: Total Fertilized Total Rock Phosphate Total Fertilizers area thru PY4 applied thru PY4 applied-thry FY4 (ha) (t) (t) 100% Paddy 14,600 NA NA 20% Foodcrops 2,800 NA NA 70% Groundnuts 5,900 NA NA 23,300 4,500 6,100 Average kg/ha applied 190 260 5.15 Seed production. With support from the seed service of MDRH and ISRA, the project developed a sound seed production program through contracts with growers. This program enabled the project to meet the requirements for rice and maize. In 1975 the project produced 18 t of improved seeds. How- ever, the seed service was up to now unable to pay the last part of the premium owed to multipliers and this could jeopardize future seed production. C. Prices and Inflation (Annex 1, paras 6 and-7) 5.16 While the project costs were significantly affected by inflation (para 4.27) farmers were protected by subsidies from increases in the prices during the disbursement period. All fertilizer was sold at CFAF 12 per kilo; rock phosphate was given free. In 1975 fertilizer prices were raised to CFAF 16 per kilo. Meanwhile, the price of seed followed the official price of crops, which increased as follows (CFAF/kg): - A.19 - 1971 1972 1973 1974 1975 Groundnuts 17.5 23.1 29.5 35 41.5 Millet 14 25 25 30 30 Maize 14 25 25 30 35 Paddy 21 25 25 41.5 41.5 D. Number and Selection of Participants (Annex 1, para 1) 5.17 As shown below, the number of project participants well exceeds appraisal estimates: 1972 1973 1974 1975 1976 /1 Number of participants: actual 3,900 12,200 24,200 23,600 34,900 appraisal estimate 4,000 8,000 12,000 16,000 20,000 L Under Sedhiou II Project. It should be noted that a large number of the participants were women (57% in 1974 and 40% in 1975). This'success stems from the attractiveness of grey soil rice, and from the influx of refugees from Guinea Bissau,-who are traditional rice growers. The subsequent departure or cessation of refugee arrivals accounts for the levelling off of the number of project participants between 1974 and 1975. 5.18 Project management applied the following criteria in the selection of participants: (i) full-time farming with good results, (ii) availability of rice land, (iii) membership in a Project cooperative, (iv) good credit record, and (v) willingness to follow the technical directives of project staff. E. Secondary Goals and Unexpected Effects 5.19 Apart from the major unexpected effect of the project in the conver- sion of grey soils into areas of high agricutural pay-off, another unexpected effect has been the.response of women to extension work, their readiness to procure inputs on credit ("credit special"), and the enhancement of their prestige resulting from their ability to feed their families and improve their cash situation. Moreover,-the success of grey soil rice has induced the men of the area, who traditionally have shied away from lowland food production, to take an interest in rice growing, and thus to work with the women of the project. 5.20 The benefits of the project have extended to the refugees from Guinea Bissau temporarily .settled in the south of the project area until 1975. According to some technicians, these refugees, who would have been used to cultivating grey soil in their country, were the first promoters - A.20 - of the.valuable grey soil rice production in Casamance. After their depart- ure from the project area, there is still evidence that a segment of the population of Guinea Bissau living near the Sedhiou district relies on the project area to procure food, which is still scarce in Guinea Bissau. 5.21 Among the various projects operating in Casamance, the Sedhiou project proved to be the most successful and to have the greatest impact in its area. The regional development of Casamance now in preparation will follow the pattern set by the project. VI. ECONOMIC RATE OF RETURN 6.01 The economic rate of return for the project (Annex 3) has been calculated on the basis of the data in Annex 1 and on the following assump- tions: (a) the project was assumed to have an economic life of 15 years, 1972-87, the average expected life of the main capital investments; (b) actual exchange rates have been used for 1972-75; that for 1976 (which increased from CFAF 225 to CFAF 245 at year-end) has been averaged at CFAF 230; Bank's estimated rate of CFAF 245 has been used there- after; (c) project costs, net of estimated taxes are actual amounts until 1976; beginning 1977, i.e. after the development period, expenditures have been limited to staff necessary to supply extension services and are expressed in 1977 terms; (d) incremental farm inputs have been priced at estimated full cost; (e) benefits are calculated on the net incremental crop production. The production without the project cor- responds to the actual participants' production cal- culated in terms of area cultivated and yields without the project as anticipated at appraisal; and (f) economic farmgate prices are based on IBRD world market prices, actual and forecasts (see Annex 3, Tables 2 to 5, for details). 