Senegal - Feeder Roads Project

Document of The World Bank FOR OFFICIAL USE NLY Report No. 4810 PROJECT PERFORMANCE AUDIT REPORT SENEGAL FEEDER ROADS PROJECT (LOAN 1221-SE) December 7, 1983 Operations Evaluation Department This document has a restricted distribution and may be used by recipients only In the performance of their offcial duties. Its contents may not otherwise be disclosed without World Bank authorization. FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT SENEGAL FEEDER ROADS PROJECT (LOAN 1221-SE) TABLE OF CONTENTS Page No. Key Project Daa............................. 11 Highlights........................................ ........... iv PROJECT PERFORMANCE AUDIT MEMORANDUM Project Implementation. ................................. 2 Institutional luat..................3 Economic impact ............... ..................... . 6 Project Financing....................................... 8 Tables 1. Traffic on Select Feeder Roads.......................... 10 2. Estimated and Actual Costs.....*....................... 11 3. Comparative Maintenance Expenditures.................... 12 4. Equipment Purchases.......................... 13 Annex Borrower Comments.................................... 14 PROJECT COMPLETION REPORT introduction.......................... 17 Project Preparation and Appraisal....................... 18 Project Implementation........................... 20 Institutional Development............. ...... 29 Economic Reevaluation................................... 31 Bank Performance.................................. .... 34 Concluions............ ........ . ...... 35 ables 1. Design 37 2. Construction 38 3. AgriculturalP39 4. Schzdule of Disbursements..............,. 40 5. Alocation t Loan rd................ 41 Agiuua Pro dutin........... *. 39- 6. Project Costs........,.. .. ...... .. ...... .. ...............* 42* 7. Assumption for the Reevaluationt...... 43 8. Economic Reevaluation (with maintenance)...... 44 9. Economic Reevaluation (without maintenance)............ 45 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. PROJECT PERFORMANCE AUDIT REPORT SENEr-^L FEEDER ROADS PROJECT (LOAN 1221-SE) PREFACE The following is a performance audit report on a feeder roads project in Senegal implemented with the assistance of a Third Window loan approved in March 1976 (Loan 1221-SE, US$6.6 million). The last disbursement was made in September 1981, at which time the Loan Account was closed, The attached Project Completion Report was prepared by the Western Africa Region based on consultants', Government and Bank reports, and informa- tion collected by a completion mission in December 1981. An Operations Evaluation Department (OED) mission visited Senegal in early December 1982, to audit the project. The assistance of the Government and the Ministry in charge of public works, in particular, is gratefully acknowledged. OED has reviewed the Appraisal and President's Reports, the Minutes of the Executive Directors' Meetige at which the project was approved, Bank records and files and the PCR. Based on these and the field visit, the audit concludes that while the PCR reviews most issues arising from pro- ject implementation, certain elements of this project's experience are only indirectly covered in the PCR. The Audit Memorandum expauds on these aspects and arrives at a similar conclusion about the project as the PCR, namely that, most likely, the project had a negative rate of return. The draft audit report was sent to the Borrower for comments; their views are included as an Annex to the Project Performance Audit Memo- randum and referenced in the text. - ii - KEY PROJECT DATA Appraisal Actual Item Estimate Reestimate Total Project Cost (US$ million) 6.6/a 11.0 overrun (W) - 66 Loan Amount (US$ million) - 6.6 Disbursed - 6.6 Cancelled - 0.0 Repaid to July 1983 - 0.1 Outstanding to July 1983 - 6.5 Project Completion Date 12/79 06/81 Proportion Completed by Expected Date (%) 100 36 Proportion of Time Overrun (%)- 40 Economic R-te of Return (%) 13.7 negative/b Cumulative Estimates and Actual Disbursements (US$ million) FY77 FY78 FY79 FY80 FY81 FY82 Estimate 3.5 4.9 6.3 6.6 - - Actual 1.1 2.4 3.6 5.0 6.4 6.6 Actual/Estimate (%) 31 49 57 75 97 100 OTHER PROJECT DATA Original Item Plan Actual First Mention in Files or Timetables 03/73 Government's Application - Appraisal 05/75 Negotiations 02/76 Board Approval 03/76 Loan Agreement Date 03/76 Effectiveness Date 07/76 06/76 Closing Date 12/79 09/81 Borrower Republic of Senegal Executing Agency Ministry of the Equipment Fiscal Year of Borrower July 1 to June 30 Follow-on Project None /a The project as appraised would have cost US$10.1 million. However, only 68% of the construction and less than 50% of the maintenance goals were actually achieved. The cost estimate figure has, therefore, been ad- justed accordingly. /b The PCR offers two rates of return, +14.5% and -11%, depending on whether maintenance is or is not carried out on the roads. The audit found that the roads were poorly maintained and for this and other reasons concludes that a negative estimate is more accurate (PPAM, paras. 13 to 15). - iii - MISSION DATA Month/ No. of No. of Man- Date of Mission Year Weeks Persons Weeks Report Identification 02/75 2.0 2 4.0 03/75 Appraisal 05/75 2.5 3 7.5 03/76 Post-appraisal 09/75 1.0 2 2.0 03/76 Supervision I 04/76 1.0 1 1.0 05/76 Supervision 1I 09/77 1.0 1 1.0 11/77 Supervision III 02/78 1.0 1 1.0 05/78 Supervision IV 05/78 1.0 1 1.0 08/78 Supervision V 08/78 0.5 1 0.5 08/78 Supervision VI 11/78 0.5 1 0.5 11/78 Supervision VII 12/78 1.0 1 1.0 12/78 Supervision VIII 02/79 1.0 1 1.0 03/79 Supervision IX 04/79 1.0 1 1.0 04/79 Supervision X 07/79 1.0 1 1.0 07/79 Supervision KI 10/79 1.0 1 1.0 10/79 Supervision KII 12/79 1.0 1 1.0 12/79 Supervision XIII 01/80 1.0 1 1.0 01/80 Supervision XIV 06/80 3.0 1 3.0 07/80 Supervision KV 12/80 1.0 2 2.0 01/81 Supervision XVI 04/81 1.0 2 2.0 05/81 Completion 11/81 1.0 3 3.0 12/81 CURRENCY EXCHANGE RATES Name of Currency (Abbreviation) Franc CFA (FrCFA) Appraisal Year Average US$1 = FrCFA 214 Intervening Years- Average US$1 = FrC?A 227 Completion Year Average US$1 = FrCFA 269 - iv - HIGHLIGHTS The project was one of a series of new type highway projects de- signed in the mid-1970s in direct support of agricultural development. It concentrated on making feeder roads more readily available and rural transport more efficient, as a means of increasing agricultural production and the standard of living of the rural population. The project comprised (i) the creation and development of the institutional framework and capacity to deal with the provision and maintenance of feeder roads, and (ii) the construction of about 1,000 km, and maintenance of about 1,2.0 km of feeder roads. The project was expected to tackle the feeder roads problem at its two critically weak points simultaneomsly: the institutional process of selecting and deciding on a yearly work program and the process of physically executing the agreed plan. The institutional set up was to be improved by the creation of an Interministerial Consultative Committee (ICC) to decide, at the political level, on a yearly feeder roads construction program developed from alternatives compiled and evaluated by the Directorate of Studies and Program- ming (DSP) of the Ministry in charge of public works. To improve the physical execution the project was to help create a Feeder Roads Subdivision (Bureau des Pistes de Production, BPP) under the Road Maintenance Division of the Public Works Directorate of the Ministry of the Equipment, and within it, three construction brigades to do the work. The project was estimated to cost the equivalent of US$10.1 million, which would be financed by a US$6.6 million Bank loan and a US$3.5 million Government contribution of which an estimated US$1.6 million were taxes and duties and US$1.9 million were local costs. The loan,was made in March 1976, and the project was expected to be completed in December 1979. In September 1981, when 68% of the construction target and less than 50% of the maintenance objective had been achieved, loan funds ran out and the project was stopped. By then a total of US$11.0 million equivalent had been spent (US$6.6 million financed by the Bank loan and US$4.4 million by the Government). Senegal's agricultural production had not increased during the construction period and the standard of living of the rural population had hardly changed. The main reasons for the poor agricultural performance were: poor weather, including drought but also rains in January when the harvest was being stored for processing, contributed to low production and poor groundnut yields; depressed international markets; the paralyzation and subsequent bankruptcy of ONCAD (the government groundnut purchasing monopoly); the serious operational and financial difficulties of regional development organi- zations (SODEVA, SOMIVAC, SODEFITEX and SAED); and uneven and often bad domestic pricing, credit, cooperatives and other agricultural policies were not conducive to maximizing production. Moreover, throughout most of the project's life Senegal was affected by extraordinarily hard financial and economic circumstances which produced a breakdown. of the entire economy and made ne"essary a redefinition of priorities by the Government and all its external partners. Because of these exogenous factors the economic return of the project would, in any event, have been well below appraisal estimates. - v - However, the audit concludes that even if these exogenous circum- stances had not existed a number of project-related factors would have pre- vented the project from having a significant impact. For example, the project's provision for subsequent maintenance of the roads and its assessment of the human and institutional needs to implement it were underestimated. The rate of return of the project is estimated to be negative, mostly on account of the short life of the roads caused by their relatively poor quality and their lack of subsequent maintenance. The audit highlights the following points of interest in connec- tion with this project: - structuring execution of the project with a rather short start-up period and optimistic output targets over just about half the time the consultants had recommended caused serious implementation problems and produced a lack of concern for quality, which contri- buted to the project's negative economic impact (PPAM, paras. 4, 5 and 14); - the project's design made limited allowance for the institution- al and practical difficulties of implementing a comparatively large project on a tight schedule in a country with scarce human resources; this was the cause of much delay and many difficulties (PPAM, para. 6); - although institution building was a major objective, the assessment of the staffing constraints and the institutional arrangements devised to implement the project failed to generate any lasting institutional impact (PPAM, paras. 7 to 12); - the comparatively low priority afforded, at least initially, to feeder roads contributed to the poor economic impact of the project (PPAM, para. 13); and - the project made no special provision for the subsequent mainte- nance of the roads although most of its economic benefits were expected to be derived not so much from the construction of the roads as from maintaining them in good condition during their expected economic life (PPAM, paras. 13 and 15). PROJECT PERFORWMACE AUDIT MEMORANDUM SENEGAL FEEDER ROADS PROJECT (LOAN 1221-SE) Introduction 1. The project was based on a consultant's survey of more than 6,000 km of roads and tracks, which identified about 1,300 km of feeder roads to be improved. The consultant proposed that the work be carried out by a new Feeder Roads Subdivision (BPP) within the Road Maintenance Division of the Directornte of Works in the Ministry of Works and Urbanism over a period of five years. In February 1975, a Bank identification mission agreed with the consultant's report and requested that staffing and equipping details be worked out to implement the project in three rather than five years. The project was appraised in May 1975, and because there was no institutional mechanism to deal with feeder roads in the future, the creation of one was included in the project. The consultant's survey was used to define the first year work program but for the second and subsequent years the institutional mechanism of the project would take over to define what was to be done. To this effect the project foresaw the creation of an Interministerial Consulta- tive Committee (ICC) chaired by the Minister of Works and in which the Minis- tries of Rural Development, Planning, Interior, and the government-run ground- nut marketing monopoly ONCAD would be represented. The Directorate of Studies and Programming (DSP) of the Ministry of Works, which had been recently created under an earlier Bank project, would act as the technical secretariat to the Committee to screen and assess feeder road proposals put forward by the ministries and to monitor execution of the project. It was expected that technical assistance experts provided to DSP in that project would assist it in developing the technical and economic criteria to screen and select the projects to be included in the subsequent annual work programs. Nevertheless, during negotiations nominal criteria for the selection of roads were agreed on and the loan made provision for consultants to assist in project monitoring. 