Tunisia - Fifth Highway (Rural Roads) Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 12468 PROJECT COMPLETIOh REPORT REPUBLIC OF TUNISIA FIFTH HIGHWAY (RURAL ROADS) PROJECT (LOAN 2108-TUN) NOVEMBER 8. 1993 FILE Copy Report No: 12468 Type: PCR Infrastructure Operation Division Country Department I Niddle East and North Africa Region This document has a resticted distibution and may be used by recipients only in the performance of their officdu duies. Its contents may not otherise be d'iclosed wibout Wodd Bank authorizaton.  CURRENCY EQUIVALENTS Currency Unit = Tunisian Dinar (DT) EXCHANGE RATE;S Appraisal Year Average 1 DT 0.46 Intervening Year Average 1 = DT 0.81 Completion Year Average 1 = DT 0.93 GLOSSARY BNT - Banque Nationale de Tunisie (National Bank of Tunisia) CTV - Cellule Territonale de Vulgarisation Agncole (ExtLnsion Centers) DGPC - Direction Generale des Ponts et Chauss6es (Highway Department) DPV - Direction de la Production Vegetale (Crop Production Departnent) ERR - Economic Rate of Return GR - Genie Rural (Rural Engineering Department) ICB - International Competitive Bidding LA - Loan Agreement MA - Ministry of Agriculture MPW - Ministry of Public Works MTC - Ministry of Transport and Communications SAR - Staff Appraisal Report REPUBLIC OF TUNISI FISCAL YEAR January 1 - December 31  FOR OFFICIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A Office of Director-General Operations Evaluation November 8, 1993 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Tunisia Fifth Highway (Rural Roads) Proiect (Loan 2108-TUN) Attached is the "Project Completion Report on Tunisia - Fifth Highway (Rural Roads) Project (Loan 2108-TUN)", prepared by the Mtddle East and North Africa Regional Office. Part 2 summarizes the findings of a borrower-produced PCR. Project objectives were to improve the standard of living in rural areas, raise agricultural production, enhance rural communication and strengthen institutions with rural development mandates. Rural road works (road improvement and maintenance), complementary investments in extension services and irrigation, farm credit and technical assistance were main components. Project implementation and attainment of objectives suffered from complex and unclear objectives, inadequate preparation and appraisal and weak institutions. Implementation took five years longer than originally planned for a total of almost ten years. Major physical targets were achieved or exceeded but institutional development aspects, including road maintenance, were cut back to varying degrees. The PCR is frank but lacks basic information about project outcomes and implications, the reault of inadequate monitoring and evaluation arrangements and poor supervision performance. As a result, overall project outcome cannot be rated. Institutional development is rated as negligible, and sustainability as uncertain. The project will be audited. Attachment This docxanmnt has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents .ay not otherwise be disclosed without World Bank authorization.  FOR OFFCIAL USE OiLY TABLE OF CONTENTS PREFACE................. I ........ i EVALUATION SUMMARY .....................................ii PART I: PROJECT REVEW FROM THE BANK'S PERSPECTVE .......... .. 1 ProjectIdentity ......................................... 1 Background: The Agrcultual Sector, TheTransportSectorandThe HighwaySubsector ......... .. 1 ProjectObjectvesand Description ............................. 3 ProjectDesipnandOrganization ............................... 6 Projectpl nati .................................... 9 MajorPectResults ................ 18 ProjectSustnability ........... .... 24 Bank's Perfo.ance ................ 25 Bonrower's Peorance ............... 27 Relations between The Bank and The Borrower ................. 28 Consultant Services .......... ............................ 28 Project Documention and Data ................. I . ........... 29 PART II: PROJECT REVIEW PROM THE BORROWER'S PERSPECTNVE ...0... PART III: STAIISTICAL INFORMATION ........................... 32 IBRD MAP No. 24898 This document has a restricted distribution and may be used by recipients only in the performance of their ofcial duties. Its contents may not otherwise be disclosed without World Bank authorization.  PREFACE This Project Completion Rport describes the preparation, appraisal, and implementation of a Fifth Highway Project in Tunisia, for which Loan 2108-TUN in the amount of $35.5 million was signed on May 14, 1982. The loan was scheduled to become effective by September 13, 1982, but became effective nineteen months later on April 24, 1984, due to delays in issuance of the decree rearding a govemmental committee to coordinate the vanous activitieq under the project. The loan was closed on October 31, 1991, after four extensions. Ultimately, $31.5 million were disbursed and $4.0 million were cancelled. The Evaluation Summary and Parts I and m of the PCR were prepared by the nfrastructure Operatons Division of Country Departmenit I, Middle East and North Africa Region, on the basis of the Staff Appraisal Report, the Loan Agreement, supervision reports, correspondence between the Bank and Tunisian authonties, intenal Bank records, and a working draft prepard jointly with the Highway Directorate (Direction Gen6rale des Ponts et Chaussdes), Ministry of Equipment and Housing, Tunis as well as other information gathered during a PCR mission in April-May 1993. The Govenment of Tunisia has prepared a separate Project Completion Report on the basis of the same working draft. Part II of the present report summariM3 the main differces between the two documents.  EVALUATION SUMARY Project Objectives and Experience Gainedh (i) The Fifth Highway (Rural Roads) Project was a continuation of activities undertake in the Third Highway (Rural Roads) Project to improve rural roads and upgrade the agncultural areas surrounding those roads. In most respects, the two projects had adopted simiar physical designs, time-frames and analytical and administrative approaches. The number of rural roads to be improved was roughly equal (1,000 km in the Third and 1,200 In in the Fifth) and the projects had similar execution p&iods of 8 and 10 years, respectively. (ii) L. each project, iurl roads were selected and justed on the basis of combined transport and .gricultural benefits and costs, and both projects sought to generate agricultur benefits by ensuring the provision of agricultural investments and extension services (para. 4.02). While the Third Project actually included Bank financing to support agricultural credit, the Fifth Project only obligated the Govenmment to ensure that credit would be available. On the other hand, the Fifth Project made extensive efforts, as did the Third, to facilitate extension services though provision of physical ficilities for the agets and through training. (iii) Similar intearental committees were used in both projects to coordinate various project activities. However, the Fifth Project required a much more detailed planning and monitoring process, with budget projections and imlementation schedules for all governmental functions, both in the Ministry of Public Works and in the Ministry of Agriculture. Moreover, the change in agricultural activities was to have been studied, planned and executed in detail, wnth comprehensive monitoring to ensure that plans were caried out and to provide a basis for impact assessment. (iv) The Fifth Project also served as the vehicle for a number of activities related to the rural roads program (para. 3.02 (b) to (M (i), (v), (vii), and (1i) ). Among them were a program of rural road maintenance, improvement of roads feeding into the rural roads improved under the project, construction of other rural public works such as irrigation schemes, studies of rural road construction and maintenance techniques and of a follow-up rural roads project, and technical assistance and taining to carry out the various road and agricultural components. All of these, except the rural road maintenance program, received financing under the loan. In addition, there were activities unrelated to ruml roads, also financed under the loan: preparation of a classified road rehabilitation program and procurement of equipment for Tunisia's vehicle testing program as a contribution to road safety (para. 3.02 (f) (vi) and (g)f. Results (v) in general, execution of the rural roads components of the project was more satisfactory than execution of the agricultural ones. The first extraneous components, the study of a program to rehabilitate classified roads was mostly executed under another project (para. 5.27) and resulted, under tis Project, only in a modest technical assistance consultancy to help prepare bidding documents for 312 km of rehabilitation works. The second extneous component, procurement of vehicle inpection equiptnent, was held up for many years due to delays in designing the component by the agency in charge, fte Mnin:try of Transport and Communications (paras. 5.03(b) and 5.12). As a result, this component took up nearly double the amount budgeted for it under the Project.  - lii - (vi) Construction of the Project nrral roads was largely completed by 1988, two years behind schedule, though supplementary works continued through 1992 (para. 5.01 and 6.03). While cost performance is difficult to establish accurately due to design changes and changes in the value of the currency, oveall costs were generally within the parameters established at appraisal. Hov ever, rural road maintenance posed some problems. Maintenance, especially routine maintenance, was not carried out systematically (para. 6.02) and even though budgets for routine maintenance have been steadily increasing (para. 5.14), money may have been diverted to other works. Spot improvements on *oads feeding into project roads were carried out mostly in parallel with the work on the corresponding project roads though achievement fell considerably short of the projected volumes (para. 5.1S). (vii) Construction of extension centers (CTVs) was delayed more than coi-truction of rural roads, and although a substantial number of the CTVs was completed by 1988, the last six were not finished until 1992 due to hisufficient budgets (para. 5.17). The agricultural component suffered most seriously from inadequate implementation of planning and monitoring elements. The Coordinating Committee failed to have regular meetings and the -'anous supportig consultancies were utilized to an extent much less than foreseen at appraisal. Timetables and action plans for the govemorates and the extension agreements were not prepared or submitted to the Bank nor were seminars held for local officials (para. 5.07). After 1985, the two parties agreed that the Govemment need only to submit the annual budget of the agency in charge of the agricultural component. All the extension agents received some training and had introduced extension contacts in various degrees but data is lacking to quantify the extent of the adoption of new farming techniques and to compare the current use of such techniques with projections at appraisal. The record discloses repeated references to institutional problems in Tunisia's extension system (such as parallel extension services dealing with different crops), which was described as unduly fragmented and inhibited in scope. These problems have recently been alleviated through a consolidation of extension services. The construction of the CTVs also contributed to this administrative improvement (para. 6.16). The improved roads in combination with the CTVs have substantially helped the funtioning of the extension service. (viii) The appraisal used a complex procedure to establish an ERR for each rural road in the first year's program, based on both traffic volumes and agricultural production and not on simplified methodologies that the Bank had developed (para. 6.10). The complexities were due not only to conceptual problems but extraordinary data demands (paras. 6.12 and 6.13). A study is currently under way to establish the impact, including the ERR, of a sample of the Fifth Highway Project roads, but its results were not available at the time the present report was finalized. Further, the conceptual and data collection complexities are so great that they cast doubt on the utility of the procedure, particularly for roads directly affected by Tunisia's increasingly urbanized and affluent economy (para. (xii) and para. 6.14). Nevertheless, the vigorous traffic growth on most project roads points to a most likely favorable economic outcome (para. 6.15). (ix) Large scale use of consultants was contemplated at the time the loan was made, but only about a third of the funds for this purpose was disbursed; much of this shortfall occurred either because they were not sufficiently defined at appraisal or because the Government did not view them as necessary or helpful (para. 6.0S). However, the Project benefited from the continued involvement of the consulting joint-venture (Tunisian and French) that had prepared the Project with financing from the Third Highway Project, and that, as a  - iv - group or as separate consultants, assisted in implementation. The same joint-venture is now carrying out the impact study and the preparation of the proposed Seventh Highway Project. Sustainability (x) The experience of this Fifth Project suggests that it is only partly sustainable. The roads themselves will require more systematic maintenance if they are to latq to their projected term, though they are built well enough to last for some time even in the face of insufficient maintenance. The maintenance of improved roads remains a problem as maintenance budgets are in general insufficient, fungible between categories of roads, and most personnel concerned with impaved roads consider them not to be real roads. On the evidence of the Fifth Project, the systems of identifying and building rural roads in conjunction with agricultur initiatives have not become firmly rooted in the Tunisian environment and will have to evolve (pan. 7.01). TIis experience contrasts somewhat with that under the Third Project where the linkage between road and agncultural investments was considered to be desirable (even though the provision of agricultural credit was regarded as inadequate) and where several key Tunisan officials strongly spported the integrated approach. Conclusions and Lesons Learned (xii) General. The Project succeeded in achieving its main physical objectives, the construction of improvements on nearly 1,200 km of rural roads and the constructing and equipping of 55 CTVs. The use of a single construction contract to cover work at several scattered sites is administratively advantageous, contributed to the result and should be considered for use elsewhere (para. 8.03). Ancillary physical objectives were only partly met (feeder roads branching off the project roads and vehicles for use by extension agents) or were not subject to full reporting (rural infrastructure works and farm investments). However, the Project fell short in the implementation of its non-physical components and was deficient in the use of consultants toward that end (paras. 