Burundi - Third Highway Project

Document of The World Bank FOR OFFICIAL USE ONLY i~~i~.N-. 'n J,l~ B i *1 Report No. 9 73 4 PROJECT PERrORMANCE AUDIT REPORT BURUNDI THIRD HIGHWAY PROJECT (CREDIT 1132-BU) JUNE 24, 1991 This document has a restrictd distribution and may be used by recipients only in the performance of CURRENCY EQUIVALENTS Currency Unit: Burundi Franc (FBu) 1980 US$1.00 90.00 1981 90.00 1982 90.00 1983 93.00 1984 122.70 1985 126.00 1986 114.17 1987 123.12 1990 155.00 ABBREVIATIONS BOR - Back-to-Office Report DGR - Direction Generale des Routes (General Directorate of Roads,MTP) EDF - European Devolopment Fund ERR - Economic Rate of Return ICB - International Competitive Bidding IDA - International Development Association INTRACO - International Transport Company MTP - Ministry of Public Works MTPT - Ministry of Transport, Posts and Telecommunications OECF - Overseas Economic Cooperation Fund, Japan OED - Operations Evaluation Department, World Bank OTRABU - Transport Office of Burundi p.a. - per annum PR - President's Report PCR - Project Completion Report PPAR - Project Performance Audit Report PPF - Project Preparation Facility SAR - Staff Appraisal Report SR - Supervision Report UNCDF - United Nations Capital Development Fund UNDP - United Nations Development Programme USAID - United States Agency for International Development FISCAL YEAR OF THE BORROWER January 1 - December 31 FOR OFFICIAL USE ONLY THE WORLD SANK Washington, D.C. 20433 U.S.A. Office of Directoreneral Operatwens Evaluat June 24, 1991 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Performance Audit Report on Burundi Third Highway Project (Credit 1132-BU) Attached, for information, is a copy of a report entitled i"Project Performance Audit Report on Burundi Third Highway Project (Credit 1132-BU)" prepared by the Operations Evaluation Department. Attachment This document has a -^ricted distnbution and may be used by recipients only in the performance of thpir off -ial dutip contents m,y not otherwise he d-closed without World Bank authorization. FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT BURUNDI THIRD HIGHWAY PROJECT (Credit 1132-BU) TABLE OF CONTENTS Page No. Preface .................................................... i Basic Data Sheet ............................................. ii Evaluation Summary .......................................... iii I. INTRODUCTION ........................................... 1 Background ......................................... 1 IDA support for transport .......................... 2 II. THE THIRD HIGHWAY PROJECT ............................... 5 Preparation ........................................ 6 Project implementation and results ................. 6 Total cost and financing ........................... 9 Economic rate of return and sustainability of project results ............................ 9 III. FINDINGS AND CONCLUSIONS ................................ 9 Foreign assistance ................................. 9 Access to the sea ................................. 10 Lessons for the future ............................. 12 Annex 1 The Bujumbura-Rugombo Road: Reconstruction and Procurement Issues ................................. 14 Appendix Comments on the draft PPAR ............................. 19 This document has a restricted distribution and may be used by recipients only in the performance n thoir nf'-' I dut'- ite ennfrnte r v pnt nt w* s d* -Id * d withr* t Wnrld l1 at ri t' L PROJECT PERFORMANCE AUDIT REPORT BURUNDI THIRD HIGHWAY PROJECT (Credit 1132-BU) SPREFACE 1. This is the Project Performance Audit Report (PPAR) on the Third Highway Project (Credit 1132-BU) of SDR 20.5 million equivalent, signed on December 17, 1981, and declared effective on March 17, 1982. The Closing Date was extended from December 31, 1984 to June 14, 1985 and the Credit was fully disbursed. The project was co-financed by the aid agencies of Japan (US$2.0 million equivalent), Belgium (US$0.3 million equivalent), France (US$0.4 million equivalent), and by the United Nations Capital Development Fund (US$1.6 million equivalent). 2. The Infrastructure Operations Division of the Bank's South- Central and Indian Ocean Department prepared a Project Completion Report (PCR) which was published on December 30, 1988. The Bank's Operations Evaluation Department (OED) had audited the Second Highway Project (OED Report No. 6255 of June 9, 1986) which urged that future IDA assistance be devoted to simple, clearly-defined objectives. The February 1990 Transport Sector Project is geared to sector policy reform and auditing the Third Highway Project offered the opportunity to investigate whether conditions are now appropriate for a shift in IDA's strategy towards the complexities of sector policy reform and away from clearly-defined targets, such as maintenance and staff development. The OED mission that visited Burundi in May 1990, discussed the project with the Borrower and with officers of cofinancing agencies, inspected roads improved under the project, and gratefully acknowledges the assistance and hospitality of the General Directorate of Roads (DGR). OED prepared the PPAR by studying the full set of project files, including the Staff Appraisal Report (SAR), the President's Report (PR), the Loan Agreement, and the PCR, transcripts of the Executive Directors' meeting which considered the projects, and published material on the economy and the country. 3. According to standard procedure, OED sent copies of the draft PPAR to the Government and cofinanciers for comments. Comments received as of May 16, 1991, are in the Appendix. iii - PROJECT PERFORMANCE AUDIT REPORT BURUNDI THIRD HIGHWAY PROJECT (Credit 1132-BU) KEY PROJECT DATA Actual or Appraisal Current 1EaE cta&in Estimate Total Project Cost (US$ million) 32.10 27.801/ Savings 13.4% Credit Amount (SDR million) 20.50 20.502/ Disbursed (6/30/84) 20.50 14.83 Disbursed (6/30/85) 20.50 Cancelled Date Physical Components Completed 6/30/84 12/31/84 Proportion Completed by: Appraisal Completion Date (%) 75 Actual Completion Date (%) 100 Economic Rate of Return (%) 19 18.5 OTHER PROJECT DATA Original gam Plan. .Gual First Mention in Files or Timetable August 78 Government's Application Negotiations 3/31/81 3/19/81 Board Approval Date 4/21/81 4/21/81 Credit Agreement Date 2/15/81 12/17/81 Effectiveness Date 2/25/81 7/17/82 Closing Date 12/31/84 6/14/85 Borrower Republic of Burundi Ministere des Travaux Publics, de l'Equipement et du Logement (MTP) Executing Agencies Direction Generale des Routes (DGR) of MTP and Ministere des Transports, Postes Telecommunications (KTPT) Fiscal Year of Borrower January 1 - December 31 Follow-on Project Name Fourth Highway Project Credit Number 1583-BU Amount (SDR million) 18.9 Credit Agreement Date 6/14/85 I/ Excluding taxes D In US$ million equivalent: 25.00 and 21.4. iv- MISSI0NLDATA Month/ No. of Staff Report Misions Year WeekA Peraons 1ees Dat Preparation2/ 9/78 n.a. 1 n.a. 11/09/78 PreparationJ/ 2/79 n.a. 1 n.a. 4/ 6/79 Preparation]/ 6/79 n.a. 1 n.a. 8/ 7/79 PreparationW/ 10/79 n.a. 1 n.a. n.a. Preparation 1/80 1.0 1 1.0 4/ 1/80 Preparation,/ 2/80 1.4 1 1.4 5/23/80 Appraisal 6/80 2.7 2 5.4 4/ 1/81 IoAL: Seven mission totalling (N.A.) staff weeks over 21 months. Supervision 4/82 1.4 2 2.1 7/13/82 Supervision 9/82 0.8 1 0.8 11/15/82 Supervision 4/83 0.7 1 0.7 6/17/83 z:upervision 11/83 1.4 1 1.4 12/19/83 Supervision 5/84 1.4 1 1.4 5/24/84 Supervision,/ 10/84 0.6 1 0.6 12/ 3/84 Supervision 1;&/84 1.0 1 1.0 4/10/85 Supervision 9/85 1.3 2 1.9 2/18/86 Supervision-/ 3/86 0.3 1 0.3 12/31/86 Supervision4/ 10/86 0.1 1 0.1 12/31/86 Total: Ten missions totalling 10.3 staff weeks over 54 months. I/ In conjunction with supervision of Highway Maintenance and Second Highway Projects. 4/ Accounting and auditing. PROJECT PERFORMANCE AUDIT REPORT BURIMDI THIRD HIGHWAY PROJECT (Credit 1132-BU) EVALUATION SUMMARY 1/ Project preparation 1. Preparation started in was approved in August 1982, work August 1978, took one and a half began in February 1983 and was years, and was difficult (paras completed in September 1984, well 12-16). Complications arose within schedule. Quality was good because IDA fund shortages and and Government accepted the works intricate co-financing in October 1984. Final total cost arrangements (PCR, paras 2.04, was US$11.4 million, about 33% 2.08, 2.09, 2.11, 2.13), led to lower than the appraisal estimate, successive revisions of the and only 3% higher than the project scope (PCR, paras 2.10, initial contract amount (para 18). 