2. Execution of the project was to be carried out by the proposed new Subdivision through three construction brigades established under the project. The BPP would be housed in a new building that was to be constructed with loan funds. The loan would also finance equipping of the brigades, materials and supplies necessary to carry out the work, and technical assistance to st-er- vise the work and train BPP's new staff on the job. To ensure proper execu- tion of the project and to giNe the new BPP adequate rank, the Bank required it to be headed by an engineer and to be staffed by at least one other engi- neer and two administrative assistants. BPP's work brigades would similary be staffed with suitably qualified foremen, operators, mechanics and workmen in adequate numbers. - 2 - 3. The project, thus, appeared to have tackled the feeder roads problem simultaneously at its two critically weak points, the planning process of selecting and deciding on a yearly construction program and the process of physically carrying out what was agreed. However, the appraisal report was primarily devoted to the physical aspects and contained no assessment of whether the Government could absorb the proposed institutional planning structure or had the staff and expertise to do so. The project's lack of success in effectively resolving the non-physical aspects of the probl.em were the main causes of its poor impact. Project Implementation 4. In retrospect, it appears that having cut back the project period from the five years recommended by the consultant to three years was not in the project's best interest. The audit agrees with comments expressed by BPP staff, that the decision required compressing into little more than half, the achievement of goals that were largely designed for almost twice as long. This meant that the yearly construction targets had to be set unrealistically high for Senegalese circumstances and, therefore, ended up working to the detriment of project objectives.!/ This happened not so much because the high targets were not an incentive to increase productivity, but because, in the drive to achieve the target, quality objectives were disregarded. In turn, the low quality implementation of the roads was one of the main causes of their short life and limited economic impact. While not contemplated at Appraisal, the PCR views the project as having had two phases: the first from mid-1976 to early 1979 and the second extending to mid-1981; it notes that during the second phase considerable improvements were made (PCR, paras. 3.10 et seq). 5. Most roads were built without proper ditches or drainage structures; a site inspection revealed that on most of them the thickness of the gravel surfacing was thin and that compaction was inadequate. For these reasons weathering and traffic have caused the roads to deteriorate quickly and revert to conditions just marginally better than those of similar unimproved feeder roads. The situation, which seems to have been more a consequence of hasty construction rather than an inadequacy of the construction standards (PCR, Table 1), has been aggravated by the lack of maintenance, although the Region 1/ Bank staff have commented that the project would have required no more than 300,000 to 400,000 cubic meters of regravelling in the three year period. However, since BPP had, at best, a fleet of 20 six cubic meter trucks, each truck would have had to complete 6 hauls every day, each day, during the three year period, just to procure the gravel, a diffi- cult task if hauling distances were long. - 3 - states that there are now increased maintenance activities. During implemen- tation, the three construction standards expected to be used were reduced to just one; the audit is of the opinion that this action was positive as it made it easier to train and organize the work crews. 6. Also, the shorter project period with high output expectations did not allow adequate time to complete activities which had to precede the execu- tion of any actual work. The project schedule was compressed unrealistically. Before the work could begin the Subdivision had to be created, equipment for the brigades had to be procured, offices for BPP had to be constructed, technical assistance had to be employed and personnel at all levels, from top management to operators and workmen, had to be recruited and trained; for all this, the schedule. allowed six months (Appraisal Report, para. 3.12). In actual fact, while the BPP was created before the loan was approved and most of the procurement was accomplished on time, the recruitment of local staff could not be achieved on schedule because Senegal did not have personnel of the qualifications and experience demanded by the project.V/ For instance, recruiting a Chief for the BPP with the engineering qualifications demanded by the Bank proved difficult, and eventually a young, inexperienced, engineer was appointed to the position. Similarly, while the technical assistance experts were recruited and commenced their assignment on time, no suitable counter- parts to work with them could be found. Also, severe difficulties were experienced in recruiting brigade chiefs, operators, mechanics and foremen. The audit has found that the limited allowance made in the project for start- up, and particularly for staffing the new BPP, together with what proved to be an optimistic assessment of qualified personnel available in the country, were the main causes of thc initial delay in project execution, although the Region points out that delays are to be expected when projects are to be executed by force account. In retrospect, it appears that little regard was paid in project design to the difficulties of managing a new program of a comparatively large size, on a tight schedule, in a country with scarce human resources. Institutional Impact 7. Institution building was one of the primary objectives of the project. This comprised (SAR, para. 4.02) the creation of an Interminis- terial Committee (ICC), the use of the Ministry of Works' Studies and Program- ming Directorate (DSP) as its technical secretariat, and the creation of a Bureau des Pistes de Production (BPP), a new Subdivision of the Ministry's 1/ Bank staff have commented that not allowing for "spot improvement" among the construction standards was a shortcoming that could have been avoided. 2/ Bank staff have indicated that while this is true, the lack of front end loaders (PCR, para. 3.04) was the cause of considerable delay. -4- Road Maintenance Division. It was hoped that the creation of these institu- tions would ensure that the building and maintaining of feeder roads would become a permanent, on-going feature of Government activity. Yet, as relaLed in the PCR and observed by the audit, building and maintaining feeder roads has not become a permanent feature of Government activity. On the contrary, as soon as project funds ran out it ceased altogether. 8. The audit questions whether, in the context of national priorities of which Central Governments have to care about, the concern for feeder roads was significant enough to have warranted the institutional approach adopted in the project. The Bank may have been mistaken in its assessment of the signals emanating from the Government on the subject of feeder roads. Perhaps the problem would have been best tackled at the local level which is much closer to the problem itself. In the end, the institutional arrangements of the project do not seem to have lived up to expectations../ 9. While the Interministerial Committee was created as agreed, it met only twice. Part of the reason was that its technical secretariat, the DSP, which was supposed to screen, evaluate, and present the ICC with a set of feeder road construction alternatives, never did so. DSP, as it turned out, could not carry out its assigned function in this area because it was understaffed and overburdened with other more pressing duties, among them the preparation under a parallel Bank project of a national transport plan. The economic and technical criteria which DSP was supposed to develop to screen and evaluate the feeder roads proposals were never prepared because no pro- vision was made in the project to assist DSP to do it and because no local staff sufficiently experienced were available to carry out the task..;/ In the beginning, the matter was not of pressing urgency because it took more than three years to carry out the first year construction program that had been defined on the basis of the consultant's study; however, as construction was completed and the definition of subsequent work programs became necessary, the problem came to a head. With the Bank's concurrence, a foreign consultant was hired to do the job and he did the task alone as no counterpart staff were available. Thus, no one in DSP was trained on how to prepare an annual program or what methodologies to use to evaluate and screen feeder roads. 10. Parallel to the ICC-DSP arrangement, the project required the creation of the BPP, a new Subdivision of the Maintenance Division in the Directorate- of Works. At -the time, --both- an earlier -(Second--Highway Project) - and a parallel (Third Highway Project) project were involved with the Mainte- nance Division. Therefore, the Bank knew that the Division and indeed the 1/ Bank staff have indicated that records of meetings with Senegal's Prime Minister, starting in late 1978 and extending through 1980, show that feeder roads had a very high priority both on political and economic grounds. In addition, they considered institution building to be, of necessity, a long term development. 2/ The Government has noted that the "impact study" included under the project would have filled some of these gaps but that it was stopped as a result of poor Bank follow-up of the funds allocated to do it (see Annex). - 5 - indeed the Directorate to which it reported, were going through a very diffi- cult transitional period; both being underfinanced, understaffed, their scarce personnel poorly qualified and their facilities (both housing and workshops) in poor condition. There were also shortages of supplies and materials, and in addition, three years of effort under the Second Hi&hway Project had not produced visible improvements. Under these circumstances, it does not appear to have been a particularly good administrativc strategy to have burdened the Division with the addition of a new Subdivision to compete for resources, for the use of equipment and facilities and most importantly, for the staff. Although initially the creation of BPP was inconsequential, particularly because of the delays in hiring the chief of the unit, attitudes * changed as BPP's chief was appointed. The new appointee's actions were limited by the complex and cumbersome administrative procedures and his authority and responsibility within the Division were left unclear. Moreover, his lack of experience prevented him from operating adequately and competing effectively for funds, equipment, materials, supplies and even accommodations, as construction of the new BPP building was delayed. Successive Bank super- vision missions which called attention to the deteriorating performance of the unit offered little concrete help in solving problems.!/ In fact, although it was the Regional Kission to Western Africa's responsibility to supervise the project, during the first critical year of the project (1976) and also the subsequent year (1977), only one supervision mission reviewed the rojRct each year. Supervision was increased in 1978 but problems persisted..y Tne lack of support and a shortness of local funds to pay, particularly, salaries and other local costs produced discouragement and low morale set in within the BPP. 11. In 1979, poor performance eventually caused the Bank to threaten to cancel the Loan, to which the Government responded by replacing the project manager, not by an engineer as agreed with the Bank, but by an experienced road technician. Simultaneously, the Feeder Roads Subdivision was made administratively and financially autonomous and to report to the Director General of Public Works. Much to the Bank's satisfaction problems began to 1/ The audit mission was informed, and the record shows, that Bank missions did not avail themselves of the help of the "Comite de Relance des Investisments", a permanent committee of the Office of the Prime Minister set up for the purpose of "unblocking" project implementation when problems arose. 2/ Bank staff do not agree that more frequent supervision from Abidjan would have made a significant contribution as each supervision mission went as far as it could in finding ways to improve the situation. It was also noted by Bank staff that as of late 1978, project problems were raised by the Bank's Resident Representative in Senegal at his monthly meeting with the Prime Minister who was also President of the Comite de Relance des Investisments. - 6 - sort themselves out; according to the PCR, these actions ended a period of unsatisfactory administrative dependrqce and a second phase of project execu- tion commenced. BPP had been raised in rank and accorded special status. they were allocated proper housing, exclusive space in workshops, all neces- sary equipment, were even given new radio equipment to keep daily communica- tions between the field brigades and headquarters. In short, BPP wai given preferential treatment throughout. However, this ixclusiveness created resentment towards the unit in the Ministry, and the arrangement could, in the end, only be sustained as long as foreign funds were available to finance it.1/ In fact, as soon as loan funds ran out activities ceased and the BPP found itself without an immediate role. Yet, salaries and other social costs have had to be continued and have become a significant burden to the Govern- ment. Moreover, because most of BPP's equipment was used up, and in par- ticular its tiuck fleet performed badly and had to be written off, all that is left of the BPP is its large, albeit well trained, staff and the building that houses it. 12. At Appraisal, the Bank seems to have understood that feeder roads occupied a higher priority than they were accorded in Senegal's transport sector, and, thus, overstated the institutional approach to tackling the problem. Also, by trying to implement such a comprehensive project and failing, the Bank has left the country with the problem of deciding what to do with the defunct BPP and its staff. One alternative is to assimilate the staff to the Ministry's regravelling brigades which are active throughout the national road network../ Economic Impact 13. That feeder roads could not have enjoyed a high comparative pri- ority in the scale of transport investments was clear even at appraisal, as not one of the roads in the over 1,000 km proposed for improvement under the project had traffic volumes above 20 vehicles a day, and most had volumes in the order of 5 to 10 vehicles a day. This contrasts sharply with the 300 to 900 vehicles per day carried, at the time, on national roads and highways and which were also in poor condition and receiving scarcely any 1/ Bank staff have indicated that the strategy of separating the BPP from the main stream of the Ministry was essential because otherwise BPF would have continued to be accorded a very low priority. 