Guinea - Rice Development Project

Docmet of The World Bank FOR OFFICIAL USE ONLY Rq.t No. 7238 PROJECT COMPLETION REPORT REPUBLIC OF GUINEA RICE DEVELOPMENT PROJECT (CREDIT 952-GUI) May 16, 1988 Africa Regional Office This document has a restricted distribution and msy be usetd by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  REPUBLIC OF GUINEA RICE DEVELOPMENT PROJECT (CR. 952-GUI) PROJECT COMPLETION REPORT ABBREVIATIONS AND ACRONYMS AfDF - African Development Fund BNDA - Banque Nationale de Developpement Agricole (National Bank for Agricultural Development) BSD - Bureau of Strategy and Development FAC - Fermes Agro-Communales (Agro-Communal Farms) FAPA - Ferme Agro-Pastorale d'Arrondissement MDR - Ministry of Rural Development ONADER - Operation Nationale pour le Developpement de la RiziQulture (National Rice Development Agency) ORG - Operation Riz Gueckedou (Gueckedou Rice Operation)  T* WOUD ANK T W_she~Wn. D.C. 2033 U.S.A. 0~ d OW._~ -w h^_ Nty 16, i988 MEORANDUMN TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report: Republic of Guinea Rice Development Project (Credit 952-GUI) Attached, for Information, is a copy of a report entitled "Project Completion Report: Republic of Guinea - Rice Development Project (Credit 952-GUI)" prepared by the Africa Regional Office. Under the modified system for project performance auditing further evaluation of this project by the Operations Evaluation Department has not been made. Attachment This docunoot hu a mtrid distdbutin snd may be un by eipbst oa in the porfost of their offici dutia Its contents may not olhfwi be disebed wbout WaN Bi uw_ona.  yO OFICAL ONLY REPUBLIC OF GUINEA RTCE DEVELOPNENT PROWiCT (CR. 952-GUI) PROJECT COMPLETIONI REPORT Table of Contents rM No. Preface ................ . .......... Basic Data Sheet ............ .. .-...-.. ..... ..... . ii Evaluation Sumary .{..... . . ... .. ........... iv I. BACKGROU CD................N D.... 1 II. FORMITIN.....................ON.... 1 A. Origin and Preparation ........... ...... 1 ! ~~B. Appraslr a i sa................... 2 C. Targets and Goals. ...... .. ................. 3 III. IMPLEMETENTATIONo ... ................... 4 A. Effectiveness and Start-up.oo.o.o.o.............p. 4 B. Reviios..i..... 4- C. Physical Progress 5 1. The National Componentoooooooo............oo 5 2. The Pilot Rice Operation...... .-...-...... 6 Do Reporting.;"oo.000 0..0 ..0 000. . 9 E. Procurement.oo..o.o.o... ...00.. ............... 9 F. Project Costs........ ............... 10 G. Financing and Disbursement ....................... 11 IV. AGRICULTURAL INPACT.o....oooo.o.o . .. ... . o.. .oo.. . . 13 A. Productiono..o _ 13 B. Technological Change... ........... 15 V. FINANCIAL AND ECONOMIC RESULTS. o.o..o.oooo.o.o..o .. o 16 A. Farmer Incomes..n....o ............. ..16 B. Economic Reevaluation.....o...ooo....... ... 17 VIo INSTITUTIONAL PERFORMANCE . . ..... . ..... 18 A. Executing Agency.g 18 B. Accounting and Financial Performance.o ... oooooo 19 Co Trann.. a.....i n i n g ... 000.00.. 20 D. Government Performance ....... . ... ... . .0.0. 20 E. Bank Performance. .... . . .. ..... * ..... 21 This document has a restricted distribution and may be usod by recipients onlY in the peornance of their official duties. Its contents may not otherwise be dioslsed without Worid Bank authoation.  VII. SPECIAL ISSUE......................... 21 A. Political and Macro-Economic Climate.............. 21 ANNMS 3-1 Project Cost by Category and by Yer 3-2 Project Cost by Station 3-3 Financing of Project Components 3-4 Schedule of Disbursem_nts 4 Key Indicators 5 Economic Costs and Benefits 6 Financial Situation after Project Completion Maps Guinea IBRD 14163R Pilot Project Area IBRD 14164  PRDJECT COMPLITION REPORT REPUBLIC OF GUINEA RICE D3VRLOWMVCT PROJECT (CREDIT 952-GUI) This is a Project Completion Report (PCR) for the Rice Development Project for whlch Credit 952-GUI In the *um of US$ 10.4 million was lent to Government with the objective of developing the overall planning capability for the rice subs-ctor including auxillary support for related activities in the subsector. The Credit was approved on September 18, 1979, becsme effective on April 14, 1980 and closed, fully disbursed, in June 1986. The project was cofinanced with the African Development Fund (AIDF) which provided US$ 4.5 million equivalent. The PCR was prepared by the Africa Regional Office following a country visit in January 1987 and is based upon a review of the Appraisal Report (Report No. 2389, dated Aug- t 17, 1979), the President's Report (Report No. P-2516 dated August ', 1979), and the Credit Agreement of November 2, 1979. In accordance with the revised procedures for project performance reporting, this PCR was read by the Operations Evaluation Department (OED) but the project was not audited by OED staff. The draft PCR was sent to the Borrower and cofinanceir on February 23, 1988. No comments were received. S ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~  - ii - REPUBLIC OF GUINEA RICE DEVELOPKENT PROJECT (Cr. 952-GUI) PROJECT COMPLETION REPORT BASIC DATA SHEET Key Project Data Appraisal Actual or Item Expectation Current Estimate ------------------------------------------------------------__------------- Total Project Cost (US$ million) 21.0 24.6 Overrun () - 17 Credit Amount (US$ million) 10.4 10.4 Disbursed (US$ million) 10.4 10.4 Date Physical Components Completed 12/84 12/85 Proportion Completed by Above Date (x) 100 100 Proportion of Time Overrun (%) - 20 Economic Rate of Return (X) 13 11% Financial Performance Good Worse Institutional Performance Good Similar OTHER PROJECT DATA Original Actual or Item Plan Revision Estimated Actual First Mention in Files or Timetable - 4/76 Government's Application 12/77 Appraisal 3/78 - 5/78 Negotiations 4/79 5/14/79 7/79 Board Approval 7/10/79 - 09/18/79 Credit Agreement Date - - 11/02/79 Effectiveness Date 31/1180 17/3/80 04/14/80 Closing Date 06/85 12/85 06/86 Borrower Governrent of Guinea Executing Agency Operation Nationale Pour le Developpement de la Riziculture (ONADER) Fiscal Year of Borrower January 1 - December 31 Follow-on Project Name Gueckedou Agricultural Development Project Credit Number 1635-GUI Credit Amount SDR 6.7 million Credit Agreement Date: December 27, 1985  MISSION DATA sent month/ No. of No. nf Staff- Date of Item by year Weeks Persons weeks Report __________________.._______________________________________________________ Identifi.ation RMWA 2/77 0.4 1 0.4 04/13/77 Preparation RMWA 9/77 0.8 1 0.8 11/08/77 RMWA 12/77 0.5 1 0.5 01/03/78 RMWA 3/78 1.0 1 1.0 03/13/78 Appraisal HQ 5/78 4.0 4 16.0 08/17/79 Post-Appraisal HQ 10/78 2.0 1 2.0 10/24/79 8.7 20.7 Supervision I HQ 11/79 1.2 1 1.2 12/20/79 Supervision II HQ 12/79 