Uganda - Fourth Highways Project and Railways Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 19518 PERFORMANCE AUDIT REPORT UGANDA FOURTH HIGHWAY PROJECT (CREDIT 1803-UG) AND RAILWAYS PROJECT (CREDIT 1986-UG) June 23, 1999 Sector and Thematic Evaluations Group Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. Currency Equivalents (annual averages) Currency Unit = Country Currency (Ugandan Shilling, USh) 1996 USh 1000 = US$100 Abbreviations and Acronyms IAP = Initial Action Program KRC = Kenya Railways Corporation MLG = Ministry of Local Government MOF = Ministry of Finance MOT = Ministry of Transport MOW = Ministry of Works MPED = Ministry of Planning and Economic Development MWHC Ministry of Works, Housing, and Communication PRRMP = Pilot Rural Road Maintenance Program RAFU = Road Agency Formation Unit RIP = Railway Investment program SWARP = South West Agriculture Rehabilitation Project TRC = Tanzania Railways Corporation URC = Uganda Railways Corporation UTC = Uganda Transport Company Fiscal Year July 1 - June 30 Director-General, Operations Evaluation Mr. Robert Picciotto Director, Operations Evaluation Department Ms. Elizabeth McAllister Manager, Sector and Thematic Evaluations Mr. Gregory K. Ingram Task Manager Mr. Antti Talvitie The World Bank FOR OFFICIAL USE ONLY Washington, D.C. 20433 U.S.A. June 23, 1999 Office of the Director-General Operations Evaluation MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Uganda: Performance Audit Report Fourth Highway Project (Credit 1803-UG); Railway Project (Credit 1986-UG). Attached is the Performance Audit Report (PAR) on two Uganda transport projects: Fourth Highway project (Credit 1803-UG, approved FY87); and Railway project (Credit 1986- UG, approved FY89). The Operations Evaluation Department (OED) prepared the PAR. The Bank commitment to the two projects was US$25 million, of which US$2.2 million was canceled. The physical objectives of the highway project were directed to protecting investments in roads and to help improve maintenance planning and operations of rural feeder roads. The railway project sought to implement a program of improvements in Uganda Railway Corporation's management and organization, and to begin rehabilitation of the Kampala-Kasese line. To achieve these objectives, the projects had civil works, technical assistance, and institutional and policy development components. The two main roads were rehabilitated as planned, the ERRs ranged from 17 to 61 percent versus 30 to 63 percent at appraisal. However, there were long delays in completing the civil works and the planned benefits from the early years did not materialize. The ERR calculations also exclude the uneven quality of the works and the early repairs that were necessary to sustain the benefits from the roads. Technical assistance to develop capability for good road management was not achieved. The Audit rates the Fourth Highway project as follows. Outcome is rated as marginally satisfactory (rated satisfactory in the ICR); institutional development is rated as negligible (as in the ICR); and sustainability as uncertain (as in the ICR). While the civil works components of the limited objectives the project pursued were achieved, the rating for outcome is downgraded because the two main roads selected for rehabilitation were completed with delay and deficiencies in procurement and contractor performance. The pilot rural roads maintenance program suffered from the shortcomings in counterpart personnel and consultants' lack of performance. Those difficulties, although anticipated, were not resolved satisfactorily during the project and have continued until today. No tangible institutional development took place because of changing priorities, and reduced and irregular consultancy presence and imperfect functioning. Sustainability is rated as uncertain because of the insufficient quality of the civil works and because, institutionally, there is little to be sustained. Bank performance in the Fourth Highway project is rated as satisfactory (satisfactory in the ICR). The Audit upgrades the rating for the Railways project outcome to marginally satisfactory (marginally unsatisfactory in the PCR); sustainability remains rated as uncertain, and This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. -2- June 23, 1999 institutional development remains rated as substantial. The outcome is upgraded because the project did achieve its principal objective of improving the URC management and organization and performance for track and locomotive maintenance, the latter being funded by a donor in parallel. The rehabilitation of the Kampala-Kasese line was wisely shelved at mid-term and project efforts focused on institutional development. Sustainability of the project remains uncertain because several critical issues remain unresolved: trade facilitation through Kenya and Tanzania to improve service and reduce the wagon requirements; track maintenance and rehabilitation east of Kampala to increase train speeds; and, wagon maintenance. Overall, the Bank's performance in the railway project was satisfactory. Of course, there are several aspects where performance could be improved. The audit recommends that the Bank more clearly articulates its work on institutional capacity and policy development, and on reducing rent-seeking. The Bank should continue helping GOU to begin manage its roads on commercial principles and improve decentralized accountability, especially for maintenance; and, develop an improved institutional framework for the feeder roads network. For the railways, the Bank now has the global experience which it should use to help commercialize the Uganda railway operations, improve trade facilitation and focus the investments in the physical plant. Attachment 1 Contents Principal Ratings................................................................................................ii Key Staff Responsible ............................................................................ . ..............ii Preface.............................................................................................................. v 1. Introduction and Background ......................................................................................... 1 Country Context..................1........................... Transport Sector Context .......................................... 1 Bank Role ...................................... .......... 2 2. Project Objectives and Preparation ............................................................................ 3 Railways.................... ....................... ..... 3 Adequacy of Project Designs and Preparation........................ 6 3. Project Implem entation............................................................................................... 7 Highways and Feeder Roads .......................................... 7 Railways.......... ......................................7 Procurement and Technical Assistance Issues......... .................. 9 4. Project Results and Bank and Borrower Performance............................................16 Achievement of Objectives and Results ................................ 16 Sustainability ................................................. 17 Bank and Borrower Performance. ................................. 17 Ratings .......................................... ....... 18 5. Lessons Learned.............................................................................................................. 19 Annexes A. Basic Data Sheet .................................................... 20 This report was prepared by Mr. Antti Talvitie (Task Manager), who audited the project in June 1998. Mr. William B. Hurlbut edited the report. Ms. Romayne Pereira provided administrative support.  111 Principal Ratings Fourth Highway Project (Cr. 1803-UG) ICR OED/ICR PAR Outcome Satisfactory Satisfactory Marginally satisfactory Sustainability Likely Uncertain Uncertain Institutional Development Partial Negligible Negligible Bank Performance Satisfactory Satisfactory Satisfactory Borrower Performance Unsatisfactory Unsatisfactory Unsatisfactory Railway Project (Cr. 1986-UG) OEDIPCR PAR Outcome Marginally unsatisfactory Marginally satisfactory Sustainability Uncertain Uncertain Institutional Development Substantial Substantial Bank Performance Satisfactory Satisfactory Borrower Performance Satisfactory Satisfactory Key Staff Responsible Fourth Highway Project (Cr. 1803-UG) Task Manager Division Chief Country Director Appraisal S.L. Kathuria Jonathan Brown C. Madavo Midterm N.A. N.A. N.A. Completion S.L. Kathuria Stephen Weissman F. Colaco Railway Project (Cr. 1986-UG) Task Manager Division Chief Country Director Appraisal Haley Goris Jonathan Brown C. Madavo Midterm Haley Goris Stephen Weissman F. Colaco Completion Haley Goris Stephen Weissman F. Colaco  V Preface This is a Performance Audit Report (PAR) on two transport projects in Uganda: the Fourth Highway Project (Cr. 1803-UG) and the Railway Project (Cr. 1986-UG). The Bank commitment to the two projects was US$25 million, of which US$2.2 million was canceled. A Project Completion Report (PCR) for Cr. 1803-UG was published on June 1, 1995 (Report No. 14554). A PCR for Cr. 1986-UG was published on July 5, 1994 (Report No.13300). The Operations Evaluation Department (OED) prepared this report based on a review of the President's Reports, Staff Appraisal Reports, PCRs, transcripts of Board proceedings, project correspondence files, Bank documents on other transport projects, and other Bank material. In June 1998, an OED mission traveled to Uganda to hold discussions with officials of the appropriate ministries and representatives of private and civil entities. OED also reviewed the projects with Bank staff. The PAR had four objectives: (i) to examine project preparation, including borrower participation; (ii) to assess institutional development, managerial improvements, and government commitment to reform; (iii) to evaluate project outcomes and the physical condition of the investments; and (iv) to draw lessons for the future. Following standard OED procedures, copies of the PAR were sent to the borrower for comments. No comments were received. 