2/ Bank staff have pointed out that this proposal was agreed with the Government, but would not have been necessary had a Second Feeder Roads Project been approved: the efficiency and effectiveness of BPP's per- formance would have guaranteed its continued financing under a second project, - 7 - attention..!/ The difference in priorities was masked by the economic analysis of the Appraibal Report which showed the project to have an average overall rate of return of 14% although the construction works under it yielded a return of just 7%. The main benefit came from an expected improvement in, and continued ainquate maintenance of the roads in the project (Appraisal Report, Annex 7). Yet., the project made no particular provision for the ex-cution of this critical element and, in the end, failed because of the lack of it. There was just a covenant in the Loan Agreement (Section 4.04) requiring the Government to allocate funds for the maintenance of the roads and to increase the amount year after year to take into account changes in costs and in the requirements of the roads; but no special arrangements were worked out to ensure future maintenance. In retrospect, this serious short- coming, together with the known difficulties of improving maintenance in an overall maintenance deficient country, where years of trying under an earlier Bank project had not been successful, made the project particularly weak and risky. 14. The PCR has presented two rate of return calculations, one assuming proper maintenance, the other assuming no maintenance. The audit agrees only with the second and just in the sense that it is a hypothetical reflection of what would have happened had the project been implemented under ideal condi- '-ions except for the maintenance. The calculation yields a negative rate of return and is based on (i) slightly higher traffic figures than those reported by BPP to the audit mission (PPAM, Table 1 vs. PCR, Table 7) - the difference is most likely due to the different timing, and consequently information available to the completion and audit missions; (ii) a somewhat high traffic growth rate - there has been hardly any growth in traffic and the rate was set deliberately high to compensate for what appeared to be an abnormally low base year (1981); (iii) a road deterioration function which assumed that without maintenance roads built in Eastern Senegal and the Thies regions would deteri- orate in 6 and 5 years respectively - the roads have deteriorated much faster; (iv) vehicle operating cost savings assessed at an average of FrCFA 123 per vehicle kilometer - too high since total operating costs for an average truck on a poor unpaved road in Senegal are about FrCFA 130 per vehicle kilometer;. / and (v) a rather long economic life for the roads - their rather poor con- struction has caused them to have a very short life.l/ I/ Bank staff have taken issue with the implied notion that trunk roads should be taken care of before feeder roads. Their view is that a bal- anced main/feeder road development strategy is the most viable approach. The virw assumes, of course, that there are enough resources to invest in both types of roads. 2/ Bank staff noted that vehicle operating cost savings were assessed at FCFA 60-119 per vehicle kilometer, depending on the state of degradation of the road section in the "without project" case. Operating costs for an average truck on a badly degraded track were estimated by consul- tants at about FCFA 280/km. 3/ The Government has commented that the rate of return estimate appears to be based on a rather simplistic view of the consequences of the poor maintenance (see Annex). - 8 - 15. In retrospect, the project has yielded a negative rate of return because: traffic levels have been extremely low (which is not directly related to the project itself but the consequence of exogenous factors over which it had no control); no maintenance has been done to preserve the roads; their unit costs were high (about US$15,000 per km);L/ and the quality of the construction was poor. The short life of the roads did not allow suffi- cient benefits to accumulate to offset the high costs. Project Financing 16. Throughout the PCR there is frequent mention that one of the prin- cipal reasons for the project's shortcomings was the lack of counterpart funding. This does not appear to have been the case in overall terms. While the Bank expected the Government to provide US$1.9 million equivalent in three years (excluding taxes estimated at appraisal to be US$1.6 million, Tables 2 and 3), it, in fact, ended up providing US$4.4 million in five years (Tables 2 and 3). However, as is evident from Table 2 there was a serious underestima- tion of labor costs, which rose by almost 200%. Since this was a purely local cost, not financed by loan proceeds, the Government appropriations for the project had to be almost fully used for salaries leaving little for materials and other project requirements. Thus, a situation developed under which it was difficult to physically implement the project. In fact, there is evidence that at times, appropriations were not enough to cover the full salary bill. As a consequence, project implementation, particularly in the early years, suffered considerably. 17. But the administration of the cash and its flow also created prob- lems. The administration of the available cash was directly related to the administrative experience of those in charge of the project and to the ade- quacy of the administrative mechanisms set up to access the funds. The cash flow, on the other hand, was directly related to the project's financial structure and funding sources. 18. The administrative experience of those in charge of the project was, as mentioned, limited; and, it was this lack of experience that prevented setting up, early on, adequate administrative channels to handle the funds that were being made available through the Ministry of Public Works' budget. For instance, the lack of an imprest account for the feeder roads prevented separating out funds allocated to the project from those allocated to general 1/ Bank staff disagree that US$15,000 per km, allowing for the residual value of the building and other minor unrelated expenses, is a high cost. - 9 - maintenance and other equipment in the budget. The situation was aggravated by the bureaucratic resentment towards a Subdivision that was expected to use, in its first year (Table 3), almost as many resources as the whole Division to which it reported; in itself a poor administrative practice./ 19. The financial structure of the project also created difficulties. By cutting the project period from five to three years, it became necessary to program the provision of funds in as few years. This meant that the Bank had to expect the Government to allocate for feeder roads an inordinately high proportion of its highway budget. Not only were the amounts large, but the Bank's disbursement practices made the situation appear as even more diffi- cult. Customarily the Bank reimburses against incurred expenditures (includ- ing taxes and duties); the Government, thus, expected to have to budget for 100% of the costs, spend the funds, and then be reimbursed the 65% of the costs covered by loan proceeds. This procedure was likely to create a cash flow problem which, in the eyes of the Government, could be alleviated only by a Bank contribution to set up a revolving fund to finance current project expenditures. The Bank, on occasion, assists in setting up imprest or revolving funds; in this case, however, it was not prepared to grant the assistance and it appeared to the Government that it had to face up to the cash flow problem alone, at a time when overall public finances were being seriously eroded by depressed world markets for Senegal's main export. 20. In actual fact, instead of implementing the project over three years, the Government implemented it over five as originally planned. This produced a much smoother cash flow (see Table 3), one that interfered much less with other Government commitments. Also, since an important proportion of Bank loan proceeds was disbursed as direct payments to suppliers, the Government's cash flow problem became easier to manage. A revolving fund for local expenditures only was set up in 1979; it essentially involved the pay- ment of salaries and did not address the problem mentioned in the preceding paragraph. Yet, the establishment in 1979 of the imprest account for Govern- ment financed BPP expenditures significantly contributed to expediting pro- ject implementation. 1/ Bank staff have noted that comparing the first year investment under the project with the Maintenance Division's budget is not appropriate because project investments included the purchase of equipment for the BPP, whereas equipment used by the Maintenance Division is budgeted for under the Equipment Division of the Ministry of Works. - 10 - TABLE 1 SENEGAL FEEDER ROADS PROJECT (LOAN 1221-SE) Traffic on Three Selected Feeder Roads (AADT 1980, 1981) 1980 1981 Light Heavy Total Light Heavy Total Vehicles Vehicles Traffic Vehicles Vehicles Traffic Bambey-Gawane 19 2 21 18 1 19 Kaffrine-Delby 13 5 18 12 2 14 Sandinier-Saretening 1 2 3 4 - 4 Source: Traffic Surveys by SONED December 1982 SENEGAL FEEDER ROADS PROJECT (LOAN 1221-SE) Estimated and Actual Costs (FrCFA and US$) APPRAISAL ESTIMATE ACTUAL COSTS US$ million at US$ million at -------FrCFA m------- ----1US$ = 225 FrCFA--- ----FrCFA m--------- --1US$ - 227 FrCFA-- Local Foreign Total Local Foreign Total Local Foreign Total Local Foreign Total Building 37.4 27.5 64.9 .2 .1 .3 60.0 23.8 83.8 .3 .1 .4 Equipment 60.2 562.8 6Z3.0 .3 2.5 2.8 57.5 510.6 568.1 .3 2.2 2.5 Materials 97.8 709.4 807.2 .4 3.2 3.6 145.2 614.4 759.6 .6 2.7 3.3 Tech. Asst. 12.8 101.0 113.8 .0 .4 .4 82.0 299.2 381.2 .4 1.3 1.7 Contracts 21.9 87.6 109.5 .1 .4 .5 - - - - - - Salaries 196.0 - 196.0 .9 - .9 581.0 - 581.0 2.5 - 2.5 Impact Study 29.0 24.9 53.9 .1 .1 .2 Transport 19.7 35.0 54.7 .1 .2 .3 CEREEQ 16.3 - 16.3 .1 - .1 Sub-total 426.1 1,488.3 1,914.4 1.9 6.6 8.5 990.7 1,507.9 2,498.6 4.4 6.6 11.0 Taxes 360.0 - 360.0 1.6 - 1.6 - - - - - - Total 786.1 1,488.3 2,274.4 3.5 6.6 10.1 990.7 1,507.9 2,498.6 4.4 6.6 11.0 N2 Source: Audit Mission based on Ministry of Planning and Ministry of Works data. December 1982 SENEGAL FEEDER ROADS PROJECT (LOAN 1221-SE) Comparative Highway Maintenance Expenditures and Project Financing (FrCFA million) Expenditures in 73/74 74/75 75/76 76/77 77/78 78/79 79/80 80/81 Feeder Roads /a - - - 1,201 537 533 - - Feeder Roads b - 40 40 558 358 409 507 668 Maintenance Division /c 1,178 1,387 1,500 1,704 1,699 1,995 n.a. n.a. Mtce & Eqpmt Directorate /c 1,312 1,508 1,674 1,990 2,284 2,550 2,438 3,427 Ktce & Eqpmt Directorate d n.a. n.a. n.a. 2,410 2,594 5,230 2,619 3,527 /a Expected expenditures per Appraisal Report: 1976/77 1977/78 1978/79 Cost Taxes Cost Taxes Cost Taxes Government 193 162 115 97 118 101 Bank 846 - 325 - 314 - 1,039 162 440 97 432 101 /b Actual expenditures per Ministry of Planning "Fiche de Suivi" and Loan disbursements: 76/77 77/78 78/79 79/80 80/81 Total Government 200 160 160 220 250 990 Bank 358 198 249 287 418 1,510 558 358 409 507 668 2,500 t4 /c Actual expenditures excluding capital expenditures per FCRs Second and Third Highay Projects. /d Actual expenditures including capital expenses. SENEGAL FEEDER ROADS PROJECT (LOAN 1221-SE) Appraisal and Actual Equipment Purchases (FrCFA m) PRESENTLY APPRAISAL EXPECTATION ACTUAL EXECUTION AVAILABLE No. of No. of Type of Equipment Pieces Cost Pieces Cost Bulldozer (200 IP) 3 74.7 3 55.75 1 Graders ( 130 HP) 9 19.6 9 91.97 4 Compactors 3 36.9 3 15.74 1 Tractor w/ 45t low bed trailer 1 20.0 1 21.25 1 Dump truck (6 m3) 18 117.O 18 109.44 1 Tank truck (8 m3) 9 85.5 9 60.30 4 Service trucks (2.5 ton) 6 24.0 6 18.78 3 Water pump (50 m3/hr) 6 3.6 6 2.16 2 Roller (600 kg) 3 3.0 3 3.11 - 250 1 Cement mixer 3 2.7 3 2.44 - Pick-up trucks 3 7.5 6 7.23 3 Station wagons 5 12.5 5 11.46 1 Tank truck (4 m3) 3 8.11 Camp equipment (sets) 3 2,7) Electrical generators (4 KVA) 3 3.6 3 1.30 2 Water tank on trailer (1 m3) 3 1.5 3 1.37 - Various tools - 3.0 - .74 - Service station trucks 3 9.9 3 32.89 2 6 agricultural tractor and compactors - - 6 56.15 4 Front end loaders - 3 24.01 2 Spare parts (5.9%) 3Z.4 36.59 578.2 568.06 Source : Audit Mission December 1982 - 14 - ANMEX Page 1 Republic of Senegal No. 01285 VINISTRY OF THE EQUIPMENT Directorate General of Public Works Dakar, July 20, 1983 The Director General of Public Works to The Director Operations Evaluation Department World Bank 1818 H St. N.W. Washington D.C. Dear Sir : Following are the comments generated by the Project Performance Audit Report on the Feeder Roads Project (Loan 1221-SE) 1 - Substitute Ministry of Public Works for Ministry of the Equipment or Minisiry in charge of public works. 