Rwanda - Fifth Highway Project

Document of The World Bank FO ORMCIAL USE ONLY MICROFICGHE COPY Report Nto. 10302-RW Type: (PCR) WEILLE, J./ X31708 / ORDD3/ T-9102 Re=tNo. 10302 PROJECT COMPLETION REPORT RWANDA FIFTH HIGHWAY PROJECT (CREDIT 1250-RW) FEBRUARY 4, 1992 Infrastructure Operations Division Country Department III Africa Regional Office This document bas a restricted distribution aid mas y b sed bS recipiens ol in the performance of their official duties. its contents may not otherwise be disclosed without World Bank authorization. , ~ ~~~~ .icoew u Wol Ban auhrzain  Currency Unit - Rwandese Franc (RF) US$ 1.00 in 1982 - RF 92 1983 _ 99 1984 - 104 1985 - 93 1986 - 84 1987 - 73 1988 - 77 1989 - 78 ABBREVIATIONS AND ACRONDMS AfDB - African Development Bank BADEA - The Arab Fund for Economic Development FED - The European Development Fund KFAED - The Kuwait Fund for Economic Development ERR - Economic Rate of Return IDA - International Development Association MGERWA - Magasins G6n6raux du Rwanda NF - Ministry of Finance NPW - Ministry of Public Works UNDP - United Nations Development Program SAR - Staff Appraisal Report GOVERNMENT OF RWANDA FISCAL YEAR January 1 - December 31  FOR OMCIAL USE ONLY THE WORLD BANK Washington. O.C. 20433 US.A. :e i Oitect~ver't Opwaum Ev4lwtumn February 4, 1992 HEHQRANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Rwanda Fifth Highway Project (Credit 1250-RW) Attached, for information, is a copy of a report entitled "Project Completion Report on "Rwanda Fifth Highway Project (Credit 1250-RW)" prepared by the Africa Regional Office with Part II of the report contributed by the Borrower. No audit of this project has been made by the Operations Evaluation Department at this time. Attachment IThis document has a restricted distribution and may be used by recipients only in t;e performance | of their official duties. Its contents may not otherwise be disclosed without World Bank authofization.  FOR OMCIAL USE ONLY PROJECT COMPLETION REPORT RWANDA FIFTH HIGHWAY PROJECT (CREDIT 1250-RW) TABLE OF CONTENTS Palo No. Preface ..... . . .................................. i Evaluation Summary ..ii PART U: PROJECT REVIEW FROM IDA'S PERSPECTIVE 1.0 Project Identity... 1 2.0 Background.. 1 3.0 Project Objectives and Description .. 1 4.0 Project Design and Organization ................... 2 5.0 Project Implementation ................................. 3 6.0 Project Results ......................... 4 7.0 Project Sustainability ................................. 5 8.0 IDA Performance .......... .......... 5 9.0 Borrower Performance ....................... 8 10.0 Project Relationship ....................... 9 11.0 Consulting Services ....................... 9 12.0 Project Documentation and Data .................. 9 Note 1: Further Detail and Comments on the Economic Evaluation in the SAR ............................. 11 PART II: PROJECT REVIEW FROM BORROWER'S PERSPECTIVE 1. Introduction .12 2. Project Objectives ................... 12 3. Design and Organization of the Project .13 4. Project Execution .13 5. Results of the Project .13 6. Comments on Part I of the Project Completion Report for the Fifth Highway Project and Confirmation of Part III . 13 EMR III: STAT-ISTICAL INFORMTxION 1. Related Bank Loans and/or Credits .14 2. Project Timetable .15 3. Credit Disbursements .16 4. Project Costs and Financing .16 5. Status of Covenants .19 6. Use of Bank Resources .21 r This document has a restricted distribution and may be used by recipients only in the performance Lof their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  - - PROJECT COMPLETION REPORT RWANDA FIFTH HIGHWAY PROJECT (CREDIT 1250-RW) PREFACE This is the Project Completion (PCR) for the 17ifth Highway Project, in which Rwanda, for Credit 1250-RW in the amo mnt of SDR 23.3 million was approved on May 20, 1982. The Credit was closed on December 31, 1988, for years behind schedule. the Credit was fully disbursed and the last disbursement was on July 17, 1989. The PCR was Jointly prepared by the Infrastructure Operations Division of the South Central and Indian Ocean Department of the Africa Regional Office (Preface, Evaluation Summary Parts I and III), and the Borrower (Part II). This PCR is based inter alia on the Staff Appraisal Report, the Credit Agreement, Supervislon reports, correspondence between the Bank and the Borrower, and internal Bank memoranda.  - ii - PROJECT COMPLETION REPORT RWANDA FIFTH HIGHWAY PROJECT (CREDIT 1250-RW) EVALUATION SUMMARY PrEQi@ct Ob_jectives and Scpv_e_. 1. The Fifth Highway Project shared the objectives of fostering economic integration of the southwest region with the rest of the country through more reliable transport and that of reducing transport costs to road users. Ther main project item consisted of a 53.5 km section of the 156 km Butare-Cyangugu road. UNDP shared in the financing of the Fifth Highway Project on account of its technical assistance component. Four additional sections totalling 88 kms of the same road were simultaneously constructed under financing by four other donors, the European Development Fund (FED), the African Development Fund (AfDB), The Arab Fund for Economic Development (BADEA) and The Kuwait Arab Fund for Economic Development (KFAED). 2. The Fifth Highway Project was of a classical type which the World Bank Group has been financing ever since its establishment, involving the construction of a road between two given points, under the rules of international competitive bidding, with construction supervision being assigned to consultants chosen according to IDA guidelines on procurement. 3. In addition to its main construction component, the project featured a -limited amount of technical assistance involving institution building for transport planning and highway maintenance; and lastly, vehicle weight control. The main project component was well prepared. ImDlementation Exoerience 4. There were no significant variances in project implementation. Due to the good prices obtained for the construction of the road section between Butare and Kitabi that represented the main project component, a saving of about US$6 M was made over project cost estimates appearing in the SAR. The savings were utilized to start off the rehabilitation of the Kigali-Gatuna road. The full rehabilitation of the Kigali-Gatuna road was financed under the subsequent Sixth Highway Credit. 