Burundi - Fisheries Development Project

Document of The World Bank FOR OMCIAL USE ONLY Reoft N.. 5683 PROJECT PERFORMANCE AUlDIT REPORT BURUNDI FISHERIES DEVELOPNENT PROJECT (CREDIT 626-WU) June 3, 1985 Operations Evaluation Department 'This document has a restricted distibutio and may be used by recipints 1uy in the perfemance of their officisa dutie Us cotents may mot odberwise be disclosed without World Bacnk _nhwraton-  WEIGrTS AND MEASURES Decimal Metric US 1 kilogram (kg) 2.205 pounds (lbs) 1000 kilograms (kg) 1 quintal (qt) 1 metric ton (mt) 2,205 pounds (lbs) I hectare (ha) 2.471 acres (ac) 1 kilometer (km) 0.62 miles (m) ABBREVIATIONS ADF Abu Dhabi Fund BNDE Banque Nationale de DEveloppement Economique BRD Banque de la Republique du Burundi FAO Food and Agricultural Organization of the United Nations GDP Gross Domestic Product GNP Gross National Product LOA Loan Administration Department OMS Operational Manual Statement p.a. per annum PPAR Project Performance Audit Report RDC Regional Development Company SAR Staff Appraisal Report SUPOBU Societe -Usine de Poissons de Burundi- UJNDP United Nations Development Programme CURRENCY EQUIVALENTS Curreacy Unit Francs Burundais (FBu) Appraisal Year Average US$1-00 = FBu 78.35 Intervening Years Average US$1-00 = FBu 90.00 Completion Year Average US$1.00 = FBu 90.00 At time of audit US$1.00 = FBu 122.08  FOR OMCIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT BURUNDI FISHERIES DEVELOPMENT PROJECT (CREDIT 626-BEU) TARETF 3 CDNTENTS Page No. Preface .............. ..............i............... Basic Data Sheet i.......... ........... .. ..i. Righlights .. ................. ............ ........ iv Project Implementation Photographs PROJECT PEFORMANCE AUDIT ERANUM I. INTRODUCTION ..1 Program Context 1 National Setting and Economy . 2 Agriculture.. 3 Fisheries.. . 3 II. PROJECT SUMKARY ..................**......*......***.....w. 4 Background and Project Formulation. . . . 4 Project Implementation.... 6 Project Administration. ..... 7 Impact and Sustainability... .. 9 III. PROJECT ISSUES...... 13 A. Fisheries Resources Research on Lake Tanganyika 13 B. Project Processing and Start-up 14 C. Project Formulation ..14 D. Signing and Effectiveness ..17 E. Disbursements ..20 F. Fisheries Statistics ..26 G. Impact on the Project by Other Bank-Sponsored Activities ..26 Tables: 1. Burundi - Lending Operations .. 29 2. Agricultural Operations in Burundi . . .30 3. Project Processing Steps and Dates.. 31 4. Delays in Signing and Effectiveness . . .32 5. Estimated Total Catches, 1970-1983 . . .33 b. Industrial Fishing Sector Data, 1981-83 . . .34 7. Artisanal Fishing Sector Data, 1981-83 . . .35 8. -Coutumi6re- Fishing Sector Data, 1981-83. 36 9. NJumbers of Fishing Boats Built by SUPOBU. 37 10. Organigram of Fisheries Administration in Burundi ...38 This doumnuot has a restricted distribution and mabe used by recipients only in the perfonnance ofI their offil duties Its cootents may not otherwise be discosed without World Bank authorization.  TABLE OF CONTENTS (cont.) Page No. Annexes: 1. Project Description (Schedule 2 of the legal documents) . .... 39 2. Withdrawal of the Proceeds of the Credit (Schedule 1 of the legal documents) . ................... 45 Attachments: L. Comments from BNDE ........ &............ 47 2. Comments from SUPOBU ...... .................................. 49 PROJECT COMPLETION REPORT I. Introduction ............................................. 55 II. Summary Description of the Country and Sector Setting .- 55 III. Project Formulation and Description ...................... 58 IV. Project Implementation ............ ....................... 70 V. Government Request for Extension of the Credit Closing Date ........................................... 84 VI. Institutional Performance ................................ 86 VII. Financial Performance .................................... 92 VIII. Lessons for the Future ................................... 95 Annex A: Table A 1 - Status of Credit Reimbursement for Fishing Units as of August 1982, Kitaza Center ......... .......... 99 Table A 2 - Status of Credit Reimbursement for Fishing Units as of August 1982, Magara Center ................... 100 Table A 3 - Status of Credit Reimbursement for Fishing Units as of August 1982, Rumonge Center 101 Table A 4 - Semi Industrial Fishing Unit Iridina Operating Results, 1982 ............................ 103 Table A 5 - Semi Industrial Fishing Unit Mwisero- Operating Results, 1982 ............................ 104 Table A 6 - Fishing Statistics, Monthly Average for 1981 and April-June 1982 .................................... 105 Table A 7 - Schedule of Disbursements .......................... 106 Table A 8 - Regional Development Society, Balance Sheet ........ 107 Table A 9 - Regional Development Society, Income Statement 108 Table A10 - Regional Development Society, Sources and- Application of Funds ............................... 109 Table All - Table 3 - Statement of Patrimonial Status, 12/31179 ........................................... 110 Table A12 - Table 3 - Statement of Patrimonial Status, 12/31/80 ........................................... ill Table A13 - Table 3 - Statement of Patrimonial Status, FY81 112 Table A14 - Mianagement Funds Status as of 12/31/79 .113 Table A15 - Managewent Funds Status as of 12/31/80 .114 Table A16 - Management Funds Status as of 12/31/81 .115 Map IBRD 11587 R  PROJECT PERFORMANCE AUDIT REPORT BURUENDI FISHERIES DEVELOPMENT PROJECT (CREDIT 626-BU) U,, PREFAC This is a performance audit on the Fisheries Development Project in Burundi, for which Credit 626-BU in the amount of US$6.0 mifl3on was approved in May 1976. The actual credit closing date was May 17, 1983, almost a year later than envisaged at appraisal, and the umdisbursed balance of about US$2.0 million was cancelled. The audit report consists of an audit memorandum (PPAM) prepared by the Operations Evaluation Department (OED) and a Project Completion Report (PCR) dated July 26, 1984. The PCI was prepared by the Eastern Africa Regional Office on the basis of a mission to Burundi in October 1983. The audit report is based on a review of the Appraisal Report, No. 977a-BU dated March 22, 1976, the President's Report, No. P-1807-BU dated April 8, 1976, and the Credit and Project Agreements of June 11, 1976. Identifications preparation and supervision reports, correspondence with the Borrower, IDA memoranda and other documents contained in relevant files have also been reviewed, and IDA staff, Government officials and others associated with the project have been interviewed. Reference has also been made to an OED report entitled, Harvesting the Waters - A Review of Bank Experience with Fisheries Development, OED Report No. 4984 dated March 13, 1984. An OED mission visited Burundi in September 1984. The information obtained during the mission was used to test the validity of the conclusions of the PCR. The audit finds the PCR excellent in most respects in its coverage of project implementation experience and concurs with the PCR's principal findings. Points discussed in the audit memorandum have been selected because of their importance to this and other Bank-supported projects. The draft report was sent to the Government and cofinancier for comments on February 19, 1985. Comments received from BNDE and SUPOBU have been attached to the audit memorandum. The valuable assistance of Government officials and staff of the fisheries company is gratefully acknowledged.  PROM= PEzuIouAm AM=l umPOr BURDEI FISHERIES. DEWEIEr PROM.T (CDIT 626-) BASIC DA SHEST IKEY PUDIECT D*AT Actual or Actual a K 4Appraisal Estimateda o Appraisal Esindmate Actual Estimate Project Costs (US$ million) 8.6 6_0 70 Credit unt (CUSS millIoe) 6.0 4.0 67 Cofinacing (US mlle.) 1.0 1.0 100 Date Physical Components Completed 07180 06/79 Proportion then Completed - Su losber of Units Constructed/a. 600 fb 100 /c 16 Amount of Credit Disbursed by RC to Fishermen (Millio FE.) 59.3 7.5 13 ount of Credit Recovered by RDC (Milon FBc) 67.4 2.5 Id 0 No. of Fishsng Centers Constructed 5 4 80 NoD of FLsig Stations Constructed 5 1 20 Economic Rate of Return (K) 20.0 Negative - Financial Rate of Return (Z) 34.6 Negative - Number of Direct Beneficiaries 3,000 900 30 Cumulative Estimated and Actual Disbursemets (US$ mllion) FY76 FY77 Frh PY79 FP80 F781 Fys2 Appraisal Estimate 0.6 1.9 2.7 4.1 5.3 6.0 - Revised Estlmte - 0.6 2.0 4.1 5.3 6.0 - Estimated Actual - 0.6 1.4 2.1 2.9 3.6 4.0 Actual as Z of Appraisal Estimate - 32 52 51 55 60 - Date of Final Disbursement May 17. 1983 Canceled US$2.04 illion (05/17/83) OTHER PRO3ECr DA Actual or Expected Estimated Actul Flrst Nentioned In Bank Files n.a. 01/31/74 Identification/Preperation n.s. 03175 Appraial U.S. 09/75 Negotiations n.a. 03/03n6 Board Approval U.S. 04120/76 Credit Agreeent Date U.&. 06/11/76 Effectiveness Date 06130/76 02122177 Credlt Closlig Date 06/30/82 05/17/83 Project Completion 12/31/81 Not completed Borrower Government of Burundi Executing Agency SUPOBU. RDC witb assistance from BNDE Cofinancier Abu Dhabi Fund for Arab Economic Development Preceding Project None, but there vas a UliDP Project /e Follow-on Project None  mm DM betweem Date Passin s.- of Mondays In Speciallxatioss hxfomnTypes of (So.Iyr.) Ohaths) Persoa Field A lisee noting _I Trad P_roem /k Identification & Prepzrl- ID 02/75 n 1 14 DllK (9/75 a. 3 45 A,Z,D supervision Im 10/76 u-a 25 A,S 2 n. SNP IDl 0877 8 1 7 A 2 2 N,T IDA a2/8 6 2 26 A.C 2 3 1,?.S IDA 0717 5 1 7 A 2 2 HI,FT ID& 02/79 7 2 11 A, 2 1 F,8,T IDA 08179 6 4 20 AACj,D 2 2 FP3T.P IDA OL/50 5 3 21 A,C,D 2 2 F,l,T,P IO& 05t80 4 1 4 A 2 2 FYT IDA 01/81 8 1 7 A 2 2 7,3,P IDA 08181 7 2 is A.A 2 2 7,3,? IDA 0282 6 4 10 A.A,B.3 2 2 l,.?P MDA 10182 8 1 8 3 3 2 1,F ID Supervision Effort 11 (Staff Weeks) by Fiscal Tar FY77 FY78 FT79 Fn80 FY1 FT82 FY77-FY82 10.7 16.7 16.8 35.0 5.9 18.2 103.3 /a A unit cocsdsts of two boats (canoes) formin a catmran. However. each boat can be purchased ad umd individually. lb The appraisal report did not speify the nuber of boats ibich imald be contructed. but estimated 450-600 nits. The target for the boat contraction progr was later set at 600 anits. /c An additional 250 boats bcd bee supplled by the earlier PADIUDP project. 7d The inout of credit actualy recovered repremets 33Z of the m_nt actually disbursed. Te PPlA pares. 12, 36-35. 7T Appraslal report lssued 03/2n/76. T hbese are approximate figures. as the project us often supervisd along with other projects wlthout always clear indication of bow niny days re spent on each project. /h A - Financial Analyst; 3 - Agrlc. Economist; C - Flsheries Specilist; D - Consultant. 1i 1 - problem free or mlnor problem; 2 - moderate problem; and 3 - Najor problem. I 1 - Improving; 2 - stationary; and 3 - deterlorating. Pk S - Staffing; M - Managemnt; P - Polltical; T - Technical; F - Financial 1l Includes Inputs by supervising projects divisLon. AGX and cosultants.  -iv-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ PROJET PERADIT REPORT BURUNDI FISHERIES DEVELOPMENT PROJECT (CREIT 626-BU) H{IGHIGHlT:S the main objectives of the project were to rai,se the standard of living of artisal fishermen along Buradi' L e Tanyika shore by increasing their fish production, and to market fish oghut the country, thus raising proteln intakes and thereby improving the nutritioa of rural consumers. IM addition, the project almd to produce a regional development plan for the lakeshore area and to initiate pilot development activities. Although artisanal fish production in 1983 (8,400 at) was the same as at the start of the project in 1977, the effect of the expulsion of Zairian fishermen in 1978 was to halve available fishing effort. In con- sequence, the catch for 1979 was only 4,700 at (PPAM, Table 5). The project, therefore, has undoubtedly assisted in restoring artisanal fish production in Burundi and in the process has contributed to raising the standard of living of many local artisanal fishermen. Efforts to promote fish consumption in rural areas did not succeed, partly because of inadequate project management, but also because total fish supply, which has now stabilized at around 11,000 at per annum, is barely enough to satisfy the demand for fresh fish in Bujumbura. Thus, even had management been more effective, the result would have been much the same so far as rural fish supplies are concerned. The regional development component was not carried out, although project funds were used for a soil survey and for two small area studies which appear to have had little or no impact on the Lake Tanganyika coastal region. Fewer boats than originally intended were built and most of these proved to have a shorter working life than the sub-loan repayment period so that loan recovery was very poor and probably would have been so even had the credit scheme been better designed and more effectively managed. Headquar- ters facilities, four outlying fishing centers and one station were con- structed, compared with a planned five ceaters and five stations, but even those that were built quickly proved uneconomic and were soon closed down. Extension services were developed and benefited an estimated 900 fishermen, but this was only 302 of the appraisal target. A number of local staff received training in Burundi and others overseas, in such matters as account- ing, credit operations, general administration, boat construction and engine maintenance, etc. Unfortunately, many of the trainees left to seek work elsewhere after training.  The generally poor performance of the project can be attributed to a number of factors, pointing to important lessons for the future: (a) project preparation was defective in terms of both time and techni- cal expertise and failed to take fusll account of the inadequacy of local managerial and technical skills and there was a need for a simple and phased development program (PCR para. 3.10 and PPAM paras. 23, 41-47); (b) project design was overly complex in relation to the management and institutional capabilities available. The iaplementing agency, SUPOBU, remained extremely weak throughout the project period (PCR para. 4.40, PPAN paras. 23 and 45); (c) project appraisal was apparently adequately staffed, but rather hastily arranged and failed to adopt a sufficiently critical approach to the resolution of project issues (PCR para. 4.40, PPAM paras. 45-47); (d) the project received insufficient technical assistance to offset the adverse impact of poor management. Sufficient funds were available, but only about one-third were utilized, partly because of recruitment difficulties, but mainly because of Goverament/SUPOBU procrastination and disinclination to use the services of expatriates (PCR para. 4.44, PPAM para. 23); (e) IDA supervision was weaker than it should have been in terms of staff input and was especially short of fisheries expertise. Suit- able boat-building expertise was finally introduced, but too late to be of much help to the project, although it did result finally in the production of an excellent fishing vessel. Supervision also suffered from IDA reluctance to act against Government and SUPOBU disinclination to apply the appropriate performance rating (PCR para. 4.41, PPAM paras. 24-25); (f) IDA's system for verifying disbursement requests was demonstrated to be defective in failing to prevent the disbursement of substan- tial sums for purposes not provided for in the legal documents (PCR para. 6.20, PPAM paras. 55-60); (g) initial disbursement delays could have been avoided had a staff member from the Loan Administration Department accompanied an early supervision mission to explain Bank procedures to the local staff concerned (PCR para. 6.20, PPAM paras. 65-66); and (h) the impact on the project of other Ban-sponsored developments, in particular the reconstruction of the lakeside highway from Bujumbura to Nyanzalac (PPAM paras. 70-72).  PROJECT PERFORMANCE AUDIT NERANDUK BURUNDI FISEERIES DEVELOPHENT PROJECT (CREDIT 626-BU) I. INTRODUCTION Program Context 1. The Fisheries Development Project was the Bank's third credit for agricultural development in Burundi. The two earlier credits were made for coffee improvement. 1/As of June 30, 1984, a further three agricultural development projects had been approved.2/ 2. Total Bank lending to Burundi, as of June 30, 1984, for 24 projects, amounted to just over US$200 million. As can be seen from Table 1, the Bank's lending program is quite diversified. Agriculture accounts for 262 (6 projects, US$52.73 million), followed by roads with 24Z (5 projects, US$49.25 million), and education 20Z (3 projects, US$40.8 million). 3. Implementation experience with the agricultural projects has not been unfavorable (see also PPAM, Table 2). While the Fisheries Project fell considerably short of meeting its objectives, the results of the two coffee improvement projects were generally good. Under the Second Coffee Improve- ment Project, the coffee development component exceeded appraisal estimates in both quantity and quality; however, in contrast, the other components either fell well below targets (foodcrop and livestock development) or were not implemented (rural development). The Forestry Pioject has been progres- sing satisfactorily; targets will be met and implementation completed in time. The two rural development projects have been seriously delayed due to funding problems, difficulties in arranging for technical assistance, and management problems, but the Kirimiro Rural Development Project is now well underway, and management and funding of the Ngozi III Integrated Rural 1/ See also PPAR - Arabica Coffee Improvement Project (Credit 147-BU), OED Report No. 1531 dated March 21, 1977; PCR - Second Coffee Improvement Project (Credit 593-BU), OED Report No. 4576 dated June 23, 1983; and Impact Evaluation Report, Burundi: First Arabica Coffee Improvement Project, OED Report No. 3314 dated February 1, 1981. 2/ Forestry Project (Credit 918-BU, approved May 24, 1979); Kirimiro Rural Development Project (Credit 1169-BU), approved June 16, 1981; and Ngozi III Integrated Rural Development Project (Credit 1192-BU), approved December 8, 1981.  -2- Development Project have improved.3, A second forestry project is being appraised and a pilot settlement schame, including development of coffee and foodcrop production in the Muyinga province, is under preparation; preparation is also underway for a national agricultural services project.4/ 4. The present audit report is the fifth OED report evaluating experi- ence with project implementation in Burundi. Experience with these projects reveals certain common aspects. In particular, most of the projects experi- enced difficulties and delays in recruiting technical expertise from abroad, exacerbated in some cases by an evident reluctance on the part of the Burundi authorities to use expatriates, or to replace them even when the initial incumbents left early (Water Supply, First Coffee and Fisheries Projects). At the same time, project preparation and appraisal missions have often over- estimated the managerial capacity of local organizations to implement project activities (Water Supply, First and Second Coffee and Fisheries Projects) which were sometimes over-designed and unrealistically complex, resulting in failure to undertake or complete all components-indicating a need for a more critical approach to appraisal. Incompatibility between Bank reporting and accounting requirements and those practiced locally had been a cause of dif- ficulty and at times dissension (Water Supply and Fisheries Projects), made worse by breakdowns or at least long delays in communication in both direc- tions and consequent misunderstandings between the Bank and the borrower (Water Supply, First Coffee and Fisheries Projects). Socio-economic condi- tions and producer incentives were noted as having been inadequately con- sidered during preparation and appraisal of the First Coffee Project, and the same thing appears to have happened with the Fisheries Project. On the other hand, the First Coffee and Road Development Projects produced significant institutional improvements to the benefit of their subsequent projects, unfortunately in marked contrast to the Fisheries Project. National Setting and Economy5! 5. Burundi is a small land-locked country ia central eastern Africa. It covers a territory of about 27,800 km2 and has a population of about 4.4 million inhabitants. Population is expected to increase at 2.7X p.a. during the current decade, but with a density of about 158 per km2 Burundi is 3/ See Status of IBRD/IDA Projects in Execution (as of September 30, 1984) SECM84-921 of October 31, 1984. 4/ See Monthly Operational Summary of Bank and IDA Proposed Projects (as of January 31, 1985) SECM 85-173 of February 20, 1985. 5/ See President's Report No. P-3960-BU dated March 22, 1984 (Third Techni- cal Assistance Project) and Burundi Economic IMemorandum No. 4784-BU dated December 26, 1984.  -3- already the second most densely populated country in Africa.6/ Population pressure has led to soil erosion, declining soil fertility, and deforesta- tion. 6. With an estimated GNP per capita of US$240 (1983), Burundi is one of the world's least developed countries. The overall literacy rate is low, morbidity is high, and basic social infrastructure is stili lacking. Limited natural resources, population pressure, scarcity of qualified and experienced manpower, and isolation from main trade routes, explain the country's poverty. 7. Burundi has been quite successful in obtaining aid from the inter- national community which has financed a substantial part of domestic invest- ment. Economic performance during the third plan period (1978-82), was deemed satisfactory by the Bank. Value-added grew only slowly, but food production barely kept pace with population growth. The secondary and ter- tiary sectors performed better, but due to a 42Z decline in the external terms of trade, per capita gross domestic income fell by over 1.1% per annum. Agriculture 8. Nearly 95% of Burundi's popu'ation is rural, living on scattered homesteads. Agriculture, mostly smallholders who grow subsistence crops and coffee, contributed 60% of GDP and 90% of exports. A few large farms grow sugarcane, tea and quinine. Of the approximately 2.3 million ha of land available for agricultural use, three-fourths are presently used, with the remainder being of low soil fertility or situated in areas prone to diseases (see also PCR paras. 