2.13). 4. Imorovement of about 360 km of secondary and tertiary roads Project implementation and results --- Work-by force account started in March 1982 and produced 201 km. 2. In April 1981, the Board The two brigades which carried out apptv?, a project aimed to assist the improvements were not fully road construction, improvement, operational at the planned and maintenance. IDA fund starting date due to delayed shortages led to a nine-month equipment deliveries. Most of the delay of the effectiveness date. roads were improved to standards The Credit Agreement was signed in higher than those planned at December 1981 and became effective appraisal. The quality of works in March 1982. The project was was good and total cost was US$2.5 completed on December 31, 1984, by million, as against US$3.2 million which time 94% of Credit funds had estimated at appraisal (paras 19- been disbursed. Compared to the 20). Work by contract involved appraisal forecast, the project heavy earthworks, drainage was completed six months behind structures, inlet and outlet schedule; road construction was structures, and ditch lining on completed with a four-month delay about 8 km of RIG 7. Work quality (paras 15-17). was good and the total cost was US$1.9 million (para 21). 3. Construction of 39.4 km EAuiment.orocurement was to of the Ngogi-Junction RIG 14 road replace worn-out items for one section to two-lane bituminous construction brigade, and to equip paved standard --- The contract fully a second brigade. IDA was to finance about 52% of the L/ In this summary, references to the PPAR are shown as (para ...) and references to the PCRs corresponding to the Ports Rehabilitation Project and to the Highway Rehabilitation Project are shown as (PCR. para ...). viii equipment and the Japan Overseas 7. Rairs on RN 5 (para Economic Cooperation Fund (OECF) 28) --- Under the Second Highway the remaining 48%. OECF equipment Project, the contractor who built was delivered in February 1982. the Bujumbura-Rugombo road (RN 5) The United Nations Capital went bankrupt. Soon thereafter, Development Fund (UNCDF) agreed to sections began to fail, Government finance IDA's share, the Credit decided that strengthening of was aaended, and t'he released IDA selected segments was neces3ary, funds were reallocated. Imperfect and signed a US$4.4 million communications between UNCDF and contract with another contractor. the Government led to delays and IDA agreed to finance 70% of the most of the equipment was total contra.t cost. The Credit delivered in the last few months Agreement was amended in April of 1984 (para 22). 1984, and the work was satisfactorily executed for a 5. Road maintenance (para total cost of US$4.5 million. 23) --- In 1984, severe financial difficulties led Total cost and financin (para 29) Government to fire one hilf of the maintenance labor force and to cut 8. Total actual cost, down drastically on budget includin- taxes, was US$30.8 allocations for maintenance million. operations. Mechanized brigades After deducting the cost of c:v!S not work regularly, labor- remedial works for RN 5 (not btrd brigades were short of included in the original project wovers, and roads were not description), actual cost would be properly maintained. US$26.3 million, as against US$35.0 million estimated at 6. Consulting services appraisal. The difference is due (paras 23-27). Wks to currency realignments (64%) and supervision -- Consultants lower prices (361). performed very well. In addition, they improved the design during Economic-rate of return and works execution, thus helping keep sustainability-of-gxoJect benefits the cost below the appraisal (para 30) estimate. Road improvement and maintenance programs -- Consultant 9. Physical work done under performance was satisfactory. the project was economically Staff training -- About 200 justified. The Audit concurs with foremen and 100 equipment the PCR which gives an 18.5% operators were trained under the reestimated rate of return, project. Consultant performance compared to 191 estimated at was satisfactory. Planning and appraisal. Given that no Rrogarmming -- One transport subsequent road construction and economist helped MTPT study the maintenance operations, can be external transport routes and mounted without external overall sector coordination, but assistance, and that Govarnment no lasting results seem to have functions under severe financial emerged from these exercises. constraints, sustainability of Another economist helped MPW carry project b6nefits will depend on out the economic evaluation of the continued financial and technical Fourth Highway Project. assistance from abroad. ix Conclusions and recommendations 1 is perceived as a disturbance 10. The Audit feels that IDA of comfortable practices. and the Aid Community ought to be more decisive in their efforts to Staff training is useful and ensure that Burundi (as well as must continue, but its Uganda and Rwanda) has safe and effects will not be visible cheap access to the sea (paras 34- for a long time. 36). Further, the review occasioned by the audit of the - Strengthening of transport Third Highway Project has yielded planning ia general, and some broad conclusions regarding staffing of DGR in project preparation, project particular, will be extremely dcumentation, and donor long-term propositions. coordination (para 37). Lessons Given that salaries are for the future can be summarised higher in the private sector, as follows: DGR will not retain the necessary numbers of 11. Financial and technical experienced and competent assistance (para 39) engineers and man&gers. Strengthening of road For the foreseeable future, transport will depend on the expatriate staff will have to volume of financial remain in DGR in the same assistance -- and on the positions and in the same quality of technical numbers as in 1990. At the assistance -- from abroad. same time, the initiative to the Transport Sector Project Sound use of such assistance to experiment with hiring of will, in turn, require that local staff under contract exchanges among Donors be with reasonable salaries more frank, and far more deserves, in the audit's integrated than in the past. view, the Government's full It is an open question, who commitment and support. among the Donors will play the indispensable part of Conditions and covenants coordinator. attached to future projects ought to reflect essential If IDA undertakes to perform concerns (e.g., cost this task, it must allocate recovery, competition among considerably more staff operators, increased role for resources than it has done so the private sector, far (para 39). scrupulous auditing of project accounts; proper 12. Proiect inception (para 40) awarding of cont:acts) and ought to be strictly Physical objectives ought to observed. be simple and measurable. 13. Project impleym.atLo (para Institutional obj4ctives 41) ought to be modest. The civil service is too big to As much as possible of the be flexible, and any change physical work (both in new construction and in x maintenance) ought to be IDA supervision missions turned over to the private ought to focus on essentials. sector. Most of the Failure to submit correctly uupervision will have to be audited accounts, and done by tec'nical assistance diversion of project staff. As long as daily resources to non-project allowances remain low, DGR purposes, ought to be staff will be unwilling to go regarded as sufficient cause to the field for site for immediate suspensior of inspections. disbursement. Management, works supervision, and training tasks must be kept separate. BU53.500 Expatriate staff assigned to one activity must not be expected to engage in another. xi 1. The OED Mission to Burundi preceded a visit to Rwanda where the Fourth Highway Project was audited. Lessons for the future are remarkably similar for the two countries. PROJECT PERFORMANCE AUDIT REPORT BURUNDI THIRD HIGHWAY PROJECT (Credit 1132-BU) I. INTRODUCTION Background 1. Burundi (27,800 km2) was poor and undeveloped when it became independent in 1962 and has since changed but little.1/ Population (4.9 million in 1986) is growing at about 3% p.a., and without substant4ve measures to reduce fertility, the growth rate will accelerate to 3.1% during 1990-2005 and to 3.3% afterwards. Total population would reach 12 million by 2015, doubling again in the next twenty years. 