5.07 and S.25). The Project, also, is only partly sustainable because road maintenance is still deficient on some rural roads and because an effective road maintenance system and integrated rural road - agricultural development program, stll, do not have an adequate institutional basis (para. 7.01). (Xiii) Project Scope. The integrated approach used in this Project presupposed certain essential requirements which were not met: (a) a sense of ownership in the Project by all participating government agencies (para. 8.04 (b)); (b) effective mechanisms for coordination among these agencies and within the Bank; (c) intensive Bank supervision (para. 8.02); and (d) clear articulation of the benefits of the integrated approach (para. 8.04 (a)). (xiv) Quality at Entry. The Project suffered from serious defects jk initiQ. The selection of roads occurred without an adequate road inventory (para. 8.04 (c)). The road designs were inappropriate because in most instances additional improvements were added shortly after the inital work was completed (pan. 6.02) and Project analysis was too academic and execution procedures were exceedingly complex (paras. 3.02 and 6.14). In addition, many Project components were not fully thought through and inadequately laid down in the Project *~~~~~~~-1 .-- r. __ '__] A - * ;e +2osa a h  - v - (xv) Project Execution. This complex integrated Project also fell short of expectations because of inadequacies in execution including: (a) too limited Bank supervision and discontinuities in Bank staff assigned to the Project, especially for tfie agricultural components (pam. 8.01); (b) inadequate information flows and monitoring caused by non-implementation of the Project component (monitoring and auditing consuitancy (para. 5.24) ) specifically directed to these needs; by too limited reporting requirements (para. 12.04); and, generally, by the Bank's acquiescence in insufficient production of information on the Project (para. 12.03); and (c) insufficient coordination of Bank activities in support of the Project.  REPUIUIC OF TUNISIA fIHIGHWAY (RURAL ROADS) PROJECT (LOAN 2108-TUN) PROJECT COMPLETION REPORT PART k PROJECT REVIW FROM TH1 bANK'S PERSPECTIVE I: PROJECT IDENTITY Name: Fi Hghway (Rural Roads) Project Loan Number: 2108-TUN Loan Amount: $35.5 million RVP Un t: MENA Country: Republic of Tunisia Sector. Transportation Subsector. Highway 11: BACKGROUND: THE AGRICULTURAL SECTOR, THE TRANSPORT SECrOR AND THE HIGHWAY SUBSECTOR 2.01 In the face of rapid urbanization (from 36% in 1960 to 54% in 1990) Tunisia's gicultural sect has experenced only slow growth over the past seval dedes. The avge value of the gross domestic product from agriculture rose by about 10% over the decade of the 1980's, with considerable vadaions from year to year due to naturl fctors such as weather ad locusts. As agrculture grew more slowly than the economy as a whole, its share in the GDP, fell from almost 18% in 1976 to around 15% in the mid 1980's. It appears that the producin of several traditional crops such as cerals and olive oil had only slow growth or none at all in comparison with vegetables and fruits which epanded more rapidly: Average Annual Production (thousands of tons) 1974-1980 1982-1986 1987-1991 Cereals 990 1170 1408 Olive oil 107 98 112 Tomatoes 296 378 500 Potatoes 117 143 215 Citrus 188 194 240 This production was achieved with a slightly declining absolute number of persons engaged in the sector (a decrease of about 4% between 1975 and 1988). Moreover, while the tota population contnued to rise, the prorton of agrcultural workers in the labor foe feU steadily from 379O in 1975 to 24% in 1988. These data point to shifts in the nature of agricultral producdon, to rising productivity in the sector and to improvements in fte conditions of rura life though not a reversal of the long term exodus from the farms to the cities.  - 2 - 2.02 In contrast to agnculture, the value of output in the transport sector grew significantly during the 1970's and 1980's keeping pace with the growth in the overall economy which expanded rapidly in the 1970's at 7 to 8% per year, with rises in manufacturing and tourism, as well as sharp increasts in Tunisia's oil revenues. In the 1980's average overall growth remained positive but averaged only 3.6% as the sectors which had driven the economy in the prior decade experienced reduced demand and lower prices. By 1988, transport represented 4.5 % of GDP as against 4.4 % in 1976. 2.03 The overall growth in transport has also been marked by an increase in road transport's share of surface movement of both passengers and freight, and by the mid-1980's road transport accounted for more than 90% of all passenger traffc and, if captive phosphate traffic is excluded, for nearly 70% of freight movements. These developments are evidenced by increases in traffic density on the paved network (10% annually between 1977 and 1992). The vehicle fleet also expanded rapidly from 71,000 in 1972 to 258,000 in 1984 and to 350,000 in 1992. While the length of the network increased only marginally, its condition improved from roughly 7,500 km of paved roads around 1970 to over 9,000 kan in 1984. By then the network consisted of about 17,000 km of classified roads (roughly half of them paved) and 20,000 km of unclassified roads (about 85% unimproved earth tracks). 2.04 The most important issues regarding Tunisia's road system in the early and mid 1980's were: (a) About two tirds of the paved roads were 5.5 m or less wide, according to a Road Maintenance Study of 1984, financed under the Tunisia Fourth Highway Project, and many of them needed to be widened to carry their current levels of traffic; (b) Maintenance, which was decentralized in 1981 needed to be strengthened, parficularly through better organization (e.g. by designating a specific agency to maintain rural roads), improved central support to regional road units and greater provision of equipment and training; (c) Regional and local roads needed to be improved in support of the Government's overall program of rural and regional development, which in turn supported Government objectives for increased agricultural production and more even income distrbution. The first two points are being addressed by the Highway Maintenance and Rehabilitation Project (Sixth Highway Project, Loan 2896-TUN of January 15, 1988), which is currently under implementation while the last point represented the principal thrust of the Fifth Highway (Rural Roads) Project. Two new projects are being prepared, covring respectively rural roads and a road improvement program. 2.05 Public sector investments in Tunisia have oeen guided by a seres of five year plans, including the Fifth (1977-1981) immediately preceding the Fifth Highway Project, the  -3 - Sixth (1982-1986) roughly paralleling the original Project peiod and the Seventh (1987-1991) which covered the extended Project period. Most of the Project rural roads were improved duing the Sixth plan which increased the Govemment's emphasis on rural development through the expansion of the rural roads program throughout Tunisia. While the share of transport in total investments has steadily decreased from 17.6 % in the Fifth Plan to 8.3% in the Seventh Plan, the share of roads and road transport in total transport investments has gone up from 27% in the earliest peiod to 52% in the latest period (actual investments for the Fifth and Sixth Plans and planned investments for the Seventh Plan). Investments in roads (excluding road tra1i.prt) show a rising trend for the Fifth and Sixth Plans from an annual average in real 1985 DT of DT 27.1 million in the first period to DT 35.6 million in the first four years of the second period. In the years 1985-86, the Project accounted for roughly 12-15% of all road investments. The total roads budget (capital, maintenance, administration) averaged DT 39.7 million (in constant 1985 DI) in the Fifth Plan and DT 50.1 million in the first four years of the Sixth Plan, with recurrent maintenance accounting on average for DT 12.6 million (or 25%) of the average budget in the latter period, a 28% increase over the average in the Fifth Plan. This works out to DT 1,378 per kin of paved road in 1984. 2.06 The Ban's policy concerns regardng T:nisia's transport sector have focussed on issues of productivity, government reguion, pricing and taxation; many of these concerns were voiced as early as the 1960's and continue in the current dialogue in a proposed Transport Sectr Development Project. Some of the road tanwsport industry and enteprises in the railway, maritime and aviation sectors are all state owned and the Bank has continually emphasized the need for improved productivity of these enterprises. In the roads sector the Bank has advocated grater flexibility in the management of public sector firms and greater involvement of pnvately owned opers, processes which are being implemented, albeit slowly. Road transport pricing is thought to be too rgidly controlled while road user taxation, though ap iat in overall amount, is thought to be biased in favor of cetain categories of vehicles (heavy vehicles). Improvements in the review of transport investment proposals have also been a long term concern, focussing particularly on the creation and strengthening of institutions for this purpose. There is now greater emphasis on addressing road safety and envirnmntal problems although of al the above broader policy issues, the Project only focussed on road safety. 2.07 A summary of prior and cunt Bank Group operations to assist Tunisia's road sector is shown in Table A: Related IDA Credts and Bank Lean. m. PROJECT OBJECTIVES AND DESCRIPTION 3.01 The Project adopted a similar design strategy and set of d&jectives as the Third Highway (Rural Roads) Project, which provided for rural roads' improvement in conjunction with programs of investment, credit and extension to support agricultural activities in the zones 41  - 4 - of influence (defined as approximately one km on either side of the improved road) of the roads. However, there were some differences between the two projects: (a) whereas the Third Project covered eight governorates, the Fifth included seven of the same govemorates plus twelve others, thus covering nineteen of Tunisia's twenty three governorates; 0) the investment activities in each governorate under the Fifth Project were covered by a detailed Action Plan and all activities under the Project were to be specified yearly through annual Action Plans; (c) the Third Project provided for an actual allocation from a Bank loan (Second Agricultural Credit Project) for credit in support of farmers in the zones of influence as well as credits from various other identified sources. The Fifth Project only made estimates of the credit likely to be required in each area and specified that Banque Nationale de Tunisie (BNI) would ensure that credit would be available as required. (d) the Fifth Project added the following elements: (i) a speca program to improve selected feeder roads which branched off the rural roads improved under the Project (though some feeder roads were included in the Third Project); (ii) a program of modest rua infrastructure works (e.g. small dams) to be designed and executed by the Ministry of Agriculture's G6nie Rural (GR) or Rural Egineering Department; (iii) consulting assistance for design and execution of the Project and study of general issues related to rurl roads; and (iv) procurement of vehicle inspection equipment to promote road safety. The. Fifth Project thus intended to have a more comprehensive investment program and greater surveillance in the process of execution than the Third. 