Sri Lanka - Road Passenger Transport Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 8011 PROJECT COMPLETION REPORT SRI LANKA ROAD PASSENGER TRANSPORT PROJECT (CREDIT 994-CE) AUGUST 7, 1989 Infrastructure Operations Division Asia Region Country Department I 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.  ABBREVIATIONS CTB - C-ylon Transport Board GOSL - Goverrment of Sri Lanka IDA - International Development Association MOT - Ministry of Transport PCR - Project Completion Report PR - President's Report RTBs - Regional Transport Boards SAR - Staff Appraisal Report SLCTB - Sri Lanka Central Transport Board  TM WORLD BANK FOR OFICIAL UU ONy WashMton. D.C. 20433 U.S.A. Opvauim d vaetiun August 7, 1989 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Completion Report on Sri Tanka Road Passenger Transport Project (Credit 994-CE) Attached, for information, is a copy of a report entitled "Project Completion Report on Sri Lanka - Road Passenger Transport Project (Credit 994-CE)" prepared by the Asia Regional Office. No audit of this project has been made by the Operations Evaluation Department at this time. Attacnment This document has a restricted distribution and may be used by ecipients only in the performance of their offlcia duties. Its contents may not otherwise be discksed without Word Dank anthorimzs4L  * OR OMCIL USE ONLY SRI LANKA PROJECT COMPLETION REPORT ROAD PASSENGER TRANSPORT PROJECT (CREDIT 994-CE) TABLE OF CONTENTS Pas. No. PREF ACE ..... .... ............. BAIC DATA SHEET ....... ... .. ......too.. it EVALUATION SteiiARY ................... .... ...to. v I. INTRODUCTION ...... ........ . 1 II. PROJECT IDENTIFICATION. PREPARATION AND APP_AISAL ............ 2 Origin ............... ....*.*.*...**..*... I Preparation and Appraisal . .. ............ . ...... 3 Project Objectives .......... .. ........ . .*.e 4 Project Description .........4...... 0 ..* 4 II:. PROJECT IMPLEMENTATION .......................... .. .... 5 Procurement of Buses .......... ...... 5 Improvement of Depot Maintenance and Operations ............ 7 Project Costs .................... ,+.............. ........... a Finatcial Sources .......................................... 9 Performance of Consultants ....... . ........................... 10 IV. OPERATING PERFORMANCE ........... .. ........................... . 11 The Bus Fleet .......................................................... 11 Bus Operations ...................... .,... 12 V. FINANCIAL PERFORMANCE ....................................... 13 Fares ................. ...... 13 Financial Results of the RTBs . ....................... 14 Financial Results of the SLCTB ................. .. 15 j VI. INSTITUTIONAL PERFORMANCE AND DEVELOPMENT .................... 16 Institutional Arrangements .......... ....... 16 Institutional Achievements ......................... 16 Institutional Shortfalls ............................................ ... 17 VII. ECONOMIC REEVALUATION .................. is.**********..... 18 Consmers' Surplus Approach ........ . . . . . . .......... ............... 19 Comparative Cost Approach ................................... 21 This document has a restricted distribution and may be used by ecipeats oy in the performance of their official duties. Its contents may not otherwise be disclosed witbout World anik authorization.  VIII. THE ROLE OF IDA ....................e.* ........... ...... 22 IX. CONCLUSIONS.. . . . . .. ** .................... 23 Project Justification and Objectives .......... .............. 23 Project Context and Scheduling .............................. 23 Project Implementation and Operating Outcomes ........... .... 24 Sustainability of Project Benefits ......................... .. 24 ANNEXES I Cumulative Disbursements ActnaL and Estimated .... .......... 23 2 Summary of Major Covenants and Status of Compliance ........ 26 TABLES table 1 Actual and Original Estimates of Project Costs .. .......... 28 Table 2 Bus Fleet Major Charact-ristics .................... ........ 29 Table 3 Selected Operating Indicators - RTBs ....................... 30 Table 4 Estimated operating Results - Combined RTBs .............. .. 31 Table 5 Revenues and Expenses - SLCTB (Estimates) ........... ...... . 32 Table 6 Estimates of Price Elasticities of Demand ................ ... 33  PROJECT COHPLETION REPORT SRI LANKA ROAD PASSENGER TRANSPORT PROJECT (CREDIT 994-CE) PREFACE This report presents a review of the genesis and implementation of thWi Road Passenger Transport Project financed under IDA Credit No. 994-CE in the amount of USS53.0 million equivalent which was made to the Democ-itic Socialist Republic of Sri Lanka in March 1980. The credit became effective in October 1980 and was closed in November 1985. twenty nine months later than expected at appraisal. At the closing of the credit, a total of US$51.6 million had been disbursed and the balance of US$1.4 million was cancelled. The Sri Lanka Central Transport Board (SLCTB) was the implementing agency on behalf of the government and nine Regional Transport Boards fRTBs), which were the principal beneficiaries of the credit. The report is based on a Project Completion Report (PCR) prepared by SLCTB in April 1987, information obtained from a re iew of the Staff Appraisal Report (SAR), supervision missions, and other reports in the project file. In accordance with the revised procedures for project performance reporting, this PCR was read by the Operations Evaluation Department (OED) but the project was not audited by them. The draft PCR was sent to the Borrower on June 9, 1989 for comments. No comments were, however, received.  PROJECT COMPLETION REPORT SRI LANKAs ROAD PASSENGER TRANSPORT PRUJECT (CREDIT 994-CE) Basic Data Sheet KEY PROJECT DATA Appraisal Actual or Current Item Expectation Estimate Total Project Cost (USS million) 86.4 69.4 lt Underrun (1) _ 19.7 Credit Amount (USS million) 53.0 53.0 Disbursed 53.0 51.6 Cancelled - 1.4 Date Physical Components Completed 6183 486 Proportion Completed by Above Date (Z) 21 45 97.4 Proportior of Time Underrun/Overrun (Z) 100 18Z.9 Economic Rate Return (X) 20 Negative 3/ Financial Performance Break-even by 1983 Deficit l/ Cost in US dollar Poquivalent much less than appraisal estimate due to a considerable depreciation of the Rupee during implementation. 2/ Proportion of disbursement to amount of credit. 3/ Negative economic rate of return based on the same methodology used during appraisal. However, using a "consumers' surplus" and "comparative costl approach, the yield would range between 20-23%.  OTHER PROJECT DATA Orixinal Revised Actual Flrst Mention in file in Timetable 1J Government Application - 12177 11 End of Negotiations - 2/29/1o Board Approval - 3/1S161 Credit Agreement - 41166e0 Effectiveness 7114/80 10/27/80 101271/0 Closing Date 6/30183 12131184 111301/3 Borrower The Democratic Socialist Republic of Srl Lanka Iipl e*xting Agency Sri Lanka Central Transport Board Fiscal Year of Sorrower January 1-to December 31 Follow-on Project None L/ Originally, a project was conceived as a combined road maintenance and road transport project. (Per Addendum dated 6/8/79 to July 1976 Project Brief). During appraisal of that project in October 1976, however, it was found that the transport component had not been prepared to a stage where it could be approved and it was therefore decided to divide the combined project into two and process each separately. The Road Maintenance Project was approved in May 1979 and the Road Passenger Transport Project was appruved in March 1960.  -Liv- HISSSION DATA Honth/ Number of Number of Total Date of Year Weeks Persons Hanteeks Reoort Identificotion 12177 NMA N/A N/A N/A Preparatios 12178 N/A N/A N/A N/A Appraisal 8179 N/A NIA N/A 02/22180 Supervision 1 11180 2.0 1 2.0 12124/80 Supervision 2 2181 2.8 4 11.2 03106181 Supervision 3 7/81 3.0 3 9.0 07124;81 Supervision 4 :2181 0.4 1 0.4 12104181 Supervision 5 7/82 2.8 1 2.8 07/28182 Supervision 6 5/83 2.0 3 6.0 05123183 Supervision 7 10/83 3.4 3 10.2 12109/83 Supervision 8 3184 1.0 2 2.0 03/08/84 Supervision 9 3/85 2.2 1 2.2 03/20/85 Supervision 10 7/85 2.2 1 2.2 07/02/85 CompLetion 2/18 1.6 1 1.6 03106/66 22.6 49.6 Hean mv/year 9 COUNTRY EXCHANGE RATES Name of Currency (abbreviation) Re Appraisal Year Average Exchange Rate USS1 - 15.6 Intervening Years Average US$1 - 24.0 Completion Year Average USS1 = 28.0 I~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ U  -V- SRI LANX ROAD PASSENGER TRANSIJRT PROJECT (CREDIT 994-CE) EVALUATION SUM*MRY PROJECT BACKGROUND 1. Tho political and socio-economic setting against which the Road Passenger Transport Project was identified and prepared is discussed in detail in the February 27. 1980 President's Report (PR) paras 2, 10. 21, 32, 36 and in the February 22. 1980 SAl. paras 4.01-4.02. The first credit for a Highway Project in Sri Lanka was approved by the Board in 1966 and was subsequently cancelJed as the government wanted to change the project approach and scope. Between 1960 and 1976 annual per capital gross national income rose by 0.9Z. The slowdown in the 1970-1977 pertod was attributed in the PR to L... a combination of factors, including inadequate investw,nt. poor management of the economy and a policy enviroiment unconducive to growth and investment.... Policy changes introduced in 1977 following the clection of the United National Party were aimed at resuscitating the economy and to increase employment. Policy changes were encouraged in 1978 by an International Monetary Fund standby arrangement for SDR 93 million. 2. The genesis of the Road Passenger Transport Project can be traced to a 1976 update of a 1966 Transport Sector Report. There had been no Bank-assisted operations for 11 years. An IDA mission visited Sri Lanka in December 1977 and recomuended assistance to road passenger transport and for road maintenance and rehabllitation. initially, the idea was to have one project cater for both. Event-iwlly, two separate projects were financeds one for road maintenanr ..nder the Department of Hlghways, and another for passenger transport < r the Ministry of Transport (MOT). 3. In the late 19709. bus passenger transport services were provided by nine RTBs, which operated bus services over the entire island. supported and coordinated by the SLCTB. RTNs and the SLCTB together were the Isrgest employer in Sri Lanka with a staff of about 61,000 in mid-1979. The PR noted (para 36) that both the SLCTB and the RTBs were e.erstaffed and that labor productivity was low. Demand for public passenger transport wae high. In 1978, the Government of Sri Lanka (GOSL) allowed private operators to operate buses on selected routes. The Road Passenger Transport Project (appraiseel in August 1979 and approved by the Board in March 1980) was designed to provide assistance to SLCTB and RTBs with a view to improving bus service in the country. PROJECT OBJECTIVES 4. The objectives of the project were to (SAR, para 4.03). (a) restrain the growth of oil imports; (b) provide a more reliable bus service which would offer the public a more acceptable alternative to the use of private carns and  -vi- (C) decentralize the control of bus operations to financially viable RThs. 5. To achieve these objectives, the project would seek to: (a) improve the fuel efficiency of the bus fleet by the purchase of new vehicles, and by improving maintenance standards; (b) obtain a higher utilization of existing buses and a longer life; and (c) end subsidies for the busn operations and to pave the way for a change in the system of financing capital development by lOOS grants, so that from 1983 the RTBs would provide part of the new capital needs from internal cash generation and borrowings. PROJECT COHPONENTS 6. The components of the project were (SAR, para 4.04): (a) the restoration of the capacity of the SLCTB to operate an efficient unit overhaul and parts exchange service. and a complete bus overhaul service; (b) the purchase of chassis and materials, and assembly of 2,120 new buses for use by the RTBss (c) the improvement of bus main..enance facilities and practices in the RTBs, and improvement in bus operations: (d) low cost traffic engineering measures designed to improve bus flows; and (e) technical assistance for the execution of the project and institution building. 