0.5 1 0.5 01/16/80 Supervision III HQ 10/80 1.2 1 1.2 11/26/80 Supervision IV HQ 03/81 2.6 2 5.2 04/24/81 Supervision V HQ 9/81 2.2 1 2.2 10/16/81 Supervision VI HQ 06/82 2.8 2 5.6 08/04/82 Supervision VII RMWA/HQ 11/82 2.4 2 4.8 02/18/83 Supervision VIII RMWA 07/83 3.0 2 6.0 08/03/83 Supervision IX RMWA 11/83 1.4 2 2.8 01/25/84 Supervision X RMWA 5/84 1.2 3 3.6 06/08/84 Supervision XI RMWA 10/84 0.8 4 3.2 01/11/85 Supervision XII RMWA 4/85 2.0 1 2.0 05/25/85 Supervision XIII RMWA 10/85 1.4 3 4.2 11/08/85 Supervision XIV ENWA 03/86 1.2 2 2.4 03/17/86 Completion RMWA 01/87 0.8 1 0.8 24.7 45.7 COUNTRY EXCHANGE RATES Year Appraisal Year Average (1978) Exchange Rate: US$1 - Sylis 20 Intervening Years Average (1979-84) US$1 - Sylis 22.95 Completion Year Average (1985) US$1 - Sylis 22.47  - iv - REPUBLIC OF GUINEA RICZ DEVELOPMENT PROJECT (Credit 952-GUI) PROJECT COMPLETION REPORT EVALUATION SUMMARY Introduction 1. The Rice Development Project, which is the subject of this PCR, was identified in a review of the rice subsector undertaken by funds provided from the Bank Group's first project in Guinea's rural sector. The project as appraised, however, represents a considerably modified version of the initial Government request that proposed the project as supporting a highly mechanized production system end rice cultivation by farmer collectives. Obiectives 2. The Rice Development Project which was to be Implemented over a five-year period was designed to: (a) improve national planning for Guinea's rice subsector; (b) upgrade rice research; (c) assess ways to improve rural incentives; and (d) implement a pilot rice development project in Guinea's high potential forest region to test and develop technical improvements as a basis for a follow-up project in the same area. The project was prepared by consultants under the IDA-financed Daboya project (Credit 569-GUI). 3. The project, as appraised, had two main components, a national and a pilot prolect component. The National Component consisted of: (a) the establishment of a specialized Rice Development Agency - ONADER, within the Ministry of Agriculture for planning and coordinating rice development in the country; (b) a study of ways to improve rural incentives; and (c) applied rice research in three different ecological zones. The pilot project component, Operation Riz Gueckedou (ORG) consisted of: (a) low cost irrigation works on 700 ha of selected swamplands; (b) extenion, training and credit services for about 2,700 farm families and 12 mechanized cooperatives; (c) mechanized seed multiplication farm and processing plant for production and distribution of improved rice seed to farmers; (d) construction of project offices, a workshop and staff housing and (e) construction and maintenancetkm fices,of feeder roadst Implementation Experience 4. The project became effective on April 14, 1980 two and a half months a.fter the original effectiveness date. Physical implementation of  the project started at a rather slow pace and all project components had considerable start-up difficulties, except for the rural incentives study, owing mainly to inexperience of project management and technical staff, problems in obtaining office space, delays in procurement and bankruptcy of the first firm selected to execute the construction program. The problems faced in the construction program led to a considerable cost overrun for this component. As a result the program was scaled down and the completion of che seed treatment plait ws d-layed, so that it started operations only in 1984. Project execution picked up, however, from the beginning of 1984 and the project was completed by the end of 1985, one year behind schedule. 5. Total project cost was estimated at Sylis 420 million (US$21.0 million) at appraisal. Actual costs have been estimated at Sylis 562.7 million (US$24.6 million), 172 above the appraisal estimate. This amount however includes about US$1 million of prefinancing for the follow-up project (para 3.23). IDA financed about 431 of project costs compared to a provision of 501 at appraisal. The African Development Fund (AfDF) financed about 18% ot project costs compared to a provision of 23Z at appraisal and Government financed the remaining 39% compared to a provision of 271 at appraisal. The lower share of IDA financing was due to the fact that Government decided to finance a 1001 of fuel and lubricants instead of sharing the financing 50Z each with IDA as projected at appraisal. Government also financed larger shares in construction and in the purchase of supplies and spares than projected. The larger share in construction was due to the cost overrun, whilst most of the spares and supplies bvught were purchased in local currency on the parallel market not reimburseable by IDA. Results 6. The pilot component of the project has been particularly successful in stimulating agricultural development in the Gueckedou region. Particularly impressive in the ORG component has been the level of farmer acceptance of, arl enthusiasm with, the project's technical package and their willingness to construct, in their swamplands, simple water control systems with their own family or hired labor. At appraisal, only half of the swampland improvements had been envisaged this way, with project machinery beirng used for the other half. In practice, almost all swamplands were developed using labor-intensive methods. There has also been double-cropping (about 151) and permanent settlement of some farmers on swamps abandoning shifting cultivation. These are extra achievements not foreseen at appraisal. 7. The physical targets of the pilot project for upland cultivation and feeder road construction and maintenance were surpassed, and targets in terms of production under swamp conditions were largely met. Total incremental production from project participants is estimated at 6,800 tons of paddy rice (1501 of appraisal estimates) with an estimated additional  - vi - 650 tons attributable to incremental production. The incremental production came from seed released from the project to farmers outside of the project area, through farmers trad4ng seeds voluntarily and through direct sales of seed by the project. 8. Average income per manday for families participating In the project, increased from about 65 Sylis (US$3.25) in the without project situation to abeut 340 Sylis (US$14.9) with the project under upland conditions, and from about 160 Sylis (US$8.0) to about 510 Sylis (US$22.3) under swamp conditions. These are considerably higher than appraisal expectations. The economic rate of return (ERR) at appraisal wa estimated at 12.82 for the pilot component. Using similar assumptions as at appraisal but a far lower rate for seed release (para 5.05), the ER has been reestimated at 10.92. The lower rate is due principally to lover world market prices for rice than used at appraisal and conservative assumption about seed release. 