2. Only about 150,000 people are in wage employment, and the rest depend on subsistence agriculture. Limited production for the market creates a heavy dependence on imported consumer goods. Coffee exports generate what little foreign exchange Burundi earns. The labor force is large, untrained, and growing. Road transport is the main internal transport mode. The road network consists of about 4,000 km of National and Provincial Roads (of which about 1,000 km have a bitumen surface, about 1,000 have a gravel surface) and a little more than 2,200 km of Communal Roads (all of them earth-surface). The port of Bujumbura has recently been extensively improvdd and the Bujumbura International Airport has the capacity to accommodate increased traffic. Transport connections to the Indian Ocean, over which the country can trade with the rest of the world, are long and inefficient.Z/ More than half of total capital formation is financed by external aid. 3. Between 1983 and 1986, financial conditions improved thanks to better coffee prices. In 1987, the situation worsened because coffee revenues declined and there were no compensatory measures in place. Also, and despite the fact that financing of development projects had been on concessional terms, the public debt-service ratio (which averaged 4.4% of exports during 1972-76), rose to 26% in 1986 and exceeded 38% in 1987. 4. The country continues not to attract much external capital on commercial terms, partly because it does not have any resources worth 1/ In 1965, GDP per capita was US$48; in 1976. it was US$120; in 1986, GNP per capital was US$240. Subsistence agriculture continues to account for over 60% of GDP at factor cost. Please see Bunm4i -- Structural Adjustment and Development Issues (South-Central and Indian Ocean Department, Africa Region, January 20, 1988). 1/ For a detailed discussion, please see the PPAR on Burundi: Second Highway Project (OED Report No. 6255 of June 9, 1986). 2 exploiting and partly because of the instability of its export earnings, which depend on fluctuating coffee prices. External aid has to be on grant terms and to include a large share of local cost financing. If development projects are economically viable, substantial volumes of foreign aid can be justified both for construction and for maintenance. On the other hand, if economically unjustifiable projects are built, not only scarce resources are wasted but additional recurrent obligations are added to a maintenance budget that is already tight. Government has not always seen the connection between new works and their future maintenance, possibly because foreign aid has, so far, been relatively ample, and possibly because it may believe that donor-financed investments can be maintained ad infinitum with donor resources. The Bank Group continues to advocate that completed projects are net additions to the country's capital, and ought to be maintained with domestic resources.a/ IDA support for transnort 5. IDA support for roads in Burundi dates back to 1957. For about thirty years, objectives were straightforward (improve the classified network and strengthen maintenance), and so were the means for their pursuit (technical assistance; staff training; provision of maintenance equipment; improvement of road sections with an acceptable rate of economic return). 6. IDA has now enlarged the scope of its concerns. In 1990, it proposes to support, at the transport sector level, the Government's macroeconomic policy by reducing overall transport costs; by promoting sectoral integration through encouragement of private initiatives and reform of the regulatory framework; by making a positive contribution to the financing of Government expenditures and a balanced budget; by supporting the development of coffee exports; by consolidating and improving road maintenance practices, institutions and funding; by achieving a better balance between maintenance and new construction; and by developing sectoral planning capabilities within both the Ministry of Public Works (MPW) and the Ministry of Transport, Posts and Telecommunications (MTPT).A/ Considering the economy of Burundi, the country's absorptive capacity, and its total dependence on external aid, these are ambitious targets. They require, on the Government's side, a finely-tuned administration and, on IDA's side, large allocations of quality staff time for monitoring and supervision. 1./ This was the issue behind the prolonged discussions, conducted during the preparation of the Second Highway Project, over the construction (to be financed by Germany and the Republic of China) of the economically unjustifiable Gitega-Gihofi and Bujumbura-Rutovu roads. Economics, and IDA advice, notwithstanding, the two roads were eventually built. A/ Staff Appraisal Report. Republic of Burundi: Transport Sector Project (February 28. 1990), para 2.01. 3 7. In 1986, OED published an Audit on the Second Highway Project and urged that future IDA assistance continues to focus on simple, clearly- defined objectives. Auditing the Third Highway Project offered the opportunity to explore whether conditions are now appropriate for shifting towards multiple packages aimed at sector policy reform. A brief summary of the experience with IDA-financed projects provides the proper perspective for such an exploration. 8. Between 1957 and 1990, the Board of Directors approved five transport projects for Burundi. IDA assistance was first extended in 1957 to the Belgian Trust Territory of Ruanda-Urundi. In 1970, a credit was granted for the financing of detailed engineering of the Bujumbura-Nyanza Lac road plus a study for the improvement of road maintenance. The credit amount was later included in 1974 First Highway Project. 9. The First Highway Project was appraised in 1971, but the 1972 civil war depopulated the area where the project was to improve a major road. Reappraisal was done in 1973, road improvements were dropped, and emphasis was placed on maintenance and betterment of existing roads.5/ The project became effective in August 1974 and all physical components were completed in August 1978, except for an office building which was not included in the original scope and necessitated a two-year extension of the closing date. The Project had mixed results. Due to cost overruns, fewer pieces of equipment could be purchased than expected at appraisal and no spare parts were procured. The recurrent budget had insufficient funds for maintenance, and loan funds for betterment were used to supplement the budget. Construction of a workshop and depot was virtually complete when the loan became effective, and the corresponding loan funds were used to renovate the MPW building in Bujumbura and the Training School building in Gitega. In both instances, construction was delayed and, for two years, held up credit disbursement. Technical assistance, and the associated staff training, suffered from lack of suitable counterparts, poor continuity, and discipline problems. In addition, training was hampered by lack of funds, slow deliveries of materials, and delays in procuring equipment./ On the g/ The June 1983 PCR on Burundi-First Highway Project (Credit 467-BU) states (para 2.08) that the project consisted of: (a) a four-year highway maintenance program, including technical assistance to the Department of Roads and Bridges, encompassing staff training at all levels; procurement of maintenance, workshop and training equipment. spare parts, fuel and materials, and the construction of a mechanical workshop and storeroom at Bujumbura and of a small maintenance depot at Gitega; (b) consultants' services for preparation of a master plan for road development; feasibility studies and detailed engineering and updating the report for preparing the project; and (c) refinancing of Credit S11-BU. g/ A debate of form over substance led to a two-year delay in the procurement of equipment. While the Borrower intended to disqualify the lowest bidder for tipper trucks on the grounds of lack of after-sales service, and to award the contract to another company it had long and good experience with, IDA insisted that the contract be awarded to the lowest bidder. It took twenty-eight months after the 4 other hand, the road plan, the studies and the engineering were eventually completed, and formed the basis for the Second and Third Highway Projects. All of the above could, in good faith, be attributed to the teething troubles of a first operation, implemented under difficult conditions. 