Tunisia - Second Agricultural Credit Project

Report No. 1133-TUN FRLE Copy Appraisal of a Second Agricultural Credit Project Tunisia October 22, 1976 Projects Department Europe, Middle East and North Africa Regional Office FOR OFFICIAL USE ONLY Document of the World Bank This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURR mCY EQUIVALENTS US$1.00 = D 0.417 D 1.000 = US$2.40 WEIGHTS AND MEASURES Metric System ABBREVIATIONS BCT = Banque Centrale de Tunisie - Tunisian Central Bank BDET = Banque de Developpement Economique de Tunisie - Development Bank of Tunisia BNT = Banque Nationale de Tunisie - National Bank of Tunisia BPDA = Bureau du Plan et du D6veloppement Agricole - Agriculture Plan and Development Office CLCM = Caisse Locale de Cr6dit Mutuel - Mutual Credit Bank CNEA = Centre National des Etudes Agricoles - National Center for Agriculture Studies COCEMO = Coopgrative Centrale de Motoculture - Central Cooperative for Tractor and Harvesting Services FOSDA F Fonds Spgcial de Dgveloppement Agricole - Agriculture Development Fund INAT = Institut National Agronomique de Tunis - National Agriculture Institute INRAT = Institut -ational de la Recherche Agricole fde Tunisie - Tunisian National Research Institute OMVWM - Office de Mise en Valeur de la Vallee de la Medjerda - Medjerda Valley Development Agency ONC = Office National des Cer'ales - National Grain Agency ONEP Office National de 1'Elevage et des Paturages - National Livestock and Pasture Agency ONH = Office National des Huiles - National Edible Oil Agency ONP = Office National des P8ches - National Fisheries Agency ONV = Office National des Vins - National Wine Agency OTD = Office des Terres Domaniales - Government Land Agency SCM = Societe' de Caution Mutuelle - Mutual Guarantee Society SCMV = Societe' Civile de Mise en Valeur - Farmer Association for Joint Farm Development SONAMO Socigte Nationale de Motoculture - National Company Providing Tractor and Harvesting Services STIL = Societe Tunisienne de l'Industrie Laitiare - Tunisian Company for Milk Industries FISCAL YEAR January 1 - December 31  APPRAISAL OF FOR OFFICIAL USE ONLY A SECOND AGRICULTURAL CREDIT PROJECT TUNISIA Table of Contents Page No. SUMMARY AND CONCLUSIONS ............................... i-ii I. INTRODUCTION .......................................... 1 II. COUNTRY BACKGROUND AND THE AGRICULTURAL SECTOR ........ 1 A. Country Background ............................... 1 B. The Agricultural Sector ............ .............. 2 C. Previous Agricultural Projects ................... 5 III. AGRICULTURAL CREDIT ...... .................. 6 A. General .......................................... 6 B. Banque Nationale de Tunisie (BNT) .... ............ 9 C. Caisses Locales de Credit Mutuel (CLCM) .......... 11 IV. PERFORMANCE UNDER THE FIRST AGRICULTURAL CREDIT PROJECT .......................................... 12 V. THE PROJECT ........................................... 16 A. Objectives ....................................... 16 B. Description ...................................... 16 C. Cost Estimates ................................... 18 D. Financing ........................................ 20 E. Procurement ...................................... 21 F. Disbursement ..................................... 21 VI. PROJECT IMPL.EMENTATION ................................ 22 A. Organization and Management ...................... 22 B. Lending Policies and Procedures .................. 22 C. Monitoring of Development Results .... ............ 25 D. Accounts and Audits .............................. 26 VII. BENEFITS AND JUSTIFICATIONS ........................... 26 A. Production and Marketing ......................... 26 B. Financial Benefits ............................... 27 C. Economic Benefits ................................ 29 VIII. AGREEMENTS REACHED AND RECOMMENDATIONS . . 29 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  Table of Contents (Continued) ANNEXES 1. Agricultural Sector Appendix 1 - Date Palm Plantations Appendix 2 - Olive Oil Production 2. Agricultural Credit 3. Banque Nationale de Tunisie (BNT) 4. Performance Under the First Agricultural Credit Project (Credit/Loan 263/779-TUN) 5. Representative Models 6. Project Cash Flow Statement 7. Estimated Schedule of IBRD Disbursements 8. Economic Rate of Return of the Project 9. Terms and Conditions of Sub-Loans  APPRAISAL OF A SECOND AGRICULTURAL CREDIT PROJECT TUNISIA SUMMARY AND CONCLUSIONS i. Agriculture provides about 18% of Tunisia's GDP and employs over 50% of the active population. Exports of agriculture products amounted in 1974 to D 100 million representing some 25% of Tunisia's total exports. The agricultural sector has lagged behind overall economic growth with a growth rate of 3.8% during the 4th plan (1973-76) as compared to 7.7% for the whole economy. One of the Government's main objectives during the 5th plan (1977-1981) is self-sufficiency in the main agriculture products and a balanced foreign trade in other food products. ii. Agricultural credit is provided mainly by the Banque Nationale de Tunisie (BNT). Several Government agencies are also involved in the provi- sion of short-term credit, while investment credit is mostly financed (through BNT) by special funds provided by the State budget or by foreign aid. At most 50,000 Tunisian farmers, i.e., 15% of all, have access to institutional credit, most of them large or medium-size farmers. Agricultural credit policy is char- acterized by the great number of institutions and programs involved and the diversity of financing terms under various funds. This framework, while allowing a certain flexibility of credit provisions, has had detrimental effects both in institutional and economic terms which were compounded by easy lending terms and conditions (low interest rates, high grant element). Recent developments show some progress in terms of definition of target groups and of interest rate levels. iii. BNT, a multi-purpose bank established in 1959 and controlled by Government, is the only institution ext2nding medium and long-term credit to agriculture in Tunisia. Funds relent by BNT for agricultural invectments are mainly Government, bilateral aid and Bank/IDA funds, and, only in a small pro- portion, BNT's own funds. BNT is well managed, has an adequate accounting system and competent staff, and has since its origin consistently shown a satisfactory profit. iv. Bank/IDA involvement in Agricultural Credit in Tunisia started with the First Agricultural Credit Project (Loan/Credit 779/263, US$8 million). The Loan/Credit which became effective in March 1972 was to finance part of Banque Nationale de Tunisie (BNT) lending to commercial farmers for investment in grain farm mechanization and dairy farming and to Societe Tunisienne de l'Industrie Laitiere for development of date palm plantations; financing of technical assistance to BNT was also included. The Project implementation during 1972 through 1974 was slow due mainly to competing funds on more con- cessional terms available to potential subborrowers. Once this issue was resolved project implementation and disbursements were satisfactory and as of October 1, 1976 total Loan/Credit disbursements reached US$7.4 million and the remaining US$0.6 million were fully committed.  - ii - v. This report appraises a Second Agricultural Credit Project for which a Bank loan of US$12 million equivalent is proposed. The Project would support a two-year lending program for on-farm development of small and com- mercial farms, for selected agroindustries and a four-year lending program for small holder date-palm plantation development. Banque Nationale de Tunisie (BNT) would be the Borrower and would be responsible for Project implementation. Whereas the First Credit Project aimed principally at in- creasing agricultural production through credit to large farmers and to a Government controlled company (STIL), the Second Project aims also at the improvement of the living conditions of small farmers with so far little or no access to institutional credit, and at the participation of associations of small farmers in pursuing development of modern date plantations. vi. The total cost of the Second Agricultural Credit Project is esti- mated at about US$26.3 million equivalent of which about 45% would be in roreign exchange. Financing of investments would come from the following sources: subborrowers would on average contribute 23% of investment cost, the Tunisian Government would contribute through grants for small farmers investment and for date plantation development 10%, and through loans for date plantation development an additional 10% of the cost, BNT would con- tribute from its own funds 12% and the Bank 45% of the Project cost. vii. The range of items to be financed under the Project is varied and would be purchased by individual subborrowers over a period of 2 years (4 years in the case of date plantations) and would not be suitable for bulk procurement or international competitive bidding. Therefore most of the procurement under the Project would be through normal commercial channels, and local bidding in the case of deep well drilling. viii. The financial rate of return of subproject as established on the basis of representative models would range between 11% and over 50%. The overall economic rate of return is estimated at 34%. With assurances ob- tained from BNT and the Tunisian Governinent during negotiations, the Project would be suitable for an IBRD loan to BNT of US$12.0 million equivalent for 12 years including a grace period of 4 years.  APPRAISAL OF A SECOND AGRICULTURAL CREDIT PROJECT TUNISIA I. INTRODUCTION 1.01 The Government of Tunisia and BNT have requested a Bank loan to help finance agricultural credit for investments in small and large scale grain and mixed farming, for dairy farming, for development of new date-palm plantations to be owned by groups of small farmers and for selected agroin- dustries. The Borrower for the Bank loan would be the Banque Nationale de Tunisie (BNT). BNT is already the Borrower under the First Agricultural Credit Project (Credit 263-TUN, US$3.0 million and Loan 779-TUN, US$5.0 million) which finances investments generally similar in purpose. This Credit/Loan is now fully committed and is expected to be fully disbursed by the end of 1976. 1.02 The total Project cost would be US$26.3 million of which the re- quested Loan of US$12 million would cover the foreign exchange cost estimated at approximately US$12 million. The balance of the Project cost would be financed by (i) subborrowers contributions, (ii) BNT's own funds, (iii) Gov- ernment grants in the case of small farmers and date plantations and (iv) Government loans in the case of date palm plantation development. 1.03 The Project was prepared with the assistance of two FAO/IBRD CP missions in September 1974 and September 1975, and was appraised in November 1975 by Messrs. J. Cerych and H1. Lenot (Bank) and G. Clement and A. Sommer (consultants). Mlessrs. Merghoub and Bertelsmeier were responsible for the project through negotiations and final processing. Mrs. Anne McLaughlin checked figures and tables and helped with calculations. II. COUNTRY BACKGROUND AND THE AGRICULTURAL SECTOR A. Country Background 2.01 Tunisia, with an area of some 164,000 km2, is the smallest and least endowed of the North African ltaghreb countries. From a climatic and agricultural point of view the country can be conveniently divided into three parts: Northern Tunisia (20% of land area); Central Tunisia (20%) and Southern Tunisia (60%). The population (5.6 million in 1975) concentrated in Northern Tunisia and along the eastern coastline is growing at about 2.3% annually. Average per capita GNP for 1976 is estimated at about US$754; the  - 2 - absolute poverty level for 1976 is estimated at US$191, the relative poverty level at US$192. 1/ B. The Agricultural Sector General 2.02 Only in Northern Tunisia, where average rainfall exceeds 350 mm, is grown a wide range of rainfed crops, dominated by wheat and barley. Yields are often depressed when fluctuations in the amount and the seasonal distribu- tion of rainfall result in drought or floods. Central Tunisia agriculture, with average rainfall of 200-350 mm, is characterized by rainfed tree crops (predominantly olives and almonds), marginal cereal production, and sheep grazing. Southern Tunisia, where rainfall drops below 200 mm, has extensive grazing with some date and vegetable crops cultivated in the oases. 2.03 Some 53,000 km2 (5.3 million ha), or 33 percent of land area, is classified as cultivable land; the remainder is range land suitable for exten- sive sheep grazing, low-yielding forest and desert. The cropped area is about 3.1 million ha including 1.3 million ha under perrenial crops, but only 110,000 ha, or 3.5 percent are equipped for irrigation. Irrigation is mostly confined to the Lower tMedjerda Valley, the Nebhana area near Sousse and Cap Bon, and the oases of Southern Tunisia. Principal irrigated crops are tomatoes, peppers, potatoes, melons and citrus. Off-season vegetables are grown in small quanti- ties along the coast near Sousse and Zarzis and on the island of Jerba. Areas in vegetables increased considerably during the 1960's. 2.04 The State owns 815,000 ha of land cultivable or suitable for agri- culture while the private sector owns as much as 4.5 million ha. Legislation exists for land redistribution and consolidation in public irrigation areas, transfer of public land to private farmers, and distribution of part of the collectively-owned land to individuals. 2.05 In 1975 agriculture contributed about 16 percent of GDP, employed a little over 50 percent of the labor force, and provided some 20 percent of total exports (principally olive oil, citrus and other fruits, and vegeta- bles); however, agricultural exports excluding olive oil represent only a small part (15%) of total farm output. Agricultural imports (mainly cereals, edible oils, dairy products livestock and sugar) have increased sharply over the past decade and represented about 27 percent of total imports in 1975 (estimate). The net trade balance in agricultural products was persistently negative over the last decade. Price regulations are in effect for a limited number of com- modities (including cereals, milk, edible oils, and sugar) and retail mark-up on all foods is Government-controlled. 1/ Based on estimates for 1975 adjusted assuming the current inflation rate of 8% p.a. and the actual GNiP growth rate of 10% p.a.  - 3 - 2.06 Mlany factors constrain intensive agricultural development. Insuf- ficient rainfall causes irrigation to play an important role in diversifying the basically extensive rainfed agriculture. The potentially irrigable area is limited to 200,000 ha, mostly in tiorthern Tunisia where rainfall is already above average and where incremental benefits from irrigation development are lowest. Other constraints are inadequacies in the agricultural extension services (Annex 1), the credit system (Annex 2) and, in some areas, ready access to marketing centers. 2.07 Grain Farming and Mixed Agriculture: Rainfed grain farming is the principal farming practice used in Northern Tunisia. In areas where annual rainfall is below 500 mm alternate-year cropping is still practiced. Where adequate water supply is available livestock production and dairy farming is sometimes combined with grain farming. The principal grains produced are soft and hard wheat and barley. After a record harvest in 1972 (1.2 million tons) the grain production remained satisfactory in 1973 and 1974, and reached in 1975, another record year, 1.3 million tons. The general increase in produc- tion compared to the years prior to 1972 is due largely to greater use of fertilizers and selected seeds. The area under grain cultivation has also increased by some 15% during the period covered by the present Plan (1973-76), reaching in 1975 about 1.5 million ha. The production of vegetables and tomatoes on irrigated land also increased substantially. 2.08 Tree Crops: Yearly tree crop production fluctuates round 1 million tons, of which olives, even in a mediocre production year, represent about 50%. The average annual production of olives for the period 1973-75 was 480,000 tons with important yearly variations depending largely on the rain- fall (1972: 900,000 tons, 1973: 375,000 tons). Citrus fruit production is stable, averaging some 100,000 tons/year. 2.09 Livestock and Dairy Production: The current estimate of the exist- ing livestock inventory /1 in Tunisia is as follows: 1973 1975 Cows /2 335,500 384,000 of which pure breed 17,500 24,000 Sheep (female) /2 2,555,000 2,790,000 Goats (female) /2 569,000 765,000 /1 Source: ONEP and USAID Survey. /2 No data available for male animals. The milk production per pure breed animals is estimated to be on average about 2,400 1 per year as compared to only 450 1 for local breeds. In 1975 Tunisia's milk production of 260,000 tons of milk had to be supplemented by imports of some 130,000 tons equivalent in powder to meet the increasing demand. The  consumption of milk and milk products is expected to reach 520,000 tons by 1980 and to meet this demand Tunisian milk production would have to grow annually by 12%, which appears highly unlikely. The main reason for the relatively low milk production was, until January 1976, the uneconomically low price of milk (D 0.065/1) set by the Government. Only two categories of farrmers could op- erate profitably or at least without loss under these conditions: (a) very large and modern dairy farms, and (b) farmers who could sell their production directly to consumers at unregulated prices (about D 0.100/1) generally much higher than the Government price. In January 1976 Government increased the milk price to D 0.090/1; this measure improves the necessary impetus for more investment and growth within dairy farming. 2.10 Date Plantations: Tunisia's date production is concentrated in the semi-desertic southern and southwestern part of the country. The bulk of the date production still comes from traditional oases which cover some 10,500 ha. Characteristic of these traditional oases are the high density of trees, out- dated methods of cultivation and mediocre yields. The system of land tenure is marked by an extreme fragmentation of ownership, so that in some cases the fruit of a single palm tree may belong to more than one family. Mlodern plan- tations irrigated by artesian wells cover an additional area of about 1,400 ha, of which 815 ha were planted since 1972 by the Societe Tunisienne de l'Industrie Laitiere (STIL) under the First Agricultural Credit Project. With the exception of one plantation dating back to the late 1920's, all plantations are under 14 years of age. The production of the last 5 years varied accord- ing to climatic conditions and availability of ground water, between 39,000 and 58,000 tons annually of which the praised Deglat variety represented about 20%. Date exports, nearly exclusively the Deglat variety, amounted during the same period in average to about 5,800 tons per year. 2.11 Agro-Industries: Excluding olive oil mills, there are presently in Tunisia about 300 enterprises which can be classified as agro-industries. They employ full time about 10,000 people and account for about 1% of GDP. The characteristic features of the sector are: (a) preponderance of flour milling, canning and edible oil industries in terms of turnover and invest- ment, (b) large number of small enterprises, (c) concentration around Tunis and coastal urban centers, (d) technical inadequacy of the often obsolete machinery and plant, (e) inadequate technical training and, (f) limited, if any, coordination between agricultural and industrial producers. 2.12 Olive Oil Production: Between 1968 and 1974 olive oil exports rep- resented on average about 50% of the value of all agricultural exports and were in value terms the most important export commodity after crude oil and phosphate. There are about 1,100 olive oil mills in Tunisia and their nominal extraction capacity of 180,000 tons/year is theoretically sufficient to cover the country's requirements even in good production years. Many of the mills however, are ancient, have run down equipment with low productivity and pro- duce second-rate oil. M[oreover the geographical distribution of the mills is such that oil extraction capacity is inadequate in the North while there is a surplus capacity in the South. Due to further extensive new plantations in the North, this imbalance will reach much greater proportions if new mills are not constructed in the North.  -5- Development Objectives 2.13 The Government's Fourth Plan (1973-76) aimed at an annual increase in agricultural output of 6.2% over the level reached at the end of the Third Plan (1969-1972) but an annual growth of only 3.8% is being attained. The policy under the Fourth Plan was to emphasize directly productive investments including rehabilitation of existing irrigation infrastructure, encourage a higher level of investments by the private sector, and direct more resources to production of livestock, fish, fruits and vegetables. Access to agricul- tural credit was to be facilitated in order to increase investments and to reach small farmers. Extension services were to be strengthened to reach more medium and small farms. The Plan also contained specific proposals for in- creasing security of land tenure and proposed to enforce existing legislation for sequestration of underutilized land. 2.14 The preparation by Government of Tunisia's 5th Development Plan (1977-81) is well advanced. The principal objectives of the Plan will be: (a) Full employment of the additional rural and urban working population; (b) Self-sufficiency in food production; and (c) Continuous increase in the living standard of the population. 