9. With respect to the impact of the national project, the consultants study on rural prices served as a key element in subsequent agricultural policy discussions with Government. Agricultural research in the project represented, for over five years, the only ongoing and active research activities in the country. ONADER itself as an institution only partially fulfilled its objectives. Its role as a rice sector planning end coordinating unit was limited and its main activity before the change in Government in 1984 was directed towards keeping the project's field components going. It started from mid-1984 to get involved in Ministry activities especially in socio-economic studies and investment review. 10. The main problems faced by the project was in the area of accounting and financial management and credit recovery. This resulted from the absence of adequately trained staff and lack of experience of these staff, compounded by the fact that the Guinean administration had difficulties initially in accepting expatriate assistance in this area. Expatriate help recruited in 1984 for the project was, however, ineffective and poor record-keeping by this financial administrator on commitments, led to overestimation of funds available. This resulted in a serious overcommitment since project management, under the mistaken impression that money was still available in the credit, purchased items for the two follow-up projects, Gueckedou and First Agricultural Services Project that became effective in the third quarter of 1986. Based on the financial situation that had been submitted to it, IDA did not raise any objections to the purchases to prepare for these two follow-up projects. This was, however, due to the fact that IDA supervision concentrated more on technical and administrative matters. A more active supervision of the accounting and financial situation could have prevented the overcommitment of funds made in the project. For the pilot project credit component also, poor recording of credit has led to a recovery rate of less than 50S. The fact that funds from credit recovery were placed in a blocked account and were not expected to be recycled for additional credit also did not encourage vigilance by the project authorities.  - vil - Findinas 11. The main findinSg of the project Include: (a) The project was able to demonstrate to the Government that a suallholder orientation was a more effective and a les costly approach to dovelopment than collectives or state farmrs. (b) The project's success can be attributed to the particular attention paid to avoidisg parallel organizational structures. The merger of ONADnR into the Ministry of Rural Development helped to create a single planning unit for the Ministry. The integration or ORG into the regional Service of Rural Development gav2 it responsibility for all rural development activities in the Gueck-dou prefectures. (c) The project wa able to successfully test simple technical packages that promoted low cost yield Improvements using simple technology (such as improved seed, fertilizer, simple water control and crop husbandry Improvem_nts). This generated impressive farmer acceptance and enthusiasm for the technical packages. The technical packages are being applied on a large scale in a follow-on project (Gueckedou Agricultural Development Project, Credit 1635-GUI). (d) IDA supervision in the area of accounting, financial management and credit recovery was weak since supervision efforts concentrated on technical and administrative matters. The problems faced by the project In these areas, which resulted from lack of trained and experienced staff, and later ineffective expatriate assistance, could have boen minimized by more active IDA supervision that focussed on accounting, financial management and cost recovery.  REPUBLIC O0 GUINEA RICE DEVELOPMENT PROJECT Completion Report I. BACKGROUND 1.01 In the period just after independence in 1958, Guinea adopted an ideology that favored a direct transition from traditional rural production to a modern socialist state emphasizing collective ownership of the means of production and self-sufficiency. In an attempt to increase the productivity of rural labor, the Government promoted collectivization and mechanization using state farms, neglecting and alienating in the process, a large majority of the rural population who were individual smallholders. Practically all public investments in the rural sector were allocated to collective agriculture and mechanization. These however did not lead to any significant improvements in agricultural output. 1.02 In 1977, Government asked IDA to help finance a program that would help develop the country's rice subsector. IDA agreed to finance a project that would improve national planning for the rice subsector, upgrade rice research, assess ways to improve rural incentives and develop a pilot rice production activity based on smallholders in the high potential forestry region of the country. The project was appraised in May 1978, signed on November 2, 1979 and declared effective in April 1910. An African Development Fund (AfDF) credit cofinanced the project. The appraised project however represented a considerably modified version of the initial Government request and the project as prepared by consultants, that had proposed highly mechanized production systems and rice cultivation by farmer collectives. II. FORMULATION A. Origin and Preparation 2.01 Under the first project financed by the Bank Group in Guinea's rural sector, the Daboya Pineapple Project, funds were provided to do a review of Guinea's rice sector and identify areas suitable for development projects. A consulting firm involved in the execution of the Daboya project was asked to do the review and identify a possible rice project. The consultant report issued in December 1976 identified four sites for rice projects, two of which were selected for detailed project preparation. Feasibility studies for these two sites were completed in .Tnly 1977. 