Chad - Lake Chad Polders Project

of Document The WorldBank FOROMCIALUSEONLY Report No. 6751 PROJECT.PERFORMANCEAUDIT REPORT CHAD LAKE CHAD POLDERS PROJECT (CREDIT 592-CD) April 24, 1987 Operations Evaluation Department This documenthas a restricted only in the performance and maybe used by recipients distribution duties.Its contentsmaynot otherwisebedisclosedwithoutWorldBankauthorization. of their official  CURRENCYEQUIVALENTS CurrencyUnit CFAF 1974 (at appraisal) US$1 = CFAF 225 1976-79(estimated averageof disbursements) US$1 = CFAF 230 WEIGHTS AND MEASURES 1 kilometer(km) m 0.6215 mile 1 centimeter (cm) 0.033 feet 1 hectare (ha) = 2.47 acres I metric ton = 2207 lbs 1 kilogram(kg) 2.207 lbs ABBREVIATIONS ADF or AFBF - African Development Fund CCCE - CaisseCentralede Coop ration Economique (FrenchDevelopmentAid FundingAgency) ERR - Economic Rate of Return FAC - Fonds d'Aide et de Coop ration (French Bilateral Aid Agency) FAO/CP - Cooperative ProgrambetweenFAO and the Bank GDP - Gross Domestic Product GMT - Grands Moulins du Tchad (MillingCorporation of Chad) GNP - Gross National Product OED - Operations Evaluation Department PAP - Projet d'Am nagement des Polders (PoldersDevelopment Project) PCR - Project Completion Report PPAM - ProjectPerformance Audit MeF.orandum PPAR - Project Performance Audit Report PPF - Project Preparation Facility RMWA - RegionalMission in WesternAfrica SODELAC - Soci t de D veloppement du Lac (Lake DevelopmentCorporation) USAID - United StatesAgency for International Development  FOROFFCIALUS ONLY THE WORLD BANK Washington,D.C 20433 USSA OiMwe of Detminor-Cfirwm Ope.atk4nEvaualiten April 24, 1987 MEMORANDUMTO THE EXECUTIVEDIRE(- RS AND THE PRESIDENT SUBJECT: Project PerformanceAudit Report - Chad Lake Chad Polders Project (Credit592-CD) Attached,for information, is a copy of a report entitled"Project PerformanceAudit Report - Chad Lake Chad Polders Project (Credit 592-CD)" preparedby the OperationsEvaluationDepartment. 4ttachment This dccument has a restricteddistributionand may be used by recipientsonly in the performance of their official duties. Its contents may not otherwisebe disclosed withoutWorldBankauthorization.  FOR 0MCIL USE ONLY PROJECT PERFORMANCE AUDIT REPORT CHAD LAKE CHAD POLDERS PROJECT (CREDIT 592-CD) TABLE OF CONTENTS Page No. Preface i .... BasicData Sheet ........... *.. **...**f..*****.. **. ii Evaluation Summary ...... iv AUDIT MEMORANDUM PROJECT PERFORMANCE 1. Project Background ............... I Introduction ....... . .................... .... I Bank Activities in Chad .2............................. ... 2 Setting .. *.,****a. a **.. 3 *..,**...* * Project Formulation ........................................ Project Objectives............... * 7 1I. Project Implsmentationand Outcome 7 ......................... Start-up and Effectiveness ........................ . 7 Project Implementation ................................... ,. 8 Project Results, Outcome and Sustainability 9 9............ III. Findings and Issues ......................... **......,*. 10 Overview 10 Issues 10 (a) Lending Pressure 10 (b) Suspension and Supervision 11 (c) Final Disbursementand Closing 11 (d) Technical Feasibility ............ 14 Appendix: Government Comments ........... ......................... 17 PROJECT COMPLETION REPORT I. Background ............................. 25 II. Preparation and Appraisal ............ 26 ILI. Imple-nentation 29 IV. Agricultural and Economic Impact 31 V. Performance ................................... 32 VI. Special Issues - Lessons Learned ..... 33 PCR ANNEXES: 1. Key Indicators 2. Disbursements 1!a2: 11447 This document has a restrcted distribution and may be used by recipientsonly .n the performance of their officialduties. Its contents may not otherwise be disclosedwithout World Bank authorization.  PROJECTPERFORMANCE AUDITREPORT CHAD LAXECHADPOLDERS PROJECT (CREDIT 592-CD) PREFACE This is a performance audit of an irrigation project in Chad, for whicl.Credit 592-CD in the amount of US$5.0 million was approved on Novem- ber 11, 1975. The project was cofinancedby the African DevelopmentFund (ADF) for US$5.0 million equivalent, the French bilateral agency, Fonds d'Aide et de Coop ration (FAC) for TJS$2.0 million equivalent,and the United StatesAgency for International Development (USAID)for US$1.0million. The projectwas expectedto be completedby June 30, 1979 and the credit to be closedby December31, 1980. However,the Association suspended creditdisbursements in May 1979due to political and securityproblems which made projectimiplementationand supervision impossible. Subsequent IDA mis- sions concluded that resumptionof project activitieswas not viable anrL recommended reallocationof the balanceof projectfunds to support the min- istries responsiblefor agriculturein their rehabilitation and drought- alleviation activities. The recommendation was approvedby tle IDA Boardof ExecutiveDirectorsin May 1985. The credit Closing Date was extended to June 30, 1985,but finaldisbursements were made on May 30, 1985. The audit report consists of an audit memorandumprepared by the Operations Evaluation Department(OED) and a Project CompletionReport (PCR) dated May 31, 1985 preparedby the Western Africa ProjectsDepartment. Data collected during an IDA review mission in February 1984 wer.:. used in preparingthe PCR. The audit report is based on a review of the Appraisal Report (No. 828a-CD dated October 16, 1975), the President'sReport (No. P-1702-CDdated October 30, 1975), the Credit Agreementdated November 26, 1975 and the PCR. Project files, includingcorrespondence with the Borrower and internalBank memorandaon project issues,have been consulted and Bank staff associatedwith the projecthave been interviewed. The draft report was sent to the Government and cofinanciers for commenton July 18, 1986. Commentsreceived from the Governmentof Chad have been appendedto the audit memorandum. The audit concurs with the principalfindings of this PCR and shares a number of its conclusions. The points discussedin the audit memo- randum have been selectedbecause of their importancefor this and other Bank-supportedprojects.  - ii - PROJECT PERFORMANCEAUDIT REPORT CHAD LAKE CHAD POLDERS PROJECT (CREDIT 592-CD) BASIC DATA SHEET KEY PROJECTDATA Appraisal Actualor Actaal as % of Estimate EstimatedActual Appraisal Estimate Total Project Cost (USe million) 13.0 8.2 631 Credit Amount (US5 million) 5.0 5.0 100t /a Cotinancing:(USSmillion) - African Development Fund (ADF) 5.0 n.a. - - French Aid (FAC) 2.0 n.a. - - USAID 1.0 n.a. - Date PhysicalComponentsCompleted 06/78 irrelevant/b - Proportion of PhysicalComponentsCompleted(M) 100 32 /c EconomicRate of Return (%) .. negative InastitutionalPerformance good inadequate AgronomicPerformance good inadequate Number of Beneficiaries 1,200 none STAFF INPUTS (staff weeks) PY75 FY76 FY77 FY78 FY79 FY80 FY81 FY82 FY83 FY84 FY85 Total Identification/ Preparation 15.0 4.3 3.4 22.7 Apppraisal 4.7 2.5 59.9 27.7 0.4 95.2 NegotiaLions 0.2 9.9 10.1 Supervision 3.1 17.5 10.4 17.9 10.6 2.5 9.3 71.3 Total 19.7 6.8 63.3 27.9 13.4 17.5 10.4 17.9 10.6 2.5 9.3 199.3 DISBURSEMtNTS CUMULATIVE FY76 FY77 FY78 FY79 FY80 ... FY85 /d Appraisal Fstimate (US$ million) 0.4 2.0 4.0 4.4 5.0 - Actual (US$ million) 0 1.4 2.1 3.8 3.8 5.0 Actual as X of Estimate(X) 0 57 54 86 76 100 Final Disbursement May 30, 1985 MISSION DATA Date No. of Mandays Specializations Performance Types uf Mission (mo./Yr.) Persons in Field Represented/e Rating/f Trena/g Problems/h Identification (FAU/CP-RiWA) 1968 n.a. n.a. Preparation(RMWA)/i 02/72 1 2 ae Preparation (FAO/CP-R?WA)LJ 05/72 3 42 ae,i,a Preparation (FAO/CP)/k 11/72 7 3 i Appraisal/I 10/74 6 108 f,a,l,s,t,i Post appraisal/m 03/75 2 It f,i Post appraisal 06/75 3 15 f,i,l SupervisionI/n 04/76 1 4 f I I Supervisioni1 11/76/o 1 6 f I I SupervisionIll/n 07/77 1 7 f 2 2 M Supervision I VL 01/78 1 4 (+2) i (X)/ 2 2 M,F Supervision V 04-05/78 3 21 i,i,f 2 3 M,F,P/r Supervision VI/s 12/78/t 2 10 i,a 3 2 F,M,P7F Supervision V117u 02/807V 3 IZ i,i,e 3 1 F,P Supervision VIII 02/84 3 1 a,e,i 3 3 F,T,M SupervisionIX/w 10/84 2 n.a. a,f -/x - Total  - i.i - BASIC DATASHEET OTHER PROJECTDATA Borrower Republic of Chad Executing Agency Soci t de D veloppement du Lac (SODELAC) Follow-on Project None Original Actual or Current Project Dates Plan Revisions Estimates First \ention in Files 1968 1968 Negottations/j 09/08-12/75 09/08-12/75 Board Approval 11/11/75 11/11/75 Credit Agreement Date 11/2b/75 11/26/75 Credit Effectiveness Date 02/26/76 02/26/76 Closing Date 12/31/80 0/30/85 /a US$3.8 million was disbursed against project-related activities and the balance of credit funds (US$1.2 million) were disbursed against the rehabilitation and drought-alleviation activities of the ministries responsible for agriculture. /b Project buildings and assets were deutroyed or stolen during the civil war. 7c This represents the weighted average of the physical infrastructure that was either fully or partially completed, before being destroyed. /d Project disbursements were suspended in May 1979 and credit funds uere reallocated and fully disbursed on May 30, 1985, when US$1.2 million was disbursed against other activities (see footnote /a). /a ae - agricultural economist, i - irrigation engineer, a = agronomist, f -financial analyst, 1 loan officer, a soil specialist, t - transport specialist. /f 1 - Problem-Free or Minor Problems; 2 - Moderate Problems; 3 - Major Problems. I1 Improving; 2 - Stationary; 3 - Deteriorating. /h F F Financial; M - Managerial; T - Technical; P - Political; 0 - Other. /i Mission to Paris to discuss preparation with consultants. 7 The loan officer was in Chad at the same time. Mission to Paris to review the feasibility study with consultants. 7l A team of two ADF staff also participated in the appraisal. 7m Updating of cost estimates witn consultants in Paris. 7i; Mission also made stopover in Paris for discussionts with consultants and cofinanciers and in Abidjan for discussion with ADF. /o A disbursement officer subsequently visited Chad in December 1976. 7j; Mission also had discussions with consultants in Paris. 7q A financial analyst reviewed auditing arrangements for two days. /r Described in supervision report as "security situation may deteriorate". 7i t Mission jointly with FAC and USAID. 7F A general programs/project mission visited Chad to determine what operations could be combined; however no such proposals were made for this project. /u Part of a broader review mission that visited Chad at that time. 7_ A general agricultural projects review mission by the agricultural projects deputy division chief took place in April/May 983. /w Part of a larger review mission in which the lawyer, loan officer, and disbursetment officer also participated. 7; No supervision report prepared, only BTO. 7jy ADF and FAC sent observers to the negotiatiotns.  - iv - PROJECT PERFORMANCE AUDIT RE?ORT CHAD LAKE CHADPOLDERS PROJECT (CREDIT 592-CD) EVALUATION SUMMARY Introduction Chad has a very limited agricultural resource base, and Govern- ment's strategic options for developing the rural sector are very re- stricted. There is a real limit to the level of increasesin farm incomes that rainfed agriculturaldevelopmentprojects can generate. Government believed,therefore,that it would be necessaryto develop the irrigation potential of Lake Chad and the riversof the Lake Chad basin. The Lake Chad Polders Project (Credit 592-CO)and the Sategui-Deressia IrrigationProject (Credit489-CD), also financedby IDA, were the first stagesof suichprogram. Objectives The project,which was the Bank Group's fourth operationin Chad's agricultural sector, aimed at: (i) promotingcotton and wheat production throughrehabilitating and completing irrigationand drainagenetworksin the Guini polder and constructing the irrigationand drainage networks of the Berimpolder; (ii) establishing a commercialagricultural development unit to settle farmerson newly developed land; (iii)providinga packageof services to the farmerswho would take up cultivation of the polders; (iv) construct- ing staff houses and service centers;and (v) providingtechnical assistance to projectmanagement. Implementation Experience The project'sobjectiveswere not achieved because of political conditions, adverseclimaticconditions,shortcomings in technical design and poor managementperformance. Productiveactivitiestook place only in the 1978/79cropping season. Most of the projectworks were not completed and those that were completed were destroyedduring the civilwar. Results The projecthas no noticeableagronomicor economicimpact,and the project'seconomicrate of return (ERR)is considerednegative. Sustainability Given destruction and deterioration of the physicalinfrastructure, this is not an issue.  -v - and Lessons Findilngs The major findings and lessons that can be derived from this project are the following: (a) the fact that governmentsexplore alternativefinancingarrange- ments might be a reflectionof the high priority they attach to per se to obtain project proposals,rather than an unwillingness Bank financing(PPAMpara. 11); (b) lending and resource transfer considerations were of averriding importancein Bank decisionmaking (PPAM paras.22-24); (c) there is limited impact of the Bank's review process on the scope and contentof projects, once appraised(PPAMpara. 24); (d) the Bank's response to security problems which hamper project implementationis not always consistent and there is extreme reluctanceto cancel IDA credits and close inoperativeprojects (PPAM paras.25-26); (e) resolution of Chad's arrears problems through a package of of IDA to existingcreditsopened the way for resumption amendments to the country(PPAM paras.27-33); -assistance (f) the proposedtechnology but not only in retrospect was inadequate, even at the time of appraisal(PPAMparas. 34-40).  - 1 - AUDIT MEMORANDUM PROJECTPERFORMANCE CHAD LAKE(YHADPOLDERS PROJECT (CREDIT 592-CD) I. PROJECTBACKGROUND Introduction 1. The projectwas the Bank Group's fourth projectin Chad's agricul- tural sector, followinga livestock(Credit 309-CD),drought relief (Credit 445-CD) and irrigation(Sategui-Deressia, Credit 489-CD) project (see also PPAM paras.3-4). The projectobjectives were not achievedbecauseof polit- ical conditions,adverse climatic conditions, shortcomingsin technical design and poor managementperformance, and in the end, the projectfailed because most of the project works were not completedand those that were completedwere destroyedor deteriorated during the civil war. Despitethis poor projectexperience,a number of interesting issues arise,and these are furtherexploredin this auditmemorandum. 2. This project is similar to a number of other irrigationprojects the Bank funded in the early 1970s in Western Africa, such as the Senegal River Polders Project,'/ the Cameroon SEMRY Rice Projects, 2 / the Niger NamarigoungouIrrigation Project, 3 / and the Mauritania Gorgol Project.4/ All of these were high-cost irrigationprojects and to a certain extent reflected the belief that "engineering" could solve developmentproblems. Part of the rationalefor them was that given arid climates and suscep- tibility to drought, irrigationprojects were an essential element of a "droughtproofing"strategyand would provide for additional food security. All these projectshave given rise,however,to equity issues:given resource constraintsshould so much developmenteffort be concentratedon so few beneficiaries in such a limited area? In addition,the replicability of these investments has been severelylimited,while their sustainabilitywill requiresubstantial furtherefforts. 1/ PPAR, Senegal- River Polders Project (Credit 350-SE),OED Report No. 277 dated December26, 1979. 2/ PPAR, Cameroon- SecondSEMRY Rice Project(Loan 1512-CM/Credit 763-CM), OED Report No. 5156 dated June 25, 1984 and see also PPAR, Cameroon- SEMRY Rice Project (Credit 302-CM),OED Report No. 2054 dated May 12, 1978. 3/ PPAR, Niger - IrrigationProject (Credit 851-NIR),OED ReportNo. 6275 dated June 23, 1986. 4/ PPAR, Mauritania - Gorgol EngineeringProject (Credit S-16-MAU)OED - ReportNo. 3134 dated September18, 1980.  -2- Bank Axtivities in Chad 3. Until recently 5 / no new projects had been funded by the Bank since 1978, but over the perfod 1968-78, 13 projects were funded involving IDA credits totalling about US$80 million (for details see table below). Two credits have been fully cancelled and only four credits coutinue to be dis- bursed. Cancellations amounted to US$19.20 million and disbursements to about US$50 million. About US$6 million has been repaid. BankLeing to Chad (as of 08i31/86) Credit Credit Approval Closing OED Project Name No. Amount Date Date Disbursed Cancelled Report No. (UST-miTtion) (US$million) Highway Maintenance 125 4.55 07/68 06/78 4.55 - 1883 Education 126 2.23 /a 08/68 06/76 2e23 - 2616 Education II 251 2.71 7b 04/71 06/79 2.69 0.02 2616 Livestock 309 2.20 02/72 09/78 2.20 - 3373 Drought Relief 445 2.00 11/73 12/78 2.00 - 2807 *Sategui-Deressia 489 15.50 /c 03/74 06/87 15.00 - - Highway II 490 3.50 06/74 06/80 3.42 0.08 5608 Lake Chad Polders 592 5.00 11/75 06/85 5.00/d - /e *Rural Fund 664 12.00 11/76 06/88 8.07 - - *Sahelian Zone 739 1.90 09/77 06/87 1.38 - 5169 /f Livestock II 783 11.50 03/78 03/83 - 11.50 5370 *Education III 811 8.30 05/78 12/86 2.97 - - Highway III 840 7.60 07/78 12/82 - 7.60 5608 Fmrgency Cotton 1716 15.00 06/86 04/89 - / HighwayMaintenance 1728 20.00 07/86 06/89 - - - 78.99 '4.51 19.20 * Disbursing Dollar Credits /a Original credit amount US$1.8 million and supplementary credit of US$0.40 million approved 08/73. /b Original credit amount US$2.21 million and supplementary credit of US$0.50 million approved 08/73. /c Original credit amount US$7.50 million (fully disbursed) and supplementary credi.t of US$8.00 million approved 08/76. /d US$3.8 million was disbursed for the project and the remainder in 198 for rehabilitation and drought alleviation activities. /e Thisreport. 7T Pass-through PCR. 7g As of 08/31/86 not yet effective. 5/ An Emergency Cotton Program was approved in June 1986 and a Highway Haintenance Project in July 1986.  - 3 - 4. The prevailing economic and political circumstancesseriously affected the scope of Bank activitiesin Chad and continued to dominate the Bank/countrydialogue. As a result of. the civil war, all Bank project activitiescame to a standstillin June 1979. A limited resumptionof disbursements was authorizedfor two agriculturalprojects: Sategui-Deressia (Credit 489-CD)and the Rural Projects Fund (Credit 664-CD),and the Board was informed about these partial resumptionsin May 1983.6/ Priority activities that couldbe carriedout under these two projectsas well as the Third EducationProjectwere identified in early 1984 and were iully resumed in June 1984. Substitution Credit Agreementsfor these three projectswere signed on May 28, 1985.7/ Followingsettlement of arrears(see PPAM paras. 26-33), Bank activitieshave resumed and two Bank operations--Emergency Cotton and HighwayMaintenance--have recentlybeen approved. Setting 5. Even without the additionaldisadvantages of civil war and polit- ical upheaval, Chad faces unusually difficult handicaps in its economic development, given its extremelypoor naturalresources,a harsh climate,and the high cost of transportation to and from the outside world, because the main centers of activity in this landlockedcountry are located 1,500 to 2,500 km from sea ports. 6. Chad is a very large country (1.3 million km2) but sparselypopu- lated, rangingfrom around 8/km 2 in the south to almost zero in the nomadic north. Its population of about 3.8 million at the time of appraiselis now estimated at almost 5 million (1985) and its populationgrowth rate is estimatedat 2.4% per annum. Life expectancyis low (43 years) and infant mortality is high (184 per 1000 live births). The present GDP per capita estimate of US$80 is not much differentfrom the per capita income levels estimatedin the early seventies. Chad is one of the poorestcountriesin the world and remainsclassified by the United Nationsas one of the 25 least developedcountries. 7. Agricultureis Chad's largest sector, accounting for an estimated one-halfof GDP and for almost all of the exports,providing a livelihood to most of the country'spopulation. Eighty percentof Lhe populatiou lives in the rural areas, with about half consistingof groups of semi-nomadic and nomadicherders living in the Sahelianzone and the other half livingin the relativelyhigher rainfallSudanianzone in the south where crop farming is the prevalentactivity. Cotton is Chad's most importantcash and export 6/ Chad: PartialResumptionof Disbursements,IDA/SecM83-193 dated May 17, 1983and IDA/SecM83-193 (REV)dated May 26, 1983. 7/ For more details see Status of IBRD/IDAprojects in Execution,SecM85- 1174 dated October31, 1985,and SecM86-443dated April 30, 1986.  - 4 - crop,8 / followedby livestock, a poorlyexploitedasset. Foodcrops(millets and sorghums) are grown as subsistenceand paddy is grown in the floodplains of the Logone (as well as sorghum behind recedingfloods). Only a limited area had been equippedfor controlled irrigati)n. Except for wheat, Chad is self-sufflcient in foodgrain productiononly in "normal"years. 8. Chad's developmentpotential lies in its rural areas. However, progressin the rural economy is slow. As the President's 9 / noted, Report "even assuming politicalstabilityand normal rainfall,it will take many years and great effort to restore the productivecapacity destroyedby years of rebellion and droughts." That predicament is as true today as it wes ten years sgo. 9. At the time of project preparationand appraisal,there was no stated sector strategy and no explicit targets and goals (except for cotton). Agricultural investmentsthat were undertaken reflected the ability of particular agenciesto obtain funding and the different ositionsof the various foreignaid agencies about what was good for Chad.1?