6.02 Based on these assumptions, the economic rate of return is 22.7% as against 18.5% estimated at appraisal. With a 10% reduction in benefit, to take into account an overstatement of yields and production (see para 5.03), the return would still be 17.5%. - A.21 - 6.03 While the higher return reflects the unquestioned success of the project with rice production from the very beginning, and also groundnut production which is now going well, it is also because the bulk of the in- frastructure envisaged was not constructed. Moreover, the investments which were made were delayed up to PY4 and PY5. VII. SPECIAL ISSUES A. Weather and Risk 7.01 .The rainfall pattern followed closely or exceeded the expected pattern for three out of four of the project years (attached graph). In 1972, however, a relatively severe drought occurred (750 mm as against an average of 1,150 mm). The drought severely damaged upland crops and rice production failed on about 20% of the fields which proved to be a blessing in disguise as it was a major factor in reorienting project activities towards grey soil rice at the expense of upland rice, thus increasing production and minimizing the-risks from drought. B. Sector Influence 7.02 The project was designed to promote Government's goal of rice import substitution. Before the project Casamance imported some 9,000 t of rice and the Sedhiou District 1,500 t annually. Rice importation in the District decreased to about 1,200 t in 1972/73 and to 600 t in 1975/76. The trend away from marketing through Government channels for urban consumption was encouraged by Government's low official prices for milled rice. Rice was substituted for millet as staple food among farm families, and to some extent paddy was stored at the farm as food insurance. In its dialogue with Govern- ment, the Bank had been pointing out the dangers of Government's policy of subsidizing urban rice consumption, with the resulting disincentive for expansion of rice production by the rural sector for the urban market. In 1974 Government eliminated the subsidy for imported rice, which had caused serious financial problems, while discouraging rice production. In 1975 the official retail price of white rice was CFAF 100/kg and the official farmgate price of paddy, CFAF 41.5/kg. C. Integrated Development 7.03 No attempt was made to design the project with a view to promoting the integrated development of the area. PMU was given responsibility for a district of the Casamance region, parallel with projects in other districts. The project was intended to focus on rice, emphasizing rice at the exclu- sion of groundnuts and millet, to an even greater extent than envisaged at appraisal. There was a deliberate effort to supervise individual fields rather than whole farm operations which proved in a first stage to be suc- cessful. - A.22 - 7.04 The Casamance region has been characterized by a mosaic of projects, with sometimes conflicting strategies and policies. The first project should be considered as a necessary first phase leading to more integrated agricul- tural development, which is being planned out under the follow-up project. D. Replication 7.05 Because of its high pay-off, grey soil rice development fully jus- tified replication and other projects also cashed in on the potential of these soils. It is expected that all the available grey soils of Casamance will soon be devoted to rice. At the same time, the project set the stage for the development of upland crops, irrigation and livestock, as the result of pilot activities-in these areas, together with new research developments. E. Settlement 7.06 In 1974 the project started to provide extension services and credit to 150 farmer settlers of the abandoned, mechanized parastatal farm of SEFA, and to promote animal traction instead of tractors to grow groundnuts and millet. Though initial yields were rather low, they have been high enough to induce settlers to stay in the area--a significant change from the previous experiences. The on-going settlement of farmers in the area under Sedhiou I is going to be encouraged and intensified under Sedhiou Il. VIII. CHANGES IN THE SECOND PROJECT 8.01 The second project, while moving towards diversification and inte- gration of agricultural development, reflects the experiences of the first project by: (a) completing the development of grey soil rice production; .