1 1. Introduction and Background Country Context 1.1 The World Bank began lending to Uganda's transport sector with two projects in the mid- 1960s. Then, during the prolonged political instability and civil violence of the late-sixties to the mid-eighties, Bank activity in the transport sector nearly ceased. During that time, civil service and social norms eroded, and private and human capital fled the country. When the audited projects were firmed up in the mid-eighties, the National Resistance Movement had established relatively peaceful conditions! The government had begun to install governance structures to facilitate social and economic activities and growth. Uganda underwent large changes between 1987 and 1993, when the audited projects were being implemented. Capacity building, commercialization, privatization, and other policies were initiated and pursued to improve administrative structures and the productivity of export industries. Nonetheless, administrative capacity remained fragile and the competitive position of Ugandan firms was weak. This is still generally true. Transport Sector Context 1.2 Five ministries affected Ugandan transport between 1987 and 1993: The Ministry of Transport (MOT) was responsible for policy, regulation, and coordination. It also was the parent ministry for the Uganda Railways Corporation (URC). The Ministry of Works (MOW) was responsible for planning, constructing, and maintaining highways. The Ministry of Local Government (MLG) had similar responsibilities for feeder roads. The Ministry of Planning and Economic Development (MPED) worked on overall investment planning. The Ministry of Finance (MOF) looked after financing and external aid. A large share of the transport service sector in Uganda was government-owned or -controlled. Another important factor is the predominance of agriculture: it was, and is, the most important industry. Ninety percent of Uganda's population of 15 million lives in rural areas. Both the rural population and agriculture industry were severely affected by the poor condition of the feeder road network. Assistance was sorely needed. 1.3 In the 1980s the transport sector of Uganda had a surfeit of institutions with thin capacity. Since then some "streamlining" has reduced the number of institutions involved in the sector. Recently, the MOW, MLG, and MOT were reorganized into the Ministry of Works, Housing, and Communication (MWHC). The Feeder Roads Department of the MLG and some departments of that ministry having to do with housing were transferred to the MWHC. Donor coordination continues to be done through the MOF, which has units to coordinate donors by sector. The transport unit of the MOF has made a 10-year roads plan with very ambitious goals and a budget of US$1.5 billion. Donors and international financial institutions have accepted the physical plan- 1. In preparation of the Fourth Highway project, during the Obote fl government in 1985, the project costs ranged from US$75 to US$ 125 million. The SAR was updated after the NRM government had established itself in 1986 with a program cost of US$60 million from which the project of US$21.8 million was extracted. 2 -with a significant financing gap-on condition that the road administration is transformed into an efficient, autonomous organization, Road Agency (RA). The first step has been the organization of a road agency formation unit (RAFU) to begin managing roads. 1.4 The Fourth Highway Project and the Railways Project were both necessary and important to rebuilding Uganda following its period of wars. Both had as a primary objective the improvement of institutions. The civil works components of the projects, meanwhile, aimed at protecting or recapturing past capital investments. Although those components were straightforward, they required-and received-much supervision time. The projects had numerous difficulties, ranging from preparation due to social disorder, backlogged needs and procurement to changed policies and donor coordination. The Transport Rehabilitation project (Credit 2587-UG), approved in 1994, continues the process begun with the two audited projects and the largely concurrent Third Highway project-approved in 1984 and closed in 1992- and gives particular emphasis to the sector's institutional reforms. Bank Role 1.5 After a 10-year hiatus, the Bank restarted its dialogue with Uganda on multi-modal transport issues in 1979. The main issues were easily identified: the condition of the infrastructure and the need for institutional development. At that time, the Bank saw itself as the leading institutional adviser to the government. The physical components were a vehicle to involve the Bank in the Ugandan transport sector under the new government, help protect and rehabilitate investments, and coordinate with donors. However, the Bank's real sectoral concerns were macroeconomic and institutional: international trade, transport administration and policy, organization and management ofgovernment institutions, financial and economic viability of government enterprises, and training and capacity building. As always, the Bank was not alone in working to rehabilitate Uganda's transport sector: several project memoranda note that loans and credits from other sources were available to the country. The Bank's comparative advantage, though, lay in its technical expertise in institution building, technical assistance, and training.2 Hence, the emphasis on these concerns in the highway and railways projects. 2. World Bank: Uganda Transport Sector Memorandum, 1983. 3 2. Project Objectives and Preparation 2.1 The two audited projects were prepared during a time of political uncertainty in Uganda. This, and the country's recent emergence from civil conflict, undoubtedly affected the design and preparation of the projects for limited objectives, still relevant, that were not explicitly connected to the Bank's broader concerns as expressed in the Transport Sector Memorandum of 1983. 2.1 The highway project had three objectives: to protect investments in roads, to help improve maintenance planning and operations of rural feeder roads, and to strengthen road institutions responsible for project implementation. Three components were intended to help meet those objectives: the rehabilitation of two trunk road sections, a pilot rural road maintenance program (PRRMP), and technical assistance and consulting services to help the MOW and MLG implement the project and conduct transport planning studies. 2.2 The method for the large road projects was clear: use international competitive bidding, but it was less clear for the other components. The PRRMP was a small, notional program to begin organizing maintenance of rural roads in Uganda. Uganda's feeder roads-and the highways, too-had been neglected for more than a decade. Institutional capacity was almost nonexistent. Agriculture, the most important industry, was severely affected by the poor condition of the feeder road network. Domestic crop distribution was important and agriculture products were the main exports. Attending to these complex issues in a still insecure political situation required intimate local knowledge. The task of creating the PRRMP-from rudimentary ideas for workshops, road camps, and equipment to rural road maintenance program-and the technical assistance component, were assigned to expatriate road engineers hired as part of the project's technical assistance program. Their TORs did not concentrate on institutional development to carry out their tasks, but focused on physical results on the ground. This probably was appropriate given the conditions in the country. The main roads component, the PRRMP, and the technical assistance centered their attention on physical objectives: procure main roads, equipment, and tools; build workshops and camps for the feeder roads and undertake some maintenance on them; and employ expatriate consultants to assist in the preparation and implementation of the PRRMP, to interact with the contractors and consultants for the two large road contracts, and to begin institutional development by means of road inventories and the like' Railways 2.3 During the preparation period, the IDA judged railways to be the "backbone" of Uganda's transport system and therefore vital to its transport sector. Three other notions should have also figured prominently in the railway project design. First, URC, by virtue of its comparative advantage in long-haul bulk traffic, could have been expected to concentrate its activities on the international routes. Second, the wagon ferry terminals at Port Bell and at Jinja offered Uganda some flexibility and competition in its outlets to the sea both to Mombasa (Kenya) or, especially, to Dar es Salaam (Tanzania). Third, the ferry services were URC's 3. SAR, paras 4.02, 4.03, and Annex 7. 4 preferred routes (at the time) to the harbors.' These considerations would have been a sound foundation for evaluating the project's physical components. In the end, however, the project's objectives and components were driven more by donor interests than by such matters as ways to improve the international routes, and the role of ferry services; however, the rehabilitation of the Kampala - Jinja road (in the Fourth Highway project) was an appropriate action to improve the land transport connections to ports. 2.4 Two, substantially donor-driven objectives emerged for the railway project: to implement a program of improvements in URC's management and organization, and to begin rehabilitation of the Kampala-Kasese line, pending major rehabilitation expected to be financed by a donor country. In spite of Uganda's difficult political economy situation, the audit believes that the government was short-circuited in the consultation process and the project was not informed enough by conditions on the ground. However, the Bank did take appropriate corrective actions at the mid-term review. 