/Given Chad's limitedagricultural resourcebase, Government's strategic optionsfor devel- oping the rural sectorwere, and still are, limited. Rainfed projectswere regardedas high3stpriority, but Government believedthat over the long term it would be necessaryto develop Chad's considerable irrigationpotential. This was seen to bring a range of benefits from: (i) diversifyingcrop production;(ii) better utilizationof the country's livestock resources through fodder production; (iii) evolvinga high technology agriculture; and (iv) insuranceagainst the disruptiveeffects of drought. It was realized that the cost of irrigationwas high and that replicability would be limited, as a countrylike Chad did not have the financesto fund such high unit cost investments and as it would be inequitable to expend large sums of money for the benefitof relativelyfew farm families. Nevertheless, the pressure to developthe country's irrigationpotential was increasing, stimulated in Chad by the recentdroughtsand within the Bank by lendingpressure (PPAM paras. 22-24). Also, Government believed that making a start with irrigation 8/ Productionamounted to around 100,000 tons in the 1960s; 174,000 tons in 1975/76; 125,000 tons in 1977/78; 85,000 in 1979/80; and more recently,in 1983/84, it was back again at 156,000 tons, dropped to 90,000 tons in 1984/85 and is expected to amount to 138,000 tons in 1985/86. However,the recentdrop in international priceshas seriously affectedthe financial viabilityof Chad's cottonsector as well as its Government,as it puts 80% of Governmentexportsand 25% of its revenues at risk. 9/ Report ard Recommendationof the Presidentto the ExecutiveDirectors on a ProposedCredit to the Republicof Chad for the Lake Chad Polders Project,ReportNo. P-1702-CD dated October 30, 1975 (para.9). 10/ Chad, Agricultural Sector Memorandum, Confidential Draft Report No. 1438-CDdated February16, 1977.  - 5 - projectswas needed as this was the only way Chad could begin to obtain the expertisein irrigationtechnology,the trained personnel,and the needed institutional development, In retrospect, as the project experienceshows, none of this was vehieved. ProjectFormulation 10. The project was identifiedin 1968. At Governmentrequest, FAC financedand a French consulting firm carriedout: (i) the installationand operationof a 70-ha pilot farm since 1969;and (ii) a feasibility study completedin late 1972. The results obtained from the pilot farm, as well as the preparationand conclusions of the feasibilitystudy.,were reviewedin early 1973 and favor- abl-ycommented upon by the Bank's RegionalMission in Western Africa (RMWA) and judged acceptable for a Bank project. However,because of the relative noveltyof the projectand because the feasibility study raised some serious issues, RMWA recommended a pre-appraisalwhich would look into a number of criticalissues(listedin the PCR para. 6).1I/ 11. However, it took the Bank a long time to decide to appraise the project. In the meantime, given the high priorityGovernment attachedto the project,it also approached other agenciesto financethe project,including FAC, ADF and STA (Soci t des Techniques Appliques), chargedwith implemen- tation of Libyan aid. STA subsequently became involved in a substantial 12 redesign of the project. / The main reasons for the slow processingof this projectwere a viciou circle 13 / of perceiveddoubts about the inten- tion of Governmentto request IDA7financing, but also uncertainties about 11/ The PCR inadvertently gives the impressionthat these issues were the result of the May 1972 review mission. In fact, they were formulated during RMWA's review of the feasibilitystudy (RMWA memo No. 51 dated February6, 1973). 12/ The STA proposalsenvisagedan interlinkedrather than separate water supply system to improve drought-proofing. But costs were also substantiallyhigher and SODELAC thereforedecided in mid-1974 that it did not preferthis solution. 13/ The Bank was slow in respondingto project proposals,and thus, sen- sibly, Government explored alternative fundingarrangements, resulting, however,in furtherdelays in Bank processing, etc. project size, cost 1 4/ and justification, and most importantlyIDA staff constraints (the BaiW was already involvedin the Sategui-Deressia Project, but more importantlythe projects division concerned covered many other Sahelian and Anglophone countriesand was increasingly involvedin a substan- tially expandedagricultural development program in Nigeria). In the mean- time, however,FAC continued to fund the pilot operations. 12. The project was finally appraised in October 1974 and the major issues raised in the subsequentissues/decision process related to SODELAC reorganization and IDA commitment to furtherdevelopment of Lake Chad Polders (PCR para. 7).15/ Substantialreservations regardingthis "high cost and high risk" protect were raised during the internal review process, but management ultimately decided to go ahead with the projectas appraised(PPAM paras. 22-24). Further problems arose when estimates had to be revised because of cost updates preparedby the consultants (post-appraisal mission in March 1975) and further delays occurred because of a military coup in April 1975. Another post-appraisal mission was mounted in June 1975. Nego- tiations subsequently took place in the second week of September1975 (PCR para. 8). 13. The project was approved by the Board on November 11, 1975. The projectwas presentedto the Board as a high-cost projectwhich, althoughnot replicableon a large scale, would have the followingadvantages: (i) as a technically and economicallysound projectit would provideinsurance against drought;(ii) it would be a pole of development in the north;(iii) it would raise technological standardsin the projectarea; and (iv) it would attract sizeable soft loans. In addition to a questionabout the "recruitment" of settlers, there were other questionsrelating to potentialimplementation problems, given the Bank's experiencewith delays in the education and Sategui-Deressia projects. It was pointed out that because of high con- tractor cost, the Sategui-Deressia Project had been reformulated and would use force account construction. Force account construction would also be used under this project. However,the polder project was much simpler than the Sategui-Deressia project which, as mainly a drainage project, required substantial earth moving. The educationprojectshad been delayed because, (Credit489-CD)was facing 14/ Given that and the fact that Sategui-Deressia substantialcost overruns at the time, the Bank, in mid-1974, was seriouslyconsidering credit for the Lake Chad Polders if an engineering Projectwould not be more suitable. 15/ In addition to the Berim and Guini polders, the consultantshad also studied the feasibilityof developingthe adjacent Mandi polder (1600 ha). IDA decided to consider development of this polder only after detailed engineeringwas completedand the settlementof the Berim and Guini poldershad provensuccessful. At that time,detailedengineering and drying of the polder was estimated to cost US$2.5 million and further development, including water distribution and drainage facilities, US$17.5million. access channeland land development,  - 7 - in the past, parliament had only met irregularly(and the resultingratifica- tion delays causedeffectiveness and start-updelays)and lack of counterpart funds hampered further progress. This was rectifiedwhen a supplementary credit was approvedas well as the principleof 100% externalfinancing(as was the case for the proposedproject). 16 / ProjectObjectives 14. The objectivesof the project were to: (i) increasecotton and wheat productionthrough rehabilitating and completingthe irrigationand drainagenetworksof the Guini polder (370 ha net area) and constructing the irrigation and drainagenetworksfor the Berim polder (800 ha net area); (ii) promote evolutionof a high technology agriculture as insurance against the disruptiveeffects of drought; (iii) promote the developmentof an under- privilegedregion in the northwestern part of Chad; and (iv) settleherdsmen who were migratingto the region followingthe droughtsof the early 1970s. In additlon to the two polder components, the project also included:estab- lishing a commercialagriculture section; providinga package of farm ser- vices; construction.)f service centers and houses for staff; agricultural research;and substantialtechnicalassistancefor project implementation research(PCR para. 9). 15. The projectwas to be carried out by the Lake Region Development Corporation(Soci t de D veloppementdu Lac - SODELAC) through a project unit created for that purpose. Over the four-yearproject implementation period (1976-79),the project was expected to cover 1,200 ha and to reach about 1,200 farm families who would be allotted plots in the polders. Projectcosts were estimatedat US$13.0millionequivalent(PCR para. 4) and the financingplan includedIDA (US$5.0million),ADF (US$5.0million), USAID (US$1.0million)and FAC (US$2.0million). AND OUTCOME II. PROJECTIMPLEMENTATION Start-up and Effectiveness 16. The CreditAgreement was signed on Novem,er26, 1975,shortlyafter Board approval. After having fulfilledthe conditionsof effectiveness in time (PCR para. 10), it became effective three months later (02/26/76). However,late arrival of expatriatestaff causeddelays in preparingbidding documentsand subsequent contractaward (PCR para. 12). As a result,works started one year behind schedule. The original schedule preparedby the appraisalmissionwas as follows: issuingbid documents/invitations: June 15, 1975 bid opening: July 30, 1975 bid award: August 15, 1975 start construction: January 1, 1976 16/ Summaries at Meetingsof the Executive of Discussions SD75-54 Directors, 'atedDecember4, 1975. The latter had already been chang, .d to April 1, 1976 in the appraisal 17 / but the first supervision report, mission, in April 1976, noted that bid invitations were now only expectedto be issuedin May 1976. ProjectImplementation 17. As reportedin the PCR, IDA activitiescame to a standstill in May 1979.18/ By that time only part of the smaller Guini polder was under irrigationand few houses had been built (PCR para. 13 and Annex 1). For project activitiesfinanceduntil 1979, estimatedproject cost amounted to US$8.2 million (compared to the US$13.0 million appraisal estimate). However, 63X of project costs only financed 32X of expected physical infrastructure (PCR paras. 