5. The project took seven years to complete instead of the programmed three. About half of the extension in the project schedule was caused by the addition to the proje_t of the rehabilitation of a part of the Kigali-Gatuna road. A variance in the project completion schedule of two years net of any increase in project scope is quite common in many countries for projects of this kind and therefore considered as acceptable.  - ili - 6. Overall project performance was satisfactory. Concerning its main road construction component, the project exceeded its goals well within the original financial envelope. In this respect there was considerable progress over the record of previous highway projects in Rwanda. Concerning its other components, under previous highway projecto progress in institution building had been slow. Little if any progress can be reported under the Fifth Highway Project as well (paras 9.06 to 9.08). Similarly no progress can be reported on vehicle weight control (paras 5.05 and 5.06). These shortcomings do not relate to the main substance of the project which was concluded successfully. However, in the broad context of the long term IDA intervention in the sector, and in strategy terms, the unaddressed issues were important. The project's ERR is 13X as compared with 16X estimated in the SAR (paras. 6.02 and 6.04). Project Sustainabilitv 7. In physical terms, provided that the Borrower adequately maintains the constructed roads, the project is sustainable. In institutional capacity terms, project sustainability is uncertain: for the foreseeable future institutional capacity will continue to depend upon financial assistance and upon the quality of external technical assistance. 8. In economic terms, the project is sustainable and most likely to maintain an acceptable level of benefits throughout its economic life, as long as technical assistance remains available to close the gap in institutional capacity. If financiallv unaided, the Borrower is not likely to maintain the project road in an acceptable way. The associated risk factor can be addressed mainly by financing project maintenance requirements through a suitable external donor. Over the long run, cost recovery in financial terms is a central issue upon which there is no progress to report. The subject was not explicitly covered in the project but requires urgent attention in future projects to be financed by IDA. IDA Performance 9. The main strength of IDA's performance showed at the project preparation stage. IDA performed well during the preparation of the construction component (para 8.02) which in financial terms accounted for 90% of the project cost. The painstaking preparation effort fully justified itself during project implementation. The main weakness was in project supervision in both quantity and scope. Findinas and Lessons Learned 10. The classical IDA approach to project design and implementation, namely: the identification of a road project for IDA financing; its preparation by consultants; its execution by contractors under IDA guidelines on procurement; and the assignment of construction supervision to equally competent consultants, is a well tested recipe that often produces satisfactory results. 11. At the project preparation stage, institution building objectLves should have been more clearly spelled out than they actually were and so should have been the scope of the project's institution building component itself. These should have be;r. accompanied by a statement of criteria of performance against which the specific actions could be monitored. During project implementation the project's institution building component should have been monitored vigilantly. Failure to submit quarterly progress reports and other financial data as required under the credit agreement should have been regarded as sufficient cause for applying the remedies available to IDA under the credit agreement.  - iv - Borrower Perfgrmance 12. The main strength ot the Borrower's performance was that it let the pro be implemented without interfering with the consultants and contrac' that were hired for this purpose. However, there were Involvements with ps ;ent matters which demonstrated a conspicuous lack of awareness of IDA guide_ "es on procurement. In '11 such instances, IDA staff in charge of the project were able to provide the necessary assistance to the Borrower to ensure the respect of the guidelines. The Borrower's record of compliance with covenants was poor, especially with regard to implementation reporting, financial and otherwise. Consultants 13. The consultants who carried out construction supervision performed their job well. The Borrower tended to use the consultants employed under technical assistance as a management resource, especially in the day-to-day sense of the term rather than in the proper sense of technical assistance which is the transfer of know-how and institution building that is expected to occur following such transfer. Donor Coordination 14. A conspicuous and continuous lack of adequate donor coordination during project implementation was a weakness involving relationships between the Borrower and the donors, and among the donors themselves. Donor coordination was a matter of concern under previous highway projects also. It continued to be an elusive goal under the Fifth Highway Project as well. The experience of the Fifth Highway Project highlights the need for a much increased level of donor coordination, especially at the project supervision stage (para 10.2). Imorovements in Pro ect Performance 15. Compared to previous highway projects, there was considerable improvement in project performance as far as the highway constru_tion component of the Fifth highway Project was concerned. Other than this, no improvements were noted in Borrower's performance.  : PRUT 1; PROJECT REVIEW FROM IDA 'S PERSPECTIVE  PROJECT COMPLETIQN REPORT RWANDA FIFTH HIGHWAY PROJECT (CREDIT 125Q-RW0 PART It PROJECT REVIEW FROM IDA'S PERSPECTIVE 1.0 Project Identity - Project Name: Fifth Highway Project - Credit No: Cr. 1250-RW - Rvp Unit: Africa Region, Country Department III - Countrys Rwanda - Sectors Infrastructure - Subsector: Highways 2.0 Back0round 2.01 The project was designed to serve the three main sector development objectives adopted by the Borrower and agreed to by IDA. These were to: (i) improve external transport connections; (ii) upgrade the main road network to facilitate the country's economic and social development, and its administrative integration: and (iii) improve the serviceability of the road network through better maintenance. 3.0 Project Obiectives and Descriotion 3.01 Soecific orgiect obiectivelt They consisted of (i) promoting agricultural development in the country's southwestern region; (ii) fostering economic integration of this region with the rest of the country through more reliable transport; and (iii) reducing transport costs. 