2.5 The railway project components encompassed several activities: * strengthening the organization and management of the URC-particularly its track repair and renewal unit, accounting department, and supplies management and telecommunications services-by providing technical assistance and training; * preparing a detailed design to accommodate wagon repair activities and a list of machine tools and fixtures for the Nalukolongo Workshop; * carrying out a study for a staff training program (financing of this component was transferred to the African Development Bank during project implementation); * repairing and improving staff living quarters in Kampala, Nalukolongo, and along the URC's western line toward Kasese; and * acquiring materials, tools, equipment, spare parts, telecommunications supplies, and vehicles and carrying out minor civil works to repair and rehabilitate the Kampala-Kasese line. An Initial Action Program (IAP) to strengthen the organization and management of the URC was agreed during negotiations and made a covenant to the credit agreement. The agreement included the following:' * developing general organization and management; * building up URC's track renewal and repair capability; * introducing monitorable performance targets; * improving labor productivity; * improving financial management, including clearance of the large arrears dating from 1982; and * improving materials management. 4. SAR, para 1.20. Port Bell was decided, built (by DANIDA) and opened during the Credit's implementation. The interviewed industry professionals indicated that improvements in land-transport connections were neglected. The rail link from Kampala to Malaba and the road to Mombasa (in Kenya) both suffered from the lack of maintenance. 5. SAR, Annex 11. 5 2.6 These objectives and components were based on common beliefs about how to make railways more efficient. These beliefs, and the IAP, were not analyzed and studied in the context of the URC during project preparation. Rather, they reflected an abstract model of what a good, autonomously managed railway company should be. It also was assumed that IAP would automatically be implemented by empowering the URC as an autonomous entity. 2.7 Participation by both borrower and partners in the preparation and design of the Railways Project was superficial. Partner coordination meetings did not discuss URC's problems and how to solve them. Rather, they were formal meetings in which partners delivered prepared views and proposed investments they could support. Little technical analysis was done. What analysis was done rested on assumptions about how quickly URC could be corporatized, and on traffic growth. The inertia inevitably encountered when trying to corporatize government departments was not viewed as a risk because corporatization had been made a covenant in the loan agreement. The demand on the rail system was insufficiently analyzed. The sole identifiable client for Kampala- Kasese line was a cement factory, the rest of the demand, it was assumed, would develop with agriculture. The expected ERR for the line was 34 percent and the expected FRR 24 percent (see Chapter 4 for results). Both expected rates of return rested on a wishful assumption of strong traffic demand for agricultural produce, but would not have helped the "concentration of URC's activities on the international routes." Thus, the investment in the Kampala-Kasese line was ill suited to a railway that was to be corporatized and commercialized. The rationale for rehabilitating it rested on the willingness of one donor to provide for its financing, not on its economic viability. 2.8 More important, the URC's problems and operating environment were too narrowly considered in project design, especially regarding international traffic, access to ports, and cooperation with neighboring Kenya and Tanzania railroads-precisely the issues IDA had identified as constituting URC's comparative advantage. The URC's institutional capacity, its competitive position, and its relations with Kenya and Tanzania regarding both traffic facilitation and rolling stock and its maintenance were known but ignored. Managerial actions that could improve service, reduce costs, and increase efficiency were not researched. For example, why did the system have excessive wagon requirements? (After the Kampala-Kasese line, the rolling stock procurement was the second largest component in the Railway Investment Program). Could the long turn-around time of trips to Mombasa be reduced? And what could be done to reduce derailments on the most trafficked sections of the system (Kampala to Kenya border). During implementation, but not in preparation, rolling stock maintenance and labor redundancy were adequately addressed. The project preparation reflected the needs and views of donors more than it did the reality of the situation in Uganda's railways. 2.9 Closely associated with the project was the (medium term, 1989-93) Railway Investment Program (RIP), drawn up together with the donors and the URC. The program identified investment needs, totaling US$1 51 million, and potential financing sources for half that amount. This ambitious program to replenish, upgrade, and modernize URC's physical facilities and rolling stock directed more than US$100 million to rehabilitating the Kampala-Kasese line and purchasing 13 locomotives and 700 wagons. The economic viability of RIP was not analyzed. It 6. SAR, Annex 12 7. Africa: The Great lakes Corridor Study. The World Bank. Africa Technical Department (1990) 6 mirrored needs, not demand. Several partners (donors) questioned whether the rail capacity was not already more than sufficient for the expected demand and whether a lowest total transportation cost alternative was being pursued, but these concerns were not addressed. Adequacy of Project Designs and Preparation 2.10 The IDA, coordinating with donors, designed for Uganda projects that consisted of highways and roads of economic significance, minor investments on railways, and support for institutional strengthening. This was appropriate under the circumstances. Nonetheless, both projects' designs suffered from the lack of knowledge how to begin institutional and physical development in Uganda's development stage. On the other hand, even now, it is difficult to prescribe a "knowledgeable" plan. 2.11 The engineering preparation of the large civil works for the highway project was appropriate, but the feeder roads pilot program and institutional capacity development components did not factor in the counterpart personnel capacity. Implementation, discussed in the next chapter, was slow and costly to the borrower. In retrospect, it is probable that the risks in the highway project's technical assistance component should have been thoroughly reconsidered after the violent change in government and given the unsatisfactory experiences in the implementation of the Third Highway project's technical assistance component. 2.12 The railway project was under-prepared. The Kampala-Kasese line was not thoroughly analyzed and the institutional component was structured around a conditionality to corporatize the railway. While the idea may have been appropriate, a borrower emerging from a civil war should not have been expected to have the capacity to accomplish it-or the LAP-without considerable guidance and help. 2.13 The destruction of physical capital and reduced human capital resulting from the civil war made these projects important for Uganda. Both tried to quickly restructure Uganda's institutions and pursued implementation of civil works, but several limitations, partly the result of the country conditions, reduced the potential of the projects. First, although the two projects were concurrent and the interdependency of road and rail was acknowledged in Bank reports, the projects were implemented independently without reference to the transport sector plan. Second, especially for the railway project, the voices of stakeholders within Uganda were heard but little during project design and implementation. The donors were the main architects of that project. The government officials were secondary actors. Institutional issues were not tackled during preparation. Third, the significance of transport operations in the neighboring countries was not factored into project design, especially in the railway project. 2.14 Changes as large as those envisioned by these projects cannot be implemented quickly. The corporatization of the URC is a case in point. It took nearly three years longer than planned to approve a change which was more in name than in function. Institutional changes, such as corporatization of a government agency, require large investments of time and effort with the client. New and complex ways of procuring even small items are time consuming. 7 3. Project Implementation Highways and Feeder Roads 3.1 The two main highways of the Fourth Highway project were built, but with delay. PRRMP and the maintenance of rural roads remained problematic. According to the ICR, about 190 km of rural/feeder road was rehabilitated, compared to a targeted of 108 km. But "procurement of civil works was cancelled ... because the feeder roads that were scheduled to be maintained were in such a state of disrepair that they needed complete rehabilitation rather than maintenance..." Consequently, no road camps or small workshops were built because of delayed plans and a change in government policy, late in the project, to carry out feeder roads maintenance by contract in the future. Some road maintenance and workshop equipment was, however, procured just before the project closed. Half the planned US$1 million was disbursed. Because of delays and changes in government policy, and declining road maintenance budgets- of which the Bank regularly admonished the government-it is probable that PRRMP had but little effect on road maintenance capability. The causes of delays centered on procurement and contractor and consultant performance. 