Cameroon - Cocoa Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 5140 PROJECT PERFORMANCE AUDIT REFORT CAMEROON COODA PROJECT (LOAN 1039-CM) June 18, 1984 Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the performance of their alicild duties. Its contents may not otherwise be disclosed without World Bank authoriEation. WEIGHTS AND MEASURES I hectare (ha) = 2.47 acres - 1 kilometer (km) = 0.624 miles 1 square kilometer = 0;39 square miles 1 kilogram = 2.204 pounds I metric ton 2,204.6 pounds 1 liter - 1.057 U.S. quarts ABBREVIATIONS CCCE - Caisse Centrale de Coopfration Economique (French Development Bank) DPR - Department of Public Roads FAC - Fonds d'Aide et de Coop&ration (French Aid Agency) FONADER - Fonds National de Dveloppement Rural (Rural Development Fund) IRAT - Institut de Recherche Agronomique Tropical (French Research Institute for Tropical Agriculture) IRCC - Institut de Recherche du Caff et du Cacao (French Coffee and Cocoa Research Institute) LPAC - Lutte Phytosanitaire et Anti-Capsid (Campaign against Cocoa Pests) ME - Ministry of Equipment OED - Operations Evaluation Department PCR - Project Completion Report PPAR - Project Performance Audit Report RKWA - Regional Mission in Western Africa SODECAD - Sociftf de Dfveloppement de Cacao (Cocoa Development Company) SOCOODER - SociftE Cooperative de Dfveloppement Rural (Rural Development Cooperative Agency) SPC - Service des Pistes CacaoyZre (Cocoa Feeder Road Unit) USAID - United States Agency for International Development ZAPI - Zone d'Actions Prioritaires Intfgrfes (Priority Zones for Integrated Development COUNTRY EXCHANGE RATES Name of Currency (Abbreviation) CFA Franc (CFAF) Year Exchange Rate Appraisal Year Average US$1.00 = CFAF 250 Intervention Years' Average US$1.00 = CFAF 230 Completion Year Average US$1.00 = CFAF 290 PROJECT PERFORMANCE AUDIT REPORT FOR OMCIAJL USE ONLY CAMEROON COCOA PROJECT (LOAN 1039-CH) TABLE OF CONTENTS Page No. Preface ..............................................................0. 1 Basic Data Sheet ........... .................................... iii Highlights ...................................................... v PROJECT PERFORMANCE AUDIT MEMORANDUM I. PROJECT SUMMARY ....................... .......1 Introduction ..... ... ............................ I Project Formulation .. .......... ............... .. 3 Project Implementation ............ ..................... 4 Project Administration ...................... ...... 6 Project Impact oo..... oo-............ .................... 7 II. MAIN ISSUES ....... ........... . . ..... 8 A. Cocoa Development in Western Africa ..... ......... 8 B. Compliance with Loan Covenants ..................... 10 C. Cameroonization ..................................... 10 D. Feeder Road Development .................. ....... 11 E. Future Perspective ................................... 14 Table 1: Lending for Agriculture .................................. 17 Table 2: African Production of Raw Cocoa ......................... 18 Table 3: Disbursements by Category ..........o.................... 19 Table 4: Expatriate Staff ........................o................ 20 Organigrams ........................................... ............ 21 Supplement - Compliance with Loan Covenants - Case Study ........... 23 Attachment - Comments from Cofinancier oo....... ............. 35 PROJECT COMPLETION REPORT I. Introduction ..............o..................... o 41 II. The Cocoa Sector ............. ........ 00 . ....... ... 41 III. Project Formulation ............. .................. . 42 IV. Project Implementation ................ ..... 48 V. Project Justification o .... o ............. .. .......... .... 58 VI. Institutional Development ..... ........... . .......... . 60 VII. Borrower's and Bank's Performance ...o...... ......... 61 VIII. Impact of the Project - Lessons Learned .................... 62 IK. Conclusions o ..... . -....................... ............. 62 Annex 1 - Estimated and Actual Costs Ma> IBED 10805 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. J PROJECT PERFORMANCE AUDIT REPORT CAMEROON COCOA PROJECT (LOAN 1039-CM) PREFACE This is a performance audit on the Cocoa Project in Cameroon, for which Loan 1039-CM for US$6.5 million equivalent was approved in September 1974 with the Republic of Cameroon. The project was co-financed by the French bilateral agency - Fonds d'Aide et de Coopfration (FAC) - which provided a grant of US$1.0 million equivalent, and Caisse Centrale de Coopfration Economique (CCCE), which made a loan of US$5.5 million equivalent. The project was expected to be completed by December 1980 and the Loan account to be closed in June 1981. The closing date was extended to December 31, 1981, and the Loan was fully disbursed following a final disbursement on May 10, 1982. The audit report consists of an audit memorandum prepared by the Operations Evaluation Department (OED) and a Project Completion Report (PCR) dated April 5, 1983. The PCR was prepared by the Uestern Africa Regional Office. A draft report dated November 1982, "Rapport d'Achevement du Programme Cacao, 1974/5-198112", prepared by FAC for the Sociftf de DEveloppement du Cacao (SODECAO), the executing agency, was used in preparing the PCR. In view of the extensive supervision of the project in its late stages, and the availability of this FAC report, no special completion report mission was mounted. The audit report is based on a review of the Project Appraisal Report, No. 462a-CM dated August 1, 1974, the President's Report, No. P-1485-CM dated August 21, 1974, the Loan and Project Agreements dated September 18, 1974, and the PCR. Project files, including correspondence with the borrower and internal Bank memoranda on project issues, have been consulted, and Bank staff associated with the project have been interviewed. The OED report, Operational Policy geview, Compliance with Loan Covenants, Report No. 4090 dated Septembe: 1, 1982, Supplement: Selected Case Studies, which contained a chapter on the Cameroon Cocoa Project, was also reviewed. An OED mission visited Cameroon in November/December 1983 (the same mission also audited the Second Cocoa Project in Ivory Coast - Loan 1069-IVC). The information obtained during the mission was used to test the validity of the conclusions of the PCR. The detailed analysis of project implementation by the audit mission has been documented in a separate report; the audit memorandum summarizes the recommendations and conclusions of this working document, which is available (in French) from OED upon request. The audit concurs with the principal findings of the PCR and shares its conclusions. However, the audit regrets that more detailed quantitative information was not included in the PCR. The points discussed in the audit memorandum have been selected because of their importance for this and other Bank-supported projects. A copy of the draft report was sent to the Borrower and co- financing agencies on April 6, 1984 for comments. 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 the Republic of Cameroon, the executing agency, as well as project farmers visited in the project area, is gratefully acknowledged. PROJECT PgRFOANCE AUDIT BASIC DATA SHEET CAMEROON COCOA PROJECT (LOAN 1039-C_) KEY PROJECT DATA Actual or Actual as 2 Its Original Plan Current Estimate of Appraisal Estimate Total Project Cost (USSaillion) 23.8 55.5 233 Loan Amount (USS million) 6.5 6.5 100 Co-financiug (US$ uillion equivalent) - Food d'Aide et do Cooperation (France) 1.0 1.0 100 - Caisse Centrale do Cooperation Economique (France) 5.5 5.5 100 Dte for Completion of Physical Components 12/80 06/81 106/a Proportion Completed by Target Date (%) 100 98% Economic rate of return () 26 negative Institutional Performance good marginal Agronomic Performance good sarginal/b Nmber of Beneficiaries n.&. CWLATIVE DISBURSEMENTS FY75 FY76 FY77 F78 FT79 FT80 Fr81 1182 Appraisal estimate (US$ million) 0.8 1.7 2.7 4.0 5.4 6.5 6.5 6.5 Actual (USS million) 0.3 1.4 2.4 3.2 4.6 5.6 5.6 6.5 Actual as 2 of appraisal (%) 38 82 89 80 85 86 86 100 Loan Terns: US86.5 million, 20 years, 5 years grace, 8Z interest Final Disbursement: May 10, 1982 Principal Repaid: USSO.86 million) Principal Outstanding: US$5.64 million) as of September 30. 1983 Borrowers Obligation:/d USS 4.9 3 million) Final Repayment Due: 1010119 4 PROJECT DATES Original Plan Revisions Actual Conception in Bank 1972 1972 Negotiations 06/74 06174 Board/Approval 0O/74 0/05/74 Loan Signing 08/74 09/18/74 Bffectiveness Date 12/18/74 02/28/75 02/26/75 Closing Date 06/30/81 12/31/81 12/31/81 HISS ION DATA Month/ No. of Days Specializations Performance Types of Mission I Tear Persons in Field Represented l a If Trend Problems /h Identification 1972 - - - - - - Preparatioql 1972/3 - - - - - - Appraisal 10-11/73 5 100 A,A,E.F.R - - - Supervision I 12/74 2 10 NE 1 2 - Supervision II 05/75 2 14 Z,FO 2 2 F,T.M Supervision III/k 11/75 1 5 F 2 2 N Supervision IV 06/76 1 7 F 2 2 N,T Supervision V 12/76 2 20 F.C 3 2 T..P Supervision VI /1 06/77 4 44 F.A,E,H 2 2 T,N,P Supervision VII 11/77 2 24 F,T 2 1 T,P Supervision VIII/a 04/78 1 13 T 2 2 M,T,P Supervision IX/n 06/78 3 24 F,T.YP / Supervision I 06/79 2 18 F.D 3 2 N.P.T Supervision XI 12/79 1 2 T 3 2 M,T,P Supervision XII 04/80 2 20 T,E 3 2 H,T,P Supervision XIII 06/80 3 36 A,A,E 3 1 M.T.P Supervision XIV 03/81 2 21 A.T 3 1 T,P Supervision XV 10/81 1 6 A 2 2 F Supervision Total OTHER PROJECT DATA Borrouer Republic of Caaroon Executing Agency Socitf de Dfveloppement du Cacao (SODECAO) Fiscal Tear of Borrower July 1 - June 30 Preceding Project None Follw-on Project None Footnotes on the following page. - iv - la From date of signing. /b Target yields not achieved. 7- Number of beneficiaries lower than implicit at appraisal. Ti includes exchange rate adjustments as per Statement of Loans dated September 30, 1983. /e A " Agriculturalist; E - Economist or Agricultural Economist; F - Financial Analyst; FO - Forester; R - Roads Engineer; T - Treecrop Specialist; C - Cocoa Specialist; H - Marketing Specialist; D - Deputy Divison Chief; and TP - Young Professional. /f 1 - problem-free or minor problems; 2 - moderate problems; and 3 - major problems. S 1§ improving; 2 stationary; and 3 - deteriorating. /h M - ManaBerial; F =Financial; P - Political; and T - Technical. /i Does not include separate missions by programs staff or specialists to review certain components (such as feeder roads). The cofinancing agencies had staff participate in most missions. The feeder roads component has been supervised separately but continuously throughout by staff from the feeder road unit in Abidjan. / Preparation by consultants, supervised by RMWA (Regional Mission in Western Africa). /k in addition, a roads engineer supervised the feeder road component in December 1975 and again in May 1976. 7T Special in-depth review mission. I In-depth review of technical aspects. 7-n Supplementary supervision necessitated by exceptional problems. /o go ratings; BTO only. -v - PROJECT PERFORMANCE AUDIT REPORT CAMEROON COCOA PROJECT (LOAN 1039-CM) HIGHLIGHTS Cameroon, with a population of about 9 million, is one of Africa's most diverse countries. It has a wide range of climatic zones, ecological conditions, population densities, ethnic groups and traditional cultures. Until the advent of rapidly rising petroleum output, agriculture had been the mainstay of Cameroon's economy, but the sector still offers opportunities for development. Cameroon's agriculture is dominated by smallholder farming. Progress in raising smallholder productivity has been slow, but continuing efforts are being made to identify more successful approaches. Cameroon is among the world's leading cocoa producers, fifth after Ivory Coast, Brazil, Ghana and Nigeria. Production increased from about 35,000 tons in 1946 to around 110/120,000 tons at present. The Cocoa Project was the Bank's fifth lending operation for agriculture in Cameroon. The project was to rehabilitate about 35,000 ha of existing cocoa and to plant up to 15,000 ha with high-yielding hybrid varieties. Extension support, training, cooperative development and adaptive research were also included, and 950 km of rural roads were to be improved. The project was to be executed by a newly established cocoa development agency, SODECAD (Soci6t& de Dfveloppement du Cacao), which was to be equipped and staffed under the project. Project implementation during the first two years was reasonably satisfactory, but in time a series of problems seriously affected project performance. The project was essentially completed within the time-frame envisaged, but at double the cost. Moreover, targeted yield levels were not reached to any significant degree, and incremental production was therefore minimal. Under these circumstances, the rate of return at completion has been assessed to be negative. The project suffered from substantial organi- zational and institutional weaknesses and serious technical problems. Proj- ect design included over-optimistic assumptions regarding yield improvements; the technical packages proposed under the project were not adopted to any great extent or, because of socio-economic conditions, were adopted in substantially modified form which negated their yield effects. The project agency was maladministered, and its weaknesses were pervasive. This resulted ultimately in de facto suspension of disbursements; while this action caused reorganization and improvement, this came too late to be of substantial benefit for project implementation. Management and organizational weaknesses were compounded by erratic Government policies with regard to cocoa industry pricing, farm inputs and credit. The broad objectives of the project were to improve the livelihood of farmers in the project area by substantially -vi - increasing cocoa production, and to demonstrate that the development and rehabilitation of the cocoa sector was possible by replicating the project or its successful components. The project failed in these objectives. Nevertheless, although the price was high, the following lessons have been learned from this experience. Planning for further cocoa development in Cameroon has changed, and for the immediate future, smallholder cocoa development would be integrated as a component in rural development projects (PPAM, paras. 22 and 36-37, and PCR, paras. 8.01-8.03). While institutional development did not take place during the project period, following its reorganization, SODECAD began operating more efficiently and is today considered a strong development organization (PPAM, paras. 16-18 and 21-22; and PCR, paras. 4.06-4.09, 4.22-4.23, 4.27-4.28, 6.01-6.03, 7.03 and 9.02). Another lesson is that more attention needs to be given during appraisal and supervision to all aspects of the succession of expatriates on the basis of realistic assessments, and that the concept of designated counterparts is not an effective way to arrange for such succession (PPAM, paras. 27-29). Other points of interest are: - a substantial number of covenants were Included in the legal agreements, but these were only partly adhered to; compliance with loan covenants was specifically studied by OED in the framework of a special study (PPAM, paras. 25-26 and Supplement; and PCR, paras. 3.22 and 4.34-4.35); - coordination of feeder road development in the project area was not very effective (PPAM, paras. 12 and 30-35; and PCR, paras. 3.19-3.20 and 4.03-4.05); and - many aspects of this project confirm experience with other cocoa development projects in Western Africa (PPAM, paras. 23-24). - 1 - PROJECT PERFORMANCE AUDIT MEMORANDUM CANEROON COCOA PROJECT (LOAN 1039-CM) I. PROJECT SUMMARY Introduction I. Cameroon has a population of 9.1 million and covers an area of 475,000 km2. It is one of Africa's most diverse countries, with a wide range of climatic zones, ecological conditions, population densities, ethnic groups and traditional cultures. Its main opportunities for development lie in the expansion of agricultural, livestock, and forestry production; the exploitation of energy and mineral resources;1/ and the processing of agricultural, forestry and mineral products for domestic consumption and export. 2. Cameroon's main economic centers are separated by large areas of low population density, and the country's port facilities and transport net- work also serve landlocked Chad and the Central African Republic. Over the past fifteen years, Cameroon has experienced accelerating output and popula- tion growth. Per capita GNP rose between 1966 and 1980 at a -annual rate of about 3.1 percent and was estimated at US$880 in 1982. Overall, the balance of payments has not posed any particular problems, apart from some short-term deterioration during the 1974-76 period.2/ 3. As of end FY83, Bank/IDA lending to Cameroon amounted to about US$800 million for 43 projects. Of these, 18 projects, with a total lending amount of US$313 million (40 percent) supported agricultural development (see Table 1).3 Within agriculture, lending by subsector was as follows: I/ Cameroon became an oil producer in 1978 and is estimated to have pro- duced some 6 million tons of crude petroleum in 1982. In 1981, Cameroon began producing refined petroleum products. 2/ For further details see the latest President's Report (P-3483-CM, dated March 10, 1983, Third Douala Port Project). 3/ For further details see PPAR Cameroon Second Semry Rice Project (Loan 1512-CM/Credit 963-CM), DED Report under preparation. -2- AGRICULTURAL LENDING BY SUBSECTOR (FY67-83) Number of Total Amount Subsector Projects of Lending % (US$ million) Treecrops 8 172.40 55 Rural Development 5 62.80 20 Irrigated Rice 2 32.70 11 Livestock 2 27.70 9 Forestry 1 17.00 5 Total 18 312.60 100 4. Agriculture4/ has traditionally been the mainstay of the Cameroonian economy, directly accounting, until recently, for about 70 per- cent of employment, 70 percent of export earnings and 35 percent of tax revenues. However, with rising petroleum output, the relative position of agriculture is quickly changing. Between FYs 1978 and 1981, the output of the primary sector (agriculture, forestry, livestock and fisheries) fell from about 32 to 29 percent of GDP. In 1978, agricultural exports accounted for 78 percent of all export earnings (cocoa and coffee alone accounted for 53 percent). But oil exports rose from 1 percent of total exports in 1978 to 52 percent in 1981 and 63 percent in 1982, while agricultural exports fell to 37 percent in 1981 and 28 percent in 1982. These trends are expected to continue. Further, if present urbanization trends persist or intensify, urban food needs will be growing strongly, so that agricultural productivity will have to increase sharply if greater reliance on food imports is to be avoided and employment opportunities created for a fast-growing labor force. 