Tunisia - Fifth Highway (Rural Roads) Project

Document of -4 . J The World Bank FOR OFFICIAL USE ONLY Report No. 3722-TUN STAFF APPRAISAL REPORT REPUBLIC OF TUNISIA FIFTH HIGHWAY (RURAL ROADS) PROJECT February 18, 1982 Projects Department Europe, Middle East and North Africa Regional Office This document has a restricted distribution and may b.e used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Currency Unit = Tunisian Dinar (DT) DT 1 = US$1.95 I' DT 1,000 - - US$1,950 DT 1,000,000 = US$1,950,000 US$1 = DT 0.51 WEIGHTS AND MEASURES International System (S.I.) British/U.S. System 1 meter (m) - 3.28 feet (ft) 1 kilometer (km) = 0.62 miles (mi) 1 millimeter (mm) = 0.039 inches (ins) 1 hectare (ha) = 2.47 acres (ac) 1 metric tonne (t) = 2,205 pounds (lb) ACRONYMS AND ABBREVIATIONS BNT - Banque Nationale de Tunisie (National Bank of Tunisia) CCDA - Commissariat Central du Developpement Agricole (Central Agricultural Development Commission) CNEA - Centre National d'Etudes Agricoles (National Agricultural Studies Center) CRDA - Commissariat Regional de Developpement Agricole (Regional Agricultural Development Commission) CTV - Cellule Territoriale de Vulgarisation Agricole (Extension Centers) DERV - Direction de l'Enseignment, Recherche, et Vulgarisation (Department of Training, Research and Extension) DPA - Direction de la Production Animale (Livestock Production Department) DPC - Direction des Ponts et Chaussees (Highway Department) DPSAE - Direction de la Planification et des Statistiques Agricoles et Economiques (Planning and Statistical Department) DPV - Direction de la Production Vegetale (Crop Production Department) FOSDA Fond Special pour le Developpement Agricole ,.(Special Fund for Agricultural Development) FOSEP - Fond Special pour l'Encouragement de la Peche (Special Fund for the Promotion of Fishery Investments). GR - Genie Rural (Rural Engineering Department) MA - Ministere de l'Agriculture (Ministry of Agriculture) MPF - Ministere du Plan et des Finances (Ministry of Planning and Finance) MPW - Ministere de l'Equipement (Ministry of Public Works) MTC - Ministere des Transports et des Communications (Ministry of Transport and Communications) OC - Office des Cereales (Cereal Board) PDR - Programme de Developpement Rural (Rural Development Program) SCM - Societe de Caution Mutuelle (Mutual Guarantee Association) REPUBLIC OF TUNISIA FISCAL YEAR January 1 - December 31 1/ The Tunisian dinar is floating. As of September 30, 1981, the rate was DT l=US$1.95. This has been taken as a reasonable average rate for the conversion of project costs. -  FOR OFFICIAL USE ONLY TUNISIA STAFF APPRAISAL REPCORT OF FIFTH HIGHWAY (RURAL ROADIS) PROJECT Table of Contents Page No. I. TRANSPORT IN THE ECONOMY A. Geographic and Economic Setting . . . . . . . . . . . 1 B. The Transport System ...... ... I. .. .. .. . 2 C. Transport and Rural Development . . . . . . . . . . 3 D. Previous Transport and Agricultural Projects . . . . . 4 II. ROADS AND RURAL DEVELOPMENT A. Road Network . . . . . . . . . . . . .. . . . . . . . . 7 B. Road Administration . . . . . . . . , . . . . . . . . 9 C. Organization of Agricultural Services . . . . . . . . 10 D. Rural Road Program . . . . . . . . . . . . . . . . . . 13 III. THE PROJECT A. Background . . . . . . . . . . . . . . . . . . . . . . 17 B. Objectives . . . . . . . . . . . . . . . . . . . . . . 17 C. Description . . . . . . . . . . . . . . . . . . . . . 18 D. Cost Estimates . . . . . . . . . . . . . . . . . . . . 22 E. Financing . . . . . . . . . . . . . . . . . . . . . . 26 F. Implementation ................... . 27 G. Procurement ..................... 27 H. Internal Accounting and Auditing Proceadures . . . . . 29 I. Implementation Schedule and Monitorin,. . . . . . . . . 29 J. Disbursements ................... . 31 K. Environmental Impact . . . . . . . . . . . . . . . . . 34 IV. ECONOMIC EVALUATION A. General Framework . . . . . . . . . . . . . . . . . . 36 B. Sub-Project Review ............ .... . . 36 C. Benefits and Beneficiaries . . . . . . . . . . . . . . 37 D. Economic Analysis ................. . 39 E. Results ..................... 40 F. Project Risk . . . . . . . . . . . . . . . . . . . . . 41 V. AGREEMENTS REACHED AND RECOMMENDATIONS . . . . . . . . 42 This report is based on the findings of an appraisal mission in June 1981 composed of Roy Knighton (Economist), Peter Long (Engineer), Cynthia Cook (Sociologist), and Jean de Gryse (Financial Analyst, Consultant) and a post-appraisal mission in October 1981 composed of Peter Long, John Doolette (Agriculturalist), and Suzanne Morris (Systems Analyst). This document has a restricted distribution and rnay be used by recipients only in the performance of] their official duties. Its contents may not otherwise be disclcsed without World Bank authorization.  Page No. ANNEXES 1. Previous Bank Projects . . . . . . . . . . . . . . . 43 2. Agricultural Extension Programs . . . . . . . . . . 47 3. Agricultural Credit . . . . . . . . . . . . . . . . . 59 4. Road Design Standards . . . . . . . . . . . . . . . 67 5. Details of Economic Evaluation Methodology: Sub-Project Review . . . . . . . . . . 68 6. Guidelines for Ongoing Preparation, Implementation, and Monitoring . . . . . . . . . . 79 7. Framework for Rural Road Impact Study . . . . . . . 83 8. Indicative Action Plan . . . . . . . . . . . . . . . 89 9. Project File Documents . . . . . . . . . . . . . . . 94 CHARTS 1. Organization of the Ministry of Public Works . . . 97 2. Organization of the Ministry of Agriculture . . . 98 3. Project Implementation Organization . . . . . . . . 99 4. Project Implementation Schedule: General ... . . . . . . . . . . . . . . . . . 100 5. Project Implementation Schedule: Technical Assistance . . . . . . . . . . . . . 101 MAPS Tunisia : Fifth Highway Project (Rural Roads) IBRD 15916 North IBRD 16145 Center IBRD 16144 South  I. TRANSPORT IN THE E(CONOMY A. Geographic and Economic Setting 1.01 Tunisia's overall development strategy is reflected in the objec- tives of the Fifth Development Plan (1977-81) which are concerned with strengthening and diversifying the economy, particularly through employment creation and emphasis on rural and regional development. However, since 1977, the growth of domestic production has slowed and employment creation has failed to meet planned targets. Although a rise in crude oil prices has helped keep the financial situation manageable, recent policy orienta- tion has shifted away from industry towards much greater emphasis on agri- culture and the rural sector as well as on achieving a more balanced system of regional development. The Government is currently preparing the Sixth Development Plan (1982-86) which is expected to include a new orientation in its policy for rural and regional development by strengthening and co- ordinating support facilities such as transport and agricultural services and by providing better incentives for increased economic activity. 1.02 The population of Tunisia is about 6.6 million (1980) and is grow- ing at the rate of about 2.5% per annum; about half of the population lives in rural areas. There is significant migraticin into the cities and into the densely populated coastal strip extending from Bizerte in the north to Sfax in the south, an area which accounts for over half the total popula- tion. In terms of natural regions, Tunisia cE.n be divided into three parts (Maps IBRD 15916, 16145, 16144). Northern Tunisia, an area of mountains and broad intermontane valleys with average rcinfall ranging from 400 mm to over 1,000 mm and a population density of about 100 inhabitants per kmi2, has favorable conditions for winter rainfed cereal crops. Central Tunisia with average rainfall between 200 mm and 400 nm and a population density of about 50 inhabitants per kn2, provides a poorer agricultural base com- prising cereal production for home consumption at lower yield levels than in the north, tree crops, mainly olives and almonds, and extensive live- stock production. In the sparsely populated areas of southern Tunisia, agriculture is only possible with irrigation and is limited primarily to oases producing dates, vegetables and some fodlder. On the whole, rainfall varies greatly from year to year and yields o" rainfed crops are particu- larly sensitive to these fluctuations. However, in these marginal condi- tions the application of appropriate agricultural techniques can both increase and stabilize yields. The rural areas are best served by the road network in the relatively densely populated valleys of the northern region. The density of the road network decreases progressively through the central and southern part of the country. 1.03 Gross Domestic Product (GDP) has been increasing at the rate of 6.5% per annum in constant prices since 1976, resulting in a 1980 per capita GDP of about US$1,330. This growth ra:e was somewhat less than the 8.6% average annual increase achieved over the 1969-76 period. This dif- ference is explained by a shift in the underlying growth factors during  - 2- the second half of the 1970s. In particular, agricultural output has fluc- tuated since 1976 partly as a result of adverse weather conditions, and partly because of weaknesses in agricultural support services, particularly extension activities and credit. However, favorable weather in most areas in 1980 has helped to re-establish production levels for most crops. Although agriculture currently accounts for a declining share of GDP--15% in 1980 compared with 24% in 1976--about half of Tunisia's population continues to live in rural areas. B. The Transport System 1.04 The transport system comprises some 17,000 km of classified roads of which half are paved, and 14,000 km of unclassified tracks, serving about 100,000 vehicles; 1,600 km of standard and narrow gauge railways; five major commercial ports and one national shipping company; four air- ports handling international traffic, and two airlines, including Tunis Air, the national carrier. Road transport is the dominant mode for both domestic freight and passenger transport and its share of total demand has been increasing steadily over the years. Roads now carry about 90% of passenger traffic and 85% of freight traffic, excluding minerals. However, the role of the railway in the Tunis-Sousse coastal corridor is greater than these figures would suggest and rail passenger traffic in the corridor has been increasing steadily over the past decade. 1.05 The Ministry of Transport and Communications (MTC) is responsible for overall organization of the transport sector, including the planning of rail, sea, and air transport and the coordination and regulation of land transport and establishment of transport rates. Subsectoral planning is tr,e responsibility of the various modal agencies including the Highway Department (Direction des Ponts et Chaussees, DPC) of the Ministry of Public Works (MPW). Inadequate subsectoral planning capabilities and insufficient overall coordination have resulted in some distortions in the implementation of transport investments in recent years. High priority has been placed on several major construction projects, as well as on the development of air transport and maritime shipping with corresponding high foreign exchange expenditures and limited short-term benefits. The overall result has been a delay in urgent programs for rehabilitation and main- tenance of road and railway infrastructure with implementation of only part of the initial transport sector plan at a higher cost than envisaged for the entire program. Although most of the line agencies are aware of these problems, the Ministry of Planning and Finance (MPF) has been unable to allocate the investment budget on a rational basis as there is no prior arbitration of investment proposals by MTC. For this reason, the Govern- ment has recently established a planning unit in MTC with technical assistance financed under the Bank's Fourth Highway Project. This unit is providing a closer investment analysis and continuous monitoring of developments in the transport sector.  - 3 - 1.06 During the 1960s, traffic levels were moderate and transport investments focussed on minor improvements to existing infrastructure. However, increasing diversification of the economy during the 1970s, particularly the emphasis on agricultural production and regional develop- ment, has since stimulated traffic demand. The balanced development of the transport network, particularly through improvements to regional and local roads, has now evolved as an important consideration in support of the Government's development policy. As such, trarsport continues to play a key role in economic development, and investments in the sector are now shifting to provide the necessary impetus to rural development and to market-oriented activity in general. C. Transport and Rural Development 1.07 As a key sector in the Tunisian economly, agriculture receives high priority in the country's development strategy,, Investments in agriculture increased 40% in real terms during the 1970s compared to the 1960s and the sector currently accounts for about 14% of total private and public invest- ment expenditure. Although output has stagnated somewhat recently and there have been considerable fluctuations from year to year, agricultural value added grew on average at 3.6% per year in real terms during the 1970s. This growth has been due in part to the development of farmer credit facilities and to favorable agriculturaL prices and in part to an increasing use of improved technology. While ex-farm gate prices for many staple crops are controlled by the Government, there is also a private market for most commodities where the price is subject to market condi- tions. In addition, for many crops such as fruit and vegetables, there is only a floor price outside of which there are no additional price restric- tions. The total land area is 16.4 million ha, of which about 5.2 million ha are arable, 34% in cereals, 33% in tree cro's, 21% fallow, the remaining 12% under vegetables, industrial and forage crops. With some 350,000 farms, there is an average of about 15 has of arable land per farm. 1.08 Efficient road transport is critical to rural and agricultural development as it provides a key input for expansion of agricultural production, communication patterns and improviag rural mobility. Public road transport in Tunisia is provided by two national companies, one providing passenger transport, the other freight transport, as well as by 12 regional transport companies providing both passenger and freight ser- vices. In addition, many firms and industrial enterprises also operate own-account fleets which are integrated with specific commercial activi- ties. Small trucks--up to 3.5 tons capacity for general cargo and up to 5 tons for agricultural produce--fall outside restrictive regulatory policies and, as a result, there is a large number of private owner- operators. For several years, the Bank has been working with MTC to simplify regulatory policies, but progress has been slow. The small private truckers provide most of the transport services in rural areas and, where competition is keen, benefits accruing to them are often passed on in part to users.  - 4 - 1.09 Road transport services cannot be provided efficiently unless the quality of road access is improved. Where access conditions are poor, extension, credit and marketing activities are unable to operate effec- tively and the timely delivery of critical inputs to the farmer is often hindered. Social welfare is affected by inadequate access to schools, health, and administrative facilities. In effect, the condition of the road network provides a major constraint to rural and agricultural development in Tunisia. Although adequate in coverage, much of the primary network requires rehabilitation and modernization. Recurrent maintenance funds have been used mostly to maintain the paved primary road network. Most of the unpaved network was never properly engineered or constructed, and routine grading cannot maintain a satisfactory level of service on such roads. In addition, the DPC is not administratively responsible for maintaining unclassified roads, even though some may be relatively important for serving agricultural areas. With the limited resources at its disposal, the DPC has therefore developed the practice of grading important unpaved roads, both classified and unclassified, after the end of the wet season just before the heavy traffic flows of the harvest season. Although this practice ensures evacuation of the harvest, the unpaved network is steadily deteriorating because of the lack of periodic mainte- nance and improvements. Many tracks are muddy and impassable during winter. Average transport costs in rural areas are therefore high, and the level of service so low that it has a significant effect on the provision of services such as agricultural extension, access to schools, and so on. The Government is sensitive to the effect of this situation on the quality of rural life, and the consequent exodus to the cities particularly along the coast. It is now giving high priority to the improvement and main- tenance of the rural road network. Bank assistance in an initial program covering t..e -LTpruv=L-eLL.. and mainteiiatgce of 1,100 km of rurai roads is being provided through a Rural Roads Project (Loan 1601-TUN, 1978) which is being implemented through 1983. The proposed project would constitute a follow-up to the First Rural Roads Project extending the rural road program to all agricultural areas of the country. D. Previous Transport and Agricultural Projects 1.10 To assist the Government in achieving its objectives in the transport sector, the Bank has participated in nine transport projects since 1964. The First and Second Highway Projects (Loan.746-TUN, 1971, US$24 million and Loan 1188-TUN, 1976, US$28 million), and the First Railway Project (Loan 606/Credit 150-TUN, 1964, US$17 million) focussed on modernization and rehabilitation of primary and