Togo - Maritime Region Rural Development Project; and Rural Development Project in Cotton Areas

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 5137 PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) AND RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) June 18, 1984 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. WEIGHTS AND MEASURES 1 kilometer = 0.62 miles 1 square kilometer (km2) = 0.39 square miles I hectare (ha) = 2.47 acres 1 kilogram (kg) = 2.2 pounds 1 metric ton (Ct) = 2,205 pounds CURRENCY EXCHANGE RATES Name of Currency (Abbreviation): CFA Franc (CFAF) Exchange Rates (Maritime PCR) Appraisal Year Average US$1.00 = CFAF 225 Intervening Years Average US$1.00 = CFAF 220 Completion Year Average US$1.00 = CFAF 250 Exchange Rates (Cotton PCR): Appraisal Year Average US$1.00 = CFAF 245 Intervening Years Average US$1.00 = CFAF 236 Completion Year Average US$1.00 = CFAF 340 FISCAL YEAR Government January 1 to December 31 SOTOCO April 1 to March 31 PRODERMA January 1 to December 31 ABBREVIATIONS AND ACRONYMS FOR OFFICIAL USE ONLY CCCE - Caisse Centrale de Coopfration Economique (French Development Bank) CCM - Commission Consultatives des Marchfs (Bid Award Commission) CFDT - Compagnie Frangaise pour le Dfveloppement des Fibres Textiles (French Cotton Development Company) CIDA - Canadian International Development Agency CNCA - Caisse Nationale de Crfdit Agricole (National Agricultural Credit Agency) CY - Calendar year DEESA - Direction des Enquates, Evaluations et Statistiques Agricoles (Directorate of Surveys, Evaluation and Agricultural Statistics) DRDR - Direction Rfgionale du DEveloppement Rural (Regional Rural Development Directorate) EDF - European Development Fund ERR - Economic Rate of Return FAC - Fonds d'Aide et de Coopfration (French Bilateral Aid Agency) FAO/CP - Bank-FAO Cooperative Program GDP - Gross Domestic Product GNP - Gross National Product IBRD - International Bank for Reconstruction and Development ICB - International Competitive Bidding IDA - International Development Agency IFAD - International Fund for Agricultural Development IRAT - Institut de Recherches Agronomiques Tropicales et des Cultures Vivrieres (French Research Institute for Tropical Agriculture) IRCT - Institut de Recherches du Coton et des Textiles Exotiques (French Cotton Research Institute) IRHO - Institut de Recherches pour les Huiles et Olagineux (French Fats and Oils Research Institute) K&E - Monitoring and Evaluation MEU - Monitoring and Evaluation Unit MPW - Ministry of Public Works MRD - Ministry of Rural Development OED - Operations Evaluation Department OPAT - Office des Produits Agricoles Togolais (Agricultural Products Marketing Agency) PCA - Project Credit Agency PCR - Project Completion Report PPAM - Project Performance Audit Memorandum PPAR - Project Performance Audit Report PRODERMA - Projet de Diveloppement Rural de la Rfgion Maritime (Maritime Region Rural Development Project) PY - Project Year RD I - First Rural Development Project in Cotton Areas RD II - Second Rural Development Project in Cotton Areas RMWA - Regional Mission in Western Africa SAR - Staff Appraisal Report SNPR - Service National des Pistes Rurales (National Feeder Road Agency) SONACAM - Sociftg Nationale de Commericialisation (National Trade Company) 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. ABBREVIATIONS AND ACRONYMS (cont.) SORAD - Soci;tg Regionale d'Aminagement et de Diveloppement (Regional Development Agency - later DRDR) SOTED - Soci-t6 Togolaise d'Etudes de Dfveloppement (Office for Development Studies) SOTOCO - Soci6tf Togolaise du Coton (Cotton Development Agency) SRCC - SociftE Nationale pour la Rfnovation et le Diveloppement de la Cacaoyare et de la CafEiare Togolaise (National Cocoa and Coffee Development Agency) T&V - Training and Visit (Extension System) UNDP - United Nations Development Program PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) AND RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) TABLE OF CONTENTS Page Nu. reface ............................................................ i ,asic Data Sheets .................................................... iii tighlights ......................................................... vii PROJECT PERFORMANCE AUDIT MEMORANDUM I. PROGRAM CONTEXT. ... ...............1............... II. PROJECT SUMMARY - MARITIME PROJECT .................... 2 Background and Project Formulation ........... .......... 2 Project Implementation and Impact .................... 4 Project Administration .............................. 5 III. PROJECT SUMMARY - COTTON PROJECT ......................... 6 Background and Project Formulation ................. 6 Project Implementation and Impact -.................. 7 Project Administration ................................ 8 Second Phase Project ............................... 9 IV. MAIN ISSUES .............................................. 10 A. Feeder Road Development ............................... 10 B. Monitoring and Evaluation ............................... 13 C. Institution Building ......... .................. 16 D. Togolization ............................................ 18 E. Input Subsidies and Price Policy ........................ 20 F. Sustainability .......................................... 23 Tables: 1. Togo Lending Program ......................................... 25 2. Agricultural Lending to Togo ................................. 26 3. Audit Reports Issued ........................................ 27 4. Feeder Road Construction/Improvement in Togo ................. 28 5. Cotton Project Supervision Reporting on Pricing .............. 29 6. Cotton Input-Output Prices ................................... 30 TABLE OF CONTENTS (Continued) Page No. Annexes: 1. Background Data Maritime Project ............................. 31 2. Background Data Cotton Project ............................... 45 3. Feeder Road Project Statistics ............................... 51 4. Project Description - Second Cotton Project .................. 54 5. Ex-post Evaluation by Cofinancier ............................ 56 Attachments: 1. Comments from Cofinancier ...................................... 59 2. Comments from DEESA ............................................ 60 3. Comments from CNCA ............................................. 61 PROJECT COMPLETION REPORT (MARITIME PROJECT) I. Introduction ............................................. 65 II. Agricultural Sector ........................................ 65 III. Project Formulation .................................... 68 IV. Project Implementation ..................................... 73 V. Project Justification ...................................... 84 VI. Institutional Development ................................ 85 VII. Borrower and Bank Performance ............................ 85 VIII. Impact of the Project .................................... 86 IX. Conclusions ............................................... 87 Annex: Estimated and Actual Costs PROJECT COMPLETION REPORT (COTTON PROJECT) I. Background ........................................ ... 91 II. Formulation ................................................ 92 III. Implementation ............................................ 95 IV. Financial Aspects ......................................... 99 V. Agricultural Aspects .-.................................... 101 VI. Institutional Aspects .......... ........... .......... . 105 VII. Economic Results ......................................... 107 VIII. Follow-Up Projects ....................................... 110 IX. IDA Performance .......................................... 111 X. Conclusions .............................................. 113 Annex: Project Completion Details (Tables 1-9) Maps: 18010 (PPA) Togo Main Crops and Climatic Zones 11886 (PPA) Maritime Region Ecological Zones 17808 (PPA) Maritime Project Area 17809 (PPA) SOTOCO Institutional Infrastructure 16296 (PPA) Second Cotton Project PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) AND RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) PREFACE This is a performance audit of two rural development projects in Togo. Credit 638-TO, in the amount of US$9.5 million, was approved in June 1976 for the Maritime Region Rural Development Project. With the agreement of the Borrower, US$1.7 million was cancelled in February 1982 when it became evident that further expenditure on farm development was not justified; US$50,000 had earlier (1980) been cancelled because of misprocurement. The Maritime Project was originally expected to be completed in June 1980 and the Credit closed by December 31, 1981. In fact, the project was not completed until the end of 1982, and the Credit was closed in December 1982; the small outstanding balance was cancelled. Credit 741-TO, in the amount of US$14.0 million, was approved by IDA on September 13, 1977 for the (first) Rural Development Project in Cotton Areas. It was substantially completed, as anticipated, in December 1981 and final disbursements made in July 1982, one month after the original credit closing date. In 1980, US$0.79 million was cancelled because .of misprocurement. On November 23, 1982, IDA approved another credit of US$23.5 million for a Second Rural Development Project in Cotton Areas (Credit 1302-TO). Both projects were cofinanced by Fonds d'Aide et de Coopfration (FAC)-the Maritime Project with a grant equivalent to US$2.25 million, and the Cotton Project with a grant equivalent to US$5.1 million. FAC has evalu- ated both projects: "Evaluation Economique de l'Op6ration Dfveloppement Rural de la Rigion Maritime au Togo" (January 1983) and "Evaluation Economique de l'Activitf Coton au Togo' (April 1983). Both reports are available in OED and are briefly summarized for ease of reference in Annex 5. The audit report consists of an audit memorandum prepared by the Operations Evaluation Department (OED), a Project Completion Report (PCR) on the Maritime Project dated August 15, 1983, prepared by the Western Africa Regional Office and a PCR on the Cotton Project dated August 31, 1983, pre- pared by the Regional Mission in Western Africa (RMWA). The audit report is based on a review of the appraisal reports (Maritime: No. 1100a-TO dated May 21, 1976; Cotton: No. 1494-TO dated August 26, 1977), the President's Reports (Maritime: P-1823-TO dated May 24, 1976; Cotton: P-2112-TO of August 31, 1977), the Credit Agreement for the Maritime Project dated June 16, 1976 and the Credit and Project Agreements for the Cotton Project dated October 5, 1977, the PCRs and the evaluation reports prepared by FAC. Correspondence with the Borrower and internal Bank memoranda on project issues as contained in relevant Bank files have also been consulted, and Bank staff associated with the project have been interviewed. - ii - An OED mission visited Togo in November 1983. Discussions were held with project and Borrower officials. The information obtained during this mission was used to test the validity of the conclusions of the PCR. The audit concurs with the principal points made in the two PCRs, but regrets that only limited quantitative information was annexed to these PCRs. Points discussed in the audit memorandum have been selected because of their importance to this and other Bank-supported rural development projects. The draft report was sent to the Borrower and cofinanciers for comments on April 30, 1984. Comments received have been attached to the audit memorandum. The valuable assistance provided during the preparation of this report by officials of the Government of Togo and the executing agencies is gratefully acknowledged. * - 111 - PROJECT PERPORMANCE AUDIT BASIC DATA SHEET G000 MARITIME RECION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) KEY PROJECT DATA Appraisal Actual or Actual as Z of Expectations Current Eatmate Appraisal Estimates Total Project Cost (US$ million) 15.7 12.7/a 81 Credit Amount (1SS million) 9.50 7.757l 82 Co-financing (US$ million) Fonda d'Ade at de Coopiration (France)lc 2.25 2.25 100 Date Physical Components Completed 06/30/81 12/31/82 130/d Proportion Completed by Target Date (M) 100 n.a./e Economic Rate of Return (Z) 34 1-4 Institutional Performance good fair Agronomic Performance good poor Number of geneficiaries 20,000 12.000/f CUHULAT:vE ESTIMATED AND ACTUAL D1SBURSEMNTS FY77 FT78 FY79 FTSO FT81 FT82 FY83 Appraisal estimate (US$ million) 1.0 3.0 5.0 7.1 8.8 9.5 9.5 Actual (USS million) 0.5 1.3 2.5 4.4 5.8 6.9 7.75 Actual as X of appraisal (%) 50 43 50 62 60 73 82/b Date of Final Disbursement: July 1982 PROJECT DATES Original Plan Revisions Actual Conception 1972 1972 Identification Report 1973 12/04/73 Negotiations 05/77 05/77 Board Approval 06/76 06/06/76 Cred2* Signing 06/16/76 Effectiveness 09/14/76/ 11/15/76 10/26/76 Closing 12/31/81/h 12/31/62/1 12/31/82 MISSION DATA Sent Date No. of Dayal Specializations/k Performance/l Types of/ Nision b (Month/Year) Persons in Field Represented Rating Trend /m Proft1ems Identification/Preparation Identification ) FAO/CP 1968 - - Identification )o RMA 12/70 - - Preparation ) CIDA 1971 - - Reconnaissance / EQ/D1WA 11/12/72 3 21 I,E.A Identification ) RRIA 05/73 1 6 A Identification -)Lr RHWA 06/73 1 7 A Preparatian ) RNWA 01174 1 6 A Preparation _- RMA 02/74 1 3 A Preparation ) RSIA 05/74 1 7 A Preparation )/t RNUA 05/74 1 2 A Preparation )W-- RA 06/74 1 3 A Preparation ) RKWA 08/74 1 2 A 57 Appraisal Preappraisal HQ 06/07/75 2 42 Appraisal HQ 10/11/75 4/u 92 A,E - 034 A,E.IC Supervision Supervision I HQ 10/76 3/v 1 A,E.A/w 1 2 none Supervision 11 EQ 05/77 17 19 A,T 2 2 n,o/ Supervision III HQ 10/77 2 14 A,E 2 2/ Supervision IV HQ 04/78 2 11 A,E 2 2 O& Supervision V HQ 10/78 2 8 F.YP 2 2 Obb Supervision VI HQ 02/79 2 19 A,F 2 2 M.Or. Supervision VII HQ 10/79 2 10 A.F 2 2 ,.07d0 Supervision VIII HQ 03/80 3/ee 14 A,F.E 2 2 07 Supervision II HQ 11/80 27ff 25 A,YP 3 3 T.M.0 Supervision X HQ 02/81 37a 15 D.A.E 3 2 T.M,0 Supervision II HQ 03/82 1 6 A 2 2 T.K.0 Completion HQ 06/82 1 5 A - - - Completion RQ 06/83 1 5 P - - - Total 169 OTHER PROJECT DATA Borrower Republic of Togo Executing Agency Initially: SORAD (Socilti Rhgionale d'Amnagement at de Dfveleppement) Subsequently: PRODERHA (Projet de DevEloppenent Rural de la RAgion Maritime) Preceding Project None/h Follow-on Project None Footnotes on following page. - iv - /a Project implementation halted, except for some minor components, when technical package was not widely adopted by farmers. b On March 5, 1980 IDA cancelled US$50,000 for aiuprocurement; on February 23, 1982 IDA cancelled US$1.7 million am the project vas being terminated; an outstanding balance of USS13,487.99 was cancelled in September 1983. le See also PPAN Chapter 4. 7d Time overrun calculated from date of signing. 7 Not comparable, part of Credit cancelled. 7? Farmers reached by the extension service. i Credit Agreement Section 6.02. h Credit Agreement Section 2.04. 7? Extended through Program Department telex dated March 10, 1982. Change in Bank's computerized accounts, however, only made in September 1983. / Not duration of mission, but total field days for all mission members combined. Ik I - Irrigation engineer; A- Agronomist; E - Economist C - Credit Specialist; T - Treecrop Specialist; F - Financial Analyst; YP - Young Professional; D - Deputy Division Chief; and P - Project Generalist. /1 1 - problem-free or minor problems; 2 - moderate problems; 3 - major problems. /n 1 - improving; 2 - stationary; 3 - deteriorating. 7n F - Financial; M - Managerial; T - Technical; P - Political; and 0 - Other (clarification in footnotes). 7o Identification of irrigation development in the valleys of the Sio and Baho rivers. L CIDA - Canadian International Development Agency. /q Mission to convey IDA reservation regarding irrigation proposals; mission suggested outline for rural development project. lr Missions to discuss outline and to suggest practical steps for preparation. is Missions to discuss identification report and make logistical arrangements for preparation. 77 Missions to review preparation work by consultants. 7u In addition the loan officer participated during the first week of the appraisal. 77 The M&E component was discussed in December and a two-day review mission of the component took place in March 1977. 7w One of the agriculturalists was an OPS advisor. 77 A two-person IDA mission visited Togo in May 1977 to discuss disbursement procedures. i Procurement probleme. Iz Severe drought; procurement problems. 7a Effects of severe drought on physical results, credit recovery. 7b Credit recovery. 7c Credit recovery and Monitoring and Evaluation Unit. 7id Price and yield variations and credit recovery; monitoring and evaluation. 7e In addition a consultant (monitoring and evaluation) financed from credit proceeds, visited Togo at the same time. 71? The draft animal component of this and the Cotton Project was reviewed separately by a Bank livestock specialist. A follow-up mission took place in June 1981. l1t Although French bilateral assistance had supported SORAD earlier with funding and especially technical assistance. -v - PROJECT PERFORMANCE AUDIT BASIC DATA SUEET TOGO RURAL DEVEDPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) KEY PROJECT DATA Appraisal Actual or Actual as 2 of Expectations Current Bastimate Appraisal Estimates Total Project Costs (US$ million) 26.0 26.8 103 Credit Amount (US$ million) 14.0 13.2/a 94 Cofinancing /c (US$ million) Fonds d'AYe at de Cooperation (France) 5.1 5.1 100 Date Physical Components Completed 12/31/81 12/31/81 - Proportion then completed (Z) 100 80 80 Economic Rate of Return (Z) 32 11 34 Institutional Performance good good Agronomic Performance good fair/d Number of Direct Beneficiaries (Incremental) 34,000 18,800 55 CUULATIVE DISBURSEMENTS FY78 FY79 FY80 FY81 FY82 F133 Appraisal estimate (US$ million) 1.45 4.85 8.25 11.65 14.00 14.00 Actual (US$ million) - 1.80 3.88 9.17 13.15 13.21 Actual as % of appraisal (2) - 37 47 79 94 94 Date of Final Disbursement: July 23, 1983 PROJECT DATES Original Plan Revisions Actual Identification Report 11/18/75 Negotiations 06/18-20/77 Board Approval 06/77 09/13/77 Credit Signing 10/05/77 Effectiveness Date 01/03/78/e 04/03/78/ 03/17/78 Closing Date 06/30/827-f 12/31/8Y 12/31/82 MISSION DATA Sent Month/ No. of Handays Specializations Performance Types of Mission By Year Persons in Field Represented li Rating L Trend k Problems .1 Identification EMMA 9-10/75 3 29 E.A.F Preparation FAC 10-12/75/a 2 - Preparation) EMMA 04/76 - 1 4 E Preparation)- RWA 08/76 1 3 E Appraisal/o HQ/RMWA 10/76 6/E 105 F,A,E,I,H,T 141 Supervision I HQ 10/77/i 2 10 A,E 1 -/r Supervision II EQ 04/78 2 11 A,E 2 2 N Supervision III HQ 10/78 2 7 A,F 2 1 F,M Supervision IV EQ 05/79 1 9 A 2 2 M,F Supervision V HQ 09/79 2 8 A,E 2 3 F,M,T Supervision VI RQ 03/80 2 16 A,F 3 2 F Supervision VII HQ 11/80 2 16 A,E/s 2 1 F Supervision VIII BQ 03/1 3 11 E,D,Y,P 2 1 F Completion RMMA 01/83 1 4 E - - - Completion EMMA 03/83 1 2 E - - - 96 OTHER PROJECT DATA Borrower Republic of Togo Executing SZcncy SOTOCO Preceding Project None Follow-on Project: Name S-cnnA Rural Develnpnent Project in Cotton Areas/t Credit Number 1302-TO Credit Amount (US$ million) 23.5 Date of Board Approval 11/23/82 Footnotes on following page. - vi - /a US$792,000 cancelled because of misprocurement. 7 FAC's actual contribution amounted to CPAF 1,250 million as expected at appraisal. 7; In addition FAC had provided preproject funding to SOTOCO amounting to CFAP 95 million (US$.04 million) in 1975 and CFAF 175 million (US$0.7 million) in 1976. /d Good for cotton; poor for foodcrops. 7 Credit Agreement Section 6.03 7T Credit Agreement Section 2.04 7 Telex dated December 23, 1977. h Telex dated April 14, 1982 7? E - Economist; A - Agronomist; F - Financial Analyst; I - Irrigation Engineer; H - Highway Engineer; T - Transport Economist; D - Deputy Division Unit: TP - Young Professional. / - problem free or minor problems; 2 - moderate problems; 3 - major problems Ik 1 - improving; 2 - stationary; 3 deteriorating 71 F - Financial; H - Managerial; and T - Technical 7m Feasibility Study prepared by consultants financed by FAC. 7; Short mission to discuss Cotton Project as prepared. 7- The Agricultural Projects Division Chief and the Loan Officer also were in the field during part of the appraisal. Zi Includes 3 staff members of RHWA who participated part time in the appraisal (Irrigation Engineer, Highway Engineer and Transport Economist). A short follow-up visit by a financial audit took place in November 1977. r No further ratings given in SPV report. 7-s A livestock specialist reviewed the animal traction component of this and the Maritime project separately. * 7? For details see SAR Second Rural Development Project in Cotton Areas Report No. 3920-TO dated October 21, 1982. - vii - PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) AND RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) HIGHLIGHTS As of June 1983, Bank lending to Togo for 18 projects amounted to US$200 million- Initially, emphasis was on infrastructural development and agriculture, but structural adjustment and technical assistance have received greater attention more recently. Bank interventions in the Togolese agricul- tural sector have covered the most important agricultural export crops (coffee, cocoa and cotton); geographically, they have covered the totality of the national territory. The Maritime Region Rural Development Project (Mari- time Project) was the Bank's second, and the Rural Development Project in Cotton Areas (Cotton Project) the Bank's third, agricultural development project in Togo. The Maritime Project was intended to increase food production through the application of improved technology and intensified cropping under the difficult climatic and soil conditions in the Maritime region. Project start-up was speedy but implementation thereafter was mixed. Until the early 1980s, IDA's perception of project implementation was not especially nega- tive, but by that time it was realized that the project's technology was accomplishing little. Major project components were then terminated, but the project closing date was extended to allow for the completion of a few small and more successful components. The project was, in general, well- administered. Input supply and credit functions were established, but not as envisaged at appraisal, and ceased to exist at project completion. In any case, whatever was achieved in terms of institutional strengthening has been negated by the failure of the project. The poor results, especially with regard to crop and farm develop- ment, were inter alia due to rainfall extremes, failure of smallholders to widely adopt the technical packages proposed, absence of basic data, and lack of any demonstration effect. As such, the experience demonstrated that smallholder farming as envisaged by the project was not viable on a large scale in the Maritime region. In the absence of monitoring and evaluation data, increases in yields and production are difficult to ascertain, but cer- tainly have been small. At completion, the rate of return was, therefore, reestimated in the range of 1-4%. Only a few small components were success- ful: vegetable production and marketing, coconut field trials and seedling distribution and rural dispensaries. In retrospect, it would have been much better if, rather than a "traditional" rural development project, a pilot/ rural development fund type of approach, emphasizing research and institr- tional strengthening, had been implemented. The Cotton Project aimed at improving living conditions and bene- fiting more directly some 53,000 farm families, basically through improved and expanded cotton and foodcrop production, infrastructural development and - viii - project management support. Despite substantial financial and organizational problems, SOTOCO (the project executing agency) managed to more than triple cotton production, reaching about 90% of appraisal objectives. As with other cotton-based rural development projects in Western Africa, foodcrop produc- tion received less attention, and although precise data are lacking, foodcrop production is estimated to have reached only half the appraisal targets. Infrastructural development was substantial, and SOTOCO's feeder road development capacity is probably the most efficient in the country. In the end, the project had substantial impact on cotton production and rural incomes. However, the rate of return is reestimated at 11Z, far below the appraisal forecast of 32%. This is mainly due to much lower actual and (currently) forecasted cotton prices than were estimated at appraisal, and to a limited extent, to a shortfall in production. No follow-up project in the Maritime region has materialized, either with Bank financing or assistance from other donors. Two decades of foreign assistance in the Maritime region have come to an end without any satisfactory impact, a sad result. A Second Rural Development Project in Cotton Areas was approved by the Bank in November 1982. As in earlier audits, these two projects reveal difficulties with institution building and effective counterpart training. In addition, they raise concerns about feeder road development and monitoring and evaluation of agricultural development in Togo. Feeder road development was a component under both projects as well as other agricultural projects in Togo. To reduce the proliferation of feeder road programs, a clear strategy was developed during appraisal: a feeder road project would be financed by the Bank and a feeder road agency established which, at the end of the agricultural project period, would take over equipment and responsibility for feeder road development in Togo. In fact, the inverse happened. The feeder road agency had not become a very effective organization, and is now acting as a sub-contractor for SOTOCO and other agricultural projects. The proliferation of feeder road development responsibilities continues. Long-term institutional arrangements for feeder road construction and maintenance in Togo remains a major issue (PPAM, paras. 29-37). Monitoring and evaluation was envisaged for both projects basically by a central unit in the Ministry of Rural Development. While ideas for M&E were sound, they lacked detail, a major cause for subsequent implementation delays. In both projects, M&E was a failure (PPAM, paras. 38-46; Maritime PCR, paras. 4.12-4.13; and Cotton PCR, paras. 3.53-3.54). M&E institutional development initially grew in the direction of a national evaluation system. Regretfully, more limited project concerns prevented continuing development in this direction; instead, such development changed course and has even worsened under the Second Cotton Project (PPAM, paras. 45-46). Institutional arrangements in the agricultural sector were extreme- ly complex when these two projects were appraised. They still are. These institutional arrangements hardly changed during project implementation (PPAM, paras. 47-50). Similarly, input subsidies and price policy were of - ix - major concern to IDA from the beginning of project preparation. However, not only did IDA have to reduce its posture regarding the resolution of this issue during project negotiations, but substantial progress in reducing input subsidies was not made during the project implementation periods. Only recently has Government been more receptive to changing its input/output pricing policies (PPAM, paras. 56-61). OED Annual Reviews have often pointed out that projects are ineffective vehicles to ensure policy changes. Certainly the experience with these projects seems to confirm this. However, projects are part of a wider Bank program. Given the absence of substantial improvements in sector policy making, with the exception recently of price policy, the judgment must in fact be that, overall, the Bank program has been ineffective during the period that these projects were implemented. Togolization was achieved under the Maritime Project and was un- doubtedly facilitated by changing the concept of designated counterparts into placing staff in line positions for further promotion upon proven perfor- mance. Hardly any Togolization took place in SOTOCO. In fact, there are more expatriate positions under the second than under the first project. However, realism is required in this respect . It does not seem to make much sense to seriously deplete middle management levels (the backbone of SOTOCO's organization) for purposes of Togolization. Furthermore, the second project is more complex, involves more components and, therefore, requires more expertise (PPAM, paras. 51-55). In terms of sustainability, the Maritime Project was basically a failure and, therefore, there is nothing to be sustained. As mentioned above, a different project design might have worked, but the sustainability of an alternative design is a hypothetical question. The Cotton Project was basically successful due to: (i) a well-adapted technical package; (ii) remunerative producer prices; and (iii) good institutional support. Although there is room for improvement in efficiency, project achievements are likely to be sustained, especially with the continued support of the second phase project. The major lesson from this experience is that technology transfer will only take place if the technical package is well-adapted, not only to the physical but also the socio-economic environment, and that remunerative producer prices are key to continued farmer interest in cash crop develop- ment. Other points of interest are: - advantage of technical assistance umbrella contracts over a multi- tude of individual contracts (PPAM, para. 55); - the Bank should be careful to see that its macro and micro objec- tives are harmonized to the extent possible (PPAM, para. 50); - price consultation clauses have not been adhered to, confirming other OED experiences that such clauses have generally not been very successful (PPAM, para. 59); and - more detailed "engineering- of M&E is required during preparation and appraisal (PPAM, para. 41). PROJECT PERFORMANCE AUDIT MEMORANDUM TOGO MARITIME REGIONAL RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) AND RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) I. PROGRAM CONTEXT 1. The Maritime Region Rural Development Project (Credit 638-TO),l/ approved in 1977, was the Bank's second, and the Rural Development Project in Cotton Areas (Credit 741-TO),2/ the Bank's third agricultural development project in Togo. The first agricultural development project, approved in 1974, was the Cocoa and Coffee Development Project (Credit 503-TO).3/ The Banks' interventions in the Togolese agricultural sector have covered the most important agricultural export crops4/ (coffee, cocoa and cotton) and geographically, the totality of the national territory. Second phases have been approved for the Cocoa and Coffee Project5/ and the Cotton Project.6/ No follow-up activities to the Maritime Project have been financed by IDA or by other aid agencies. 2. Bank lending to Togo (as of June 30, 1983), for 18 projects, has amounted to about US$200 million (see also Table 1). The sectoral composi- tion of the lending program was as follows: Loan/Credit Percentage Subsector Number of Projects Amounts of Total (US$mln.) (%) Agriculture 5 67.00 33 Highways 5 48.20 24 T.A./Engineering 3 13.38 7 Industry 2 9.20 5 Education 1 11.00 5 Water Supply 1 12.00 6 Structural Adjustment 1 40.00 20 18 200.78 100 1! Hereafter referred to as Maritime Project. 2/ Hereafter referred to as Cotton Project or RD I. 3/ See also, Project Performance Audit Report - Togo First Cocoa and Coffee Development Project (Credit 503-TO), OED Report No. 4552 dated June 15, 1983. 4/ Togo's major export commodity is phosphates. 5/ Second Cocoa and Coffee Development Project (Credit 945-TO), approved in June 1979 in the amount of US$14 million. 6/ Second Rural Development Project in Cotton Areas (Credit 1302-TO), approved in November 1982 in the amount of US$23.5 million. - 2 - 3. Initially, emphasis was on infrastructural development (highways) and agriculture. More recently, structural adjustment and related technical assistance activities have received greater attention. Project implementa- tion, in general, has been satisfactory, but there have been problems. The Maritime Project clearly failed to reach its objectives; difficulties encoun- tered with counterpart funding have hampered the Feeder Road and Fourth High- way Projects, and technical and financial problems encountered in the CIMAO Project required a supplementary restructuring effort. Implementation notes on agricultural projects are in Table 2.7/ 4. In addition to the two projects audited in this report, four proj- ects8/ have been reviewed by OED (Table 3). There is one common theme among the various issues discussed in these reports, namely difficulties with institution building and lack of effective counterpart training. To date these remain areas of concern. Il. PROJECT SUMMARY - MARITIME PROJECT Background and Project Formulation 5. The Maritime Project has its origins in project initiatives for- mulated in the late 1960's. At that time, Government was interested in irri- gation in the valleys of the Sio and Haho rivers. After review by RMWA of the identification report, IDA conveyed end 1972 its reservations regarding irrigation development in the Sio and Haho valleys to Government, but indicated that a rural development project in the Maritime Region would be a .possibility./ Subsequently, the French bilateral aid agency (FAC) financed land capability and land use mapping in the area, and UNDP financed the feasibility study. Substantial changes in project design (Maritime PCR, para. 3.01) were made during appraisal, which took place in 1975. Of major concern during further project processing were the institutional framework (PPAM paras. 47-50) and the input subsidies (PPAM paras. 56-61). The latter also was a major point of discussion during the negotiations (Maritime PCR, para. 3.14) in May 1975. 