2.15 Self-sufficiency in food production is defined as (i) reaching pro- duction levels of grains and livestock (milk and meat) to meet expected in- ternal consumption, and (ii) reaching in other agricultural products a ba- lanced foreign trade. Under the maximum growth projection, and based on the latest production evaluation, this would mean producing in 1981 1.8 million tons of grains (estimates for 1976, 1.3 million) 170,000 tons of meat (1976, 105,000 tons) and 600,000 tons of milk (1976, 268,000 tons). Agricultural exports should reach D 79.5 million (D 60.3 million, 1976), and agricultural imports should not be higher than D 94 million (1976 D 81 million). Expressed in yearly growth rates these objectives are for grains 6.7%, for meat 11.1%, for milk 17.5%, for agriculture exports 5.7% and for agriculture imports 3% only. The Government tentative estimates of investment in agriculture needed for reaching this growth over the period of the Plan would be D 507.0 million (US$1.2 billion). While the objective to reach self-sufficiency in grain pro- duction and to increase agricultural exports by 5.7% yearly appears feasible in case of rigorous application of the Plan, the prospects of increasing milk and meat production to the projected levels seems highly unlikely. C. Previous Agricultural Projects 2.16 The Bank/IDA financed its first agricultural project in Tunisia, the Cooperative Farm Project, in 1967 (Loan/Credit 484/99-TUN, US$18.0 million reduced to US$9.2 million). Following major changes in Tunisia's agricultural  -6- policy in 1969, the original project was substantially revised and then com- pleted in 1973. 1/ In 1971, the Bank/IDA approved a loan/credit for the Agri- cultural Credit Project (Loan/Credit 779/263 TUN, US$8.0 million) now fully committed and expected to be fully disbursed by the end of 1976 (para 4.01 through 4.14). IDA also finances a Fisheries Project (Credit 263-TUN, US$2.0 million) aimed at development of Tunisia's inshore fisheries. Implementation of the Project is satisfactory, all funds are committed and are expected to be disbursed by the first quarter of 1977. In December 1974, the Bank ap- proved a loan for an Irrigation Rehabilitation Project in thie Medjerda Valley and Nebhana (Loan 1068-TUN, US$12.2 million): the loan became effective in September 1975 and Project execution is in its early stages. III. AGRICULTURAL CREDIT A. General (Annex 2) 3.01 The Tunisian banking system comprises the Central Bank, 12 commer- cial banks (the two most important 2/ being Government-owned), 43 local mutual credit banks and several other financial institutions. 3/ The banks are con- trolled by the Central Bank through a variety of measures: rediscount rate, rediscount ceilings, regulation of interest rates, prior approval of loans, reserve requirements, ratios (liquidity, solvency, medium-term financing, etc.). These control measures reflect concern over important issues such as banking liquidity, incentives to finance investments, financing of seasonal activities and encouragement of savings; the control system however has been too rigid and bureaucratic to be well-adapted to the new economic policies of encouraging private investment and establishing an open market economy. The detrimental effects of the low interest rate structure on the economy in terms of resource mobilization and allocation have been detailed in a recent Bank economic report (No. 962-TUN) and in the IFC-sponsored work on the financial sector. Agricultural Credit 3.02 The Ministry of Agriculture is involved together with the banks, especially Banque Nationale de Tunisie (BNT) and the Caisses Locales de Credit Mutuel (CLCMs), in the planning and in the provision of agricultural credit. 3.03 Short-term credit is granted to larger farmers by BNT, or, to a small extent, by other commercial banks, and to the smaller farmers by the 1/ Project Performance Audit Report circulated to the Board Januarv 8, 1976. 2/ Societe Tunisienne de Banque (STB) and Banqle MNationale de Tunlisie (BNT). 3/ One of whiich, Banque de Developpement Econom ique de Tunis je (RDTET), is 1 DFC bank Borrower.  CLUMs and by several Government agencies under specific programs (usually credit in kind, as part of an input package) for a given crop (grain, olive oil) or for a given region. In 1975, the banking sector provided about D 8 million to 12,000 farmers, and Government agencies about D 4 million to 40,000 farmers. The sectoral distribution of short-term credit does not fol- low the national priorities or the contribution of each crop to total produc- tion but is more influenced by commercial banking criteria: e.g., grain farmers, often being wealthier, receive more than half of the short-term credit extended by the banks, while contributing about 20% of total agricul- tural production. About 90% of all loans were disbursed in Northern Tunisia. 3.04 Investment credit for agriculture is mostly financed by special funds provided by the State budget or by foreign aid; the banks themselves invest very little of their own funds in medium- and long-term loans to agri- culture as loans to industry and commerce are safer and more profitable. Each special fund has its own lending terms, its own appraisal procedures and its own target group. Government-financed grants are often given in addition to loans. In 1975, BNT (through which almost all loans and grants including FOSDA are channelled) granted about 14,000 investment loans to farmers and 1,300 loans to cooperatives for a total amount of D 14 million including grants. More than one-third of the loans were for agricultural machinery and another third for tree plantations. About 50% of the loans were disbursed in Northern Tunisia. 3.05 Credit to agroindustries is provided by a great number of banks. Short-term credit, which includes also rediscountable credit for financing stocks of olive oil, wine and grain held by state agencies, amounts to about 20% of all short-term credit to manufacturing industries, or 7% of all short- term credit. Investment credit is provided almost entirely by three banks, STB, BNT and BDET; they also provide equity capital for some agroindustrial companies. 3.06 FOSDA. The most important source of investment credit for agricul- ture is the Special Fund for Agricultural Development (FOSDA), a Government fund financed by a yearly budget appropriation and channelled exclusively through BNT. The funds are used for direct investment grants and for loans to small- and medium-size farmers. The lending conditions are revised each year, and in the past have provided for low interest rates (3%-4.5%) and maturities related to the life of the investment, ranging from two years (rabbit breeding) to 25 years (pistachio trees). All the repayment risk is carried by the State budget; recoveries in the past have been around 65%. In 1975, FOSDA loans were made to about 13,000 private farmers for D 4.7 million (plus D 1.3 million in grants) and to about 94 cooperatives for D 800,000 (plus D 57,000 in grants). With an average loan amount to private farmers of D 662, the FOSDA clientele tends to be composed mostly of small size farmers. 