2 - The appraisal estimate of total project costs is given on page (ii) as US$ 6.6 million but on page (v) it is US$10.1 million. The two figures differ from the estimate of US$ 8.5 million given in the PCR and in the Appraisal Report of March 1976. 3 - Page 1: there has never been a Division of Highways within the Ministry of Public Works and Urban Development. Also, the Ministry's name is Ministry of Public Works and Urbanism not Urban Development. 4 - The report makes no mention of the impact study; that compo- nent received great attention from the Bank because it would have filled gaps with regards to the definition of feeder road selection criteria, but the component was cancelled be- causeof the poor monitoring by the Bank of the availability of funds, which was carried out without taking into account the itemized allocation. The report should have brought out these facts. - 15 - ANNEX Page 2 5 - The calculation of the 11% rate of return appears to bi based on a very simplistic approach to the consequences of poor maintenance over the life of the roads. Please excuse the delay in furnishing you this reply, it was due to the fact that the document, of which only one copy was provided to us, was examined by the various departments involved in monitoring the project. Very truly yours, /s/ Moussa TAMBAlOU 一一／ご＿_ メ乙りJのれ：メA・k - 17 - PROJECT COMPLETION REPORT SENEGAL FEEDER ROADS PROJECT LOAN 122L-SE) 1. INTRODUCTION 1.01 Senegal's transport system is among the best developed in West Africa. Extensive port and railroad facilities were first established when Dakar was the administrative and commercial center of French West Africa. The Port of Dakar is still a major deepwater port serving the Sahel Region and an important bunkering point for north-south sea traffic, while the 1,030 km railway line, of which 660 km to Mali, has lost some of its earlier importance for goods traffic due to the railway's steadily declining opera- tional capacity and efficiency, and higher competition for improved alterna- tive routes. 1.02 Development of the road network between Dakar and the regions of the country became a priority objective after Independence in 1960. At the start of the 1980s, the country had a total of 13,700 km of classified roads, some 3,230 km of which are paved. It has thus more paved roads than the Ivory Coast (with about 3,000 km of paved roads), and considerably more than Cameroon (about 2,000 km), despite the fact that has only about one-fourth and one-half of their respective GNPs and a smaller area. 1.03 Bank Group involvement in the road sector started rather late and has essentially focussed on highway maintenance and feeder road construction. The First Highway Project (Credit 198-SE, US$2.1 million, 1970) financed feeder roads, road maintenance equipment and a study for the improvement and maintenance of the primary road network. The Second Highway Project (Credit 366-SE, US$8.0 million, 1973) continued to provide equipment, technical assistance and training for road maintenance and financed a first phase pavement strengthening program, as well as studies for a subsequent phase strengthening program and for the improvement of feeder roads. 1.04 Based on these studies, a Third Highway Project and a Feeder Roads Project were negotiated in 1976 (Loans 1222-SE, US$15 million, and 1221-T-SE, US$6.6 million). The former continued the pavement strengthening program, improved the organization of highway maintenance and equipment repair opera- tions in the Directorate of Public Works (DPW), .1/ provided 5upport to the National Soils Laboratory (Centre Experimental de Recherches et d'Etudes pour 1'Equipement--CEREEQ), included an evaluation of the local construction industry and provided technical assistance to improve the transport planning capacity in the Ministry's Directorate of Studies and Programming and to carry out preinvestment studies. The Feeder Roads Project is the subject of the I Reorganized in March 1979 and renamed General Directorate of Public Works. - 18 - present report. During its implementation, a Fourth Highway Project (Loan 1810/Credit 993-SE, US$10/28 million) was approved in 1980. It continues assistance to road maintenance through a training program, a study to improve maintenance planning and pavement strengthening, provides for upgrading CEREEQ, and for construction of an 85 km two-lane paved road between Louga and Dahra, co-financed by OECF, Japan. 1.05 Several Bank-supported agricultural projects also included sizeable feeder road components. The more important of these were located in the southern and southeastern regions of the country, where agricultural potential is good and population density comparatively low. The Casamance Rice Project (Credit 252-SE, US$3.7 million, 1971), as well as two settlement projects in eastern Senegal (Credit 254-SE, US$1.35 million, 1971 and Credit 570-SE, US$2.0 million, 1975) included road construction units that operated with varying efficiency. All three projects experienced difficulty in attaining their goals, including some 250 km of roads in Casamance and 166 km in eastern Senegal and to obtain acceptance by DPW for maintenance of them. 1.06 The present report is based on consultants' progress reports, Government reports, Bank supervision missions and on information collected during a project supervision/compLetion mission in December 1981. II. PROJECT PREPARATION AND APPRAISAL Preparation 2.01 The project under review was a "new" type of highway project focus- sing exclusively on feeder roads construction and maintenance in Senegal. It was prepared under the Second Highway Project by a French consulting firm, whose terms of reference were to review the country's feeder road requirements over a ten-year period, to formulate an appropriate strategy in feeder road development and to propose a project based on a priority list of roads to be improved over a first five-year period. 2.02 After reviewing data on current agricultural production and plans for future development of the sector, the consultants surveyed some 6,000 km of roads and tracks to assess their serviceability, as well as actual and estimated future traffic. In a report submitted the end of September 1974, they proposed a five-year improvement program for some 1,330 1cm of feeder roads, to be carried out by a subdivision to be specially created in the Road Maintenance Division of DPW. The Division had itself been established only in 1973 to execute the four-year maintenance program financed under the Second Highway Project. 2.03 In February 1975, an identification mission discussed the report with the Government and the consultants. To ensure that the proposed project could be kept reasonably responsive to changing agricultural development needs and would not exceed available Bank Group financing, it was decided to reduce the project period from five to three years. The consultants were - 19 - asked to develop detailed proposals for staffing and equipping the proposed new feeder road unit and to examine the scope for involving small local contractors in the execution of the project. Project Appraisal and Processing 2.04 The project was appraised essentially as prepared in May 1975. Project processing was done in parallel with the Third Highway Project: negotiations took place in Dakar, February 9-11, 1976; Board presentation in Washington, March 16, 1976; and signing of the Loan Agreement and four side letters on March 31, 1976. Project Objectives 2.05 The Feeder Roads Project as appraijed had two main objectives: (a) to support agricultural deve.lopment through the improvement of about 1,000 km of feeder roads and the maintenance of an additional 250 km con- structed under earlier IDA agriculture projects; and (b) to establish the institutional framework for continuous expansion and proper maintenance of the feeder road network. In order to achieve this, the project foresaw: (a) the creation of a feeder roads subdivision, Bureau des Pistes de Production (BPP) under the Road Maintenance Division (RMD) of the Directorate of Public Works (DPW) including the establishment of brigades to do the work, aiming at efficient execution of feeder road construction and maintenance; (b) the development by the Directorate of Studies and Programming, (DSP), of technical and economic criteria for defining annual road construction programs from proposals submitted by various Government agencies; and (c) the establishment of an Interministerial Consultative Committee (ICC) to review and approve these programs. 2.06 The project aimed also at developing the capabilities of small contractors who would construct BPP headquarters and drainage structores, and haul laterite for the brigades. It further foresaw a study of the tarh- nical and economic feasibility of intermediate technology methods for feeder road construction, and to explore the possibilities of obtai,.*,ng local contri- butions from rural communities for feeder road maintenance. Project Description 2.07 To meet its objectives, the project provided financing for a three-year program of feeder road improvements. Funds were to cover: (a) purchase of equipment and spares for three construction brigades; (b) purchase of supplies and materials for operations included under the project; - 20 - (c) construction of headquarters for the Feeder Roads Subdivision (BPP) and; (d) technical assistance to BPP for (i) assisting in programming and implementing the proposed improvement and maintenance program and (ii) monitoring the project. 2.08 In contrast to traditional highway projects, which finance the construction of defined road sections, mostly by t:ontractors, the project provided for a three-year program of some 1,000 km of rural road improvements to be carried out by force account. Of the proposed three-year improvement program 400 km were to be only summarily improved (POA standard) with a 6 m reshaped roadway and spot gravelling as required; 500 km would be one-lane roads (PA1 standard) with a 7 m shaped roadway and a 4 m gravel wearing course; and 100 km would be two-lane roads (PA2 standard) with an 8 m shaped roadway and a 6 m gravel wearing course of 15-20 cm thickness depending on soil quality. Standards are described in Table 1. In addition, the project was to provide maintenance of 250 km of low-trafficked gravel feeder roads not covered by the Second Highway Project, as well as initial maintenance on the project roads. 2.09 Only the "first-year" construction program was firmly defined, while "second" and "third-year" programs were indicative and subject to modification in the light of changing priorities in the agriculture sector. It was expected that construction would start with three work brigades in December 1976, and that each brigade would produce some 110 km annually. The firmly established first-year program consisted of 106 km of POA roads; 239 km of PAl roads and 7 km of PA2 roads. A mechanism needed to be developed that would ensure application of selection criteria acceptable to the Bank in the establishment of subsequent road improvement programs (para. 3.20-3.21). Project Cost and Financing 2.10 The project was estimated to cost US$8.5 million equivalent, net of taxes and duties. -Foreign exchange costs were estimated at US$6.6 million equivalent or 77% of total project cost. In addition to local costs of US$1.9 million equivalent, taxes and duties were estimated at US$1.6 million equiva- lent, both to be financed by the Government. A Third Window Loan of US$6.6 million was approved to cover 100% of foreign costs. III. PROJECT IMPLEMENTATION 3.01 July 29, 1976 was the expected date of Loan effectiveness. Inasmuch as the Government had supplied the requisite legal opinions well ahead of this date, the Loan was actually declared effective on June 15, 1976. - 21 - Establishment of BPP and ICC 3.02 The BPP was established legally by Decree No. 00447 of January 1976 which defined the organizational structure of the Directorate of Public Works as a whole and included the BPP as a subdivision of the Road Maintenance Division. As agreed during project negotiations, the ICC was established by Decree No. 76600 of June 1976. This committee included representatives of various Government and regional development organizations responsible for agricultural development, public works, and transport. Its task was to review annually the indicative rural roads program in the light of various agencies' development programs and to bring changing priorities to the attention of BPP. The program would be submitted to DSP for technical and economic review. The ICC's final recommendations for each annual construction program would be based on the results of these studies, which applied criteria acceptable to the Bank. Final approval of the annual program by the Prime Minister was required before it would be carried out. Project Start-up 3.03 Before project construction and maintenance work could begin, a number of preliminary actions were required. Equipment for the brigades needed to be procured; offices for the newly created BPP had to be con-tructed; technical assistance had to be employed to help establish the unit and support local staff during the early years of project execution and personnel had to be recruited and trained. Although initial implementation of these activities was satisfactory, difficulties soon appeared, when the requisite local finan- cing for the project did not materialize. 