5. Cameroon's agriculture is dominated by smallholder farming, accounting for about 93 percent of total agricultural output, mainly food crops for local markets, and cocoa, coffee and cotton for export. The remaining 7 percent of agricultural output is produced by plantations. Over- all, progress in raising smallholder productivity has been slow, but con- tinuing efforts are being made to identify successful approaches. 6. The cocoa project was the Bank's fifth lending operation for agri- culture in Cameroon (PCR, para. 1.01). OED has already (1981) reviewed implementation experience with this project, but in the specific framework of a review of compliance with loan covenants.5/ For ease of reference, the relevant case study has been appended to this audit memorandum as a supplement. 4/ See also President's Report (P-3305-CM, dated May 6, 1982) on the Oil Palm and Rubber Consolidation Project. 5/ Operational Policy Review, Compliance with Loan Covenants, OED Report No. 4090, dated September 1, 1982. - 3 - 7. Cameroon is among the world's leading cocoa producers.6/ Production increased from about 35,000 tons in 1946 to 65,000 tons ten years later, to a level of around 110,000 tons in the seventies.7/ Concern with the relative stagnation of cocoa production in the early seventies led Government to formulate substantial development plans for the sector (PCR, paras. 2.01-2.07). Production increased to over 120,000 in the early 1980s but has dropped sharply recently because of drought (see Table 2).I Project Formulation 8. Under terms of reference approved by the Bank's Regional Mission in Western Africa (RMWA), consultants prepared the feasibility study during 1972-73. Preparation work was financed by the French bilateral aid agency, FAC. Later, both FAC and CCCE assisted in project financing under parallel arrangements (PCR, para. 3.13). The draft study was reviewed by RMWA staff during missions in March and July, 1973, and the feasibility study was finalized in September. Appraisal took place in October/November 1973. Some reservations were expressed about project design, but on the basis of the results of earlier attempts to improve cocoa productivity, a consensus was reached that the project should proceed (PCR, paras. 3.01-3.03). 9. Negotiations took place in June 1974 and the project was approved by the Board on September 5, 1974. The Loan and Project agreements were signed on September 18, and the Loan was declared effective, after one extension of the due date, on February 26, 1975. 10. The project as approved would be carried out over a seven-year development period and would involve:9/ (a) rehabilitating about 35,000 ha of existing cocoa and the new planting of up to 15,000 ha with high-yielding hybrid cocoa; (b) strengthening extension services to help farmers control disease and Improve their cultivation practices, and providing farmers with necessary inputs and credit facilities; (c) providing training for some 230 extension workers and 2,000 new cocoa farmers; (d) improving 950 km of rural roads; 6/ Cameroon ranks fifth after Ivory Coast, Brazil, Ghana and Nigeria. 7/ Analysis of the World Cocoa Market, World Bank Staff Commodity Working Paper No. 8. 8/ Preliminary production estimates for 1982/83 are for just over 100,000 tons and for 1983/84 below 100,000 tons. 9/ Project description taken from appraisal report No. 462a-CM, dated August 1, 1974, para. 3.07. -4- (e) adaptive research on cocoa and food crops; (f) equipping and staffing a new agency-Soci&t& de DEveloppement du Cacao (SODECAO)-to. carry out the project and take over responsibility for the development of cocoa production in the project area; (g) employing consultants to: (i) advise on the organization of SODECAO; (ii) evaluate the progress and impact of the project; and (iii) prepare a follow-up project; and (h) encouraging the formation of farmers' groups to participate in cocoa marketing and the supply of inputs and credit. Project Implementation 10/ 11. Project implementation during the first two years was reasonably satisfactory, but a series of problems subsequently seriously affected project performance. These are described in detail in Chapter IV of the PCR The most important points can be summarized as follows: (a) the decision to entrust SODECAO with the implementation of the feeder road component led to substantial delays and cost overruns; (b) after the project had been underway for a year, Government combined grades I and II into a single export grade. However, Government subsequently reduced the price differential between grade and sub-standard cocoa and also neglected to maintain the cocoa price at a competitive level with foodcrop prices; (c) credit became a problem when a newly created cooperative agency became responsible for marketing. As a result, the SODECAO/ FONADER11/ link to marketing was severed. SOCOODER12/ was not willing to get involved in etdit recovery; and (d) FONADER experienced funding problems, and partly as a result input deliveries were late. 10/ This as well as the next two sections in this chapter are adapted from the PcR. l1/ FONADER is the National Rural Development Fund under this project responsible for input procurement (PPAM, para. 16). 12/ SOCOODER is the Rural Development Cooperative Agency. - 5 - 12. Road development suffered seriously when SODECAO rather than DPR became responsible for this component. Additional equipment had to be pur- chased at substantially higher costs than envisaged. A road program needed to be prepared and a road unit set up. All this delayed the start-up of road work to late 1976. Even thereafter road construction remained inefficient. In the end, only 342 km were reconstructed compared to 950 km estimated at appraisal and at four times the appraisal cost estimate (PCR, paras. 4.03- 4.05). By contrast, cooperative development was good. By December 1976, 283 pre-cooperative centers had been established, well over the appraisal target of 180. These functioned well until affected by institutional changes in the cooperative structure introduced by Government in 1977 (PCR, para. 4.10). Training has also been successfully executed under the project. However, until placed under proper management, the extension service was of limited effectiveness (PCR, para 4.28). Similarly, until new seed gardens were established as part of the SODECAO reorganization, the provision of seed was a chronic problem (PCR, para. 4.30). 13. Research was well executed under the project, and various trials were undertaken. However, the research program has been criticized as being too academic. In any case, research activities had limited relevance for the ongoing project (PCR, para. 4.31). Effective monitoring and evaluation with feedback for project implementation did not take place (PCR, para. 4.32). Given the serious problems encountered during project implementation, no follow-on project was prepared.131 Neither is a single crop/cocoa project deemed desirable; any further support for cocoa development will be inte- grated as a component of rural development projects. 14. The project was implemented more or less within the time-frame envisaged at appraisal, but at more than double the originally estimated cost (see basic data sheet and PCR, paras. 4.25-4.26). Substantial efforts were made for cocoa rehabilitation through intensive crop maintenance and protection. However, the exact area covered under this component was never known (PCR, para. 4.13); furthermore, it became clear that pest control measures were inadequately applied, and that due to the labor requirements involved, proper application became uneconomic. As a result, yield increases due to rehabilitation have been marginal at best. SODECAO records indicate that 13,645 ha were newly planted, somewhat less than the appraisal objec- tive. However, it has become clear that the prospective yields of 800 kg/ha could not be achieved under everyday smallholder conditions given prevailing technical and economic constraints. A supervision mission, in early 1981, estimated that at best only one third of the trees would yield at levels estimated at appraisal. No data is available on plantations abandoned or destroyed, and in the absence of detailed monitoring data it is difficult to assess incremental production due to the project; in any case, it is thought to be minimal. Furthermore, such production increases might have been partially offset by production foregone as a result of the project. As the PCR indicates (PCR, para. 5.02-5.03), incremental production within the 13/ As the cofinancier rightly points out, in the meantime, SODECAO is being supported with Government budget resources (Attachment). - 6 - project area has only been marginally higher than that outside, another indication that project production impact has been minimal. Under these circumstances, the economic rate of return has been assessed to be negative. 15. The project suffered from substantial organizational and institutional weaknesses and serious technical problems. The latter relates to the fact that the technical packages proposed under the project were not adopted to any great extent or adopted in a substantially modified form which negated their yield effects. Partly, this is a reflection of the substantial labor requirements involved in applying these packages in an area where there are serious labor constraints. In retrospect, labor requirements and availability were severely underestimated at appraisal (PPAM, para. 20). The institutional objective of the project was development of the cocoa sector. No positive impact occurred during the project implementation period, especially since some of the sector interventions by Government were counter-productive. Project Administration 16. As envisaged at appraisal, project organization was quite complex. As the existing ZAPI (agency responsible for integrated agricultural development in a specific region) in the project area was weak, a specific cocoa development agency (SODECAO) was created in February 1974 to implement the project. In addition, FONADER (Rural Development Fund) was responsible for the procurement of farm inputs and their delivery to SODECAO, which was responsible for onward delivery to farmers; capsid spraying was under the responsibility of a special program, LPAC (campaign against cocoa pests) under the Ministry of Agriculture; rural road improvements were to be executed by the Department of Public Roads (DPR) but under the aegis of SODECAO; and research came under the responsibility of two specialized French research institutes: IRCC and IRAT14/ (PCR paras. 3.09-3.10). 17. In fact, the above arrangements proved to be too complex. Rather than being executed by DPR, the road component was executed by the feeder road unit in SODECAO, and no coordination took place with the Ministry of Equipment. The relationship between FONADER and SODECAO has been poor, and FONADER experienced delays in ordering and distributing farm inputs. Furthermore, with regard to credit, the situation became aggravated when Government decided to create a new cooperative agency responsible for marketing. This severed the links between credit, extension support and marketing. For a substantial time, credit distribution and recovery were low as a result. The situation only improved when a new credit agreement was signed between SODECAO and FONADER in July 1981. 18. SODECAO was created to implement the project, and did not previously exist. It should, therefore, be recognized that staffing, training and organizing a new agency is a challenging job under the best of 14/ IRCC, the French Coffee and Cocoa Research Institute, and IRAT, the French Research Institute for Tropical Agriculture. - 7 - circumstances. initially, and as envisaged, the Center-South ZAPI was gradually but effectively integrated in the SODECAO. Serious management problems became apparent thereafter. Senior management positions were vacant during much of the time, or the incumbent was temporary in place, or the persons appointed were deemed unqualified and/or unsatisfactory by the Bank. Not surnrisingly, this resulted in lack of coordination, weak administration and ptar organization. Together with inadequate accounting and poor financial control, this significantly hampered project progress. After two in-depth reviews in 1977 and 1978, the Bank requested Government to reorganize SODECAO (PCR, paras. 4.06-4.09). Government appointed a management committee to implement organizational changes, but the process of reorganization and reform proved cumbersome and slow. When, in addition, serious financial irregularities became known, the Bank intervened and made arrangements for de-facto suspension of disbursements from October 1979 onwards. A new management team was installed in May 1980 and an audit completed thereafter. However, only after the signing of a satisfactory credit agreement between FONADER and SODECAO in August 1981 were disbursements resumed (PCR, paras. 4.21-4.23). With new management, SODECAO began to operate much more efficiently and is presently regarded as an agency with good financial and technical management (PPAM, para. 22). However, this came too late to be of substantial benefit for project implementation. Project Impact 19. The broad objectives of the project were to improve the livelihood of the farmers in the project area by substantially increasing cocoa production and cocoa exports, demonstrate that the rehabilitation of the entire cocoa sector could proceed effectively by either replicating the proj- ect or adopting its successful components. The project failed to reach these objectives to any extent.15/ 20. The primary cause was the over-optimistic assumption made in proj- ect design concerning yield improvements which could be obtained on small- holder farms from the intensive application of fungicide to control the black pod fungus, which is responsible for reducing yields of much of the cocoa crop in Cameroon. Because of the climatic conditions, the disease is much more serious in Cameroon than in other Western Africa cocoa-producing count- ries. It became apparent that the yields necessary to justify the twelve sprayings annually could only have been obtained with a more intensive pro- gram (PCR, para. 9.01). The practicability of giving such optimum treatment over the entire cocoa zone was substantially over-simplified. 21. A second and important factor was the maladministration of the project executing agency. Its weaknesses were pervasive - failure to recruit and retain capable management, absence of discipline, financial irregulari- ties, lack of field coordination, inadequate extension services and others. This was compounded by erratic Government policies with respect to cocoa industry pricing, farm inputs and credit. 15/ The cofinancier assesses project results and project design from a different perspective (Attachment). - 8 - 22. Nevertheless, although the price was high, lessons have been learned from this experience. Cameroon's planning for cocoa development has changed. Rather than crop intensification and new planting of cocoa, the approach is now to rehabilitate and replant existing cocoa to maintain pro- duction at existing levels. For the immediate future, smallholder cocoa development would be integrated as a component of rural development projects16/ (PCR, paras. 8.01 and 9.04). Furthermore, after appointment of a new management team in 1980, SODECAO began to operate more efficiently. Today, SODECAO is considered a strong development organization with generally good technical and financial management. Too late to benefit implementation of the Cocoa Project, it nevertheless now provides a strong focal point for future rural development in the Center-South province (PCR, paras. 4.23 and 9.02). For the time being, however, the Bank is no longer directly involved with SODECAO or cocoa sector development, but a future Center-South Rural Development Project is being considered.17/ II. MAIN ISSUES A. Cocoa Development in Western Africa 23. As can be seen from the table below, the Bank has financed nine cocoa projects in Western Africa to date. A further eight projects contained 16/ A similar change of approach took place in the Ivory Coast, where after the Second Cocoa Project (see PPAR, OED Report under preparation), smallholder cocoa development was integrated as a component in the Center-West Agricultural Development Project (Loan 2167-IVC). 171 However, its preparation has not advanced sufficiently to allow its inclusion in the monthly operational summary of Bank and IDA proposed projects. -9- cocoa components.18/ Eight of the nine cocoa projects have been completed; the only ongoing project is the Second Cocoa and Coffee Project in Togo.19 Cocoa Projects in Western Africa Country Project Loan/Credit No. PPAR No. 1. Cameroon Cocoa Ln. 1039-CH /a 2. Ghana Eastern Region Cocoa Cr. 205-GH 3526 3. Ghana Ashanti Region Cocoa Ln. 1181-GH /b 4. Ivory Coast Cocoa Ln. 686-IVC 1899 5. Ivory Coast Second Cocoa Ln. 1069-IVC /c 6. Nigeria Western State Cocoa Ln. 764-UNI 1859 7. Nigeria Second Cocoa La. 1045-UNI 3957 8. Togo Cocoa and Coffee Cr. 503-TO 4552 9. Togo Second Cocoa and Coffee Cr. 945-TO /d /a This audit. lb PCR to be circulated to the Board shortly. /c PPAR under preparation. T7- Project being implemented. 24. Experience with cocoa development in Western Africa has been analyzed in more detail during the audit of the Ivory Coast Second Cocoa Project. The PPAR for that project (OED Report under preparation) is being circulated separately; the PCR on the Ghana Ashanti Cocoa Project (OED Report No. 5092 dated 05/24/84) has also recently been distributed. 18/ Cameroon - ZAPI Integrated Rural Development Project (Cr. 776-CM); Ivory Coast - Center West Agricultural Development Project (Ln. 2167-IVC); Liberia - Lofa County Agricultural Development Project (Cr. 577-LBR); Liberia - Bong County Agricultural Development Project (Cr. 700-LBR); Liberia - Second Lofa County Agricultural Development Project (Cr. 1242-LBR); Sierra Leone - Integrated Agricultural Development Project (Cr. 323-SL); Sierra Leone - Second Integrated Agricultural Development Project (Ln. 1138/Cr. 568-SL); and Sierra Leone - Third Integrated Agricultural Development Project (Cr. 1094-SL). 17/ However, cocoa rehabilitation is still supported, sometimes as a component in agricultural development projects, or as a separate project. A cocoa rehabilitation project in Equatorial Guinea has been appraised. A third cocoa project in Ghana is envisaged. However, the immediate needs of the Ghana cocoa sector are taken care of by the recently approved Export Rehabilitation Project. - 10 - B. Compliance with Loan Covenants 25. A substantial number of covenants were included in the Loan Agreement (PCR, para. 3.22), but the Borrower only partly met these loan covenants (PCR, para. 4.34). Compliance with loan covenants under this project was specifically studied by OED in the framework of a special study.20/ As it is relevant for a better understanding of the project, the case study on this topic prepared for the special study has been appended to this audit memorandum as a supplement. 