3.2 In sum, institutional development under the project was less successful than envisioned at appraisal. Poor and unstable consultancy and changed government priorities, partly originated by the Bank, all slowed down institutional development. Railways 3.3 Easing transport bottlenecks on principal import/export routes and upgrading rolling stock maintenance capabilities were partly achieved through donor-funded projects. The repair of the Kampala-Kasese line was shelved at midterm evaluation because traffic demand was low and the donor funding for complete rehabilitation of the line fell through. From midterm onward, institutional development was given increased emphasis. 3.4 Positive outputs from the IAP helped strengthen the organization and management, and railroad maintenance of the URC. Most IAP components were carried out on schedule. The exception, the Railway Act establishing the URC as a commercially managed organization, was delayed three years and had only a minimal effect on the project. However, the LAP failed to arrest the decline of URC's traffic. In 1992, Bank contributed to the formulation of an emergency action program intended to arrest the decline in traffic, market URC services more effectively, and carry out various other initiatives. The management and staff training was partially successful, but although performance did improve, URC was not financially viable at loan completion. 3.5 The management of the railway project, like that of the highway project, was sometimes untidy. Much time in this project was devoted to procuring inputs, dealing with the procurement 8. ICR, Appendix B, Borrower Contribution, Part B: Rural Feeder Roads Program. 9. The initial feeder roads maintenance program was submitted for IDA review in late 1990 and the proposal for road camps and workshops two years later. Nearly a year later, in July 1993, this proposal was withdrawn as unnecessary because most feeder roads were to be maintained by contract in the future. 8 process, and managerial coordination with donors trying to secure financing for RIP rather than to overseeing results and evaluating whether the project was responding to demands and needs. Late in the railway project, for example, after the corporatization of URC, a substantial donor wrote in a letter to the government of Uganda (copied to the Bank): "What has disappointed us.. .is the failure of URC to improve its overall performance.... The outstanding economic function of URC consists in providing transport for Uganda's foreign trade in cooperation with the neighbouring Kenya Railways Corporation (KRC) and Tanzania Railways Corporation (TRC). However, so far URC has not only failed by far to make full use of its market potential. After a temporary rise until 1990 the company has even incurred losses in cargo volume....This development was caused by URC's own deficiencies and...[those] of the neighbouring railway corporations....We also see a major cause for the low competitiveness.. .in their restricted commercial freedom and capacity as a result of their dependence on government administrations." In other words, the project was not consistent with the outstanding economic function of URC, and the corporatization alone did not bring about the desired results. To do so, it should have addressed, among other things, the provision of rail service-including trade facilitation--east of Kampala and from northeast Uganda to Mombasa and Dar es Salaam ports. 3.6 In the same letter the donor questions the neglect of the land routes in favor of the link over Lake Victoria: "We have been observing that due to the overall drop in transport volume, but also because URC has steadily reduced traffic on rails in favour of the waterways (after Port Bell was opened), the capacities of the Ugandan locomotive fleet and the new Nalukolongo workshop are being used only to an unsatisfactorily low degree. This is causing considerable losses to URC and the Ugandan economy..." This comment gives the supply-side dimension to the problem: inefficient use of resources. The round trip from Kampala to Mombassa took about 28 days; today the URC wagon turnaround time in KRC is 26 days (so the Kampala-Mombasa- Kampala run is probably the same 28 days). The distance from Kampala to Mombasa is only 1300 kIn. The round trip should take no more than 10 days, the current performance target being 12 days. Reducing the turnaround time would have an immediate impact on availability and demand for locomotives and wagons-halving the turnaround time would halve the wagon requirement and so on. Instead, the URC bought new locomotives and wagons and service to customers was not improved.'0 The same donor who wrote the letter proposed the privatization of the new Nalukolongo workshop, which has been done, and that the workshop's trained staff serve purposes other than URC locomotive rehabilitation, including KRC, TRC, and wagon maintenance. 3.7 Clearly, therefore, insufficient attention to the problems and capacity of the URC during project design affected project implementation. Furthermore, the physical component of the project was not well integrated with the technical assistance. The midterm restructuring and the resulting emergency action program helped, focusing URC management attention on marketing, billing and accounting, marine maintenance, supplies management and procurement, and improvements followed. But experience shows that project design needs to include components that address immediate concerns and that technical assistance and management training are most effective when they deal with those concerns first. 10. The Bank initially opposed building the Port Bell terminal, considering it excess capacity. However, the donor- funded Port Bell provided an alternative and a better route to Dar es Salaam port, and hence improved competition to Mombasa. It was a worthwhile investment. But it should not have been a substitute for improving the all-land connection to Mombasa and for much-improved trade facilitation. 9 Procurement and Technical Assistance Issues 3:8 The results of both the Fourth Highway project and the Railways Project were diminished by two problems in implementation: procurement and technical assistance. Those problems arose because of limited borrower capability. The two issues are examined below in the context of the highway project. Procurement 3.9 The bids for the Mbarara-Ishaka and Kampala-Jinja roads, the two main roads rehabilitated in the project, were received February 12, 1987. After several extensions of bid validity, the contract for the former was signed February 15, 1988, and for the latter February 10, 1988, about one year after the bid closing date. A long exchange of letters, between the government and the Bank preceded both awards. That exchange included the affected Bank Executive Directors. 3.10 The three lowest bids for the Mbarara-Ishaka road were within 2 percent of each other. Two bids (T and C) contained errors that required interpretation and clarification and uncertainties and questions regarding taxes. Contrary to the bidding documents, one bidder (T) had two sets of costs for some procurement items, lump-sums and unit costs, and had arithmetic errors in calculating the bid amount. Two of the bidders (T and C) had letters of discount included in their bids; the discount offer for the third contractor (S) was "found" three days after bid-opening and was not, of course, read at the bid opening." The bid evaluation consultant accepted all letters of discount as valid and, without showing it to the government, submitted the bid evaluation report to the Bank with S as the lowest bidder. In. that evaluation the unit costs, not the lump-sums, were used in calculating T's bid. In April 1997, a month later, MOW informed the Bank that they had not had chance to "peruse" the evaluation report and that S's discount letter was not valid because it was not rubber stamped as the others and recommended the award to contractor C. However, in a change of mind, the Government concluded in July that the S's discount offer was valid and recommended that it be awarded the contract. Meanwhile, T was protesting the acceptance of S's discount offer and claiming to be the lowest bidder when the lump-sum quote was used. 3.11 A lengthy correspondence followed including the Executive Directors. The government continued to favor contractor S and claimed that the multiple cost items and confusion about taxes were errors of commission to mislead the government, "a deliberate attempt by the bidder to have two options in his bid."l2 The consultant evaluating the bids continued to recommend that the second lowest (undiscounted) bid (S) be accepted because it was error-free, had unambiguously included the taxes as advised, and contained no suspiciously low unit costs.13 The government concurred with the Bank but demurred: the lowest (calculated) bid, that of contractor 11. S's discount offer was dated February 23, 1987, not February 13 the bid opening day. A consultant's engineer claimed to have seen the discount offer on February 16, 1987. 12. Telex to the Bank, Sept 25, 1987. 13. As late as December, 1987, the consultant claimed that "[contractor TI had not included relevant taxes", however the Government sought clarification and T contraindicated that claim. 10 T, was chosen. 14 In its acquiescence the government said that it, in addition, had reservations about the firm's ability to mobilize and do the work. These reservations were not totally without foundation. The same company was also the lowest bidder for the Kampala-Jinja road, underbidding the engineer's estimate by 30 percent. In July 1987, the bid evaluation consultant for the latter project-not the same one as for the Mbarara-Ishaka road-had expressed to the government a fear that the low-bid contractor "may not be able to complete the works for the quoted amount" and that the "time of mobilization is likely to be much longer." He cited the contractor's lack of mobilization in Uganda or any neighboring countries and the fact that similar de facto parastatals exhibited delays in mobilization and beginning physical works. In another communication the government told the Bank that "past performance in Uganda and ability to commence work promptly... [are both] Government priorities at the moment." 