3.04 Bidding documents for project equipment were prepared at the time of project preparation and approved by the Bank in January 1976, i.e., before negotiations. Procurement of project equipment thus was timely, the bulk of it being delivered by December 1976 as foreseen at appraisal. However, three front-end loaders and six agricultural tractors and rollers, which were originally to be provided to the project from existing stock in DPW's Central Equipment Division, were not provided and had to be ordered in late 1976. Due to delays in contract signing and lack of counterpart financing, they were not delivered until November 1977; their non-availability seriously hampered start-up of works by the brigades. 3.05 Bids for construction of the BPP office headquarters were received in May 1976, but contracts were awarded only in April 1977, again mainly because of the shortage of counterpart funding. The building, which also was to accommodate DSP, was ready for occupancy by BPP only in November/December 1978. Until that time, project personnel occupied temporary offices at the Road Maintenance Division. 3.06 The contract for technical assistance to the project was awarded to a French consulting firm and the team, consisting of one road engineer (Chief of Mission) and one chief mechanical engineer, took up service two months after project effectiveness, on August 15, 1976. The initial contract foresaw a total of 48 months of consultancy input, mainly to help establish the new feeder roads unit, organize the execution of the initial work program - 22 - and provide on-the-job training for local personnel. A short-term assignment (three months) of an economist was also foreseen, to assist in preparing subsequent programs. 3.07 Recruitment of local personnel was, however, difficult. The ap- praisal foresaw employment of one road engineer (Chief of BPP), one mechanical engineer counterpart to the expatriate chief mechanic, and two administrative assistants at BPP headquarters; and 85 foremen, operators and mechanics for manning the three construction brigades and the maintenance brigade. Part of the contingency was to be transferred from DPW; new recruits were to receive training under the programs set up for maintenance personnel under the Second and Third Highway Projects. The training arrangement did not work as expected, because the project recruits could not be accommodated at the requisite time; thus training of project personnel was done initially by the main suppliers of equipment, and subsequently under the project itself. 3.08 A young Senegalese graduate engineer was recruited by December 1976 as Chief of BPP, but no suitable counterpart for the chief mechanic could be found before the technical assistance contract expired. Difficulties were encountered in the recruitment of brigade chiefs as well, mainly because Government rates for experienced technicians were well below those offered in private industry at the time. Recruitment of equipment operators, mechanics and foremen, of which more than double the number foreseen at appraisal were finally hired, took until mid-1977 to complete. Many of the recruits had only limited job experience and required extensive training. Their performance suffered also from delays (over three months at times) in salary payments. Many of the experienced or trained staff left the project during its first year of operation; BPP found itself still short of competent personnel in July 1978, when the initial technical assistance contract was drawing to a close. 3.09 Some of the above problems were again a consequence of the non- availability of counterpart funds, which was aggravated by the fact that payments to BPP personnel, initially handled through the Road Maintenance Division, were, on occasion, delayed. in an attempt to give the project at least a minimum of local operating funds, a revolving fund in an initial amount of CFAF 15 million was established with Government funds in June 1977. It took until March 1978 for it to become effective. Initially, it was not administered by BPP, and could not be used to pay contract staff. First Phase Execution 3.10 Two clearly distinct phases of actual project execution can be distinguished. The first one extended from July 1976 to the end of February 1979. By early 1977, BPP had acquired the physical means for becoming a functioning entity. Its failure to become effective, however, was a result of: (a) dispersed and confused administrative responsibility for different aspects of project execution; (b) academically qualified but inexperienced project leadership and inexperienced brigade personnel; and, most damagingly, from (c) the continuous shortage of counterpart funds resulting in slow equipment delivery, rupture in supplies of fuels and spares to the sites, loss of experienced personnel and low morale of remaining labor. Construction - 23 - output which, perhaps somewhat optimistically, had been expected to reach around 600 km during the first two years of operation stood at 140 km at the end of February 1979; some 220 km of other roads had been maintained during the last eight months of this project phase. Frequent supervision missions had drawn the Government's attention to the poor and even steadily deteriorat- ing performance of the project and suggested practical measures to reverse this trend, to little or no avail. 3.11 With little response from either the Ministry of Finance with respect to counterpart funding, or the Ministry of Works with respect to project organization and administration, the Government was informed in March 1979 by RMWA that, in the absence of decisive remedial measures to ensure conformity with the Loan Agreement, a recommendation would be made to Bank headquarters to consider cancellation of the Loan. The prospect of loan cancellation resulted in bringing the project to the Prime Minister's personal attention. Soon after, the problem of counterpart funding was resolved. An experienced DPW technician replaced the young engineer as project manager. He was given full administrative authority on all aspects of the project and the BPP, although nominally still a subdivision of the Road Maintenance Division, reported directly to the Director General of Public Works. An official note of July 6, 1979, issued by the Director General of Public Works and specifying organizational relationships between the different divisions in DPW, showed the BPP alongside RMD and stated that BPP "is directly responsible for its own administration"* This arrangement ended a more than two-year period of friction and unsatisfactory administrative dependence of BPP on the Road Maintenance Division (although a ministerial ordinance of September 1979 continued to show BPP as a subdivision of RMD in the reorganization of the Ministry). BPP was given control over its advance fund which was increased from CFAF 15 to 30 and subsequently to 50 million, when payment of salaries for contract staff out of this fund was authorized. Second Phase Execution: Road Works 3.12 Having at last acquired a workable organizational set-up, a modicum of operating funds and an experienced and competent manager, the project entered a second phase (March 1979 - June 1981), during which construction performance approached, and at times even exceeded, appraisal estimates. Over the 28-month period until exhaustion of loan funds, a total of 540 km of roads was improved, and initial maintenance was carried out on some 440 km of project roads. A number of actions by project management were instrumental in the achievement of these results. 3.13 In order to reduce the frequent periods of complete immobilization of the brigades because of a breakdown of one piece of heavy equipment (such as a loader or dozer), the initial three light brigades were reorganized into two "hezvy" brigades. The new head of BPP, being a seasoned DPW employee, and having direct support from the Director of Public Works was better able to obtain regional public works equipment, which often needed only minor repairs, to keep BPP brigades operating during equipment emergencies. Since equipment repairs by DPW's Central Equipment Division were extremely slow, a facility for repair and maintenance of project equipment by project personnel was set - 24 - up in parallel in available space at the Central Equipment Division; it also served as a base for regular training cycles for all brigade mechanics. (The DRM staff quite resented what they saw as preferential treatment to BPP.) Radio equipment was purchased to provide daily communication between head- quarters and brigades in the field. This means of contact became particularly important in the last years of the project, when failure of the aging equipment was becoming more frequent and timely provision of spares and expert instruc- tion to brigade mechanics were essential to keep up brigade productivity. Some of the equipment purchased for the project proved no, appropriate under Senegal field conditions and had to be modified at considerable expense. The truck fleet especially performed poorly despite continuous maintenance efforts. Except for the personnel carriers, the fleet was written off completely by the end of the project; during the last year of operation, it functioned mostly with reconditioned vehicles recuperated from among existing DPW equipment, and with private truckers doing the hauling. 3.14 To provide the necessary technical backing to BPP during this period, the Government requested, and the Bank agreed, to continued technical assistance. The contract was extended first for a ten-month period from March 1979 through December 1979; and subsequently for another eighteen-month period from January 1980 through June 1981. Throughout this time the team consisted of one road engineer (Chief of Mission); one mechanical engineer, responsible for spares procurement and supervision of repairs on site and in the central workshop; and one mechanical engineer/trainer. ThILs small team worked very effectively with rather than alongside, or instead of, their Senegalese counterparts and contributed much to forging the BPP into a well-functioning unit. 3.15 The actual program of rural road construction executed under the project differs in several respects from the indicative program established at appraisal on the basis of 1973/74 consultants' surveys and reports. Firstly, the project's regional coverage is wider than foreseen at appraisal: the ICC's suggestions resulted in the inclusion, in the latter years of project execution, of roads in the Fleuve region; in the groundnut and Casamance regions, some proposed sections were dropped and new ones intro- duced. Of the total 680 kilometers constructed under the project, 400 km or about 60% were roads identified for priority upgrading at appraisal; the remaining 280 km were among those subsequently evaluated (Table 2). 3.16 In addition, maintenance was carried out on three non-project roads in 1978, and on 15 of the 33 roads improved under the project during 1979-1981, Secondly, the characteristics of the roads constructed differed from those proposed in the appraisal report. None of the improved sections were left at POA standard. In fact, all roads were constructed to a basic two-lane gravelled standard between the PAl and PA2 standards defined at appraisal (Table 1). The 97 km Kolda-Velingara Road, the main section included in the appraisal program for improvement to PA2 standard, was paved under a diffe:ent program. Of the 400 km of roads to be summarily improved (POA standard), well over half were re-evaluated during 1980 and replaced by other road sections proposed by ICC that had a higher rate of return. The majority of the remain- ing POA sections were located in regions with sandy soils, and continuous gravelling was necessary on most of them to ensure year-round viability. - 25 - The choice of a uniform gravelling width for all project roads evolved empirically: correct application of different standards is not easily con- trolled by project management, and uniform standards, once brigades have been trained to use them, will result in higher construction outputs. The study of the technical and economic feasibility of constructing rural roads in Senegal by intermediate technology methods and comparison of them to the equipment- intensive methods which had been adopted for the project was dropped. 3.17 Thirdly, no new structures were built, and the brigades carried out minor culverting works mostly with recuperated material and temporary bridges using local wood and stone. Civil contractors were used only for laterite hauling: on road sections where the hauling distance was more than 45 km, project vehicles were not sufficient to provide materials as required. While haulage was easily contracted out, there were very few local contractors in Senegal capable of carrying out construction, and the project could not provide the close supervision they would have needed. Effective supervision would have required additional BPP headquarters staff, and, under the circum- stances, project management correctly decided to concentrate on improving project brigade efficiency first. 3.18 At BPP headquarters, additional administrative staff were employed and up-to-date financial and cost accounting introduced. Progress reporting, introduction of new contracts, and processing of withdrawal applications were handled accurately and expeditiously. Annual Construction Programming 3.19 An important aspect of the project was the proposed redefinition and evaluation of indicative annual work programs which was to have been carried out jointly by the Borrower (ICC, DSP) and consultants. DSP, while practically without local staff other than the Director, disposed of some 144 man-months of technical assistance for transportation planning under the Third Highway Project that would, inter alia, carry out the bulk of the programming work, while the project itself provided for only three months of an economist's support to assist BPP. Since DSP was mainly devoted to preparing a national transport plan when the rural road evaluation was finally required, a somewhat more extensive economist input than originally foreseen (5.5 months) had to be financed by the project. 