14-17). Disbursementshad reachedUS$3.8million (see table below). Disbursements (as of 12184) Category Description Amount (US$) I Cement,equipment, vehiclesand chemicals 1,106,011.92 2 Operating expensesand local starf salaries 1,468,712.45 3 Technicalassistance 795,813.31 4 Corsultant servicesand scholarships 33,983.88 - Revolvingfund 400,000.00 3,804,521.56 17/ SAR: Appraisalof Lake Chad Polders Project, Report No. 828a-CDdated October 16, 1975 (para. 4.17). The same SAR also mentioned that transportof supplies and equipmentto the Bol area was only possible by barge between December and March, when the lake level was high. The distance by road is much longer (305 km) and the road is only passable by four-wheel-drive vehicles. In 1978, official transport prices were CFAF 18,000/tonby road comparedto CFAF 7,000/tonover the lake. 18/ In fact, expatriatesand Chadian staff and their families had been evacuatedby helicopterfrom Bol on February 16, 1979. Expatriates subsequently left Chad includingthe projectdirectorand accountant on March 16, 1979. An IDA mission in May 1979 concludedthat no meaningful project activitieswere possible and disbursements were subsequently suspended on July 18, 1979,as reportedto the Board.  - 9 - 18. Nevertheless, at that time the future of the project was already very bleak. Substantial delayshad occurreddue to enormouslogistical prob- lems. Access to the Bol area was limited, even over the lake. Costs had been reestimated and indicated a cost overrunof US$5 million over the US$13 million appraisalestimate to complete the project. The cost overrun was largelydue to inflation, implementation delays and underestimated unit costs especially for land levelling and transport. Navigabilityof the lake access channel to the Bol area and the very unsatisfactory situation regarding repair of equipmentremained major issues to the end. Also, from the agron- omic point of view, attemptsat double croppinghad failed. In addition, project managementby the SODELAC project unit was unsatisfactory, largely due to the fact that the SODELAC Director-General failed to delegate any authority, and this sloweddown projectactivities and resulted in the resig- nationof two projectmanagers(PCR paras. 19, 24 and 25). 19. Due to politicaland securityproblems,which made project imple- mentationand supervisionimpossible, the Associationsuspended disbursements in May 1979. In February 1984, an IDA mission reviewed FAC-prepared pro- posals to revivethe projectthrougha two-yearrehabilitation programof the Guini polderbut decided against thia for various reasons (PCR para. 4).19/ A subsequentIDA mission to Chad, in September1984, recommendedthat the creditbalance(US$1.2million)be used to supportthe ministries responsible for agriculture, and this was subsequently accepted by th! Board (but see also PPAMparas. 26-33). ProjectResults,Outcome and Sustainability 20. Projectactivities took place only in the 1978/79croppingseason, and on a limitedscale (PCR para. 21 and Annex 1). Subsequently, the equip- ment, and what was constructed,was either destroyedor deteriorated during the civilwar. The projecttherefore had no noticeableagronomicor economic 20/ is not an issue. impact. Under thesecircumstances, sustainability 19/ In its comments, Governmentpoints out that "all the studiesmade before and after the civilwar bear out the economicfeasibility of the project despitethe negativeassertions of World Bank experts." In this context they argue that the report "is too severe and does not take into consideration the favorable views of its financial partners or the government's declared intention to make this project one of its priorities" (seeAppendix). 20/ In its comments, Government "recognizes that certain constraints hindered harmoniousexecution of the project; however we regret the Bank's prematuredeparture." In fact Governmentfeels "that the Bank has abandonedthe project,leavingbehindunfinishedbusiness. However, this Lake Chad Polders Project is viable, and the Government has included it among its prioritiesas a means of achievingfood self sufficiency nationwide" La the regionand subsequently (see Appendixfor more details).  - 10 - III. FINDINGS AND ISSUES Overview 21. The project is an obvious failure, but not only because of its destruction and deteriorationduring the civil war. If that had not happened it would have been a failure because of hydrological circumstances: the water level in the lake was much below that needed for gravity irrigation. And even if that had not happened,the wheat/cottonrotation proved to be technically and agronomicallyinfeasible (PCR para. 21). Thus, the project's proposed technologywas clearly inappropriate (PPAM paras. 34-40). Also, project institutional arrangementswere clearly inadequateand could have seriouslyhampered effective project implementation (PPAM para. 19). In retrospect, a strong case can be made that the project would not hcve been successful even in the absenceof securityand political problems. Issues (ti) Lending Pressure 22. Lendingpressurehas been documentedin some OED reports 21 /. Also audit results have shown that the pressure to lend, and strong mvernment persistence, largelyexplainwhy the Niger Namarigoungou Irrigation Project, despite substantialreservations, ultimately reached the Board.22 / This Lake Chad Polders Project providesanother example of the overridUinimpor- tance of lendingconsiderations in Bank decisionmaking. From the beginning this was an importantaspect. As noted in an internal communication in January 1974: "Top managementhad allocatedimportant IDA resourcesto Chad and withoutprojects such as a viable Lake Chad PoldersProject,I do not see how we can transferto Chad the IDA resources allocatedfor the country." 23. This became even more clear during the reviewprocessfollowing the Bank's appraisal(PPAM paras. 10-13). The issues/decision process already gave rise to substantial oppositionto further processingof this project, given its high risk and costs, 2 3 / the seriousfinancialand managerialprob- lems of SODELAC,and the ability of nomadic herdsmento take up the rela- tively sophisticated farming methods proposed under the project. It took considerable time to argue the variouspoints of view, but ultimately manage- ment at the highest levels decided to go ahead with the project for the 21/ Annual Review of Project PerrormanceResults 1985, OED Report No. 6417 dated September30, 1986,paras. 2.49-2.53. 22/ PPAR, Niger IrrigationProject (Credit 851-NIR),OED Report No. 6275 dated June 23, 1986. 23/ At that time estimated (in 1974 dollars) at US$5,500/hafor civil works but US$12,500/ha other projectcosts. includlng  - 11 - following reasons: (i) the project seemed technicallyand economically sound; (ii) it fitted into the Bank's lending strategyfor the Sahel; (iii) there were no alternatives for substantialinvestment in Chad's agricultural sector; (iv) Governmentand other donors attachedhigh priority to it; and (v) a relativelymodest IDA credit would attract sizeable soft funds. 24/ Even at the Board (both in the presentation and documents)it was acknowl- edged that this was a high-costproject unlikelyto be replicable, but that for variousreasons(see also PPAM para. 13) the projectshould be pursued. 24. As in the case of Niger, it appears in retrospectthat it would have been better if the Bank had not pursued the project. This experience also confirmsthe relatively limited impact of the Bank's review process on the scope and contentof projects,, 25 / once appraised. and Supervision (b) Suspension 25. In 1979,IDA suspendeddisbursements for all its activitiesin Chad as politicaland security probleVas made supervision, and to a large extent implementation, impossible. Sudh action is clearly in line with existing 2 b/ However, the Bank does not always react similarlyin legal agreements. such situations. It took a much more lenient positionwhen, for example, securityproblemsseriouslyhamp red the implementation of the Lower Adiabo Development Project in Ethiopia.* C/ Both cases also reveal extreme reluc- tance on the part of the Bank, e en in extreme circumstances,to cancel IDA creditsand close inoperative pro ects. and tlosing (c) Final Disbursement 26. It is acknowledgedthat!theclosingof inoperative projectsand the cancellationof remainingcredit balancesis never an easy decisionand that a suitablebalanceneeds to be maJntainedbetweenoptimism and pessimism. It 24/ This was correct given the 'cofinancingattractedto fund the project. However, these funde were not fungibleas they were linked to specific project procurement. In fact one could argue that they were therefore "lost" to fund alternativedevelopments. It is true, of course, that Chad would gain some foreign exchange as a result of local cost financing. 25/ Limited impact of criticism during project review was raised as an issue in the PPAR on the Nigeria Lafia and Ayangba Agricultural DevelopmentProjects(Loans 1454-UNI OED ReportNo. 5972 and 1455-UNI), dated December16, 1985 (par4s.68-76). 26/ General Conditions Applicable to Loan and Guarantee Agreements (Section6.02). 27/ PPAR, Ethiophia- Lower Adiabo Development Project (Credit 516-ET),OED ReportNo. 3971 dated June 17, 1982 (paras.23-36).  - 12 - is clear that in 1979and 1980no one could have predicted how long the civil disturbances would last. However, it is also clear that when time continues to pass, projects should be closed, especially in the case of this project and the SahelianZone Project. Regionalstaff argue that the real question by 1984 was not any more whether IDA should close inoperative projects, but how resourcescould be pulled togetherto help Chad recover from civil war and drought. The audit sympathizes with that position, which was correctat that time, but could only be implementedbecause of indecision earlier. The audit remains convincedthat with the benefit of hindsightit is clear that some creditsshouldhave been cancelled earlier. 27. The PCR indicatesthat followingan indepthreview IDA decided not to resumeprojectactivities but recommendedthe use of the balanceof credit funds to support the ministriesinvolvedin agricultural rehabilitation and droughtalleviation activities. FollowingBoard approval,such disbursements were made in May 1985 (PCR paras. 4 and 18). While correct, the story is somewhat more complicatedthan narrated in the PCR given that the Bank pursued its dialogue with the country and resolved Chad's increasingly alarmingarrearsvis-a-visIDA under pArticularly difficultcircumstances. 28. Arrears dated from June 1979 and had become an increasinglyirri- tating issue, not so much because of the amounts of money involved (these were small in relationto the financial magnitudesin the Bank's accounts), but because of the principlesinvolved. In accordance with its accounting and auditingprinciples, overdues (initiallyby more than 60 days, later by three months) ef principalinstallments and servicechargesare footnetedin the IDA financialstatements. From 1981 to 1984 these were largely due to Chad arrearsand they were so recordedin footnotesin the respective World Bank annual reports. However, all four IDA financial statements also mentioned: "It is not anticipated that these delays in payments,including those due from Chad, will result in any loss to the Association." At the Bank side therewas irritation at the fact that thesearrearswould not allow any liftingof suspensions and hamperedfutureBank activities. On the other hand, Governmentfrom time to time round the Bank penny pinching and only interestedin getting paid "while children died from malnutritionin the field and peoplewere killed in the streetsbecauseof the insurgencies." 