(b) applying the information gained during the first phase to start new water control and rice-mill programs; (c) combining the experience gained on upland crops with recent research data to improve upland crop production; (d) promoting improved livestock husbandry; (e) settling about 400 farms in SEFA area; (f) including a monitoring and evaluation unit; - A.23 - (g) including the preparation of a master agri- cultural plan for the Casamance region in the follow-up project, in recognition of the lack of coordination among agricultural development projects in the region; and (h) reorganizing the relationship between ONCAD and PMU. These steps will bring about an improvement of the situation but not with respect to: (a) the timely delivery of inputs to PMU by ONCAD,and (b) the overall inefficiency and necessary financial studies of ONCAD. It was agreed therefore that Government would commission a study of the feasibility of decentralizing the procurement of farm implements and inputs. IX. GOVERNMENT AND BANK PERFORMANCE 9.01 An outstanding feature of the Casamance Rice project was the flexibility in project implementation, whereby a major shift in emphasis was made as early as FY1. The Bank was able to support this shift--because of close supervision and adequate information provided by the project's monitoring system. This confirms the need for close monitoring and super- vision during implementation and flexibility in the design of agricultural development projects. 9.02 The main weakness in the Bank performance was the lack of depth of project preparation. This led to an underestimation of costs at apprai- sal, especially for infrastructure, and an over estimation of the upland rice potential, although the appraisal mission did recognize the weakness of the data base. 9.03 The identification mission by RMWA suggested that the project be carefully prepared by a team of five which would need to spend 4-6 weeks in the field. A subsequent FAO/CP identification mission suggested a 5 man-month study that was to be carried out by a consulting firm before completion of project preparation. This study however did not materialize and further preparation was limited to a two man mission by FA0/CP which spent one month in the country, of which 8 days were spent in the field, and submitted a very brief back-to-office and final report. This report was the basis for the appraisal mission, and the latter found it possible to finalize its field work because of the help and information it received from the then operational SATEC project in the Kolda-Velingara area (next to Sedhiou District). - A.24 - 9.04 Several issues were efficiently tackled during the preparation: location of the project, project authority, rice prices. Another issue was the suitability of the project for upland rice development, which was doubtful for several reasons: (i) sociological obstacles, (ii) competition with groundnuts, and (iii) over-optimistic yield forecasts. 9.05 The appraisal mission raised the issue of project organization, the need for further study of the upland rice potential and the hydrology of the valleys, the difficulties attendant to land clearing and large scale organization. 9.06 Other developments were not adequately foreseen: the high road building costs, the non-implementation of the rice mill program, and the channeling of the rice output through the traditional rather than official marketing, its use for family consumption and the restocking of traditional paddy storage depleted by the recent droughts. 9.07 Supervision took place twice a year, as envisaged. The main prob- lem with supervision was the lack of continuity of the IDA personnel result- ing from staff turnover. Yet the missions addressed the problems in a pro- ductive manner; arrangements for the effective channeling of project funds, monitoring the switch of emphasis to grey soil rice, and the switch to road building by force account, postponement of the rice mill component and of irrigation works. Supervision missions did try to tackle, but without success, the rice marketing problem. BENEGAL CASAHANCE RICE PROJECT ((REDIT 252 SE) Completion Report Yearly Project Progress 1972i1973 1973-1974 - 1974-1925 1975-16 197-19771 A EA EA E A T l 1. Number of Participants Families without draft oxen NA 500 NA 1,000 NA 1,500 2,500 2,000 2,500 FamIlies with draft oxen 100 500 400 1,000 - 1000 1m0 22000 2 Total Families HA 1,000 NA 2,000 NA 3,000 4,000 4,000 100 5.000 Actives 2/ NA 4,000 NA 8,000 NA 12,000 NA 16,000 20,000 Participante / 3,895 4,000 97 12,169 8,000 152 24,188 12,000 202 23.577 16,000 147 Men 2,308 4,902 10,456 1o,244 Women 1,587 7,267 13,732 13,333 2. Crop Development (ha) Swamp paddy / 190 300 63 487 750 65 2,026 1,135 179 2,415 .1.495 162 1.860 Grey soil paddy 508 200 254 1,692 460 368 3,500 785 446 2.979 1,120 266 1,440 Upland paddy 111 400 28 296 1.220 24 181 2.680 1 2 4.740 6 Sub-total paddy 809 900 90 2.375 2.430 98 5,707 4,600 124 5,684 7,355 77 9,990 Foodcropa (maize, millet) 25 1,450 2 738 2.450 30 1,485 3,275 45 1,741 3,825 46 4,680 Groundnute 20 . 1 1,203 3,180 _ 8 _A> .11. 