3.20 Due to that three years went in carrying out the firmly established "first-year" construction program, the program review process involving ICC and DSP was initiated at the end of 1979 only. DSP invited the Governors (regional authorities) to submit lists of priority road sections for upgrad- ing. From over 2,000 km submitted, some 900 km were preselected and examined by DSP and ICC, who agreed on 600 km to be retained for the program. However, the selection was based on summary screening criteria only (groundnut traffic flows and population in the road influence area). At the Bank's request, an economic evaluation was subsequently carried o- by technicians financed under the project. This review included the 600 km J2 roads) retained by tl-! DSP and ICC, 320 km (10 roads) submitted by the Governors and rejected by the ICC, and 245 km (6 roads) from the original 1974 program and proposed to be dropped. - 26 - The final 1980-81 program for implementation was established from the results of this evaluation; only some 500 km showed an economic rate of return above 10% and were retained. 3.21 A comparable exercise was carried out in the Fall of 1980 to determine an overall program for a proposed follow-on feeder roads project. To avoid the long process of eliminating purely "political" candidate roads and focus more directly on the needs of agriculture, regional development authorities rather than Governors were invited to submit proposals. For a number of roads that were to serve specific agriculture projects, economic benefits were measured in terms of net incremental value added, attributable to both agriculture and road investment. Again, the economic analysis was carried out by the consultant; but since the various regional development organizations had evidently come to accept the economic evaluation process in the establishment of their respective road improvement priorities, the Director of DFW had arranged for a local economist to be employed by DSP for execution of this work in the future. Study of Road Impact 3.22 To provide more reliable information on the socio-economic impact of rural road improvements on the populations served, the project included a study for monitoring the performance and impact of a sample of roads constructed under the project. It was to be carried out by consultants over a three-year period. Supervision of this component vas to be the responsibility of DSP. The study contract was awarded tG a local consulting firm in 1977; as the loan agreement only provided for financing of foreign costs under Category 2, it was amended in October 1978 to finance 90% of the total cost. Contract signing, however, did not take place until March 1979. 3.23 During the difficult start-up years of the BPP, the study was of minimal ccncern to the subdivision and a low priority item for DSP. Thus, in designing and carrying out the first phase of the study (baseline surveys), consultants had no effective inputs from DSP. Field work for the base surveys was carried out according to the contract schedule and a draft report on the results was presented in October 1979. 3.24 On review by ti-e Bank it was found that the study had focussed on monitoring the agricultural production impact of three roads, all of which served areas already intensely cultivated and where road improvement had been econnmically justified on the basis of vehicle operating cost savings to normal traffic. Given the major annual variations in agricultural production (Table 3) due to the varying intensity and timing of the rains, the monitoring of minor production changes due to improved access to market would have required much more intensive study coverage than actually provided. Following discussions between DSP/BPP, the consultant and the Bank, it was therefore agreed to reorient the study to what could be expected to be measured with acceptable reliability--i.e., the traffic effects of the road improvement. Traffic counts on the selected road sections were to be carried out three times per year over a three-year period; a brief summary of results was to be presented at the end of each calendar year. The household and village - 27 - surveys in the road influence areas were also redesigned to concentra:e on a smaller number of more reliable measurable indicators of change, and the consultants agreed to re-analyze the baseline surveys in terms of the newly chosen indicators. 3.25 Cooperation between consultant and DSP/BPP in the redefinition of the study was good; the traffic counts were carried out on schedule, and the updating and reworking of baseline results initiated as foreseen. Difficulties and delays occurred, however, in processing the agreed contract amendment reflecting the changes in study coverage; it was approved in :March 1981 only. In the absence of payments for substantial services already rendered, the consultant was deficient in producing reports: results of the traffic counting had to be collected in preliminary form by the last Bank supervision mission, and the re-draft of the baseline surveys has not as yet been received. Limited socio-economic analysis can be expected; at best, the study will provide information on traffic for a small sample of rural roads. Implementation Schedule 3.26 By the end of 1979, the original target date for project completion, the project had constructed the BPP offices, procured all brigade equipment, and improved some 380 km of roads. A total of US$4 million equivalent or just under two-thirds of the loan proceeds had been disbursed at that time. Loan closing dates were deferred to December 1980 and subsequently, to December 1981, however, all loan funds were committed well before June 1981 and the final payments made on September 16, 1981 (Table 4). The actual period of project execution thus was five years rather than the three-and-a half years foreseen at appraisal; and a total of 680 km of roads and tracks were improved, compared to the 630 km to standards PA and 390 km to be summarily improved, as appraised. Project Costs 3.27 Appraisal estimates and actual project costs (at current prices) are summarized in Table 5. The road improvement program carried out differed considerably from the indicative programs established at appraisal and cost 23% more. The local coitribution to total project cost was more than double the original estimate. 3.28 Actual costs of equipment purchased for its operation were lower than estimated: US$2.3 vs. $2.7 million equivalent for equipment, but some equipment was rented which was not foreseen; US$3.0 vs. US$3.4 million equivalent was spent for spares and supplies; materials purchases were minimal and amounted to only a third of appraisal estimate. Almost half the amount foreseen for drainage and culvert works by contractors, which were not con- tracted or not done was paid to contractors for hauling materials. The office building for BPP was constructed at 60% of the cost estimated at appraisal. 3.29 By contrast, personnel costs were considerably higher than foreseen from CFAF 198 million to CFAF 550 million: the cost of local head-office personnel was estimated at CFAF 11 million annually for a period of three - 28 - years; actual costs averaged over CFAF 16 million annually for a period of five years. Brigade personnel costs were estimated to average about CFAF 55 million per year for three years; actual expenditures were just under CFAF 94 million annually for five years. Partly, the difference is explained by higher than estimated increases in salary levels; minimum salaries were increased several times during the implementation period, and last by a substantial 25-30% in January 1980. But mainly the difference is due to much higher staffing levels than originally proposed. The appraisal provided for unrealistically low numbers both of administrative and brigade personnel. Actual office staff of around 40 people (including messengers, gardeners, building guards, etc.) and brigades' staff (including hand laborers) averaging 100-120 each are perhaps generous. 3.30 Proportionally even higher was the cost overrun for technical assistance: actual cost of US$1.73 million equivalent compared with an appraisal estimate of US$0.5 million. This reflects the fact that man/months provided to BPP was 139.5 rather than the 48 foreseen at appraisal. The cost per man/month (including overhead and all local allowances) also increased considerably during the latter years of project implementation and averaged US$12,400 equivalent. The cost of the road impact study (US$320,000 equiva- lent, actual) had not been detailed in the appraisal report and it is not clear whether it was included under the "technical assistance" category. 3.31 In summary, the project carried out construction of 680 km instead of construction of 630 km and summary improvements to 390 km and a similar proportion of its maintenance objectives at 123% of the estimated cost, taking 142% of the originally foreseen project period to do so. The actual overrun of US$1.9 million equivalent is mainly attributable to the lengthening of the project implementation period which, in turn, was caused mainly by insufficient and irregular local financing during the critical years of project start-up. Not having provided timely requisite financing initially, the Senegalese Government in the end contributed twice the original estimate. Consultants and Borrower Performance 3.32 Consultants to BPP during the first project phase (July 1976 February 1979) performed as satisfactorily as circumstances permitted: due to lack of counterpart funds, absence of trainable counterpart personnel and poor administrative arrangements for the newly established BPP, they were not in a position to carry out fully the tasks assigned them in their terms of reference. By contrast, the three-man team of technicians assigned to the project during the second project phase (March 1979-June 1981) operated very successfully in an organizationally much improved, if not perfect environment. Towards the end of the project, technical assistance and local counterpart personnel worked as a well-integrated team whose overall level of competence had grown considerably over the project period. 3.33 Short-term consultant services provided for assisting in the develop- ment of an annual program evaluation scheme and in the preparation of a second fecder roads project (5.5 man-months in all) were timely and of good quality. - 29 - 3.34 Consultants carrying out the socio-economic impact study gave a somewhat mixed performance. Without seeking guidance from either BPF/DSP or the Bank, on their terms of reference, they embarked on an ambitious base-line surve7 which could not possibly provide short-term measures of road impact over a three-year period. Following a Bank review of base-survey results, the thrust of the consultants' work was changed to concentrate on traffic surveys, and transport availability and cost, on sample roads. The field work was carried out satisfactorily. However, reporting has been very poor and no more than preliminaL, results have been communicated so far. I/ 3.35 The Borrower's performance under the project varied from very bad to good over the five-year project implementation period. All procedures up to project effectiveness were handled with exemplary dispatch. But once given existence on paper, the newly established BPP for more than two years received neither the organizational support inside DPW, nor the necessary counterpart funds required to carry out its mandated tasks. Only the prospect of project cancellation finally redressed the situation which then remained satisfactory through completion of the project. IV. INSTITUTIONAL DEVELOPMENT 4.01 Among the main objectives of the project was the creation of the capability, inside DPW, responsible for the evaluation, programming and execution of rural road construction and maintenance (para. 2.05). As in the majority of countries in the West Africa Region, main road construction and improvement in Senegal had been pursued vigorously after Independence and absorbed all the scarce manpower and funds of the road sector. The responsi- bility for planning and implementing rural road i-nprovements was scattered among mult various regional/local institutions and development projects, all working to their own priorities, standards and schedules, and none suitably equipped to ensure maintenance of the roads they constructed. The project--by creating and financing a feeder roads unit in DPW--was to provide a techni- cally competent unit to be responsible for all rural roads works. The BPP and DSP were to establish an inventory of existing rural roads and tracks, deter- mine their adequacy for the traffic they were expected to support, and define the road improvements needed to eliminate transport bottlenecks impeding agricultural and rural development. The DSP would establish annual construc- tion programs using consistent criteria to determine the economic viability and development priority of proposed road improvements. 