Tunisia - Fisheries Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 3753 PROJECT PERFORMANCE AUDIT REPORT 'rUNISIA FIRST FISHERIES PROJECT (CREDIT 270-TUN) December 31, 1981 Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. ABBREVIATIONS BNT - Banque Nationale de Tunisie (National Bank of Tunisia) CGP - Commissariat General a la Peche (Fisheries Authority) ERR - Economic Rate of Return FOSEP - Special Fund for the Encouragement of Fishery Investments (Fonds Special d'Encouragement a la Peche) FPD - Fisheries Project Division ONP - Office National des Peches SCN - Societe de Construction Navale SMAG - Salaire Minimum Agricole Garantie (Minimal Guaranteed Agricultural Wage Rate) FISCAL YEAR January 1 - December 31 FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT TUNISIA FIRST FISHERIES PROJECT (CREDIT 270-TUN) TABLE OF CONTENTS Page No. Preface ............................................................i..... Basic Data Sheet ...................... ............................ ii Highlights ............................................................ iv PROJECT PERFORMANCE AUDIT MEMORANDUM I. SUMMARY ........1...................................... 1 II. MAIN ISSUES ............................................... 6 A. General Project Objectives and Results ............... 6 B. Inadequate Preparation and Appraisal ................. 8 C. Supervisory Problems ................................. 9 D. Project Concept - Replacement of Obsolete Vessels .... 9 E. Loans for Vessel Construction ........................ 12 F. Credit for Inshore Fishermen - General Problems ...... 13 Annex 1 - Ministry of Plan and Finance Comments .................... 15 Annex 2 - BNT Comments ..... ....................................... 17 PROJECT COMPLETION REPORT I. Introduction ................................................ 21 II. Formulation .............................................. 21 III. Implementation ........................................... 26 IV. Institutional Performance .................................... 32 V. Financial Performance ................................... 40 VI. Project Impact and Rates of Returns ........................ 44 VII. Bank Performance ............................................ 48 VIII. Conclusions .............................................. 50 Annexes 1. Report by Commissariat a la Peche (CGP): September 1980 53 2. Table: Situation of Boats Financed under First Fisheries Project as at September 11, 1980 ...... 61 3. Table: Disbursement ............... ....... ........ 62 4. Table: Recovery of Loan Principal and Interest as of September 11, 1980 ........ ...................... 63 Map This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  PROJECT PERFORMANCE AUDIT REPORT TUNISIA FIRST FISHERIES PROJECT (CREDIT 270-TUN) PREFACE This is a performance audit of the Tunisia First Fisheries Project for which Credit 270-TUN was approved on September 14, 1971, in the sum of US$2.0 million. The final disbursement in respect of this Credit was made on January 17, 1980. The audit report consists of an audit memorandum prepared by the Operations Evaluation Department (OED) and a Project Completion Report (PCR) dated July 10, 1981. The PCR was prepared by the EMENA Regional Office on the basis of a country visit in August-September 1980 and a PCR prepared by the Borrower (not attached). The audit memorandum is based on a review of the Appraisal Report (No. PA-91a) dated August 30, 1971, the President's Report (No. P-985) of September 2, 1971, the Credit Agreement dated September 24, 1971, and the PCR; correspondence with the Borrower and internal Bank mem- oranda on project issues as contained in relevant Bank files have also been consulted and some Bank staff associated with the project have been interviewed. An OED mission visited Tunisia in August 1981. The mission held discussions with officials of the Fisheries Authority (CGP), the National Bank of Tunisia (BNT), the National Fisheries Office (ONP) and others. The infor- mation obtained during that mission was used to test the validity of the conclusions of the PCR and to consider some additional aspects of the project. A copy of the draft report was sent to the Borrower on November 9, 1981, for comments. A telex from the Ministry of Plan and Finance stating that they have no comments is attached as Annex 1. Comments received from BNT are attached in Annex 2 and have been footnoted in the PPAM. The audit generally finds the PCR thorough and accurate with respect to the project's principal achievements and shortcomings. The audit memoran- dum discusses the project achievements in general and the change in project concept, and focusses in particular on preparation and supervision and on loans for fishermen. The valuable assistance provided by the Government of Tunisia and the many Government officials and others met during the preparation of this report is gratefully acknowledged.  - ii - PROJECT PERFORMANCE AUDIT BASIC DATA SHEET TUNISIA FIRST FISHERIES PROJECT (CREDIT 270-TUN) KEY PROJECT DATA Appraisal Actual or Item Expectation Current Estimate Total Project Cost (US$ million) 3.1 2.9 Underrun (%) - 6 La Credit Amount (US$ million) 2.0 2.0 Disbursed December 31, 1976 2.0 1.3 Disbursed December 31, 1980 2.0 2.0 Outstanding January 17, 1980 2.0 2.0 Physical Components Completed 06/76 12/31/79 Proportion Completed by Appraisal Target Date (%) 100 60 Proportion of Time Overrun (%)- 74 Institutional Performance - Worse than expected Financial Performance - Worse than expected Economic Rate of Return (%) 36 43 Cumulative Disbursements (US$ million) FY72 FY73 FY74 FY75 FY76 FY77 FY78 FY79 FY80 Appraisal Estimate 0.03 0.24 0.79 1.54 2.00 - - - - Actual 0 0 0.02 0.40 1.20 1.40 1.70 1.91 2.00 Actual as % of Estimate 0 0 3 26 60 70 85 96 100 Final Disbursement Date: 01/17/80 OTHER PROJECT DATA Original Actual or Item Plan Revisions Current Estimate First Mention in Files - - 04/23/68 Government's Application - - 07/24/70 Negotiations - - 07/20/71 Board Approval - - 09/14/71 Credit Agreement Date - - 09/24/71 Effectiveness Date 11/24/71 02/24/72 05/24/72 Closing Date 12/31/76 06/30/77 07/31/78 12/31/78 12/31/79 Last Disbursement 12/31/76 - 01/17/80 Borrower The Republic of Tunisia Executing Agency The Republic of Tunisia/Fisheries Directorate/Bancue Nationale de Tunisie Fiscal Year of Borrower January 1 - December 31 Follow-on Project Name Second Fisheries Loan Number 1746-TUN Amount (US$ million) 28.5 Agreement Date 07/20/79 /a Project was scaled down from 335 boats to 190 boats because of cost overruns of 57% per boat financed. - iii - MISSION DATA Month/ No. of No. of Man- Date of Item Year Weeks Persons weeks Report Preparation by CP 05/69 3 4 12 11/10/69 Preparation by Bank 06/70 1 2 2 08/06/70 Appraisal 12/70 3 5 15 08/30/71 Supervision I 12/71 1 2 2 01/07/72 Supervision II 05/72 0.5 2 1 06109/72 Supervision III 11/72 1.5 2 3 12/12/72 Supervision IV 05/73 1 2 2 06/06/73 Supervision V 10-11/73 2 2 4 01/08/74 Supervision VI 06/74 1 3 3 07/12/74 Supervision VII 09/74 1.5 1 1.5 09/30/74 Supervision VIII 11/74 1.5 3 4.5 01/27/75 Supervision IX 05/75 1 3 3 06/23/75 Supervision X 11/75 1 2 2 01/09/76 Supervision XI 06/76 1 3 3 07/14/76 Supervision XII 02/77 2 3 6 04/29/77 Supervision XIII 06/77 1 1 1 07/08/77 Supervision XIV . 11/77 3 2 6 12/08/77 Supervision XV 06-07/78 1 2 2 08/05/78 Supervision XVI 10/78 0.5 2 1 12/12/78 Supervision XVII 09/79 1 3 3 10/09/79 Completion 08-09/80 2 2 4 07/10/81 COUNTRY EXCHANGE RATES Name of Currency (Abbreviation) Dinar (D) Year: Exchange Rate: Appraisal Year Average (1971) US$1 = DO.525 Average 1973 US$1 = 0.434 Average 1975 US$1 = 0.389 Average 1976 US$1 = 0.422 Average 1977 US$1 = 0.431 Average 1978 US$1 = 0.414 Average 1979 US$1 = 0.410 Completion Year Average US$1 = DO.394 /a This mission also prepared the second fisheries project and supervised the agricultural credit project. - iv - PROJECT PERFORMANCE AUDIT REPORT TUNISIA FIRST FISHERIES PROJECT (CREDIT 270-TUN) HIGHLIGHTS The project which was the first IDA operation in the fisheries subsector in Tunisia was the first step for modernizing Tunisian fishing fleet consisting of 4,700 registered boats. The project, which was based on a 1969 fleet composition survey, was to provide credit to fishermen for replacement of obsolete inshore craEt by 335 fully equipped motorized boats. The project also aimed at reinforcing institutions responsible for the fisheries subsector and for credit to fishermen and at providing overseas training. The main project objective was to enable project-financed boats to increase production of inshore demersal fish by about 55% through better fishing gear, better boat efficiency and an increase in the number and duration of fishing trips. The project has only partially succeeded in achieving its original objectives. Boat unit costs turned out to be about 57% more than appraisal estimates mainly because of lack of detailed design as a basis for cost esti- mates, project start up and implementation delays and inflationary pressure in the wake of the 1973 oil crisis. The total number of boats had to be reduced to 190 and most of the planned improved design features had to be eliminated to cut costs. Frequent engine breakdowns were experienced with excessive repair times resulting in some resistance among fishermen to buy the new boats. The last boat was sold in January 1980, three years after the target closing date. In September 1980, 151 boats were fully operational, 12 boats had sunk and 24 were immobilized and partly abandoned by their owners. Production per boat was less than half of appraisal estimates. Despite the low catch figures, more than half of all boats had acceptable (21-31%) rates of return because fish prices rose faster than inflation; for the same reason, the project's ERR is higher (43%) than estimated at appraisal. Recovery perfor- mance for boat loans has been poor mostly because of the lack of an effective recovery system and technical difficulties encountered by fishermen. The Second Fisheries Project, approved in 1979, has taken account of the lessons learnt from the first project. The following lessons are of particular interest: - the need for effective and thorough project preparation particularly with regard to engineering design and costings in fisheries project (PPAR, paras. 25-29 and PCR, para. 7.02); - v - - because of the lack of fisheries expertise within the Bank fisheries project necessitate more external technical expertise during super- vision than has been the case hitherto (PPAR, paras. 30-33 and PCR, para. 8.03); - sale of boats to people who had no previous experience of boat ownership was a major departure from boat replacement objectives and increased the risk of misuse and default (PPAR, paras. 34-38 and PCR, para. 3.13); - with inshore/artisanal type fisheries, it is counter productive to attempt to force the pace of technical progress unless the program follows a period of trial and demonstration (PPAR, paras. 39-49); and - the need for firm and appropriate arrangements regarding cost recovery from inshore fishermen (PPAR, paras. 52-54 and PCR, paras. 5.04-5.08). PROJECT PERFORMANCE AUDIT MEMORANDUM TUNISIA: FIRST FISHERIES PROJECT (CREDIT 270-TUN) I. SUMMARY 1. The project was appraised in December 1970, following which a US$2.0 million Credit was approved by the Board in September 1971. The Credit was fully disbursed in January 1980, some three years later than the original target Closing Date of December 1976. 2. The initial project idea came from the Tunisian Office National des Peches (ONP) in 1967, in the form of proposals for the replacement of 3,000 fishing boats out of a total of 4,700 registered non-mechanised craft with new motorized units. These proposals were studied by an FAO/CP reconnaissance mission in 1968, an FAO/CP preparation mission in May 1969, a Bank preparation mission in June 1970 and finally by a Bank appraisal mission in December 1970. 3. Based on a 1969 fleet composition survey which showed that, of the 2,400 active boats comprising the Tunisian inshore fishing fleet, about 600 were mechanised but that at least 300 of these were 10 years old or older and in urgent need of replacement,' the number finally retained in the project by the appraisal mission was 335 boats of 10.65m length overall. Smaller craft were excluded as being uneconomic. 4. The appraisal cost estimates were based on a preliminary boat layout prepared by a naval architect at the preparation stage. The Appraisal Report (AR) specified that detailed design would have to be undertaken by a consul- tant naval architect luring project implementation, design criteria being given in the AR. The Lack of detailed design as a basis for cost estimates and the absence of a qualified naval architect, who could have worked out more precise figures, in the appraisal team were the main reasons for a serious underestimation of unit costs. This was aggravated by implementation delays and the inflationary pressure in the wake of the oil crisis of 1973. Boat unit costs finally turned out to be some 57% more than appraisal estimates. The total number of project boats had to be reduced further to 190, only 57% of the envisaged number, because of their cost overruns, and at the same time most of the improved design features which were originally intended also had to be eliminated to cut costs. The outcome was a 10.4 meter boat which differed from the traditional craft only insofar as it had an electrically started Perkins diesel engine in place of the more familiar and locally manufactured, hand-started Tubukh diesel. 1/ Appraisal Report, page 4, para. 2.12. - 2 - 5. The project represented the first IDA operation in the fisheries subsector in Tunisia. It followed a Bank/IDA co-operative farm project (Loan 484-TUN/Credit 99-TUN), which was signed in 1967 but ran into difficulties in 1969 when Government made membership of farmers in collective co-operatives compulsory. Following strong antagonistic reaction by farmers, this policy was reversed one year later, and co-operative membership was again made voluntary. However, these events had negative effects on the co-operative movement and also on the existing fishermen's service co-operatives. Fisher- men were afraid of collectivization of their boats in production co-opera- tives, of which they were genuinely distrustful, and therefore discontinued their membership. As a consequence, nearly all fishermen's co-operatives ceased to exist and are only now just starting to be re-established, 10 years later. 6. The demise of fishermen's co-operatives resulted in important changes in the project's proposed organization and management during prepa- ration: originally it was intended to entrust ONP with project implementation and to oblige project boatowners to sell their catch through co-operatives to ONP, which would retain a certain percentage of the catch value as debt service. Since co-operatives no longer existed at the end of 1970 when the project was appraised, ONP was rejected as the executing agency because the Tunisian Government, in agreement with the appraisal mission, decided that boatowners should be free to market their catch wherever they wanted. No need was seen under these circumstances to retain ONP in the project management, which was transferred to the Fisheries Directorate, a department within the Ministry of Agriculture. However, in retrospect, it proved to be a serious deficiency that no firm arrangements linking catch marketing and debt service were made. ONP still markets more than 50% of all catches from coastal fisheries and could have been instrumental in avoiding at least some of the loan recovery problems which developed as the project progressed. 7. According to the AR, the main project objective was to enable project-financed boats to achieve production of inshore demersal fish of about 4,900 tons per annum, of which 2,650 tons would be additional to production without the project., This was based on an estimated annual catch per project boat of 14.5 tons, compared to an average catch of 9.6 tons/year by similar sized traditional craft. The increase was explained as the effect of using better fishing gear, better boat efficiency and an increase in the number and duration of fishing trips. Issues were indicated to be procurement, sub-loan conditions and project financing. The organization, management and recovery of sub-loans were not mentioned as issues, despite the fact that warning signs were already present in the poor recovery performance of loans to private fishermen under the existing Special Fund for Fisheries Development (FOSEP).i/ 8. Project funds were passed on by the Tunisian Government to the Banque Nationale de Tunisie (BNT) under a subsidiary loan agreement. BNT was the executing agency for the financial part of the project: (a) upon request 1/ AR, page 7, para. 2.25. - 3 - of the Fisheries Directorate it handled contract financing for the procurement of marine Diesel engines and boat hulls; (b) it appraised sub-borrowers' creditworthiness based on technical recommendations from the Fisheries Direc- torate; (c) it concluded the sub-loan agreements with the new boatowners; and (d) it was responsible for recovery. IDA funds were obtained by BNT at 3% annual interest and onlent to sub-borrowers at 6%. Loan applicants were asked to contribute 10% equity, and the Government gave them an investment grant of 15%. Sub-loans covered three quarters of the investment costs, of which IDA refinanced 67.5% and the Government 7.5%. The default risk was shared in equal parts by the Government and BNT. 9. International bids were to be invited simultaneously for complete boats and for marine Diesel engines. Local bids were to be obtained from Tunisian shipyards for the hulls only, including engine installation, but not for the supply of engines. 10. Project management was to be in the hands of the Fisheries Project Division (FPD), to be created in the Fisheries Directorate prior to effective- ness and incorporating three qualified consultants: a fisheries expert as project manager, a naval architect for boat design and construction supervi- sion and a master carpenter for construction supervision and maintenance training. Another condition of effectiveness was the alignment of the conces- sionary terms granted by FOSEP to the terms specified in the project. Again in retrospect, it is unfortunate that there was no inclusion of expertise on the organization of fisheries credit during preparation and appraisal and that the consultants employed to facilitate project management did not also include at least a few man-months specifically to advise and assist BNT in setting up realistic and effective arrangements to service its sub-loans. 11. Effectiveness and project start-up were considerably delayed. While the FOSEP lending terms were quickly adjusted to the project terms, it took eight months to recruit suitable consultants. The credit was declared effec- tive after a consultant contract was signed in May, 1972. The project manager and the naval architect started work in Tunisia in October 1972. However, the procurement process for the boat engines did not start until May 1973, and for the hulls until February 1974, for two reasons: (i) upon recommendation of the naval architect, the procurement pro- cedures were modified to reflect a change in the bidding sequence; bids for hulls were obtained only after a decision on the engine award had been made so that its specifications could be incorporated in the tender for hulls; and (ii) the new boat design prepared by the naval architect was rejected for cost reasons, and the decision was made to return to the traditional Mediterranean design with some minor improvements. 12. A disagreement. arose between IDA and the Government during engine procurement over the recommendation made by the Fisheries Directorate to award the -engine contract to a local firm which assembled in Tunisia engines of Danish components, even though their bid after allowing for the 15% prefer- ence margin was still 47% higher than the lowest evaluated bid. After lengthy discussions, during which cancellation of the Credit Agreement was considered, IDA's view prevailed and the contract was awarded to the lowest evaluated bidder. Unfortunately, frequent engine breakdowns were experienced with occasionally excessive repair times, which resulted in some resistance among fishermen to buy the new boats. Some owners even abandoned their immobilized boats when they had no money to pay for repairs. In September 1980, 151 boats were fully operational, 12 boats had sunk and 27 were immobilized!/ and partly abandoned by their owners. Every effort must be made to improve this situation, which would seriously affect the attitude of fishermen toward acquisition of boats in the Second Fisheries Project. The Government and BNT have committed themselves to this objective. 