3.07 Actual access by farmers to both short-term and investment credil is limited: at most, 50,000 Tunisian farmers (15% of all) have access to insti- tutional credit (banks, CLCMs and Government agencies). Small farmers' access to credit is even more limited partly because of loan securitv requirements (lack of title to Land) but also because of the inability or unwillingness of  BNT, the most important provider of credit, to develop a small farmers' lend- ing program: BNT is a commercially oriented bank and has been reluctant to engage in such a program where both costs and risk are high. The Government and BwNT's management have just begun to address the need to formulate credit policies which allow for credit to reach small farmers in an effective manner and still be recollected. In the past the few projects or institutions spe- ciallv geared to small farmers (the CLCMs and some Government agencies for short-term credit, the IJorld Food Program (IJFP) and some bilateral projects for investment credit) were either ineffective, being hampered by institu- tional difficulties (e.g., the CLCMs, para 3.21 ff), or limited in scope, sometimes financing no more than pilot projects. 3.08 The great number of institutions involved in agricultural credit and the diversity of financing terms under various funds allow for a certain flexibility in credit terms which could be adapted to target groups; however this framework has had detrimental effects both in institutional and economic terms, which were compounded by the easy lending terms and conditions (low interest rates, high grant element). In institutional terms, the control of credit is difficult because of the multiplicity of credit-giving institutions outside the banking system, and the precise delimitation of the roles of the various institutions involved is not always clear. In economic terms, the availability to farmers in some cases of several credit sources under differ- ent lending terms has often led to a competitive situation, with a high de- mand only for the funds with low rates, cursory project appraisal, little supervision and low repayment pressure. The situation has been only partly corrected by administrative decisions on eligibility of farmers for specific funds. The subsidized credit policy, particularly interest rate subsidies and investment grants, constitutes a heavy burden on the budget and it dis- courages banks from providing funds to agriculture (especially investment loans). The low interest rate level encourages capital-intensive invest- ments. 3.09 Under a covenant of the Irrigation Rehabilitation Project (1068- TUN), the Government (tMinistry of Agriculture) agreed to review its agricul- tural credit policy and institutional framework and to submit its study and recommendations to the Council of Mlinisters not later than October 31, 1975, so that the decision on implementation of the recommended changes could have been taken by end of November 1975. The study efforts have led to a new set of lending policies and procedures for FOSDA. Bank staff have reviewed the drafts of decrees containing the new measures at various stages since January 1976; Government hopes to pass the decrees by late 1976. The draft recommend- ations represent concrete progress in at least two important respects: (i) an income-based definition of small farmers who would be eligible for concessionary financing under FOSDA aimed at insuring eligibility to smaller farmers, and (ii) an increase in interest rates from 3-4.5% to 6% for all types of FOSDA lending.  The maximum loan amounts per ha or per head of livestock, the proportion of investment grants and farmers' contributions, and the repavment terms have also been revised. The need to reduce the number of credit institutions and programs and to improve coordination between them has been recognized but a concrete plan for achieving this has not yet been proposed. Government has a study under way to make proposals on this problem. B. Banque Nationale de Tunisie (BNT) (Annex 3) 3.10 BNT was created in 1959 to provide agricultural credit and to perform all normal banking and financial operations. BNT is also the second most important commercial bank in Tunisia, providing short-term loans to commerce and industry. Its paid-in capital reached D 4 million in November 1975, of which 25% is held by the Government, 44% by Government-controlled companies, 17% by the Caisses Locales de Credit Mutuel (CLCIMs) and 14% by private shareholders. It is administered by an 11-man board, two of whom are private members. Its current President was appointed in 1974 and previously served as Acting General Director of BDET. 3.11 BNT's internal organization appears, in general, satisfactory, although the decision-making process should be reviewed with emphasis on decentralization. The need for decentralization has been recognized by BNT's management and more delegation of authority to branches is planned. The staff is adequate for BNT's present operations but would need to be expanded if BNT was to take a substantially more active role in credit to small farmers and agro-industries. 3.12 BNT's lending and granting procedures vary depending upon the source of funds. BNT staff appraises itself loans made from BNT's own resources (almost all short-term) and from IBRD/IDA and IJSAID funds. A consequence of the project implementation covenants under the First Agricultural Credit Project was the improvement in procedures used by BNT staff for the appraisal of investment projects financed by the Bank/IDA with, as result, high repay- ment performance on these sub-loans. On loans made from other funds, BNT only performs a credit-worthiness study of the prospective borrower and the lending decision is taken outside BNT by Government. Supervision procedures are also less stringent for these loans than for those under the First Project. BNT's charges above the interest rate are small (at most D 10 per application). 3.13 BNT makes short-term loans out of its own resources to large private farmers (D 4.9 million in 1975) and to cooperatives (D 0.8 million in 1975). The recent creation of the Mutual Guarantee Societies (SCMs) will allow small farmers to borrow directly from BNT, being jointly and severalLy liable as members of a SCti. 3.14 On its own resources, BNT has practically made no medium- and long- term loans to agriculture but only to industry and commerce. BNT's agricul- tural loans have been financed by funds for investment credit from budgetary sources or from foreign borrowing or aid, which are channelled through BNT.  - 10 - Total agricultural medium- and long-term lending amounted in 1975 to D 11 million (80% of which went to private farmers). Investment grants are often provided in conjunction with these loans (D 3.1 million in 1975). BNT's equity portfolio amounts to about D 1.5 million, with investments mostly in other banks and finance companies (43%) and agroindustries (22%). 3.15 The resources used by BNT for lending are broken down into normal and special resources. Normal resources totalling D 109 million at end 1975 consist of capital, free reserves, deposits, and rediscount facilities. Spe- cial resources amounting to D 42 million in commitments at end 1975 include BNT's borrowings (from USAID and IBRD/IDA) and managed funds from budgetary resources (including FOSDA and some eight other funds) or from foreign aid; BNT does not normally bear the repayment risk on managed funds but does on foreign borrowings. 