7/ Implementation notes on other projects can be found in the most recent President's Report (Structural Adjustment Project), No. P-3534-TO dated April 26, 1983 and in the Status of IBRD/IDA Projects in Execution docu- ment dated November 3, 1983 (SECM 83-1010, 11/03/83). 8/ Highway Maintenance (Credit 131-TO), Second Highway (Credit 450-TO), Cocoa and Coffee Development (Credit 503-TO) and Third Highway (Credit 693-TO) Projects. 9/ The Bank's full indentification report was completed in 1973 and distri- buted to Government in early 1974. - 3 - 6. After minor debatelO/ the Board approved the project on June 8, 1976. The Credit Agreement was speedily signed thereafter (June 16, 1976)11/ and the Credit became effective on October 26, 1976.12/ 7. The project as approved would be the first phase of an integrated rural development program of the Maritime region. The project would cover the entire region, but areas of concentration would be selected to benefit about 20,000 farm families over the five-year period of 1976 to 1981. It would be carried out primarily by SORAD Maritime and would involve: (a) Farm and Crop Development. Providing farmers with inputs, equip- ment, and extension and credit services to: (i) improve and diver- sify the production of the principal upland crops (maize, cassava, groundnuts, cowpeas and cotton) on some 18,500 ha for 12,000 families; (ii) develop about 3,000 ha of lowland rainfed rice for 6,000 families, and provide about 100 power tillers and 75 village rice mills; (iii) develop 25 ha of vegetable production for 250 families, and rehabilitate irrigation facilities on 20 ha of exist- ing vegetable production; and (iv) renovate 600 ha of smallholder coconut for 2,000 families. (b) Farmer Support Services. (i) Strengthening SORAD Maritime by pro- viding additional staff, staff training, buildings and vehicles, expanding seed multiplication, and establishing an input supply unit; and (ii) establishing an agency of CNCA to meet project credit requirements. (c) Infrastructure. Constructing or improving 300 km of roads, 15 dis- pensaries, 20 wells, 2 earth dams, 20 village markets and 150 village stores. (d) Studies and Field Trials. (i) Carrying out an engineering study for the gravity irrigation of 1,000 ha in the lower Zio valley for rice production; (ii) conducting field trials on techniques to re- store the productivity of depleted soils, and to test the resis- tance of coconut hybrids to Kaincope disease; (iii) establishing a project monitoring and evaluation unit; and (iv) providing 10/ Questions were raised regarding the need for IDA to approve annual road plans, any changes in organization of farmer support services, and any new "non-project" activities; and whether rainfall shortages in certain seasons had been taken into account in economic calculations. 11/ The letter setting out disbursement procedures was sent soon thereafter (June 23, 1976). Measures to be taken for effectiveness were spelled out in a formal memorandum (Aide Memoire) dated May 6, 1976 and handed to the Togolese delegation during negotiations. A repeat and reminder was sent by letter dated August 16, 1976. 12/ After one extension, from September 14 to November 15, 1976 to allow IDA time to receive and review the various documents in the mail at that time. consultant services to make investment proposals for the area of North Togo to be cleared of onchocerciasis. Project Implementation and Impact 8. Project start-up was speedy and senior staff were in place in early 1976. The organization of the project was, therefore, well underway by the time of the first supervision mission, thanks to the financing of the start-up phase in the first half of 1976 by FAC. Implementation thereafter was mixed (for chronological description, see Maritime PCR, paras. 4.02- 4.07), but until 1980, IDA's perception of project implementation was not especially negative. By that time, however, it was realized that project technology was accomplishing little. Following a detailed in-depth review in January 1981, it was decided to terminate the farm development components as of June 1981, and in February 1982, US$1.7 million was cancelled. Only a few small components13/ were continued with the remaining credit funds, and the closing date was extended by one year (until December 31, 1982) to allow for their satisfactory completion. 9. Farm and crop development under the project were failures. The extension service did reach a large number of farmers (about 12,000 compared with an appraisal estimate of 18,00014/), but the area reported under improved upland cultivation was only 3,600 ha, much below the 18,500 ha appraised target; rice development covered 750 ha compared with a 3,000 ha target. The vegetable component was successful, as were coconut plantings: 600 ha were planted (on target) but mainly by larger farmers. Less than 150 smallholders were affected by this program compared to the probably over-optimistic appraisal objective of 2,000 (Maritime PCR, paras. 4.23-4.25). 10. Infrastrucutural development (roads, wells, dispensaries, stores) under the project was much below appraisal estimates (Maritime PCR, paras. 4.17 and 4.18 and Annex 1, Table 1). In addition, the SORAD office building in LomE was expanded, a building for the CNCA agency constructed in TsEviE and a training center established in AgbatopE. The seed multiplication program was beset with many difficulties (Maritime PCR, para. 4.26). Mechanized land clearing and plowing was also envisaged under the project, but only limited surfaces were treated each year (Maritime PCR, para. 4.27). 11. A French consulting firm carried out a study of the Sio Valley in 1979 and confirmed the possibility of the development of a 300-hectare scheme for rice irrigation. Field trials on foodcrops by IRAT and tests on coconut by PRODERMA, with technical assistance from IRHO, were carried out satis- factorily. Monitoring and evaluation was not effective (PPAM, paras. 38-46). Consultant services for the onchocerciasis study were replaced by 13/ In addition, it was agreed that a a socio-economic study of the Maritime region would be undertaken by consultants (PPAM, para. 5.19). 14/ SAR estimate of beneficiaries of upland crop and lowland rice development. -5- preparation of a rural development project in the Bassar region. Preparation was done by SOTED,15/ assisted by consultants. However, the Second Cotton Project subsequently replaced the Bassar Project (PPAM, paras. 27-28). 12. The project reached a substantial number of farmers and was espe- cially successful in establishing farmer associations (Maritime PCR, para. 4.15), but without an acceptable and adequate technical package, the produc- tive impact of the extension effort was limited. It has now become clear that although the risks of agricultural development in marginal areas with uncertain climate had been recognized during appraisal,16/ the project was implemented as a "normal" rural development project. The absence of effec- tive monitoring and evaluation aggravated the lack of learning about project risks. In retrospect, it would have been much better, if, rather than a "traditional" rural development project, a pilot/rural development fund type of approach emphasizing applied research and institutional strengthening had been implemented. Project Administration 13. At appraisal it was envisaged that the project would be carried out by SORAD Maritime and by CNCA for the credit component. The role of SOTOCO in the project area would be limited to cotton input supply, marketing and technical advice. The project was in fact administered as envisaged at appraisal. Although organizational changes took place during the proj- ect,17/ this did not affect project implementation. The project (PRODERMA) was, in general, well administered. 14. The input supply unit was established but never became the autono- mous self-sustaining cooperatively owned input supply organization envisaged at appraisal. The reasons for this were: low turnover; high costs; sub- sidized input prices; and weak cooperative structure. The unit, therefore, basically became a small division in the project organization (Maritime PCR, para. 4.16). The project credit agency (PCA) was semi-independent from both CNCA and the project. PCA recorded amounts of goods and services provided by PRODERMA and collected repayments on loans it did not provide; it was not seriously involved in reviewing credit applications. Credit operations have been limited and credit recovery poor, with losses being financed from 15/ A Togolese parastatal consulting firm. 16/ "The assumptions concerning yield increments at farmer level may be subject to substantial error and, as a consequence, so may the assump- tions regarding farmer response to innovations promoted under the proj- ect..... the risks involved in the project must be considered substantial (Maritime SAR para. 8.03). 17/ SORADs were replaced by "Organismes REgionaux de Promotion et de Production VivriZre in 1977 and later by "Directions Rfgionaux de 1'Animation Rurale et de l'Action Coop-rative"; these in turn were replaced by the present regional offices of the Ministry of Rural Road Development (DRDR). - 6 - PRODERMA working capital. PCA has ceased to exist. Building and equipment have been transferred to CNCA which has now established a branch office in Ts6vif. Both this project and the cocoa-coffee projects have established separate project credit agencies only partially linked to CNCA. Senior CNCA staff argue that, in retrospect, these arrangements are unsatisfactory. In future, CNCA wishes to either be fully involved in and responsible for credit operations or not involved at all; a sensible point of view. 15. Four expatriates filled senior positions during most of the project period, but Togolese staff successfully replaced these expatriates in late 1981 (PPAM, paras. 51-55). Counterparts training has been successful in most cases. PRODERMA's train'ng program for extension workers and farmers associ- ations has also been successfully implemented. 16. However, whatever was achieved in terms of strenLthening the proj- ect agency has been negated by the failure of the project and the related cut-off of funds. External funding is no longer available to the Maritime DRDR, and due to general Government financial constraints, local resources are extremely limited. Vehicles have been returned to the Government's over- all vehicle pool, substantial numbers of staff have had to be laid off and further reductions are expected in 1984. Under these circumstances, the Maritime DRDR will find itself extremely constrained to operate effectively. III. PROJECT SUMMARY - COTTON PROJECT Background and Project Formulation 17. In 1966, FED and FAC began providing substantial aid and technical assistance for a cotton program aimed at switching from the Mono variety to the Allen variety. In March 1974, a specialized agency - SOTOCO - was estab- lished to promote cotton development. An RMWA mission, in October 1975, identified the project, for which a feasibility study was being prepared at that time with FAC assistance. The feasibility study became available in early 1976. Project concept remained essentially unchanged during further project processing. Appraisal took place in October 1976 followed by nego- tiations on July 18-20, 1977. Outstanding issues were resolved (Cotton PCR, paras. 2.21-2.33) although, in retrospect, not always effectively.18! The Board approved the project on September 13, 1977 after minor debate.19/ 18/ For example, SOTOCO's financial situation remained difficult throughout the project period. 19/ Only one question was raised about the training of local staff and Togo- lization of positions staffed by expatriates; the reply that Togoliza- tion would be achieved in two years was, in retrospect, somewhat optimistic (PPAM, paras. 51-55). - 7 - The credit was signed on October 5, 1977 and became effective on March 17, 197820/ (Cotton PCR, para. 3.11). 18. The project, as approved, aimed at improving living conditions and benefiting more directly some 53,000 farm families. The project's components (Cotton PCR, paras. 2.41-2.42) can be grouped as follows: (a) improved and expanded cotton and foodcrop production, by providing supporting services, farm inputs, credit, as well as a seed multiplication system and animal traction; (b) infrastructure, by building and improving 700 km of feeder roads and providing for additional storage capacity;21/ and (c) project management support, by strengthening SOTOCO through pro- vision of infrastructure, vehicles, equipment and additional staff, as well as project monitoring and evaluation. Project Implementation and Impact 19. Despite substantial financial and organizational problems, SOTOCO managed to more than triple cotton production, reaching about 90% of appraisal objectives in terms of surface and production (Cotton PCR, paras. 5.11-5.14 and Annex 2, Table 1). As a result, incomes for a large number of small farmers increased substantially (Cotton PCR, para. 7.22 and Table 8). Although foodcrop development was a major objective, the project's main emphasis remained on cotton, an experience similar to other cotton-based rural development projects in Western Africa.22/ As a result, although very detailed information is available on cotton production, only imperfect in- formation is available on related foodcrop production. But, on the basis of available sources of information, foodcrop production is estimated to have reached only half the appraisal targets (Cotton PCR, paras. 5.21-5.25). 20. Yearly procurement of inputs caused serious problems on many occasions (Cotton PCR, para. 3.24), and inputs usually arrived late, and in 20/ The date was extended once from the originally scheduled date of January 3, 1978 to April 4, 1978 to allow for fulfillment of outstanding conditions of effectiveness. 21/ Well construction was not part of the project as this was financed con- currently under a FED village well construction program. It should be noted that the follow-on project (PPAM, Annex 4) again includes a well component. Ginning capacity expansion was not included under the proj- ect either, as this was a separate program. In fact, additional capacity was created with CCCE financing but somewhat later than ex- pected, creating problems in the interim (Cotton PCR, para. 3.43). 22/ See, for example, Ivory Coast: Cotton Areas Rural Development Project (Loan 1077-IVC), PPAR, OED Report No. 4568, dated June 20, 1983. more than one case, nearly too late.23/ Seed production was well organized for cotton but certain constraints had to be overcome for foodcrops (Cotton PCR, para. 5.24). The animal traction component initially progressed more quickly than expected, but physical progress subsequently suffered from weak and disorganized management (Cotton PCR, para. 3.52). Animal traction com- ponents were included in many agricultural development projects in Togo, each following different approaches and using different credit terms. Recently, however, a national effort has been made to establish a common approach which is also supported under RD II. 21. Feeder road construction started late, and about 70% of the appraisal construction target was achieved (Cotton PCR, para. 3.51). In addition, a substantial track minimum maintenance program, not envisaged at appraisal, was also supported by SOTOCO (Cotton PCR, para. 3.51; and PPAM, paras. 29-37). Other infrastructure constructed under the project included a new headquarters building in Atakpam-, a main workshop and garage in Lama Kara, regional offices and housing, sector offices and storage facilities, and substantial numbers of village stores. Vehicles and equipment were also provided. 22. The project has had a substantial impact on cotton production and thus rural incomes. However, the rate of return is reestimated at 11%, much below the appraisal forecast (32%). To a certain extent this is due to a shortfall in production, but the major factor is the recent changes in cotton price forecasts, resulting in much lower forecasted cotton prices than those prevailing at appraisal (Cotton PCR, paras. 7.31-7.34). Project Administration 23. Project implementation was the responsibility of SOTOCO. The audit believes that in the face of severe difficulties SOTOCO did, in general, a credible job in managing the project. Initially, organization and management left much to be desired: there was a high turnover of expatriate staff and the field service was judged to be far from effective. SOTOCO also suffered the usual start-up delays which, for example, affected the feeder road compo- nent and the flow of disbursements. It also took substantial time before the technical assistance contract between CFDT and SOTOCO was signed. From the beginning, efforts got underway to improve SOTOCO's field service. As a result, substantial improvements in management, training and organization of the field service were achieved. It is probably SOTOCO's field organization more than anything else which is responsible for the substantial increase in cotton production. 24. However, the encouraging progress in the field was not matched by developments elsewhere, and by mid-1979, the project faced serious problems: (i) there was a crisis in project management and the working atmosphere among senior staff was strained; three expatriate staff had re- signed and the Director General had been removed from office; 23/ For example, only because drought delayed planting (by about one month) in 1980 was it possible to fertilize the majority of the cotton area sown. - 9 - (ii) the project was in bad financial shape, basically due to a lack of Government contributions; and (iii) infrastructure to cope with increased cotton production was inadequate both in terms of transport and ginning facilities. These managerial and technical problems were resolved relatively soon, but the financial situation continued to worsen. As a result, the project was rated a problem project (status 3) in April 1980. At that time, the total shortfall in Government financing was estimated at CFAF 2.4 billion (US$11.6 million). 25. SOTOCO faced financial problems throughout the project period. This was due in part to delayed disbursements by the cofinanciers caused by the inability of SOTOCO's financial department to process withdrawal applications in time,24/ but mostly to substantial arrears in Government contributions. These arrears have been financed by SOTOCO by delaying pay- ments to major suppliers, and short-term borrowing. This entailed heavy costs in terms of interest or tenders with built-in markups to cover finan- cial costs of delayed payments. Also, during most of the period, the "barame"25/ did not fully cover SOTOCO's marketing costs, and the resulting losses had to be written off against SOTOCO's capital. Recently, however, improvements have been made and the latest "bar?me" is based on the principle that it should cover all SOTOCO costs except those financed by the external donors. As a result, SOTOCO has become much less dependent on transfers from the Government Treasury. In addition, arrangements have been made under RD II to increase SOTOCO's reserves.26! 26. Contrary to reasonably efficient project management practices, none of the monitoring and evaluation activities and arrangements as envisaged at appraisal were realized (Cotton PCR, paras. 3.53-3.54; PPAM, paras. 38-46). Second Phase Project 27/ 27. A second phase project (Second Rural Development Project in Cotton Areas) was approved by the Board on November 23, 1982. Credit 1302-TO, in the amount of SDR 21.7 million (at the time of approval equivalent to US$23.5 24/ Due to accounting inefficiencies and lack of appropriate liaison with Government departments in Lomi. Improvements have, however, subse- quently been made. 25/ "Barame is the yearly agreed cost structure between Government, OPAT, and SOTOCO which determines the price to be paid to SOTOCO for each kilo of cotton lint delivered to OPAT. 26/ Reserve increases would be obtained by SOTOCO's retaining farmer repay- ments against quantities of fertilizer financed by external sources. 27/ See Staff Appraisal Report, Second Rural Development Project in Cotton Areas, Report No. 3920-TO dated October 21, 1982. - 10 - million), was signed on February 18, 1983 and became effective on July 29, 1983. RD II would continue the expansion of cotton production but would put more emphasis on foodcrops than had RD I. Greater usage of animal traction would also be promoted. SOTOCO would remain the project executing agency and its efficiency would be upgraded through staff training and technical assis- tance. Associated infrastructure (feeder roads, village water supply, buildings) would also be provided. In addition to its production objectives, the project would aim to build up SOTOCO's managerial and field services capacity. The project would also provide support to Government's national seed multiplication program. 28. Project implementation started off efficiently. Most senior staff positions were filled quickly and detailed training plans for local staff were drawn up. First year production targets are expected to be met and prospects for satisfactory project implementation generally exist. IV. MAIN ISSUES A. Feeder Road Development 29. The Maritime Project road program as envisaged at appraisal "would be carried out over PY2-4 and would encompass 300 km, namely: upgrading 90 km of feeder to secondary roads, constructing 90 km of new feeder roads and rehabilitating 120 km of existing feeder roads. SORAD would also be equipped to maintain this network as well as existing roads" (Maritime SAR, Annex 10, para. 3). Actual road construction amounted to 166 km (55% of appraisal estimates) over the project period (Maritime PCR, para. 4.15), and main- tenance operations covered 279 km of roads (close to appaisal estimates). Only two roads (16 km) were constructed under contract__/, and most road construction and all maintenance was done by SORAD's engineering unit on force account. 30. Under the Cotton Project, "the main objective of the proposed feeder road component in the project area would be to start as quickly as possible to improve and subsequently maintain 694 km of feeder roads, in the most cost-efficient way consistent with quality. This can (in the short term) only be obtained by substantially reenforcing the engineering section of SOTOCO and commissioning SOTOCO to do the works on force account" (Cotton SAR, para. 3.15). Actual road construction amounted to 498 km (72% of appraisal estimate) over the project period (Cotton PCR, para. 3.51). The shortfall is explained by the late arrival of equipment and expatriate staff which caused substantial start-up delays (about two years). Thereafter, the road construction program was well executed. In addition, a minimum road maintenance program was set up for cotton-extraction roads covering, by the end of the project period, more than 1800 km annually. 28/ At appraisal it was envisaged that secondary roads (70 km) would be constructed by private contractors. - 11 - 31. However, serious issues arose at appraisal regarding long-term institutional responsibility for feeder road development. "The fact that SOTOCO, an agricultural development agency, would have to construct the required feeder roads points to the absence of a Governmental agency with overall responsibility for feeder road construction and maintenance. The result is a proliferation of various feeder road programs carried out by dif- ferent bodies" (Cotton SAR, para. 3.15). All three IDA-financed agricultural development projects included feeder road construction and maintenance. In addition to the two projects mentioned above, the first Cocoa-Coffee Develop- ment Project (Credit 503-TO) included construction of 50 km of new roads, rehabilitation of 110 km of feeder roads in pocr repair and maintenance of all roads built by the National Cocoa and Coffee Development Agency (SRCC).29/ The three projects together covered construction and rehabilita- tion of 1,154 km of feeder roads and their maintenance. 32. During preparation, SOTOCO had prepared a preliminary inventory of feeder roads totaling 1,442 km. At the time of appraisal, a clear strategy for feeder road development in Togo had been developed, consisting of the following elements: (i) a separate feeder road project would be appraised focusing mostly on organization and maintenance with the final goal of establishing a centralized efficient agency in this sector; (ii) the feeder road inventory of SOTOCO would be divided almost equally between the Feeder Roads Project and the Cotton Project; and (iii) when the feeder road components of the agricultural projects had been completed, the equipment purchased under these projects would be transferred to the feeder road agency.30! 33. The Feeder Roads Project (Credit 810-TO) was approved by the Board in May 1978. The project consisted of a four-year program of improvement and maintenance for 1,000 km of feeder roads, technical assistance to establish the Feeder Roads Agency and implement its program and short-term consultancies to study a follow-up feeder roads program (Project statistics are in Annex 3). 29/ Actual feeder roads construction of 196 km exceeded appraisal estimates by 22%, and an average of 74 km were maintained each year. See PPAR, Togo First Cocoa-Coffee Development Project (Credit 503-TO), OED Report No. 4552 dated June 15, 1983. 30/ Feeder Roads Project SAR, Report No. 1944-TO dated May 2, 1978, para. 3.04 and 2.09, but this was also made clear in the Cotton Project SAR (Annex 8 para. 10): "If during the project life Government institutes a centralized agency in this sector, then SOTOCO's feeder road department, with its equipment and staff, would be transferred to it, with the pro- vision that feeder roads in the project area would be built as a matter of priority within the agency's overall program." - 12 - 34. The project had made a promising start, albeit with some delays, but took a dramatic turn for the worse after two years. Due to outside interference, project management became very difficult. Following a period of informal suspension, management improved, but serious financial problems subsequently beset the project. At end 1982, the project was being wound down. Only about 530 km of roads have been constructed due to lower than expected output from the road brigades, institutional difficulties, and a shortage of local funds which was alleviated only in part by an increase in disbursement percentages. The uncor-mitted funds were, at that time, reserved for maintenance of feeder roads already constructed under the project. But lack of counterpart funding prevented SNPR from using these funds and carrying out this maintenance31/, and at that time it was envisaged not to extend the closing date and to cancel the undisbursed balance.32/ However, the latest development (1984) is that the closing date has actually been extended and that SNPR has been asked to resume feeder road construction (about 100 km) with the remaining credit funds. In the meantime, the future role of SNPR is being studied (PPAM, para. 37). 35. At present, two well managed feeder road units exist: one under SOTOCO and another under SRCC. Rather than handing over responsibility for feeder road development at the end of the project to SNPR as envisaged (PPAM, para. 32), these units are actually being strengthened under follow-on projects: - Second Cocoa and Coffee Development Project (Credit 945-TO). Under this project, 300 km of feeder roads would be constructed or rehabilitated. The project's maintenance program would cover the 160 km of roads constructed under the first project and those con- structed under the second project.33/ - Second Rural Development in Cotton Areas Project (Credit 1302-TO). Under this project, 950 km of feeder roads would be constructed and 2,100 km would receive periodic maintenance.34! 31/ In 1983, lack of counterpart funds did not permit SNPR to implement construction or maintenance under the Feeder Roads Project. However, under management of a newly appointed highway engineer seconded from the Public Works administration, SNPR has carried out the construction of about 100 km of feeder roads as subcontractor to SOTOCO. Regional staff point out that -such performance accomplished without help of technical assistance indicates that SNPR's technical capability is satisfactory.- 32/ Government has recently submitted a feasibility study for a possible second feeder roads project. This study is being reviewed within the framework of an overall review of institutional responsibilities for feeder roads development (PPAM, para. 37). 33/ Second Cocoa-Coffee Project SAR, Report No. 2408-TO dated June 4, 1979, para. 4.10. 34/ Second Cotton SAR, Report No. 3920-TO dated October 21, 1982, para. 3.24. - 13 - 36. By contrast, SNPR, while technically competent, is still not very effective as an institution, and it is being supported by somewhat unusual means. Thus, under the Second Cotton Project f or example, SOTOCO's feeder roads unit would do all periodic maintenance, but it would only construct half the roads and subcontract the other half to SNPR. This is inverse of what was intended: the national unit now acting as a subcontractor to and under the supervision of SOTOCO. 37. While there is no doubt about the effectiveness of the SOTOCO road unit (and for that matter the SRCC road unit), the long-term viability of these arrangements can be questioned. At present, the responsibility for feeder road construction and maintenance remains scattered35/; funding is almost exclusively provided by external donors. Without the artificially (externally)-funded subcontracting arrangements, SNPR would already have ceased to exist. Thus, physical implementation of feeder road construction (short-term) is well under control. However, in terms of institution-build- ing (long-term), achievements have been limited, and what will happen to the feeder road network after the project investment period(s) is unclear. Ultimately, however, a link must be made with the national road construction and maintenance strategy. The region, of course, realizes that this is important, and an inter-divisional working group has been established to study this situation and make proposals to rationalize the three main respon- sibilities for planning, constructing and maintaining feeder roads.36! B. Monitoring and Evaluation 38. Because of the riskier nature of the Maritime Project (PPAM, para. 12), the appraisal suggested building into the project an early warning system to detect problems which might develop and allow either corrective action or changes in relative emphasis. A Monitoring and Evaluation Unit (MEU) was to be created within the Ministry of Rural Development as a nucleus for a unit operating at the national level. The unit would determine base- line data and organize sampling as the project developed, prepare quarterly progress reports and also comprehensive annual reports. Under the project, provision was made for 30 man-months of consultant support to help design the monitoring program and assist in preparing guidelines on data collection and analysis (Maritime SAR, para. 4.25). 35/ For example, the IDA-appraised and supervised, and IFAD-financed Notse Rural Development Project (Report No. 3908-TO dated July 13, 1983) en- visaged reconstructing 195 km of feeder roads - to be subcontracted to SNPR and annually maintaining about 300 km of feeder roads - to be sub- contracted to a SOTOCO road brigade. 36/ Regional staff now have come to "believe it necessary to centralize the planning of feeder roads and to assure the proper integration of feeder roads planning with agricultural project and sector planning. There is also a need for rationalized execution and supervision of feeder roads construction and maintenance, although actual construction could either be carried out by one or more public agencies or by local contractors with due consideration to rationalization of equipment and spare parts." - 14 - 39. MEU was to be established by November 30, 1976. Before that date, the Minister of Rural Development appointed his top technical advisor as head of MEU with the intention of calling in a consultant to advise on the estab- lishment of the unit and to assist from time-to-time in the analysis of the data. IDA regarded this as an excellent choice "providing he can be relieved of most of his duties at the Ministry and providing he receives adequate support" (October 1976 Supervision BTO). In retrospect, it was overopti- mistic to assume that that would materialize, especially given the seniority status of the advisor's position (Maritime PCR, para. 4.12). 40. At that time (October/November 1976), the Cotton Project was appraised. The appraisal mission, and rightly so, linked its monitoring and evaluation proposals to the then existing situation. It was envisaged that MEU, created and supported under the Maritime Project, would have overall responsibility for Cotton Project monitoring and evaluation, and the Cotton Project provided only for a technical officer based in AtakpamE, a vehicle, equipment, and short-term consultancy services. 41. While the ideas about monitoring and evaluation formulated at appraisal were sound, they lacked detail. In practical terms, therefore, decision-making about designing the details of surveys - what to measure and in what format to report - was left for the project implementation period. In retrospect, this was a major reason for delays. This experience confirms other OED findings37/ which all have led to the conclusion that more -engineering" of M&E is required during preparation and appraisal. As the Maritime PCR rightly points out (para. 4.13), it was not consistent, given the risky nature of the project, to place prime importance on the need for farm data and to rely on M&E to obtain it early in the project, yet leave M&E to be organized and staff to become trained during implementation. 