3.12 The review of Mbarara-Ishaka road bid evaluation continued even after contractor T was chosen and the contract signed in February 1988. The consultant explained to the Bank in a March 1988 letter, under the heading "Double quotation for Lump-sum Items", that the uncertainty [in evaluating T's bid] was due in part "that one of the bidders [the winning bidder] had entered different amounts for the same items in the 'rate column' and the 'amount column'. In this area our project staff did not know of the flexibility of Bank procedures which is in the interest of the Client..." and in part resulted from "surprisingly low quotations for daywork rates and pay items when compared to the engineer's estimate. The order of magnitude caused suspicion amongst our project staff as to whether taxes and duties were included at all leading to enquiries to the two bidders for whom it could be questioned."15 Later, the consultant reversed his earlier opinion, and states that these inquiries were not justified for tender evaluation purposes. A Bank memo'6 states that the consultant "admitted that the firm had made mistakes in bid evaluation... [and] promised to advise the Government of its errors" and the Bank "does not intend to take further action in this matter as the firm has admitted its errors." 3.13 In short the government and the consultant wanted to choose the bidder that had the lowest error-free bid and would be able to mobilize promptly. The presence of the two lump-sum and unit cost estimates for some items should have been sufficient for the rejection of T's bid. In addition, the fact that the government was concerned with the firm's prompt mobilization and ability to perform, and that it was a parastatal certainly are factors not weighing in favor of bid acceptance. 3.14 The selected company [T] had promised to employ a grader at the road site at the commencement of the contract March 1, 1988. That did not take place. In May 1989, when 80 percent of the contract time had elapsed, only 5 percent of the physical work had been done. Ultimately, after hand-wringing and threats of contract termination, the contractor accepted two 14. In a telex, November 1987, the Permanent Secretary writes: "...We note however that you [the Bank] do not agree with our views. Although we are at variance with your recommendation, we have to yield to your decision in order not to bog down the project any further. We would however still like to put on record that our recommendations were based not only financial considerations but also on technical and practical considerations which came to light during bid evaluation and after prequalification. We remain very pessimistic about [the contractor name] competence and ability to undertake and mobilize promptly for a project of this nature. Given that it will be their first time in east Africa.." the telex goes on to itemize the specific concerns and anticipates delays. 15. Letters dated March 7 and March 25, 1988. In Bank's opinion the rates were not unusually low. 16. Bank memo to files, April 8, 1988. 11 subcontractors, already mobilized in the country, to help complete the road works two years later than planned, and agreed to pay (some)liquidated damages. In the minutes of a November 1991 meeting to discuss "stage of work and outstanding items," the Managing Director of the company attributes some of its problems to "[a] changing political situation... [which] has resulted in managerial problems... .Originally [the company] was a parastatal company with full government backing. It now was a privatized firm... in the transition period it was difficult to get foreign funds released."17 3.15 The Bank may have been right on procedures, but the delay in Mbarara-Ishaka road procurement was costly. According to the benefit-cost analysis, the (discounted) benefits lost to users in the first two years were close to US$3 million. Furthermore, works normally commence within 2 weeks to 2 months after bid closing, the one year long bid evaluation period added 0.5- 1.0 million to the benefits lost. This is 50 percent of the construction costs of the road. The transaction costs, not included in these figures, were also high. The government was within its rights, and not inconsistent with Bank procurement guidelines, to reject a bid that contained duplicate prices. In view of the explicit concern for the contractor's timely mobilization, performance and independence there was a need for reverification of the prequalification criteria and eventually requiring a higher performance security, especially when there was a 'clean' bid within one percent. 3.16 The Kampala-Jinja road bids presented slightly different problems. The three lowest bidders were 33 to 22 percent below the engineer's estimate of approximately US$8 million. The government and its consultant recommended the second-lowest bidder partly because the low-bid contractor was a parastatal, and partly because, in their view, the low bidder would not be able to complete the works for the quoted amount or to mobilize quickly. Furthermore, the Kampala- Jinja road is the busiest section of the northern corridor in Uganda and, in the government's words, "we would not like to take chances where delays appear obviously inevitable."" The Bank called the two bids "competitive." No additional performance security was required. The correspondence between the Bank and the government took a long time and the low-bid company withdrew its bid in September 1987. The second-lowest evaluated bid was chosen. Even that bid, at 22 percent less than the engineer's estimate, was apparently too low. In 1991, one year after the contractor maintenance period, the road began to show signs of longitudinal and transverse cracking. A stretch of 800 meters had to be strengthened and resurfaced. The repair costs estimated at about US$0.8 million (done by the same consultant that supervised the original works!) brought the total cost of the road rehabilitation to about US$7.2 million, or within 10 percent of the engineer's estimate. A protracted procurement process followed for the repairs-no institutional learning had taken place in the earlier experience-and the repair works were not completed before the loan closing December 1994. The MOWTC road database shows that the road was resealed again in 1998 at a cost of US$2 million.'9 17. Minutes of a meeting November 26, 1991. It might be desirable for the Bank to adopt clear guidelines when and where parastatal companies can bid for works. The situation described above is not an isolated incident. 18. Telex to the Bank August, 1987. 19. Neither the strengthening nor the resealing are accounted for in the ex ante or ex post cost-benefit calculations. This shows them to be unexpected repairs. 12 3.17 This case shows, once again, that very low bids may not be 'competitive' in the true sense of the word. A very low bid should be examined closely and the reasons for low unit costs ascertained. If a clear explanation cannot be presented to satisfy such concerns, a higher performance security should be required. The delay in awarding the contract, resulted in a loss of benefits-approximately US$2 million over and above the repair costs of the new road. Both procurements show that long delays in contract award and contractor performance result in transaction costs and large benefit losses to the users, totaling in these cases in excess of US$5- 6million.20 These costs should not be ignored in procurement or ex-post benefit cost analyses. As observed in numerous other projects, the costliness of procurement delays would indicate a need for timely technical assistance to avoid them. Technical Assistance 3.18 Two technical assistance issues stand out: (a) Why did the consultants fail to develop (road) plans and procedures that would have client ownership? (b) Why did they fail to train local practitioners who could take over after a training period? Unfortunately, the project files do not support such an inquiry and reconstruction of facts-always a formidable task-is impossible. The ICR for the highway project mentions as lessons learned that "closer supervision would have brought to light the inaction on the part of the borrower/ implementation ageny(ies)/consultants" and "the terms of reference of technical assistance staff should include a detailed work program with verifiable inputs and outputs for each individual staff." But whether awareness of inaction or better terms of reference would have led to better results cannot be known?' It is clear that the processes followed in procurement, technical assistance, and capacity development did not resolve all these difficulties because they continue to persist. 3.19 Often the issues of technical assistance and project preparation and design were viewed as problems unrelated to the processes and procedures to execute them?2 There was not enough assessment of the available technical capacity and the extant processes followed in Uganda. The client felt that much time was spent on studies and getting approvals and that the Bank's 20. In spite of the delays, according to the ICR, the disbursements ran ahead of the planned. In part this is due to the slow disbursement profile used, but mostly to the implementation schedule. According to the SAR, Chart II, the execution of work was planned to start ten months after bid closing. This is an unreasonably long time and few contractors extend bid validity beyond three to four months. In general, disbursements in Uganda's projects are long and preceded by long procurement periods. It applied to every procurement in the audited projects. In a contemporaneous Uganda project, the South West Agricultural Rehabilitation project, agricultural feeder roads were the largest project component. It was implemented in reduced form two years behind schedule and asynchronously with the agricultural component, thus seriously undermining the benefits of that project. The local professionals claimed that the Bank's lengthy procedures-for appraisal, then for planning and prioritizing, then for employing design consultants, then for Bank review of designs, and then for procurement of contractors-was ill-suited to prevailing conditions in Uganda. By the time the contractor mobilizes, the roads are no longer in the condition they were at the design stage, and a new design and rehabilitation, not maintenance, is required. The lengthy process results in large transaction costs, loss of benefits, frustration, and disillusionment with aid. 21. There is much evidence against these beliefs (however commonly held) in Uganda and elsewhere. For example, awareness of performance indicators did not improve performance; consultant plans for rural roads were not implemented; awareness of the contractor's inaction did not make him active despite numerous letters. The reasons for resistance were elsewhere. 