3.02 Project Comoonents. The project consisted of: (a) Construction to paved standard of the Butare-Kitabi section (53.5 kme) of the Butare-Cyangugu road (156 kms). (b) Consulting services for (i) project preparation under a PPF, and construction supervision under (a); (ii) engineering studies for the rehabilitation of the Kigali-Gatuna road (80 km.); (iii) pre- investment studies; and (iv) technical assistance to the Ministry of Public Works (MPW). (c) Procurement of materials and supplies for the routlne maintenance of the KigalL-Gatuna road over a two-year period. (d) Procurement of weighing scales for control and enforcement of vehicle weight regulations.  - 2 - 4.0 Proiect Desian and Organization 4.01 The project had a clear cwnceptual foundation. It followed in the wake of the four highway projects that IDA had previously financed in Rwanda. To dat IDA-financed highway projects in the country have alternated between two ma project types, namely institution-building-cum-maintenance and construction-cum- rehabilitation projects. The Second and Fourth highway projects were of the firas kind whereas the First, Third and present Fifth were mainly of the second. 4.02 The Fifth Highway Projec', was of a classical type which the World Bank Group has been financing ever since establishment, involving the construction of a road between two given points der the rules of international competitive bidding, with construction supervi64on being assigned to consultants also chosen according to IDA guidelines on procurement. 4.03 Of the full length of the 153 km Butare-Cyangugu road, a length of 123 km8, between Butare and Ntendenzi, required upgrading. IDA financed the section between Butare-Kitabi with a length of 53.5 kms. The road sections other than the one included in the Fifth Highway Project totalling 69.5 kma were financec by FED, AfDB, BADEA, and KFAED (See Part III, Section 5B). The overall project size including ite sections financed by other donors was large for Rwanda. Its financial envelope represented the equivalent of seven years of thA country'li annual highway construction budget, financed from both internal and externa; sources. .4.04 The ambitious economic objectives of the project were matched by the size of the total investment. However, how the project would impact on the development of Rwanda's Southwestern region was not clearly spelled out in the SAR nor in other available project documents. The project objectives were agreec upon by all the relevant parties. The project was not innovatL a in any way, in a Rwandese, African or broader context. 4.05 Pre2aration of the oroieot. The project was well prepared except f its small technical assistance component. In particular, much care was accord to the preparation of an adequate design responding to the difficult terrain anc soil conditions that prevailed over the road's alignment. IDA staff who preparec the project were keenly aware of the difficulties p-evious highway projects in Rwanda had run into, causing large escalations to project costs. To avoid the repetition of such experience they took special care to assure an adequate soils investigation capability both within the consultant's team and in project preparation missions mounted by IDA. 4.06 Cost estimates reflecting modest design standards were carefull prepared. A special contingency allcwance was introduced into cost estimates in addition to the standard quantity contingency allowance of 10% reflecting the spec'Al engineering risks that applied to the project. The economic evaluation of t ,r, project prepared by the consultants was carefully monitored by IDA staff in charge of project preparation. These special efforts proved themselves most useful during project appraisal and implementation, as explained in the sections that follow. 4.07 Roles and Resoonsibilities. They were clearly defined in the SAR. Borrower's inadequacy during implementation in fully playing its role as definec in the SAR cannot be attributed to definitional shortfalls, but to institutional capacity and an apparent lack of sufficient interest on its part.  -3- 5.0 Proiect Xmolementation S.01 Variances in imolementation. There were no significant variances. Due to the good prices obtained for the construction of the road section between Butare and Kitabi that represented the main project component, a saving of about US$ 6 million was made over project cost estimates appearing in the SAR. The savings were utilized to start off the rehabilitation of the Kigali-Gatuna road. Originally, only the financing of routine maintenance was programmed under the credit. The full rehabilitation of the Kigali-Gatuna read waa financed under the subsequent Sixth Highway Credit. 5.02 The project took seven years until completion instead of the programmed three. About half of the extension in the project schedule was caused by the addition to the project of a part of the rehabilitation of the Kigali-Gatuna road. A variance in the project completion schedule of two years net of the increase in project scope is quite common in many LDC's for projects of this kind, and therefore, considered as acceptable. 5.03 Risks: The SAR identified four main risks: (i) work interruptions; (ii) cost overruns caused by (i) and/or by other factors; (iii) the risk of the project institution building component not attaining its intended goals; and (iv) weak enforcement of axle load regulations. The first two risks mentioned were considered by the SAR to be major risks. However, no significant interruptions occurred during project execution and as explained in para 5.01, construction costs turned out to be lower than estimated in the SAR. The third and fourth risk were considered to be of a low order of probability, although at implementation stage they turned out to be substantial. 5.04 Concerning the institutional risk, the SAR did not specify any goals or criteria of performance against which the implementation of the project's technical assistance component could be evaluated. The project implamentation record itself did not describe project achievements in institution building. Indirect evidence available suggests that little i' any institution building occurred during the life of the project (For further discussion see Section 9 on Borrower performance). 