29. It should be noted that while the amount of arrears was small in relation to Bank financialmagnitudes,it was substantialin the Chadian context where the Governmentbudget at the time amountedto only about US$20 million equivalent. In fact, with increasingarrears a vicious circle was soon in operation. For Chad, in order to pay these arrearsit needed foreign exchange,but for the Bank to providesuch foreign exchangerequiredpayment of these arrears. Under these circumstances it is understandablethat the Bank and the Government lookedfor solutions to this dilemma. 30. Disbursementshad been suspendedon July 18, 1979. And as projects could no longer be implementedin many cases, therewas an increasedfeeling that credits for such projectsshouid be cancelled on project grounds. On the other hand, Programsstaff in general felt that it was neitheropportune  - 13 - nor urgent to cancel. Throughout,missions continued to visit Chad to explore the possibilities for: (i) payment of arrears; (ii) restarting existingprojects;and (iii) designingnew projects. By end 1982, arrears had reachedabout US$1.4million but limitedresumption of disbursements was 28 / followinga first repayment authorized of arrears in early 1983. 31. In early 1984, a mission was mounted to identify priority activities that could be supported under ongoingagricultural projects. That mission was seen as crucial, and at the highest managementlevels it was hoped "that the mission would bring back specific proposals for the settlementof arrears." The returning mission reported in March 1984, that the situationin Chad warrantedfull resumptionof IDA activities. As a result, the bank pursued proposals under which funds under two inactive agricultural projectswould financeexpenditures incurredby the Ministry of Agriculture in 1984 (thus retroactivefinancing)and so provide foreign exchangefor Government to clear outstanding paymentsto IDA. Government was requestedto preparethe necessary withdrawal requestsand agreed in a letter dated March 26, 1985 that IDA could apply the proceedsof these withdrawals directlyfor the paymentof all outstanding balances. 32. A package of measures was subsequently 29 / submittedto the Board basis that consisted on a no-objection of: (i) substitution agreements for Credits 489-CD (Sategui-Deressia), 664-CD (Rural Projects Fund) and 811-CD (EducationIII) to allow for the financingof restructuredactivities; (ii) amendments to Credits 592-CD (Lake Chad Polders) and 739-CD (Sahelian supportfor the ministries Zone) to allow for operational responsiblefor agriculture; and (iii) harmonization of repayment dates and currency of repayments, including Credits 445-CD (Drought Relief) and 490-CD `Highway 28/ Chad - PartialResumptionof Disbursements dated May 17, IDA SecM83-193 1983 and a revisiondated May 26, 1983. 29/ Chad: Resumptionof Operationsand Revised Amendmentsto Development Credit Agreements(CteditsNos. 445-CD,489-CD,490-CD, 592-CD,664-CD, 739-CD and 811-CD), IDA/R85-177dated May 3, 1985 and revisedMay 17, 1985. 30/ Although these amendmentswere made, the Bank's operationalfiles on these two projectsdo not containany documents relatedto or references these amendments. However, the Legal Department stated that the relevantlegal files did containthese documents.  - 14 - 33. As no objections or questions were raised, the proposalswere approved and the amendmentswere signed May 28, 1985. Settlementwas immediately and directlymade in IDA accounts. Governmentand IDA agreedto implement these transactions directly in IDA's books rather than by transferring funds to the Chad financialauthorities and back. Having thus settled the arrears,the suspension of disbursements was formally lifted.31 / This packoge of amendmentsto existing credits resolved Chad's arrearage problems vis-a-visIDA and enabled renewed assistanceto the country. In retrospectthe 'uestion can be raised as to whether this packagemade sense and was thus an appropriateuse of funds. Clearlythere is no doubt that the packagewas the essential elementin restoring the relationshipwith Chad and the only method to enable the Bank to consider new developmentlending. However, on the other hand it would not have been possible in the audit's view to justify the retroactivefinancingof recurrentexpenditure for the Ministry of Agricultureas a separateproject proposal. Operational staff believe that sufficientnotificationwas provided to the Board. In the audit's view the Board should have been informedmore explicitly about the use of offsetting arrangementsfor the settlementof arrears. (d) TechnicalFeasibility 34. The technical package proposed under this project consisted of wheat/cotton rotation,a considerably more complex farmingmethodologythan that which the proposedsettlers were practicingor introducedin other irri- gation projects in Western Africa (rice in Cameroon SEMRY and Niger Nameri- goungou and rice/vegetables in Senegal Polders). Insufficientagronomic experiencewas provided by the Bank during the preparationand appraisal process.32 / It is clear that the croppingpattern of wheat/cottonwas both technicaliyand agronomically difficultto implement,and that was demon- stratedeven during the limitedtime the projectwas operational (PPAMpara. 18). While it is undoubtedly true that experimentshad confirmedthe feasi- bility of the proposedcroppingpatternand vields,once again the experience provesthat the researchstationis not the same as the farmer's field. 35. To what extent, as the PCR wonders (PCR para. 30), the desire to imaximizeeconomic returns on the project did take precedenceover sound agronomic choice of cropping pattern is difficult if not impossibleto prove. It is a fact, however,that the rate of return during projectprepa- rationwas boostedby: the high yields;the substantial transportelementin the wheat importsubstitution price;and the premium for qualitycotton. dated June 4, 31/ Chad, Restorationof WithdrawalRights, IDA/SecM85-212 1985. 32/ In addition,the effectiveness on the appraisal of the agriculturalist by his limitedknowledge missionwas constrained of the Frenchlanguage.  - 15 - 36. In addition to agronomicproblems,hydrologicalconditionsalso adversely affected the project. At appraisal,the principleadopted for irrigation was gravity (in at least nine years out of ten). However,when the February1984 supervision mission reviewedthe project,it concludedthat since 1972, hydroclimatic conditionshad deteriorated drastically,to the extent that pumpingwas requiredin 1979 and would have been requiredduring part of the followingseasons if the project had continued, In February 1984, the water level of the lake was 3 m below the level at which gravity irrigation was possible,and under these conditions, even pumpingwould have been impossible. This, in retrospect, raises the questionas to whetherthis hydrological behaviorcould have been predicted. 37. The mysteriousdesert lake,Lake Chad has alwaysbeen the object of attentionby explorersand scientists, and since the early sixtieshad been the objectof intensive study by the French researchorganization ORSTOM.33 / Lake Chad is all that remainsof a large inland sea which in ancient times extended over an area of some 350,000km2 and drained a catchmentarea of about 2.5 millionkm2. Its main tributaries came from the northernmountains of the Air, Tibesti and Ennedi. Between 5,400 and 4,000 B.C. there was a drastic change in climatethroughoutthe region. The rivers from the north dried up and the residual lake was formed on an elevated portion of the western ridge of the old basin at the point of dischargeof Chari River. There has been no drastic change since that time, only oscillations in the 34 water levels. / These oscillationsfit into cyclical patterns of long duration. The present period of relativedryness, which started some 200 years ago, is similarto that which prevailedbetween 1,000and 400 B.C. 38. In addition, however,the SAR also noted that "at present the lake is below the level it reaches in a one year in a hundred frequency. It is estimated that it will take about two years of normal rainfall over the lake's catchmentarea to raise it to the one year in ten probability, and four to five yeats to raise it to its mean level" (SAR Annex 6, para 1.3). In addition,appraisalgraphs (SAR Annex 6, Table 1) also clearly show a general declining trend in water levels and dramatic drops since 1965. Earlier, In 1972, it was already reported that not enough water entered the canals of the pilot projectand that additionalpumpinghad been needed. How under these circumstances gravity irrigationcould have been proposed is unclear. The audit'sview is that too much attention was focusedon average occurrence(1 in 100 year occurrence)rather than probable occurrenceand trends. However,to what exteat these trendswill remain permanentand how 33/ Lake Chad, Ecology and Productivityof a Shallow TropicalEcosystem, edited by J.P. Carmouze, J.R. Durand and C. L v que, Dr. W. Junk Publishers,The Hague 1983. 34/ Nevertheless,the changesin water levels are importantgiven the lake bank's slight slopes, which makes the lake area and volume very sensitiveto changesin water levels.  - 16 - aridityand droughts they relate to a generaltrend,if at all, of increasing in the Sahel, is the subjectof much 35 debate. / 39. Again this raises the questionof whether economic considerations caused over-optimismin hydrological assumptions. This is conjecturaland no proof exists. However, as the February1984supervision mission pointedout, the appraisalrate of return was very sensitiveto variationsin costs and benefits,and if pumpinghad been consideredrather than gravity irrigation the resultingrate of returnwould probablyhbve been only marginally accept- able. 40. The appraisal clearly also underestimated what to do about the logistical access to the projectarea (see also PPAM para. 16, footnote5 and paras. 18 and 21). It was clear that lake transport was far more economic than road transport to the Bol area (wherein additionto this projeLtother polder developmentswere also envisaged in the near future). Before appraisal,two sand barriers in Lake Chad had been exposed, effectively blocking transport to Bol for large parts of the year. However, the appraisal only noted that a channel excavated across the barriers might improvetrafficand irrigation accessto Bol. Throughout projectimplementa- tion this barrierproved a formidable obstacle. Some of the prolificgrowth of aquaticplants (due to low lake levels)was clearedfrom time to time,but the dredgingof a channel(estimates varied betweena few hundredthousandto more than half a million dollars)was often contemplated but never executed. Given its essentialimportance, improvedwater and transportaccess to Bol shouldhave been an integral part of the project.36 / 35/ Desertification in the Sahelianand Sudanian Zones of Western Africa, World Bank 1985. 