Ghana - Road Rehabilitation and Maintenance Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 13314 PERFORMANCE AUDIT REPORT GHANA ROAD REHABILITATION AND MAINTENANCE PROJECT (CREDIT 1601-GH, AFRICAN FACILITY CREDIT A-1-GH AND JAPANESE GRANT 1601-1-G) JUNE 30, 1994 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 Currency Unit: = Cedis c US$1.00 (1985) = 53 Cedis (SAR) US$1.00 (1986) = 89 Cedis US$1.00 (1987) = 154Cedis US$1.00 (1988) = 203 Cedis US$1.00 (1989) 270 Cedis US$1.00 (1990) = 326 Cedis US$1.00 (1991) = 367 Cedis ABBREVIATIONS AfDB African Development Bank BHC Bank for Housing and Construction DANIDA Danish International Development Agency DFR Department of Feeder Roads DUR Department of Urban Roads ERP Economic Recovery Program GHA Ghana Highway Authority IDA International Development Association MOF Ministry of Finance and Economic Affairs MOTC Ministry of Transport and Communications MRH Ministry of Roads and Highways NFRR&MP National Feeder Roads Rehabilitation and Maintenance Project OPEC Organization of Petroleum Exporting Countries PAR Performance Audit Report PCR Project Completion Report PMU Project Management Unit RR&MP Road Rehabilitation and Maintenance Project SAR Staff Appraisal Report TA Technical Assistance TR-I Transport Rehabilitation Project TR-II Second Transport Rehabilitation Project WFP World Food Program UNDP United Nations Development Program USAID United States Agency for International Development FISCAL YEAR January 1 - December 31 FOR OFFICIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A. Office of Director-General Operations Evaluation June 30, 1994 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Performance Audit Report on Ghana Road Rehabilitation and Maintenance Project (Credit 1601-GH, African Facility Credit A-l-GH and Japanese Grant 1601-1-GH) Attached is the "Performance Audit Report on Ghana - "Road Rehabilitation and Maintenance Project (Credit 1601-GH, African Facility Credit A-l-GH and Japanese Grant 1601-1-GH)". The project's overriding objective was to support the government's economic recovery program (launched in the mid-eighties) by removing obstructions to road communication. Specific goals were to strengthen road agencies, revitalize road maintenance, and upgrade transport planning and coordination. Implementation was initially constraint by government difficulties of local cost financing. Otherwise, implementation went rather well. Compliance with covenants was good. The project set a process in motion which led to rehabilitation of road infrastructure, reconstitution of sector agencies and capacities and introduction of new system of road financing. In combination with the three follow-up lending operations for roads, it provided long-term support to the economic recovery program. The project showed the feasibility of radically changing pace and direction in the sector, and it established guideposts for continuing sector improvements. The positive results are attributable to several factors: a favorable macro-economic climate; a committed and cooperating government; and full-scale and high-quality Bank assistance. The audit rates overall project performance as satisfactory, institutional impact as substantial and sustainabiity (mid-term) as likely. H. Eberhard K6pp for Robert Picciotto Attachment 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.  FOR OFFICIAL USE ONLY PERFORMANCE AUDIT REPORT GHANA ROAD REHABILITATION AND MAINTENANCE PROJECT (CREDIT 1601-GH, AFRICAN FACILITY CREDIT A-1-GH AND JAPANESE GRANT 1601-1-GH) TABLE OF CONTENTS Page No. PRE FA C E . . . . . . .. . . . . . .. . . .. . .. . .. . .. . .. . . BASIC DATA SHEET .........................ii EVALUATION SUMMARY ..................... iv 1. Sector Background 1.........................1 2. Sector Strategy and Prescriptions 2................2 General Framework 2.........................2 Road Rehabilitation and Maintenance Project 3 Follow-up Lending 4.......................... 4 Transport Rehabilitation Project 4................4 Second Transport Rehabilitation Project 5 National Feeder Roads Rehabilitation and Maintenance Project 6...............................6 3. Project Implementation 6...................... 6 4. Project and Program Results 7...................7 5. Project Lessons ............................10 This report was prepared by Albert Weckerle, Senior Evaluation Officer. Ms. Helen Watkins provided administrative support. 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 wvithout World Bank authorization.  i PERFORMANCE AUDIT REPORT GHANA ROAD REHABILITATION AND MAINTENANCE PROJECT (CREDIT 1601-GH, AFRICAN FACILITY CREDIT A-1-GH AND JAPANESE GRANT 1601-1-GH) PREFACE This is a Performance Audit Report (PAR) for the Ghana Road Rehabilitation and Maintenance Project (RR&RM). The project was supported by Credit 1601-GH for US$ 40 million and African Facility Credit A-l-GH for US$ 11.9 million equivalent. Both credits were fully disbursed. Several donors other than IDA cofinanced the project. A Project Completion Report (PCR) was issued as Report No. 10747 on June 26, 1992. The PAR was prepared by the Operations Evaluation Department (OED). To prepare it, OED staff reviewed the President's Report, Staff Appraisal Report (SAR), PCR, legal documents, transcript of Board Proceedings, project correspondence files, Bank documents on other highway projects and other material. In November 1993, an OED mission travelled to Ghana where it held discussions with officials and representatives of these agencies and parties: Ghana Highway Authority (GHA), Depart- ment of Feeder Roads (DFR), Department of Urban Roads (DUR), Ministry of Roads and Highways (MRH), Ministry of Transport and Communications (MOTC), Ministry of Finance and Economic Affairs (MOF), Bank of Housing and Construction (BHC), road users, construction companies and the United States Agency for International Development (USAID). OED also reviewed the project experience with Bank staff in the field and at headquarters. The PCR, which was of good quality, dealt primarily with implementation issues. To deepen the PCR analysis and extend it in new directions, the PAR aimed at: elaboration of the context, in which the Bank launched the RR∓ analysis of the Bank's contribution to reconstituting transport, after its almost complete collapse; and identification of factors bearing on project outcomes. RR&MP was the first in a series of four integrated road projects to support Ghana's economic recovery program (ERP) and rebuilding of the transport sector. The three post-RR&MP projects are still under imple- mentation. The PAR, therefore, also dealt to some extent with the RR&MP follow-up lending. Following standard procedures, OED invited borrower comments on the draft PAR. None were received.  