2.02-2.04). Important issues in agriculture relate to productivity increases, soil fertility and erosion control. Foodcrop production is necessary to ensure supplies for the fast growing population, while other possibilities for the generation of export revenues are limited and require rationalization of the country's traditional export potential and diversification into new export crops. Fisheries (PCR paras. 2.05-2.11) 9. While there are a few smaller lakes in the north, Lal-e Tanganyika is obviously the most important in terms of the country's fish potential. In recent years annual fish catch has averaged about 10,000 tons. UNDP-funded research (PPAM paras. 35-38) provided provisional estimates on the resource base with a potential annual yield of about 24,000 tons, but recent indica- tions suggest that there has been a decline in the abundance of ndagala.7/ Three types of fishing are practiced on the lake, according to 1983 data supplied by the Service des Peches: 6/ Rwanda, with almost 200/km2, is the first. 7/ A pelagic sardine-like fish.  -4- (a) industrial fishing (by about 19 purse seineg/ units); (b) artisanal fishing (nylon liftnets operated from 670 catazarans); (c) -coutumiare fishing (60 single canoes using scoopnets and beach seines). 10. There are two institutions linked to the fisheries subsector. The Department of Waters and Forestry of the Ministry of Agriculture9/ is offi- cially responsible for the fisheries subsector, but in fact concentrates mainly on data collection. Even with this emphasis, fish catch statistics are regarded as only indicative in terms of -catch quantities, but can show trends with reasonable accuracy. In addition, as part of a UNDP project, SUPOBU (SociEte 'Usine de Poissons de Burundi') was created in 1973 as a Regional Development Corporation (RDC)10/ to prowote fishing in Burundi through processing and marketing of fish, construction and sale of fishing craft and sale of fishing gear. SUPOBU was also responsible for implementing the IDA project (see also PCR paras. 2.05-2.11). II. PROJECT SUKMARY Background and Project Formulation (PCR paras. 3.01-3.32) 11. At completion of a preceding UWDP project, discussions started between Government and IDA for possible financing of a fisheries project. A 81 A purse seiner is a fishing boat using a net which, when set to surround a fish shoal, hangs vertically in the water and is then -pursed- by drawing the footrope and bottom of the net together. 91 See organigram of the fisheries administration in Burundi, PPAM, Table 10. 10/ Decree-Law No. 1180 of 1971 empowered the Minister of Agriculture to create para-statal Regional Development Corporations (RDCs) in cases where the Republic executes projects in the form of public enterprises within the agricultural sector, e.g. for land development, crop and animal husbandry improvements and the establishment of processing enterprises for agricultural products, etc. RDCs would have separate legal identity and authority to carry out commercial, industrial and financial activities in accordance with standard business practices. In addition to SUPOBU, other RDCs have been established, notably for coffee and tea development, and have proved reasonably successful, dealing as they do with products less vulnerable to rapid spoilage and operating within already well established and regulated marketing systems. The Third Technical Assistance Project (Credit 1456-BU) includes provisions to assist upgrading the managerial capacity of RDC personnel.  -5- one-person identification/preparation mission v.isited Burundi for about two weeks in Harch/April 1975. This mission concluded that fisheries development merited high priority. The project was appraised in September/October 1975 by two IDA staff and a consultant (for more details see PPA1 paras. 41-43). Negotiations took place in Washington during Narch 3-8, 1976 and the project was approved by the Board on April 20, 1976 under special procedure arrangements. 12. The main objectives of the project as approved were to raise the standard of living of artisanal fishermen along Burundi's Lake Tanganyika shore by increasing their fish production, and to market fish throughout the country, thus raising the nutritional levels, particularly protein intake, of rural consumers. In addition, the project aimed to develop a regional development plan for the lakeshore area and initiate pilot development activities there. 13. The project was to be funded by IDA (US$6.0 million), the Abu Dhabi Fund for Arab Economic Development-ADF (US$1.0 million), the Government (US$0.7 million equivalent) and the fishermen (US$0.9 million equivalent). SUPOBU was the implementing agency. 14. The project consisted of eight major components (see also PPAM, Annex 1 for a more detailed description of each component): (a) construction or expansion, equipping and operation of five fishing centersll/ and five fishing stationsl2/ and establishment and operation of hea 7---rters facilities; (b) construction of 400 fishing units and their sale, on credit, to fishermen as well as sale of fishing equipment and materials; ,c) equipping and operation of five catamaran purse seine units; (d) collection, transportation, storage and sarketing of fresh, smoked and dried fish and for that purpose the construction of a main depot and a freezer/cold store unit in Bujumbura, and four secondary depots, as well as the preparation of fish marketing studies and plans; (e) providing extension services and the constructlon of houses and offices; (f) staff training both in Burundi and overseas; llJ Fishing centers provided facilities for smoking, mechanical drying, sun drying racks, storage space for dried fish, bulk fuel storage and o.ffices. 12/ Fishing stations were similar to centers but without facilities for mechanical drying or smoking.  -6- (g) studies and trials concerning the fisheries subsector and the pollution of Lake Tanganyika; and (h) study and pilot activities of an integrated development program of the coastal project area. Project Implementation 15. Implementation experience with these eight project components varied. The first entailed construction or expansion, equipment and opera- tion of five fishing centers and five fishing stations and establishment and operation of headquarters facilities. Four centers and one station were constructed or expanded and the headquarters facilities were established. However, the fish drying kilns, probably the most important equipment in the centers, ceased operation within two years after installation because of lower than anticipated revenues in relation to operating costs, caused by shortage of raw material and low demand for the rather expensive end product. 16. The second project component called for the construction of fishing units and their sale, on credit, to fishermen as well as the sale of fishing equipment and materials. According to supervision records, about 200 wood plank boats, equivalent to 100 fishing units,13/ of traditional design were supplied to fishermen, but as they were mostly made with local softwoods for lack of other timber, they were not durable and only a small percentage of the credit has been reimbursed. With respect to improving artisanal fishing boats, after some unsuccessful attempts with fiberglass boats, a new type of laminated wood boat was introduced on an experimental basis which appeared to be suitable to the fishermen's needs. However, very few were built during the project period and even fewer were sold to fishermen because of lack of managemPnt expertise by SUPOBU, insufficient knowledge of market potential, high cost relative to the fishermen's purchasing power, lack of a functional credit scheme and absence of -salesmanship on the part of SUPOBU staff. Finally, fishing materials and equipment were supplied to fishermen by RDC, but for the most part this activity ceased as soon as the credit was closed and IDA disbursements were discontinued. 17. The third component called for equipping and operating five catama- ran purse seine units. One unit was inherited from an earlier UNDP project and a second was built with project funds. However, both encountered techni- cal and management problems resulting in substantial financial operating losses. At the suggestion of IDA, RDC has been trying to sell these boats, so far unsuccessfully; at the time of audit, one of the vessels was noted to have sunk, but the other was still operating. 13/ There is a large discrepancy between the numbers of fishing units supplied by SUPOBU, according to supervision records (100), and the numbers reported by SUPOBU Management to the audit mission (232 wood plank and 45 laminated fishing units) excluding boats sold to Rwanda and the present unsold stock of 47 laminated hulls (see Table 9 for further details). In the limited time available for audit it was not possible to verify either claim (see also PPAM paras. 28-29).  18. Storage, processing, and marketing of fish throughout the country were to be developed. Lower than projected fish quantities were processed by SUPOBU so that by 1979 processing facilities were operating at only 15Z of capacity. Since 50Z capacity utilization was the estimated requirement for breaking even, SUPOBU incurred substantial financial losses and processing was greatly curtailed. Currently, about 40-50 t annually are processed and marketed by RDC, primarily in Bujumbura. Marketing in the rural areas of the country was beset with problems including supply, distribution logistics, and high prices, and most commercial activity outside major urban centers was quickly abandoned. 19. Extension services were to be developed to serve about 3,000 fish- ermen in the project area. It is estimated that about 900 fishermen were reached although others probably benefited through informal sessions and discussions. 20. RDC staff were to be trained in Burundi and overseas in accounting, credit operations and other project-related matters. Although training did not reac the level originally envisaged, several staff members underwent training in Belgium in management and administration and one person in lami- nated wood boat-building techniques. Training in Burundi included some 15 candidates who attended an 8-month extension training course and other SUPOBU staff who received on-the--job training in boat building, engine maintenance and management. Unfortunately, many of the trainees left SUPOBU for employ- ment elsewhere after completion of the training. 