13. Construction of hulls in Tunisian shipyards went more smoothly than engine procurement; in fact, the institutional performance of the shipyards under close supervision of a CP naval architect was very satisfactory. All 190 boat hulls were completed in about four years, but construction speed slowed down when demand for project boats began to decline in 1976 as a consequence of the emerging problems with the engines. To revive interest.in boat acquisition, FPD started approving sub-loans for applicants who did not possess the necessary professional experience and at the same time accepting financing of the boatowner's 10% equity contribution by the Special Fund for Rural Development. Although especially the latter fact was discussed by supervision missions, no objections were raised by IDA because the detrimental effects on repayment performance were not foreseen. By September 1978, all hulls were constructed, but, because some engines were missing and because of lack of demand, the last boat was not sold until January 1980, three years after the target closing date. 14. Recovery performance for boat loans has been very poor and was characterized, at least until the last two years, by inadequate co-opera- tion between the Fisheries Directorate and BNT in establishing an effective recovery system. The main objective for the Fisheries Directorate was an increase in fish production, and they were in consequence less concerned about the financial soundness of the project. BNT in turn believed that the cen- tralized recovery system which produced more or less acceptable results in agriculture would also work in fisheries and failed to recognize essential differences between the two areas of activity. No special staff was recruited for recovery purposes, and no endeavours were made in the first three project years to establish new and more appropriate procedures. Only in 1978, while negotiating a Second Fisheries Project and after the recovery rate had fallen to 11%, could BNT be persuaded to recruit seven loan recovery officers. Besides the lack of attention paid to recovery, the attitude of boatowners toward their repayment obligations was also an important factor. Due to long and sometimes excessive down-times, many boatowners either did not have the 1/ According to the Region, 11 of these were repaired by August 1981. - 5 - funds to take care of their debt service or used the technical difficulties as an excuse to avoid payment. Legal proceedings took excessive time, and one attempt to auction off a boat under court order failed when the boat was sold to the only bidder at about 20% of its market value. 15. Performance has improved slightly since 1980 with greater awareness of the consequences by all concerned. Representatives of the Fisheries Directorate (now called the Commissariat General a la Peche - CGP) and BNT have been meeting regularly in Tunis to concert their efforts. Since 1978, the collaborative process has been extended to the certain regions so that the local BNT and CGP staff, who are in direct day-to-day contact with fishermen, can work together more effectively. 16. Unit costs for the project boats with price contingencies were estimated at appraisal at D4,210, while actual unit costs were D6,630, 57% more. In dollar terms, the difference is 88% due to a 38% depreciation of the average dollar exchange -rate against the dinar during the project period. As a consequence, only 190 boats could be constructed and sold instead of 335. Total catch of all project boats in 1979 is estimated at 780 t. The reason for this disappointing result is the reduction in numbers of boats built and also the frequency of engine breakdown, which reduced the production per boat to about 6.5 t per annum, less than half of appraisal estimates. Despite the low catch figures, it is estimated that slightly more than half of all boats had acceptable financial rates of return, between 21 and 31%, because fish prices rose much faster than inflation. This same reason led also to the surprising fact that, despite the low production performance, the project's expected economic rate of return is higher than estimated at appraisal: more than 40% if calculated without benefits derived from wholesale and retail activities but with shadow pricing of labor cost as in the AR. If marketing benefits are taken into account (as was done for the subsequent Second Fisheries Project), the ERR increases to about 80%. It is estimated that about 250 new jobs have been created; 300 was the estimate at appraisal. Figures about improved employment opportunities are inconclusive. About 160 man-years of employment have been created in the shipyards. 17. Among the general lessons for the Bank are: (i) the realization that fisheries projects are not easy to manage and necessitate the deployment of more technical expertise during supervision than has been the case hitherto; the lack of in-house expertise means, in consequence, that supervision costs will be high;./ (ii) the need for careful scrutiny of client government requests to limit the use of consultants, to guard against premature termination and ensure reasonable continuity of project management; 1/ Project Completion Report, para. 8.03. - 6 - (iii) the need for firm and appropriate arrangements regarding cost recovery, particularly in cases involving sub-lending to inshore fishermen;1/ (iv) the need to plan all procurement activities carefully at appraisal, so that appropriate specification and adequate maintenance are assured; and finally, (v) the need to adhere more strictly to the terms of project agreements, especially as regards project objectives and to permit variations only after careful prior consideration of the reasons for and implications of such changes. 18. The Second Fisheries Project has taken account of the lessons learnt from the First Project so far as they were already apparent in 1979. The implementation lessons are permanently applied as the Second Fisheries Project progresses. II. MAIN ISSUES A. General Project Objectives and Results 19. The objectives of the Tunisia Fisheries I Project are clearly set out in paragraphs 22 through 26 of the President's Report (No. P-985, dated September 2, 1971). The main objective was to enable project-financed boats to achieve production of inshore demersal fish of about 4,900 tons per annum, of which 2,650 tons would be additional to production without the project. This target would be achieved by: (i) providing credit (through BNT) to fishermen for replacement of obsolete inshore craft by about 335 fully equipped motorized boats, including fishing gear; (ii) establishing a Fisheries Project Division in the Ministry of Agri- culture to manage the project, staffed by an expatriate fisheries specialist (project manager), naval architect and master carpenter/ shipwright and Tunisian counterpart staff, whose functions were to design and cost the new project fishing vessels, incorporating design features making them significantly better than existing traditional craft in terms of sea-worthiness, trip duration and production capacity; to supervise their construction; to organize short term training courses for the fishermen who would be using the new craft and to undertake technical appraisal of loan applicants; (iii) reinforcing BNT's staff to carry out financial appraisals of loan applicants and ensure satisfactory loan repayments; and 1/ Project Completion Report, para. 7.04. - 7 - (iv) providing overseas training for selected Tunisian staff in the Fisheries Project Division of the Ministry of Agriculture. 20. In the event, because the project was inadequately prepared partic- ularly as regards vessel costs and was therefore seriously underfinanced, only 190 vessels were built, and even these had none of the improved design features which were to have been provided, except for engines fitted with an electrical starting system in place of the more traditional hand-starting mechanism. 21. The Fisheries Project Division was established, and the project manager and naval architect were appointed, in October 1972, but the post of marine carpenter was deleted. Tunisian counterpart staff were also appointed, but frequent staff transfers and substitutions and the departure of the project manager after 21 months just when the project was entering its active implementation phasel meant that management lacked continuity and was therefore not as effective as originally anticipated. In particular, super- vision and inspection of vessel construction in the boatyards was not as thorough as was planned, and would have been worse but for the inclusion by the Bank of an FAO/CP naval architect in several of the supervision missions. The training courses for fishermen takin over the new vessels were far from adequate, either in duration or content_,/, and there appeared to be little or no follow-up to ensure that the new owners were using the vessels properly. 22. Although the project aimed to reinforce BNT's staff to manage the fishermen's loan scheme, BNT did not want either financial support or provi- sion for technical aid to assist it in setting up an effective and relevant mechanism. Until 1980 only one BNT official was assigned to this role. Since then seven additional loan officers have been appointed, but even though the recovery of arrears has improved slightly, the situation is still far from satisfactory. 23. The provision for overseas training for FPD staff, which was con- tained in the original budget, had perforce to be deleted when the inadequacy of the total provision to build the boats was realized. This is unfortunate, because there is no doubt that many Tunisian fisheries staff would have bene- fitted greatly from the opportunity to broaden their experience by studying the fisheries of other countries and thus increased their own technical expertise. 24. Thus, although the PCR points to some brighter results, such as a re-estimated economic rate of return of 43% compared with an appraisal target of 36%, and to the creation of 250 new jobs from 190 boats compared with the 1/ Project Completion Report, para. 4.05. 2/ Project Completion Report, para. 4.06. -8- original target of 300 jobs based on 335 boats, these figures only serve to cloud the reality, which is that, in several important respects, the project has either failed or has at best only partially succeeded in achieving its original objectives. With the interests of future fisheries projects in mind, the most crucial aspects are considered in greater detail in the following sections. B. Inadequate Preparation and Appraisal 25. The FA0/CP preparation mission in May 1969, which comprised a fisheries generalist, an economist, a regional fisheries officer and a naval architect, apparently did a fair job in identifying where the needs for replacement boats lay and in reducing the earlier overambitious ONP proposals to a project of more manageable proportions. The naval architect produced, as a guide only, a preliminary general arrangement drawing and provisional costing of a vessel which the team thought might be suitable, subject to more investigation, detailed design and accurate costing. 26. The Bank appraisal mission which visited Tunisia for three weeks in December 1970 consisted of four staff members and one consultant (fields. of expertise not stated) but did not include a naval architect. The mission used the CP preparation boat sketch and costing as the basis for their calculations and concluded that 250 smaller craft would not be economic, so reducing the number to be built under the project to 335 of the longer (10.65 meters l.o.a.) class. 27. But instead of requiring the Borrower to carry out the detailed boat design and costing study before finalizing the appraisal report and presenting the project for Board approval, it was decided, as in most agricultural projects financed at that time, to leave this work until after the project became effective, with the result, as described in the PCR, that the project was underfinanced from the outset and thereafter had no chance of achieving its objectives. 28. In fairness, it is understood that formal arrangements to advance funds under a Project Preparation Facility (PPF) were not in force until some time after 1970, but there must have been some alternative means of ensuring that economic evaluation of project proposals could be based on accurate and up-to-date cost information. Irrespective of the size of a project, the basic issue is that it should not be allowed to emerge from the appraisal phase until it has been fully prepared and costed. 29. It must be acknowledged that even had this been done, the budget would still have suffered, in company with many others, from cost inflation consequent on the 1973 oil crisis, but the project would have had a better chance of achieving its objectives if the post-appraisal estimates had been more soundly based. -9- C. Supervisory Problems 30. A total of 18 supervisory missions were undertaken during project implementation, between January 1972 and September 1979, at roughly six- monthly intervals. The frequency of visits was thus quite adequate, and the general continuity of staff undertaking these missions was satisfactory. 31. In this case, the key post was that of the project manager, who served for 21 months from October 1972, but who left just when the project became fully active after a long series of procurement delays. The Bank should have resisted more strongly the government's proposed reduction in the length of his appointment from 24 to 18 months, and insisted on his replace- ment by a person of equivalent calibre for at least a further period of 24 months, in view of the delays which made it clear that the project could not possibly achieve its target closing date of December 1976. 32. Technical input thereafter was limited to five occasions when the supervision missions were accompanied by an FAO/CP naval architect, without whom it would not have been possible to identify and rectify numerous defi- ciencies in boat construction. In retrospect, it would have been useful to include other specialists on occasion, such as a marine engineer to check on engine installation and the quality of other equipment. 33. Even so, such missions rarely spent more than one to one and a half weeks in the country, and much of that time would be used for meetings in Tunis. It might therefore be advisable to include financial provision in such projects for a number of short-term technical consultancies of two to four weeks duration, which should be long enough for the experts concerned to gain an appreciation of the situation in the field and identify incipient problems early enough to brief subsequent supervision missions on the action needed to effect solutions. D. Project Concept - Replacement of Obsolete Vessels 34. Although the project's objective was simply to increase produc- tion, all project documentation from identification through preparation and appraisal up to and including the President-s Report, refers to the provision of new boats by the project as being the means whereby old and worn out craft can be replaced, thereby restoring and improving the livelihood of the fisher- men concerned!!. It follows that sub-lending was originally intended to be restricted to those fishermen who owned and operated one of the old boats. 1/ CPS staff points that from the outset, the Bank accepted the view of ONP that worn-out vessels were the binding constraints to productivity. No alternative approach to improving productivity was seriously considered. Other major constraints to productivity, such as the demand for fish and the adequacy of marketing arrangements were not explored. A better grasp of these issues may not only have resulted in sounder cost recovery procedures but may have strengthened the basic economic rationale and design of the project. In sum, a major deficiency in the project pre- paration effort was in the area of economic analysis, namely the analysis of basic development constraints which needed to be tackled to improve the productivity of fishermen. - 10 - 35. Thereafter the term "replacement" hardly figured at all in the supervision reports or file correspondence. Indeed, in the PCR (para. 6.07), an analysis of the 190 new boatowners notes that only 57 had previously owned boats, without stating how many of these were motorized, and also records that no more than 38 obsolete boats were actually replaced. 36. It must therefore be assumed that at least 150 old boats which were originally intended to be replaced are either still in operation 10 years later or in the meantime have finally gone out of service, leaving their owners and crews presumably without the means to carry on fishing. 37. The real issue, however, lies in the fact that the relaxation by BNT, towards the end of the project, of standards for sub-borrower appraisal to permit the sale of project boats to people who had little previous experi- ence of boat operation and maintenance represented a departure from agreed project lnding policy and a significant increase in the risk of misuse and defaulti The record provides no indication that this policy change and its implications were proposed, studied or approved in any formal manner, or indeed that the change received more than cursory retrospective examination by supervisory staff in view of what was seen as the over-riding need to find buyers for unsold boats. 38. The experience points very clearly to the need in future to adhere much more strictly to the terms of a project agreement, and to permit varia- tion of those terms only after careful prior consideration of the reasons for and implications of such changes. Flexibility in response to changing circum- stances is all very well, but unless the response is regulated, it can only lead to waste of scarce resources. 39. Secondly, the experience re-emphasizes the lesson that, with inshore/artisanal type fisheries in particular, it is counter-productive to attempt to force the pace of technical progress by providing numbers of boats or equipment of unfamiliar types, unless the program follows a period of trial and demonstration with two or three prototypes and concentrating on a few fishermen to develop a feeling of confidence. Thereafter, successful use of the equipment by such fishermen serves as the best possible demonstration to the fishing industry as a whole and can trigger a snowballing response as awareness spreads*to create a new and growing demand for the equipment. 40. In this case, the objectives of replacing old boats and thereby increasing production from underexploited fish stocks in no way required that the new vessels be significantly different from the ones to be replaced; some improvements in design (e.g. the addition of cabins) were proposed, however, to enable the new boats to operate further from shore and for somewhat longer periods before returning to port. In view of the large number of replacements that were proposed in a relatively short time with no built-in opportunity for a trial and demonstration period, it would perhaps have been prudent to deliberately minimize the changes. Thus, there is a stronger case perhaps for 1/ BNT notes that it was not responsible for the selection of fishermen. - 11 - the Tunisian view regarding engine selection than they actually presented/Y, but it is a matter which should have been resolved before taking the decision to go for international tender. More importantly, if the engine tender specification had been based on any reasonable survey of what the fishermen were accustomed to using and on the knowledge, after the government deleted the cabins, that the engines were to be used in flush-decked craft with minimal protection from sea water and spray, it should have stipulated a mechanical starting system rather than electric. 41. As a general rule for future guidance, manual or hand-crank starting systems are the least complicated and therefore particularly suited to arti- sanal conditions. They are perfectly suitable for Diesel engines up to about 40 hp in hot climates, but rather less power in colder areas. Spring loaded inertial mechanisms are available for starting engines from 30 hp upwards, and there are hydraulic pressure loaded inertial starters for engines from about 50 hp to 2,000 hp. Compressed air starters cover a similar power range, and of course electric starters are perfectly satisfactory for small and large engines, provided that all the electrical components and circuitry can be properly protected under cover from sea water and moisture in a wheel- house or similar structure. 