ii PERFORMANCE AUDIT REPORT GHANA ROAD REHABILITATION AND MAINTENANCE PROJECT (CREDIT 1601-GH, AFRICAN FACILITY CREDIT A-1-GH AND JAPANESE GRANT 1601-1-GH) BASIC DATA KEY PROJECT DATA Original Plan Actual Total Project Costs (US$ million) 121.8 132.5 Overrun (%) 8.8 Credit Amounts Disbursed: 1601-GH 40.0 40.0 A-1-GH 11.9 11.9 Estimated Economic Rate of Return (%) 26-73 30-67 Physical Components Completed 12/89 12/90 Physical Components Completed by Target Date (%) 100 NA OTHER PROJECT DATA Original Plan Actual Identification 06/80 Appraisal 09/84 09/84 Negotiations 02/85 05/85 Approval 05/85 06/85 Agreement 06/85 Effectiveness 11/85 Closing 06/90 06/91 Borrower Republic of Ghana Executing Agencies Ghana Highway Authority (GHA) Department of Feeder Roads (DFR) Bank of Housing and Construction (BHC) Follow-up Projects Transport Rehabilitation Project, C1858-GH Second Transport Rehabilitation Project, C2192-GH National Feeder Roads Rehabilitation and Maintenance Project, C2319-GH Fiscal Year/ Borrower January 1- December 31 iii STAFF INPUTS (staff weeks) Stage of Project Cycle Planned Revised Final Through Appraisal 82.3 Appraisal through 91.3 Board approval Supervision 123.4 Other 14.0 Total 311.0 MISSION DATA Days Performance Month/ No of in Specialization Rating Dates of Stage of pro]ect Cycle Year Persons Field Represented Status Report Identification 6/80 2 4 HE, EC - 7/30/80 Preparation I 10/80 3 4 2 HE, EC - 12/15/80 Preparation II 3/81 4 3 2 HE, EC, IS - 5/21/81, 5/28/81 Preparation III 6/81 1 2 HE - 7/27/81 Preparation IV 7/81 1 3 EC - 9/ 9/81 Preparation V 12/81 1 3 HE 1/29/82 Preparation VI 1/83 2 2 HE, EC - 2/18/83 Preparation VII 4/83 1 4 HE - 5/ 3/83 Preparation VIII 8/83 2 3 HE, EC - 9/27/83 Preparation IX 11/83 2 2 HE, EC - 11/23/83, 2/ 3/84 Preappraisal 2/84 4 6 HE, EC, FA, ME - 5/14/84 Operational 7/84 1 2 DDC - 8/20/84 Appraisal (a) 9/84 6 20 HE, EC, FA, ME, - 12/26/84' TS, SO Appraisal (b) 10/84 3 5 DDC, 6E, SC 10/29/84 Supervision 1 6/85 1 na HE - 7/ 2/85 Supervison 2 10/85 3 7 2 HE, ME 1,2,1,2 11/27/85 Supervision 3 1/86 1 11 500 2 2, 2. 2 419/86 Supervision 4 2/86 1 3 FA - 4/28/86 Supervision 5 5/86 4 9 HE, EC, ME, C 2,2,2.2 6/20/86 Superision 6 11/86 1 7 ME - 12/12/86 Supervision 7 1/87 1 5 HE, cis 3,2,2.2 3/10/87, 3/20/87 Supervision 8 3/a7 1 3 PA - 4/10/87 Supervision 9 3/88 3 10 DDC, HE, EC, TS 3,2,2,2 4/30/87 Supervision 10 1/88 1 6 HE 3,2,2,2 3/15/88 Supersion 11 10/88 2 5 HE, EC 2,2,2.2 10/ 1/88 Super-ion 12 10/88 1 4 C - 1/ 3/89 Super,ision 13 2/89 1 4 TIS - 3/ 6/89 Super-ision 14 3/89 3 7 2 HE, EC 2,2,1,2 5/26/89 Supervision 15 10/89 3 na HE, EC, ME 1,1'1.1 12/28/89 Supervision 16 4/90 3 14 HE, EC, TRS 2,1.1,2,2 7/10/90 Supervision 17 10/90 3 7 E, EC, TRS 2,1,2,2,2 1/ 2/91 Supervision 18 1/90 2 3 HE, TRS 2,1,2,2,2 3/25/91 Supervision 19 4/91 1 3 EC 2,1,2,2,2 6/22/91 Supervision 20 6/91 3 7 HE, PC, TRS 1.1,1,2,2 7/17/91 C = Consultant CIS Construction Industry Specialist EC = Economist FA Financial Analyst HE = Highway Engineer IS Industrial Specialist ME = Mechanical Engineer PA Procurement Advisor SOO = Senior Operations Officer TRS Transport Specialist TS = Training Specialist Combined Issues Paper and Decision Memorandum. iv PERFORMANCE AUDIT REPORT GHANA ROAD REHABILITATION AND MAINTENANCE PROJECT (CREDIT 1601-GH, AFRICAN FACILITY CREDIT A-1-GH AND JAPANESE GRANT 1601-1-GH) EVALUATION SUMMARY Background 1. The situation in which the Ghana had also caught up with Ghanian transport sector found itself in the early contractors, to whom GHA had gradually eighties was grim. Years of macro- shifted a fraction of routine maintenance economic mismanagement had left it in and a large part of periodic maintenance. such poor shape that it had become a The volume of works was way down, liability to the national economy. In road government payments for works were transport, by far the most important mode erratic, and most equipment was immobi- in the country, all system parts were lized for want of repair. The road trans- defective. Road infrastructure, consisting port industry languished because of poor of trunk, feeder and municipal roads, had road infrastructure and general economic deteriorated so much that it had lost its conditions. capacity to carry even the reduced levels of traffic of the early eighties. Some roads 3. Funds for sector investment and had become unable to carry any traffic at operation was extremely scarce and hardly all. Of the 14,100 km of trunk roads in sufficient to meet government employee 1985, about 6,500 km (46%) had suffered payrolls. Resource mobilization in the sec- moderate or severe failure. A sharp tor yielded little. Vehicle operators paid increase in the failure rate was imminent fuel taxes, but both taxes and total receipts under prevailing economic and sector were small. The road maintenance conditions. In feeder roads, the situation budget could finance no more than a small was even worse as the majority was not fraction (20% ) of needs. Funding for road motorable even during the dry season. rehabilitation and new construction was even more precarious. [paras. 1-6] 2. The Ghana Highway Authority (GHA), the agency in charge of trunk Objectives and Design roads, had almost ceased to function. Wages for government employees were 4. Macro-economic and sector condi- below costs of living, causing high tions called for a donor-led, comprehen- absenteeism, low productivity and sive, intensive and prolonged sector over- widespread demoralization. The situation haul. The task entailed rehabilitation of in the Department of Feeder Roads physical infrastructure, reconstitution of (DFR), the Ministry of Roads and sector agencies, support to private contrac- Highways (MRH) and the Ministry of tors, improvements in infrastructure Transport and Communications (MOTC) financing, and improvement in conditions was comparable. Economic conditions under which the road transport industry could recover. Sector transformation was V to take about fifteen years. The Bank Implementation took quickly a leading role in donor assis- tance. In 1985, it approved the Road 7. As soon as implementation started, Rehabilitation and Maintenance Project it became clear that government could not (RR&MP&MP), the subject of this PAR. fully meet its financial obligations. The Between 1987 and 1991, it approved an- Bank came quickly to the rescue with other three projects, with all four forming additional funding (African Facility Credit an integrated package. All post-RR&MP A-1-GH). Even so, the government experi- projects are still under implementation. enced financial difficulties, which contributed to a one year project delay. 