secondary roads, and renewal of track and purchase of railway equipment. The First Highway Project was a broad success. However, the rehabilitation component was cut back substantially because of high inflation and dollar devaluation over the implementation period. Difficulties were encountered with right-of-way acquisition and with road transport regulation and road user taxation improvements (PPAR No. 2772, December 26, 1979). Implementation of the Second Highway Project was delayed because of problems of property acquisi- tion and particularly because of a persistent lack of local financing. The  - 5 - project is now making satisfactory progress following an agreement with the Bank on a revised financing and execution plan. Both highway projects included technical assistance which encouraged the development of several competent local consulting firms which now work closely with DPC. The First Railway Project was completed successfully (PPAR No. 2109, June 23, 1978). Two small loans in the port sub-sector (Loan 380-TUN, 1964, US$7 million, and Loan 573-TUN, 1969, US$8.5 million) provided support facilities for port operation, mainly dredging, maintenance operations and cargo handling equipment. Both these projects were implemented successfully and had an important institution building impact (PPAR No. 1049, February 26, 1979). Another snLall loan provided financing for a gas pipeline project (Loan 724-TUN, 1971, US$7.5 million) which was satisfactorily completed within one year. A Thiird Port Project (Loan 1797-TUN, 1980, US$42.5 million) is assisting the Government in modernizing berth facilities at the main ports of La Goulette and Sfax and is providing support facilities for improved port opetations. The Third Highway (Rural Roads) Project (Loan 1601-TUN, 1978, US$32 million) initiated the Government's program for improving rural roads by providing financial assistance for an initial slice of about 1,100 km of roads together with some coordinated agricultural inirestments in a cross-section of the country's agricultural areas. Progress with the road component has been good throughout. The agricultural components suffered an initial start-up delay, but the Government is now acce.Lerating implementation and completion is expected on schedule at the end of 1983. At the same time, continuity in the Government's highway program and balanced development of the highway network are being provided through the Fourth Highway Project (Loan 1841-TUN, 1980, US$36.5 million) which emnphasizes rehabilitation of primary and secondary roads and improved maintenance over the whole network, as well as through the Kuwait Fund, wliich has assisted with the construction of several high-standard high-volume roads. Progress with the Fourth Project is very good. In addition, the close collaboration between the Bank and the Government in preparing the project has enabled DPC to establish its own unit for routine preparation of rehabilitation programs. Details of Bank transport projects are in AnneK 1. 1.11 The Bank has supported Tunisia's agricultural sector through credit, irrigation and fisheries projects. Recently greater emphasis has been placed on rural development with the recently approved Northwest Rural Development Project. Bank group lending for agriculture in Tunisia began in 1967 and to date eleven projects have been financed or approved total- ling US$243.7 million of Bank/IDA funds. The eleven projects include three irrigation projects (FY75, FY79 and FY80), two fisheries projects (FY72 and FY79), a cooperative farm project (FY68), three agricultural credit pro- jects (FY71, FY77 and FY80), the Northwest Rural Development Project (FY81), and the Grain Distribution and Storage Project (FY82). 1.12 Performance of Bank-financed agricultural projects has been mixed, reflecting institutional constraints in the sector. The physical compo- nents in the First Irrigation Project concerned the rehabilitation of two separate irrigation areas and implementation w'as satisfactory. The Second  -6- Irrigation Project (Sidi Salem Multi-purpose Project), which involves new installations to expand the areas of irrigation, is moving satisfactorily although there has been a delay in the construction of a major canal. The third (Southern Irrigation Project) has started up satisfactorily. The Tunisia Cooperative Farm Project signed in 1967 provided financing for the development of agricultural production cooperatives as part of a ten-year national plan to implement agrarian reform. A major land reform policy change in 1969, when the attempt to convert all agricultural land into cooperatives was abandoned, caused a revision of the Project. The Bank's Project Performance Audit Report No. 968 confirms that a number of objec- tives were attained, but those concerned with the elimination of subsis- tence farming and the incorporation of private land owners into production cooperatives had to be abandoned. The First Fisheries Project was com- pleted after considerable delays and with a history of low loan recovery by the National Bank of Tunisia (Banque Nationale de Tunisie, BNT). The pro- blems were addressed under the Second Fisheries Project after agreement was reached on a strategy for improving recoveries. The First Agricultural Credit Project provided US$8 million to finance lending by BNT for farm investments in machinery, dairy development, date palm plantations and an agro-industrial component for dairy products. The loan/credit was fully disbursed 34 months after the original closing date. The major issue was that Bank Group funds lent by BNT had a higher interest rate than Govern- ment funds lent by BNT, thus affecting farmers preference and making dis- bursement slower. Uniformity of lending terms, regardless of source of funds, was pursued under the Second Agricultural Credit Project (US$12 million). The Third Agricultural Credit Project (US$30 million) which became effective in June 1981 will continue to emphasize the decentralization and acceleration of loan processing procedures, particu- larly through strengthening BNT's regional staff. Credit for farmers in the zones of influence of roads being improved in the Third Highway Project (Rural Roads Project) is being provided by the Second and Third Agricul- tural Credit Projects. Details of Bank agricultural projects are in Annex 1. -  - 7 - II. ROADS AND RURAL DEVELOPMENT'! A. Road Network 2.01 Tunisia's road network is shown in the table: Km Surface Classified Unclassified Total Paved roads 9,100 - 9,100 Unpaved roads 8,600 9,000 17,600 Forestry tracks - 5,000 5,000 TOTAL 17,700 14,000 31,700 Of the unpaved rural network, about 2,000 km or less than 10% are improved with some drainage, earthworks and gravel. Traffic volumes on the paved classified network average about 1,100 vpd and have been increasing in recent years at about 10% annually. Traffic on the unpaved rural road network ranges from only a few vehicles per day to over 300 vpd with an average between 50 and 100 vpd. Traffic consists mainly of pick-ups for most commodity and passenger mcvements, as well as light trucks and agricultural equipment. 2.02 At independence in 1956, Tunisia inherited a diversified and relatively well-developed road network which was adequate to support the moderate traffic through most of the 1960s. Tbme road system was assessed in 1969 under the Transport Survey financed by UNDP with the Bank as Executing Agency, followed in 1971 by the Highway Maintenance Study financed under Engineering Loan S2-TUN. These studies, together with the modernization and bridge rehabilitation studies also financed under the Engineering Loan, formed the basis of the First Highway Project (Loan 746-TUN) which focussed on modernization works and an extensive program (1,900 km) of rehabilitation and road widening designed to meet the demands of rapidly increasing traffic volumes. Although planned as a continuing program item under DPC's capital bucLget, the rehabilitation scheme was cut back to 1,000 km of roads because of rapid inflation and dollar devaluation. More seriously, funds were not made available to continue the program under the Fifth Development Plan (1977-81). By this time, the Bank's Second Highway Project (Loan ].188-TUN) was assisting DPC in improvments to about 200 km of key primary routes, but this program also suffered delay because funds were diverted to other major construction projects such as the Bizerte Bridge. /1 This chapter is oriented towards the rural road network. A more detailed description of the whole road netwiork is contained in Staff Appraisal Report No. 2810-TUN of April 1980, and in a Draft Transport Sector Memorandum.  - 8 - 2.03 Faced with the complete lack of capital budget funds for rehabi- litation work on the primary network in the second half of the 1970s, the DPC built up its force-account surface-dressing capability in an effort to retard deterioration of the network. The rehabilitation program was rein- stated in 1981 with financial assistance provided through the Bank's Fourth Highway Project (Loan 1841-TUN). Road Maintenance 2.04 Significant improvements in the overall road maintenance system, particularly on the paved network, have been made since the mid-1970s. These began with a reorganization of the maintenance services in 1975 based on the recommendations of an organization and maintenance study carried out in 1971, financed under the Engineering Loan (S2-TUN, 1969). The Third Highway Project (Loan 1601-TUN, 1978) only covers rural roads (see para. 2.08), but the Fourth Highway Project (Loan 1841-TUN, 1980) continues the institutional development of the road maintenance system with emphasis on the improvement of management organization and methods, and by supporting a training program. A detailed review of maintenance tasks, crews and methods is being carried out under the project. 2.05 With the developing capability of the maintenance organization during the second half of the 1970s, attention was naturally first con- centrated on the paved network with its higher traffic volumes. As noted in para. 1.09, the unpaved network had never been the subject of a well- organized improvement and maintenance program. The unclassified network had been nominally the responsibility of the Ministry of Agriculture (MA), althouga that M.LL6LE-y had been neither iunded nor equipped to discharge this responsibility. The classified network, including unpaved classified roads, was the responsibility of DPC. However, DPC had neither the main- tenance budget, nor the staff and equipment necessary to maintain the unpaved roads. 2.06 In 1974, the responsibility for the construction of all unpaved rural roads (pistes agricoles), without distinction whether they were classified or not, was assigned to DPC. The DPC, with Bank assistance, immediately began preparation of a rural roads project, resulting in the Third Highway (Rural Roads) Project, which became effective in early 1979. 2.07 Although responsibility for rural road construction was thus clearly assigned to DPC, together with the corresponding capital budget, the question of maintenance of unpaved roads was not resolved. DPC con- tinued to be responsible for maintenance of classified roads but without a commensurate budget, while no specific agency appeared to be responsible for unclassified roads. As an interim measure, the DPC requested and received a capital budget allocation of DT 2 million (US$5 million) to purchase road maintenance equipment, specifically to maintain unpaved roads. This equipment was operated with an ad hoc mixture of financing, the operators and skilled labor salaries being met from the DPC recurrent budget, and operating costs for fuel, supplies, and some unskilled labor  -9- by the governors from Rural Development Program. (Programme de Developpe- ment Rural, PDR) funds. As noted in para. l.OS, DPC diverts much of its maintenance effort to the unpaved network for a concentrated period just before the annual harvest, and lightly grades the most important classified and unclassified roads. Little maintenance is carried out on the roads during the rest of the year. Since the roads are mostly of natural soil without earthworks or drainage, tLe surface becomes loose sand in summer or mud in winter, depending on the soil type. Successive blading of these surfaces and wear from traffic has lowered the roads below the level of the surrounding terrain, aggravating the situation. In this condition, normal maintenance cannot provide a satisfactory level of service. 2.08 Although the maintenance organization has been concentrating on the paved network, the Third Highway (Rural RoaLds) Project focussed atten- tion on maintenance of the unpaved network. Under that project the Bank is financing the improvement, to sound engineering standards, of about 1,100 km of rural roads. Once the roads are iriproved, normal maintenance can then assure a satisfactory level of service. It is also essential to avoid losing the capital invested in the road i.mprovement. The Government gave a formal assurance to the Bank that roads improved under the project would be subsequently maintained. To assist wi.th this maintenance, the Bank, under the project, financed the procurement of equipment for rural road maintenance brigades. 2.09 In order to assure maintenance of the improved roads and to fulfill the maintenance covenant, the DPC, withi support from the Coordi- nating Committee, requested and obtained from the MPF a new item in the 1981 recurrent budget. This item is entitled Article 60 bis, and is specifically for the maintenance of roads improved under the rural roads program. Together with the Fourth Highway Project which is addressing improved maintenance of the whole network, the capability is now being developed to maintain both the paved and improved unpaved roads to a satisfactory level of service. In order to su])port these levels, real increases of 10% annually in the recurrent maiatenance budget (Article 60) have been established, together with allocatioas to Article 60 bis com- mensurate with the maintenance tasks required for the increasing kilome- trage of improved rural roads each year. Current road user revenues from vehicle registration fees, taxes on fuel, lubricants and tires, and import duties on vehicles and spare parts are estimated at over DT 80 million annually, which is more than twice the investment and maintenance expendi- tures on road infrastructure. An assurance was obtained from the Govern- ment during negotiations that adequate recurrent budget levels will be made available for rural road maintenance activities and that these levels will be reviewed as part of each annual plan of operations (para. 3.25). B. Road Administration 2.10 DPC has about 3,500 employees including 55 engineers and 360 technicians and comprises six headquarters divisions (programming and organization, maintenance and operations, major road works, studies,  - 10 - applied research and laboratory, and equipment maintenance and operation) and a central administrative and accounting service (Chart No. 1). DPC operates in the regions through 19 road subdivisions, one for each pro- vince. The subdivisions carry out routine maintenance operations with force account brigades, while light periodic maintenance is carried out by national brigades controlled from headquarters. 2.11 Planning, design, preparation of contract documents, and tender calls are the responsiblity of the studies division of DPC. The division has design groups capable of handling small projects and routine items while the preparation of large projects is carried out by consultants. DPC is currently attempting to institutionalize the major function of analysis and planning, particularly for routine rehabilitation programs. Competent domestic consultants are available who prepare medium-sized projects with their own staff, and joint venture with foreign associates for large or specialized projects. 2.12 Construction and major periodic maintenance works are normally carried out under unit price contracts awarded after competitive bidding. Many contractors, both foreign and local, operate in Tunisia and provide efficient competition. Small contracts (of the order of US$1 million) are usually won by Tunisian contractors, while larger contracts (over US$5 million) are frequently won by joint ventures of foreign and local contractors. Construction supervision is carried out by DPC engineers with the assistance of the laboratory, except for large special projects where consultants are used. Although supervision is generally satisfactory, experience gained under the First Rural Roads Project has shown that supervision undertaken by the sub-divisions needs to be reinforced to provide the flexibility required with rural road construction. The Government has recognized and is addressing this problem under the ongoing project, and intends to provide increased use of engineering consultants on an intermittent basis to supplement its own services for the proposed project (para. 3.17). C. Organization of Agricultural Services General 2.13 Agricultural Services are organized at the provincial (Gouver- norate) level in Regional Agricultural Development Commissions (Commis- sariats Regionaux de Developpement Agricole - CRDAs). A Central Commission in Tunis (Commissariat Central du Developpement Agricole, CCDA) provides overall coordination (Chart No. 2). CRDAs provide agricultural extension for farmers, undertake forest management and planting, assist the BNT to appraise credit applications and coordinate public organizations involved in agriculture at the provincial level. In practice, the CRDAs have been limited by cumbersome administrative procedures and a lack of autonomy in dealing with investments and project implementation, but recent legislation has given them greater autonomy with the aim of increasing their capacity to handle investments and to speed up project implementation. These  - 11 - changes are in line with the Government's efforts to decentralize economic activity and to step up public investments in agriculture during the Sixth Development Plan (1982--86). 