7. At appraisal, the project was expected to be completed by December 31. 1982. The performance of the SLCTB and RTBs was considered satisfactory in view of the constraints under which they had been operating (PR. para 63). Physical c.u%onents were similar to previous works executed by SLCTB and were not expecced to present a.ny particular problems (PR, para 51). The accounting system of SLCTB and the RTBs was found to be outmoded and accounts were seriously in arrears. This concealed a growing operating loss until the position was revealed by a consultant report in August 1979. Consultants were engaged under the project to assist with the work of updating the accounting records and system (PR, para 52). Honitoring would be conducted through quarterly progresJ reports by the SLCTB (PR. para 55). IMPLEMENTATION EXPERIENCE AND RESULTS 8. Experience with both objectives and with the instruments for their pursuit has not been encouraging.  -vii- (a) Objectivest {i) reduced growth of oil imports: not achieved because of the changing conditions during the life of the project: (ii) provision of more reliable bus services: achieved to some degree. but not as foreseen in the SAR. The more reliable service was actually provided by private operators (PCR, para 36); and (iii) decentralization of bus operations to financially viable RTBss responsibility 38s delegated to the ITBs, but they did not become financially viable (PCR, paras 42-44). (b) Instruments: (i) improved fuel efficiency of the bus fleet: achieved by default. The RTBs were forced to reduce services, scrap overage buses, and concentrate on new buses with better designs--including fuel efficiency. However. fuel costs escalated between 1980 to 1985 (PCR, para 45). (il) higher utilization of existing buses and a longer life: achieved by default. Competition from private operators caused an eventual retrenchment of routes and services. The new busect purchased under the project had a life expectancy of 12 years, and the maintenance program was intended to increase the reliability of the buses (PCR, para 53); and (iii) end subsidies for bus operations: not fully achieved (PCR, para 41). COMPONENTS 9. Experience with. procurement and implementation has not been encouraging. (a) Bus overh2ul service: not discussed in the PCR; (b) procurement of buses: A total of 1.618 buses were actually procured as against 2,120 proposed at appraisal (PCR. paras 16-21); (c) improvement of depot maintenance and operations: A total of 15 depots were completed as against 22 proposed at appraisal (PCR, paras 22-23); (d) low cost traffic engineering measures: .'ropped (PCR, para 14d)5 and (e) technical assistances A total of 220 staff-months was financed under the project; benefits are modest (PCR, para 32).  -viii- COSTS 10. At project completion (PCR, para 24), reported actual expenditures amounted to US$69.4 million (as against the appraisal estimate of US$86.4 million). 'Reported actual expenditure" figures should be treated wtith caution because of imperfections in the accounting systems and procedures of SLCTB and the RTBs (PCR, paras 26, 37, 41, 43, 48). PROJECT COMPLETION 11. The SAR (paras 4.19-4.20) states that all Executing Agencies (SLTCB, MOT, the Colombo Municipal Council, and the Department of Highways) were deemed capable of executing the project and that the bulk of the project's components were similar to previous works executed by SLCTB and should present no particular problems. The PR (para 50) expected the project to be completed by December 31, 1982. In fact, the project was completed in June 1986 (PCR, Basic Data Sheet) because of the problems related to procurement. SLCTB management, and labor issues (PCR, para 15). OPERATING AND FINANCIAL PERFORMANCE 12. By the end of 1985, bus services had been curtailed below their levels at the beginning of the project, even though the size of the fleet had remained the same (PCR, para 36). 13. Bus operation statistics are inconsistent and somewhat unreliable as operating records were not consistently maintained (PCR. para 37). 14. Financial evaluation of RTBs was made "... more complex and unreliable by the deplorable state of the accounting of the entire operations of both the SLCTB and the RT1sm (PCR, para 41). The '...management of the centralized workshops and warehouse were weak..." (PCR, para 47). The "...actual financial situation of the SLCTB...remains largely unknown. Since the inception of the project, until official closing, it was never possible to establish a true financial picture of the SLCTB for any of the past years.... Three financial audits were carried out..., but the Auditor General could not approve the accounts" (PCR, para 48). ECONOMIC REEVALUATION 15. Economic rate of return estimated at appraisal by using fare receipts as a proxy for benefits resulting from the provision of road passenger transport services under the project: 20t (SAR, paras 6.01-6.04). Economic rate of return reestimated after project completion by using the same methodology: negative (PCR, para 58). The economic reevaluation used in the PCR is unrelated to the appraisal methodology. tt is based on (a) consumer surplus, and (b) comparative cost (PCR, para 59). The two approached yield economic rates of return in the 20-23Z range (PCR, para 76).  SRI LANKA PROJECT COMPLETION REPORT ROAD PASSENGER TRANSPORT PROJECT (CREDIT 994-CE) I. INTRODUCTION 1. The transport system and facilities of Sri Lanka comprise about 25.300 km of roads under the Department of Highways and about 41,000 km under the local authorities; about 1,450 km of railway tracks most of which consist of single track on broad gauge; one main and two minor parts: and seven airports of which one handles international traffic. Inland transport services are provided by the Ceylon Government Railways, public bus transport boards consisting of the SLCTB and nine TB3s, and an increasing number of private road passenger and goods operators. Road transport is by far the most important activity--accounting for over two thirds of the seotor compared with the railways which plays a relatively insignificant role accounting for about 4Z. 2. For over two decades prior to 1979 the public sector maintained a virtual monopoly over regular bus service in Sri Lanka. In June 1978 the Government of Sri Lanka (GOSL) reorganized the public bus service by decentralizing the former Ceylon Transport Board (CTB) to form the SLCTB and the nine RTBs. In April 1979, GOSL made a major policy move by lifting restrictions imposed by the previous government in the operation of regular bus services by private operators. with a view to relieving the chronic overcrowding of buses operated by the RTBs, and eliminating the long linss of passengers at bus stops caused by inadequate and unreliable bus schedules. The lifting of restrictions to operate private bus service was also intended to encourage competition in the passenger transport industry. 3. The bus services provided by the RTBs were inadequate and poorly managed in the face of growing demand for passenger service. The problems of the public bus services were compounded by the fact that more than a quarter of the fleet was overaged and a majority of the buses needed major overhauls. Thus in the 1979-83 investment program prepared by GOSL, a need was foreseen for providing adequate funding for maintenance and rehabilitation of the public passenger service in the light of its predominant role in the transport sector. A road passenger transport project proposal, the main thrust of which was the renewal and augmentation of RTB fleet, was the basis for the project for which IDA provided funding under this credit. 4. In early 1959 all privately-owned bus services in Sri Lanka were nationalized and the CTB was created to operate road passenger transport as a monopoly. In June 1978 the CTB was dissolved and replaced by nine public bus operators--the RTBs, and one central service organization--the SLCT8. Among the responsibilities of the SLCTB were the provision of goods and services to the RTBs and coordination among them including personnel deployment, review of bus fares, general planning and procurement.  -2- 5. Since project exscution coo enced in 1980. major institutional changes supported by IDA were implemented with the goal of providing greater operational autonomy of the RTas, and reduced control by the SLCTS. Functions such as operations of the central and regional workshops, procurement of routine materials and spares were subsequently devolved to the MTs. The centralization of these functions at the SLCTS (which proved to be a major operational handicap for the RTBs) was not only a strain on the already weak management of the SLCTB, but also. important repair work and parts supply had to be referred to Colombo from the nine regions for SLCTB approvai before action could be taken. This caused serious delays ln bus repairs, low bus availability, and high workshop downtime. 6. The financing of the project was based on the following plant (a) IDA to provide a credit to GOSL in the amount of US$53.0 million on standard terms; (b) GOSL to provide US$34.3 million equivalent in local currency to cover local costs of the project including taxess (c) IDA funds were to be on-lent to SLCTB and to each of the nine RTBs under a subsidiary loan agreement with each agency in respect of goods and services received under the project; (d) GOSL contribution to the project was to be treated as equity towards the capital of each beneficiary agency; and (e) the funds from the credit earmarked for the traffic engineering component were to pass to the MOT as a grant. 