42. Both the Maritime PCR (paras. 4.12-4.13) and the Cotton PCR (paras. 3.53-3.54) clearly describe the failure of the monitoring and evalua- tion components of the two projects, and there is no reason to disagree with their conclusions. A Togolese economist was initially selected to become tle Cotton Project's technical officer, but after six months' training in Ibadan, he returned to another position in SOTOCO. The position of technical officer was left unfilled, and supervision reports gave only passing reference to the Cotton Project's monitoring and evaluation component. In any case, SOTOCO's data collection system, certainly as far as cotton production is concerned, is quite good, and for this it did not need an additional evaluation system. In retrospect, one can question if separate monitoring and evaluation func- tions were really needed. Project level monitoring is more properly the domain of management as is the preparation of quarterly and annual reports on the basis of existing management information systems.38/ In-depth farm sur- veys can, of course, always yield interesting results, but it is not quite 37/ See, for example, Rural Development Monitoring and Evaluation, an Assessment of Experience in Northeast Brazil, OED Report being finalized. 38/ Under RD II such reports are to be prepared by SOTOCO management on the basis of contributions by work unit heads. - 15 - clear if such surveys would have been required to measure the achievements of this particular project. In any case, the technical officer would not have been equipped to execute such surveys. 43. Under the Maritime Project, MEU was established in time (end 1976), but its program had not yet been established to any degree. Attempts by both the Bank and consultants to prepare a program for the unit have been frag- mented, incomplete and without adequate follow-up. The February 1979 super- vision mission recommended, therefore, that a team of consultants be hired, consisting of an evaluation unit specialist and a data processing specialist, to help improve the situation. In the meantime, MEU became fully staffed by end 1977 with the appointment of one supervisor, three controllers and 14 enumerators. On a trial basis, surveying began in 1978, mainly to train the enumerators. A consultant visited the project in February 1980, one year after a recommendation for such a visit had been made. The consultant recom- mended inter alia that a permanent expatriate specialist be hired to head MEU and also that the staff of DEESA39/ be merged with MEU. As it turned out, the technical advisor of the Ministry of Rural Development had very little time for MEU activities (PPAM, para. 39); by end July 1979, the Director of DEESA was given overall authority of MEU. The expatriate arrived in June 1980, but there were considerable initial administrative difficulties in arranging accommodations, transport and contract signing. In 1980, MEU car- ried out a survey on land use and the report on this survey was issued in May 1981. A similar survey was carried out in 1981; however, it did not include data on yields at different stages of crop intensification, simply because such a yield survey, in the absence of the expatriate who was on leave, was not executed. The contract of the expatriate was extended to assist in processing the data of this survey. He left Togo in the middle of 1982 and submitted his report on the second survey in late 1982. (It should be remem- bered that the project was substantially scaled down in June 1981 and completed a year later). 44. After the November 1980 supervision, the Bank agreed to finance a socio-economic survey to determine specific development actions that could be envisaged in the project area. This study was carried out in the second half of 1981 by SOTED with the support of an expatriate consultant. The survey covered the Western part of the project area, the Ave region, and the depleted coastal plateau. The main conclusions of the survey were that about 2,500 ha of bottomland rice could be developed in the Ave region, and about 3,000 ha of old coconut plantations, damaged by the Kaincope disease, could be replanted in the coastal zone. No further follow-up has taken place. 45. The end result of all these M&E activities is the production of two studies on geography, the land use surveys, and a socio-economic study gathering dust on the shelves. Unfortunately, there have been personality conflicts between expatriate and Togolese staff. In the beginning, PRODERMA 39/ DEESA - "Direction des Enqutes, Evaluation et Statistiques Agricoles" (Directorate of Surveys, Evaluation and Agricultural Statistics in the Ministry of Rural Development). - 16 - management was not very supportive of the evaluation function either, and by the time they needed data to justify any follow up actions it was too late. The absence of a full-time unit head was also a serious constraint. Unfor- tunately also, the expatriate head of MEU followed up on the proposal to merge DEESA's staff with MEU's. But the Bank had, at that time, changed its mind and recommended against such a merger. There has always been a conflict between a more project-oriented evaluation and a more regionally-oriented evaluation effort within a national framework. The Director of DEESA, in a case study prepared for a Bank-sponsored Regional Workshop on Monitoring and Evaluation of Rural Development Projects in Western Africa (Cameroon, March 1983), argued that a national agency should have overall responsibility for project evaluation. In fact, this was also the objective at appraisal when MEU was seen as the nucleus of a national evaluation system. Regretfully, more limited project objectives prevented institutional development in this direction. This trend is reinforced under the Second Cotton Project which supports a separate monitoring and evaluation unit. While the RD II apprais- al report makes it clear that to the extent possible cooperation with DEESA is to be achieved, it also indicates that the monitoring and evaluation unit, staffed with an expatriate specialist, would be independent and directly responsible to SOTOCO. By contrast, under the Bank-supervised and IFAD- financed Notse Rural Development Project, an M&E unit, under the supervision of an expatriate specialist, would be established and attached to DEESA. 46. In the audit's view, the appraisal objective of working towards a national evaluation capability was the correct approach, especially given the relatively small size of the country. Regretfully, project-oriented priori- ties prevented development in this direction. Such impact evaluation capa- city would of course not negate the need for within project-monitoring which is essentially management information. At present there are different approaches towards evaluation and different and sometimes overlapping insti- tutional arrangements under the many rural development projects being implemented in Togo, even between projects supervised by the Bank. In this situation the audit recommends that the Bank considers sending an M&E specialist to Togo to make an overall assessment of evaluation efforts from both the institutional and technical (survey techniques) points-of-view. Given present resource and staff constraints in Togo, any streamlining that can be achieved would be helpful. C. Institution Building 47. An assessment of institution building effects as a result of imple- menting these two projects shows mixed results. Compared to objectives established at appraisal, inverse institution building took place with regards to feeder road development (PPAM, paras. 29-37) and monitoring and evaluation (PPAM, paras. 38-46). As an institution, the Maritime DRDR is probably in a weaker situation than it has been in the last twenty years. Certainly whatever strengthening took place during the project period has now been negated (PPAM, para. 16). SOTOCO has become a more effective develop- ment institution, especially now that it has been put on sounder financial footing (PPAM, para. 25) and with the continued support provided under the Second Cotton Project (PPAM, paras. 27-28 and Annex 4). - 17 - 48. Institutional arrangements in the agricultural sector in Togo have been and are still extremely complex. There are two ministries responsible for agricultural development - the Ministries of Rural Development and Rural Equipment -- with many overlapping functions. In the field, there is an overlap between the regional offices (DRDRs) of the Ministry of Rural Development and the crop-specific specialized agencies. In the absence of a clear overall agricultural and rural development policy, rural development projects and rural services have been badly coordinated. Fragmentation, duplication and inefficiency have resulted and have become worse as the number of projects increased. In the field of agricultural research there is an almost total lack of coordination. Four ministries and a dozen agencies are involved. In addition, each project has invariably its own applied research component. Five French research institutes are also represented in Togo. Similarly, in the field of seed multiplication most projects have their own components. Lack of coordination has resulted in duplication of efforts; seed quality has sometimes been poor. Animal traction has been promoted under many projects each following a somewhat different approach. There was anarchy in the setting of terms and conditions upon which animals and equipment were provided to the farmers (PPAM, para. 20). 49. Even as the Cotton and Maritime Projects were prepared and appraised, Bank sector and economic work pointed to the difficult situation of the institutional set-up in the agricultural sector in Togo and the lack of an overall policy for the sector. However, although there were organizational and name changes during the period that these projects were implemented, the situation basically remained the same. As such, these projects were an insufficient instrument to effect policy and institutional changes. Recently, steps have been taken under the Second Technical Assistance Project (Credit 1270-TO)40/ for remedial action, the leverage to a large extent being a substantially worsened economic situation in Togo and a Bank-supported structural adjustment program. In March 1983, a rural development planning unit and a research coordination unit were set up in the Ministry of Rural Development, each staffed with a resident expatriate. Both units have started energetically, but it remains to be seen how effectively these units will be integrated in the rural development decision-making structure. In addition, a major rural development study by consultants, also financed under the Second Technical Assistance Project, is underway. This study would review recent evolution and sector potential and constraints, and evaluate institutional arrangements, support services and the policy environ- ment, as well as completed and ongoing projects. In retrospect, the fact that such a major study was needed at this time seems to indicate that the Bank's sector work in Togo has not been very effective in the past decade. 40/ See President's Report for a Second Technical Assistance Project, Report No. P-3345-TO dated May 28, 1982. - 18 - 50. It has often been said that projects are ineffective vehicles to ensure sector policy changes.411 Certainly the experience with these proj- ects seems to confirm this. However, projects are part of a wider Bank pro- gram. Given the absence of substantial improvements in sector policy-making, with the exception of price policy (see PPAM, paras. 56-61), the judgement must in fact be that, overall, the Bank program has been ineffective during the period that these projects were implemented. Various parts in a program should be mutually reinforcing. However, as experience has shown, this is not always easy to achieve. Under the structural adjustment program, severe constraints have been imposed on budgetary expenses and recruitment of Government staff. On the other hand, under the Second Cotton Project, sub- stantial strengthening of SOTOCO and the implementation of much expanded project activities is envisaged. In the end, SOTOCO might find itself con- strained at the micro level to achieve Bank-supported project objectives because of constraints, also Bank-supported, at the macro level. The Bank should be careful to see that its macro and micro objectives are harmonized to the extent possible. D. Togolization 51. Under the Maritime Project, "each internationally recruited staff would be assigned a Togolese deputy who would be trained to take over the responsibilities of the incumbent" (Maritime SAR, para. 5.08). However, the concept of designated counterparts was, at Bank suggestion (during the first supervision mission), modified; it was suggested that rather than designating counterparts from the beginning, Togolese staff be placed in line responsi- bilities in the various departments and that the proven most competent staff be subsequently selected for succession. Togolization in PRODERMA was accomplished by the end of the project period. In most cases, deputies had been given sufficient authority early on so as to gain experience while working for the project. In the case where this was not so, it led to substantial frustration and was the failure of the expatriate to promote staff development. 52. While Togolization was not specifically mentioned in the Cotton Project SAR, it was discussed during the Board meeting. Replying to a question as to how local staff would be trained and when they would take over from expatriates, Bank staff replied that training facilities existed in Togo and that expatriates would be replaced after a two-year period. This was, in retrospect, an overoptimistic assumption. The appraisal envisaged inter- national recruitment for seven positions.42/ In fact, two department heads and five technicians were expatriate (Cotton PCR, para. 3.22). Under the Cotton Project, no Togolization took place; rather, one could almost argue the reverse. 41/ See, for example, Eighth Annual Review of Project Performance Audit Re- sults, OED Report No. 4101 dated September 9, 1982. 42/ Technical Manager, Financial Manager, Deputy General Manager, Civil Engineer, Chief Mechanic, Head of the Feeder Road Unit, and Senior Training Officer. - 19 - 53. For the Second Cotton Project, internationally recruited staff were anticipated for 12 positions,43/ and signing of contracts involving at least six positions44/ was a condition of effectiveness. The reason for the larger number of expatriates needed under the Second Cotton Project is basically the increased complexity of the projects and the fact that trained manpower is still in short supply in Togo; furthermore, given salary and location constraints, SOTOCO is still finding it difficult to lure capable staff away from more renumerative jobs in the capital. But Togolization is again a major objective under the Second Project: "All internationally recruited positions would initially be in line management roles. However, a deputy (in some cases two), with clearly defined responsibilities, would work closely with each expatriate staff. The deputy staff would be expected to gradually take over responsibility during the latter part of the project. Prior to his eventual departure each expatriate staff member would be expected to assume an advisory role to facilitate this process" (RD II SAR, para. 4.08). 54. During the audit it has become evident that Togolization is of major concern to SOTOCO. Candidates have been identified for each position and training programs designed. However, in the audit's view, one should be careful not to make Togolization an objective per se. Most candidates at present have middle level management positions in SOTOCO. The training and promotion of such staff might deplete SOTOCO's already meager middle level management, which is the cornerstone of its successful field service. Furthermore, the replacement program in itself is meager: there is only a one-to-one relationship without alternatives. Thus if anything happens, such as an accident or a returning trainee taking another position, the replacement cannot take place. While the audit certainly does not argue against Togolization, it wishes to underline that realism is needed in this respect and that Togolization be considered in the overall framework of staff development and management needs at all levels in SOTOCO. 43/ Internationally recruited staff are anticipated for the following posi- tions: the Directors of the Agricultural Production Department, the Finance and Administration Department and the Industrial Department; the heads of the engineering division and the monitoring and evaluation unit; the chief of the extension service; the chief of the farmer groups and credit service; a cost accountant; a seed production agronomist for SOTOCO; an agronomist for the national seed production component; and two workshop managers. This total of 12 positions would aggregate to 42 man-years plus consultancy visits amounting to 22 months for the animal traction component and 24 man-months of consultancy services for the preparation of a possible third phase project and for other needs that may materialize during project execution (RD II SAR, para. 4.05). 44/ The first six mentioned in footnote 43 above. - 20 - 55. The experience with these projects has also shown the advantages of technical assistance umbrella contracts which cover recruitment for all positions, rather than a multitude of contracts with individuals. Even with the high turnover of expatriate personnel in certain years, arranging for replacements did not prove to be a major or time-consuming complication. As a result, RD II provides for the same arrangements. "The resident expatriate staff would be recruited if possible through a single contract with a consul- tancy firm or group in order to benefit from the back-up services provided under such an arrangement" (RD II SAR, para 4.06). E. Input Subsidies and Price Policy 56 At the time of appraisal of both projects the level of input sub- sidies was of some concern to IDA. At that time, fertilizers were subsidized at almost 75% of costs and insecticides at 50% of cost. Togo's policy was based on an infant industry argument for promoting modern agriculture through subsidization of inputs. Furthermore, the amount of inputs used was relatively small and concentrated on cash crops with export revenues more than covering the costs of these subsidies. 57. The Maritime Project pre-appraisal mission, therefore recommended that the issue of subsidization not be pursued at that time, but subse- quently, during the review process, it was strongly suggested that the issue of subsidization be pursued. Consequently, it was decided that during nego- tiations assurances should be obtained regarding a gradual decrease of these subsidies and annual consultations about progress in this matter. The draft legal documents reflected IDA's intention for the Borrower to follow a price policy aimed at progressively reflecting the full cost of farm inputs in the price of such inputs to farmers in accordance with a timetable satisfactory to IDA. The Togolese delegation rejected these proposals and instead a con- sultation clause was agreed to.45/ However, at the same time, "the Associa- tion indicated, and the Togolese delegation noted, that in connection with the proposed Cotton Development Project, a timetable for the progressive reduction of input subsidies would have to be discussed during the prepara- tory stages and agreed upon no later than during negotiations for that project.-46! 58. Therefore, the issue had now to be resolved in the framework of the processing of the Cotton Project, but in the end a timetable was also not agreed to. Discussions during preparation and appraisal of the Cotton Proj- ect with the Togolese Government were rather inconclusive. "The Decision Meeting therefore agreed that the Association should write a letter to Government reminding it of IDA's insistence on a plan for the discontinuation 45/ Development Credit Agreement, Section 4.04: "For the purpose of fol- lowing a price policy aiming at progressively reflecting the full cost of farm inputs, such as fertilizers and pesticides, in the prices charged to farmers, the Borrower shall not less than once a year ex- change views with the Association on prices applicable." 46/ Official Minutes of Negotiations, dated May 6, 1976. - 21 - of the subsidies and inquiring about Governments' intentions. In the mean- time. tf. mission would work out a scheme for both decreasing input subsidies antd increasing producer prices to be discussed before negotiations. OPS s:;ff indicated that OPS would insist that the presentation by Government of a reasonable plan for phasing out of subsidies should be made a condition of negotiations of the proposed credit." (Decision memorandum, December 17, 1983). However, the appraisal mission clearly did not favor requiring a rigid timetable for the elimination of the subsidies as a condition of the project. It argued that 4n the case of cotton, lower than market prices for inputs did not result in a net subsidy to producers as the cost of these subsidies was more than compensated by Government-controlled producer prices being kept lower than world market prices. After further discussion of the issue, the decision meeting position was scaled down to simply sending a letter to Government reminding them of the discussions during the Maritime Project negotiations and asking their intentions as to future levels of input subsidization. Subsequently, some thought was given to requiring a timetable as a condition of Board presentation, but this was also dropped.47! Finally, as under the Maritime Project, only an assurance was obtained on annual consultations.48/ 59. During supervision of the Maritime Project, only scant attention has been paid to input/output pricing issues and the related annual consulta- tions. Much more attention has been given to these issues during supervision of the Cotton Project, (see PPAR Table 5). Project records and files do not indicate that the once yearly exchanges with the Association in fact took place. Neither was IDA successful during the period that these projects were implemented in progressively reflecting the full cost of farm inputs, such as fertilizer and pesticides, in the prices charged to farmers. (see PPAR Table 6). In the formal sense, therefore, these price consultation clauses have not been adhered to. Under the First Cocoa and Coffee Project (Credit 503-TO), a similar non-adherence occurred.49/ In fact, this experience as 47/ Two other suggested conditions of Board presentation also disappeared before Loan Committee review. The modification of SOTOCO?s statutes to extend its activities to foodcrops in cotton areas became an assurance (first to be implemented by December 1978, later within a reasonable time limit). An increase in SOTOCO's equity became a CFAF 275 million payment by Government to SOTOCO as a condition of effectiveness. 48/ Development Credit Agreement, Section 4.01: "For the purpose of fol- lowing a price policy aiming at progressively reflecting the full cost of farm inputs such as fertilizers and insecticides in the prices charged to farmers for said inputs, the Borrower shall at least once a year, exchange views with the Association on such costs and p7ices. 49/ "Government never adhered to Section 4.05 of the Credit Agre,!ment which provided for annual consultations between the Bank and Goverument prior to the setting of farmgate prices for coffee/cocoa" PPAR, IED Report No. 4552, dated June 15, 1983. - 22 - well as other OED findings50/ confirm that such price consultation clauses have generally not been very successful. The Bank might therefore be well advised to carefully consider if such clauses should be inserted in future legal documents for specific agricultural projects. 60. However, this does not warrant the conclusion that the Bank was not concerned about input/output pricing policies; it was and still is. Early in the project period, substantial concern was already expressed on this issue. "An additional, and crucial, element is to obtain satisfactory farmgate prices for the different crops at the beginning of the marketing season. Our action in this field should be close and thorough. Raising the matter cursorily and at infrequent intervals, as the timing of our supervision missions necessitates, is simply not enough. We should even consider physically helping the Ministry of Planning to create and staff an economic analysis unit with specific responsibilities to make recom- mendations for farmgate and inputs prices and subsidies, which would be consistent with Togo's development goals and, whenever possible, with our recommendations of what we think is necessary for the successful outcome of our project, from the financial and economic point of view" (Internal Memo dated November 4, 1977). Substantial progress, however, was not achieved during the project implemen- tation period, despite substantial attention to pricing issues in the Bank's economic and sector work, the results of which were even published by the Bank.51/ More recently, Government has been more receptive to changing its input/output pricing policies. Consultants financed under the First Techni- cal Assistance Project (Credit 930-TO)52/ undertook an extensive analysis of agricultural prices in Togo in 1982/83:3/ Government agreed in 1982 to set up a price analysis and monitoring unit in its agricultural products market- ing agency (OPAT); this unit is now fully operational and in the process of updating the analysis prepared by the consultants. Furthermore, in the framework of the Structural Adjustment Credit (Credit 1365-TO),54/ Govern- ment has agreed again, as it had done earlier under the Second Cotton 50/ See, for example, Eighth Annual Review of Project Performance Audit Results. 51/ Agricultural Pricing in Togo, World Bank Staff Working Paper No. 467, July 1981. 52/ See also President's Report on a proposed credit to the Republic of Togo for a Technical Assistance Project, Report No. P-2503-TO dated May 31, 1979. 53/ Louis Berger International, Inc., Etude des Prix Agricoles, Rapport Final, June 1983. 54/ See also President's Report on a proposed Credit of SDR 36.9 million to the Republic of Togo for a Structural Adjustment Project, Report No. P-3534-TO dated April 26, 1983. - 23 - Project, to phase out subsidies on fertilizers for cotton and foodcrops over a five-year period beginning in 1983. In that year, as a first step Government increased the cotton fertilizer price from CFAF 45/kg to CFAF 60/kg..5/ 61. Thus, in the end Government and the Bank came to an agreement on the right objectives for input/output pricing, albeit much later and via a different route than originally expected. Undoubtedly the severe budgetary crisis in Togo has also been a major factor in influencing Government to make changes. F. Sustainability 62. At project completion, a reestimate was made of the project's costs and benefits, partly on the basis of actuals during the completed investment period and partly on the basis of expectations of future development. These implicit assumptions are therefore the major determining factors in sustain- ing the project's impact. What are these factors in the case of the Maritime and Cotton Projects? 63. The Maritime Project did not achieve its objectives as defined at appraisal; the project was a failure and basically, therefore, there is noth- ing to be sustained (PPAM, para. 16). A different project design might have worked (PPAM, para. 12), but the sustainability of such alternatives is a hypothetical question, hardly relevant for purposes of this audit. 63. The Cotton Project was basically successfully implemented and the question of its sustainability is therefore relevant, although substantially obscured by the second phase project. It will be this second phase project which will be the major reason why the First Cotton Project achievements will be sustained. Rate of return re-estimates have been done for a ten-year period only; given the continued coverage of this period under two projects there is a good chance that they can be achieved as estimated under these circumstances. Nevertheless, certain factors remain important: (i) the technical package for cotton development is well adapted and accepted and there should be no problem with the continued transfer of this technology; this is less so for foodcrop development, and extension will need to improve its advice regarding foodcrop pro- duction even to existing cotton farmers; (ii) remunerative producer prices are key to continuing farmer interest in cotton production. With the price analysis unit in OPAT now fully operational, there is good reason to believe that this will continue to be closely followed-up, especially important in view of Government's commitment to phase out fertilizer subsidies; 55/ Simultaneously it increased the cotton price from CFAF 65/kg to CFAF 75/kg. - 24 - (iii) any major differentiation in world market prices from those pro- jected for the ex-post project analysis will have a major effect on the viability of the project one way or the other; and (iv) continued institutional support is needed but is certainly assured under the present circumstances. 64. Therefore, at the present stage, there is little doubt that the Cotton Project's achievements can be sustained. Nevertheless, there is room for improvement in efficiency. SOTOCO should continuously be engaged in improving its effectiveness as a cotton organization; any such improvements can be translated into higher revenues for Governme.t or increased faragate prices or be used for building up reserves. Although projections indicate positive financial and economic margins on cotton production in Togo, they are not so large as to allow a less vigorous approach to efficiency improve- ments. - 25 - Table 1 PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) AND RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) TOGO LENDING PROGRAM (as of June 30, 1983) Loan/Credit Amount Credit Loan Approval Date Project Name Loan/Credit No. SDR US$ US$ - Highway Maintenance/a Cr. 131 - 3.70/c - Oct. 1968 - Second Highway/a Cr. 450 - 8.70 - Nov. 1973 - Cocoa-Coffee/a Cr. 503 - 6.00 - Jul. 1974 - Cimao Regional Clinker Ln. 1296 - - 3.50/b June 1976 - Maritime Region/a Cr. 638 - 9.50 - June 1976 - Third Highway/a Cr. 693 - 10.00 - March 1977 - Cotton Areas/a Cr. 741 - 14.00 - Sept. 1977 - Feeder Roads Cr. 810 - 5.80 - May 1978 - Technical Assistance Cr. 930 - 2.20 - June 1979 - Cocoa-Coffee II Cr. 945 - 14.00 - June 1979 - First Education Cr. 1018 - 11.00 - May 1980 - Fourth Highway Cr. 1139 16.40 20.00 /d - May 1981 - Phosphate Engineering Cr. 1169 4.80 5.70 /d - June 1981 - Power Engineering Cr. 1190 1.80 1.98 /d - Nov. 1981 - Second Technical Assistance Cr. 1270 3.10 3.50 /d - June 1982 - Cotton Areas II Cr. 1302 21.70 23.50 7 - Nov. 1982 - Cimao Restructuring Cr. 1326 5.30 5.70 /d - Feb. 1983 - Lome Water Supply Cr. 1354 10.90 12.0o /d - May 1983 - Structural Adjustment Cr. 1365 36.9 O/e 40.00 /d - May 1983 197.28 3.50 /a Completed and audited project. /b In addition a US$49.5 million loan was made to CIMAO, jointly and severally guaranteed by Ghana, Ivory Coast and Togo. /c Including exchange rate adjustments, credit amounted to US$4.32 million. 7d US$ equivalent at time of approval. /e Of which SDR 18.45 from regular IDA funds and SDR 18.45 from the IDA special fund. Tablo 2 sEMr fgpnC Ano1 nEu T1 MARIT ma L acA. ughå=a r pacCmT (CEDT 6-ro) am EAl DhlVU··. P~BZUU IN C01208 AEA1 (CE 1T 741-TO) SCI2I11~Lm~SU TO T=11 1. Pirat Coce-Coffee Dlk peet Prwet Credic 50-:. USS6.0 S1li . & - Jo! 23. 1974 Apprelel Ieport 352-TO. 06/26/74 v17 dls=ereed: 9836.0 .1111ea s - Aug 06. 197^ Preaidec*e eport P-1481-T0. 07/10174 £ - 1~r 12. 1975 FPAR. OE lepert 4552. 06/15/83 CL - Dec 31, 1940 The project mo. co prmotc doelopemat of eodrm ---11b1er coco and coff.. prod.ct-o throgh .tr~egrheot~g of project Ie.tlrutone. Cr-en.3 o atemein aeta ad agricultural credit pereomet. ad to expond applid reearth. Abeet 6.500 --Ilold~r- ar. epected to b pro~ided vith teetal aIatr-- cr"ilt eed grante. Total Pteoteo--~eeld eo~et to 4.400 ha of eree. ad 4,000 he of coffeo. yteldiug at full proderion 4.400 t of coco. and 4.800 t of dry coffea bee~.. In addition 50 lm of feeder rod ~od be contreuted ad 110 k. roas rehbhiliated. C~oetAote iaald be eptoyed to recommod leproved e-eleerettve e~ructures and -mooag procedre. - wLL m waluete project lpect ad lrepare a followr-on project. Coffo platIge of 5.740 ha exceeded apprae 1 emtimt bet coco. platlge of 2.164 be fell re chem 5X belo. epprwtkl margets. AboC 14.000 -ellboldier. beefttted fra the credlt progra. h. road coeetruction/reabilltation co~pooeet elmo exceeded eppralal emtma-te. hral project coete ere omly 4.6Z oboe ettate* bet ----el productio racbed oely 1.200 t coca ad 3.500 t coff e . the ecoiatc rate of reurn he been recIculam d at 2m ee coupared to the appre~eal wetmate of 33C. The dileappttleg performae hae been aty due to tho ~meatiefmctory coco. develop- t. 2. ekriti. eMgim mural De~lopent Projct Credit 63-TO: USS9.5 in1-e A - Ja E0. 1976 Appreetal Eport 1t0-T0. 05121/76 DIsbermed: 8S7.75 illen 1 - Ja 16, 1976 freetident'* Mport P-1823-TO. 05/24/76 Cancelled: 1.75 Sillen £ - Oct 26, 1976 Prag, thle OED Teport CL - Dec 31. 1981 The project ~o to proIde farmer. with iopu.t, equipeot asd extenatom nd credit erricen to Improve .ed di~etlfy eplmd crop. develop lo1eod ralofed rlce ad oegetabee, and rehbtltate cocenete; *trengthen 90AD (FnSA ) ad etmblteb a credlt agency: contruct roS, wlle, dLapeeearee. Tillage arhete and store: carry ot agronole roeoerch. ämd Irrigation engiaeerlg stedy and propare Invtent proposalm for aroe to be cleared of o~cbocerclaele. The project felled to eet It beetc objectie to eny eeb.etattal degrea. The reetleated rate of rater= at coplotico ~o 1-42 compared to e ppraleel =tlte of 34. Soe adt~r coupomet of the project ~ere relatiely eoceefut: ~egotable prodectioo ad arkettg. reteutton dam for domestic umter eupply. coconut field trial., ad rerat dipa- earle.. 