42. The decision to employ international competitive bidding for engine supply, with the aim of providing power for the boats at least cost and, as it transpired, coincidentally providing fishermen with an alternative choice in a market which had until then been a virtual monopoly for the locally manufac- tured engine, was no doubt admirably intentioned and in keeping with Bank lending policies and procurement guidelines. However, it did not foresee the conservatism of inshore fishermen and their lack of technical expertise in regard to mechanization and failed to provide for the kind of training needed to familiarize fishermen with the new engines. No doubt had the project manager continued in post or had he been replaced, more effective action would have resulted. 43. The agent acknowledged that the electrical starter system was unsuitable but claimed that, in accordance with the original specification, it was supplied on the assumption that it would be positioned inside a wheel- house. He claimed to have recommended substituting either a dynamo or an alternator having an integral regulator instead of the components originally specified and agreed, but neither course was adopted, probably because of the additional expense involved. 44. The record indicates that the agent cooperated not unreasonably in maintenance efforts by providing staff and spares for a mobile unit operated in conjunction with CGP and in establishing stocks of spares in several centers. No doubt this could have been even better, but whether such a service could have been justified in commercial terms is open to question. 1/ PCR, para. 3.05. - 12 - Although a number of cases involved serious problems such as engine seizure, it is also significant to note that in some instances, often involving a long journey for the mobile team, the cause of engine failure was found to be quite trivial, e.g., failure to start caused by battery draining resulting from water ingress causing a short circuit, or even only a blown fuse or loose battery terminal. It was claimed that only five cases necessitated cylinder head removal. The most serious cases resulted from engine misalignment caused by the use of unseasoned eucalyptus for engine bearers, which in several instances resulted in fracture of the rear-end drive shaft. E. Loans for Vessel Construction 45. Although as stated in the AR the ultimate project objective was to increase fish production, at least insofar as the Fisheries Directorate was concerned, in essence the project was really directed at providing a line of credit to the skipper/owners of obsolete vessels to enable them to replace their old craft with new ones. 46. The project aim might therefore have as easily been achieved simply by making the funds available through BNT for the fishermen concerned to make individual applications, without any intervention in bulk procurement of engines and boats, leaving it to the applicant to stipulate to the boatyard what size and type of craft and engine he wanted, within perhaps some frame- work of guidelines as to costs and specifications. 47. Since in such a case each boat would have been built as a one-off job, unit costs would no doubt have been higher, but the fishermen would have inspected their boat during construction and would therefore choose the nearest yard to their own base for fishing operations. 48. The approach adopted, namely of deciding beforehand on boat specifi- cation, bulk purchasing of engines and equipment, and building boats on con- tract, certainly had the advantage of reducing unit costs and facilitating maintenance but tended to ignore the fishermen's essentially conservative and individualistic nature. Each fisherman has his own ideas as to design, layout and equipment, and although a "standard" boat may satisfy some of his personal requirements, it is a near certainty that he will find something to grumble about, no matter how careful the designer may have been to ascertain fisher- men's views beforehand. With luck, the unpopular aspects may be fairly minor features, which the fishermen can either modify without undue expense or learn to live with. In this instance, however, the complaints centered around a crucial feature, namely the engine, and in particular on its electric starting system. 49. The effect was a drop in demand from the experienced fishermen and owners of old boats whom the project was intended to assist, and in conse- quence, when confronted with the prospect of large numbers of boats going unsold, project management succumbed to the temptation of selling to other applicants who had insufficient professional experience to meet the original - 13 - criteria. It should have been apparent from the outset that, if experienced fishermen had difficulty in maintaining these engines, the problems would simply multiply when boats were put into inexpert hands, and it should there- fore occasion no surprise to find that this did happen. F. Credit for Inshore Fishermen - General Problems 50. In recent years, there have been several excellent studies into problems associated with the provision of credit to fishermen for the purchase of new boats and equipment. FAO in particular has published a number of reports on the subject aimed at providing fisheries administrations and lending institutions with the guidance necessary to avoid the more obvious pitfalls and to assist in setting up workable schemes. After the unexpected demise of cooperatives suddenly removed the best and most obvious means of channeling the credit to the fishermen, the authorities in Tunisia and the Bank still decided to venture into a loan scheme which took little account of the nature of inshore fishing or of the existing studies of the related problems. 51. Inshore fishermen operate in the main on a daily basis or overnight and so land their catches for sale several times each week. In consequence, their cash receipts are frequent and in small amounts, and unlike farmers they rarely if ever handle large sums of money. In developing countries especially, fishermen are rarely more than semi-literate and therefore unable to maintain account books. This situation frequently gives rise to a belief amongst fishermen that they are poor people who earn very little and who thus occupy a very subordinate role in society. The view is often shared by other people, not least by their own children, who may have received better educa- tional opportunities, so that it is not uncommon for a son to undervalue his father's worth and show reluctance to carry on his father's occupation. This is unfortunate because knowledge of fishing grounds, fish migrations and other seasonal changes which has been acquired over generations of experience and passed down from father to son, is at the root of any successful inshore fishing industry, and very little of this information, which is the real collateral for a loan, finds its way into the records of scientific knowledge of the fishery. In reality, if accurate accounts could be kept, experienced fishermen's earnings over a year frequently add up to amounts that a farmer might well envy, and their role in society is much more important than they realize. 52. One of the benefits of a well organized fishermen's loan scheme is that it can provide the accountancy role which the fisherman cannot perform, and so help to create an understanding of his worth. However, it can only do this if the organization extends to the level of personal day-to-day contact with the fisherman to facilitate loan repayments in the form of daily deduc- tions of a proportion of his income from catch sales. In other words, the bank has to establish itself on the quayside at the point of sale, and bank staff must set out to create a relationship based on mutual respect and confidence, such that the fisherman's automatic first point of call after - 14 - receiving his catch value is the bank, and not the neighborhood cafe. The bank should be a place where the fisherman can be encouraged to discuss his experience and his problems and to seek advice. 53. A fisherman's only means of earning is when he is out at sea. It follows that he cannot afford to take time off to travel any distance to pay loan installments, and for this reason as well it is essential that the bank must establish facilities on the quayside. In addition, the scheme must take account of the fact that the quantity and variety of fish caught will vary markedly from day to day and seasonally, and it is thus useless to expect that a fisherman can pay a fixed amount each time. The bank's own accountancy system must be organized to accommodate fluctuations in payments, and this is easily done if the daily payments are credited to a savings account from which periodic deductions can be made for credit against his loan account. 54. Finally, the loan scheme and bank organization must be able to cater for the fisherman's need, at certain times of the year when fish migrate to another section of the coast, to follow the fish and base himself temporarily at another fishing harbour. Unless the bank has another branch there, it must also be prepared to move on some temporary basis.E/ 1/ BNT notes that since fishermen were free to sell their catch everywhere it was not possible to benefit from ONP's and traders' assistance for recovery. A monthly recovery system was also impossible to establish because of the lack of adequate facilities of BNT and Fisheries Direc- torate. Better cooperation between BNT and Fisheries Directorate for cost recovery procedures and selection of fishermen would have been desirable. Since recovery laws were not enforced, more appropriate arrangements for recovery should have been worked out. Slow rates of recovery were mostly due to (i) technical shortcoming of boats, (ii) poor selection of fishermen, (iii) lack of extension and monitoring systems and (iv) the socio-political context prevailing during project implementation. In addition, the recovery rates mentioned in this report might well be somewhat misleading since they do not take into account that (i) outstanding debt is the only element taken into account for calculating the recovery rate, regardless of interests which are first repaid and (ii) the BNT's central accounting system does not immediately record repayments made to local agencies. - 15 - ANNEX 1 Mr. Shiv S. Kapur Director Operations Evaluation Department World Bank 1818 H Street N.W. Washington, D. C. 20433 Dear Mr. Kapur: Thank you for your letter of September 11, 1981 together with the Audit Report for the First Fisheries Project, Loan No. 270. We are pleased to report that the DG PR has no additional comments to make. We thank you for consulting us and would be happy to be further involved in the preparation of such reports in the future. With best wishes, M. A. Souissi Director General of Projects Ministry of Plan and Finance Tunis, Tunisie  -17 - ANNEX 2 FORM NO. 788 Page 1 of 3 (1-74) IM LANWAGE SEVCES DISION CONTROL NO. - ATE: December 2 , 19 1 ORIGINAL LANGUAGEt * DEPT : t CABLE FROM: BANQUE NATIONALE DE TUNISIE (= BNT) TO: INTBAFRAD WASHINGTON TELEX NO. 24250/SM HEREWITH REPETITION FIRST IDA FISHERIES PROJECT, 270-TUN COMMENTS ON EVALUATION REPORT 1. BNT WAS NOT RESPONSIBLE FOR SELECTING THE FISHERMEN. THE SOLVENCY STUDY ENTRUSTED TO IT CAN HAVE NO BEARING SINCE IT IS A MATTER OF NEW CLIENTS. 2. IT WAS NOT POSSIBLE TO INVOLVE ONP OR THE FISH TRADERS IN LOAN REC- OVERY SINCE THE FISHERMEN HAD COMPLETE FREEDOM IN MARKETING THEIR PRODUCTS. 3. MONTHLY RECOVERY WAS ALSO DIFFICULT TO ENVISAGE SINCE IT NECESSITATES SPECIAL STRUCTURES TO BE PUT IN PLACE IN THE FRAMEWORK OF THIS PROJECT AT THE LEVEL OF BNT AND IN PARTICULAR OF THE FISHERIES DIRECTORATE. CLEARER DEFINITION WOULD ALSO RAVE BEEN NECESSARY OF THE RELATION- SHIPS TO BE ESTABLISHED BETWEEN THESE TWO AGENCIES FOR EXECUTION OF THE PROJECT AND THEIR COLLABORATION IN CONNECTION WITH LOAN RECOVERY AND SELECTION OF BENEFICIARIES. 4. NO MEANS WERE PROVIDED FOR EXERTING PRESSURE IN THE EVENT OF NON- PAYMENT, PARTICULARLY SINCE THE CURRENT LEGISLATION RESPECTING LEGAL PROCEED- INGS WAS NOT APPLIED. 5. THE LOW LOAN RECOVERY RATES ARE ATTRIBUTABLE TO A NUMBER OF FACTORS, INCLUDING THE FOLLOWING: ANNEX 2 Page 2 of 3 - APPRAISAL ERRORS WITH RESPECT TO FISHING TECHNIQUE. - SELECTION OF THE FISHERMEN. - CHOICE OF EQUIPMENT. - LACK OR INADEQUACY OF EXTENSION AND MONITORING SERVICES. - FAILURE TO APPLY CERTAIN MEASURES ADOPTED IN 1979 BY MUTUAL AGREEMENT BETWEEN BNT AND CGP. - THE DIFFICULTY OF TAKING LEGAL ACTION HAVING REGARD TO THE LEGISLATION IN FORCE. - THE SOCIO-POLITICAL CONTEXT IN WHICH THE PROJECT IS BEING EXECUTED. 6. THE ACTUAL LOAN RECOVERY RATES ARE NEITHER REAL NOR MEANINGFUL, FOR THE FOLLOWING REASONS: - VERY SLOW REPAYMENT IS TYPICAL OF LOANS FOR THIS TYPE OF CREDIT. - THE RATE IS CALCULATED ON PRINCIPAL ONLY, WHEREAS INTEREST IS COLLECTED IN PRIORITY. - NO ACCOUNT IS TAKEN OF MATURITY DELAY FOR LATE DELIVERY OR IMMOBILIZATION OF VESSELS. - THERE IS OFTEN A TIME LAG BETWEEN RECEIPT OF REPAYMENTS BY BNT AGENTS AND THEIR POSTING TO THE APPROPRIATE ACCOUNTS. - THE AMOUNT OF ARREARS IS CONSIDERABLY INCREASED BY THE LARGE NUMBER OF VESSELS OF UNSUITABLE DESIGN. - THE RECOVERY RATE IS CALCULATED BY REFERENCE TO THE-CUMULATIVE AMOUNT OF MATURITIES FALLEN DUE (EXCLUDING TIME EXTENSIONS) AND THEREFORE CANNOT BE MEANINGFUL FOR A GIVEN YEAR. I_THE IMPROVEMENTS DO NOT BECOME APPARENT UNTIL AFTER THE LAST MATURITIES HAVE FALLEN DUE. ANNEX 2 19 Page 3 of 3 7. GRANTING OF NEW LOANS FOR REPAIRS OUT OF FOSEP FUND WITH ARRANGEMENT FOR SETTLING ARREARS. 8. THE LOCAL AGENCIES SHOULD PARTICIPATE IN LOAN RECOVERY IN COLLABORATION WITH BNT, THE CGP RETENTION OF 25 PERCENT OF DAILY PRODUCTION. (Sentence evidently defective.-Translator). 9. (None -- perhaps part of 8 above belongs here.- Translator). 10. THERE IS ALWAYS A LINK BETWEEN PRODUCTION, PRICE AND LOAN RECOVERY. YOURS SINCERELY, BANQUE NATIONALE DE TUNISIE NATIONAL BANK OF TUNISIA  - 21 - TUNISIA PROJECT COMPLETION REPORT FIRST FISHERIES PROJECT (CREDIT 270-TUN) I. INTRODUCTION 1.01 The Credit Agreement does not envisage the involvement of the Borrower in the preparation of the Project Completion Report (PCR). Nevertheless, the Government's Fisheries Authority (CGP)/1, established in 1980 as part of a reorganization of the Fisheries subsector, prepared a completion report which is reproduced in Annex 1. This PCR is therefore based on: (a) CGP's completion report; (b) the Associations's (IDA) supervision efforts; and (c) the findings of a completion mission that visited Tunisia in August/September 1980, eight months after the Closing Date. 1.02 The PCR focuses on: (a) the Project background to understand the original Project idea and its evolution; (b) the procurement aspects; (c) the difficulties with recovery of sub-loans; and (d) the lessons learned for the implementation of the follow-up Project. (Fisheries II, Loan 1746-TUN signed July 1979). II. FORMULATION A. Identification and Preparation 2.01 The initial Project idea came from Office National des P9ches (ONP), which in 1967 proposed the replacement of 3,000 fishing boats out of 4,705 registered boats without engines by new motorized units. This was considered by ONP to be of extreme urgency to increase production after a proper marketing organization had been set up in the form of service cooperatives which were to sell the fish caught by their members to ONP. The first FAO/CP reconnaissance /l Commissariat Ggndral : la P&che -22 - report of 1968 contended that the figure of 4,705 registered boats included also a large number of inactive boats, and that a more likely figure for the active ones was about 3,000. It was therefore proposed to envisage the construction of about 750 motorized fishing boats or 1/3 of the existing inshore fishing fleet. The report predicted an annual catch of 5.5 t/year/boat resulting in a total production of 4,125 t and an incremental production of 1,500 t. ONP was envisaged as executing agency which would receive financing from IDA, construct the hulls in its shipyards and sell boats to fisheries cooperatives. Two points are of special interest in this first proposal (a) it was at this time envisaged to use locally assembled boat engines, and (b) subborrowers' repayments were to be effected by way of deducting them from the value of fish delivered by cooperatives to ONP. The proposal in the report to use locally assembled engines was modified in a letter from FAO/CP to the Bank which said that the Government would also agree to import engines from abroad if this meant a substantial reduction in the cost of the Project. Marketing of fish through cooperatives which would sell to ONP became impossible in the wake of the experience in 1969 in the agricultural sector, when Government had made membership of farmers in collective cooperatives compulsory, but reversed its decision one year later. Thus the establishment of effectively operating fishermen's cooperatives was not possible due to the genuine distrust of fishermen against such organizations. Not until 1979 were new attempts made to spread the cooperative idea in the context of the Second Fisheries Project. 2.02 Preliminary results of a survey obtained in August 1968 showed an active fleet of less than 3,000 boats of which 700 were estimated to be motorized. In a follow-up report, it was brought out that the annual catch of combined longline/gill netters of about 9 to 9.5 m overall length was 7 to 11 t and in accordance with these figures, the envisaged production of Project boats should be higher than 5.5 t per year. The FAO/CP preparation report issued in November 1969, proposed a Project consisting of 600 fishing boats (250 of 6.5 m length and 350 of 9.5 m length at the waterline) phased over five years. The mission consisted of a fisheries generalist, an economist, a regional fisheries officer and a naval architect. All the 600 Project boats were intended to be a replacement of existing boats catching demersal /1 fish. The incremental production was therefore estimated to be only 2,000 t/year out of a total production of 5,200 t/year. The assessment of demersal fish resources seems to have been correct. As research completed in 1977 and 1978 shows that there are substantial stocks of demersal fish still unexploited on the total continental shelf which permits the conclusion that also the zone reserved for coastal fisheries (up to a water depth of -50 m) have still potential reserves. Investments in purse seining operations by lamparo fishing at the time seemed to be premature because for the species caught by lamparo fishing techniques, mainly sardines, as for pelagic /2 species in general, resource studies were not available. Moreover, they would have required simultaneous investments in shore facilities such as freezing tunnels and cold store equipment, which are now being made available under the Second Fisheries Project (Loan 1746-TUN). /1 Demersal fish are bottom water fish, e.g. perches and breams. /2 Pelagic fish live in open waters and migrate, e.g. tuna, sardines, mullets, anchovies. -23 - 2.03 In the preparation report, the idea of limiting procurement to locally assembled engines was dropped and ICB procedures were proposed for engines and fishing gear while boat hulls were envisaged for construction in local shipyards. It was explicitly said in the report that the basic design of existing traditional boats would be retained with improvements such as: (a) shaping the stern in a way that would allow for more working space on deck; (b) adding crew accommodation for a crew of four for the larger type boats; and (c) adding an insulated fish hold. 2.04 Due to overriding cost considerations, the above improvements were later dropped during execution of the Project. Other Project features proposed in the preparation report but not retained at appraisal because they were deemed to be too costly, not important enough or impractical were the following: (a) the construction of two ferro-cement boat prototypes; (b) 12 man-months consultant time for a training program to train 14 Tunisian instructors who in turn would train the 600 boat skippers; (c) emphasis on boat: maintenance to be carried out through delivery of spare parts with the boats for storage in the cooperatives, repair of boats in the cooperative workshops and the ONP central workshop, periodical inspection of maintenance and cancellation of loans for fishermen who do not maintain their engines properly; and (d) marketing of full boat catch through ONP marketing organization and retention of 20% of catch value from payments to fishermen. Features proposed at preparation and retained at appraisal were: (a) 36 man-months of expert time, 12 months for a naval architect and 24 months for a master carpenter; (b) the choice of BNT as financial agency to process sub-loans; and (c) loan conditions including financing of 80% of investment costs by loan, 15% by Government subsidy and 5% by downpayment of fishermen. Loan duration eight years and interest 3%. It was proposed however, to give the subsidy only to those fishermen who would acquire a boat for the first time (mostly retained). 