Tunisia - Second Agricultural Credit Project

Document of The World Bank FOR OFFICAL USE ONLY RePw N. 5401 PROJECT PERFORMANCE AUDIT REPORT TUNISIA SECOND AGRICULTURAL CREDIT PROJECT (LOAN 1340-TUN) December 28, 1984 Operations Evaluation Department This document has a restricted distribution and my be used by recipients only i. the perforgmance ef their official duties. Its contents may not otherwise be disclosed withot World Bank aoiaon. WEIGHTS AND MEASURES Metric System ABBREVIATIONS BNT - National Bank of Tunisia (Banque Nationale de Tunisie) BNDA - Agricultural Development Bank ERR - Economic Rate of Return FOPRODI - Special Fund for Industrial Promotion FOSDA - Agriculture Development Fund (Fonds Sp6cial de Dgveloppement Agricole) FRR - Financial Rate of Return GA - Guarantee Agreement GOT - Government of Tunisia IBRD - International Bank for Reconstruction and Development LA - Loan Agreement MOA - Ministry of Agriculture OED - Operations Evaluation Department PCR - Project Completion Report PPAM - Project Performance Audit Memorandum PPAR - Project Performance Audit Report RS - Regional Services SAR - Staff Appraisal Report SCMV - Small Farmer Association SONAMO - Agency Provi'-ng Tractor and Harvesting Services (Socift6 Nationale de Motoculture) STIL - Tunisian Company for Milk Production (Socifti Tunisienne de l'Industrie LaitiZre) FISCAL YEAR OF BORROWER Calendar Year FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT TUNISIA - SECOND AGRICULTURAL CREDIT PROJECT (LOAN 1340-TUN) TABLE OF CONTENTS Page No. Preface ............................................................ i Basic Data Sheet ---- ............................................... ii Highlights ......................................................... iii PROJECT PERFORMANCE AUDIT MEMORANDUM I. STMARY .................................................. 1 A. The Project...................................1......1 B. Implementation Delays ..................... ...2......2 C. Investments and Disbursements by Category............. 3 D. Project Impact and Benefits............... .......... 6 E. Other Results.............................. ....7......7 II. ISSUES ...................................... ............ 9 A. Credit Policy .9.....................................9 B. Institutional Development .................. .......... 12 C. Monitoring and Evaluation and Measurement of Project Benefits. ........................................... 13 D. Conclusions. ......................................... 15 Annex 1 Comments from Borrower .-..................... .... .....17 PROJECT COMPLETION REPORT I. INTRODUCTION ............................................. 27 II. PROJECT FORMULATION ........................................ 28 III. PROJECT IMPLEMENTATION ...................................... 29 IV. PROJECT IMPACT ............................................... 42 V. LESSONS LEARNED AND CONCLUSIONS .......................... 46 Annex 1 BNT's Summarized Balance Sheets, Income and Expenditures Statements, and Sources and Application of Funds ......... 49 Annex 2 Disbursements ........................... .............. 52 MAP 18707 This document has a restricted distribution and may be used by recipients only in the performance of their ofrcial duties. Its contents may not otherwise be disclosed without World Bank authorization. - I - PROJECT PERFORMANCE AUDIT REPORT p TUNISIA - SECOND AGRICULTURAL CREDIT PROJECT (LOAN 1340-TUN) PREFACE This is a performance audit of the Tunisia Second Agricultural Credit Project, for which Loan 1340-TUN in the amount of US$12.0 million was approved on November 23, 1976. The Closing Date was extended from December 31, 1980 to December 31, 1982. However, the Bank agreed to honor all disbursement requests received -mntil June 30, 1983. The last disbursement was made on April 5, 1983. As total disbursements on June 30, 1983 amounted to US$11,901,360, the balance of US$98,640 was cancelled on that date. Considerable delays in disbursements were also experienced by the First and Third Agricultural Credit Projects. The First Agricultural Credit Project (Loan 779/Credit 263-TUN), in the amount of US$8.0 million, was approved on July 1, 1971, and closed on July 22, 1978, almost three years after the original Closing Date [see Project Performance Audit Report (PPAR), First Agricultural Credit Project (Loan 779/Credit 203-TUN), OED Report No. 2497 dated May 11, 1979]. The Third Agricultural Credit Project (Loan 1885-TUN) in the amount of US$30.0 million, was approved on June 26, 1980. As of April 9, 1984, disbursements were about 30Z of the expected level. The Banque Nationale de Tunisie (BNT) has been the Borrower and executing agency for all three projects. The audit report consists of an audit memorandum (PPAM) prepared by the Operations Evaluation Department (OED) and a project completion report (PCR), which was prepared by Agricultural Division II of the Europe, Middle East and North Africa Regional Office and is based in part on a completion report prepared by the Borrower. The audit memorandum is based on a review of relevant Bank documents, including minutes of project presentation to the Board and interviews with Bank staff who have been associated with the project. Proiect documents reviewed include: the Appraisal Report (No. 1133-TUN) dated October 22, 1976, the President's Report (P-1932-TUN) of November 3, 1976, the Loan Agreement dated December 17, 1976, the PCR of March 1984, and memoranda on project issues as contained in relevant Bank files. A draft version of this report was sent to the Borrower for com- ments on September 19, 1984. Comments which were received have been taken into account in the report and are reproduced as Annex 1 to the PPAM. On the basis of this procedure, the audit finds that the PCR gener- ally covers the main features of the project experience. The audit memoran- dum elaborates on problems of credit policy, institutional development, moni- toring and evaluation, and the related topic of measuring project benefits. PROJECT PERFORMANCE AUDIT REPORT TUNISIA - SECOND AGRICULTURAL CREDIT PROJECT (IAN 1340-TUN) BASIC DATA SHEET KEY PROJECT DATA Appraisal Actual or Actual as I of Estimate Estimated Actual Appraisal Estimate Total Project Costs (US$ million) 26.3 24.6 /a 94 Loan Amount (USS million) 12.0 11.9 99 Daze Board, Approval 06/76 11/23/76 - Date Loan Agreement - 12/17176 - Date Effectiveness 10/76 07/19/77 - Closing Date 12/31/80 12/31/82 - Economic Rate of Return (2) 34 33 /b - Financial Rate of Return (2) 11-50 neg.-46 - Number of Direct Beneficiaries Ic 1,250 5,825 465 Number of Subloans 957 4.063 4Z5 CUMULATIVE DISBURSEMENTS FY77 FT78 FY79 FT80 FY81 F82 FT83 Appraisal Estimate (USS million) 2.7 8.3 11.4 11.8 12.0 12.0 12.0 Actual (USS million) - 1.7 3.6 6.7 9.4 10.3 11.9 Actual as 2 of ESciMAte 0 20 31 57 78 86 99 Date of final disbursement: April 5, 1983 MISSION DATA Date No. of Nandays Specializations Performance Types of (mo./yr.) Persons in Field Represented/d Razing le Trend/f Problemsa Identification 01/75 - - - - - - Preparation JI975 l 5 205 A,C.E,I,E - - Appraisal 11/75 4 84 A.E.F.I - - - Sub-total 289 Supervision 1 03/77 2 22 li C,E 2 2 N,T Supervision 2 06/77 1 12 1i C 2 2 1I,T Supervision 3 11/77 2 3s 71 C,E 2 2 NT Supervision & 05/78 1 9 It E 2 2 H,T Supervision 5 11/78 2 8 C,E 2 2 H,T Supervision 6 0/79 1 20 A C 2 2 H,T Supervision 7 03/81 3 15 A,C,E 2 1 H,T Supervision 8 10/81 1 5 E 2 2 H.P Supervision 9 07/82 1 7 E 2 2 a.P Supervision 10 12/82 2 7 /1 E.F 2 2 H.p Sub-total 143 Total 432 OTHER PROJECT DATA Borrower: Banque Nationale de Tunite (BNT) Executing Agency: Banque Nationale de Tunisle (BN) Fiscal Year: January 1 - December 31 Name of Currency (Abbreviation) Dinar (0) Currency Exchange Rate: Appraisal Year Average: US51.00 - D 0.429 Intervening Years Average: USS1.00 - D 0.430 Completion Year Average: US$1.00 - D 0.591 Preceding Project: Follow-on Project: Name First Agricultural Credit Third Agricultural Credit Loan/Credit Number Loan 779/Credit 263-TUR Loan 1885-TUN Loan/Credit Amount (US$ N) 5.0/3.0 30.0 Date Board Approval 07/01/71 06/26/80 PPAR Number 2497 - Ia D 10.6 million at average exchange rate during the Intervening years if US$1.0 - D 0.430. 71 See PPAN pars. 24. 7c Defined as recipients of sub-loans in the case of mall and commercial farmers; in the came of sub-loans to SCNVs anl cooperatives, there was an average of 50 beneficiaries per sub-loan. /d A - Agriculturist; C - Credit Specialist; E - Economist; P - Financial Analyst; I - AgroIndustries Specialist; L - Livestock Specialist. Ie 2 - moderate problems. if I - improving; 2 - stationary. N M - Managerial; T - Technical; P - Political. h FAO/CP. 71 Combined with supervision of other project(s). - i1i - PROJECT PERFORMANCE AUDIT REPORT TUNISIA - SECOND AGRICULTURAL CREDIT PROJECT (LOAN 1340-TUN) HIGHLIGHTS The project supported a program to increase agricultural production which would help meet rising domestic demand and reduce huge agricultural imports. In contrast to the First Agricultural Credit Project (Loan 779/ Credit 263-TUN), the project provided investment credit to small/medium farm operators who, up to 1976, had limited access to institutional credit. Project investments were expected to total about US$26.0 million, of which Loan 1340-TUN would finance US$12.0 million. Banque Nationale de Tunisie (BNT), the Borrower and executing agency, would finance US$3.1 million equivalent, sub-borrowers US$6.0 million, and the Government US$5.2 million, half of which would consist of grants and the other half as loans to sub-borrowers. Main investments to be undertaken through the project were: light agricultural equipment, seeds, dairy cattle, and well and irrigation improvements by small/medium farmers; heavy agricultural equipment, dairy farm development and irrigation infrastructure by large farmers; date palm plantation development by cooperatives; and olive oil mills by agro-industries (PPAM paras. 1 and 2, and PCR para 2.04). Project implementation was delayed by two years, but the overall delay was about two and one-half years as Board presentation was delayed five months due to a controversy over credit policy between the Government and the Bank. Further, the final disbursement took place four months after the closing date. Delays were caused by problems of meeting the effectiveness conditions by the Government, competition from alternative sources of funds, and Government policies (PPAM paras. 7-10 and PCR para. 3.01). Although the investment pattern at completion was somewhat different than envisaged at appraisal, the project generally appears to have achieved its objectives. However, due to incomplete development of an investment monitoring system that was a component of the project, precise information on project accomplishments is not available. According to estimates, the project achieved its overall incremental crop and livestock production objectives (PPAM para. 23 and PCR paras. 4.01 and 4.02). Project beneficiaries have been estimated at 5,800 at completion compared with 1,250 estimated at appraisal (PPAM para. 25). This increase was primarily due to smaller loans being made to more small/medium farmers than was envisaged. The reestimated rate of return is 33%, about the same as the appraisal estimate; however, this estimate is conditioned by the lack of basic project performance data noted above (PPAM paras. 24 and 49). - iv - Other points which may be of special interest are: - institutional development of BNT was significant; administration was reorganized and decentralized, professional staff was expanded, and appraisal procedures improved (PCR paras. 4.06 and 4.07); - although BNT's lending level expanded considerably during the project period, its agricultural lending operations have not been profitable (PCR paras. 4.08 and 4.09); - the Bank's efforts to change credit policy in Tunisia have not been significantly successful (PPAM paras. 30-45); and - monitoring and evaluation of credit projects is especially difficult due to the dispersed nature of investments, but a better designed monitoring and evaluation system could improve results (PPAM paras. 49-52). PROJECT PERFORMANCE AUDIT MEMORANDUM TUNISIA - SECOND AGRICULTURAL CREDIT PROJECT (LOAN 1340-TUN) I. SUMMARY A. The Project 1. The principal objective of the Second Agricultural Credit Project was to increase agricultural production to meet rising domestic demand for food and other agricultural products in Tunisia and to reduce the huge agricultural production deficit. An innovative feature of this project, which was supported by a Bank loan of US$12.0 million, was that it would provide investment credit to small/medium farmers1/ who up to that time (December 1976) had little access to institutional credit.2/ Project investments were expected to total about US$26.0 million, of which 30% would be made by small- and medium-sized farms. Further, these farms were to benefit from 32% of the Bank loan which would assist in the purchase of light agricultural equipment, seeds, dairy cattle, and well and irrigation improvements. Other investment credits were to be granted to commercial farmers, collectively owned smallholder date plantations, and agro- industries. Investments by commercial farmers were expected to includeheavy agricultural equipment, pure bred dairy stock and related physical facilities, and construction and improvement of irrigation infrastructure. Date palm and plantation development was expected to consist of groundwater irrigation, drainage development and land levelling, land preparation, planting and maintenance (first seven years) of date palms, windbreaks, pumping stations, tractors, miscellaneous agricultural implements and storage facilities. Agro-industries were expected to invest mainly in olive oil mills. 2. The First Agricultural Credit Project3/ had supported essentially the same types of investments as the Second Project - grain, dairy, and date palm development - but only for commercial farmers and agro-industries. 1/ The term "small farmers" is used in the appraisal report but "small and medium farmers" is used in the Loan Agreement, likely a compromise due to the difficulty encountered in agreeing on the size definition of small farmers. 2/ The limited access to investment credit by small/medium farmers was through FOSDA and bilateral assistance projects. Further, the Irrigation Rehabilitation Project (Loan 1068-TUN), which was approved December 19, 1974, earmarked US$2.1 million for small farmer credit. 3/ See also Project Performance Audit Report, Tunisia First Agricultural Credit Project (Loan 779/Credit 203-TUN), OED Report No. 2497 dated May 11, 1979. - 2 - However, date palm plantations were developed by STIL (Socifta Tunisienne de 1'Industrie Laitiare), a semi-private industry with much experience in date production. 3. At full development, incremental annual production of the project was expected to total 50,000 tons, consisting of grains, fruits, vegetables, beef, dates and olive oil and some miscellaneous livestock. The value of this incremental output would be around D 6.5 million, in 1975 dinars. Further, the project was expected to provide increased incomes to about 1235 farmers, of which about 960 would be smallholders, including 300 SCMV members (Socift6 Civil de Mise en Valeur - Farmer Association for Joint Farm Development). About 475 of the smallholder beneficiaries were estimated to be in the poverty group. The project, through investments in commercial, smallholder and cooperative farms, was expected to create 1,300 man-years of family employment, and 350 jobs at olive oil mills. 4. The Borrower and executing agency of the project would be the Banque Nationale de Tunisie (BNT), as it was for the First Agricultural Credit Project. At the time of project appraisal, BNT was the second most important commercial bank in Tunisia, providing mainly short-term loans to commerce and industry. A start was made under the First Agricultural Credit Project to improve BNT's lending activities to agriculture. An objective of the Second Project was to continue to improve BNT's efficiency as an agency for providing agricultural credit by fully staffing the Project Unit within BNT that was established under the First Agricultural Credit Project. 