7. A factor which strongly influenced the project execution was a prolonged civil disturbance beginning in 1983. This disturbance caused increasing damages to property in the northern and northeastern regions of the country. This unfortunate disruption had an adverse effect on passenger transport in those areas and led to severe weakening of the operational and financial performance of the RTBs generally. Estimates to assess the extent of damae to the public passenger transport indicate that property damage in excess of USS15.0 million was sustained by the Sri Lanka Transport Boards including the destruction of about 900 buses and several depots. In those areas where the impact of the civil unrest was less severe, cowpetition from the private bus operators had diverted traffic from the RTBs prompting a forced, but necessary reduction in the RTB regular bus services. II. PROJECT IDENTIFICATION. PREPARATION AND APPRAISAL Origin 8. Based on the updated Transport Sector Report of 1966. an IDA mission visited Sri Lanka in December 1977. Two separate projects were identifieds one for road maintenance under the Department of Highways, and  -3- another to address the need for assistance to the Sri Lanka road passenger transport system. The mission also recommended to GOSL that adequate funding be provided in the immediate program for both road maintenance and rehabilitation of the public bus service. As further work was required for preparation of the road passenger project, IDA approved a Project Preparation Facility (PPF) in December 1978 to assist the SLCTB prepare a suitable project. Preparation and Appraisal 9. The project was prepared as a product of cooperation between the SLCTB and consultants hired under the PPF; it was subsequently appraised by IDA in August 1979 followed by a Board approval in March 1980. The project was aimed at expanding the scale of bus operation of the transport boards by about 302 by 1983 consistent with the demand for passenger service at that time. To achieve this target, a program of investment consisting of substantial additions of new buses to the fleet and major overhauls of existing vehicles was defined. In addition, provisions were made for improving the vehicle maintenance facilities by reequipping the workshops and supplying critical spare parts which had been out of stock for a considerable time. Furthermore, technical assistance was to be provided in the areas where management capabilities were inadequate or did not exist in both the SLCTB and the RTBs. 10. During project execution. events which had not been foreseen at the time of apprais%1 affected the project design in terms of its objectives and implementation logistics. The liberalization of trade which GOSL announced in 1979, coupled with the attractive fiscal incentives offered to the private bus operators, encouraged a major influx of private passenger vehicles into the country. This affected the RTBs and claimed a rapidly growing share of rhe passenger market. This trend of market share loss of the RTBs was accertuated by the freedom with which private bus operators were left to choose their routes. Thus it became imperative for the RTBs and the SLCTB to reassess the project objectives, particularly with respect to the scale of operations of the RTBs. The loss of market share had meant that RTBs buses were increasingly forced to run with fewer passengers because private operators' services were perceived by the public to be better and more reliable than those of the RTBs. 11. As a result of the strong competition, the foicus of the project objectives was directed away from expansion of operations and placed instead on fleet consolidation and efficiency improvements. The initial actions involved freezing services at their 1980 levels, and subsequently scrapping and removal from the fleet of the overaged buses. Procurement of new buses with better design features also become an objective to reduce driver fatigue and resultant absenteeism caused by driving very old buses requiring a great deal of manual labor to maneuver the steering wheel and to shift gears. Also. operating and financial data collection and processing systems were introduced to provide management with a mechanism for monitoring operations which would assist in decision making. In retrospect, the outcome of these adjustments in passenger service resulted in elimination of overcrowding and staff retrenchment as a result of the reduced role of SLCTB and individual RTBs. Moreover as there was no longer justification for expanding the fleet size .n the face of an increasing number of private operators, focus was turned to replacements with buses of better design..  -4- Project Objectives 12. The main project objectives were the support of government policies to: (a) restrain the growth of oil imports; (b) provide a more reliable bus service which would offer the public an acceptable alternative to the use of private cars; and (c) decentralize the control of bus operations to financially viable Regional Boards. To achieve these objectives. the project sought to: (a) improve the fuel efficiency of the bus fleet by the purchase of new vehicles and by improving maintenance standards; (b) obtain a higher utilization of existing buses and a longer life: and (c) end subsidies for bus operations and to pave the way for a change in the system of financing capital development by 1002 grant, so that from 1983, the RTBs would provide part of the new capital needs from internal cash generation and borrowing. 13. rXnder the prevailing social-economic conditions at the time of appraisal, the project objectives were appropriate and well conceived. Despite the changes which had to be made in operational strategy occasioned by the private operators, project objectives generally remained unchanged, although approaches to the objectives were altered. For example. the reduction of the level of bus service in most cases was aimed at reducing operating losses of the RTBs. Project Description 14. The project was designed to include the following major components: (a) Reepuipping SLCTB Workshops Restoration of unit exchange and overhaul facilities of the SLCTB. This component was designed to improve the output and quality of the central and provincial workshops operated and owned by the SLCTB for the benefit of the RTB buses. It was envisaged that at project completion the workshops would be adequately equipped to recondition units, (e.g., engines and gear boxes) and sub-units (e.g'.. wiper and starter motors). In addition, it was envisaged that the workshops would be capable of carrying out complete bus overhauls including bodywork needed for the entire fleet of the RTBs. (b) Purchase of New Buses This component provided for the purchase of bus chassis and body materials for local assembly at the Ekalla factory of SLCTB located in the suburbs of Colombo. In addition, it provided for purchase of a small number of complete buses with enhanced features to be tried out in service. (c) Improvement in Bus Maintenance and Operations This component involved construction and equipping of 22 new depots and purchase of equipment for new as well as existing depots and improving bus operations.  (d) Low-cost TraffLc Enuineerins Measures- This component (whith was later -Iropped due to aaniuistrative difficulties imposed by the involvement of several agencies), was aimed at smoll measures such &s provision of bus lanes, relocation of bus stops, minor improvements on pavement and curbs, and regulations concerning back-loading and unloading. (e) Technical Assistance This component was aimed at providing consultancy services and training in the areas of production engineering, procurement. accounting, traffic operations and management reorganization. III. PROJECT TMPLEHSNTATION 15. Project execution (which was exposed to a series of problems throughout) took 61 months to complete from the day of credit effectiveness compared with the appraisal estimate of 36 months. Principal constraints included protracted delays in making decisions on procurement of the buses; a high turnover in the management of the SLCTB which occasioned deferrals of important implementation decisions; and labor related issues at the central workshop and the Ekalla bus body assembly plant. Procurement of Buses 16. This component of the project involved procurement of about 2,100 bus chassis units and related body materials for assembly in Sri Lanka. The first order for single-deck bus chassis under the project consisted of 660 units for assembly at the SLCTB Ekalla assembly plant. At the time of the award of tender, substantial time had been lost by the SLCTB in meeting conditions of credit effectiveness and in completing the bidding and evaluation processes. As a result, the first delivery of bus chassis was taken more than one year after the date set at appraisal. In the second order for 400 chassis units, the bid documents provided that the winning bidder could be given an option, at the discretion of the SLCTB, to include an additional 200 units under the same specifications as those of the main bid for the 400 units. The objective of this option was to induce SLCTS.to purchase a small number of completely built buses equipped with 'novel' features which had not been tried previously. The features that were considered as experimental included: (a) rear-mounted engine; (b) automatic transmission; (c) a stronger bus body to withstand overloading; and (d) a combination of these features. The intention was to run such buses on test basis, monitoring the performance of each major feature against those of the completed buses assembled to existing specifications. These novel features would be considered for inclusion in the specifications for the next order should the experiments prove positive. 17. While the procurement of the 400 chassis Units progressed well. SLCTB delayed in purchasing ten completely built buses with novel features. This delayed the delivery of the further 200 chassis units which was contingent upon the purchase by'SLCTB (using prudent shopping as a 1 I'e for X , l I ) of i , buses w d l feature'.  l8. Meanwhile, the bus chassis which were being delivered under the first and second orders began accumulating at the Ekalla factory as the assembly was progressing at less than the expected rate. Appraisal estimates assumed an assembly bus production of 75-100 per month. Actual monthly output fluctuated widely, but generally fell between 30 and 50 buses, depending on a variety of adverse factors at the factory and in the SLCTB. At the Ekalla factory (which was constantly plagued by staff absenteeism and a lack of incentive to produce) management was almost non- existent and there was little effort to follow a systematic production schedule. SLCTB later purchased the ten complete buses with novel features from Japan and the United Kingdom. Following this decision by SLCTB. the suppliers who won the contract for the 400 chassis units were advised of their right to deliver the 200 additional chassis under the same specifications and contract. 19. Pressured by the accumulation of chassis arriving at Ekalla and faced with the risk that storage space could run out. SLCTB realized that the slow production at the Ekalla factory was delaying project implemeintation. SLCTB then requested IDA to approve the procurement of completely built buses in the next order to speed up the number of buses to the fleet--replacing a growing number of overaged buses. By the closing date IDA had approved the contract for the supply of 200 built buses: not surprisingly there was a balance in the credit in excess of US$Z9 million of uncommitted funds. GOSL requested IDA for an extension of the closing date to permit completion of various components of the project which were at different stages of implementation. 