10. While the First Highway Project was experiencing problems, the Second Highway Project was approved in 1978.Z/ Physical output did not come up to appraisal expectations.l/ However, components related to staff training, technical assistance, improvement of road maintenance operations, and construction and operation of the MPW Soils and Materials Laboratory, were satisfactorily implemented. In spite of the reduced scope, and largely because of the extra expenditures necessitated by remedial work for the Bujumbura-Rugombo road, total cost exceeded appraisal estimates by about 30% (from US$23.97 million to US$31.05 million). Works execution encountered difficulties due to Government procedures, lack of experienced staff, imperfect supervision by consultant and IDA staff, and political conflicts in neighboring countries which interrupted, for long periods of time, the arrival of supplies and equipment. On the positive side, road maintenance improved, and staff training as well as institution building, were beginning to remove some of the factors responsible for poor results under the First Project. 11. The Third Highway Project, approved in April 1981, is the subject of this Audit. The major road financed under the Second Project began to fail in 1983 and the Third Project was amended in April 1984 to finance remedial works. Even so, funds did not suffice and additional financing was included in the 1985 Fourth Highway Project, which also supported a 3-year road maintenance and strengthening program. In early 1990, when the Transport Sector Project (TSP) was approved, the Fourth Project was about 2 years behind schedule. first bid was opened before IDA withdrew its objection. Z/ Credit 773-BU for US$14 million of February 1978. !/ The 65 km Bujumbura-Rugombo Road was expected to be completed in 30 months. The original contract was awarded in September 1978. Execution encountered considerable difficulties, the quality of work was poor, and remedial work had to be financed under the Fourth Highway Project. Works by force account to improve the classified network produced 72% of the planned output (207 km instead of 289 km). Works by contract to improve the classified network produced 22.6 km of road improvements (instead of the planned 149.5 km): 10 wooden bridges were replaced (instead of the planned 22); and 68.000 m2 of steep grades were stabilized (5,000 m2 more than planned) but no surface treatment was applied. 5 II. THE THIRD HIGHWAY PROJECT 12. This was a simple project in terms both of physical work and of institutional objectives. Nevertheless, ten years elapsed between inception and PCR publication. The following listing of milestones hints at many of the reasons why even simple targets take long to achieve in Burundi. September 1978: Preparation mission. February 1979: Preparation mission. June 1979: Preparation mission. October 1979: Preparation mission. January 1980: Preparation mission. February 1980: Preparation mission. February 1980: Project Brief. March 1980: Bank approval of a Project Preparation Facility (PPF) advance. April 1980: Report on difficulties in implementing projects in Burundi. April 1980: Preparation in the ank of form letters that Goverment must send to the Dank for the PPF advance. June 1980: Revised Project Brief. June 1980s Project Appraisal. July 1980: Issues Paper. September 1980s Letter from Goverment to IDA noting that the project Is different from *the desires of the Government". December 1980s Yellow Cover Staff Appraisal Report (SAR) review, Including changes In project financing. March 1981: Project negotiations. April 1981: Board approval. December 1981: Credit signature. March 1982: Credit effective. December 1983- February 1984: Procurement problem. April 1984: Project amended. June 19841 Bank Headquarters-Resident Mission exchange. paecmberv 1i8: Project completion. early 1985: Beginning of drafting of the Project Completion Report (PCR). May 1988S PCR elements still missing. December 1988: Pr published. 13. Preparation missions over one and a half years were necessary because Government was not geared up to deliver planning and engineering work that met IDA standards. The work had to be done by consultants, who had to be paid, but there was no money, and IDA's Project Preparation Facility had to be used. The Borrower had to submit an official request but, apparently, not even this was feasible. Consequently, IDA drafted formal letters addressed to itself, sent them to the Borrower, and the Borrower sent them back to IDA. 14. Communication difficulties must have led the June 1980 appraisal mission to put together a project with which the Borrower took exception in September. Further complications arose because of the need for co- financing, and because IDA fund shortages delayed project effectiveness. Then, in 1983, failure of the road financed under the Second Project necessitated remedial works. After considerable debate between IDA Headquarters and the Resident Mission in Bujumbura, the Third Project was amended in April 1984, and part of the remedial works were carried out. Meanwhile, the Fourth Project was being prepared and some of that credit went for further remedial works of the failed road. In greater detail, experience with the Third Highway Project can be summarized as follows. 6 Preparation 15. Project preparation began in August 1978. IL was difficult and long because of high inflation, leading to increases in estimated costs (PCR, para 2.07), and to successive revisions of the project scope (PCR, paras 2.10, 2.13). Appraisal took place in June 1980 and the Board approved the project in April 1981. 16. The IDA credit (US$25.0 million equivalent) was supplemented with cofinancing by Japan (US$2.0 million equivalent), the United Nations Capital Development Fund (US$1.6 million equivalent), Belgium (US$0.3 million equivalent), and France (US$0.4 million equivalent). Cofinancing was necessary because of cost increases (PCR, paras 2.03 ff). Financing arrangements were complicated (PCR, paras 2.04, 2.08, 2.09, 2.11, 2.13). Project implementation and results 17. In April 1981, the Board approved a project aimed to assist road construction, improvement, and maintenance. IDA fund shortages led to a nine-month delay of the effectiveness date. This, however, did not prevent preparatory work from continuing. The Credit Agreement was signed in December 1981 and became effective in March 1982. The project was completed on December 31, 1984, by which time 94% of Credit funds had been disbursed. Compared to appraisal forecasts, the project was completed six months behind schedule; road construction was completed with a four-month delay. (a) Construction of 39.4 km of the Ngozi-Junction RIG 14 road section to two-lane bituminous paved standard 18. After a first round of international competitive bidding (ICB) prices were about 30% higher than estimated at appraisal. Standards were reduced, the road was redesigned, and a second round of ICB awarded the contract (US$11.1 million, or about 24% higher than the appraisal estimate) to a Chinese firm. The contract was approved in August 1982, work began in February 1983, and was completed in September 1984, well within schedule. Quality was good and Government accepted the works in October 1984. Final total cost was US$11.4 million, which is about 33% lower than the appraisal estimate, and only 3% higher than the initial contract amount. (b) Improvement of about 360 km of secondary and tertiary roads 19. Up to 30% of the works (110 km) were to be done by contract and the rest by force account. Actually, only 8 km were done by contract, and 201 km by force account. The project provided for the procurement of road construction equipment. 20. Work by force account started in March 1982 and when the project was ended in December 1984, 201 km had been improved. This amounts to 7 about 6 km per month, as against 10 km estimated at appraisal.2/ The comparison is not meaningful for two reasons. First, because the two brigades which carried out the improvements were not fully operational at the planned starting date due to delayed equipment deliveries. Second, most of the roads were improved to standards higher than those planned at appraisal. The quality of works was good and the total cost was US$2.5 million, as against US$3.2 million estimated at appraisal. Cost per km was US$12,437. 