5. RR&MP's overriding objective was Otherwise, implementation went rather to support the government's economic well, with the volume of physical works recovery program (ERP) by removing ob- coming close to the appraisal targets. The structions to road communication. Specific project demonstrated the feasibility of goals were to strengthen road agencies, labor-intensive feeder roads construction. revitalize road maintenance, and upgrade The prescribed steps to enhance institution transport planning and coordination. The and capacity building were taken. The components were: a program of trunk road Road Fund started to function as planned. rehabilitation and maintenance; a pilot Compliance with covenants was good. program of feeder roads rehabilitation and [paras. 34-38] maintenance; and a program of institution and capacity building, aimed at public sec- Results tor agencies and private sector contractors. 8. RR&MP set a process in motion 6. Because of weak borrower institu- which led to rehabilitation of road infra- tions, much implementation responsibility structure, reconstitution of sector institu- rested with a huge contingent of technical tions and capacities and introduction of a assistance (TA), totaling 700 staff-months. new system of road financing. In combina- The figure implied an average of 15 ex- tion with the three follow-up operations, it perts in the country throughout implemen- provided long-term support to the ERP. tation. Total project costs were about US$ RR&MP showed the feasibility of radical- 122 million. The government's obligation ly changing pace and direction in the was about US$52 million (43%), which sector, and it established guideposts for was a high figure for a cash and resource continuing sector improvements. strapped country. Five donors subscribed to the remainder, including the Bank 9. RR&MP's maintenance component (IDA). Creation of a Road Fund, com- was the first rapid defense against bined with a steep increase in road user progressive road deterioration. It bought taxes (on fuel), was a condition of time to organize large-scale rehabilitation effectiveness. [paras. 7-17] under the follow-up projects. Elimination Vi of the rehabilitation backlog is in full user taxes, was an important step toward progress, but it will take much longer than securing adequate road maintenance fund- had been expected in the RR&MP ing. However, the funding situation appraisal. RR&MP's rehabilitation remains very tight. With the current component allowed resumption of good inflow of IDA and other external funding, communication between the country's two the situation is manageable. The long-term most important cities. All expost economic outlook is clouded, unless the economy rates of return were attractive. has high rates of growth and domestic resource mobilization can be increased. 10. In institution and capacity building, RR&MP initiated the most important 13. The prospects for RR&MP changes. With the help of TA, the project sustainability are good for about the next made the road agencies, in particular five years. The sector has much momen- GHA, functional again. It launched the tum, not the least due the Bank's post- long-term task of training and developing RR&MP road operations. As they are still agency staff, which handles a large under implementation, they can continue volume of work. As for now, the agencies stabilizing, transforming and developing have recruited a core of competent and the sector. The presence in the sector of motivated staff. However, unattractive the Bank and other donors, which all have conditions of government employment their own reputations at stake, would af- remain a major institutional problem . ford some insurance against near-term Another institutional issue is dependency calamities. Prospects for long-term on external agency, in particular World sustainability are not necessarily worse, Bank, ideas, expertise and drive. Few but they are uncertain in view of the sector management or policy improvem- future state of the national economy and ents of significance are initiated without the nature and volume of future donor external participation or blessing. assistance to the sector. 11. RR&MP made a successful start in 14. RR&MP's positive results are at- reviving a domestic construction industry tributable to several factors: a favorable which can undertake a wide range of road macro-economic climate; a committed and works, and it prepared the ground for an cooperating government; and full-scale and even more intensive industry support by high-quality Bank assistance spanning four the follow-up projects. This permitted the projects so far. The ERP produced an government to launch ambitious civil environment conducive to sector change, works programs and to contain, if not it brought Ghanian emigres back to the restrict, the size of force account units. country, and it improved fiscal conditions RR&MP's experience with labor-based in the sector. ERP also attracted donors feeder roads construction, by contract, was that, confident that efforts would not be so good that DFR could start similar wasted, were willing to assist the road sec- schemes country-wide. tor. 12. RR&MP's creation of the Road 15. With RR&MP, the Bank provided Fund, combined with the increase in road timely, relevant and intense sector assis- Vii tance. It responded expeditiously to a Solutions should deal with the ma- grave situation, and it synchronized and jor sector issues which bear on integrated the project with support to the achievement of the overriding ob- ERP. Design of RR&MP was thoughtful jective - creation of an efficient and comprehensive, and while scope for and effective road transport variation in project detail existed, little system. scope existed for variation in thrust. During the entire project cycle, the Bank * A good estimate of time needed for followed good practices of project putting a deteriorated sector back management, consistent with its own on its feet is fifteen years or rules and regulations. Bank staff time more. Rebuilding of institutions, allocated to RR&MP was at least double capacities and physical infrastruc- of that of regular project budgets. ture, and creation of adequate systems of financing are all long- 16. OED rates RR&MP as follows: term undertakings. Overall outcome - satisfactory; institutional development - substantial; and * Rebuilding distressed road sectors sustainability (medium term) - likely. The requires donor support exceeding PCR-based ratings were identical. [paras. normal assistance levels. However, 39-50] an unusually strong donor engage- ment carries the risk of making borrowers overly donor dependent. Lessons Once sector rebuilding is well un- der way, donors should pursue a 17. The experience with RR&MP, and well defined and stated policy of with the follow-up lending, suggests that gradual disengagement. dysfunctional road sectors can be revived, provided conditions are right and efforts 18. The rebuilding of the Ghana road are substantial and sustained. Lessons are: sector should be revisited, once the three ongoing projects are implemented. All * Road sector changes should be four ERP projects formed a lending pack- based on a clear set of long-term age, whose impact cannot be fully evaluat- sector objectives and targets, pre- ed separately. When revisited, process and sented upfront at initiation of the results should appear with greater clarity, first changes. This sets an agenda and a better assessment of sustainability which provides orientation for all should be possible. At that time, it should sector parties. also be clear whether there will be a need to continue with high volumes of donor assistance. [paras. 