1at prtly du. to rai.fall etreeoe aS Margleal co~ditlo~e in the project ea, faernre fall to midoly adopt the far Ioput p.ckage propoeed der the project. Actuat Ireamen ta prodction ad ylelde ha~o been tonal. A euch, the experienco dezonstrated that wall>older farelng me etmaged by the project no not vinble. 3. mkral Deeloment Proj~ct in Cotton Are. Credlc 741-TO: US814.0 Sllen A - Sep 13, 1977 Appralsal eport 1490-TO. OB/26/77 DIsbered: S13.21 eilt1n - Oct 05. 1977 Preeidente. eport P-2112-TO. 08/31/77 Caelled: 1S8 0-79 illion e - akt 17, 1976 PFAX, thi~ OED Sport CL - Dec 31, 1982 The projoct mo. to ~eprovo living coditta in the projet area en a broad otal. by eap~edte cotton productiof (20.000 be) and fod erop productIo (maen 13,700 ha, morghau 7500 ha. gro-aoe 2.700 ha) to be acbierod throagh stre"sth~e 10g of the exocuttog egeecy* he.dqaarter. ad field mervice; proelom et Ieproed seda, fertillzer aod I~eecticld=a ad marotlag ~erlcoe, Ieproed ned infrantructure (700 be of feod r rea), and Introdctlon of animal trection en a pllot beta. Dm, to Cormat-e fiuancial proble the project did not teitially recetme on tim. all tho ede necemmry to curry ot Ito ativltie. efflrceotly. Dempito them. diffIcullIes it organe»d lt field egente effictently and they ecceeded in pro~ldeng eceeeary service. to frmar.e. Ae a roemt, corten predet ta roo etroogly derlag the projert ipieapl etca period and reched 9S of ite objecttvoe. SOTOCO no Ies macmeaeul te food crop pr~eotloe. Fer=er. ceob lem- and Cooer~et foreign cchanga earonige eo iocre--ed «e anLIcipated. Tho EK U tae tited et Z4E agaet the appratmel reportC 32Z. the project hae butlt up en offectiro onocting agrn cepable of dollverteg the ervicee rered uLth' the help of a etrong expatriae eam. Deceue of the tec~«clt and flaelal support It vecelvod. the effticeneee wIth ~hich it carrie, out tre acttvitles and flaally the gotreee tt hem achievd. 50TOCO be. beca5h a poerful age~cy mithin the agriculturel institutloal frame~rt. 4. Secod Con-Coffe. Delopment Project Crodit 945-TO: us3514.00 ettlion A • Jun 28, 1979 Appreel eport 240T2. 06)04/79 Diembred (10/30183): MSS 3.76 eti. - Jul 23, 1979 ProeldenC Sport P-2521-TO. 06/01/79 E - Jul 11, 1980 CL- Doc 31. 1985 Replantteg aboet 7.500 ha of coffoe and 4,000 ha of cocoa by eo 17.000 emallaolder farm.re. ttrodertag Improvod food- crope, pro~idiag plantala earkatiog feciltte, conducing applied roarch, co~.trctl.g abet 300 be of feeder ned., catructilg lou-coet etorage epace tn village. .tregthenteg the project exceting egency (SoEC) e bd 1 -aerng etadtes. Coffea replantinga cootlnue et leele hlgher thao appratal eetie.tea bot the 192/83 dro~ght *erlously eftected plent- Ing. The cocoa replantIng compomet han failed. Alternetive cocom rebobilltation progreom are techeicelly feaethlo, bot floau~ally attractio and constrained by loglatlcal probleme. Creditr rocorry for both coco. ad coffe. have forther decliaod to lovet ia the 29 range. the edequecy of prire. need. fartber terea. SM la m aofficlautly -eeoyge orgenliation bot little be. beoen done eo far to loet. and trata Togolee contrperte. 5. Socood Exral Dovlopment Project in Cottan Aro Credit 1302-TO SDR 21.70 ilion A - Nov 23. 1982 Apprateal Eport 320-TO, 10/21/82 <eqevalet et epproval to M5523.50 milleon) 5 - Yeb 1M, 1983 Preeldent'. report P-33W-TO. 11/04/62 DIbureed (10/30/83) SPi 0.71 dilln E - .11 2. 1983 (equvalleer et dehurecent US 0.75 Ufleon) CL - ov 01, 19M Upgradig S0T~C. the project'. eemcutieg gency. ProvIdieg farme with tecbtcel laetece end Iepute to expaed cot- ta prod=ctioa d food crope (aioe. eorghm. gronedeute, peddy as c~' p~s); eultiplcatin ad ditributtion of lprored ftd crop eede nupport to notooaet ed mltiplictio. progra; applied remearch en cotmn, fte crope ad farm eytem, eaäl helth and pronotten of o~ traction (1125 palre); reconstruction of eo 950 km end periodlc elute- nance of 2.100 k of foedr roede; contructien of 400 village rolls. The firet project h- letd a eolid hage for the coatiaecion of epeowtoa t cotton produCtion, ad cottoe eeed proder- tio l at proeet over apprais~l etmat.e. aoor recorded food trop prodctrion te behId appreleal wtLutem - ede are to hort eppty e me m varliet. Introduced are mnaeiteble for local use. At the beg onlag of the n@coed project SO0CO bad been reorgaeixed ta four deprtmencr and a now =o compeet -~g-ee t te.. par la plac-. So far the project hab mode e good ert. - 27 - Table 3 PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) AND RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) AUDIT REPORTS ISSUED Date of Board Distribution Report No. Report Title/Loan No. See. Memo Date A. Project Performance Report Issued 1649 06/24/77 Highway Maintenance 77-527 06/24/73 Credit 131 3987 06/30/82 Second Highway 82-651 07/15/82 Credit 450 4552 06/15/83 Cocoa-Coffee 83-629 06/21/83 Credit 503 4904 01/23/84 Third Highway/a 84-67 01/31/84 Credit 693 B. Project Performance Reports Under Preparation Maritime Region Rural ) Development (Credit 638) ) ) this report Cotton Rural Development ) (Credit 741) ) /a Pass-Through PCR. -28 - Table 4 PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) AND RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) Feeder Road Construction/Improvement in Togo Project Name Feeder Road Construction Planned Actual Achieved Cocoa/Coffee I ( Credit 503) 160 196 Maratime (Credit 638) 300 166 Cotton (Credit 741) 694 498 Feederroads (Credit 801) 1,000 530 Cocoa/Coffee II (Credit 945) 300 -/a Cotton II (Credit 1302) 950 -/a /a Project implementation underway. -29 - Table 5 PROJECT PERFORMANCE AUDIT MEMORANDUM TOGO MARITIME REGIONAL RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) AND RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) COTTON PROJECT, SUPERVISION REPORTING ON PRICING Status Report Supervision Legal Covenants Other References (Section 4.01) 1. 10/77 n.a. mentions possibility of cotton price increase 2. 04/78 n.a. mentions that cotton price has been increased 3. 10/78 being met, no reference no action needed 4. 05/79 being met, suggests further price no action needed increase 5. 09/79 being met, no reference no action needed 6. 03/80 being met, makes detailed calcula- no action needed tions of the amounts of subsidies involved in producing cotton 7. 11/80 condition met, high subsidies are heavy heavy input burden; suggestion analysis subsidies; letter of this to be made by to Borrower Government 8. 03/81 (condition met)/a agricultural input and output prices are being examined; input prices have been increased but there is still a 50Z subsidy /a Implicitly; SPV Report only lists conditions not being met. - 30 - Table 6 PROJECT PERFORMANCE AUDIT MEMORANDUM TOGO MARITIME REGIONAL RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) AND RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) COTTON INPUT-OUTPUT PRICES Year Cotton Fertilizer 1977/78 60 15 1978/79 60 15 1979/80 60 15 1980/81 60 15 1981/82/a 65/b 45/c 1982/83 65 45 1983/84 75 60 /a Last project year. 7'? Tripling fertilizer prices announced in March 1981. /c Cotton price increase announced in December 1981. - 31 - ANNEX I PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) PROJECT DATA Table 1. Infrastructural Development Table 2. Project Staffing Table 3. Extension Effort Table 4. Farmers and Associations Reached Table 5. Coconut Plantations Table 6. Road Construction and Maintenance Table 7. Wells Constructed Chart Project Organigram at Appraisal Chart Project Organigram at Present - 32 - Annex 1 Table 1 PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) INFRASTRUCTURAL DEVELOPMENT Component Appraisal Target Actual 1. Roads 300 km 166km/a 2. Dispensaries 15 4/b 3. Wells 20 6 4. Earthendams 2 1/c 5. Village Stores 150 187w 6. Village Markets 20 none /a In addition, 279 km were reportedly maintained. lb Now operated by Ministry of Health. /c In addition, a small dam was built for an experimental animal traction station. /d Instead of village production stores, input supply stores have been constructed (construction of three more financed by FAO). - 33 - ANNEX I Table 2 PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) PROJECT STAFFING 1979 1980 1981 1982 1983 Senior Staff 40 40 39 40 39 Other Staff 55 55 50 49 49 Extension Workers 115 107 108 65 63 Support Staff Budget/a 37 42 70 17 16 PRODERMA b 136 166 108 65 33 383 410 375 236 200 la Financed from internal budget. /b Financed from project funds (FAC continued to provide some funding after IDA disbursements stopped). PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) Extension Effort Food Crops No. of No. of Extension Farmers Surface Improved Surface Under Advice Fertilizer Year Sectors Sub-Sectors Workers Reached Under Advice Practices Rice Cotton Distribution ------------------------ ha --------------------- (tons) 1976 5 25 122 - - - - - - 1977 5 25 122 7,052 6,890 1,273 873 232 500 m 1978 5 25 129 8,209 10,577 2,817 452 336 682 1979 5 25 186 9,727 11,336 1,884 1,086 708 700 1980 5 25 136 10,359 14,299 2,201 900 740 800 1981 5 25 141 11,740 16,716 3,192 754 460 682 (1982) (13,942) (21,984) (722) (852) (1,072) Source: PRODERMA Rapport de Synthese - 35 - ANNEX 1 Table 4 PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) Farmers and Associations Reached Number of Farmer Groups Number of Farmers Year Pie-Cooperation Cooperatives Total Organized Individuals Total 1977 249 168 417 4,815 2,237 7,052 1978 277 184 461 4,967 3,242 8,209 1979 367 185 552 6,642 3,085 9,727 1980 402 205 607 6,299 4,060 10,359 1981 442 256 698 7,440 4,300 11,740 (1982) (288) (270) (558) (6,979) (6,963) (13,942) Source: PRODERMA Rapport de Synthese. PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) Coconut Plantations 1977 1978 1979 1980 1981 Total No ha No ha No ha No ha No ha No ha A. Smallholders Credit (Project) - - 16 27.4 3 4.1 - - 1 5.0 20 36.5 Credit (Caritas) - - 14 17.0 34 54.0 11 12.0 - - 59 83.0 Cash Sales 14 9.5 14 21.4 15 11.8 14 17.7 13 15.4 70 75.8 - --- - - 1 Total 14 9.5 44 65.8 52 69.9 25 29.7 14 20.4 149 195.3 B. Enterpreneurs Credit (Project) - - - - - - 3 8.9 1 5.8 4 14.7 Cash Sales 22 35.0 26 47.7 20 51.0 27 88.4 39 93.0 134 315.1 Total 22 35.0 26 47.7 20 51.0 30 97.3 40 98.8 138 329.8 C. Project Plantation - 3.5 - 6.5 - 8.9 - 8.9 - 8.2 - 36.0 D. Outsiders /a 6 3.5 17 12.0 21 9.5 18 19.2 23 4.8 85 39.0 TOTAL 42 51.5 87 132.0 93 139.3 73 145.1 77 132.2 372 300.1 /a Sales of coconut seedlings (cash only) to people from outside the region. a Source: PRODERMA Rapport de Synthese. - 37 - ANNEX 1 Table 6 PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) Road Construction and Maintenance Sector 1976/77 1978 1979 1980 1981 Total A. Construction TSEVI 37 - 5 5 - 47 TABLIGBO 10 11 19 12 - 52 VO 16 - - - - 16 ANEMO 4 21 11 10 5 51 TOTAL 67 32 35 27 5 166 B. Maintenance TSEVI - 19 19 19 41 98 TABLIGBO 16 16 17 12 8 69 VO - 16 16 10 10 52 ANEMO - 4 10 32 14 60 TOTAL 16 55 62 73 73 279 Source: PRODERMA Rapport de Synthese PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) Wells Constructed No. Name Depth Yield Situated Near Purpose (m) (mThour) 1. Agoenyive 156 100 Pumpsation Irrigation 2. Topkli /a 65 600/b Seed Farm Water Supply 3. Badougbe 106 10 Togoviele Road Water Supply 4. Vo Kossidome 65 12 In Front Primary School Water Supply 5. Mome Hounkpati 138 10 Behind Primary School Water Supply 6. Aqbodjemi 63 100 Trial Field Irrigation and Water Supply /a Very low yield; therefore no pump and reservoir inst-lled. /b This is expressed in liters/hour. 1r,I TOGO PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT Project Organization at Approlsol Techn-cal ar-.1 Maretng Aoncies (5OTOCO. 1T . -Rm- - P io MRnger ATOGOCNA TOGOGAN. OPAT etc ) stra ochnal Deecutrsa Engneenng A.dministroleveo ln ectinicof Dl,ectorIIs'En'~fn Clewt Ag9rCV II II I- Vegetobie Pice Speco!st WorksTrainjg FsTn Accounng ntenance I asoSco I Sub-Sectors I -----Admninistrat"v Control ---Technicac3 Supoor F*UdExtenSion--- Maetn L -_-_--_-_--_-_--_-_--_-_--_-__mlie Form FrnityWcW Bar* - 26306 PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (Credit 638-TO) Project Organigram at Present looæ .loroom ----------...----------.----.------- Pr,eInag, .CNCA So10C0, OPAT I ~DeputyPsoisetI Pro tion RuralEngin As~sont Asstnt Sector Sed Roduction DMsmn Agrcultira MechantohanDMsi n LOME, TSM tGO, Vegeabs DMon Pubile W~,s & Roads kr%mvnn! DMan \GA. ANEHO Coconut DM*ln MochanrcoaiWak~opDM~o Lraining DM~lon Z~Vnes L.-. I- -Suw Farer PROJECT PERFORMANCE AUDIT REPORT TOGO RURAL DEVELOPMENT PROJECT IN COTTON AREAS/ RAPPORT D'EVALUATION RETROSPECTIVE TOGO PROJET DE DEVELOPPEMENT RURAL DANS LES ZONES COTONNIERES Financial Returns to Cotton Sector/Coton Rentablite Financlere CFAF/Kg Fiber/ FCFA/K0 do Fibres Producer PcO, FPri au- Producteurs SSOTOCO Coits/ 900 -Couts SOTOCO OPAT Coos/ CoutsOPAT 800 - toan Repoyrrionti E flemboursoment do Profs Total Earnings/ inancief - - finocit SuVusl veu ota Financial Surplus/ Surplus Financier - \Slu r 700- Total Costs/ Couts Totaux - - Extefnat Financialsl Loan Repayrments/Remboursernents de Prets- Financoments Exteirnes OPAT Costs/ CoulsOPAT Export Earnings/ - DevenuslD'Exportation 400 - SOTOCO Costs Including/Cou's SOTOCO y 'ompris - - - Extension/Encodrement - Subsidy/Subvention - Morketing/Commercialisation 200 Production Subsidy/ Subvention a la Production 100 - Producer Price/ Prix aux Producteurs 0 1982-83 1984-85 1989-90 1994-95 Source LoUis fiwgw Interinationa. Los Prix Agricoes ou Togo Wed Bank - 26310 PROJECT PERFORMANCE AUDIT REPORT TOGO RURAL DEVELOPMENT PROJECT IN COTTON AREAS/ RAPPORT D'EVALUATION RETROSPECTIVE TOGO PROJET DE DEVELOPPEMENT RURAL DANS LES ZONES COTONNIERES Economic Returns to Cotton Sector/Coton Rentablilte Economique CFAF/Kg Fiber/ F CFA/K ;de Fibre Procducer Coirl/Couts cies Product"&r SOTOCO Costs/Couts 5O1OCO 8w OPAT Costs/Couts OPAI WSurplus ( # )o Dm elhcil (-)/ F 7SurpluS .) ou Deficit(C-) Economic Surplus/Deficit/ Surplus/Defic(-) S Total Costs/ Economlque \ Couts Totaux SupuDfci( 00- OPAT Costs/Couts OPAl 4 Export Earnings/----- ------- Revenus d'Exportotlon 500 - 400 -. 400 1SOTOCO Costs including Extension Subsidy Marketing/ Couts SOTOCO y Compris - - - Encodrement Subvention 300- Conninerciallsation 200- 100- Producer Costs/- Couts des Producteurs 1982-83 1984-88 1989-90 1994-95 Sorce Louis Borger Inteirriational Los Prix AgrceO Su Togo World Bank - 26309 TOGO RURAL DEVELOPMENT PROJECT IN COTTON AREAS Project Organigram at Appraisal Ministry of Rural Developwant Monitoring SOTOCO and General Manager Evaluation .0 - Unit Technical . Manager Finance and Commercial CI Engineering Administration and Marketing Training DMsio Region i and Mechanical Unit Plateaux Regions Centrale Region Kora Region Savanna Region sectors IL sectors sectors Sectors Feedew Roads Unit Sub-Sectors Sub-Sectors Sub-Sectors Sub-Sectors Extension Agennts Extenson Agents Extension Aents World Bank - 26308 TOGO SECOND RURAL DEVELOPMENT PROJECT IN COTON AREAS SOTOCO OrganIgram Administrative Council General Director Monitoring and Evaluation Unit Agricultural Production Finance and Administraton Industrial Department Marketing Department Department Department ZZ II SERVICES SERVICES SERVICES: SERVICES: ENGINEERING DMSION: Training & Extension: - General Accounting - Foctodes Management - Cotton Marketing - Roads - Extension - Analytical Accounting, - Transport Management - Tubewells - Training Management Information - Workshop & Garage - Buildings - Former Groups & Internal Audit - Credit - Inventory Management - Animal Traction - Procurement & Payments Research & Seed Production: - Personnel Management - Research - Seed Production SOTOCO Field Services Savanna Koro Central Plateau Region Region Region Region World Bank - 26307 - 45 - ANNEX 2 PROJECT PERFORMANCE AUDIT REPORT TOGO RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) Background Data Cotton Project Table 1. Cotton Production under the Project Table 2. Feeder Road Construction Table 3. Cotton Production in Togo Table 4. Financial Benefits from Cotton Production Table 5. Analysis of Value Added Chart Finance Returns to Cotton Sector Chart Economic Returns to Cotton Sector Chart Project Organigram at Appraisal Chart SOTOCO Organigram - 46- ANNEX 2 Table I PROJECT PERFORMANCE AUDIT REPORT TOGO RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) COTTON PRODUCTION UNDER THE PROJECT Appraisal Actual Number of Surface Production Number of Surface Production Year Farmers (ha) (tons) Farmers (ha) (tons) PYO 76/77 19,300 11,680 6,720 18,724 8,960 6,310 PY1 77/78 23,100 13,790 10,200 13,220 6,740 4,205/a PY2 78/79 29,500 17,560 14,160 25,972 15,867 13,512 PY3 79/80 35,700 21,600 18,770 40,696 26,310 20,030 PY4 80/81 44,400 26,360 23,940 46,171 29,130 23,862 PYS 81/82 53,000 31,660 29,210 38,092 23,840 21,201/b PY6 82/83/c 53,000 31,660 31,460 44,061 27,316 27,480 PY7 83/847 53,000 31,660 32,540 n.a. 30,188 28,397/a (29,200)/d (27,620)/d /a Drought affected production. 7b Increase in input prices. /c Interim year between First and Second Projects. Td Estimates for Second Cotton Project PY1. - 47 - ANNEX 2 Table 2 PROJECT PERFORMANCE AUDIT REPORT TOGO RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) FEEDER ROAD CONSTRUCTION PY2 PY3 PY4 PY5 Feeder Road Construction (km) 102 181 214 197 Feeder Road Maintenance (km) 110 102 283 497 Minimum Maintenance (km) - 700 1,500 2,000 - 48 - ANNEX 2 Table 3 PROJECT PERFORMANCE AUDIT REPORT TOGO RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) Cotton Production in Togo Year Surface Production Yield /a 1964/65 6,223 1965/66 7,893 - 1966/67 9,262 - 1967/68 - 9,897 - 1968/69 - 3,977 - 1969/70 - 4,973 - 1970/71 - 5,561 - 1971/72 26,093 7,065 640 1972/73 22,992 6,026 620 1973/74 26,671 8,786 720 1974/75 29,603 10,778 655 1975/76 31,274 9,772 486 1976/77 18,958 7,063 700 1977/78 11,740 4,537 624 1978/79 26,467 14,365 843 1979/80 26,310 20,374 761 1980/81 26,130 23,862 810 1981/82 29,480 21,201 886 1982/83 23,318 27,450 1,005 1983/84/b 30,188 28,397 - /a Bou variety only. /b Estimate. PROJECT PERFORMANCE AUDIT REPORT TOO RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) Financial Benefits from Cotton Production 1982/83 1984/85 1989/90 1994/95 Total Per Ton Total Per Ton Total Per Ton Total Per Ton (CFAF '000) Cotton Lint (CFAF '000) Cotton Lint (CFAF '000) Cotton lint (CFAF '000) Cotton Lint Production (ton) cotton 27 000 35 000 45 000 54 000 cotton lint 11 200 14 500 18 000 22 400 A - Producer price 1 754,2 156 027 2 271,1 156 627 2 928,9 156 627 3 508,4 156 627 B - SOTOCO costs 4 996,2 446 085 5 788,3 399 190 6 796,5 363 448 7 684,7 343 066 C - OPAT costa 751,5 67 094 957,8 66 056 1 206,8 64 535 1 428,1 63 753 D - Export revenues 6 467,9 577 488 9 554,4 658 922 12 481,1 667 439 14 842,8 662 625 E - Surplus (D - A - B - C) ()1 034,0 (-)92 318 537,2 37 049 1 548,9 82 829 2 221,6 99 179 F - Support of international aid agencies 1 856,4 165 750 1 856,4 128 026 - - - - G - Provision for debt repayment (-)357,6 (-)31 929 (-)409,0 (-)28 270 (-)532,8 (-)28 492 (-)532,8 (-)23 786 H - Surplus net (E + F - G) 464,8 41 503 1 984,6 136 868 1 016,1 54 337 1 688,8 75 393 Source: Louis Berger International Inc. Etudes des Peux Agricoles au Togo. PROJECT PERFORMANCE AUDIT REPORT TOGO RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) Analysis of Value Added /a (CFAF million and percentages) 1981/82 1977/78 Before Project Value % 1977/78 prices 1981/82 prices/b % Production Value 3,947 100 788 1,151 100 Inputs 1,384 35 295 431 37 Value Added 2,573 65 493 720 63 of which: Salaries 860 22 177 259 22 o Agricultural Salaries 229 6 43 63 5 Taxes less Subsidies 188 5 -128 -187 -16 Gross Revenue Farmers 845 21 349 509 44 Gross Revenue Enterprises 187 7 14 20 2 Gross Revenue SOTOCO 264 5 38 56 5 /a Source: Evaluation Economique de l'Activilf Cotton on Togo. 71 Assumes inflation at 10%/year. UlN - 51 - ANNEX 3 PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) AND RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) Feeder Road Project (Credit 810-TO) Statistics Table 1. Implementation Progress Table 2. Supervision Ratings - 52 - ANNEX 3 Table 1 PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) AND RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) Feeder Road Project (Credit 810-TO) Statistics Implementation Progress km of roads km of roads constructed or constructed or improved by improved by Date Public Work Feeder Road Unit Total (month/year) Plan 400 600 1,000 11/79 40 15 55 04/80 116 25 141 08/80 127 110/a 237 11/80 127 124/a 251 05/81 200 156/a 356 07/81 210 181/a 391 09/81 210 188/a 398 11/81 214 226/a 440 03/82 206/b 232/a 438 10/82 195/b 293/a 488 /a of which 28 km executed under contract by SOTEXHA (Socift6 Togolaise de 1'Exploitation de Materiel). /b Some construction which did not form part of the agreed program and was not funded under the credit deleted from the statistics. - 53 - ANNEX 3 Table 2 PROJECT PERFORMANCE AUDIT REPORT TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) AND RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) Feeder Road Project (Credit 810-TO) Statistics Supervision Ratings Mission Dates Status /a Trend lb Problems /c September 1978 2 2 F,M November 1978 2 1 F,M January 1979 2 2 M March 1979 2 1 M July 1979 1 2 M November 1979 1 2 M April 1980 3 3 P,M October/November 1980/d 3 1 P,M January 1981 - - -/e May 1981 3 1 M,P July 1981 2 1 M September 1981 2 1 M November 1981 2 1 M March 1982 2 3 M,F June 1982/f - - -/e September 1982 2 3 M,F February 1982 2 1 F October 1983 - /a 1 = problem-free or minor problems; 2 = moderate problems; and 3 = major problems /b 1 = improving; 2 = stationary; and 3 = deteriorating 7-c F = Financial; M = Managerial; P = Political /d Combined supervision report on two missions - one in October, one in November. /e No supervision report prepared, only BTO. 7T Until and including that mission supervised by RMWA -feederroad unit in Abidjan; afterwards supervised from headquarters. - 54 - ANNEX 4 Page 1 PROJECT PERPORMANCE AUDIT REPORT TOGO RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) Summary Description Second Phase 1, The above general objectives would be achieved through a five-year project comprising the following activities: (i) upgrading of key departments within SOTOCO including provision of technical assistance and vigorous recruitment and training (on the job and short-term training fellowships) of Togolese personnel. Such institutional strengthening would include upgrading of the financial department and the industrial (cotton marketing, process- ing and transport) department; (ii) strengthening of SOTOCO's extension service including further emphasis on polyvalent training, whole farm planning, refinement of the recently established training and visit system, establishment of small farmer groups for extension and cooperative activities, improved input supply and primary cotton marketing activities, and decentralization of responsibilities; (iii) multiplication and distribution of improved food crop seeds; (iv) applied research on cotton, food crops and farming systems; (v) encouragement of animal traction through training, animal health and credit for purchase of oxen and equipment; (vi) promotion of bio-degradable pesticides to reduce losses of tradi- tionally stored food crops; (vii) provision through credit or cash of farm inputs and equipment rela- tive to the above, including streamlining of distribution and accounting procedures; (viii) reconstruction of some 950 km of feeder roads and periodic main- tenance of some 2,100 km of existing feeder roads so as to provide basic access for extension, input supply and marketing activities; 1/ SAR, Second Rural Development Project on Cotton Areas Report No. 3920-TO dated October 21, 1982, para 3.02. - 5- ANNEX 4 Page 2 (ix) establishment of 400 village tubewells to support agricultural development in agronomically suitable areas where agricultural pro- ductivity is constrained by the difficulties encountered and sub- stantial time involved in obtaining water for household use; (x) additional storage capacity and minor infrastructure to increase annual ginning capacity of existing factories and provision of trucks for transport of inputs and cotton products; (xi) establishment of a project monitoring and evaluation unit; (xii) additional buildings (field stores, regional offices, workshops, houses for staff, expansion of the headquarters office); (xiii) consultancy services for the preparation of a possible third phase project; and (xiv) support to the Ministry of Rural Development's national seed multi- plication program including the creation of a small seed farm to maintain and produce breeders seed for food crops, a seed quality control unit attached to the seed farm to monitor the quality of seed produced at all multiplication stages in Togo and creation of a coordinating committee and secretariat to coordinate countrywide production. - 56 - ANNEX 5 Page 1 PROJECT PERFORMANCE AUDIT REPORT TOGO MARATIME RURAL DEVELOPMENT PROJECT RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 638-TO) AND (CREDIT 741-TO) Ex-Post Evaluation by Co-Financier Introduction 1. FAC has undertaken ex-post evaluations of both the Maritime and Cotton Projects. These evaluations were undertaken by teams of experts and included field visits for periods of about three to four weeks. Full ex-post evaluation reports were published in early 1983 and are available from OED. For ease of reference, short resumes are given below. Maritime Project 1/ 2. "The PRODERMA operation has resulted in physical and economic achievements that are marginal, reflecting poor design and poor management." 3. "Those who designed the operation planned to carry out a regional development program in a region that is extremely complex, given its diver- sity of production and processing activities and an economic environment dominated by urban and frontier phenomena, using exclusively a technical approach that consisted in splitting up rural production into as many sectoral components as there were products. The project therefore became, particularly from the farmers' standpoint, a mixture of mini-operations con- cerning individual products. Thus, it was destined to produce the results that this type of approach usually yields when it is badly handled, i.e., farmers do not follow the technical advice given to them." 4. "In view of the large number of mini-operations, it was not pos- sible to handle each one efficiently, hence the uneven results of the proj- ect. Other contributing factors were a very poor understanding of the far- mers' production systems, a traditional "top-to-bottom" research set-up, all of which leads one to questit-' this kind of project that is supposed to be integrated but is in fact not. It does not take into account the totality of the regional development issues which should, on the contrary, aim to inte- grate all human activities within a given region." 5. "It is not the regional approach to rural development that is to blame for the poor results, but rather the way in which it has been handled. I/ From "Evaluation Economique de l'Opfration Developpement Rural de la Rfgion Maritime au Togo" (January 1983). - 57 - ANNEX 5 Page 2 Cotton Project 2/ 6. "The overall picture of this project that emerges from the analysis is still one that raises many questions, however." 7. "Admittedly the project produces a large foreign exchange surplus and this is its only really tangible result; at the same time the Government account just breaks even, but its equilibrium seems extremely precarious, since it has been achieved only by freezing farmer income for the past four years. 8. "The economic effects of the distributed income should also be taken into account when ansessing this project; even so one must proceed with caution because of the fungibility of income; a policy to obtain monetary returns from food products by exporting them to neighboring countries would not fail to distribute substantial amounts of income either. Only the in- comes of employees can be taken into account with any certainty as incre- mental, non-replaceable incomes that can be attributed to the project; their volume is such that their effects are still very limited." 9. "From the point of view of the country's farming systems, the project does not seem to have had any marked effect. The level of distri- buted income per farm is low, and the extent to which this crop has been incorporated in the farming patterns is still marginal. One cannot help wondering to what extent cotton-growing only reflects an opportunity to these farmers which receive extension advice, an opportunity which will disappear if the extension agent leaves." 10. "The fact of the matter is that this project has had no structural impact on the socioeconomic environment. It has not led to the establishment of grower associations able to take over some of the work handled by SOTOCO, which at present can only solely guarantee the continuation of the innova- tions introduced and hence of rural development per se; admittedly such a development might lead to some weakening in the efficiency of the mechanism established, which SOTOCO cannot allow, since it is a prisoner of its goal of maximizing production." 11. "SOTOCO has indeed accomplished the primary task assigned to it, that of maximizing cotton production. It has not had any significant impact on rural development, but this was not part of its mission." 12. "Further prospects can be forecast on the basis of past results. The continuation of a high level of cotton production depends on the survival of the mechanism that made it possible, because it is hard to imagine how a different direction could be chosen at this point. From the point of view of the economic results it seems clear that cotton revenues cannot finance diversification because there is no surplus. If other project components were to be developed on a large scale they would have to rely on their own 2/ From "Evaluation Economique de l'Activit6 Coton au Togo" (April 1983). - 58 - ANNEX 5 Page 3 resources; where these would come from is not clear, if they were not to overburden the Government budget." 13. "In the final analysis SOTOCO has not had any major impact on Togo's development in terms of rural development or the Government's finan- cial situation." Audit Observations 14. The audit's finding and those of the above ex-post evaluation report on the Maritime Projest, are similar. There is no doubt that the project was a failure. To a large extent this was, in retrospect, the result of project design: a "traditional" rural development project was designed and implemented in a marginal region. In the absence of substantial responses to the technical packages introduced, activities had to be curtailed. Some relatively minor project activities were successfully implemented, but is is doubtful if these can be sustained in the face of severe budget constraints. Without any major donor funding, DRDR is being substantially scaled down both in terms of staffing and operating funds. After more than 15 years of foreign aid interventions in the Maritime region, this is a sad result (PPAM, paras. 2.13 and 2.18). 15. The French ex-post evaluation report on the Cotton "roject used the effects method in its economic analysis.3/ Basically, this method determines the effects a project has on the economic environment through an analysis of value added. This provides indications of how much value added goes to Government, to SOTOCO, to the farmers, etc. (PPAR, Annex 2, Table 3). Due to data constraints, this analysis had to be limited to only one year (1981/82). However, various developments have taken place since. While farmgate prices for cotton remained constant at CFAF 60/kilo seed cotton over the four-year period 1977/78-1980/81, resulting in a farm income decrease in real terms during that period, they were increased to CFAF 65 (1981/82- 1982/83) and more recently to CFAF 75. Price projections made by consultants also indicate that financial surpluses will be derived from the cotton sector in Togo (PPAR, Annex 2). 16. The audit's assessment of the Cotton Project is, therefore, more optimistic than the views presented in the FAC ev-post evaluation report. 3/ See, for example, Rpublique FranSaise, Ministare de la Coopfration, The Methodology of Planning, Manual of Economic Evaluation of Projects, The Effects Method, by Marc Chervel and Michel le Gall; and also, Rfpublique Frangaise, Ministere de la Coopfration, Mthodologie de la Planifica- tion, Analyses Critiques des Mfthodes d'Evaluation de Projets, Volumes I and II. - 59 - ATTACHMENT 1 PROJECT PERFORMANCE AUDIT MEMORANDUM TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) AND RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) COMMENTS FROM COFINANCIER UNOFFICIAL TRANSLATION Telex No. 091 Prom: Embassy of France, Lome, Togo To: World Bank, Washington Date: May 11, 1984 FOR MR. SHIV KAPUR (OEDOD) FROM THE FAC RESIDENT REPRESENTATIVE IN LOME (TOGO) THANK YOU FOR SENDING US THE PPAR ON PRODERMA AND SOTOCO. OBSERVATIONS AND JUDGEMENTS BY FAC DEPARTMENTS HAVE ALREADY BEEN REFLECTED IN TWO REPORTS WHICH HAVE BEEN TRANSMITTED TO THE BANK. SINCERELY YOURS, FAC RESIDENT MISSION, LOME DEGALLAIX OED Observation These reports have been summarized in Annex 5 and were consulted for the preparation of the PPAM. As mentioned in the Preface, they are available in OED. - 60 - ATTACHMENT 2 PROJECT PERFORMANCE AUDIT MEMORANDUM TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) AND RURAL DEVELOPMENT PROJECT IN COTTON APEAS (CREDIT 741-TO) COMMENTS FROM DEESA UNOFFICIAL TRANSLATION Telex No. 89650 From: Lome, Togo To: World Bank, Washington Date: May 30, 1984 FOR MR. SHIV KAPUR REPORT SATISFACTORY. REGARDS, GBEBLEWOO [Direction des EnquAtes, Evaluations et Statistiques Agricoles (DEESA) ] - 61 - ATTACHMENT 3 PROJECT PERFORMANCE AUDIT MEMORANDUM TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) AND RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) COMMENTS FROM CNCA UNOFFICIAL TRANSLATION Telex No. 611 From: Caisse Nationale de Credit Agricole du Togo (National Agricultural Credit Agency) To: Mr. Shiv Kapur, Director, Operations Evaluation Department Date: June 4, 1984 WITH REFERENCE PROJECT PERFORMANCE AUDIT REPORT OF TOGO MARITIME REGION RURAL DEVELOPMENT PROJECT (CREDIT 638-TO) AND RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT 741-TO) WE DO NOT HAVE ANY PARTICULAR OBSERVATIONS TO ADD. SINCERELY YOURS, DAUBAN DIRECTOR GENERAL, CNCA, TOGO C2- €¤ Ó - 63 - TOGO PROJECT COMPLETION REPORT MARITIME REGION RURAL DEVELOPMENT PROJECT Credit 638-TO August 15, 1983 WEST AFRICAN PROJECT DEPARTMENT Agricultural Division D 万が卼Jでなノrプイ ー乃フー - 65 - TOGO Project Completion Report Maritime Region Rural Development Project Credit 638-TO I. INTRODUCTION 1.01 The project was the Bank Group*s second lending operation in the agriculture sector. The first, the Cocoa/Coffee Development Project, Cr. 503- TO, US$6 million, 1974, was successful. 