5.05 The risk of implementation weaknesses in enforcing axle load limits was not adequately addressed. Considerable procrastination occurred in first procuring and then installing the weighing scales. They were installed in December 1988 after a five year delay and only six months before project completion. According to information obtained from the last supervision mission of the subsequent Sixth Highway Project, the weighing scales have still not been put into use. The Borrower is now claiming that the scales procured are inadequate and wishes to procure new scales with "better" specifications. However, since recently, most heavy traffic is routed through Kigali and weighed at the MAGERWA installations for import control purposes. Ways are being explored to extend this weight control to an axle load control, which in time would make the frontier weighing unnecessary. However, this requires better linkage and coordination betwe6n MPW and MF, the latter being responsible for MAGERWA. 5.06 The substantial risk associated with the axle load issue could have been foreseen. It is common knowledge that enforcing axle load limits is a difficult task even among relatively developed LDC countries. 5.07 Actions/decisions affecting iroiect imolementation. The only important decision was to make a start with the rehabilitation of the Kigali-Gatuna road through the use of the credit funds that became available following the savings in the construction costs of the Butare-Kitabi road. The decision was acted upon. No other decisions were needed to keep the bulk of the project on course.  -4- 6.0 Proiect Results 6.01 Proiect obiectives. Project objectives did not change during implementation. Concerning its main component which was road construction, the project exceeded its goals through the addition of the partial rehabilitation of the Kigali-Gatuna road to the project scope. Concerning its institution building aspects there is llttle progress to report (see section 9 on Borrower performance). Concerning its axle load enforcement aspects, see paras 5.05 and 5.06) the project failed to attain its objective. 6.02 Proie8t EBB. For purposes of ERR determination, the SAR divided the IDA-financed road section into two sub sections. For each one of them it calculated a separate ERR. There were sound reasons for doing this as explained in Note 1. In the SAR, the 35 km Butare-Kigeme sub section showed an ERR of 18% and the second sub-section between Kigeme and Kitabi with a length of 18.5 kms, showed an ERR of 13%. The methodology used in calculating the ERR was not adequate, especially concerning the ERR for the second section. The methodology issue is discussed in paras 8.04 & 8.05 and in Note 1. 6.03 The data available are insufficient to recalculate ERR's for the two individual sub-sections . The Borrower did not supply IDA at any time through project completion with traffic counts on any of the sections of the project road financed by other donors which affect the ERR for the Kigeme-Kitabi sub-section. On construction costs, there is only a broad reference to total cost, in an aide memoire from a donors conference held in 1989. A total cost of about FRw 4.01 billion (about US$ 48 million equivalent) is reported with no further detail. The total cost is for the road section financed by IDA and the four other sections each of which was financed by a separate donor. 6.04 For the Butare-Kigeme sub-section, allowing for the construction costs which turned out to be 9% lower than projected, assuming an annual traffic growth of 7% as thl n was done in the SAR, an induced traffic elasticity of one and applying the correct methodology, the revised ERR would be roughly 15%. Doing the same for the 18.5 km Kigeme-Kitabi sub section, the ERR is, also roughly, about 9% (see Note 1 for further detail). The ERR for the first sub-section is thus satisfactory, while it is not for the second sub-section, if estimated without taking into account the overall result of the new Cyangugu road access. 6.05 Institution buildino: Under previous highway projects, progress in institution building had been slow. No progress can be reported under the Fifth Highway Project, in the sense that at project completion MPW required as much technical assistance as it did when the project started (see further detail under para 9.08). 6.06 Physical environment. The Butare-Cyangugu road crosses the environmentally protected Nyungwe forest. During construction some damage was caused to wild life, probably on the endangered species list but the contractors were early on made aware of the environmental concerns. Quantifiable data on the subject are not available. At the same time considerable replanting of trees took place in particular on the section financed by IDA. 6.07 Social environm_tM . The project had a positive impact on the incomes of the population of the road's zone of influence. No data are available for a more specific assessment.  -5- 7.0 Proiect Sustainabilitv 7.01 The sustainability issue can be evaluated at four different levels namely technical, institutional, economic and financial. At the technical level, provided that the Borrower regularly maintains the constructed road section and resurfaces it every seven years as stipulated in the SAR, the project is sustainable. In institutional capacity terms, project sustainability is not certain. For the foreseeable future institutional capacity will depend on financial assistance and on the quality of the technical assistance from abroad (see section 9 below, on Borrower performance). 7.02 In economic terms, the project is sustainable and most likely to maintain an acceptable level of benefits throughout its economic life, provided that technical assistance remains available to close the gap in institutional capacity. If financially unaided, the Borrower is not likely to maintain the project road as and when required. The associated risk factor can be addressed mainly by getting financing for the project's maintenance requirements from an external donor. Cost recovery in financial terms is a central issue. The subject was not explicitly covered in the project but it requires urgent addressing in future projects to be financed by IDA, in part by improved cost recovery under the road fund and in part by better donor coordination. 8.0 IDA Performance 8.01 Overall assessment. The main strength of IDA's performance showed at project preparation stage. The main weakness was in project supervision, both quantity and scope. 