51-53] PERFORMANCE AUDIT REPORT GHANA ROAD REHABILITATION AND MAINTENANCE PROJECT (CREDIT 1601-GH, AFRICAN FACILITY CREDIT A-1-GH AND JAPANESE GRANT 1601-1-GH) 1. SECTOR BACKGROUND 1. 1. The situation in which the Ghana transport sector found itself in the early 1980s, was clear and grim. Years of macro-economic mismanagement and dislocations had left it in such poor shape that it had become a liability to the national economy, barely performing minor functions and facing the prospect of deteriorating even more. While most other economic sectors did not do better, poor transport weighed heavily on an economy that critically depends on the efficient flow of agricultural and other produce. Deficient transport constrained what was left of the national market economy and severely curtailed opportunities for exports. All transport modes were in a crisis. In road transport, by far the most important mode in the country, all system parts were defective whether belonging to the public or private sectors. 1.2. Road infrastructure, consisting of trunk, feeder and municipal roads, had deteriorated so much that it had lost its capacity to carry even the reduced levels of traffic of the early 1980s. Some roads, while still on the books, had become unable to carry any traffic at all, or only at prohibitively high costs. Of the 14,100 km of trunk roads in 1985, about 6,500 km (46%) had suffered moderate or severe failure. A sharp increase in the failure rate was imminent under prevailing economic and sector conditions. Analysts placed the length of failed roads by 1989/1990 at 8,600 km (61%) or more if underlying trends would not be reversed. In feeder roads, the situation was even worse as the majority was not motorable even during the dry season. 1.3. The Ghana Highway Authority (GHA), the agency in charge of trunk roads, had almost ceased to function. Due to the exodus from government service or even the country, some 65 % of its engineering positions were vacant. Unskilled labor was in surplus. Wages for government employees were below costs of living, causing high absenteeism, low productivity and widespread demoralization. The situation in the Department of Feeder Roads (DFR), which is responsible for feeder roads, was comparable, if not worse as its institutional capacity was never at a par with that of GHA. The Ministry of Roads and Highways (MRH) and the Ministry of Transport and Communications (MOTC) had their own share of institutional malaise. MRH is the umbrella organization for roads. MOTC regulates road transport and has transport planning and coordination functions. 1.4. Economic conditions had also caught up with Ghanian contractors, to whom GHA had gradually shifted a fraction of routine maintenance (5%) and the larger part of periodic maintenance (75%). Foreign contractors still handled the bulk of major civil works. Domestic contracting capacities had shrunk precipitously. The volume of works available was way down. Government payments for completed works were erratic. Equipment was immobilized as old pieces were worn out, and contractors could purchase neither spare parts nor new pieces, for lack of capital and foreign exchange. The banking sector could provide little relief. Support to the Bank for Housing and Construction (BHC) through the Third Highway Project (Credit 2 1029-GH, approved in 1980 and closed in 1986) was beneficial on the margin, but not much more. Compared to real needs, BHC had little to lend, and it used outmoded loan procedures. 1.5. Capital for investment and operation was extremely scarce in any sector, and budget revenues were hardly sufficient to meet government employee payrolls. Resource mobilization in the road sector yielded little. Vehicle operators paid fuel taxes, but both taxes and total receipts were small. Whatever the receipts, they disappeared into the government coffers in an environment where all sectors suffered severe underfunding. The road maintenance budget in the early 1980s could finance no more than a small fraction (20%) of needs. Funding for road rehabilitation, to make up for existing or anticipated road failures, was even more precarious. Room to finance road improvement and new construction was virtually nonexistent. 1.6. The road transport industry, composed of many private-sector and a few public-sector operators, languished because of poor road infrastructure and general economic conditions. Business was down, vehicle operating costs were up and few imports of vehicles and spare parts were available. The fleet of privately owned and roadworthy vehicles was in decline. 2. SECTOR STRATEGY AND PRESCRIPTIONS General Framework 2.1 Macro-economic and sector conditions called for a donor-led, comprehensive, intensive and prolonged sector overhaul. Following a frank discussion on the country's economic recovery program, the Bank took a leading role. In consultation with government, other donors and international consultants, it designed sector reform, documented in a 1985 sector strategy paper and in a series of project appraisal reports. It arranged for a substantial flow of IDA resources into the sector, helped secure lending from other donors, and applied its institutional experience in rescuing transport sectors. Compared to other countries, Ghana received an unusually intensive Bank treatment. 2.2 The strategy to revive, rehabilitate and reorient the highway sector was based on improvements on all fronts. It aimed at: a rehabilitating physical infrastructure, a reconstituting sector agencies, a supporting private contractors, o establishing an improved system of infrastructure financing, and o creating conditions under which the road transport industry could recover. Transformation of the sector was to take some fifteen years. The accent was initially on the restoration of physical infrastructure, almost in the style of an emergency operation. Subse- quently other concerns, especially institution building, received more attention. 