22. This conclusion is drawn from the TORs for the technical assistance engineers and the design of the project components. 13 procedures were cumbersome and designed to exclude-by TORs and prequalification-both the local consultants and the local contractors. For example a case cited in which the studies and approvals took so long that the feeder roads component was out of phase-four years late-with the rest of the project and resulted in loss of benefits in many markets? 3.20 There are, however, technical successes that are yet to be felt in the practice of road management. These successes were not results of the audited project. For example, Uganda now has an operational maintenance management system (MMS).4 This will help make the evaluation methods responsive to local needs, not the needs of the donors who have had their own methods and wanted data tailored to them. Another maintenance management system was built up in the Third Highway project, but no capacity was developed to support it in that project or in the audited Fourth Highway project. To sustain good road management, the database for all classified roads, including feeder roads, systematic road condition surveys, and traffic counting schemes must be installed. A data support unit should be created to establish MMS as a management tool. These activities appear to have been part of the Fourth Highway project but apparently were not undertaken." 3.21 But this is not enough. Management training should ensure that MMS is an integral part of annual program planning and used in day-to-day decisionmaking. Program management and monitoring from the viewpoint of the road administration's central management, using the MMS, needs training and practice. In order for the road administration's districts to make link-specific (engineering) proposals and decisions requires training at the district level. This training should include usage of HDM-like project model, simplified to meet the local data and human resources, project management for mechanized routine maintenance, and project supervision and management by contractors. 3.22 While the railways project made progress on institutional and managerial development, much remains to be done before the URC can become a commercially viable company. However, there are positive results: improved performance, and within the Bank and URC there is an understanding of priorities: continue commercialization through concessioning and management training; address trade facilitation, especially at borders and ports; reduce the workforce, making redundancy payments if necessary; rehabilitate and improve maintenance of commercially viable track and wagons; and improve wagon turnaround time. These could form the core of a new railways project. 23. A recent audit for an agricultural project in Uganda calculates that delays in the implementation of its feeder roads component reduced the ERR from 17 percent in the ICR to 3 percent, and even that estimate was optimistic and based on the assumption that the roads yield benefits without maintenance. Of course, the ERR calculations do not take into account the benefits lost due to delays. Because the ERR calculations are internal to the project timeline, delays show up as savings in the cost of capital while taking no account of the benefits not received in real timeline. For this reason, delays increase the ERR and, furthermore, ERR should not be used to decide project timing. 24. As part of that MMS development, MWHC now has a database for classified roads. But it is not up to date and has no regular traffic counting system attached to it. There also is a database for feeder roads, but it is updated mostly in response to donor needs to analyze and choose the roads it intends to finance. 25. The preparation of "inventories of road networks" and the "reviewing and evaluating highway inventories ... traffic counts.. and other data" were in the TORs for the rural roads maintenance engineer and (two) road design engineers (SAR Annex 7). There is no evidence that such activities were carried out even though it would have logically followed from the MMS development in the Third Highway project. 14 3.23 Institutional and managerial development in the Fourth Highway project achieved small results. A substantial reorganization is, however, being undertaken: MWHC is to become responsible for (macro) planning, policy, and regulation; and, the Road Agency (RA) is to become responsible for planning and design, construction, rehabilitation and maintenance, and program and project management. In broad terms, the RA is an autonomous entity responsible for managing Uganda's roads. 3.24 The client expressed a desire that the Bank would coordinate projects better. It was indicated that Uganda is doing its own portfolio reviews by looking all the projects at the same time to synchronize them. The task managers/TTLs were seen too concerned with the implementation of their own project, not seeking to get maximum impact of all the projects. 3.25 There is a concern for capacity building within government administrations and agencies, and for local consultants and contractors. When positions and works are advertised for worldwide competition requiring 10-15 years of experience and sizeable financial capability, local capacity cannot be expected to develop and motivation, the paramount factor in its development, is discouraged. 3.26 To gain an understanding of the views of the Ugandan private sector on the issues of transport sector priorities and governance, and the reasons for the lack of success in creating a local construction industry (the Third Highway project had a component intended to do so), the audit mission interviewed two people: a representative of an Ugandan trade association and the president of a local construction firm who also was a board member in an Ugandan manufacturing association. The most interesting points made by those interviewed follow:26 * For the trade association the issues are clear: improved trade facilitation to aid exports, and better roads overall; especially important was the speeding-up of rail service between Mombasa and Kampala. The association felt that both MWHC and URC were inefficiently run and the quality control of works was poor. The procedures for tender are lengthy and complicated. * The local construction company that was interviewed had been registered in Uganda for 29 years and has more than 2,000 employees, 1 percent of whom were foreigners.27 Nearly all the engineers were native Ugandans, worked well, and were paid well. The company also has good relationships with the tribes because of a tribal distribution of its workforce. * The company had not worked in Bank-supported projects, but was interested in roads and civil works: roads, bridges, retaining walls, basic civil engineering for which it has competence.28 * The company could be effective and operational on feeder roads right away because it already owns the necessary equipment and has access to materials from its own quarries; it also would be interested in maintenance contracts. 26. Although both were single-person interviews, they represent many voices (those of the associations). What is singular about the interviews is that they succinctly state the conclusions of this audit. 27. The company president who was interviewed was a foreign national who has lived in Uganda for 30 years. 28. Once the company had acquired the bidding documents. Calculating backwards they deduced that they could not deliver at the rate required in the bidding documents and, therefore, did not bid. 15 * The company was also interested in building main roads if the lot sizes are in the range of US$1-3 million. In the past, they could not compete because the lot sizes were too large and because the "foreign parastatals" underbid them by 30-40 percent.9 * The interviewee estimated that with lot sizes ofUS$1-3 million, four to five local construction companies would bid for the works.0 3.27 These interviews, as well as others conducted for this audit, raise the following issues, questions and comments: * The Bank may enforce its procurement guidelines too narrowly and that makes the procurement process costly and slow, at least in the beginning. Is that process too complex and demotivating for newcomers? In the early stages, many countries are not ready for state-of-the-art procedures. It might serve development better to incrementally improve the countries' procurement processes with the aim of gradually meeting the Bank's procurement guidelines. * Corruption may be a problem." Does the Bank's procurement style and processes inadvertently favor non-transparency and rent-seeking behavior? The way this works in Uganda is unclear. The factors most commonly suspected by the interviewees were prequalification and the quality of works. For prequalification, relationship-building is more important than technical and financial competence. In quality, corrupt practices range from no work done for (shared) payment to inferior work quality for full payment and shared use of the "surplus." In some cases budget money may be diverted to personal use because accountability processes are lacking. In any case, rent-seeking behavior is a complex problem that has two parties: the taker and the giver. In any solution, the motivations of both parties and their incentives and means to end must be addressed. * Should all consultant and construction works include a local partner in a major capacity? 