6.490 _2,9 Total 854 3,850 22 4,316 8,060 54 10,230 12,675 81 11.542 17,670 65 22,640 1 No hard data avilable for actual performance. Beat eatisate shown. This figure concerna all active people including young and old 3/ This figure concerne only the registered participants and thereby excludes most of the young and the old even though they may be economically active. Second paddy cropping not included because negligible. lA -197l 19 14 A 1974-1975 I -1976 Total thru PT4 1976-19n A E ( A i g IBA s y A . ). Fare Outpute 3.01 Average Yield (t/ha) Paddy 1.6 1.1 145 3.9 1.2 325 3.7 1.3 285 3.2 1.4 229 1.5 Foodcropa (maize, millet) 1.5 0.8 188 2.1 0.9 233 1.7 1.0 170 1.5 1.2 125 1.2 Groundnuta (unahelled) 2.1 1.1 141 2.3 1.15 200 1.9 1.2 158 1.7 1.3 131 1.4 3.02 Production (t). Paddy Total / 1,317 1,000 132 9.253 2.950 314 21,373 6,000 356 18,200 10,200 178 50,143 20,150 249 15,000 .Without project / 400 400 1,220 800 2,575 1,240 2,478 1.680 6.673 4,120 2,160 Incremental 917 600 8,033 2,150 18,798 4,760 15,722 8,520 43,470 16,030 287 12,84o Foodcrops Total 37 1.150 3 1,547 2,250 69 2,561 3,400 75 2,611 4.500 58 6.756 11.300 60 5,800 Without project 1,200 1,200 3,660 2,400 7,781 3,750 7,523 5,100 20,164 12.450 6,600 Incremental (1,163) (50) (2,113) (150) (5,220) (350) (4,912) (600) (13,408) (1,150) (495) (800) Groundnuta Total 42 1,650 3 2,827 3,650 72 5,783 5.900 98 6,897 8.300 83 15.549 19.500 73 10,850 Without project 1,500 1,500 4.575 3,000 9,566 4,650 9.293 6.300 24,934 15,450 8,100 Incremental (1,458) 150 (1,748) 650 (3,783) 1.250 (2,396) 2,000 (9,385) 4,050 24 2.750 i/ Second paddy cropping not included because negligible y Cultivated areae/without project . Sedhiou I appraisal estimates for active x number of registered participanta Yielda/wJ thout project f (Sedhiou I apprainal eatimatea) . -Paddy Py1 1.00 P12 1.00 PT3 1.03 P14 1.03 PY5 - 1.08 t -Foodcrope 0.80 0.80 0.83 0.85 0.88 -Groundnuta 1.00 1.00 1.03 1.08 1.05 1972-197 1973-1974 1974-19" 19?3-1976 .A I A E A E A E 4. Activities per participant Number of participants 3.895 4,000 12,169 8,000 24,188 12,000 23.577 16,000 4.01 Area cultivated per participant (ha) Paddy 0.208 0.225 0.195 0.304 0.236 0.383 0.241 0.460 Foodcropa 0.006 0.363 0.060 0.306 0.o61 0.273 0.0'A 0.239 oroundnute 0 0.375 0.100 0.398 0.126 0.400 0.i a.leo6 Total 0.219 0.963 0.355 1.008 0.423 1.056 0.49 1.105 4.02 Production per participant (t) /2 /2 Paddy 0.338 0.250 0.760 0.369 0.884 0.500 0.772 0.638 - Foodcrope 0.009 0.288 0.127 0.281 o.i06 0.283 0.111 0.281 / Groundnuts 0.011 0.413 0.232 0.456 0.239 0.692 0.293 0.519 /1 Labor/Year/Participant (average 200 mandays/ha) Actual: about 100 mandays Eatisates: about 220 mandaya /2 Calories/D/Purticipant (Average 3,000 calories/kg) - 1975-76 - Paddy and Foodcrops Actual: about ?,300 (4200 calories/person) Estimates: about 7,500 (4300 calories/person). 1972- 3 197-1974 1974-1975 197-196 Total 5. Farm input & Field Equipment 5.01 Seasonal Inputs Improved seeds (t) Paddy 61 148 274 236 719 Foodcrope 1 3 7 18 Groundnuta x x x x x Rock phosphate (t) 515 651 1,907 1.473 4,546 Fertilizers (t) 2] 331 1,051 2,355 2.339 6,076 Pesticides (t) 3/ 5.7 5.3 10.7 20.2 41.9 " (1000 0.3 0.1 0.6 0.8 1.8 5.02 Small equipment Local hoe - - 602 30 632 Sickle 2,507 12 2,519 Provided by ONCAD 2/ UREA; 8-18-27; 8-14-181 10-21-21 / Il21; HEPrACHLilRE; LINDANE; CARBARYLI CAPTAGRANOX / TIMUL 35; HINOSAN; ZITHIOL l72-973 1973-1974 1974-1975 i975-1976 Total A E A E % A E,AE 5 .03 Draft oxen & Implements Pair of oxen on credit 21 250 294 250 453 250 398 250 1,165 1,000 17 Pair of oxen supplied from the local herd 81 160 166 250 67 250 1. ? Total Dair of oxen 101 500 454 500 619 500 565 500 1.739 2,000 87 UCP plow 68 500 344 500 537 500 515 500 1,464 2,000 73 Ridging plow - - - - - - 1 - 1 - EBRA plow 2 - 3 - - - - - 5 - Super ECO Seeder 2 500 101 500 553 500 699 500 1,355 2,000 68 Hand seeder - - - - - - 37 - 37 - - SINE hoo 48 - 56 - 177 - 159 - 440 - Western hae - - 1 - - - - - 1 ARIANA Equipment - - - - 1 - - - 1 - ARARA frame - - - - 1 - 34 - 35 - Ox cart 1 - 8 - 155 - 210 125 374 125 504 Donhey cart - - 24 - 64 - 168 - 256 - 5 .09 Field Equipment Metal drum - - - - 53 - 436 - 489 - Paddy pedal thresher - - - - 100 - - - 100 - Hand aprayer 100 200 48 350 - 450 - A0o 148 1,300 Power Sprayer 4 - - - - - - - - Winche - 60 - 12 - 28 - - - 100 Spare cable - 180 - 36 - 84 - - 300 1972-19 19Z1974 1974-1975 195-1976 1L9-197 A ki A k s A k s A B 6. Price of farm outpute (CFAFAIg) 6.01 Average annual production price / Paddy 25.0 21.6 25.0 22.3 41.5 22.9 41.5 23.6 24. Foodcropa - 14.4 - 14.8 - 15.3 - 15.8 16.! mkize 25.0 - 25.0 - 30.0 - 35.0 - illet 25.0 - 25.0 - 30.0 - 30.0 - - Groundnute 23.1 18.0 29.5 18.6 35.0 19.1 41.5 19.7 20.3 6.02 Average annual economic prices / Paddy 18.2 17.0 34.4 16.8 56.0 16.6 33.9 16.4 j6.2 Foodcropa 23.5 14.0 25.9 14.0 42.4 14.0 38.1 14.0 14.5 Maize 24.1 - 26.5 - 44.1 - 39.2 - - Millet 23.2 - 25.6 - 41.5 - 37.6 - - Oroundnute 41.6 18.3 58.2 18.3 98.6 18.3 62.1 18.2 18.2 1 Anotraian1 nrice estimates (Paddy - 21.0; Foodcrone = 14.0; Groundaute - 17.5) adjusted for inflation by the following coefficients as foreseen at appraisal; PY1: 1.03; PY2: 1.0609; P13: 1,0927; PY4: 1,1255; I'Y5: 1,1592 2J Actual prices see Annex 3 tables 2, 3. 4 & 5. 7. Price of farm inputs & Field equipment 7 .01 Seasonal inputs (CFAF/kx) Improved seeds Paddy 31.3 31.3 51.9 51.9 Maize 31.3 31.3 43.8 43.9 Millet 31.3 31.3 37.5 37.5 Groundnute 28.9 36.9 43.8 43.8 Rock phosphate 21 0 0 0 0 Fertilizer (CFA"Ag) 1/ 12 12 12 16 Pesticides 2/ 0 0 0 0 7 .02 Small equipment (CFAF) 4/ Local hoe 800 800 Sickle 317 317 1/ Actual coots included interest of credit (2 ) 2/ Fully subsidzed 1/ Partly subsidized 4/ Actual costs included interest of credit 1972-1973. 