3.16 BNT's profitability has been good and a sizeable profit has been generated every year. The spread between interest received and cost of re- sources has been around 3% of average total assets and operating costs have been kept low (1.5% of average total assets). As a result, dividends paid have been quite large. BNT's total assets are projected to increase at about 12.5% per year for the next five years, and net income after tax at about 10.5% per year. BNT's financial profitability is expected to remain satis- factory. 3.17 BNT's financial structure, although generally satisfactory, seems to point to undercapitalization, despite the 1975 capital increase to D 4 million. The building-up of reserves has not kept pace with the growth of deposits and lending. Other standard financial ratios appear satisfactory. The financial covenants under the First Project on BNT's financial condition are generally complied with, except for the solvency ratio - capital and re- serves as percentage of sight and fixed term deposits - which is still less than the required 10%, but has improved substantially by December 31, 1975. 3.18 BNT's main lending risk is linked to its exposure in the agricul- tural sector (50% of its outstanding loans and participations) and is there- fore compounded by output fluctuations (bad harvests) and by political and organizational difficulties. The overall provisions against losses estab- lished by BNT are large and sufficient to cover its risks. A better manage- ment of the provision for loans under litigation would be advisable, in order to follow more closely the changes in the portfolio of loans under litigation. However, the provision for miscellaneous risks is in any case large enough to provide sufficient protection. 3.19 BNT s recovery procedures on loans made from its own resources ap- pear thorough and methodical, at least during the early stages of collection efforts during which about 75% of the recoveries are made. Many old loans, on which there are no more repayment prospects, are however still carried on the books and not written off against provisions. Repayment rates vary from 92% (BNT short-term to cooperatives, IBRD/IDA First Agricultural Credit Project) to 50% (German Fund for Irrigation) or even 32% (Special Fund for Financial Assistance to Cooperatives), depending upon the efforts BNT makes for recovery.  - 11 - 3.20 BNT has a reasonably good accounting organization as well as a man- agement information system. Its internal and external auditing need to be improved (see para 6.16). C. Caisses Locales de Credit Mutuel (CLCM) (Annex 3) 3.21 The CLCtIs are rural credit cooperative societies, whose purpose is to promote savings and to provide short-term credit, under a limited mutual guarantee system, to rural people, and, among others, small farmers. Managers are BNT staff members seconded to the CLCMs. The first CLCMs were created in 1963 in the Cap Bon as part of Government efforts to extent mutual credit; the first nine received a state grant. In 1965-66, their management was taken over by BNT but they remained separate entities. They have been plagued by numerous problems: poor management, inadequate assistance by the Government, poor repayment (due to the farmers' unwillingness to repay what they considered as Government grants and to the lack of efforts on the part of the CLCMs' staff and Directors to recover loans overdue) and most of them were in a dismal fi- nancial position. BNT attempted to correct the situation through training and financial assistance, but the situation did not improve much and was further aggravated by the cooperatives' situation (although the loans made to coopera- tives were ultimately repaid by the Government) and the poor harvests in 1967- 69. The CLCMs now play a relatively minor role, collecting small deposits mostly for BNT's benefit (almost half of the CLCMs' assets are in deposits with BNT or in BNT shares, yielding a low return of about 3.2%) and providing short-term loans to agriculture and commerce. 3.22 As a whole, the CLCMs have lost money every year since 1968, al- though in 1974 about 22 CLCMs made a profit as compared to only 12 in 1970. The CLCM deficit is due essentially to their low return on loans (2.7% to 3.2% on loans to agriculture) and on other assets, especially assets with BNT. Recoveries are low, and arrears account for close to 50% of the loan portfolio; contrary to what could be expected in a mutual credit system, neither the other borrowers nor the CLCM Directors exert pressure on the delinquent borrowers. The future of the CLCMs is uncertain. Their close contact with small farmers is very valuable; however, for them to play a use- ful role, drastic financial and administrative measures need to be taken. The creation of the Societes de Caution Mlutuelle (para 3.13) may imply that the CLCMs are expected to dissolve or be completely absorbed by BNT, although Government intentions are far from clear. The on-going credit study and re- view should help to clarify the issues and should ultimately result in their solution.  - 12 - I\V. PERFORMANCE UNDER THE FIRST AGRICULTURAL CREDIT PROJECT Loan/Credit 779/263-TUN US$5.0 m/3.0 m Project Description 4.01 The Loan/Credit was made to finance part of Banque Nationale de Tunisie (BNT) lending to commercial farmers for investment in grain farm mechanization and dairy farm development in Northern Tunisia and for date palm plantations to be undertaken by the Societe Tunisienne de l'Industrie Laitiere (STIL) in the desert areas of South Western Tunisia. The Project included financing of technical assistance to strengthen BNT's loan appraisal division, olantation management of STIL and vehicles for technical services. 4.02 BNT was to lend to subborrowers not more than 70% of the subprojects' cost and IDA/Bank was to reimburse BNT 75% of these subloans thus contributing 52.5% of the subproject cost. The subloans were to be made at an interest rate of 8% p.a. for a period varying from 6 to 15 years according to invest- ment category, with grace periods between 1 and 8 years. However, in line with Government's stated policy to restore confidence in the private sector and to encourage commercial farmers' investment, IDA/Bank agreed during nego- tiations to Government's request to allow a temporary Government interest sub- sidy of 2% so that the resulting interest rate to subborrowers would be only 6/%. The on-lending rate of 6% remained in effect throughout Project implemen- tation. 