2.02 The preparation reports for the two sites, influenced strongly by the then Government policy, proposed highly mechanized production systems. Upland rice normally cultivated by smallholders, was to be grown on state farms using force acccunt with participation of individual farmers only as  -2- outgrowers, and bottomland development was to be done mechanically by farmer collectives. In the proposed design of the project, the traditional smallholder who in practice had cultivated the uplands, had been neglected although the terms of reference for the study specifically asked the consultants to propose simple and inexpensive methods of rice production through improved service to farmers. The Bank, while not completely rejecting the proposals contained in the preparation report, concluded that the project proposals were technically and economically unsatisfactory and would lead to a low rate of return. The Bank therefore asked for a reformulation of the project that would put the main emphasis on developing services for smallholders. B. ApDraisal 2.03 A four man appraisal mission visited Guinea in May 1978. At the outset of appraisal, Government indicated its interest in moving away from state farms and hence requested the appraisal mission to reformulate the project which, as prepared, had envisaged an important component for state farms. The .aission therefore redesigned the project to include measures that would assist Government in its overall rice development objectives as well as a pilot project for smallholder rice development in a part of Gueckedou region in the forest zone of Guinea. The pilot project would include simple and replicable improvements to existing fauming practices of the two basic production units in Guinea. The smallholder and the collective but with emphasis on the former. 2.04 The appraisal mission raised a number of issues as regards the project. Aside from the general economic issue of currency overvaluation and centralized decision making, the mission found that: (a) production incentives were very low when interpreted in terms of Government official output prices, though parallel market prices and unofficial exports led to real economic benefits for the farmers that would be involved in the project; (b) Government still considered swampland development as a preserve for collectives; and (c) that expatriates were generally regarded with suspicion but owing to the shortage of technical expertise in Guinea a large number of expatriates would be necessary. 2.05 A decision meeting held to consider the issues raised by the appraisal mission decided that since there was not enough knowledge on costs of production and incentives to make immediate demands for altering Government pricing policy, the pricing and incentives study proposed by the appraisal mission should be carried out rapidly and a dialogue established with Government on prices. The meeting also agreed to the incorporation of certain collectives in project execution provided they meet certain performance criteria and accept an effective technical package to augment their productivity. Proposals for including expatriates in the project were also accepted, but the meeting asked that the number be kept to the minimum necessary. 2.06 Government agreed to the reformulated project and the proposals of the appraisal mission on project components with some minor changes in emphasis. These changes however did not affect the final project design.  -3- C. Targets and Goals 2.07 The final project design consisted of two main development activities; a national component aimed at assisting Government to overcome institutional and technical constraints impeding rice development and a pilot project in a part of Gueckedou region in Guinea's high potential forest zone. 2.08 The National Component consisted of: (a) the establishment of a specialized Rice Development Agency - Operation Nationale pour le Developpement de la Riziculture (ONADER) within the Ministry of Agriculture for planning and coordinating rice development and for acting as executing agency for the project's field level activities; (b) a study of rural prices and incentives; and (c) applied rice research at three stations: Kamalo, near Gueckedou in the forest region, Baro in Upper Guinea, and Koba in Maritime Guinea, each representative of one of the three main rice growing ecological zones in Guinea. 2.09 The Pilot Rice Development Proiect, Operation Riz Gueckedou, (ORG) consisted of: (a) low cost irrigation works on 700 ha of selected swamplands; (b) extension, training and credit services for about 2,700 farm families and 12 mechanized cooperatives; (c) mechanized seed multiplication farm and processing plant for production and distribution of improved rice seed to farmers; (d) construction of project offices, a workshop and staff housing; and (e) construction and maintenance of 63 kms of feeder roads. 2.10 The National Component, through ONADER, was to provide the Ministry of Agriculture with an appropriately staffed and equipped sentral office for planning and guidance on rural sector development with particular emphasis on rice. It was also expected to restart an applied research program, limited during the project period to rice, but aimed at developing a program capable of providing appropriate agronomic recommendations and demonstrations upon which Guinea's medium and longer term agricultural development could be based. In addition the component aimed at studying producer incentives and providing Government with recommendations on prices of agricultural produce and inputs. Provision was also made in the component for preparing a possible project in Upper Guinea in the Niger/Tinkisso valley. 2.11 The ORG, unlike the Bank's first venture in Guinea's rural sector, the enclave Daboya Pineapple Project, set out to make inroads on development of what was then and still is regarded by the Bank as Guinea's key resource in the rural sector, smallholder agriculture. As such the design of this pilot component was a novelty for Guinea, whose main emphasis was on collective agriculture. Collectives or state farms were substituted by a smallholder orientation; manual labor was favored over mechanization; and the technical package promoted low cost yield impro- vements using simple technology (improved seed, fertilizer, simple water control, and simple crops husbandry improvements) and was based on proven experience in similar ecological areas. The key concept was for development of technically and economically viable methods for improving productivity whi v' could, if successful, be readily applied to a larger number of farmers. This approach also involved establishment of the usual  - 4 - range of support services in the pilot project area (management, research, extension, seed multiplication, feeder roads, extension, training, credit, monitoring and evaluation) many of which were themselves novelties in terms of practical implementation for Guinea. 