3.02 As approved by the Board, the Project included: (a) A te-year program (1983-1985) of rural road improvements totalling some 1,200 km, of which 400 km had been appraised when the Project was approved and the rest were to be agreed by the Bank in two annual increments (Table B: Rural Road Improvements as Plroposed at Appraisal and as Fnally Executed); (b) Spot improvements to feeder roads branching off the improved rural roads: the spot improvements were not specifically idendfied but were estimated at one paved ford for each 4 kn of feeder road or one culvert  to each 1.5 km of feeder road. Spot improvements were carred out in only nine governorates or feeder roads of only 23 of the improved rural roads ( Table C: Spot Improvements on Related Feeder Roads); (c) A road maintenance program for the rural roads improved under the Project: this program was not specified at a isal except with respect to annual cost ($7.5 million equivalentj but it was to have been descrbed in detail in action plans to be submitted annually; (d) Complementary investments related to agnculture in zones of influence of each rural road to be improved: while these investments were not hsted in full detil at appraisal, they included: - 55 buildings for aetension centers (CTVs) (valued at $1.7 milion equivalent), - equipment for these buildings, (valued at $0.2 mifflion equialent), - vehicles for the 55 CTVs (valued at $0.9 million equivalent), - rural infrstrucre schemes, which were sometimes also desied as rigation schemes, small scale irrigation schemes, land dninage works and small irgation dams, and which were to be submitted to the Bank for review prior to implementation even though the Bank was not to finance any of them (valued at $4.5 million equivalent); (e) Credit from BNT at various terms to farmers in the influence areas of the rural roads subject to improvement. The purposes of these credits were not spefled out but sample lists of goods to be financed were prepared. The aggeg amount of BNT credit included in the Project costs was $8.8 million equivalent ($2.9 million equivalnt short term and $5.9 millon medium and long term) and BNT credit represented about two- thrds of credit from all sources (none of them from the Bank) included in Project costs. (f) Consulting services to assist in: (1) the feasibility, design and implementation of ruml road improvements; (ii) preaation of guidelines for spot improvements; (ii) monitng, auditing and impact evaluation of the Prroject; (iv) a study of constuction and maintenance materials and techniques for unpaved roads in Tniia; (v) preparin of a foLow-on project (vi) p of a classified road rehabilitation program; (vii) preparion of extension and credit programs for the second ant third years of the Project; and  - 6- (viii) training of extension agents, including overseas traimng missons; (g) Procurement of equipment for vehicle inspection stations: no list of equipment was prepared at appraisal and the list eventually filanced was subject to lengthy discussions between the Bank and the Government. This Project element was intended to support a broader highway safety component in the Fourth Highway Project, a component which itself remained to be precisely defined at that time (see Report No. 8648, Project Completion Report for the Fourth Highway Project, Loan 1841- TUN, March 18, 1990). 3.03 Implementation of the Project was to be the responsibility of an interagency Coordinating Committee, including rpresentatives of the Ministries of Public Works, Agriculture, Plan, Transport, Communications, and Interior and BNT (this Committee had already been established under the Third Highway Project). The Direction Gdn&6ale des Ponts et Chaussees (DGPC), that is, the Highway Department, was to be responsible for all activities, mcluding technical assistance, relating to roads as well as the socio-econonic impact study, while the Ministry of Agriculture (MA) would have charge of all activities related to agriculture, mcluding technical assistance for that purpose. The procurement of vehicle testing equipment would be handled by the Directorate of Land Twanspot (DTI) in the Ministry of Transport and Communications (MTC). BNT would administer agricultual credit with the assistance of MA extension staff. The Project was to extend from mid-1982 to mid-1987, about the same implementation period as originally predicted for the Third Highway Project. This period was comparable to the profile of disbursements under the Tunisia First Highway Project, but about two years less than the historical profile for projects in Tunisia generally. IV. PROJECr DESIGN AND ORGANIZATION 4.01 The Project was idendfied during a mission to Tunisia in November 1978 and was prpared by the DGPC with assistance from French and Tunisian consultants financed from the Tunisia Third Highway Project (Loan 1601-TUN). The consutants prepared a 17 volume report (1980 study) on 2,714 km of rural roads. The roads were chosen for study jointly by representatives of the Ministries of Public Works, Agriculure and Interior from among 18,000 km of rurl roads adned by DGPC. Some of these also were part of the system of 17,700 km of (paved and unpaved) classified roads. The list was further refined to 212 road segments after consultations between the consultants and the Tunisian authonties. Engineering, traffic and agronomic data for each road were analyzed and economic rates of return (ERR's) were calculated using a model called MAGON which was a simplified version of the Bank's Rural Road Model. Appraisal took place in June 1981 after two and one half years of preparation. 4.02 The Project was to be a continuation of the rural roads lending in Tunisia supported by the Third Highway Project (Loan 1601-TUN) which was under way in 1978 and was not completed until 1986. (The PPAR for the Third Project, Report No. 7182, was issued on April 11, 1988). From the beginning, the Fifth Project, similar to the Third one, was to  -7- include financing and other activities in support of agriculte in the roads' zones of influence. However, doubts were raised wthin the Bank whether the agricultural components of the Third Project were being properly executed and whether similar components should be part of the Fifth Project. The Government of Tunisia was opposed to discontinuance by the Bank of the integrated approach in its rural roads lending to Tunisia because this would be contrary to the Govenment's approach in its overall nrral development program. The Goveement also was not !nterested in using the Bank's 'new approach" to rural roads lending. Although the Bank detemined that agricultural support activities in the Tlird Project were not as deficient as had appeared initially, the decision was made to exclude aU agricultural credit from Bank's financing under the proposed Fifth Project though such credit would continue to be part of the Project. 4.03 As Tunisia's r roads program was very large and could not be financed by the Bak alone, co-financing was envisioned for the Fifth Project from the start. However, the Government refused to invite any co-financing from commercial banks on the grounds that this method of financing should be aeserved for projects of a directly commercial chaacter. Nor could co-financing be raised from any other source. Accordingly, the program financed by the Bank was scaled back from 5 to 3 years. Road maintenance equipment in support of keeping the roads maintained was originally included for financing under the Fifth Project, but when co- financing became unavailable some unused funds under the Fourth Project were reallocated for this purpose. 4.04 Negotiations took place early in 1982 followed by loan signaue in May. Effectiveness, which had been foreseen for September 1982 did not occur until Aprfl 1983; this was due to delays in issuance of the decree establishing the govemmental committee to coordinate the various actvities under the Project. This was, in fact, the same committee as the corresponding committee under the Third Highway Project but had to have its jurisdiction expanded to cover the new project. The planned project timetable and the planned cumulative disbursements under the loan are shown in Table D: Project Timetable and Table E: Projected and Actual Cumulative Loan Disbursements rectively. 4.05 Project costs as originally estimated are shown in Table F: Project Costs, and correspond to capital costs of the improvement works on 1,200 km of roads and related agricultural support investments over the six years 1982-1987. Of the estimated DT 53 million, roughly 20% represented transfer taxes. The planned pattern of Project financing is shown in Table G: Project Fiancing. Recurrent budget requirements in support of the Project were esdmated to equal about DT 2 million per year by the sixth year of implementation. 4.06 The bulk of civil works procurement under the loan was to be executed under the Bank's normal ICB rules. Rural road improvement works would be aggregated into lots valued at a minimum of DT 500,000 each and lots would in tun be bid in groups of at least DT 3 million, with bidders having the option of bidding on one or more lots in each group. CTVs and spot improvements would be procured under local competitive bidding. Equipment procurement would be mostly by locally advertised bidding or, in the case of vehicle inspection equipment, after negotiation with suppliers of equipment ihat would be compatible with wbat was already in use in Tunisia.  -8- 4.07 The specific items to be included under the Project were spelled out only in part. Annex 2 to Schedule 2 of the Loan Agreement ()A) showed in the lrst year of the program, improvement (upgrading or construction) of 400 kn in 33 listed sections of rural roads located in eight governorates. An additional 800 km of about 70 sections were to be agreed between the Bank and Government from time to time for the second and third years of the program on the basis of feasibility studies and detailed engineering. The 55 extension centers and related equipment to be included under the Project were not identified, but it was noted that the extension centers would serve the agricultural areas around the roads to be improved under the Project, nor was there any indication as to the location of spot improvements. Some of this information would have emerged from submissions under a covenant (LA, Section 3-03 (n)) which required the Borrower to submit annually a plan of activities to be carried out in the following year. The nature of vehicle inspection equipment for five inspection stations also was not specified. 4.08 Loan covenants covered the usual requirements related to Project execution (fabin- H: Status of Covenants). 4.09 Project design and organization were amended on several occasions: (a) In June 1985, the reimbursable percentage of expenditures for rural road works was raised from 40 to 52%. After the loan was made, the value of the DT fell significantly in relation to the US dollar and with that change the orginal reimbursable percentage of 40% would only have covered part of the foreign exchange costs. (b) In December 1986, the LA was amended to establish procedures for the use of loan withdrawals on the basis of statements of expendiures. This procedure was intended to apply to contracts valued under $100, 000 equivalent for agicultural extension buildings and spot road improvements. The amendment was supposed to facihitate disbursements and overcome a serious lag in utilization of the loan. (c) In January 1989, the provision of vehicle inspection equipment was reduced from the original five stations to two at Sfax and Tunis. This change resulted from a decision by the Tunisian authorities to concentrate vehicle inspection activity at the two most heavily used stations. However, in the end the equipment financed by the Bank under the Project was used at six stations and, while the Bank gave its concurrence to this, the LA was not formally amended. (d) In March 1989, the procurement requirements of Schedule 4 of the LA were amended to allow local competitive bidding on rural road works financed under the loan, provided that no preference would apply, that bid and performance guarantees would be equal for every bidder, and that bids would be opened in public. At this point all of the original rural road  -9 - works under the Project had been completed and this amendment was specifically designed to allow financing of pavement on eight previously financed (under the Third or Fifth Projects) road sections where traffic had increased beyond expectations. Contracts would not exceed $400,000 equivalent individually or $2,000,000 all together. The original requirement for ICB thus was not considered necessary. V. PROJECr IPLEMENTATION Timing of Implementation d Extent of Completion 5.01 The Project, which was scheduled for completion at the end of 1986 (with loan closing a year later), actually was finished five years behind schedule at the end of 1991 (with loan closing 3 months later). However, delays varied among individual components, with rural road works, the major part of the Project in terms of money, being minimally delayed while other elements were delayed up to five years (Implementation Schedules I and DE and Table Lo Project Expendkures by Year). Similarly, completed rural road works met the target of the programmed 1,200 kn while other Project elements fell short of targets in various degrees, as shown in Table F: Project Costs. 5.02 All of the rural road works were tendered by September 1985. By September 1983, all of the first group of 400 km (the first group) was completed, and by December 1984 al of the second group of 400 km (the second group). Most of the third group of 400 kn (the third group) was completed by early 1986 as scheduled but two roads were delayed to 1987 and 1988 respectively due to the failure of the original contractor. Table J: Rural Roads Inprovement Contracts summarizes the information on contracts and contractors in the three groups of rural road improvements. Later, additional rural road works were included to pave eight sections improved under this Project and the Third Project. 