2.14 In addition to the CRDAs, special development authorities (offices) have also been created to handle specific agricultural projects, mainly those concerned with irrigation or with particular development strategies designated for well-defined areas. The latter include the Northwest Rural Development Area, where the Bank and other donors are firLancing projects, and the Central Tunisia Area, where a project is being financed by USAID and other important investments are planned for the future. As project development authorities are financially accountable, they offer several advantages for projects whe.e cost recovery is an important element. This approach is also effective in areas where there is a clearly identifiable development strategy and where integrated project activities may extend over parts of several diEferent provinces. This is the case with the Northwest Rural Development Project, which is concerned with improving agriculture in marginal, watersaed areas and which extends into seven provinces and CRDAs. Agricultural Extension 2.15 Agricultural extension is provided by MA through three depart- ments, namely: the Department of Training, Research and Extension (Direction de l'Enseignement, Recherche, et Vulgarisation, DERV) which is responsible for training and rural communicatiDns; and the Crop Production Department (Direction de la Production Vdgetale, DPV) and the Livestock Production Department (Direction de la Production Animale, DPA) both of which supervise farm-level extension operations. At the provincial level, agricultural extension agents are attached to the CRDA offices. Extension is also provided by the semi-autonomous Cereal Board (Office des Cereales, OC), by certain project development authorities, particularly in irrigated perimeters, and some special interest groups. With effective technical agricultural personnel, at the present time, of about 1,100 engineers and technical specialists and about 5,000 technicians, and assuming some 60% are engaged in so-called extension, there seems to be an oversupply of extension personnel. When account is taken of staff required for adminis- trative and other duties, this staffing represents an effective average density of one agent for 100 to 150 farmers. An acceptable density would be one agent for 250 to 400 rainfed farmers and one agent for 50 to 100 farmers in irrigated areas. In reality, there is duplication of effort as agents from different institutions contact farmers about the same subjects and there is considerable involvement of agents in non-extension activi- ties. An overall review of the extension organization is provided in a draft Agricultural Sector Memorandum and a suumary is in Annex 2. 2.16 Weaknesses in agricultural extension services, particularly a lack of facilities and equipment, the rigid financial framework imposed by a highly centralized system, and lack of specific training and management of agents, have been partly responsible for the slow growth in agricultural  - 12 - production and to low rates of use of recommended inputs. For example, fertilizer use, particularly for smaller farms, has stagnated on the whole since 1974/75 at about 4-5 kg of nitrogen and 5-6 kg of phosphate per ha of crop land per year. On average, this is about one-fifth the level required for optimum net value of crop production in Tunisia. The Government was aware of these problems and under the Fifth Development Plan (1977-81) started to reinforce agricultural extension activities by focussing extension activities through new extension centers (Cellules Territoriales de Vulgarisation Agricole, CTV) and developing new extension messages to introduce improved cropping practices (Annex 2, para. 8). The First Rural Roads Project has assisted the Government in this program by financing equipment and buildings in those areas served by project roads and by initiating a training program for extension agents. This assistance fits fully within a broad national scheme for improving agricultural extension as defined by the MA, and which is expected to be given high priority under the Sixth Development Plan (1982-86). 2.17 In its continuing efforts to improve agricultural extension services, the Government is currently seeking to increase the efficiency of extension activities. This is a priority area under an on-going Agri- cultural Sector Survey which is being prepared with Bank assistance. Problems which are being addressed include: (i) the poor coordination of extension activities of the CRDAs, OC, and various project development authorities, (ii) the large amount of time spent by extension agents on administrative and credit matters, (iii) the absence of links between research and extension, (iv) inadequate methods and communications skills practised by extension agents, and (v) poor management and follow up of extension programs. Steps have already been taken to strengthen the CRDA system and to improve coordination in agricultural development in the provinces. Under the system now evolving, extension departments are being set up in the CRDAs to direct all extension work, except in irrigated perimeters and areas served by development authorities, where the CRDAs would only be responsible for coordination. The extension departments would receive direct technical inputs from the central Department of Training, Research and Extension (DERV), the Crop Production Department (DPV) and the Livestock Production Department (DPA). DERV is currently upgrading its agricultural extension training program which could con- stitute a component of possible Bank assistance under the proposed Fifth Education Project. Discussions are also continuing on possible expansion of the Bank's current involvement in agricultural extension with a view to its participation in a national agricultural research and extension pro- gram. In this way, this proposed highway project would provide continuity of Bank involvement in agricultural extension in Tunisia and an effective bridge until such time as extension projects can be developed. Agricultural Credit 2.18 Agricultural credit is provided primarily by the Government-owned BNT. BNT also provides loans to other sectors. Financing for agricul- tural credit is provided from BNT's own resources, rediscount facilities of  - 13 - the Central Bank, World Bank loans, Government grants, and bilateral sources. BNT provides medium- and long-term loans, as well as short-term loans for annual inputs on which guarantees are usually provided by local mutual guarantee associations (Socidtes de Caul:ion Mutuelle, SCM). 2.1.9 In addition to BNT's normal agricultural credit funds, the most important source for medium- and long-term credit is the Government- financed Special Fund for Agricultural Developmnent (Fond Spdcial pour le Developpement Agricole, FOSDA). In spite of Baink efforts over a long period, credit procedures and loan terms and c)nditions still vary with the type of fund. For medium- and long-term loans made by BNT under FOSDA, loan appraisals are decentralized and are made by MA agents, not by BNT which carries no risk but serves as the Government's agent in disbursement and recovery of loans. The faster appraisal process for FOSDA credit and lesser requirements for their distribution and recovery have resulted in farmer preference for these funds and a relative lack of interest in credit finianced by BNT from its own funds. However, as there is an overall shortage of agricultural credit, BNT is able to be selective in cnanneling its funds to credit-worthy farmers. This tends to concentrate medium- and long-term agricultural credit in the hands of the larger farmers. On the other hand, the short-term credit program operates effectively and provides for critical seasonal inputs such as fertilizer, pesticides, seeds and hired equipment services to a wide range of farmers. 2.20 The standardization of credit procedures and conditions has been one of the main objectives of the current Second Agricultural Credit Project. BNT has already taken steps to speed up loan processing proce- dures by streamlining its own organization. The Agriculture Credit Department has been split into two, one handling administrative matters and the other handling technical matters including training of staff in BNT's 52 branches. The need to continue decentralization of credit procedures and to strengthen BNT regional staff will be pursued under the Third Agricultural Credit Project. In the meantime, improvements in the pro- vision of agricultural credit have followed from legislative action simplifying procedures for obtaining land titles necessary for long-term credit and a reclassification of certain long-term credit items to medium- term credit. However, the Government is continuing to seek ways to further improve the credit system and is considering such options as opening up agricultural credit to other banks and the creation of a specialized credit institution. Such reforms are likely to be introduced as part of the Sixth Development Plan (1982-86) and the proposed project would benefit from these institutional changes. A review of agricultural credit in Tunisia is provided in Annex 3. D. Rural Road Program Objectives 2.21 Until 1975, rural road improvements consisted of a modest program financed by the MPF under PDR and small scale feeder road improvement schemes included as integral parts of some agricultural projects. In 1975,  - 14 - the Government asked the Bank to assist in financing a nationwide program of improvements to rural roads, including complementary investments necessary to achieve forecast development potentials in the areas served. The First Rural Roads Project was presented to the Board in June 1978. The project provided for a five-year program covering the improvement of about 1,100 km of roads in 8 out of Tunisia's then 18 provinces. The Government has recently requested extension of the project to three additional provinces. The program responds to a basic need to improve accessibility in rural areas and is in line with the country's emphasis on regional and rural development. 2.22 Given the relatively advanced level of agricultural development throughout most of the country, the improvement of rural communications plays a key role in supporting the agricultural sector and in the develop- ment of market-oriented activity in general. Moreover, in serving mainly small rainfed farms with an average of about 10 to 15 ha, the rural road program is addressing an important section of the rural population which has hitherto been somewhat neglected by the Government's development programs and projects. The First Rural Roads Project alone serves an estimated 250,000 persons representing 7% of the rural population. The areas tributary to project roads cover 500,000 ha of cultivated land or about 10% of the total cultivated area in Tunisia. 2.23 The main objective of the rural roads program is to implement integrated investment packages combining road improvement with a strengthening of agricultural extension services and provision of agricultural credit. The investment packages were developed for roads where road user savings alone justified improvement as well as for those which could only be justifie, v.lLi. comxpiementary agricultural invest- ments. Incremental needs identified under the project take into account existing facilities and services and are fully integrated with overall national programs. Coordinated action for the investment packages is the responsibility of a Project Coordinating Committee which acts regionally through MPW sub-divisions and the CRDAs. The First Rural Roads Project includes financing for road construction and maintenance operations, and for strengthening extension services, and draws on funds provided under the Second and Third Agricultural Credit Loans to BNT for the credit component. Rural Road Project Progress 2.24 Progress on the road construction component of the project has been excellent and 800 km of the road improvement program have now been committed. All of the highway maintenance equipment planned for procurement under the project has been delivered, run in, and is being sent to MPW subdivisions responsible for maintenance operations. A special recurrent maintenance budget for rural roads (Article 60 bis) has been introduced and has been set at a satisfactory level for 1982 (para. 2.09).  - 15 2.25 Although initial progress on the agricultural extension component was slow, considerable progress has now been made in procuring equipment, constructing extension centers and in training extension agents. In 1980/81, construction started on 23 extension centers, and about 50 vehi- cles and miscellaneous extension equipment were procured. MA proposes to complete the balance of the program in 1982/E3, comprising the construc- tion and equipment of about 50 extension centers designed to serve the remaining areas under the project. 2.26 Extension agents have contacted all farmers in the project areas for agricultural credit and project monitoring purposes. Information from these surveys is now being used to update extension guidelines for each road. The implementation of short-term credi.t in the areas served by the project has been successful and amounts disbursed are exceeding the incre- mental needs identified under the project. Eey inputs financed by short- term credit include fertilizer, seeds, pesticides and custom services. In many areas the project has been the vehicle :Eor introducing the use of short-term credit for the first time. OveraLl it had been estimated that less than 5% of the farmers in the zones were borrowing for seasonal pro- duction needs. For 1980, the first year that project actions could have influenced short-term credit distribution, 1,560 loans totalling about 550,000 Dinars were made in the zones of influence of roads in Kef, Siliana, Nabeul and Kairouan. This is estimated to be a three-fold increase. On the other hand, the provision of medium- and long-term credit for on-farm investments has been slow, partly because of slow processing by BNT but also because of the risk in lending to the smaller farmers served by the project. The first medium- and long-term loans attributed to the project activities were made in the same four provinces in 1981 when BNT approved 560 loans, estimated to be a two-fold increase over previous levels. Some improvements were made in early 1981 by simplifying procedures for obtaining land titles which are necessary for long-term credit and reclassifying certain long-term items, particularly forage crops, as medium-term items. The Government is also continuing to seek ways to improve disbursement of agricultural credit including speeding up loan processing procedures through decentralization (para. 2.20). Monitoring and Evaluation 2.27 Considerable progress has also beeii made with the monitoring and evaluation system for the project. The CRDAIs and BNT are now reporting project implementation information on a regular basis and this is being coordinated by the Central Agricultural Development Commission (CCDA) in Tunis assisted by the Planning and Statistical Division (Direction de la Planification et des Statistiques Agricoles et Economiques, DPSAE) of MA. With the assistance of the National Agricull:ural Studies Center (Centre National d'Etudes Agricoles, CNEA), MA has also made progress in setting up a monitoring system to measure agricultural impact. Similarly, DPC will undertake traffic surveys to monitor variou.3 aspects of rural mobility and use of services. Consultants will coordinate traffic and household surveys for twelve sample roads improved under the ?roject, and will develop a framework for a socio-economic impact study (paras. 3.45-3.46).  - 16 - Future Role of the Rural Road Program 2.28 Following its experience with the First Rural Roads Project, the Government is now committed to the integrated approach of reinforcing agricultural extension in the zones of influence of the roads that are constructed. The Government considers this approach to be the only way to ensure that planned agricultural development takes place on time to get the full benefit of the road development. The first project has also high- lighted some of the weaknesses in agricultural extension and credit services. 2.29 Under the Sixth Development Plan (1982-86), the Government plans to increase its emphasis on rural development, irrigation schemes, agri- cultural extension and agricultural credit. While rural development and irrigation projects would be concerned with specific project areas, the rural roads program would continue to address development in the main- stream of Tunisia agriculture throughout the country, namely the small- and medium-sized rainfed farms where there is a large potential for future increases in agricultural production. Both approaches--national and regional--are complementary and require close integration with overall programs for the improvement of agricultural extension and credit.  - 17 - III. THE PROJECT A. Background 3.01 The proposed project is a follow up to the Third Highway Project (First Rural Roads), Loan 1601-TUN. It continues the same general approach developed for the first project, that is, combining road improvements and complementary agricultural investments in integrated programs for specific rural roads and the agricultural areas they serve. Some modifications have been made based on lessons learned from implementation of the first project. The project was identified in July 1977 and has been prepared by the DPC and MA with the assistance of consultants and Bank preparation missions. B. Objectives 3.02 The project would improve the level of access and increase agri- cultural production in rural areas in Tunisia. Moreover, it would serve to consolidate the general approach and program for rural road improvements initiated under the first project. Specifically, the project is designed to; (i) raise the level of service p:rovided by project roads to a minimum "all weather" standard, commensurate with forecast traffic; (ii) increase agricultural production within the zones of influence of the project roads; (iii) improve standards of living in rural areas including better access for the provision of social and administrative services; (iv) improve the capability of the road maintenance organi- zation so that project roads continue to provide the desired level of service; (v) enhance the benefits of the project roads by improving selected connecting feeder roads in the zones of influence;l/ (vi) strengthen the institutional framework by continuing to assist project preparation, implementation and monitoring, together with some applied research; and 1/ The term feeder road in this context is used to designate rural roads of a lower class than those being improved under the project. They are generally access or spur roads which feed into the project roads.  - 18 - (vii) continue Bank involvement in a comprehensive road safety program begun under the Fourth Highway Project. C. Description 3.03 The project would consist of: (i) a three-year program (1983-85) of rural road improvements tentatively covering about 100 roads totalling some 1,200 km; (ii) a program of complementary investments extending over the same period for agricultural extension and over five years for credit services (1983-87); (iii) a program of improved rural road maintenance; (iv) spot improvements to feeder roads; (v) equipment for four vehicle inspection stations in provincial centers and one in Tunis; and (vi) technical assistance for project preparation, implemen- tation and monitoring; agricultural extension training; research on unpaved road materials, construction techniques and maintenance; and a study of the socio-economic, agricultural and transport impact of rural roads. Rural Road Improvement Program 3.04 Improvements would consist of upgrading, particularly of drainage and surfacing, generally following existing alignments and conforming to the standards of Annex 4. These standards have been developed in the light of experience with the first project in Tunisian conditions using local materials. The standards were reviewed during appraisal and were discussed and confirmed with the Government during negotiations. Generally, roads will be improved and maintained to one of two specified standards taking into account forecast traffic volumes and existing road conditions. The lower standard would provide a six-meter road platform gravelled to a depth of 200-250 mm (or more over very weak soils) and would be adequate for traffic volumes up to about 100 vpd. The higher standard would provide a nine-meter platform gravelled over a six-meter width to a depth of about 300 mm and would generally apply to traffic volumes in excess of 100 vpd. However, in cases where the existing road platform is of adequate width or where the road provides a through link in the network, the higher standard may be proposed even though volumes are somewhat less than 100 vpd, always providing that the higher standard is economically feasible. In addition, for a few roads with very low traffic volumes and limited future potential, improvements would be made using a thinner pavement structure over a five-meter width.  - 19 - 3.05 Tentatively itncluded in the three-year program are about 100 roads totalling some 1,200 km. A firm work program hlas been established for the first year as well as provisional programs for the second and third years. Programs are subject to modification according to needs and priorities during project implementation (para. 4.02). Iin defining an indicative overall program, due attention was given to regional balance in selecting roads as well as to economic priorities (para. 4.17). Complementary Agricultural Investments 3.06 These investments would follow the same general lines as those now being implemented under the first project. Twa main elements are required to achieve the goal of increased production in the zones of influence of the project roads, first to improve the extension services to farmers, and second to improve their access to credit. 3.07 The Bank would assist by financing the foreign exchange costs of comstructing the equivalent of about 55 CTVs, and equipping them with demonstration and extension equipment. This would ensure more rapid implementation of the national extension plan in the areas served by the project. Government would assign agents from within the existing cadre who have special aptitude for extension both to these and to some existing centers which serve the project areas. The number of extension agents required for the project areas is estimated to be about 100. The method by which these agents would operate and be managed is discussed in Annex 2, para. 8. Detailed extension programs built on the priority production objectives for each zone have been prepared for roads scheduled for con- struction in the first year, and preparation will continue for roads in the second- and third-year programs. 3.08 Provision of adequate agricultural credit to farmers is essential for the success of the project. For the first project, credit was included as an integral part of the project, although it was separately financed. Short-term credit was wholly financed by BNT, while the foreign exchange component of medium- and long-term credit was financed under the Bank's Second and Third Agricultural Credit Loans to BNT. For the proposed pro- ject, agricultural credit needs have been identified in detail (Project File Documents B3-B18), and their provision is an important element of the project. However, as with the first project, credit financing would not be included under the loan. Agricultural credit would be administered by BNT but credit financing would not be restricted to any particular credit program or source. During negotiations, the Government gave assurances that BNT would make adequate credit available for the project in accordance with the needs shown in Annex 3. 3.09 While the program of investments for extension services would be implemented generally concurrently with road improvements, the needs for additional credit for farms served by the improved roads would continue for a long time into the future. However, the major increase in investments can be expected to follow during the first few years after roads are  - 20 - improved. To allow for major credit needs generated by the improvement of roads in the last year of the program (to be completed in late 1986), credit needs have therefore been estimated to the end of 1987. Overall credit requirements for individual roads would be refined within the preparation of detailed annual programs including systematic farmer surveys and an analysis of representative farm models (para. 3.43). 3.10 In addition to on-farm investments which would be eligible for credit financing, project preparation has identified modest rural infra- structure works as part of investment packages on certain project roads. These investments include the construction or rehabilitation of small irrigation schemes and small dams for irrigation purposes. Although such infrastructure works could qualify for credit financing through farmer associations, experience under the first project has shown that associa- tions have not materialized in all cases. However, the MA considers these investments to be important for development, and they have been included as part of the project but with the Government providing full financing. MA's Rural Engineering Department (Genie Rural, GR) would be responsible for preparation and implementation, and each scheme would be subject to detailed design and a minimum economic justification. These works would be included as part of the annual action plan to be submitted for Bank review (para. 3.43). Improved Maintenance Program 3.11 The significant improvements in the overall road maintenance system made since the mid-1970s have been described in Chapter 2 (paras. 2.04-2.09). A comprehensive review of maintenance is being carried out under the Fourth Highway Project, and the methods, procedures and capability to maintain unpaved roads will be further developed under the proposed project. Improved maintenance operations will be specified as part of annual action plans, to be reviewed and commented upon by the Bank, and for which funding will be obtained under the special recurrent budget Article 60 bis (para. 3.25). Spot Improvement Program 3.12 The roads being improved under the project (para 3.04) are generally the more important links in the local road network. Feeder roads connecting with these roads carry less traffic and require a lower level of service. However, feeder roads are frequently impassable for extended periods because of mud and water at a few low points along these roads. Small drainage structures such as paved fords or culverts could render the whole road passable and enhance the overall level of accessibility in the areas served. 3.13 For several years most DPC subdivisions have had programs for spot improvements of this type, financed by provincial funds under the PDR. However, these funds have tended to be allocated on an irregular basis and the program is not well coordinated or supervised. Funds were also frequently used instead for prestige-type improvements, including lengths  - 21 - of bituminous surfacings, without sufficient a':tention to economic priorities and technical needs, such as drainag,e, to preserve the investment. Diversion of crews from routine maintenance tasks compounds the problem. 3.14 The project therefore includes a program of spot improvements on feeder roads, essentially for drainage structures at low points. These, improvements would be on roads located within the zones of influence of the roads improved under the rural roads program. To avoid dissipating the funds on a few expensive projects outside the scope and spirit of this item, a criterion has been agreed that expenditures should not exceed DT 500 per km of feeder road (US$1,000/km). This limit would be adjust- ab:Le for inflation and in the light of experience gained with such works during implementation of annual plans. An overall budget amount for this item has been estimated on the basis of 1.5 km of feeder roads per km of road improved under the rural road program. The global budget for each province will then be allocated to feeder roads ranked by the population served by each road. The proposed program for each year would be included in the action plan to be submitted to the Bank for review and comment (para. 3.43). The improvements would be constructed by small local con- tractors, experienced in masonry and concrete work. The program should also have an important institution building effect at the local level by implementing a systematic program of spot improvements, linked to local development plans and priorities, using explicit socio-economic and techn- ical criteria for project selection and programming. The procedures and criteria to be used in implementing the spot improvement program will be developed by consultants during project implemLentation (para. 3.17). Road Safety Component 3.15 In common with most developing countlies, road accident rates in Tunisia are much higher than in industrialized countries at comparable levels of motorization, and do not appear to be declining towards such rates. Accident rates are, of course, related to road and vehicle conditions, driver characteristics, and enforcement orientation and intensity. In the context of highway projects, initial involvement specifically oriented towards traffic safety was begun under the Fourth Highway Project} with the supply of road marking equipment and traffic signs, and the implementation of an annual program of installation. 3..16 While not particularly oriented towards rural roads, it is desirable to maintain Bank involvement in traffic safety and to continue the dialogue with the Tunisia Road Safety Association (L'association tunisienne de la prdvention routi&re) as well as with Ministries con- cerned. An element has therefore been included in the project to assist the Ministry of Transport and Communications to complete its national coverage of vehicle inspection stations. Passenger cars are tested once a year and commerical vehicles twice a year. The project includes equipment for four vehicle inspection stations located in Kebili, Sfax, Tataouine and Tozeur, and one heavy commercial inspection station in Tunis.  - 22 - Technical Assistance 3.17 The project would provide for about 324 man-months of consulting services. These services would: (i) assist the DPC as needed with the preparation of feasi- bility studies, detailed designs and contract documents, and with supervision of improvements to roads under the project; (ii) assist the MA in the preparation of detailed extension programs and credit needs for the zones of influence of roads in the second- and third-year programs; (iii) assist the DPC and MA in the preparation of a project covering the next phase of rural road improvements; (iv) assist the DPC to prepare guidelines for regional engineers to develop a program of spot improvements; (v) assist the DPC and MA with project auditing and with monitoring, implementation and evaluation of the socio- economic, agricultural and transport impacts of the project; (vi) assist the MA with training, including instructors for extension agents, and short training missions overseas; (vii) assist the DPC to carry out a study of materials, construction and maintenance techniques for unpaved roads in Tunisia; and (viii) assist the DPC with the preparation of a program of road rehabilitation. 3.18 Terms of reference for technical assistance described in sub- paragraphs (i) and (ii) will be similar to those for studies already prepared for first-year implementation items. Those for subparagraph (iii) will be similar to those for the current project, and for sub-para- graph (vi) will be similar to the training financed under the Third Highway Project. Copies of all of these are available in the project files. Outline terms of reference for subparagraph (iv) are available in the Third Highway Project file, and for subparagraph (v) are in the project file (Working Paper C2). A preliminary draft of terms of reference for subparagraph (vii) is available but cannot be completed until the mainte- nance study being carried out under the Fourth Highway Project is further advanced. Item (viii) is similar to studies carried out for the prepara- tion of the Fourth Highway Project. Terms of reference for all of the above were discussed and confirmed with the Government during negotiations.  - 23 - D. Cost Estimates 3.19 Table 3.1 shows a summary of investment costs which are estimated at DT 53.5 million, equivalent to US$104.3 million. Of this the foreign exchange component is estimated to be DT 22.4 million (US$43.7 million) and the local cost component DT 31.1 million (US$60.6 million), including DT 11.6 million (US$22.7 million) in taxes and duties. The front end fee on the Bank loan would be in addition to the above costs. It is US$525,000 and would bring the total foreign component to DT 22.7 million (US$44.2 million). In addition to these investment costs, the Government will provide annual recurrent costs for road maintenance activities, agri- cultural extension services and a project control unit as shown in Table 3.2. The cost estimates have been discussed and confirmed with the Government during negotiations.  -24- PROJECT (DST SUMMARY TABLES Table 3.1: inyestment Costs I/ % of I (DDNAR '000) I (US$ '000) | Foreign % of Total Local Foreign Total Local Foreign Total Ewchange Base Cbsts RCOD WDR 16,021.8 10,681.2 26,703.0 31,243.8 20,829.2 52,073.0 40.0 69.5 BUIIDINES 626.2 268.4 894.6 1,221.2 523.4 1,744.5 30.0 2.3 F4JIPMENT 41.4 70.5 111.9 80.7 137.5 218.2 63.0 0.3 IGATION WDRKS 1,379.5 919.7 2,299.2 2,690.2 1,793.4 4,483.6 40.0 6.0 SIDRT-TEREDIT 587.6 587.6 1,175.2 1,145.9 1,145.9 2,291.7 50.0 3.1 MEDIUM- AND IfOe-THEM CREDIT 1,863.3 1,552.0 3,415.3 3,633.7 3,026.5 6,660.2 45.4 8.9 SPOT IMPROVEMENIS 1,079.0 531.5 1,610.5 2,104.2 1,036.4 3,140.6 33.0 4.2 TECHNICAL ASSISTANCE 354.4 1,063.1 1,417.5 691.1 2,073.2 2,764.2 75.0 3.7 VERiaC TESrINU STATION 3QUIPMENT 120.2 204.8 325.0 234.5 399.3 633.8 63.0 0.8 VEHICLES 268.3 178.9 447.2 523.2 348.8 872.0 40.0 1.2 OVERSEAS MISSIONS - 48.0 48.0 - 93.6 93.6 100.0 0.1 Total BASELINE CDSIS 22,341.9 16,105.6 38,447.5 43,568.4 31,407.2 74,975.5 41.9 100.0 Piysical Cbntingencies 2,234.2 1,610.6 3,844.7 4,356.8 3,140.7 7,497.6 41.9 10.0 Price Contirgencies 2/ 6,502.7 4,704.6 11,207.3 12,680.9 9,174.3 21,855.2 42.0 29.1 Total PROJECT CDSIS 31,078.8 22,420.8 53,499.6 60,606.1 43,722.2 104,328.3 41.9 139.1 Fmort End Fee on Bank Loan - 269.2 269.2 - 525.0 525.0 TUrAiL CDST 31,072.0 22,678.5 53,750.6 60,593.0 44,225.0 104,818.0 Table 3.2: Recurrent Costs!- % of I (Dr '000) - l (US$ '000) | Foreign % of Total Local Foreign Total Local Foreign Total Exchange Base ODsts Road Maintenance 3,750.0 - 3,750.0 7,312.8 - 7,312.8 - 71.2 Extension Services 1,406.5 - 1,406.5 2,742.7 - 2,742.7 - 26.7 Project Control Unit 108.0 - 108.0 210.6 - 210.6 - 2.1 Total Baseline Costs 5,264.5 - 5,264.5 10,266.1 - 10,266.1 - 100.0 Price CoatiugemciesY! 1,709.5 - 1,709.5 3,333.7 - 3,333.7 - 32.5 Total Project Costs 6,974.0 - 6,974.0 13,599.9 - 13,599.9 - 132.5 1/ Base cost estimates are at January 1982 prices. 2/ Price increases for foreign and local costs are as follows: 8.5% in 1982, 7.5% in each year 1983 to 1985, and 6% in 1986 and 1987.  -25- 3.20 Estimates for civil works for road im.provements and extension buildings are based on unit prices for contracts in Tunisia, particularly recent tenders under the first project, adjusted to January 1982 levels. Quantities are based on substantially completed engineering studies for the first-year program of works and preliminary estimates of the quantity of each cost item required for works in the secor,d and third years of the program. 