1973-1974 *1974-1975 19751276 A E A E $A EA 7.03 Draft oxen & implementa (CPAF) Pair of oxen 33.655 20,600 33,655 21,380 33.655 21,854 60.511 22,510 UCY plow 12,424 12,424 12,424 13,071 Ridging plow 3 - 613 EBRA plow 24,730 24,730 - Super EO seeder 14,297 14,420 14.297 14,866 14,297 15,298 15.042 15.757 Hand seeder . - 11,269 SINE hoe 7.997 7*997 7.997 9*512 Western has - 7997 7,997 9,512 ARIANA equipment - - 46,075 49,140 ARARA frame 6,595 6,585 6.585 6,28 ARARA plow 7,440 ) 25.750 7,440 ) 26,725 7.440 ) 25,316 7.827 28,138 Canadian equipment - ) - ) ) 7.521 ) Ox cart 32,480 32.342 32,480 39,980 32,480 40,670 34,12 42,094 Donkey cart 29,092 29,092 29,092 30,608 7 .04 Field equipment (CFAF) etal drums 3,660 3,952 Paddy pedal thresher - - NA Hand sprayer 0 0 Power aprayer 2/ 0 - Winch - i80,40o - 192,242 - 196,686 - 202,590 Spare cable - 27,810 - 28,863 - 29,503 - 30,389 I/ Anoraisal coet eatimates adjusted for inflation by the following coefficients as foreseen at appraisal: 171 = 1.03; PY2 . 1.0609; PY 3 - 1.0927; PY4 - 1.1255 Actunl costs included interest of credit. / Fully eubsidized. 19?2 19 q74 Total A E Å EZ3. % E A E % A E Total coat of farm inpute (CFAF million) Seasonal inpute & vmall eguipuent Improved aoeda 1.9 6.0 14.4 12.5 34.8 Rock phoophate } - - F.rtiliz*ra 4.0 12.7 28.6 37.4 82.7 Peeticidea - Small equipment 0.1 - 1.0 0.1 1.2 Sub-total (cumulative) 6.o 15.6 18.7 34.8 44.0 58.4 50.0 85.8 118.7 194.6 (incremental) 6.0 15.6 12.7 18.6 25.3 22.5 6.0 25.7 50.0 82.4 Draft ozen o.8 5.2 9.9 5.3 15.1 5.5 26.9 5.6 52.7 21.6 Ift.1emente 1.3 20.1 ^7.0 20.7 24.5 21.3 29.1 27.2 61.9 89.3 Total ith eeaaonal inputa at S,teidited pricen (cumulative) 8.1 4.09 35.6 60.8 83.6 85.2 106.0 118.6 233.3 305.5 (incresental) 8.1 4.09 29.6 44.6 64.9 49.3 62.0 58.5 164.6 193.3 lertiliter Subaidy element (cuaulative) 6.0 13.6 19.1 32.7 42.9 53.7 56.1 77.3 124.1 177.3 (incremental) 6.o 13.6 13.1 18.7 23.9 20.0 13.2 22.0 56.2 74.3 Tota) with neaeonal Inputa at r.--r.-ena1:i Zed pricea (cu^ulatlve) 14.1 54.5 54.7 93.5 126.5 138.9 162.1 195.5 357.4 482.8 (1rcremental) 14.1 54.5 42.7 63.3 88.8 69.3 75.2 80.5 220.8 267.6 J r'..n W 't ont r.ten a.junted for inflation by tite following coefficienta aa foreoaen at appraeisal: ' - J///), P ', a 1./)27; PY4 1.1255 * 'rredit. * ~ . </~ pe.1 fo.r ,,c ,jtti<ta. 9. Farm Credit (CAf million) (at subsidized prices for seasonal imports) Medium tore credit 2.1 25.2 16.9 26.0 39.6 26.7 56.0 32.9 Short tere credit cumulative 6.0 15.6 18.7 34.8 44.0 58.4 50.0 85.8 Total credit 8.1 40.8 35.6 60.8 83.6 85.1 106.0 118.7 Farmera' repayment To be repaid by farmere Outstanding 0.6 - x - 3.2 - NA - Medium term 0.6 - 4.0 5.2 11.6 10.7 NA 16.5 Short term (cumulative) .0 15.6 18.8 pA.8 44.0 2E 4 NA 85.8 7.2 15.6 22.8 40.0 58.8 69.1 80.3 102.3 Repaid by farmere Outstanding 0 0 x NA Medium term 0.6 3.5 10.5 NA Short term lCumulative) 2.4 15.9 NA 3.0 19.4 50.2 68.2 % Repayment 41 86 85 85 1/ Appraiail estimntes adjusted for inflation by the following coefficients as foreseen at appraiaal: PY 1 - 1.03; PY2 = 1.0609; PY3 = 1.0927; PY4 - 1.1255 10. Droject ~>t Brenkdown (CFAP millior.) 0t.egory I Equirmnt 7.2 20.9 87.8 28.6 3-4 147.9 58.8 514 åtegory II Infraetructure 0.1 3.0 50.9 63.7 5.1 122.8 334.? 37 Category 111 Staffing 69.7 96.3 134.9 199.6 111.4 611.9 43?.9 140 Category IV Operating conta 24.3 34.1 49.5 60.2 61.6 229.7 55.2 416 Category V Contingencie .. - 163.3 Unallocated Total 101.3 154.3 323.1 352.1 181.5 1.112.3 1P49.9 io6 Cumulative 101.3 255.6 578.7 930.8 1,112.3 % Total projectcoet oatiate 10 24 55 89 106 1/ Eatimaten L亡屹．Xllli。兀 華 1972 19711-1974 19,4il975 197 -1976 A PLT A E 9 A 12. Project Director Agronomist x Agronomist Assistant- 2 2 1 2 1 2 Agriculturol technical engineer (ITA) - - 2 - 2 - 2 Livestock technical engineer (ITE) - - Technical agent (Zone Chief) 6 6 12 9 12 12 12 12 Field agent (Extension worker) 64 60 120 72 147 100 147 100 Training Officer x I 1 1 1 1 1 1 1 Assistant training Officer I I - Rural Engineer x I 1 1 Assistant rural engineer 2 2 2 2 2 2 2 2 Topographer 1 1 1 2 1 2 1 2 Financial Analyst x I I I I I I I I Accountant I I I 1 1 1 1 1 Agro Economist - - - - - 1 - I Cooperative inspector -1 1 1 1 1 1 1 1 Cooperative marketing officer 2 2 2 1 2 1 2 Cooperative management officer 2 3 2 3 2 3 2 3 Cooperative supervisor 5 5 11 7 12 10 12 10 Planning Officer - I I I I I I I Administrative officer - 1 1 1 1 1 1 1 Pico mill mechanics - 2 - 6 - 11 - 17 Clarks. Stura keepers 2 2 26 4 26 6 26 6 92 95 190 119 217 161 217 166 x exontri-ste I/ Rural Works Engin-er CiTP) 21 Plus 20 man-montha per )ear and per rice mill &a Tort-time staff A 1972-1973 1973-19Z4 ' A '1974-19 A 1975-1'976 13. Activities parextension worker 13.01 Number of extension workere 64 60 120 72 147 100 147 100 147 Number of villages involved 193 