Central African Republic - Highway Projects

Document of The World Bank FILE COPy FOR OFFICIAL USE ONLY Report No. 1819 PROJECT PERFORMANCE AUDIT REPORT CENTRAL AFRICAN EMPIRE FIRST HIGHWAY PROJECT (CREDIT 146-CA) AND HIGHWAY MAINTENANCE PROJECT (CREDIT 199-CA) December 13, 1977 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.  FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT CENTRAL AFRICAN EMPIRE FIRST HIGHWAY PROJECT (CREDIT 146-CA) AND HIGHWAY MAINTENANCE PROJECT (CREDIT 199-CA) Table of Contents Page No. PREFACE PROJECT PERFORMANCE AUDIT BASIC DATA SHEETS HIGHLIGHTS PROJECT PERFORMANCE AUDIT MEMORANDUM I. Background 1 II. Project Results I III. Main Issues 3 IV. Future IDA Support for Highway Construction 7 V. Conclusions 8 Annex Guidelines for Assessment of Training Needs ATTACHMENT: PROJECT COMPLETION REPORT 1. Introduction A.1 2. The Project ) (Cr. 146-CA) The A.2 3. Project Implementation ) First Highway A.3 4. Economic Re-evaluation ) Project A.3 5. The Project ) A.5 6. Project Implementation ) (Cr. 199-CA) The A.5 7. Economic Re-evaluation ) Highway Maintenance A.8 8. Borrower's Performance Project A.9 9. The Role of IDA A.10 10. Conclusions A.11 Annex A General Comments and Observations Tables 1. Costs and Basic Data for Calculation of Economic Return, First Highway Project 2. Schedule of Accumulated Disbursements, First Highway Project 3. Highway Maintenance Project Status and Cost Estimates 4. Costs in Local Funds, Highway Maintenance Project 5. Highway Maintenance Field Operations 6. Schedule of Accumulated Disbursements, Highway Maintenance Project 7. Recomputation of Economic Return, Highway Maintenance Project Maps 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. & PROJECT PERFORMANCE AUDIT REPORT CENTRAL AFRICAN.EMPIRE FIRST HIGHWAY PROJECT (CREDIT 146-CA) AND HIGHWAY MAINTENANCE PROJECT (CREDIT 199-CA) Preface This report presents a performance audit of the Central African Empire (CAE) First Highway Project, for which Credit 146-CA in the amount of US$4.2 million was closed in 1975, and Highway Maintenance Project, for which Credit 199-CA in the amount of US$4.3 million was closed in 1976. The memorandum is based on the attached Project Completion Report (PCR) prepared by the Bank's Western Africa Regional Office on July 23, 1976, discussions with Bank staff, review of project files and minutes of the Board of Executive Directors' meetings which considered the projects, and a one-week mission to CAE by OED staff in May 1977. The Department of Public Works, Equipment, and Land Development provided assistance to the mission and comments on the draft audit, both of which are grate- fully acknowledged. On the basis described above, the audit accepts the principal conclusions of the PCR. However, it expands on possible future reduc- tions in benefits from road construction under the Highway Project due to neglect of maintenance operations. The audit also discusses the lack of improvements in maintenance and of adequate planning of the training program under the Maintenance Project. In addition, it comments on measures which seem essential for IDA to take before supporting future highway construc- tion projects in Central African Empire.  PROJECT PERFORMANCE AUDIT BASIC DATA SHEET CENTRAL AFRICAN EMPIRE: FIRST HIGHWAY PROJECT (CREDIT 146-CA) KEY PROJECT DATA Appraisal Actual or Item Expectation Current Estimate Total Project Cost (US$ million) 5.6 4.8 Underrun (. - 15 Credit Amount (US$ million) 4.2 4.2 Disbursed and Outstanding to 10/31/ 77 - 4.8-- Date Physical Components Completed 1/72 6/72 Proportion Completed by Above Date (7. 