36/ Government (seeAppendix) has stated that "the bottleneckrepresentedby transportation will be solved thanksto the channelbeing constructed as part of the Mamdi project,and above all to the non-floodable road from N'Djamena to Bol."  Appendix Page 1 - 17 - E-291/87 October 9, 1986 Fr(Chad) OED TS:bas Translation of incoming telex dated September 29, 1986 from Baniara Yoyana, Minister of Agriculture and Rural Development, Ndjamena, Chad, to Otto Maiss, Director, Operations Evaluation Department, IBRD. No. 072/MADR/DG/86. Further to Evaluation Report on Lake Chad Polders Project (Credit 592-CD), I have the honor to present the following comments. At the time work on the project stopped following the halting of IDA disbursements in May 1979, all the project components had not been completed. Altogether a budget line of US$4.8 million remained unused, which 1/ included US$1.2 million in IDA funds. The volume of works completed was l9S for surface developments and 38% as regards infrastructure. Of the planned 1,200 farmers, 220 were allocated parcels. The project activities only lasted one agricultural season and on a very small scale; however, encouraging results were achieved with cotton and potatoes, for which record yields were obtained. The facilities suffered little damage during the civil war. All the studies made before and after the war bear out the economic feasibility of the project despite the negative assertions of the World Bank experts. In this context the evaluation is too severe since it does not take into account the favorable opinions of the financial partners or of Chad. We recognize that certain constraints hindered harmonious execution of the project; however, we regret the Bank's premature departure at a time when excellent prospects were opening up for this region. The Chad Government is exploring other resources for accomplishing the objectives set. We will send you detailed comments by regular mail. Regards, Baniara Yoyana l/ Auditor's note Please remember that these US$1.2 million were subsequently disbursed for other activities (see basic data sheet and PPAM paras. 26-37). )INV7T SI aOvd SIHI  Appendix Page 2 - 19 - E-386/87 November3, 1986 French (Chad) OEDDl JCB:bas Republicof Chad Ministere et du Developpement de l'Agriculture Rural Secr6tariatd'Etat DirectionGenerale Ref: 1060/MADR/DR/86 October 8, 1986 N'Djamena, From: and Rural Development The Ministerof Agriculture To: Mr. Otto Maiss, Director,OED, World Bank,WashingtonD.C. Dear Sir: Audit Report on the Thank you for sending the ProjectPerfcrmance Lake Chad PoldersProject (Credit592-CD). are our commentson the PPAR: Following 1. were: The projectobjectives (aa) to rehabilitate and drainagenetworksin and completethe irrigation the Guini polder (370 ha net area); (bb) to constructan irrigation and drainagenetworkfor the BERIM polder (800ha net area); (cc) to install1,200 farmersto cultivatethe newly developedlands,and nomadicherders fleeingthe country's to sedentarize arid zones; (dd) to providetechnicalservicesto the farmersto help them with (cotton/wheat double-cropping rotation); researchat the Matafo experimental (ee) to conductadaptiveagricultural farm; (ff) to furnishconsultantassistance; to build three seryicecentersand 16 houses for staff; (hh)[sic] (ii) to contributeto the country'seconomictakeoffthroughpilot investmentsin this frequentlyoverlookedarea of the country's Northwest.  Appendix Page 3 - 20 - 2. Execution It must be admittedthat as of the date of cessation of project activities, following the suspension of disbursementsby IDA in May 1979, none of the projectcomponents had been completed: (aa) therestill remaineda $4.8-million line of credit,of which $1.2 millionwas accountedfor by undisbursed IDA funds; (bb) only a small part of the Guini polder,220 ha net area of the planned1,200ha, had been developed; (cc) five out of the 16 houseshad been built; (dd) from the agronomicpoint of view, the cottonyield for the 1978/79 cropping seasonalone was 35% higher than Lhe targetfigure. A yield of 25 quintalswas recordedfor potatoes. The wheat-growing campaigndid not materialize becauseof the disturbances.Only 220 settlers receivedplots,againsta planned 1,200. The experimental research activitieshad only just gotten under way. The low water levelof Lake Chad, due to the persistent drought in the Sahel, meant that the lake's irrigation potential was unusable; (ee) the total cost of activitiesfinanced was $8.2 million,against $13 millionestimated at the time of projectappraisal; (ff)which would represent38% of the volumeof infrastructure to be executed,takingaccountof the high cost of transportation. 3. Contextof projectexecution (aa) the feasibility studiesapprovedby the Bank attestedto the project's economicviability; (bb) after the disturbancesin the projectarea, the FAC studies carried out in 1984 and resumedby the ADB in September1985 confirmedthe project's viability,despitesome damage; (cc) there is a very large materialpotentialthat can easily be recovered at littlecost,were the project to be reactivated; (ee) this ADB Report is favorableto continuation of the project,and recommends that the availablebalanceof credit fundsbe used to develop350 ha net area; (ff)becauseof errors of design, the projectexperts installedthe pumpingstationbeyond the free waters of the lake, causing the irrigationprogram to fail for lack of water;  Appendix Page 4 - 21 - (gg) the projectmanagement institutionalframework was acceptedby all parties,being one of the prerequisitesfor projectexecution; (hh) theremay be certainindividualproblems,which shouldnot have the slightesteffecton projectexecution; (ii) the Bank's closingdate is premature,since no tangibleresulthas been obtained,whereasall componentswere incomplete and were sharedamong the variousdonors,who experienced some delays in theirworks planning; (kk) a largenumber of orderswere not filledby the merchants; droughtin the projectarea was a negative (11) lastly,the persistent factor. 4. Sustainability (aa) the projectactivitieslastedfor only one croppingseasonand were very limitedin scope; sufferedlittle from the civil war; (bb) however,the equipment (cc) significantyieldswere obtainedwith cottonand potatoesdespite regarding the assumptions a cotton/wheatrotation; was relatively (dd) the cost of transportation high comparedto initial estimates; (ee) the inadequateknowledge patternsin the lake and of hydrological its affluentswas a negativefactor. 5. Conclusion (aa) in this setting,the Bank's PPAR is too severeand does not take into considerationthe favorable views of its financialpartnersor of the government's declaredintentionto make this projectone of its priorities; (bb) we feel that the Bank has abandonedthe project,leavingbehind unfinished business; (cc) the bottleneck representedby transportation will be solved thanks to the channelbeing constructed as part of the Mamdi project,and above all to the non-floodableroad from N'Djamenato Bol. (dd)we regret the Bank's premature departurewhen this regionfaces such excellentprospects; (ee)we acknowledge when it considersthat a the Bank'ssovereignty projectis not "viable";  Appendix Page 5 (ff) however, this Lake Chad Polders Project is viable, and the government has included it among its priorities as a .neansof achieving food self-sufficiency in the region and subsequently nationwide. Isl Yoyana Minister of Agriculture and Rural Development  - 23 - PROJECTCOMPLETION REPORT CHAD:LAKECHAD POLDERS PROJECT (Credit 592-CD) May 31, 1985 WesternAfricaRegionalDepartment Agriculture D  z "'NCrIm It Cr' H~~~~OM  - 25 - CHAD LAKECHAD POLDERSPROJECT I. BACKGROUND 1. The Government of Chad had requested`DA assistance to develop 2,800ha of irrigated agricultural development tor the promotion of cotton and wheat production in the Bol area, usingwater of Lake Chad, and to financean initialstage of 1,200ha. This requestwas based on a feasibility study financed by Fonds d'Aide et de Cooperation (FAC),and was reviewed by an IDA appraisal missionwhich visitedChad in October 1974. On November26, 1975 the Association approved a Credit of US$5.0million for the project, which was the fourthBank Group operation in Chad's agricultural sector. 2. Chad has a very limitedagricultural resourcebase and Government's strategyoptionsfor developing the rural sectorare very restricted. There is a real limit to the levelof increases in farm incomesthat rainfedagricultural development projectscan generate. While rainfedagricultural projectsshouldcontinueto be allotted high priority, Government believedthat in time it would be necessary to developthe irrigation potential of Lake Chad and the riversof the Lake Chad basin. The rationale was that, given Chad's arid climateand the susceptibility of most of the countryto drought,an irrigation program could solve the critical problemsfacedby the economyas a whole mnd the rural sectorin particular. Government believed,therefore, that therewas a case for making a start in irrigation in spite of its cost becauseit offeredChad an opportunity to begin to obtain the expertise in irrigation technology, the trainedChadianpersonnel, and the evolution of institutions needed to exploitits irrigation potential effectively and in the long run to improve the economic base of the country. The Lake Chad PoldersProject and the Sategui-Deressia Irrigation Projectalso financed by IDA were the first stagesof such program. 3. The objectives of the projectwere to: (i) encourage expansionof cottonand wheat production throughrehabilitating and completingthe irrigationand drainage networksof GuiniPolders (370 ha net area) and constructing the irrigation and drainage networksfor Berim Polders (800ha net a,ea); (ii)promoteevolution of a high technology agricultureas insuranceagainstthe disruptive effectsof drought;(iii)promote the development of an underprivileged region in the northwesternpart of Chad; and (iv)settleherdsmen who would migrate to the region followingthe droughtsof the early 1970s. Over the four-year projectimplementation period (1976-1979), the projectwas expectedto cover 1,200ha and to reach about 1,200farm families who would be allottedplots in the polders. 