21. Work by contract involved heavy earthworks, drainage structures, inlet and outlet structures, and ditch lining on about 8 km of RIG 7. The work was done by an Italian contractor, its quality was good and the total cost was US$1.9 million, which amounts to US$240,000 per km. 22. EquiRment _rocurement was to replace worn-out items for one construction brigade, and to equip fully a second brigade. IDA was to finance about 52% of the equipment and the Japan Overseas Economic Cooperation Fund (OECF) the remaining 48%. OECF equipment was delivered in February 1982. The United Nations Capital Development Fund (UNCDF) agreed to finance IDA's share, the Credit was amended and the released IDA funds were reallocated.10/ Bids for UNCDF equipment were called in March 1983, proposals were opened in April, and equipment was scheduled for delivery by December 1983. Imperfect communications between UNCDF and the Government led to delays and, at Government's request, an IDA mission visited UNCDF Headquarters in New York to settle the matter. Most of the UNCDF equipment was delivered in the last few months of 1984, which is to say at the end of the project. (c) Road maintenance 23. In the early 1980s, the four regional maintenance centers in Bujumbura, Gitega, Bururi and Ngozi carried out both manual and mechanized maintenance of earth and gravel roads, as well as patching and sealing of paved roads. In 1984, severe financial difficulties led Government to fire one half of the -sintenance labor force and to cut down drastically on budget allocations for maintenance operations. Spare parts were not available, equipment was not maintained, and repairs were put off. As a result, mechanized brigades could not work regularly, labor-based brigades were short of workers, and roads were not properly maintained. (d) Consulting services 24. Works supervision -- The consultants who designed the reconstruction of the Ngozi-Junction RIG 14 road were selected to carry out works supervision, with the assistance of Italian consultants responsible 2/ Supervision Report of July 1. 1983, Annex III, para 6. .L/ Reallocation covered cost overruns on road construction; operating costs for a regraveling brigade; services of an additional expert for road maintenance; and transport planning and coordination. 8 for materials and drainage control. Consultant supervision was very well done. Furthermore, the consultants improved the design during works execution, thus helping keep the cost below the appraisal estimate. 25. Road improvement and maintenance programs -- Consultants continued services begun under the Second Highway Project. Their performance was satisfactory. 26. Staff trainina -- Consultants continued services begun under the Second Highway Project and provided adequate training for maintenance personnel: about 200 foremen and 100 equipment operators were trained under the project. Consultant performance was satisfactory. 27. Planning and orogramming -- The project envisaged that UNDP would finance technical assistance for transport planning and programming to the Ministry of Transport, Postal Services and Telecommunications (MTPT) and to the Ministry of Public Works (MN). When UNDP financing became unavailable, IDA agreed to provide the necessary funds and the Credit Agreement was suitably amended. One transport economist helped MTPT study the development and utilization of external transport routes and overall sector coordination but no lasting results seem to have emerged from these exercises; IDA has persisted and the same assistance, with the same objectives, was provided under Highways IV. Another economist helped MPW carry out the economic evaluation of the Fourth Highway Project. (e) Reairs on RN 5 28. Under the Second Highway Project, a Belgian contractor supervised by German consultants built the Bujumbura-Rugombo road (RN 5). He went bankrupt. Soon thereafter, sections began to fail.11/ Government decided that strengthening of six segments (totalling about 18 km) were immediately necessary.12/ Work would involve a new sub-base and base courses plus a double-surface treatment on six stretches totalling about 18 km; local repairs on four stretches totalling about 19 km (10.5 km of which would have to be strengthened later on); and reconstruction of drainage structures. Government signed a US$4.4 million contract with an Italian contractor and IDA agreed to finance 70% of the total contract cost. The Credit Agreement was amended in April 1984, and the work was satisfactorily executed for a total cost of US$4.5 million. Total cost and financing 29. Total actual cost, including taxes, was US$30.8 million. IDA financed US$21 million, Belgium US$1.0 million, Japan, US$2.0 million, France US$0.3 million, UNCDF US$2.2 million, and the Government US$4.3 11/ More details are provided in Annex 1: The Bujumbura-Rugombo Road -- Reconstruction and Procurement Issues. 12/ Strengthening and repair of four stretches (about 16 km) would also be needed but not as urgently and the Fourth Highway Project provided the necessary financing. 9 million. After deducting the cost of remedial works for RN 5 (which was not included in the original project description), actual cost would be US$26.3 million, as against US$35.0 million estimated at appraisal. The difference is due to currency realignments (64%) and lower prices (36%) Economic rate of return and sustainability of prolect benefits 30. Physical work done under the project was economically justified and the Audit concurs with the PCR (paras 5.01 and 5.08-5.08) which gives an 18.5% reestimated rate of return, compared to 19% estimated at appraisal. Road improvements and maintenance operations have generated vehicle operating cost savings translatable into longer vehicle lives, fewer spare parts, and lower fuel consumption. Given the shortage of foreign exchange which plagues the country, these are important benefits. Given that no subsequent road construction and maintenance projects (e.g., the Fourth Highway Project and the Transport Sector Project), could have been mounted without high volumes of external assistance, and also given that Government functions under severe financial constraints which often lead to effective suspension of maintenance operations, sustainability of project benefits will depend on continued financial and technical assistance from abroad.1/ III. FINDINGS AND CONCLUSIONS Foreign assistancel4/ 31. Like all previous and subsequent operations, the Third Highway Project was prepared and implemented with foreign assistance. Financing aspects apart, Burundi cannot yet rely upon its own human resources for the design and supervision of road construction and maintenance. IDA, and the Donor Community as a whole, have tacitly accepted that, in the foreseeable future, strengthening road transport will require additional foreign aid. Financial assistance will have to cover more than three fourths of construction and maintenance costs. Technical assistance will have to cover virtually one hundred percent of the resources required for planning and implementation of construction and maintenance.1/ Regardless of 13/ For comments from the Secretary of State with responsibility for planning, please see Appendix 3. LA/ For comments on paras 31. 32, and 33 from the Secretary of State with responsibility for planning, please see Appendix 3. 11./ The 1990 Transport Sector Project makes it clear that foreign assistance is indispensable for capital investment, for recurrent expenditures, and for the actual running of the road sub-sector. Implementation of this US$90.8 million project will require US$65.2 million from IDA. the African Development Bank, OPEC, and the aid agencies of Belgium. Germany, France. Japan and Italy. Government is expected to contribute US$25.6 million equivalent but the SAR acknowledges that there is a risk of local fund shortages. 10 this manifest dependence on external aid, the formal process for securing donor commitments is assumed to be the outcome of a "policy dialogue". 