21. Planned studies of the fisheries sub-sector, fish resources, and lake pollution were not realized and the project's regional development com- ponent was also not carried out. Project funds were used to finance a soils study by the Government agricultural research agency (ISABU), and for two studies by a local consulting firm on the development of the Rumonge and Nyanza Lac areas. These studies resulted in the creation by Government of new RDCs for Rumonge and Nyanza Lac, although their existence to date appears to be on paper only, and their proposed reiationship to SUPOBU remains obscure. Project Administration 22. As envisaged at appraisal, the Ministry of Agricullure and Livestock had general responsibility for implementing the project and coordi- nating project activities. Most of the project fisheries components were delegated to SUPOBU, but the Bureau Technique d'Etudes in the Presidency was responsible for coastal region development activities such as the studies referred to in para. 21 above and also assisted in project coordination. The National Development Bank (BNDE) was the channel for project funds and, at least initially, also assisted in financial control. Unfortunately, growing difficulties with project implementation resulted in BNDE's withdrawing alto- gether during the later part of the project period (for more details see BNDE comments, Attachment 1).  -8- 23. The preparation and appraisal missions correctly identified insti- tutional weakness in SUPOBU and questionable Government commitment as major sources of risk. Despite this forewarning, the project as planned proved too complex for the lidited managerial capacity in SUPOBU, and despite efforts to build-in safeguards by means of technical assistance, missions failed to per- suade Government that these problems were of a critical nature, as the out- come clearly shows. From start to finish the project suffered from local administrative and managerial ineffectiveness, if not outright Incompetence, and from Government's procrastination over the recruitment of the few key expatriate posts provided for in the agreement, which was also reflected in a hostile working environment in which people could achieve little. The cali- bre and suitability of some of the consultants were also suspect. In conse- quence, essential managerial decisions were frequently delayed or not acted upon and much time and effort was wasted on irrelevant or inconsequential matters. 24. IDA supervision was weaker than it should have been, given the mul- tiplicity of problems, and there was reluctance on the part of IDA to take appropriate action when repeated recommendations to SUPOBU or the Government were ignored (PCM paras. 4.41 and 6.17). Project performance should have been rated -3' by at least the third or fourth supervision mission, but this was not done until the final mission, by which time it was too late to do any good. Thus, IDA's own inertia compounded the consequences of local inade- quacy instead of generating corrective action. 255. The region agrees that IDA's supervisory role raises some important issues, and following review of the audit report they formulated these issues as follows: -(a) the need for very intensive start-up supervision to constitute a 'project launch' and work out the imple- mentation details which were hard to pin down before- hand. This is a practice in many cases now and the Project would have been a good candidate; (b) the need for more frequent missions during project implementation with representation of more disciplines in a sustained fashion. This would add up to a higher supervision input than current averages but could be justified by the need for outside support in circur- stances like those facing the project; (c) need for more effective action on management problems 'lessons' here are, in our view, twofold: specialized staff inputs from IDA to address management issues would have helped resolve problems; more important, however, would have been more effective IDA action in securing senior staff changes, always a tricky issue; and  -9- (d) the relatively low priority of this project was a factor and should be recognized. Although the project was problematic, it was not of critical importance to Burundi's agricultural development strategy.' While the audit generally agrees with these lessons, it falls to understand the logic that priority considerations somehow are an explanatory factor for weak supervision and troublesome project impimatation. Priority consid- erations are certainly valid in deciding if to proceed with a project, but once that stage is passed it is surely in the interest of the Bank and the Borrower to do everythiLg reasonable to ensure a successful outcome of approved projects. Impact and Sustainability 26. The project fell considerably short of meeting its objectives. It did, however, generate some positive impact for both major groups of intended beneficiaries, i.e., artisanal fisbermen and consumers. Benefits to artisa- nal fishermen included the supply of fuel, equipment and materials extension services, and training. Unfortunately, when the credit was closed most of these activities were severely curtailed due to SUPOBU's inability to con- tinue independently. After some false starts the project unit developed a new boat which could potentially meet the needs of local artisanal fisher- men. It is made mostly of local materials, is durable, and can be operated easily and economically since an engine is not necessary. Questions with respect to market potential, financial viability and credit scheme require- ments still need to be answered, but in terms of technical innovation this boat represented a promising start. Fish consumption in rural areas has remained low for lack of supplies, aad rural consumers still rely on mer- chants for their limited supply of fish, whereas SUPOBU was to have set up a rural marketing chain. However, drying racks were introduced at project centers and stations, and fishermen and merchants now use these to dry fish rather than drying on the ground; this provides rural consumers with a more hygienic product. 27. The disappointing performance of the project may be attributed to a number of factors, including the following: (a) Project design was rather complex in relation to the management and institutional capabilities available. The implementing agency, in particular, remained extremely weak throughout. (b) Limited IDA staff resources were committed to the project, espe- cially during identification, preparation and to a certain extent during supervision. (c) Events external to the project seriously affected its implementa- tion, including the aftermath of a civil war, a cholera epidemic and the repatriation of many Zairian fishermen operating until then in Burundi.  - 10 - (d) Technical assistance was inadequate or insufficient. Although suf- ficient funds were available, only about one third was utilized as SUPOBU chose to employ technical assistants only during the first half of the project period with the exception of one consultant who was present throughout. Unfortunately, there was dire need for technical assistance throegbout the project period, because of the general lack of qualified and experienced local staff. (e) The monitoring system was lacking. Although reporting requirements were communicated to SUPOBU before the credit became effective, these were not complied with. No annual work programs were pre- pared. An IDA supervision mission introduced some simplified reporting forms in the fourth year of the project which SUPOBU pre- pared regularly. However, this was perhaps too late and, in view of SUPOBU's weak accounting and financial systems, of questionable reliability. (f) The country's isolated location affected comunications and pro- curement. Identifying names and addresses of potential suppliers, exchanging correspondence, receiving bids, clearing them with IDA and placing an order was, at best, a very time-consuming process. Delivery of goods through Tanzania or Kenya-Uganda-Rwanda also pre- sented logistical difficulties, especially during the war between Tanzania and Uganda. 28. According to figures supplied by SUPOBU, from 1978 until April 1983, when the credit was finally terminated, SUPOBU built and distributed on credit, 464 wooden boats, 14 g.r.p. (fiberglass) catamarans and 90 laminated hulls, thereby contributing a total of 291 artisanal fishing units and asso- ciated fishing gear to a section of the fishing industry which only a few years earlier had been virtually destroyed by the civil war. As shown in PPAM, Table 9 these figures do not include 245 metal pirogues supplied by the earlier UNDP/FAD project, the rusting remains of which can be seen on most fish landing beaches14/ (see also PPAM para. 16). 14/ As noted in para. 16 there is considerable discrepancy between sets of figures regarding boat building by SUPOBU. The region agrees that .there has been some confusion on how many boats were actually built by SUPOBU due to incorrect use of the terms 'boats' vs. 'unit' and also due to sloppy record keeping by SUPOBU. However, there is obviously double counting when .UPOBU contends that 464 wooden boats were built (232 units), in addition to the 245 metal boats supplied by FA0. SUPOBU's records show 232 credit agreements for sales of fishing units. Since SUPOBU inherited most of the FAO boats and demanded repayment of the credit from their owners and since almost all SUPOBU constructed boats were sold on credit, it is not possible that 232 credit agreements cor- respond to 232 units plus the 245 boats inherited from FAO.-  - 11 - 29. Notwithstanding that very few of the credit purchasc agreements have been fuUly repaid-in fact only 7 out of 232 agreements for which records are available there is no doubt at all that this assistance came at a time when the artisanal fishery very badly needed help. Assuming that each unit employs 5 crew, nearly 1,500 fishermen plus their families (say some 7,500 people in total) have been provided with a means of regaining their liveLihood and independence who otherwise would by now have been in desperate straits. Ancillary workers such as fish traders, SUPOBU staff and boat repair workers, etc., will also bave benefited. 30. Processing facilities at the outlying fishing centers, which operated for only a short while, were rendered almost im-diately redundant for a reason which was not, and possibly could not have been foreseen prior to appraisal, namely the new and much easier marketing opportunities opened to fishermen by the reconstruction and tarmac surfacing of the Bujumbura- Rumonge-Nyanza Lac road. Fish can now be landed and delivered in fresh condition to Bujumbura Market within two hours or so, whereas previously the long journey over bad and sometimes impassable roads, from Rumonge and beyond, precluded fresh sales, except locally, and most of the catch had to be processed. 31. As regards sustainability, short of another national calamity, there seems no reason to suppose that the fishermen who have been assisted and rehabilitated will revert. Bujumbura is clearly not over-supplied with fish as yet, so their market outlets seem quite secure. Fish is also reported to be in extremely short supply at most of the inland urban and rural markets_5/ (but for more details see PPAM paras. 70-72). 