4. Projectcostswere estimated at US$13 million,net of dutiesand taxes. The foreignexchangecomponent was estimated at US$10.4million or 80 percentof total costs. Principalcomponents were: irrigation works  - 26 - construction and supervision, US$5.5million;farm development; US$0.9 million;agricultural research, US$0.9million;ProjectUnit overhead, US$1.2million;other,US$0.1million;and contingencies, US$4.4million. The IDA creditof US$5 millionfinanced38 percentof total projectcosts and coveredsome 48 percentof the foreigncosts. The remainder of project costswere met by the AfricanDevelopment Fund (AFDF),US$5 million or 38 percentof totalproject costs;USAID,US$1 million,8 percent;and FAC, US$2 million, 16 per-tent.IDA financing would be passed on by Government to SODELACas a loan to be repaidover 50 years startingin 1979 and bearingan interest rate of 0.25%. The Projectwas executedby Socigtgde Development du Lac (SODELAC) througha ProjectUnit createdfor that purpose. Projectimplementation startedon January 22, 1976 and by Decem- ber 1978 only 5 houses out of 16 had been built and 220 ha in the Guini Poldershad been developed. Due to politicaland securityproblems, which made projectimplementation and supervision impossible, the Association suspended disbursement in May 1979. In February1984, an IDA mission visitedChad and reviewedthe prospects of resumingprojectactivities on the basis of a PAC financedstudy "MissionBilan et Programme de Relance SODELAC" which proposeda two-year programaimed at rehabilitating the completed part of the Guini Polders (250ha), including provisions for the necessarytechnical assistanceand SODELACoperating costs. The study estimated the cost of the rehabilitation programat CFAF 760 millionor aboutUS$1.8million. Based on its review and analysisof the program,the IDA missionconcluded that the programwas not viable on the groundsof low economic rate of return,technical uncertainties, doubtsabout the adequacy of farmers' benefits,and uncertainty concerning the mechanisms and level of cost recovery. It therefore recommended to IDA not to resume project activities, and a subsequent IDA mission to Chad in September1984 recommended the use of the balanceof credit funds to supportthe Ministries responsible for Agriculture in their rehabilitation and drought-alleviation activities. While the Government welcomed such a supportat this time, it would like to pursue irrigation investments in the projectarea at some futuredate. AND APPRAISAL II. PREPARATION Chronology 5. In 1968 an FAO/RMWAidentification missionvisitedChad and recommended a 2,800ha poldersirrigation projectin the Bol area of Lake Chad. After reviewing the identification report,the Government of Chad requested and obtainedFAC financing for the installationand operation of a 70 ha pilot farm in the alreadydried up Guini polder and for the preparation of a feasibility study for the identifiedproject. FAC employeda Frenchconsulting firm to performboth tasks. The operation of the pilot farm startedat the end of 1969 and the studybegan in October 1971.  - 27 - 6. In May 1972,a joint FAO/RMWApreparation missionwas sent to Chad to reviewthe preparation of the feasibility study as well as the resultsobtainedin the Bol-Guini experimental farm and pilot project. The mission concluded that becauseof the very high yields achieved on the experimental farm and the pilot project,the projectwas worth considering for IDA financing in spite of the high investment costsper hectare. The mission also requested furtherpreparation work to resolvekey issues relatedto: (i) attainable and maintainable farm yields; (ii) the need for fertilizer application duringthe first few years of cultivation; (iii) Government policy regarding production input subsidies;(iv)optimumsize of the averagefamilyfarm; (v) SODELACcarryingout other activities than the management of the proposedproject;and (vi) the possibility of reducingthe cost of projectengineering works throughforce accountor international competitive bidding, rather than throughhigh-cost local contracting. 7. The Projectwas appraised in October )74 by a team of IDA and AFDB staff. Discussions took placewith FAC and CaisseCentralede Cooperation Economique(CCCE)in Chad and in Paris. The two issuesraised at the Decisions Meeting on November26, 1974were: (i) SODELACReorganization. Government was requested to resolve SODELAC's financial difficulties prior to negotiations. Government agreedand confirmed that proposalsfor reorgani- zationof SODELACwould be submitted to IDA for consideration by December1974. If Government failedto do so prior to negotiationsor if the proposedreorganization was not acceptable to IDA, a separate entitywould be createdto carry out the project. Subsequently, the latter solution was adoptedbecauseit would allow timely processing of the project,even though it was not desirable from an institution-building standpoint. (ii) FurtherDevelopment of Lake Chad Polders. Therewere severalconstraints which could affectprojectimple- mentationincluding difficulty of access,harsh living conditions and the need to attractpopulation to the project area. The minimumsize of an initialdevelopment of the polderswhich would fullyjustifythe investment, and the buildupof an organizational framework was about 2,800ha (or the development of the three poldersof Guini,Berim and Mamdi). However,to minimizethe financial and operational risks the projectincludedthe development of Guini and Berim poldersonly. Development of Mamdi would be considered only after the engineering design for Mamdi had been completed and the settlement on Guini and Berim polders had been proven successful. Furtherdevelopment of Lake Chad Polders also faced the issuesof whether capital intensive and sophisticated water development schemesshould be pursuedin preference to lower cost agricultural development projects; and whetherpriorityshouldbe given  - 28 - to development in spareselypopulatednorthernChad rather than in the populoussouthernregions. 8. Negotiations were held from September 8-12, 1975. Agreement was reachedthat the projectunit, Projectd' Amenagement des Polders (PAP), would be relatively autonomous from SODELAC. Further,severalminor changeswere made in the draft legal agreements, relatinginter alia to the date for reviewing epidemiological studies,and the execution and delivery of the Subsidiary Loan Agreement, which was deletedas a condition of Credit Effectiveness.Boardpresentation took place on November11, 1975 and the Credit Agreement was signedon November26, 1975. ProjectDescription 9. The projectoriginally consistedof: - rehabilitating and completingthe irrigation and drainagenet- works of Guini polder (370ha net area); - constructing the irrigation and drainagenetworksfor Berim Polder(800 ha net area); - establishinga Commercial Agricultural DevelopmentSectionunder the projectwhich would ease the way for farmersettlement, and be responsiblefor the firstyear of operations on all newly developedland, and for the basic cultivationsduring the follow- ing year; - providinga packageof services to the farmerswho would take up cultivationof the poldersafter the initialfirstyear of operationsby the CommercialAgriculturalDevelopment Section, includingresettlement assistance,training,extension, credit and marketing; - constructing 3 servicecentersand 16 houses for projectstaff; - expandingadaptiveagricultural researchat the Matafo Research Station;and - providing264 man-monthsof consultant assistance to Project d'Amenagementdes Polders (PAP)for projectimplementation and 108 man-monthsof technical assistancefor Matafo Research Station.  - 29 - III. IMPLEMENTATION Effectiveness 10. The conditions of effectiveness were as follows: - approvalof AFDF Creditby AFDF board and approvalof PAC grant by FAC committee; - notificationof IDA that Governmenthad paid its outstanding share of SODELAC'scapitaland the subsidies budgetedfor SODELAC In 1972-74; - appointment of ProjectUnit Manager;publicationof the decree restating SODELAC'eauthority to organizesettlementof farmers on newly developedpolders;and - promulgationof long-term contractbetweenSODELACand Cotonchad and GrandsMoulinsdu Tchad (GMT)respectively. These conditionsof effectivenesswere fulfilledon time and the Credit became effectiveon February26, 1976 as scheduled. Procurement 11. Pquipment,cementand vehicles were procuredthrough international competitivebidding (ICB)in accordance with Bank guidelines. The same French consultingfirm that had preparedthe projectwas chosento providethe technical assistance after internationalconsultation and negotiations on contractconditions. Start-upof Works 12. Due to delays in the preparation of biddingdocuments and in awardingcontracts, irrigation works startedone year behind schedule. The delay resultedfrom the late arrivalof expatriate staff:the appraisal reportassumedthat the expatriate staff would have been appointed and would have startedwork shortlyafterprojectnegotiation in the fall of 1975, but the contract with the consulting firm was not signeduntil April 1976 and the expatriate staff only resumedduties in May 1976. PhysicalProgress 13. BeforeIDA activities came to a standstill in May 1979, 220 ha of Guini Polderwas under irrigation. Only five of the sixteenhouses estimatedat appraisalwere built. Key indicators(Annex1) show project accomplishmentsas comparedto the appraisal estimates. These works were carriedout at a relativelyslow pace due to technical difficulties.The project'smost seriousand persistent problemwas the transportation of  - 30 - constructionmaterialswhich had to pass throughthe channelfrom Bol to Lake Chad. From an agronomicalstandpoint, therewere also difficulties relatedto double cropping(thelate cotton crop precludedsowingwheat at the optimalperiod). Costs,Financingand Disbursements 14. ProjectCosts.The appraisal reportestimated projectcost net of tax at CFAF 2,925million,equivalent to US$13 millionat the prevailing rate of CFA 225 per US dollar. The cost estimateincluded physical contingencies of 15% on irrigationworks and 5% on all other projectitems. Price contingencies appliedon baselinecostswere compounded as follows: Annual Rates 1975 1976 1977 1978 1979 Cost of Civil Works 6 14 12 10 10 Cost of Other Items 12 10 8 7 7 15. Actual costs have been difficultto establish becauseimportant projectcost data were destroyed duringthe civilwar. An attempthas been made to estimateactualcosts by using actualdisbursements made by IDA and PAC up to June 1979 and stagesof completion of AFDF and USAID-financed components. For the projectactivities financed up to June 30, 1979,the estimated actualcosts and appraisal estimatescompareas follows: Appraisal Estimated Actual as X of Estimates Actual Cost Appraisal ____ ------- US$ '000… Irrigation Works 5,490 4,687 85 Commercial Unit Investment 900 1,258 139 Agricultural Research 860 791 92 Project Unit Overhead 1,230 1,458 119 Other 140 N.A N.A Contingencies 4,380 - - TOTAL 13,000 8,189 63 16. Actual expendituresdeviateconsiderably from the originalcost estimates. Althoughan estimated 63% of the totalprojectcostswere spent, only 32% of the expectedphysicalinfrastructure was completed. The significant differences betweenactualexpenditures and originalcost estimates are partly attributableto the unexpected high cost of transportation, land levellingand operations and maintenance. The delays in the start-upof projectoperations, high inflation, and the subsequent effects of the civilwar also contributed to cost overrun. 