32. The Audit believes that transport project financing in Burundi emerges out of monologues by individual donors aho decide what they can offer and then proceed to justify their a-priori decisions. Lip-service is always paid to the benefits -- and difficulties -- of aid coordination bu t, for a long time, aid fragmentation was agreeable to donors because, besides facilitating export of goods and services, it allowed bilateral and multilateral agencies to reassure their constituents by pointing at specific projects and activities they supported. For its part, Government did not mind aid fragmentation. Indeed, it encouraged it, partly on the grounds that it inspired competition among donors. From the viewpoint of evaluation, a disadvantage of fragmented aid is that it is difficult to monitor. 33. By the late 1980s, the aid community began to have second thoughts about the effectiveness of its assistance. Reservations arose not simply from the Burundi experience. The modest results that development aid has had on a Continent-wide scale provoked questions about whether familiar formulae for the delivery of assistance should not be revisited. There is a strong possibility that both the absolute volume, and the means of channeling foreign assistance to Burundi might soon undergo major modifications. If this comes to pass, it will be to the advantage of all parties concerned to be aware, sufficiently ahead of time, that the aid climate I..s changed. The one area where Donors must get together as soon as possible is to help Burundi gain cheap and safe access to the sea. Access to the seal6/ 34. Burundi is landlocked and access to the Indian Ocean has long been a matter of concern for Government and IDA1l/ alike. IDA views regarding access to the sea were initially set out in a 1978 Transport Sector Memorandum. They were amplified further in the 1980 ReRort on the International Transportation Bottlenecks Affecting Rwanda and BurundL.1/ They were taken up once again in the 1990 Africa -- The Great Lakes Corridor StuLtdy./ 16/ For comments from UNDP on this section, please see Appendix 1. I/ Part of the funds made available to Burundi through an Industrial Development and Finance Credit ()31-BU. US$3.4 million. 1978) were onlent to INTRACO (a parastatal trucking company), to purchase trucks to haul petroleum products and other merchandise from Mombasa. IS/ Eastern Africa Projects Department. December 1980, in two volumes. The February 27, 1980 Project Brief (para 1.11) on the Third Highway Project expressed the hope that this Study would help determine future Bank Group lending for transport in the area. 12/ Infrastructure Division. Africa Technical Department. March 1990. 11 35. Between the 19701 and 1990, the donor community, with the Bank Group frequently taking the lead, extended assistance to littoral countries so that transport users in Uganda, Rwanda, Burundi, Malawi and Zambia might receive better service. Numerous projects in Kenya and Tanzania were justified in part through assertions that they would also serve transit traffic.2&/ Still, the quality of transport services through littoral countries was not ideal when the East African Community was in operation, and has not much improved since.21/ 36. When aid recipients receive funds that are, at least in part, expected to help neighboring countries survive and prosper and through negligence, inefficiency, procrastination, or worse, reduce the volume n4 value of traffic, then the donors not only have the right but the obligation to switch from dialogue to strict conditionality. The donor community ought to ensure that the rights and interests of lr.d-locked countries are safeguarded, and that further discussion -- and, especially, discussion among land-locked and littoral countries -- can no longer serve as a substitute for action.22/ Admittedly, land-locked countries are not strong enough to demand treatment for their transit traffic in accordance with international procedures. Consequently, action ought to be taken by the donor community. Considering the size of extornal assistance granted to Kenya and fanzania, it is inconceivable that if aid agencies decided to act together they would not succeed in enforcing international procedures and, in particular, that transit traffic should not be held up because of real or alleged formalities at ports, at points of transshipment, or at the frontiers.22/ 2/ Specifically: justification of improvements to the Nairobi-Mombasa road, to the port of Dar es Salaam, and to the Tanzania Railroad are always accompanied by lengthy expositions of the benefits for landlocked countries to the west. 2J/ See, for example, the PPAR on East African Community-Third Railway Project (Loan 674-EA) (OEV Report No. 4533, dated June 3, 1983). The November 6, 1985 SAR oi. the Sixth Highway Proiect devoted an entire section to international transport (paras 1.09-1.12) and employed complicated syntax to conclude that nothing much had happened. The 1986 PPAR on Burundi -- Second Highway Project devoted an entire section to sea access. L2/ This is acknowledged in page 4 of the June 7, 1989, comments on Rwanda: Transport Sector Project. Project Brief which say: "Unfortunately, the Northern Corridor Transit Agreement is not beginning to facilitate the transport flow of exports and imports. So far it has been a failure and has even complicated matters, as documented in the Great Lakes Corridor Study." Z3/ When this observation was advanced in the PPAR on Burundi -- Second Hihway Project, IDA staff took strong exception and wrote to OED as follows: "We do not believe that it is realistic to assume that the Bank and the donor Community can persuade the transit countries to provide better transport services to their neighbors than to their nationals. We do not think it is appropriate either and we should not give false expectations to the landlocked countries through such statements". 12 Lessons for the future 2a/ 37. The review occasioned by the audit of the Third Highway Project has yielded some broad conclusions which might be taken into account with advantage during preparation of future interventions: (a) Project preparation was invariably hurried to produce a credible package. However, credibility is harder to achieve in the case of multi-donor projects because each donor wants his assistance to pursue different objectives. (b) Projects have been approved on the basis of documentation and analyses that were not irreproachable. (c) IDA and Donor Agency Missions interact with consultants (who possess statistical and other information) and with Ministers (who possess political power). Statistical and political support are important but actual project work is done at the medium and lower levels of the civil service. Few IDA and Donor Agency staff take time to get to know the civil service and this does not bolster institution-building. (d) In the Aid Community, the sense of impending cut-backs is palpable. If this comes to pass, there will be need for far more openness and collaboration among Donors than has so far been the case. 38. Suggestions for the future can be summarized under the following three headings: 39. Financial and technical assistance25/ For the foreseeable future, strengthening of road transport will depend on the volume of financial assistance -- and on the quality of technical assistance -- from abroad. Sound use of such assistance will, in turn, require that exchanges among Donors be more frank, and far more integrated than in the past. It is an open question, who among the Donors will play the indispensable part of coordinator. If IDA undertakes to perform this task, it must allocate considerably more staff resources than it has done so far. 24/ The OED Mission to Burundi preceded a visit to Rwanda where the Fourth Highway Project was audited. Lessons for the future are remarkably similar for the two countries. 15/ For UNDP comments on this section. please see Appendix 1. 