2.12 The main crop to be improved during the pilot phase was rice. Based on official input and output prices, smallholder net returns were expected to increase from 52 Sylis (US$2.60) per man-day to 85 sylis (US$4.25) per man-day fr t.pland rice and from 94 Sylis (US$4.70) per man-day to 146 Sylis (US$7.30) per man-day for swamp rice. For the collectives involved in the project their net incomes were expected to increase from an average per cooperative of 21,000 Sylis (US$1,050) to an average of 600,000 Sylis (US$30,000). III. IMPLEMENTATION A. Effectiveness and Start un 3.01 The credit was approved on September 18, 1979 and became effective on April 14, 1980, two and a half months after the original effectiveness date. The four conditions of credit effectiveness were (a) the appointment of suitable directors for ONADER and ORG; (b) signature of a contract with a consulting firm or group for technical assistance services; (c) opening of a local currency account and payment by Government of its initial contribution into this account; and (d) the meeting of conditions precedent to the effectiveness of the African Development Fund Loan Agreement. At the date initially fixed for effectiveness, January 31, 1980, the conditions involving the appointment of directors _or ONADER and ORG as well as signature of contract for technical assistance had been met. However delays in meeting AfDB conditions, creating the local currency account and issuing a legal opinion for the credit led to a postponement of the effectiveness date to mid-March. The project eventually became effective on April 14, 1980. 3.02 Some start-up activities were initiated before effectiveness. Contract for the price study was signed by February 20, 1980. Bidding documents were prepared for the civil works component of the project and bids called for by the time of effectiveness. Despite these preparations, work on the construction did not start until February 1981 due essentially to delays in selecting the firm for the construction and some apparent financial difficulties of the selected firm in getting things off the ground. The price study started towards the end of 1980. Apart from the price study then that got off to a relatively good start all other project components including general field level activities had considerable start-up difficulties owing mainly to inexperience of project managers and technical staff, problems in obtaining office space, and delays in equipment and vehicle acquisition. B. Revisions 3.03 There were no major changes made in project design during implementation though some changes in scope were made. The project had envisaged a preparation study for a rice project in upper Guinea. This was  - 5 - however dropped in favor of preparing a second phase for the Gueckedou pilot project plus small pilot phase activities centered around the two research stations at Koba and Baro. The decision to drop preparation of the rice project was taken because several development agencies had already started projects in the upper Guinea area, about the time the study was due to start. 3.04 The firm that had originally been awarded the contract for construction went bankrupt only about six months after construction started. This situation was perhaps due to poor evaluation of bids by Government in not verifying the financial soundness of the firm despite IDA request to Government and AfDF, which financed the component, to review the firm's financial situation. A new company was selected and the higher cost of the new contract led to a considerable scaling down of the civil works program. AfDF had initially indicated a willingness to give a supplementary credit to cover the higher cost but changed its mind later (para 3.08). The appraisal report had proposed the construction of; an office for ONADER and three houses for expatriates in Conakry; 2 housing units, a simple office and laboratory and storage at each of the research stations in Guinea maritime (Koba) and upper Guinea (Baro); and at Gueckedou, a project headquarters, including an office block, 17 staff housing units, a workshop, hangars for equipment storage, training centre, small laboratory and a seed processing plant. At project completion only part of the Gueckedou complex including the office block and training centre, laboratory and seed processing plant had been constructed. Five of the 17 housing units were also constructed at Gueckedou. 3.05 Since the initial stages of the project no special effort was made either by project management or by IDA supervision missions tooprovide any support for the Agro-communal farms (FAC) included in the project at Government insistence. Apart from some credit for spare parts given out to them by the project in 1982, no special programs were set up for them either by way of extension, training or seed distribution. Government emphasis on these collective farms, the FAC declined since 1980 in favor of another form, the District Agro-Pastoral Farms, FAPA that were supposed to be more autonomous. From about 1983 then, the component to support the Agro Commercial farms was dropped. C. Physical Progress 1. The National Component 3.06 Prices and Incentives Study. The rural prices and incentives study started in the second half of 1980 and proceeded according to schedule. The first phase of the study was completed and discussed and the second phase field work started by April 1981. The field work was finished by about April 1982 and the final report submitted in February 1983. The report provided an important data base for a comprehensive Agricultural Sector Review Report (GUI-4672) prepared by the Bank beginning with a sector mission in November 1982. The report also served as a key document in policy discussions to encourage support for Guinea's smallholders. 3.07 Research. Research tr .als at all three stations, Baro, Kamalo (near Gueckedou), and Koba started in 1981. The lack of experience of research staff was quite evident from the conduct of trials during the first two years. Even with the support of ONADER's Agronomy Department  -6- research at Baro and Koba did not yield readily applicable technical packages. A combination of often inappropriate objectives (inadequate orientation of most trials to farmer needs), logistical difficulties (lack of fuel, labor and spares) and eco-climatic setbacks (floods and droughts) retarded progress towards the developi.ient of sure technical packages. 