20. In May 1983, IDA sent a mission to review the status of the project in detail and decide whether or not an extension of the closing date was warranted. Although all the relevant information required to make a comprehensive assessment of the status of implementation was not available, the mission was nevertheless able to draw some conclusions about the project and the prospects of project objectives. The mission findings were summarized as follows: (a) at the June 30, 1983 original closing date, only about 50X of the project would have been completed. The much needed fleet replacement was drastically slowed down by the inability of the Ekalla assembly plant to achieve reasonable production targets; (b) the bus operations of the RTBs in general had deteriorated due to diminishing availability of their buses--compounded by a fierce competition from an increasing number of private bus operators. This raised the question of whether it was prudent to continue supplying more buses to the RTBs without an effort to reduce operations to a manageable level; (c) the construction of the new depots-and workshops and improvement of existing ones had not been completed because GOSL had declined to provide funding for local costs. These facilities were considered vital for the improvement of the bus maintenance and service facilities of the RTBs; and  (d) the GOSL had made a definite commitment to curb the-management turnover which had plagued the SLCTS. The new management which took office in February 1983 (the fourth since project implementation commenced'd were more responsive than the previous ones *nd appeared to understand the project objectives; additionally they were aware of the changing circumstances in the passenger transport service. 21. With this background, IDA approved extensions to the closing date in stages with a view to permitting more time to assess the competence and enthusiasm of the new managemnt, and to encourage GOSL and SLCTB to carry out a program of reforms consistent with the objectives of the project. The first extension of six months was to permit project expenditures for which IDA had given itz approval prior to the closing date. Subsequent extensions were granted upon willingness of SLCTB to accepts (a) an agreed program of scrapping about 1,400 buses which were crowding the service and maintenarre facilities of the RTIs; (b) procurement of more buses which would have specified novel features; (c) decentralization of provincial workshops and handing them over to the RTs; and (d) reorganization of the materials management and purchasing functions of the SLCTB- These actions were closely monitored by IDA, and to a large measure the objectives were achieved. Vnder the the action plan agreed between IDA and SLCTB, 150 additional fully built buses (with new specifications which had not yet been tried in the RTBe) were purchased under international competitive bidding procedures. A total of 1,618 buses were eventually procured under this project. Improvement of Depot Maintenance and Operations 22. The number of new depots proposed at appraisal was reduced from 22 to 15 in recognition of the need to scale down the operations of the RTBs--a change necessitated by the rapid expansion of the private bus I operations. This component faced a serious setback when GOSL decided that the responsibility for funding the local costs of this component should be placed with the RTBs rather than from a government budget. The execution was thus delayed for a while as the RTBs were unable to provide the needed funds. Realizing that the dispute between the SLCTBIRTBs and GOSL over the funding of this component could delay execution indefinitely, IDA advised SLCTB to make a request through GOSL to IDA to transfer funds from an expenditure category in the credit so that the civil works remai4ing under this component could be completed. With the availability of this funding, fifteen new depots were completeW fully equipped and became operational together with some existing depots. 23. The combination of construction of new depots, provision of related equipment and tools, and the scrapping of the unserviceable buses is expected in the long term to improve the standard of maintenance wLich  -8- in turn would permit a better balanced geographic distribution of the bus services offered islandawlde. Project Costs 24. At the completion of the project, repocted actual expenditures amounted to US$69.35 million equivalent in both local and foreign costs. This represents a cost underrun of 20Z against an appraisal estimate of US$86.4 million. Table 1 (page 28) shows a consolidation of the actual and estimated total cost of the project and is sum arized below under major components as follows: Component Expected Actual Z of Expected 1. Spares and workshop equipment 5.99 17.13 286 2. Buses and assembly 42.26 41.20 98 3. Depot construction and equipment 10.59 8.21 78 4. Low-cost traffic engineering 0.39 5. Consultantsltraining 2.24 2.81 125 6. Contingencies 24.91 -- Total 86.4 69.4 80 25. The cost figures are denominated in US dollar equivalents. using the exchange rates prevailing at the time of each transaction, covering the period December 1980 to July 1986. While thare was an overall project cost underrun of about 202. the underrun related to the local costs was nearly 5OZ of the original estimate. The principal factor which contributed to this cost underrun was the massive devaluation which the Sri Lankan Rupee (Rs) underwent during the project period. At the time of project appraisal, an exchange rate between the US dollar and Rs of 15.6 to a dollar was applied. But rapid depreciation of the Rs followed almost immediatety, bringing the exchange rate to nearly Rs 21 to the US dollar oy the end of 1961. In July 1986 when IDA made the final disbursement from the credit, the exchange rate stood at 1.. 27, a depreciation of over 402 during the project period. Another major contributing factor to the cost underrun relates to the reduction in the scope of the project in such areas as the number of new depots to have been constructed (reduced from 22 at appraisal to 15) and the number of buses to have been (reduced from 2.120 to 1,618). Furthermore. the entire component under low-cost traffic engineering measures in the project cost estimate was dropped from the project due to administrative and organizational difficulties chat were encountered from the beginning. 26. Because of the accounting problems created by the non-existence of any consistently applied and accepted procedures in the SLCTBS the costs associated with the bus assembly and workshop operations have probably been understated. Specifically, SLCTB did not have the capability to apportion  -9- lobor and overhead costs of its operations which were in=orporated in the ex-asuembly prices of the buses which were transferred to the RTBs se long- term loans. Although efforts wece made to install and operate sound accounting systems. no noticeable improvements were realized during the. project life, largely due to management turaover and non-availability of qualified accounting personnel in the MTs and SLCTS. 27. Looking at other specific components. it is noteworthy that there were cases of cost overruns, particularly in the case of spares and units where the original estimate was almost tripled. The underlying reason for the cost overrun in spares was that the bus operations were crippled by the lack of spare parts and engine and gearbox units (which had been underestimated at appraisal). Therefore, SLCTS was compelled to supplement its units and parts exchange arrangem-nt with a direct infusion of new spares and units. With the approval of IDA. more spare parts and engine and gearbox units were purchased to revitalize the fleet. 28. The costs under the coLjultant and training component were exceeded due to an enhanced training program which involved both top and second level management personnel of both the SLCTB and RTBs. The traLning program was expanded to include visits to selected countries to observe bus manufacturing and operations. In addition. second level management personnel were sent overseas for tral k (pars 52). Financial Sources 29. Financing of the project was accomplished through the IDA credit and an allocation from the GOSL's budget. The actual and original expectation of financing is suumasrized belows Source of Funding Original Plan Actual Rs M US$ M* 8R5 N4 US$ M S IDA Credit 827 53.0 58 1,342 51.62 74 GOSL 535 34.3 37 622 23.92 34 SLCTB 70 4.6 5 (161) (6 191) () 1,432 91.7 100 1,803 69.35 100 * An exchange rate of Rs 15.6 and Rs 26 to a US dollar respectively for original plan and actual figures were applied in the conversion. ** SLCTB was unable to contribute, and in fact were subsidized. 30. Funding for the project was expected to come from IDA, GOSL and SLTCB's surpluses. But the funding contribution from SLCTB did not materialize because of operating deficits which absorbed funds of about US$6.19 million equivalent. The IDA credit provided nearly three quarters of the total project cost of USS69.3 million equivalent against an original estimate of about 61X based on the estimated project cost of US$86.4 million at appraisal. The increase in the share of the IDA credit in the actual cost of the project is explained by:  - Q- (a) the effect of the depreciation of the ts against the US dollar, during the implemntation period. which resulted in reduced dollar equivalents of the local components of the project cost; (b) the fact that more (than originally estimated) fully built buses were purchased, reducing the cost gssociated with local assembly; (c) the cost of additional spares and units; and (d) the unforeseen use of credit funds for completing civil works at several depots (whose construction stopped temporarily due to non- availability of local funds). 31. As the completion of the depot construction was considered highly beneficial in achieving efficiency improvements in the transport boards. IDA encouraged SLCTB to utilize the uncommitted funds from the credit for this purpose. About USS1*38 million of the credit remained unutillied at the completion of the project aad was cancelled, making a fully disbursed total credit of US$51.6 million. GOSL made several attempts to persude IDA to permit the use of credit savings for more spare parts. However. IDA decided that the overall needs of the public transport could be ,.I_essed in more detail in the context of a transport study. The findings of the now completed study are defining the spewzfic needs of the public bus service and is being used as a basic for preparation of other transport projects. Performance of Consultants 32. The need for the services of consultants was foreseen at an early stage of this project. At the preparation stage, consultants were needed to prepare the project so that important preliminary work to ready the project for appraisal was completed on time. The consultants for project preparation were funded from a PPF provided by the IDA. Other project implementation responsibilities which required the use of consultants included: drawing up of bus specifications and preparation of bidding documents; bus route costing study; preparation of SLCTSB accounts; and design of an inventory management system. A total of 220 staff-months of both local and expatriate consultants was expended during the entire period of pro_*ct execution. However, most of the staff-months had already been utilizeA by the original closing date of credit of June 30. 