2.3 The Bank built its support to the sector around a series of lending operations, four so far since 1985: (i) RR&MP (the subject of this PAR), approved 1985; (ii) Transport Rehabilitation Project (TR-I), Credit 1858-GH, approved late 1987; 3 (iii) Second Transport Rehabilitation Project (TR-II), Credit 2192-GH, approved late 1990; and (iv) National Feeder Roads Rehabilitation and Maintenance Project (NFRR&MP), Credit 2319-GH, approved late 1991. In addition, support came through macro-economic initiatives. TR-I and TR-II had large railway components, but the main focus was on roads. All post-RR&MP projects are still being imple- mented. Expected closing of TR-I is mid-1994. The closing date of TR-II is late 1996, and that of NFRR&MP, mid-1997. Road Rehabilitation and Maintenance Project 2.4 RR&MP's overriding objective was to support the government's economic recovery program (ERP) by removing road transport obstructions to external trade and domestic agricultural traffic. It placed the main accent on trunk roads, but also targeted feeder roads. Specific project goals were: to strengthen road agencies, revitalize and improve road main- tenance operations, and upgrade and rationalize transport planning and coordination. Important elements in these goals were staff development through improved staff incentives and training, and increased use of the private sector, in particular for road construction and maintenance. 2.5 Project components were: o a program of trunk road (including bridge) rehabilitation and maintenance; o a pilot program of feeder roads (including bridge) rehabilitation and maintenance; and o a program of institution and capacity building, aimed at all important road sector entities, including public sector agencies (GHA, DFR, MRH), BHC and private sector contractors. 2.6 The trunk road program arranged for stepped-up road maintenance and rehabilitation, reversing a trend of diminishing efforts for the upkeep of roads. In maintenance, the target was coverage of 30% of actual needs, which was a modest, though significant beginning. Maintenance expenditures were to rise by 60% above pre-project spending levels. In reha- bilitation, the major task was reconstruction of the all-important road (105 km) between the country's two major cities (Accra and Kumasi). 2.7 The feeder roads component sought to prepare the ground for a rational, feasible and coherent subsector development program. While it also aimed at maintenance of a significant length of roads, its key element was testing of labor-intensive road rehabilitation by contract. Answers sought were the scope for low cost construction and rural employment generation. 2.8 The institutional development program entailed strengthening of all road agency functions. Efforts to reconstitute a good human resource base, through staff training and food aid, were critical means to this end. Food aid, given to all employees, was essential in view of government inability to grant more than minimum wages. The training component targeted many agency employees from nearly all professions. BHC received institutional support for its role as the leading bank in financing construction equipment. The construction industry was part of capacity building for its critical role in executing road works. Widening a similar policy predating the project, the government envisaged all future construction and maintenance to be done by contract, except road emergency works. TA, equipment financing through BHC and award of actual contracts were the vehicles to develop the domestic construction industry. 4 2.9 Because of weak borrower institutions, much implementation responsibility rested with a huge contingent of technical assistance (TA), totaling 700 staff-months. The figure implied that an average of 15 experts would be in the country throughout implementation. A higher borrower profile in project management and reduced roles for TA might have been desirable for the sake of institutional growth and project ownership. But project planners felt such an approach was not possible as the risk of an unwanted change in project content or scope would have been too high. TA was a very visible element in all project components. 2.10 Total project costs were US$121.8 million, with the trunk road rehabilitation and mainte- nance program accounting for the largest share of project expenditures. TA was about US$5.3 million. The government's obligation was US$52.2 million (42.9%), which was a high figure for a cash and resource strapped country facing demands from all sectors. Five donors subscribed to the remainder, US$69.6 million (57.1%), in a demonstration of international backing of Ghana's economic recovery program: (a) World Bank (IDA), US$40.0 million (32.8%); (b) African Development Bank (AfDB), US$18.1 million (14.9%); (c) World Food Program (WFP), which financed food aid, US$8.1 million (6.7%); (d) Japanese Government, US$2.0 million (1.6%); and (e) United Nations Development Program (UNDP), US$1.4 million (1.1 %). 2.11 Two conditions of credit effectiveness signaled where the project had to make instant progress, in order to contain the risk of implementation delays and unwanted changes in project scope or content: creation of a Road Fund, combined with a steep increase in road user taxes (on fuel); and engagement of consultants for the Project Management Unit (PMU). The Road Fund had to overcome stiff opposition in government (Finance) and the World Bank. Opponents argued that earmarked funding is fiscally undesirable. The declared purpose of the Fund was financing of road maintenance and rehabilitation, and the new fuel tax and other road user taxes would provide its income. The PMU needed to be in place before implementation started, and its staff needed to be of international caliber. Project sponsors took no chances to jeopardize implementation through weak project management. PMU's functions were project coordination and management of individual components. GHA, DFR, MRH and BHC were the implementing agencies. Planned implementation period was four years. Follow-up Lending Transport Rehabilitation Project 2.12 TR-I's formal objectives were to help sustain and consolidate the momentum of Ghana's economic recovery program, and to complement and reinforce RR&MP. Specific tasks were: to continue progress initiated under RR&MP on road maintenance and rehabilitation; to continue RR&PM's institution and capacity building, aimed again at GHA, DFR, MRH and MOTC; and to raise productivity in the road transport industry and develop new rural transport technologies. 