3.21 The cost estimates for equipment for extension services are based on recent purchase prices in Tunisia updated to January 1982 levels. 3.22 Credit needs, both short-term and invrestment credit are based on feasibility study estimates of agricultural inputs for the zones of influ- ence of roads in the provisional three-year program. Unit prices are current costs in Tunisia for inputs and equipmLent updated to January 1982 levels. 3.23 Estimates for spot improvements are based on sample studies on roads included in the Third Highway Project. They are roughly equivalent to either one paved ford for each 4 km of feedler road, or one culvert every 1.5 km. These quantities are multiplied by unit prices updated to January 1982 levels. 3.24 The technical assistance cost estimat:es are based on recent rates for consulting services in Tunisia, considerinig the likely mix of local and expatriate consultants and experts. The average man-month costs are expected to be about US$8,500 plus travel, office space, secretarial help, etc. A limited amount of short-term, high-levrel expertise is planned to assist with preparation, monitoring, training and research. This assist- ance is estimated to cost about US$11,000 per man-month including salaries, fees, international travel and subsistence. rhe project also provides for about 24 visits overseas by Tunisian officiaL3 in connection with the organization of and attendance at agriculturaL training courses as well as with visits related to research into unpaved road construction and main- tenance; for these missions 100% of the costs of international travel and overseas subsistence is included for a total of DT 48,000 (US$94,000). 3.25 Recurrent costs are provided from the annual budget allocations to DPC and MA, for administration and operations. The costs are based on current allocations plus increased supplies, operating costs and personnel necessary to support the expanded operations ander the project. The global recurrent budget needs were discussed and confirmed at negotiations, and an assurance was obtained that funds, facilities and services and other resources to operate and maintain the roads and extension services financed under the project will be provided promptly as needed. The estimated annual needs for the project will be included in the action plans to be suabmitted to the Bank for review and comment in preliminary form by April 30 and in final by September 30 each year (para 3.43). Recurrent budget requirements for fuel and supplies (Article 60 bis) for rural road maintenance are estimated in current prices at DT 200,000 in 1982 increasing to DT 1.5 million in 1987. Incremental operating expenditures  - 26 - for agricultural extension services under the project are estimated in current prices at DT 120,000 in 1983 increasing to DT 530,000 in 1987. The operating costs of a project control unit are estimated at DT 19,000 in 1982 and DT 26,000 in 1987 (Annex 8). 3.26 Physical contingencies of 10% of base cost have been allowed on all project component items. Contingency allowances for price escalation were calculated for each year of the planned implementation schedule (Chart Nos. 4 and 5) and overall represent 26% of base line costs plus physical contingencies. Price escalation for the capital items is estimated at 8.5% in 1982, 7.5% in each year from 1983 to 1985, and 6% in 1986 and 1987. 3.27 Foreign exchange components have been calculated for each project item. For road construction civil works, the breakdown between foreign and local (including tax components) was calculated for each major item in the bill of quantities. Under the first project, civil works have been and are being procured under ICB. Lots were sized to be attractive to local contractors, and tender calls grouped to be of interest to international contractors. While many foreign contractors prequalified, only a few submitted bids, and Tunisian contractors won all contracts under that project. This is probably connected with the dispersed nature of the works and is likely to be repeated under the proposed project. For the calculation of the foreign exchange component, it has therefore been assumed that all contracts would be won by Tunisian contractors. E. Financing 3.28 The proposed financing is shown in Table 3.3. The Bank would finance under the project all foreign exchange costs of the capital items, except those of short-term and medium- and long-term agricultural credit and the minor-rural infrastructure works to be constructed by GR as col- lective investments (para. 3.10). The Government will finance the local Table 3.3: Prject Finarrnin'/ US$ Million MPW MA M5C BN Grant 2/ Farmers Bank % Bank Total Road Works 42.6 - 28.9 40 71.5 Extension Buildings - 1.7 -- - - 0.7 30 2.4 Extension Equipment - 0.1 - - - 0.2 63 0.3 Extension Vehicles - 0.7 - - - 0.5 40 1.2 Irrigation Works - 6.6 - - - - - 6.6 Short-term Ondit - - 2.9 - 0.6 - - 3.5 Med.,long-term Credit - - 5.9 2.0 2.3 - - 10.2 Spot Inrovents 2.7 - - - - 1.3 33 4.0 Vehicle Inspection Station Equipment - - 0.3 - - - 0.5 - 0.8 Tecchnical assistance 0.7 0.2 - - - 2.8 75 3.7 Oierseas Missions - - - - - 0.1 100 0.1 Fmnt End Fee on Bank Lan - - - - - 0.5 100 0.5 Total 46.0 9.3 0.3 8.8 2.0 2.9 35.5 - 104.8 1/ Including distributed contingencies. 2/ Governmett subsidy.  -27- component of all capital items (except credit financed from BNT's own resources), as well as foreign and local costs of the annual recurrent budgets in line with the proposed annual programs of DPC and MA (para. 3.25). During negotiations, the Government requested that the 1.5 percent front end fee on the Bank loan (US$525,000) be financed out of the loan. F. Implementation 3.29 Implementation will be the responsibility of a Coordinating Committee. A committee was established to implement the First Rural Roads Project and is now functioning successfully. Project implementation organization is shown on Chart No. 3. Membership of the committee will include the Ministries of Public Works, Agriculture, Plan, Transport and Communications, and Interior and the BNT. The: Chairman will be the Minister of Public Works and the Vice-ChairmanL the Director of the Highway Department. 3.30 The Coordinating Committee will be established by Presidential Decree and will meet quarterly. The various agencies in charge of execution will report to the committee. DPC will be responsible for implementing: (i) rural road improvement works; (ii) rural roads mainte- nance; (iii) the program of spot improvements on feeder roads; and (iv) technical assistance for road aspects and coordination of project preparation, implementation and monitoring, research on unpaved road design and maintenance, and coordinating the socio-economic impact study. MA will be responsible for: (i) the improvement of exttension services including the training of agents and the procurement of extension buildings, equip- ment and vehicles; (ii) the minor rural infrastructure works; and (iii) technical assistance for agricultural aspects of project preparation, implementation and monitoring. BNT will be responsible for administering agricultural credit with the assistance of MA extension staff for the preparation of loan applications in the field, 3.31 At the regional level, DPC responsib-lities including construc- tion supervision will be carried out through the regional engineers (chefs d'arrondissement), and MA responsibilities th:rough the regional commis- sions (CRDAs). In order to improve communical:ions and coordination, regional start-up seminars will be organized izy DPC and MA to discuss and explain project objectives, scope, authority .and responsibilities. Holding such seminars was discussed and confirmed with the Government during negotiations (Annex 6). G. Procurement 3.32 Under Tunisian law (D;cret No. 74-75i du 27 juillet 1974 portant sur la-rdglementation des marches publics) the opening of bids submitted in response to a tender call (appel d'offres) is conducted by a commission. The law specifies that the meeting shall not be open to the public. Assur- ances were obtained from the Government durin; negotiations that, to conform with the Bank's interpretation of the Guidelines that bids should be opened in public, representatives of bidders would be admitted to the Opening.  -28- 3.33 Road construction contracts totalling about US$72 million are based on unit prices and would be awarded after international competitive bidding in accordance with the Bank's Guidelines for Procurement. Roads would be grouped in lots for bidding. Contractors would be prequalified to undertake either single or several lots. Except where it is geographi- cally impracticable to group widely scattered roads, lots would be sized at a minimum of DT 500,000 (US$1 million), suitable to be undertaken by regional contractors. Several lots would be grouped for tender calls to a value of about 3.0 million dinars (US$6 million) so as to make them more attractive to larger contractors, both Tunisian and international, who would be permitted to bid for a group as a whole. Adding more lots to a group to provide a larger package value would not be practical because of the dispersed nature of the contract works. During negotiations, the arrangements concerning lot sizes, grouping for tender calls and bidding for a group as a whole were discussed and confirmed with the Government. 3.34 other works include spot improvements to feeder roads and buildings for extension centers. These are small scattered contracts and would be most efficiently procured by local competitive bidding (LCB) pro- cedures which are satisfactory to the Bank. The contracts would be of the order of US$20,000-40,000 with a total value of US$4 million for spot improvements and US$1.6 million for buildings, and would provide useful development opportunities for small local contractors who are experienced in masonry and concrete slab work. 3.35 Equipment to be procured under the project with Bank financing consists of vehicles for agricultural extension services (estimated cost US$ 1.7 million); agricultural equipment for extension demonstrations (estimated cost US$ 200,000), and audio visual equipment for training (estimated cost US$ 5,000). To the extent practical, equipment would be grouped into packages for tendering. Packages estimated to cost more than US$100,000 would be procured under ICB following the Bank guidelines. Packages estimated to cost less than US$100,000 (but with a total value not exceeding US$500,000) would be procured on the basis of Government LCB procedures, which are satisfactory. 3.36 Technical assistance for the various project items would be pro- vided by qualified consultants firms or individuals as appropriate, engaged in accordance with the Bank Guidelines for the Use of Consultants (August 1981). 3.37 The size and nature of the contracts for the various items determine contract review procedures. For rural road improvements, all bidding packages would be subject to prior review of the procurement documentation; for the spot improvements to feeder roads and buildings for extension centers, an initial review will be made of standard tender docu- ments and of the procurement program. Thereafter, a random post review will be made of bid evaluation, award decisions and signed contracts. Prior review of road improvement contracts will cover 92% of the total value of works financed under the project.  -29- 3.38 Most of the equipment eligible for Bank financing will be pur- chased in large packages during the three years of project implementa- tion. Two tender calls by MA are expected, orLe for vehicles, and one for agricultural demonstration equipment. The documentation for these major tenders will be subject to prior review by the Bank. 3.39 As the project is a sector approach and the program for the second and third years is not yet firmly fixed, some minor additional procurement of equipment may be expected later in the project. The procurement documentation for this equipment would be similar to that which would have been previously approved for the large packages. The additional list of equipment would be agreed by the Bank, but packages valued at less than US$100,000 would be given a post review of bid analysis, award and signed contracts. Audio visual equipment with a total value less than US$5,000 would also be subject to post review. Based on the expected size of the bid packages, these procedures will result in prior review of about 90% of the total value of goods financed under the project. H. Internal Accounting and Auditing Procedures 3.40 The internal accounting systems in the Ministries of Public Works and Agriculture have been reviewed and found l:o be adequate for the pur- pose of certification of accuracy and compila:ion and control of project- related financial transactions, including those connected with the road maintenance activities and extension services requirements. Separate accounts will be maintained to record expendi:ures made under the project, the main items of these accounts corresponding to the loan categories. The systems are described in a Working Paper in tie Project File (Document No. Cl). 3.41 Government control procedures, as presently applied, do not produce the necessary audit reports. Althouga data are available and easy to trace, there is at present no procedure for the compilation and produc- tion of timely reports. The Government has proposed that the Contrdle G'ndral des Finances fulfill this role, and tne independent status of the Contr8le Gdndral des Finances is acceptable to the Bank. These arrange- ments were discussed and confirmed at negotiations. I. Implementation Schedule and Monitoring 3.42 Project implementation schedules are shown in Charts No. 4, General, and No. 5, Technical Assistance. The General chart shows that the implementation basically covers three annual programs beginning in each of the years 1983, 1984 and 1985. The charts have been prepared in some detail, based on experience with the pilot project. The annual work program of 400 km is about the same as the maximum rate attained under the first project and is within the capacity of the local contracting industry. The scheduled beginning of each activity can be reasonably achieved under Tunisian conditions, barring unforeseen delays. In the case of the road improvements, preparation of tender documents for the first group is well advanced, and contracts could be let somewhat ahead of  -30- the schedule shown, but a start in January 1983 is considered reasonable for planning purposes. A total construction time of 24 months has been allowed for each group. This is much more than the contractual time allowed, but experience has shown that Tunisian contractors working on rural roads almost invariably exceed the maximum. The finishing dates of each group are therefore realistic rather than target dates. The project is expected to be effective in mid-1982 and execution of the Bank-financed components would be completed by the end of 1986, a total of about 4-1/2 years. These schedules were discussed and confirmed with the Government during negotiations. 3.43 Elements of the first-year program and annual budgetary require- ments for the project as a whole are summarized in Annex 8. Annual plans of operations will be submitted to the Bank in preliminary form by April 30, and in final form by September 30 each year. The Bank's comments can then be taken into account by the Government in determining both capital and recurrent budgets to be allocated to various items for the subsequent year. The overall program, the elements of the first-year action plan and the preparation and submission of action plans for subsequent years were discussed and confirmed with the Government during negotiations. Guide- lines for ongoing preparation, implementation and monitoring of project activities are in Annex 6. An example of an integrated action plan with extension guidelines and implementation data for a road in the first-year program is in the Project File (Document No. B22). 3.44 There are two aspects to project monitoring, firstly to follow the physical and financial progress of the project, secondly to evaluate project outputs. A system for monitoring the delivery of agricultural extension and credit services in the road influence zones has been developed by DPSAE under the first project. Quarterly progress reports are prepared and submitted to the Bank through the Coordinating Committee. A central project monitoring and coordination unit has been set up under the supervision of DPSAE to handle the flow of data from the first project roads and to assist in preparing detailed action plans for the roads to be improved under the proposed project. In addition, agricultural extension agents assigned to the first project influence zones have undertaken a comprehensive census of farms within their area of responsibility. The processing and analysis of these data by the project monitoring and coordination unit will provide direct feedback to the design and imple- mentation of agricultural action programs in connection with future road improvements, as well as generally improve the Ministry's capacity to manage agricultural and rural development program planning. 3.45 The system will be expanded to include operational data which will show whether the project is meeting its objectives in the longer term. Basic data on agricultural impact would be drawn from a socio-economic impact study (para 3.46). In addition, DPC intends to establish a separate monitoring system including traffic counts and measures of surface condi- tion to assist in the planning of road maintenance activities and to assess their effectiveness. During negotiations the project monitoring system was discussed and confirmed with the Government.  -31- 3.46 The socio-economic impact study will be carried out on a sample of roads improved under the First Rural Roads Project and is also planned for a sample of roads to be improved under the proposed project (Annex 7). This impact study will make use of traffic and agricultural data collected for project monitoring purposes, but will go beyond this to address the full range of rural development impacts resulting from the introduction of road improvements together with agricultural support services, including changes in household income levels and consumption patterns, in personal mobility and communications, in levels of service provision and utiliza- tion, and in migration patterns. The study will also monitor changes in traffic mix following road improvement as well as the use of animals in rural transport. The impact study will include measures of baseline data and measures of change one year after completion of project roads in the sample. 3.47 In addition to the quarterly reports and the impact study, the Government will also prepare a project completion report (PCR) within six months of the Loan Closing Date. During negotiations, assurances were obtained from the Government that it will prepare and submit quarterly progress reports and a PCR. J. Disbursements 3.48 The proposed loan of US$35.5 million would be disbursed as follows:  -32- Category uS$ M Percentage of Expenditures 1. Road works 21.0 40% 2. Extension buildings 0.5 30% 3. Extension equipment 0.15 100% of foreign expenditures or 63% of local expenditures (represent- ing the foreign component). 4. Spot improvements 1.0 33% 5. Extension Vehicles 0.35 100% of foreign expenditures or 40% of local expenditures (represent- ing the foreign component). 6. Equipment for vehicle 0.4 100% of foreign expenditures or inspection stations 63% of local expenditures (representing the foreign component). 7. Technical Assistance 2.1 75% 8. Overseas missions 0.1 100% 9. Unallocated 9.35 10. Front End Fee on Bank Loan 0.55 Amount due Total 35.5 3.49 The annual work program covering the improvement of 400 km of rural roads is about the same rate as the maximum attained under the first project and is within the capacity of the local contracting industry. The most frequent constraint on expeditious execution of highway projects in the past has been the delay with which the MPF has made funds available. In addition to the formal commitments under the loan, the high priority accorded by the Government to the rural road improvement program should ensure that suffi- cient funds will be made available in time to execute the program according to the agreed schedule. The disbursement schedule has been developed from the General Implementation Schedule (Chart No. 4). In the case of the Technical Assistance Schedule (Chart No. 5), the various studies have been shown to be carried out at reasonably early target dates. Experience has shown that studies, particularly those not directly linked to construction, often slip behind forecasts dates. The schedule has therefore been main- tained for implementation monitoring of studies, but the corresponding disbursements have been assumed to be spread over the project implementa- tion period to the end of 1986. The amounts involved in studies are a minor part of the whole project cost, the major costs in the project being for  -33- road improvements. The time slice nature of the project, including monitoring by means of the submission and execution of annual programs, and the realistic estimates of construction time should help ensure that project disbursements follow the forecast schedule. Based on experience in Tunisia, an average time of nine months has been allowed from execution to disbursement. The forecast schedule has been compared with historic experience. Unfortunately, the statistical base for the available historic profiles is limited. Although average profiles for all EMENA transport projects and all Tunisian projects are close together, with the transport projects showing 100% disbursed about 7-1/2 years after Board presentation and the Tunisia projects after about 8 years, the range encompassed by these averages is large; for example from about 3-1/2 to over 11 years for transport projects. More recent projects show somewhat shorter profiles. The forecast schedule for this project shows initial disbursements slower than the historic averages, only 6% being disbursed after two years compared with an historic 17%. The forecast catches up with the historic at 50% of total disbursement after about 3-1/2 years, and is expected to be completed somewhat more quickly in 6 years. The more rapid disbursement in later stages is expected to result from the time slice nature of the project. The First Highway Project, the only completed Highway Project in Tunisia, although of a different nature, disbursed 50% after about 4 years, and 100% after 6-1/2 years. The disbursement schedule is therefore considered to be a realistic forecast. During negotiations, the disbursement schedule was discussed and confirmed with the Government.  -34- Table 3.4: Disbursement Schedule IBRD Fiscal Year Cumulative Disbursements and End of Quarter at end of Quarter (US$000 Equivalent) FY83 September 30, 1982 500 December 31, 1982 500 March 31, 1983 500 June 30, 1983 600 FY84 September 30, 1983 800 December 31, 1983 2,400 March 31, 1984 4,100 June 30, 1984 5,800 FY85 September 30, 1984 7,500 December 31, 1984 10,300 March 31, 1985 13,300 June 30, 1985 16,300 FY86 September 30, 1985 19,200 December 31, 1985 21,900 March 31, 1986 24,600 June 30, 1986 27,300 FY87 September 30, 1986 30,000 December 31, 1986 31,400 March 31, 1987 32,800 June 30, 1987 34,100 FY88 September 30, 1987 35,400 December 31, 1987 35,500 K. Environmental Impact 3.50 The project would enhance the quality of the rural environment by improving the standard of the existing roads, most of which are earth tracks, thus giving better access to markets and social services. Most of the improvements would be along existing alignments, and there would be very  -35- little need, if any, to displace dwellings or to consume additional agri- cultural land for road rights-of-way. The pl.anned improvements will have relatively little impact, either positive or negative, on present low levels of air and noise pollution. The location an<l design of drainage structures will be planned so as not to aggravate problems with storm run-off and to reduce the risk of flooding. Care will be taken during construction to mitigate, insofar as possible, potential short-term negative impact on adjacent farm lands and families.  -36- IV. ECONOMIC EVALUATION A. General Framework 4.01 The improvement and maintenance of rural roads in Tunisia responds to a basic need to improve accessibility in rural areas and is in line with the country's current emphasis on rural and regional development. As part of its program to correct regional and urban/rural imbalances, the Govern- ment attaches high priority to the rural road program. Through the program the Government is: (i) providing and maintaining the necessary road system to support agricultural development and hence increase living standards in rural areas; and (ii) ensuring adequate access in rural areas for the provision of social and administrative services. By combining rural road projects with a spot-improvement scheme for connecting lower-class feeder roads, the proposed project will help to integrate local road networks, thereby providing savings to existing traffic as well as improved access for the rural population. Furthermore, improved accessibility will increase the efficiency and effective range of administrative and commer- cial services including agricultural extension and credit, medical and veterinary coverage, marketing services and input and equipment suppliers. B. Sub-Project Review 4.02 Under studies financed as part of the First Rural Roads Project, some 2,500 km of roads were subject to a detailed economic evaluation. These roads were selected by the Project Coordinating Committee from an initial long list of roads prepared for each province by a regional committee. Some of the roads studied have been included under the First Rural Road Project in accordance with the sector or program approach adopted for that project. A similar approach has been retained for the proposed project with appraisal being on the basis of a firm first-year program of about 400 km of roads as well as tentative programs of similar magnitude for the second and third years. The latter may be subject to modification depending on changes in needs and priorities as the program is implemented. Revised annual programs for rural road improvement will be presented to the Bank for prior review, with each sub-project prepared according to the requirements and procedures outlined below. 4.03 In submitting sub-projects for inclusion under the second and third years of the program, the Government would be required to provide: (i) a feasibility report for each road including preliminary engineering and construction costs, estimates of present and future traffic and an assessment of agricultural development as well as the complementary investments necessary to achieve the forecast production increases; and (ii) extension guidelines, farm budget analyses and implementation data for the agricultural component.  -37- 4.04 The criteria for Bank approval of sub-projects are that: (i) the quality of preparation is acceptable and in line with the agreed general approach; (ii) the ER of the investment package is at least 10%; and (iii) the financial rates of return of representative farm models are sufficiently high for faLrmers to undertake the proposed on-farm investments and to be able to take advantage of credit facilities. 4.05 Details of the socio-economic evaluation methodology are provided in Annex 5 and guidelines for the preparation and implementation of integrated programs combining road and agricultural investments are in Annex 6. During loan negotiations, an assurance was obtained from the Government that it will follow the agreed procedures and criteria in preparing and submitting sub-projects for Bank approval. C. Benefits and Beneficiaries 4.06 The economic benefits quantified under the proposed project flow from two sources: transport cost savings and incremental value added on agricultural production. The project roads fall into two categories: those which would be economically justified on transport cost savings alone, and those which would be justified only by including increases in agricultural production and incomes in their influence zones. However, the project has been planned in such a way that both types of benefits are expected to be realized in every case. The underlying assumption is that all-weather access is a necessary precondition for the successful delivery of services to stimulate agricultural production and ensure timely marketing. 4.07 Transport cost savings accrue to people who own, operate, and/or travel in motorized vehicles on rural roads. Theoretically, paid pas- senger transport is provided by the public transport enterprises organized at the national and regional level, but in practice, their services rarely reach rural roads. Collective taxi services operate out of the major towns and provide service on improved rural roads. Informal passenger transport service is provided by other vehicle operators, and conventional charges are levied. Most road vehicles are owner-operated. Thus, user cost savings on passenger transport will accrue initially to the vehicle owners, although savings will be passed on to other users through increased competition following road improvement. 4.08 Approximately 70% of all freight transport on rural roads is carried out using vehicles owned by farmers: tractors, light trucks, carts, and pack animals. Less than 30% is carried in heavy trucks, which would be more likely to be owned by middlemen, cooperative societies, or government agencies. Farmers who do not own trucks generally can arrange,  -38- individually or in groups, to rent one from a more affluent neighbor when there is a significant quantity of goods to be moved. Again, the benefits of transport savings on freight traffic accrue initially to the public sector or to the vehicle owners. 4.09 The real benefit of rural road improvements to the small farmer is not that the monetary costs of travel and transport decrease, but that services are provided at all. Significant improvement in access, time savings and improvements in safety, comfort and convenience enable the farmer to obtain more inputs, market additional products, and have access to services that would be otherwise unavailable. These gains are parti- cularly significant when all-weather or nearly all-weather access is provided to communities which would otherwise be isolated for months at a time. In other words, small as well as large farmers benefit from the provision of regular motorized transport at a reasonable cost. 4.10 Value added in agriculture is generally calculated as a direct function of production increases. It may derive from the introduction of new crops, the cultivation of new land, or the intensification of farming on existing land. The value added is assumed to accrue to the population of the impact zone either in the form of a direct consumption increase or in the form of additional income from marketing surplus production. In most of Tunisia, the potential for introducing new crops is critically dependent on irrigation, and the potential for bringing new lands under cultivation is extremely limited. Most of the anticipated increases will therefore come from the introduction of new or improved technologies to increase yields on presently productive land or simply from the more timely and appropriate use of agricultural inputs. While certain technologies are more likely to be made available to, and to be adopted by, larger farmers first, and by households with a non-farm income source permitting them to accept a certain minimal level of risk, improved techniques will also benefit the small farmers. It is on the small and medium farms that there is a major potential for increases in productivity, but these will be achieved only to the extent that the new technologies are independent of scale constraints or appropriately modified to be available to farmers with small areas and modest means. Some of these constraints can be overcome through farmer associations or equipment hire facilities. 4.11 An unknown but probably small proportion of the rural population is lancdless. Many more own parcels of insufficient size to support a family. An analysis of detailed data from the First Project shows that in one province 5% of households in the road impact zones are landless and another 35% have farms of 5 ha or less. These farmers are dependent on the rural labor market to provide them with cash income to meet their household needs. Rural road improvements increase the scope and range of opportuni- ties to find work and thereby will help to raise wages in the rural labor market. In other words, rural road improvements increase the probability that landless laborers will be able to capture a share of any increase in agricultural value added.  -39- 4.L2 In relation to social services, planners responsible for the provision of health and education services in rural areas of Tunisia indicate that their actions are closely coordiiiated with the development of the rural road network. Recommendations for road improvements are discussed by line agency representatives at the provincial level before they are forwarded to Tunis for technical and economic evaluation (para. 4.02). Progress in implementing road imnprovements is monitored by the agencies and used as a basis for planning future service delivery programs. The rural roads program thus complements and Enhances the chances of success of other Bank-financed projects to meet the needs of rural residents in Tunisia including, specifically, the Health and Population Project which is directed toward decentralization of health services to rural areas. 4.13 In summary, the combination of road inprovements with the imple- mentation of a detailed action plan to promote increased agricultural production in each impact zone will result in increased farm incomes, including the incomes of farmers below the rural poverty level, as well as provide support for the goal of self-sufficiency in major food products. Although insufficient time has elapsed since the implementation of the First Rural Roads Project to provide detailed statistics on road impact, observations and interviews in the project areas indicate substantial initial responses such as in the demand for credit and the use of fertilizers, as well as a major improvement in the perceived quality of life in these areas. It is expected that the project will have a broad economic impact on the rural population through the demonstration effect of new technologies introduced in the road impact zones. Residents will also benefit from improved access to services, either through their increased mobility or through that of service providers. Finally, the amenities associated with road improvements will increase-the quality of life in rural areas and slow the rate of rural-urban migration. Both the economic and social effects will be carefully measured through the impact study, which will cover roads from the First Rural Roads Project and the proposed project (para. 3.46). D. Economic Analysis 4.14 In the economic evaluation, net-of-tsx costs have been used for road construction and maintenance as well as for buildings, equipment and operating expenditures for agricultural extension services. As several agricultural inputs are subsidized at about half their economic costs, the latter have been derived from world market prices adjusted for transport and handling in Tunisia. Similarly, ex-farm prices for agricultural produce reflect current world market prices with the exception of non- tradeable commodities where local market prices are used. 4.15 The approach used in the economic evaluation is based on an in-depth analysis of the rural sector and focusses on:  -40- (i) the characteristics of agricultural production in terms of farm size distribution, crop areas and yields as well as on-farm consumption of food crops; (ii) existing average annual traffic volumes; (iii) the probable distribution of benefits among transporters, traders and farmers in the light of agricultural pricing, marketing and transport conditions; (iv) the probable response of farmers to an increase in ex-farm prices or the relative change in accessibility following road improvement; and (v) estimates of the complementary inputs necessary, such as extension services and agricultural credit, for the fore- cast increases in agricultural production. 