84- Proportion of Time Overrun (%)- 17 Economic Rate of Return (7) 14 14 OTHER PROJECT DATA Original Item Plan Revisions Actual- First Mention in Files - 10/65 Government's Application - - Early 1966 Negotiations 1/69 - 1/27-31/69 Board Approval 3/69 - 3/25/69 Credit Agreement Date 4/69 - 4/3/69 Effectiveness Date 6/31/69 - 6/11/69 Closing Date 3/31/7_.L. 3/31/74 and 12/31/74 7/7_;5 Borrower Central African Empire Executing Agency Department of Public Works Fiscal Year of Borrower January 1 - December 31 Follow-on Project Name Highway Maintenance Project Credit Number 199-CA Amount (US$ million) 4.3 Credit Agreement Date 6/19/70 MISSION DATA Date Month/ No. of No. of of Full Item Year Weeks Persons Manweeks Report Identification 7/66 n.a. 1 n.a. 8/66 Preappraisal I 10/67 1 2 2 11/67 Preappraisal II 2/68 1 1 1 2/68 Appraisal 10/68 1 3 3 3/69 Supervision I 10/69 n.a. 1 n.a. 11/69 Supervision II 6/70 1 1 1 7/70 Supervision III 3-4/71 1 1 1 6/71 Supervision IV 10/71 1 2 2 11/71 Supervision V 1-2/72 1 1 1 2/72 Supervision VI 4/72 1 1 1 5/72 Supervision VII 5-6/74 1 1 1 6/74 Supervision VIII 11/75 2 2 4 1/76 COUNTRY EXCHANGE RATES Name of Currency CFA franc (CFAF) Year: 1969 Exchange Rate: US$1 = CFAF 247 1970-72 US$1 = CFAF 249 1973 US$1 = CFAF 223 /1 An amount of US$652,000 was not needed for expenditures related to this project and was transferred to the Highway Maintenance Project (Credit 199-CA of June 1970 for US$4.3 million). /2 Includes exchange adjustment of US$600,000. 7T Proportion disbursed. As shown in Credit Agreement. 7 Final disbursement date.  PROJECT PERFORMANCE AUDIT BASIC DATA SHEET CENTRAL AFRICAN EMPIRE: HIGHWAY MAINTENANCE PROJECT (CREDIT 199-CA) KEY PROJECT DATA Appraisal Actual or Item Expectation Current Estimate Total Project Cost (US$ million) 5.4 5.7 Overrun (%) - 6 Credit Amount (US$ million) 4.3 4. /1 Disbursed and Outstanding to 10/31/77 - 4.6/2 Date of Completion of Procurement and Feasibility Study 12/74 1/75 Proportion of Expected Maintenance Achieved at Completion (%) 100 10-20 Proportion of Time Overrun () - 4 Economic Rate of Return (%) 28 -20/3 Institutional Performance - Slightly better than at appraisal OTHER PROJECT DATA Original Item Plan Revisions Actual First Mention in Files - 10/69 Government's Application - - n.a. Negotiations 5/70 - 5/4-8/70 Board Approval 6/70 - 6/11/70 Credit Agreement Date 6/70 - 6/19/70 Effectiveness Date 4/21/70 12/31/70, 2/15/71, 3/15/71, 6/10/71 4/15/71, 5/15/71, and 6/30/71 Closing Date 6/30/75 12/31/75 2/76/5 Borrower Central African Empire Executing Agency Department of Public Works Fiscal Year of Borrower January 1 - December 31 MISSION DATA Date Month/ No. of No. of of Full Item Year Weeks Persons Manweeks Report Appraisal 10/69 3 2 6 5/70 Supervision I 6/70 1 1 1 7/70 Supervision II 3-4/71 1 1 1 6/71 Supervision III 10/71 2 2 4 11/71 Supervision IV 1-2/72 1 1 1 2/72 Supervision V 4/72 1 1 1 5/72 Supervision VI 5-6/74 1 1 1 6/74 Supervision VII 11/75 2 2 4 1/76 Total (I-VII) COUNTRY EXCHANGE RATES Name of Currency (Abbreviation) CFA franc (CFAF) Year: 1970 Exchange Rate: US$1 = CFAF 278 1971-76 US$1 = CFAF 227 /1 Plus US$652,000 not needed for expenditure related to the First Highway Project (Credit 146-CA of April 1969 for US$4.2 million), which was transferred to this project. /2 Includes exchange adjustment of US$300,000. 73 Based on educated guess assumptions and application of the appraisal report methodology. /4 As shown in Credit Agreement. /5 Final disbursement date.  PROJECT PERFORMANCE AUDIT MEMORANDUM CENTRAL AFRICAN EMPIRE FIRST HIGHWAY PROJECT (CREDIT 146-CA) AND HIGHWAY MAINTENANCE PROJECT (CREDIT 199-CA) Highlights The First Highway and Highway Maintenance Projects in Central African Empire were to concentrate, respectively, on expanding and main- taining the highway network. The straightforward First Highway Project was implemented successfully, although the need for retendering delayed completion for five months. Lower unit prices than expected resulted in a 15% underrun in the actual project cost. Of the actual cost, 96% was devoted to road construction and the balance to purchases of maintenance equipment, spare parts, and fuel. The economic rate of return is re- estimated to be 14%, the same as the appraisal estimate. The Highway Maintenance Project was complex and too ambitious. Only 10% to 20% of the planned maintenance program and a limited amount of training were accomplished. This unsatisfactory outcome is primarily