32. SUPOBU was gravely damaged by externally and self-inflicted wounds and is down but by no means out. Staffing has been trimmed to five senior 15/ The region agrees that the new road has certainly facilitated transport of fish to Bujumbura. However, even before the new road existed, catches of industrial fishing units landed throughout the coastal area had to be transported to Bujumbura. The law prohibited their sale except in Bujumbura. The project's objective was to market fish in the rural areas and construction of the new road to Bujumbura would have had a very small impact on this operation. The fish preservation program was based on the desirability of supplying fish to more remote rural areas and to compensate for the characteristic monthly and seasonal ups and downs in fish production. In practice, the situation was different from that anticipated at appraisal because reduction in fish catches increased the importance of the Bujumbura market for project output (and hence demand for fresh fish) during the project period. It was this more than transport factors that forced the smoking/drying activities to be curtailed and ultimately stopped." Whatever the case, the audit nevertheless considers that this effect should have been foreseen earlier, in time to modify or defer some of the expenditure on redundant fish processing facilities in the stations/centers.  - 12 - and eighteen junior personnel (most of the juniors are watchmen to safeguard outlying property) and a new Director took over in April 1984, who has clearly done a lot to revitalize such activities as SUPOBU can reasonably still continue. 33. Since the project ended in April 1983, SUPOBU has built a further 63 laminated hulls. Credit sales were stopped as of April 1984, despite the existence of a large stock (presently 47) of unsold boats. Nevertheless, a few boats have been sold for cash to Lake Tanganyika fishermen and an order is about to be dispatched for 20 of the hulls for an FAO project on Lake Kivu. SUPOBU's boat builder argues that his facility is capable of producing 4 laminated hulls per week, provided that the necessary market demand can be created (see elsewhere in the PPA& for more details about prices and cost-effectiveness). An information and advertising campaign could probably generate a number of cash sales, but it is clear that such a scale of produc- tion cannot be realized unless a new and better organized credit scheme can be established. 34. SUPOBU is still a long way from being the cost-effective and prof- itable enterprise that was envisaged at the start of the project, but the Government is clearly determined to persist with this parastatal approach, and appears to accept that much of the money lent to and through SUPOBU willI never be repaid. Nevertheless, the outstanding credits to fishermen have not been abandoned and renewed efforts are being made to recoup as much as possi- ble, using, in addition to SUPOBU's limited resources, the Government's district administrative services to trace and assist the borrowers to repay their debt. 35. Thus, although the project as a whole either failed or fell seri- ously short of attaining its various over-optimistic objectives, and was clearly defective in terms of the very limited IDA effort devoted to its formulation and supervision, especially in regard to fisheries expertise, it has imparted some significant benefits to the fishing industry in Burundi and some of the facilities created by the project are still in operation, albeit at low levels of utilization. In particular, SUPOBU retains the capacity to build durable artisanal fishing boats which have proved their worth in use and which are also finding export market outlets. Hany fishermen have been assisted and, after all the vicissitudes of past civil war and epidemics, the total catch from the lake has stabilized at around 10,000-12,000 at per annum and may be expected to increase in due course. Even the redundant fishing centers may find a future use if current discussions succeed in the establishment of viable fishermen's cooperatives, to which ownership of the facilities can be transferred. It would not be the first instance of an apparently disastrous experience leading to a much more favorable outcome, e.g., the First Fisheries Project in Indonesia, and it may be worth taking a fresh look at the Burundi situation in a few years time.  - 13 - 1rI. PROJECT ISSUES A. Fisheries Resource Research on Lake Tanganyika 36. A three-year, UNDP-funded FAO fisheries research preject on Lake Tanganyika was extended in early 1974 for a further three years and expanded to provide a substantial development component, in response to the need for rehabilitation of the non-industrial fishing sector which was severely damaged during the 1972 civil conflict. At a total cost of US$890,000, (UNDP contributed US$708,500), the development component provided fishing boats and gear, three fishing centers-including basic fish processing facilities, training and a marketing study, etc. The development component ended in late 1976 having achieved most of its planned targets thanks to an enthusiastic team of local and expatriate staff led by a particularly able project manager.16/ The IDA project was intended to build on and expand this experience. 37. The research component, for which UNDP contributed US$150,000, con- tinued to concentrate on studies to quantify the biomass and to determine its relative composition and dist ribution. Investigations were restricted to Burundi waters, which comprize only 7Z of the total lake area; but regular contact was maintained with a parallel UNDP research project covering Tanza- nia's 41X of the lake, and based at Kigoma. Both projects benefited from the sharing of research findings and data, etc. Principal conclusions,17/ as at the end of 1976 when both projects ended, were that in general the pelagic fish stocks were underexploited and the fishery was therefore capable of expansion. The position was complicated, however, by the discovery of cyclic variations in abundance of ndagala (Stolothrissa and Linothrissa spp) and .-ukeke (Laes and Luciolates spp),L8/ with a periodicity of 6-8 years. At the present stage it would appear that the predatory mukeke are dominant, so causing a decline in the stocks of ndagala which can be expected to continue to a point where the prey stocks become insufficient to support the predator population, which will in turn collapse. Rapid recovery of ndagala can then 161 FAO Terminal Report Fl/DPJBDl/73/007; Burundi Fisheries Development in Lake Tanganyika, Project Conclusions and Recommendatioas, Rome, 1977 (see also PCR para. 3.02). 17/ See: FAO Terminal Report FI:DP/BDl/73/020; Burundi, Fisheries Research on Lake Tanganyika; Project Conclusions and Recommendations, Rome, 1977; and FAO Technical Report Fl:DP/URT/7l/012, Tanzania, Lake Tanganyika Fishery Research and Development Project; Fishery Biology and Stock Assessment, Rome, 1978. 18/ Ndagala is a pelagic sardine-like fish of greatest importance Co the industrial and artisanal fishery. Mukeke refers to four species of predatory perch-like fish, which prey on ndagala and are also of commercial importance.  - 14 - be expected because of its fecundity, fast growth rate and short life cycle (up to 19 months) which contrasts with the much longer life cycles of the predator species. Nevertheless, by the end of 1976 a great deal of research remained to be done.19/ 38. The several years of UNDP-funded fisheries research does not appear to have had any lasting institution building effect, since research effort virtually ceased with the ending of the UNDP project. The Service des Peches still possesses basic laboratory facilities and some equipment and its staff also includes a few people having appropriate scientific qualifications, but they are mostly engaged in other duties. The future viability of the Lake Tanganyika fishery will depend to a great extent on effective monitoring of changes in the resource base and its environment, and it is therefore unfortunate that IDA did not insist on the implementation of proposed pollution studies and their effects on fish, since this might have concenr- trated more attention on the importance of continued research effort. B. Project Processing and Start-up 39. In summary, the actual processing timetable (more details in Table 3) reads as follows: Identification/Preparation 03/75 Appraisal 09/24 - 10/10/75 Negotiations 03/03 - 08176 Board Approval 04/20/76 Signing 06/11/76 Effectiveness Date 02/22/77 Closing Date 06/30/8220/ 40. As the timetable indicates, project processing in the Bank was rather quick, but this apparent bureaucratic efficiency was, in the audit's view, at the expense of the quality of project preparation and design (PPAR paras. 41-47). Furthermore, delays started almost immediately, beginning with a delv in declaring the project effective (PPAR paras. 48-52). C. Project Formulation 41. Discussions first started in January 1974 between the Government and IDA for a project to continue and expand on the UNDP/FAO development project after its completion at the end of 1976. An identification/prepara- tion mission was undertaken by an IDA economist who visited Burundi for two 19/ Proposals were drawn up during 1978/80 for a regional project in which all four riparian states (Burundi, Tanzania, Zaire and Zambia) could participate. Unfortunately, this important project still awaits a donor to finance its implementation. 20/ Final Disbursements were made 05/16/83.  - 15 - weeks during 4archlApril 1975. A project outline was drafted which repre- sented a compromise between some idealistic and elaborate proposals by Governaent and a much more cautious and phased approach advanced by UNDP project staff (PCR paras. 3.04-3.10). The mission raised three major issues: (a) the political and social sensitivity of the project area, which was still suffering from the aftermath of the L972 civil war; (b) institutional weaknesses, shortage of trained manpower and ques- tionable Government commitment at the implementation level; and (c) the need to ensure that project benefits accrued to artisanal fish- ermen rather than to other, less needy segments of the population. 42. In addition, the mission cited pollution of the lake as a longer- term problem requiring study and also the need to establish a system of regional cooperation between the four riparian states borderiag Lake Tanganyika, on such matters as fishing rights, research programs and environ- mental control. There was insufficient time for detailed analysis of any of the risks involved and, as noted in para. 3.10 of the PCR, it appears that the preparation mission was too optlmistic in assuming that the various con- straints to the project could be overcome. The mission was also too short in terms of time and professional expertise, and was perhaps misled by the successful record of the UNDP project without recognizing how much of that success depended on the efforts of the particular project manager. 