4.16 The economic evaluation of each road covers a 20-year period, allowance being made for major periodic maintenance or rehabilitation during the project life. In estimating the benefit streams over time, account has been taken of the absorptive capacity of the various farm types for improved technology. Generally, increased agricultural production is assumed to be achieved over a period of five to ten years, with growth thereafter being in line with long-term projected agricultural growth rates. Transport cost savings are estimated by applying unit vehicle operating cost reductions to forecast passenger and non-agricultural traffic while increased revenues for transporters carrying agricultural products are estimated based on the relative changes in costs and rates following road improvement. Benefits to generated passenger and non- agricultural commodity traffic have not been included and the quantified benefits are therefore underestimated. E. Results 4.17 A provisional three-year program of improvements to 1,200 km of rural roads was selected based on the economic evaluation of over 2,500 km of individual roads. A minimum acceptable ER of 10% has been applied for each individual road. In defining priorities under the program, prefer- ence was first given to those provinces not covered under the first project, mainly the coastal provinces of central and southern Tunisia. In defining the balance of the program under the second and third years, emphasis was given to projects in the relatively poor areas of central and west-central Tunisia and to the northern provinces where there is consi- derable scope for increasing agricultural production (Maps IBRD 15916, 16145, 16144). In all cases, due attention was given to establishing appropriate groups of roads suitable for contract lots. 4.18 The ERs of the 100 individual roads included in the three-year program range from 10% to about 50%. The overall ER for 33 first-year roads is estimated at 21%. For first-year roads, benefits from transport  -41- cost savings represent about 40% of total project benefits and about half of the project roads could be justified on the basis of transport cost savings alone. Even under the most conservative assumption of a 25% reduction in overall benefits combined with a 25% increase in project costs, the overall ER of the first-year progra.m is not expected be less than 15%. Table 2 in Annex 5 provides key parameters of the economic evaluation for each individual road. 4.19 Social benefits would represent a significant addition to the quantified economic benefits. In serving over 200,000 rural inhabitants or about 6% of the rural population, the benefits would include improved access to social, administrative and commercial services, thus increasing the general well-being and mobility in rural areas. Through the socio- economic impact study, it is hoped to improve the selection and evaluation of roads by including social criteria on a syatematic basis. In the meantime, the Government is examining the use of a social rate of return method which would involve the application of income distribution weights to the benefit streams in order to establish relative priorities among justified sub-projects (Project File Document No. C14). F., Project Risk 4.20 The main area of risk associated witli the proposed project is the timely implementation of complementary agricuLtural investments. A delay in such investments would of course delay projected increases in produc- tion and agricultural benefits would be lower than forecast. However, a delay of one or two years would not have a significant effect on the ER of individual roads, particularly as transport cost savings are often an important share of total benefits. Moreover, the economic justification is based on relatively modest increases in agricultural production, part of which should be realized from the increased eEficiency of existing agricultural extension and credit services foLlowing road improvement. 4.21 Implementation of the First Rural Roads Project has done much to familiarize both Government services and the rural population with the oojectives and overall concepts of the rural roads program. Assistance provided for agricultural extension services is now better integrated w:ithin national plans and the Government is nDw strengthening the insti- tutional framework. At the same time, the Government is currently reassessing the organization of agricultural zredit and the project would benefit from improvements in the system. Although there is a need to make medium- and long-term agricultural credit more readily available to the small and medium farmers served by the project roads, most of these farms will, through the project, have access to short-term credit for critical seasonal inputs such as fertilizer, pesticides, and equipment hire. To the extent that these inputs are available and. that better road access and imnproved extension services will ensure their timely delivery and use, it should be possible to attain the modest production increases assumed under the proposed project.  -42- V. AGREEMENTS REACHED AND RECOMMENDATIONS 5.01 During loan negotiations,agreement was reached on various issues referred to in the report, including the following matters of particular significance: (i) The Government will cause BNT to make available adequate short-term, and medium- and long-term agricultural credit (para. 3.08); (i) The Government will formally establish a project Coordi- nating Committee (para. 3.30); (iii) The Government will admit representatives of tenderers for ICB procurement to the bid opening (para. 3.32). (iv) The Government will implement the project in accordance with a three-year action plan, and will submit annual action plans to the Bank for review and comment (para. 3.43); (v) For individual road sections for the second- and third- year programs, the Government will submit technical and economic studies, and agricultural extension guidelines, in accordance with agreed procedures and criteria (para. 4.05); and 5.02 Agreement having been reached on the issues in the report, the project is suitable for a Bank loan of US$35.5 million equivalent to the Government of Tunisia for the standard term of 17 years, including a 4-year grace period.  -43- Annex 1 TUNISIA STAFF APPRAISAL REPORT OF A FIFTH HIGHWAY (RURAL ROAD',) PROJECT Previous Bank Proje,ts A. Transport Sector l. Loan 380-TUN (1964, US$7.0 million, Republic of Tunisia, First Port Project) was for a project, the major com?onents of which were dredging and refilling, the construction of berths, transit sheds and other infrastructure at La Goulette and consull:ing services. The pro- ject was satisfactorily completed. 2. Loan 573-TUN (1969, US$8.5 million, OEfices des Ports Nationaux Tunisiens, Second Port Project) also included a major component for dredging, as well as the provision of dredges and cargo handling equip- ment, provision of a grain berth, training and consulting services. The project was satisfactorily completed (PPAR No. 1049, February 26, 1979). 3. Loan 606/Credit 150-TUN (1969, US$17.) million, SNCFT, Railway Project) was for a project, the major componen:s of which were the renewal of 352 Iam of track, purchase of rolling stock, improvement of workshops and consulting services. The project was delayed by about two years due to the disastrous floods of 1969 and 1971 and shortages of funds resulting from slow payments by the railwiay's principal customer (Sfax/Gafsa Phosphate Co.) prior to 1973, when the Government assumed the debts of the phosphate company. The project was satisfactorily com- pleted (PPAR No. 2109, June 23, 1978). 4. Loan S2-TUN (1969, US$0.8 million) wa3 for engineering studies for hignways included in the First Highway Project. It was satisfacto- rily completed. 5. Loan 724-TUN (1971, US$7.5 million, SDcidt4 Tunisienne de 1'dl4ctricite et du Gaz, Gas Pipeline Project) was for a project which included procurement and installation of pipelines, compressors, and ancillary services. The project was satisfactDrily completed in 1972. 6. Loan 746-TUN (1971, US$24.0 million, First Highway Project) was for.partial financing of construction and improvement of 274 km of high- ways, resurfacing and rehabilitation of 1,920 km of roads, reconstruc- tion of 51 bridges and consulting services. The project encountered difficulties with the rise in costs at the time of the oil crisis.  -44- Annex 1 In addition, further studies showed that the Tunis-Turki section should be constructed as a four-lane freeway rather than as the three-lane road included in the project. The Board, therefore, approved exclusion of the Tunis-Turki road from the project; this road was later financed by the Kuwait Fund. The disbursement rate was reduced on the remaining items. All other construction items and consulting services were satis- factorily completed (PPAR No. 2772,' December 26, 1979). 7. Loan 1188-TUN (1976, US$28.0 million, Second Highway Project) was for the improvement of about 225 km of primary and secondary roads, and consulting services to assist in a transport coordination study and to assist with the preparation of a Rural Roads Program. Construction has been completed on two lots of GP8 Tunis-Bizerte and is nearing completion on MC 28/27 in Nabeul. It is in progress on GP6, Beja- Jendouba, GP5 Mornaghia-Medjez el Bab and the GPl Sousse bypass. Both studies under the project are complete. The project has encountered persistent problems with delays caused by insufficient provision of local funds by the Ministry of Planning and Finance, but is now expected to be satisfactorily completed by the end of 1982. 8. Loan 1601-TUN (1978, US$32.0 million, Rural Roads Project) was for the improvement of about 1,100 km of rural roads; the procurement of road maintenance equipment, and agricultural facilities and equipment; the provision of investment and short-term credit; the improvement of extension services; and technical assistance to the Ministries of Public Works and Agriculture. Construction is nearing completion on four lots in the north and complete or underway on roads in all eight of the provinces originally included in the project. Implementation of the complementary agricultural investments and extension services suffered an initial lag, but is now proceeding satisfactorily. The road main- tenance equipment has been procured. Technical assistance for the preparation of a second tranche has been completed. General project progress is satisfactory. 9. Loan 1797-TUN (1980, US$42.5 million, Third Port Project) provides for the modernization of port facilities at La Goulette and Sfax. The major components of the project include specialized faci- lities for roll-on/roll-off vessels, storage areas, transit sheds and warehouses, as well as additional cargo handling equipment and repair and maintenance facilities. Project progress is satisfactory. 10. Loan 1841-TUN (1980, US$36.5 million, Fourth Highway Project) is aimed at improving the management of the Tunisian highway network through maintenance, rehabilitation and strengthening the institutional framework. The first year of a three-year action plan is being imple- mented, including routine maintenance and rehabilitation, road mainte- nance equipment is being procured and a training program is being established. Project execution is satisfactory.  -45- Annex 1 B. Agricultural Sector 11,. Loan 484/Credit 99-TUN (1969, US$18.0 million, Cooperative Farm Project) for a total amount of US$18.0 million was subsequently reduced to US$9.8 million. Following major changes in Tunisia's agricultural policy in 1969, the original project was substantially revised in 1970 and completed at the end of 1973 with US$8.8 million being canceled. The Project Performance Audit Report which was circulated to the Board on January 8, 1976, discussed the difficulties of Tunisian agriculture under the collective system in effect until late 1969 but confirmed that productivity, production, working incomes and institutional viability objectives were attained. Two other objectives, elimination of sub- sistence farming and incorporation of private land owners into produc- tion cooperatives, were abandoned. 12. Loan 779/Credit 263-TUN (1971, US$8.2 million, First Agricul- tural Credit Project) was to finance part of the National Bank of Tunisia's (BNT) lending to commercial farmers for investment in grain farm mechanization, dairy farming, and date palm plantation. Imple- mentation was slow initially but the loan was fully disbursed by July 1968, 34 months after the original closing date. The PPAR indicated that mechanization had not yielded the benefits expected and that more attention should have been devoted to the agro-industrial component which suffered major cost overruns. 13. Credit 270-TUN (1971, US$2.0 million, First Fisheries Project) was for a project aimed at development of Tunisia's inshore fisheries and was completed at the end of 1979. Implementation encountered con- siderable delay, and the lending institution (BNT) experienced low loan recoveries. 14. Loan 1068-TUN (1974, US$12.2 million, Irrigation Rehabilitation Project) is for a project to rehabilitate and improve the use of exist- ing water resources as set out in the primary stages of the Water Master Plan. This project covers the Lower Medjerda Valley and Nebhana. The Lower Medjerda Valley component includes about 20,000 ha of which 75% is already equipped with irrigation and drainage facilities; the remainder lacks secondary and tertiary distribution networks. The Nebhana compo- nent includes the rehabilitation of existing irrigation, drainage and road networks over about 5,000 ha in the Nebhana area. The loan became effective in September 1975 and the physical implementation is satisfactory. 15. Loan 1340-TUN (1976, US$12.0 million, Second Agricultural' Credit Project) became effective July 19, 1977, and aims at supporting a two-year lending program for on-farm development of small and commercial fanns and for agro-industries and a four-year lending program for small  -46- Annex 1 holder date/palm plantation development. The loan, which also serves credit needs under the Rural Roads Project (Loan 1601-TUN), is nearly fully disbursed. 16. Loan 1431-TUN (1977, US$42.0 million, Sidi Salem Multipurpose Project) is the first phase of execution of the Northern Tunisia Water Master Plan. It includes the irrigation of 10,600 ha in Testour/Medjez el Bab and Nabeul areas, prevention of the decline of 6,000 ha of citrus in Nabeul, improvement of agricultural production on 32,800 ha in the Lower Medjerda Valley, helping meet potable and industrial water requirements, reduction of periodic flood damage, generation of elect- ricity and the improvement of rural roads. Implementation is satisfactory. 17. Loan 1746-TUN (1979, US$28.4 million, Second Fisheries Project). This project became effective on May 14, 1980. Agreement was reached on a strategy to improve recoveries of subloans. Project pro- gress is satisfactory. 18. Loan 1796-TUN (1980, US$25 million, Southern Irrigation Project) covered irrigation systems in oases in the desert provinces of the south. The project has started up satisfactorily. 19. Loan 1885-TUN (1980, US$30 million, Third Agricultural Credit Project) was designed to continue the institution building begun under the First and Second Credit Projects. It became effective on June 24, 1981. 20. Loan 1997-TUN (1981, US$24 million, Northwest Rural Development Project) is financing a 5-year time slice of a 15-year development pro- gram in the northwest region of Tunisia. It will improve the income, nutrition, health and education of the population in the project area while reducing soil erosion and the consequent decline in agricultural productivity. The loan will become effective in January 1982. 21. Loan 2052-TUN (1981, US$42 million, Grain Storage Project) is designed to create a modern bulk distribution system in the import circuit from port to mill. The major components are: rehabilitation, expansion and construction of grain silos, construction of rail sidings and purchase of hopper cars, and technical assistance. The project is due to become effective in January 1982.  -47- Annex 2 TUNISIA STAFF APPRAISAL REPORT OF A FIFTH HIGHWAY (RURAL ROADS) PROJECT Agricultural Extension IPrograms A. Principles 1. What is needed to induce improved agricultural production is described in many ways. Simply said, farmers regardless of the size of their land holding generally increase their productivity when four prerequisites are met. These are as follows: (i) an improved farming system: whereby a combination of materials and practices that is clearly more productive and profitable, with an acceptably lower level of risk, than the one currently used, become.s available to the farmer; (ii) instruction of farmers: whereby the farmer is shown on his own farm or nearby how to put the practices into use and made to understand why they are better; (iii) availability of inputs: whereby the required inputs, and if necessary, the credit to finance-their purchase are avail- able to the farmer when and where he needs them and at a reasonable cost; and (iv) availability of markets: whereby the farmer has access to a nearby market that can absorb the increasing supply of produce without excessive price drops. None of these is more or less important than t:he other and they inter- relate. Trying to attribute more importance to one than the other is like trying to evaluate the impact of one alone, not only difficult but probably illogical. However, a strong dependence exists for the second, third and fourth prerequisites on all weather access to the point that its absence could prevent their impact. 