2.05 At the time when. the preparation mission visited Tunisia in 1969, the Government had not yet decided whether ONP or its Fisheries Directorate should be the main executing agency. Mission suggestions at that time favored ONP because of its experience and the availability of a marketing network. A possible conflict of interest because ONP owned shipyards which would possibly construct part of the boat hulls was described as main obstacle to making ONP the chief executing agency. Numerous IDA comments to the report were discussed at length with CP. Some points were highlighted, which proved to be important for Project organization: - 24 - (a) There was no legislation to compell fishermen to sell their catch to ONP. This led to the situation that retention of money at the source was not envisaged in the appraised Project. It was later repeatedly proposed by IDA supervision missions as a consequence of the low recovery rates; (b) Fishermen had in recent years been increasingly distrustful of cooperative organizations because they feared vessel collectivization. With voluntary membership, members would no longer have this concern. However, at the time fishermen were distrustful of cooperatives also because of fear of corruption within the cooperatives. Therefore fishermen withdrew from the cooperatives and most of them were dissolved. It was correctly predicted already at this time that inclusion of independent fishermen might result in problems with loan recovery; (c) The transport costs for imported boats were estimated to be $1,000 which would hardly justify import of the hulls; and (d) More information on resources were given and it was stated that although pelagic fish resources were more abundant than demersal fish resources, a bankable project would not be available in the near future due to the underdevelopment of the domestic consumer market and to the difficulties experienced in exporting canned pelagic fish. 2.06 Shortly before appraisal, CP indicated that following the strong reaction in Tunisia against compulsory cooperative membership, a number of fisheries cooperatives had been dissolved, the Director of ONP had been replaced and that now fishermen would try to market their catch outside ONP channels. This then led to the development that ONP was entirely excluded from Project management with the consent of the Tunisian Government. In the official Government request for appraisal of the Project dated July 24, 1970, it was proposed to allow fishermen to sell their catch freely but that granting of a sub-loan would be contingent on the fishermen's approval to turn over their catch to fish traders approved by BNT in order to make direct deductions from the sales proceeds. This shows that at that time, the recovery difficulties which had to be expected from free marketing arrangements were correctly assessed. This realistic attitude was shared by an IDA preparation mission in August. However, no satisfactory solution for controlled marketing was found at appraisal. B. Appraisal 2.07, The terms of reference for the appraisal mission were issued in November 1970 for a team of four Bank/IDA staff members and one consultant. They included an examination of data on fish resources and their development potential, and an assessment of the suitability of the existing and proposed boats and fishing gear. However, the appraisal mission did not include a naval architect and used the same boat sketch that was contained in the preparation report for appraisal and cost estimates. At the return of the mission, the Project composition was still unchanged, i.e. about 600 boats of 6.5 m and 9 m length at the waterline. Issues were indicated to be procurement, sub-loan conditions, and Project financing. Organization, management and recovery of sub-loans were not mentioned -25 - as issues. During preparation of the SAR, it became clear to the mission that the boats needed larger engines than originally estimated, 20-25 HP for the small boats and 30-40 RP for the larger ones. Investment costs for boats were then found to be so high that the financial and economic returns of the smaller boats were no longer considered to be satisfactory. In discussions in Tunisia preceeding the formal credit negotiations in Washington in July 1971, the Tunisian Government reluctantly agreed to the exclusion of the smaller type boats. In retrospect, the decision of the appraisal mission to exclude the smaller type boats seems to have been correct not only because of the lower profitability of these boats but also because experience has shown that the trend has been to larger boats and it may have been difficult to sell the smaller ones. 2.08 The Project as finally retained in the SAR and the credit documents had the following features: (a) Provision of 335 fishing boats of 9.5 m length at the water line or 10.65 m overall length intended to replace an equal number of obsolete boats in four years. The number of 600 boats was reduced because of financing constraints; (b) The design criteria for the boats to be incorporated into the detailed design by a consultant naval architect would be a cut-off stern to increase propeller efficiency, centered engine and cabine and wide beams to improve handling at sea and enable fishing from the stern as well as the bow, lowered decks to protect the crew adequately and insulated hold to improve the quality of landed fish; (c) Sub-loans to fishermen not exceeding 75% of investment cost would be made by BNT at 6% interest for a duration of eight years with fishermen contributing 10% and the Government 15% as subsidy. A condition of Board presentation was introduced that draft legislation should be introduced, that would make lending conditions provided under the Special Fund for Fisheries Development (FOSEP) identical to those provided under the Project; (d) With respect to procurement, it was stipulated that the following bids would be invited simultaneously: (i) bids for complete inshore fishing boats by international competitive bidding (ICB); (ii) bids for inshore fishing boat hulls including costs of installing marine Diesel engines by local competitive bidding (LCB) from Tunisian boatyards; and (iii) bids for marine Diesel engines to be installed in Tunisian hulls constructed in Tunisia by international competitive bidding (ICB). No indication is on record that these procurement procedures were seriously challenged by the Tunisian Delegation during negotiations. (e) It was agreed with BNT that for each 100 active fisheries sub-loans, one loan officer would be employed on a permanent basis. This seemed to be adequate at the time to cover loan appraisal of which the technical part had to be done anyway by the Fisheries Directorate and for loan recovery. No specific structure for loan recovery was proposed, BNT was believed to be capable of recovery through its own regular system; - .26- (f) The Credit Agreement contained under Category 3 an amount of US$70,000 for spare parts. This amount was never used for the import of spare parts, because spare parts were ordered and stocked by the engine supplier; and (g) Project management was to be in the hands of the Fisheries Project Division (FPD) newly to be created within the Fisheries Directorate, a department of the Ministry of Agriculture. FPD would hire for the initial Project operations three qualified experts with Tunisian counterparts: a fisheries specialist, a naval architect and a master-carpenter. C. Targets and Goals 2.09 Based on an estimated number of 130 trips of 48 hours duration each and on the use of 30 pieces of trammel net, the annual catch rate was expected to be 12.7 t in the first year of operation and 14.5 t thereafter. The SAR compared this to a catch rate of similar size boats in Tunisia of 9.6 t/year, based on 110 trips each of 36 hours duration using 25 pieces of trammel net. The increased production was explained by the increased number of nets (20%), the increased number of trips (18%), the increased duration of trips (33%), and the improved boat efficiency based on higher speed, improved handling at sea and more working space on deck (29%). Based on these assumptions, total envisaged catch of the new boats was estimated to be about 4,900 t of which about 2,650 t/year would be additional. The Project would also offer improved income opportunities to some 1,300 fishermen and generate about 300 new jobs. These figures were based on a crew of five fishermen per boat. Under the expected award of contracts to local shipyards, utilization of existing boat construction capacity would be used for about three years and generate a total of about 225 man-years of employment for supervisory, skilled and semi-skilled jobs. III. IMPLEMENTATION A. Effectiveness and Start-up 3.01 The Project was approved by the Board on September 14, 1971. The Effectiveness Date, originally set in the Credit Agreement for November 24, 1971, was twice postponed, first to February 24, 1972, and then to May 24, 1972, mainly due to the difficulties in the appointment of consultants, the creation of FPD and the promulgation of the FOSEP decree. The long search for an appropriate Project manager, naval architect and master-carpenter was conducted by the Fisheries Directorate with the assistance of IDA, UNDP and FAO and resulted finally in a contract with a French consultant. Accordingly, the Credit was declared effective on May 24, 1972. However, the signature of the contract did not lead to an immediate start-up of the Project. The envisaged Project manager resigned in July shortly before he was supposed to start work, and a new candidate had to be agreed upon in August 1972. The Project manager commenced his work in Tunisia in October 1972 and the naval architect made his first visit to Tunisia in the same month. Start-up of the major Project component, the construction and sale on credit of coastal fishing boats was however further delayed because the naval architect proposed and IDA approved modifications in -27 - the boat layout, and in the procurement procedures. He also proposed a stronger engine for the boats. Procurement could not then be started until necessary changes had been made in the Credit Agreement in June 1973. B. Revisions 3.02 Three revisions were made to the Credit Documents, two in 1973 and one in 1974. The first revision modified the procurement procedures set out in Schedule 3 to the Credit Agreement. The new version of the Project procurement procedures introduced in April 1973 and made official in June 1973 was limited to (a) international tender for marine Diesel engines only followed by (b) the international tender for hulls only. In this way, a decision on the engine award could be made prior to the determination of final specifications for the hull, so that the exact engine dimensions could be used as a basis for the hull design. At the same time, the new procurement procedures eliminated tendering for complete boats because a need for importation of complete boats was not foreseen. During the procurement process, it became evident that for price reasons foreign shipyards had no chance of competing against Tunisian shipyards for wooden boat hulls. The same would have been true if bids had been invited for fully equipped boats., 3.03 The second revision modified the disbursement percentage for consultants' services from 55% of total expenditure to 100% of foreign expenditures. At the beginning of 1973 BNT had pointed to a financing gap of 2.5% in the Credit Agreement which was corrected in a third revision in February 1974 when the Government's confirmation was received that section 3.09 of the Credit Agreement and Section 2.06 (b) (iii) of the Project Agreement would be revised to reflect an increase in Government supplementary loan funds from 5% to 7.5% of investment costs. 3.04 The Project Description, which specified that 335 boats would be financed, was not amended to reflect the reduction to 190 boats when the amount of expected cost overrun was known. C. Procurement Boat Engines 3.05 Following the revision of the procurement procedures, tender documents for the procurement of the boat engines under ICB were approved by IDA and bids were invited in May 1973. The tender called for the delivery of about 250 marine Diesel engines with a first option of 45 to 55 HP engines and a second option of 65 to 75 HP engines. The second option was included upon recommendation of the Project's naval architect who wanted to equip 25% of all Project boats with larger engines for operation along the northern coast to give them a higher speed and better maneuverability. On the recommendation of IDA, a price escalation formula and some other minor modifications were incorporated in the tender documents before they were published. Some 40 engine suppliers purchased tender documents and in June 1973, 14 bids were obtained. It took the Fisheries Directorate and FPD one full year until a contract was awarded to the lowest evaluated bidder, a foreign firm represented by a Tunisian firm. The original - 28- bid validity of 90 days had to be extended several times. The reason for this extraordinary delay was mainly a disagreement between the Fisheries Directorate and IDA regarding the contract award. A careful bid analysis was prepared by FPD which showed that after giving the 15% margin of preference allowed for in the Credit Agreement to the bid submitted by a Tunisian firm, its bid still was not the lowest evaluated bid. However, the Director of Fisheries recommended that the contract be awarded to a Tunisian company which assembles engines of Danish components. This bid was about 47% higher than the lowest evaluated bid. The Director of Fisheries presented the following arguments to justify his recommendation: (a) The technical specifications of the tender documents gave advantage to the lowest evaluated bidder. This argument was repeated later in a complaint of the Danish engine manufacturer to the Executive Director, although there was no evidence that could have substantiated this claim; (b) The price of the locally assembled engines could be reduced by excluding from the contract a number of optional extras which were required in the tender specifications. However, these extras belonged to the standard set in the lowest evaluated bid, but had to be paid for separately under the local bid; (c) The Tunisian fishermen were used to the marine version of the engine assembled locally of which about 1,200 were operating along the Tunisian coast. Furthermore, there were a number of stationary engines of this type whereas the engine made by the lowest evaluated bidder was known only in its version installed in a certain tractor, but not in its marine version; (d) The locally assembled engine was considered to be much sturdier, required less maintenance and furthermore had a manual starting system which, in case of a breakdown on sea, would become very important; (e) The fact that engines of the lowest evaluated bidder were not known to the Tunisian fishermen could lead to the Project boats becoming unacceptable to fishermen who were conservative by nature and would buy only an engine with established reputation; and (f) It would be in the interest of the Tunisian economy that the engines be assembled in Tunisia. IDA considered these arguments and came to the conclusion that they could either not be substantiated or did not carry enough weight to justify the award proposed by the Borrower. The Director of Fisheries was informed accordingly in October 1973. Still in correspondence to IDA he was reluctant to accept the responsibility for purchasing the engines. He proposed instead a compromise which suggested the cancellation of the tender and retaining the three most qualified suppliers in accordance with fishermen's wishes. This proposal was rejected by IDA as being unacceptable on the grounds that bids were responsive and therefore they should not be rejected. Moreover, the compromise would very likely result in an increase in engine and hence boat prices. This IDA position communicated to the Fisheries Directorate at the end of December 1973 was again discussed in the Fisheries Directorate for several weeks. Discussions were then extended to a higher level in Government and at that stage also the possibility of cancellation of the Credit Agreement was examined. Finally, the decision was reached to honor the ICB stipulations of the Credit Agreement and contract - 29- negotiations with the lowest evaluated bidder were taken up in February 1974 which led to award of contract on June 14, 1974. No order was placed for the 65 to 75 hp engines probably because of cost considerations. 3.06 The adequacy of the weightings given in bid evaluation to the factors cited by the Directorate of Fisheries is a debatable issue. Especially the repercussions of the fact that the engine offered by the lowest evaluated bidder was not known to fishermen and that the fishermen were not technically qualified to perform a demanding engine maintenance, were underestimated. On the other hand the justifications given by the Director of Fisheries, had they been accepted by IDA, would have resulted in admitting for competition only one engine brand that was known to fishermen in Tunisia and rejecting all others. If that was the Borrower's aim, this should have been proposed at Project appraisal, during negotiations or at the latest when the procurement procedures were revised in June 1973. Another possibility would have been for the Project to keep out procurement altogether and leave the choice of engine and boat type to the fishermen and their procurement to the shipyards. However, at that time (as now) it was considered that there was insufficient competition in the Tunisian market to accept such a procedure. Had an official request been made to IDA to permit a negotiated contract for boat engines, this probably would have been rejected because the primordial concern for Borrowers and IDA was and is today cost effectiveness brought about by adequate competition . As the experience shows the Project difficulties which developed later have to be attributed mainly to deficiencies in training of skippers and after-sale service. Under the circumstances, IDA could not have arrived at a different decision because ICB procedures had to be followed and because the engines proposed by the Borrower were at the time about D 1,000 more expensive than the engines proposed by the lowest evaluated bidder, which would have further increased the cost overrun of the Project boat price to 81% over appraisal estimate instead of the 57% which actually resulted (para. 5.02). The remark made in the completion report of the Commissariat G4n4ral A la P9che (CGP). "It would have been better if IDA had followed our recommendation that the engine offered by the Tunisian firm be selected. After all, it was well placed in the ranking of the bids received. Tunisian fishermen were already familiar with it and the manufacturer's after-sale service was well organized and well distributed along the entire Tunisian coast," represents an over-simplification. It may have been possible though to introduce at least some more flexibility into the tender documents by leaving room for fishermen's preferences and award contracts to two suppliers in the order of bid ranking, in this case the engines of the lowest evaluated bidder and other engines of an Italian manufacturer would have been selected, but not the engines of the Tunisian firm which ranked in 6th place. 