NA 331 NA 484 "A 524 NA NA Number of participanta 3.895 4,000 12,169 8,000 24,188 12,000 23,577 16,000 147 Total area cultivated (ha) 854 3,850 4,316 8,060 10,230 12,075 11,542 12.690 91 Area cultivated in Paddy (be) 809 900 2,375 2,430 5,707 4,600 5,684 7.355 77 Paddy production (t) 1,317 1,000 9,253 2,950 21,373 6,000 18,2oo 10,200 178 Pair of draft oxen (cumulative) 101 500 555 1,000 1,174 1.500 1,749 2,000 87 13.02 Villages per Extension worker 3.0 NA 2.8 MA 3.3 NA 3.6 HA NA Participants per Extenaion worker 60.9 66.6 101.4 111.1 164.5 120 160.3 160 100 Total area cultivated per Extension worker (ha) 13.3 64.2 36.0 111.9 69.6 126.8 78.5 126.9 62 Area cultivated in Paddy per Extension worker (ha) 12.6 15.0 19.8 33.8 38.8 46.0 38.7 73.6 53 Paddy production per Extension Worker (t) 20.6 16.7 77.1 41.0 145.4 6o.o 123.8 102.0 121 Pair of oxen per Extension Worker 1.6 8.3 4.6 13.9 8.0 15.0 11.9 20.0 60 1972-1973 1975-1974 1974-1975 1975-1976 A E % A E 5 A E .E 14. Rice Mille & Storage Rice milla (unit - 2 - 4 - 5 6 Cumulative - 2 - 6 - 11 2 17 Storage (a2) - 370 - 740 - 925 - 1,110 Cumulative - .370 - 1,110 - 2,035 - 3,145 Number of village stores 11 - 51 - NA - NA - (Cumulative) 11 - 62 NA 67 15. Road Program (ka) Construction - - - 244 23 - 29 - (Cumulative) - - - 244 23 244 52 244 Maintenance - - - - NA 244 NA 244 16. Irrigation and Drainage Works (ha) Gravity NA 7.0 NA 5.0 NA 15.0 NA Pumping - NA NA NA -N Tota 10~ 100 M7 lZu (Cumulative) - 100 30.4 200 41.1 450 87.40 800 17. ;ereal Marketing (t) Paddy marketed by project 174.0 639.0 1,056.0 772.0 Maize marketed by project 4.7 9.3 23.4 41.7 Millet marketed by project 0.4 2.0 2.2 4.4 Rice Imports for :c'hiou District 1,530.0 1,598.0 1,171.0 75.0 1/ Purchase of 2 rice sills of which only one operating (Bounkiling) 2/ Of which 60.40 ha under pumping irrigation 27.00 he under gravity irrigation 19?2-1973 19pI-974 1974-1975 1975-1976 1mroact of the Project in the Sedhiou District ("Department" of Sedhiou) 18.1 Villages Number of villages involved in the project 193 331 484 524 Number of villages of the District 770 M 770 770 % Project/District 25 43 63 68 18.02 Participants Number of participants 3,895 12.169 24,188 23.577 Number of actives in the District 94,000 96.000 98,000 Loo,000 % Project/D iatrict 4 13 25 24 18.03 Total area cultivated Project 854 4,316 10,230 11,542 District 90,000 92,000 94,000 96,000 % Project/District 1 5 11 12 18.04 Paddy production Project 1,317 9.253 21,373 18,200 District 15,560 21,900 41.544 33,851 % Project/District 8 42 51 54 18.05 Draft oxen Pairs of DO provided by the Project 101 454 619 565 Pairs of DO provided by ONCAD in the District 129 114 258 426 % Project/District 78 398 240 133 I Rough estimate:- 1975 total population of the Sedhiou District was about 200,000 - annual population growth rate is eatimated at 2.2% SENEGAL CASAMANCE RICE PROJECT (CREDIT 2 5 2 - S E) -Completion Report FARM BUDGET Y 4 - no oxen 1971 prices (actual) 1975 prices (actual) Total Tota Ton CFAF/Kg Value CFAF/Kg Value ------------------CFAP------------------- Production Groundnut 1.75 ha @ 1.1 t/ha 1.925 17.5 33,680 41.5 79,880 Foodcrops 1/ 1.25 ha 0 0.9 t/ha 1.125 14,0 15,750 31,7 35,660 Swamp rice 0.5 ha @ 2.5 t/ha 1.250 21.0 26,250 41.5 51,875 Grey soil rice 1.0 ha @ 3.0 t/ha 3.000 21.0 63,000 41,5 124,500 4.50 138,680 291,925 Production Cost Seed Groundnut 100 kg x 1.75 ha 0.1750 17.0 2,975 51.9 2/ 9,080 Foodcrops 1/ 10 kg x 1.25 ha 0.0125 15.0 190 39.6 2/ 495 Paddy 60 kg x 1.50 ha 0.0900. 25.0 2,250 51.9 2/ 4,670 Fertilizer Groundnut 150 kg x 1.75 ha 0.2625 12.0 3,150 16 4,200 Foodcrops 1/ 200 kg x 1.25 ha 0.2500 12.0 3,000 16 4,000 Rice 400 kg x 1.50 ha 0.6000 12.0 7,200 16 9,600 18,765 32,045 Family Consumption 3/ 18,900 40,600 Net Revenue 101,025 ?19,280 1/ 2/3 Millet + 1/3 Maize 2/ In 1976 prices (Sedhiou II appraisal) 3/ (Sedhiou I appraisal) 0.2 t groundnut 0.8 t foodcrops 1/ 0.2 t paddy SENEGAL CASAMANCE RICE PROJECT (Credit 252-SE) Completion Report FARM BUDGET Y 4 - with oxen 1971 prices ( actual) 1975 prices (actual) Production Ton CFAF/kg Total CFAF/kg Total Value Value ------------------------- CFAF------------------------ Groundnut 2.7 ha at 1.9 ton/ha 5.130 17.5 89,775 41.5 212,895 Foodcrop 1/ 1.3 ha at 1.9 ton/ha 2.470 14.0 3h,580 30.7 78,300 Swamp rice 0.5 ha at 2.5 ton/ha 1.250 21.0 26,250 41.5 51,875 Grey soil rice 1.5 ha at 3.0 ton/ha 4.500 21.0 .9h,500 41.5 186,750 -97 245,105 529,820 Production Cost Seed Groundnut 100 kg x 2.7 ha 0.270 17.0 h,590 51,9 2/ 11,010 Foodcrop 10 kg x 1.3 ha 0.013 18.0 195 39.6 2/ 515 Paddy 60 kg x 2.0 ha 0.120 25.0 3,000 51.9 ?J 6,230 Fertilizer Groundnut 150 kg x 2.7 ha 0.405 12.0 4,860 16 6,480 Foodcrop 1/ 200 kg x 1.3 h 0.260 12.0 3,120 16 4,160 Rice 400 kg x 2.0 ha 0.800 12.0 9,600 16. 12.800 25,365 44,195 Debt Service (equipment, pair of oxen) + OM 2/ 17,790 35,900 Family Consumption ./ 18,900 40,600 Net Revenue 183,050 409,125 C 1/ 2/3 Millet + 1/3 Maize. 2/ In 1976 prices ( Sedhiou II Appraisal). 