26. In particular, the case study reviewed three covenants in detail: (a) appointment of key staff; (b) training of counterparts; and (c) agricultural practices.21! From a detailed review of management appointments, it is clear that there have been misunderstandings and lack of effective communication (to say the least) between the Bank and Government regarding the appointment of key project staff. Different perspectives on what was needed did not help either. A related issue, counterpart training, only received limited attention (PPAM, paras. 27-29). As noted above (see also PPAM, paras. 14, 15 and 20), the agricultural practices proposed under the project turned out to be inadequate. C. Cameroonization 27. An integral part of project design was Cameroonization. In fact, it was envisaged that local staff would be trained to take over the functions of expatriates within periods ranging from three years (average period) to six years. As it turned out, this was over-optimistic. Not only was more time required, but partly due to the requirements of the feeder road component more expatriate assistance than envisaged at appraisal was also needed as the table below shows (see also Table 4). 20/ See OED Report, Operational Policy Review, Compliance with Loan Covenants, Report No. 4090 dated September 1, 1982, especially its Supplement - Case Study No. 2. 21/ Section 3.01 (a) of the Loan Agreement required the Borrower to carry out the project in conformity with appropriate agricultural practices. - 11 - Expatriate Personnel (Staff-months) 74/75 75/76 76/77 77/78 78/79 79/80 Appraisal 90 90 90 36 24 12 Actual 91 130 132 125 78 59 28. The appraisal gave only limited attention to the practical implica- tions of its Cameroonization proposals. As a result, although counterpart training was envisaged, no funds for such training were included under the project. This was cited by Cameroonian officials as a major reason why such training had in fact not been provided. Counterparts and counterpart training also received limited attention during supervision. In retrospect, it is questionable whether the concept of counterparts was the best solution for efficient replacement of expatriates. It certainly did not work in this case, and the effectiveness of this method has been questioned in other cases.22/ Finally, in 1980, the concept of counterpart staff was abandonned. A supervision mission in late 1980 reported that "it has been agreed in principle not to appoint formal counterpart personnel. The eventual succession to posts held by expatriate technical assistance staff will be assured through structural development and promotion of national managers". 29. The lessons from this and other experiences are that more attention needs to be given during appraisal and supervision to all aspects of succes- sion of expatriates on the basis of realistic assessments, and that the concept of designated counterparts is not an effective vehicle to arrange for such successions. D. Feeder Road Development 23, 30. Feeder road development was an essential part of the project to ensure efficient logistics related to input/output delivery. When the 22/ For example, under the Togo Maritime Region Rural Development Project (Credit 638-TO), the concept or designated counterparts was, during the first supervision, at Bank suggestion, modified; rather, Togolese staff were placed in line responsibilities and the most competent staff selected for succession. Togolization under this project was completed by the end of the project period (see PPAR, OED Report under prepara- tion). 23/ CCCE financed a consultants review of feeder road development in cocoa areas. This section is based on that report which is available from OED upon request, "Remise en Etat de Pistes Cacaoyeres dans la Region du Centre Sud Cameroonais-Bilan des Action mene6s dans ce domaine depuis 1975 par la SODECAO et d'Autres Organismes - Proposition d'Evolution des Modes d'Intervention," January 1981. - 12 - project was prepared, Government therefore requested consultants to prepare an inventory of feeder road rehabilitation needs in the cocoa areas. This inventory covered secondary feeder roads in the cocoa region in its totality, and tertiary feeder roads in the project area itself. The feeder road inventory covered 3,267 km as follows: - secondary feeder roads, of which 2,361 km - roads recently rehabilitated 710 km - roads urgently in need of repair 958 km - roads needing repair subsequently 693 km - tertiary feeder roads 906 km 31. The 906 kilometers identified became the feeder road rehabilitation component of the project. A parallel feeder road project was approved by CCCE, which envisaged the rehabilitation by SODECAO of 287.6 km of feeder roads in an adjacent cocoa area (the Lekie and Mefou departments). 32. Originally it was envisaged that feeder road development under the overall program control of SODECAO would be executed by the Department of Public Roads (DPR). During negotiations, however, this was modified (PCR, paras. 3.19-3.20): SODECAO would be responsible for the execution of feeder road development with assistance from DPR's cocoa roads unit (SPC). SODECAO established a feeder road unit in 1975, and it was envisaged that this unit would be responsible for work planning and programming in close collaboration with DPR and supervision of work execution. Actual rehabilitation was still to be done by SPC on the basis of annual agreements between SODECAO and SPC. However, no agreement could be reached between these two agencies. As a result, SODECAO took over full responsibility for feeder road development and rehabilitation including executing the works on force account. But, as SODECAO's feeder road unit had not originally been set up to do this work, substantial delays occurred, while the unit adapted itself to its new task (PPAM, para. 12). 33. However, SODECAO was not the only agency engaged in feeder road development; SPC and other agencies also undertook feeder road rehabilita- tion. Overall feeder road development in the project area and in the adja- cent cocoa area during the project period is estimated as follows: - 13 - Feeder Road Development SODECAO Adjacent Project Area Area Total Program as appraised 906.0 287.6 1,193.6 Program Execution by SODECAO 342.4 174.6 517.0 by others 264.0 87.0 351.0 Subtotal 606.4 261.6 868.0 Not Part of Program la by SODECAO 50.6 20.7 71.3 by others 266.6 204.2 470.8 Subtotal 317.2 224.9 542.1 Total 923.6 486.5 1,410.1 /a Rehabilitation of roads not part of the inventory included in project proposals. 34. Although SODECAO's feeder road unit faced many problems, the feeder roads it rehabilitated are of good quality, and it is generally regarded as an excellent feeder road unit, especially after it was reorganized as part of SODECAO's overall reorganization in 1980 into a separate well-structured feeder road division (see organigram). However, design was in excess of traffic requirements, and roads were rehabilitated to standards unnecessarily high for this type of road.24/ As a result, they were not classified as rural roads and were not included in the normal maintenance program supported by DPR. SODECAO, therefore, also had to ensure maintenance, although that is not the proper role of a development agency. 24/ The cofinancier does not agree with this assessment (Attachment). - 14 - 35. As the table above shows, SODECAO's feeder road unit only executed just over 40 percent of all feeder road rehabilitation that took place in the project area during the project implementation period. The remaining 60 per- cent were executed by other agencies, mainly SPC. This is a somewhat curious situation: two agencies responsible for feeder road rehabilitation in the cocoa areas must imply duplication of efforts. While to a certain extent there are short-term advantages if projects can manage their own feeder road investment programs, problems frequently occur. If many units undertake feeder road investments, coordination is essential. Furthermore, arrange- ments are needed for maintenance after the investment phase. Both these con- ditions did not exist in Cameroon. In any case, short-term project efficiency is mostly at the expense of long-term institutional develop- ment.25/ It is a pity that the Bank was apparently unable to reconcile its approaches towards feeder road development under the various projects it supported. On the one hand, it pursued a long-term institutional strategy under its Feeder Roads Project (Loan 1494/Credit 749); on the other hand, it pursued short-term expendiency when it agreed that SODECA0 could expand its feeder road unit to be responsible for all aspects of feeder road development. In retrospect, one wonders if stronger Bank support could not have resulted in implementation of SODECAO/SPC agreements regarding feeder road rehabilitation in the cocoa areas. E. Future Perspective 36. Sustainibility of project results, in this case, is not an issue. The reason is that the project has been a failure. During project imple- mentation there were serious management problems, but, more importantly, the technical packages were not adopted as originally designed, as they were not attractive under the socio-economic conditions under which smallholders had to implement, in an area where labor has become a serious constraint, the rehabilitation and planting activities supported under the project. As a result, productivity increased to a limited extent, and the returns on investment are therefore negative. If the project would have faced only management problems, then the recent improvements in this field could have had a positive effect on future returns,26/ but as the problems were substantially technical, it is unlikely that returns on investments can be improved to any extent. 25/ On this subject see also PPAR, Togo Maratime Rural Development Project (Credit 638-TO) and Rural Development Project in Cotton Areas (Credit 741-TO), OED Report under preparation. 26/ An example of such turn-around is the Indonesia Fisheries Project (Credit 211). This project was beset by problems in project design, procurement, construction, maintenance and operation of both the ships and shore facilities. At audit, the returns on this investment were estimated to be negative; but about two years later, project performance had improved considerably as a result of more effective local management. - 15 - 37. Nevertheless, on the basis of the lessons learned from this experi- ence, the point to address is still the future. To a certain extent, the Bank is removed from this effort as it is not directly involved in finaucing cocoa investments in Cameroon for the time being. SODECAO, however, must meet this challenge. After its reorganization, it is now probably in a better position then ever before to do so. il, - 17 - Table 1 PROJECT PERFORMANCE AUDIT REPORT CAMEROON COCOA PROJECT (LOAN 1039-CM) Lending for Agriculture (FY76-FY83) Board Approval Lending Audit Project Name (Abbreviated) Year Source Amount Status (completed) 1. CAMDEV (Rubber) I FY67 IBRD/IDA 18.00 yes 2. Oil Palm I FY69 IBRD 9.60 yes 3. SEMRY Rice FY72 IDA 3.70 yes 4. Livestock FY74 IBRD 11.70 yes 5. Cocoa FY75 IBRD 6.50 /a 6. Niete Rubber FY75 IDA 16.00 yes 7. Plaine des M'Bo R.D. FY77 IDA 2.00 yes 8. SOCAPALM II FY77 IBRD/TW 25.00 /b 9. Rural Development Fund FY77 IDA 7.00 - 10. Camdev (Rubber) II FY78 IBRD 15.00 /c 11. SEMRY Rice II FY78 IBRD/IDA 29.00 /d 12. Zapi Integrated R.D. FY78 IDA 7.80 - 13. Western Highlands R.D. FY78 IDA 8.50 - 14. HEVECAM (Rubber) II FY80 IBRD/IDA 31.50 - 15. Livestock II FY80 IDA 16.00 - 16. Northern Province R.D. FY81 IBRD/IDA 37.50 - 17. Forestry FY82 IBRD 17.00 - 18. Oil Palm/Rubber Cons. FY82 IBRD 50.80 - 312.60 /a This report. 7-b PCR being prepared. 7E PCR received by OED. /d PPAR under preparation. PROJECT PERFORMANCE AUDIT REPORT CAMEROON COCOA PROJECT (LOAN 1039-CM) African Production of Raw Cocoa /a ('000 Metric Tons) 72/73 73/74 74/75 75/76 76/77 77/78 78/79 79/80 80/81 81/82 /b 82/83 /c Cameroon 107 110 118 96 82 107 106 124 120 122 106 Gabon and Congo 6 8 7 7 6 6 6 6 6 6 6 Ghana 418 350 377 397 320 268 250 285 258 225 178 Equatorial Guinea 10 12 15 12 6 7 8 5 8 8 10 Ivory Coast 181 209 242 231 230 304 312 379 403 457 355 Liberia 2 3 3 3 3 3 4 4 4 4 5 Nigeria 241 215 214 216 165 205 137 169 155 181 156 Sao Tome and Principe 11 10 7 8 6 6 8 7 8 8 8 Sierra Leone 6 8 5 6 6 7 7 11 10 8 10 Togo 19 16 14 18 14 17 13 15 16 11 10 Zaire 5 5 4 4 5 4 4 4 5 4 Other Africa 22 8 4 3 3 4 4 8 8 8 8. Total Africa 1,028 954 1,011 1,001 845 939 .859 1,017 1,000 1,043 856 Total Central and South America 288 398 442 410 392 458 510 491 528 520 515 i Total West Indies 42 43 40 41 42 42 47 35 44 50 53 Total Asia and Oceania 39 53 56 60 59 64 73 79 92 113 109 WORLD TOTAL 1,397 1,44 1,549 Y 512 1,338 1,503 1,1 91,622 1,664 1,726 1,533 /a Source: Cocoa Statistics, Gill & Duffus. T Estimate. 7T Forecast. - 19 - Table 3 PROJECT PERFORMANCE AUDIT REPORT CAMEROON COCOA PROJECT (LOAN 1039-CH) Disbursements by Category Category Description Allocation Actual Disbursements 1 Administration costs of SODECAO $ 2,780,000.00 $ 4,287,869.15 2 Equipment and vehicles 940,000.00 1,402,195.63 3 Sprayers and Fungicides 780,000.00 809,935.22 4 Unallocated 2,000,000.00 - - 0 - TOTAL $ 6,500,000.00 $ 6,500,000.00 -20- Table 4 PROJECT PERFORMANCE AUDIT REPORT CAMEROON COCOA PROJECT (LOAN 1039-CM) Expatriate Staff 74-75 75-76 76-77 r77-78 78-79 79-80 (Staff e ths) Appraisal Estimates Deputy Director General 12 12 12 12 12 - Senior Inspector 12 12 12 6 - - Chief Accountant 12 12 12 - - - Training Director 12 12 12 - - - Chief Engineer 12 12 12 - - - - Chief Training Section 12 12 12 - - - Research Officer /a 18 18 18, 18 12 12 Total 90 90 90 36 24 12 Actual Director General - - - 4 4 - Deputy Director General 9 12 12 12 12 5 Financial Director lb 12 12 12 12 12 6 Senior Inspector 11 12 12 - 12 2 - Training Director 12 12 12 12 - - Chief Engineer 12 12 7,5 - - - Chief Training Section 12 12 12 12 - - Research Officer 9 12 12 12 12 12 Chief Input Supply /c - - - - - - Chief Agro-economic Studies 2 12 12 9 - - Chief Soil Surveys - - 3 12 - 12 Chief Workshop - 12 12 12 12 12 Mechanic /d 12 18 14 - - - Road Unit - Chief Field Unit /e - - - 8 12 12 - Chief Field Unit71 - 4 12 8 - - Total 91 130 132.5 125 78 59 /a One permanent position-and part-time consultants. lb Instead of Chief Accountant. 7F Expatriate from October 1981 onwards. /d Dutch volunteer working for both SODECAO and ZAPI. W~ For feeder road development in the project area. 71' For feeder road development outside the project area. k 一訌－ 一小 PROJECT PERFORMANCE AUDIT REPORT CAMEROON COCOA PROJECT (Loan 1039-CM)/ RAPPORT D'EVALUATION RETROSPECTI'VE CAMEROUN - PROJET CACAO (Pret 1039.CM) Sector Organlzatlon/Organigramme du Secteur Warehouse/ Garage/ Sector Chief/ Magasin Secteur Garage Secteur Chat Secteur Rufoli Roads UnIt/ tiglklg Entretien Piste, Deputy/ Adjoint Cocoa Pest Controll Extension/ Applled Remarch/ Tralning/ Lutte Ant1capside Vulgarisation Expedrnentatlon Forniation dAccompagnomnt Chief Operafor/ Chiel Controitef/ Supervisms/ supelvIsms/ Training Officer/ Chef Machiniste Chef Testeur Assistants Assistants Fomialeur Secteur operalors/ Contfollers/ Extension WMers/ Exionsionworkers/ Machinistes Testeurs Encadreurs Encadreurs Forniers/ [:F,,or'iers/ Laborers/ Farrners/ Planteurs P anteurs Manoeuvres Planteurs Wodd Mnk - 2M - 23 - Supplement Page 1 PROJECT PERFORMANCE AUDIT REPORT CAMEROON COCOA PROJECT (LOAN 1039-CM) COMPLIANCE WITH LOAN COVENANTS - CASE STUDY 1/ EXPERIENCE WITH COVENANTS: PROJECT EXECUTION 1. As the project was to be implemented by a new and untested agency, the Bank felt strongly that the success of the project depended crucially on the quality of managers and technicians appointed to. SODECAO's key positions. The Project Agreement therefore specified the management structure of SODECAO and determined that appointments required the concurrence of the Bank. Another covenant called for the employment and training of local counterparts who would gradually replace managers that had been recruited internationally. 2. During project implementation serious disagreements arose between the Borrower and the Bank concerning these covenants. Feeling that these differences caused major management problems the Bank at one point suspended disbursements. A. Covenant: Appointment of Key Staff 3. Section 3.01 (b) of the Project Agreement (PA) states that "SODECAO shall employ .... the following personnel, whose qualifications, experience and terms and conditions of employment shall be satisfactory to the Bank: a General Director, a Deputy General Director in charge of technical affairs, a Senior Inspector, and a Chief Accountant.- Experience . 4. Initially, Section 3.01 (b) seemed to pose few problems. At the time of Board presentation (September 5, 1974) the Government had already appointed a Cameroonian General Director and an expatriate Deputy General Director whose qualifications and experience were satisfactory to the Bank. The Senior Inspector and Chief Accountant were appointed before the end of January 1975. 5. Difficulties started when the Deputy General Director died unexpectedly in December 1974. The Government taplaced him with a French national and then Director General of Agriculture. The Bank reluctantly agreed to the appointment, but only for a six-month probationary period. A Back-to-Office Report of May 1975 commented on the new appointee and said that "it is doubtful that he has (the) management ability required 1/ Taken from OED Report, Operational Policy Review, Compliance with Loan Covenants, Report No. 4090 dated September 1, 1982, more especially its Supplement - Case Study No. 2. - 24 - Supplement Page 2 6. SODECAO was aware of the Bank's opposition to the Deputy Director General but did not agree that management experience was a crucial requirement for the post of Deputy Director General. They felt, instead, that the Director General was responsible for management and that the Deputy Director General was in charge of technical matters. SODECAD considered the new Deputy Director General qualified in that respect as he had 30 years of experience of agriculture in Cameroon and had worked as Director General of the Department of Agriculture. The SODECAO Board, consequently, confirmed him as Deputy Director General. Although the appointment and confirmation was made without Bank approval, the Bank did not formally disagree to the confirmation. 