2.13 The project included a full-scale feeder road rehabilitation and maintenance compo- nent, corresponding to DFR's entire work program for two years. For GHA, the project fi- nanced a two-year time slice of its program as well. Without external aid, the SAR argued, the government could finance no more than prevailing levels of maintenance, which would entail a 20-year period for clearing maintenance backlogs and stabilizing road conditions. With the new project, the expected period was five years only. 5 2.14 Institution and capacity building in GHA and DFR focused on road maintenance. In MRH, focus was on project management. In MOTC, it was on sector planning, policy formu- lation and policy implementation. Support to DFR included a staff training program, an extension of the RR&MP assistance to GHA. 2.15 The pilot transport program arranged credit for modernization of private auto repair shops, and for development of low cost non-motorized vehicles. Several hundred repair shops were the expected beneficiaries. Among the new technologies, which the program sought to test, were bicycle-trailers, a form of transport used widely in Asian countries. 2.16 Implementation of the roads components hinged again on a large contingent of TA, though the project's total number of staff-months planned, 376, was well below the RR&MP figure of 700. Planners expected local consultants to meet almost one-half of the total TA needs. Part of TA was reserved for extending PMU services to TR-I, for project management and coordination. 2.17 Estimated total project costs were US$213.8 million. The financing plan envisaged that domestic sources (government budget and Road Fund) would furnish US$61.9 million (27.8%). The shares of external sources were: (a) IDA US$60 million (27%); and (b) other co-financiers (not identified in the SAR), US$100.5 million (45%). Project implementation had a schedule of over six years. However, several components, in particular GHA and DFR's two-year road maintenance programs, had a much shorter period. Second Transport Rehabilitation Project 2.18 TR-II's general objective was again to sustain, and even accelerate, Ghana's economic recovery program. Among the specific objectives were: removal of transport constraints to exports, agricultural production and labor mobility; private sector development; and alleviation of rural poverty, combined with self-development of rural women. Main tasks were continuation of road maintenance and rehabilitation, institution and capacity building (road agencies and private contractors), and enhancement of rural transport. 2.19 The road maintenance and rehabilitation program sought to bring Ghana closer to complete elimination of a backlog of work. But implementation experience under TR-I had shown that the task was larger than expected before, and that it would need extension even beyond TR-II. The TR-II program included a major effort on trunk roads, and a smaller effort, making use of labor-intensive methods, on feeder roads in one Region. 2.20 The institution and capacity building component dealt again with GHA and MRH, to improve their task and project management functions, and with MOTC, to aid it in policy analysis and formulation. Staff training for all three agencies was an integral part of the component. Support to private contractors consisted of training and technical advice for running their business. 2.21 The rural transport component involved, besides improvement of rural infrastructure, promotion of non-motorized transport vehicles. Complementary project items were studies, TA and funding for other forms of rural development support. 6 2.22 Implementation still relied on TA, in particular for PMU. But staff-months, about 170, were way down from the high levels of the two preceding projects. PMU received an allotment of about 120 s-m, and road training units, one of about 50 s-m. TA for PMU was, in essence, a way to deal with low government salaries, which made it difficult to fill government staff positions. On similar grounds, the road agencies also contracted increasingly with local engi- neering, management and business consultants. 2.23 Estimated total project costs were US$230.4 million. The financing plan, specified these contributions, besides the government's share (budget and Road Fund) of US$68.4 million (29.7%): (a) IDA, US$96.0 million (41.7%); and (b) other donors, which were not yet firmly identified in the SAR, US$66 million (28.6%). The scheduled project implementation time was five years. National Feeder Roads Rehabilitation and Maintenance Project 2.24 The NFRR&MP was, again, to support Ghana's economic recovery program. In addition, it served as lending support to the Medium Term Agricultural Development Strategy (MTADS), a joint government-Bank initiative to stimulate agriculture and rural development. Main components were a feeder road maintenance and rehabilitation program, and an institution and capacity building program aimed at DFR and the local construction industry. An additional component was a scheme to enhance rural mobility. 2.25 The maintenance and rehabilitation component covered a combined total of some 5,000 kin, which was a large part of the 21,000 km feeder roads network. The institution and capacity building component comprised development of all DFR functions, and various forms of support to the local contracting industry. It included training for DFR and contractors. The rural mobility component entailed schemes to promote use of non-motorized vehicles. 2.26 The TA became again a major factor in project implementation, specifically in institution and capacity building. Including TA for studies, the total TA figure was over 700 staff-months, about the same as that under RR&MP. About 320 s-m were set aside for DFR management and line positions. 2.27 Estimated implementation period was about five years. Estimated total costs were US$106.0 million, of which the government had a share of US$8.4 million (8.0%). Other contributors were: (a) IDA, US$55.0 million (51.9%); (b) United States Agency for International Development (USAID), US$21.0 million (19.8%); (c) Danish International Development Agency (DANIDA), US$15.0 million (14.1 %); (d) Organization of Petroleum Exporting Countries (OPEC), US$5.0 million (4.7%); and (e) Japanese Grant Fund, US$1.6 million (1.5%). 3. PROJECT IMPLEMENTATION 3.1 Shored up by the vast number of TA experts, which appeared in important positions such as PMU, implementation of RR&PM almost went on as planned. The initial going was slow, but subsequently, the pace accelerated. In the end, the project had a time overrun of one year, due largely to government difficulties to finance its share of costs. That the government's share might be 7 too high became obvious soon enough. The Bank came quickly to the rescue with African Facility Credit A-1-GH for about US$11.9 million equivalent. The new credit raised the IDA share of appraised costs to almost 43%, and lowered that of the government to about 33%. 