3/ (Sedhiou I Appraisal): 0.2 t groundnut 0.8 t foodcrop 1/ 0.2 t paddy SDNXAL CASAHANCE RICS PROJECT (CtEDIT 252 SE) Completion Report Economic Rate of Return 1972 r1974 i 16 i 1M78 lg 18o-1986 Incremental Production (t) Paddy 917 8,033 18,798 15.722 11,868 10,000 10,000 10,000 10,000 Foodcrope (1,163) (2,113) (5,220) (4,912) (5,570) (5.000) (5.000) (5.000) (5.000) Oroundnuto (1,458) (1.748) (3.783) (2,396) 3.255 3,000 3,000 3,000 3,000 Prices (CFAF/t) Paddy 18 34 56 34 28 34 36 t0 44 Foodcrope 24 26 43 38 41 46 46 47 47 Groundnuts 42 58 62 63 67 59 60 60 Value of incremental production (CFAF million) Paddy 17 273 1,052 535 332 340 360 00 440 Foodcropa (28) (55) (224) (187) (229) (230) (230) (235) (235) Oroundnuta (61) (101) Jm) (19) 0 M.7M 180 180 Total value of incremental production (72) 117 453 199 308 311 307 345 385 Coset (CFAF million) Project coata 5 101 154 323 352 182 125 125 125 125 Estimated taxee 2020 10 10 10 10 Project coets net of taxes 86 134 278 302 162 115 115 115 115 Fare input coats 14 55 127 162 80 50 50 50 50 Total coats not of taxes 100 189 405 464 242 165 165 165 165 Net Benefite (CFAF million) (172) (72) 48 (268) 66 146 142 180 220 / ee Anrex 1 rnre ?/' 'ee Ai.nex w,age 7/ See Annex l page 5 1/ 3ee Annex 1 nni,e 8 - aR 23.75% Without any re4uction in benefit - ERR = 17.55% with a 10% redction in benefit BENIAL CASAMANCE RICE PROJECT (CREDIT 252 SE) Completion Report Price Structure for Paddy 1994 zz n41976 1977 1978 918 18a Reference Price Y (S/E) 147 350 542 363 254 300 319 353 391 3901 Discount for lower quality (40%) 88 210 325 217 152 180 191 211 235 234 Freight 20 21 22 23 25 27 27 97 27 27 Insurance (2.5 of FOB) 2 8 4 4 6 6 Value CIF Dakar ($) 110 236 355 245 181 212 222 243 268 267 Exchange rate 252 223 241 210 230 245 245 245 245 245 Value CIF Dakar (CFAF) 27.720 52.628 85.555 51.450 41,63o 51,940 54.390 59,535. 65.660 65.415 Port, Custom, Handling 2,500 2,900 3,600 4.1wo 4,400 4,800 4.900 4,900 5,000 5,000 Transport Cost to Project Area 1,900 2,200 2,700 3.100 3,300 3.500 3,600 3,700 4.000 4,ooo (384 km) / Value of RIce, ex mill 32,120 57,728 91,855 58,650 49.330 60,240 62,890 68,135 74,660 74,415 Value of Paddy equivalent 20,878 37,523 59.705 38,122 32,065 39,156 40,878 44,288 48,529 48,370 Milling Costa 2,300 2,700 3.300 3,800 4.000 4,300 4,300 4,300 4.300 4.300 Buying and Collecting - 400 400 400 400 400 400 400 400 400 400 Economic Faragate Price of Paddy 18,178 >4,423 56,005 33.922 27,665 34,456 36,178 19,588 43,829 43,670 (CFAF/t) 1/ Current terse up to 1976; Constant 1977 prices thereafter 2/ FOB Bangkok, 5% broken. Source: Economic Analysis and Projection Department, Commodity Trade and Price Trend (1976), and Commodity Price Forecasts, November 1976. 3/ Eatimate based on compounding fundinga of 1971 appraisal miesion by Bank international inflation index. E/ 8 CFAF/ka in 1975 (See Appraisal Report of Second Sedhiou Rice Project, No. 1094-SE, of Jun. 4, 1976, Annex 20, Table 3). Adjusted for inflition before this year and until 1977. Milling yield: 65% / CFAF 3,800/t in 1975 (See Appraisal Report mentioned in note 5 above, same table). Adjusted for inflation before this year and until 1977. SENWXAL CASAMANCE RICE PROJECT (CREDIT 252-SE) Completion Report Price Structure for Groundnut. 1222 195 1976 1978 v2 -0 i2- Reference Price (S/) and Value ex-Dakar ?/ (1/t) 261 393 607 452 419 420 372 377 378 Exchange rate 252 223 241 210 230 245 245 245 Value ex-Dakar (CFAF) 63,162 87,639 146,287 94,920 96,370 102,900 91,140 92,365 92.610 Equivalent (unshelled (70%) 44.213 61,347 102,400 66.444 67,459 72,030 63,798 64,655 64,827 Transport and handling 2/ 2,600 3,100 3,800 4,300 4,600 5.000 5,000 5,000 5,000 Economic Faragate Value 41,613 58,247 98,600 62,144 62,859 67,030 58.798 59,635 p9,827 1 Cirrent terms up to 19761 constant 1977 prices thereafter. j2 Nigerian shelled c.i.f. Europe. Excess of transport coat. over shipoing from Nigeria to Dakar only assumed to equal cutom duties and handling expenses, thus reference price unchanged. I/ CFAF 4.300 in 1975 (See Report 1094-aS of June 4, 1976, Annex 20, Table 4). Adjusted for inflation before this year and until 1977. CASAHANIE RICB PROJECT (CREDIT 252 SE) Completion Report Price Structure for Maize 1972 197 im1. 1 197? 9 1M 12M Reference price - S/t 62 73.6 132 119.7 112 120 124.9 126.6 125.6 125.9 CFAF 15,624 16.410 31,822 25,137 25,760 29,400 30,600 31,020 30,770 30.800 Freight, Insurance, Port, Custom 3,863 6,959 8,518 9,753 11,377 13,130 13.130 13,130 13,130 13,130 Value ex-Dakar 21,.487 23.369 40.330 34,890 37,137 42.530 43.730 44,150 43.900 43.930 Transport and Handling 2,600 3,100 3,800 4,300 4.600 5,000 5,000 5,000 5.000 5,000 Economic Farmgate Value 24,087 26,469 44,130 39,190 41,737 47,530 48,730 49,150 48,900 48,930 Current terms up to 1976 - Constant 1977 prices thereafter. SENIAL CASAMANCE RIS IROJEC? (GREDIT 252 SE) Completion Report Price Structure for Sorghum 1972 1M 1974 i 1976 122 1978 i2 12- i2A1 Reference price - S/t 58.6 69.6 121 111.9 105 110 111 112 111 110.1 CFAF 14,768 15,520 29,161 23.500 24,150 26.950 27,195 27,44o 27,195 26,950 Freight. Insurance.Port, Oetom 5,863 6,959 8,518 9,753 11.377 13,130 13.130 13,130 13.130 13,130 Value ex-Dakar 20,631 22,479 37,689 33,253 35.527 40,080 40,325 40,570 40,325 40,080 Transport & Handling 2,600 3,100 3,800 4.300 4,600 5,000 7,000 5,000 5.000 5,000 Economic Farugate value 23,230 25.580 41,490 37,553 40,127 45,080 45,325 45.570 45.325 45,0 1/ Current terms up to 1976; constant 1977 prices thereafter Foodcropa . 