43. The project was appraised during a two-week visit to Burundi in September/October 1975 by two IDA staff members aad a fisheries consultant, timing of the appraisal having been largely dictated by the approaching end of the UNDP project. As noted in the PCR (para. 3.11), the project remained basically unchanged following appraisal from what was proposed in the prepa- ration report. The issues paper repeated two of the general issues raised during preparation namely, potential difficulties arising from institutional weaknesses and shortages of trained manpower, and second, the need for con- tinuing vigilance to ensure that the primary beneficiaries would be the small-scale fishermen. In addition, seven project-related issues were cited: (a) the proposed interest rates for sub-borrowers, disputed by Government; (b) the number of technical assistance posts, which Government wished to reduce; (c) the need for effective market planning during implementation; (d) procurement procedures which departed from standard international competitive bidding practice; (e) audit procedures to be undertaken by BRD (the central bank) rather than BNDE (the national development bank);  - 16 - (f) sales of fishing gear which could generate cash flow problems f or SUPOBU, unless they quaLified for prompt IDA disbursement; and (g) pollution of the lake, for which the mission proposed additional research and Bank support for any resultant practical remedial action. 44. A decision meeting held in October 1975 concurred with the appraisal mission proposals, agreeing that the project manager should be a Burundi national and that the appointment of qualified counterparts be made a condition of effectiveness. 45. The PCR (para. 3.15) notes that while all these issues were rele- vant, a further three could (and perhaps should) have been added: (a) project complexity, involving a widely differing range of activities; (b) the difficulty of designing and implementing credit schemes for artisanal fisheries; and 5c) the inadequacy of information on the resource base. En retrospect, the key issues proved to be project complexity and the insti- tutional weakness of SUPOBU and other entities connected with the project. Patrly because of the limited staff time devoted to identification and prepa- ration, the mission underestimated the impact of these two issues, and did not give sufficient consideration to their resolution. In this regard, how- ever, the appraisal mission, which did have more staff time, also failed to adopt a sufficiently critical approach and so the last opportunity to rectify the omission was lost. 46. In the audit's view, IDA should never have sent a preparation mis- sion unaccompanied by any fisheries experts, especially in the case of a first fisheries project dealing with an almost totally unknown situation. At best this should have been regarded as a reconnaissance or identification mission only. Having identified a situation where the local fishing industry had been largely destroyed, where organizational and managerial expertise for a fisheries project was virtually non-existent (UNDP said so and the fact was acknowledged in the preparation report as a major issue), and where Govern- ment appeared to want an unrealistically complex and multiple-component proj- ect, it seems in the audit's view premature to have recolmmended such an early appraisas, notwithstanding the forthcoming end of the UNDP project. A more rational approach would have been to recoacend a series of short- erm prelim- inary studies to cover inter alia: (a) a review of the resource situation and state of knowledge; (b) a study of fishermen's needs in terms of numbers and types of boats and fishing gear, building on UNDP experience;  - 7 - (c) a study to design an appropriate credit scheme for artisanal fish- ermen using locally available expertise, e.g., BNDE; and (d) a study to determine the most appropriate organizational framework to promote fisheries development, reviewing direct governmental, parastatal, commercial and cooperative approaches as alternatives or in combination. This would have undoubtedly delayed matters, but would have been better, in the audit's view, than risk the disaster that in fact happened. 47. Regretfully, this is another case of a fisheries project where the audit record indicates that project forsulation is too often undertaken by missions containing insufficient expertise, working to timetables that some- times allow insufficient time to do the job properly. 21/ D. Signing and Effectiveness 48. The Board approved the project on April 20, 1976221 (special pro- cedure). The Credit and Project Agreements were signed on June 11, 1976, the delay mainly caused in obtaining BNDE authorization for the embassy personnel to sign the documents on their behalf. 49. Because of onlending and cofinancing arrangements, there were a subscantial number of conditions of effectiveness:23/ (a) execution and delivery of the Project Agreement on behalf of BNDE and SUPOBU; (b) execution of a subsidiary loan agreement between Government and BNDE; (c) execution of a subsidiary loan agreement between BNDE and SUPOBU; (d) signing and effectiveness24/ of the loan agreement with ADF; 21/ Harvesting the Waters, a Reviev of Bank Experience with Fisheries Development, OED Report No. 4984 dated March 13, 1984 (para. 3.19). 22/ The appointment of the General Manager of SUPOBU was a condition of Board presentation, but had been fulfilled as the General Manager, whose qualifications were satisfactory, actually participated in the negotiations. 23/ In addition to execution and delivery of the Credit Agreement as specified in Article 12.01 of the general conditions. 24/ Conditions precedent to the effectiveness of that agreement, except for the effectiveness of the IDA agreement, needed to v'e fulfilled.  - 18 - (e) establisbment of a special account for SUPOBU and an initial deposit of $100,000; and (f) appointment of the expatriate administration and finance officer and his counterpart./5/ 50. The date by which the Credit Agreement was to become effective was originally set at September 9, 1976 (3 months after signing) but was extended four times 26/ ultimately to February 28, 1977. The ADF Loan Agreement was declared effective on February 18, 1977 and the IDA Credit Agreement on February 22, 1977. 51. Conditions of effectiveness were discussed with the Burundi delega- tion during negotiations and confirmed in a formal letter of the Legal Department dated August 8, 1976. In addition to the first supervision mis- sion in October 1977, a Bank lawyer visited Burundi (October 14-17) to review effectiveness matters. It took a long time before many of the conditions were met, but the most important delays were caused by the recruitment and appointment of the (expatriate) administration and finance officer and also by the initial deposit in the special SUPOBU account. 52. The retrospective question is whether these delays could have been avoided. An answer, however, is only relevant in a wider context. The audit, therefore, has reviewed experience with signing and effectiveness of all projects in Burundi approved until end FY84 (see Table 4 for details).27/ 53. On average, it took two months between Board approval and signing.28! As of August 30, 1984, the two latest29/ out of 23 projects had not yet become effective. For the remainder it took an average five months 25/ An earlier suggested condition of effectiveness--appointment of the fisheries officer-was deleted during negotiations as UNDP project officials were still in the country. 26/ First to November 15, 1976, then to January 6, 1976, then to January 31, 1977, and finally to February 28, 1976. 27/ Similar reviews of country experience were made in PPAR, Ethiopia - Coffee Processing Project (Credit 290-ET), OED Report No. 3976 dated June 21, 1982; and PPAR, Ivory Coast - Second Cocoa Project (Loan 1069-IVC) OED Report No. 5190 dated June 29, 1984. 28/ In cases where substantial delays occurred (Third Highway, Credit 1132 and Kirimiro Rural Development, Credit 1165), this was due to lack of IDA commitment authority. 29/ Ruzizi II Regional Hydroelectric Project (Credit 1419), approved 12/06/83 and signed 04/04/84; and Third Technical Assistance Project (Credit 1456) approved 04/17/84 and signed 07/17/84.  - 19 - after signing before credits became effective or an average excess of almost two months over the time interval allowed in the agreements.30/ The record also shows a slight upward trend in the time needed to make credits effective, but more importantly the record shows important differences between sectors: on average, agricultural credits took 4.2 months in excess of the time envisaged in the agreements to become effective, against just under one moath excess for the other sectors combined. As Table 4 shows, agricultural credits are among those with the heaviest effectiveness conditionality. But a review of files also indicates that in some of the other sectors (such as education), conditionality is brought forward (i.e., senior staff appointments before Board presentation). It appears that cross- fertilization between sectors with regard to application of effectiveness conditionality could be beneficial. 54. The effectiveness experiences reviewed here compare favorably with earlier OED-reviewed experiences.31/ Nevertheless, the conclusions of these earlier reviews as well as another OED study32/ on the subject still remain valid, namely: (a) the importance of a realistic timetable for.meeting conditions of effectiveness, included in the project start-up plan as discussed during negotiations and reflected in the agreed minutes; (b) setting dates of effectiveness realistically as required in any case by Bank guidelines;33/ and (c) codification of Bank experience and guidelines regarding effective- ness and start-up in a separate OMS. 30/ Only a few (5 out of 21) credits became effective within the time frame envisaged in the agreements. 31/ Burundi's two-month average excess over what was provided in the agree- ments compares to Ethiopia's 2.9 months (OED Report No. 3976 op. cit.) and Ivory Coast's 5.1 months (OED Report No. 5190 op. cit.) 32/ OED Report No. 813 dated July 22, 1975, Management Policy Review: Delays in Effectiveness. 33/ OMS 3.30, para. 4 (Disbursements) - "In sum, conditions of effectiveness should not be viewed as isolated events but should be considered integral parts of the planned sequence of events agreed well in advance with the Borrower/Guarantor"; and para. 6, "It is therefore necessary in each case to mnake a realistic estimate of the time necessary to fulfill the conditions of effectiveness and to specify an appropriate date on this basis. In aost cases, this can best be done by drawing up at an earlier stage an agreed start-up plan showing all steps to be taken by each party and making .i realistic estimate for each step."  - 20 - Given that the last report on the subject34/ was done almost 10 years ago, OED might wish to consider a new review of project start-up experience. E. Disbursements 55. The disbursement schedule (PCR para. 7.03) included a category to cover 9O0 of total operating costs for SUPOBU fisheries extension, fisherman training, local training and fishing trials. The amount allocated to this category was US$100,000. In fact, more than US$400,000 was disbursed under this category.35/ 56. By the end of the disbursement period regional project staff dis- covered that IDA had been disbursing for SUPOBU salaries beyond what was specified in the disbursement schedule. Apparently what happened, according to information in the operational files, was that a few years earlier IDA had agreed, on a temporary basis, to fund SUPOBU's monthly payroll as SUPOBU was at that time extremely short of cash. However, once agreed, this practice continued and it appears that disbursements for the SUPOBU payroll had been made under Category 3, but mostly under Category 2 (civil works),36/ over the period September 1977 to April 1983 for an amount which has been tenta- tively estimated in the US$600,000-800,000 range. Approximately 15-20Z of total disbursements went for salaries, most of which were not justifiable under the terms of the Credit Agreement nor the intent of the project. Proj- ects staff felt that lack of monitoring and coordination with the Loan Admin- istration Department (LOA) was at least partly to blame for this. LOA staff countered that the use of operating expenditure categories in disbursement schedules does not allow much control because of lack of definition and that the use of statements of expenditures would prevent effective control in any case. 34/ OED Report No. 813 dated July 22, 1975. 