3.2 The total volume of physical works came close to the appraisal targets. Achievements under rehabilitation and periodic maintenance (trunk roads) corresponded to plans. In routine maintenance (trunk roads), there was a shortfall of some 20%. In the pilot rehabilitation scheme (feeder roads), there was a shortfall of 40%. Otherwise, the pilot scheme showed the feasibility of labor-intensive construction, with costs just below capital-intensive methods, good workmanship and attractive scope for employment generation. Total actual project costs were 9% above appraisal estimates, which was the result of two factors--higher than expected unit costs for civil works contracts, and--lower than expected scope of physical works. Local currency depreciation and an unanticipated pace of road deterioration caused the unit cost increases. 3.3 Implementation managers in Ghana took the prescribed steps to enhance institution and capacity building. They recruited and deployed consultants. They organized staff training programs. They distributed food aid, and they furnished support to BHC and domestic private contractors. The project, through BHC equipment financing, management training, the award of small-scale contracts and other measures, put an initial contingent of contractors on its feet. The volume of contractor work corresponded to appraisal plans. 3.4 The Road Fund started to function as planned, except that revenues generated did not reach expected levels because of low volumes of traffic. Contractors received regular Fund payments, which helped alleviate the late payment problem that had plagued the industry before. The successful working of the Fund demonstrated the feasibility and utility of such a financing instrument. 3.5 Compliance with covenants in the Development Credit and Project Agreements (with GHA and BHC) was good. At project completion, the borrower had fully met almost all covenants. The exception was a small set of covenants of minor importance, which it had only partially met. 4. PROJECT AND PROGRAM RESULTS 4.1 RR&MP set a process in motion which led to rehabilitation of road infrastructure, reconstitution of sector institutions and capacities and introduction of a new system of road financing. In combination with the three follow-up operations, it provided long-term support to Ghana's economic recovery program. RR&MP showed the feasibility of radically changing pace and direction in the sector, and it established guideposts for continuing sector improvements. 4.2 RR&MP's maintenance component, consisting of routine and periodic maintenance operations, was the first rapid defense against progressive road deterioration. It bought time to organize rehabilitation under the follow-up projects, which involved arranging financing, surveying road conditions and planning for works execution. With TR-I, the task of road rehabilitation got into high gear, though elimination of the rehabilitation back-log has been and is being stretched out well beyond the target date of late-1990s that RR&MP had hoped for. On current estimates, the road rehabilitation may extend until about 2005, which would imply a time span of 20 years, measured from RR&MP. Reasons for extension of the schedule are higher than originally expected unit costs of construction, due to underestimates of works needed, and financial constraints. Reconstruction of the Accra-Kumasi link under RR&MP allowed resumption of road communication at reasonable costs between the country's most important cities. Bridge rehabilitation was particularly beneficial, as deteriorated structures did or could do much harm to traffic. According to PMU calculations, trunk road maintenance yielded an 8 economic rate of return (ERR) of 67% ex-post, bridge rehabilitation yielded 88%, and reconstruction of Accra-Kumasi, 30%. The numbers highlight the desolate condition of roads before improvement. 4.3 It was in the area of institution and capacity building where RR&MP initiated the most important changes. With the help of TA, the project made the road agencies, in particular GHA, func- tional again. It launched the long-term task of training and developing agency staff. As for now, the agencies have recruited a core of competent and motivated staff, which handles the large volume of donor-supported infrastructure works. Unattractive conditions of government employment remain a major problem, however. The way around this issue, procurement of TA and consulting services (such as for PMU in GHA), works only up to a point. Limits to reliance on consultants are: key agency management tasks, such as policy initiation and formulation, ought to be handled in the agencies them- selves, because they are better placed to represent agency and government interests; and competent management of contracted services requires competent agency managers. Contracting out agency func- tions also is not easy on agency budgets. Unit costs of contracted staff are invariably much higher than those of agency staff. 4.4 Another institutional issue is agency dependency on donor, in particular World Bank, ideas, expertise and drive. Few sector management or policy improvements of significance are initiated without donor participation or blessing. While this may have been unavoidable in the early years of Ghana's economic recovery, it should be a matter of concern now. The longer this kind of government- donor relationship continues, the more will dependency become ingrained in the sector. The concept of project and sector ownership, which plays a large role in the current design of Bank assistance, implies a fair measure of borrower-distancing from the Bank, at least after a succession of lending operations. One quantitative indicator of borrower dependency is Bank staff time made available to projects. In the last 8-10 years, as the Bank mounted four road sector operations, it spent the larger part of 950 staff-weeks (s-w) on Ghanaian road transport (with a smaller part on rail transport). That is 95 s-w or more per year on average. The yearly statistics have shown a tendency to drop, but they were still significant in the last two years. They were 71 s-w in FY 92, and 58 s-w in FY 93. Throughout this period, other donors also provided technical support of a nature comparable to that of the Bank, though on a smaller scale. In addition, Bank projects financed the large volume of TA services. 4.5 RR&MP made a successful start in reviving a domestic construction industry which can undertake a wide range of road works, and it prepared the ground for an even more intensive industry support by the follow-up projects. This permitted the government to launch ambitious civil works pro- grams and to contain, if not restrict, the size of force account units. RR&MP's experience with labor- based construction, by contract, was so good that DFR could start similar schemes country-wide on a much larger scale. Apart from contributing to rural infrastructure improvement, labor-based con- struction provided rural labor with income earning opportunities. Many women took up road construction employment. Success in labor-based construction was due to development of good works organization and procedures and training provided to small-scale contractors. Another factor was low rural wages which made work by labor competitive with equipment-based construction. 