1.02 About 80% of the population of about 2-3 million depends on agricul- ture for a livelihood. The sector accounts for about one-third of both GDP and export earnings and is the most important in the country. However, following some years of growth in the period 1960-67, agriculture activity tended to stagnate through 1968-74, the years immediately prior to the project; production of major crops declined, exports levelled off, and the maintenance of an adequate supply of domestic foodstuffs was becoming a problem. Cocoa and coffee accounted for the bulk of agriculture exports, followed by cotton, groundnuts and copra. The food staples, yams, cassava, maize, rice and sorghum, accounted for 80% of agriculture output. 1.03 Thus the Third Develolvent Plan (1976-80) env4aaged substantial improvement in agriculture performance including rehabilitation of the cocoa/coffee subsectors, (Credit 503-TO), a major prograrn to increase cotton production, and a program to attain self-sufficiency in food crops in the eighties. The objectives were to be accomplished by increasing yields and expanding the area under crops. In support of them the Government had already provided incentives including increases in the producer prices for export crops, improved credit and support services, and the provision of subsidies, particularly for fertilizer. It was in this context that the project evolved. At the same time, increased urban demand caused food crop prices to rise and improved farmer incentives to grow food crops. II. AGRICULTURE SECTOR 2.01 From north to south Togo may be divided into five agriculture regions: (i) savannah characterized by irregular rainfall, afflicted by onchocerciasis in some areas but over-populated in areas free of it, producing mainly food crops; (ii) Kara region, partially mountainous, with adequate rainfall, but overpopulated with resultant depleted soil fertility, and with substantial areas affected by onchocerciasis; (iii) central, with little population because of a lack of drinking water; (iv) -Plateau area, with the best agriculture potential, the main coffee/ cocoa area, cotton and food - 66 - crops, and intensive maize cultivations; and (v) the coastal Maritime region, with an area of 6,000 sq. km, including the capital, Lome, and the project area. 2.02 The project encompassed the entire Maritime region which contains 800,000 people of which a quarter of a million lived in Lome. The region was and remains a net importer of food, amounting to some 50,000 tons annually, mainly maize and yams. It was held therefore that the demand for food should provide an adequate incentive to farmers to produce food crops. 2.03 The Maritime region is in turn, ecologically diversified. Its northern section experiences high rainfall and two crops can be grown annually. Sparsely populated, it has good agricultural potential despite limited soil fertility. The coastal plateau can also produce two crops in favorable seasons, however the second depends on the "second" rainy season, which is short and erratic. It produces mainly food crops and is highly populated. The depleted coastal plate&u is also an area of food crop production and is characterized by large population, marginal rainfall, over- cropping, and soil fertility depletion. In addition there are three river valleys with potential for irrigated crops and a depression crossing the coastal plateau with potential for rice. The region was broadly characterized by a food deficit, overpopulation and soil depletion, and coconut disease (Kaincope) which had drastically reduced coconut production, the only cash crop on the coast. Farmer support services were inadequate, as were the rural roads and water supply. 2.04 The Societes Regionales d'Am6nagement et de D'veloppement (SORADS) were responsible for agricultural development in the country's five adminis- trative regions. However they were weak and lacked adequate management, trained manpower, proper financing and a well defined programs of activity. Specialized agencies assumed responsibility for cash crop activities, for example, Sociite Togolaise du Coton (SOTOCO) for cotton. The Caisse Nationale du Credit Agricole (CNCA) was responsible for agricultural credit, channeled mainly through the SORADS. SORAD Maritime, responsible for agricultural development in the region relied on growers (farmers) associations to deliver support services. It had a weak and thinly spread extension service. It was also responsible for seed production. 2.05 The prime objectives of the project were thus to assist in the agricultural development of the Maritime region by improving and diversifying crop production and strengthening agricultural services for some 20,000 smallholders to whom the project package of inputs and services would be provided during 1976-81. The balance of the farmers in the region, about 40,000 would not be involved, except to the degree the demonstration effect of a successful project might have an impact on their cropping practices. - 67 - III. PROJECT FORMULATION 3.01 The project was identified by an RNWA (Regional Mission in West Africa) visit in 1973. A feasibility study, financed by UNDP, was carried out by consultants in 1974. An IDA pre-appraisal mission in June 1975 modified the composition of the project substantially because it considered the feasibility study inappropriate insofar as it contained a number of components which were not integrated within an overall project concept, included some elements unsuitable for IDA financing and proposed subsidies which were likely to be unpalatable. The modifications made by the mission, did employ however the basic data contained in the feasibility study. The mission added the coconut component to the project. Crop Programs 3.02 The project contained some fifteen components, five of which comprised individual crop developments including agricultural research and seed production in the different ecological areas of the region, vegetable and coconut components, roads program, water supply, storage and marketing, and dispensary components, and in the realm of studies, irrigation engineering, field trials, and a development study of an area freed of onchocerciasis. 3.03 In the uplands the project made available credit and extension services to expand the adoption of improved farm techniques already introduced by SORAD Maritime. In the savannah participants would be provided with cot.on seed, fertilizer and pesticide by the project to increase cotton production following techniques successfully introduced in conjunction with its rotation with maize and groundnuts. The project provided maize hybrid seed and fertilizer and like inputs for groundnuts. However, the main thrust of the project was in the coastal plateau. In this area of already high productivity per unit area for food crops, marginal rainfall and suffering soil depletion, SORAD had already been encouraging improved practices. Here the project would encourage farmers to use fertilizer, emphasizing maize during the long rainy season, and stress rotation with groundnuts and cowpeas, to combat soil depletion. 3.04 On the depleted soil of the coastal plateau the project would test minimum tillage and mulching techniques and, if viable, promote their exten- sion with 1,000 farmers during the last two years of project implementation. In the lowland, tractors would be employed to clear and plough new lands the cost of which would be borne by the 6,000 farmers selected by SORAD to cultivate rice. The farmers would be organized in growers associations, and provided with the necessary inputs. Extension and supply services were to be provided to 250 vegetable growers and two technical support centers created near Lome and in the Zio River Valley for demonstration purposes, extension training and variety testing. The existing vegetable center at Agoueve was to be rehabilitated including the replacement of diesel by electric pumps. In an attempt to rebuild the copra industry, 600 ha of coconut seedlings grown from hybrid varieties to resist Kaincope disease would be planted in a field trial for the benefit of 2,000 smallholders. - 68 - Infrastructure 3.05 The project provided for the construction of 90 km of new feeder roads, the upgrading of 90 km of feeder roads to secondary road quality, and the improvement of 120 km of feeder roads to improve access to crop areas and facilitate marketing. Domestic water supply, a major problem in much of the project area, was to be improved by the construction of 20 tubewells equipped with pumps and two small dams for water retention. Materials for the cons- truction of 150 village stores and of 20 village markets to store farm inputs and promote marketing of the products of the harvested crops was to be financed by the project, as was the provision of 15 six-room dispensaries to serve about 10,000 people within a short radius of them. The expansion of SORAD headquarters, construction of a training center and vehicles and other transport were to be financed. 3.06 The project was to be carried out by SORAD Maritime. The posts of Technical Director, Financial Director, Training Officer and Research Agronomist were to be financed by the project. Togolese deputies to these officials would be trained to take over the posts. Support services were to be strengthened with project assistance. About 143 extension works were to be employed in addition to the existing 122 to improve the extension worker per farm ratio to 1:150 for upland crops, 1:100 for rice and ecsonut and 1:40 for vegetables 1/. A training program was to be provided and implemented for existing lower level staff, including extension, and new staff of all levels. An Input Supply Unit would be created and have responsibility for establishing input needs, and procuring and distributing such supplies. 3.07 As set forth in the Appraisal Report, SORAD Maritime "activities included work with rice, coconuts and vegetables, rural roads maintenance and credit"' in addition to responsibility for "rural education to train leaders and develop farm organizations, land improvement, maintenance of dams, canals, etc., development of rural artisanal activities and regional organiza- tion of cooperatives, agriculture credit, crop extraction and marketing." Assurances were obtained that no new "non-project" activities would be undertaken in the region within the period, and SORAD would review its activities, including "non-project" ones by June 30, 1977, and discuss the findings with the Association. Through these measures, "by the end of implementation, the overall development program in the Maritime region should be satisfactorily oriented." 3.08 There was little basis for this last expectation since providing assistance and advice to the 20,000 farmers who were given the inputs package could not be expected to induce the remaining 40,000 farmers, who received neither advice nor the input package to adopt a similar cropping and input regime. Further the means of accomplishing the many objectives were too limited, or put another way, the project was over ambitious and too extensive in scope. Indeed, the project was questioned at the review stage. The then Central Projects Department raised the following points : 1/ In fact, the nimber of extension workers did not exceed 120 during implementation. - 69 - - the information on farm and family size was very thin; - the assumption that farmers would swallow the entire package of inputs in one bite was questionaile; - the information on farm models, budgets, areas, cash flows, and so forth, was confusing; - that field level ..taff would be unlikely to have the experience needed to do the ..ob, given that their duties looked more complex than those of the average extension worker in West Africa; - there was an air of expediency in creating an input supply unit attached to SORAD with the objective of enabling grower associations to commercially take over input supply; and - no feel was given for the probability of adverse variations in the parameters tested. Project Description 3.09 In summary, the project was intended to accomplish the following over 1976-81: - Improve and diversify production on 18,500 ha of upland crops; - Develop 3,000 ha of lowland rainfed rice and 25 ha of vegetables; - Rehabilitate irrigation facilities for 20 ha of existing vegetable production; - renovate 600 ha of smallholder coconut holdings; - Provide support services and infrastructure for 20,000 families. Cost Estimate 3.10 The estimated cost of the project was US$15.7 million including taxes of US$1.8 million, of which the foreign component was US$9.3 million or 67%, net of taxes. The costs included physical contingencies of 10% on equipment and farm costs and 15% on infrastructure, and price increases amounting to 26%, based on annual inflation of 11% in 1976 and 9% thereafter. Estimated and actual costs are given in Annex 1. Financing Plan 3.11 Fonds d'Aide et de Cooperation (FAC) France, co-financed the project with a grant of USS2.2 million equivalent, and Government and farmers provided the balance, as follows: - 70 - IDA FAC Government Farmers Total ---------US$ million equivalent--------- On-Farm Costs 0.40 0.10 0.28 0.33 1.11 Medium Term Credit 0.37 0.09 0.15 - 0.61 SORAD 2.84 0.83 0.81 - 4.48 CNCA 0.44 0.11 0.10 - 0.65 Infrastructure 2.25 0.42 1.13 - 3.80 Onchocerciasis Study 0.30 - - - 0.30 Other Studies 0.52 0.10 0.03 - 0.65 Contingencies 2.38 0.60 0.96 0.16 4.10 Total 9.50 2.25 3.46 0.49 15.70 Total (Net of Taxes) 9.50 2.25 1.68 0.49 13-92 Percentage 68 16 12 4 100 IDA Credit funds were transferred from Government to SORAD as a grant. Procurement 3.12 International competitive bidding in accordance with Bank guidelines was to be employed for goods exceeding US$50,000, involving aggregate con- tracts of US$3.4 million, while rural works, buildings and other goods were to be let after competitive bidding according to local procedures. The latter category of procurement met, as will be seen, with problems. Goals and Targets 3.13 The broad purpose of the project was to substantially increase agri- cultural production in the Maritime region, overcoming the problems of over- population, soil depletion, Kaincope disease, poor farmer support services and inadequate infrastructure. In the process it was hoped to overcome the problem of low income by roughly doubling net farm incomes. Direct benefits, in the form of increased crop production, were expected from an increase in the yield/ha in the order of 50% on existing crops, in some cases with an increase in area, and the establishment of two new crops, cotton and rice, over a sizeable area. The crop production of the 20,000 farmers was expected to increase from 60,000 to 80,000 tons, that is, by 20,000 tons. The targets and actual results are given in Table 1, para 4.22. The economic rate of return was estimated at 37% overall and, for the major components, food crops and cotton, was also about 37% Loan Negotiations 3.14 At negotiations in May 1977 the only issue was the subsidization of inputs, fertilizer and insecticides by Government, which met the bulk of their costs. The Association proposed a timetable for the graduated reduction of subsidies. The Government refused on the grounds that only 20% of the farmers enjoyed subsidies and it considered it necessary to introduce the other 80% to subsidies to encourage them to use such inputs. It was left that a timetable - 71 - for their reduction would be further considered in conjunction with the nego- tiations for the Togo Cotton Rural Development Project. Important Covenants 3.15 The following were the more important covenants in the Credit Agreement: - funds to purchase farm inputs to be lent to CNCA by Government at a 5.5% interest rate for a term of 15 years with 7 years grace; - establishment of a project account (in effect a revolving fund) for SORAD Maritime for prefinancing goods and services for the project; - establishment of a revolving fund for CNCA for the purchases of fertilizers and pesticides, and its annual reimbursement with an amount equal to the subsidy Government would have paid had it fully financed such inputs; - employment, and maintenance in office, of project officials (listed in para 3.06) with qualifications satisfactory to the Association, and appointment of qualified nationals to serve as deputies; - preparation by December 31, 1976, of a training program and its review by the Association; - design of a field trial program and arrangements for satisfactorily carrying it out; - an agreement by December 31, 1976 with Institut de Recherches pour les Huiles et 0leagineux (IRHO) covering the details and budget of the coconut trials program); - establishment within three months of the Credit Agreement date of an Input Supply Unit, and review thereafter by June 30, 1980; - establishment of a Project Monitoring and Evaluation Unit (ME) by November 30, 1976, and employment therefor of consultants satisfactory to the Association, and provision of a program within three months of their employment; - SORAD Maritime not to undertake additional activities beyond the scope of the project, and the review of SORAD's activities with the Association by June 30, 1977; - purchase of ten power tillers (for field trials of minimum tillage) and the reporting of their performance to the Association, and agreement on subsequent purchases and spare parts; - prior to construction of roads, wells and dams to seek Association approval for the annual detailed work programs, including design standards and, for wells and dams that operating and maintenance standards should be met under appropriate arrangements; - 72 - - SORAD Maritime to construct and equip the dispensaries, and staff, maintain and supply them; - for the purpose of following a price policy aimed at progressively reducing farm input subsidies the Borrower would exchange views with the Association on prices applicable each year; - the appointment cf the Administrative, Financial and Technical Directors, and the establishment of the project account and the Revolving Fund were conditions of effectiveness. 3.16 The Credit was approved by the Board on June 16, 1976 and became effective October 26, 1976. IV. PROJECT IMPLEMENTATION 4.01 The project was not successful and most targets were not achieved. In fact, well toward the end of the project it was realized that little or no increase in crop production was being obtained and with the agreement of the Borrower, a substantial amount of the Credit was cancelled. The project was a little unique in that the Association's perception of it through implementa- tation, until the limitations were recognized, was not especially negative. The following chronological description helps to understand why this was so. 4.02 During 1976, the first year of implementation the results were mixed. The filling of senior positions, procurement, training and some components of the project commenced promptly. The vegetable marketing and coconut compo- nents proceeded well, a credit unit was organized and the ME unit was said to have been organized and to have commenced. Management appeared reasonably adequate. However, the basic data on crop production and the area under cropping based on extension service figures (pre-project situation) was considered inadequate and/or unreliable. For example, it appeared that of the 1,200 ha which were supposed to be under rice cultivation, no more than 300 ha actually were. It was thought that better estimates could be obtained from the early project experience. There was doubt about the quality of the topographic survey made for irrigation works and of the suitability of the area selected to gravity irrigation development. It was found that some of the well sites proposed in the north were difficult geologically and that their boring would require drilling equipment and entail costs which were not justified. It was therefore planned to leave further well development in the area to FED (European Development Fund) which already had a well drilling program underway nearby. The study of development of the area freed of onchocerciasis was complicated by the prospective activities of FED and, instead, the use of the funds for studies of other areas freed of the disease was put under consideration. 4.03 Field trials on depleted soil were well managed and produced even- tually improved crop growing recommendations. The unit cost of a dispensary, estimated at CFAF 4 million (US$18,000) at appraisal, appeared to be about CFAF 14 million (US$60,000). Problems arose in the procurement of fertilizer - 73 - in amounts worth less than US$50,000. 4.04 In the spring of 1977 there was no short rainy season, a situation which resulted in a shortage of seed. The entire year saw such a severe drought that (project) rice output was negligible, seed production poor, and other crops averaged about 50% of normal expectations. The drought wiped out the impact of the project on crop production up to that point. The areas under crop cultivation were about 75% of appraisal estimates except for rice, 65%, and cotton, 30%. The experience with mechanization demonstrated that power tillers had a limited use for irrigation rice, 4-wheel tractors were better adapted to annual land preparation, and the high cost of mechanization did not justify its use for annual cultivation of rice. The thinking at this juncture was that the project ought mainly to promote drought resistant crops, groundnuts, cowpeas and scrghum, and to focus on the development of bunds in valleys and flood.plains for the growing uf rice. 4.05 The results of the drought, still experienced in early 1978, coupled with the absence of monitoring and evaluation work and weak management of extension services, made it impossible to reliably assess project impact. It was estimated that the project was reaching about 7,000 farmers but that aside from the vegetables and coconut components, it was in poor shape, it being necessary inter alia to import the bulk of the seed from other regions. With crop production at much lower levels than expected and credit based on voluntary repayment, farmers were falling far behind in credit repayment, and there was little incentive for them to become interested in the establishment of village storage facilities. In 1978, rainfall was adequate but unevenly spread, and the harvest produced mixed crop results. As a consequence of the large maize crop, maize, which had sold for CFAF 180 per kg prior to harvesting, cost as little as CFAF 15 per kg immediately thereafter. Planting targets and yields were reached for other crops except cotton and rice, for which about 35% of the planting target was attained, with an average rice yield but a low cotton one. 4.06 The following year, 1979, there was too much rain in some areas but crop production was generally good, although there was a suspicion that it was over-estimated. The seed and groundnut crops were poor, rice fair and yields of the other crops good. It was thought that the project had by this time reached 8,200 farmers versus 13,000 estimated at appraisal. It was toward the end of 1980 when it was realized that project technology was accomplishing little and that the risk of mounting a food crop project in such an uncertain and marginal climatic area had more than been proved on the basis of the following: - of about 15,000 ha under project control only 20% had been sown with improved seed and received fertilizer applications during 1980; - the bottomland rice area under planting was 1,000 ha compared with the PY4 expectation of 2,500 ha; while 520 had been developed under the project, the net increase was only 170 ha because much of the newly developed area under rice was destroyed by heavy rains; - seed production was highly variable and far less than the demand; - maize yields had been less, in general, than in pre-project years; - 74 - - only the coconut and vegetable programs were proceeding satisfacto- rily. 4.07 Following this a desk review of the project experience was made by the Association in February 1981, which concluded that while the infrastructu- re components had been moderately successful, the project had only a limited effect in terms of its the adoption by farmers, area put under cultivation and number of beneficiaries, and would fall much short of reaching the benefits estimated at appraisal. Further, that the more important negative factors were the lack of an appropriate technical package for food stuffs and the absence of cash crops and, to a lesser extent, managerial and technical deficiencies. Finally, the report observed that adverse climatic conditions through PY1 - PY4 had resulted in poor harvests which obviated any demonstra- tion effects. Cancellation of US$1.7 million of the Credit 4.08 In the circumstances the Association proposed, and Government agreed initially to terminate the crop production components as of June 30, 1981, and cancel the corresponding US$2.1 million from the Credit, the amount cancelled being changed subsequently to US$1.7 million. Government raised no objections to the Association preposal and evidently had similar feelings about the project's continued usefulness. Several smaller components were to continue, financed by funds left in the Credit for the purpose specifically: the Atti- Atouvou dam construction would be completed and seed production, applied research on food crops, maintenance of experimental coconut fields and technical assistance to cotton would be continued until December 1981. This agreement was made on the understanding that FAC would continue to provide full funding of applied research, but subsequently FAC was unable to do so. Consequently, to enable the research to continue, and to make a socio-economic study of the development potential of the Maritime Region, the Closing Date was extended to December 31, 1982 and the amount of the Credit cancelled reduced to US$1.7 million. This amount was cancelled from the Credit, February 23, 1982. The wisdom of terminating production in mid-1980 was further confirmed. During the first half of 1983 there was no rainy season and the climate once again was disastrous to the crops. Management and Organization 4.09 As the project progreseed it became apparent that most of the management and institutional activities were somewhat weak and ineffective. This was attributed in part to the Togolese Director General who, though technically and managerially competent, nevertheless exercised poor control and little discipline over activities and was somewhat removed from project administration. This may have been the case because the Director General was approaching retirement and tended to avoid situations which Eight be contro- versal or lead to personnel difficulties. Training 4.10 Construction of a training center was completed in 1977, and the original training program completed satisfactorily in 1978. At that time a national replaced the expati.iate who had been in charge. However, inasmuch as - 75 - farmer associations, cooperatives and extension agents were to receive training a new expatriate was hired for a year to layout and implement a program under the administration of the expatriate Technical Director. The program continued through 1980, during which 20 new extension workers were recruited and trained, and refresher courses given to sector chiefs and extension workers. About 30 embryonic farmers associations were trained and refresher courses given their officers. The training program also encouraged new associations of farmers with similar crop interests, for example, rice production, processing and marketing. Extension Services 4.11 Originally, the focus was on the orientation of extension staff inherited from SORAD and retraining of existing staff, emphasizing field measurement. Sector and sub-sector chiefs were trained as noted and technical and credit documentation explained. Improved forms for recording farm data were adopted at the suggestion of the Association, but when they were put in use it became apparent that many subsector chiefs were inadequate, and little use was actually made of the forms. Emphasis was then placed on the importance of properly using the information obtained. However, over time, the sector and sub-sector chiefs showed little willingness to complete the forms to avoid being put under the close control of the Technical Director. This was a manifestation of the failure of the Technical Director to hold monthly meetings with his sector-chiefs, which was in turn a reflection of his frustration with the Director General who had refused to take disciplinary action against sector-chiefs. This laxness was considered one of the main causes of the ineffectiveness of the extension service. Project Monitoring and Evaluation (ME) 4.12 The ME unit had a long history and was the subject of several initiatives and considerable Association assistance, but did not function effectively until close to project end, when it was reaching the point of usefulness. This was unfortunate as the initial lack of base line data and subsequent failure to develop it was an important weakness. ME, with three sections and eight enumerators was under the administration of the Technical Adviser to the Minister of Rural Development. The Technical Director was too occupied with other duties to provide real direction to ME, and consequently it languished. It was still not fully operational by 1979 when the Association advised the Borrower to hire an evaluation unit specialist with competence in data processing, which took over a year. An expatriate consultant took charge in June 1980 and organized activities were initiated, including the first base line survey on land use and crop production in the project area. Subsequently, a socio-economic survey was made in 1981. 4.13 It was not entirely consistent to place prime importance on the need for farm data and rely on ME to obtain it early in the project, yet leave ME to be organized and become trained during implementation. Following the first supervision mission in the fall of 1976, the Association studied the monitoring and evaluation situation in the light of the specific objectives and ecological limitations of the project and prepared a brief report which was sent to the Borrower. It proposed an 18 man force of enumerators and recommended that, as the main thrust of the project was in the Coastal Plateau area, the evaluation unit's program sho'.d be centered in it. It also listed - 76 - the main themes that required ME research, identified as the principal hypotheses on which the project was based. Ideally, this appreciation would have been developed as an integral part of project formulation and the basis of ME established well before it was. Yet is is not clear what insights ME might have provided had this been the case. While it would likely have assessed the amounts of, and variations in crop production and the extent of the areas receiving the input package, this information in crude and inconsistent form was available, and the uncertainties of precipitation were a matter of record. Organization 4.14 In October 1977, SORAD Maritime and the other regional SORADS were replaced by an "Organisme R6gional de Promotion et de Production Vivri5re". In due course, Government explained that pending formal arrangements being worked out the project would be administered by the same management and staff as SORAD Maritime under Projet de D&veloppement de la R-gion Maritime (PRODERMA). A communication from Government advised the legal basis, which the Association found satisfactory. The arrangement continued through the entire project. Growers Associations 4.15 The creation of growers associations was to be encouraged under the project for the purpose of providing support services, and it was hoped to combine them at the village level eventually to became cooperatives. Progress was achieved, and ultimately there were about 700 groups averaging 10 members, reaching 7,500 farmers. But a main activity, credit, was suspended due to the high level of farmer non-repayment (para 4.21). The farmers did benefit from the training program for bottomland rice cultivation provided through the associations (para 4.10) but the formation of cooperatives became moot with the fading of the project concept at the farmers level. Input Supply Unit 4.16 The Input Supply Unit (para 3.05) established to provide project inputs operated on a very limited basis. When its role with respect to being converted to a commercial basis was reviewed in 1980 PRODERMA argued against it, among other reasons because the farmers association movement had insuffi- ciently matured to assume the responsibility. The unit's operation was characterized by low volume and high operating costs and altogether it appeared somewhat of an anomaly in the scheme of things. Infrastructure 4.17 The infrastructure components of the project were not provided to the extent planned, but those built have proved to be well justified: - The 150 village stores provided for were judged to be unnecessary early in the project. The farmers had lost interest in them as one of their main functions, the storage of crops, appeared of little use in view of the disappointing crop experience. However, eventually, thirteen supply stores were built in 1981 at important locations in the project area to sell seed, fertilizer and insecticide. They are - 77 - functioning well. - The 15 dispensaries planned were scaled down to four which fill a very important role. A dispensary visited in 1983 is treating 4,000 cases, including 125 births a year, serving over 50,000 people within a radius of 20 miles. The reduction in number was due to the much higher cost encountered, some four times that estimated at appraisal. - Of the 20 pump equipped tubewells planned for village water supply, five were constructed in 1981, and one large water retention dam, instead of the two small ones planned, was completed in 1981. The scaling down was mainly the result of higher costs than anticipated. The water supplies have been useful. In particular, at the retention dam located in an area otherwise without water during the dry season, which feeds a block of 15 stand pipes, 300 people coming for water were counted in one hour. The facility was invaluable during the 1983 drought. 