35/ LOA has provided additional details about procedures regarding category overdraft authorization as follows: -The first withdrawal application (US$108,611.05 equivalent) submitted against Category 3 (operating expenses) involved an overdraft against the original allocation. In order to process this application, Disbursement Division staff contacted the Project Officer to request authorization to overdraw the category. At that time, our procedures did not require that we obtain written authorization, but it was our practice to notify Projects staff that we would continue to process applications against the overdrawn category unless we received their instructions to the contrary. We have been unable to locate any documents indicating that we should discontinue process in applications against this category. (In 1981, our procedures were amended to require that we obtain written approval in order to exceed category allocations; we also request an indication of the extent to which we may overdraw the category concerned pending formal realloca- tion of funds, and enquire about the likely date of this reallocation. These changes should prevent any recurrence of a problem such as this.)" 36/ The amount allocated to this category was US$300,000; in fact more than US$500,000 was disbursed under this category.  - 21 - 57. Given this situation, the audit investigated further but, within the PPAR time frame, a full review of all records was not feasible. Never- theless, on the basis of records reviewed, some interesting conclusions can be drawn which might be helpful for the Bank's ongoing efforts to increase disbursement efficiency and speed up resource transfer (see PPAM para. 59). A few points are worth mentioning first: (a) About US$4 million dollars was disbursed under this project. A total of 337 withdrawal authorizations were prepared by LOA to cover 476 disburseme-nt requests prepared by the Borrower. Many disbursement requests were for amounts less than a few thousand dollars.37/ (b) As a full reconstruction of disbursement records was not possible, the audit suggests that the internal audit department be requested to undertake a specific review of this (and other) projects not so much from the accounting perspective38/ but to review the adequacy of administrative procedures. 58. The records show in fact that a large amount was disbursed for salaries, initially under Category 3 and later under Category 2.39/ From 1980 onvards statements of expenditures were received frequently (a] st monthly). These were clearly marked as both staff salaries for parts A(i), A(ii), A(iii), D(i) and E(i)40/ and for disbursement Category 2. Supporting documents included the SUPOBU payroll marked to indicate which staff were related to these project items. While it is true that salaries can be covered under civil works (force account) ,41, the regular occurrence of these requests related to specific parts of the project description should have given rise to questioning of the validity of these requests earlier. However, the legal documents were also somewhat confusing. While the project's financing plan is clearly limited to investments, except for a 37/ In fact, the lowest withdrawal Authorizations (W/A) were for the equivalent of US$13.64 (W/A 288), US$15.19 (W/A 277) and US$17.41 (W/A 201); 45 W/As were each for the equivalent of less than US$1,000. 38/ Both internal and external auditors review disbursement records as a matter of course during their account auditing. 39/ Category 2 covered civil works under parts A through E (see Annex 1). 40/ See Annex 1. 41/ LOA indicated that -as the applications clearly specified Category 2 (Civil Works), our staff assumed that these payments were for force account works rather than operating costs.  -22- small amount to cover SUPOBU's fisheries extension service, the legal project description42/ referred to construction, equipping and operation:43J of fishing stations, fishing centers and the SUPOBU headquarters. 59. While various factors convene to allow for plausible explanations of this situation, a few conclusions useful for wider application can never- theless be drawn. First, aside from the debatable question of whether the Bank should use credit funds to relieve cash flow constraints of executing agencies, contrary to what was envisaged in the credit documents, obviously such decision-making should be formaUy recorded. It is precisely the absence of such records which, in the audit's view, led to continued dis- bursements for something which was apparently intended to be only temporary. Second, it seems that coordiration between the departments concerned regard- ing disbursement approval can be improved. Extensive disbursement files are available in LOA and project staff should be encouraged to consult those from time to time. Monthly reports on disbursement status are being prepared, but it would be helpful if these were distributed to the project officers con- cerned (and also the project executing agencies) as a matter of course. Third, desk review in LOA of disbursement requests can obviously only be limited.44/ Nevertheless, a little vigilance should have identified a potential problem. Fourth, the use of statements of expenditure requires supervision to look into the validity of such requests by reviewing the supporting documentation in the field. In fact, this was not done and ORD experience indicates that it is often the case. Supervision missions are simply not equipped nor allowed the time to do this work. The Bank should be more realistic in its assumptions regarding what is administratively feasible in such cases. Fifth, while the increased use of statements of expendi- ture45/ has increased administrative efficiency in handling disbursements, it has weakened internal control. Measures should be taken to improve inter- nal control both in the Bank and in the executing agencies. Six-h, in this context, LOA's point that -operating expenditure- categories are especially difficult to control because of lack of definition is well taken. 60. In the meantime, it should be noted that the Bank has undertaken an indepth review of its disbursement procedures (the latest review before that had been undertaken in 1966). Two task forces, one for the review of dis- bursement policies and one on disbursement procedures and systems were con- stituted. These task forces have concluded their reviews and issued their 42/ Schedule 2 of the Credit Agreement (see Annex 1). 43/ Audit emphasis. 44/ As the audit discovered, it is clearly impossible to determine the justification of many items in statements of expenses in relation to the project such as postage, fuel for vehicles, social security, dental care, etc. 45/ And more recently revolving funds.  - 23 - reports and recommendations. Among the recommendations are increased respon- sibility for expenditure control with the borrowers, more delegation of approval authority to the Bank missions in the field, increased use of state- ments of expenditure and revolving funds but with adequate internal and external control at the executing agency level, and higher thresholds for individual statement of expenditure disbursement requests. Following further review, these recommendations will be incorporated in revised OMS and other guidelines for withdrawal of proceeds of IBRD Loans and IDA Credits. 61. This is also another example of disbursements occurring after the closing date, a topic covered in earlier audit reports.46/ The closing date was set at June 30, l982.47/ The PCR (paras. 5.01-5.15) describes exten- sively the Bank's decision-making process in relation to Government's request to extep' the closing date from June 30, 1982 to June 30, 1983. In summary, the Bauk judged SUPOBU's program of activities for 1982/83 as unsatisfactory and would consider extension only provided certain Bank recommendations and modifications were implemented. Despite numerous interactions, divergencies of opinion remained, and on March 9, 1983, IDA informed Government of its decision not to extend the closing date and to cancel the remaining credit balance not later than April 30, 1983. The latest disbursement was made on May 17, 1983, for an application which was received in the Bank on April 17, 1983, the remaining balance of just over US$2.0 million was cancelled with effect from April 30, 1983 and the Bank's loan account was closed.481 62. While the closing date was not formally extended, the Bank contirr- ued to disburse against project expenditures. These disbursements did not only cover payments for normally delayed disbursement requests, but also apparently for expenditures incurred after the formal closing date. While the former was acceptable (there are always delays in processing disbursement 46/ PPAR, Ethiopia - Addis Ababa Dairy Development Project (Credit 269-ET), OED Report No. 4001 dated June 23, 1982; PPAR, Ethiopia - Second Livestock Development (Credit 269-ET), OED Report No. 3977 dated June 21, 1982; and PPAR, Ghana - Livestock Development Project (Credit 500-GH), OED Report No. 3972 dated June 16, 1982. These are three audits in which the topic of disbursements after the closing date has been discussed as a specific issue. This does not warrant the conclur sion that this is a problem in only a few cases; in fact, disbursements after the closing date are much more frequent. 47/ "The Closing date shall be June 30, 1982 or such other date as the Association shall establish. The Association shall promptly (audit emphasis) notify the Borrower of such later date." (Development Credit Agreement Section 2.04). 48/ However, cancellation and closing were not mentioned in any of the monthly reports to the Executive Directors on Bank and IDA operations.  - 24 - requests), the latter was not (at the time Bank practice49/ was that the closing date was the cut-off date for eligibility of the project expendi- tures;50/ however, more recently eligibility rules were released to allow disbursements for expenditures after the Closing Date5ll). 