8.02 Identification and prenaration. IDA performed well during the preparation of the project's construction component which accounted for 90% of the project in financial terms. The painstaking preparation effort fully justified itself during project implementation. This not withstanding, how the project fitted into sector investment priorities was not explained. This is noteworthy especially because as mentioned earlier, the project was particularly large for Rwanda. 8.03 Appraisal. At appraisal, the mission had a well prepared project to appraise. The appraisal mission could have profitably spent more of its time to firm up the project's technical assistance component. The need was to establish a link between past IDA efforts at institution building in the sector and the project, to formulate institution building goals and to establish performance indicators for monitoring purposes during implementation. 8.04 In the SAR there are repeated references to the unity of the total project, including its sections financed by other donors, of which the IDA- financed section formed an integral part. Accordingly, there was a need to appraise the project road as a whole. Despite this, the economic evaluation in the SAR confined itself to the section which IDA financed. The evaluation took into account, only to a limited extent, the totality of the project as a self- contained entity. There is room for considerable improvement in the methodology used in the economic evaluation appearing in the SAR. Some of the assumptions made in projecting traffic lack substantiation. The feasibility studies underlying the economic evaluation are missing in the files. Therefore, the substantiation issue cannot be further commented upon, except to state that traffic projections do not appear to be plausible especially for induced traffic estimates. For further details on the economic evaluation appearing in the SAR, see Note 1.  8.05 Since the project involved several donors, it would have been most useful to have established a suitable interdonor coordinating mechanism. This was not done (see also para 10.03). 8.06 SuDervision. Fourteen supervision missions were mounted from the date of credit effectiveness through project completion, extending over a seven year period averaging a mission every six months. The missions were staffed with an average of 1.6 professionals. In all but one of the missions, an engineer participated, with the balance of participation being provided by an economist and a financial analyst in equal proportions. 8.07 The number of days spent in the field averaged 6.7 days per mission. However, during this period of time, a mission had to look after an average of 2.6 projects, at various stages of the project cycle. Thus, roughly 2.6 working days only were available on average for supervising a single project during a given mission. That this amount of time is little for effective supervision cannot be disputed, even for projects which are relatively problem-free. Moreover, insufficient preparation by the Borrower prior to supervision missions resulted in less than optimal use of scarce time. 8.08 The inadequacy of supervision time is also reflected in staff time spent on the project as registered in MIS records. Of a total amount of staff weeks spent on the project over a ten year period from its identification through its completion, preparation through appraisal accounts for an input of 68 staff weeks; appraisal through project effectiveness 12.4 staff weeks, and supervision only 29.4 staff weeks. This allocation appears to be lopsided even when allowing for how well the project was prepared and for the fact that its implementation was relatively problem-free. 8.09 During project implementation, a major modification occurred in the project's scope when the partial rehabilitation of the Kigali-Gatuna road was added to it. A credit amendment was made to allow for this addition. There is no record of the estimated economic rate of return for this addition to the project, although had one been prepared it would most probably have been satisfactory. Under the Sixth Highway Project, the rehabilitation of this road was to be completed at an estimated rate of return of 20%. 8.10 The supervision effort mounted by IDA for this added component left a list of project-related subjects inadequately attended to. Most notably, there are only scattered references in the supervision reports on how the institutional building component of the project was faring. The available references are not sufficient to evaluate how this particular project component progressed. Technical assistance under the project was mostly financed by UNDP. The assistance terminated in December 1985. 8.11 There is no record in the files of the agreement signed between IDA and UNDP on technical assistance work scope nor a description of the obligations of each side. Missing also is evidence on how the consulting company, recruited to provide the technical assistance, was selected and whether the candidates for the TA slots were submitted to IDA for approval. Equally unavailable in the files are progress reports and a completion report prepared by the consultants on technical assistance featuring the standard scope and information normally found in such reports. There is no correspondence in the files showing that IDA requested from UNDP or MPW any information concerning the progress of this component. 8.12 There is no record in project supervision reports indicating that the supervision missions tried to come to grips with the Borrower's lack of compliance on reporting requirements. The issue persisted from the date of project effectiveness through project completion. There is no record of audit  -7- reports on project finances. There are scattered references to problems associated with project accounting and accounts. In a particular supervision report it is stated that both accounting and auditing procedures practiced in the country leave much to be desired. i/ 8.13 There is little indication in the files that the supervision missions concerned themselves with getting information on how implementation was proceding on those sections of the road whose construction was financed by other donors. The construction of these sections had a direct bearing on the economic justification of the project. There are no records indicating that the construction of these sections was actually completed and the completion dates. Supervision reports tended to present a rosy picture of conditionality observance. A more detailed presentation on the last subject is presented further below. 