5. Procurement under the project by farmers would be through local channels, but procurement of machinery and equipment for oil mills would be made by selecting the best of three offers. Local competitite bidding was required for procurement of construction of deep wells for plantation development. Most of the goods for on-farm development was considered to be too varied to be suitable for bulking and international procurement. 6. Investment results were expected to be monitored on a representative sample basis. To this end, BNT would set up a monitoring system within 12 months of effectiveness. Further, the Borrower had agreed to provide suggested procedures for designing an appropriate monitoring system within nine months after effectiveness. B. Implementation Delays 7. The project was delayed even before approval. The project was appraised in November 1975 and was scheduled to be presented to the Board in June 1976. Differences between the Bank and GOT developed on credit policy, including interest rates. Another contentious issue was lending for small farms which GOT did not favor under the lending terms advocated by the Bank. As a result, the project was not presented for Board approval until November 19, 1976. 8. Loan effectiveness was delayed by four months. The main reason for this delay was a condition that required GOT to issue and publish decrees revising the terms and conditions by which special agricultural development - 3 - funds (Fonds Sp&cial de D6veloppement Agricole - FOSDA) were on-lent by the Borrower (BNT). Fulfilling these terms and conditions required significant changes in the interest rate and other lending terms existing for FOSDA at that time. Although the new conditions and terms finally decreed for FOSDA were not exactly comparable to those of the Loan Agreement, they were accepted by the Bank as being sufficiently comparable as to not have a detrimental effect on the timely disbursement of the loan. 9. Another condition of effectiveness required the Borrower to furnish the Bank with: "(i) a report of the audit of its acounts and financial statements for fiscal year 1975 of such scope and in such detail as requested by the Bank, and (ii) certified copies of its audited financial statements for such year." This report was received by the Bank; however, it did not contain the standard audit certification because of the limited scope of the audit. Nonetheless, the Bank accepted the report as satisfying Loan Agree- ment Section 7.01 because BNT's management agreed that a full audit would be done in compliance with standard audit practice by an auditor acceptable to the Bank. This decision was taken before the first effectiveness date; thus, compliance with this condition did not delay loan effectiveness. 10. In sum, the project was implemented with an official delay of about two years, but the overall delay was about two and one-half years. The final Closing Date (December 31, 1982) took place two years after the original Closing Date, but disbursements, as agreed by the Bank and the Borrower, were continued for another six months until June 30, 1983, based on commit- ments as of December 31, 1982. The final disbursement was made on April 5, 1983. An undisbursed balance of US$98,640.42 was cancelled on June 30, 1983 C. Investments and Disbursements by Category 11. For on-lending purposes, the project was divided into four categories: I - Sub-loans4/ to small and medium farmers for general farm development and to SONAMO (Sociftf Nationale de Motoculture) and to service cooperatives for agricul- tural machinery. II - Sub-loans to commercial farmers for general farm development. III - Sub-loans to SCMVs (Sociftf Civil de Mise en Valeur) for collectively owned date palm plantations. IV - Sub-loans for agro-industrial enterprises for food processing, packing and storage. 4/ All sub-loans were medium- and long-term, i.e., 3 to 20 years depending on type of investment. For example, sub-loans for deep tubewells were for 20 years with a two-year grace period. 12. Sub-loans to small/medium farmers got off to a slow start due to a lack of staff within the Project Unit who could apply the new system of evaluation of sub-loan applications which was initiated under the proj- ect.5/ It took about twelve months for the first disbursement to be approved by the Project Unit. Because of a shortage of technically competent staff within BNT, it was decided that sub-loan applications from small farmers would be technically appraised by the Regional Services (RS) of the Ministry of Agriculture (MOA), whose personnel had to be trained in sub-proj- ect appraisal procedures required by BNT and the project. RS was reluctant to appraise sub-loans under terms which were not exactly the same as those of FOSDA (Fonds Sp cial de D veloppement Agricole). This reluctance also slowed sub-loan approval and disbursements. It was only in May 1978 that MOA agreed to appraise sub-loans under lending terms and conditions which were in accor- dance with the Loan Agreement. 13. There was an abundance of sub-loan applications from the small/ medium farmer group in 1979 and through the first half of 1980; thereafter, there were few sub-loan applications because of the availability of FOSDA funds with simpler lending procedures and conditions, contrary to the letter and intent of the Loan Agreement. In December 1982, in order to speed up disbursements after the Closing Date had been extended twice, the Bank per- mitted the transfer of sub-loans already approved by BNT for financing by FOSDA funds. 14. Sub-loans to SONAMO (Soci t Nationale de Motoculture) and service cooperatives were made without difficulty and accounted for 21% and 15%, respectively, of total investments under Category I. These totaled D 3.5 million compared with D 3.3 million expected at appraisal (including contingencies). Disbursements by the Bank under this category amounted to US$4.6 million vis-a-vis US$3.9 million6/ allocated to Category I at approv- al. The additional funds for this categ--y came from a transfer from Cate- gory III. 15. On the basis of sub-loans, the number of beneficiaries under Cate- gory I was about 4,860 compared with 960 estimated at appraisal, when lending to small/medium farms was expected to finance investments in integrated farm development which would encompass irrigation development, cultivation equip- ment, livestock, and stables. However, lending to small/medium farmers was generally for single investments, and the average amount of sub-loans was about one-sixth of the expected amount. This together with sub-loans made to 30 Cooperative Production Units (approved as beneficiaries after the Loan Agreement was signed), which included an average of about 30 farmers, accounts for the increased number of beneficiaries. The incomes of the small/medium farmer group were expected to increase by about D 1,000 (more than double their average base level) at full development through project 5/ At appraisal, the small/medium farmer group was defined as farms with incomes of not more than D 600. On delays in implementation, see also Comments from Borrower, Annex 1, paragraph reference (12). 6/ The exchange rate averaged US$1 = D 0.430, over the implementation period. investments. Due to the lack of an adequate monitoring system (see PPAM paras. 49-52), no information on the impact of project investments on the incomes of this group is available. 16. Sub-loans to coaercial farmers (Category II) were made promptly and without difficulty because of BNT's prior experience, and because loans to this group of farmers were mainly for mechanization, which were ineligible for financing by FOSDA under its new terms and conditions. Loans for farm machinery accounted for 82% of lending under this category. This result was essentially in line with appraisal expectation. Expected total disbursements under this category were exceeded by 10%. There is also no information available on income improvements resulting from these investments (PCR paras. 3.22 and 3.23). 17. Sub-loans to SCMVs (small farmer associations) for development of date palm plantations (Category III) experienced a considerable delay. At completion, only 58% of the expected disbursements were made. Lending under this category only began during the second quarter of 1981 (six months after the original Closing Date). The initial delay was related to compilation of legal documents required for establishment of SCMVs, performance reviews, and lack of extension services. Delay was also due to the Government's opposition to creation of new date palm plantations by farmer associations. However, in 1980 this policy was changed, and the creation of eight SCKVs was approved. Sub-loan requests, in the amount of D 2.0 million for six SCMVs, were processed quickly by RNT. However, implementation of these investments was hampered by difficulties in obtaining drilling rigs, suitable contractors, and supplies of suitable date palm seedlings. As of August 1983, none of the six SCMVs had been completed. The uncompleted works were eventually financed by the Third Agricultural Credit Project (PCR paras. 3.09 and 3.10). 18. After it became apparent in December 1982 that the reduced (see para. 14 above) amount of funds available under this category would not be disbursed before the Closing Date, the Bank approved conversion by BNT of D 243,642 in grants it had already made to SCMVs to sub-loans eligible for reimbursement under Category III. This action resulted in additional dis- bursements by the Bank of US$202,000 under this Category. 19. Sub-loans to agro-industries (Category IV) were made without any difficulty; however, no sub-loans were made to olive oil processing plants as expected at appraisal. Instead, project funds were used to finance food, drink, seed and feed processing facilities, and cold storage. The reasons for the change to financing of these facilities (which was permitted under the Loan Agreement) were: requirements for oil mills were overestimated, agro-industrial sub-projects did not qualify for FOPRODI (Special Fund for Industrial Promotion) subsidi,s which some prospective borrowers insisted on getting, and BNT financed some oil mills out of its own funds for reasons unknown. Actual investments under this category were about 25% less than anticipated at appraisal (including contingencies) (PCR para. 3.02). 20. Only one minor amendment was made to the Loan Agreement which per- mitted sub-loans to production cooperatives (see para. 15) to be reimbursable under the project. Such cooperatives had been excluded from the original - 6 - project because they were not well managed. Upon the Borrower's request, the Bank later agreed to include such cooperatives if they were judged to have good performance and management. 21. A number of covenants were not complied with or were only partially complied with by the Borrower. With respect to reporting and audits, quar- terly and annual reports were generally complete and timely, but financial and audit reports were usually late (up to one year) and had received only qualified approval of the auditors. The reason, reportedly, for their delayed arrival at the Bank was that the auditors could not finalize their audit before BNT's general assembly had approved BNT's financial statements which usually occurred six months after the end of the fiscal year. A more serious problem occurred with Covenant 3.08 of the Guarantee Agreement under which the Government agreed to ensure that credit for purposes similar to those of the project would be made available on such terms and conditions that would not materially affect the carrying out of the project by the Borrower. Funds provided through FOPRODI at terms and conditions more favorable than those of the project competed with lending to agro-industries (Category IV) and, similarly, FOSDA funds competed with lending to small/medium and commercial farmers. This competition was partly responsible for the delay of disbursements under Category I.7/ 22. Under the Loan Agreement (Section 3.05), the Borrower agreed to design and establish an investment monitoring system within 12 months of the signing of that agreement. The objective was to obtain information on the effective use of project inputs and outputs. This system was designed with the guidance and approval of the Bank, and implementation was first initiated in 1978 with about 10 medium and large farmers. The program was interrupted in 1979 but resumed in 1980 with a sample of 25 farms and extended to 58 farms in 1981. Inadequate information was collected by this system (see PPAN paras. 49-52 and PCR paras. 3.22, 3.23 and 4.05).8/ D. Project Impact and Benefits 23. Due to BNT's limited success in establishing an adequate monitoring system, there is only indicative information on the project's impact. On the basis of limited inform2tion available and appraisal of sub-loans by MOA and BNT, 12 representative farm models were developed. Estimates of total incre- mental annual production were made by weighting the models by the actual num- ber of sub-loans made to each type of farm represented by these models. These resulting production estimates at full development compare with those made at appraisal as follows: 7/ See also Comments from Borrower, Annex 1, paragraph reference (21), item 5. 8/ The Borrower disagrees with the assessment, see Annex 1, paragraph reference (22). -7- Appraisal Estimate Estimate At Completion (tons) (tons) Grain 13,155 16,000 Fruits and Vegetables 22,700 - Vegetables - 19,490 Fodder - 2,000 Milk 6,500 10,970 Beef 370 776 Lamb and Mutton - 150 Wool - 8 Dates 3,000 3,100/a /a At full development (18 years); however, total investments were not financed under this project. On commercial farms, large yield increases did not materialize. These farm- ers generally increased land cropped and continued their traditional cropping practices rather than using more intensive production practices as was expected at appraisal. Most of these investments were in tractors. Yield increases on small/medium farms were less than expected too, since credit was mainly used for single investments rather than for integrated development packages as assumed at appraisal. 24. Estimates of the financial rates of return (FRR) to the 12 farm models range from negative (a commercial livestock farm model) to over 50%. The re-estimates of FRRs for agro-industrial investments range from negative (fruit and vegetable processing) to 20%, and that for date palm plantations is 14%. The re-estimated FRR for the entire project is 33%. Because the estimates of the economic rates of return (ERR) for the project's components were generally higher than the financial rates for the Second and Third Proj- ects, the PCR concludes that the same relationship would likely hold for the re-estimated ERR for the Second Project, and thus the ERR is not less than 33%. However, the PCR notes (para. 4.05, footnote) that no separate calcula- tion of the project's ERR was made because of lack of actual results of proj- ect investments. Because of this, and because of input subsidies, Government control of product prices, and the shadow pricing of labor at 20% to 50% of the minumum wage,9/ all of which affect the relationship between the FRR and ERR, the audit questions the validity of the PCR's assumption about the re-estimated ERR.10/ 9/ This is difficult to justify since investments were made i large farm m-chines. 