7. SODECAD had pointed out that even though the Deputy Director General was an expatriate and although he was appointed to a position included in the project, his salary was not paid from the Bank loan. This caused the Bank to argue that since he was "being paid out of local funds, consequently the position of technical manager, as foreseen in the appraisal report, is vacant." Bank staff, therefore, recommended appointment of an additional Technical Director. This was not accepted by SODECAO on the grounds that (i) the post had not been envisaged in the Project Agreement; and (ii) the technical expertise, as represented by the new Deputy Director General, was seen as adequate. A February 1977 supervision mission agreed with SODECAD that "the project contains no provision for a separate Technical Director." Instead, it recommended the creation of another post under the title of "Director of Field Operations." The subsequent supervision mission of August 1977 reported that -it is assumed and accepted by Cameroonian authorities that the Director of Field Operations would be an expatriate." The Director General, however, who was later interviewed by OED staff, felt that at that time he was "compelled to create an additional expatriate post that was neither included in the project nor justified". 8. In November 1977, the Bank staff interviewed two candidates and on December 28 proposed one of them for the new post. The Director General of SODECAO submitted this candidate's CV to the SODECAO Board, but the Board judged him to be underqualified. The Board then appointed another person to the post, an act which further strained the relationship between the Bank and SODECAO. 9. A Government enquiry took place in 1978 and the result was that the Director General was transferred out of SODECAO and the expatriate Finance Director and his local co-Director were removed. Consequently, the Government appointed a new Acting Director General. The Government did not consult the Bank about this appointment. The Bank considered the new man to be "relatively inexperienced" but did not formally object to the appointment. 10. In November 1978, the Acting Director General retired for medical reasons. The Government, again without consulting the Bank, then appointed the Deputy Director General as Acting Director General. The Bank had originally been opposed to the appointment of this person to Deputy Director General and felt that now his replacement was -of the most urgent nature." The Bank recommended two candidates for the job of Director General. The Government rejected both because they wanted an "executive with University - 25 - Supplement Page 3 background in Business Administration or Economy with experience in agricultural projects." The Bank maintained that their two candidates had the right profile and expressed in a July 1979 telex that the Government had set "very high standards". 11. Meanwhile, the Government had appointed a group of consultants to prepare a study on the reorganization of SODECAO. The consultants' report suggested that the Director General had too many interlocutors which prevented him from concentrating on policy matters. They recommended that the Director General be relieved from day-to-day responsibilities by a secretariat which was to be a service department with no executive powers. The consultants attributed SODECAO's managerial problems to the existence of a Deputy Director General and recommended that the post be abolished. They also recommended that the cumbersome administrative/financial department be split into more manageable units, and that less people report directly to the Director General. 12. The Bank noted that the Government while accepting many of the Consultant's recommendations, decided to retain the post of Deputy General Manager and not to split the administrative and financial functions. 13. In July 1979 the Government by Presidential Decree set up a Management Committee to enquire into the affairs of SODECAO. The Committee was headed by the Technical Assistant to the President and undertook a thorough review of SODECAO including the report and recommendations of the consultants. On October 19, 1979, the Bank informed the Borrower that it would take action if by November 20 the Government did not -take immediate steps to implement the consultants' recommendations, including rapid nomination of a Director General.- 14. On November 2, 1979 (i.e., well before the deadline date set by the Bank) the Management Committee summoned the representatives of the co-financiers including the Bank's Resident Representative and formally informed them of the decisions taken by the Government on the recommendation of the consultants. The Committee also informed them of the candidate the Government had identified for the post of Director General. In the absence of any formal objection the Cameroonian authorities felt assured that the actions taken by them were considered by all* the three co-financiers as adequate under the circumstances. The Government subsequently proceeded to negotiate a contract with the candidate and his release for appointment as Director General. 15. While the Management Committee was working out the details for implementing its recommendations, Bank staff held discussions with a Cameroonian delegation that had come to Washington for the negotiation of another project. The discussions were apparently not conclusive and no minutes were recorded. The Borrower's representatives felt that they had explained the decisions taken by their Government and had indicated the action being take to implement them. While some Bank staff held that the Government was taking action to remedy the situation, others gained the impression that the Government was not prepared to act on the reorganization and the appointment of a Director General. The legality of suspending - 26 - Supplement Page 4 disbursement was informally discussed with the Legal Department and it was noted that "whilst it might be held to have contravened certain sections of the agreement (e.g., 3.01 and 3.02), measures have already been put in hand to remedy the infringement.*2/ The Bank, however, decided to ask the borrower to refrain from applying for further disbursements.3! 16. The Government was informed by the Bank's Resident Representative on November 22, 1979 of the suspension of disbursement until the Government had decided upon definite measures to redress the SODECAO situation,l/ the Government was also asked to refrain from applying for further disbursement. The Government was surprised about what it considered the Bank's unilateral decision, especially since the Bank representative had not raised any objections at the November 2 meeting when the Government's actions had been announced.5/ It is interesting to note that the other co-financiers did not follow the Bank's example and continued to disburse. 17. In March 1980, a letter was sent to the Borrower which defined the measures which had to be taken to enable the Bank to resume disbursements. One of these conditions was that a new Financial Manager had to be appointed who had qualifications and experience satisfactory to the Bank. The Bank letter also told the Borrower that according to the Loan Agreement, all key positions must be approved by the Bank. 18. An October 1980 supervision mission noted that a new management team had been installed in May 1980, and a number of key posts had been staffed without prior consultation with the Bank. The mission noted that it had "reservations regarding the competence of some of the appointees." A follow-up letter was sent to the Minister of Economy and Plan which said that it was regrettable that the Bank was not allowed to give its opinion on the candidates which bad been nominated, as had been agreed to in the Loan Agreement. In April 1981, however, Bank staff reported that substantial improvements had taken place and that the new management team which had been installed since May was "progressively transforming SODECAD into an effective development agency." a 19. By June 1981, the Bank considered that there were several remaining actions which needed to be implemented, and that if they were undertaken by July 31, 1981, disbursements against expenditures incurred after June 30, 1980, would be resumed. If they were not, steps to close and cancel the loan 2/ Office Memorandum dated August 7, 1978. 31 The documents recording the agreement not to send disbursement applications were apparently not cleared by the Legal Department. 4/ It should be pointed out that Bank concern about the appointment of key managers reflected wide concern about the general financial situation of SODECAO. 5/ Bank staff maintains that the possibility of suspension was discussed at length with Government at various levels. - 27 - Supplement Page 5 would be taken. These actions primarily concerned the signing of a Ministerial Decree and an agreement between the main institutions involved with the project defining the role of each institution and the project's agricultural credit procedures. On July 27, the documents were signed and, after a period of 19 months, the Bank resumed disbursements of the loan and extended the closing date to December 31, 1981. By this time no Cameroonian held a senior management post. Comments 20. Despite a number of missions there was surprising lack of effective communication between Bank staff and the Borrower. Throughout the years of project implementation the Borrower felt that Bank staff was more interested in filling vacant positions with candidates of their choice - all of them expatriates - than in developing the country's own management capability. The Borrower also considered candidates nominated by the Bank to be very expensive and often professionally mediocre and less than competent. Bank staff, on the other hand, appear to have been primarily concerned about SODECAO's serious management problems and objected to the appointment of management personnel without previous Bank approval. B. Covenant: Training of Counterparts 21. Section 3.01 (c) of the Project Agreement stipulates that "SODECAO shall promptly employ counterparts to the internationally recruited members of the management personnel, referred to in paragraph (b) of this Section, and shall make appropriate arrangements for the training of said counterparts to enable them to gradually assume the positions of said internationally recruited management personnel." 22. Although the covenant does not specify any time period within which this program would be implemented, it was understood during appraisal and specifically stated in the President'g Report that .... local staff would be trained to take over the functions of expatriates after periods ranging from three to six years." Experience 23. It is difficult to reconstruct the history of compliance with this particular covenant. The Supervision Missions of 1975 and 1976 do not raise the issue of counterpart personnel. In fact, the first mention of the topic is in June 1977, when a Supervision Mission reports that an important aspect of the new position of a Director of Field Operations would be to -develop management skills in his deputy and in other, lower level Cameroonian personnel with whom he works." An indirect reference to the covenant is made in a May 1978 supervision report which said that the Government -may feel it necessary to increase rather than diminish the numbers of internationally recruited staff in key position.- 24. Afterwards, many of the Bank's references to the issue of counter- part staff are obscure. For example, in the January and February 1980 super- vision reports, covenants 3.01 (b) and 3.01 (c) are treated as one - -The - 28 - Supplement Page 6 employment of key staff and counterpart personnel." The only comment made in the February report is that "A new General Manager has been appointed and recruitment is in hand for other key staff." 25. Finally, in an April 1980 supervision report, the Bank makes its first direct reference to counterpart staff with the sentence "The capability of counterpart staff is being reviewed." In October 1980, another supervision report comments that formal counterpart personnel will not be appointed: "The New Organigram has not yet been finalized but it has been agreed in principle not to appoint formal counterpart personnel. The eventual succession to posts held by expatriate technical assistance staff will be assured through structural development and promotion of national managers." This comment is repeated, word-for-word, in an April 1981 supervision report, the last reference to counterpart staff. Comments 26. The language used in the final supervision reports suggests that from the beginning the covenant has not been complied with, and that SODECAO did not "promptly employ counterparts to the internationally recruited members of the management personnel." The fact that the Bank did not raise the issue of compliance and eventually opted effectively for abandoning the concept of local counterparts, indicates that Bank staff did not feel seriously committed to the covenant. 27. The Cameroonians apparently had a different perception of the issue. Although generally supportive of the concept of counterparts they maintained that "without any provision for funds in the loan agreement ... SODECAO could (not) train personnel for management positions." EXPERIENCE WITH COVENANTS: PROJECT OBJECTIVES 28. Important technical specifications of the project turned out to be deficient. This created problems for the Borrower who was obligated by the Loan Agreement to follow these specifications. Covenant - Agricultural Practices 29. Section 3.01 (a) of the Loan Agreement requires that "The Borrower shall carry out the project ... in conformity with appropriate ... agricultural practices ..." 30. The "appropriate agricultural practices" had been defined in the appraisal report. They were based on consultants' recommendations and specified that: (a) "The main goal of the rehabilitation program would be the control of black pod." This requires "... five to six sprayings in the rainy season, apart from the six sprayings in the dry season. Significant results can be obtained only if a full spraying program is carried out." - 29 - Supplement Page 7 (b) New plantings would be established at a density of 1,600 trees per hectare. It was recommended that the trees be grown under permanent shade and that no fertilizer be used. Experience 31. Initially, supervision missions insisted that the Borrower apply the recommended agricultural practices, particularly intensive spraying. Yet, the appropriateness of the technical recommendations began to be questioned when it was found that cocoa production did not increase as expected, and its quality did not improve. 32. Early in 1977, a supervision mission was made aware of the fact that "... the old age of the trees (in established plantings) made black pod treatment a pretty risky proposal..." However, no action was taken by the Bank at that stage. 33. The issue was finally brought to the attention of the Bank's management during the Fall 1979 Problem Projects Review when it was noted that "... it is now apparent, after two-thirds of the project period has elapsed..." that the assumption that intensive spraying in existing plantations would increase yields by 200 kg per hectare "O..was overoptimistic and will not be attained." The Review concluded that "...it has since been realized that it is not economical for the farmer to spray more than 4 to 6 times a year." From then on the Bank supported the Borrower's proposal for a reorientation of the project objective away from rehabilitation of old trees to new plantings. 34. But the Bank's recommended agricultural practice for new plantings also was not without controversy. Some Bank staff were skeptical about these recommendations noting a number of discrepancies between the technical package recommended in Cameroon and another one suggested in comparable conditions in the Ivory Coast. There were differences in (a) spacing of trees - 1,600 trees per hectare vs. 1,300; (b) use of fertilizers - none vs. fertilizing young trees for three years; (c) planting of trees in permanent shade being mandatory in Cameroon but not in Ivory Coast; and (d) seed requirements of 60 pods/ha in Cameroon as againpt 65 pods/ha in Ivory Coast. Comments 35. It is now generally recognized in the BaUk that the agricultural practices recommended by the consultants were inappropriate. The fact that this could happen reflects not only negatively on the consultants but also suggests some weakness in the appraisal process. Where examination by the Bank's headquarters staff of such recommendations shows them to be questionable, the terms of reference of appraisal missions should specifically require full examination of them in the appraisal report. 36. Supervision missions are responsible for early identification of emerging problems, most of which affect compliance with covenants. While this apparently happened in the case of the cocoa project (see para. 32 above) the follow-up seems to have been less than adequate. As and when such -30- Supplement Page 8 problems are identified subsequent missions should focus more specifically on the dimensions and possible causes of the problems. 37. The Borrower has suggested that the inappropriateness of the technical package could have been established much earlier if there had been a mechanism for appeal to any panel of experts independent of Bank staff involved in the designing and enforcing of covenants. - 31 - Supplement Attachment Page 1 AIDE-HEMOIRE %April 23, 1981) GIVEN TO OED MISSION TO CAEROON 1. For proper assessment of the problems encountered in complying with the main contracts for the appointment of expatriates, one must take into account the circumstances leading up to the approval of the World Bank loan of US$17 million for the cocoa project. 2. Cameroon, as the sixth leading cocoa exporter in the world, has developed a comprehensive system of extension services for farmers and a marketing system used by foreign exporters. The latter purchase cocoa either directly or through intermediaries with low prices paid to the farmer. Because cocoa is grown on small plots of 2-3 ha with family labor, the farmers have been in a weak bargaining position and often have been paid prices lower than those fixed by the Government for their product. To ensure a proper return to farmers, the Government has supported the formation of regional missions with the participation of CINAML/ as advisor. Because this effort to organize farmers was successful, the Government issued two decrees in 1972 establishing two Regional Development Companies for the Priority Integrated Action Zones (ZAPIs) of the Center-South and the East, headed by an expatriate (CINAM) and a Cameroonian counterpart. For the Center-South ZAPI three Entreprises de Progres Local (EPLs) were established, under the direction of CINAM emperts assisted by Cameroonian personnel who are not assigned responsibility. The "Integrated Development Program" of the Center-South ZAPI was based chiefly on an increase in high-quality cocoa production and on monopoly marketing within the zone, in order to obtain a higher cash income for the farmers concerned. 