5.03 (a) Implementation of the main physical component for agriculture, the construction of CTVs, went safisfactorily after some delays at the start. By late 1986 less than half of the CTVs had been built, by 1988 this number had risen to only 43 and the projected number of 55 CIVs was not reached until 1992 (Table K: Progress in Consticton of CTVs). Another 41 CTVs were built during this period financed soldy from the Government budget, which were also to serve the farmers in the roads' zones of influence. In addition, there were 13 CTVs, which were in rented buildings and were also linked with these rural roads. Therefore, there are 109 CTVs among the 400 Centres de Munwon Agricole, most of them built in the last 15 years, which serve the rural roads under the Project. Training programs for extension agents began in 1983 and were concluded in 1984 and 1985, even though expenditures extended through 1988. Equipment procurement for most CTVs was mostly completely by 1984 and the procurement of motor vehicles, though mostly finished by 1987, continued through 1990 but was never completed (Table L: Construction of CTVs and Table M: Expenditures for  - 10- CTVs, CTV Equipment and Vehicles and Extension Iaining and Studies). Similarly, spot improvements on feeder roads wee only partially executed (Table C: Spot Inprovements on Related Feeder Roads) and no report is available ngarding completion of rural infrastructure (which was part of the Project but not to be financed under the loan) other than an indication of a global expenditure of $3.4 million equivalent. (b) The component related to vebicle inspection equipment suffered the most serious diversion from the schedule; it was subject to repeated redesigns and was finally not completed until 1992, covering six stations instead of the original five (later amended to two) stations (Table N: Equipment for Vehicle Inspection Stations). (c) Technical assistance and fellowships (overseas missions) are summarized in Table 0: Expenditures under Project for Techal Asstance (including Traing and Overseas Misions). There were seven consultancies, six with the Ministry of Public Works and one with the Ministry of Agnculture, which includes the extension agents training contracts with the Ministry of Agriculture and the overseas missions. The study of constructing techniques of unpaved roads was delayed by nearly two years, but the impact study was delivered only a few months beyond target. The study to prepare a follow on project of rural road improvements (Part H rf) and riii) of the Description of the Project in the LA) was delayed due to problems in procurement and eventually was transferred for financg under the ongoing Sixth Highway Project. (d) The reasons for delays of various components was due mainly to failure of coordination among participants in the Project. 5.04 As a result of the partal execution of the Project (i.e. the purchase of vehicles and use of technical assistance) and as a result of the change in the exchange rate of the doLar in rlation to the Tunisian dinar (from $1 = DT 0.46 at appraisal to $1 =DT 0.93 currently), about $4 million of the loan was not used and this resulted in cancellation of $3 million in April 1990 and $1 million at loan closure. Coordbiat Ip aon of Rura Road Improents and of Supporting Project Elements. 5.05 More than 70% of the Project's baseline costs were for rural road construction and associated consulting services. Mos. of the remainder was for elements which were intended primarily to help stimulate agricultural production in the roads' zones of influence. To ensure that the supportng elements were properly camred out, the SAR included vanous bar charts, budget projections and action plans. The Project was to be executed in three groups, beginning in the first, second and third years of the Project, each extending over three years and each consising of one year of rural road design and two years of rurl road constuction. During each three year peiod parallel programs of rura road maintenance, spot road improvements and  - 11 - agricultural promotion should have taken place; how--ver, many of these activities were shown to continue for several years beyond the three years associated with each group. The budgetary consequences of all these activities were originally incorporated into capital and recurrent budget requireirents for the Project as a whole for each of the six years of Project execution (Tables 2(a) and 2(b) of Annex 8 of the SAR). The capital budget requirements became the "overall program" (Secdion 3.03 (b) of the LA and Suipplemental Letter No. 1). 5.06 Detailed implkmentation of the Project was to be carried out under various action plans and annual programs. These included: (a) An action plan for extension work for each iiidwiidual road, based on an agricultural survey made by the extension agents of all farmers served by the road and specifying 'key farm parameters which affect absorptive capacity for new techniques and the ability to qualify for new credit". The action plan would include road by road, extension guidelines, yield targets and credit needs, as well as a framework for project monitoring and evaluation (Parts C and D of Annex 6 of SAR). (b) An action plan for each governorate covering all rural road improvements which the Project would finance in that governorate and related support activities under the Project (spot road improvements, CTV buildings and equipment, minor irrigation works, and agricultural credit). (c) Annual programs consistent with the overall program (Supplemental Letter No. 1 contained action plans for the eight governorates included in the first group) which would include the preparation and updating of road feasibility studies and the preparation of integrated action plans (Part B of Annex 6 of SAR). The foregoing provisions of the SAR were incorporated in Secton 3.03 of the LA, which called for: in subsection (a), execution of the Project in accordance with the agreed overall action program; in subsection (b), the Borrower, through its Project Coordinating Committee, to submit annually, for Bank comment, a plan of project operations consistent with the overall action program; and in subsection (c), submission to the Bank of feasibility reports for rral road improvements in the second and third groups, and extension guidelines, farm budget analyses and implementation data for extension to be carried out in each year of the Project. 5.07 The Bank's file does not disclose how the foregoing provisions were implemented. (a) While all rural roads in the second and third groups appear to have been included in the 1980 study, the file contains no reference to their formal inclusion under the Project or to updating of the feasibility conclusions of the 1980 study. (b) While a Tunisian firm was engaged about the beginning of 1986 to make surveys of agricultural potential and prepare agriculturai development programs for Project roads in 13 governorates, the results of its work are not on file.  - 12 - According to the MA, the firm prepared a report for each road in the second and third groups and by the end of 1987 had completed nearly 70 reports which were mai3tained only at the Ministry. Neither does the file contain action plans for te govemorates responsible for the sections of rural roads in groups two and three. The file does contain copies of budget requests made by the Direction de la Production VEgeta1e (DPV), or Crop Production Department, for the years 1984, 1985, 1987 and 1989, which provide justification for various expenditures by DPV nationwide on construction and equipping of CTVs, training of extension agents and operating costs (such as motor fuel) in support of the Fifth Highway Project.- According to the MA the two parties agreed to substitute these national budgets for the action plans for each governorate referred to in para. 5.06 (b) above. Nor does the file indicate whether seminars were held in each of the governorates to fmiliarize local officials and extension agents with the Project, to brief them on the road - by - road action plans or to explain the farm surveys extension agents would have to make (Part D of Annex 6 of SAR). Accordng to the MA, the surveys by extension agents referred to in para. 5.06 (a) above were not made because they were inappropnate to the educational level of the agents and would require contacts with large numbers of farmers thus diverting the agents from other essential work. Finally, there is no record of formal submission to the Bank of annual programs or action plans for individual roads. In fact, it was reported in early 1987, that the Project Coordinating Committee had not met for over a year. 5.08 The abser^,e of documentation on action plans and annual programs suggests that the supporting elements were not implemented in accordance with the movell progam" and that monitoring and evaluation could not be done on a systematic basis. Some information was collected on credits and extension provided in a sample of the zones of influences. ITnis information permits us to estimate that the credits given in those zones of influences were in the order of DT 4 million and were adequate. Civil Works in General 5.09 Detailed designs and contract documents for the first group of mral roads improvements were prepared under a consultancy financed from the Tunisia Third Highways Project. Similar documentation for the second and third groups of rural roads improvements were made available under further contracts with the same consultants financed from the current loan. All these works were executed by DGPC after international tendering. Spot improvements of feeder roads rumnning into mral roads improved under the Project were also to be executed by DGPC according to procedures and criteria to be developed by consultants financed from the loan. However, only a small part of the projected spot improvements was carried out and DGPC never engaged consultants under the Project to develop special procedures. Therefore, DGPC executed the works under its normal procedures. Supervision of rural road works was to be done by DGPC with the assisnce of consultants while the arrangements for supervision of spot improvemets was to be specified later and, therefore, no special procedures were used. In general, aU work was supervised by the Regional Directorates  - 13- of DGPC under the general responsibility of DGPC's Central management. Rural road improvement works were almost aU awarded to local firms with the exception of a bidder from Hungary. Spot improvemenits were executed by small local firms experienced in masonry and concrete work. 5.10 Rural infrastructure works (such as minor irrigation) as well as the design and feasibility studies for those works, were to be executed by the G6nie Rural (GR) of the MA. While financing out of the Bank loan was not available for these works, GR was to submit the details of each scheme for the Bank's review before implementation (SAR, Annex 6, para. 9). However, it appears in the files that GR did not submit any such designs or studies and specific infrastructure works that could be attrbuted to the Project were reported only in an aggregate amount. 5.11 The construction of CTVs was to be the responsibility of the MA. No special provisions for the design and supervision of the CTVs were included in the Project documents and procurement was to be carried out under local competitive tendering procedures. The bidders were provided with building plans and bills of quantities. 5.12 The design of vehicle inspection stations for which equipment was to be procured under the Project presented major difficulties as it depended on the developraent of a road safety program. An initial report on this program financed under Tunisia Fourth Highway Project was submitted in early 1984. A final decision by the Government whereby the inspection statons became the responsibility of the governorates was not made until early 1990 at which time another two years would have been needed to design the stations. Instead it was decided that equipment under the Project would be procured for four (later reduced to two and finally increased to six) existng stations to replace units which had become worn out. Rural Road Improvements and Mainte 5.13 The rural road improvements were largely carried out on the roads listed in Annex H to Schedule 2 of the LA (the initial 400 km) and on other roads tentatively identified at the time of appraisal (the remaining 800 kin). However, some roads, on the basis of traffic, were further improved after the initial work prinarily by widening them or by providing them with bituminous surfaces and a part of these additional works were financed under the loan. The supervision of the rural road improvement works was done by the regional and central administaions of DGPC in accordance with their normal procedures and without the assistance of consulants. 