17 Financing. The originalfinancing plan was adheredto. APDF financedthe irrigationworks built by local contractorsand the equipment  - 31 - requiredfor their construction, the procurement of PVC secondary distribution pipes, and the constructionof staff houses and service centers. USAID financedagricultural research componentfor cereals. AFDF and USAID financingwas parallelto IDA and FAC, who financed pari passu (71%-29%) all other projectitems. 18. Disbursement.Annex 2 comparesappraisal estimateswith actual disbursement.Disbursement over time startedslowlyrelativeto appraisal estimatesdue to delays in the start-up of projectactivities. Disbursementdid not exceed 86% of appraisal estimates throughthe implementationphase till creditsuspension in May 1979. In May 1985, IDA disbursedthe balanceof credit funds (US$1.2 million)to the Ministries responsiblefor agriculture in supportof their rehabilitation and drought-alleviationactivities. ProjectManagement 19. The Projectd' Amenagement des Polders (PAP)was responsible for projectmanagement with the GeneralDirectorof SODELACas its president. The PAP was to be managed by an expatriate projectmanagerand staff according to an established organization chart. In reality,the General Directoref SODELAC, who was relatively inexperienced, failedto delegate authority to the projestmanager;from the start of the projecthe got closelyinvolvedin projectexecution and took all decisions himself. Thus, an ambiguous and conflicting situation prevailed regarding responsibilities of the GeneralDirectorof SODELACand the project manager. In practice,the GeneralDirectorimposedhis views and it was extremely difficult for the projectmanager to challenge him or have projectexecution under control. This resultedin dilutionof responsibilities, ineffective use of the technical experts(expatriates) providedto SODELAC,and delayswhich led to unsatisfactory project performance. That also explained why two successive projectmanagers resignedtheirposts. IV. AGRICULTURAL AND ECONOMICINPACT 20. AgriculturalImpact. The objectives of the projectwere to increasecottonand wheat production throughthe construction and rehabilitationof poldersin the Bol area of Lake Chad region,resettle herdsmenin the projectarea, and promote the evolution of high technology agricultureas insuranceagainstthe disruptive effectof drought. At appraisal,the projectwas expectedto cover 1,200ha and to reach 1,200 families. Unfortunately,these projectobjectives were not met, mainly becauseof politicaland climaticconditions. 21. As indicated in Annex 1, productiveactivities took place only in the 1978/79croppingseason. This was followed by civil war which resulted in damages and losses involving lootingof equipment, suppliesand furniture,destruction of houses and offices,and theft of cars and trucks.  - 32 - Meanwhile, therewas deterioration of equipment and irrigation works due to lack of maintenance. The situation was compounded by drasticdeterioration of hydro-climatic conditions. By 1984,after the secondmajor drought,the water level of the Lake was 3 meters below the level at which gravity irrigation is possibleand even pumpingwould be difficult, if not impossible, under the circumstances. The crop rotation of wheat-cotton as envisaged duringproject appraisal proved to be technically and agronomically unfeasible. The projectis, therefore, considered to have littleor no agricultural impact. 22. EconomicImpact. The appraisal reportestimated the ERR of this projectat 17%. In its assumptions, the agricultural researchcomponent and the cost of consultingservicesto PAP were excludedsince the benefits they would generatewould be attributable to development activities beyond the scopeof the project. A shadowexchangerate of CFAF 270 per USdollar was used to take accountof the prevailing importand exporttaxes and the then weak Chadianbalanceof paymentsand low level of foreignreserves. The sensitivity analysisindicated that the rate of returnwould drop to 11.5%if costswere to increase by 25%. 23. Becauseof the civil war and coupledwith changesin hydro-climatic conditions, this projecthas no noticeableagronomic or economicimpact. No attemptis made to recalculate the ERR of the project, which in any case would be negative. V. PERFORMANCE Consultants 24. Under the circumstances,any assessmentof consultants' performance would be speculative. Qualificationsof the expertsprovided were good and somuiof themwere exceptionallymotivatedto ensure the successof the project. But the two successive projectmanagers were never given the opportunity since the GeneralDirectorof to take responsibility, SODELACdid not delegateauthority. Borrower 25. After some delays in the start of the project,becauseof Government delay in finalizing contractswith the consultants,implementa- tion proceededsmoothly until managementproblemsbetweenthe ProjectUnit and the GeneralDirectorof SODELACsloweddown projectactivities. The problemswere reportedto Government officialsbut no actionswere taken to resolveit. Moreover, local projectstaffwere weak and Government failed to assignmore competentstaff,while refusingto recruitexpatriates as substitutes.  - 33 - Co-financier-AFDF 26. The generalprojectmanagementproblemsalso affectedthe implementation components, of AFDF-financed which were furtherset back by confusionaboutmethodsof procurement and delays in the disbursementof funds. Bank 27. After an identificationmisoionin 1968,the Bank sent a prepara- tion missionin 1972. Identification and preparation was performed primarily by FAO/RMWA. Appraisal took place in October 1974. The Bank sent nine supervision missions,an averageof two missionsper year from 1976 to 1978,one in 1980, and two in 1984. The totalnumber of man-weeks spent in the fieldwas 17.9. Four of the missionshad the participation of an engineer. The continuity was good with the same projectofficer participating in all missionstill after 1978. 28. While the technical supervisionwas done competently,the Bank could in hindsighthave taken a strongerstand in the face of the project's management problems. In particular, the Bank did not clearlydefine the rolesof the ProjectUnit and SODELACbeforeprojectimplementation started, resulting in much frictionbetweenthese two institutions. However,supervision missionsand follow-up lettersoften urged Government to take measureswhich could improveoverallprojectperformance. The response was not encouraging,particularly with respectto the performance of SODELAC. VI. SPECIALISSUES- LESSONSLEARNED 29. The main issue of this projectwas the organization of SODELAC. Once IDA made a decisionto createa separate Unit within SODELAC to carryout the project,adequate effortsshouldhave been made to carefully delineate the functions of the ProjectUnit and SODELAC Directory. Unfortunately, in an attemptto speed up projectprocessing, a number of loose ends were not tightened, which showedup duringproject implementation. If the roles of SODELACand ProjectUnit had been clearly definedat the outset,a number of thesemanagement problems which hindered projectimplementatior might have been avoided. 30. There are also questionsabout the soundness of projectdesign. The projectdesignwas highly sophisticated and very costly to maintain. The croppingpatternof wheat-cotton was both technically and agronomically difficult to implement. Irrigated cottonis not cost-effective in Chad, becausecottoncan be grownmore cheaplyin the South. One therefore wonderswhether the desire to maximizeeconomicreturnson the projectdid not take precedence over sound agronomicchoice of croppingpattern.  - 34 - 31. It shouldbe recognized that the projectarea has good agricultural potentials, given a soundlydesignedproject.Small traditional irrigation schemes, similarto the schemecurrently financedby USAID in the projectarea or by IDA in the SateguiDeressiaarea, are worth considering for futuredevelopment in the Lake Chad region. It shouldbe equallyrecognized that IDA's decision to scale down projectsize from 2,800 to 1,170ha was sound; in hindsight, projectsize shouldhave been even smaller. The idea of cofinancing the projectwith other donors also proved to be beneficial. Becausethe projectwas cofinanced with other donorswho are activein Chad, therewas a systematic approachto the development of Lake Chad Polders and the experience of the projecthas convinced the donors that any futureprojectin the area shouldnot depend on theoretical models but on real chancesof success. May 31, 1985 WAPAD  -35 Annex_I CHO LAXE CHAD POLDERSPROJECT (Credit 592-CD) REPORT PROJECTCOMPLETION 2/ Key Indicators Appraisal 2 of Appraisal ----Reali.ed-- -- tstim e- ---Estimat es- 1977/78 1978/79 1977/78 1978/79 1977/78 197S/79 1. Cultivated area (ha) Cotton - 220 2032 270 0 81 Wheat - pm 50 520 0 0 Potatoes pm 50 - - - _ Maiz - PM - - - - 2. Yields (t/ha) (forecast) Cotton - 3.5 4.5 2.6 - 134 Wheat - - 3.1 2.6 - 0 Potatoes 25 _- - - 3. Production(t) Cotton 770 90 715 0 108 Wheat - 155 1,330 0 0 Potatoes PM 3/ 1,250 - - - Maize - pm - - - - 4. No. of direct beneficiaries 220 300 -3 Characteristics campaign, of the 1978/79agricultural as comparedto appraisal estimates,can be summarizedas follows: with cottonwas 812. (a) Area cultivated (b) Wheat was not grown. (c) Cotton yield was 35% higher,resultingin an 82 higherproduction. was 73%. (d) Number of benefi:iaries 1 / Credit was suspendedin 1979 and project activities were never resumed. 1/ iO fia of cottonon experimental farm. was lost becauseof inadequate 3/ A large portionof the production means of transportation. 4/ 1 ha per beneficiary.  - 36 - Annex 2 CRAD LAKECHAD POLDERS Credit 592-CD) PROJECTCOMPLETIOidREPORT Disbursement Table IDA FiscalYear Actual Disbursements I/ Appraisal 2/ Actual Semester Apprai.sal 1976 1st - 400 N.A 2nd 400 NA 1977 1st 625 1,000 63 2nd 1,136 2,000 57 1978 1st 1,413 3,000 47 2nd 2,143 4,000 54 1979 1st 3,262 4,200 78 2nd 3,800 4,400 86 1980 1st 3,800 4,600 .83 :nd 3,800 . 5,000 76 l/ Creditwas suspendedin 1979and the balanceof credit funds (USS1.2million)were disbursedagainstthe rehabilitationand activities drought-alleviation of the Ministriesresponsible for Agriculture. at 12/31/80 2/ ProjectClosingDate was estimated at prolectappraisal. MAO SECTION  IBRD11447 _ -(0 -ra CHAD MARCH 1975 ,,--0 Ira CHIiA D "Margoum LAKE CHAD " -w' "' POLDERSPROJECT APPRAISAL REPORT " 'I / >-. Ngorangou AREA PROJECT Ns [ m1 "\ -3Dried -\ Polder Polder to be Dried Dunes …-- Rood (Inset) Boundaries Internotional N L- t.- .''I (;.. -- / I Detail \ shown 11448 onIBRD L1BYA 0 201, 4( Y ~~ ~~ ouo 0 ~ ~ ~ ~ ~ ~ ~~~~~~-. o Y8ua -VS ) \ -200 N 20 o N IG R i Bogourmi / C H A D j Kotokorom \' / ; Iool,on Z 0~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~1 NIGERIA .amcnD .N I~~~~ 0 ,C ,- ./ ., -_ \, . KILOCVETEr,I CENTRAL .fCAMERt'CNL _. AFRICAN REPUBLIC I' ____________________________________________________________________________________ / 0O