3.08 Construction. Despite the fact that bids had been called for the construction program before effectiveness, the program could not proceed according to schedule. Questions were raised after bid evaluation In July 1980, about the financial health of the firm proposed by Government to do the construction. Government, after discussions with the African Development fund (AfDF) that financed this component, informed IDA in September 1980 that they were satisfied that the proposed firm could carry out the construction. The contract was therefore signed in November 1980 and work started in February 1981. Construction work was halted around the last quarter of 1981 because of financial difficulties and general incompetence of the entrepreneur. A new firm was selected and a new contract signed in June 1982. The higher cost of the new contract combined with an underestimation of construction cost at appraisal led to a scaling down of the construction program that had been envisaged. The African Development Fund (AfDF) that was responsible for financing civil works had agreed in December 1982 to prepare a supplementary credit of US$2.5 million to complete the initial construction program and also to enable continued financing of technical assistance. AfDF however, did not follow up on this supplementary credit despite repeated requests from Government. The reduced construction program led to difficulties for ONADER throughout the project period especially in finding appropriate office space. The delays in the program for Gueckedou also meant that seed treatment could not start until the 1984 season when the construction was finished. 2. The Pilot Rice Operation 3.09 General. Field level activities at Gueckedou started in 1981 as envisaged at appraisal but under extremely difficult conditions. Since most of the initial equipment arrived only in late 1981, swampland development, seed multiplication and feeder road construction for this first year of field level implementation was on a rather low scale. Most of these activities started in earnest only in 1982. Farmer participation, for example, was below appraisal targets up to 1983, before picking up from 1984. Extension service performance was disappointing over the first two years owing to poor initial training and lack of experience of extension staff. Some delays were also encountered in the land development and road I construction program during the first two years due in part to lack of equipment but also largely due to inexperience of staff and for the swamp development program, due to initial reluctance of some of the farmers to I participate in it. From 1983 when farmer suspicion of Government intentions had abated, the swamp development program proceeded satisfactorily. 3.10 Agricultural Extension. During the project, an extension service was set up and towards the last two project years organized in line with the Training and Visit System. This represented the first attempt in any part of the country to organize a regular extension system for smallholders. Initially, the system was quite inefficient because of poor  - 7 - training and a clear lack of experience of the extension staff. With the introduction during the last two years of the Training and Visit system as well as consultant support in training, the quality of extension did improve. The system permitted a first attempt to introduce systematic agricultural extension for smallholder farmers in Guinea and to organize farmers into voluntary groups and is generally agreed to have helpad in reaching a larger number of farmers. 3.11 Seed Production. Improved seed was to be multiplied on a 400 ha mechanized seed production farm with seed from the farm being processed (cleaning, drying, dressing & bagging) at a seed treatment plant that was to be constructed during the first two years of the project. Using manual techniques, the seed farm cultivated some 12 ha of upland rice and 2.5 ha of swamp rice in 1981. After equipment was received in 1982, about 104 ha of land was cleared for upland rice but only 90 ha cultivated. The maximum area cultivated by che seed farm was 160 ha in 1984. In general, yields from the mechanized seed farm were lower than expected at appraisal and also much lower than what farmers obtained on their own fields, averaging about 700 kg/ha before treatment, compared to an average of 1,300 kg/ha expected at appraisal and about 1,500 kg/ha being obtained by farmers on their own fields. As of the 1984/85 season, the project realized that production from the seed farm was too expensive and did not yield enough to meet all the demand coming from farmers, it therefore selected set71 outgrowers as contract farmers who receive special extension advice on seed production. However the lack of a clear policy on seed pricing before the liberalization of the agricultural pricing system in Guinea beginning towards the end of 1985, prevented a full use of the contract farmer system. The follow-up Gueckedou project has proposed a fuller use of the contract farmer with the seed farm producing only foundation seed. . 3.12 The seed treatment plant came into operation in early 1984 and was used for processing the seed produced during the 1983/84 season. This plant represents the only seed conditioning unit existing in Guinea. Even though it has a capacity of about 450 tons/year, it has been processing an average of about 100 to 120 tons/year over the last three years (including 1986). A seed project being prepared by IDA proposes to slightly modify the origi.al design of the plant to be able to process about 2.000 tons per year to serve the whole of the forest zone of the country. 3.13 Input SupplY and Asricultural Credit. Provision of agricultural credit in Guinea, during the project period, was the responsibility of the National Bank for Agricultural Development, BNDA, but because of lack of personnel and absence of field offices it was unable to fulfill this role. Under the pilot phase of the project, ORG was responsible for input supply and credit management. It signed an agreement with BNDA, through which the latter seconded a credit agent to ORG, to be responsible for agricultural credit administration. The agent worked with the management of ORG and had two credit agents per zone. Credit terms were one year for short term credit (fertilizer, seed and implements), and three years for medium term credit of agricultural equipment and labor costs for land development. All credit attracted an interest rate of 7%, 3 points below the BNDA credit rate of 10%. 3.14 The credit unit, was attached to the extension unit and field extension agents distributed inputs as part of their extension work and recorded all credit given out for submission to, and control by, the credit  ~0 unit. Except for credit for labor used in land development, all other credit was given out in kind, and a Government policy dictated that credit was to be paid back in kind, mostly through rice. Credit recording and collection under the pilot phase was not satisfactory. In the initial project years, though recording of credit given out was done satisfactorily, repayments in kind were normally not recorded and hence it has been difficult to establish the correct reimbursement rate. Farmers were also reluctant to repay in kind, since quantities of rice needed to repay credits were calculated based on the official rice price, which was considerably lower than the parallel market price. Available information shows that the average repayment rate over the entire project period was about 41%, for both short and medium term credit. Funds from credit recovery were put in a blocked account and no provision was included in the project for re-using these funds for additional credit. 