13 40. Project inception. Physical objectives ought to be simple and measurable. Institutional objectives ought to be modest. The civil service is too big to be flexible, and any change is perceived as a disturbance of comfortable practices. Staff training is useful and must continue, but its effects will not be visible for a long time. Strengthening of transport planning in general, and staffing of DGR in particular, will be extremely long-term propositions. Given that salaries are higher in the private sector, DGR will not retain the necessary numbers of experienced and competent engineers and managers. For the foreseeable future, expatriate staff will have to remain in Dl in the same positions and in the same numbers as in 1990. At the sate time, the initiative to the Transport Sector Project to experiment with hiring of local staff under contract with reasonable salaries deserves, in the audit's view, the Government's full commitment and support. Conditions and covenants attached to future projects ought to reflect essential concerns (e.g., cost recovery, competition among operators, increased role for the private sector, scrupulous auditing of project accounts; proper awarding of contracts) and ought to be strictly observed. 41. ProJect implementation. As much as possible of the physical work (both in new construction and in maintenance) ought to be turned over to the private sector (that the domestic construction industry is still in its infancy presents a challenge in itself and deserves patient attention and assistance). Most of the works supervision will have to be done by technical assistance staff. As long as daily allowances remain low, DGR staff will be unwilling to go to the field for site inspections. 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 correctly audited accounts, and diversion of project resources to non-project purposes, ought to be regarded as sufficient cause for immediate suspension of disbursements. 14 Annex 1 THE BUJUMBURA-RUGOMBO ROAD: RECONSTRUCTION AND PROCUREMENT ISSUES Out of the funds for Credit 1132-BU, an amount of US$4.5 million (not included in the original project scope) was devoted to repair work on the Bujumbura-Rugombo Road. Details on the implementation experience with this road are given in the PPAR on Burundi: Second Highway ProJect (OED Report No. 6255 of June 9, 1986), from which the following are extracted. Design work completed by consultant. It called for a natural gravel basecourse, stabilized with 2% cement (prudently increased to 3% in the bill of quantities since 2% is difficult to mix evenly); double surface dressing, applied over a 5.5 m width, as well as bituminous priming on the 1.5 m shoulders. 191& During the bid period, prospective bidders were informed of certain changes, including a shortening of the contract period to 18 months, and an increase of the paved width from 5.5 to 6.0 m (with reduction of shoulders to 1.25 m). The lower of the two bidders proposed the omission of cement in basecourse construction and the provision of "mechanical stabilization", which involves mixing of two gravels. The price differential between the two solutions was modest (about 15% lower for "mechanical stabilization") but the award committee accepted mechanical stabilization partly because of cement shortage at the time. The low bid also included a 7.6% price reduction which made it even more attractive. IDA off ad no objection and contract award was made to the low bidder. Work was to have started in late 1978 but was delayed because of the Uganda-Tanzania war and did not get underway until mid-1979. 1972-1980 The contractor worked in a satisfactory way under the consultant's supervision. At the same time, the contractor submitted claims for delays he had suffered on account of the Uganda-Tanzania war but Government both denied the claim and refused to submit it to arbitration. 1980-1981 The contractor underwent a change of management in late 1980 and, in early 1981, so did the consultant. Through 1981 the project was under "new" management. 15 Early 1982 First time that IDA Supervision Reports mention minor problems: "sweating" of the bituminous wearing course. Several localized failures started to occur, beginning along the edges of the road and spreading inward, probably caused by ingress of water into basecourse and lower layers. Second quarter of 1982 The contractor went bankrupt. His personnel left Burundi, leaving the plant and equipment behind.*and a road which seemed to be about 99% completed with only 1.5 km of surfacing remaining to be done. The contract for supervision had expired and the consultant left. Prior to leaving, however, the consultant had refused to accept certain road sections due to the poor surfacing, and had defined the completion and repair works that were necessary. The November 15, 1982 IDA Supervision Mission Report contains no reference to technical failures of the road. September 1982 Aware that repairs '&ad to be done and that surfacing of the remaining 1.5 km had to be completed, Government asked two firms to make prope als for the work. Only one responded, submitted a priced proposal (which was accepted) and the work order was issued on September 25, 1982. The contract was totally financed by Government and supervision was done by Ministry staff. October 1982-April 1983 As repair work progressed, road failures increased at an alarming rate and, by early March 1983, it became obvious that a very large amount of work had to be done. On March 10, 1983, Government issued an additional work order for patching about 12,000 square meters of road, using approximately the same unit - tces as in the contract of September 25, 1982. Work was suspended in early April when Government funds ran out. May-August 1983 It was now that Government decided to conduct an investigation of the causes and possible remedies of the failures. This was done by the National Laboratory which issued three reports dated May 19, June 9, and August 11, 1983. The Laboratory recommended a two-phased approach. First, localized deep patching had to be done as necessary over the whole length of the road; a single seal surface dressing would have to be applied on the existing surfacing of selected sections (totalling 10.5 km); and a new basecourse and surface dressing would be applied over 18 km, including 1.0 km where a drainage subbase layer would also L3 applied. Second, an overlay would have to be applied on 5.6 km and the embankment would have to be raised on another 5.6 km. Based on its recommendations for the first priority work, the Laboratory prepared a bill of quantities of works to be performed. The contractor who had already worked on the road submitted a priced 16 proposal for this in July 1983. The estimated cost of the work was about US$4 million equivalent. The question was where the money would come from. In late July 1983, and while en route froy Madagascar to Rwanda, the IDA project officer was asked to stop off in Burundi to discuss the matter of financing. Between July 25-27, he reviewed the Laboratory reports, surveyed the failures, found that a state of emergency existed, reviewed the draft contract, found that prices were reasonable, reviewed the Credit Agreement and its Amendment, found that road maintenance and its execution by contractor were covered, that sufficient funds existed in the Credit to cover IDA's participation, and that the contractor was in the best position to carry out the works urgently, efficiently and at a reasonable cost. Consequently, he gave verbal approval for the work to proceed. On the strength of this verbal approval, Government asked the contractor to submit an amended proposal, which he did on August 2, 1983. The order for the work was issued on August 4, 1983 as a second amendment to the work order issued on September 25, 1982. IDA guidelines dictate that written confirmation has to be issued by IDA Headquarters for any commitment made by Project Officers. In this case, no such confirmation was issued. October 1983-March 1984 Early in October 1983, an inquiry from the IDA Resident Representative led to a review in IDA Headquarters of the contract for the repair works and this raised the following questions: (a) "whether proper procurement procedures were followed for a contract amounting to US$4 million"; (b) "whether the issue was properly dealt with by the supervision mission and properly treated in the supervision report"; (c) "whether the practice could be misinterpreted by the Government that deviations from the Bank's normal procurement procedures is not a serious matter"; and (d) "the amount provided under the Third Highway Project was not intended for the repair work of a recently built highway and in an amount of such magnitude, and whether the matter was properly reviewed by the management within the Bank". A meeting held in the Bank on January 10, 1984 to review the procurement and disbursement issues associated with the contract for the repair works concluded that: (a) "The