3.07 The order placed with the lowest evaluated bidder in June 1974 called for delivery, storage and supervision of installation in a maximal period of three years of 250 marine Diesel engines with propeller shaft and propeller. The contract contained a price escalation formula for the second and third years. The number of engines on order was later reduced to 160 and in July 1975 increased again to 190. The first batch of 20 engines arrived in September 1974 and was installed in the boats in October 1974. Within one year, 70 engines were delivered, the remaining 120 engines in 1975 and 1976. - 30 - Procurement of Hulls 3.08 The procurement process for hulls started in February 1974. The reason for this delay was a change in the boat concept, i.e. the switch from a new boat design based on the Mediterranean boat forms but with major improvements which required a larger boat hull (see below) to an existing design also based on the traditional Mediterranean boat but with minor improvements only. Here the question must be raised whether the design aspect was adequately prepared and appraised. The preparation report provided a sketch of a Project boat prepared by an FAO naval architect, which was considered sufficient for appraisal. Detailed boat design was to be made by the Project's naval architect. He started design work early in 1973, still with the terms of reference to design a new boat. The major improvements of the new boat as compared with existing traditional boats were: (a) a transom stern instead of the pointed stern traditionally used in Tunisia in order to simplify construction and provide better working space on deck; (b) a wheel house arrangement; (c) an insulated fish hold under deck; (d) a manually operated winch for retrieving of nets; and (e) sleeping quarters for the crew. 3.09 According to these design objectives, the Project vessel was necessarily larger than traditional boats. When the design was nearly completed, a new cost estimate showed that the construction of the hull in accordance with this design would lead to total boat cost without gear of D 7,500 as compared to D 3,225 reflected in the SAR. This was in part the effect of inflation, but also of an underestimation of cost at appraisal. After a reevaluation of the boat design it was proposed by FPD and approved by IDA that the Project boat would have to be based on the design of traditional boats though retaining as many of the intended improvements as possible. In June 1973, it was still recommended to retain the transom stern, the wheel house arrangement and the insulated fish hold, but to drop the sleeping quarters and to use partially the local Eucalyptus wood instead of Iroko wood. Later on, also these improvements were sacrificed in order to keep boat cost as close as possible to the SAR estimate. A new boat layout was then prepared by a Tunisian naval engineer on the basis of the traditional forms. Detailed design and working plans with all necessary calculations were never produced, which led to the fact that the hull versions built by the three Tunisian shipyards are somewhat different from one another because each yard made some small modifications. Also here the question must be asked how could the delay and the waste of consultant's time in producing an entirely new boat design which later was rejected could have been avoided. It must be noted that the terms of reference of the Project's naval architect were not clearly defined and the alternatives in boat design and specifications were not fully considered at the preparation and appraisal stages. -31 - 3.10 In order to exclude those Tunisian shipyards - which in the opinion of FPD were too small to participate in the construction program - from participation in the bidding process, a prequalification request was published in the first half of 1973. This was an unusual procedure for an order of that magnitude. Prequalification is normally applied when large or complex contracts are involved which was not the case for an order of 200 boat hulls for coastal fisheries. The desired purpose could have been achieved much easier with a post-qualification procedure. Fifteen shipyards asked to be prequalified. Six requests arrived after the Closing Date and were disregarded. Of the six shipyards which were finally qualified, four were Tunisian, one Yugoslav and one German. It took another five months until March 1974 before bids for the boat hulls could be invited, because first a decision with respect to the award of contract for the boat engines had to be made. Finally in March 1974, bids from three Tunisian shipyards and from two foreign shipyards were received. The bid analysis showed that two Tunisian shipyards presented the two lowest bids. The third Tunisian offer was not considered because a bid bond had not been included. A Yugoslav and a German bid were much higher and played no role in evaluation. It was proposed to award contracts to both Tunisian shipyards, 105 hulls to the first and 75 hulls to the second with 30 hulls to be delivered in 1974, 80 in 1975 and 70 in 1976. 3.11 The recommendation was approved by IDA in April 1974 and contracts were awarded in May 1974. The contracts called for the delivery of the first fully equipped boats by November of the same year. The contract price for the hulls to be built by the two shipyards were D 3,800 and D 3,860 respectively. Prices were firm until end of May 1975. After this date, a price escalation formula was to be applied. Actually, however, both shipyards maintained their prices. Only the last ten hulls were 10% more expensive. The first contract was split in two parts to be executed in shipyards in Bizerte and Sfax. While the first had no start-up difficulties and completed a first series of 15 hulls by the end of 1974 as envisaged, the second supplier did not live up to his delivery promises because of problems with the relocation of his shipyard in the port area of Sousse. During this first construction period, IDA twice in September and November 1974 sent a naval architect from CP to supervise progress of work together with the counterpart naval architect of FPD, the Tunisian master-carpenter and the Section Chief for training and sales promotion of FPD. These visits proved to be extremely important, in that they gave valuable technical assistance to the shipyards and the Tunisian supervision team. The work done by the two shipyards of the first supplier was found more satisfactory than the quality of hulls built by the second supplier (para. 4.14), for a while it was even considered to reallocate part of the original order. Under constant supervision, things got better however, although weaknesses remained which needed permanent attention. About half way through the construction program, the speed of construction slowed down considerably because a lack in demand developed and the yards could not any longer accomodate the number of completed boats in their tiny private harbors. Upon insistance of Bank supervision missions, the shipyards were officially requested to take immediate measures to catch up with the delay in construction. Progress of construction was fastest in the first year where about 30% of all hulls were completed. The remaining 70% were then about equally spread over the following three years until the second half of 1978. - 32 - D. Demand for and Sale of Boats to Fishermen 3.12 Through the whole period of Project preparation and Project start-up, reports indicated that demand for boats was strong. When it actually came to signing sub-loan agreements, the demand was not overwhelming but cautiously optimistic, probably because of the natural prudence of fishermen, who first want to see how the boats would perform. Demand increased substantially thereafter as they recognized, that the first boats performed well. 3.13 Demand was strong in 1974 and 1975 but fell off sharply when word spread among fishermen that (a) the engines purchased required much more maintenance than the engines assembled locally with which they were familiar, leading to higher maintenance cost; (b) the electric system including the electric starter was not well suited to maritime use and led to breakdowns on sea; (c) the engine was also subject to mechanical breakdowns, mostly seizing of crankshaft bearings, possibly due to too small a discharge of the oil pump; and (d) the arrangement on deck did not leave enough working space aft. However, many applications were based on the expectation to get a boat in the form of a Government grant and could be used to make quick profits. Many were also made by people without any fisheries background and had to be discarded using the strict selection criteria adopted in the beginning of the implementation period. Indications of a declining demand appeared in January 1976, three months after the first mentioning of engine servicing problems. At that time, the Project management agreed to make a major sales promotion effort coupled with attempts to improve the after-sale service. But the frequency of engine breakdowns increased despite all endeavors to improve service facilities, demand then remained low up to the end. By mid 1976, the demand situation was so poor that the Fisheries Directorate started approving boat sub-loans for applicants who did not have the necessary professional experience. An additional step taken was the acceptance of a financing scheme under which the 10% equity contribution of the boat owner was financed by the Special Fund for Rural Development. Both measures may well have increased the number of applications but proved to be detrimental to the Project as a whole. It will take now drastic and determined measures to improve the performance of the existing boats in order to bring them to the production standard set at appraisal. 3.14 By the end of 1975, i.e. in the first year about 50 boats were delivered, followed by the delivery of 37 boats in the second year, 47 boats in the third year, 45 boats in the fourth year and 11 in the fifth year. The last delivery was made in December 1979. A total of 190 boats were delivered. IV. INSTITUTIONAL PERFORMANCE A. Government 4.01 The Project was implemented by FPD and BNT although several other Government agencies or departments were involved in fisheries with partly overlapping responsibilities (ONP, the Merchant Marine Department of the Ministry of Transport and Communications and the Department of Harbors and Aerial Bases of the Ministry of Equipment). - 33 - 4.02 In 1970 the contribution of the fisheries subsector to GDP at constant market prices was about .6%, and represented only about 4% of what the agricultural sector as a whole contributed. In line with this small contribution of the fisheries subsector to the economy Government paid little attention to its development. A first attempt to devote more attention to the fisheries subsector had been made with the institutionalizing of FOSEP in 1968, and the proposed Fisheries Project was supposed to be another step in the right direction. Government had finally recognized the potential in the fisheries subsector and decided to encourage investments to increase production. However, no decision was taken to reorganize the fisheries subsector because either the need for such reorganization was not perceived at the time or because in the wake of the sweeping changes in the orientation of economic policies in September 1969, no one wanted to introduce more changes. With respect to Project organization, the decision was taken to entrust the Fisheries Directorate with the Project management which, in retrospect, appears logical. No link was established between marketing arrangements and debt servicing. Although cooperatives no longer existed and thus it was not possible to implement the idea of marketing through cooperatives, at least an obligation of ONP should have been retained to cooperate with BNT in concluding marketing arrangements with the owners of Project boats in areas where ONP represented the only possibility for marketing of catch. Also BNT did not introduce firm arrangements despite their bad experience with repayments from individual fishermen under FOSEP loans which was specifically mentioned in the SAR. 4.03 At IDA's suggestion, a Fisheries Project Division (FPD) was established as main executing agency within the Fisheries Directorate. In accordance with the stipulations of the Credit Agreement, FPD was to be manned by: (a) A Chief of the Project, a qualified and experienced expert in inshore fisheries together with an experienced and qualified counterpart; (b) A qualified and experienced expert in naval architecture and a counterpart; (c) A qualified and experienced expert in naval carpentry; and (d) Three officers qualified and experienced in inshore fisheries operations. The appointment of all experts and their counterparts with the exception of the master-carpenter was a Condition of Effectiveness. 4.04 Already by the end of 1971, FPD had been created, the counterpart to the Project manager had been nominated and the decree modifying the FOSEP terms and conditions had been drafted and signed by the Minister of Agriculture. This quick action shows the commitment of the Tunisian Government to the Project. At the end of March 1972, the Director of Fisheries proposed to conclude a contract with a French consultant that included the services of all three experts which was a condition of effectiveness. IDA gave its approval in principle, but indicated reservations with respect to the reduction of the Project manager's stay in Tunisia from 24 months envisaged in the SAR to 18 months. The Director of Fisheries replied that in the Government's opinion a duration of 18 months for the Project manager would be sufficient, but that the duration of his contract could be extended if this would prove necessary. The contract was finally signed on May 22, 1972 and the Credit Agreement was declared effective on May 24, 1972. - 34 - 4.05 Staffing of FPD was complete in November 1972, after IDA had approved the CVs of the Tunisian counterparts to the Project manager and the naval architect. The original team appears to have worked well together, but staff in FPD was frequently exchanged for internal reasons. The expatriate Project manager stayed for 21 months of service after an extension of three months, just when the Project entered into its active implementation phase. He was temporarily replaced by his former Tunisian counterpart who was transferred soon after and another Project manager was appointed. He stayed on for only a year and was replaced by someone else. The master-carpenter to be provided by the Project consultant who was originally envisaged for the job was no longer available. In April 1974, the Director of Fisheries proposed to waive the employment of an expatriate naval carpenter with the justification that: (a) The number of boats to be constructed had been reduced from 335 to 190 which needed less supervision; (b) The two shipyards entrusted with the construction of the hulls had enough expertise and were employing well qualified expatriate carpenters; and (c) The Tunisian counterpart to the Project's naval architect would be perfectly capable of supervising the yards. 4.06 IDA agreed to this reasoning and the position was dropped. Unfortunately, in 1975, the Tunisian naval architect was transferred to ONP with the result that the Project was without naval architect during the actual construction period. Even the position of the Director of Fisheries was held by three different persons and was vacant for several months between new appointments. All this led to a lack of continuity in Project implementation which started already with the fact that none of the members of the Tunisian team that negotiated the Credit in Washington stayed on for the implementation of the Project. This lack of continuity had a negative effect on the Project, in that the actions undertaken by FPD were not part of a coherent Project action plan. Instead, they were undertaken when the need arose, usually as remedial measures as in the case of reinforcing the after-sale service of the engine supplier representative by a mobile unit staffed by FPD personnel, and often only upon insistance of IDA supervision missions. Instead of carrying out an elaborate training to take account of the lacking knowledge of most of the new boatowners in maintenance procedures for the engines and the navigational requirements of skippers, only short courses usually of less than half a day were given to boatowners when they came to the boatyard to take delivery of the boats. 4.07 Another negative aspect in the performance of FPD was the lack of cooperation with BNT. Upon recommendation of IDA a Coordinating Committee was appointed to discuss sub-loan approval, approval of supplementary loans for boat repair or fishing gear, and the recovery efforts by BNT, which met fairly regularly in the first three years of Project implementation but then the meetings became more sporadic. Coordination of efforts by Fisheries Directorate and BNT agents was imperative from the beginning to discuss the merits of sub-loan applications and to keep a joint close contact with new boatowners regarding recovery arrangements, but this coordination was generally lacking, particularly in the field. - 35 - 3. Banque Nationale de Tunisie 4.08 The institutional role of BNT in Project implementation did not change at appraisal or during impLementation as have other components. BNT was the only institutional source for agricultural credit in Tunisia and had also some experience in lending to the fisheries subsector since the establishment of FOSEP in 1968. It was therefore only natural to retain BNT as the channel for the funds made available under the Project. 4.09 The terms and conditions for on-lending to fishermen were changed several times however. Under the FOSEP regulations, sub-loans to fishermen were made at 5% interest for a duration of 6 to 10 years. Although a fisherman's equity contribution of only 5% was envisaged at that time, BNT restricted its lending to 80% of investment costs so that actually 20% of equity had to be financed by the fisherman. In 1970, however, the FOSEP regulations were modified and a 15% subsidy participation of the Tunisian Government was introduced and the interest rate was lowered to 3%. During negotiations of the Project, agreement was reached that Project sub-loans should be made at 6% interest for a duration of eight years including three months grace period with an equity contribution of fishermen of 10%, and a Government subsidy of 15%. In order to avoid competition between FOSEP funds and IDA funds, the FOSEP decree was modified as a Condition of Effectiveness to reflect identical on-lending conditions. The preparation report suggested that similar to the handling of FOSEP loans, BNT would not assume any default risk and that therefore the margin of 2% BNT received on FOSEP loans oustanding at the end of the year should be lowered. To get BNT more interested in a proper loan administration, the spread for BNT was increased to 3% under the condition that BNT assume 50% of default risks, the other 50% to be borne by Government. At appraisal it was calculated that under these arrangements and with a loan from Government for 15 years including a grace period of 5 years, BNT would be left with a sizeable surplus for reinvestment for the first 11 years which would be sufficient to cover the default risks. Despite the initial experience with the repayment performance of fishermen who had received loans from FOSEP, BNT was apparently optimistic in the beginning of Project implementation that the forecast profit could be attained. As the situation stands now with 80% of arrears, it seems very likely that BNT will accumulate substantial losses from the Project operations. Given the extremely low recovery rates it will take considerable time and effort to avoid larger write-off of receivables. 4.10 It was BNT's task to appraise the creditworthiness of applicants who were already technically approved by FPD and to make the final decision whether the loan request should be granted or rejected. This did not present major problems to BNT's Technical Division and for the first two years after Project signature, BNT was actively and enthusiastically involved in the promotion of the Project. The subsidiary loan agreement concluded between the Ministry of Finance and BNT was drafted by BNT and signed three months after credit signature. BNT recommended in February 1973 opening an account funded with D 100.000 from FOSEP funds, from which BNT could prefinance the installment payments to be made to the engine supplier and the shipyards in accordance with progress of construction. The recommendation was approved by the Ministry of Finance. 4.11 BNT seemed to loose much of its initial enthusiasm when it became apparent that the proper supervision of the Project repayment performance needed a much heavier personnel input. Cooperation with FPD deteriorated and BNT - 36 - apparently considered that the reason for inadequate repayment performance was the lack of cooperation of regional fisheries personnel with their branch personnel. Similar arguments were used by the Fisheries Directorate which complained that BNT did not cooperate with its regional personnel. Only in 1978 prior to negotiations for a Second Fisheries Project could BNT be convinced to recruit seven loan/recovery officers and to post them in BNT branches along the Tunisian coast. 4.12 Since the beginning of 1980 when these measures were implemented, the coordination between the Fisheries Directorate and BNT has improved, but it is not adequate. For example, the Fisheries Directorate and the present Fisheries Authority (CGP)/l , favored more rescheduling of loans and the approval of supplementary loans for engine repairs or fishing gear. This was a point of disagreement during the Project because BNT rigidly followed its loan regulations which do not allow approval of sub-loans for a client as long as any amounts from previous loans are overdue. Only in the last year of Project implementation when the maintenance situation had deteriorated very rapidly, BNT officials were ready to discuss loan rescheduling or the approval of supplementary loans when they were faced with the necessity of losing the sub-loan principal totally or coming to an amicable agreement with loan defaulters which would allow them to resume operations. 