2 millet + 1 Maize 23,516 25,876 42,370 38,099 40,664 45,897 45,793 46,763 46,517 46.36) 3 ANNEX 4 Page 1 SENEGAL CASAMANCE RICE PROJECT (CREDIT 252 SE) COMPLETION REPORT Difficulties Encountered According to the Project Director 1/ A. Deficiencies of the Appraisal Report (a) Technical, financial and economic aspects: 1. serious underestimation of certain costs especially vehicles, infrastructure, roads, buildings, and water control; 2. omission of certain expenses such as paddy pedal threshers, seeds, labor; 3. false assumptions such as meeting family consump- tion needs, definition of farm models; 4. lack of knowledge about the extensive areas of grey soils, and optimistic projections regarding the rice production potential of upland areas. (b) General 5. inadequate complementary studies on feeder roads, rice mills and water control schemes because of the lack of data; 6. inadequate exploration of procurement procedures. B. Difficulties in implementing the first project: (a) With respect to production 1. design of an effective extension service hampered by an inadequate knowledge of the local environment especially about farm structures (size, organization, locatic&); 1/ October 1975 Se:hiou II Feasibility Study (Annex 6). ANNEX 4 Page 2 2. development of ox traction retarded by inadequate stump removal; 3. development of swamp rice cropping retarded by the lack of water control; 4. development of grey soil rice and lowlands retarded by a strict application of the traditional land tenure system; -5. progress of intensification and mixed farming retarded by various social constraints (specialization of tasks according to sex, lack of ability of credit to women, separation between livestock and crop production) and tradition; 6. field extension workers not sufficiently interested in training; 7. improvement of agricultural output hampered by the small size and scattering of individual rice fields as well as the over estimation of labor capacity of farmers; 8. inefficient overlapping due to the lack of coordina- tion between the various departments and projects acting in Casamance (Chinese missions, ONCAD, Seed Service, SODEFITEX, etc); 9. difficult follow-up of farm activities in the remote areas because of the lack of roads; 10. reluctance in producing food beyond family needs because of the lack of efficient cereal marketing network, lack of balance between crop prices and lack of post harvest equipment. (b) With respect to the "Programme Agricole" 11. development of ox traction slowed down because of the low price of oxen; 12. agricultural activities hampered and farmers discouraged because of the late ONCAD delivery of inputs and the ONCAD failure of supplying certain items; 13. mechanization notorized by the poor quality of some of the equipment (carts, paddy threshers) and the lack of service; ANNEX 4 Page 3 14. difficulties in recovering credit because of the reliance on government assistance, especially as a result of the drought; 15. execution of the agricultural program slowed down by the lack of storage facilities and their remote- ness from the utilization centers. (c) With respect to marketing 16. the quantities of rice marketed by farmers were low in regard to estimates and actual production because of: - low producer prices until 1973 relative to groundnut and white rice price; - the importance of family consumption; - poor organization of the official marketing network especially because of late and irre- gular financing; - existence of a more lucrati e parallel market; C. Difficulties in financing 17. project implementation retarded by the lack of start-up financing and the difficulties in implementing a revolv- ing fund; 18. extremely slow procedure in refurnishing the revolving fund because of a liquidity problem due to a prior control by the Minister of Finance. D. Other difficulties 19. project implementation slowed down and consequently higher costs - because of the inadequacy of procure- ment procedures; 20. project implementation slowed down because of the difficulties in harmonizing Government and Bank procedures. ANNEX 5 SENEGAL CASAMANCE RICE PROJECT (CREDIT 252SE) COMPLETION REPORT Monthly Rainfall in mm (SEDHOIU) 1972 to 1975 460- 400- 4W0 350- 300 250 200 100 so - *(; l l 0\ I:1/ J F M A M J J A S O N 0 S F M A M J A S 0 N 0 Ttal 1972 - - - - 12 63 184 289 138 63 - - 749 ............ 1973 - - - - 4 140 240 480 215 47 - - 1126 1974 - - 15 54 322 355 334 53 7 - 1140 1975 - - 1 7 440 375 450 80 - - 1353 .AVERAGE OVER 48 YEARS 1370 World Bank-17047  六