1983. Much of the proc rsment work which followed after June 30, 1983 until the final closing date was carried out by S'iTU's omn staff who gained considerable experience from the consultants--particularly in the preparation of bus specifications and production of bid documents. 33. Other consultant assignments were generally of short-term nature. and hence, there was not sufficient time for SLCTSB counterparts to gain from the experience of the short-term consultants. The route costing study was carried out and completed in four months. After the departure of the consultants, little effort was made by SLTSB management to review the recommendations, largely due to the rather rapid turnover in the management at the time. The accounting studies were designed to assess the liquidity position of the SLCTS and MTe. on two occasions when it was felt that SLCTS  -11- has been withholding funds which GOSL had granted as operational subsidies to the RTs . Both of theie short-term specific purpose studies which were called for in the moments of crisis did little to improve the overall accounting and financial functions of the RTBuISLCTB as there were acute staff shortages in these disciplines. The overall impact of technical assistance was therefore mixed. Except in the case of procurement where som experience appears to have been gained. little or no expertise was passed on to the staff of SLCTB and RTBs through this component. IV. OPERATING PERFORMANCE 34. Operating performance of public transport was influenced by fleet characteristics, maintenance and traffic scheduling practices in force, competitive forces in the passenger scene. financial standing and management capability and disposition of the RtBs and the SLCTB. Also, about a quarter of the fleet was composed of buses with ages ranging fruA twelve to twenty years. The major characteristics of the bus fleet are shown in Table 2 Qage 30) and a brief summary is presented below: 1980 1985 Change Z Average number of buses (in the fleet) 7,322 7,335 0 Average age of the buses (yeers) 7.9 7.7 -2.5 Buses operated (average per day) 5,670 4.875 -14.0 New buses (1980-85) 1,618 Old buses scrapped and withdrawn (1980-85) 2,125 i The Bus Fleet 35. When the project was appraiaed in 1979, there was a fleet of approximately 7,300 hases. Of this total, around 6.300 buses were in actual daily service to meet the time schedule of bus routes. The remaining buses were used as back-ups or replacements when repair or maintenance became necessary. Several buses were later scrapped and withdrawn from the service in 1980 due so frequent treakdowns and mounting maintenance costs. Implementation of various components of the project was aimed at modernizing the fleet and eventually increasing the leve' of service by about 222. However, the rapid deterioration in the quality of service and the looming uncertainty about the finances of the kTBs forced a change in the approach to fleet modernization. The shifts that were made in the operating plans included: (a) a phasing out of over 2,600 overaged buses (these buses were the principal focus of blame for the excessive overcrowding at the central workshops and in most of the regional bus depots; the scrapping was intended to free up workshop space as well as substantial maintenance resources);  -1Z-2 (b) an immediate reduction in the timetable requirement (i.e., the nuubz of buses required daily to make trips on scheduled routes) tg a level no higher than that of 1980 to reduce operating deficits of the RTBs which resulted in turn from the growing loss of the passenger market share to privately operated buses; (c) a reduction in the combined workforce of the RTBa and SLCTS by about 10-132; and (d) devising a route licensing system for the private operators which would encourage a free and fair competition between the private and public bus operators. 36. Emphasis in the operations thus changed from the expansion of passenger service as originally envisaged, to consolidation of services tthrough selective reduction of the timetable requirement, and improvement in the availability and utilization of buses by concentrating bus maintenance on the vehicles which were less susceptible to frequent breakdons. By the end of 1985, services had substantially been curtailed below their levels at the beginning of the project. even though the size of the fleet had remained generally the same. Bus Operations 37. Selected operating indicators are presented in Table 3 of this report. These data are not wholly consistent nor are they adequately reliable as operating records were not consistently maintained. This was particularly the case with regard to the daily service requirement plan und buses actually made available for operation for the day. Nevertheless, it records the trend for increasing route lengths and the declining number of passengers. Major highlights from Table 3 have been extracted and are p,esented below in brief summary which compares these indicators at the beginning of the project execution in 1980, and the results at the completion of the project. 1980 1985 Total route length (km) Million 46.7 60.4 Passenger-km scheduled Million 19,291.4 19,050.0 Passenger-km operated Million 22,224.3 14,720.0 Passengers carried Million 1,846.8 988.5 Passenger per bus-km 3.6 2.5 Load factor 79.7Z 69.72 Vehicle utilization (kmlday) 249.0 219.0 Employees per bus 8.6 7.9 38. With emphasis on the curtailment and consolidation of services, as mentioned above (para 36), the scale of operations went down considerably. The private bus operators were forcing a radical change on the bus services  of the RTIs who were not fully prepared to meet the unexpected competition - in the passenger service. Thus overall passenger-km operated by the public road transport declined from 22,2 billion in 1980 to 14.7 billion in 1985. a shrinkags of 33.82 largely due to a waning market share of the RTSe to the more aggressive private bus operators. The number of passengers carried declined from about i :47 million passengers in 1980 to about 988 million in 1985--a drop of 46.5S. RTIs were left with no choice other than selective withdrawal of services. This reduced operated bus-km from 516.6 million in 1980 to about 402 million in 1985. The decision to reduce bus services was not an easy one. Having operated under the lax market conditions as a monopoly, the RTKs perceived their goal V.o expand bus service to every accessible locale on the island--and were almost oblivious to operating costs. IDA tried to convince the SLCTS and the RT8s that increasing kms of operated buses was not competitive and was adding to the mounting operating deficits. It was the latter reason which eventually caused the SLCTB and the RTBs to retrench as creditors became lees patient when funds become unavailable to pay operating costs. V. FINANCIAL PERFORMANCE Fares 39. At appraisal, the future finances of the RSTs/SLCTB were predicated on the policy that passenger services would be provided to permit these agencies to become financially viable. To achieve this objective the government would determine the scale of fare increases each year on the basis of annual financial forecasts prepared by the RTKS. At the time of the project appraisal, it was determined that RTS operations had been incurring larger financial deficits than had previously been estimated. and that a program of fare increases to reverse the trend was necessary. In pursuit of this goal. the Development Credit Agreement specified that GOSL would permit increases in the bus fares necessary to eliminate the financial deficits of the RTBs by 1983. Financial forecasts were therefore based on assumptions that permitted a financial break-even before or at the latest by 1983. 40. During project execution, a number of fare increases were approved by GOSL and put into effect by the RTBs. The first fare increase was in March 1980 when there was a belief that the deficits of the RTBs Vere reaching alarming proportions. As RTBs tried to assess their financial performance amidst crude accounting and financial records, it becam evident that more frequent and much higher fare increases were inevitable to achieve the goal of a financial break-even position by 1983. An uneven series of fare increases, one of 61? in November 1980, another of 242 in March 1981 and the last increaue of 172 in August 1983 were swiftly approved by GOSL largely to ward off an immirent possibility of having to substdize RTB deficit from the severely constrained central budget. 41. Despite all these rather generous increases in the bus fares there was no indlcation tiat the financial deficits of the RTBs were abating. The financial evaluations were made more comples and unrellable by the deplorable state of the accounting in the entire operations of both the  -14- SLCTB and the RTBs. IDA insisted on many occasions on the appointment of a financial controller, but'a premature departure by the first financial controller. followed by a sudden death of his successor. meant that little was done to improve accounting procedures. In addition. private operators were penetrating the passenger market, which left the role of the RTBs in a state of disarray. Often private operitors offered lower fores to the public than those of the RTBs. This act neutralized further arguments for fare increases and provided the government with a strong case against providing subsidies to the RTBs. It was thus demonstrated that fare increases per se vas not necessarily a cure for the financial ailments of the RTBs. Financial Results of the RTBs 42. The fare increases which were administered during 1980183 period proved that the financial problems of the RTBs had little to do with the level of fares, and resolving them therefore called for more drastic action. To illustrate this, the record provides a striking comparison. While average fares went up three times during the period, average revenue per bus per year increased by about one and three quarters. Average operating cost per bus, which was consistently higher than average revenues, increased by about 802. 43. A serious problem which effectively neutralized the revenue yield of the fare increases was the level of pilferage in the fare collection. and the rampant fraud linked with the tampering of the ticket machines by bus conductors and non-paying passengers--some of whom were in collusion with the conductors. Several measures were taken to curb the pilferage and stop the growing number of non-paying passengers. The measures included a ticketing system that functioned as a lottery with weekly pay-offs made to the winning tickets. The aim was to lure passengers into buying a ticket (to reduce non-paying passengers) not only for a ride, but also for a chance to win a lottery drawing. The other measures included spot checks on the buses on route by a team of 'flying squads' equipped with jeeps (procured under the project), and procurement of specially designed tamper- proof ticket machines. Although benefits were achieved through these measures, it was difficult to evaluate specific savings because of the inadequacy of the overall accounting systems of both the SLCTB and RTBs. 44. However, some crude figures reveal that some noticeable improvement was achieved. Average revenue per passenger more than doubled- -increasing from 7 cents in 1980 to 16 cents in 1985. Operating losses of the RTBs declined from Rs 488.8 million in 1982 to about Rs 316.6 millior. in 1985 (Table 4). This is a reduction of over 351 in the deficit in three years which, considering the competitive circumstances was not only welcome, but also encouraging. Looking at another indicator, revenue per average journey increased by more than one and half times--from 86 cents in 1980 to 222 cents in 1985, thanks largely to the fare increases of 1980 and 1983. 