3. The initial concern of the ZAPI was to enstere that the farmers produced higher-grade cocoa and receive the full price fixed by the Govern- ment. During this period the ZAPI sent the cocoa colleeted to exporters for sale abroad. In 1973, however, the ZAPI began to export cocoa directly, finding that the total profit from this operation was substantial. For this reason, it quickly became an active competitor with e.tablished exporters, who continued to export cocoa from other regions* 4. The ZAPI operated successfully in the East. In the Center-South, however, its success was only partial owing to the lack of funds to pay farmers promptly and to strong competition from major exporters in the important region of the Center-South that includes the capital city of Yaounde. 1/ Compagnie d'Etudes Industrielles et d'Amfnagement du Territoire. -32- Supplement Attachment Page 2 5. An analysis of the ZAPI's activities was undertaken by SEDES,2/ a French firm based in Paris, in 1971. Instead of advising on what should or could be done to overcome deficiencies so as to make the ZAPI more efficient and operationally sound, SEDES recommended that the four ZAPI sectors of the Center-South be replaced by a new organization, SODECAO. This agency was established by the Government, with a board of directors and a general directorate made up of a Cameroonian director general and eight expatriates; its purpose is to help farmers organize with a view to increased cocoa pro- duction and sales to exporters, thereby eliminating local intermediaries, but without a monopoly on exports in the region. 6. Although it did not seem very promising to leave responsibility for ensuring higher prices to farmers from foreign exporters to a national agency headed by a Director General and supported by Technical Departments made up almost entirely of expatriates, the Government of Cameroon reluctantly agreed to this because of the active participation in the project by an inter- national financial institution like the World Bank, whose stated policy is to assist in developing the local potential of its borrowers. SODECAO had been given impetus by an agreement concluded with the Bank Appraisal Mission to the effect that one director general would be a Cameroonian and that from the outset local personnel would be trained to replace some of the expatriates after periods of three to six years (Appraisal Report, paragraph iv). 7. Placing our trust in the World Bank's commitments and in its ability to assist in developing local capacity to direct the project, the Cameroonian delegation signed the legal documents proposed by the Bank and presented upon its arrival in Washington one day before the negotiations. 8. Only later was it possible to examine these documents in greater detail in Yaounde, whereupon it was found that: First, there was no provision - as had been agreed in the minutes or correspondence, that the director general would be a Cameroonian. In point of fact, the corresponeing section of the Project Agreement (3.01 b) mentioned management positions, including that of the director general, as being open to -international recruitment" which is the Bank's term for expatriates. Second, the specific undertaking as to the period for the training of local personnel to enable them to replace the expatriates in three to six years was omitted. Instead, the document referred to SODECAO as making "appropriate arrangements" for the training of counterpart personnel to -gradually assume the positions" of the international management personnel recruited (Project Agreement, Section 3.01 c). It was not clear how, without any provision for funds in the loan agree- ment for such "appropriate arrangements", SODECAO could train personnel for management positions. 9. In view of this background, SODECAO has been substantially and increasingly called upon by the Bank to appoint expatriates, particularly because of the absence of a like interest in implementing the agreement as regards the training of local personnel for key management positions. 2/ Societe d'Etudes pour le DEveloppement Economique et Social. -33- Supplement Attachment Page 3 10. The fact that the person approved by the Bank for the position of assistant technical director general - with responsibility for the technical aspects of cocoa production - was an expatriate, who admitted that he had never seen a cocoa plant before coming to Cameroon, has not helped to allay the misunderstandings. 11. SODECAO has continued to resist the constant pressure to comply with the project agreement by filling all positions with expatriates at exorbitant costs", and even with unqualified candidates which it has consis- tently rejected as mediocre and incompetent. 12. At the same time, SODECAO feels that for nearly all of the local appointments that it has been able to propose, premature and rather vague questions have been asked as to whether the candidate had "adequate manage- ment experience." This was noted especially in the steadfast opposition of the Bank supervision mission to the appointment of an expatriate with service in the Directorate of Agriculture as well as 30 years' experience in agri- culture in Cameroon. SODECAO noted that even though he was an expatriate and appointed to a position included in the project, his salary - which incidently was low in comparison to the average salary of over CFAF 1 million per month paid by the Bank and approved for "experts" - was not paid from the proceeds of the Bank loan, but rather from the French Government budget (about CFAF 800,000 per month). 13. As a result of these misunderstandings, the Bank unilaterally ceased disbursement. This point is mentioned not to criticize the Bank, but to learn from the problems caused by inappropriate contracts and the inability to include provisions that could help in carrying out the stated policy of the World Bank: to develop local capacity by drawing upon the experience of this project. We would suggest, for future projects: (a) The full text of- loan documents should be sent to the Government well in advance of negotiations, so that the ministries concerned may examine all of these implications. (b) The minutes of negotiations should be recorded so that the con- clusions reached may be fully understood by both parties, especially regarding points that are not specified in any contract. (c) Loan agreements calling for the appointment of expatriates should be specific contracts providing that: (i) The minimum academic qualifications and experience required for each position are accepted by the Bank and the Government before any haphazard recruitment begins. (ii) When local personnel meeting the required academic qualifica- tions are available, but lack the minimum experience required, they will be appointed on an interim basis and a qualified expatriate adviser will be assigned to supervise them until they have gained the requisite experience. *34- Supplement Attachment Page 4 (iii) If local personnel with the required academic qualifications are not available, expatriates with the minima acceptable requirements or higher qualifications and more experience should be engaged. (iv) In all cases where expatriates are appointed, a detailed and phased program of training can be arranged between the Bank and the Government, with sufficient funds for its implementa- tion, so that local personnel will be trained in time to replace the expatriates at the end of the contract. (d) Before unilaterally invoking provisions that give the Bank, and not the borrower, the means of complying with agreements reached autually, and before the disbursement of funds by the Bank is halted, the Bank should send the borrower a formal statement indi- cating the reasons for its decision. This would enable the borrower to explain its view of the situation. 14. Only when the management of the Bank has reviewed both sides of the question is it possible to decide on the feasibility and correctness of the recommendations of its specialists. In doing so, the Bank should attach greater Importance to the priority objectives of the project than to the procedures for achieving them, or general contracts such as that requiring the borrower and not the Bank to show "due diligence" in the execution of projects. - 35 - ATTACHNENT PROJECT PERFORMANCE AUDIT MENDRANDCK Page 1 CAMEROON COCOA PROJECT (LOAN 1039-CM COMMENTS FROM COFINANCIEE Translated from Fr.nch Caisse Centrale de Coopration Economique (CCCE) Cameroon Office Mr. Cooper World Bank Resident Representative Yaoundi No. 739/PaD/cd/D May 23, 1984 Subject: Cocoa Project Performance Audit Report Letter dated May 11, 1984 signed John H. Malone, Jr. Dear Mr. Cooper: I have studied the referenced documents with great interest. Please find below my comments and suggestions on the Cocoa Project PPAR. Reference to French (1) Points of detail version of - draft PPAR. -Page 10, last line (French text): "cacaotiers" should be replaced by All suggested "cacaoyers." corrections -Page 24, first line (French text): I would suggest the following wording made in French in the French "Halgre les nombreux problkmes auxquels a iti confronte le version but service de ... -1similar -Page 24, line 8 (French text): "classies" instead of "classifiies." changes were -Table 3, column 3: "dollars Etats-Unis"/instead of "dollars Etats-Unie"/. not required -/First/ Supplemental (Annexe), page 9 (French text): the "seed in English requirements" probably relate to "cabosses" (pods) rather than to version. "graines" (seeds). -/Second/ Supplement (Pi&ce jointe),-page 1, para. 22, line 22 (French text): as a "monopole" relates to exclusive selling rights, it would be preferable to use the term "exclusivit& de la commercialisation." (2) Basic points In CCCE's mind, SODECAO was more of a (full-scale) research and development project than a productive enterprise. The balance sheets drawn up* in 1980 clearly showed the value of the lessons that could be drawn from such an experiment for the formulation of subsequent projects under the Cocoa Plan, which formed the distillation of these partial analyses. * Agronomy, cooperatives, roads, training, etc. PPAR comments Many of the CCCE comments are understandable against their notion that this was a pilot project. However, appraisal documents nowhere indicate that the project was designed as such. The audit, therefore, regards the project as a failure. If the project had been designed as a pilot project, results would, of course, have been assessed in a different light. - 36 - ATTACHMENT Page 2 It accordingly seems harsh to me to say (p. 10, para. 13 of French; PPAK, p. 5, para. 13 of English) that "no follow-on project was prepared." The fact para. 13 is that, in view of the serious problems of all kinds encountered by SODECAO, the Government bore out of its budget the costs of the structure until the latter, having demonstrated its regained efficacity, could legitimately claim to act as executing agency for new projects. Similarly, it does not seem quite accurate to assert that the PPAM. project was a "failure" (page 15, para. 21 of French; page 7, para. 21 of para. 21 English) or that "none of its objectives were reached" (page 14, para. 19 of para. 19 French; page 7, para. 19 of English). It is the actual estimating of these. objectives during the appraisal which must be questioned. e aui mission It seems that the yield figures adopted, which are attainable under does not believe optimum operating conditions, were used solely to justify an acceptable rate that the appraisal of return. Increased yield esti- mates deliberately While they are modest (of the order of 300 kg/ha on average for for rate of return SODECAO's cocoa area as a whole), yields have been increased slightly whereas calculation purposes. there has been a definite deterioration in the nonproject areas, where no treatment whatsoever has been done for several years and where the plant as indicated in PPAM material arrived in poor shape at best. (para. 14) A study is under way to quantify SOCECAO's precise impact in comparison with the nonproject areas or those brought into it only recently. The appraisers, of PPAM, (Page 19, para. 26, last line, of French; page 10, para. 26, last course, expected the para. 26 line, of English) "the agricultural practices proposed under the project agricultural prac- turned out to be inadequate"; this did not, however, prevent the appraisers of tices proposed to be the project from rating them as good. good; but as it turned out, they were inade- It is apposite to quote here a comment by Jean Grimaldi,* namely quate. that "cocoa is an antiagronomic crop": - cocoa does not respond to fertilizers. - production does not follow a "logistic law" over the life of the plantation. - the precise causes of disease and pest attack (black pod fungus and capsids) are not known. The practices that were promoted were developed empirically on station and adapted to the economic and ecologic conditions obtaining in Cameroonian cocoa production. . See coments nest - ineffective use of insecticides: the technique employed by SODECAO page (systematic first treatment of all plots, second only on those plots where an infestation test proves positive) seems rea;onable. If there is another more economical technique that gives comparable results, the writers of the report ought to specify it. * Who has devoted 30 years of his life to cocoa produttion in Cameroon, first as researcher and subsequently as Director of SODECAO (in which capacity he is the target of numerous criticisms in the report). - 37 - ATTACHMENT Page 3 - similarly, the control of the black pod fungus in Cameroonian cocoa Comts trees requires ifitensive application of fungicide (para. 20, page 14 of The audit and PCR clearly French, page 7 of English). which is highly labor-intensive. Research point out that the technical undertaken in plant health laboratories with the assistance of IRCC experts to package an proposed at develop micronizable molecules has not been successful. There can be no doubt appraisal turned out to be that application of ULV techniques in cocoa plantations would produce very Inadequate in practice. positive results, as has already happened with cotton. It might have been valid In theory, but that is - (page 24, para. 34 of French; page 12, para. 34 of English): IRD not relevant. a 3 reproaches SODECA0 with having applied unnecessarily high standards in feeder roads. There is a minimum standard below which one cannot go if one wants subsequent maintenice to be possible. Moreover, these standards were unquestionably judged too modest by the, Miiotry of Infrastructure (Department of Public Roads)o since it does not consider them worthy to be classed as rural roads which would have qualified them for maintenance out of the Department*&a operating budgeto Finally, on page 9 of the Supplement (French text -- page 7 of Ci rents English) there are a cirtain number of assertions that are hard to understand It should be remembered or run counter to what is stated in the main text: that the Supplement ias ta1-- verbatim from an - para. 32: "the old age of the trees (.)made black pod treatment earlUer OED document and a pretty risky proposal"(?) simply Included for ease - parh. 33: "Intensive spraying in existing plantations would of reference. Retrospective increase yields by 200 kg/ha" (in itself far from the target figure cments or changes can, of 800 kgthal; is this referring to an additional yield? therefore, not he ade Nor can the effect of the two forms of pest and disease control be under themi conditions. separated? It is difficult for a plantation that does not receive any treatment to exceed a yield of 0 kg/ha of very poor quality. - para. 33: "reorientation of the project objective away from rehabilitation of old trees to new plantings," whereas on page 15, parea. 22 of the French (pag 7, pita. 22 of the English) it is stated that "Rather than crop intensification and new planting of cocoa, the approach Is now to rehabilitate and replant existing I would appreciate it if you would transmit these coements to the Operations Evaluation Department if you consider they could serve to improve the final report. Yours, etc. 1sf J. Charpentier Copy: CCCE Paris coetarcagscn ．•一叔＿＿一、．.-.．一－一A“· ：化 ·一多g一’ l 汕＇兮e&＇為斤·：- 唱鳥 CAMEROON CAMEROOK COCOA PROJECT PROJECT COMPLETION REPORT (Loan 1039-CM) April 5, 1983 (Revised May 25, 1984) Western Africa Projects Department Agriculture 1) .C~) - 41 - CAMEROON COCOA PROJECT (Loan 1039) PROJECT COMPLETION REPORT I. INTRODUCTION 1.01 The project was the Bank Group's fifth lending operation for agriculture but the first in the cocoa subsector. It was preceded by the Camdev (palm oil) Plantation Project, Ln 490/Cr. 100, US$11.0 million, in 1967; the Socapalm Oil Palm Plantation Project, Ln 593 US$7.9 million, in 1969, with a supplemental lending operation, Ln 866, US$1.7 million, in 1972; the Seary Rice Project, Cr. 302, US$3.7 million, in 1972; and the Livestock Project, Ln 983-CK, US$11.6 million, in 1974. 1.02 About 75% of Cameroon's population of approximately 8 million depends on agriculture for its livelihood. When the cocoa project was formulated in 1971, agriculture accounted for 40% of GDP and 70% of the value of exports, principally cocoa, coffee and cotton. In addition to the traditional sector which accounted for 90% of agricultural production, there were large state or foreign owned plantations, mainly palm oil and rubber. 1.03 While the agricultural sector was fairly well organized, agricul- ture development was limited by weak Government institutions and services in the sector. Stabilization funds existed to regul"te prices for the cash crops and there was an agricultural credit mechanism to provide loans for export crops. II. THE COCOA SECTOR 2.01 In 1973 Cameroon was the world's fifth largest producer of cocoa. It produced 110,000-tons of dried beans, about 7% of world production, and had cocoa planted over an area of 380,000 ha. The estimated average yield, thought to be 300 kg/ha, 1/ was low due mainly to black pod, a fungus disease which destroys cocoa pods. Black pod spreads rapidly under high humidity and relatively low temperatures; it is severe in Cameroon where these conditions exist, compared with other West African countries. In,fact, it was estimated that about 50% of the annual cocoa crop was lost to black pod. On the other hand, anti-capsid control was fairly effective. 