4.03 The Project cost was estimated at US$15.0 million and the Loan/ Credit of US$8.0 million was to cover the estimated foreign exchange cost of the Project. Project Implementation 4.04 The Project became effective on January 25, 1972 and by April 1972 some 85 applications for a total of D 275,000 under categories Mechanization and Dairy Farms were approved, and drilling of the first 4 wells and also planting of 80 ha of date palms was completed. Project implementation re- mained satisfactory through 1972 with the exception that Government did not strictly enforce the stipulation of the Loan Agreement requiring BNT to charge at least the same interest rate on all loans similar to those under the Proj- ect and instead disbursed to farmers eligible under the Project loans from FOSDA funds on more advantageous terms. During 1973 and 1974 loan disburse- ments for farm mechanization and dairy farm components dropped far behind schedule due to BNT's tendency to use the lower cost and risk-free FOSDA funds to finance such investments and to the decreased viability of milk production in the face of low milk prices fixed by Government and the high cost of im- ported colJs. 4.05 The issue of competing FOSDA funds was settled by end .January 1975 when Government officially instructed BNT that no FOSDA funds be used for  - 13 - farmers qualifving for a Bank-financed loan. Government took this step only after the Bank informed it that unless the issue was satisfactorily resolved by the end of January, the Bank would cancel the part of the Loan allocated for Dairy Farm Development. 4.06 During 1975, and 1976 disbursements under the Farm Mechanization component have been satisfactory. Lending for Dairy Farms improved only marginally because of lack of interest in new dairy investments due mainly to low milk prices. To help alleviate a shortage of milk and milk production in Central Tunisia, the Bank agreed in July 1974 to BNT's request to include in the Project the financing of a Milk Processing Plant based principally on imported milk in powder form to be built by STIL in the Sousse Region. The Date Palm Plantation subproject, after slow implementation during the second half of 1972 and 1973 made good progress in 1974 and 1975, and is now nearly completed. 4.07 The following table shows a summary evolution of BNT's disburse- ments under the three categories during the Project period and the overall disbursement improvement in 1975 and 1976. 1972 1973 1974 1975 1976 /2 Total Farm Mlechanization No. of Approved Loans 206 101 123 210 53 693 NTo. of Dis- bursed Loans 167 113 59 212 88 639 Disbursements D 629,000 D 433,000 D 239,000 D 793,000 D 659,000 D 2,753,000 Dairy Farms No. of Approved Loans 65 24 13 31 5 138 No. of Dis- bursed Loans 38 21 6 27 7 99 Disbursements D 71,000 D 48,000 D 16,000 D 244,000/1 D 30,000 D 409,000/1 Date Palm Plantation Disbursements D 97,000 D 199,000 D 405,000 D 298,000 D108,000 D 1,107,000 /1 Including D 159,000 for the Milk Processing Plant. /2 As per June 30, 1976. 4.08 In compliance with the Loan Agreement, BNT set up a Project Unit composed of one expatriate agricultural credit expert, 4 Tunisian agricul- tural engineers and supporting staff. The performance of the Unit, whose job consisted of loan application evaluations, supervision of subprojects and ex- tension services to subborrowers, was satisfactory throughout Project imple- mentation. Bank disbursements under the category Technical Assistance were limited to the foreign exchange part of the expatriate expert's remuneration, all other expenses relating to the Project Unit were born directly by BNT.  - 14 - 4.09 As of October 1, 1976, total Loan/Credit disbursements reached US$7.4 million and the remaining US$0.6 million were fully committed. The Loan/Credit is expected to be fully disbursed by the end of 1976. Project Implementation 4.10 The effective investments financed under the Project as compared to the appraisal estimates are shown below: Disbursed Subloans Appraisal Estimate June 30, 1976 Average Total Average Total No. of Loan Loan No. of Loan Loan Loans Amount Amount Loans Amount/I Amount No. D D'000 No. D D'000 Farm Mechanization Harvesters 240 5,308 1,274 115 5,570 ) Crawler Tractors 135 5,652 763 45 5,140 ) 2,475 Wheel Tractors 465 .2,062 959 299 2,975 ) Implements 600 642 385 973 564 ) No. of Subloans and Total Loan Amount 840/2 - 3,381 639/2 2,475 Dairy Farms Water Supply Development 150 2,146 322 14 665 9 Farm Implements 150 56 84 ) 31 1,100 34 Agr. Implements - - 28 ) Dairy Heifers 1,800 183 329 908 228 207 Milk Processing - - - 1 - 159 No. of Subloans and Total Loan Amount 150/2 763 99/2 - 409 Date Palms 1 1,197,000 1,197 1 1,107,000 1,107/3 Development of 900 ha, 18 wells __ __ ___.___ 1 1,197,000 1,197 1 1,107,000 1,107 BNT Total Lending 991 5,341 739 3,991 /1 Average loan amounts are based on projects approved as opposed to disbursed. /2 Loans were expected to and did include more than one of the enumerated cate- gories of investment. /3 813 ha were planted as of Dec. 31, 1975 - when the Program is completed the area will cover 886 ha; 13 artesian wells have been drilled.  - 15 - 4.11 In physical terms the Farm Mechanization component remained behind the estimates due to slow Project implementation in 1973 and 1974. The number of disbursed loans as of June 30, 1976 for harvesters represented 48%, for crawler tractors 33% and for wheel tractors 64% of estimates. The total number of subloans was 75%, and the aggregate amount of subloans disbursed was 74% of appraisal estimates. 4.12 The Dairy Farm component is to be considered the least successful of the Project. The reasons for this were (a) competing financing by FOSDA funds, (b) artificially low price of milk parallel to steep increase in prices of imported heifers, and (c) structure of investment other than pro- jected (little investment in wells and irrigation - Table in para 4.10). 4.13 The Date Palm Plantation component was by far the most successful in its implementation. Out of the estimated 900 ha, 813 were already planted as of September 30, 1975, and additional 73 ha are being planted during 1976. In all 13 deep wells were drilled with a total artesian yield of 633 1/s. The total cost should reach about D 1.85 million when the subproject is fully completed. 4.14 BNT's financial and institutional performance under the Project was generally satisfactory although a few features still need to be reviewed and improved (auditing procedures, financial covenants (para 3.17 and Annex 3)). Project Impact 4.15 BNT's records of operations do not enable evaluation of physical and financial investment results at the subproject level, nor do statistics permit ex post analysis of the subproject results. However, if a judgment is based on the findings of the Bank's supervision mission and on reports of BNT's Project Unit staff the large majority of subprojects under the Grain Farm Mechanization component are successful both from the technical and the financial point of view. It appears that the expected economic rate of re- turn on investment (22%) has generally been attained. Dairy farming was suc- cessful mainly in cases where the farmers were able to sell their produce on the free market, however, the recent (January 76) price increase of milk should improve financial viability of those subprojects where farmers sell their products at Government prices. The rapid increase in date prices (1970/71: D 130/ton, 1975/76: D 280/ton-Deglat variety) will more than compensate the increases in operating cost. Final judgment on the financial and economical success of date palm plantation component will however only be possible in about 12 to 15 years after the plantations have reached full production.  - 16 - V. THE PROJECT A. Objectives 5.01 While the First Project aimed principally at increasing grain and milk production through mechanization and modernization of large farms in Northern Tunisia, and at improving BNT's capability as an agricultural credit institution, the Second Project would not only aim at increasing agricultural production to meet Tunisia's rising demand and to diminish the huge agricul- tural production deficit, but would also contribute to extension and improve- ment of agricultural investment credit to small farmers who so far have had little or no access to institutional credit. The Project would contribute to further development of the desertic Southern Tunisia by helping to finance date palm plantations to be established by small farmers associations; it would also improve Tunisia's export position in olive oil by establishing small but modern and economical oil. mills allowing for production of premium quality olive oil. Continuity in improvement of efficiency of BNT as an agricultural credit institution would also be an aim of the Project. B. Description 5.02 The Project would finance part of BNT's medium- and long-term lend- irig for agriculture and would consist of four major components: (a) Medium- and long-term lending to small farmers for on- farm investments; (b) Medium- and long-term lending to commercial farmers for on-farm investment for grain and dairy production, for overall farm development, and for vegetable and fruit production; (c) Medium- and long-term lending for investment in collec- tively owned smallholder date-palm plantations; (d) Medium- and long-term lending to agro-industries, prin- cipally olive oil mills. Small Farmers 5.03 This category would include grain, dairy and mixed farming, small orchards and vegetable growing. Investment would be in agricultural equip- ment such as ploughs, seeds and few cultivators, in dairy cross-breed heifers, in draft animals and in well and irrigation improvements. Investment in tractors and heavy agricultural machinery would not be financed under this category except in case of investment loans to semi-public organizations  - 17 - such as Societe liationale de Motoculture (SONAIO) or cooperatives providing tractors and harvesting services to small farmers. The large majority of borrowers is expected to come from the Northern part of Tunisia, the north- eastern coastal areas and from around the larger consumption centers, but qualified applications from the rest of the country will also be considered. Criteria to be applied for qualification as a small farmer are described below in para 6.04. SONAIO or cooperatives providing tractor and harvesting services to small farmers would qualify under this component after Bank's prior approval of specific investment and operating plans. About 660, or 70%, of the total number of Project sub-loans and 32% of the total amount of Project sub-loans are expected to be disbursed under this category. Commercial Farmers 5.04 The category would include grain farms, dairy farms, fruit and vege- table producers and mixed farming. Investment is expected to be mainly in tractors, harvesters and heavier agricultural machinery and equipment, in pure breed dairy heifers, stables, and milking facilities, and in construction and improvements of shallow wells and irrigation infrastructure. As in the case of small farmers, most of expected subborrowers will come from the North and the North-east, where rainfed grain production is concentrated and where the aquifer often permits construction of shallow wells and irrigation, although qualified applications from the rest of the country will also be considered. About 275, or 30% of the total number of Project sub-loans, representing 27% of the total amount of loans are expected to be disbursed under this category. Date Palm Plantations 5.05 The beneficiaries of loans under this category would be the Societes Civiles de Mise en Valeur (SCMV), production cooperative-type farmer develop- ment associations (Annex 1, Appendix 1) investing collectively in date palm plantations in the Djerid and Nefzaoua regions. Under the First Agricultural Credit Project, STIL has completed planting of over 800 ha of date plantations in the same two regions. The technical success of STIL's venture resulted in the creation by local smallholders of over a dozen SClIVs eager to follow STIL's example. Three of them have already received FOSDA loans for planta- tion development and are proceeding with planting. The membership of the SCMVs varies between 10 and over 50 persons and land holdings depending ultimately on the yield of the wells are generally around 2 ha per member (as much as 4 ha in the earliest SCMVs and about 1 ha in the later ones). 5.06 Investment would be for ground-water development (deep artesian wells) with irrigation and drainage network, land levelling and land prepara- tion, planting and maintenance (first 7 years) of date palms and windbreaks, pumping stations, tractors, miscellaneous agricultural implements and storage facilities. As under the First Credit Project, Government would bear the cost of dry bore holes. Plantations would be in the Djerid and Nefzaoua areas in South Western Tunisia, which appear to be most appropriate climatically and have proven substantial artesian aquifers. The Project would finance four  - 18 - plantations of between 50-75 ha depending on the yield of each well. About 21% of the total amount of subloans are expected to be disbursed under this category. Olive Oil Mills and Other Agro-Industries 5.07 Important new plantations of olive groves in North and Central Tunisia in recent years will increase Tunisia's olive production from the present average of some 550,000 tons to about 900,000 tons by 1981. The increased production will result in an overall severe shortage of oil extrac- tion capacity in the two regions, and some one hundred new mills (capacity of 30 tons/day each) will be needed to meet the requirements. The required and expected increase of olive oil production will result in increased availabil- ity of olive cake and the need for several mills specialized in olive cake oil extraction. The growing fruit and vegetable production based mainly on new irrigation will ca]l for extension of existing processing facilities and/or construction of new ones. 5.08 BNT has in the past few years financed from its own funds on average six to eight olive-oil mills each year and the Project would help finance 18 such mills over a period of two years. Other agro-industries such as an olive cake oil extraction mill and/or fruit and vegetable processing facilities could also be financed under this category after Bank's prior approval. About 20% of the total loan amount are expected to be disbursed under this category. C. Cost Estimates 5.09 The Project cost estimates are based on recent BNT experience, on current prices (Octover/November 1975) of the respective investments and the representative models of production units examined during appraisal (Annex 5). Total Project cost is estimated at D 11.0 million (US$26.3 million) of which approximately 45%, or US$12 million equivalent, would be in foreign exchange. These estimates are summarized in the following table:  -19_- Project Cost Foreign Local Foreign Total Local Foreign Total Exchange D -------000