10/ See also Borrower's interpretation of the validity of the reestimated ERR and project benefits, Annex 1, paragraph references (23) and (24). - 8 - 25. On the basis of the number of sub-loans that were made, the number of project beneficiaries at completion is estimated at about 5,800 compared with just 1,250 estimated at appraisal. Only 616 sub-loans were made to com- mercial farms; hence, the bulk of the beneficiaries are small/medium farms (see para. 15). The main reason for the increase in the number of beneficia- ries is that the average loan amount to small/medium farms was much less than expected due to the investment pattern noted above. E. Other Results 26. Performance of the Borrower. BNT lending to agriculture almost tripled during the 1976-82 period. About half of the BNT lending was for short-term loans, while project and special funds such as FOSDA were used for medium- and long-term lending. Despite expanding its lending, BNT's agricul- tural operations have not been profitable. A special study of BNT's opera- tions for 1979 and 1980 determined that losses on agricultural lending amounted to about D 1.0 million each year. This figure, however, results from using a conservative reserve for bad debts, so actual losses were likely higher. 27. BNT's sub-loan recovery record remained poor over the project peri- od and in 1982 appeared to have detericrated, especially for funds provided by FOSDA and the Second Agricultural Credit, which were 52% and 66%, respec- tively. This generally poor recovery rate is due to regulations that do not provide BNT with incentives to recover loans financed by special funds. In contrast, its recovery of loans provided by its own funds (mainly short-term lending to large farmers) averaged about 90%. BNT attributes this high recovery rate to the need for obtaining new loans each season, and the fact that national grain marketing agencies can withhold marketing proceeds to repay any outstanding loans held by BNT. 28. Positive steps were taken during project implementation to strengthen BNT's organizational structure. Staff in charge of credit operations was expanded by about 70% and the credit department was reorganized. 29. Bank Performance. The Bank was instrumental in getting BNT to pro- vide funds for investments by small/medium farmers at more economic rates of interest and to conduct more vigorous appraisals of sub-projects. However, the Government's lack of commitment to changing its practice of providing funds to small/medium farmers through BNT at concessionary terms and rates was not fully recognized by the Bank until late in the implementation peri- od. The Government supplied FOSDA funds at terms and conditions which com- peted with Second Credit Project funds, and this contributed to the extended period of disbursements. Although some progress has been made, the Bank has not been able to get the Government to change its general policy on terms and interest rates for lending to agriculture. -9- II. ISSUES A- Credit Policy 30. As already pointed out, both the First and Second Agricultural Credit Projects for Tunisia experienced considerable time overruns; the First Project, 34 months, and the Second Project, 27 months. Moreover, the Third Project is now far behind its estimated disbursement schedule. Only about 30% of the expected level of disbursements for that project had been achieved as of April 1984. The Third Project (Loan 1885-TUN), which was approved on June 26, 1980 and became effective on June 24, 1981, has been listed as a problem project since September 1982. A common reason cited as a cause for the delays in implementation and consequently disbursement of all three proj- ects has been competition from other sources of credit, principally FOSDA, at interest rates and terms much more favorable to the farmers than those per- mitted under the Bank loans. In BNT's view, more important factors contributing to project delay are: other sources of financing (in addition to FOSDA), which are better suited to the needs of farmers in well defined areas than are Bank projects covering the entire country, the Bank's frequent disregard for certain factors that are inherent in the country's socio-economic situation, and the time required to put project mechanisms and procedures in place at start-up. 31. A major issue between the Bank and the Government at the time of negotiations of the Second Agricultural Credit Project was the uniformity of terms and conditions for all institutional agricultural credit in Tunisia. Further, sub-projects which were financed by special funds (e.g., FOSDA) were subject to little financial and economic appraisal or supervision by BNT. In addition, these funds included a grant portion as well as low interest rates. In the Bank's view, FOSDA was not providing sufficient credit for small farmers. The Bank sought to increase lending to small farmers, but only with higher interest rates and adequate subproject appraisal and super- vision. 32. Because of the Bank position on lending conditions to small farm- ers, the Government wanted the project to provide funds only for on-lending to large farmers and agro-industries. The Government planned to provide sub- stantial FOSDA funds for small farmer credit on its traditional subsidized terms. The Government considered terms and conditions of lending .o small farmers to be a sensitive internal political issue in which it preferred not to involve the Bank. However, the Bank continued to promote credit policy reforms in the direction of what it believed would lead to greater effective- ness and efficiency in lending to agriculture. 33. Eventually, after numerous discussions, the Government acceded to most of the Bank's conditions for agricultural credit policies. These policy discussions, as already noted, led to delays in starting formal negotiations, originally planned for May 1976 but actually held in September 1976, and in presenting the project to the Board, which was delayed from June 1976 to October 1976. - 10 - 34. Eventual compromises reached during negotiations included: inter- est rate at 6% for small/medium farmers, 7% for commercial (large farmers) and 8Z for agro-indus tries. The Bank had proposed 6% for small/medium farm- ers, 8% for commercial farmers and 9% for agro-industries. BNT had been lending to small farmers at interest rates from 2.OZ to 4.5% together with a large grant element. Under conditions of the Loan, small/medium farmers, in- dividually and as members of cooperative associations, would continue to receive sub-loan grants of 10% to 40% (LA11/, Schedule 4, Section C.1). It was also agreed (LA 7.01(a)) that all loans made by BNT out of the proceeds of FOSDA be under revised terms and conditions which would reflect the specific terms and conditions detailed in paragraphs B.1, B.2 and C of Schedule 4 of the IA which stipulated lending conditions for small/medium farmers. 35. Ln the Guarantee Agreement (Section 3.08), this covenant stipulated "The Guarantor shall take all steps necessary on its part to ensure that credit for purposes similar to those of the Project shall be made available on such terms and conditions as shall not materially and adversely affect the carrying out of the Project by the Borrower.- 36. As already noted (para. 8), compliance with this covenant delayed the effectiveness of the loan from March to July, 1976. A supervision back- to-office report of July 1977 noted a number of differences between the revised terms and conditions for FOSDA funds as reflected in published "D&crets- and "Arrates" and those detailed in the Loan Agreement. The dif- ferences were not considered significant enough to affect disbursement under the project. 37. Early supervision missions did not detect any signs or receive any information indicating that under actual terms and conditions and availabil- ity FOSDA funds were competing with lending under the project to the small/ medium farmers. As late as September 1979, a supervision mission expected the full amount of the loan to be disbursed by the original Closing Date of December 31, 1980. It was noted, however, that there had been no sub-loan commitments for date palm plantations and that funds for the Category (III) probably would have to be reallocated. Thereafter, there was a gap of 18 months in supervision, essentially because the Third Agricultural Credit Project was being appraised and and processed for approval during that peri- od. In the meantime, the Closing Date of the loan had been extended by six months to June 30, 1981. The latter supervision mission of March 1981, while recommending the extension of the Closing Date to December 1982, did not recognize that FOSDA loans were competing with lending to small farmers under the project. 38. It was not until the October 1981 supervision mission that the com- petition between FOSDA loans and Bank supported sub-loans was noted. Prior- ity was being given to sub-loan applications which would qualify for FOSDA financing by the Regional Services of the Ministry of Agriculture, which was charged with the technical appraisal of loan requests. This indicated that 11/ Loan Agreement. - 11 - there was considerable and, possibly, increasing Government allocation of funds to FOSDA. Commercial farmers, too, had access to FOSDA loans. Because these funds excluded lending for mechanization, they did not compete with sub-loans to commercial farmers under the project. Sub-loans for mechaniza- tion represented 82% of the amount of sub-loans made to commercial farmers, (see para. 16). 39. It is not intended to imply that competition from FOSDA was the only cause of delay in project implementation. The point is that the Govern- ment did not, to a large degree, support the Bank's principles of credit policy: that beneficiaries should pay the full cost of credit or at least a positive rate of interest, sub-loans should only be made for economically sound enterprises or investments, and credit institutions should be operated on a financially sound basis--they should bear the risk of their lending operations in order to promote greater efforts to recover overdue loans. 41. Government policies, especially provision of alternative sources of credit at subsidized terms, including lack of enforcement of repayment of loans, and on terms conflicting with those agreed to under Bank financing, also affect implementation of the Third Project.12/ As of December 31, 1983, the original Closing Date of the project, disbursements had just reach- ed 24%. The Bank had indizated that it would not extend the Closing Date unless the Government and BNT took measures to improve loan recovery and reduce subsidies available to large farmers. The Bank agreed to an extension of the loan to December 31, 1984 when the Government proposed to undertake the following measures (April 9, 1984 Supervision Report, Annex 3): (i) an increase in interest rates for large farmers and agricultural investment projectsl3/ from 6.0% p.a. up to 7.5% and 8.0 p.a.14/, these investments and beneficiaries being moreover ineli- gible for FOSDA investment subsidies. For small and medium farm- ers, Government proposed to evaluate the impact of an increase in lending rates on financial viability of typical farmers before tak- ing any action; (ii) the presentation to Parliament for approval of a law aimed at granting to financial institutions the Privilage d'Etat for recovery of loans extended on budgetary resources or guaranteed by the state; 12/ In BNT's view, the Third Credit Project was delayed due to slowness in completion of the Second Project (which in turn was due to the factors mentioned) and appreciation of the U.S. dollar exchange rate vis-a-vis the Dinar; see Annex 1, paragraph reference (41). 13/ Defined as Category C in the 1982 law of promotion of investments in agriculture. 14/ 7.5% for investments in livestock and fruit tree plantations. L - 12 - (iii) increased financial incentive for financial institutions to lend to the agricultural sector, through the assumption by a guarantee fund (Fonds National de Garantie) of part of the risk of default on agricultural loans; and (iv) the strengthening of the management and financial situation of the Socift6s de Caution Mutuells (SCMs). 42. These measures were to be taken in February 1984, but up until April no significant results were apparent.15/ Note that item (i) above implies that large farmers were receiving investment 3ubsidies, which was contrary to Section 3.08 of the Guarantee Agreement. 43. In the audit's view, the Bank's efforts to change agricultural credit policy and to establish an economically viable agricultural credit system in Tunisia has only met with limited success. Interest rates have been raised, but are still negative in real terms. The Government continues to provide alternative and competing sources of credit at subsidized interest rates together with investment grants to both small/medium and commercial farmers. 44. The Bank has been patient in its attempts to get the Government to accede to reforms in institutional credit. Although the Government agrees to make changes which the Bank believes to be in the right direction, it con- tinues to revert to earlier practices.16/ 45. It has been noted in Bank reports that one of the reasons for the Government's conCinued subsidization of agricultural credit is that prices of agricultural products are controlled at low levels to hold down the cost of living. While the Bank has taken note of this fact, there is no evidence that it has attempted to promote its credit policy in the broader and more realistic setting of appropriate or incentive pricing for agricultural out- put. Under such a regimen, farmers would be better able to afford higher credit charges. B. Institutional Development 46. A major objective of the Bank was to develop BNT into the foremost Bank providing credit to agriculture in Tunisia, especially to small farmers. In addition to the three agricultural credit projects, BNT has been the Borrower and executing agency for the Northwest Rural Development Project (Loan 1997-TUN, US$24 million) and also has managed part of the funds of the Second Fisheries (Loan 1746-TUN, US$28.5 million) and Rural Roads (Loan 1601-TUN, US$32 million) Projects. However, during the implementation of the Second and Third Agricultural Credit Projects (1981), the Government 15/ BNT notes, Annex 1, paragraph reference (42), that measures adopted by the Government are now being implemented. 16/ BNT states, without being specific, Annex 1, paragraph reference (44), that with the new agricultural regulations, radical changes have been made in the system. - 13 - created: (a) an Agricultural Development Bank to finance investments in agriculture; (b) a fund for lending to small farmers in marginal areas and rural communes; (c) an agency for promotion of investments in agro- industries. All of these institutions compete with BNT in some respect.17/ According to a Bank supervision report (April 9, 1984), loans by BNDA to large farmers (Agricultural Development Bank) were competing successfully with those provided under the Third Project because BNDA provided loans at more favorable overall terms. 47. Recovery of sub-loans by BNT has been a continuing concern of the Bank. The recovery rate on BNT's own funds has averaged about 90%; recover- ies for the Bank-supported projects have ranged from 40% to 80Z but have averaged under 70%, while for FOSDA they averaged about 55%. In light of the covenants under the Guarantee Agreement (GA), there is no incentive for BNT to make special efforts to ensure high recovery rates. In essence, the GA (Section 3.06) for the Second Project stipulates that the Government under- write a large part of the losses due to non-repayment of loans by small/ medium farmers (parts A and C of the Second Project). Similar conditions were provided for under the Third Agricultural Credit Project. The apparent purpose of these covenants was to ensure the financial viability and promo- tion of agricultural credit operations by BNT despite a poor record of sub- loan recovery. Partly as a result of these subsidies, BNT has remained financially viable. As noted, preliminary studies indicate that BNT is los- ing money on its agricultural credit operations. Further detailed studies to be done under the Third Agricultural Credit Project have not yet been completed. 