8.14 However, beginning in mid 1987, a marked improvement can be observed in supervision reports. Most of the references in the files concerning donor coordination, technical assistance, and financial reporting matters date from this time through project completion two years thereafter. However, with the bulk of the project having been implemented by then, it probably would have been difficult to make up for time past, in particular given the limited amount of staff time available for project supervision. 8.15 None of the shortcomings mentioned above relate to the main substance of the project which was concluded successfully and can only be deemed a success. From a strict project view point, lack of reporting in IDA's supervision of project implementation did not produce effects detrimental to the project. However, in the broad context of Borrower-IDA relations, the record of IDA interventions in the highway sector, past, present, and planned; and in IDA strategy terms for the sector, the issues overlooked were important. 8.16 Main lessons of experience relevant to other IDA financed Droiects. The classical IDA approach to project design and implementation, namely the identification of a road project, for IDA financing; its preparation by competent consultants, its execution by contractors under IDA guidelines on procurement, and the assigning of construction supervision to equally competent consultants; is a well tested recipe that more often than not produces satisfactory results. Such an approach can be contrasted with institutional building and other program projects in environments not ready for them. The results in these cases are often less than satisfactory unless a special effort is made during project preparation and a considerable amount of vigilance is practiced during their supervision combined with an equal amount of reporting requirement. 8.17 At the project preparation stage institution building objectives ought to be clearly spelled out and so should the scope of the project's institution building component. These should be accompanied by a statement of criteria of performance against which they can be monitored. During project implementation special attention should be devoted to a project's institution building -omponent as this is a subject where the probability of Borrower delinquency is at its highest. There is a Borrower letter directed to IDA requesting comments on the proposed TORs for the project audit which the letter claims is appended to the letter; but the TORs are missing from the files and so is the audit report itself, assuming one was submitted. Also, there is no record of any comments made by IDA on the TORs. No annual budget statements and budget execution on MPW activity re highway maintenance and construction was received from the Borrower on the highway sector during the whole period of project implementation.  -8- 8.18 "Management, works, supervision, and training tasks must be kept separate. Expatriate staff assigned to one activity must not be expected to engage in another. IDA supervision missions ought to focus on essentials. Failure to submit quarterly progress reports and other financial data as required under the credit agreement must be regarded as sufficient cause for immediate suspension of the project".ZI/ 9.0 Borrower Performance 9.01 Main strenaths and weaknesses. The main strength of the Borrower's performance was that it let the project be implemented without interfering with the contractors and consultants that were contracted for the purpose. However, there were instances of Borrower's involvement on procurement matters which showed a conspicuous lack of awareness of guidelines on IDA procurement. In all such instances, IDA staff in charge of the project were able to provide the necessary assistance to the Borrower to secure compliance. 9.02 During Rroiect oreparation. The Borrower was mainly involved in the search for overseas financial assistance to cover persisting gaps in the financing of the overall road project. Otherwise he contributed little to the preparation of the project other than making demands related to project composition. At no point did the Borrower offer any substantiation as to why he was proposing a particular project scope, and -hy he subsequently wished it to be modified. After technical review, IDA accepted some of these demands and declined some others. 9.03 During implementation. The Borrower approved consultants and contractors invoices on time using for this purpose the technical assistance made available to him under the project. There is no record of problems in this respect. The Borrower showed commendable initiative in asking IDA to add to the project the partial rehabilitation of the Kigali-Gatuna road which was financed out of credit savings. 9.04 There is no evidence that the Borrower much concerned himself with project aspects related to institution building. In all other project implementation aspects, the Borrower was largely passive. However, as the technical assistance provided under the project functioned effectively, negative consequences did not ensue from this state of affairs. This evaluation holds over the short run and in a strict project context, given the fact that the bulk of the project was a road construction project. 9.05 Proiect conditionality compliance. Of the ten covenants entered into (cf Part III, 7), the Borrower fully complied with two. He complied with two others after a delay of four to five years. A fifth covenant concerned the employment of qualified consultants. The evidence in the files shows that the covenant was complied with except for TA to MPW. There is no evidence showing that IDA approved or disapproved the qualifications of the candidates for the TA. This may suggest that the matter was not brought to its attention in written form. The remaining five covenants were either not complied with, or there is no evidence in the files confirming their compliance. 9.06 The denloyment of consultant resources. The consultants who did construction supervision performed their job well. The consultants that provided technical assistance also performed well when allowing for the tendency of the / The quote is from the Project Performance Audit Report on Rwanda Highways IV of March 26, 1991. It is quoted ad verbatim as it fully applies to the Fifth Highway Project also.  - 9 - Borrower to use the consultants employed under technical assistance as a management resource, especially in the day-to-day sense of the term rather than in the proper sense of technical assistance which is the transfer of know-how and the institution building expected to occur following such transfer. Despite this limitation in TA use by the Borrower, some know-how may have percolated. No hard data are available to evaluate how much of this happened. According to more recent supervision missions, little transfer of know-how seems to have occurred. According to these missions, there is little change to report on the Borrower's attitude to technical assistance use. 9.07 The circumstances surrounding the provision of technical assistance described in para 9.06 often occur in several developing countries. Even under the best prepared technical assistance program, once implementation begins, the pressure to deemphasize the main purpose of TA in favor of assistance in day to day management of the organization often becomes stronger. A large degree of vigilance is required on the part of the supervising IDA staff to at least keep such prerssure low since its total elimination is not a realistic prospect. 9.08 As mentioned earlier, the project's technical assistance component did not feature concrete goals and measures of performance. However, IDA supervising staff did not seem to be aware of the problem referred to in paras 9.06 and 9.07, and even if they had been, they spent too little time in the field to be effective in this respect. A standard end of assignment report on technical assistance is not in the files. There is no reference to one either and presumably none was prepared. 10.0 Proiect Relationship 10.01 The written record does not contain any direct reference to Borrower- IDA relations. However, some procurement problems arose during project implementation on such subjects as guidelines, project reporting and compliance with cred.t covenants. The fact that the Borrower does not adequatel1 prepare the information required to get the most out of supervision missions and does not manifest an active interest in the missions' supervision work during their visits to the country constitutes indirect evidence that the Borrower-IDA relationship needs further strengthening. 10.02 A conspicuous lack of adequate donor coordination throughout project implementation affected relationships between the Borrower and the donors, and between the donors themselves. Donor coordination was a matter of concern under previous highway projects. It continued to be an elusive goal under the Fifth highway project as well. For future projects financed by several donors such as the Fifth Highway Project the following steps are recommended: a. At project preparation stage all participating donors should make a deliberate effort to coordinate their plans and programs. b. Donors should insist that the totality of the project incorporating several donor components is jointly reported upon according to a satisfactory format to be agreed upon between themselves and the Borrower. c. At implementation stage they should meet regularly, at least twice a year, to jointly monitor project progress. 11.0 Consultina Services. For a discussion of the subject, reference is made to Section 9 on Borrower performance.  - 10 - 12.0 Proiect Documentation and Data 12.01 The strengths and weaknesses of the lecal acreements. The legal agreements were drawn in clearly stated and satisfactory fashion except for two instances. Section concerning the commitment to maintain an adequately staffed and functioning planning unit (section 4.02 of credit agreement), was not detailed enough as far as the work scope of the unit was concerned. 12.02 Section 5.01 (h) of the credit agreement requires the Government to submit project reports including indices for measuring performance. The scope of these reports was spelled out in Annex III of the SAR but not in the Credit Agreement. The project files yield no information as to the reasons that might have caused this omission. 12.03 The Staff Appraisal Reoort. It had limited usefulness as a reference for IDA and/or the Borrower during project implementation. It spelled out the progress reporting requirements, a draft TOR for a transport planning expert and one for a senior maintenance engineer. However, the technical assistance context, goals and measures of performance were not specified. Neither was the economic analysis sufficiently detailed for monitoring purposes. The SAR lacked sufficient information on road sections of the broader road project which were being financed by other donors. 12.04 Prolect Comoletion Report (PCRI PreRaration and data. The Borrower supplied some of the complementary data needed for the preparation of Parts I&III of the PCR just as the PCR was about to be finalized. Part of the data supplied was already available, such as traffic on the IDA-financed road sections. The most critical data requested concerned traffic counts on the whole length of the project road including each of the sections financed by other donors and their construction costs. Traffic data for these sections was not supplied. The only traffic information given was for the roads in Rwanda, some of which contradictory, and all of which was insufficient to establish general traffic growth trends in the country since traffic counts were only carried out every other year. It could not be used in this PCR. 12.05 Proiect documentation. None of the project documents listed in Annex V of the Staff Appraisal Report (SAR) could be found in the files. The files of the preceding Fourth Highways Credit were also checked and these documents did not appear in them either. There are no working documents in the files for project Economic Rates of Return (ERRs). They could only be obtained from the private files of the IDA economist who prepared them. The quarterly progress reports which the Borrower was expected to submit to IDA were also lacking. There are no audit reports or any other financial reports prepared until 1988 (for 1987), when the first auditors were appointed. With one exception involving the addition to the project of the rehabilitation of the Kigal.-Gatuna road, the credit amendments referred to in the files could not be found either. During implementation, project files should be checked regularly to make sure that all the important project documents are present.  - 11 - NOTE 1 FURTHER DETAIL AND COMMENTS ON THE ECONOMIC EVALUATION IN THE SAR 1. The total distance from Butare to Cyangugu is 156 kms of which a 33 km section from Cyangugu was in the process of rehabilitation at the time the project was appraised. Total project length was thus 123 kms of which IDA would finance a 53.5 km section and the balance of 69.5 kms was split between four different donors. In the SAR the economic evaluation limits itself to the section financed by IDA which was evaluated as two sub sections. The first is the Butare- Kigeme sub section of 35 kms and the second the Kigeme-Kitabi sub section of 18.5 kms. The reason for the subdivision is that traffic drastically falls off west of Kigeme in the unpopulated area of the Nyungwe forest. 