4.6 RR&MP's creation of the Road Fund, combined with the increase in road user taxes, was an important step toward securing adequate road maintenance funding. A small portion of the tax receipts is earmarked for the Fund. So far, the Fund provided financing for periodic maintenance. The other categories of public sector expenditures for roads (in particular for routine maintenance, rehabilitation, improvement and new construction) had to rely on other sources: either general government budgets, or donor financing, or both. The funding situation is very tight. Options are limited, though increases in road taxes and earmarking for the Fund are being contemplated. With the current inflow of IDA and other donors' funding, the situation is manageable. The long-term outlook is clouded, unless the economy has high rates of growth and domestic resource mobilization can be increased accordingly. 9 4.7 The prospects for sustainability of RR&MP are good for about the next five years. The sector has much momentum, not the least due to the Bank's post-RR&MP road operations and the quality of macro-economic management. As all three post-RR&MP projects are still under implementation, they can continue stabilizing, transforming and developing the sector. The presence in the sector of the Bank and other donors, which all have their own reputations at stake, would afford some insurance against near-term calamities. Prospects for long-term sustainability are not necessarily worse, but they are uncertain in view of: the future state of the national economy; the future quality of sector management by government; and the nature and volume of future donor assistance to the sector. 4.8 RR&MP's positive results can be attributed to several factors: a favorable macro-economic climate; a committed and cooperating government; and full-scale and high-quality Bank assistance. All factors were interlinked. Another factor, already discussed, was reinforcement of RR&MP by three projects coming in rapid succession, with all four projects forming an integrated lending package. 4.9 The economic recovery program produced an environment conducive to change, from which road transport benefitted. As it began to take hold, it caused many Ghanians to return to their country, which they had left for better employment and business opportunities. Some returnees became available again for the transport sector, either for work in private companies (road construction or consulting) or road sector agencies. Progress on the ERP produced economic growth and improved fiscal conditions, which allowed domestic financing of increasing shares of road maintenance. ERP also attracted donors that, confident that efforts would not be wasted, were willing to assist the road sector. 4.10 The government carried its share of project preparation and implementation to the best of its abilities. When implementation problems arose, it had the resolve to confront them. Relations with the Bank were professional and maintained in a spirit of mutual trust. The government wanted the project to succeed. It understood its lending obligations, and it set out to honor them. 4.11 With RR&MP, the Bank provided timely and relevant sector assistance. It responded expeditiously to a grave situation, and it synchronized and integrated the project with support to the ERP. The Bank was instrumental in defining a long-term sector agenda and in rallying donor support to the sector. It established that sector changes were needed on several fronts and that there were no quick and easy fixes. It also impressed on the borrower that maintaining infrastructure is central to good sector management. Design of RR&MP was thoughtful and comprehensive, and while scope for variation in project detail existed, little scope existed for variation in thrust. In all phases of the project cycle, the Bank followed good practices of project management, fully consistent with its own rules and regulations as stipulated in its operational instructions (Central Project Memoranda, Operational Manual Statements, and others). The Bank proved its sector commitment with staff budgets and costs that were unusually high, but justified by the circumstances of 1985. On project preparation, it spent 174 s-w, and on supervision and miscellaneous, another 137 s-w for a total of 311 s-w. Most Bank road projects in other countries, even when much larger and facing equally complicated situations, are usually budgeted for half these amounts or less. Other donor agencies also incurred costs for project administration, thus reducing the burden for the Bank. But it was the Bank which shouldered the largest share. 4.12 The audit rates RR&MP as follows: Overall outcome satisfactory institutional impact substantial sustainability (medium term) likely The PCR-based ratings were identical. 10 5. PROJECT LESSONS 5.1 The experience with RR&MP, and with the follow-up lending, suggests that dysfunctional road sectors can be revived, provided conditions are right and efforts are substantial and sustained. The message is important not only for Ghana, but other countries as well whose road sectors have disintegrated and ceased to perform. 5.2 The Ghana experience offers these lessons: 0 Road sector changes should be based on a clear set of long-term sector objectives and targets, presented upfront at initiation of the first changes. This sets an agenda which provides orientation for all sector parties. o Solutions should deal with all major sector issues which bear on achievement of the overriding objective--creation of an efficient and effective road transport system. o A good estimate of time needed for putting a deteriorated sector back on its feet is fifteen years or more. Rebuilding of institutions, capacities and physical infrastructure, and creation of adequate systems of financing are all long-term undertakings. o Rebuilding distressed road sectors requires donor support exceeding normal assistance levels. Demand on donors can be substantial, both in terms of lending volume and donor staff expertise and time. However, a stronger than usual donor engagement carries the risk of making borrower agencies overly donor dependent. Once sector rebuilding is well under way, donors should pursue a clearly defined and stated policy of gradual disengagement. 5.3 The rebuilding of the Ghana road sector should be revisited, once the three ongoing projects are implemented. All four ERP projects formed a lending package, whose individual components cannot be fully evaluated separately. When revisited, impact should appear with greater clarity, and a better assessment of sustainability should be possible. At that time, it should also be clear whether there will be a need to continue with high volumes of donor assistance.