4.18 The feeder road program provided for the construction/improvement of 300 kn of roads, 150 kam by force and 120 km by contractor. Altogether 166 km (55%) of the program was completed as follows: Road Program Length Completed (k) 1976/77 67 1978 32 1979 35 1980 27 1981 5 TOTAL 166 The roads constructed were of good quality but received little maintenance afterwards. The annual road programs were reviewed by RMWA, and in at least one instance, the proposed construction of a sizeable length of road was rejected because of the high cost and the lack of economic justification. Procurement 4.19 Initially some difficulties were experienced with the procurement of fertilizer, vehicles and equipment. Procurement of fertilizer was centralized and project and non-project fertilizer was generally ordered under one contract, a fruquent West African country procedure. The Borrower and SOCODER tended to avoid strictly competitive bidding or follow standard bidding procedures, especialy for smaller contracts. However, a series of such situations, involving amounts usually less that US$50,000, were one way or another resolved, for example, diversion of the entire amount of fertilizer to the non-project area thereby avoiding the need for the Association to refuse - 78 - to finance it. Matters improved but eventually Government did involve itself in a serious misprocurement, the purchase of over US$2 million of fertilizer and insecticide of which about US$750,000 was for the Togo Cotton Project (Cr. 741-TOG), US$75,000 for the project, and the balance for non-project activities. The award was made to the highest bidder, a local firm (which, of course, imported these supplies) and, in fact, the insecticide for the cotton project was of a formula of only limited effect and not entirely satisfactory for the use. Consequently, in 1980 the Association cancelled US$792,000 from the Cotton Credit and US$50,000 from the project Credit. 4.20 One reason the Borrower had for awarding the contract to its choice of supplier was that 60% of the cost of the contract was paid from the Bcrrower's own resources and not by the Credit. Similar procurement situa- tions have been experienced in other countries and it is evident that procuring Bank Group-financed and Borrower-financed fertilizer under one contract in the interests of simplicity, economy, and to ensure that the Borrowed will indeed provide his as well as obtain the Association's share of fertilizer, may give rise to this problem. Pricing and Credit 4.21 The seasonal loans administered by the Agence Credit Projeu (ACP), a unit of CNCA, were only one-quarter of the appraisal target of US$1.1 million. This reflected the poor results of the crop program, the small area put under planting, the limited uptake of improved technology and the small number of groups qualified to obtain credit. After the drought, and poor harvest, in 1977, over 80% of seasonal credit had to be rescheduled. However even with this aid, credit recovery was very limited because of the continuing poor crop years, with about 75% of credit repayment remaining outstanding at project end. Long term farm credit for coconut planting was refused by CNCA at the outset of the project on the grounds that credit to coconut smallholders presented too great a risk during the early production stage. Fortunately, CARITAS, a charitable organization, which had provided funds and/or credit for various activities in the country, gave credit to coconut smallholders for about 2 years. Finally, CNCA did extend credit for coconut planting, but only to large scale farmers with other sources of income. In general, ACP was a fair organization although its credit agents were not always qualified. However, the expenses of ACP were disproportionate to the volume of credit it handled. In conjunction with the decision to discontinue major project activity, credit to farmers was discontinued and beginning with the 1981 season they paid cash for inputs. ACP subsequently focused on credit recovery as its main activity. 4.22 Project Financing As the project progressed PRODERMA's expenses fell considerably behind appraisal estimates, as did disbursements from the Credit, and by the end of 1978 both were about 45% of the forecast amount. The provision of funds to PRODERMA by Government to cover its share of project cost and operating expenses was adequate and usually prompt. There were, however, delays in reimbursing PRODERMA for the Government subsidies on inputs and for other reimbursements, aggregating about US$50,000 in mid-1979. Government eventually settled such outstanding accounts. At that time, it was estimated that carried to completion the project would experience a cost underrun of - 79 - over US$2 million. This reflected, of course, that most project activities were carried out at a level well below that foreseen. Project Results, Coconut and Food Crops 4.23 The lack of baseline data at the outset of the project coupled with the failure of ME to provide reliable crop estimates as the project progressed led to a serious problem. It was not really possible to estimate incremental production and relative project success. Indeed, the information that is now available is hedged by many reservations concerning consistency and accuracy and the data given in Table I below must be seen as an assembly of figures culled from those available in order to provide a basis for examination of the results. A major inconsistency was in the PRODERMA figures, which indicated that in 1981 the extension service reached 11,700 families and 16,700 ha (in fact, the appraisal target) of upland crop, whereas the area improved by PRODERMA was only about 3,500 ha or 14% of the 16,700 ha target. The area under rice cultivation (lowland), 750 ha, was about 13% of target. However, the information on yields is sparse. The foregoing figures do not entirely coincide with those in Table 1 below, given the various footnotes. - 80 - PROJECT AREAS TABLE 1 STATUS AS OF END 1980 Number Participating Area Under Yield Farmers Cultivation ha 1/ Kg/ha Maize 3,250 (15,800) 3/ 770 4/ Cassava Groundnuts 120 (2,930) 3/ Cowpeas 10 (3,900) 3/ Rice-Rainfed 750 (3,000) 3/ 900 7/ Cotton 800 (1,800) 3/ Total 10,000 (20,500) 3/ 4,930 1/ (27,400) 3/ 9,000 Z/ Rice-Bottomland 1,940 (5,000)3/ 680 1/ 6/ (2,500) 3/ 900.1/ 1,010 2/5/ 1/ Provided improved technology-fertilizer; in case of cotton, improved seed. 2/ Covered by extension services. 3/ Appraisal targets. 4/ Average, 1977-80. / Pre-Project, 960 ha. 6/ Net Increase, 170 ha. 7/ 1980 Yield. - 81 - 4.24 Vegetables The vegetable component was successful. At the end of the project the situation was as follows: Vegetable market Growers 90 Production 227 tons Revenue per 1/2 ha plot CFAF 800,000 (US$2,500) Benefit CFAF 270,000 (US$800) Vegetables growers in project area receiving extension services 260 (Appraisal target, 330) Fertilizer use 93 tons Seed 280 kg Recovery of water changes 100% At the vegetable farm near Lome, the diesel power pumps were replaced by electric pumps, and a tubewell was provided for vegetable irrigation. A modern vegetable marketing center was constructed in Lome and served as a convenient and successful outlet for the vegetables produced. Thus project vegetable market activities were concentrated in the Lome area, an area in which there was a large demand for fresh vegetables and ready payment was available. In effect, an already existing need was met. A small vegetable center at Tovegan was abandoned due difficult marketing conditions, and the idea of opening one at Tokali given up. 4.25 Coconut The coconut plantings component was also successful. Against targets of 600 ha and 2,200 sallholders, 605 ha were planted of which only 195 ha was held by 149 smalholders, larger farmers planting the remainder. This situa- tion resulted from the restrictions on credit (Para 4.21). The plantings aided by credit were poorly maintained while the larger holdings were well kept-up. By mid-1983, 750 ha were under planting. While generally very successful, the persistent dryness during 1980-82 resulted in depletion of the 1982 crop at the 40 ha Bayime Plantation to 15% of normal level. There are some traces of Kaincope disease, which are being monitored carefully. The nursery at Baguida whi.ch provided the young hybrid seedlings continues to operate satisfactorily at the 10 ha research farm. The coconut trees in their fourth year, irrigated for the first two, are growing normally and experiencing a full crop. So far, therefore, the coconut field trials aimed at demonstrating that hybrid plantings would resist the Kaincope disease have been a success. 4.26 Seeds The seed program was afflicated by the difficult and erratic clima- tic condition: in PY only 146 tons was produced overall against a target of 672 tons. Seed for maize varied between 20T and 95T annually, rice 12 and 56, cowpeas 2 and 12, and groundnuts 3 and 10, againist fifth year targets of respectively 383, 100, 17 and 172 Groundnut and cowpeas seed production was very small due to putting under cultivation a substantially smaller area than - 82 - planned, lacking sufficient quantities of the foundation seed supplied by the Institute de Recherches Agronomiques Tropicales et Cultures Vxvrieres (IRAT). The seed multiplication areas for maize and rice were as planned. Seed production, although it did not meet all demands, was a positive project feature. 4.27 Mechanization The objective was to open 3,000 ha for bottomland rice and 6,000 ha for maize. For the latter, mechanization was field tested in minimum tillage and mulching techniques to rehabilitate depleted soil areas. Sufficient inputs were provided for 1,000 farmers on the coastal plateau. PRODERMA's figures indicated that more than 5,000 farmers participated, however, the number may be exaggerated. While mechanization was only to be used to open bottomlands the smallholders tried to do the annual plowing using mechaniza- tion, but were unable to repay the credit they obtained to do it. In any case only 500 ha was opened, and many problems were experienced with mechanized plowing. Long delays occurred due to breakdown of equipment which could only 6o& rai-rd at the central garage, thp ncattered lctio= of pldt, -ifferent ploughing operations being necessary on the same plot, and the unsuitability of some equipment, necssitating its replacement by better adapted equipment. The implications of mechanization may become clearer from a survey that was initiated toward the end of the project. 4.28 Research The project research program was successful. The work program of IRAT was well oriented and adapted to the needs of smallholder agriculture. IRAT screened many maize/groundnut varieties, introduced some new ones into farming practices in the project area, possibly solved the agronomic problems of depleted soils, and consistently updated all crop growing recommendations. They also produced at a reliable and sufficient rate the basic seed require- ments. The results of IRATs work are given in several reports available in the West African Region files. The results were employed gradually in con- junction with the project, it being realized that ordinarily the results of research on crops conducted under the auspices of a project are not known soon enough to have a substantial impact on the project within its disbursement period. 4.20 Studies The study (para 4.02) of the development possibiilties of areas freed from onchocerciasis was not carried out. As late as the end of 1978 it was agreed to have a study in the Bassar area of a small rural development project. However, it was eventually concluded that the effort was misplaced and the idea was dropped. The irrigation engineeering study of the lower Zio river valley (para 4.02) was eventually concluded with the finding that 300 ha of rice land could be developed through irrigation. Project Cost 4.30 The tabulation of actual and estimated costs is given in CFAF francs because most costs were incurred in that currency or in French francs. Actual cost was 2,870.6 million CFAF vs 3,531.5 million CFAF at appraisal, or 81% of - 83 - estimate. The underrun reflected in part that the production elements were terminated before the end of the project, the cancellation of US$1.7 million from the Credit being a manifestation. Detailed costs are given in Annex 3, however, the main figures are as follows: % Actual vs. Estimate Actual Estimate CFAF millions Farm Inputs 378 231 61 Buildings 36 107 295 Vehicles and Eluipment (PRODERMA) 304 328 108 Operating Costs (PRODERMA) 191 634 332 Infrastructure 1,244 344 26 Ti general, the cost underrans are understandable in the context of the history of the respective ccmponents. The overrun on the operating costs of PRODERMA was mainly from the larger than estimated costs to operate vehicles, primarily due- # a nearly five-fold increase in fuel price as the result of the oil - - ion. 4.31 Covenant Experience By and large the covenants were met. This is not to say that the objectives were always achieved. Thus few dispensaries of those agreed were constructed, but for lack of funds. Of interest was that under a 1980 decree the fertilizer subsidy was set at a level of 50% of delivered cost, constituting a major cost saving. V. PROJECT JUSTIFICATION 5.01 The economic costs and benefits of the project are difficult to identify. The supervision reports gave some indications but they are not precise and not complete. Sometimes they are inconsistent. However, an economic analysis has been done using the best available data. The proiect costs and benefits are valued at 1975 constant prices as in the appraisal report and are adjusted as follows: - Costs of inputs are recalculated by actual planted areas and the cost of credit is restricted to the number of participants. The cost of infra- structure was reduced to 26% while the other cost items are valued at 81% of the appraisal estimates, reflecting the relative weights of actual expenditures (Annex 1). - The economic benefits are recalculated using the actual hectarage planted to individual crops in each project year. Due to incompleteness of the data, two assumptions are made in calculation of the economic rate of - 84 - return (ERR). (a) that level of production at PY5 would continue until the end of project life PY30; (b) that level of production at PY6 until the end of project life would at least be the average of what has been achieved during PY3-PI5, and remain constant onward. The first assumption gives the minimal rate of return of less than 1%, while the second assumption has an improved rate of return of about 4%, both of which are substantially lower than the appraisal ERR of 37%. VI. INSTITUTIONAL DEVELOPMENT 6.01 The project included a number of what amounted to institutional development initiatives: strengthening of SORAD Maritime, establishment of village level cooperatives, investment in and upgrading of training, and the establishment of a commercial input supply system, vegetable production and marketing, coconut development and dispensaries. The vegetable, coconut, dispensaries and training initiatives were, on the whole successful, although only a part of the facilities planned were completed. However little institu- tional building of a mainstream agriculture quality was achieved as with its manifest weaknesses, the project was not a vehicule for institution building purposes. VII. BORROWER AND BANK PERFORMANCE 7.01 The Borrower's performance generally was satisfactory. The Government supplied all the funds required, was cooperative, generally met the covenants and supported the Association. It did create a misprocurement situation and, ideally, could have provided a superior PRODERMA management and administration, but the main failings of the project were not its fault. 7.02 The Association's performance during project implementation was satisfactory as well. It mounted 11 comprehensive supervision missions and provided a considerable amount of technical assistance and guidance; the poor outcome of the project was not the result of the Association's deficiencies in this period. Rather, the results can be attributed in at least in some measure to its role in project formulation and, later, appraisal. As indicated by the results, the crop production elements, the mainstream project thrust, were not adapted to application in the Maritime area. To the degree this could have been recognized a priori, the Association pre-implementation judgment may be questioned. However, it may be considered that the project was something of a pilot operation, intended to demonstrate whether the region was suited to such a cropping regime. - 85 - VIII. IMPACT OF THE PROJECT 8.01 Regardless, important lessons were learned from the project. By any standard, it was found that the amount of rainfall and its timing are so erratic in the Maritime region as to preclude the reliable economic production of food crops and cotton on a smallholder crop intensive basis, that is, employing improved technology and farm inputs. It might be speculated that with a series of "ifs", such as organization, disciplined planning, precision timing, massive crop storage and the like, such cropping might be exploited as an adjunct to the (main) growth of food crops in other areas so that the Maritime regions annual production variations could be absorbed on something resembling an economic basis. But clearly there is no prospect at this time. 8.02 Thus further agricultural development efforts along these lines in the region must be of a relatively piecemeal kind. A small scale follow-up project under discussion tentatively includes maize, cotton, bottonland and irrigated rice but with the realization that rainfed rice would be at best marginal and irrigated rice possibly unattractive. Hybrid type coconut development might be adopted, depending on further experience with the field trials thus far successful. The Borrower visualizes a project comprising irrigated rice and irrigated food crops utilizing retention dams to provide water, another 2,000 ha of Kaincope resistant coconut, village water supply also using dams, and dispensaries, all of which reflect the successes of the project. 8.03 A second lesson confirmed the need for adequate project preparation. The project was marked by a lack of adequate information in the first place and by the fact that ME was not an organization in place to make a base line survey or evaluate results at the start of implementation. The importance of this limitation is all the more evident when it is realized that the Bank Group's only prior agriculture operation in the country was the cocoa and coffee credit of 1974. 8.04 The appraisal report included an unsually descriptive section on "risks" to the effect that the project was the first major Bank Group initiative in upland arable crop production in a bimodal rainfall area of West Africa where food crop development had generally been neglected and therefore yield assumptions and those concerning farmer response might be in substantial error; and while rainfall was better than in the north, the benefit of the rainfall regime might be illusory; therefore, the technology base for the project was less firm than for other arable crop projects in the West Africa. How much risk should be taken is obviously a matter worthy of full consideration before committing a project. 8.05 The counterproductive results of having too many components was demonstrated. It cannot be said that the crop production component would have been successful if there had been fewer components. But it was evident that supervision missions were in danger of being swamped in their efforts to deal effectively with them, to the detriment of the main thrust and objectives of - 86 - the project. IX. CONCLUSION 9.01 The project failed to meet its goals by a wide margin. The project was predominantly intended to increase food crop production by means of improved technology under climatical and soil conditions relatively untested in West Africa. The poor results reflected the extremes of rainfall experienced, lack of any demonstration effect, failure of the input package to be used by farmers on a wide spread basis, inadequate farm data information and weak extension services. In any case, the project demonstrated that crop intensive smallholder farming as conceived by the project was not viable. Unfortunately the poor information both of a without project and with project nature makes it difficult to provide quantitative results. However, the re- evaluated ERR of probably 1-4%, compared with the estimated ERR at appraisal of 37%, does provide an indication of the project's impact. 9.02 Because of the minimum achievements during the first three-quarters of project implementation, financing of the project crop production activities was terminated in June 1981, although various studies and technical assistance related activities continued to be financed until the (extended) Closing Date, December 31, 1982. As a consequence, an amount of US$1.7 million of the Credit was cancelled with the agreement of the Borrower on February 23, 1982. Because of misprocurement an amount of US$50,000 was also cancelled during implementation. 9.03 However, the project was partly successful in that the minor components, although not completed to the extent planned were mainly successful. Field trials of Kaincope disease resistant hybrid coconut plantings are proving successful but will require more time to firmly demonstrate the case. A fresh vegetable growing and marketing system in the Lone area likewise succeeded. The provision of village water supply and the introduction of dispensaries in the project area both proved highly valuable and there is a demand for more. 9.04 While there were some weaknesses in the execution of the project, and while both national and expatriate direction were not as good as they might have been and extension services and project monitoring and evaluation were poor, the Borrower and the implementing agency were cooperative and reasonably efficient and project results are not attributable to maladministration. Rather the problem was mainly one of an overoptimistic assessment of the possibilities of the cropping regime when the project was formulated and appraised. 9.05 The project had little institutional impact on the principal agencies involved or on the agriculture sector. It was positive, however, in so far as it has provided patterns for establishing village water supplies, fresh vege- table marketing and dispensaries. - 87 - Annex 1 ESTIMATED AND ACTUAL COSTS % Actual Appraisal 1/ Actual 2/ va Estimate ------CF 3 milions------ Farm Inputs 378.6 231.1 61 Credit, Farner Associations 206.8 4.3 2 PRODERMA Buildings 36-3 107.1 295 Vehicles and Equipment 304.9 328.5 108 Operations and Administration 915.5 1,493.7 163 ACP 186.5 93.2 50 Infrastructure 1,244.0 344.6 26 Studies 259.0 268.5 105 TOTAL 3,531.6 2,871.0 81 1/ Including taxes and contingencies. 2/ Including taxes. 一g才一 戶添、戶分命乙～斤 - 89 - T 0 G 0 RURAL DEVELOPHENT PROJECT IN COTTON AREAS (CREDIT 741 - TO) PROJECT COMPLETION REPORT December 29, 1983 WESTERN AFRICA,,PROJECTS ACRICULTURE DIVISION A. lk - 91 - TOGO RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CR. 741 - TO) PROJECT COMPLETION REPORT I.BACKGROUND 1.11 Togo is located on the Gulf of Guinea and covers an area of 56 000 km bordered by Ghana, Benin and Upper Volta. The country is divided into five administrative regions, from north to south: Savannes, Kara, Central, Plateau and Maritime. Population is estimated at 2.5 million, growing at 2.5% per annum, of which 2 million (80%) are rural, with heavy concentrations in the south and in the Kara region. Average GNP per capita was US$ 340 in 1981, about twice that of the rural population. The agricultural sector accounts for 80% of employment, 30% of GDP and 30% of export revenues. About 11% of the geographic area is cultivated by 265,000 families (7.5 persons per family) who grow food and cash crops on about 2 ha on average. Food crops (maize, sorghum, millet, cassava, yams and beans) cover about 450 000 ha and industrial crops (oil palm, coffee, cocoa, cotton) about 160 000 ha. Animal husbandry has some importance in the northern and central regions. 1.12 Except for most of the Maritime Region, Togo is ecologically well suited for cotton and food production. Until the early 1960's, Togo grew only the Gossipium barbadense variety Mono in association with other crops with average yields of 100 kg/ha. In the late 1960's Togo, with the assistance of the French Fonds d'Aide et de Coopgration (FAC) and the European Development Fund (EDF), started to switch over to the Gossipium Hirsutum variety Allen and then Bou grown in pure stands with average yields of 700 Kg /ha in 1976, due to the new variety's responsiveness to fertilizers and insecticides. As a result, production increased steadily in the early 1970's but the upward trend was halted during the period 1975-1977 mostly as a result of prevailing drought conditions. 1.13 In 1972 a FAC financed study carried out by the Compagnie Francaise pour le DEveloppement des Fibres Textiles (CFDT) recommended that a cotton development project managed by a mixed public - private company in charge of technical assistance to farmers, primary marketing, processing and export, should be set up to increase cotton production from 10 400 tons in 1973/74 to 23 000 tons in 1977/78. When Government established the Socift6 Togolaise de Coton (SOTOCO) in 1974 it retained most of the study's technical recommendations but SOTOCO became a state company, with export operations reserved for the Office des Produits Agricoles du Togo (OPAT). 1.14 The first 1/ Rural Development Project in Cotton Areas (RDI) was implemented from 1977 to 1982 against a rapidly changing national financial background which could not be foreseen at appraisal but which had a crucial impact. Before 1974, the Government had relied on export oriented private sector initiative - with rock phosphate accounting for 76% of total exports - pursued cautious fiscal policies and maintained 1/ : It was followed by the Second Rural Development Project in 1982. - 92 - external and internal financial equilibria. The price increase of rock phosphate after the industry's nationalisation in 1974, led to a doubling of Government revenues in 1975 and an ambitious development policy which gave primacy to direct state intervention in most sectors. When prices of rock phosphate fell in 1976, most state enterprises did poorly, and the 1976 and 1977 crop years were much below average due to persisting drought conditions, the Government budget experienced severe financial difficulties which, combined with lax fiscal discipline, severely affected the Govern- ment's capacity to honor its financial commitments towards RD1. This point is essential in understanding difficulties experienced during project implementation. II. FORMULATION 2.1 General 2.11 In October 1975, the Regional Mission in West Africa (RMWA) undertook a general identification mission to Togo and focused on the possibility of a rural development project in cottcn growing areas. This orientation responded to the Government's high priority towards such a project which was in an advanced stage of preparation. When a FAC financed feasibility study became available in early 1976, RMWA endorsed the project concept as well as its major objectives. It considered that the study was well documented and contained most of the necessary data. During a five year period, the proposed rural development project was to expand cotton and food crop production, improve infrastructure, strengthen the executing agency, expand ginning capacity, expand ox-cultivation and provide credit for inputs. While this concept remained substantially unchanged, during appraisal which took place in October 1976, the cotton production objective was reduced from an incremental 30 000 tons to 21 500 tons and IDA agreed that ginning facilities would be financed separately by Government. Some of the issues were settled before and others during negotiations which took place from July 18 to 20, 1977. 2.2 Issues settled before Negotiations 2.21 Institutions. Prior to SOTOCO's establishment in 1974, the Ministry of Rural Economy and its local agencies, the socift6s d'am6nage- ment et de dEveloppement (SORAD) had been in charge of both food crop and cotton development. By charter, SOTOCO's activities were limited to cotton development. Based on past performance, the appraisal mission recommended that SOTOCO and not the SORADs should be entrusted with project implementation, an argument which management.accepted. Finally, after an initial insistence on an amendment to SOTOCO's charter enabling it to promote food crops, IDA realized that amendment procedures would be long and threatened to delay project processing. It therefore satisfied itself with general assurances that the necessary steps would be taken within a reasonable time. 2.22 Subsidies: Government believed in the mid 19?Us that inputs should be subsidized probably because (i) several crop failures were attributed to insufficient use of insecticides, and (ii) the bad crop - 93 - years 1976 and 1977 had reduced farm incomes to such low levels that farmers were thought to be incapable of purchasing inputs. The Government therefore, without advising IDA, eliminated in December 1976 the farmers' contribution towards the cost of insecticides. At the decision meeting stage IDA management wanted an agreement on phasing out subsidies as a condition of negotiations or Board presentation. But tie appraisal mission recommended a more flexible approach: it argued that (i) phasing out of subsidies would neither change the farmers' nor the Government's cash flow position if the producer price was increased correspondingly, (ii) subsidies were more than compensated by Government taxation; but on the other hand, input subsidies reduced farmers' risks resulting from yield variations and encouraged them to adopt modern cultural practices required for project success. Management accepted these reasons prior to negotiations and agreed to an annual review of the level of producer prices and input costs and subsidies, a position to which Government had also agreed in the Maritime Region Development Project. 2.23 Input procurement Based on the SORAD experience, when inputs were procured by a Government agency and arrived usually late, IDA staff recognized the need to ensure timely arrival of inputs. This was to be achieved by ensuring that SOTOCO was in charge of input procurement and that it had the necessary funds at its disposal when they were required. To this effect, a Special Account was to be set up in SOTOCO's name and funded by Government, the cofinanciers and farmers. The initial deposit of CFAF 300 million by Government to the Special Account was a condition of effectiveness. There is no evidence that this issue was discussed at any great length at negotiations. However, project files suggest that the problem was solved prior to negotiations and an understanding -.ached with Government. A memo on file indicates that IDA thought that the operation of the Special Account by SOTOCO was not incompatible with the channeling of all procurement through the Government's official procurement agency. 2.3 Issues settled at Neotiations 2.31 SOTOCO Finances One of the major reasons for the SORADs' failure in cotton promotion had been their lack of resources. Initially SOTOCO experienced similar problems as the ad hoc Government subsidies it was to receive were usually late and remuneration of services by OPAT under the barZme, equally late and insufficient to cover its costs. The appraisal mission therefore devised a system which would phase out Government subsidies to SOTOCO by 1981 and ensure that OPAT payments under the bar-me were sufficient to cover SOTOCO's costs. To start with, the Government was to provide progressively declining annual subsidies - the first payment of CFAF 275 million was to be a condition of effectiveness - and OPAT was to pay SOTOCO an additional CFAF 16 per kilogram of seed cotton delivered to its wills. 