32 To stimulate local private sector interest and planning, MWHC should publish a realistic five-year plan indicating what works and likely lot sizes will be up for tender, when and how, and circulate it for information. * In order to increase management accountability, improve the cost-efficiency and quality of works, and reduce possible corruption, performance indicators with benchmarks, including the trend in road condition, need to be developed, monitored, and publicized. This is as important as the road management systems (MMS) now being adopted. 29. The company president added wryly, "of course, there were problems with these contracts; sometimes the job was not finished, often the quality was poor, and there could have been corruption." 30 "Slicing," or smaller lots, may in some cases be economical and could be allowed to be decided by the bidding process. However, many small contracts require local capacity to manage contracts. If such capacity is absent, as may be the case in Uganda, costly delays will ensue. Thus, "slicing" decisions should be made in the context of sectoral and national priorities to have the roads in service quickly, or more gradually to develop national capacity to manage and supervise contracts, and learn new technology. These considerations should be part of (Bank-supported) project design. 31. Development News, in an article on September 11, 1998, includes a reference to a donor's report that wams the Ugandan government that its "funds could dry up if it fails to take action against corrupt officials." 32. In the United States, for example, government-supported projects are required to include minority (sub)contractors. 16 4. Project Results and Bank and Borrower Performance Achievement of Objectives and Results 4.1 The outcome of the Fourth Highway project was marginally satisfactory. In making this judgement it is recognized that the project was identified, appraised and implemented in difficult circumstances. There was a change of government and it took two years from appraisal to approval. The situation and personnel had changed -were changing- reducing the quality at entry. Nonetheless, the civil works components of the limited objectives the project pursued were achieved. The two main roads selected for rehabilitation were completed albeit with delay and deficiencies in procurement and contractor performance. The PRRMP suffered from the shortcomings in counterpart personnel and consultants' lack of performance. Those difficulties, although anticipated, were not resolved satisfactorily during the project and have continued until today. The quality of the civil works was uneven and some work required early repairs. The pilot rural roads program did not achieve all its objectives. No tangible institutional development took place because of changing priorities, and reduced and irregular consultancy presence and imperfect functioning. 4.2 The outcome of the Railway project was also marginally satisfactory. The Railways Project relieved transport bottlenecks on principal import/export routes and improved track and locomotive maintenance through donor-funded projects. The interim rehabilitation of the Kampala-Kasese line was shelved at the midterm review. Funding shortages, meanwhile have delayed sustained results from the assistance to strengthen the URC's maintenance management. The results of the project training component were positive. Clearly, URC is now better managed, more aware of its problems and realistic about priorities. Yet it is not a financially viable corporation. Four of the seven performance targets set in the SAR for 1993 were achieved. The shunter locomotive availability and two ferry-related indicators were not achieved because of a required mechanical overhaul and accident repair. The improved performance has not been completely sustained. In 1997, only one performance target is still achieved (the shunter performance target not achieved in 1993!) and performance on four indicators has declined since 1993. But, excepting one ferry performance indicator, all others are better than before the project. In addition, and what have been sustained achievements, the number of employees has been reduced from 7,500 to 3,500 at loan closing and to 1,960 at present, with corresponding improvement in staff productivity, and the Nalukolongo workshop has been modernized and privatized (after project completion). 4.3 Rates of return for the trunk roads were somewhat lower than estimated at appraisal: the return on the Mbarara-Ishaka road, estimated at 30 percent, was actually 17 percent and the return on the Kampala-Jinja road, estimated at 63 percent, was actually 61 percent. The reduced returns were primarily because of lower than expected traffic volume. As discussed earlier, the ERRs do not reflect the benefits lost because of procurement delays and the supplementary maintenance actions; no ERR was calculated for the PRRMP. The Railway Project's small physical component, the Kampala-Kasese line, did not attain the economic benefits anticipated at appraisal and was shelved at mid-term evaluation. There were many reasons for this, including lower than anticipated traffic, lower than anticipated investment from a donor, major devaluation of the shilling, changing priorities, and difficulties to secure counterpart financing. 17 Sustainability 4.4 The lack of a transport policy for Uganda makes sustainability very hard to evaluate. In 1979 discussions with Uganda the Bank identified a need for a major overhaul of the country's transport sector. The 1983 Transport Sector Memorandum recommends improvements in "transport administration and policy making" and in data systems and planning; coordination between Uganda and Kenya was seen important, "particularly as it applies to rail transport." There was a need establish URC as "a financially autonomous agency with necessary increase in capitalization." These problems continue to persist, although there is, in the Bank side, a clear view that the railway need to be concessioned, and a better administration and an autonomous management is needed in the road sector; the latter having been recently established. However, and this may be due to the unstable situation at the time and thinness of human resources in Uganda, both projects' designs suffered from the lack of a product or products and from the lack of an incremental plan for institutional development that took into account Uganda's development stage. Both of these also reflect the Bank of that time. Nonetheless, several operations later, institutional capacity and policy formulation have not yet substantially improved. The institutional restructuring that took place was the introduction of some commercially oriented management practices in the URC with consequent improvements in URC performance. Little progress has been made on rural roads and rural public transport (which has been ignored), both of which are critical in a country where 90 percent of the population is rural. There are hopeful signs with the recent (initial) formation of the Road Agency and some movement toward improved management of the URC. Finally, the significance of operations in neighboring countries have not been factored into Uganda strategy. Ugandans are paying for inefficiencies in Kenya and Tanzania irrespective of how efficient they could become within their borders. For all these reasons sustainability has been rated as uncertain in both projects. Bank and Borrower Performance 4.5 Bank and Borrower performances are harder to rate than the projects themselves. These were first transport projects since the Museveni government assumed power (the Third Highway project was started under the previous government but most of it was implemented during the new government). There was instability in Uganda, and the institutional capacity and human resources were thin. The development of Borrower capacity is centrally dependent on the Bank and its consultants' ability to impart knowledge and skills. These obstacles to development cannot be overcome easily, let alone in one project supervised at a distance. 4.6 In the Fourth Highway project the main roads component was well prepared, but its implementation had very high transaction costs and severely eroded its benefits. The design for the feeder roads pilot program and institutional capacity development components did not work as planned. Both components suffered further because the institutional problems involved and processes needed for their solution could not be clearly defined under the then prevailing circumstances in Uganda. Much supervision time was devoted to procurement and too little time to feeder roads and institutional development, both necessary for economic and social improvement. The performance of technical assistance consultants was inadequate and their country presence unstable; there are several references in Bank communications to inaction on the part of the consultants. On the positive side, the project files contain the task manager's reflections and periodic assessment of project progress. The concept is excellent and worth emulating. Why, though, did management not act when important project objectives or 18 recommendations made by the task manager or in the Transport Sector Memorandum were not being achieved? Overall, given the unstable and difficult circumstances in Uganda at the time, and the thinness of the client's human resources, Bank performance was satisfactory in the Fourth Highway project. 