3.15 An external audit of all Phase I project activities done in early 1985 had as one of the principal tasks to establish the true situation of agricultural credit during this first phase. Because of the poor credit recording system, the audit report issued in July 1985 estimated that it would be difficult to obtain reimbursement for over half of all credits given out before December 31, 1985. The follow-up project has proposed a reorganization of the credit unit making the financial director directly responsible for credit administration. Staff have been recruited whose main responsibility would be: to study farmer credit requirements, prepare, issue and control farmer credit, distribute inputs to farmers, and recover and collect all credit repayments. Extension agents will no longer be involved in input distribution or credit recovery. Farmer repayments are to be used as a revolving fund for future input purchase. 3.16 Monitoring and Evaluation. A small monitoring and evaluation unit was to be set up at Gueckedou to monitor the project's impact and collect data for the follow-up project. The unit 4as to be made up of local Guinean staff assisted by the expatriate agro-economist of ONADER. The unit in Gueckedou was never able to get off the ground. The initial difficulty was presumably the reluctance under the Sekou Toure regime, to provide adequate information and farmers unwillingness to let the authorities know exactly what they were doing. Hence there was a lack of support from project management as well as farmers themselves for the unit's work during the initial years. The main problem of the unit however has been the rather poor technical support it obtained throughout the project f:-stly from the ONADER Agro-Economic unit and secondly from a short-term consultant financed by the project to help get the unit started. The intervention of the consultant beginning in mid-1984 (about three visits) only led to additional confusion for the monitoring and evaluation staff, who besides not having any high-level statistical training had very little experience in data collection and analysis. The consultant introduced a very comprehensive but rather theoretical management information system, for monitoring and evaluating all aspects of the project. The system required a level of sophistication in data collection and interpretation that none of the local staff had; Thus despite over six man-months of consultant services, the unit was neither able to monitor nor evaluate meaningfully any project activity. When supervision missions realised that the local staff could not apply the system proposed by the consultant, they made proposals for simplifying it. These proposals , however came too late to provide any results. Steps have been taken during the second phase to define more clearly the output required from the M&E  _9- unit. Its main activities would concentrate on monitoring the effects and quality of project interventions and performance of the recommended technical messages. A visiting consultant, well-versed in developing simple indicators of project performance will assist the unit from time to time. 3.17 Road Construction. The main objective of the road component was to upgrade and maintain 63 km of key access roads. Because of a lack of equipment and operating means for the Gueckedou subdivision of the Ministry of Public Works, the rehabilitation and maintenance of all the key secondary and tertiary arteries became the responsibility of the Rural Works Department of ORG. The road constru.tion capacity was built up between 1981-83 through equipment acquired during the project and a bulldozer given to it by the Gueckedou subdivision of the Ministry of Public Works. Through simple bridges, culverts, land clearing and filling up holes, the Roads Unit of ORG was able to upgrade several roads and tracks, linking major district centers, to all weather standards. The project rehabilitated a total of some 247 km and maintained some 195 km during the project period. Most of the equipment is satisfactorily maintained and repaired in the ORG workshop established by the Project. D. Reportint 3.18 In addition to annual reports that were generally submitted on schedule to IDA, the project also put out annually reports on the research results for the past year and the research program for the coming year. Visiting supervision missions were also provided progress reports on the project. ONADER, in its role as a planning agency, did specific studies from time to time on which it prepared reports. The quality of theoannmal and progress reports however were not uniform and generally leave gaps in information. In general while the reports do contain detailed information on work programs, staff levels and composition and activities of departments, they do not provide any insight into project impact in terms of yield and production increases or for example the quality of land development works. Neither do they give any comparisons between achievements and objectives and costs of various activities. Despite consultant support for monitoring and evaluation, the unit has not been able so far to measure project impact or monitor the reach of extension activities (para 3.16). E. Procurement 3.19 Procurement of goods constituted one of the key problem area of the project. Serious balance of payments difficulties in Guinea over several years accompanied by Government controls on trade and price levels, including a serious overvaluation of the local currency resulted in shortages of all types of goods in the country. Thus most requirements of the project had to be imported. Because of the lack of adequate training and experience of project management and financial staff, they found it difficult to plan forward as required by the procurement procedures, leading often to delays in acquiring much needed goods for the project. In general however goods and services that should be acquired under ICB were done according to acceptable procedures. The main underlying cause of procurement difficulties was the lack of forward planning. This was  - 10 - compounded before 1984, by unwieldy local administrative and procurement procedures requiring clearances and involvement of several ministries in bid document preparation and evaluation. The system changed after 1984 when the Government abolished the central bureau of Projects at the presidency responsible for overseeing all procurement. 