63. The argument is not a legal one, as the closing date is defined as -the date specified In the Development Credit Agreement as of which the Association may by notice to the Borrower terminate the right of the Borrower to withdraw from the credit account any amounts therefore unwithdrawn.'52/ Neither does the audit take issue with the Bank's decision-making process (the Bank after due consideration simply decided that continued support for the project was no longer warranted). At issue here is simply consistency and clarity in the administrative application of Bank decision-making on this point. A few points are worth making in this context. First, after the legal closing date, disbursements are at the discretion of the Bank. As a consequence, borrowers would be uncertain as to the Bank's further commitment to the project. Second, changes in the legal closing date have to be reported to the Executive Directors. There appears to be somewhat of a bureaucratic reluctance to do so because these extensions are negatively per- ceived. Third, while from the operational point of view flexibility is desirable, from the bookkeeping standpoint more mechanical treatment is preferable. 64. It might well be that bookkeeping and legal requirements have become too much intertwined. On one hand, there is the legally defined closing date in the Credit Agreement, linked to estimated project completion date. On the other hand, there is the actual closing of the accounts in the 49/ For more details on -informal- and -formal- guidelines on disbursements after the Closing Date, see the PPAR on the Ethiopia Addis Ababa Dairy Development Project, op. cit. 50/ -It is the Bank's practice to process all withdrawal applications to cover expenditures incurred prior to the closing date-, Senior Vice President's memo of December 1982. 51/ -As stated in para. 9 in my 1982 memo, disbursements may continue to be made for a period of maximum 6 months after the Closing Date without formal postponement thereof. In such case, the Borrower should be advised of the period within which all withdrawal requests must be received by the Bank. Although in most instances disbursements during this period would be for expenditures incurred prior to the Closing Date, withdrawal requests for eligible project expenditures Incurred after the Closing Date should also be lowered, provided, of course, that such requests are received within the period specified to the Borrower- SVP memo of January 1984. 52/ General Conditions Applicable to Development Credit Agreements (Article 2.01, 2.12).  - 25 - books of the Bank. As OED Annual Reviews of Project Performance Audit Results show, project implrmentation intervals are invariably longer than estimated at appraisal. In the case of this project, the de Sure closing date remained June 30, 1982; the de facto closing date was May 17, 1983. There are many good reasons why proiject accounts should be kept open after project completion,53/ although administrative efficiency and resource transfer objectives would require expediency in this matter. Nevertheless, it seems that more clarity between de ure and de facto closing might be helpful.54/ Given the fact that present practice seems open to divergent interpretation by staff and might appear arbitrary to borrowers, the audit suggests that the Internal Audit Department be requested to study these administrative matters in more detail. 65. It was also noted (PCR para. 6.20) that there were substantial delays in disbursements during the early period after project start-up, which created serious difficulties in SUPOBU's liquidity position. These difficul- ties were resolved satisfactorily after IDA staff from the Disbursement Division visited Burundi and explained IDA procedures. Identical situations have been observed in other cases during OED audits, e.g., Panama First Fisheries.55/ It would seem that the explanations of IDA procedures which are given during negotiations and the instruction manuals concerned are either not adequate or, more likely, are not given to the people who are ultimately responsible for preparing disbursement requests. 66. Such delays are clearly avoidable, and with this in mind it is recommended that in all cases, except those involving follow-up or subsequent projects to the same executing agency, a Disbursement Division representative should accompany the earliest supervision mission possible as a matter of course. This would ensure that those concerned with project implementation were fully conversant with IDA requirements and also that the Disbursement Division was properly briefed on any unusual features of the local sys-ems for financial management.56/ 53/ Retention payments, for example, cover relatively small amounts outstanding for an extended period after physical completion. A possible alternative in such cases, which would speed up disbursements, might be the use of Bank guarantees. 541 A possibility might be to define the closing date more as a completion date after which project expenditures would not be ineligible for disbursement under the loan/credit. 55/ PPAR, Panama - (First) Fisheries Project (Loan 784-PAN), OED Report No. 2578 dated June 29, 1979. 56/ See also PPAR, Maldives Fisheries Project, OED Report No. 5587 dated April 5, 1985.  -26- F. Fisheries Statistics in Burundi 67. Any country desirous of developing its fishing industry must have a system for the collection and analysis of catch and fishing effort data capable of yielding reasonably reliable information on catches, catch rates and seasonal or cyclic changes and trends in the fishery. This would provide a measure of the impact of development activities, serve as a basic input to its research effort and provide the basis for a system of resource management and regulation of the fishery. In para. 2.07, the PCB casts some doubt on the accuracy of data collected by the Service des Peches in Burundi, despite the high proportion of departmental staff (18 of 25) who have this responsibility. 68. Even though there was a limited amount of time afforded the audit to evaluate this data collection system, it was concluded that the system deserved more favorable co=ment than that provided in the PCR. Data collec- tion for the industrial fishing sector is based on a total count (by weight and fishing trip) of all catches delivered to Bujumbura.57/ The widely dispersed nature of the artisanal and coutumiare- sectors makes it a practi- cal impossibility to attempt a total count and instead the system depends on twice yearly total counts of all fishing vessels and fishermen, and on a sample count by seven enumerators of catches at various fish landings. Appropriate weights are then applied to convert the sample data into overall monthly and annual catch figures as well as catch per unit effort figures. 69. The system is based on and closely follows FAO recommendations for a frame survey approach to artisanal fish catch data collection. While no sampling system can be lOOX accurate - no doubt this one could also be further refined and improved - it does, in the audit's view, compare very favorably with systems used by many other countries. In particular, it is capable of showing ttends in the fishery with a very fair degree of reliabil- ity. Summary catch data for the years 1970-83 and a more detailed analysis of the industrial, artisanal and coutumiere sections of the fishery for 1981, 1982 and 1983 are provided in Tables 5, 6, 7 and 8. G. Impact on the Project by Other Bank-Sponsored Activities 70. As noted in para. 30, a number of fishing centers were established along the lakeshore where fish could be hygienically processed by drying and smoking, in order to ease fishermen's marketing problems. These facilities were rendered almost immediately redundant following the reconstruction and tarmac surfacing of the Bujumbura-Rumonge-Nyanza Lac road (see attached map), which closely parallels the lakeshore. By means of the new road, fishermen and traders were enabled to deliver catches in fresh condition to Bujumbura Market within two hours or so after landing, whereas previously the long journey over bad and sometimes impassable roads meant that most of the catch had to be processed. 57/ It is a legal requirement that all industrial vessels must deliver all their catches to Bujumbura for first sale.  - 27 - 71. Detailed engineering studies for reconstruction of the lakeshore road were financed under IDA Credit S-il-BU during 197001 and further updated under the First Highway Project, Credit 467-BU which was approved in February 1974 and which led to Kuwait Fund and African Development Bank fund- ing of the reconstruction work. Road works were in progress during 1976 when the Fisheries Project was under negotiation. The appraisal mission (SAR, para. 2.25 and elsewhere) noted the road construction plan but does not appear to have considered its probable effect on fish marketing patterns. Common sense should indicate that fishermen would prefer to sell their catches as fresh fish, if at all possible, rather than incur tht' additional labor and cost of drying and smoking the fish. Thus, the plans for fishing center development and fish processing facilities might have been modified or phased in at a much later date.58/ 72. This experience points to the need for closer attention to parallel or cross-sectoral plans for development by the Bank and other agencies during the course of project preparation, so as to take full advantage of such developments and avoid the risk of unnecessary adverse impacts and waste of scarce financial resources. It is understood that the third Technical Assis- tance Project, Credit 1456-BU contains provision for training and upgrading the capacity of RDC staff and management. It is hoped that this work will also benefit SUPOBU. 58/ The Region does not agree with this assessment. As pointed out in PPAM para. 31, footnote 15, they felt that (i) the appraisal did take account of the road construction plans; and (ii) factors other than the road forced curtailment of smoking/drying activities.  - - ^ : :t - f '- :: -: -E :- :; - : : w - 8t- : -- -: - - - - r s z S.r- :; ' :- . .ss-: ' ' < '    '- = - X - f  ' ff  Table 1 29 - PROJECT PERFORMANCE AUDIT MORANDUM BURUNDI FISHERIES DEVELOPMENT PROJECT (CREDIT 626-BU) BURUNDI - LENDINGS OPERATIONS Abbreviated Loan/Credit Approval Disbursement/ Proje-t Name Number Aount Date Audit Status Port and Highway Ln 165 /a 4.80 (4.80)/b June 1957 FD & repaid Bujumbura Water Supply Cr 85 1.32 ( 1.29) Mar. 1966 FD (792) /c Coffee Improvement Cr 147 1.93 ( 1.93) Apr. 1969 FD (1531) Highway Engineering Cr SO0l 0.38 ( 0.08) June 1970 FD (4582) Highway Maintenance Cr 467 5.07 ( 5.07) Feb. 1974 FD (4582) Seco_d Coffee Improvement Cr 593 5.20 ( 5.20) Nov. 1975 FD (4576) Technical Assistance Cr 613 1.50 ( 1.50) Feb. 1976 FD /d Fisheries Cr 626 6.00 ( 3.96) Apr. 1976 FD 7 Education Cr 679 10.00 (10.00) Feb. 1977 FD T7- BNDE Cr 731 3.40 ( 3.09) July 1977 - Highway II Cr 773 14.00 (14.00) Feb. 1978 FD l Technical Assistance II Cr 917 2.50 ( 2.49) May 1979 FD Forestry Cr 918 4.30 ( 3.49) May 1979 - Education II Cr 976 15.00 (13.94) Jan. 1980 - Urban Cr 1049 15.00 ( 6.58) June 1980 - Telecommunications Cr 1058 7.70 t 4.74) July 1980 - Highway III Cr 1132 25.00 (13.02) Apr. 1981 - Nickel Expl. Engineering Cr 1154 4.00 ( 1.97) May 1981 - Kirimiro Rur. Dev. Cr 1165 19.30 ( 3.89) June 1981 - Ngozi III Rur. Dev. Cr 1192 16.00 ( 2.32) Dec. 1981 - Local Construction Industry Cr 1230 5.20 ( 0.98) Apr. 19