5.14 (a) Part D of the Description of the Project provided for the implementation of a maintenance program of 'all improved' rural roads. Section 4.04of the L staes that the Borrower shall cause all the 'Rurl Roads' (that is, roads improved under the Project) as well as other rural roads offering an appropriate level of service to be maintained and to provide the necessary budgetary and other resources to that end. Table 2 ) ofAnnex 8 of the S4R indicates a projected recurrent road maintenance budget for 10,000 km rising from DT 187,000 in the first year to  - 14 - DT 3,750,000 in the last year. The exact coverage of roads to be achieved by this budget is not indicated. (b) Table P: Mlaintenance Expenditures for Rural Roads under the Project, other Improved Rural Roads and All Roads, 1983-1992, shows actual performance of various road maintenance budgets in Tunisia in the period 1983 - 1992. This indicates that all road maintenance expenditures have roughly doubled over the decade in nominal terms and gone up by about 50% in real terms. The expenditures for rural roads have gone up much more steeply (by a factor of nearly eight) and, for rural roads under this Project, to DT 1.1 million in 1989 and to an average of nearly DT 2.4 million in *:990-92, or about DT 2000 per km of such roads. However, it appears that the expenditures for rural road maintenance include periodic maintenance and minor improvements; separate data is not available to separate routine from periodic maintenance on many roads. While the Bank's supervision reports do not indicate which roads were visited during supervision missions, a mission in late 1988 which inspected rural roads in four governorates reported that the roads they saw seemed to have had no routine maintenance. More information on routine maintenance of rural roads may emerge from a study currently under way. Meanwhile, it will be difficult to compare budget performance for rural road maintenance, particularly routine maintenance, with targets at appraisal. Spot Improvements on Feeder Roads Serving Improved Rural Roads 5.15 DGPC had responsibility for selecting and executing minor improvements on roads feeding into the rural roads being improved under the Project. Over 90% of these expenditures occurred in the years 1984 through 1987 and nearly all were for culverts and small bridges. Of roughly 20 spot improvements, about one half are associated with rural roads improved under the first group, about one quarter with roads under the second group and the rest with roads not part of the Fifth Highway Project. The total value of these works is about DT 0.9 million, roughly 60% of the amount foreseen at appraisal. No information is available as to the reasons for the final content (location and amount of the works) of this Project component. Minor Irrigation Works 5.16 This element was to include small scale irrigation, irrigation dams and land drainage works. CTV Buildings 5.17 The phasing of CTV construction is shown in Tables K and L and the location of these works and the rural roads to which they relate is shown in these tables and in the map. The most active period of construction was in the five years 1984-1988 when an average of eight units per year was built annually. Four govemorates had only one unit built while three  - 15 - govemorates accounted for one third of the units. The general design which was somewhat adjusted in each region provided for six rooms on one floor, including a meeting room and a kdtchen, with a total area of about 150 n2. The construction, equipping and staffing of CTVs apparently was not done in coordination, as equipment purchases and training preceded the construction of many CTVs. Consequently, extension agents were in their posts and had to depend on rental accommodations. CTV Equipment 5.18 This element represented le.s than 10% of the cost of the buildings. It included: (i) demonstration and testing equipment for extension agents and (ii) office fumiture and equipment. Nearly all of this equipment was procured in 1983 and 1984, before any of the CTVs were built. 5.19 Though veiicks were only one of the items of equipment for use by extension agents, this category was treated separately in Project cost tables and in the LA's List of Goods. The Project envisioned roughly one vehicle per CTV which represented about 20% of the cost of CTVs. About one third of the vehicles were procured two years before any CTVs were built and over 90% were delivered by 1986 when only 28 CTVs had been completed. Agricultural Credit 5.20 Short, medium and long-term credit was to be provided, without resort to the Bank loan, to farmers in the roads' zones of influence, in amounts up to nearly DT 7 million equivalent, and the Borrower agreed that BNT would provide the necessary credit given that it might not be available otherwise. The Thrd Highway Project had had difficulty in mobilizing agricultural credit for farmers, but this difficulty was not expected to continue through the Fifth Highway Project. In addition to BNT financing, capital requirements were to be met by the farmers themselves and the Government (through a subsidy) for some medium and long term financing. Total BNT credit commitments over six years were to be nearly DT 4.5 million and farmer and Government finance was to total about DT 2.5 million over the period. No systematic evaluation of credit needs was made as the detailed annual programs described in para. 3.09 of the SAR and referred to in Section 3.03 (c) of the LA were not prepared nor their related farm surveys or representative farm models. As a result, data was not generated on the credits actually utilized in the Fifth Highway Project. However, a sample survey has been msvz for several governorates on the basis of which it was estimated that BNT provided credits to farmers in the zones of influence in the amount of DT 4 million over the period 1986 to 1992. While one may conclude that BNT credits were made available in the area which was sampled, the sample is too constricted to judge whether credit was adequate in all zones of influence. Moreover, no data is available on the financing (by the farmers and the Government) and the adequacy of the overall financing provided can, therefore, not be determined.  - 16- Consultg Servces - General 5.21 Consultant assistance was to phy a crucial role in implementation of the Project. It was to support the process of rural road design (including economic feasibility determination) and construction, execution of various other programs, and monitoring and evaluation of project results. In addition, corsultants were to assist in preparation of two follow-up projects both for rural roads and for rehabilitation of classified roads. The DGPC and various units of the MA were to be the principal beneficiaries of the consultants' work. Consulting Services - DGPC: Design and Supervision of Runl Road Improvements 5.22 A contract was signed in April 1981 for the design of 410 kn of roads in eight governorates intended for construction under the Project. This contract was with the same group of French and Tunisian firms who had done the feasibility studies and was financed from the Third Highway Project loan. Extensions of this contract were signed with the same consultants for the design (including deterination of economic feasibility) of the remauiing 800 km of rural road improvements, even though the Technical Assistance Implementation Schedule (Chart 5 of the SAR envisioned two separate consultancies for the second and the third groups. Consulting Services - DGPC: Design of Spot Inprovements on Feeder Roads Serving Improved Rural Roads 5.23 The work concerned primarily the removal of difficulties created by water at low points in a feeder road's alignment. The consultancy was to provide guidelines for DGPC regional engineers in developing programs of spot improvements. No terms of reference were ever drafted for this consultancy (though outline tems of reference were mentioned in the SAR) and no one was ever appointed to it. The file does not indicate the reason for this omission or whether it was ever discussed but the reduced size of this part of the Project may have reduced the need for the conultancy. Consulting Services - DGPC: Auditing and Monitoring and Inpact Evaluation 5.24 This consultancy was not implemented as originally foreseen. No specific consultancy related to auditing and monitoring was described in the SAR although an outline terms of reference was mentioned in the SAR. While auditing and monitorng was refeTred to in the Technical Assistance Implementation Schedule (Chart 5 of the SAR) and in Part H (1) (7) of the Descnption of the Project, it was not developed or put into effect. The impact evaluation was to utilize the traffic and agricultural data collected in Project monitoring and, in addition, consultants would do surveys on a sample of roads from the Third and Fifth Highway Projects. These surveys would address the full range of nual development impacts, including household income, consumption, personal mobility and communication, provision and utilization of services and migration patterns, in addition to traffic and agricultural data. An initial Impact Stt dy report was made by some of the same group of French and Tunisian consultants who had done the feasibility studies, but on the basis of one survey in 1986 concerned primarily with traffic counts and agricultural production, not a series of surveys. But the Impact Study made use of  - 17- other data as well, such as the 1984 Census. Currently another study is under way by the same group of consultants to prepare a further rural roads p;,ject for Tunisia (the Seventhff Highway Project) and this includes a esurvey of the Fifth Highway Project roads covered in 1986 plus seven other roads from that Project. As in 1986, the current study is limited in scope and primarily concerned with traffic and agriculture. Some of the findings of the 1986 survey and some preJmiary results of the current study are discussed in Chapter VI below. Consling Services - DGPC: Study of Techniques for Unpaved Roads In Tunsa 5.25 This study was executed by two of the French and Tunisian consultants who were part of the consortium that had done the feasibility studies for the Project. The study made detailed laboratory analyses on a sample of 26 unpaved roads and reviewed the geotechnical methods used in their construction. The study concluded that there were indeed deficiencies in the construction of the roads (e.g. insufficient fines in the aggregate) and that the most important problem of unpaved roads in Tunisia was the implicit belief that only a paved surface made a real "road'. The lack of adequate maintenance was identified as a problem of equal imporance, pardy related to the derogatory view of unpaved roads. The study recommended observation of tighter technical standards for construction of rural roads and rationalizaon of maintenance by concentrating authonty in the govemors and by setting up dedicated budgets, personnel and equipment for maintenance of unpaved roads. The Bank's file does not indicate how the study's recommendations were followed up. Consulting Services - DGPC: Preparatory Study for Next Phase of Rural Road Improvement 5.26 A Bank mission in late 1988 prepared draft terms of reference for this study for consideraion by the Govemment; this was originally scheduled for the latter part of 1982, and by the end of 1989, proposals had been received and evaluated. A British consultant was given the highest evaluation to which the Bank agreed. However, the Central Tendering Board decided on cost grounds to give the contract to another bidder. The Bank disagreed, refused financing under the loan and proposed a rebid (para. 9.03). A high level Bankl mission, a year later, which was to resolve a number of procurement issues in Tunisia resulted in an agreement to rebid this work on a modified basis and to finance it under the Sixth Highway loan because insufficient time remained before the Fifth Highway loan would be closed. In addition to this principal study for additional rural road improvements, two smaller studies covering about 600 km of rural roads for inclusion in the next rural roads project (the proposed Seventh Highway Project) had been contacted in 1984 and were financed under the Fifth Highway Project. Consulting Services - DGPC: Preparation of a Program of Road Rehabilitation 5.27 The study of a program to rehabilitate classified roads was financed under the Fourth Highway Project and the Fifth only financed the preparation of bidding documents for 312 km of road rehabilitation works.  - 18 - Consulting Services - MA: Preparation of DetaUed Extension Programs and Credit Need Projections 5.28 Though this work is described under Agricultural Extension in the Description of the Project Part B of the LA, it is not included under Part H, Technical Assistance and Fellowships. However, the Technical Assistance Implementation Schedule envisioned two separate consultancies for the second and third groups that were to be executed in 1982 and 1983; the record does not indicate how the first group was to be covered. In any event, one consultancy was arranged in late 1985 with a Tunisian firm (one of the group which had done the feasibility studies) to study the zones of influence of roads in all three groups. Most of the consultant's reports, which resulted from this work, are only available in the files of the MA. It also seems that the 'systematic agricultural surveys of aU farmers served by the project roads" (para. 3 of Annex 6 of SAR), which were to serve as baseline data for preparation of action plans, were not carried out. The consultants thus did not have some of the material that was to underlie their work. Consulting Services - MA: Training of Extension Agents 5.29 This activity was scheduled to occur over four years (1983-1986). It was implemented through a contract made with a Tunisian individual in June 1983. The contract covered 25 one week courses (five series of five one week courses) in communication techniques for 87 extension agents extending over 14 months. The quality of this training was reported to have been very effective. Fellowships 5.30 Part H (2) of the Description of the Project of the LA provided for short training missions overseas for staff of DGPC and MA. The file indicates that three such missions were arranged through French entities for visits to France for 25 agricultual engineers of the Direction de la Production Vegetale (DPV). No other activity of this type was carried out under the Project. VI. MAJOR PROJECT RESULTS Achievement of Project Objectives 6.01 The Project objectives as stated in para. 3.02 of the SAR referred to the specific rural roads to be covered by the Project and their zones of influence, to the organization responsible