4.13 Another deficiency was discovered only very late during Project implementation. BNT requested the boatowners to insure their boats against damage from accident and loss, and did not agree to boat delivery before the insurance policy was deposited in the credit files. Unfortunately, the insurance premium was paid only for the first year in many cases, and BNT did not supervise whether fishermen paid their premiums also for subsequent years. This led to the fact that from 12 boats which have sunk only six were covered by insurance. It would be advisable to include the insurance premium for the duration of the sub-loans into the loan amount so that BNT would be able to pay the premium directly to the insurance company. Another possibility would be to halt issuance of fishing licences until the insurance premium is paid. C. Boat Yards and Boat Maintenance 4.14 The performance of the boat yards was excellent. At preparation, without a detailed assessment of the boat construction capacity it was stated that the existing yards would be capable of constructing 600 boats in five years. One of the two main and most qualified yards later did not participate in the tendering process. The SAR specified that sixty-five boats were to be built in year 2, 125 in year 3, and 145 in year 4, which were subsequently found to be reasonable estimates of construction capacity. This was the first time in Tunisia that standardization in boat construction for such a large series had successfully been adopted. Moreover, the Project boat was attractive to fishermen because: /1 Commissariat G6n6ral A la PAche. -37 - (a) Construction time was much shorter than for boats constructed in accordance with the individual wishes of boatowners; (b) The firm prices Eor the Project boat became more and more attractive as time passed by and inflation rose; (at the end of the Project period, the boat was 23% cheaper in constant prices); and (c) The financial terms were also liberal because there was no monetary ceiling to the 15% subsidy or the sub-loan as in the case of boats financed only by FOSEP. 4.15 The boat layout finally adopted had the following objectives: (a) Eliminate all unproductive investments, i.e. sleeping quarters and wheel house, but retain improvements which would increase productivity, i.e., increase the number of trammel nets; (b) Improve quality of construction as compared to traditional constructions which would contribute to the boat solidity; and (c) Introduce modifications only if they had no influence on the investment cost which had been estimated to have a minimal rate of return. 4.16 Contracts with the twQ shipyards were signed in May 1974, and immediately thereafter the shipyards started construction under close supervision of both the Project's naval architect and the naval architect from CP. In the first two years a series of recommendations were made, which pertained to different aspects: (a) Construction material to be used; (b) Quality of construction; (c) Modifications in boat design; (d) Boat testing and delivery to the boatowners; and (e) Maintenance problems. Quality of construction was a problem in the beginning in the shipyard in Sousse because the shipyard used timber which was insufficiently seasoned resulting in cracking of construction members. Another problem was incorrect cutting of timber and too large clearances between construction members. Upon request of FPD, the shipyard reinforced a considerable number of frames, beams and keelsons. 4.17 Over time also the construction performance improved. All shipyards learned to construct simultaneously series of four to six boats in various stages of construction progress. The speed of construction dropped in the last two years of Project implementation because the yards just did not have enough space to have completed boats moored in their tiny harbors. - 38 - 4.18 A thorough inventory of all Tunisian shipyards at the time of Project completion showed that today boat building capacity of all Tunisian shipyards combined has grown to about 227 crafts in the following categories: - Trawlers 19 - Lamparos 18 - Boats of 9 m length 132 - Boats of 14 m length 58 Shipyards employ the following staff: - Master carpenters 40 - Qualified carpenters 121 - Skilled workmen 137 - Caulkers 42 - Apprentices 187 This was brought about at least in part through the encouragement and the incentives provided by the Project, but also by increased demand from fishermen which shows that the subsector is considered by fishermen a viable source of income. Maintenance 4.19 Most problems that developed during Project implementation including specifically the repayment performance of fishermen, can be traced back either directly or indirectly to poor maintenance performance. It required a determined and well coordinated effort from engine supplier, boatyards, FPD and fishermen to organize a properly functioning after-sale service, but that effort was never made. Fishermen did not respond to training proposals possibly because nobody compensated them for the loss of productive time. FPD started well by devising a four-day training program which would have covered all aspects of boat operation and boat maintenance, but did not follow through when reluctance of fishermen became apparent. The Tunisian representative of the firm supplying the engines, who had indicated in his bid that there was a whole series of their representatives along the coast, was initially of the opinion that these would be more than sufficient to serve the 250 marine Diesel engines expected to be sold under the initial contract. This turned out to be a serious misjudgement probably because the firm's representative had no experience with fishermen and their attitude toward technical progress. When problems developed he was not inclined to make large additional investments to create a service network in the major ports that was directly operated by him. The independent local agents did not want to buy and stock the spare parts needed for the marine version of the engines. Upon request of fishermen, local agents were ready to order specific spares from Tunis but only after having received a down-payment to be sure that the parts would be purchased after their arrival. During the first two years after boats had been commissioned, it happened several times that spare parts were not available even in Tunis and had to be flown in causing fishermen to lose weeks of precious operating time. Another factor that aggravated the situation was the fact that no factory trained mechanics were available. The firm's representatives were usually only sales outlets for parts. Therefore the mechanics who were asked by boatowners to do repairs and servicing of engines were usually automechanics, not properly qualified to work with marine Diesel engines. - 39 - 4.20 Very soon after the first Project boats had been sold, the engines started to develop problems. The most frequent problems with the engines were seizing of crankshaft bearings probably due to a malfunctioning of the oil pump, but the exact cause has not yet been determined. Still more frequent were problems with the electrical system, which was not really designed for maritime use, and not sufficiently well protected from sea-water. When this led to breakdowns of the electric starter, the fishermen complained vigorously that no manual starting device was available. Moreover, the locally produced batteries were of poor quality and insufficient to ensure lighting of the boat at night during fishing operations, and subsequent starting of the engine. As a consequence, fishermen were afraid to switch the engine off at sea which led to much higher fuel consumption. A number of measures was taken to improve the situation, among them the establishment of a mobile workshop manned with a mechanic and a naval carpenter from FPD, and an electrician from the firm's representative in Tunis. This mobile workshop made regular trips to all fishing ports and undertook repairs and servicing on a no-charge basis except for spare parts which had to be paid for. All the measures taken however were directed toward curing of symptoms and did not attack the root of the problems. It would have been FPD's responsibility to organize and supervise the following program: (a) To familiarize the new boatowners with the operation of the new engine and its maintenance and service requirements; (b) To install a dense network of outlets for spare parts in the fishing ports where spare parts would be sold at reasonable prices; (c) To have trained mechanics available in these places who would be capable of servicing engines and making small on board repairs; and (d) To have workshops in the ports with the largest boat population which would be qualified to undertake larger repairs and engine overhauls. The statement in the CGP report: "Despite the efforts made and goodwill shown by the firm's agent, after-sale service was far from adequate, boats were sometimes out of commission for weeks because spare parts were not available" is not exaggerated, but makes no mention of FPD's role and responsibilities. 4.21 At the time 'of the completion mission in September 1980, the situation of the 190 boats was as follows (Annex 2): - Number of boats financed 190 - Boats sunk 12 - Boats immobilized 27 - Boats under repairs 7 - Boats in operation 144 The boats that have sunk were either involved in accidents or were caught in heavy weather. According to later information received from CGP, the number of immobilized boats had been reduced somewhat by January 1981. 4.22 Measures to get the immobilized boats back into operation consist mainly of the following actions: - 40 - (a) Individual negotiations with each of the boatowners which should result in agreements to have the boats repaired with complimentary loans by BNT, debt rescheduling or as a last resort sale of boats to new owners; (b) The creation of a special fund from which boat repairs can be prefinanced; and (c) Close supervision of the repairs to be undertaken in order to avoid similar problems. It is expected that it will take some time to complete these measures. Average repair costs have been estimated to be D 2,800. The commissioning of the 10 new fishing ports in 1982 and 1983 presently under construction with slipways and workshops may be an important step toward a permanent improvement of future operation which will reduce the risk of operation damage. D. Consultants 4.23 Originally three consultants were to help implement the Project: a Project manager, a naval architect and a master-carpenter. The Project manager and the naval architect started work in October 1972, more than one year after Credit signature (para. 3.01). The Project manager was a well qualified fisheries expert who did a satisfactory job and was commended for his performance by the Fisheries Director. Due to the delays in agreeing on the boat design and in procurement of boat engines, his assignment was terminated five months before the first boats were delivered to fishermen, in July 1974. 4.24 The naval architect was also a qualified professional who prepared a good boat design which through no fault of his own, was rejected as being too progressive and too costly. It is understandable that he was unhappy about the decision taken but he nevertheless continued his work and contributed to the preparation of the tender documents for hulls. Also he left in mid-1974. He was replaced by a Tunisian naval architect who was transferred after one more year and then left the Project without a naval architect. V. FINANCIAL PERFORMANCE A. Costs 5.01 Total Project cost at appraisal was estimated at US$3.1 million of which US$2.0 million, or about 64% were expected to be foreign exchange. This estimate was based on the construction of 335 boats estimated at US$2.69 million including customs duties and contingencies, supporting services at US$365,000 and spare parts at US$72,000. The actual total cost was US$2.93 million, slightly lower because of the currency fluctuations of the various currencies involved in the financial transactions. The following table shows a comparison of estimated and of actual costs: - 41 - Total Cost Item A>praisal Actual Appraisal /c Actual - (D'000)------- --.----(US$'OOO)----- I. Fully equipped boats 1,410.4 /a 1,309.9 /b 2,686.7 2,868.4 /e II. Supporting services 191.6 26.7 364.9 61.8 If III. Spare parts 38.0 0.0 72.4 0.0 Total 1,640.0 1,336.6 3,124.0 2,930.2 Foreign Exchange Costs Item Appraisal Actual Appraisal /c Actual --------('000 D)------ ----- (US$'000)------ I. Fully equipped boats 954.0 /d 795.7 1,817.2 1,938.2 /e II. Supporting services 59.3 26.7 69.9 61.8 /f III. Spare parts 36.7 0.0 112.9 0.0 Total 1,050.0 822.4 2,000.0 2,000.0 /a For 335 boats including D 111,000 customs duty and D 143,400 price contingency. /b For 190 boats including about D 50,000 for individual gear sub-loans including duties and taxes. /c At the rate of exchange appraisal date of US$1.0=D 0.525. /d for 335 boats including D 94,400 price contingency. 7e At average rate of exchange compiled through Project implementation of US$1=D 0.411 /f At average rate of exchange compiled through Project implementation of US$1=D 0.433 - 42- 5.02 The comparison becomes meaningful only when unit costs for fully equipped boats are compared. At appraisal the unit cost was estimated to be D 4,210 (US$8,020) including contingencies and customs duties. The average actual unit cost turned out to be about D 6,630 (US$15,100) about 57% more than the expected price in terms of Dinars or 88% in terms of Dollars. Two reasons exist for this discrepancy: (a) a serious underestimation of unit costs at appraisal which was due to detailed engineering not being available as a basis for cost estimates and (b) much higher inflation than expected. The difference between the two percentage figures is explained by a 38% depreciation of the Dollar against the Dinar. B. Disbursements 5.03 Disbursements started in the first half of 1974, two years later than estimated at appraisal. The last disbursement was made in January 1980, three years later than estimated at appraisal. The reason for the delays have been discussed in previous chapters. More detailed information on disbursements is presented in Annex 3. C. Recovery Performance 5.04 The background that led to the extremely poor recovery performance has been described in the earlier chapters. No firm recovery arrangements were proposed in the SAR, and it was left entirely to BNT to devise suitable procedures. The Project Agreement contained a covenant that the duration of sub-loans would be eight years including three months grace period, and that repayments of principle and payments of interest should be made in monthly installments of a uniform annual amount through a deposit account at a BNT branch office. Hence sub-loan agreements specified the amounts due each year and requested sub-borrowers to pay installments of about 10% of the annual amounts during ten months of the year. This system proved to be an unrealistic approach which probably was based on BNT's experience with large farmers who have current accounts with BNT and are in need of annual short-term credit to finance agricultural inputs. This was not the case for artisanal fishermen who sold their catch to any wholesaler or organization which, in their opinion, offered them a fair price and.was conveniently located. Hardly any of the Project boatowners had a current account with BNT. The only way to achieve good recovery results would have been to oblige the fishermen to agree to a deduction from the value of his landed catch wherever and whenever the catch was sold. This, of course, would have required an intensive supervision in all fishing ports where the catch was disembarked or agreements with ONP, wholesalers, municipal markets or local agents of the Fisheries Directorate to act as collection agents for BNT. In other words, a highly decentralized system of supervision needed to be established. Instead, BNT did not make any attempt to change its usual system of centralized loan and recovery administration. Under this system, computer processed notices were mailed to clients shortly before maturity dates through postal services. In many cases they did not reach the addressees because fishermen had moved to other working areas. In case of non-payment, reminders were again sent from Tunis. After a certain time, the files were transferred to the Legal Department where legal proceedings were initiated. - 43 - 5.05 The first sub-loan contract for Project boats was signed in December 1974. In June 1976, an IDA supervision mission reported BNT's non-compliance with the stipulations of the Project Agreement requesting monthly installment payments. BNT was informed accordingly but contended that under the contract model, fishermen were indeed requested to make monthly payments. At the same time, BNT admitted its inability to supervise fishermen's compliance with these terms. In February 1977, for the first time recovery figures as of December 31, 1976 were examined, which showed a recovery rate for loan principal of 35%. It became apparent at that time that payment record was specifically bad in Zarzis, in the far south, a situation that has not changed to-date. The recovery rate for loan principal fluctuated from 35% in December 1977 to 58% in May 1978, 39% in August 1978, and finally to 11% at the end of September 1978. These figures were based on computer printouts received from BNT and were not in all cases correct as later experience has shown. On occasion of the negotiations of a Second Fisheries Project, BNT agreed to undertake a special effort to improve the recoveries and signed a formal commitment to employ seven recovery agents, and to place them into the most important fishing ports along the coast. BNT also committed itself to undertake all necessary actions to reach a 30% recovery rate by 6/30/79, 40% by 12/31/79, 55% by 12/31/80, 65% by 12/31/81, 75% by 12/31/82 and 80% by 12/31/83. The recovery rate was to be understood as the ratio of loan principal and interest received to loan principal and interest due and overdue. BNT recruited the seven recovery agents in the first half of 1979, and made big efforts to improve its contacts with boatowners and its cooperation with the local agents of CGP. As result, recovery rates improved somewhat but much slower than had been hoped in 1978. At the time of the completion mission, BNT prepared for the first time a complete recovery statement that included interest accrued and paid and amounts under litigation. Annex 4 contains the details of this statement and shows separate recovery rates for loan principal and interest and for the normal portfolio and the portfolio under litigation. The total recovery rate at that time was a disappointing 20%. 5.06 Besides the lack of firm repayment arrangements and BNT's weak efforts to recover loan interest and principal, there are a number of other reasons which have influenced the attitude of boatowners. The first and principal reason is connected to the technical difficulties which have plagued the Project boats from 1975 onwards. Many boatowners had extended down-times due to breakdowns of the electric system or the engines which led to serious revenue losses and inability to repay their debts.. When BNT started to reinforce its collection efforts, it became apparent that another group of boatowners had used the technical deficiencies of the Project boats in general as an excuse to avoid payment. This group of boatowners must receive special attention from BNT agents and it is in this group where examples could be set by CGP which would lead to a discontinuation of subsidies on fuel and as a further step, the non-renewal of fishing licences. As experience has shown, legal proceedings are not an appropriate way to avoid defaults. A court order is needed to foreclose on the mortgaged boat and a second court order to permit its sale in an auction. In a first experience of this type, it took about two years until a boat could be auctioned in the port of 4ahdia. The only bidder was a relative of the previous owner, who acquired the boat for D 650. If no bid at all is received under Tunisian legislation, BNT would become the owner automatically. Any other measure that can be taken short of legal proceedings is therefore preferable. - 44 - 5.07 Yet another group of boatowners was made up of a large percentage of those 108 fishermen who had received financing of their 10% equity contribution from the Rural Development Fund. Most of these have perceived that contribution as a Government grant rather than a loan, and some, in absence of serious loan recovery efforts by BNT, have gradually begun to regard BNT loans as grant, too. These fishermen had not contributed any equity toward the acquisition of their boats, and were therefore less inclined to spend the necessary amount of money to keep them in good operating condition. They were also not too concerned about losing their boats in legal proceedings after loan default since they had not invested at all in them. When they ran out of money for boat repair, they frequently abandoned the boat not feeling that they had lost anything. They felt no sense of responsibility for the boat nor did they have any sense of ownership in it. It would be of little advantage to grant supplementary loans for boat repairs to them, it would rather have to be envisaged to arrive at a negotiated agreement that would transfer ownership to a more responsible person. 