2.32 The Togolese Delegation had no objection to making the payment of CFAF 275 million a condition of effectiveness (in addition, annual government contributions were confirmed in a side letter dated October 5, 1977), but it refused to commit itself to putting any figure on the proposed increase in the barZme. Agreement was reached on a general - 94 - wording whereby the Borrower undertook to set the remuneration for SOTOCO under the Bargme at a level which would cover the cost of marketing and extension activities. If it had been applied, payments under the barBme and annual government subsidies would have made SOTOCO financially independent by 1981. The Government's refusal to commit itself to any particular figure suggests with hindsight, that it had become aware of financial difficulties ahead. 2.33 Producer Prices During project preparation IDA staff had been aware that returns per acre and per manday were usually higher for food crops than for cotton. Therefore they recommended that the producer price should be raised to CFAF 60/kg to make cotton cultivation more competitive. Once again, the Togolese delegation would not agree to putting a figure on the producer price. It agreed instead to set the farm gate price at a level sufficient to provide adequate incentives to seed cotton growers and to consult annually with IDA on the matter. 2.4 Project Description 2.41 The proposed project which would be carried out over five years (including the FAC-financed start-up period), would endeavor to improve living conditions in the area on a broad scale. Basic Infrastructure would be improved, both for transport and the supply of drinking water, the latter under an EDF-financed project. Traditional farming would be improved by introducing better cultivation methods, including use of oxen, and promoting higher yields through the integration of food crops with cotton into a simple, well-balanced rotation. 2.42 The project would be executed by a Government agency, SOTOCO and would reach some 53,000 farm families - over 400,000 people - representing about 36% of the rural population in the project area in PY5. The main project features would be: (a) strengthening SOTOCO, by providing: (i) buildings and equipment for SOTOCO headquarters and field offices, storage facilities, housing, vehicles and equipement; and (ii) additional manage- ment and field staff, training and operating expenses; (b) providing incremental on farm inputs on seasonal credit to expand improved cotton and food crop planting; (c) establishing a seed-multiplication system including equipment and vehicles and carrying out adaptive field trials; (d) building and improving about 700 km of feeder roads and providing for their maintenance; (e) providing oi-cultivation training including demonstration equipment and credit to farmers for acquisistion of ox-drawn equipment, and (f) supporting project monitoring and evaluation. - 95 - III. IMPLEMENTATION 3.1 Start-up 3.11 The Board approved the credit on September 13, 1977 and it was signed on October 5,1977. Two crucial conditions of effectiveness were that the Borrower provide a CFAF 275 million subsidy to SOTOCO for FY 1977 and deposit CFAF 300 million in a Special Account for input procurement. CNCA certified in January 1978 that SOTOCO had two accounts on their books showing a balance of CFAF 575 million in SOTOCO's favor. IDA considered this as evidence that the conditions of effectiveness had been met, and declared the credit effective on March 17,1978. Bank missions found out in 1978 that despite CNCA certification, Government had not deposited the CFAF 575 million in SOTOCO's accounts and that the conditions of effectiveness had therefore not been met. Bank supervision missions spent the next two years trying to get the Government to provide these funds and to meet SOTOCO's financial needs. 3.12 In line with prudent practice, the SAR had planned on early procurement of inputs, equipment, vehicles, and construction of major facilities. These schedules proved to be optimistic as the appraisal mission had overestimated SOTOCO's implementation capacity resulting from the inexperience of its personnel - expatriate as well as Togolese -, its poor administrative infrastructure and the complexity of the institutional environment it had to contend with. As a result, procurement operations and construction activities suffered from initial delays which remained permanent concerns during project implementation. 3.2 Organization and Management 3.21 In what follows, a distinction must be drawn between problems afflicting SOTOCO's field services which did well nearly from the start (para 3.31) and headquarter staff which had to deal with the financial problems and which were in the limelight as long as the national financial problems persisted. They were slowly brought under control after the new general manager and key expatriates took over in 1979/80. 3.22 As appraised, SOTOCO's general Managcr and his deputy were to supervise four departments (Technical, Commercial, Financial and Engineering). In fact, the Engineering Department was not established but the relevant units (workshop, transport fleet and road construction unit) reported directly to the General Manager. Except for the Commercial Manager, the department heads and five technicians were expatriates seconded from CFDT and BDPA whose services SOTOCO retained for a total of 28 man years. Experienced Togolese were in short supply as salaries were not sufficient to lure capable officers away from the private sector and the capital city Lome. Turnover among the expatriates was high until 1979 due in part to inadequate working conditions, office facilities and housing, and lack of schools and afterwork activities at Atakpame. 3.23 During the initial two years financial management was the weakest link in SOTOCO's management: initially the financial department - 96 - relied essentially on a single expatriate. Local staffing remained subsequently insufficient; the quality of accounting clerks and their office accomodation was poor. Despite efforts to find a suitable Togolese chief accountant, uncompetitive Government wage levels did not enable SOTOCO to fill this slot. The accounting systems were characterized by the lack of cost accounting, internal audit, inventory and cash control. Significantly, SOTOCO used up 5 financial directors in the early years.As a result, financial statements were chronically late, withdrawal requests to cofinanciers were not made on time, which exacerbated cash flow difficulties. Those sta:ements that were actually produced were unreliable and misleading as accounting procedures changed over time, inventories were inaccurate, accounts insufficiently defined. The situation improved significantly in 1980/81 after the arrival of the incumbent financial director and the new team 3.24 Input Procurement. One of the major problem areas was SOTOCO's role in the procurement of fertilizers and insecticides. In the appraisal mission's original view SOTOCO was to be the procurement agency which, to enable it to carry out its task, was to dispose of a Special Account (see para 2.23). In fact the Special Account was never established and SOTOCO's involvement in procurement remained marginal. This may have been fortuitous in view of SOTOCO's subsequent management and financial problems (para 3.23) since inputs were effectively procured by the Ministry of Rural Development and prefinanced from Government sources. SOTOCO's role was confined to determining their input requirements, participating in bid evaluation,acting as a contact with IDA, receiving,storing and distributing inputs. While the procurement process did not work the way it was supposed to, inputs did usually arrive, though on more than one occasion they were nearly too late. SOTOCO took the precaution to build up stocks which enabled it to supply outlying or inaccessible areas that would have suffered most from late input deliveries. In one year, because of irregularities, IDA declared misprocurement in March 1980 and cancelled US$ 792 000 representing its share of the 1979 tender for 5 680 tons of fertilizer and 400 000 litres of pesticides 3.3 Field Services 3.31 Field services were the backbone of SOTOCO's operations based on a strong vertical (Technical Manager assisted by Training Officer) and horizontal (regional) structure. During the project implementation period the number of agents increased from less than 500 to more than 700. Most of the agents had been taken over from the SORADs. They were initially poorly trained, and lacking direction and motivation. It was perhaps one of the project's major achievements to transform them throu3h proper training and a system of sanctions and rewards into an efficient cadre that performed all the services necessary to reach project objectives by providing technical advice, inputs, credit and purchasing seed cotton against cash. Thcy achieved a remarkable 96% credit recovery rate. 3.32 Field agents concentrated essentially on cotton, and on food crops only to the extent required for integration into the cotton rotation. They - 97 - delivered each year technical messages to farmers and checked on their execution. They succeeded in getting most cotton growers (90-100%) to maintain average fertilizer rates at more than 200 kg/ha, to increase insecticide application rates by 50%, as well as planting densities, which account for a substantial share of the production increases. Messages on land preparation, sowing dates and weeding were less generally adopted. Although results in food crops were less impressive (para 5.21-5.23), overall field management was good. 3.33 The project had contract agreements with IRCT and IRAT for the production of foundation and breeder seed and applied research. For cotton, seed production and multiplication proceeded without difficulties in cooperation between IRCT, SOTOCO, the East Mono cooperative, and farmers from the Savannes region. For food crops the problem proved to be more difficult (para 5.22) 3.4 Marketing and Processing 3.41 Initially SOTOCO was only supposed to purchase seed cotton from farmers and to sell it to OPAT which took care of ginning. In 1978/79 the increase of cotton production under the project hit SOTOCO unprepared. Since neither funds, nor marketing teams, nor storage,nor ginning capacity were sufficient, SOTOCO lost 1500-1800 tons of seed cotton due to storage in the open air, probably one of the causes for the early departure of the incumbent general manager. Under the new general manager and after a strong intervention by IDA the situation improved considerably (para 9.22) Financing was provided more regularly, marketing teams were organized and 400 storage sheds built, a remarkable achievement. Due to insufficient ginning capacity (para 3.43), stocks had to be carried over from one year to the next, entailing high financial costs and delayed cash flows for SOTOCO and country. 3.42 Transport proved to be serious bottle neck. Instead of procuring trucks and trailors progressively, SOTOCO purchased them in bulk in PY4 only. When they arrived SOTOCO, was not in a position to use them fully due to lack of maintenance facilities, qualified drivers and management capacity. These problems were compounded by high transport costs on certain routes due to the isolation and dispersion of some cotton farmers (para 3.51) and the location of the new mill at La Kara (para 3.43), and Government requisitioning. Despite repeated IDA recommendations, SOTOCO did not hire a consultant to advise on transport management. This problem is only being solved under RDII. 3.43 The lack of sufficient ginning capacity was another crucial constraint. The appraisal mission had found that the Government was making satisfactory provisions for sufficient ginning capacity. Indeed, it had already procured machinery for the new mill at Lama Kara in the North and had initiated a study on the extension of the Notse ginnery in the South. For these reasons, the appraisal mission did not expect ginning capacity to be a constraint to project development. However because of financial difficulties Government did not succeed in finding suitable contractors for the construction of civil works in Lama Kara and the - 98 - ginnery became fully operational only in late 1981 after it had been refinanced by CCCE. In the meantime insufficient ginning capacity required storage at field level and entailed spoilage and financial losses (para 3.41). This problem is being addressed under RDII. 3.5 Other Components 3. 1 Feeder roads. Due to late arrival of expatriate personnel -.nd equipment, the program started with delays. Nevertheless, the road unit built 500 km of feeder roads, about 70% of the appraisal's 700 km con- struction target. The underachievement on new roads was compensated by an active road minimum maintenance program involving by PY4 some 1800 km compared to an appraisal estimate of less than 500 km. Road standards were appropriate relative to traffic volume; good management and intensive equipment usage resulted in low direct unit costs. Nevertheless, a EMWA road consultant questioned the economic justification of some of the roads, some of which were running parallel to existing ones. He questioned whether some roads had been properly evaluated in relation to the production they would have to carry. These criticisms were not accepted by SOTOCO who asserted that the roads were necessary to extend cotton production throughout their territory. 3.52 Animal traction. After a promising beginning with this pilot component it turned out that organisation and supervision of the mobile demonstration teams put in place were weak and disorganized and that the component's impact on the farm population was negligible. Compared to an appraisal target of 350, only 136 Savannes Region farmers had been equipped by October 1980. In mid 1981, SOTOCO claimed that it had no technical expertise in animal traction and that other projects were better suited to carry out the project component. In 1982, SOTOCO conceded that only 18 of the planned 23 demonstration centers still carried on some kind of activity due to deaths of animals and farmers' lack of interest. Other projects have not fared much better in Togo but efforts are underway on a national level to streamline procedures, organize provision of animals and equipment, train field agents and develop a consistent national policy. RDII will make a major contribution to this effort. 3.53 Monitoring and Evaluation. Under the project a Monitoring and Evaluation Unit was to be established under the supervision of the Maritime Region Rural Development Project and the Ministry of Rural Development. A project officer with training in agriculture would collect farm level data, monitor physical project progress, alert project management to possible problems and cooperate with project management on quarterly and annual progress reports. In the event neither the national level nor the project MEU were set up. SOTOCO felt that data collected by their field agents in the course of their work was sufficient. 3.54 A monitoring and evaluation specialist who participated in a RDII supervision mission in June 1983 found that each of SOTOCO's departments more or less monitored its own activities as the need arose. These data were not easily accessible and insufficiently evaluated. They focused on cotton production but gave scant emphasis to food crops and to - 99 - production systems. Field agents kept relatively precise records on cotton growers for operational purposes but efforts to analyze a sample of these records have not yielded meaningful results so far. 3.55 Reporting The appraisal mission had prepared a detailed checklist of items to be covered in quarterly and annual progress reports. These data would serve as a management tool and a major input for the preparation of a follow-up project. In fact, starting in 1978, SOTOCO produced a series of annual reports on all aspects of cotton cultivation, interspersed here and there with data on food crops. These reports were internal documents of the Technical Department, valuable as such, but limited in scope, although this department was SOTOCO's most important one. What was missing was information on other aspects such as finances, and a synthesis. However, given SOTOCO's difficulties, supervision missions were probably right in not pushing more strenuously for SOTOCO's compliance with reporting requirements, especially since SOTOCO was always helpful in producing the necessary data for supervision missions. 3.56 SOTOCO was requested to prepare a completion report. It supplied a series of notes on various project aspects,but their data base is weak. D.Salvy, the former technical director, has produced a final report on cotton including comments on seed production. 3.57 SOTOCO's first three fiscal years were audited in May 1981 and the fourth in November 1981 by a local auditing company. The fifth year which covered the period from September 1, 1981 to March 31, 1982 is being audited now. In view of the quality of SOTOCO's book keeping in the early years the "commissariat-aux-comptes" type audit that was performed by the local auditing form was probably not sufficient to make sufficient impact. Under the new project auditing arrangements including choice of firm are now being received. IV.FINANCIAL ASPECTS 4.1 Project Period 4.11 The FAC grant covered a five-year period starting on JanLuary 1, 1977 and the IDA credit, a four-and-a-half year period starting on October 1, 1977. From January 1, 1977 until September 30, 1977, FAC financed salaries of key personnel and operating costs, and Government,other operating costs and inputs. Starting on October 1, 1977. IDA, FAC and Government cofinanced investment and incremental inputs, salaries and other costs. The financing period does not coincide with the agricultural year and only partially with SOTOCO's fiscal year. This difficulty is compounded by SOTOCO's book keeping practices which makes it arbitrary to reconcile the various periods and to draw valid conclusions from such reconciliation. This report considers the years 1977/78 as PY1 and 1981/82 as PY5 in line with official RDI documents for which project completion was on December 31, 1981 and the closing date on June 30, 1982. - 100 - 4.2 Project Costs. 4.21 The following table summarizes project costs incurred from October 1, 1977 to March 31, 1982, as given by SOTOCO: TABLE 4.1 : Project costs, in CFAF million Appraisal Actual Difference Act./AE Estimate in Z Investments Buildings 737 671 (66) 91 Vehicles and equipment 1,255 918 (337) 73 Road equipment 233 141 (92) 61 Sub-Total 2,225 1,730 (495) 78 Operations Fertilizers 409 339 (70) 83 Insecticides 494 356 (138) 72 Salaries 1,129 859 (270) 76 Other costs 2,124 3,350 1,226 158 Sub-Total 4,156 4,904 748 118 T 0 T A L 6,381 6,634 253 104 The total project cost in US$ equivalent is US$ 26.8 million (Annex Table 1). 4.22 Actual project costs were within 4% of appraisal estimates. The project spent 20% less on investment and, except for "Other costs" operated generally at a lower than anticipated level. On the other hand, the project spent nearly 60% more on "Other costs" due mainly to rises in transport and financial charges. 4.3 Financing 4.31 General SOTOCO received during the implementation period only 89% of the funds deemed necessary at appraisal but Government and FAC supplied the balance of their commitments in 1982/83 after project completion. The shortfall in the IDA contribution was due to a cancelation of US$ 792 000 in March 1980 and exchange rate losses. The last IDA payment took place in July 1982. - 101 - TABLE 4.2 Project Costs and Financing (in FCFA millions) 1 2 3 4 5 TOTAL Sources IDA - 386 964 1 392 311 3 053- FAC - 138 317 513 174 1 142 4/ 31 GOVM4NT - - 240 300 500 430 1 470- Total - 764 1 581 2 405 915 5 665 Project Costs 590 1 309 1 596 2 101 1 038 6 634 Difference (590) (545) (15) 304 (123) (969) Additional Financing OPAT 5/ 199 265 155 789 1 570 2 978 Balance (391) (280) 140 1 093 1 447 2 009 Cumulative Balance (391) (671) (531) 562 2 009 1/: Difference due to cancelation of US$ 792 000 and exchange losses. 2/: FAC provided its final contribution of FCFA 109 million in February 1983. 3/: Government provided its final contribution of FCFA 170 million in 1982/83. 4/: From the Government's Investment budget (B.I.E). 5/: Covers OPAT remuneration under the bargme, an annual OPAT subsidy of FCFA 110 million and SOTOCO credit recovery from farmers. 4.32 The Table confirms the project's initial shortage of working capital which was only alleviated because of ongoing OPAT financing. The positive balance of CFAF 2 billion covers the "with and without" project situation. It includes transport costs from the ginnery at La Kara to Lome of CFAF 500 million for PY4 and PYS, accumulated reserves of CFAF 800 million while the balance of CFAF 700 million covers the without project situation of CFAF 140 million annually. V. AGRICULTURAL IMPACT 5.1 Cotton Development 5.11 Considered globally, cotton production appears to have stagnated from the mid 1960's to the mid 1970's at a level which the appraisal mission put at 6 700 tons. However, clear trends appear when the distinction is made between the traditional Gossipium barbadense variety Mono (para 1.12) and the improved Gossipium Hirsutum variety Bou. The production of the low yielding variety Mono declined from a maximum production of more - 102 - than 9000 tons in 1968 to a few hundred tons in the early 1980's. The production of the high yielding Bou variety shows the opposite trend: it increased from 182 tons in 1965/66 to 9586 tons in 1974/75 and reached 27,450 tons in 1982/83. Drought years interrupted this upward trend twice: the first time in 1972/73 and the second time in 1976/77. Drought usually threatens the farmers' food supply first. Farmers then grow food crops on areas normally reserved for cotton until their food sufficiency is ensured. The effect on cotton areas may in certain cases be delayed by one year or two depending on farmer anticipations. 5.12 The switch over in the early 1970's from the low yielding Mono to the high yielding Bou variety laid the groundwork for the success of cotton production under RDI. SOTOCO organized an efficient field service which succeeded in getting most cotton growers to maintain high fertilizer and adopt high insecticide application rates, and follow proven agricultural practices (paras 3.31-3.32). It is true that production reached its lowest point (4200 tons) during PY1 but this low production was due to the after effects of the preceding drought years. It is likely that PYI production would have been even lower in SOTOCO's absence. 5.13 While cotton production was right on target from PY2 to 4 it suffered shortfalls in PY1 and PY5 for reasons outside project management's sphere of responsibility: PY1 production was affected by the preceding drought years (para 5.12) and PY5 by rises in the cost of fertilizers although the reduction (10%) was not great in relation to the tripling in fertilizer price (para 7.14). To better appreciate project results, it seems judicious to consider the actual PY5 production level (21 200 tons) as an accident compared to preceding (23 900 tons) and following (27 500 tons) levels and to use an average of 25 700 tons as a more appropriate PY5 figure. Under these assumptions, the project would have achieved between 85% and 98% of its objectives as shown by the table (also Annex Table 3). Comparison between Objectives and Adjusted Production Data. AE 1/ 2/ Adjusted Objecti es Achieve- PY PY PY PY - ments 4 6 5 5 in % N Planters 46,171 44,061 45,116 53,300 85 Production,tons 23,860 27,450 25,655 29,210 88 Yields, kg/ha 819 1 005 909 923 98 Areas, ha 29,130 27,318 28,224 31,660 89 1/ :SEDES ; 2/ : SOTOCO; 3:/ : SAR; AE = Appraisal Estimates 5.14 During project implementation, cotton production increased by 236% from 6300 in PYO. Yields increased from 700 kg/ha to 900 kg/ha, the number of participating farmers from 18700 to 38100, and average cotton area per farm from 0,48 ha to 0,62 ha. The cotton area a farmer grew to cotton was relatively small compared to total cropped area: 25% in the Plateaux, 15% in the Centrale, 16% in the Kara and 11% in the Savannes - 103 - Region. 44% of the production increase came from a doubling of participating farmers, 33% from a nearly 30% increase in yields and 23% from an increase in the cotton area per farm. The relative importance of regions did not change significantly with the Plateaux region producing more than 50%, the Central region about 25% and the Kara and Savannes regions sharing the balance. 5.2 Food Crops 1 5.21 By PY 5, SOTOCO was to have developed 29,200 ha of stage II - and 15,030 ha of stage III food crop areas (maize, sorgho, groundnuts). In the absence of a MEU, it is difficult to precisely evaluate food crop yields and planting achievements, and PCR findings are based on input supply and cotton area planting data and the field observations of supervision missions and SOTOCO staff. The SEDES report concluded that in view of the quality of data and in-depth analysis of the food crop component was not possible. It also questioned whether the areas under improved varieties really met all the criteria stipulated in the SAR as food crop fertilizer supplies had been very limited. Subsequently, the SEDES report did not assume any benefits from food crops. 5.22 Whilst acknowledging the imperfection of information on food crops, particularly yields, the PCR believes some clear benefits have accrued from this component. It is clear that participating stage II farmers, all used improved seed and fertilizer at least in their first year, as it was an obligatory SOTOCO package and that by year 5, 6,600 ha of grade III plantings, 45% of appraisal targets had been achieved, Annex Table 4. In the stage II situation, maize planted targets were almost achieved, groundnut targets exceeded but sorghum targets only 20% achieved. This was mainly due to the unavailability of an improved sorghum variety acceptable to the farmers. The PCR also doubts that appraisal yields expectations for the stage III situation were realised. 5.23 Incremental food crop benefits for the stage 2 situation assumed increasing yields of from 700 to 1,200 kg/ha for maize and from 500 to 800 kg/ha for sorghum due to the residual effect of cotton fertilizer on local maize and sorghum varieties and an improvement in cultural practices. Although the SEDES report assumed no residual benefit from cotton fertiliza- tion on subsequent food crops, IRCT research carried out since 1965 on many sites in Togo shows an average 14% food crop yield increase due to residual fertilizer. The PCR has thus assumed a 14% yield benefit over traditional yields for food crops following cotton; estimated at 14,000 ha at PY 5. No benefit has been taken for possible unmeasured improvements in cultural practices. PCR has thus estimated a total incremental food crop production at PY 5 of approximately 4,500 tons, 46% of the PY 5 appraisal estimation of 9,800 tons (Annex Table 5). 1/ Stage II involving integration into cotton rotation, timeliness of operations, correct spacing and plant population. Stage Ill: Usage of improved varieties, seed dressing and fertilizer applications (excepting groundnuts. - 104 - Sur.ary of Yields Kg/ha Appraisal Estimate PCR Actual Estimate WIO Project Stage 2 Stage 3 Stage 2 Stage 3 Maize 700 1,200 1,800 800 1,300 Sorghum 500 800 1,300 600 900 Groundnuts 400 - 1,000 - 900 5.24 Under the project, the Institut de Recherche Agronomiques Tropicales et de Cultures Vivriares (IRAT) was to produce breeder, foundation and SOTOCO commercial seeds with the help of contract farmers. Seed production suffered from a number of constraints: potential contract farmers were inexperienced, isolation of multiplication areas was difficult, adequate equipment and infrastructure for harvesting, treatment and seed storage were non existent in early years of the project. Nevertheless, the project produced considerable quantities of maize and seed (750 tons maize seed and 285 tons sorghum seed from PY 1-5), Annex Tables 6 and 7. Relatively little was actually distributed as seed and the bulk of the SOTOCO production was sold for consumption. The improved seed varieties of sorghum multiplied have not proved acceptable to the project area farmers. In the case of maize, the composite La Posta has been accepted in the northern areas of Togo whilst NH1, a relatively poor yielding short cycle hybrid maize, is the only variety thus identified by research which is accepted by farmers in the southern bi- modal rainfall zones. A Positive aspect is that SOTOCO has mastered seed multiplication techniques and has developed the necessary infrastructure to produce significant quantities of food crop seeds in the future for Togo. 5.25 IRAT has undertaken variety screening, basic crop nutrition, plant population, spacing and crop protection trials with a view to developing a balanced rotation with cotton. 5.3 Relative Profitability 5.31 There has always been concern that food crops were more pro- fitable than cotton and that, as a result, farmers would prefer growing food crops to planting cotton. In fact, on the one hand, food crops face limits to their expansion and, on the other hand, there is considerable complementarity between cotton and food crops. 5.32 While. the yield reductions from later planting and labor profiles may preclude some food crop expansion, it would seem that not only the certain market for cotton paid at known times is attractive but that food crop marketing is uncertain. This latter factor may be important for any large scale food programs. Togo is regarded as normally self-sufficient in food. Thus, in two or three years out of five, farmers who concentrate solely on food production could find it difficult to market their produce and could suffer heavy storage losses. Togo exports food crops from time to time but the Government does not always permit export with the result that the farmer cannot be sure of this market. Given the present marketing - 105 - uncertainties, the farmer may therefore be more interested in increasing food crop yields to release land or labor for cotton growing than producing large food surpluses. The interest in improved seed and the fact that the higher potential agricultural areas have the largest percentage of land down to cotton suggest that the farmer believes that cotton offers the more certain method of turning his surplus labor after subsistence into cash. This may change in future. In its policy dialogue with Government, IDA has stressed the potential for food crop production and exports through promotion of storage facilities and lifting of the export ban on food exports. These measures would enable farmers to improve their income, reduce interseasonal price variations and contribute to food security. The Monitoring and Evaluation Unit of the second project should examine the husbandry practices of cotton and non cotton farmers to see whether in fact cotton farmers have a better level of food crop husbandry and whether in fact the farmer believes there is a limit to the marketing of surplus food crops. 5.33 The uncertainties of the degree of effectiveness of the food crop component must lead to speculation about the relative importance of cotton and food crops in the smallholder's economy and to whether total food crop production per se was capable of a much faster rate of expansion than was achieved (and hence the chances of achieving targets in the first project). The evidence so far gained suggests that cotton and food crops have a degree of complementarity. On the face of it, food crops, maize in particular, have a higher net return per hectare and per labor day than cotton. However, maize yields (and returns) decline quite quickly the later the crop is planted. Cotton, on the other hand, can withstand later planting and, in some areas, is planted second after the maize. It has a somewhat different labor profile to maize using more labor later in the season. The two crops grown in combination therefore tend to optimise labor. This is particularly so on the plateau area where maize is planted as a first cycle crop with cotton following as a second cycle crop. 5.34 Cotton is a major cash crop. Whilst farmers clearly give subsistence food crops their highest priority, they also grow maize and other food crops as cash crops. And there is some evidence that they weigh the profitability of the two. Thus, in periods following drought or following input cost increases affecting cotton the number of farmers growing cotton declines and one only presupposes that they grow food crops instead. Yet if the weather in the preceding season is good resulting in good food crop harvests (or when there are cotton price increases) more farmers grow cotton presumably at the expense of food crops rather than continue to grow food crops for cash (despite the apparent profitability). The hypothesis raised in the above paragraph would seem to fit the available yearly production data and various observations made during the course of the project. VI.INSTITUTIONAL ASPECTS 6.1 Institutional Environment SOTOCO is one of the family of CFDT sponsored cotton development companies in French speaking West Africa. They were originally single crop oriented production companies supported - 106 - by single crop oriented research institutes but have become involved in food crop development. CFDT usually takes equity participations, has a management contract and provides to these companies a series of valuable services, in particular technical assistance and export marketing. When SOTOCO was set up as an exception to the general rule, CFDT's role was confined to supplying personnel and technical advice. 6.02 Since SOTOCO did not have full control over the cotton sector (para 1.13) it had no assured revenues but depended instead on the liquidity, good will and cooperation of other agencies: the Ministries of Finance, Rural Development, Rural Equipment, OPAT, CNCA and the cofinanciers IDA and FAC. This complexity taxed SOTOCC's financial and management capacity to the full and made its task extremely difficult. 