4.02 The preceding section on project execution has implicitly traced the performance of BPP as an "institution". For an initial two-year period when DPW assigned neither sufficient funds nor any experienced manpower to its newly acquired task of rural road construction and maintenance, the BPP 1/ We are not certain whether they have finally been paid, a factor Lhat was holding up completion/release of reports. - 30 - appeared condemned to flounder. Only after the unit became administratively autonomous under an experienced local manager and was given reasonably adequate operating funds did it also become productive. With some additional equipment and personnel selected by the new manager, output over the next 28-month period reached 540 km of road construction, and initial maintenance was carried out on all project roads as required (some 440 km). 4.03 Two ingredients can be identified as instrumental in this trans- formation: (a) direct control by the unit over a minimum of operating funds (local cost revolving fund for salary and small operational expenditures); and (b) local management experienced in both the technical and administrative aspects of the job, who set production goals, clearly communicated them to staff (all contractual), followed up on performance, and did not hesitate to remove poor performers. The unit thus effectively operated more like a "contractor", oriented not to making a profit but to achieving a maximum amount of construction output with given funds. 4.04 The BPP did not become involved in other feeder road projects being executed at the time. The most important of these was the UNSO-sponsored construction of the 26 km Sambadia-Djiffere road in Sine-Saloum; at the time the UNSO-brigade was being set up (1977) it was foreseen that it should be constituted as a fourth brigade of BPP. However, in the early difficult years of BPP's establishment, the UNSO brigade started functioning separately, under expatriate management. It built roads of much higher standard than the roads constructed under the project, at a per km cost several times the one under project. With the completion of this initial program, it could have been integrated into the then efficiently functioning BPP; this was not pursued, however, since the Government feared to lose the financial support (by Dutch and German aid) to the brigade, if it were taken out of the UNSO framework. 4.05 BPP brigades had succeeded in attaining the expected yearly output. However, since September 1981 when project funds ran out, they functioned for a while at a reduced level with financing from the Road Fund, and in 1982, as the Road Fund has not been replenished given the overall country budgetary crisis, they came to a stop as did the rest of the Road Maintenance Division activities. The BPP brigades have been nominally reinstated under the RMD, to execute both feeder road construction and periodic maintenance of main roads depending on the priorities and funds. A second feeder roads project, requested by the Government, was not pursued, as the overall decline in agricultural production made it difficult to justify more feeder roads when the existing ones were underutilized and not maintained; it is expected, though, that the brigades could assist in feeder road works needed by future agricultural projects in new areas. These works, however, would require major equipment renewal and continued on-the-job training for some categories of personnel (mainly mechanics). The Sixth Plan (1981-1985) documents assume that BPP is to carry our road infrastructure programs for the various regional development projects planned throughout Senegal. - 31 - 4.06 The planning capability had also developed. DSP had made progress in a continuous updating of the rural road inventory and in evaluating annual improvement priorities. Even the ICC, while a somewhat cumbersome mechanism, had proved valuable in providing DSP/BPP access to, and some influence over, the various institutions concerned with rural road improve- ments. It provided a forum for discussion and agreement on consistent screen- ing procedures needed for establishing rural road improvement pricrities--and, on balance, worked well and was accepted by regional authorities. 4.07 A local consultant also benefitted and acquired some experience from the project. Domestic contractors were not employed as foreseen; their development should be pursued under other projects, as the potential looks good. An agreement, spelled out in a side letter to the Loan Agreement, by the borrower to explore the possibilities of obtaining voluntary contribu- tions in cash or labor from rural communities for feeder roads, was abandoned. As long as the project was functioning badly, it was not feasible to sollicit local contributions: and when it later worked well, the uncertainties sur- rounding the future of the BPP effectively precluded any initiative in this respect. V. ECONOMIC RE-EVALUATION 5.01 The indicative three-year road construction program covering some 1,000 km of different standard roads had been evaluated at appraisal to yield an average economic return of 12%; the maintenance of some 250 km of non- project roads was estimated to yield over 20%, and the project as a whole would have an economic return of about 14%. Benefits taken into account were mainly savings in vehicle operating costs to normal traffic, growing at 4% p.a. Only for two roads, which were to serve planned vegetable growing areas not previously accessible by road, a combined return to the agriculture and road investment based on net incremental value added was calculated. These roads, totalling 36 km, were part of the second-year indicative program and were not included in the actual construction program. The main beneficiary of the project was expected to be ONCAD, the organization then charged with the marketing of groundnuts and with provision of inputs such as improved seeds, fertilizers, pesticides and equipment to area farmers. The farmers would benefit mainly by more reliable access to markets rather than by cheaper prices of inputs and higher prices for crops, since agriculture farm-gate prices are officially and generally uniformly set by the Government. Benefits from institution building were considered to be substantial, although they were not quantified for inclusion in the economic calculus. 5.02 The same methodology as at appraisal was used in the re-evaluation. Due to inadequate cost accounting for the early project years, road construc- tion and maintenance costs during the five-year period of project execution could not be satisfactorily separated and a joint return to both activities is estimated. The economic analysis includes all costs incurred by the project with the exception of those for the impact study. The very heavy technical - 32 - assistance costs associated with the first unproductive phase of the rural road improvement program have thus been fully charged to the project. The substantial residual value of the equipment expected at appraisal turned out on the contrary to be quite low at project end. 5.03 Based on actual project experience, maintenance requirements over the rest of the project life (10 years) are estimated separately for two climatic zones. Routine maintenance is assumed to cost CFAF 40,000/km/p.a.; spot regravelling, estimated to cost CFAF 260,000/km, would be required every third year in heavier rainfall regions mostly in the southern half of the country and every fifth year in the dry zones in the north. 5.04 Traffic data were derived from DPW 1975-1981 traffic counts for secondary and regional roads and from the 1980-61 traffic counts on the project roads covered by the impact study (for details of the assumptions underlying the economic analysis, see Table 7). The pattern is one of steady, and on most regional and departmental roads, very steep, trafric growth (doubling to quadrupling) between 1975-1978, a slight decline in 1979, and a steep decline (down to 1975 levels in many cases) for 1981. The 1981 decline was the result of a dramatic drop in that year of groundnut production below the 300,000 ton level, compared to average annual production of around one million tons. Traffic patterns were further distorted by tne dissolution of the giant state marketing agency, ONCAD, in the Fall of 1980. ONCAD's transport responsibility was to be devolved to a number of successor agencies, but confusion and delay ensued, and input distribution as well as crop collec- tion were severely disrupted. The 1981/82 crop season has been a good one, however. Groundnut production is expected to be around the 800,000 ton mark, and in December 1981, crop collection was betng organized quite efficiently directly by the oil mills, relying on private, hired transport exclusively. Only the provision of inputs to farm cooperatives is to be handled by one of the ONCAD successor agencies, SONAR, who has inherited the best part of the ONCAD vehicle fleet. Thus, for the 1979-81 drought years, traffic is taken as constant at tne-low-1981 level when production of main products was about half the usual tonnage, although traffic did not lecrease as much; in 1982, traffic is assumed to grow at 10% to reflect the upswing in agricultural production in the 1981/82 crop season, also due to better prices to farmers and hopefully, better organized transport; 1983 growth is assumed at 7%; for the remainder of the project life, only a 4% growth rate (as at appraisal) has been used to account for the likelihood of lower traffic, drought years. 5.05 The economic analysis distinguishes between zones for which vehicle operating costs are different: the dry/sandy north (Fleuve, Louga, and northern Thies); the groundnut basin (Thies/Diourbel/Sine Saloum) and northern Senegal Ori!ntal; and the south (Casamance and southern Senegal Oriental). Vehicle operating costs used in the analysis are those prepared by consultants under the project. For existing roads on which maintenance only was carried out in the early years of project implementation, full vehicle operating cost savings are taken for the first year after maintenance was carried o,it. For the low trafficked road in eastern Senegal, deterioration would take place without maintenance over a five-year period and reduce vehicle operating - 33 - cost savings to a fraction of those in the first year as follows: year 2, 90%; year 3, 70%; year 4, 40%; year 5, 10%. For the heavily trafficked road sections in the Thies region, deterioration would occur over a three-year period as follows: year 2, 75%; year 3, 40%; year 4, 10%. Future mainte- nance of the roads would be the responsibility of the regional maintenance authority, as the BPP is integrated into the general RAMD, and it is hoped that works will be carried out to prevent deterioration, at least on the heavily trafficked roads. 5.06 With these assumptions, the project is estimated to yield a global economic return of 14.5%, practically as at appraisal (Table 8). This is due to various compensating factors: traffic is much lower; unit vehicle operating savings have more then doubled the 1975 levels, as roads deterio- rated further and operating costs increased also with inflation; and construc- tion costs were some 50% highe.* than the appraisal estimate. If only half the cost of technical assistance were charged t .ne project--on the assumption that institution-building costs should be spread over a longer-term program horizon--the project's rate of return woald rise to 16%. There is a high probability that traffic will be at least 50% higher than assumed levels, should peanut production return to previous levels; if so, the return would be considerably higher. However, there is a risk that maintenance will not be performed since DPW does not have the institutional financing capacity to assure this will be done, and if so, the economic rate of return would be negative. Assuming that the roads would deteriorate at the rate indicated in para. 5.05, the economic rate of return would be -11% and the investment wasted (Table 9). 5.07 The fact that only one standard -aa applied to all roads, irrespec- tive of traffic, could mean that the cost of some roads was more than can be justified. Of the 400 km that were to be summarily improved, only 80 were improved, to higher standards and in the areas with higher traffic; but 217 km were rehabilitated in Casamance, where traffic in general is lowest and vehicle operating savings per se do not justify the inveitment. However, it is premature to draw a conclusion on the effects and justification of the roads, given the lack of appropriate traffic data and the possibility that agriculture may recover and new projects be started. It is recommended that for this type of project, where the impact may be fully appree l.ed only some years after its completion, the PCR be postponed, or followed up some years later. 5.08 Benefits from road improvement would primarily accrue to the oil mills or to some parastatal agencies; the farmers would mostly benefit from cheaper and faster personal transport, as tariffs for groundnuts and their t-ansport continue to be administered. The rates have been simplified - 34 - compared to earlier schedules applied by ONCAD; 1/ as the rates now include a fixed charge per ton, it is difficult to say whether transporters are paid more or less than before in real terms however, the variable charge seems to follow closely the operating costs, including taxes, on the various road types, established by DSP (which is consulted for the rate levels), so that the transporters would not benefit much from the road improvements. VI. BANK PERFORMANCE 6.01 The Senegal Feeder Road Project is the first one to have been executed from a Bank field office--the Resident Mission in West Africa (RMWA)--from preparation through completion. During preparation and super- vision, proximity to the Borrower has had the obvious advantage of making frequent and close consultatious possible. Performance in the later years of project execution indicates that the institutional arrangements (BPP and ICC) were correctly defined and appraised. The fact that all administrative actions required from the Borrower before loan effectiveness were accomplished on, or ahead of, schedule, points to the conclusion that the new project concept developed for rural roads was appreciated and actively supported by the higher level planning authorities in the country. However, it seems that the Bank did not realize that this appreciation was initially not equally strong at the working level in the Ministry of Works (MOW), or when funds were required. The willingness and perhaps the capacity of the MOW to accord the project sufficient priority were over-optimistically assessed. 