48. More vigorous methods of appraising sub-loan requests have been established by BNT, but there is still a staffing problem. A start has been made to train and establish an effective agricultural extension service in BNT and MOA (Third Project). Some training of BNT staff has been done, but after much delay due to jurisdictional disputes. In sum, despite consider- able growth in lending and improvements in lending procedures, BNT needs further improvements in lending criteria and recovery rates if it is to come up to the Bank's expectations for it. C. Monitoring and Evaluation and Measurement of Project Benefits 49. The PCR notes (para. 4.05, footnote 2) that no separate calculation of the ERR (in addition to the FRR) was made because of the lack of reliable data concerning the results of project investment. Estimation of project benefits and rates of return are notoriously difficult for agricultural credit projects because of the number of beneficiaries, their dispersion, range of farm sizez, and varying production methods and types of products 17/ However, according to BNT, the National Agricultural Development Bank was established to finance integrated projects of the scale in which BNT and commercial banks cannot become involved because of the nature of their resources, Annex 1, paragraph reference (46). - 14 - produced. Usually, estimates of ERRs for credit projects are made by assuming some idealized farm models (as is done at appraisal) augmented by fragmentary information of on-farm results and then aggregating model results by the number of sub-loans made to farms (industries) and possibly adding to the cost-benefit stream an overhead charge related to implementation of the project. As a result of this idealized procedure, rates of return usually turn out to be very favorable. In many cases, these may be legitimate rates of return because many credit projects introduce,through investment credit and sub-project supervision, input packages (high-yielding seeds, fertilizers, pesticides, irrigation, etc.) that yield large productivity increases. However, in all cases, to certify the true benefits of credit projects, it is necessary to undertake detailed sample surveys of projects' beneficiaries in order to obtain the required information on the cost-benefit streams with and without the project. To make good estimates of the without project cost-benefit streams, it is necessary to obtain information on beneficiary farms some years before the project is implemented and on the basis of this information estimate what would likely occur without the project. Collecting such information requires good statistical sampling design and ample resources, which are not usually available.-18 Nevertheless, the achievement of high standards in measuring benefits of credit projects is difficult and costly. Other problems of measuring benefits of credit projects, as noted in a recent OED report,19/ are accounting for overhead cost, additionality, substitution and diversion. 50. As noted in paragraph 22, the Loan Agreement included a covenant (Section 3.05) under which the Borrower was to prepare not later than nine months after the Agreement Date, a proposal for an investment monitoring system (based on sampling) to estimate the benefits derived from sub-loans by sub-borrowers and then to establish such a system not later than 12 months from the date of the Loan Agreement and to continue to operate it there- after. Questions were raised at Board presentation of the efficacy of delay- ing implementation of a monitoring system so long after approval, and about the possible success of such monitoring system. Bank staff stated that an attempt would be made to start the monitoring system earlier but admitted that slow progress had been made in establishing good monitoring systems because of, inter alia, inadequate data bases. 51. The monitoring system of project investments got off to a slower start than expected and then had an up-and-down history. BNT prepared a pro- posal in 1977, with assistance from the Bank, and initiated a program in 1978 with a sample of 10 (large) sub-borrowers. Some farmers in the sample were reluctant to participate because they did not want to disclose farm financial information. No monitoring was done in 1979 because the engineer in charge left the project. A replacement was recruited and trained and monitoring was resumed in 1980 with a sample of 25 farmers and expanded to 58 farmers in 1981 (PCR para. 3.22). The same monitoring system was to be continued under the Third Project, but the audit could find no information on its present status. 18/ See also Comments from Borrower, Annex 1, paragraph reference (49). 19/ Ninth Annual Review of Project Performance Audit Results, OED Report No. 4720, September 16, 1983, Vol. 2, Chapter 5, Supp. 2. - 15 - 52. In conclusion, more effective and efficient methods must be developed to monitor the implementation and measure the results and benefits of credit projects. Terms of reference should be drawn up at appraisal which would provide programs of data collection and analysis which would follow scientific procedures. Further, beneficiaries should be required to partici- pate in data surveys as a condition of receiving subloans. Otherwise, useful information will be lost and uncertainties about results will remain. D. Conclusions 53. Changes in credit policy (or any other price policies involving subsidies) are very difficult to effect when such changes are contrary to country economic development philosophy and when vested interest of long standing exists. As has been pointed out by OED previously,20/ this issue might best be addressed at the ministerial level rather than through indivi- dual projects. However, such change is unlikely unless serious government financial constraints exist. In the case of the First, Second and Third Credit Projects in Tunisia, other sources of credit funds were Pavailable. 54. The design and program for monitoring the impact of credit projects need to be agreed to by all parties involved, preferably at the time of proj- ect approval, in order to obtain accurate measures of project progress and benefits. Otherwise, the uncertainties which many times surround the bene- fits of credit projects will continue to exist. 20/ Project Performance Audit Report, Yugoslavia First Agricultural Credit Project (Loan 1129-YU), OED Report No. 5100, May 29, 1984. 一たニー一 メルンノ汐介ノ泌レ才 ち - 17 - ANNEX 1 Page I PROJECT PERFOM4ANCE AUDIT MEMORANDUM TUNISIA - SECOND AGRICULTYRAL CREDIT PROJECT (LOAN 1340-M) COME ENT S FROM BORROWER Translatea- THE NATIONAL BANK OF TUNISIA from French Banque Nationale de Tunisfe The President and Managing Director Tunis, November 15, 1984 Mr. Yukinori Watanabe Director, Operations Evaluation Department World Bank Washington, DC Reference: Your letter of September 19, 1984 Subj*ct: Project Performance Audit Report - Second Agricultural Credit Project (Loan 1340-TUN) Dear Mr. Watanabe: Enclosed please find our observations concerning the Project Performance Audit Report for the Second Agricultural Project. Sincerely yours, Isl Mohamed Ghenima ANNEX 1 -18 - Page 2 TECHNICAL DEPARTMENT Tunis, November 2, 1984 PROJECT PERFORMANCE AUDIT REPORT Chapter 1 - Summary A. The Project (1) Small/medium farmers had access to investment credit through FOSDA or other bilateral assistance projects (FSA, FAI, KfW). At the time of preparation of the project, however, the financing of small/medium farmers thrcugh the network of Local Mutual Credit Funds (CLCMs) was included among the principal objectives. B. Implementation Delays (7) We do not find the second point of contention between the Tunisian Government and the Bank very explicit. (10) Disbursements continued in 1983, based on financial commitments through December 31, 1982. C. Investments and Disbursements (12) Subloans made to small/medium farmers had a slow start because of the problems we encountered with the Ministry of Agriculture, related mainly to the introduction of a new system of evaluation. Furthermore, regional offices of the Ministry of Agriculture do not take part in the approval of loans, which falls within the sole purview of the National Bank of Tunisia. The regional offices were willing to assist us in the evaluation process, but they had little motivation and inadequate preparation for this task. Thus, "differences in loan terms and conditions" can by no means be cited to explain the delays found. (13) BNT is not empowered to grant loans from FOSDA or from other bilateral aid funds or those of supervised projects. The fall in demand that began in 1980 was due to the implementation of a number of supervised projects better geared to the needs of farmers because their scope is limited to specific areas. Moreover, their appraisal procedures are simpler and the lending process is decentralized. (15) The cooperatives that were approved as beneficiaries after the signing of the loan agreement are the "cooperative production units." We believe, also, that our monitoring system has enabled us to compile sufficient information to assess the impact of the project and to determine the economic rate of return (see below). ANNEX 1 Page 3 - 19 - (17) The constraints that have delayed the implementation of the program of Small Farmer Associations (SCMVs) are described in detail in our project completion report, paragraph 4.3.2. The Ministry of Agriculture is by no means opposed to their establishment; rather, the delays are related to the performance reviews, the compilation of legal documents, and the lack of extension services. Financing for the uncompleted works was provided by the Third Agricultural Credit Project without the participation of FOSDA. (19) With regard to agroindustries, the requirements for oil processing plants were overestimated in our opinion, given the foreseeable development of agricultural production and particularly that of domestic and export markets. R Apart from the competition from other banks that provide funding for this sector, the main problems that we have been able to identify are: - The legislation in force, which requires that the banks meet a percentage of medium term lending from their own funds; - Competition from FOPRODI with its more advantageous conditions, especially as regards self-financing. (21) The quarterly and annual reports to the Bank on project implementation have always been filed within reasonable periods (one to two months). Conversely, the financial statements and audit reports are the ones that have not always been sent in time, owing to the auditors' delays in completing their work. As for compliance with the Covenant 3.08 of the Guarantee Agreement, we feel that the Bank was already aware of the terms and conditions of loans made in Tunisia, and that the audit report understates the importance of these differences where they do exist. Moreover, the wording of the covenant does not imply a matching of these conditions with those of the Bank. Finally, we consider it difficult to assess the impact of these conditions on project execution when all other factors involved are taken into account. In our opinion, the main problem lies in the large number of sources of financing, which are managed with greater flexibility when mechanisms are more or less independent, while terms and conditions are similar. (22) To the contrary, the monitoring system designed by BNT and implemented after approval by the Bank has enabled us to gather highly useful and interesting information for analysis of the post-investment performance of farms. ANNEX 1 Page 4 - 20 - Furtnermore, the procedures were subjected to several field tests, and the monitoring program was introduced after an appropriate informational campaign since this is not a very common practice in Tunisia. The interruption of the program in 1979 was due to the departure of the engineer in charge. It naturally took some time to train his replacement, who was a young man just out of university. D. Project Impact and Benefits (23) The impact of the project has been assessed in a very precise and detailed manner (see our project completion report). We deliberately limited the number of representative models in order to make the necessary comparisons with the models indicated in the performance report. Moreover, it wculd have been impossible to monitor all of the models because of the diverse natere of the farms and investments financed, especially since the projecu covers the entire country. As in the case of the appraisal, the project performance report must be based on realistic and feasible approaches. As to the integrated projects of small/medium farmers, attention must be called to the incompatib!-lity of this formula with current practices and procedures. In addition, the low incomes earned from these farms would never have sufficed to cover cie self-financing requirements. (24) The economic rate of return could have been calculated precisely from the information compiled during evaluation of the project impact and the results of investments. Moreover, we believe that the approach taken in the project completion report is quite realistic since both lines of credit have had the same conditions and have financed the same types of investments. E. Other Results (27) It is entirely normal that recovery rates would change very little and even Zall during poor agricultural years. One must be aware of the accruing amounts due, which when they decrease are nonetheless added to the total delinquent amount, despite any recoveries. The high rate of recovery of seasonal loans from own funds is explained by the fact that the borrowers must return to BNT each year to obtain the new credit they need to begin the season. ANNEX 1 Page 5 - 21 - Furthermore, grains are marketed mainly through national agencies which pay the proceeds to BNT, making it possible to concrol the entire operation and to apply withholding at the source. Thus, BNT is obliged to make additional efforts to recover sums due on investment credits financed from other funds. (28) The office responsible for managing World Bank loans has been in e-istence since the first credit project. There has been only an internal reorganization within BNT. (29) Small/medium farmers can obtain loans with similar conditions and rates from all funds. It is. rather, in the case of commercial operations that differences exist, particularly as regards interest rates and subsidies. That component has not caused any special problems. Furthermore, the Bank was aware of these differences, which are stated clearly in the FOSDA documents. ANNEX 1 Page 6 - 22 - CHAPTER II: ISSUES A. Credit Policy (30) The delays in implementation and disbursement of the three agricultural credit projects are explained not only by competition from FOSDA or other sources whose conditions are deemed more favorable. We must call attention to other factors that in our opinion are even more decisive: - Other sources of financing which are better geared to the needs of farmers in well-defined areas than Bank Projects covering the entire country; - The start of any project takes sorne time at the outset to put project mechanisms and procedures in place. The delay therefore involves all other ongoing projects. - The Bank's projects frequently disregard certain factors that are inherent in the country's socioeconomic situation. For that reason the projects are often poorly suited or overestimated in terms of financing possibilities. They are likewise more restrictive to the farmer (rules for the timing of investments) and for BNT (risks). - Variations in the exchange rate of the dollar have often been responsible for the lag in disbursements. For example, the third project was appraised assuming a rate of D 0.4 = $1.00, whereas the present value of the dollar is more than twice that amount. - The risk assumed by BNT compels it to be highly prudent, in contrast to other projects financed by government funds or the like (bilateral assistance). (31) At the time of project preparation FOSDA had a major impact on small/medium farmers. In BNT's view, "t was rather the expansion of operations of the Local Mutual Credit Funds whose members were small/medium farmers. Lacking resources, those institutions could only finance short-term loans. The changes that occurred thereafter were the result of the new approach taken by the authorities, which was