6.03 Structural Constraints SOTOCO had twin objectives: it was a public service oriented state company but operated also a commercial venture requiring a profit oriented approach. These objectives were not always compatible. The RDII preparation mission explored the extent of these contradictions with Togolese authorities who were perfectly aware of their existence. It proposed long term solutions (para 6.08) but concluded that there were hardly any ready answers in the short term, besides operational improvements. The dilemma was that a public service oriented behavior appeared to be irrational from the point of view of a commercial venture, and vice versa. Thus, with the farm population's welfare in mind, SOTOCO had the tendency of looking for technical solutions irrespective of costs. For instance, the need to supply inputs to farmers in a timely fashion led SOTOCO to advocate their procurement and storage one year in advance which involved high financial costs. A commercial venture would have tried to avoid these costs by procuring inputs annually in time for planting. This was in fact a recommendation made by all IDA missions but the Government procurement machinery could not operate under these conditions. 6.04 SOTOCO also faced drawbacks because it was a state company: In the first place SOTOCO was subject to Government guidelines on recruitment, salaries and fringe benefits. It could not offer attractive salaries to qualified nationals so that it never succeeded in hiring a national chief accountant, for instance. Despite the cost difference, it is still easier for SOTOCO to hire expatriates rather than nationals. Partially as a result of this constraint SOTOCO has a high calibre expatriate team but no national counterparts prepared to take over. In fact, the follow-up project, RDII, calls for an increase in the number of expatriates but has included a major training and fellowship component to try and redress the situation. 6.05 A second drawback was that SOTOCO had no access to financing by mixed sector Banks, access which would have been valuable at a time when Government financing did not come forward. Mixed public-private companies had shown interest in equity participation in SOTOCO's commercial activities. One of the reasons why the Government excluded such participation was that it did not wish other than public involvement in a company which was also promoting food crop production, a national priority. - 107 - 6.06 Approach to Rural Development: SOTOCO's involvement in food crop promotion has always been a point of debate. Prior to SOTOCO's existence, the Ministry of Rural Economy, through its local agencies, the SORADs, had been in charge both of food crop and cotton development. Subsequent efforts consisted in taking responsibilities away from the Ministry of Rural Economy and the Ministry of Rural Development, its successor organization. The first step was the creation of SOTOCO in 1974 which was put in charge of cotton development; the second one was the technical necessity to make SOTOCO also responsible for promotion of food crops in rotation with cotton, thereby further weakening the existing Ministries. Probably as a result of the perceived danger of further dismantling existing ministries, SOTOCO did not receive a legal mandate to promote food crops: its charter was never changed to include them officially in its activities, although this did not prevent SOTOCO from becoming active in this field. 6.07 In 1981, the responsibility for food crop development became a major issue during RDII preparation. Having noted SOTOCO's limited success in food crop development, the Minister of Rural Development, who is also the president of SOTOCO's Board of Directors, questioned SOTOCO's competence and interest in food crop production, and insisted that food crop production was a national priority goal that could not be left to a state company that was, for all practical purposes, single crop oriented as shown by its results. Instead he wanted SOTOCO to take care only of technical aspects of cotton production under the supervision of his ministry's regional development agencies, successors to the SORADs. VII. ECONOMIC RESULTZ 7.1 Farmer Benefits 7.11 Results: The SAR estimated that farmer benefits would on the average result from a 50% increase in the net value of crop farm production from CFAF 76 000 to CFAF 113 000 per annum and a doubling of net cash income from CFAF 32 000 to CFAF 64 000. More than 75% of this increase would come from improved food crops. There is a major difficulty in fully ascertaining results: the lack of hard farm level data. Indeed, SOTOCO did not focus on the farm unit as such and its records on food crops are incomplete (para 3.54). 7.12 For new plantirs, the average cash income per farm growing 0.62 ha of cotton rose by CFAF 29 200 in PY5 and for old planters by CFAF 15 700. The cash income probably explains to a large degree the farmers' continuing interest in cotton production (Annex Table 8). 7.13 Relative Profitability. The Identification and Appraisal missions had found that food crop production was more profitable for farmers than cotton production and IDA had always urged that Government should make cotton production more attractive by adjusting the producer price of cotton upward, taking into account progressive elimination of input subsidies. - 108 - 7.14 During project implementation, Government adjusted farmer incentives several times. The first change came in December 1976 - after the departure of and unknown to the appraisal mission - when the Government abolished the farmer contribution towards meeting the cost of insecticides. In 1977 - as a result of IDA insistence during negotiations - the Government increased producer prices from CFAF 48/kg to CFAF 60/kg. The result was a 50% increase in net returns to labor for cotton farmers which, once the effects of the 1976-1977 drought years had worn off, may have set the stage for the spectacular increase in cotton production over the following years. The second adjustment in 1981 which took place in two stages was initially less successful. The Government announced a tripling of fertilizer prices in March 1981 which caused a reduction in the number of farmers, probably the marginal ones, and a 10% drop in production in 1981/82 season. The subsequent increase in producer prices in December 1981 by CFAF 5/kg of seed cotton only partly compensated for the preceding cost increase and reduced returns per manday to the 1978/79 level at current prices, and to pre - 1977/78 levels in real terms, assuming a 10% annual rate of inflation. Nevertheless, the price increase was sufficient to induce in 1982/83 a 16% increase in the number of farmers, an 11% increase in yields and an almost 30% increase in total production. 7.15 Despite adjustments in the producer price of cotton, returns to food crops continue to be higher than returns to cotton, which has not prevented more and more farmers from growing cotton. The explanation of the apparent paradox is probably in the behavior pattern of farmers who grow both food crops and cotton. In the first place the cotton area is small compared to the food crop area in rotation with cotton: 25% in the Plateaux, 15% in the Central, 16% in the Kara and 11% in the Savannah region. It is even smaller if compared to the farmers' total area under food crops. Secondly, there is not always competition between food crops and cotton but a significant complementarity (Section 5.3). 7.2 Government Revenues 7.21 All project costs were expected to be recovered by PY4 and to steadily increase over the next six years. Two studies have recently examined those aspects: one by LBI (see PREFACE) and another one by SEDES. Both firms analyzed the cotton sector in 1981/82, the only year for which sufficient data were available. Neither report included an analysis of the Togolese textile industry. 7.22 In February 1983, LBI examined the cotton price structure, excluding taxes and transfer payments, from the farm level to the world market. It found that in 1982/83 total cost per kilogram of fiber amounted to CFAF 669 but fetched only CFAF 557 on the world market, a deficit of CFAF 92/kg. - 109 - Cost and Revenue per kg of Cotton Fiber at 1982/83 World Market Prices CFAF/kg % Farmer 157 23 SOTOCO 446 67 OPAT 67 10 Total Cost 669 100 Income 557 Deficit 92 Source Louis Berger International INC, Etude des Prix Agricoles, Rapport final, vol 2: Coton, Juin 1983, Table 5.1, P.66 LBI held out the hope that revenues could possibly cover costs and could become positive in 1985 if yields and world market prices rose as anticipated. 7.23 The SEDES report is less pessimistic than the LBI report. It estimates the total value of cotton production from the 1981/82 season at CFAF 4.0 billion including value added of CFAF 2.6 billion (65%) and imports of CFAF 1.4 billion (35%). It considers that the cotton sector account is globaly balanced but only at the expense of a fall in the farmers' real income during project implementation. It also notes that the share of farmer income from cotton production is about the same (21%) as that of salaries of SOTOCO employees (22%) who seem to be the real project beneficiaries. In 1981/82 SOTOCO employees numbered less than 1200 and participating farmers 38,000. 7.24 The weakness of the two reports is that they have a narrow data base (1981/1982). Extrapolations from their findings may therefore be premature, particularly since the two reports do not reach the same conclusions: the LBI report concludes at a deficit and the SEDES report at a global balance of the cotton sector account. Nevertheless, one conclusion can be made, namely that the cotton growers' share is only 23% while SOTOCO's and OPAT's shares reach 67% and 10%. respectively. Assuming that the cotton account was globaly balanced in 1981/82 a change in relative shares could only be achieved at the expense of the other beneficiaries or through a net transfer from Government. If the cotton sector was unbalanced in 1981/82 and 1982/83 as suggested by LBI, it has actually received a net transfer from Government. An obvious policy prescription for the future would be to make an effort at reducing SOTOCO and OPAT costs and raise producer prices correspondingly. 7.3 Economic Benefits 7.31 The appraisal mission estimated that direct benefits would result from the incremental output of cotton, maize, sorghum and groundnuts - 110 - and lead to an economic rate of return of 32% over the project's 10 year life. Risks included yield responses, quality of extension staff, logistic transport problems and farmer adoption rates of new agricultural practices. The SAR suggested that the corresponding ERR was typical of projects introducing high productivity techniques on annual crops grown traditionally with low yielding methods- 7.32 The PCR recalculated tLa ERR along appraisal assumptions. It updated costs and benefits to 1982 levels with the help of MUV indices for foreign exchange and GDP deflators for the local currency component. Under these conditions, the recalculated ERR over 10 years was 29% against the Appraisal's 32%. This shortfall is due to lower than expected food production. A price comparison shows a major change has taken place in the meantime in IDA's price forecasts: the July 1983 forecast reaches only 75 to 90% of the 1977 forecasts. If the latter is used to value project output, the ERR drops to 11% over a 10 year project life period. 7.33 It is interesting to compare the ERR of the first and second projects. The RDII appraisal has calculated its ERR over a 20 years' span. It has also assumed that benefits would not any more rise after project implementation. Under these same conditions, RDI's ERR is 13% compared to RDII's 21%. 7.34 The sensitivity analysis shows that, as presently designed, the project is not economically viable if it is only based on cotton and that food crop benefits are essential for its success. However, Cotton production is based on a sound technological basis and can certainly be maintained at the existing levels. Food crop production levels are likely to be improved under the follow-up project and would also benefit RDI. Similarly, under the follow-up project, control of tiansport and financial costs would be significantly increased, benefiting once again RDI. With an increased emphasis on food crops, the project would become a full fledged rural development project rather than remain a cotton production project. VIII. FOLLOW-UP PROJECT 8.11 The follow-up project RDII, which the Bank approved in October 1982 would continue the expansion of cotton production but would place more emphasis than under the first project on food crops grown by participating farmers. Greater usage of animal traction would also be promoted. SOTOCO would continue to be the project executing agency and its efficiency would be upgraded through staff training and provision of technical assistance with particular emphasis in redressing past weaknesses and upgrading of field services. Associated infrastructure required to support these objectives (feeder roads, village water supply, some buildings, would also be provided. In addition to its production objectives, the project would aim to build up SOTOCO's managerial and field services capacity so as to provide the future option for a possible restructuring of SOTOCO into a specialized cotton agency and a separate nationally managed extension service. The project would also provide support to Government's national seed multiplication program. - 111 - 8.12 Although RDII has a more secure base, it faces similar problems. But IDA has learnt from past experience and has tightened the conditions Government is supposed to meet: - Subsidies: While it did not touch the principle of full insecticide subsidies, IDA obtained Government commitment to phase out fertilizer subsidies by 1987. Producer prices: a price analysis unit in OPAT was set up to study all cost elements, and submit annual reports. Price levels would then be determined in consultation with IDA. This is a more formal procedure than under RDI. - Input Procurement: Inputs under RDI were procured by Government and pre-financed from their own sources, with SOTOCO playing a marginal role. Under RDII, inputs are definetely to be procured by SOTOCO and financed by OPAT, farers and cofinanciers. - Finances: As under RDI, the RDII appraisal mission has attempted to ensure that SOTOCO would be financed to as large a degree as possible under the bareme. Under RDII arrangements, most SOTOCO costs are covered under the barZme so that the need for Government subsidies would be reduced. Overall, on the basis of the work done under RDI, RDII has a good change of reaching its objectives. Its management has been strengthened through a greater number of expatriates. The problem is the weakness of the national counterpart team which will require intensive training before it will be fully capable of assuming responsibilities. To this effect, RDII has an in-built major training component. IX. IDA PERFORMANCE A. Design 9.11 Bank Involvement. In 1975/76, IDA had a rather limited experience of the Togolese scene. IDA had only just granted a credit for a Cocoa/Coffee Development Project (Cr. 503 - TO, dated August 6 1974) which was to be carried out by the SRCC, a specialized single crop oriented state company. IDA was also involved in preparing the Maritime Region Development Project (Cr.638 - TO, dated June 16, 1976), to be carried cut by the SORAD Maritime, a broad based regional development agency. At this stage, IDA had not shown any preference for either type of executing agency. 9.12 In its CPP of August 1974, Programs had listed RDI as the top priority after the Coffee/Cocoa and the Maritime Region Project for both project and country reasons. - Project reasons: agriculture had been previously neglected by Government and the Bank; the project had both cash and food crop components; it would have a favourable balance of payments impact and affect a large number of farm families belonging to the poorest segments of the Togolese population; - 112 - it would contribute to a more equitable regional development by concentrating or. the more backward areas of the country. - Country reasons: past lending to Togo had been quite limited and the project pipeline did not include many suitable projects. The project identified by RMWA met the above criteria: it was high on the Bank's as well as the Government's priority list; a well documented feasibility study, project promoters and cofinanciers were available. Despite the absence of an agricultural sector survey, IDA felt confident to proceed with the project. 9.13 Financing. Preparation and Appraisal missions had identified SOTOCO financing as a major problem area and they had devised a mechanism agreed to by Government, which was to ensure that SOTOCO received necessary funds in the right amount and on time. This mechanism broke down temporarily because the Government experienced severe financial difficulties as a result of the downturn of its own cash flows (para 1.14) which also prevented Government from erecting the ginning capacities it had under- taken to finance from its own funds (para 3.43). 9.14 The question is whether IDA could have foreseen in 1976 the financial problems that have subsequently affected Togo. There are no indications in the project file that IDA had been aware of unfavorable financial developments. If it had been, it could have made alternative arrangements: (1) IDA could have insisted, for instance that SOTOCO be allowed responsibility for exports of cotton to ensure an independent income. Such arrangements would however have required profound changes in Togo's administrative set-up and would therefore have been unrealistic; (2) IDA could also have insisted on a more direct role in financing the new ginneries, the timely construction of which was essential for project success. Instead IDA satisfied itself with strong Government assurances that the ginning facilities would be available (para 3.43), assurances which turned out to be insufficient. Project staff at their level, had designed a project which subsequently suffered from financial developments beyond their control. 9.15 Institutions. Another important problem which never became severe but which was always present in the background was that of institutions. IDA accepted the appraisal mission's argument that SOTOCO was more efficient than the SORADs and that SOTOCO should therefore be used as the executing agency (para 2.21). Despite initial financial and management problems, SOTOCO organized efficiently the services it supplied to cotton growers and succeeded in the promotion of c-tton production. The appraisal mission for RD II expressly confirmed SOTOCO's success and proposed to keep it for another five years. Nevertheless, the choice of SOTOCO as the executing agency, the importance it has assumed because of cofinanciers' support may have introduced in the agricultural sector an institutional change which needs rationalising in relation to the other agencies involved in the agricultural sector. - 113 - 9.16 Optimism. In the absence of a prolonged country experience,the appraisal mission assumed that SOTOCO was capable of following tight construction schedules for buildings and feeder roads; providing extension, input and marketing services; launching animal traction; promoting food crops like cotton; making food crop seeds available like cotton seeds; obtaining funds from Government, OPAT, CNCA and cofinancers, as and when required. They erred in their belief that SOTOCO could manage all these activities simultaneously and efficiently from the start. What is remarkable is how well SOTOCO did despite all the constraints it faced. 9.2 Project Supervision. 9.21 Between projzcL igning in October 1977 and closing in July 1982, IDA supervised the project 9 times, the last one in November 1981 on the occasion of the 'D II appraisal. Staff spent on average 11.5 man days in the country for the first 8 missions for which data are available. In 1979, the project was rated 3 twice (once for status and once for trend): 2 and 3 in September followed by 3 and I six months later which improved subsequently to 2 and 1. Headquarters sent agriculturists, financial analysts and economists to supervise the project. It was assisted by an occasional RMWA road engineer who looked after the road component at infrequent intervals. 9.22 Supervision missions had to contend with management, financing and marketing problems. They came to a head in September 1979 when IDA threatened to suspend its project participation if these problems were not solved. At the time, the incumbent general manager was in prison and expatriate managers changed at frequent intervals. The Government had neither provided its financial contributions for 1978 and 1979 nor set up the Special Account for input procurement, nor provided the marketing credit for the coming season. Finally, the coming marketing season required urgent attention if repeated losses were to be avoided: marketing teams, emergency storage facilities were to be set up and efforts were to be made to finally erect the ginneries. 9.23 This energetic Headquarters intervention produced results. A new general manager was appointed who organized efficient marketing teams and emergency storage facilities; the Government obtained refinancing of its ginneries from CCCE; it started releasing funds for the project; a new technical and a new financial manager were appointed in 1980 and 1981. The RD II appraisal mission of November 1981 which also supervised the RD I for the last time found that the project did not have any major problems and rated it 2 and 1, a rating that was maintained until closing in July 1982. X. CONCLUSIONS 10.01 The project suffered initially from country wide financial problems which hampered its start-up and delayed its activities, delays which could not be completely overcome subsequently. Under these circumstances, project management bad to ccntend with problems of an unforeseen complexity which it was not equipped to handle. That it - 114 - nevertheless achieved a solid degree of success is remarkable, although it is unfortunately not easy to document in the case of food crops for lack of a MEU. 10.02 The project was conceived as a rural development project with a view to improving the living conditions on a broad scale through the expansion of cotton and food crop production. It has substantially reached its cotton production objectives but has fallen short of its food crop targets. Farmers' cash incomes and Government foreign exchange earnings have broadly increased as anticipated. Government cash flows are expected to become positive at the latest by 1985. The ERR is estimated at 21% over a 20 years life period similar to that of the follow-up project. It would, however, decline to 14% if only stage II food crop benefits are assumed. 10.03 The project has shown that SOTOCO can, with the crucial help of expatriates, efficiently organize the production and processing of a cash crop. SOTOCO concentrated on cotton development, thereby making a success of an important project component. The follow-up project, RDII, attempts to overcome constraints experienced by RDI, and opens the door for a future reorganization. T 0 G 0 RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CR.741 - TO) PROJECT COMPLETION REPORT Anticipated and Actual Project Costa in CFAF millions 1977/78 1978/79 1979/80 1980/81 1981/82 1982/83 AE ACT AE ACT AE ACT AE ACT AE ACT AE ACT Buildings 35 101 255 18 174 87 161 249 112 216 737 671 Vehicles and equipment 141 - 216 102 177 15 346 783 375 18 1255 918 Road equipment 52 - 181 127 - 12 - 2 - - 233 141 Sub-total 228 101 652 247 351 114 507 1034 487 234 2225 1730 Fertilizers - 29 114 146 97 248 90 (152) 108 68 409 339 Insecticides - 113 209 256 75 164 99 (182) 111 5 494 356 Salaries 178 32 229 118 257 182 239 304 226 223 1129 859 Other operating costs 161 315 394 542 536 888 527 1097 511 508 2124 3350 Sub-total 339 489 946 1062 965 1482 950 1067 956 804 4156 4904 TOTAL 567 590 1598 1309 1316 1596 1457 2101 1443 1038 6381 6634 Rate of exchange: CFAF/US$ 209 201 226 287 341. Cost in US$ million 2.8 6.5 7.1 7.3 3.0 26.7 Source : SOTOCO and SAR. Controller's Department for exchange rates. AE : Appraisal Estimat ACT : Actual - 116 - Annex TO G 0 TABLE 2 RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CR. 741 - TO) PROJECT COMPLETION REPORT Schedule of Disbursement (US$.1 000) Appraisal Revised Actual Actual Estimate A E after as % of cancellation 1 FY 78 3 06/30/78 4 1 450 1 450 FY 79 1 12/31/78 2 3 150 3 150 822 26 3 1 134 06/30/79 4 4 850 4 850 1 804 37 FY 80 1 2 066 12/31/79 2 6 550 6 550 2 507 38 3 2 913 06/30/80 4 8 250 7 458 3 876 58 FY 81 1 4 440 12/31/80 2 9 950 9 158 7 079 77 3 8 813 06/30/81 4 11 650 10 858 9 174 84 FY 82 1 9 663 12/31/81 2 13 350 12 558 11 209 89 3 12 552 06/30/82 4 14 000 13 208 13 149 99 FY 83 09/30/82 1 14 000 13 208 13 208 100 1/ : US$ 792 000 declared misprocured in March 1980 - 117 - Annex Table 3 TO G 0 RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT NO. 741 - TO) PROJECT COMPLETION REPORT Cotton (BOU) Production in TOGO 1976-77 1977-78 1978-79 1979-80 1980-81 1981-82 1. Production Data Number of planters 18 724 13 220 25 972 40 696 46 171 38 092 Production (t) 6 310 4 205 13 512 20 030 23 862 21 201 Area (ha) 8 960 6 740 15 867 26 310 29 130 23 840 Yields (kg/ha) 704 624 852 761 819 903 Area/Planter (ha) 0,48 0,51 0,61 0,65 0,63 0,62 2. Income (CFAF '000) Gross Income 308,502 239,924 849,592 1,210,186 1,325,224 1,328,960 Input costs 58,470 21,055 42,449 66,487 95,747 231,895 Net Income 250,032 218,869 807,143 1,143,699 1,229,477 1,097,065 Net Income/Planter 13.4 16.6 31.1 28.1 26.6 28.8 Net Income/ha 27.9 32.5 50.9 43.5 42.2 46.7 3. Technical Data Fertilized area (ha) 6 821 5 349 11 277 18 977 25 867 20 410 Fertilizer input (kg/ha) 222 231 207 184 207 224 Treated area (ha) 8 359 6 573 13 406 25 226 28 534 23 273 Insecticide input (1/ha) 10.7 15.8 15.6 17.3 14.5 18.4 Source : SEDES, Evaluation Economique de l'ActivitE Coton au Togo, Avril 1983. T0G0 RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CR. 741-TO) PROJECT COMPLETION REPORT Stage III Food Crop Areas 1979/80 1980/81 1981/82 1982/83 AE Actual AE Actual AE Actual AE Actual Maize 1 580 770 2 870 1 280 4 450 3 430 6 030 6 950 Sorghum 2 680 - 5 040 590 7 880 1 565 10 640 2 165 Groundnute 1 600 100 2 200 520 2 700 1 640 3 200 2 340 I TOTAL 5 860 870 10 110 1 620 15 030 6 635 19 870 11 455 Target Achievement 15 24 44 58 Source : SAR Annexe 4, Table 12, for appraisal estimate (AE) Synthese Agronomique, D.Salvy, April 1983, SOTOCO. - 119 - Annex TABLE 5 TO GO RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CR. 741 - TO) PROJECT COMPLETION REPORT Food Production Estimates a/ Naize PY1 PY2 PY3 PY4 PY5 PY6 Productions, tons 460 1470 1915 2880 5130 Value, F. million.E 244 77.9 101.5 152.6 271.9 Sorghum b Production, tons - 150 515 876 876 1206 Value, F. million d 96 33.6 56.1 56.1 77.2 Groundnuts Production, ton7 - 50 310 820 1170 Value, F. million!' - 350. 21.7 57.4 81.9 GPV. F. Million 34 146.5 179.3 266.1 431.0 a/ Macklin memo of October 10, 1983 br Macklin, Annex 2, Table 5 *9/ F. 53/kg 7' F 64/kg Wr F. 70/kg jsr7/EN 12/13/83 - 120 - Annex Tables 6 & 7 T 0 G 0 RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CREDIT NO. 741 - TO) PROJECT COMPLETION REPORT Table 6 Commercial Seed Requirement in tons PY 1 2 3 4 5 Cotton 500 630 760 950 1 140 Maize 30 50 80 Sorghum 9 19 26 Groundnuts 160 220 270 Source : SAR Table 7 Commercial Seed Production in tons PY 1 2 3 4 5 Cotton n.a 1 516 2 006 1 473 1 358 Maize composite 148 171 144 Maize hybrid 26 261 Sorghum 126 42 117 Groundnuts 23 17 60 Source : SOTOCO T 0 G 0 RURAL DEVELOPMENT PROJECT IN COTTON AREAS (CR.741 - TO) PROJECT COMPLETION REPORT Net Returns From Cotton per Hectare 1/ 2/ , Yields Gross Cost of Net Index Return Years kg/ha returns inputs returns 1976/77 to labour F/ha F/ha F/ha - 100 F/md 1976/77 700 33,600 11,775 21,825 100 159 1977/78 614 36,840 3,775 33,065 152 241 1978/79 859 51,540 3,775 47,765 219 341 1 1979/80 746 44,760 3,775 40,985 188 299 1980/81 810 48,600 3,775 44,825 205 327 1981/82 886 57,590 10,525 47,065 216 343 1982/83 971 63,115 10,525 52,590 241 384 ----------------------------------- ;--------------- ---------------------- ------------------------------------ 1/ : Producer price: F 48/kg in 1976/77. F 60/kg from 1977/78 to 1980/81 and F 65/kg from 1981/82 to 1982/83 2/ : Fertilizer : 225 kg/ha; F 15/kg from 1977/78 to 1980/81 and F 45/kg from 1981/82 to 1982/83 Batteries : F 400/ha. 3/ : 137 md/ha. SAR Annex 4 Table 2. Source : Synthese Agronomique par D.Salvy, SOTOCO, April 1983. Evaluation Economique de l'Activit& Coton au Togo SEDES, April 1983. T 0 G 0 RURAL DEVELOPMENT PROJECT IN COTTON AREAS CREDIT NO 741 - TO Economic Analyses Tin CFAF million) PY1 PY2 PY3 PY4 PY5 PY6 PY7 PY8 PT9 PY10 Incremental Costs F.E. Component a/ 532 1 108 1 464 1 933 1 587 1 164 1 247 1 257 1 290 1 323 Projet Local Cost Component. b/ 238 603 849 1 152 868 868 868 868 868 868 ERR Period Family Labor - 265 676 804 810 932 1 013 1 069 1 069 1 069 Years Total Costa 770 1976 2989 3889 3056 2964 3125 3194 3227 3260 10 Incremental Benefits 10 Cotton 1331 2546 3257 3590 10 Food Crops 34 147 179 266 10 Total Benefits - 1365 29-3 -34-4 3856 4 43-' 4820 5090 5206 5283 10 Net Benefits(1977 .am) (770) (611) (296) (453) 200 1470 1692 1896 1976 2023 29 10 Price adjustment F 269 287 488 919 687 488 372 272 256 10 Ajustuent IB (779) TEO) T583) T941) TI9) 783 1204 1524 1704 TI W 11 10 Benefits remain constant at PY6 level, Total Coats 770 1976 2989 3889 3656 2964 3128 3194 3227 3260 20 Benefit. Cotton 1331 2546 3257 3590 3923 392A5' 3923 3923 3923 20 Food crops 34 147 179 266 431 431 431 431 431 20 Total Benefits 1365 2=9 3 3856 4354 4354 4354 4354 4354 20 Price adjustmentdL. 269 287 488 919 687 488 372 272 256 20 Adjusted benefits 0=9 2 TW 7 3W 3W 3W 1W I 20 Ad uted et benefit (770) (860)_. ()8)) (941) (719) 703 738 788 855 86 13 20 a uV adjusted actual costs from PY 1 to 5 V GPDV adjusted actual costs from PY I to 5 a o'Assualis the same progression of benefits as in SARI dTaking account of July 1983 projections IBRD18010(PPA) U P P E R VOLTA JANUAEYI911 II. I MAUJRITANIA -le .....-.MAL.I NIGER /GUINEA, & NIGERIA FOOD CROPS N søå/ a, Mil INDUSTRIAL CROPS Rice Groundnuts - Arar/ild' Groundnurs . Arachide K A A Niat c,, ,-Niamtouuo Sorghum, Mil RJuda. Sorgho, Mi/ Yam /gnome Kara Co.E NabarN Nut Rioc r - oz CE N T.' R Maizo Sotoubo0ui Bambara Nut- /gnomø, Ric T OG O Cean RURAL DEVELOPMENT PROJECT IN COT TON ARE AS Cm»a PROJET DE PEVELOPPEMENT RURA C DANS LES REGIONS COTOVIERES P L A T CrA,, MAIN CROPS AND CLIMATIC ZONES_Oj alm PRINCIPALES CULTURES ET r - 'palmrs b ZONES CL/MATIQUES Apa Lui!. N- Pm-f ar M- T C~ossac~ -20 D 0o. do Rio 1. KILOMETER c ~. Io rO NOYSE T og o 0 TOGO ad~~f- rUG OMARITIME REGION RURAL DEVELOPMENT PROJECT PROJET DE DEVELOPPEMENT RURAL DANS LA REGION MARITIME -- - -T ECOLOGICAL ZONES/ZONE8 EOOLOGIQUES nopa Forest Reserv4s Rn~ ømeti .. p ~ Sovannah Zone &ume - 6 45 .\~Coastal Plateau ( Terres de barre) . .......) Depleted Coastal Plateau ftret d# a., ed4w (Terres de barre degraddes) [1iiI ] Lama Depression (Bado)oèmwuon wU fOsea 4 EZICocstal Sands SpWu crm R Lowland Rice Developmmnt ft==Af V Vegetables owitummfi f Irrigated Rice Development 0'z%#" Marshy Areas lwnt inendahM obligbo --90- |sohyets in Millimeters fohrte en mm Major Road% oa n. P-n Regional Boundares Limitn nnN - .,. .. ;- -- International Boundaries Frontitm f..q KILOMETERS 0 10 20 30 N .N 18 D 118 , Mar c .976 ... ...., , roAcenA NM EA 3 6,15 4 1I N 0 T?h4 PPA å11,i 1 hisNI on RGI )I- R~Jsband f~i, ~c -0 A.AT~'Y TOACAALOME I sé,0n ro Ae l I BRD 17808 (PPA JANUARY 1984 DAM5 T 0 G 0O RErENUES TUBEWELLS MARITIME REGION RURAL DEVELOPMENT PROJECT FORRAGES WAREHOUSES PROJE TDE DE VEL OPPEMENT RLIRA L DA NS I A REGION MA RI TIME MAGASINS FAO WAREHOUSES MAGASINS FøA0 DISPENSARIES DISPENSAIRES BUILDINOS B ATIMENTS TRACKS Tolpevie, Tomei Kondi PP7ES Wonougha E FEEDER ROAOS OUVERrURE DE PISTE m .Zove .o NSECONDARY ROADS Kpodil Agbolouv Kouv' Tokpli ROUTES SECONDAIRES Ep_ o_o_ MAIN ROADS .. Asgomé 'ko Kondi! N ROUTES PRINCIPALES Ahipi RAILWAYS TABIG80 -CHEMINS DE FER - RIVERS MARSHY AREAS FLEUVES ZONES INONDABLES Ga anyangan SECTOR HEADOUARTERS 6!30- -6'30' Dagbali SdÉ0E DE SECTEUR un Amegnran . .. va SECTOR BOUNDARIES gbotop4 .it,g , LIMITES DE SECTEUR TSEVIE REGIONAL SOUNDARIES G H A N A o o@ LIMITES DE REGIONAL Badia . nlaweINTERNATIONAL BOUNDARIES olo Kovis oholod Anlal Aklakau \ damie LIMITES D ETAT * isin oe Kpogamn VGN anov MALI . NIGER VAimrndgn ,N .i. P'g To ' rand.Popo 0 10 20 30 40 VOLTA Sange .eKILOMETERS ·~y. S eu n NEHO VOJ E -A doo ENi le-N IVORY TOGCN/NIGERIA --Aerwo COAST/ >ilFAmce cmwahm d~~rn ~on wld fir hib« d~ n#wdp da M C G HANA FN m*F In #a e n a =~@ afun ... -0 Ae LoMiE IBRD 17809(PPA) UPPEIR VOLTAI --' .......M. LAu NIGER i _ N M. d u -- GRI NaNd KA z RURAL DEVELOPMENT PROJECT IN COTTON AREAS PWOJE;r DE DEVEZ0PPEMENT AteLZ 7 . OA4NS ZES REGIONS C0rONIIERES SOTOCO INSTITUTIONAL INFRASTRUCTURE INFRASMUCrTURE INSIMMTONNEUE DE IA SOTOCO0 r ~Ct. Ar~ P- " A- OC ab E SNtT 5~0 soc0wal o lN COTTONAAREAS4 AINFRATISTRUCTURE 5ooc R Nl.,a ClgME O .. mirplc.i. C Ner- 0 20 , O ks o Aneibo XILOMETERS LM 'r JANUART1984 IBRD I 6296CPPA) UPPER VOLTA ANUAY 10BC MALI NIGER MO J - 1 cow VOLV '- SA A N N ""'''1 - / ATIANTGC OCEAN G. H A-N A - G H AAK- A R A b N.ran.goues.ed C N T AL TOGO / SECOND RURAL DEVELOPME s ' .4 . PROJECT IN COTTON AREAS z DEUXIEME PROJET DE DEVELOPPEMENT RURAL A DANS LES REGION COTONNIERESrI 0 Ne. A lar.r? P L A i E -A U V JeIaA.s " t~ E.shvg C- R 5d Fo-5 North C-rm,al ( ~ mN- unh S0.th C-remi ( aC~ -£.da P o.nd. M67 e . - ..drn. aI 0 23pao 40 60 !10 01 8~. .- , d I1ER w*L