45. With regard to operating costs, a number of factors affected the finances of the bus operations adversely. Average cost per passenger went up almost two and half times during the period. Similarly average cost per journey increased from 102 cents in 1980 to 264 cents in 1985; the rise in  -15- costs is explained by (a) a massivi devaluation of the Re affecting the prices of imported items iuch as spares and buses the value of the Rs went frx Re 15. 6 in 1980 to Rs 26 to the US dollar in ,85s (b) the sharp rise in -he price of fuel whose consumption accounted for 40X or so of the total costs of the bus operations: the average price per gallon of fuel went from Rs 13.5 in January 1980 to about Rs 37 in July 1985. an average increase per year of about 252S and (c) the absence of correct maintentnce practices which caused many cancelled bus trips due to either bus break downs or non-availability of equipment. Based on the estimated combined operating results of the RTBs which were made by consultants for the period 1982185, the operating deficits showed some noticeable decline. This is a wide disparity from the appraisal estimates which computed a deficit of Rs 94.5 million for 1982. and operating surpluses from 1983 onwards. The cost push effect of the fuel prices and the massive depreciation of the Rs not only perpetuated a marginal cost rate which was higher than the marginal revenue of bus operation, but it also widened operating loss per operated km. Financial Results of the SLCTB 46. The business of the SLCTB, unlike that of the RTBs, faced no competition, and in at least in some respects the SLCTB enjoyed a monopolistic position in its dealings with its principal customers--the RSBs. SLCTB derived its income largely from the mark-up of 182 on the value of goods and services it procured on behalf of the RTBs. In addition. the operations of the parts and units exchange system handled by SLCTB at the provincial workshops and the central warehouse and workshop complex at Werahera enabled SLCTB to bill such exchanged units at an arbitrary 502 the price of a new one. Under the circumstances SLCTB faced no competition and was under no pressure to improve the quality of the units exchanged (which had a high early failure rate.) 47. As shown in Table 5 (page 33), net trading profit of the SLCT9 went on a rapid decline in the 1980 to 1985 period. This was a clear indication that the management of the centralized workshops and warehouse was weak so that the mark-ups (which were modest) were not enough to cover the costs. SLCTB then conceded and relinquished the management of the provincial workshops to the RTBs, and also encouraged RTBs to do their own purchases of locally available materials without having to go through its ! controlled Werahera parts and units warehouse. 48. The SLCTB generally remained in operation without the threat of ! financial deficits. Nevertheless in 1984, its worst year. the SLCTB incurred a net operating deficit of Rs 8.1 million, compared with. peak profit of Rs 141.3 million in 1980. It is important to bear in mind that these financial statements are by no measure representative of the actual financial situatio.. of the SLCTB--which still remains largely unknown. Since the inception of the project. until official closing. it was never possible to establish a true financial picture of the SLCTB for any of the past years. The problems associated with the absence of an adequate accounting system (including shortage of qualified personnel) were so monumental that the resources allocated under the project to address tnem were totally inadequate. Three financial audits were carried out during the life of the project, but because the information was largely  -16- inconsistent and incomplete, the Auditor General could not approve the accounts. IDA considered suspension of disbursements as a measure to bring this covenant into compliance; however, this action was not pursued as it was felt it would be counter-productive. VI. INSTITUTIONAL PERFORMANCE AND DEVELOPMENT Institutional Arrangements 49. The project was designed to benefit public passenger services which consisted of nine RTBs island-wide and the SLCTB. The nine RTBs and the SLCTB emerged as the outcome of a reorganization of the former CTB in 1978 which decentralized road passenger operations to the regions, and placed certain central services such as personnel management, major procurements involving international transactions. and planning and policy formulation (including acting on behalf of RTBs in fare reviews with the government) to the SLCTB. SLCTB also was left with the responsibility for operating such central facilities as the Ekalla bus body assembly plant and the central warehouse at Werahera and workshops. The project implementation was conceived so that SLCTB would play a principal intermediary role of acting on behalf of the government, at the same time helping the RTBs with the physical execution of the project. This was the plan of action conceived in the decentralization act of 1978 which iA turn formed the institutional framework under which the execution of the Road Passenger Transport Project was predicated. Institutional Achievements 50. At the inception of the project, the benefits resulting from the devolution of operational decision making through decentralization had not yet materialized. In the course of the execution of the project, a problem surfaccl which threatened to impede not onlyr project execution but also interfered with the day-to-day running of the passenger services. In its central role for personnel deployment for itself and for the nine RTBs, SLCTB was responsible for most procurement of materials and supplies, and control of the central warehouse and provincial and central workshops. This arrangement forced the RTBs to become virtually dependent on SLCTB even on day-to-day matters. One of the institutional achievements under this project was the transfer of a number of important responsibilities from the SLCTB to the RTBs. For example, SLCTB agreed to transfer provincial workshops including their ownership and management to the RTBs. Thus many of the RTBs did not have to deal with SLCTB in Colombo whenever they needed the services that the provincial workshops were able to provide. 51. SLCTB also left to each RTB the responsibility of doing its own purchases for ordinary materials and supplies which did not involve international transactions. Therefore. the main institutional achievement in this regard was that of the GOSL permitting more operational autonomy to the RTBs whose operations were too centralized. A secondary achievement was the simple recognition by SLCTB/RTBs that they had to adopt modern managerial tools of analysis to avoid making costly mistakes. Therefore,  -17- in the latter part of the project, SLCTBIRTB decided that a computer system was necessary to handle the massive database that was continuously generated in the course of daily operations. If properly utilized. and accurately recorded, the database management system covering nine RTTs and connected to the central processors at SLCTB would have the capability of providing management with vital information needed for future decision making. 52. The training program carried out under the project was tailored to improve managerial capabilities in both the RT5s and SLCTB to help it better operate a modern passenger transport system. This goal was to be achieved by concentrating on staff at the top management. i.e.. Chairman of the RtBs and the SLCTB Chairman and his General Manager, and later expanded to include second tier managers--namely head of key departments such as traffic operations, maintenance and finance. A total of 71 persons were involved in this program. most of whom visited organizations in selected European and North American countries which were engaged either in the manufacture of transport equipment or operation of passenger transport service. The training program which was contracted out and managed by London Transport International. was well organized and well received by the beneficiaries. Institutional Shortfalls 53. Generally institutional performance in helping to achieve the objectives of the project was unimpressive. First. the new organizational structure of the public bus transport system was in its early post- inception period following the decentralization act of June 1978 from which the nine RTBs and the SLCTB were created. Also, the management capability needed to match the complexity both in the magnitude of operations and the geographic coverage, created difficulties during the entire implementation of the project. To cite a specific example, for sometime the nine RTBs had operated under direct control from the SLCTB to the extent that these boards were for all practical purposes operating as though they were regional divisions of the SLCTB. Virtually all major decisions. (contrary to the spirit of the decentralization act) were made by SLCTB with little involvement from the RTBs. In the period between credit effectiveness and February 1983. the managements of the RTBs had changed at least once and that of the SLCTB had four changes in its top management. This caused a continuous situation of management in transition which then caused decision making to be a secondary concern. Management nonetheless, stabilized beginning in 1983 when renewed enthusiasm in pursuing the project objectives became apparent from the new management of the SLCTB. 54. Secondly, the SLCTB and RTBs were no exception to the traditional staffing constraints many public enterprises normally face. i.e.. the tendency to overemploy at the unskilled to semi-skilled level. and an inability to attract qualified personnel at higher levels due to unattractive employment terms. In 1980, public bus operations (SLCTB and 'Tis combined) had 60,000 employees--an average of about eight employees per bus in the fleet. In contrast, these operations had few (if any) employees who pose ied the requisite experience and qualification directly needed in the many units of the bus operations. For instance in the entire operation involving millions of Rs in daily transactions there was no  -18- single qualified accountant in any of the nine RTBs or the SLCTB itself. Much of the labor force was unionized which rendered staff redundancies a near-impossibility. 55. Thirdly, the advent of the private operators which began in 1979, brought in a competitive environment in the passenger service for which the public bus operators were not prepared. In addition. the absence of a definite policy of the government for route licensing of the private operators in the initial period of entry into the passenger service unduly favored the private operators, as they were left with complete freedom to operate on any route any time without notice. 