to encourage a new mutual credit system (mutual guarantee companies) and to make the banking system responsible for investment credits. (41) As mentioned earlier, there has been no discernible conflict in loan terms and conditions that %uuld have affected the implementation of the Third Credit Project. It was rather the delay in completion of the second project that slowed Lhe start of the third project, at least for small/medium farmers. -23- ANNEX 1 -23 - AR Page 7 in the matter of disbursements, we point out the movement of the exchange rate of the dollar during this period, especially when the project was too ambitious considering all of the others in place. (4Z) The measures adopted by the government are now being implemented. Large farmers are no longer eligible for loans from FOSDA or similar funds, and are no longer receiving subsidies from any lender. Thus, they are being penalized since they are paying at least 0.5 to 1% more in interest than that charged on Bank funds. The sole advantage to young farmers under (35 years of age) is a reimbursable grant to cover part of their self-financing in the case of integrated projects approved by APIA. (44) With the new agricultural regulations, radical changes have been made in the system. B. Institutional Development (46) BNT was not the borrower in the Rural Roads Project (Loan 1601-TUN), and the Second Fisheries Project (Loan 1740-TUN). Only part of the funds, to be used for making loans, is managed by BNT. Moreover, the National Agricultural Development Bank (BNDA) was not established to compete with BNT, but rather to finance integrated projects of a certain scale in which BNT and the commercial banks cannot become involved because of the nature of their resources. Finally, we would point out that there is no National Popular Bank. Rather, there is a fund intended for small farmers in marginal areas and for rural communes. (48) In the area of training, BNT has had an appropriate training program since the start of the second project, which has trained both its own personnel and those of the Ministry of Agriculture. The Bank has been informed of these actions periodically and there is no legal dispute. We believe that the desired training cannot take place within the periods stipulated. Practical exercises and field work can ensure the best possible training in problems of agricultural credit. Finally, we would point out that several years ago BNT instituted a program of continuous training for all its staff. The programs are reviewed every year and the courses are taught by its own personnel in addition to seminars held by other institutions. (49) The method that you propose for performance auditing has been adopted in its entirety at our instigation. ANNEX 1 -24- Page8 We are likewise in agreement with you on the problems in making good estimates, especially since the majority of our farmers do not keep accounts. (51) It is entirely normal that the monitoring system devised by BNT and used for the first time in Tunisia would be slow in getting started for three basic reasons: - Time is needed to test the system in the field and train personnel; - The impact of the investment does not become evident until after a season; - Agriculture is a cyclical process, which means that information can be obtained only at very specific points in the year, namely the beginning and the end of the season. The withdrawal of some farmers and the halt of monitoring in 1979 had no impact on the results, inasmuch as the final sample was quite representative and the data for the 1979 season were collected in the following year. As noted, the program was suspended in 1979 because of the departure of the engineer in charge and not because of these problems. Needless to say, we had to recruit and train another expert who has specialized in this area since the period in question. (52) We believe that the monitoring procedure we have adopted is effective enough to be reviewed. In our opinion, strict. scientific procedures cannot be adopted because of the educational level of most of our farmers. Moreover, supervision of investments by BNT is i usual condition of the contract. In practice, however, it is highly difficult to enforce it given the lack of means. Rather, our monitoring system is based mainly on informing farmers and persuading them to cooperate. D. Conclusions (54) Whatever monitoring system may be adopted, we believe there is always uncertainty about the benefits of a project, for various reasons: - It is impossible to monitor all beneficiaries; - Even when the sample is sufficiently representative and the information collected is reliable, the highly diverse nature of the farms financed and the unforeseeable changes that occur on the farm mean that the results of sampling are always approximate; - The estimate of project benefits made after adding up the data is likewise uncertain, and cannot reflect the actual benefits derived from the project. - 25 - PROJECT COMPLETION REPORT TUNISIA SECOND AGRICULTURAL CREDIT PROJECT (Loan 1340-TUN) March 1984 EMENA Regional Office ^67,s- - 27 - PROJECT COMPLETION REPORT TUNISIA SECOND AGRICULTURAL CREDIT PROJECT (Loan 1340-TUN) I. INTRODUCTION Project Background 1.01 Bank/IDA involvement in agricultural credit in Tunisia started with the First Agricultural Credit Project (Loan/Credit 779/263, USt8 million). The Loan/Credit became effective in March 1972 and financed part of Banque Nationale de Tunisie (BNT) lending to commercial farmers for investment in mechanization and dairy farming, and to Sociftf Tunisienne de 1'Industrie Laitiare (STIL) for development of date palm plantations. The Project was completed in July 1978 and a Project Performance Audit Report was issued in May 1979. The Second Agricultural Credit Project is a follow-up project with, however, a change in focus towards small and medium farmers. To date, Bank's total lending to Tunisia amounts to US$1,338.5 million for 72 projects (62 loans and 10 IDA credits). In the agriculture sector 13 projects for a total loan amount of US$273.4 million have been approved from 1967 to December 1983. The Agriculture Sector 1.02 Tunisia's 6.7 million pop*--ation (1982) is growing annually at 2.5% and almost half of it is rural. GDP growth rate during the seventies has been high (about 8%) but declined to about 5% in the early eighties as a result of stagnation in agricultural, petroleum and chemical output. Agriculture's share in Tunisia's GDP declined from 24% in 1960 to 13% in 1981. Similarly, agriculture's share in total employment declined during the same period from 56% to less than 35% and its share of exports of goods from 57% to less than 10%. The decline in the relative importance of agriculture has been caused essentially by the more rapid growth of other sectors, particularly petroleum, manufacturing and tourism. 1.03 Covernment objectives in the sector include the pursuit of food self- sufficiency (defined as a balanced commodity trade in agricultural products), increased rural employment and income generation, foreign exchange earnings and savings, and economic growth. Under the Fifth Development Plan (1977-81) total investmenc in agriculture amounted to D 584 million of which 44% was for irrigation, 17% for farm machinery and 12% for livestock. Government's objectives in employment and food self-sufficiency were, however, not attained. The agricultural trade balance has worsened due to relative stagnation of exports and growth of imports and because of a deterioration in the terms of trade of agricultural goods. The basic issues to be addressed in - 28 - the Sixth Development Plan (1982-86) have not changed though the emphasis has shifted somewhat from highly capital intensive projects such as irrigation and farm machinery to labor intensive subsectors such as livestock and fruit production. II. PROJECT FORMULATION Origin, Preparation, Appraisal 2.01 The Second Agricultural Credit Project was prepared with the assistance of FAO/IBRD CP and represented a continuation of activities financed under the first project. The project was appraised in December 1975 by a four-person team. The Issues Paper dated December 9, 1975 proposed the financing of an agriculture credit project of a total cost of US30 million and Bank financing of USt15 million. It flagged as the most important issue the appropriateness of the Government's proposed lending terms and conditions in the agricultural sector (6% interest to all farmers, increase in subsidies to small/medium farmers and no interest charges during the grace period). 2.02 At the decision meeting held on December 17, 1975, it was agreed that (a) a 6% interest rate on loans to small farmers should be accepted; and (b) interest rates similar to those agreed upon under the Sixth DFC Project (8%) should apply to loans to commercial farmers and agroindustry. A decision on the loan amount was postponed until receipt of the Government's response to the Bank's proposition concerning the terms and conditions of lending to commercial farmers and agroindustry. The Loan amount was finally set at $12 million to cover a 2-1/2 year credit program instead of the 3-year program envisaged at appraisal. Negotiation and Approval 2.03 The negotiations were held only in September 1976 after lengthy preliminary discussions with the Government on the proposed lending terms to commercial farmers (interest rate of 7% p. a. was accepted by the Bank) and the definition of small farmers. Board presentation had also to be delayed until November 1976, one year after the return of the appraisal mission, because of delays in signing by BNT and the Government of a relending agreement under the Irrigation Rehabilitation Project (Loan 1068-TUN), which was a condition of Board presentation of the present Loan. Project Objectives and Description 2.04 The Second Agricultural Credit Project financed a part of BNT's medium- and long-term lending for agriculture. While the First Project financed investments by commercial farmers in farm mechanization, livestock and date palm plantations, the Second Project, in addition included an agroindustries component and aimed at improving the availability of investment credit to small farmers for on-farm development and to small farmer associations for date plantations. Out of the Loan amount of US$12 million equivalent, US$3.9 million (33%) was for small farmers, US3.7 million (31%) - 29 - for commercial farmers, UStl.9 million (16%) for small farmer associations investing in date palm plantations, and US$2.5 million (21%) for . agroindustries. BNT's Project Unit, created under the First Project and consisting of five agriculturalists, was to be supplemented by an agroindustry specialist and the Unit was to be maintained and expanded as required during . the Project implementation. A system to monitor the performance and production impact of subprojects was also to be set up. III. PROJECT IMPLEMENTATION Loan Effectiveness . 3.01 Two special conditions had to be met for Loan effectiveness: (a) agricultural credit policies and procedures agreed for the Project during negotiations would be applied to all FOSDA loans (interest rates for lending to farmers to increase from 3.5-4% to 6% for small farmers and from 6% to 7% for commercial farmers); and (b) the audit report of BNT's accounts for 1975 had been completed and submitted to the Bank. The original date of effectiveness of March 16, 1977 had to be postponed twice, mainly to allow time to the Government to meet the first of the two conditions because of delays in approving and publishing the revised terms and conditions for FOSDA. The Loan became effective on July 19, 1977. Project Cost and Financing 3.02 The cost estimates at appraisal and the actual costs are compared below. Project Cost (D'000) Appraisal Estimated Actual Difference Subprojects Cost Cost Cost (%) Small farmers 3,300 3,505 +205 +6 Commercial farmers 2,800 3,626 +826 +29 Date palm plantation 2,700 1,798 -902 -34 Agroindustry 2,200 1,670 -502 -23 Total 11,000 10,599 -401 -4 The difference between SAR projections and actual costs reflects the actual demand for loans in each category and the delay in implementation of the date palm plantation component, and the subsequent transfer of funds between categories (see paras. 3.05-3.12). - 30 - 3.03 The SAR financing plan compared to actual Project financing is summarized below. Project Financing (Z of total) SAR Actual Sub-borrowers contribution 23 24.5 Government grants 10 7.5 Government loans 10 3 BNT 12 14 Bank 45 51 Total 100 100 The substantially lower than expected financing by Government loans affected the financing from all other sources. The SAR projections of financing by Government loans was based on the assumption that Government loans would be substituted for BNT financing in the case of loans to SCMV's for date palm plantations because BNT was not authorized to grant loans with maturities as long as 15 years from its own funds, and the assumption that agroindustrial subprojects would consist mainly of oil mills where the Government finances through subsidized loans investment in olive oil storage facilities. The delays in implementation and subsequent reduction of the date palm plantation component (para. 3.09) led to a corresponding reduction of the portion financed by Government loans, and other agroindustrial ventures than oil mills were financed (para. 3.11) with no Government financing. Project Start-Up and Implementation by Category 3.04 The first subloans to be approved in the quarter following effectiveness of the Loan were to commercial farmers, and first disbursement by the Bank was made in the following quarter. Disbursement of loans to small farmers started only a year later, in late 1978. 1/ The first loan for investment in agroindustries was disbursed early in 1978. Disbursements for loans to small-farmer associations for date plantations started only in 1981. Category I - Loans to Small Farmers 3.05 Besides direct loans to small farmers, defined as those whose yearly income does not exceed D 600 (1977 prices), the Category included loans to 1/ Calendar year - as in all future references. - 31 - SONAM (SociAtA Nationale de Motoculture, a semi-public organization providing tractor and harvesting services to small farmers) and to service cooperatives providing similar services. 3.06 While BNT's lending to SONAM for purchase of farm machinery started promptly, there was a 12-month delay in disbursement of loans to small farmers. The main reason for this delay was that, because of the important number of subloan applications expected under the project and of the shortage of technically competent agricultural staff within BNT, mostly at branch level, it was decided that subloan applications from small farmers would be appraised technically by the Regional Services of the Ministry of Agriculture, who were not familiar with and had to be trained to use subproject appraisal procedures required by BNT. I/ Another reason for the initial delays was the reluctance on the part of MDA credit department to approve subloans when the terms and conditions were not exactly the same as FOSDA terms. It was only in May 1978 that the Ministry of Planning informed the Bank that the MOA had agreed to comply with the uniform lending terms and conditions agreed during negotiations. 