2. Extension is ineffective without appropriate profitable techno- logy. Frequently in examining extension systems the major weakness is found to be in the research area, and also the payoff from improving the extension service is to put research closer to the farmer. Research carried to the logical conclusion of demonstration in the farmers fields goes a long way to minimizing extension problems. Research is a vital link. Regional extension services suffer froci some or all of the following deficiencies, namely:  -48- Annex 2 (i) subject to centralized and split authority; (ii) staffed by people whose technical training and skills exceed their training, skill and aptitude in communications; (iii) burdened with statistical and regulatory demands, and responsibility for input and service supplies; (iv) oriented to mass media methods frequently due to lack of mobility, too high a ratio of farmers to agents and a reluctance to visit; (v) short of appropriate messages; (vi) infrequent contacts with research; and (vii) takes no account of the subtle cultural characteristics of the rural population being treated. 3. The needs for extension depend on the nature of the agriculture and the needs change as the agriculture changes. Hence, an extension system must evolve with the country's agriculture. In determining the approach that may now be appropriate, it is important to gauge the point of evolution of the agriculture and it may be different from sector to sector. The amount that can be invested in extension (and research) largely depends on the scope for yield development and this will vary with agro-ecological zones within the country. B. Development of Extension in Tunisia 4. Considering the way in which agricultural extension within Tunisia has developed gives a useful background to consider the present issues and the proposals being developed for the future. This is described in greater depth and detail in the agricultural sector survey 1/. Furthermore, it provides a basis for assessing the extent to which the principles described above have or can be incorporated. Before independence, agriculture was characterized by a marked contrast between the modern mechanized sector managed by colonists in which production was directed towards export, and the traditional sector of small and medium sized farmers which was a mass of people living in a subsistence economy. The first used techniques adapted from those in Europe which may not have always been in the best long-term interest of the natural soil resources but which were at the time productive. The traditional sector practiced an archaic subsistence agriculture which neither the Government nor the colonists were interested in modernizing. 1/ This survey is being carried out concurrently.  -49- Annex 2 5. In the first ten years from independence, a structure was put in place by the Ministry of Agriculture in which efforts were mainly directed to the creation of development offices, management of state lands and the staffing of production cooperatives. It was attached within the Ministry in a way in which it could not be assured satisfactory working arrangements and it did not survive long. A major developnment commenced in 1967 when the Ministry of Agriculture, realizing the importance and perhaps the vul- nerability of the sector recreated an extension service with modest means. The service was concerned with both studies and interventions. It had nine trained agriculturists (four Tunisians) at the central level. It gave priority to the provinces of Le Kef, Jendouba, Beja, Bizerte, and the region of Cap Bon. The regional staff was concerned with many aspects including extension. Mainly, the central staff visited the regional staff to discuss rural problems with a view to finding solutions. The principle was valid but the impact of the results was negligible and did not justify the cost of operation. 6. In 1968, the idea of campaign themes was introduced and at this point real extension started. At that time the mass media information activity started as a support to extension campaigns. With time, the Ministry raised the status of the extension activity and the professiona- lity of the occupation. A significant step in 1971 saw the development of professional national organizations directed towards promoting production. At a point in 1974, the Department of Training, Research and Extension (DERV) had been created; individual and group extension at the farm level was firmly established in the Department of Crop Production (DPV); offices and special groups were deeply involved in indlividual and group extension for their constituents in their particular sphkeres of influence; and pri- vate enterprises were beginning to advise their clients who were mainly large commercial farmers. So extension services in Tunisia are recent and have evolved rapidly. 7. The objectives assigned to Tunisian agriculture for the Fourth and Fifth Development Plans (1972-1976, 1977-1981:1 were to correct the food balance in which the deficit was growing dangerously and to improve the employment situation and the living conditions; of the rural population. An overview of the situation in the period of the Fourth Plan reveals three major trends, namely: (i) a multiplicity of structures and organizations concerned with instructing farmers; (ii) a multiplicity of methods and approaches used for extension; (iii) a substantial development of rieans (300 graduates, 1,100-1,200 agriculturalist, of average ability and some vehicles).  -50- Annex 2 In total, the effect on sensitizing and instructing the farmer was posi- tive, but it posed a significant problem of coordinating actions and syn- chronizing the facilities and services available. 8. In the Fifth Plan, steps were taken to resolve the above pro- blems. Actions within the mass media services were directed to broad scale sensitizing of farmers to problems, themes, important technologies and to providing technical support in the form of publications and extension aids for the individual direct contact extension. Farm level extension was to be directed towards having a single interlocutor for a group of farmers and an integrated approach by this direct contact agent to the whole farm. At the same time, the offices and groups who were individually interested in one product were expected to retract somewhat and provide specialized support that would be directed to the farmer through the direct contact agent. The strategy required that, from within the general mass of agricultural technicians, a team of extension agents be selected exclu- sively for direct contact, personalized extension. At the farm level, they would operate from work centers (CTVs), which would be located in the farmirng communities. Several CTVs would be managed by a more specialized agriculturist at the delegation level. The structure would provide, at the provincial level within the CRDA, specialized agricultural graduates to program and supervise their activities and provide subject matter support. An individual work center would provide accommodation and an office space for the agent, a small meeting room and storage. Locating the work center in the community allows the farmer easy access to the agent, but the principle method envisaged was for the agent to visit as many of his clients as frequently as possible to extend selected themes and identify farmer's problems. So that within the period of the Fifth Plan a program to develop the system of direct contact extension using work centers located within the activity sphere of the individual agent was drawn up and some aspects were started. Some work centers are being constructed (para.. 11) and agents have been nominated for them. Mostly they are located at nearby CRDA offices awaiting completion of buildings. The program at the farm level is a logical one to deal with the Tunisian situa- tion. The installation of management at the delegation and provincial level is lagging behind and this is critical because multi-purpose agents at the field level cannot operate effectively without being strongly sup- ported and regularly programmed. 9. Despite the principle of a single interlocutor, the fragmentation still exists. As of now, there are several directorates within the direct jurisdiction of the Ministry of Agriculture and a large number of organiza- tions and offices under aegis of the Ministry of Agriculture involved in extension. of those directly responsible, the DPV (which is engaged in farmer contact) and the DERV (which is concerned with mass media extension) are the two most important. The offices and organizations involved in extension include commercial offices for livestock, cereals, wine, oil, development offices such as those to cover the irrigated perimeters, the northwest and central Tunisian development projects, and special interest groups for vegetables, citrus and dates.  -51- Annex 2 C. Problems Which Exist at the Present 10, The Tunisian authorities are largely i;ware of existing problems and expect to address them in the course of the Sixth Plan. They come in four main categories, namely, organization, lilnkage with research, person- nel and means. Fragmentation results in duplia:ation of effort and contra- dictory advice. There is confusion between extension, regulatory and administrative demands put on the agents and taeir supervisors, and the agent frequently sees himself and is seen by tkie farmer to be an enforce- ment agent. The movement of extension actions is one way with little or no feedback. Contacts between research and exten3ion are infrequent, and research tends to be rather pure and without sufficient consideration of the limited financial resources of the farmers. Extension staff have rarely been selected for their communication capabilities, frequently are less competent technically than some of their -lients and mainly have been assigned their roles without regard to personal preference. They are little motivated because of lack of housing and transport, and feeble support from supervisors who are overburdened with other responsibilities. D. Bank Involvement Through First Rural Roads Project 11. The Third Highway Project (Loan 1601-TUN), known as the First Rural Roads Project, is making an important contribution to agricultural extension in Tunisia through support for the development of the work center system. It is financing the construction of work centers and equipping these as well as equipping some centers which have been set up in existing or rented buildings. The following program for developing CTVs is underway as a subcomponent of the First Rural Roads Project. Under Construction Planned Total New construction 23 51 74 Refurbishing of existing buildings 29 29 Total 52 51 103 This contribution to the planned total system of around 500 can be compared with 81 being established through other sources to this time. These work centers developed with funds under the First Rural Roads Project are appro- priate for the zones of influence of the roads improved in the project. The Government of Tunisia agreed to assign at least one extension agent to each work center. All of the agents for the centers located under con- struction have been assigned. They have been located temporarily at the CRDA or delegation office to which they are responsible and moved to the centers as they are complete. Agents have been working in their designated  -52- Annex 2 zones whether the buildings are complete or not. They have surveyed all of the farms in their zones to collect base data of farm resources. They have strongly promoted the use of credit and are beginning to extend themes outlined for them by the CRDA offices. The means of the agent, particu- larly transport and equipment for demonstration, are being provided from the loan. A further and significant contribution to the agricultural program is for training the extension agents in techniques of communi- cation. This is being provided by a consultant who works closely with DPV. 12. Progress in providing these two main items in the agricultural subcomponent was slower to begin with than road construction because locat- ing suitable agents, developing building plans, preparing construction contracts and purchasing demonstration equipment took longer to arrange. After the initial delays, progress has been satisfactory and is recorded in quarterly reports. The major issue rests now not within the activities financed from the first project but in the national approach to extension. This is currently under review as part of a sector survey. There still remains a multiplicity of agencies carrying out extension in Tunisia and the work center system is not yet responsible for all extension work. The manner in which the scope of activities of the other organizations would be retracted to use them in support of the central system is expected to be defined during the Sixth Plan. The work center concept at present lacks flexibility to adapt to variable needs of sectors or areas, although this could be improved. There are sufficient technicians with experience to develop the management network, and Tunisian proposals indicate their acceptance of the principles of methodical farm extension. They are in the process of defining their target groups and refining their production cam- paign priorities. This is expected to be implemented within the framework of changes proposed in the Sixth Plan. E. Agricultural Extension Component in the Project 13. The component proposed is simply an expansion of that being car- ried out in the first project, namely, construction, staffing and equipping a number of additional work centers which fit the national plan. For each road, a two-stage study has been done. The zones of influence of likely roads were defined and a preliminary feasibility study carried out to determine whether the road improvement could be justified. For each road retained, an execution plan has been prepared (or is in the process of being prepared). This plan describes development opportunities, crop and livestock production improvement. Because the proposed roads spread through 18 provinces, there are a variety of extension priorities according to agro-ecological zones. Production increases are based on modest yield improvements using existing knowledge, and do not depend on generation of new technology even though this is likely to occur during the life of the project. It makes a proposal for extension work centers which takes account of existing services. An example is included as Project File Document B22. In areas managed by a development authority it is proposed  -53- Annex 2 that any strengthening of extension services would fit the work center system (para. 8), but with the development authority retaining overall management in its area. This organizational structure is likely to be proposed in the concurrent sector study. 14. Table 1 sets out for the roads proposed in the first-year program the length, number of farms, and type of agriculture and calculates that 33 additional agents would be needed according to the following ratios, which have been developed with Tunisian authorities having in mind similar situations: Rainfed farming in the North (cereals, livestock) 1:250 Rainfed mixed farming in the North (cereals, treecrops, livestock) 1:200 Rainfed fanming in the South 1:400 Treecrop farming in the South 1:400 Rainfed mixed farming in the South 1:350 Irrigated farming in the North 1:100 Oasis farming 1:100 For the whole project, 100 additional agents would be needed. Table 2 reduces these numbers to estimate that the equivalent of 19 additional work centers would be needed the first year, 57 for the whole project, taking into account 1.25 agents per work center and that, on average, 30% of the number required in the national plan have already been completed. Detailed annual plans will specify exact locations of centers to be built and the distribution of project funds between new construction and additions to existing centers, taking account of existing services and the needs of individual zones of influence. The cost estimate can be calculated using the following costs for a work center unit:  -54- Annex 2 Cost of a Work Center Unit in Dinars Investment Cost Investment Cost Office, store, meeting room (85 m2) 9,555 Lodging (55 m2) 6,195 Office equipment ) Audio visual equipment) 525 Demonstration equipment 1,445 Transport - vehicle 4,200 - motorcycle 210 22,130 Recurrent Costs: Annual Cost Personnel Cost 1 extension agent 3,150 0.25 coordinator & subject matter specialist 1,050 0.15 office support 210 Running Costs Demonstration equipment 155 Transport 893 Office and Management 410 Demonstration parcels 260 Cost tables in Annex 8 summarize the quantities and costs for the overall program. The total cost for buildings, vehicles and equipment is about DT 1.5 million (January 1982 prices) with an estimated 36% foreign exchange component. 15. It is proposed to continue for two years the training component that started in the first project, and to provide for start-up seminars, which proved to be extremely successful. Provision is included for study trips overseas for extension supervisors. The individual costs of these three items are as follows: Man-months uS$ Training 27 230,000 Start up Seminars 6 50,000 Study trips 10 40,000 TOTAL 320,000  -55- Annex 2 F. Farmer Participation 16, Table 1 shows the number of farms for the zone of influence of each road in the project first-year program expected to be served by the exten- sion service. The following is the expected rate of participation from the beginning of the project for those farms estimated to be served by the extension service.