56. For a time the private operators became too ad hoc and unreliable for the public to depend on them. The government. after receiving complaints from the traveling public as well as from the SLCTBIRTBs established a directorate in the MOT to oversee an orderly operation of the private operators principally through a route licensing arrangement. The objective of the directorate was to see that while operators would have freedom of choosing routes and of determining the level of service of bus trips. they would be required to advertise their timetables. Although monitoring the operations of the private operators was never easy, a remarkable level of public satisfaction was achieved indicating that the route licensing instituted in the HOT paved the way for a more orderly operation of the private bus operators. VII. ECONOMIC REEVALUATION 57. The economic evaluation methodology adopted in the SAR was based on using fare receipts as a proxy for the estimated economic benefits resulting from the provision of road passenger transport services under the project. Based on this methodology, the SAR shows total benefits of USS102.2 million in 199Z, the last year of the analysis. and an estimated economic rate of return of 20.42. In carrying out an economic reevaluation of the project, the same methodology used in the SAR normally would be adopted to compare what was expected during appraisal with what actually happened during project implementation. 58. In this case. however, the calculation of fare receipts as the benefit of being able to make desired trips, i.e. meetin& a travel demand, yielded a negative return because fare receipts have been well below costs and. SLCTB's operational losses have been met by government subsidies for the past several years. This macro approach would not adequately isolate the effects of adding 1.618 buses procured under the project to a fleet of about 7,300 SLCTB buses. Moreover, a number of unforeseen events occurred since the appraisal of the project in 1979, making the rbefore" and *after" the project comparison inappropriate. First, there was a great expansion of privately operated buses after 1979 which captured about 50X of the market by 1986. During this period, the SLCTB fleet capacity remained more or less unchanged. Second, ethnic disturbances in three provinces substantially reduced bus operations in those areas despite a resurgence in overall economic activity.  -19- 59. For reasons stated above, therefe-, the economic reevaluation has been carried out using two different appro.nest (a) consumer surplus and (b) comparative cost. These two approaches measure the project's benefits more accurately than the estimated fare receipts. Corqumers' Surplus Alproach 60. The full benefits of the project are best reflected by the maximun amount each person would be willing to pay rather than forego his trip. The fares paid by the users in fact represent only the minimum amounts, i.e., the amounts they are required to pay, rather than the amounts they would be willing to pay. The estimation of consumers' surplus requires knowledge of the shape and slope of the demand curve for bus transportation. A rough indication of elasticities was derived from the effects of past fare increases on traffic volumes. 61. Information on the fare increases, operated seat kms and passenger kma for the buses is shown in Table 6 (page 34). It is assumed that the privately-owned buses, after their reintroduction in 1979, followed the same fare pattern as the public buses. as suggested in the S:i Lanka Transport Sector Study-I In considering the data, individual years are analyzed with and without the fare increases, to provide sufficiently short time periods that incomes, alternative means of transportation and other factors can be considered to remain relatively constant. and to some degree, support the assumption that uther things remain equal.* In the comparisons, the full effects of each price increase are assumed to occur in the year following each fare increase. 62. As shown in Table 6, there were 17.5 billion passenger kma in 1977. Fares were increased by an average of about 25S in July 1978. This was a period of rapidly itcreasing demand, attributed to the newly- liberalized economic policies of the government. If a relatively high "normal growth rate in demand of 182 per year from 1977 to 1979 is assumed, there would have been about 24.4 billion passenger kms in 1979, compared with the actual 23.3 billion. Thus, with an average increase of about 251 in fares, passenger kms decreased by about 4.5?. This indicates a low price elasticity of demand (around .18) over this price range. and suggests that people were willing to pay considerably more than they were required to paw. 63. In 1979. there were no fare increases and 23.3 billion passenger kan. Fares were increased by an average of about 451 in March 1980 and a further 752 in November L980. Total passenger kms for 1981 were 25.4 billion. Assuming the growth rate of 16S per year from 1979 to 1981 without the fare increase. there would have been about 31.3 billion passenger kms. Thus, a compound fare increase of about 1552 resulted in a decrease of about 18.81 in traffic, implying a price elasticity of demand of about 0.12 over this price range. 64. In 1982 there were no fare increases and 28.2 billion passenger km. Fares were -screased by 25Z in March 1983 and an average of about 24Z in August 1983. ..ie total for 1984 was 29.3 billion passenger kms. 1/ 1966, Vol. 3, pages 3-12.  -20- AsSUing a "normal" growth rate of 12? per year between 1982 and 1984 (similar to 1982, when there was no price increase), there would have been 33.8 blllion passenger kms in 1984. Thus, a compound fare increase of 501 resulted in a traffic decrease of 17.22. Again, the results indicate a low price elasticity of demand (0.35), although the elasticity is tending to increase at this higher fare level. 65. The buses under this project were added to the fleet between 1982 and 1986 to replace old buses, but 1984 was taken as the average year in which all buses were assumed to be added. If all riders of the publicly- owned buses had paid full fares, the total estimated revenue from private and public bus operations in 1984 would be Rs 0.188 per passenger km times 29.3 billion passenger kms, or about Rs 5,303 million. Using a base year of 1984 and assuming a price elasticity of demand of 0.5 (higher than the analyses shown in Table 6) and a linear demand curve above the 1983 faro level, the consumers' surplus would be about Rs 2,080 million or 392 of fare revenue. In other words, the total benefit of the service provided by the new buses would be about 392 greater than the fare revenue. 66. This is a simplification of a complex issue. in part because it uses a static analysis technique applied to a dynamic situation over a period when there were almost certainly shifting demand curves. However. the assumptions have generally been conservative, tending to overstate the demand elasticities and thus understate the consumers' surplus. As an example, if the 1983 fare increases had taken place one year earlier, they would not have had a greatly different effect on total passenger kms than they had in 1984. The actiual ridership in 1983 was 30.3 billion passenger k1m with an average posted fare of Rs 0.121 per passenger km, resulting in revenue of Rs 3,654 million. 67. If the fare increase had taken place in 1982. estimated ridership in 1983 would have been 13.32 less, or 26.7 billion. at a fare of Rs 0.181, resulting in revenue of Rs 4,833 million. It can thus be implied that riders would have been willing to pay additional fares amounting to Rs 1,179 million, or 32?, in 1983. It should be noted that this estimate considers only the consumers' surplus between the fare levels of Rs 0.121 and Rs 0.181, and does not include the remaining consumer surplus under the portion of the demand curve above Rs 0.181, yet it is almost as high as the estimate of 39? based on demand elasticities. This tends to confirm the conservative nature of the estimates, and therefore, a consumers' surplus of 39? of revenue is used in the subsequent analysis. 68. It is assumed that the 1,618 new buses provided under the project would have average lives of 12 years, with average utilization of 240 kms per day and 75? of the new buses in operation on any given day. With 54 seats per bus and load factors averaging 812, they would produce approzimately 4.65 billion passenger kms per year. Assuming all passengers pay the full posted fares (or alternatively, assuming that the benefits to those traveling under concessionary fares are the same as in those paying full fares), the revenue at Rs 0.181 per passenger km would be Rs 841 million per year. The total benefits, including the consumers' surplus of 39?, would be Rs 1.170 million per year.  -21- 69. The average operating cost of the publicly-owned buses in 1965 was estimated at Re 6.91 per bus km, excluding depreciation.21 It is assued that the improved maintenance, fuel efficiency and operations will reduce this by an average of St, resulting in operating costs of Rs 6.56 per bwu km and Rs 0.15 per passenger km. The operating cost at 4.65 billion passenger kmJ per year would then be Rs 698 million per year yielding annual net benefit of Rs 472 billion. 70. It is assumed that the proportions of financial and economic values are the same for both costs and benefits. With benefits of Rs 472 billion per year for 12 years, starting in 1984, the present value of the benefits as of the beginning of 1984, at a discount rate of lOZ per year, is Rs 3,216 million. It is assumed that the full expended project costs of Rs 1,941.8 million were required to obtain the benefits and operating cost levels estimated above, with the costs assumed to have been expended ln '.983. Under these assumptions, the benefit-cost ratio would be 1.7; the net present value Rs 1,274.2, and the internal rate of return 22.1Z per year. This is very close to the estimated rate of return of 20.42 shown in the SAR. Comparative Cost Approach 71. An alternative approach is to assume that, in the absence of the project, the privately-owned buses would provide the capacity which the project would otherwise have provided. Under this approach. the total costs of the service are compared with and without the project. 72. The private sector operated rather efficiently in maintaining low cost levels, but tends to operate primarily medium and small-sized buses because of capital constraints. The weighted average privately-owned bus in 1985 was estimated to have 21.5 seats, a load factor of 892. and total operating costs (including cepreciation but excluding interest) estimated at Re 4.17 per bus km. Rs 0.194 per L.z,t km and Rs 0.218 per passenger km. For the estimated 4.65 billion passenger kms per year provided under the project, the cost using privately-owned buses would be Ru 1,014. 