3.07 During 1979 and the first half of 1980, incoming subloan applications were abundant. However, individual subloan amounts were substantially smaller than expected at appraisal (para. 3.15) with a corresponding adverse impact on total disbursements. In 1981, loan applications dropped to a trickle (only 24 were received during 1981), essentially because of improved availability of FOSDA funds which were preferred by farmers because of their simpler administrative procedures, no requirement of field appraisal and no preconditions for investment in wells, livestock, etc. Subloan applications were also negligible during 1982 and BNT requested a third postponement of the Loan Closing Date to December 31, 1983. The Bank did not agree to this request, but accepted to continue to disburse until June 30, 1983 for expenditures under subloans approved prior to December 31, 1982 (telex dated January 21, 1983). To speed up disbursements the Bank also agreed (telex dated December 15, 1982) to BNT's proposal to transfer to the Project recent subloans approved on FOSDA funds but qualifying for financing under the Loan. Bank disbursement resulting from this reallocation amounted to USt535,936. By June 30, 1983, US$4,593,000 had been disbursed, nearly US700,000 more than originally allocated under this category. Category II - Loans to Commercial Farmers 3.08 Lending under this category proceeded without difficulty and at a pace corresponding to appraisal projections. This was mainly because BNT had prior experience in this type of lending, no major organizational changes were required, and most importantly because loans for mechanization were not eligible for FOSDA financing. The first subloans were approved in the quarter following Effectiveness. By the end of 1980, two-and-a-half years after Effectiveness, BNT had approved 652 subloans totalling D 2,812,000 (the appraisal estimate for the total amount of loans under the category was 1/ This also contributed to poor loan recovery (see para. 3.28). - 32 - D 2,956,000, including contingencies) and disbursed 500 loans amounting to D 2,156,000. By the end of 1981, the category was overcommitted and BNT transferred some of the pending applications to FOSDA and some to the Bank's Third Agricultural Credit Project (Loan 1885-TUN). US320,000 from category IV (Agroindustries) was also used to partly meet the substantial demand for subloans from commercial farmers. Loan funds for commercial farmers were fully disbursed by June 1981. Category III - Loans to Small Farmer Associations for Date Palm Plantations 3.09 BNT's lending under this category started only in the second quarter of 1981, some 6 months after the original Closing Date of the Loan. Initial reasons for the delay were (a) lengthy administrative and legal requirements needed to be fulfilled to establish the small farmer associations (SCMVs); (b) difficulty for SCMV members to assemble the minimum equity capital required; and (c) delay by MOA agents to provide the technical assistance for the design of the civil works (land leveling, irrigation infrastructure, drainage, etc.). A contributing factor was the apparent preference during 1978-1979 by the Tunisian authorities to allocate the underground water resources to new Government sponsored plantations rather than the SCMVs. When this attitude changed in 1980 and the Government approved the creation of 6 SCMVs, subloans were processed rapidly. BNT's commitments for the six subloans were D 2.0 million, representing a potential Bank disbursement of about US$2.5 million, or about US600,000 more than amounts available for the purpose. 3.10 It was expected that the first three SCMV plantations would be completed by June 1982, and the remaining three by April 1983. However, progress was slower than expected because (a) there was only one rig available for drilling; (b) there were difficulties to find suitable contractors to execute the irrigation infrastructure works; (c) MOA technical assistance was, at the initial stage, inadequate in relation to demand; and (d) temporarily, there was a shortage of suitable date palm seedlings. By the end of 1982, when it became clear that the funds available under the category would not be disbursed before the Closing Date, BNT proposed that FOSDA subsidies, which to date had been disbursed before the subloans, be deferred and that the subloan amounts be correspondingly increased. The additional subloans qualified for Bank disbursements. 1/ The Bank agreed to this procedure (telex dated December 15, 198?), which resulted in Bank disbursements of USt202,800. By 1/ BNT's loans to each SCMV are global loans covering all contemplated investment to establish date palm plantations (tubewell, irrigation infrastructure, land levelling, etc.). As a normal banking practice, BNT should have first disbursed subloans against the accrued expenses, and only when the agreed upon total loan amount was fully utilized, disburse the Government subsidies. Instead, BNT was considering each application as a separate investment and was first disbursing the Government subsidy and then the subloan. This procedure was changed to conform with normal banking practice in December 1982, as discussed above. - 33 - June 30, 1983, total Bank disbursement under the category amounted to only US$1.1 million, US$800,000 short of the original allocation and about USl.l million short of funds needed to complete Bank financing of these loans. The Bank agreed to finance their balance under the Third Credit Project. Category IV - Loans to Agroindustries 3.11 As in the case of lending to commercial farmers (para. 3.08 above), lending under this category proceeded smoothly, and except for the type of investments financed (see para. 3.21), according to apprai.-al projections. By the end of 1980, the original Loan Closing Date, BNT disbursed under the category D 1,169,000 representing Bank disbursements of US$2,122,000, some US378,000 less than originally allocated. The Bank agreed to BNT's proposals to finance further agroindustrial investments from funds available under the Third Credit Project (Loan 1885-TUN) and to use of unused balance of Category IV to finance loans to commercial farmers where BNT's commitments exceeded the original allocation. Amendments to Loan Agreement 3.12 The Loan Agreement was amended once, in October 1979, to allow financing under Category I of loans to production cooperatives in addition to service cooperatives. This amendment did not constitute any major change in the Project description or the Project objectives. Credit to production cooperatives had earlier been excluded because, as a group, they were assessed to have a poor overall performance. The Bank agreed to include on a selective basis those production cooperatives which met agreed criteria for good management and performance. This inclusion was justified on the grounds of being able to provide credit to a larger number of credit-worthy small farmers while encouraging institutions which were performing satisfactorily. Disbursemencs 3.13 Bank's disbursement covered 75% of BNT's disbursements under the Project. The SAR estimate of the disbursement period was 4 to 4-1/2 years and all funds were expected to be disbursed by December 31, 1980. Actual disbursements took two-and-half years longer and as of June 30, 1983, US$11,901,360 were disbursed. The balance of the loan (US98,640) was cancelled as of the same date. About five months of this total ielay of 30 months is due to delayed Board presentation 1/ and four months aue to delayed effectiveness of the Loan. Remaining delays are to be attributed to slower disbursements than anticipated for new activities by small farmers and SCMVs for date palm plantations (paras. 3.05-3.07 and 3.09-3.10). The projected and actual disbursements are summarized below and details are given in Annex 2. 1/ The estimated disbursement schedule in grey cover SAR maintained the estimate of yellow cover in spite of delayed Board Presentation. - 34 - Cumulative Disbursements (U million) IBRD FY Appraisal Estimate Actual X of Estimate 1977 2.7 - - 1978 8.3 1.7 20.0 1979 11.4 3.6 31.6 1980 11.8 7.1 60.2 1981 12.0 9.1 75.8 1982 10.4 86.7 1983 11.9 /a 99.1 /a UStO.1 million cancelled. Sublendie Terms and Conditions 3.14 Agreed sublending terms under the project were the following: Interest Term Grace Period Category Rate (YearZ) (Years) Small farmers /a 6% 5-8 2 Commercial farmers Grain farms 7% 8 1 Dairy farms 7% 8 3 Mixed farms 7% 8 1 Date palm plantations 6% 15 8 Agroindustries 8% 7 2 /a Defined as farmers whose income derived from farming does not exceed D 600/year (1977 prices). It was a condition of negotiations (see para 3.01 above) that the same lending terms and conditions as agreed for this project at negotiations would apply for all agricultural lending in Tunisia. 1/ Some discrepancies regarding 1/ Interest rates for loans to small farmers were substantially lower prior to loan effectiveness. - 35 - procedures and emphasis on repayment between Governmeue's FOSDA and the project however remained (see para. 3.29). Number of Subloans and their Size 3.15 The table below compares the actual number of subloans and their size with appraisal estimates. Appraisal Estimate Actual Category No. of Average Loan No. of Average Loan Subloans Amount /a Suboans Amount (D) (D) Small Farmers Individual farmers 660 2,390 3,396 /b 440 Cooperatives /c - - 30 13,015 SONAM /c - - 1 539,293 Commercial Farmers 275 5,230 616 4,359 Date Palm Plantations (SCMVs) 4 /d 250,740 6 312,400 le Agroindustries 18 84,610 14 83,481 /a Without contingencies. /b Including subloans transferred from FOSDA. /c Loans to SONAK and cooperatives although specifically authorized in the LA and mentioned in the SAR were not singled out in the SAR cost table. Each subloan to cooperatives represents in average about 50 direct beneficiaries. /d About 300 beneficiaries in total. te Part of the subloan is being financed by the Third Credit Project. Based on the number of subloans the actual number of beneficiaries was about 5,825 (instead of estimated 1,250), an increase of about 365%. The increase in the number of beneficiaries is partly the result of a substantially larger number of subloans than expected to small farmers and partly of the inclusion of production cooperatives under this category. The SAR estimates were based on farm models in which subloans ranged from D 955 to D 5,800, and in all cases were expected to finance integrated farm development projects comprising investments in well improvement, pump, irrigation equipment and cultivator or in livestock, stable, and small irrigation. In reality, small farmers were reluctant to invest generally in more than one item, with, as a result, a much larger number of subloans and beneficiaries (there were very few cases where one farmer benefitted from more than one subloan). - 36 - 3.16 SAR projections proved to be accurate in the case of commercial farmers and subloans to SCMVs. In case of loans to agroindustries, the numerical estimates proved to be quite accurate, though this is quite accidental as investments other than projected were financed (see para. 3.21). Investments Financed 3.17 The investments actually financed compared to the SAR estimates are summarized below: Category Small Farmers Appraisal Estimate /a Actual (D'000) (D'000) Individuals: Small irrigation 656 (27%) 653 (19%) Cultivators 195 (8%) - Livestock & stables 1,300 (54%) 920 (26%) Agricultural equipment 276 (11%) 239 (7%) Plantations - 94 (3%) Plastic greenhouses - 306 (9%) SONAM: Agricultural equipment and spares - 275 (8%) Farm machinery - 461 (13%) Cooperatives: Farm machinery - 34 (1%) Livestock - 523 (14%) Total 2,428 (100%) 3,505 (100%) /a Excluding contingencies. 3.18 Generally the investments financed and the amount of funds invested in them were in line with the SAR estimates, with two exceptions: a significant number of cultivators were not introduced as expected, and considerable investment in plastic greenhouses was made contrary to projections Technical innovation by small farmers to adopt, for example cultivators, were found to be too optimistic in view of the weakness of the existing extension services. The unexpected introduction of plastic greenhouses had not been foreseen because at appraisal, no such demand was identified (the introduction of plastic greenhouses in Tunisia is relatively recent). Subloans for livestock to small farmers were also smaller than ex-ected possibly because milk prices did not rise enough to maintain incentives for milk production. - 37 - 3.19 Category Commercial Farmers (D'O00) Appraisal Estimate Actual Livestock, stables, ensilage 382 (19%) 221 (6%) Agricultural equipment 82 (4%) - Farm machinery 1,591 (77%) 2,969 (82Z) Irrigation - 180 (5%) Greenhouses 90 (4%) 129 (4%) Plantation - 127 (3%) Total 2,055 (100%) 3,626 (100%) The SAR projections were largely in line with the actual results except for the investment in livestock, which was again less than projected because of failure of largely administered milk prices to maintain pace with other prices and thus provide incentive for production. The investment in plantations, which was not expected at appraisal, represented loans to STIL for completion of work initiated under the First Agricultural Credit Project. As under the First Credit Project, most loans for farm mechanization (60%) were for replacement of existing equipment. 3.20 Category Date Palm Plantations (D'000) Appraisal Expected Estimate When completed Wells, Pumps, Irrig. Network 541 (34%) 1,590 (48%) Land Prep. & Planting 435 (27%) 1,438 (43%) Equipment & Storage 76 (5%) 160 (5%) Plantation Maintenance 539 (34%) - Livestock - 129 (4%) Total 1,591 (100%) 3,317 (100%) All financing was close to projections except for plantation maintenance, for which there was no need. Later, studies made in 1982 by BNT and MOA confirmed that once the plantation investments are completed, maintenance - 38 - expenditures could be financed by revenues from intercropping, without recourse to loans. Subloans for the purchase of livestock, although not foreseen during appraisal, were justified and authorized because of the complementarity of livestock and date palm plantation activities. 3.21 Category Agroindustries (D'000) Appraisal Estimate la Actual Oil mills (18) 1,523 - Fruit & vegetables processing (4) - 684 Tomato juice & concentrate (1) - 120 Seed processing (1) - 32 Cold storage (1) - 135 Meat processing (1) - 37 Tuna canning (1) - 129 Olive canning (1) - 231 Animal feed (3) - 186 Soft drink production (1) - 116 Total 1,523 1,670 /a Excluding contingencies. The actual results were significantly different from projections as not a single oil mill was financed under the Project, and, on the other hand, 14 other ventures were. At appraisal, an increasing need was seen for new mills in the north and center, where the deficit in extraction capacity was estimated to reach 3,200 tons/day by 1981. To meet this shortage some 100 new mills were anticipated and BNT was already financing 6-8 each year (30 tons/ day capacity). The Project mainly provided for 18 such mills, but also allowed for financing of other agroindustrial ventures. The primary causes of this discrepancy between SAR projections and actual results are: (a) requirements for new mills in the north were overestimated; (b) agroindustrial subprojects did not qualify for FOPRODI subsidies .(para. 3.29(a)) and some potential borrowers insisted on getting these subsidies and thus refused to take Project financing; and (c) for reasons difficult to establish, BNT financed some oil mills (8 in 1977-1978 alone) from its own resources without using the Project facility. Monitoring and Evaluation 3.22 In 1977, BNT prepared a proposal for a system to monitor the impact of subloan investment at the farm level. The proposals were found by the Bank - 39 - to be complete and well designed, but possibly too ambitious in view of BNT's staffing constraints. The system started to be implemented early in 1978 on a sample of 10 sub-borrowers with the intention to progressively expand it to cover 10-15% of all sub-borrowers under the Project. However, already in the first year of implementation, it was found that farmers frequently changed their cropping pattern as compared to the one originally programmed which made the results difficult to interpret. Moreover, some farmers became reluctant to cooperate with BNT agents because of a risk of exposure of their financial situation. Monitoring was thus interrupted in 1979 and resumed in 1980 on a sample of 25 farmers and was extended to 58 farmers in 1981. 