3.20 In the area of procurement of spare parts and small items and equipment dene through international shopping, much delays and confusion were encountered. Since most items had to oe imported and there were relatively very few companies represented in Guinea, orders were often placed through the group of firms providing consulting services for the project and sometimes directly through dealers on mission to Conakry. However communication difficulties often led to delays of between 6 to 12 months even for orders placed through such channels. The work of ONADER in procurement was made all the more difficult by the fact that ORG and the research stations did not appreciate the need to adhere strictly to laid-down procurement procedures, and this complicated matters. When these outposts placed orders with ONADER, they wanted the orders filled within the shortest possible time, often unwilling to accept that certain procedures needed to be followed. Project management however sometimes took the initiative of buying some spares locally on the parallel market using local counterpart funds. This helped ease some of the problems caused by delays in international procurement. 3.21 From mid-1984, a liaison office was created for ORG, the largest "outpost", in the Conakry office of ONADER to help follow up on orders and inform ORG of the procurement difficulties. This office together with a new financial management in ONADER from 1985 did accelerate considerably acquisition of goods for ORG. It should also be noted that from this period several new firms were establishing offices in Guinea following the new Government's liberalization policies. During this period however, the objective of accelerating procurement for ORG and its follow up project, the Gueckedou Agricultural Development Project, was done at the expense of a clear monitoring of available funds leading to overcommitment of funds under the credit (paras 6.09 & 6.10). In general however, despite the delays during the early project period and slight deviations from the procurement provisions especially towards the end of the project the procurement process was satisfactory given the Guinean conditions under which the project was executed. The procurement delays did not seriously affect project execution. F. Proiect Costs 3.22 Total project costs were estimated at 420 million Sylis (US$21.0 million) at appraisal including contingencies of 77.8 million Sylis (US$3.9 million). Actual costs have been difficult to establish especially per component, in part because of poor accounting, and the fact that ONADER, which was supposed to purchase items for the other components such as ORG did not always communicate the costs of such items to the units concerned. Figures supplied by the various units (ONADER and ORG) in their annual financial reports are often not consistent with the audited accounts. Available figures indicate however that the costs of equipment, vehicles and other project inputs, construction and fertilizer, were significantly higher than appraisal, while local salaries, and fellowships and consultants have been estimated to be lower than at appraisal. A summary of actual costs and appraisal estimates compare as follows, with details in Annex 3-1 and 3-2.  - 11 - Table 3.1: C_mparison Between Appraisal and Actual Expenditure Actual Appraisal Estimated as % of Estimate Actual Appraisal US$ 030 Equipment, Vehicles, Spares, Supplies and Farm Inputs (except Fertilizer) 5.4 8.5 158 Fertilizer 0.6 0.9 148 Resident Technical Assistance 3.2 3.9 120 Fellowships and Consultant& 2.1 1.6 77 Fuel and Lubricants 4.0 3.1 77 Civil Works 2.9 4.8 166 Local Salaries 2.8 1.8 65 21.0 24.6 117 0~~ 3.23 The equipment, vehicles and spares category exceeded appraisal estimates generally because close to US$1 million of the excess served to prefinance items for the second phase Gueckedou project while about the equivalent of US$1.7 million of spares and supplies were purchased by project managerment locally on the parallel market at considerable mark up prices in dollar terms (converted in project cost at the official exchange rate). The fertilizer exceeded appraisal targets also because of some prefinancing of the second phase. Anticipating delays in the start of the Gueckedou project, management procured enough fertilizer in 1985 to cover two seasons. The Resident Technical Assistance category was over appraisal targets because of additional staff not foreseen at appraisal but included during supervision. Finally construction cost exceeded estimates because of the initial difficulties in the construction program (para 3.08) but also from the fact that Government financed a number of repairs and temporary facilities for the project before the completion of the various civil works programs. On hindsight, the acquisition of items in preparation for the second phase project proved very useful because of considerable delays by AfDF in getting the Gueckedou project effective. Thus for the whole of 1986, the new Gueckedou project operated on items acquired under the Rice Project. It should be noted though that IDA agreement for the purchases was given based on an inaccurate financial situation that the project management provided (para 6.06). G. Financing and Disbursement 3.24 IDA financed an estimated 43% of project costs compared with a provision of 50% at appraisal. Government, on the other hand financed an estimated 39% compared with a provision of 27% at appraisal. The slightly lower IDA share and higher Government share were due to the fact that Government financed a 100% of fuel and lubricants instead of the 50% projected at appraisal. Government also financed significant shares of  - 12 - construction and equipment and supplies (the latter covering supplies and spares bought locally), items for which it was expected to participate only slightly in their financing at appraisal. The initial financing plan was adhered to, in as far as foreign and local cost financing were concerned. IDA and AfDF financed all "foreign costs" 1/ and Government financed all local costs. However the initial financing plan was not adhered to in terms of percentages and categories of expenditure financed. IDA for example financed Su% of fertilizer costs, 43% of resident technical assistance costs and 79% of the costs of vehicles, equipment, spares and supplies instead of 0%, 33% and 100% respectively. In general Government share in financing different categories was highar than estimated and AfDF share lower than foreseen. Financing by category is shown in the table below with a comparison between Appraisal and actual estimates in An