73. The average operating cost for the publicly-owned buses in 1985 was Rs 7.95 per bus km, including the addition of Ru 0.77 to put depreciation costs on a replacement basis. However, this is based on the existing, overaged fleet and the resulting high maintenance costs. For the new project buses, it is estimated that the cost would be Rs 7.75 per bus km. Wlth an average of 54 seats and 81? load factors, this would result in costs of Rs 0.144 per seat km and Ru 0.177 per passenger km, for a total cost of Rs 823 million per year. Table 2 shows the age distribution of SLCTB fleet and Table 3 SLCTB operational indicators. 74. The cost saving of using public rather than private buses would be Rs 191 million per year, with a 12-year present value of Rs 1,301 million which discounted at 102 per year. In this total-.ost approach. the capital costs are included in the operating costs in the form of depreciaticen. It thus does not lend itself to the calculation of a benefit-cost ratiu or rate of return. However, the present value of the savings, Ru 1.301 2/ Transport Sector Study, Vol. 3, Table 4-1.  -22- million, can be roughly equated to the net present value of Rs 1,274 million estimated under the consumers' surplus approach. and indicates that the results are similar. 75. It should be noted that this result does not imply a judgment regarding the relative efficiencies of the public and private operators. The private operators maintain higher average load factors which reduce their costs per passenger km. but possibly at the cost of a higher degree of rider discomfort. They also have lower overhead costs, which represent a real economic advantage. In general, however. the lower costs of the private operators are offset by the more cost-effective large buses used by the publicly-owned operation. The results of the analysis. in favor of the publicly-owned buses, is thus almost entirely a function of bus size rather than of public versus private operating efficiency. 76. Both the economic reevaluation and the economic analysis contained in the SAR are fairly rough approximations. Although three different approaches have been used, the results of all three are similar, with rates of return in the 20 to 23Z range. The economic reevaluation had the benefit of additional information from the 1966 Transport Sector Study and additional years of traffic and cost data. making the task somewhat easier. However, it was not possible to make reasonable estimates of the rates of return on the individual components of the project, such as the technical assistance, traffic engineering, and improvements to workshops and depots. All of these components are considered necessary to attain the benefits estimated above, and their costs are included in the total capital costs. VIII. THE ROLE OF IDA 77. After being oat of the sector for almost a decade, IDA sent a miission to Sri Lanka in December 1977 to review the status of the sector. The mission identified the needs and recommended areas where IDA could play an important role in assisting the Sri La.xkan Government. 78. The rapid deterioration in the conditions of the road network and the moribund state of the bus fleet were seen as the transport areas where urgent assistance was needed. Two projects were identified for IDA's assistance, one of which was the Road Passenger Transport Project, under IDA Credit 994-CE. 79. IDA provided most of the funds for the project: at appraisal IDA financing was anticipated to cover about 61Z of the total funding requirement. However, due to severe budget constraints of the GOSL and higher than expected operating deficits of the SLCTB. the IDA credit eventually accounted for 742 of the total financing. Additionally, as SLCTB did not possess sufficient capability to prepare the project. a PPF was requested to engage consultants to prepare the project. Much of the supervision of the consultants, terms of reference was done by IDA. The review mission report was completed in 1978 and formed the basis for IDA's appraisal which visited Sri Lanka in August 1979.  -23- 80. IDA contributed to the policy reforms that the government undertook to rewitalize the sector aS a whole, and to road passenger service in particular. As a result, many restrictions which eliminated private bus operators in the past were removed which freed then to provide much needed paesenaer service in competition with the RTBs. IX. CONCLUSIONS Proiect Justification and Obiectives 81. Following an update to the 1966 Transport Sector Report in 1976, priorities for assistance were identified for the improvement of the road network and the road passenger service. A separate project under Credit 900-CE (Road Haintenance Project) was designed and implemented under the Department of Highways to arrest the deterioration of the road network. The road passenger service which operated under a public monopoly organization had a bus fleet of about 7,300 consisting of many overaged buses averaging ten years and older. Preventive maintenance was either not done or was erratic at best. Replacement of buses was estimated at about 22 of the fleet size as insufficient investment funds were provided. whereas growth in the passenger demand was estirated at 8Z and on the rise. At the time of appraisal, there was no indication that private operators would emerge to create the significant competition and cause such an impact on passenger service. 82. The project was thus seen as justified from the standpoint of the need to improve bus availability of the RTBs through a program of fleet replacement and upgrading the vehicle maintenance facilities. in aiming to meet the growing passenger demand, there was sufficient justification to support public organizations (SLCTB and RTBs as opposed to private operators) at the time as private operators had been absent for a number of years. Project Context and Scheduling 83. In retrospect, the project underestimated the consequences of the radical economic reforms introduced beginning in 1978 after a new government was elected into office. Among these reforms was the government policy to lift a long-standing ban on imports and operation of privately owned passenger vehicles. Further, lucrative incentives and generous tax rebates were provided to private bus operators. The response of the private sector ushered in a competitive element in the passenger service, effectively throwing the operations of the public road passenger service dad its institutions into a state of disarray. Unable to respond positively to the pressures of competition from the private operators, the SLCTBIRTBs became operationally vulnerable and were unprepared to implement the project in the manner that was originally deemed within their capabilities. 84. In addition, there are doubts whether or not the complexity, the size, and the scope of the project would have permitted a timely completion considering that both SLCTB and the RTBa were only a few months old. While  -24- RTBs had h-rdly completed moving into their newly established headquarters and branch offices in the regions. they were asked to execute various components of a complex project. The staff lacked the experience and/or understanding required to achieve the objectives. To further complicate the issue, government procedures which SLCTB and RTBs were required to follow were such that decision-making was very cumbersome and caused long delays in implementation. The appraisal implementation timetable did not adequately reflect this constraint. The level of complexity which project execution entailed in relation to the implementation cap:bility of the SLCTBIRTB suggests that project estimates were ambitious. Proiect Implementation and Operating Outcomes 85. Implementation of the project was carried out by consultants. especially during the initial period. However. practically all contracts with consultants expired within the original closing date of the credit. Much of the unfinished work which later was done through extensions of the closing date of the credit was carried out by the SLCTB!RTBs themselves. The burden of project implementatiorn proved to be too heavy on SLCTBIRTBs whose top management was at the same time undergoing continual turnover. and remained unprepared to make critical decisions on various important issues of project execution. These problems resulted in numerous delays bringing the total implementatien time to 61 months compared with the original plan of 36. Considerable delays were attributable to general tardiness of the Borrower procurement procedures. Operating outcomes in many ways fell below expectations. Operating efficiency measured by bus availability and utilization worsened; operating costs of the RTBs escalated and financial deficits did not abate even though generous bus fare increases were approved by the government and promptly administered by the RTBs. Sustainability of Proiect Benefits 86. The project demonstrates the complexities and inefficiencies of a series of public sector agencies operating bus fleets and emphasized the constant need for attention to appropriate technical specifications in fleets; to proper pricing policy and the need to control operating costs. These lessons were derived from the project but will only be sustained with a contiaued commitment to sound management principles. The Transport Sector Study has been used as the basis for preparation of other transport projects.  -25-A%L I PQJ COMPLUION RCPORT ROAD PASSDNGtE tRANSPORT PROJtCT (CREDIT 994 U) Cuulativa Disburso uts: Actual od Estimted 0US5$000) Appraisal Use. Actual FY 1980 4th Quarcer 37 FT 1961 Ist Qurcer 5,054 24d Quarter 10,031 265 3rd Quarter 20,646 4ch Quarter 26,361 2,700 FT 1982 Ist Quarter 32,076 2nd Quarter 37,791 5,310 3rd Quarter 41,593 7,810 4th Quarter 4,395 12,3350 FY 1983 Ist Quarter 49,197 2nd Quarter 53,000 16,330 3rd Quarcter 20,000 4th Quarter 21,750 FT 1984 Ist Quarter 25,500 2nd Quarter 3rd Quarter 27,200 4th Quarter FT 1981 Ist Quarter 2nd Quarter 3rd Quarter 4th Quarter 35,800 FT 1986 Ist Qua3t,45 2nd Quarcer 393 3rd Quarter 39,998 4th Quairter, 51,635 1.1 1I Actually July 2, 1986, but eudiJune regaided as cut-Off date.  PROJECT COHPLETION REPORT SRI LANKA ROAD PASSENGER TRANSPORT PROJECT (CEDT 99*-CE1 Summary of Halor Covenants and Status of Compliance Aureement Sect-on Status Desc-lPtion CeXnent5 Cretit 3.02 A Subsidiary loan agreeoment betwoen borrower and SLCTD. Credit 3.07 Complied borrower to furnish IDA with An action comuittew wOa turnOd policy stat5emnt re private by the Minister of lranaport sector role In passenger boards tO detine rols at private transport. sector &nd mes recoefinoftions to *ove*rewet. A directorate of privato conibus was estlairbnda. Credit 3.01 Partially borrower to taue measuras to Provisional tigur*s of profit and Luumpl tit r5iucS op5erting losses, lose statseent pr SpdtCu by SLEISI show a toos ot aoOut Us 311 million for 1985 compared with a laOS of as 09 omillon for 1982. Credit 3.09 Cunipl ed Borrower to arrange that each Route costing eMercibd island-wide pTS adjubt fores according to was carrlod out which loan formed SLCI0 gudelineS, tngo basis for fire adjustments. Credit 3.10 con.plied SLCTI and RI0s not to meae Any capital *mpanditures over as 50.000.000 in one financial (I! year. * X Credit 3.10 B uot udistea accunts5reports aue. Uue to the deplOrable conaltion of Complied ~~~~~~~~~the occoutaftini functions auoits of Complied hoth the SLCTS and AlTs can oistea