Bank's procurement procedures had not been followed for the review and approval of a contract amounting to about US$4 million. No bidding documents were prepared or reviewed. No competitive bidding (neither ICB nor LCB) was conducted. Contract was approved verbally by the project officer in the field without subsequent formal telex confirmation by the Division from Washington...." (b) "No one outside the Division in Headquarters was informed of the transaction (Programs, Legal, OPD Procurement Unit, and the Front Office of the Projects Department). The supervision 17 reports do not provide adequate information about the transaction." (c) "....On December 2, 1983, in response to a Loan Department inquiry (whether reallocation of funds would take place--already overdrawn on November 16 by SDR 760,000--or, if no reallocation was planned, up to what limit could the Loan Department continue overdisbursement), the Division again without prior consultation with any person outside the Division authorized the Loan Department to disburse up to SDR 3.5 million under Category 2(b), under which only SDR 400,000 were available. After the event, it appears that no attempt was made to inform anyone outside the Division of the authorization....". The meeting agreed that, in view of the fact that two thirds of the repair work had already been completed, and in the best interests of the project, the Bank should consider that a waiver be given to the Bank's procurement requirements subject to a review by the Engineering Adviser of the technical merit of the work under the contract, and a review by the Legal Department on the need for revision of the Project Description. The Engineering Adviser visited Burundi and his report (dated February 23, 1984) stated that the design of the repair works was technically reasonable and the quality of works satisfactory. The Bank reviewed disbursement of withdrawal applications under the contract and concluded that, since "Government had signed the contract in good faith and with the Association's approval", disbursement should be made in keeping with the terms of the contract. The description of the Third Highway Project, as well as Schedule I of the Development Credit Agreement were duly amended. Summary The first contractor proposed a questionable alternative to cement stabilization and agreed to finish the work one full year faster than envisaged by the appraisal team. While the contractor was encountering difficulties beyond his control, he was directed by MPW to follow major design changes. The contractor incurred financial losses, but was not allowed to submit his case to arbitration. He went bankrupt after having completed more than 98% of the work, was denied payment on his last bill and the performance security was called. It would appear that the contractor's rights and privileges under the contract may have been abused by the MPW. The supervising consultant had prepared the original road design and specifications. Nevertheless, he agreed to a technically questionable alternative to cement stabilization, and appears to have been lax in controlling laboratory testing and supervision. Monitoring by IDA staff was not what it should have been. Changes to important design features were agreed after appraisal; the mechanical stabilization solution proposed by the bidder was agreed to by IDA; 18 there was a two-year gap in supervision while the bulk of the work was carried out; and financing of the remedial work was discussed and agreed upon in an unorthodox manner. MPW was known to be short of staff experienced enough to supervise closely the execution of a major contract. IDA supervision missions became twice as important and necessary. Instead of increasing the frequency of such missions, IDA did not send any missions at all over a period of two years, precisely during the time when the bulk of the construction work was being done. IDA missions that visited Burundi later, stated that both the construction and the supervision work were "satisfactory" but annexes of the Supervision Reports contain data which contradict this assertion. BUB3.501 19 APPENDIX 1 91E0605 April 16, 1991 French (Burundi) OEDD1 PS/McM:uh Translation of Fax from Bujumbura. Burundi dated March 29. 1991 IQ: Graham Donaldson &f: Burundi: Third Highway Project. Proiect Performance Audit Renort We acknowledge receipt of the above-mentioned report. A. We find the report well written and that its recommendations are very responsive to our concerns, particularly with respect to the effort of donors to provide Burundi with low-cost maritime access. The need to overcome Burundi's landlocked position is one of UNDP's concerns. B. The success of the project depends on the efficient coordination of aid and regular consultation among donors. C. With regard to technical assistance, the main donors should provide high-caliber experts and insist that the Government make counterpart personnel available so that the latter can benefit from the former's expertise and take over the work following project completion. D. The action proposed by IDA, p. 2, item 6, concerning the Government's macroeconomic policy is also consistent with UNDP's position. SLTS (KOTTA) APPENDIX 2 20 Direction G6n6rale des Routes P.O. Box 1860, Bujumbura, Burundi. March 28, 1991 Mr. Graham Donaldson, Chief, Agriculture, Infrastructure & Human Resources Division Operations Evaluation Department The World Bank Washington, D.C. SUBJECT: Third Highway Project (Credit 1132-BU) Proiect Performance Audit Report Dear Sir: In response to your letter of February 13, 1991 accompanying the above-mentioned report, I have pleasure in informing you that, other than a few clarifications, I have no particular comments to add. I would merely mention the following: 1. On page iv, paragraph 7, the report says that the contracting company working on the Bujumbura-Rugombo highway had gone bankrupt but that it completed the work around the middle of 1984. In actual fact, after this company declared bankruptcy in 1982 the work was taken over by another contractor through 1984. 2. On page 9, paragraph 24, they were Italian, not Swiss, consultants. 3. On page 10, paragraph 28, the report mentions that the Belgian contractors had gone bankrupt but that they finished the work around mid-1984. What happened, in fact, was that after declaring bankruptcy in 1982 the company naturally stopped working. Another contractor took over the work. Very truly yours, [illegible signature] Director General of Highways Ministry of Public Works and Urban Development Copies as per original APPENDIX 3 21 91E0617 May 36, 1991 French (Bur4ndt) OEDD1 PS/Mdf4ecw ranslation of tlex. No. 12046691. dated April 17..1991. for 4fe attentitg- of Mr. Donaldeen. with reference to le&ter dated February 13 traimitting Protect Performance Audit Report on Third HighwaX Proiect 1132-RU We are pleased to Liura you of our agreement with -ho terms of the technical evaluation of the execuLotin of this project. We c4r-t share your view that the work on the Gitega-GlhuC and Bujumbura-Rutovu roads was economically "unjustifiable." The traffic gentrated -- to mention just one aspast -- was included in the most recent count&, with which you are familiar. These projects must now be sufficiently justifiable for you to plan to fund the laying of a wearing course on the GILega-Gihofi road, instead of this being funded by German financing as originally planned at the time of the TSP appraisal, the mobilization of which has run into some dela;.. We note your opinion in paragraph 13 and also Lt pointlessness of saying that bids for tho Ngozi-Junction, RIG 14, were 25% higher than at appraisal. when the -orks actually ended up costing 33% less than at appraisal. The analysis contained in paragraph 30, which state. UiaL "the Government's serious financial difficulties are resulting in the suspension of maintenance operations," is incorrect. The counterpart funds from the Government scheduled in accordance with the Credit Agreement are budgeted in a satisfactory and proper manner. But Lhe procurement procedures have been extremely unwieldy, particularly for the Fourth Highway Project. We do not share the views you express in paragraphs 31, 32 and 33, to the effect that beneficiary counLrLes take advantage of competition among donors. Kind regards, Salvator Sahinguvu Secretary of State, with responsibility for Planning