6.02 When it became clear that the project was in trouble and was unlikely to achieve its objectives, the Bank took forceful and, in this case, success- ful action at the right level in Government. 6.03 Supervision of the project from RMWA averaged between three and four visits a year throughout the project. This above-average intensity was justified given the difficulties in project start-up and the institution- building nature of the project. The same amount of supervision of the Third 1/ Instead of 10-12 different tariffs based on road conditions and trip distance, there are now only three basic distinctions: Paved Road Improved Gravel Earth Road 1974 Rate per ton/km, CFAF to collection centers 17 22 28-50 Rate per ton/km, CFAF to oil mills 12 17 - 1981 Fixed charge (CFAF/ton) 520 520 520 Variable charge (CFAF/ton/km) 25 28 44 - 35.- or Fourth Highway Project was provided simultaneously from headquarters. Also, the Bank's resident mission in Dakar proved to be a most useful permanent contact point between the Bank and the project. There is no evidence to suggest that additional supervision would have influenced project execution to any significant degree. 6.04 With hindsight, the technical assistance requirements of the project would seem to have been underestimated at appraisal. The cost of technical assistance might have been kept somewhat lower, if the Bank had insisted on negotiating a new contract rather than (for reasons of the Government's administrative expediency) approve a number of amendments to a 1976 contract with a high initial unit cost and a price revision formula which turned out to be quite favorable to the consultants. 6.05 The project was conceived as the first of a series, since, as is the case of highway maintenance projects, institutions cannot be expected to be consolidated within the period of a single project. A follow-on project was included in the lending program until early 1981, and the Government had done some preparation and submitted a formal application for a follow-on project. Based largely on the poor performance of Senegal's agriculture sector, a further feeder roads project per se cannot be justified at this time; priority of the scarce resources should be given to maintaining the existing network. Therefore, the follow-on project was deleted from the lending program (the built-up capacity was integrated into the Road Mainte- nance Division, where it will be assigned priority work - para 4.05), but future Bank agriculture projects will include feeder road components as justified, eventually to be executed by force account, and the Bank is also considering assistance for-road maintenance, including feeder road maintenance. VII. CONCLUSIONS 7.01 After a seriously inadequate start, the project constructed 600 km of 630 km foreseen of 7-8 m width, improved 80 km of 400 km foreseen and maintained some 220 km of other roads. The institutions and consultative arrangements (BPP and ICC) functioned reasonably well; and the economic viability of the project works has been confirmed with an overall 14.5% rate of return. The Government is intent on continuing rural road works through BPP as soon as financing can be found, and has been active in eliciting interest from major donors (Paris Meeting, October 1981 and Dakar Meeting, March 1982) for the provision of assistance to feeder roads programs and the sector as a whole. The Bank is considering assistance to the maintenance program, of highest priority in the sector, and which the Government is unable to finance. 7.02 Through this project and the Fourth Highway project appropriate assurances had been obtained that the project would receive adequate support to achieve its original purpose. Extremely unfavorable economic circumstances have prevented--directly (lack of justification for new roads outside an agricultural project), and indirectly (lack of financing)--fall realization - 36 - of this objective. As the agricultural sector in Senegal recovers and begins to increase its efficiency, there will be a continuing need for rural road improvements. 'the institution created to carry them out may, however, have undergone a declikne, thus requiring some effort to restore it. - 37 - tablel1 SENEGAL FEEDER ROADS PROJECT (Loan 1221-T-SF) PROJECT COMPLETION REPORT Desizn Standards : At Appraisal and Actual Apraissl Report Act-al Tracks Roads Roads without with some one two two improvement improvement lane lanes lanes (piste ordinaire) (piste o.amf1iorfe) Designation PO POA PA 1 PA 2 PA Yearly traffic (in tons) 3,000 6-7,000 <.20,000 20,000 + 6-20,000 Vehicles per day 5 10 20 20/30 Rcadway width (in meters) 6 7 8 7-8 Gravelled surface width spot (in meters) gravelling 4 6 5,5 Thickness of road materials (in cm) n.a 12/15/20 15/20 12/25 Drainage lateral lateral lateral lateral ditches ditches and distches and ditches and paved fords culverts culverts Design speed n.a n.a 60 km/hour n.a 1/ Construccion cost - per km (force account) sandy soils n.a US$14,000 Us$12,250-28,000 other soils IS$1,400 US$3500-5,250 USS 8,750-12,250 Yearly maintenance cost per km (routine and 1/6 of periodic) US$130 US$300 US$350 US$:320 1/ Ai of December 1975, without taxes. Scurce: Government, consultants and mission reports. &―너‘「『「「 샅 , &. - 39 - Tatle3, SENEGAL FEEDM ROADS PROJECT (Loan 1221-T-SE) PROJECr CMIPLETION REPORT Agricultural Production, 1973174-1980/81 Un Thousand tons) Crmp 73/4. 74/5 75/6 76/? 77/8 78/9 79/60 80/al Groumdruts-InFi. 675 994 1,450 1,182 586 9w 737 2-300 Groundruts-Cons. 18 18 25 14 8 13 9 6 Sorgho - Millet 511 ?80 r77 SO? 416 795 495 552 Mats 32 43 49 42 32 47 45 54 Paddy 64 117 1115 los 62 12B 113 so Cotton 33 42 31 45 37 33 27 22 Tomatoes 20 20 Gr,een Beans 6 6 Ortiorts 10 10 Fertilizer Dis- tributed 97 104 as 103 Estimated figures fr-om various sources. Source: Governtowt, covaultants and mission rei3orts. - 40 - Table 4 SENEGAL FEEDER ROADS PROJECT (Loan 122L-T-SE PROJECT COMPLETION REPORT EstLimated and Actual Schedule of' Disbursements PY1977-1982 IBRD CUJLATIVE DISBURSEMENIS AT END OF QUARTER FISCAL YEAR - US Thousand - AND QUARTER ESTIMATED ACTUAL FY 77 Sep 76 200 - Dec 76 1,200 - Mar 77 2,700 400 Jun 77 3.500 1,100 FY 78 Sep 77 3,800 2,000 Dec77 ,100 2, 100 Mar 78 4,500 2,30, Jun 78 4,900 2,400 FY 79 Sep 78 5,300 2,700 Dec 78 5,700 3,200 Mar 79 6,000 3,400 Jun 79 6,300 3,600 FY 80 Sep 79 6,600 3,700 Dec 79 4,100 Mar 80 4,500 Jun 80 5,000 FY 81 Sep 80 5,300 Dec 80 5,900 Mar 81 6,100 Jun 81 6,400 FY 82 Sep 81 6,600 Source: Government, consultants and miseon reports. - 41 - Table 5 SENEGAL FEEDER ROADS PROJECT (Loan 1221-T-SE> PROJECT CO1PLETION REPORT Estimated and Actual Allocation of Loan Proceeds (US$ equivalent) Loan Agreement Loan Category July 19 6 Actual I. Equipment, spare parts, supplies and materials 5,000,000 4,865,775 II. Technical assistance 400,000 1,4571025 III. Civil works by contract 400,000 277,200 IV. Unallocated 800,000 TOTAL 6,600,000 6,600,000 Source: Government, consultants and mission reports. -42- TABLE 6 ~gM tOAS PRCUECT (toan 1221-T-SE) PaDJECr C~@LTION REP0RE A1ssal Est±ates5 pr.: actufal Q:'ee: C0:2 (Corrent Prices ) ApparaieL Estimat Actual Ccit .ocal Foreign Total 4ocal Foreign Tctal Fnrc.-;' as rctal Cs ts CPAF million USt 00 St OCO -I CFAR rillicr uSt 000 LS. CCO ,t cf rital . n ÄrrfUsä equivalnert CCst Estirate A. mffc -ff e , 014 678 må 222 4 166.6 2.Sc2,2 .Älding 37.4 122.2 206.5 14:. 109.4 1, OficA Eruimfflent 4 Coration 6,9 9. 36.4 12.0 '4.3 67.7 22 'M Por%onml 33.1 - 147.1 82,9 .35.2 . tenia sicre13.0 44''.6 SOS.3 63.a 1.445.1 i.27W.2 64 24.2 ImPaCt: stýdy . - - -42.7 %3.2 3._ ::1ie . C, Rodd Contructiorn Sr.d antnarca 314,0 .,635.6 7.031,1 827.2 4.7j.1 . c-, EQUiVent Purhae 53,4 2,492.9 2,73.2 46.6 2,I03.7 2,262.9 91 EcuiPtnent Pantal - - - 2. - 126.7 . Loplies and Spaas 74.2 3,090.7 ,420.4 66.9 2,717.4 3.0192. 5ersommil 162.9 - 724.0 459.7 - 2,CES2 - 26 V<trials 23.5 52,0 156.5 11.1 34.7 -3 Vtilc. Czst aS. IL. Eq, Trupt. - . 4.1 - 18.1 6 D,Cl Rork.s Contracted 21.9 :3e, 46,7 19,7 154_4 2464 S4 Acuno.425 6.600 6.500 870 8,600 10, 430 22 78 37 63 100 Gýovrnmwnt contriution for EPP ouilding anly 2/ Head officA pErSonnel coat uatinatad at 15 % of total oersonl oSta, based on brCakdcwn for csole months. 3/ Incluse part of mats *li1g1blm for Bn financing, but rot Paid dua to exaustion of loan precands 4/ At wstiPettd excdange rate of USG 1 - CFAF 225 At avr excha~gQ rate of USS 1 . CFAF 222. Rata& actually fluctuated betwe~r CFAF 200 - 285 ouring 9-g - 1961 rhnce tne ciffererc-, in doQllar ~vnts up~re Per expe,dtturc catugory to those Boen r Table 5. 5ource: Govear~t. consultmnte and *isLaon reprt. l1 - 43 - sEMEGAL IEDER ROADS PROJECT (loan 1221-T-SE) PREzCr COmPLETION REPORT Assumptions tsed in Re-.valuatina Project Roada ZONE A / ZONE B ZONE C 7 Project Roads Constructed and Maintained (ka) 82.0 38L.5 217.9 other Roads Maintained (ka) - 45.0 173.0 State of Track Vithout ProJe4t Year 1. (Sumary) , Pa 100% 335 50% (ordinary) Po - T7 50% Year 2: PS 100% 50% 70 PO - 50% 30% Year 3: PS 100% 70% 100% P0 - 30% - Year 4: PS 100% 100% 100 PO -- Yebtale Operating Cost PLeAsener 7edole TALttw k Media Truck Savins (Dee. 1980 prices) CFAP/V*h/ft CPAPVeh/0f CPAP/Veth 100% PS ) PA (gravel) 47.58 99.13 165.9i TOS PS/30% PO- PA ( " ) 39.55 79.43 130.94 50% PS/50% PO- PA ( " ) 34.20 66.26 107.57 33% PS/5T% PO--PA ( " ) 29.73 55.28 88.06 Road Maintenance Cost zone A Zone 3 7one C Current maint.(CPAF/ka/year) 40.000 40,000 40,000 Spot reagravellin (CPM/ks) 260,000 260,000 260.000 every fitth year every third/fifth yr. every third year Reduction in voiteope. savins if no maintenace Year 2 75% 902 Year 3 402 Intermdatae 602 Year . 102 302 Year 5 102 Average 1979/81 Daily Traffic Passenger Vehicles 3 T 2 Trucs ./ 8 121/ 4 kaned Roads Ps...enger Vehicles - 28 3 Trucks i/ 421 6my Traffic Growth: 1982t 102 1983t 72 1984 S 4% I/ Fleuve/Lougs/North of Thies. ?/ Thies-South/Diourbel/Sine Saloum/Nor-th of senegal Oriental. 3/ Senegal Oriental-South/Cassmance. -/ Truck traffic is assoned to consist of 30% light trucks (8 tons), 70% medium trucks (10 tons). / reliminary results of the Impact survey gave 37 and 25 vpd on two ample roads. Estiate for sone adustad douad based on agricultural production data and aistion field eaperience. 6/ afic counts for one road eotion show 157 vpt for 1978 end 90 vpe ror 1981. N"o 0o"11t wre available for the second setion, and since it seres aler SocGlity, It is snsumd to be loer. The estieor ot 7 vpd retained for the analY1111 ia consermative. Sourcet Government, consultants ankd msion reports. - 44 - TABLE 8 SENEGAL. EDnR READ PROJXT Lca- 1221-T-BE) PRJ~CT =PEUTTI REPON T Econodic Re-avaluatton of Ytader R6ad- lfprötgd Under the Projaet Assuming Adequata Maintannes (CFA7 Millian In Dec. 1980 prtces) Kl O7MEs 00T I ITS Main- Constructed and construc Subacaunt Matenanceo tained Maintalned tion £ 298 383 173 TOTAL .£Xatat dd only BENE- BENE- ZONES Inteal km km km COSTS A B C - C VITS FTTS å c- A ,Bl 32 1c mainten. 1976 - -6,0 - - - - 761.00 - 761.00- - - - . -761.0 1977 - - 94 - - - 497.80 - - - - 497.80 - 19.67 - - - 19.67 -.78.13 1978 45 173 34 - - 116 - 357.80 - - - - 357.80 8.13 47.41 9.51 122.54 31.96 219.55 -138.25 1979 - - - 122 k10.5 50 - .;,1.00 - - - - 571.00 16.26 127.42 27.84 91.90 57,87 321.29 -249.71 1980 - - 48 - 55 51. - 511.10 - - - 511.10 27.73 265.46 47,57 49.01 45.00 434.77 - 76.33 1981 - 173 298 183 - - 11.92 - 15.37 44.98. - 72,22 43.13 319.20 59.28 13.48 70.72 505.81 433.59 1982 - - 11.92 - 15.32r 6.92 - 34.16 46.13 341.50 63.41 - 75.65 526.72 492.56 1983 - - - 11.92 - 99.58 - 6.92 ~j118.42 48.01 355.16 65.95 - 78.68 547.80 429.38 1984 -- 11.92 - 15.32-44.98 - 72.22 49.93 369.37 68.58 81.82 569.70 497.48 1985 - - -77.48 - 15.32 - 6.92 - 99.72 51.92 384.14 71.33 - 85.10 592.49 492.77 1986 - - - 11.92 - 99.58 -6.92 - 118.42 54.00 399.51 74.18 - 88.50 616.19 497.77 1987 - - -11.92 - 15.32-44.98 - 72.22 56.16 415.49 77.15 - 92.04 640.84 568.62 1988 - - -11.92 - 15.32 -6.92 - 34.16 58.41 432.11 80.23 - 95.72 666.47 612.91 1989 - - - 11.92 - 99.58 - 6.92 - 118.42 60.74 449.39 83.44 - 99.55 693.12 574.70 1990 - + 18.9 - 77.48 - 15.32- 44.98 -118.82 63.17 467.37 86.78 - 103.53 720.85 602.03 aY 50% residual valua of BP? build . RÄTE OF k~TUN - 14.5% Source: Governm~nt, congultaet. and miion raports. Table 9 SENEGAL FEEDER ROADS PROJECT (Loan 1221-T-SE) PROJECT COMPLETION REPORT Econoic Re-evAluAtinn of Peeder Roads Imvroved Under the Proiect.: Assuming Inadequate Yaintenance (CFAF Million in Dec. 1980 prices) Total Cost of Construction BENEFITS Total ear and Initial Maintenance On Roads Const- On Roads Maintain- Bene Net ructed ed fits Benefits 1 - 761.80 - - - 761.80 2 - 497.80 19.67 - 19.67 - 478.13 3 - 357.80 65.05 154.50 219.55 - 138.25 4 - 571.00 171.52 149.77 321.29 - 244.71 5 - 511.10 340.76 94.01 434.77 - 76.33 6 - 379.44 41.77 421.21 421.21 7 - 315.64 7%56 323.20 323.20 8 - 190.12 - 190.12 190.12 9 - 49.75 - 49.75 49.75 RATE OF RETURIT 10.9% Source: Government, consultants and mission reports.