3.23 The final results and usefulness of the effort proved to be rather doubtful. This can be ascribed to several causes: (a) The system as developed by BNT was a priori too ambitious and did not take into account the often low level of formal education of the farmers; (b) The farmers, in many instances, did not fully understand the purpose of the exercice and were afraid to collaborate or furnished erroneous information; (c) The level of supervision by BNT was inadequate due to staff constraints and because BNT was not convinced of the importance of the monitoring concept; and (d) The prevailing practice in Tunisia to rent land which induces huge variations in cultivated areas from one year to the next. Audit 3.24 The satisfactory audit of BNT's accounts was a major issue during implementation of the First Credit Project. BNT, as a condition of Loan Effectiveness, made new arrangements for its audit and provided the Bank with a satisfactory audit report for 1975. While its old auditors SIMM& (Socidt& Internationale Maghrebienne du Management et d'Audit) continued as B4T's auditors, they, however, had the assistance of international auditors Coopers & Lybrand. 3.25 Over the project period, final audit reports were received regularly, but six or more months later than specified in the Loan Agreement (six months after the end of the fiscal year). The reason for the delays was that the auditors could not finalize their audit before BNT's General Assembly had approved BNT's financial statements, which had to be cleared by a Comissaire aux Comptes (generally six months after the end of the fiscal year). This constraint has been taken into account in the Third Agricultural Credit Project which requires the Audit reports to be submitted to the Bank not later than four months after the approval of BNT's financial statements by its General Assembly. - 40 - 3.26 The delays in audit procedures imply that the auditors' comments cannot be taken into account in the preparation of the financial statements of the year concerned. In practice, however, the recommendations contained in the auditor's reports have been taken into account the following year, provided BNT's Board agreed with their substance. BNT has used this external audit more like a management tool to improve its operations rather than a legal obligation since its accounts are already verified by the Commissaire aux Comptes. 3.27 The auditors' reports were of acceptable standard and quality but in all cases provided only a qualified approval of the accounts. However, the number of qualifications has considerably decreased since 1976. The substantial improvement of the auditing standards over the Project implementation period should be considered an important achievement of the Project. Recovery of Subloans 3.28 BNT's loan recovery performance before the Project was, and during the Project period remained, unsatisfactory. A comparison of BNT's loan recovery under the Loan, FOSDA and agricultural lending from own resources is shown below. BNT's Rate of Recovery /a Second Agr. Credit FOSDA Own Funds Overall 1978 79% 57% 91% 76% 1979 40% 53% 85% 71% 1980 62% 55% 88% 72% 1981 68% 54% 91% 73% 1982 66% 52% 91% 72% /a Principal recovered as percentage of principal fallen due as of December 31 of each year. BNT's practice is not to write off any bad debts and overdues includes potential bad debts dating from the first years of BNT's operations. While the recovery performance under the Loan has not improved over the last three years, it is somewhat better than recovery achieved for lending using FOSDA funds, but much worse than when BNT's own resources are applied to lending activity. BNT's lending to agriculture from its own resources is essentially short term and goes to large farmers. Rate of delinquency is therefore much smaller than in the case of lending to small farmers. The generally poor performance recovery on subloans financed from special funds (FOSDA, FOSEP, Bank, etc.) can be ascribed to three factors. First, BNT bears no, or only a small part of the default risk, and efforts to achieve better - 41 - loan recovery are simply not cost effective. Secondly, BNT receives a commission for its onlending activity which is related only to its total . outstanding portfolio financed from special funds, and this commission is not linked in any way to subloan recovery levels. Incentives and instruments to achieve improved loan recovery performance are critically needed. Thirdly, . subloan appraisal is done by the MOA extension staff, who have no responsibility for loan recovery, which is BNT's but without the latter involvement in identification of the borrower and appraisal and approval of the loan. Compliance with Loan Conditions 3.29 Although BNT complied with all major Loan covenants, it had difficulties with the covenant specifying that BNT should employ a qualified engineer specialized in agroindustries. Two attempts were made (in 1978 and 1980), but in both cases the engineer left after working for less than a year. As a final solution, BNT trained two of its agricultural engineers in appraisal of agroindustrial projects. The Government did not, however, comply fully with Covenant 3.08 of the Guarantee Agreement specifying that lending terms outside the Project should not be such as to adversely affect the Project. This was true in several cases: (a) under Category IV - Loans for Investment in Agroindustry where potential subborrowers had access since December 1975 to the Government FOPRODI (Special Fund for Industrial Promotion) funds which provided loans on soft terms (12-year repayment period including 5 years of grace; interest 3% p.a.) to cover the borrowers equity contribution of up to 30% of total investment cost, and in certain cases (investments below D 75,000), an additional loan covering the remaining 70% of investment cost (10 years, grace 3 years, interest 4%); (b) under Category I -- Loans to Small Farmers through the availability of FOSDA funds. Although the financial terms and conditions under FOSDA were made consistent with those of the Project, the administrative procedures were different because (i) FOSDA appraisal and approval procedures involved little or no technical evaluation and there were no technical pre-conditions for subloans; (ii) farmers could obtain a new loan even when they were in default on previous loans; and (iii) of the widely shared perception among farmers that FOSDA loans do not have to be repaid. After 1980, when FOSDA operations were decentralized and budgetary allocations increased substantially, these differences began to disrupt Project activities and subloan applications under the category were reduced to a trickle; and (c) under Category II -- Commercial Farmers continued to have access, except for mechanization loans, to FOSDA loans with more favorable terms than Project's subloans. Loans for mechanization are excluded from FOSDA financing. As a result, subloans for mechanization purposes represented 82% (in value) of total subloans to commercial farmers (para. 3.19). - 42 - IV. PROJECT IMPACT Yields, Production and Project Output 4.01 The SAR estimates of yields and output were based on an aggregation of six farm models, a model each of a date palm plantation and of an olive oil mill. The expected increases in yields were substantial and varied between 50 and 66% of pre-project yields. The increases in production were to be achisved, in case of commercial cereal farms, through more intensive cultivation of the land, and by progressively eliminating fallow periods, and in case of small farmers, by the change in crops and cropping patterns, and more intensive use of agricultural inputs. 4.02 The expected high increases in yields in the case of commercial farmers did not generally materialize 1/ as the farmers continued their old cropping practices. Where the Project investment was a first acquisition (as opposed to replacement), they tended to rent additional land in order to make full use of their equipment. An increase in production has therefore been achieved by cultivating more land rather than by more intensive farm practices on existing farms. In the case of small farmers, yield increases were also less dramatic than expected as the farmers generally did not invest in integrated farm development projects (para. 3.15). Actual increase in yields and production was due to a change in crops cultivated and wider use of small irrigation. In order to estimate the yearly average actual incremental production due to the project, 13 representative farm models were prepared covering about 80% of the actual number of subloans and more than 90% of the total investment cost of small and commercial farmers. A weighted aggregation of the production impact (either as incremental or prevention of decline) of each model led to the following results: Incremental production 2/ of Cereals 16,000 tons Fodder 2,000 tons Dates 3,100 tons 3/ Off-season vegetables 1,130 tons Season vegetables 18,360 tons The above figures are to be considered indicative and are subject to important variations due to climatic conditions, changing crop rotation, etc. Financial and Economic Returns 4.03 At appraisal, the expected financial rate of return of the farm models ranged from 15% to over 50%. In addition, it was 11% for the date palm plantations and 29% for the oil mill. The actual financial rates of return of the 13 representative farm models, as calculated for the PCR, range from 13% I/ Based on findings of BNT's monitoring unit. 2/ Average per year. 31 At full production -- year 18. to over 50%. 1/ These estimates are in line with those made for similar investments at appraisal of the Third Credit Project in 1980 when the FRR was calculated to range between 22% (tractor replacement by a commercial farmer) to over 50 (small farmers: irrigation and well improvements, pump, pure-bred cows and pasture). 4.04 The estimates of the FRR of the actual investments in agroindustrial investments range from negative (in the case of a particular fruit and vegetable processing unit) to 20%. The FRR of the date palm plantations was calculated at 14%. The aggregation of all the models led to the calculation of the FRR for the whole Project with, as a result, an FRR of 33%. 4.05 The appraisal estimate of the economic rate of return (ERR) of the farm component was 38%, of the date plantations 13%, of agroindustries 36%, and of the total project investments 34%. These rates were in most cases several percentage points higher than the FRR of the components. The same relation between the FRR and the ERR was found at appraisal of the Third Agricultural Credit Project, where the ERR was calculated for each farm model. No factors which would reverse this relation have intervened since the appraisal of the Third Credit Project, and it is therefore safe to assume that the overall actual ERR of the Project investments is not below the FRR of 33%. 2/ Institutional Performance of the Borrower 4.06 During Project implementation, BNT's staff in charge of agricultural credit operations expanded from 310 to about 530 in 1982, and comprised at the latter date 3 agricultural engineers, 12 assistant agricultural engineers, and 33 agricultural technicians. From the organizational point of view, several positive steps were taken: the Agricultural Credit Department was reorganized and credit decisions decentralized on a pilot basis; and a new service was created which was put in charge of (a) identification and appraisal of agricultural and agroindustrial projects; (b) management of lending under IBRD projects; and (c) training of personnel. 4.07 BNT's lending to agriculture 3/ in general expanded substantially during the project implementation period. 1/ With the exception of the case of a commercial farmer investing in livestock only, where the FRR was negative. 2/ No separate calculation of the Project ERR was made because of lack of reliable data concerning the actual results of project investments (see paras. 3.22 and 3.23). 3/ Although BNT is by far the largest source of institutional credit to agriculture in Tunisia, its main activity -- and main source of income -- is lending (mainly short-term) to industrial and commercial enterprises. BNT's income from agricultural lending represented in 1978 and 1982 (first and last of Project implementation years) only 12.2% and 11.0% respectively of its total income, and 15.8% and 13.6% of interest and commissions received. - 44 - BNT's Lending for Agriculture (D Million) 1976 1978 1980 1982 Agricultural Commitments 83.6 113.4 145.0 208.0 of which Agricultural production & investment 49.8 75.0 104.0 144.4 Marketing (wheat, oil, wine) 33.8 38.4 41.0 63.6 of which under the Project - (2.2) (6.3) (7.5) Agricultural Disbursements 16.7 33.3 39.6 45.1 Short-term 6.7 15.6 20.4 24.0 Medium- and long-term /a 10.0 17.7 19.2 21.1 Number of Branches 34 58 58 61 /a Generally from special funds. Over the seven-year Project implementation period, BNT's short term lending to agriculture increased by 258% and medium- and long-term lending by 111%. This substantial increase was possible because of improvement and streamlining of BNT's organization and because of increased availability of resources. BNT's Financial Position 4.08 BNT's comparative balance sheets and income and expenditure statements during the period of Project implementation are shown in Annex 1. BNT's overall financial position remained sound over this period, BNT's total assets 1/ increased from D 239 million to D 567 million, sight deposits from D 63 million to D 224 million, and income from D 12 million to D 36 million. Net profit had not quite kept up with BNT's growth rate and increased during the same period from D 1.5 million to D 2.8 million. The ratio of interest income to total assets improved from 4.1% to 5.2% and the ratio of interest income to outstanding loans, after a decrease from 6.2% in 1977 to 5.3% in 1979, increased again to 6.5% in 1981 and 7.8% in 1982. Interest and commission paid increased from D 4.7 million to D 17.0 million during the same period, and represented from 4.9% (1977) to 10.5% (1982) of outstanding medium- and long-term borrowing. 2/ Profit before tax as a percentage of 1/ Excluding contingent items. 2/ Bonds and term deposits, banks and correspondents, special funds. equity 1/ decreased from 20% in 1977 to 16.3% in 1981 but increased again to 24% in 1982. Profit after tax as percentage of equity remained fairly constant at about 10%. 4.09 BNT's agricultural lending operations, however, have not been profitable. Under the Third Agricultural Credit Project (1885-TUN), BNT set up a cost accounting system to permit the allocation of costs of and revenues from its various activities and assess the financial viability of its agricultural credit operations. Results from the study of BNT's agricultural lending in 1979 and 1980 indicate that the institution incurred a net loss of about D 1.0 million (US$1.4 million) each year on its agricultural operations, and that an increase of about 2% of the available financial margin would be necessary to ensure their financial viability, along with other measures to improve recovery levels (para. 3.28). Current dialogue between the Bank and the Tunisian Government aims at resolving these issues. 4.10 BNT's provision for bad debts increased from D 2.7 million in 1977 to D 8.1 million in 1981 2/ covering 97% of BNT's overall bad debt risk (as estimated by the auditor -- BNT's own estimate of bad debt risk is lower and the bad debt provision covers it fully). BNT's practice is not to write-off any bad debts and, as a result, about 28% of the arrears on its books date back three years or more. The Bank's recommendation to adopt more realistic practices has not been followed. 