for maintenance of these rural roads, to the institutional framework and process for improving Tunisia's rural roads and to Tunisia's highway safety program. About 1,200 km of rural roads were to be improved to "minimum all weather" standard, and agricultural production and living standards, including access to social services, in the roads' zones of influence were to be increased. About 1,186 km of construction have been completed. Whether the levels of service currently provided by the roads remain at the standard achieved after  - 19- improvement is uncetain, but the lack of a systematic approach to routine and periodic maintenance suggests that this is unlikely. On the other hand, many roads received additional improvements such as pavement strengthening (widenng and/or overlaying), or a double bituminous surface treatment. These additions were made primarily on the basis of traffic and are shown in Table B. While some of these improvements were made under the Project, others were not. Accordingly, there is no comprehensive record of the capital expenditures that were necessary to bring each road to its present condition. Similarly, there is no overal accounting of what occurred in the roads' zones of influence. Some impact studies were carried out on selected roads in 1986 (published in 1987) and others are curren*ly under way (Table Q-1: List of Roads in lmpact Surveys). The data on the roads of the 1987 study are summarzed in Table Q-2: Slected Data cn the Sample of Roads Before the Project (1988, 1979 or 1980) and After the Project (1985 or 1986). Additional data from the impact study currently under way are smmanzed in Table Q-3: Selected Data on a Sample of Roads Before the Project (1978, 1979 or 1980) and After the Project (1992). The 1987 study suggests that agricultural production rose in aU areas but the increase was very modest in 25% of the sample. However, as the period studied was from several years before the improvement, which obviously is too long, to only one or two years hrefter, which is too short, the fuU impact on improving agricultural output could not be captured. 6.02 The Project included a study of rurl road construction and maintenance techniques for Tunisia, which made concrete recommendations for improvement of rural roads maintenance organizaions. The file does not indicate any specific follow-up on this study. Beyond this study and a covenant to maintain 'improved' rural roads (Section 4.04 of LA), the Project contributed little to help improve rual road maintenance. The SAR did set out cetain budgetary targets for road maintenance on a 10,000 km road network, and maintenance budgets did increase. However, it is difficult to establish how much was actually spent on nural road maintenance, in particular routine maintenance, especialy since maintenance funds are fungible between categories of roads. Moreover, the Project did not include any rvision of the structure of the maintenance system which remained essentially unchanged and many works continued to be more in the nature of emergency repairs rather than scheduled interventions. 6.03 Thc Project was based on an elaborate procedure designed to integrate improvements to rural roads with their related agriculture. It began with a selection process including feasibility studies, proceeded to design studies and concluded with the procurement of the construction of all roads in a particular governorate together in a single contract Concurrendy, the Project envisioned a system for stimuating increased agricultural production through ancillary public works, intensified extension, and guaranteed access to credit. The whole procedure was to be subject to comprehensive monitoring and evaluation. Some of the procedure was a repetition of what was done in an earlier Bank project (preliminar studies and road construction). However, the part of the procedure related to agricultural support was only pardy implemented and cannot be consdered to have been firmly insttutonalized. The special support provided by several Tunisian officials in the Third Highway Project did not continue when they were transferred to other duties. The Project thus only pardy met its objective of improving the institutional framework for rural roads.  - 20- 6.04 The Project was to contribute to a comprehensive road safety program that was being developed under the Fourth Highway Project. The Project did result in a supervision mission visit to Tunisia of a Bank highway safety expert in 1984 who formulated a pilot highway safety program for consideration by the Government. The file does not indicate what the expert recommended nor the impact of his work. However, the Government's vehicle testing program has continued to expand gradually so that by the end of the eighth plan (1996), there will be 23 stations in operation. The loan provided financing of equipment for six stations whose existing stock had become worn out. The acquisition of vehicle testing equipment did help to evidence the Bank's continued concern about road safety in Tunisia but the discussions related thereto seemed to have had limited impact on the Govemment's policy or programs regarding the issue. Moreover, more recent developments indicate that other types of interventions may have a greater impact on safety than vehicle testing. Adhievement of Physical Targets 6.05 Physical targets of the most important elements of the Project were mostly met or even exceeded. Rural road improvements, while slightly under the 1,200 km target, in many cases involved more extensive and costlier works thanorginally foreseen. The number of CTVs built also matched the 55 units planned. Several other Project elements, which had not been specified in physical terms but only in monetary amounts, had significant shortfalls in monetary terms; namely, spot improvements to feeder roads, purchases of vehicles, and minor irrigation works. The impact of these short falls on the Project cannot be determined. However, for vehicle inspection equipment, expenditure was nearly twice the target. Technical assistance expenditure also was below planned fnancial targets (one half if measured in DT or two thirds if measured in US$), indicating that much of it simply was not carried out. The consultancies on spot improvements, on auditing and monitoring were not implemented at all. The consultancies on the next phase of rural roads improvement and on road rehabilitation were financed under the Sixth and Fourth Highway Projects respectively. The consultancies on impact evaluation and for detailed agricultural extension programs were carried out only in part. The implementation of consultancies would perhaps have been greater if detailed (not merely outlined) terms of reference had been prepared for them at appraisal. However, the mair. reason for the non-execution or partial execution of consultancies was that the Governmtent did not view them as necessary or helpful and the Bank acquiesced in the Government's not pursuing these components of the Project. The record on Pioject implementation provides evidence that Project components that had been planned in detail at appraisal, particularly to the point of physical specification, were much more likely to be camed out than those that were subject to a general description only. Project Costs 6.06 Project accounts were kept in DT, the currency in which procurement overwhelmingly took place. Accordingly, Table F shows cumulative costs in current dinars converted into US$ equivalent at the average rate of exchange over the course of Project implementation. In the absence of complete data on agricultural credits, Project costs are shown without an estimate of such costs ($64 million equivalent). The corresponding totals at appraisal  - 21 - were $105 (with agricutural credits) and $91 (without) million equivalent. If costs are shown in this manner, the Project is seen to have cost substantially less than projected, primaily because rura road works, which were about 70% of the costs, were incurred in dinars, which lost value in terms of US dollars over the past decade. In addition, the fact that several components of the Project were not fully implemented also contributed to reduce costs. Comprehensive information on the foreign exchange content of the Project were not available as the authorities were not able to quandfy them. 6.07 While project costs were incurred over 10 years, other costs were concentated between 1984 to 1986; they fell precipitously in 1987 due to the financial problems of two contractors. If the expenditures for rural road works are expressed in constant 1982 dinars (using the consumer price index through early 1992), they would amount to DT 32.7 million, about 11% over the appraisal estimate of DT 29.4 million, including the 10% physical contingency. A similar comparison of CIV construction costs indicates that they cost only 3% more than projected, DT 1.23 million, as against the appraisal estimate of DT 1.20 million. These buildings were substantially executed as designed. For other components of the Project, comparison of the overall projected and final costs is not meaningful because these components were not carried out as projected. 6.08 Large numbers of similar units of work were procured under both rural road improvements and CTV construction. The average cost per km of improvements is roughly DT 30,000 (m 1982 Dinars) or DT 60,000 (in 1992 Dinars) or $65,200 and $64,500 respectively. An examination of per kn improvement costs of individual roads shows considerable variations. These variations reflect differences in the nature of the improvements, construction conditions and the value of the dinar and certain construction materials. The average cost of a CrV was DT 16,600 (in 1982 Dinars) or DT 33,000 (in 1992 Dinars) or $36,100 and $35,500 respectively. If the costs of CTVs built at roughly the same time are compared, they are very similar. Also, CTVs built in the early years cost close to the indicated estimates while discrepancies between projected and realized costs occurred in later years, which suggests that the original esfimates were used unchanged throughout. 6.09 Table E compares actual with projected disbursements. The relative delays in physical execution are mirored in the rates of disbursement, with road improvement disbursements reaching about 70% of their ultimate projected level (including distributed contingencies) by the beginning of 1988 while all other categories trailed behind. In the end, road improvements were within 3% of projections, while all other elements fell significantly below that level with the exception of fellowships, which were as foreseen, and vehicle inspeion equipment, which absorbed roughly double the amount of loan funds originally allocated (Table R: Evolution of Disbursements by Category). The actual rate of disbursement exceeded by about two years the typical Tunisia disbursement profile (Table E). Economic Evaluation 6.10 The economic evaluation at appraisal concentrated on the roads and related agricultural activities, which were treatd together as one single economic activity. (rhe SAR  - 22 - had not quantfied the vehicle safety equipment component). The analysis sought to segregate traffic on the project roads into passenger, non-agrcultural commodity, and agriculturAl commodity traffic. Also, tfie analysis sought to establish two types of benefits: (i) traffic benefits based on improved vehicle operating costs for passenger and non-icultural commodity flows and (ii) agricultural benefits based on the net value of increments in agricultural production in the roads' zones of influence, which would include the traffic benefits for agricultural commodity traffic. These benefits were then compared with costs to obtain an ERR. For purposes of the analysis, the following costs were considered: - the rural road improvement construction costs (but not the cost of spot improvements on roads feeding into project roads, nor the cost of related studies, supervision and overseas visits); - cost of acquisition of CrV buildings, equipment and vehicles (but not the cost of related studies, traimng and overseas visits); - cost of on-farm investments, except seasonal inputs covered in farm budgets (but not the cost of minor rumal works such as irrigation); - cost of routine and periodic road maintenance; and - cost of farm operations. The SAR does not explain the rattonale behind the inclusion of some costs and the exclusion of others. All benefits and costs were projected for a period of 21 years. 6.11 The SAR economic evaluation was based on data from the 1980 study. A swrmmary of this data for the 33 roads in the first group is shown in Table 2 of Annex of the SAR. Similar summaries, also based on the 1980 study, for the second and third groups of roads were not found in the Bank's files. Table S: Rural Road Inprovements and Related Parameters for Economic Evaluation provides some of this base data as well as the corresponding information for the present, to the extent it is available. 6.12 The original evaluation had many complexities both in concept and execution. It sought to segregate agricultural impacts within a particular perimeter and to separate the prmary impacts of agricultl activities on tnort (through transport of inputs and products) from their secondary (income) effects. Most frequently road construction costs and traffic benefits were not lined to agricultural investment costs and incremental agricultural added value. Sometimes, traffic benefits caried the road investment and sometimes agricultural benefits provided the bulk of the benefits, though in most cases road investments substantially exceeded agricultural investments. While all these vaniations were within the range of possibility, they showed tat each case was sui gneris and that special plans would have to be prepared for each case, something which was never done.  - 23 - 6.13 The data requrements also were immense. Existng as well as future traffic projections of freight traffic needed to be segrated by load type excluding generated non- agrcultural freight taffic. Passenger traffic had to be projected, though again excluding generated traffic. Maintenance (routine and perodic) had to be established for each road and corresponding sevice levels would need to be estimated, including the costs of capacity expanding investments to maintain these service levels. Crpping patterns needed to be developed, costs (both on and off-farm), harvests and prices needed to be projected for individual crops, and investments (oth on and off-farm) needed to be determined. All these proections were required to be made over two decades. Even if the monitoring system had been established as intended, the replication of this data on an apost basis would have been very difficult; the absence of monitoring systems made the task impossible. Accordingly, it was agreed some years ago that the impact evaluations would be limited to samples, but this work has only been implemented recently in connection with preparation for the proposed Seventh Highway Project. Although some data has been collected, which provided the basis for Table S, further data and calculated results, with ERRs, were not available at the time the present report was finalized. 6.14 These experiences have raised the issue of whether the existing system for selection and evaluation of rural roads is the most suitable one that could be devised. It was developed dunng a period when the country had a substantially lower per capita income and was substantially less urban. In the Northern and Easten parts of the country where most of the people live and where most of the roads under the Project were built, rura life is becoming more closely linked to the city and the role of agriculture is changing, becoming for many a secondary occupation. This suggests that the planning of rural roads may need to be based more on traffic and broader social, regional, economic land planning and transport system considerations, and less on agricultural production, at least in those areas of the country where urban influences are strongly felt. 6.15 Table S shows that most roads under the Project had a significant traffic increase which should by itself be sufficient to justify the improvements made. However, it would be difficult to quantify this point in the absence of reliable data on the condition of the road before improvement, over the intervening years and today and on the amounts spent for the road over the past ten years. Also, tffic performance may reflect not so much conditions of the individual roads but the rise in the general prosperity of Tunisia and the accompanying increase in motorization. 6.16 Table S as well as Tables Q-1 and Q-2 give some indication on the evolution of agriculture in the roads' zones of influence. These suggest significant increases. However the aggregate increases are not of the order of magnitude indicated at appraisal (on a discounted basis, group one roads at appraisal showed agricultural benefits equal to 2.4 times traffic benefits). The causal link between the road improvements and the agricultural upswing is difficult to establish. The increase in production did not follow directly upon ihe improvement of the roads because of the tee year drought of 1986-1988. In the early years after the roads were improved, extension services were hampered by the fact that they were scattered among several administrative organizations but they have recently been combined in one agency, which  - 24 - opetes under unified programs. The CIVs are managed by agricultural engiees, assisted by agriculturlist and extension agents together with the chiefs of the suborinate extension units attached to the CIVs. The estabLishment of the CTVs and the improvement of the roads has led to improved technical servces to the farmers and more frequent extension contacts. The constuction of the CIVs permitted the improvement of extension services and of the adminitrative strucue of the agency which provided them. As rgards the prvision of agriculural credit and its contribution to the growth of agnculture in the zones of influence, it was impossible to obtain complete information on the credit activities of BNT, the Government and the farmers. On the basis of a sample study of the credits provided by BNT in some of the zones of influence, it can be inferred that the provision of credit by BNT was adequate. However, the drought has seriously impacted the operations of BNT because of the farmers' inability to repay their loans and the consequent unwillingness of BNT to provide them more credit. Envonmental Impat 6.17 The environmental impacts of the Project would encompass not only the roads but related agricultural acivities. The road works were carnied out largely on existing ahgnments and therefore, had limited impact on soil erosion, slope stability or water flows. The agricutral activities probably had more significant ecological repercussions, especially when there was expansion of irrigation perimeters and greater use of ferlize and pesticides with crop inmensification. VII. PROJECT SUSTAINABIL1TY 7.01 he issue of sus_ nability is aised in two ways in this Project: the sustainability of the prcedum for selection and execution of ruml road improvement works with the attendant support activities and the sustainability of benefits generated by the road improvements. The seon, design and construction procedure for road improvements have become formalized and repeated in two project cycles (the Third and Fifth Highway Projects). They require large amounts of consuting assistance and so far have been assoiated with large infusions of extnal finance. The willin and financial capacity of the Tunisian authorities to maintain these programs on their own remains to be demonstrated. Indeed, the analysis procedures and the model are very complex and may not be fully understood. The Government, however, wishes to continue these procedures and has not so far been interested in the Bank's simplified apprach. The sustinability of road improvement benefits dqpends on adequate road maintenance and estblishment of the necessary organization to achieve this. While maintenance budgets have been increasing, the organizational strcture for rural road maintenance remains unchanged and does not provide systematic maintenance services.  - 25 - VIII. BANK'S PERFORMANCE 8.01 The Bank made a major effort to resolve the difficulties of an inherently complex project. Missions and staff inputs which supported tis effort in substantial measures are mmard in Tables T: MiNions and U: Staff Inputs. The Project involved close collaboration of several administrative structures both on the side of the Goverment and the Bank. Prtect design sought to identify and address potential problems in detail (by the use of coordinating bodies, action plans, annual programs and other measures) and to assess risks realistically. Faced with the possibility that difficulties in the execution of the agricultural support activities of the Third Highway Project would apply to the next project as well, the Project designers wanted to abandon the integrated approach and concentrate o,. r% d improvements but were persuaded otherwise by the Government and by more optimistic reports firom some Bank staff. 8.02 However, in the execution of the Project the Bank did not commit resources commensurate to the complexities of Project design and particularly of its agricultural components. Out of 28 supevision missions, only six were by or included agricultural specalists. After an inital mission within six months of loan approval, intervals between visits by agricultural specialists became lengthier, with a two and a half year gap from early 1985 and no visits at all after early 1988. As a result, failures in monitoring and in follow up went unobserved and unremedied. 8.03 Faced with the limitaion of resources, the Bank appropriately concentrated on the Project's largest and most manageable component, which was completed close to its physical targets and its time table. The Project continued the contracting system of the Third Highway Project whereby works are grouped in lots covering several roads in a governorate but the lots were considerably higher in value approaching an average of $3 million each. The experience of this contactg system appears to have been administratively convenient by consolidating a number of smaller scattered works. This system may have application elsewhere and should be examined in detail. Even though the rual roads themselves became the primary focus of Bank concem, the Bank did not expend much effort to draw attention to the fact that maintenance of these roads was insufficient or to devise means to overcome the problem. 8.04 Beside the undue complexities of the Project (further exacerbated by the highway safety component which involved a third ministry), the Project had other design weakness: (a) The attempt to mix into a single project the construction of rural roads by one Ministry with the intensive development of agnculture in the adjoining areas by another Ministry presents inherent difficulties of administrative incongruities and programmiatc clashes. Both organizations are structured appropriately for efficient delivery of ffeir respective services and programs, but have difficulty separating out small sections of road and agricultural areas for special attention. Moreover, these sections may not correspond with the highest priorities in their respective road or agricultural systems. A method to combine efficiently rural roads and rural development is desirable. Such a combined approach was  - 26 - regrded by the Government as havig been effective in the Third Highway Project though only a smaU part of the parallel investments were actually made there (Project Completion Report, Third Highway Project, Report no. 7182 of April 11, 1988). (b) The Bank also should ensure that its projects are based on a sense of ownership by the relevant agencies of the Government so that the projects do not languish of effective bureaucratic support. Such support is reported to have been actively provided in the Third Highway Project but, with personnel changes, was not maintained in the Fifth Project. (c) the selection of the initial array of roads for feasibility study from the universe of Tunisia's rural roads was made on the basis of an administative process and without the availability of a road inventory. Under these circumstances priorities, interdependencies and linkages are difficult to establish and a road's economic functioning may be unclear. Several roads turned out te '-- near urban areas and an agricultural methodology may therefore have been inapropriate for them. (d) The role of rural roads in an increasingly wealthy and urban economy like Tunisia cannot be properly defined m terms of surrounding agricultural activities om;y. Increasingly, rural inhabitants have off-farm employment and resort to agriculture only as a secondary source of income or as a social safety net. The analysis of rual roads needs to account for this new dimension and needs to concern itself with such issues as traffic hierarchies, regional and land planning, installation of public udlities, provision of social services, the structure of transport system, the control of urban sprawl and the protection of ecosystems. The use of agriculture as the primary factor of analysis in rural road projects should be limited to cases where this is the appropriate context. (e) The Project sought to use agricultural extension and credit as the major instruments to promote agricultura development in the roads' zones of influence. However, Tunisia's extension system was reported to be seriously flawed. With various project and commodity organizations having introduced their own semi- autonomous extension systems for individual crops or regions, Tunisia was said (in December 1984) to have no effective national extension service. Extension agents were said to be preoccupied with administration and other non-extension work during up to 70% of teir time and farmers were being advised by seveal agents, depending on the subject. Demonstration plots were said to be too large and too few in number. The agents reportedly lacked transport and their link to research was weak. Information on credit needs and supply, which the agents were to use, was not being collected but, according to a 1987 report, lack of credit appeared not to be a constraint. This "classical type" of extension was thought by some observers to be less effective than the rural development model where the farm is studied as a whole, not by crop or problem. Bank staff thus had misgivings about the extension effort applied in this Project and was divided  - 27 - on what would be better. In this sense the Project provided a poor test of the hypothesis that extension could help in conjunction with a better road to stimulate agricultural output. IX. BORROWER'S PERFORMANCE 9.01 While the Borrower insisted that the Project be designed on an integrated road/agriculture basis, it did not pursue the agricultural components vigorously and from the begnning, failed to carry out the procedures set out therefore in the loan documents. Perhaps the absence of action plans and other planning documents made no difference to agriculual growth but without these actions the Project could hardly have served as a fair test for the k pothesis connecting increased agricutural production with deliberate promotional activities. 