5.08 Finally, the question must be asked whether this course of events was unavoidable. Certainly, the Project management did not have any influence on the events which happened before Project appraisal and which led to the decline of the cooperative idea. In spite of these events, it would have been possible ' however to design institutional arrangements for the collection of loan funds. If this would have been accompanied by a formal commitment of BNT and the Fisheries Directorate from the beginning of Project implementation, better results could have been obtained. It was unfortunate that at least in the first years of Project implementation, this commitment did not exist. Later it proved extremely difficult to correct previous omissions in both establishing a properly functioning after-sale service and in setting up a working recovery system. VI. PROJECT IMPACT AND RATES OF RETURNS A. Production and Financial Rate of Returns for Boats 6.01 The exact production of Project boats from the start of operations until the date of the completion mission, September 1980, can only be estimated. Fishermen are free to disembark their catch in any fishing port and are also believed to disembark outside of fishing ports and sell directly to hotels or other middlemen. Part of their catch is also distributed among crew members as part of their remuneration. Moreover, the statistical services of the Fisheries Directorate were very thinly staffed and had no means of transport to cover all landing points on the entire coastline. They existed in the larger ports, but there normally during official working hours only. Thus only a certain percentage of the actual catch was recorded and in areas where there were no port facilities, it was not recorded at all. The production estimates therefore have to be based on estimates of CGP (see Annex 1), estimates made at the appraisal of the Second Fisheries Project in 1978, and at the time of the completion mission in September 1980. The portion of unrecorded catch in line with CGP figures has been estimated to be about 40%, although it may be larger in cases where the recorded figure is very small. For the remaining lifetime of the Project boats, new projections were made which were based on the assumption that in view of the expected improvements of after-sale service brought about by the installation of 13 new slipways and repair yards, 10 in new ports and 3 in existing ports along the Tunisian coast in 1982/83, the situation for the Project boats will slowly improve. - 45 - 6.02 At appraisal, it was estimated that Project boats would catch 12.7 t in the first year of production and 14.5 t thereafter which would result on a financial rate of return of 31%. The quantity of catch was calculated on 130 trips of 48 hours each. Both the quantity and the duration estimate were based on the assumption that boats would function properly and that through extensive training, fishermen could have been incited to work more. Because of the difficulties with boat operation, actual catch was lower than projections, at least in 1979, the only year for which figures were furnished. CGP indicated in its report that in 1979, 138 boats had declared their catch. If the 27 boats are excluded that work mostly in coral and sponge fishing, the average catch per boat comes to 6.5 t, about half of what was estimated at appraisal. This gives a production average of boats active in fishing operations only and excludes inactive or sunken boats. 6.03 Most of the boats nevertheless reached acceptable financial results because fish prices increased in a way that made fishing more attractive than foreseen at appraisal. The production and financial return figures for operational boats in the three ports with the highest number of boats and other ports are as follows: Port of No. of Boats Estimated Estimated Switching Attachment Production Financial Value of per Boat Based Rate of Benefits /e on 1979 incl. Return /d Undeclared Catch Zarzis 43 8.9 t 30 19 Bizerte 27 3.4 t /b 1 -19 Tabarka /a 27 1.0 t 31 20 52 kg /c Other Ports 47 6.1 t 21 15 Total of estimated operational Boats 144 /a Main benefits from coral fishing. /b The catch figure for Bizerte appears to be inconsistent with mission observations. A better estimate may be 5.0 t which would result in a 13% FRR. /c Coral. 7- 31% at appraisal. 7e Percentage by which bertefit value can be reduced before reaching a financial rate of return of 10%. - 46 - 6.04 The projected situation of the 46 boats which in September 1980 were not operational or part time out of commission because of technical difficulties is presented below: Category of Boats No. of Production Estimated Switching Boats Estimate Financial Value Rate of Return Immobilized 27 /a 16 13 Weak Production performance 7 2 t -2 -38 Sunken boats insured 6 /b 7 -6 Sunken boats uninsured 6 lb -89 -78 Total 46 /a 6 t in year 1, 3 t in year 2, 0 in year 3 and 4, reaching 8 t in year 7. /b 6 t in year 1, 3 t in year 2 then 0. 6.05 In 1979, the annual catch of all boats is estimated to have been 780 t. After immobilized boats would resume operations in 1982/83, it is expected to increase to about 960 t per year as compared to an appraisal estimate of 4,858 t per year at full development. The full catch was allocated to economic benefits. In the PCR calculations, it is assumed that all catch was incremental in comparing the "with Project" and "without Project" situations, but about 50% only of this incremental production is actually additional to the total production level at the beginning of the Project. In 1979, total fish production was 75,278 t and total production of coastal fisheries was 28,952 t. The Project catch therefore added about 1.0% and 2.7% respectively to national production. 6.06 In calculating the catch value, coral fishing of Tabarka boats has been included. It is not known how many boats were involved in shrimp fishing in the Gulf of Gabes, for them the catch value and hence the FRR would be substantially higher. In the global picture this would be balanced by the fact that some boats catch octopus which has a lower value. The projected figures do not take account of the expectation that due to the construction of new ports in the south fishing trips will be possible at any tide and the number and duration of trips will probably be substantially increased by 1983. B. Social Returns 6.07 Social returns have come close to expectations. The SAR estimated that on the basis of 335 boats being built, 300 new jobs would be created and that the Project would offer improved income opportunities to 1,300 fishermen. Addtionally, it was estimated that about 225 man-years of employment for supervisory, skilled and semi-skilled jobs would be generated in shipyards. Although only 190 boats were constructed and sold, the estimated actual number of - 47 -- new jobs created is close to the SAR estimate, the reason being that more young crew members from fisheries training schools were employed on the boats than originally envisaged. Of the 190 new boatowners, 57 had owned boats before. Another six had been skippers. The remaining 127 with a few exceptions had been crew members on coastal fishing boats including rowing and sailing boats, or on lamparo boats or trawlers. It is estimated that only about 20% of the Project boats replaced other obsolete boats because it is usual in Tunisia to keep boats afloat and in operation as long as possible. Taking into account the sunken boats and the likelihood of some boats always being immobilized, it is estimated that about 125 jobs for skippers and 125 jobs for crew members have been created in accordance with the following evolution: Project boats sold 190 Sunken boats 12 Boats replacing obsolete ones 38 Boats that may be out of operation at any one time 15 Remaining boats 125 Skipper jobs created 125 Out of three crew members, either mechanic or deckhands recruited from fisheries training schools in first employment 125 Total jobs created 250 It is further estimated that about 160 man-years of employment for skilled and unskilled workers have been created in the shipyards during the four-year construction period. 6.08 It is difficult to assess the development of incomes. At appraisal, crew remuneration was estimated at D 1.344 in 1971 prices. This corresponds to D 2.930 in 1977 prices, an increase of 118% due to inflation and higher fish prices in real terms. The actual crew remuneration in 1977 prices for boats operating in Zarzis is estimated at D 1.900, about 35% less than the projected figure. For boats operating out of.other ports the crew remuneration is lower. The amount of D 1.900 (about D 650 for the skipper, D 530 for the mechanic and D 360 each for two deck hands) still compares favorably with the minimal agricultural wage rate (SMAG) of D 1,2/day, which in 1977 would have led to an annual pay of about D 300 for a permanent worker and D 120 for a seasonal worker. The absolute poverty level in rural areas was estimated to be D 240 for a family of six, still much lower than wages earned in employment opportunities offered by the Project. C. Economic Rate of Return 6.09 At appraisal, the economic rate of return (ERR) had been estimated to be 21% for the total Project under the assumption that labor cost was fully costed in the catch share system and using a catch rate of 14.5 t per year priced at a conservatively estimated producer price of D 245t. The ERR was estimated at 36% with a shadow pricing of labor. -48 - 6.10 In ERR calculation for the PCR, the catch volume and the number of Project boats specified in para. 6.03 have been applied. The following assumptions have been made: (a) all prices are expressed in constant 1977 prices (1977 is the mid year in the five-year cycle of boat sale); (b) average producer prices for white fish and for coral were applied; (c) alternatively, the same prices were increased by the margin between retail and producer prices which is estimated to be 30% after deduction of cost and taxes related to transport, handling and marketing. These margins are benefits directly attributable to the Project. Since margins actually applied by wholesalers and retailers can hardly be controlled, they may be higher in many cases; and (d) the labor wage rate has been assessed at the minimal agricultural wage rate for the skipper and one crew member and at half the minimal wage rate for the two remaining crew members. 6.11 Under these assumptions, the ERR is 43% without and 78% with marketing benefits. This surprisingly high result is based entirely on the development of the average producer price for demersal fish. At appraisal in 1970, this was estimated to be D 245/t. Using the Tunisian consumption price index, this would correspond to a price of D 355/t in 1977 prices. This compares to an actual price in 1977 of D 467/t and of D 556/t in 1980 expressed in 1977 constant prices. As explained in the SAR of the Second Fisheries Project, the price is expected to grow by 6% annually in real terms until 1985 and by 4.5% until 1992. In addition, the PCR calculation includes the benefits from coral fishing at Tabarka which was not envisaged at appraisal. Total benefits can be reduced by 60% and 46% respectively before the ERR would reach the cut-off point of 10%. VII. BANK PERFORMANCE A. Project Preparation and Appraisal 7.01 The Project idea as described in the SAR is simple and convincing. However, the appraisal was undertaken at a time when drastic changes in the economic system of the country had just taken place. This led to the abandonment of firm catch marketing and repayment arrangements and as a result the financial success of the Project was compromised. The appraisal mission then did not pay enough attention to the consequences of this, by not raising the missing link between marketing and recovery as an issue and proposing an appropriate course of action. 7.02 Another deficiency was the fact that at appraisal no detailed engineering for a boat accepted by Government and fishermen was prepared that could have served as basis to calculate Project cost more accurately. The result was a serious underestimation of unit costs and big delays in construction start-up due to the need to prepare a new acceptable design. The increase in unit costs led to a reduction of Project boats from 335 to 190. Part of the problems could have been avoided had a qualified naval architect been attached to the appraisal mission. -49 - 7.03 The other main issues have been identified at appraisal. The organization and management aspect was properly covered. A project of such limited scope was not the appropriate vehicle to propose a complete reorganization of the subsector. It was appropriate to propose a Project unit which would operate within an existing organization, the Fisheries Directorate, instead of creating an entirely new one. The proposed on-lending conditions were adequate and it was justified to request a draft decree aimed at creating equal lending terms for FOSEP and IDA lending as Condition of Effectiveness. Catch rates in retrospect may seem optimistic because they have not materialized. In reality, however, such rates can be achieved with boats that are properly maintained and operate satisfactorily. A precondition to achieve such performance is adequate training of skippers and crews. Such results have been proven by some of the Project boats, though only as an exception to the rule. The training needs for boatowners were also assessed at appraisal and a covenant was incorporated in the Credit Agreement and Project Agreement, without specifying, however, the duration of the training, thus leaving the door open for a decline in training efforts. Under IDA's procurement guidelines, the procurement procedures laid down in the SAR and the Credit Agreement were satisfactory. No experience was available in the subsector with similar programs which would have helped predict the technical difficulties that arose later. If it was the intention of the Borrower to promote local industry for engines and hulls, a case could have been made by Government at the time of appraisal for a negotiated contract. B. Supervision 7.04 IDA's supervision effort was adequate. Supervision missions spent 52 man-weeks in the field for 18 missions, on average 6.5 weeks per year to which an equal time of work in Washington must be added. Also the continuity of mission composition and their professional background was satisfactory. IDA acted swiftly in all matters which required its attention, but did not in all matters arrive at decisions which in retrospect can be considered as correct. Some supervision efforts and attitudes deserve to be mentioned: (a) IDA should not have agreed to the departure of the Project manager after 21 months at a time when he was needed most. Also IDA's no objection to waive the recruitment of the consultant master-carpenter put a heavy burden and responsibility on IDA staff and led to the result that, during construction time, nobody in FPD was really qualified for supervision of boat construction. IDA had to put in several visits of a CP naval architect: because of this consultant's absence. This contributed greatly to the fact that hull construction went relatively smoothly and improved quality of construction; (b) The importance and serious draw-backs Qf not having a retention of money from catch values at boat landing places was recognized by IDA missions very early in the implementation period. Repeatedly, the first time in 1977, the idea was discussed with the Fisheries Directorate to introduce this procedure but. it was explained by both the Fisheries Directorate and BNT that the proposed procedure was not practical, required too much staff and was therefore too costly. Only in 1978 was the idea partly realized by negotiations with the newly founded service cooperatives. But at that time fishermen had already become used to the idea of not - 50- having to pay or that it was easy to avoid or delay payments. It will take now years of constant persuasion and education of fishermen to introduce financial discipline. A properly functioning system of retention would necessitate the cooperation of all wholesalers including ONP which still markets more than half of all catch of coastal fishing boats, and would have to be implemented with the close collaboration of cooperatives; (c) Supervision missions tended to be too optimistic regarding expected progess in Project implementation and hence the disbursement schedules were nearly always unrealistic; (d) The decline of the training efforts was tolerated and it was predicted that the payment of equity contributions from the Rural Development Fund would not have any detrimental effects and that therefore no action was necessary; and (e) The Project was correctly and repeatedly declared a problem project, first in December 1972 when it became apparent that the boat design would have to be altered until July 1974 when the number of boats to be financed was finally reduced to 190. From 1974 to 1978, the Project rating was nine times 2 and twice 1. In December 1978, it was again declared a problem Project for a second time when the recovery rate dropped to an unprecedented level. Major problems remain with the Project and it would be advisable to continue monitoring the technical and recovery aspects with the supervision of the Second Fisheries Project. VIII. CONCLUSIONS 8.01 The drastic change in agricultural policies in 1969 in Tunisia, the insufficient preparation of the design for the Project boat, the introduction of new boat engine for Tunisian fishermen, the serious deficiencies in after-sale service and training of fishermen, the lack of commitment of the Fisheries Directorate and BNT to take appropriate measures toward the establishment of a functioning loan recovery system, were all factors contributing to poor Project performance. However, most boats will continue to operate for another ten years, and against the bleak production and recovery performance, that is due to the technical difficulties and BNT's and Government's leniency, must be compared the excellent economic results of the Project. However, at the present, it looks as if Government and BNT, the two institutions which each bear 50% of the default risk, have to expect some financial losses. 8.02 The long implementation and disbursement period of slightly over eight years, three years longer than expected at appraisal made the Project a costly one for IDA in terms of staff input. The problems with the boat design and the introduction of a new engine turned it into a Project which was extremely difficult to manage. The Project objectives in terms of production and financial returns for fishermen and boatowners set out at appraisal have not been reached, and it seems clear that appraisal targets will not be met. On the other hand the Project contributed greatly to increasing and improving the local boat building capacity. The Project also laid the groundwork for the Second Fisheries Project and helped perceive the need for more port infrastructure and a reorganization of the fisheries subsector. - 51 - 8.03 The following general conclusions can be drawn: (a) Fisheries Projects are difficult to execute and need careful supervision. In view of the wide range of expertise needed, which is usually obtained by using consultants because of a lack of in-house expertise, appraisal and supervision tend to be more costly to the Bank for fisheries projects than for other types of projects. In the interest of improving the scope and continuity of Bank involvement, it would be helpful to increase the number of fisheries specialists in the Bank, preferably at the Departmental level; and (b) As confirmed already as general policy, detailed engineering as a basis for exact cost estimates should be available at appraisal or shortly thereafter in order to avoid misjudgements on costs and benefits to be expected and to avoid implementation delays which equally lead to reduction of benefits.  - 53 - ANNEX 1 Page 1 FIRST FISHERIES PROJECT (CREDIT 270-TUN) PROJECT COMPLETION REPORT Report By Commissariat A la P@che (CGP): September 1980 I. Project Data 1. Title . Fisheries Project 270-TUN 2. Beneficiary : Ministry of Agriculture (Fisheries Dept.) 3. Lender : IDA 4. Date of Signature . September 24, 1971 5. Cost (as initially estimated) a) External Financing Dinar - 67.5% of amounts to be lent to fishermen (US$1,637,000) 859,425 - Auxiliary services: Fees to 3 experts (foreign exchange component), training and vehicles (US$113,000) 59,325 - Spare parts (boat engines)(US$70,000) 36,750 - No specific allocation (US$180,000) 94,500 1,050,000 1,050,000 b) Local Financing Public Funds - 15% subsidy on purchase price of boats (FOSEP) 184,163 - 7.5% supplementary loans (FOSEP) 61,387 - Expatriate personnel: Compensation (in dinars) for 3 experts recruited under private agreement No. 72 TUN 104 21,780 - Operating expenses: . fuel, maintenance, repairs, taxes, insurances for vehicles 12,670 280,000 280,000 Private Funds - Contributed by beneficiary fishermen themselves: 10% of their loans 122,775 1,452,775 1,452,775 - 54 - ANNEX 1 Page 2 6. Conditions; - Proceeds of the credit to be onlent by the State to National Bank of Tunisia (BNT), which will handle contacts with final beneficiaries.Terms and conditions to fishermen are to be the same as those allowed by FOSEP (75% of investment figure over 8 years at 6% interest; self-financing 10%; Government subsidy 15%). 7. Expected duration : From 1971 to 1976 8. Description and Purpose of Project Study, design, financing and building and inshore fishing boats to be ceded to final beneficiaries. II. Remarks on Formulation of the Project 1. The initial Project objective was the design of a new type of inshore fishing boat modeled on the traditional Tunisian fishing craft but which would allow greater advantage to be taken of inshore fisheries resources. The original plan was that the proceeds of the credit would finance the construction of 335 boats 10.5 m overall length and fully equipped. Finally, however, it was possible to build no more than 190 boats, the-obvious conclusion being that Project cost was underestimated from the outset. 2. An error in the whole conception of the Project was the fact that under the terms of the credit agreement procurement proceedings for engines and hulls were to be conducted simultaneously. However, one type of engine differs from all other types in certain essential ways which are precisely the ones that dictate the design characteristics of the hull that is to accommodate it. It is only after the award of the "engine contract" that a naval architect can be commissioned to design a fully suitable boat. 3. Estimates of the duration of the Project failed to take account of real potential demand, which meant that the termination date had to be postponed three times. The Project was eventually completed three years late, in December 1979. III. Project Execution A. By Department of Fisheriest (1) Choice of Craft The type of craft chosen was that of the traditional "martigane", which has an overall length of 10.40 m (9 m at keel). Engine and hull were chosen after two separate bidding procedures. After numerous exchanges of views with IDA, it was decided to award the engine contract to the lowest evaluated bidder. For the boats themselves, contracts were awarded to two Tunisian boatyards. - 55 - ANNEX I Page 3 (2) Demand for Boats (a) Orders Several campaigns were organized in order to apprise fishermen of the advantages they could expect from the Project, which were the following; - Advantages flowing from supervision of construction by the technical authorities (elimination of risk for buyer, craft built in strict accordance with specifications); - Advantages flowing from mass production (quick, easy maintenance and repairs, standardization of hull and engine parts); and - Advantages as regards delivery dates (assurance of delivery at time promised; much shorter delivery periods than could be offered in the case of isolated orders). These campaigns had the desired effect, as demand for Project boats was heavy, at least initially. After the first craft was delivered, there was a noticeable drop in demand, ascribable to the following: - Frequency of engine breakdowns; and - Difficulties encountered in manipulating fishing nets, as fishermen had been used to much more space in the aft sections of their boats. Subsequently, when credit was made available through the rural development program (PDR), there was a slight improvement in demand. It must be recognized that without that additional source of credit, Project execution would have been still slower. (b) Deliveries Deliveries by GovErnorate were as follows; Governorate Total No. No. of Boats with Self- of Boats financing under PDR Medenine (Zarzis) 49 44 Gabbs 18 16 Sfax 13 9 Mahdia 8 - Monastir 4 6 (sic) Sousse 4 - Nabeul 14 13 Tunis 12 - Bizerte 45 12 Jendouba (Tabarka) 23 8 Total Boats Deli.vered 190 108 - 56- ANNEX 1 Page 4 As the above table shows, the Governorates in the northern and southern regions were the most affected by the Project -- a reflection of the Government's concern to develop fishing where untapped potential existed. (3) Fishermen's Reactions to the Project; Problems with the Engine Fishermen found it much more difficult to get used to the engine of the lowest evaluated bidder than to the Tunisian engine, which had been fitted in other inshore craft. The reasons were these: - The engine model of the lowest evaluated bidder features an electrical starting system previously unknown to the fishermen; - It requires far more maintenance than the Tunisian engine. In combination, these two factors meant that boats were very often out of commission. Steps Taken to Solve Engine Problems In an attempt to avoid prolonged periods out of commission and breakdowns at sea, it was proposed that a mannual starting system using a crank handle be fitted to each engine. Since beneficiaries could not afford the additional expense involved, the Department of Fisheries reached agreement with IDA that the cost should be covered from the allocation for purchase of nets. It was to be left as an option. At the same time, arrangements were made with the help of various specialists to provide hull and engine mintenance and assistance inspections. Printed material on practical engine maintenance procedures were given to fishermen when they purchased their boats, which was then explained and enlarged upon in Arabic by a team of instructors. Despite these measures, however, the problems persisted. After having to deal with a succession of breakdowns, some fishermen were ready to replace the engine with a Tunisian engine if the arrangement could be included under the terms of the credit. But as those interested had become insolvent from the bank's viewpoint, such replacements cold not be financed, with the result that the boats were either abandoned or turned over for a consideration to operators who were in a better financial position. It should also be pointed out that, despite the efforts made and goodwill shown by the agent, after-sales service was far from adequate; boats were sometimes out of commission for weeks because spare parts were not available. From the foregoing, it is clear that choice of a marine engine must be based primarily on the following two considerations: - The occupational qualifications of the user. As far as main- tenance and functioning are concerned, a new engine should resemble the old engine the fishermen are accustomed to as closely as possible, and all the more so if it has given proof of its reliability. 57 ANNEX I Page 5 - The provision of really effective after-sales service, monitored by the authorities and covering all fishing ports. In conclusion, it can only be said that the engine chosen gave the Project a bad reputation among fishermen. It would have been better if IDA had followed our recommendation that the engine offered by the Tunisian firm be selected. After all, it was well placed in the ranking of the bids received. Tunisian fishermen were already familiar with it and the manufacturer's after-sales service was well organized and well distributed along the entire Tunisian coast. B. Project Execution by BNT: (1) Financial Statement Drawing down of proceeds of IDA Credit 270-TUN: ----Us$--- - Financing boats and equipment (D 795,679) 1,938,191 - Technical assistance (D 26,748) 61,809 Total disbursed (D 822,427) 2,000,000 Computer Statement -----D---- - Loan funds authorized (IDA - FOSEP) 888,829 - Subsidy 177,766 - Loan funds utilized (IDA - FOSEP) 888,829 - Subsidy 177,766 - Unpaid 151,076 - Not due 533,204 - In litigation 129,624 On comparing total loan funds utilized and total loan proceeds drawn down, there is an overrun of D 4,268. (2) Observations (a) BNT Project execution is carried out by three different divisions of the Bank: - Studies and Research Division; - Loan Management Division; and - IBRD Division. There appears to be a lack of coordination among these three sections of BNT. Had BNT appointed one staff member to follow up Project execution activities in the three divisions concerned, there would have been a considerable improvement. ANNEX 1 Page 6 (b) In the early years of the Project period, the repayment record was poor, as follow-up activies -- the joint responsibility of BNT and the Fisheries authorities -- were inadequate. (c) In 1979, collection agencies were set up in the main ports so that the activities of Project boats could be monitored. This has led to a notable improvement in boatowners' repayment status. One should not be too optimistic, however, as a certain percentage of boats are inactive and will remain so long as no solution is found to the engine problems discussed above. IV. Impact of the Project (a) Inshore Fishing Yields In 1979, yields from Project boats were as follows: Governorate Yield (kg fish) Jendouba (Tabarka) 15,626 967 kg coral Bizerte 97,948 193 kg coral Tunis (La Goulette) 11,695 Nabeul (Kelibia) 32,877 Sousse - Monastir 2,140 Mahdia 14,123 Sfax - Gabes 55,917 Zarzis 194,897 425,223 1,160 kg coral These figures show declared production; they do not include the part of the catch taken for their own consumption by boatowners and crew members or any undeclared catch. If autoconsumption is assumed to be 42% of the total declared catch (the usual assumed figure) and the non-declared catch to be approximately 120 metric tons (for 32 boats), total production would then be roughly 720 metric tons, or 4.5% of the entire national inshore catch. It is quite clear that the catch obtained by the Project boats falls short of the projected figure. The main varieties of fish caught are the following: Red mullet (rouget) Sole Porgy Dentex Rascasse Grouper Perch Gray mullet Chien de mer Octopus Cuttlefish Bonito Shrimp Squid Pageot Figures on volume of catch by variety are not obtainable. ANNEX I Page 7 (b) Income of Beneficiaries Given the reduced size of the catch recorded, total income has not been as high as expected. In addition, a large percentage of earnings has gone toward repairing the frequent engine breakdowns, some beneficiaries still remaining in debt to the spare parts supplier and repair shops on top of their debt-service obligations to BNT. This stage of affairs has made it necessary for a number of them to engage in other occupations in order to meet their repair bills. Only the boats fishing out of Zarzis and Tabarka have produced a reasonable income for their owners. There is some case for believing that the low volume of the catch has been offset by increased prices for fish. (c) Fish Prices (Bulk) The impact of the catch obtained by the Project boats on fish prices has been limited. There has been a significant price rise in recent years ascribable mainly to increased demand. Simply by way of information, the accompanying table shows changes in fish prices in the Governorate of Bizerte, which has the largest complement of Project boats: Kind of Fish 1976 1977 1978 1979 ------------------- D/t----------------- Daurades 2,090 2,115 2,340 2,485 Dent4s 1,170 1,175 1,440 1,440 Loups 2,165 2,330 2,610 2,080 Marbr6s 820 980 1,115 1,315 Meroux 905 880 1,030 1,300 Mullets 1,230 1,540 1,485 1,670 Pageots 930 1,090 1,310 1,490 Pagres 1,285 1,280 1,520 1,705 Rougets 1,670 1,750 2,005 2,375 Soles 2,505 2,560 2,650 3,290 Crevettes 2,960 3,350 3,720 5,780 Langoustes 4,430 4,470 5,090 7,000 ANNEX 1 Page 8 V. Conclusion It must be recognized that at this stage a considerable number of boats are not operated as well as they might be. In addition, production figures show that boats which are the same in all other respects have catches that may vary in volume by anything up to twice or even four times. This is explained for the most part by the frequence with which some boats break down and by the lack of qualifications found among owners. It is essential that the breakdown problem be solved urgently so that boats out of commission can be returned to the fleet. If this is to be achieved, all parties will have to cooperate: owners, the supplier of the engine, the National Bank of Tunisia (BNT) and the Commissariat GAngral A la PAche (CGP). Even though the Project has not accomplished expected production goals, at least not in 1979, it has brought the following benefits: - jobs have been created; - a new type of inshore fishing boat has come into use; and - inshore fishing has increased in areas that were formerly untapped (the northern and southern coastal zones). - 61 - TUNISIA Project Completion Report First Fisheries Project (Credit 270-TUN) Situation of Boats Financed under First Fisheries Project as at September 11. 1980 /a Ricapitulation de l'Etat des Bateaux IDA Pfche au 11 Septembre 1980 /a Total all Boats without /b lb /b Ports / Known attach- Grand Tabarka BizerEe La GouletEe Kelibia Sousse Monasuir hahdia Sfax Gabes Zarzis ToLal Lous menL /Zarques TuLal Ports Recherchges Number of Boats Financed / Nombre de Bateaux financds 24 45 4 14 5 7 6 12 17 46 180 10 /c 190 Transfer / Transferrd +4 -4 - - - - - - - - - - - Sunk / Could - 6 1 1 2 - - - 1 1 12 - 12 Immobilized/ Immobilisd 1 4 - - 3 4 3 1 8 2 26 1 27 Under Repair /En panne - 4 1 - - - - - 2 - 7 - 7 In operation / En travail 27 27 2 13 0 3 3 11 6 43 135 9 Ic 144 /a Information based on CGP report and mission findings / Renseignements bas6s sur rapport du CGP et recherches de la mission /b Not visited / Non visiti /c Information not confirmed for seven boats / Renseignement non confirmg pour sept bateaux January 1981 (185E) - 62 - ANNEX 3 TUNISIA PROJECT COMPLETION REPORT FIRST FISHERIES PROJECT (Credit 270-TUN) Disbursement Table (US$ million, cumulative) Period ending Appraisal Actual Actual as % Estimate of Estimate 6/30/72 0.03 - 0 12/31/72 0.03 - 0 6/30/73 0.24 - 0 12/31/73 0.48 - 0 6/30/74 0.79 0.02 3 12/31/74 1.13 0.10 9 6/30/75 1.54 0.40 26 12/31/75 2.00 0.80 40 6/30/76 - 1.20 60 12/31/76 - 1.30 65 6/30/77 - 1.40 70 12/31/77 - 1.50 75 6/30/78 - 1.70 85 12/31/78 - 1.78 89 6/30/79 - 1.91 96 12/31/79 - 1.96 98 1/17/80 - 2.00 100 - 63 - PRWJECT C0MPLETION 0EPORT I "APPORT D'4Ck"EVEME DU PROJET AN2 4 Fr- F-.herLes Pr.j.ec (C 270-TUN) I Pr..-.e Pro-et de Pche (C.270-TUN) 0r Lon Pr-c~ ad Inte re. t a... 0f S.p,e 11. 1980 Reouvrement., des Pre2., Prore2p.l *0 Ineret 6rr829 7. 11 Septemb,., 7960 99T Branch 69.../ No No,.and Pur- Lo.n PrincoIpl/ Principal jntere.t / Intéret Tota1 Prancipal7and2onterst / Total Pr,ncipa.etInt4ret Agenc. 8NT, No. pose of Sub- D..bur,.ed/ 0.12.., Due/ Repay.ent/ Reovery Fallen Due0/ Repayment/ Reovr Pallan Duel7 *.payment 6eco,r * 0221,ooo...,2R.20P....7 S - -------------------------------------- -- -------- ----------- -------- D ------------ 12 30 £ 136.921 65,686 22.090 32 30.50 12,074 40 96,190 13,64 4 u00erte B8aerte 22 I Gb 525 525 0 0 60 0 0 585 0 0 Subt0a2 B r 12 32 137,646 66,211 21,090 32 30.064 12.04 40 96775 364 34 Soustotal B92erte L.togaton Bgsrte 02 6 7599 . 207 22,050 22,030 4,300 20 4,319 214 5 27,076 4.794 18 Cootnteux 8,2ert. Total B28erte 12 37 (35+ 20) 60.200 88,968 25.660 03 54.83 12.088 35 12,852 37,948 3L Total 81a2rte J.rb. 23 2 B 9,352 2.930 133 5 1,660 100 24 4,6j0 535 12 J.0rb L,.gation Jirb. 23 2 4.605 4,605 100 4 1,625 090 24 6,299 587 9 CooÉnoux J.rba T..2 J.rb. 13 3 14.020 7.605 332 1 3,303 090 24 70.908 1,122 70 Toota j-rb. 1.hdia 26 5 0 2.625 2,625 1.40 53 214 L46 68 2,939 1,56 54 Mahd.. L12-atton M.hd. 16 5 33,34 23,354 5,483 23 7009 2,634 03 32.363 0,120 23 C etu ad 0otal ~hd6,. 26 10 (50 59) 25.979 25,979 6,983 26 7,225 7 07M- - -,- - ,. 2 Total Mahdt. 19 93 99 59,925 18,954 4,694 25 9.692 2.203 37 20,645 7,897 29 90.. Sima 19 6 4 2.028 2,390 407 17 227 80 35 2,627 480 19 Sfaa Total S2.. 19 09 (138 + 60) 62,643 21,344 5,101 24 8.929 3.203 37 30,262 8.384 28 To.-a f-. S-ue. 0 99 4,720 2.362 0 0 2.394 0 0 3,754 0 0 Souss a-sse 20 3 0 2,575 1,575 224 4 188 112 60 7,763 336 19 sousse Ba62,2.2 9..s-e 20 4 6.295 3,935 226 6 1.,582 12 7 5,517 336 6 soustoaI Bo... L.oxgation .oss 20 2 9.357 9,350 6 0 3.944 0 0 13,301 0 0 ont....*.0 So-e Tot.2 Bouas. 20 6 (39 + 30) 15.652 23,292 224 2 5.526 112 2 08,828 336 2 Tot.a Soo.. T.b"rk 22 22 1 102,868 35.966 20,70 60 15,537 10.411 67 51,4.3 32.073 62 T.bark, T.bar6. 22 6 0 3,276 2,689 0,280 40 243 120 49 2,932 .309 45 T"bak. Total 0.ar. 22 28 (228 * 6G) 105.954 38,635 22.890 59 25.80 10.531 67 54,415 33,421 61 TotalBrk Ur. 25 35 B 262,602 60,792 4,085 7 32,360 5,251 16 95,215 9,336 10 Zara2. 2arate 25 8 0 6,232 3,420 209 9 370 30 6 5,090 509 8 Zar.0. u,b2ota Z... . 25 43 166.603 44,212 6,363 7 32.730 5.282 16 96,942 9,645 20 Sous.otal 0. L,ittion Z.r.ia 25 8 36,693 36.693 2,496 7 10,945 442 4 47.638 2,93M 6 cotent0eux 0. 0o2a1 29rsi. 25 51 (239 8) 203,596 100.905 6,859 68 43.675 5,724 13 744,580 72,583 9 Total 0. 26 16 74,708 32,70 3,000 9 16,805 2,848 17 49,585 5,848 22 bes G.be. 26 1 4 562 282 0 0 99 0 0 380 0 0 Gabee .ub0120Gebe 26 17 75,20 33,62 30 9 08.904 2,848 17 49,9 5,848 10 0 Liitoation Geb. 26 2 9,393 9,393 0 0 2.608 0 22, 0otal 4.bes Tat,a Gb,, 26 19 (IBB . 207 84,663 42,454 3,006 7 19.520 2.848 25 61,974 5.648 94 T-l Nebel 20 7 99 4,74 585 0 0 260 0 9 0 o I.bol Total .Ie 27 B 4,714 589 0 260 4 0 949 0 6 Total #ebe,l T29 8 4.707 2.765 0 0 1.063 0 0 2,828 0 0 To.al 2 oo. 29 i 4,707 2,765 0 0 2.063 1 2,828 I 0 Total7Tunt Goebalta 30 0 6 9,622 2,995 0 0 7.400 0 0 4,396 0 0 Gombal, Total rombali 30 2 8 9,622 2,995 0 6 1.401 0 0 4,396 0 0 Tot.l Gomba.1. . 1.tt. 3g 0 8 35183 16,141 5,830 5 6,54 2,306 34 22,997 9,.30 35 L.Goulette L,a Ga2.tt. 38 14 4 0,266 6.573 1,402 22 73 270 29 7,247 1,612 22 2, 7l2t btoal tla o*lett. 38 21 38,529 22,656 7,234 02 7.582 2.576 33 30,138 9,756 32 9999t.9,2. Goulette L,,gt:n La Go22.6t. 38 4 18,333 ',333 0 0 5,464 0 0 23,827 0 0 C0,6en62.u. 646440.26t Total L ~ G.12l29t 38 25 (120 • 14) 56.862 40,989 7,234 28 13.076 2.516 19 54,065 9,750 18Toa La Goulette Moasur 42 3 8 14,054 5,598 0 0 3,43 234 4 9,329 7 8onastar Nea2t0r 62 2 0 7.050 788 120 15 80 30 38 868 =32 07 6m t,r ubt.ta .o.a.6r 42 5 15,204 6,658 720 2 3.527 164 2 70,797 284 5 ~ous.a06. t'ou.at, ,L"tisati.n Men 2t. 42 2 9.4 9,45 0 0 7.653 0 0 2,087 o o Cote,t»ou Moa. Tota2 M star 42 7 (IB • 20) 24,638 26,720 120 1 6,64 164 3 22,284 24 1 Total 9onast0r 8* J.be2 499 9 B 37,3y8 15.202 1,694 77 7.570 7.064 27 22,564 0,960 23 6. J.be2 Total a. 3ebe2 499 9 * 37.378 15.212 7,64 27 7,370 0.286 17 22,584 2.940 13 Total 6. Jebe Eo..r Essaf 55 2 1 9,395 5,282 580 77 2,670 737 27 1,952 1,320 7 Koor gesar 6aans E...8 55 74 4 7.320 7.311 2,504 34 656 340 52 7,967 2,844 56 2ao.r E...7 Toti2 ... f 55 16 (28 . 144) 16,706 2.593 3,092 25 3.320 1,472 22 15,919 4,164 26 to, E..a0 T~baulo 63 2 8 4.675 2,622 0 2 7,433 0 4,355 0 0 T Ib.lb T-a T.a b.. 63 1 B 4,67 2,922 0 0 3,43 0 0 4,355 0 TaT. -b.ba .2.b,. 60 20 8 47,591 75.925 .236 4 6,930 7,755 25 20,925 0,594 70 8el,ba. TtlK bha 68 0 B 47.391 15,985 606 4 6,930 7,758 25 22,917 0.592 70 Tot.2 Kelab2a Tota2 8092. 454 5 774.54 286.339 63,456 22 138.376 40,39 30 424,878 103.839 04 Nombr total d b u Tota2 G4.r1 61 4 30,930 28,117 7,523 27 2,877 1.006 07 00,998 0.590 18 8omb.re0ota des 0al.et G,ad Tota1 3 + G 275 745,444 314,356 70,979 23 747,250 41,452 29 455,606 112.431 25 Tot. d.. . , . e . . ,0 Total L,1 gat 50 733,996 733,996 12,746 10 38.,03 2,680 7 172,599 I5,42 9 Tot*7oatnt07 G8AD T0L 245479,440 48.2 83,725 19 79,853 242 25 6 25 27 857 20 29.,b7. b - for f.lly .quipped b.-t I Prlt. pour h, . quIPPé L ' (-Losf f-n . 6 . - -nlIy/ Prt*0our*0ngLn depO hØ (316 or 340b)  긷