2. A cement plant is located beyond Cyangugu at the end of the road near lake Kivu. According to the working papers of the economic evaluation, of the plant's projected annual production of 30000 tons, 24000 tons would divert to the road once constructed. In the SAR, total benefits on account of the diverted cement traffic was pro rated to the two sub-sections financed under the credit on a km length basis. The basis was the whole of the 123 kms which required construction. 3. For the Kigeme-Kitabi subsection of 18.5 kms the SAR. showed an ERR of 13%. Without the inclusion of benefits related to diversion of cement traffic from an earlier more complex road-lake-road transport, the ERR would have dropped to about 7%. Induced traffic benefits for this sub section were calculated assuming an induced traffic elasticity in excess of five. The treatment of benefits generated by diverted cement traffic is thus of decisive importance for the ERR of the Kigeme-Kitabi section. 4. A more adequate method to follow would have been to exclude cement related benefits from the Butare-Kigeme section and remain with an ERR of 13% which is marginally satisfactory. This means that this particular sub section can be economically justified and constructed as a self contained project without linking its economic justification to that of the other sections. A single ERR estimate would then be calculated for the remaining 88 kms of the road to which all the cement related benefits would be credited, as well as other traffic on this road. 5. On the basis of cement related benefits alone, and assuming that per km construction costs are not substantially different for the non IDA-financed sections, the ERR for the 88 km length is about 4.5% which is unsatisfactory. However the addition to the benefit analysis of the normal traffic on the sections not financed by IDA would obviously have boosted the ERR of this section. Maybe, if the feasibility studies were available it would be possible to establish the ERR by applying correct methodology to this 88 km section which includes the second IDA financed sub section . Until this is done, the economic justification of the second sub section financed by IDA remains undemonstrated as does that of the 88 km section itself. 6. A variant of the above described method would have been to calculate an overall rate of return for the whole 123 km length of the project. The preferability of this method to the above described "one-two" approach is based on the premise that since the project was conceived as a single operation, not involving phases, it could be evaluated through a single rate of return. Either one of the proposed methods would have been more appropriate than the one used in the SAR.  X X H f fi N S N 3 g " g m | S :f r  - 12 - PROJECT COMPLETION REPORT RWANDA FIFTH HIGHWAY PROJECT (CREDIT 1250-RW) PART II: PROJECT REVIEW FROM bORROWER'$ PERSPECTIVE 1. Xntrodag&ion The Fifth IDA Highway Project covered by Credit 1250 RW is composed as follows: I. Paving of the 53.50 kms Butare-Ritabi section of the Butare-Cyangugu road. It. Consultants' services for: * Preparation of the project under the PPF and supervision of the paving work on the Butare-Kitabi section; * Technical studies for restoration of the Kigali-Gatuna road; * Preinvestment studies; e The program of technical assistance to the Ministry of Public Works, Energy and Water Resources. III. Purchase of materials and supplies for routine maintenance of the Kigali-Catuna road. IV. Purchase of weighing scales to monitor and enforce vehicle weight regulations. The Fifth Highway Project started in 1982 and was closed on May 30, 1989. The works listed above were carried out satisfactorily and the loan requirements were met without any difficulty. 2. Proiect Obiectives The primary objective of the Fifth Highway Project has been to promote the economic integration of the southwestern region of Rwanda with the rest of the country following construction of the E.a're-Cyangugu road, and also to link Rwanda with the other countries along the Northern Corridor following restoration of the Kigali-Gatuna road. This has facilitated commercial flows between the southwestern region and other parts of the country, as well as trade among Zaire, Burundi and Rwanda. Rehabilitation of the Kigali-Gatuna road will considerably reduce the cost of transporting goods from Kenya via Uganda. The Ministry of Public Works, Energy and Water Resources notes with satisfaction that the objectives assigned to the Fifth Highway Project have been achieved. The weighing scales were ordered in 1983 and were installed in 1989 following construction of parking areas at Kagitumba and Gatuna on the Ugandan-Rwandess border. Unfortunately these weighing scales have never operated as they are inappropriate to actual conditions in the field. A report has been prepared on this subject and will be sent to IDA for information and possible comment. It should, however, be noted that the Government intends to replace these weighing scales with others of appropriate design, and that scales for trucks have continuously been used in Kigali.  - 13 - 3. DasLSn and Organizatlon of tho Proiect The careful design and organization of the project enabled the expected results to be achieved. Thanks to detailed preparation work with the World Bank, project implementation was greatly simplifLed. 4. Project- ExScutlon The portion of the project dealing with rehabilitation of the Kigali- Catuna road benefited from an approximately $6 million saving on the cost of constructing the Butare-Kitabi section. This surplus enabled work to be continued on the Kigali-Gatuna road, which was comple -ad with funds under the Sixth Highway Project. Regular visits by World Bank missions allowed certain problems that arose in the course of project implemontation to be resolved. It should be noted that, even though work has been completed, a major problem still exists regarding construction of a bridge across Swamp B. Until this brLdge in built, traffic along the route will be subject to interruption. Supervision of the project and the work performed by the consultants were both satisfactory. The absence of training and transfer of technology components is to be regretted. We are of the opinion that IDA should back up the Government in strengthening tra!