2.02 For some years prior to the project, the Government had been concerned with the relative stagnation in co-oa production, which reflected that the average age of the trees was increasing and production per tree declining. It as apparent that this could be corrected only by massive 1/ As will be seen, this estimate appears to have been about 25% too high. - 42 - programs of rehabilitation and replanting at a very considerable cost. Moreover, the black pod problem was a prominent consideration in planning the rehabilitation of the sector. The Government sought French bilateral expertise for guidance and toward the end of the 19608 it turned to the Bank as well. 2.03 In 1967, the Government initiated a program to double cocoa production within five years in the Center-South, which was carried out by the Centre-South ZAPI, one of the integrated agricultural agencies created to foster agricultural development in an assigned region. However, extension work was largely ineffective and cocoa marketing arrangements had limited success, with the result that cocoa production increased but slightly, if at all. 2.04 Under the Third Plan (1971/72-1975/76), the Government set a cocoa production target of 150,000 tons, which called for an annual increase of 7%. However, this would have entailed improved cultivation and extension services, appropriate marketing and credit arrangements, and sound black pod control throughout the cocoa production areas. This was unlikely to be possible and the Government wisely decided to make a start in only one region which came to constitute the project area. 2.05 The Ministry of Agriculture, the Zone d'Actions Prioritaires Intigrees (ZAPI), and the Cocoa Stabilization Fund were the main institutions concerned with cocoa. To increase production the Ministry provided planting materials, free fungicides and sprayers on credit, and took responsibility for capsid control. These services improved cocoa production somewhat, out were not sufficiently well organized to make a substantial difference. 2.06 The Cocoa Stabilization Board, an autonomous agency, regulated pricing in the sector by setting annually the producers (farmers) price for cocoa and establishing a Stabilization Fund to regulate that price. By 1971, the Fund had accumulated US$52 million. Producer prices, while reasonable in the context of a low input/low ouput system of husbandry, were insufficient to justify the labor, chemical, and other inputs to improve yield levels. In 1972, the exrort price was CFAF 300/kg, comprising: producer price 100, Stabilization Fund 116, export tax 34, and transport and other costs 50. 2.07 The "Fonds National de Developpement Rural" (FONADER) was organized in 1973 as a credit agency to assist in financing rural development. III. PROJECT FORMULATION 3.01 The project was prepared by a foreign consulting firm in 1972-73, in accordance with Terms of Reference approved by the Bank. The final report was issued after taking into consideration Bank views, which raised some sub- stantial points. Some reservations about the project were expressed within the Bank. One such concern was that the project might not result in increasing production very much since its design was similar to that of some other cocoa projects in West Africa which had mixed results. Another concern was the - 43 - quality of the soils. However, a consensus was reached that the project should proceed. 3.02 By 1972, the Bank Group had mounted cocoa projects in Nigeria, Ghana, Ivory Coast, Sierra Leone (on a small scale), and Togo. However, it was too early to make a judgement of the success of these projects. Cocoa growing practice in West Africa was well known and, in a technical sense, was treated relatively routinely with emphasis placed on the less controllable aspects: organization, pricing and marketing. 3.03 The Bank's confidence that cocoa production in Cameroon could be intensified, above all by black pod control, had its origin in: (a) the success of a highly controlled USAID small-scale 500 ha scheme in which yields of 800-1,200 kg/ha of dry beans were achieved; (b) the indication from a bilateral financed scheme that farmers over a cocoa area of 8,500 ha were able to protect their crop by extensive use of fungicide (although there were warnings that the producer price structure should be sufficient to convince farmers to carry out adequate spraying to obtain first grade cocoa); and (c) the view that improved black pod protection had,largely contributed to the national increase in cocoa from 1966-1971 (although other interpretations might perhaps be put on the results: 1965 - 91,000 tons; 1966 - 81,000 tons; 1967 - 86,000 tons; 1968 - 89,000 tons; 1969 - 104,000 tons; 1970 - 100,000 tons). 3.04 The main cocoa area is on the large plateau in south Cameroon, 600 m above sea level. The conditions are adequate for growing cocoa by West African standards but far from ideal by world standards. Soils are generally deep, highly leached and well drained, but have low fertility and need fertilizer to sustain high yields. Some cocoa is planted on rather shallow lateritic soil. 3.05 The project area included 80,000 ha planted with cocoa, of which 35,000 ha were estimated to yield 300 kg/ha, and 13,000 ha to yield little or no cocoa because trees were either too ybung or neglected. The rural population of the area was 260,000, of which 200,000 comprised 38,000 families on cocoa farms. It was served by a substantial network of secondary and rural roads, aggregating 1,650 km generally in poor condition. 3.06 The establishment of a good extension service was vital. According- ly, a program to train eight instructors who, once trained, would in turn train 230 extension workers, was incorporated in the project. 3.07 To improve cocoa marketing, the prcject provided for the development of marketing cooperatives. Farmers were to oe encouraged and assisted in forming groups of 100-150 members, each of which would elect a managing committee, in areas corresponding to the 180 collection centers. About 30-50 of such groups made up a pre-cooperative, each of which was to have a storage facility. Sector stores were to be financed by the project and collection center sheds built by the farmers. - 44 - 3.08 The pre-cooperative plan was intended to remedy the marketing difficulties experienced in the cocoa sector and to elimininate the markup of the middleman, who graded, collected and paid (the farmer) for the crops, and made them available to exporters. Upon the establishment of pre-cooperatives, the farmer would do his own grading and weighing, and would market his crops through the pre-cooperative. 3.09 As the performance of the ZAPI, which was previously responsible for development in the project area, had been weak, a new area authority; "Socifte de Dveloppement du Cacao" (SODECAO) was established in February 1974 to execute the project, with specific responsibility for: (a) extension services and the supply of inputs and credits; (b) training courses for extension workers and farmers; (c) improvement of rural roads; and (d) formation of farmers groups and pre-cooperatives. SODECAO's management was to consist of a General Directo:, a Deputy General Director in charge of technical affairs, a Marketing and Credit Director, a Senior Inspector, and a Chief Accountant. 3.10 FONADER was responsible for the procurement of farmer inputs and their provision to SODECAO, the "Lutte Phytosanitaire et Anti-Caspide" (LPAC) for the capsid spraying service, the Department of Public Roads (DPR) of the Ministry of Equipment for executing, under the aegis of SDDECAO, rural road improvements, and the research institutes IFCC (now IRCC) and IRAT for conducting research. PROJECT DESCRIPTION 3.11 The project was to be carried out over the period 1974-80, and included: - rehabilitation of 35,000 ha of existing cocoa plantations through intensive crop maintenance and protection; - planting high-yielding hybrid cocoa on 15,000 ha of new land; - strengthening of extension services to help farmers control black pod disease, improve cultivation practices and provide the necessary inputs and credit facilities to them; - rehabilitation of about 950 km of rural roads; - assistance in the formation of farmers pre-cooperatives to increase the role of farmers in marketing; - 45 - - applied research on cocoa and food crop growing; and - preparation of a follow-up project. COST ESTIMATE 3.12 The cost estimate of the project was prepared by the consultants. The estimates, in 1974 prices, included a 5% physical contingency, and price contingencies as given below : Price Contingencies 1974 1975 1976 1977 & thereafter Equipment 14% 11% 7.5% Civil Works 18% 15% 12% Local Costs 10% 9.5% 8% The project was.estimated to cost USS23.8 million, of which US$14.3 foreign (60%) and US$9.5 local (40%), equivalent to CFAF 5.94 billion. The Bank loan was for US$6.5 million. Estimated and actual costs are given in the Annex. FINANCIAL PLAN 3.13 Bilateral agencies were highly involved in the preparation of the project and financed the feasibility study. The two agencies, "Ponds d'Aide et de Cooperation" (FAC) and "Caisse Centrale de Coopiration Economique" (CCCE) assisted in the financing under a parallel arrangement, as follows (US$ million): -46- IED PAC & Government Farmers Total CCCE SODCAO' S nadm stratIve cost, operation maintenance, collection centers, road program 3.72 0.69 4.87 0.04 9.32 Training -- 1.64 1.64 Pest Control -- 0.65 1.33 -- 1.98 Research - 1.66 - -- 1.66 On-Farm Inputs 0.78 - 0.91 0.44 2.13 Unallocated 2.00 1.86 2.97 0.22 7.05 Total 6.50 6.50 10.08 0.70 23-78 Share of Costs() 27 27 42 4 100 3.14 The Bank and bilateral agencies each covered 27% of the costs, the Government 42%, and the farmers 4%. The funds were to be transferred by Government, as grants to SODECAO (US$18.7 million equivalent), to DPR (US$3.0 million), and to FONADER (US$1.4 million). Procurement and Disbursements 3.15 Bank funds were to finance 58% of the cost of SODECAO's administra- tive staff and 100% for vehicles and equipment, road maintenance equipment for DPR, and sprayers and fungicides. Procurement was on the basis of interna- tional competitive bDidding with the exception of initial orders for vehicles and equipment which, to facilitate the start-up of the project, were purchased on the basis of quotations. Retroactive financing of USS150,000 for small items purchased prior to loan signature was included. Procurement was carried out without undue difficulties, except for the delay in placing orders for farm inputs (para 4.11). Goals and Targets 3.16 The broad objectives of the project were to improve the livelihood of the farmers in the project area, increase substantially cocoa production and cocoa exports, and demonstrate that the rehabilitation of the entire cocoa sector could proceed effectively by either replicating the project or adopting its most successful measures. - 47 - 3.17 The project Denefits were expected to be yield increases from 300 kg/ha to 500 kg/ha on the 35,000 ha of existing cocoa with intensified black pod and capsid control and of 450 kg/ha in the fourth year to 800 kg/ha in the tenth year on the 15,000 ha of new plantings, remaining at this level thereafter for 20 years. At full development, incremental project production was estimated at 15,400 tons, with an annual value of about US$12 million and export earnings of USS10 million annually. 3.18 Cocoa net revenues for existing farmers were expected to increase by 70%, but with an almost equal increase in labor (an aspect whose implications seem fairly evident but which appear to have been overlooked in project formulation), and for new farmers they were expected to be about 35% higher than the after-project revenue of existing farmers. The targets are given in the table in para 4.24, together with actual results. Loan Negotiations 3.19 Loan negotiations took place in June 1974. Only one important issue arose, the resolution of which had unfortunate consequences. The Bank had proposed, and understood, that the rehabilitatioi of the rural roads--a major project component--would be carried out by DPR under the broad jurisdiction of SODECAO. At negotiations, the Bank restated this understanding, but the Government responded that this was not the case and that SODBCAO would be directly responsible for the execution of the road work, with the assistance of IPR and its cocoa roads unit, "Service des Pistes CacaoyZres" (SPC), which had been created to carry out such work. 3.20 However, during the first supervision mission (December 1974) SODECAO insisted that it would carry out the roads rehabilitation by setting up its own rural roads division, backed up by the Ministry of Equipment (ME). Eventually, the Bank agreed to the change, although there were differing views: the regional Highways Division supported the DPR objectives whereas Programs agreed with Government, primarily on the basis of expediency. DPR oDjected strongly on the grounds that it had jurisdiction over the road sector, maintained the roads and had the resources, and that the Govei-nment- SODECAO arrangement would be inefficient, a view subsequently substantiated (paras 4.03 and 4.04). Important Covenants 3.21 The following were the important covenants in the Loan and Project Agreements: - SODECAO would carry out the project, assieted by the Ministry of Equipment roads division in the execution of Ihe roads component; - the Borrower would establish, after consultation with the Bank, an annual program of road rehaoilitation and adequately maintain the roads afterwards; - SODECAO and FONADER would be provided with sufficient funds annually; - 7ONADER would enter into a farmer credit administration agreement with SODECAO, the agreement being a condition of effectiveness; - 48 SODECAO would provide extension services and establish an efficient system to distribute farm inputs; - SODECAO would be provided with sufficient seeds to enable it to carry out its almual planting program (which was unrealistic in that it assumed that priority would be given to SODBCAO over the needs of the rest of the cocoa sector); within six months of Loan effectiveness the Government would prepare a program for integration of the ZAPI (activities) in the project area into SODECAO, and the ultimate substitution of SODECAO in place of ZAPI; SDDECAO would employ consultants to assist it to integrate ZAPI activities and to evaluate SODECAO's performance and the project impact; the Government would make arrangements to encourage cocoa farmers to join pre-cooperatives in the cocoa area; SODECAO would prepare within six months of effectiveness a comprehensive training program for extension workera; - SODBCAO would maintain the posts of General Director, Deputy General Director in charge of technical affairs, Senior Inspector and Chief Accountant, and fill them with appointees with qualifications satisfactory to the Bank; SODECAO would employ counterparts to expatriate management personnel, and provide them with training to enable them to assume such posts; - the appointment of the Chief Accountant to SODECAO was a condition of effectiveness; - the signing of the PAC and CCCE agreements were also conditions of effectiveness. 3.22 The loan was approved by the Board on September 9, 1974 and became effective on February 26, 1975. IV. PROJECT IMPLEMENTATION 4.01 The project was not successful and the targeted yields were not achieved. However, the project did serve a very useful purpose in demonstra- ting the intractable nature of some of the problems. It also showed that the design of the project was not suited to the conditions--even if the farmers had been able to give the 12 black pod sprayings provided for in project design, the labor entailed would have made it uneconomic--and that, pending further experience, it was not advisable for Cameroon to start on the nation- wide cocoa expansion program envisaged. While the source of many problems was -49- the Borrower, the poor outcome was due mainly to weaknesses in the project, as reflected by the unrealistically high yield expectations. Initial Stages 4.02 Performance during the first two years of implementation was fair, but problems loomed. The decision to make SODECAO, rather than IPR, responsible for rural road construction having been taken, it was found in April 1975 that additional heavy road equipment would be needed, almost doubling the cost, largely due to price inflation from the 1973 oil crisis, thus delaying the start of road work until late 1976. The cost of sprayers had also increased considerably, and substantially fewer were being purchased by farmers than estimated at appraisal. While new planting was proceeding fairly well, PONADER experienced bureaucratic delays in placing fungicide orders and there were unexpectedly high costs in distributing them. Training was satisfactory, the research program had started, and a considerable number of collection centers had been set up*following the decision to establish a larger number of smaller centers at an extra cost of US$0.5 million. Overall, a cost overrun was becoming apparent early. Rural Roads 4.03 Contracts for road equipment were awarded in November 1975 but SODECAO had neither prepared a road program nor set up a road organization. The feeder road unit of the Bank Resident Mission in Abidjan visited the project in December 1975, to help improve matters. It organized with CCCE and Government a work program and provided advice on the content and organization of a training program for road maintenance and staffing of a road unit. Road work began after a road unit was created in 1976, but ME failed to cooperate with it, a necessity for standardization and coordination. Notwithstanding, two brigades completed 100 km of rural roads in 1976. , 4.04 Road construction continued to be inefficient: vehicles and equipment were only available 50% of the time and actual road building was about 25% of what had been planned. SODECAO, lacking the cooperation of ME, eventually assumed responsibility for maintenance of the project roads. By mid-1978, only 200 km were rehabilitated out of the 950 km of rural roads programmed, due to poor organization, management and work discipline. Two years later, less than 50% of the target had been reached, at almost four times the cost estimate, i.e., US$12,000/km actual versus US$2,300 at appraisal, an excessive amount. 4.05 The roads were built to the same standards as Category A feeder roads in Cameroon, an unnecessarily high standard for this location. The construction was of satisfactory quality, but the design in excess of traffic requirements. The results provided confirmation that future agriculture access roads should be built by DPR to uniform standard in the interests of efficiency, avoiding duplication of road maintenance equipment and spare parts, and a balanced road network and vehicle fleet. Organization 4.06 SODECAO was created for project purposes by Government in February 1974 to develop and rehabilite the cocoa sub-sector; to improve the socio- - 50 - economic situation of the project cocoa region, it was provided with considerable autonomy and a substantial initial capitalization. Three experts were recruited to organize the internal structure of SODBCAO, and prepared a proposal to this effect during April-December 1974. In the early stages of the project, the Center South ZAPI was gradually integrated into SODBCAO, the Director General of SODECAO (who was appointed in 1974) assuming charge of ZAPI to expedite matters. In 1974/75, SODECAO grew to a staff of 81, eventually reaching 132 in 1976/77, of which about 60% was extension agents. 