4.11 The growth of BNT's lending was possible because of a sustained increase in deposits and an increase in Special Resources (FOSDA, FOSEP, Bank, etc.). Sight deposits increased over the period 1977-1982 from D 63.5 million to D 224.0 million, term deposits from D 29.4 million to D 48.5 million, and Special Resources from D 48.5 to D 102.0 million. BNT's sources and application of funds over the Project implementation period is shown in Annex 1, Table 3. Bank Performance 4.12 Since it was a follow-up project with BNT, the assessment of the Bank's performance needs to be made in light of strengthening achieved in BN9T's agricultural lending activity, especially, the new component dealing with small farmers. Prior to the Project, investment credit for small farmers was financed exclusively from Government FOSDA funds with little or no subproject technical appraisal or supervision of subloans. By including under the Project a substantial small farmer credit component, the Bank encouraged the Government not only to improve availability of investment credit to small farmers, but also to introduce sound appraisal methods for subprojects and better discipline concerning loan repayment by this category of farmers. This was not achieved fully because of continued and increased availability of 1/ Capital, reserves and provisions. 2/ Auditor's report for the fiscal year 1982 was not available at time of the drafting of present report. - 46 - FOSDA funds. In retrospect, it appears that during appraisal the magnitude of the problems connected with the introduction of such lending methods had been underestimated, and Government's commitment to set small farmer investment credit on a sound basis was probably overestimated at the time of appraisal and negotiations. Throughout Project implementation, the Bank maintained a consistent and firm position with regard to this component, which did finally take off and during the extended period was disbursed fully. 4.13 On the other hand, the Bank should probably have taken a firmer stand during implementation, when it became apparent that contrary to the spirit of Guarantee Agreement, the procedures and conditions (not the financial terms) of lending from other funds were more liberal to subborrowers than those agreed under the Project. This issue became a central one for the Third Agricultural Credit Project now under implementation. 4.14 The timing of Bank supervision missions was adequate and the missions reflected a reasonable level of continuity in terms of staff participation. V. LESSONS LEARNED AND CONCLUSIONS 5.01 The major lessons learned during the processing and implementation of the Project are: (a) BNT's Loan Recovery performance varies considerably from dismal (below 50% of due) for loans granted with Government resources and for which the Government bears 100% of the default risk, to poor for loans granted under the Project (65%) for which BNT bears part (25% for small farmers and 100% for commercial farmers) of the risk of default, to satisfactory (90%) for loans granted on BNT's own resources and for which it bears fully the risk of default. The Bank failed to establish adequate incentives for BNT to improve its loan recovery efforts and performance by (i) not making BNT gradually assume a larger part of the subloans' default risk, and (ii) not linking BNT's remuneration, fixed under the Project (Guarantee Agreement) at 3% of the subloans' principal amount outstanding at the end of the year, to recovery performance. (b) Role of the Ministry of Agriculture in subloan appraisal was a major contributing factor to the delay in subloan approvals, as well as in the eventual poor performance in loan recovery. Unlike under the credit operations financed by its own resources, BNT's role in appraising the sub-borrowers' creditworthiness and the proposed investments was reduced, in many cases, to providing an accounting function. On the other hand, the MOA staff, who approved subloan application, had no responsibilities for loan recovery. Such a split in the responsibility for subloan appraisal and loan recovery, and the involvement of staff outside a credit institution in purely credit functions, proved to be unsatisfactory. On the one hand, it would have been difficult to request that BNT bear the full default risk for subloans (categories A and B) technically appraised by the - 47 - services of MOA and not by its own staff; on the other hand, reduced exposure to default risk did not provide BNT with sufficient incentive to increase its recovery efforts. The lesson here is clear: an agricultural credit institution should be given full responsibility for loan appraisal and approval and bear the full default risk of its lending operations. (c) Disbursement of Subloans to Small Farmers was slow essentially because of competing FOSDA funds (paras. 3.07, 3.14 and 3.29). While considerable attention was paid to achieving equality in loan terms and conditions, other procedures and conditions, such as quality and speed of the appraisal and processing of loan applications, were not emphasized. The lesson to be learned is that it is not sufficient to have the same financial lending terms, but that all conditions such as subloan appraisal procedures and incentives provided to the lending institution to undertake its assigned tasks should be at least comparable for all sources of funds in order to ensure satisfactory performance under a Bank loan. (d) Delayed Disbursement to Small Farmer Associations (SCMVs). Considering that SCMVs were new "institutions" and there was little experience in Tunisia and in the Bank as to their effectiveness, the implementation complications which plagued this component during the early years should have been better foreseen during Project preparation. The implementation of this component could have been facilitated by greater attention before appraisal to its key design features. (e) The Monitoring System for assessing the impact of the Project investments did not produce satisfactory results (paras. 3.22-3.23). In retrospect, it appears that the Bank should have paid more attention to this issue, both from the point of view of the objectives and design of the M&E system, and of convincing the Government and BNT of the importance of this aspect of the project cycle. The lesson is that monitoring and evaluation systems work only if they are designed to provide a true service to the end user (the Borrower) and therefore win its full-hearted commitment. 5.02 In summary, despite considerable implementation difficulties, especially for the components dealing with small farmers, Project objectives set at the time of appraisal were generally ret, albeit with considerable delay. BNT has improved its lending procedures and its lending to agriculture during the Project period increased substantially. Although interest rates to farmers remained negative in real terms, they were increased substantially as compared to the pre-project situation (from 3-4.5% to 6% for small farmers, and from 6% to 7% for commercial farmers). However, at the end of the Project period, BNT, as an effective financial intermediary for agricultural lending, needed further reform on issues which are currently the subject of a dialogue between the Government and the Bank. - 48 - 5.03 Problems of low level of loan recoveries, lack of decentralization of BNT's loan appraisal procedures, inadequacy of a financial margin available to BNT for its agricultural lending operations and resolution of the interest rate issue became the objectives of attention of the Third Agricultural Credit Project, which was approved by the Board in 1980. BNT has since then set up a cost accounting system to review the financial viability of its agricultural lending operations, has created a loan collection unit to strengthen its recovery procedures, has defined a training program for its agricultural credit staff and has decentralized, on a pilot basis, the task of appraisal and approval of subloans to small farmers. Some of the other measures necessary to resolve the major issues in institutional credit are outside the responsibility and control of BNT. In particular, Government is directly responsible for establishing the policies regarding interest rates, the spread available to financial institutions and creating a climate for financial discipline and loan repayment. Increased recognition is "eing given by the Government to these issues. Some actions have recently beea taken by the Governent (privilfge d'Etat for the recovery of agricultural loans) but furtber action to resolve these issues is critically needed for the sound gromth of a viable agricultural credit system and for continued involvement of the Bank in supporting this subsector. - 49 - FFi1Er Com'fEr3Der IEn AReg I Table 1 71UMMA SED AMIU MrAL CIEDIr EMCr (on 130-1T0 ffr - gmmarized Emlen Shesta 1976-82 (D 'O0) 1976 1977 1978 1979 1980 1981 1982 AS~S AMF Camh n Hand, caae~ Back, Postal Checks 1,340 1,881 1,794 2,83 2,029 1,974 1,852 Caisse, cP, Cr B~s n Ccaeqanfct 5,075 2,714 13,907 25,407 22,198 21,295 6,028 B-nqus et rrensparne Ge , Br~ and Tne,ry Bills 16,072 20,003 22,673 34,228 39,765 46,646 53,470 Effets Publice et Autres Titres ills Di-e~e 30.948 28,875 36,586 47,784 69,862 85,312 94,682 Porteauille Ec-me Ca~n Ar==" Cnbit> 55,073 54.476 64,633 76,675 81,208 89,392 107,309 Cau3~e cmra Débiteus ~y Dat~ 93 103 76 53 2,006 4,880 2,691 Dibiteurs Dier Equity Iresmanrs 3,504 4,128 5,951 7.197 7,370 8.234 9,561 Participa:ions FiA Asts. Nn-Pr~xtimn Assets 2,198 2,586 3,114 3,727 4,092 4,452 4,597 lsmbilisatiom et No-valen net ice11 ,= 35.675 76,561 57,8 63.117 83,451 121,585 182,667 Divers • kar~l 149,978 191.327 206,579 260,991 311,981 383.770 462,917 S~u-Total Specal 1r~d Inme.rnenta 41.533 47,497 57,743 67,915 74,714 87.835 106,234 Prts mr Dotations Späciales 1l 191,511 239,274 264.322 328,906 386,695 471,605 67,151 Sundxy Ios Dier H~ Bilan Grante Dekor 45.513 57,084 111,185 128,368 155,626 235,568 247,471 D~itu Per Caution et Acptle Ingae~r Caedits 2,694 2.688 2,674 2,854 2,995 3,001 2,969 CrEmr~a de l'Etat I Recourir 1Drm 48,207 59,772 113,859 131,222 158 621 238,569 250,&0 1D CROD 1m 39718 299,046 378xr181 460128 565t316 710,174 817,591 GLwe uDL L AE A- flnD EHNDS pASS Depasia i Sight 61.249 63,481 94,777 123,003 156,871 185,436 223,956 D~pa å Vue Ho~s al Term Deposita 29,589 29,449 32,619 45,791 53,4% 54,436 48,506 Hans et Canes 1 Terme Baks atd Corrxepodta 7,598 16,482 6.393 7,655 7,130 12,693 10,558 H-BrL- et COrrespoa--n Saky Creditorm 10,630 7,400 8,096 9,536 6,919 6.106 6,999 Crediteu Diverm 14 ei1 m 23.849 57,972 4346 50,196 56,606 88,556 135,702 Dir S.totalr 132,915 174,786 186,229 236,181 281,020 347,225 425,719 Sous-Total Special Fummi Nssre Speciales EOSD 17,780 21,175 26,869 33,842 38,686 ) ) 6,010 7,340 8,658 9.507 10,754) ) 0SEP 2,758 2,761 2,765 2,765 2,769 ) ) lUD-D 6,899 6,783 7,520 9,230 10,372 1 ) -IL th 10,989 10,466 12,539 15,306 18,023 ) ) Divers Suaotal 44,436 48,525 58,351 70,650 80,602 89,862 102,083 SOU-Total Provisims Provisions ine11m~m=• Rinka 3,400 3,400 3,540 684 832) %8) Riaqueu Divers Eqity Insm ~f 181 323 574 721 793) 897> D)p~iation dem Titres W Db 2,017 2,721 3,209 4,29 8,059) Catetieux Subttal 5.598 6,444 7,323 5,534 7,190 9,924 12,790 Sous-Tatal RCe 3,129 3,618 4,80 8,657 9,711 0,223 11,785 RMve Capital 4,000 4,400 6.000 6,000 6,000 12,000 12.000 Capital Yerly Pfit1433 1,503 1 611 1 884 2,172 2,371 2774 Tfice de 'Mercice Subotal 8,562 9,521 ,?.W T' T7W 24,596 26,559 SOUs-Ttal lDDm 191.511 239,274 266,322 3290 386695 471,605 567,151 1Dum u Divers It Bilan rrea emi.. 45,513 57,084 111.85 128.368 155.626 235.568 247,471 EAme~a per caution et acceptatioa Gmern Cit åaa-~a to Collect 2,69 2,688 2,674 2,854 2,995 3,001 2,9 Etat - rGe Cre 1 Reanurir lEDL 48,207 592 113.859 131,222 156 238569 250,440 10tAL G(UD IMDL 239,718 299,046 378181 4W0128 565 3 710,174 817,591 CR~ 1lDL (2333. p.30) le. 1III j [in qja Iij mass a es as ser s a e bn- mi I I- w0a ases r 1 aran aI e imisEs PROJECT COMPLETION REPORT TUNISIA SECOND AGRICULTURAL CREDIT PROJECT (Loan 1340-TUNI BNT - Sourceas and ADplication of Funds 191n 192 19.80 ili 1982Z Net Profit 1.6 1.9 2.2 2.4 2.8 R6sultat Net Depreciation l. 1i 1. LA LA Amortissements et Provisions Cashflow 2.7 3.6 4.3 5.9 6.2 Casflow Increase in Capital 1.6 - - 6.0 - Augmentation du capital Increase in Reserves 1.4 4.0 1.3 0.7 1.6 Augmentation des r6serves Other Increases Autres Augntations Sight Deposits 31.3 28.2 33.9 28.6 38.5 06p6ts a vue Term Deposits 3.2 13.2 7.7 0.9 - Bons et Comptes & terme Special Resources 9.8 12.3 9.9 9.3 12.2 Ressources spiciales et i plus d'un an Sundry Creditors 0.7 1.4 - - 0.9 Cr6diteurs divers Miscellaneous Liabilities - 5.7 6.2 31.7 38.4 Divers Passif Decreases Diminution Liquid Funds - - 3.5 6.5 15.4 Disponibilit6s Sundry Debtors - - 2.2 06biteurs divers Bonds l-= - Bons TOTAL SOURCES 70.0 68.4 66.8 89.6 114.4 TOTAL RESSOURCES APPLICATION OF FUNDS EMELOS Increase in Lending Augmentations des engagements Own Resources 17.9 12.0 4.5 11.0 17.9 Sur fonds propres Special Resources L3A .4 .u _U l .11 -1. Sur ressources spiciales Subtotal 27.7 22.2 11.3 24.2 33.3 Sous-total Increase in: Autres augmentations Discount Portfolio - 11.2 22.1 15.4 9.4 Portefeuille escompte Equity Participation 1.8 1.2 0.2 0.9 1.3 Participation Fixed Assets 0.7 0.8 0.7 1.1 0.1 Effets Publics Liquid Assets 21.3 11.3 - - - Liquidit6s Miscellaneous - 5.3 20.3 38.1 61.1 Divers Actifs Profit Allocation (previous year) 1.5 1.6 1.9 2.2 2.4 Affectation r6sultats ann6e pr6cedente Decrease in: Diminutions Sundry Creditors - - 2.6 0.8 - Criditeurs divers Miscellaneous Liabilities 13.8 - - - - Divers passif Provisions - .LIt . - - Provisions TOTAL APPLICATION OF FUNDS 70.0 68.4 66.8 89.6 114.4 TOTAL EMPLOIS (2333E, p. 32) -52 - ANNEX 2 Table 1 PROJECT COMPLETION REPORT TUNISIA SECOND AGRICULTURAL CREDIT PROJECT (Loan 1340-TUN) Quarterly and Cumulative Disbursements (in US Million) IBRD FY Actual SAR Projections Actual as Quarter Ending Quarterly Cumulative Quarterly Cumulative Z of projection FY77 March 31, 1977 1.25 1.25 0 June 30, 1977 1.40 2.65 0 FY78 September 30, 1977 1.15 3.80 0 December 31, 1977 1.45 5.25 0 March 31, 1978 0.24 0.24 1.50 6.75 4 June 30, 1978 1.48 1.72 1.55 8.30 21 FY79 September 30, 1978 0.54 2.26 1.25 9.55 24 December 31, 1978 0.24 2.40 1.65 11.20 22 March 31, 1979 0.38 2.88 0.12 11.32 25 June 30, 1979 0.78 3.66 0.11 11.43 32 FY80 September 30, 1979 0.52 4.18 0.10 11.53 36 December 31, 1979 0.49 4.27 0.11 11.64 40 March 31, 1980 1.26 5.94 0.10 11.74 50 June 30, 1980 1.21 7.15 0.09 11.83 60 FY81 September 30, 1980 0.33 7.48 0.08 11.91 63 December 31, 1980 0.43 7.91 0.09 12.00 66 March 31, 1981 0.96 8.87 - 12.00 74 June 30, 1981 0.27 9.14 - 12.00 76 FY82 September 30, 1981 0.44 9.58 - 12.00 80 December 31, 1981 0.28 9.96 - 12.00 82 March 31, 1982 0.15 10.01 - 12.00 83 June 30, 1982 0.37 10.38 - 12.00 87 FY83 September 30, 1982 0.16 10.54 - 12.00 88 December 31, 1982 0.12 10.66 - 12.00 89 March 31, 1983 0.88 11.54 - 12.00 96 June 30, 1983 0.36 11.90 - 12.00 99 PROJACT COMPLETION REPORT TUNISIA ýECOND AGRICLTURAL CREDIT PROJECT (LOAN 1340-TUN) SI-isbursement Profile - Cåmparison of EMENA TypicÅl, SAR and actual 1 / i Typical Profile IBRD EMENA Agriéultural Credit Loans ----•---- SAR Disbursement!Profile .... *.*...*. • Actuål Loan Disbursement Profile ic ,z (Dy .4 ANNE 2 Table 3 PROJCT COMPLETI) REPORT TUNISIA ECOKID AGRICåLTURAL CRED T PROJECT dOAN 1340-TUA) Disburslment Profilé per Category (Compard to origin.1 Allocati Category - Small FaLers - egory-- G--omerc-ial--nmes--- - - --- - Category II - SCMv'siDate Palm Plantation * i ~ Category ris CaeoyV- Agroindjstries- 1 far. i '> li Ge 1 1*- :FYi *q* 1 L po , , & & - Redeyel El Guellar SFAX ý-ýKq p kon no Islands G GAFSA Mezzouna Mchares Rediryof Metlaoul El Guertor U- Sekhira, XiL- 34* T 0-2 E UR Mcihouio TOZEUR GABES +Hournt Sauk Nafta E l Harnmo Adilm KEBIII G A B E 5 Dior( Mareffi K E B l L l Mormota za(755 Douz MEDENINE Bant Keddoche. a ýan Gordane -33* ähournrossen 33-- lp T A T A 0 IJ l N E U 0 25 50 75 * National Capital l KKOMETERS i * Governorate Copitals rhis map has been prepared by The World BankS Maff exclusivety Poved Roods for the convenience 01 the Rodways readers and is exefusively for the 0 Rernada Map inlernal usa of The World Bank Arra JA se. Ports and the Inlernationai Finance Copporshon The denorninations Airports used and lhe boundarjes Shown 0 on this map do not impy, on lhe m Salt Lokes C) part of rhe World Bank and the A L G E R I A L I B Y A m Governorate Boundaries International Finance Corporation, 32t any p~ent on the legal slatus International Boundaries M any lerntory or any -ou endorsement or jcceplance of such boundaries 108 "'NIGE R CHAD 1