9.02 The Project design placed significant responsibility in the Project Coordination Committee (a simrla committee already was operating under the Third Highway Project) whose investiture with responsibilty for coordinating the Fifth Highway Project was a condition to the ffecidveness of the LA. However, the influence of this Committee on Project execution appears to have been limited pary because the Committee suspended its meetings for a consieable length of time. However, even when it met, the Committee was unable to overcome the shortcomings of its constituent agencies, especially the Ministry of Agiculture. The monetary and institutional stakes of this Ministry in the Project were a small fraction of those of DGPC. Moreover, its administrative organization did not confonn to the zones of influence associated with each road and, by offering special savices for the farmers in those zones, the agricultural staff was required to supermpose seemingly arbity dividing lines on top of the extg spheres of responsibility. 9.03 By and large procurement of goods and seices and execution of worls went without complaints from suppliers. The Bank did refuse to finance one consultancy under the loan because it considered that its guidelines had not been followed (para. 5.26). A number of Project components (e.g. feeder road spot imprvements) were not fully executed. 9.04 The road safety component, which was under the jurisdiction of MMC, suffered major delays and restructuring. One reason for this was that this component had not been designed when the loan was made and more than six years elpsed before the Govemment, after considering several alternatives in succession, was able to fix on plans for this component. 9.05 The performceof the Government regarding rura road maintenance is variously reported. There are credible indications that maintenance was deficient. However, the Project made limited provision for rural road maintenance other than budgetary projections in the SAR for maintenance of 10,000 km of roads; also the Project contained a covenant that all rural roads under the Project and "all other improved rural roads offering an appopriate level of service" should be adequately maintained. The Project did not address the institutional issue of how the various echelons of Tunisia's road administration should accommodate the maintenance of  - 28 - scattred road sections, some as short as seven kIn, in their general road maintenance systems. The task described in the covenant thus may have been difficult to fulfill. 9.06 The Borrower's performance with regard to the LA is summarized in Table H. There were significant shortfaLs in comphance by the Borrower with LA provisions. In addition to the points referred to above, the Borrower in all cases but one was late in submitting audit reports. However, land acquisition was never reported to cause a problem. X. RELATIONS BETWEEN THE BANK AND THE BORROWER 10.01 Relations between the Bank and the Tunisian authorities were friendly and cooperative throughout. Two serious issues between the parties occurred in the preparatory stages of the Project regarding the integrated design of the Project and the possibility of co- financing. The Government in both instances was able to persuade the Bank to its point of view. The disagreement on proement of certain consulting work, which involved Tunisia's Central Tendering Board, was resolved in the context of a Bank mission to Tunisia concerned with a number of procurement issues. XI. CONSULTANT SERVICES 11.01 The Project provided for a large amount of consulting services (324 staff -month or somewhat under $3 million equivalent excluding escalation). The actual amount spent for this purpose under the Project and financed from the loan was under $1 million equivalent due to the fact that many of the consultancies were not implemented. Consultants played a major role in the selection and design of ral road improvements. They took no part in supervision of the Project and this separation of the designers from any responsibility for execution may have had impacts also on the quality of the design as well as on the control of the works. The role of consultants in the design and execution of lesser works (CTVs, spot improvements on rural roads and lesser rural works) was limited. The designs and studies that were needed for this purpose were prepared by local officials of the MA. 11.02 Consultant services for this Project, except for tra work and study missions abroad, were provided by sub-assemblages from the joint-venture of French and Tunisians firms that had made the original feasibility studies of 1980. This arrangement fostered a unified approach to the work, may have helped in coordination of the participating agencies of the Government of Tunisia and could have helped to reduce costs. However, it probably also made it difficult for other firms to compete and may have precluded the infusion of divergent views. 11.03 Bank review of the work of the consultants was generally not intensive. As the consultants had become familiar with the rural roads of Tunisia because of their long work as a group on these problems, the Bank's degree of supervision may have been appropriate. However, closer supervision of the work of the consultants concerned with agriculture would have been helpful.  - 29 - XI. PROJECr DOCU?MEATION AND DATA 12.01 The SAR contained much detailed information on individual Proect components and this infomaton was extensively supplemented by material in the Project File. However, the usefulness of this documentaion was Imited as the Project proceeded from very early on without reference to this mateial (for example, the action plans and implemenation schedules). 12.02 The requirements for documentaion regarding farm budgets and credit needs wer very demanding. This may have been one cause why the provision of credit was never monitored in accordance with Project specifcations because the agricultural investments and the related records were made for the administive units of MA as a whole rather han for the zones of influence. 12.03 Data for preparation of the PCR were lacking in significant respects. The Borrower's PCR was not available at the time specified and, although the Bornower collaborated in the draRfing of this report, it did Pot produce economic data. On the basis of the Bank's file there was considerable uncainty as to which Project components had been financed under the loan or included in the Project, as to the course of the Borrower's expenditures for individual components and as to the progress of their physical execution (pars. 5.07). 12.04 A more detailed and systematic reporting keyed to the implementation program envisioned in the SAR and to the principal objectives of the Project would have been desirable. With the addition of information developed for this Project Completon Report, considerable data is now available about the Project and its impact on the nrul economy.  REI!WLIC OF TUNISIA FIf SHIGHWAY (RURAL ROADS) PROJECT (LOAN 2108-TUN) PROJECT COSCLETION REPORT PART II: PROJECT REVIEW FROM THE BORROWER'S PERSPECTIVE 1.1 The Government's view of the Project is contained in the Project Completion Report (PCR) it submitted in late June 1993 (Republique de Tunisie, Ministbre de l'Equipement et de l'Habitat, Direction G6n4rae des Ponts et Chaussdes; Rapport d'Achevement Cinqui2me Projet Roudter Routes Rurales - PMt 2108 Tun). This report is based on the draft prepared by the Bank, discussed and amended extensively during a Bank mission to Tunisia in April/May 1993. The Government's report is basically very similar to the current Parts I and II of the present PCR with some changes in emphasis and detail. It is available in the Project files. 1.2 The Govemment's report concurs with Parts I and II that the physical objectives of the most important Project elements were attained and that achievement of Project elements that were spelled out in detail at appraisal was generally better than the Project elements that were descibed only in general terms. The Govemment's report also agrees that there were serious problems of nt inisa coordination, though the Government's report notes that coordination of many Project activities took place at the provincial level. i.3 The main difference between the two documents can be summarized as follows: The Government's views on regulation of road transport and on pnce control in transport varies from the Bank's point of view; rThe Govemment's PCR mentions the existence of EERs for the second and third groups of roads; and The Governmet's PCR does not mention the absence of information in the files about roads -'d related elements. 1.4 The Government's PCR notes several facts concerning the history of the Project that the Bank's PCR does not mention, but which provides additional useful insights into the process of Project execution: There were no foreign construction contractors in this Project except one Hungarian (previously mentioned) and one Chinese (not yet referred to); Procurement of vehicles was limited by Government-wide restrictions on vehicle imports;  - 31 - Training visits for agricultural experts were made not only to France (already mentioned) but also to Australia (not yet indicated); and The third group of roads was not prepared by the original consult^iits but by another firm.  }REPUBLIC OF TUNISIA FIFTH HIGHWAY (RURAL ROADS) PROJECT (LOAN 2108-TUN) PROJECT COMPLETION REPORT PART 111: STATISTICAL INFORMATION  Table A: Related UDA Credits and Bank Loans Loan Number FY of and Title Purpose Approval Status Loon 746-TUN Improvemet and construction of about 275 km of primary and 1971 Completed; PPAR no. 2732 First Highway Project secondary roads; reconstruction and rehabilitation of 51 bridges 12/79 and culverts; resurfacing and rehabilitation of 1920 kn of paved roads; reorganization and strengthening of highway maintenance operations; improvement of arrangements for transport coordination; study of road user taxation and road tanff rens. I Loan 1188-TUN Improvement of 225 km of primary and secondary roads updating 1976 Completed; PCR no. 5647 Second Highway Project 1968 Transport Survey; preparation of rual roads program. 5185 Loan 1601-TUN Improvement of about 1,000 km of mrul oads and 1978 Completed; PPAR no. 7182 Third Highway (Rural complimnentary agricultual investments; purdas of road 4/88 t Roads) Project maintenance eqiipinelt, spwes and tools, studies of about 2,000 |_________________ _ _ km of additional selected rural roads. Loan 1841-TUN Rehabilitation of about 800 km of eaved roads; first three years 1980 Completed; PCR no. 8648 Fourth Highway Project of six year ma_itenance program including acquisition of 5/90 equipment and materials; studies and training; materials testig and trffic counting equipment; studies and entgineenng of about 1000 km of rural roads. Loan 2896-TUN Upgrading of maintenance management and rehabilitation of 800 1988 Ongoing Highway Maintenance and km of priority roads, five year program of routine and periodic Rehabilitation Project maintenance; workshop construction; acquisition of road maintenance, workshop and laboratory equipment; rehabilitation of 800 km of roads; preparation and implementation of the five year training plan; studies related to the foregoing. Souc: Bank Files  - 34 - Table B-1: Ruml Roads Improvemente as Page I of 3 Proposed at Apprasa (A) and as Fnmay Executed () La"gth Iment ttLvelI Vlidth Cot (00.0 D Coot Km(Dl) _-Ferm oraSeon Pawed_ Aoe NumblermNe . Lots A E A E (im) (ine A E A E swtans 204 U047-oEt, d'A210 2A 13,0 16.47 M4 M4 6 5 360 2S823 205 MC 171P-24 2A 8.8 8,40 N3 N3 6 161 1240 17566 40 12 20S RVSE234tfaia 28 7,0 el89 N44 N4 6 160 22657 2658 SdI Ibu Routs-GPO 2WA 2S,3 19.70 N4 N4 6 6 669 818 22490 41 623 707 Noame JA 6,2 7,44 NS N3 S 6 146 26X630 17605 35783 708 Awlsnne CP3 JA 0.5 6.06 Ns Ns 7 a 103 171.60 16*46 28364 09 Otrult easa A 7.0 S6S1 NO NS 7 100 162.0 14 266 28 060 710 CirtultToasur TA 6,6 6590 NO N3 5 126 174,97 18024 28665 711 t ed-Noua 7A 10.3 7,08 N3 NO 7 6 193 26620 1S736 37560 S07 M9aeth - Osl SC 8,0 723 N2 N2 5 S 114.0 75,7 14260 10470 608 mUe-Sra-.Zat 8C 6.7 6.37 N2 M4 6 6 1S0,2 110,3 21 66 13176 609 KettanaZIg Beranla SC 6,4 5,12 N5! N2 5 5 61,2 44,8 1507 6 770 901 bus M'Khfl OA 13,0 12,2 NO N3 a 246 16823 802 Sidi Abd _bset OA 16,2 12 NS Ns 6 S 290 22 s76 S03 Eob OA 6,0 5.9 N3 NO 6 133 2386 22167 39417 SOSAeKd Crab 9A 9,7 9.6 N4 N4 a 5 486 36876 807 KetMeaUar GA 20.1 19,6 N4 M4 6 750 37313 1602 Berbla-Touam- Naa 15t 16.1 10.o N3 NO 6 270 10 917 1503 Ucn-8t HUMsen 15B 12,6 12,e Y3 NO 6to6 196 802s 16 476 24 729 1508 Sau OthA-Menrme Kwnel 16A 21,2 20,5 4 N4 M toe 61t06 728 804,6 34s40 40996 15I06 Dou Othnm-Zaut Kontc 1%0 6.0 7,9 Ns NO 5 128 16000 1602 Bounwrde-SiJem IOA 15.3 17,82 N4 N4 a a 7 9641,02 42022 35072 1651 SOU498eura ISA 26, 24,94 NS Me 9 6 408 46E,83 16003 16401 1652 Swas i-Wd1 Bouhble ISA 10,0 10,04 N3 M4 9 6 171 246,36 17180 24737 1701 DIsbnana-Ageng 17A 16.2 15,38 M4 N3 a 5 1S9 926 13002 1702 El Amm-easg 17A 11.6 11,36 N4 Ne4 6 064 4e00 ooo 1703 TanikraBuJabOe 17A 12,2 11.98 Me H4 6 6,2 309 616 27767 34037 1704 SldlAbdeke ITA 11,7 11,27 Ns NO 6 22S S36 19060 1751 Ould Ah:ned-Cogran 17A 11,0 10,75 NO NO 6 5 180 363 17273 1501 OugtlhbG ISA S,70 S,40 N3 NO 6 126 224eo 14108 25600 1802 MaunseSdl Chanwkh 18A 11,60 11,S0 NO NO 6 144 350,47 12t22 31 256 1603 dbts-HasJ8etb 16A ,46 9.45 NO NO 0 216 309,50 22867 32751 1604 Moueae- Kantra ISA 11,S0 11,0 NO NO 6 216 2S4,s4 18 05 24s70 1605 KhUEl-Oubbigla 1SA 11,57 11,57 NO NO 6 160 269,57 14143 26028 8IToTAL-16sTRANCHE - 408 367.40 -. - - 833 12348 21804 61876  - 35 - Table B-2: Rural Roads Impements as Page 2 of 3 Proposed at Appraisal (A) and as Finally Executed (E) _ Len __ ___l