4.07 The establishment, staffing, training and organizing of a good sized state-owned entity, as part of the project itself, were major jobs and under the best of circumstances were unlikely to be very efficient. However, the developing inability to fill SODECAO's management posts with both administra- tively able and technically qualified experts made administration all the more difficult. Much of the time management posts were vacant, the incumbent temporary not qualified and/or unsatisfactory to the Bank, or the position vacated. In particular, the posts of General Director and Deputy General Director for technical affairs were de facto empty for long periods. 4.08 This situation led to a continuing weakness in SODECAO's management as manifested by poor organization, very weak administration, a lack of coordination between the various field activities, negligible discipline, and an openness to waste and corruption. Technical prGblems aside, these organi- zational weaknesses, together with inadequate accounting, poor financial control, and improper financial activities were significant brakes on progress. This was compounded, inter alia, by the confusion invoked by Government decisions regarding cooperative organization and changes in subsidization of fertilizer. This continuing state of affairs, as well as the lack of technical progress, finally led the Bank to request Government to reorganize SODECAO in 1978. When the process of reorganization and reform proved inadequate, disbursements were suspended in 1979. 4.09 In addition to management weaknesses and the roads problems, SODECAO's extension service was of mixed success (para 4.28), and its accounting and financial administration poor (para 4.27). Its main successes were the training program, which was well conceived and carried out, and the formation of pre-cooperatives (para 4.10), Vnich were highly successful. Pre-Cooperatives 4.10 By December 1976, 283 pre-cooperative centers had been established, more than meeting the appraisal target. They functioned well, and credit repayment was good. However, in 1977, Government created a new agency, the "Socit- Coopirative de Diveloppement Rural" (SOCOODER) to, inter-alia, install area cooperative systems to replace the pre-cooperatives system, and administer (farmer) marketing. The Bank was not conpulted, in contravention of the loan agreement. For a time, marketing continued to be satisfactory because the old pre-cooperative staff remained in place. 1/ 1/ There was controversy over whether such staff should continue to be paid from the loan since they were no longer employed by SODECAO and their employment status was in a kind of limbo; however, they continued to be paid for some time. - 51 - Credit 4.11 However, credit which had been a minor problem quickly became a more serious one. As implementation progressed, SODECA0 suffered a funds shortage due to Government delay in providing them and, in turn, FONADER did not receive sufficient funds in time to procure inputs on schedule. Moreover, the relationship between FONADER and SODECAO was poor. When SOCOODER assumed marketing responsibility FONADER could not dep! with SOCOODER cooperaties legally as its credit agreement was with SODECAO, and SOCOODER was unwilling to deal with credit, which was not within its mandate. Consequently, SODECAO had to try to assume the handling of credit. As a result, in 1978/79, there began a long period during which little or no credit was given to farmers to purchase sprayers and other inputs. When the Bank asked Government to reorganize SODECAO, it also insisted that a new credit agreement should be reached to put credit on a sound footing. But it was not until July 1981 that a SODECAO-FONADER credit agreement was in fact signed, toward completion of the project. Cocoa 4.12 After the project had been underway a year, the Government simplified the cocoa grading system, eliminating the differential between grades I and II and reducing the margin between grade and sub-standard cocoa. Consequently, the incentive for farmers to produce export-quality dried beans was reduced. This problem was subsequently compounded in annual producer price reviews by the failure to maintain the terms of trade at producer level between cocoa and food crops. 4.13 By the end of 1976 several anomalies became apparent. There was uncertainty, which would continue through implementation, concerning the number of hectares given black pod control: 75 fungicide packets were considered the average number required per hectare for treatment, and the hectarage treated was determined by dividing the fungicide packets provided by 75. However, a substantial amount of diseased beans did come to market, implying inadequate application of fungicide, or the ordering of fungicide sufficient for one area but its application to a larger one. This led to the speculation that farmers found it more profitable not to spray fungicide but to opt to put more land under (cocoa) planting, thus increasing cocoa revenue, as the additional labor entailed cost less than the spraying required to achieve the same production. 4.14 The area planted with new trees was mudh less than expected due to the limited, or wrong, incentives which were provided; the lack of interest of rural youth in a crop with a long gestation period, extensive labor require- ments, high disease risk, and poor growth of young cocoa trees reflected the lack of support from extension services. The provisi5n of annual payments to farmers until the maturity of the new or replaced cocoa plantations would have induced more active participation by farmers. Altogether, losses in pods during transport from the seed farm, disease losses, inexperienced extension agents, and limited financial incentives were becoming inimical to success at this stage of project implementation. 4.15 In mid-1977, reflecting concern with the project, an in-depth review mission was made, reinforced by expertise from the (then) Central Projects - 52 - Department. The Government had taken two measures to increase production: an increase in the producers price from CFAF 150 to 180 per kg, with no price differential between Grades I and II, and the provision of free fungicides for black pod. It had also moved toward further bureaucratization of cooperatives (para 4.11). Once fungicides were free, there was an upsurge in the amount used and in the number of sprayers. I_/ The impact on production was small, although new plantings increased somewhat. 4.16 Project design had assumed an existing production of 300 kg per ha average, and established an implementation target of 500 kg per ha. A sample inventory in 1976/77 implied that the average yield might be about 330 kg per ha. It was then thought that the pre-project yield must have been in the neighborhood of 225 kg/ha. It was apparent that the major reason for not achieving the targeted yield was the labor intensity of cocoa husbandry, in particular of black pod spraying, which had led farmers to limit sprayings to 4-6 per year rather than the 12 envisaged in project design. One survey indicated that 60% of trees was not sprayed at all and that 40%, in the area nearest the farmer's home, was given about 4 sprayings. 4.17 A new trend observed was in-filling (redensification), for which most seedlings purchased were apparently being used. The mission supported this trend on the grounds that it helped optimize production and revenue. However, during the following supervision mission, in-filling appeared less attrac- tive. It was estimated to have little impact on production as it would account for replacement of only 1% annually of SODECAO area trees. New planting continued to be limited, but was of good quality. 4.18 There was another retrospective review mission in 1978. Cocoa production had improved in 1977/78, reaching a level of 13,000 tons in the SODECAO area due to fungicide being free and a further increase to CFAF 220 per kg in the producer price. The amount of sub-quality cocoa had been substantially reduced, due to the fact that its price had -remained at CFAF 90 per kg, thus increasing the differential. But the SODECAO yield was still under 240 kg per ha and, in any case, production had increased substantially in the non-SODECAO growing areas as well. It was determined that the probable yield of the seven separate areas producing cocoa in the SODBCAO zone ranged from 173 to 327 kg per ha, reflecting that the average age of the trees in the areas ranged from 30-40 years. . 4.19 The conclusion that the disproportionate number of old plantations was a major factor limiting production invoked the scale and importance of the longer term cocoa sector issue, i.e., that to lower the average age of cocoa trees by 10 years, from, say, 34 to 24, would require an annual program of 1,000 ha new planting or 10,000 ha replanting over 10 years, raising the area under cultivation to 500,000 ha. 4.20 The monitoring results suggested that a second principal factor limiting production was labor and, more generally, that the yield was equally dependent on the extent of spraying, the care of the trees, and labor 1/ Because the fungicides were free of charge to farmers there was an enormous waste. Many sachets were stored because the farmers feared the subsidy might be discontinued. - 53 - limitations. Thus, a good farmer tending a small plot could contend with this, whereas with a larger plot or a less efficient farmer it was not possible. 4.21 One outcome of this review was the Bank's request for a reorga- nization of SODECAO, a SODECAO-FONADER credit agreement (para 4.11), and the establishment of a SODECAO seed farm (pare 4.30). The Government agreed in principle but the next important action taken was the establishment in July 1979 of a Management Committee, which had several built-in impediments to the effective supervision of SODECAO affairs. The Committee instituted a study of SODECAO organization, small seed farms were put underway, and the auditing of accounts from 1976/77 was initiated (para 4.27). 4.22 In October 1979, the Bank raised with Government its concerns about some serious financial irregularities in connection with the project, which had become known and pressed the strong need for house-cleaning (para 4.35). It asked that, under the circumstances, the Government should abstain from submitting withdrawal applications, and the Government agreed. This amounted, de facto, to suspension of disbursements. The Bank's intervention also reflected its concern at the minimal pace of reorganization and filling of senior positions, credit, and more broadly, the looming failure of the project. In April 1980, the Bank advised what was still required for resumption of disbursements, including completion of the audit and its acceptance by Government, appointment of a new financial manager, and measures to recover outstanding farmer credit. Finally, in May 1980, Government installed a new management team and some other posts were filled. Disburse- ments were resumed in August 1981, when a credit agreement between SODBCAO and FOIADER was signed and SODECAO's affairs had otherwise improved. 4.23 With the appointment of a new expatriate Director General, SODECAO started to turn the corner. Under the new management SODECAO began to operate more efficiently, inadequate staff was released, new staff recruited, extension services organization improved, input distribution made more efficient, and the rural road construction revitalized. Today SODECAO is considered a strong regional development organization in the Centre-South Province with generally good technical and financial management. It would likely become the project executing agency of any future rural development project in the province. PROJECT RESULTS * 4.24 The physical results of the project are given below. It is noted that project cost was underestimated, project area increased, and Government provided a large amount of money to cover the cost overruns. For these reasons, the results are not directly comparable with appraisal estimates. In this respect, the monitoring of the project was largely ineffective, and many of the physical results are of an estimated nature or open to argument. The footnotes to the table qualify the figures accordingly. - 54 - Estimated and Actual Project Results Appraisal October 1979 Completion Cocoa Production/(tons) a/ Rehabilitated farms (35,000 ha) 17,500 19,340 Unimproved farms (32,000 ha) 9,600 New planting 900 Total 28,000 23,500 b/ Cocoa Production (kg/ha) a/ Production 500 300 350 Incremental yield (1980/81) 200 New Planting (ha) b/ By new farmers 3,000 1,580 By existing farmers 12,000 11,350 Total 15,000 12,930 Credit to Farmers (CEAF '000) Amount 360 c/ 258 Unrecovered 20 118 Pre-Cooperatives Formed 180 2B3 283 Training - Number Instructors 8 13 14 Extension supervisors 19 47 48 Extension agents 211 296 301 New farmers 2,000 1,061 1,353 Rural Road Rehabilitation (km) In SODECAO zone 950 342 National Cocoa Production ('000t) 1974 1975 1976 1977 1978 1979 1980 National ('000 tons) 118 96 82 114 106 124 120 Marketed through SODECAO cooperatives Not Available 14.0 13.4 15.4 15.3 a/ The project had no effective monitoring and evaluation unit and it is not possible to establish clearly the impact of the project either on production or yields. b/ After July 1977, two additional cocoa areas were made the responsibility of SODECAO, increasing the area to 80,000 ha. The increased production is included in the results. c/ Free fungicide for the control of black pod substantially reduced the credit requirements of the farmers. - 55 - Project Cost 4.25 The comparison of actual versus estimated costs is made in france CFA because virtually all costs were incurred in that currency or in French francs. Actual cost was CFAF 10.5 billion, 77% more than the CFAF 5.9 billion amount estimated at appraisal. The reasons for the cost overrun are as follows: - administrative expenses doubled; - the cost of the rural road program was 50% more than estimated, despite only partial execution of the road programs; - vehicles and equipment were more than 40% over estimate; - personnel costs were 80% higher; - research costs were 17% higher because of personnel costs; - Government increased the scale of the project when it added additional zones to SODECAO's area of responsibility in 1977; The massive increase was offset in part by: - savings from not carrying out some of the planned construction; - savings of 26% in anti-capsid controls; - similar savings in the cost of farmer activities, cost of which was in part assumed by other agencies; - an almost 50% saving in the cost of organizing the pre-cooperatives because the process went ahead quickly and effectively; and - a saving from not preparing a successor cocoa project. Lying behind the individual overruns was the oil crisis impact on prices. Inefficient managem3nt was otherwise a main factor in increasing administrative, personnel and road program costs. Project Financing 4.26 The project was to have been financed 54% by the Bank and French bilateral agencies, and 46% by Cameroon. The actual financing was Bank/ France 30% and Cameroon 70%, reflecting that Government paid the additional costs. 1/ In fact, after some delays at the outset of the project, the Government was highly responsive to SODECAO's ever-increasing need for funds. 1/ Thus, the Bank financed only 15% of the cost. Under the circumstances it may be surprising that the Bank exercised considerable leverage, e.g., reorganization of SODECAO. On the other hand, it may also account for the relative indifference to Bank missions concerning management appointm*nts and other matters. - 56 - Indeed this seemingly ready availabilty of money could possibly have been a factor in its inefficient use by SODECAO. SODECAO"s Financial Situation 4.27 SODECAO's financial status as a part commercial, part development- oriented organization was ambiguous. It did not have a mandate to make a profit or become self-financing, but to rehabilitate the cocoa sector efficiently. The evolution of SODECAO's accounts did not progress well and at the Bank 's urging, in 1979, a Cameroonian auditing firm was hired to audit the 1977 and 1978 accounts and prepare proper accounts for these years, and to correct accounting anomalies. The job took 3 man-years and was well done. The auditors made several recommendations concerning the establishment of an internal audit department, delegation of power, need for a procedures manual, and means to improve SODECAO's accounting. It would serve little purpose to incorporate SODECAO accounts herein; however, they are available for inspection in the WAN files. Extension Services 4.28 The extension service was of limited effectiveness. Its weakness in some measure reflected the frustrations from Government's confusing and sometimes obstructive role in regard to supply of in&uts, pricing, and cooperative organization. Because of poor administration, it spent too much time collecting data at the expense of servicing and advising farmers, among other things. However, later on it was redeployed to spend more time demonstrating techniques for new agronomic processes and its impact improved significantly. Capsid Control 4.29 Prior to 1975 anti-capsid control was provided by the Ministry of Agriculture, although not very efficiently. SODECAO subsequently took over responsibility. However, control continued to be ineffective, some 7 litres of pesticide per ha being used instead of the usual 4 liters. The service was free and the main demand on the farmer was to provide labor to carry the heavy spraying equipment, which was especially awkward on small plantings where growth from poor maintenance interfeked with, and prolonged, the job. There was a lack of support from the extension services but this improved in 1980 when the responsibility for anti-capsid control was decentralised to the area sector chiefs (extension service) level. Seed 4.30 The provision of seed was a chronic problem. The NKoemovone seed farm did not meet project needs, partly because many pods rotted before delivery to farmers and, in any case, it could not provide enough seed due to management standards and poor maintenance at the farm. Much insecticide was wasted, areas treated being calculated from the volume of spray applied. At times the farm had insufficient funds to pay the labor. One suggestion was that the project should run the seed farm itself. However, the situation eventually lead, at Bank urging, to the establishment of new hybrid seed gardens at Mengang, where six gardens of 5 ha each were planted between 1979- 1981. - 57 - Research 4.31 Research was done on planting .techniques, using cocoa plants one year of age. Trials were conducted on (a) clearing and replanting; (b) inter- planting with retention of old trees; and (c) in association with food crops. The research was criticized on grounds that there was inattention to cost, over specialization and that some trials were too exotic, of a laboratory quality, and not capable of practical application and, in any case, that the farmer should be involved in agriculture trials. It was concluded from the results that: - planting in association with food crops demonstrated that plantains are best associated with cocoa; - replanting trees with plantain cover is somewhat more productive than providing it with forest shade and fertilizer; and - interplanting with retention of old trees is not advisable; The research activities had little impact on the project's evolution. In any event research results cannot be expected to have an impact on technology during the project period, as it takes too long with perennial crops to establish which of the treatments tested give the best results. Monitoring and Evaluation 4.32 A quasi-monitoring and evaluation unit ("Inspecteurs Itin-rants") existed in SODECAO's organization, but had not received training and did not become effective as a means for control and orientation of the project. As it was, there was failure to make adequate use of the information it did provide. It analyzed, for example, the generation gap between old farmers with no tools and consequent low production, and young farmers with tools, more initiative and resultant higher production. But still, it did not have the capacity to assess the project, except in the most general terms, by 1977. On project completion there was no clear picture of what had been accomplished in terms of hectarage planted, or total production and other targets, as detailed in the footnotes to the estimated results (para 4.24). Some of the project results therefore lack validity and are probably exaggerated. Follow-up Project 4.33 The loan included funds for the preparation of a follow-up cocoa project. However, another cocoa project was not prepared, given the disparate nature of the results. The Bank had come to the conclusion that a single crop project was not suited to the conditions and, instead,.the Centre-Sud Rural Development project is now under preparation. Covenant Experience 4.34 The Borrower only partially met the loan covenants. The more important failings were as follows: (a) the Government did not give responsibility for execution of the rural roads component to DPR; - 58 - (b) the Borrower was dilatory in preparing annual road programs and it was some years before action to maintain them was taken; (c) the Government in effect gutted the credit administration system of SODECAO-FONADER by creating SOCOODER to assume SODECAO marketing responsiilities; (d) the extension services were inadequate; (e) there was a chronic insufficiency of seeds which was only partially remedied by the estaolishment of several small seed farms; (f) SODECAO and/or Government were both dilatory and weak in appointing management personnel and filling the posts with qualified parties; (g) SODECAO was unable to provide sound and efficient management of the project. 