4.7 In the Railway project Bank performance was also satisfactory but not optimal on several grounds. Before the mid-term 'course correction' the project faltered. It was under-prepared and, obviously, the project design too complex for its size. The Kampala-Kasese line was not thoroughly analyzed and the institutional component was structured around a conditionality to corporatize the railway. While the idea was appropriate, a borrower emerging from a civil war should not have been expected to have the capacity to accomplish it-or the LAP-without considerable consultation. Several times early in the project, the assessment of the situation by Bank management contributed to the failure to redirect the project to appropriate objectives and positive results. The Bank did act decisively at midterm. It shelved the unprofitable Kampala- Kasese line rehabilitation and the RIP, and redirected efforts to institutional development with real results. There were tangible improvements in marketing, costing, billing, supplies management and permanent way maintenance and, most importantly, there was learning. Performance improved.3 Perhaps no more can be achieved in a small project. The strong finish merits the rating of Bank performance as satisfactory. 4.8 Borrower performance was satisfactory in the Railway project but unsatisfactory in the Fourth Highway project. In the railway project, in spite of the shortcomings noted, the borrower did execute the project and improved URC performance. For the Fourth Highway project there are many reasons for unsatisfactory performance: inadequate funding of road maintenance; delayed and costly implementation; weak results in institutional development in spite of the presence of four expatriate advisers; and finally, there may have been attempts at corruption, but actions have now been taken that seek to rectify the institutional environment that may have permitted it. There were lapses in timely processing of the audit reports in both projects; however, most other covenants were complied within reasonable time. Ratings 4.9 In summary, the PAR ratings for the Fourth Highway project are as follows: outcome (rated satisfactory in the ICR) is downgraded to marginally satisfactory. The ratings for institutional development and sustainability remain the same as in the ICR: negligible and uncertain, respectively. The ratings for Bank and Borrower performances also remain the same as in the ICR: satisfactory and unsatisfactory, respectively. The PAR ratings for the Railways project remain the same as in the PCR excepting outcome which is upgraded to marginally satisfactory (marginally unsatisfactory in the PCR). Sustainability remains rated as uncertain and institutional development as substantial-although managerially the URC is still not an autonomous organization. Bank and Borrower performances are rated satisfactory. 33. Performance indicators, PCR, Part III. Statistical Information, p. 28 and URC Annual Report 1997. 19 5. Lessons Learned 5.1 The main lessons learned relate to project design and the implementation processes: * Quality at entry should start with a shared understanding of potential problems and mutually agreed, concrete results, even if they are small. This should be based on an assessment of political economy considerations and institutional capacity. The significance of circumstances and operations in the neighboring countries should be factored in. * Project complexity must be scaled to borrower capabilities. Institutional capacity was a problem in both audited projects. It was studied during project preparation but insufficient attention was paid on the ability to produce outputs and results. Project preparation should include sufficient dialogue with the counterparts and with donors on issues pertaining to outputs and results, including anticipated problems and the project designed to deal with them. * Changing a parastatal organization, such as URC, to a market-oriented organization operating under commercial principles may be an impossible task; a contention supported by experiences elsewhere. A satisfactory outcome for commercializing a parastatal (railways) comes through privatization-concessioning (possibly coupled with "unbundling") its operation and physical plant to a private entity-not from efforts to transform it without privatization. Commercial sustainability is ensured only when there is competition, accountability, and incentives that foster ingenuity and efficiency, and allow custody of the benefits from improved productivity and performance. * During implementation, achievement of results need to be reviewed periodically with the client. The client's views should be central for both finding improved ways to achieve results and for removing institutional resistances for change. Especially important is to make sure that procurement is not delayed without institutional learning. Timely technical assistance to avoid procurement delays and ensure learning should be provided. * Procurement is a development issue. In infrastructure sectors, both in consulting and contracting, procurement should not just help the client comply with Bank procedures, it should help to develop markets and capacity in the client country. Procurement delays can be expensive for the client and prevent achievement of results. * In technical assistance, management training should have top priority. Process (on-the-job) training of responsible managers should not be equated with the development of a computer program or management systems, but with the use of management systems in sector program development and evaluation, and daily application in road or railway management. Supervision of technical assistance and civil works must focus on results and not on inputs. 20 Annex A Basic Data Sheet FOURTH HIGHWAY PROJECT (CREDIT 1803-UG) Key Project Data (amounts in US$ million) Appraisal Actual or Estimate Current estimate Total project costs 21.8 19.6 Loan amount 18.0 17.4 Cofinancing None None Cancellation 0.6 Date physical components completed 6.30.1993 12.31.1994 Economic rate of return 44% 7% Institutional performance Cumulative Estimated and Actual Disbursements FY88 FY89 FY90 FY91 FY92 FY93 FY94 FY95 Appraisal estimate (US$ thousands) 1,000 5,000 9,000 13,000 16,000 18,000 Actual (US$ thousands) 3,350 6,480 12,820 14,810 16,260 16,510 17,340 17,414 Actual as % of estimate 335 130 142 114 102 92 - - Date of final disbursement: March 15, 1995 Project Dates Steps in project cycle Date planned Date actual Identification NA December 1984 Preparation NA February/March 1985 Appraisal April 1985 May 1985 Negotiations January 20, 1986 April 14,1987 Board presentation March 25, 1986 May 21, 1987 Signing May 29,1987 May 29,1987 Effectiveness August28, 1987 September 11, 1987 Project completion June 30, 1993 December 31, 1994 Loan closing December 31, 1993 December 31, 1994 Staff Inputs (staff weeks) Stage of project cycle Planned Revised Actual Weeks US$ Weeks US$ Weeks US$ Through appraisal NA NA NA NA 26.6 NA Appraisal-Board NA NA NA NA 4.4 NA Board-Effectiveness NA NA NA NA NA NA Supervision 75.0 NA 85.0 NA 79.1 NA Completion 11.0 NA 11.0 NA 6.2 NA Total 116.3 NA Annex A 21 Mission Data Stage of project Performance rating" Staff days in Specialization staff Implementation Development Types of cycle Month/year No. of fielde skills 36represented status impact problems" persons Through appraisal 1. 12/84 1 12 HE 2.2-3/85 2 10 HE, EC 3.5/85 3 15 HE (2), EC Appraisal through 6-7/86 3 16 HE (2), EC Board approval Board approval 7/87 2 14 HE (C), EC thru' effectiveness Supervision 1. 10-11/87 1 13 HE(C) 2 1 M 2. 3/88 1 8 HE(C) 2 1 M 3.6-7/88 1 10 HE (C) 2 1 M 4.11/88 2 8 HE, HE (C) 2 1 M 5.5/89 1 10 HE 2 1 M, T 6.10-11/89 1 20 HE 2 1 M, T 7.4-5/90 1 21 HE 2 1 M, T, F 8.9-10/90 1 21 HE (C) 2 1 M, F 9.7/91 1 7 HE(C) 2 1 M, T, F 10.7/92 1 19 HE (C) 2 1 M, T, F 11.12/92 1 17 HE (C) 2 1 M, T, F 12. 9-10/93 1 11 HE(C) 2 1 M, T, F 13.7/94 1 16 HE(C) S S M Completion 2. 1/95 1 14 HE (C) - - 34. I - Problem free; 2 - Moderate problems; S-Satisfactory 35. Combined with supervision/preparation missions for other highway projects. 36. HE - Highway Engineer; HE (C) - Highway Engineer, Consultant; EC - Economist 37. M - Managerial; T - Technical; F - Financial 22 Annex A Basic Data Sheet RAILWAYS PROJECT (CREDIT-1986-UG) Key Project Data (amounts in US$ million) Appraisal Actual or Estimate current estimate Total project costs 8.8 6.283 Loan amount 7.0 5.6 Cofinancing None None Cancellation 1.4 Date physical components completed 6.30.1992 12.31.1992 Economic rate of return 34% Negative Institutional performance Cumulative Estimated and Actual Disbursements FY89 FY90 FY91 FY92 FY93 FY94 Appraisal estimate (US$M) 0.60 5.60 6.60 7.00 7.00 7.00 Actual (US$M) 0.00 0.67 1.41 2.69 5.30 5.60 Actual as % of estimate 0 12 21 38 76 80 Date of final disbursement: Project Dates Original Actual Initial Exec. Proj. Summary 09/15188 09/15/88 Appraisal Mission 11/14/88 11/14/88 Negotiations 12/19/88 12/19/88 Board approval 02/21/89 02/21/89 Credit Signature 03/10/89 03/10/89 Credit Effectiveness 11/08/89 11/08/89 Credit closing 12/3192 06/30/93 Credit Completion 06/30/92 12/31/92 Staff Inputs (staff weeks) Stage of Project Cycle Planned Revised Final Through Appraisal 49.7 45.1 Appraisal through Board 9.2 4.3 Approval Supervision 70.0 75.0 58.0 Total 70.0 133.9 107.4 Annex A 23 Mission Data Stage of Proj. Cycle (Month/year) No. of Staff days in Specialization Performance Rating trend persons field represented rating 40 perons38 39 Through Appraisal 7/88 6 10 TM, RE(C) - . Preparation ME(C), CE(C) F(C), RS(C) Appraisal 11/88 8 22 TM, RE(C) ME(C), CE(C) F(C), RS(C) E(C), SE(C) Appraisal through Board - - - - - Approval No mission Board Approval through Effectiveness Donors Conference 4/89 1 4 DC, TM, RR, - - RE(C) Follow-up 7/89 4 1 TM Initial SPN 10/89 2 7 TM, RE(C) 0-2, P-1 F Supervision Mission 1 2/90 3 7 TM,RE, F(C) 0-2, P-1 C Mission 2 9/90 2 11 RE, F(C) 0-2, P-2 P,C Limited SPN 4/91 1 3 RE Mission 3 2/92 3 10 TM, RE, F(C) 0-3, p-2 P, F, C Mission 4 1/93 2 6 RE, F(C) 0-3, P-2 P, F 38. TM: Task Manager, RE: Railway Engineer, F: Financial Analyst, CE: Civil Engineer, ME: Mechanical Engineer, SE: Signal Engineer, RS: Railway Management Specialist, E: Economist, DC: Division Chief, R.R.: Resident Representative, suffix (C): Consultant 39. 0: Overall Rating in Form 590, P: Project Development Objectives Rating in Form 590 40. F: Financial Management, C: Compliance with Covenants, P: Procurement