4-35 In the OED Report on "Compliance with Loan Covenants" (see page (i)), there is some criticism of the "surprising lack of effective communication between the Bank staff and the Borrower". It also noted that "the Borrower felt Bank staff were more interested in filling vacant positions with candidates of their choice - all expatriates - than in develjping the country's own management capability". However, the Report does remark in this conjunction that Bank staff main concern was with SODECAO's serious management problems. In fact, to a very large extent, Bank actions, with respect to management personnel, had as their underlying reasons "the favoritism, degradation of discipline, rare use of sanctions, abuse and diversions in the SODECAO recruitment process, "misuse and diversion of SODECAO's funds", and other unfavorable practices which were prevalent (from FAC Report referred to on page (i)). It is probable that OED did not have access to this information, at least in printed form, when it interviewed SODECAO management. V. PBDJICT JUSTIFICATION 5.01 The production results (para 4.24) are, as explained, beset with uncertainties and incompleteness. In the circumstances and, in any event, given the large differences between years due .to climatic factors, a 'best' estimate of project impact was obtained by fitting a trend line over the years for the SODECAO and national (non-SODECAO) productions, the difference in slope being the mean of incremental production. The exercise is given below: - 59 - '000 Tons a/ 1977 1978 1979 1980 Mean M.A.C % of Mean A. National 110.6 114.2 117.8 121.4 116.0 3 B. SODECAO 13.6 14.2 14.8 15.4 14.5 4 B. as % of A. 12.3 12.4 12.6 12.7 12.5 a/ Information for 1984-76 is available. 5.02 The results imply that production in the project area increased 1% more annually than it did nationally, less than the possible range of accuracy. Nevertheless, assuming it is accurate, this would mean that the SODECAO area produced about 2 kg/ha more than the 200-300 kg produced nationally--a negligible figure. This amounts to saying that project area production was not significantly increased over national production during the project period. The result appears to be supported by the situation during project implementation. 5.03 Project design was based on increasing production by providing adequate incentives, fungicides and other inputs, improved marketing and credit arrangements, better access roads, and improved extension services. All of these inputs and services were available on a national basis in varying degrees, and the differences lay in the fact that the project would make credit available to farmers to spray for black pod control, provide superior extension services, and generally focus project activities better. However, in 1977, the Government made fungicides free, putting project and non-project farmers on an equal footing from the standpoint of black pod control (anti- capsid control was already free). 5.04 Moreover, the insufficient differentiation between grade and ab- standard cocoa reduced the incentive to produce cocoa of high quality. The subsequent increase in the producer price and in the price differential between graded and ungraded cocoa reversed the trend, but too late to have an impact on project results. The Governient discontinued SODECAO's pre- cooperatives and undermined the farmer credit system in the project area. The SODECAO extension service failed to live up to potential, and partly because of this and the underestimated labor requirements for black pod control, the improvement expected from systematic spraying was not achieved. 5.05 Altogether, these and other negative project developments had the effect of putting the project area on much the same basis as the non-project area from the standpoint of cocoa, as suggested by the production exercise (para 5.01). 5.06 The Economic Rate of Return (ERR) was calculated at 26% at appraisal. Broken down, the ERR for rehabilitation was 32% based on an incremental cocoa yield of 200 kg/ha and for new plantings 20%, in expectation of a yield of 800 kg/ha starting in the fourth year of implementation. The risk was recognized in that the difficulty in achieving full control of black pod was acknowledged and, in the case of rehabilitation, an ERR of 10% instead of 32% was - 60 - considered possible in the event. The appraisal ERR exercise was based on a static yield in the non-project areas which constitute more than 80% of the area under cocoa in the country. 5.07 The 1% annual increase in SODECAO area production above national production over the period of implementation of 6 years amounts to a total of 6%, or about 12 kg per ha. This amount would clearly give a negative ERR. It would have to be multiplied to have any impact, and even then, would be on the borderline of producing a small positive ERR. The project must therefore be considered to have had a minimal impact on cocoa production. In view of the poor quality of information, a detailed recalculation of the ERR would seem to be of little benefit. 5.08 Nevertheless, the intangible benefits from the project discussed in Section VII and VIII, namely: (a) demonstrating the inadvisability of Government proceeding with its long term plan to expand cocoa production on a national basis through crop intensification; (b) rather showing th t this could perhaps only be achieved by a rehabilitation and replanting program to lower the average age of the cocoa stands; and (c) the results which provide valuable information for smallholder cocoa development as a crop in the context of rural development projects, are all important. VI. INSTITUTIONAL DEVELOPMENT 6.01 The prime institutional objective was the development of the cocoa subsector. To this end the project had hoped to improve agricultural institutions, establish a training program for technicians on more advanced cocoa technology and extension services, and improve and institutionalize cocoa development in the sense of packaging the design to enable the project to be replicated over the cocoa growing areas of the country. 6.02 The replication of the project was clearly obviated by the indiffe- rent results, and the mixed role the Government played in the project was generally counter-productive to institutional development. Institutionally: (a) the cocoa cooperative farmer movement, after a good start with the aid of the Government, was emasculated by further Government actions; (b) the concept, at the insistence of Goiernment, of establishing a rural road construction unit in the cocoa development agency was counter- productive and clearly inappropriate; (c) the early institutional arrangement for falmer credit experienced problems, but likely would have improved had not Government weakened the credit process by proliferating agencies to administer it without putting in place mechanics for their interaction; (d) the Government and SODECAO, by their dilatory actions in filling senior posts in SODECAO and keeping qualified people in post, weakened the training accomplishments in extension services, project - 61 - monitoring and overall project management, administration and coordination. 6.03 In the "ocoa pricing structure, the abolition of the difference between Grades I and II was rational in view of farmers' low awareness of differences in grades of export cocoa. The Government's subsequent decision to provide black pod fungicide free could have supported this, but the producer price differential between grade and sub-standard cocoa provided insufficient incentive to farmers and, therefore, both of these initiatives were wasted. The problem was subsequently worsened by Government's insensitivity in mannual price reviews to changing terms of trade which moved steadily in favor of food crops compared with cocoa. VII. BORROWER' S AND BANK' S PERFORMANCE 7.01 The Borrower' s performance was weak in many respects. While the failure of the project was of a technical origin, the counter-productive actions--and at times inaction--of Government concerning project matters, and the great weakness of SODECAO, might have led to failure even had there been no technical deficiencies. Nonetheless, the Governme.t'acted soundly and, in the interest of the project at formulation, acted expeditiously in establishing SODECAO and provided adequate funds through most of the project implementation period. 7.02 The Bank's performance has to be related to the role of the co- financing agencies, CCCE and FAC, and of USAID. The last two had conducted pilot operations on intensifying cocoa production in Cameroon, wich became the technical basis for the project. In the face of at least a few questions and uncertainties concerning the pilot operations, and others concerning the consultants' report, the Bank accepted the results and collaborated with the French agencies in the project. Whether there was anything in the Bank Group' s previous cocoa experience in West Africa, where black pod control is of less importance, which might have raised cautionary signals is not known. Nevertheless, it was unreasonable to aspume that 12 black pod sprayings would increase yields by 200 kg per ha, and even more sprayings would still have Deen uneconomic. Thus, the farmer proved right in limiting the number of sprayings. 7.03 The supervision of the project was frequent and thorough. Two special in-depth review missions, reinforced by additional expertise, were mounted at critical stages of the project, -when there was still time to orient it in a different direction. With the realization of the acuteness of the black pod labor conflict came also the realization that there was no remedy within the context of the project. Nonetheless, the Bank continued its efforts to improve the project within the constraints. One such effort was the suspension of disbursements which proved at least partially responsible for an overhaul of the SODECAO organization and left SODECAO better equipped to carry out its continuing responsibility. SODECAO is now, ten years later, considered to be a strong, well managed organization. - 62 - VIII. IMPACT OF THE PROJECT 8.01 The lessons learned from the project may be expected to have an important impact on agriculture development in Cameroon. The project demonstrated, among other things, that, notwithstanding the long-term substantial development of the cocoa subsector in Cameroon, in which an annual cocoa production of 100,000 tons has been achieved, further substantial growth would have to come from planting new areas, and not primarily from intensified production. Secondly, that prospective cocoa yields of 800 kg per ha, which had been envisaged, could not be achieved under everyday smallholder conditions given technical and smallholder economic constraints. Thirdly, that cocoa will continue to be seen in terms of yields of 200-300 kg per ha unless new areas are well chosen in terms of soil and climate suitability, and finally, that to maintain national production at existing levels will require a continuing long-term inves tment in rehabilitation and replanting to lower the average age of the national cocoa stand. 8.02 With the knowledge gained from the project, it would seem only suitable for the co-donors and the Bank to advise the country and provide guidance on the rehabilitation of the cocoa subsector, perhaps funding part of such a program. While, obviously, much is known of the Cameroonian cocoa sector, the history of the project suggests that the knowledge has not been systematically examined and analyzed. 8.03 A follow-up project may be considered by FAC "that would include all rural development operations in SODECAO's area, stressing those particular features and disparities of the subprojects and components in order to adjust to the various soil, climatic, plant disease and human constraints found in such a relatively large area". I/ The Bank is in the process of identification and preparation of the Centre-Sud Province Rural Development Project, which would encompass the SODECAO area and that of another large project, which include a substantial cocoa component. However, it would not be another cocoa project per se. IX. CONCLUSIONS 9.01 The cocoa project was generally of indifferent success and failed specifically to meet the objective of substantially intensifying cocoa production on a large scale at smallholder level. This was because the pzimary project design assumptions were overoptimistic: i.e., the ravages of black pod disease could be largely overcome, and the yield of cocoa per ha could increase to 500 kg (from 300 kg) by intensifie3 fungicide spraying without increasing the labor required beyond the smallholders capability or means. The yields necessary to justify the 12 sprayings annually against black pod disease would only have been obtained with a more intensive program of extension supervision to ensure optimum treatment. In the event, the number of treatments applied were sub-optimal and the supervision was 1/ Mean annual change. - 63 - inadequate, thereby forfeiting the objective. These technical problems were compounded by the Borrower' s administrative weaknesses. 9.02 The impact of these actions on regional and project institutions was essentially negative. The project could have led to considerable institu- tional development but did not, except to the extent that important lessons were learned from the experience. Nevertheless, SODECAO is today a strong regional organization, with generally good management, and is likely to be the focus of future agriculture operation in the area. 9.03 The Bank and the co-donors, in relying largely on the results of large pilot scale tests of the control of black pod disease as a basis for project design, are in a sense parties to the unsatisfactory project results. However, the Bank exerted great efforts through the supervision and review process to improve the outcome of the project. 9.04 The project was, to an extent, justified in demonstrating what is possible and what is not in the cocoa subsector, Cameroon's most important agricultural export commodity. The long-term wish and objective of Government had been to substantially expand cocoa production, primarily by crop intensi- fication. Originally an initial project encompassing over 100,000 ha was envisaged but, at the instance of the Bank and the co-donors, was substan- tially reduced. The results discredited the high yield possiblity. As a consequence it was established that long-term cocoa development lies in rehabilitation and replanting of existing cocoa in areas with suitable soil to lower the average age of cocoa trees significantly, in increasing the area under cocoa planting and, of course, in developing a more suitable input package to exploit the potential yield obtainable on a practicable basis by smallholders. 9.05 While the co-donors and the Bank Group have rural development projects under consideration as prospective projects in the SODECAO area, it would obviously make sense for them to provide advice, guidance, and possibly funding to Cameroon to rehabilitate and -possibly expand the cocoa subsector as discussed above. - 64 - AN1EX CAMEROON PRO. 3CT COMPLETION REPORT COCOA PROJECT (Loan 1039) ESTIMATED AMD ACTUAL COSTS Appraisal a/ Actual b/ Increase () -- -CFAF million)------ Actual vs Estimate Estimate Administrative Expenses 1,457 2,980 105 Expatriate Salaries 244 486 99 Buildings 149 -- - Vehicles and Equipment 131 383 192 Collection Centers 107 236 120 Operation and Maintenance 450 1,279 184 Rural Roads 772 2,002 159 Training 581 620 7 Pest (capsid) Control 704 863 23 Research 588 965 64 Farm Inputs 758 684 Total 5,941 10,498 a/ from FAC report (page i) bf' Including contingencies. l B R D 10805(PPA CAMERO 0 N COCOA PROJECT Proiect Location EAFIA Internainnal Tund -n - =z 0t4u N 0 -02130OA SglØc1 z i AnJANric Ganom CEAN -- Pr~ Roas National Roads Som - Cammunai RDads -------... Pr S ANMAG4 A - R l, 'IMNATE - - - - Diviio Bondre -Na l Gapirol L EK iEDivision C aias odoula~bOl GDo - --Ink~naional Bmundaries C) /U TID N N G T'J Son -a M6engDze-g Libok Song M?ang 7'Nt A PEA4 or T~ NAP -As wEEN Su.-L¥ E~VAoEc -f NDMEPROWhelA L CAPIT ALS tA VE ØEEN •MGNUGM"TED