explained, respectively, by the unavailability of sufficient local funds for recurrent maintenance expenditures and the inadequate planning of the training program. Because of the shortage of local funds, the project was scaled down three times. The smaller project was completed within the time estimated for the original project. Higher equipment prices than expected resulted in a 6% overrun in the actual project cost. Of the actual cost, 92% was allocated to a four-year maintenance program. The program included purchases of road maintenance equipment, spare parts, and materials, estab- lishment of accounting and inventory techniques, and training of local staff. The balance of the project was devoted to preinvestment studies for a main road. Because the planned maintenance improvements were not implemented, the re-estimated economic rate of return is minus 20%, com- pared with the appraisal estimate of 28%. The following points may be of particular interest: Reduction in benefits from road construction due to neglect of maintenance operations (para. 7); Detrimental effects on the project of the shortage of local funds and diversion of equipment (para. 13 and PCR, paras 6.02-6.08, 7.04, 8.02-8.04, and 9.02); Shortcomings of the "global" approach to formulation of the training program (paras 18-22); Lack of full realization of the training program (paras 23 and 24); and Need for risk analysis and workable requirements for Government action before future IDA support for highway construction projects (paras 26 and 27).  PROJECT PERFORMANCE AUDIT MEMORANDUM CENTRAL AFRICAN EMPIRE FIRST HIGHWAY PROJECT (CREDIT 146-CA) AND HIGHWAY MAINTENANCE PROJECT (CREDIT 199-CA) I. Background The Transport System 1. Central African Empire (CAE) has a simple but adequate transport system. Bangui, site of the capital, main port, and international airport, is the focal point. The dominant mode is the 22,000 km road network. Major roads radiate from Bangui, and the main internal transport axis is CAE's 1,300 km section of the Trans-African Highway (Cameroon Border-Bossembele- Bangui-Sibut-Zaire Border). Also important to this landlocked country is its access to the Atlantic Ocean via the 1,800 km Transequatorial Route, either by river and rail between Bangui and Pointe Noire (Congo) or by road and rail between Bangui and Douala (Cameroon). IDA Support for Highways 2. Two IDA credits totalling US$8.5/million have been provided since 1969 to assist highway development in CAE. Credit 146-CA of April 1969 provided US$4.2 million under the First Highway Project for reconstruction of the Bangui-M'Baiki Road and purchases of equipment to maintain the road (PCR, para. 2.01). Credit 199-CA of June 1970 made available US$4.3 million under the Highway Maintenance Project for a four-year maintenance program (PCR, para. 5.01) to ensure that the condition of primary and secondary roads would not hinder further economic development. Both projects have been completed, and their results are compared in this Project Performance Audit Report. II. Project Results Implementation 3. The Highway Project was implemented successfully and in line with the scope envisaged at appraisal. Completion was five months later than expected because retendering was required after devaluation of the French and CFA francs. Due to the resulting lower prices than expected at appraisal, a 15% cost underrun materialized, as shown in the following table: - 2 - Actual Appraisal Cost Project Item Cost Estimate Difference US$ thousand % Bangui-M'Baiki Road Reconstruction Soil and Topographic 32 Surveys Civil Works 4,469 5,160 -13 Evaluation of Bids and Supervision 80 88 - 9 Road Maintenance Equipment and Spare Parts 169 310 -45 Total 4.,750 5.oo -15 The US$652,000 IDA share of the US$850,000 project surplus was transferred to the Maintenance Project for purchases of spare parts and fuel. Although the Government undertook to maintain the road that was built in good condi- tion, maintenance of some sections appeared to be necessary at the time of OED staff's May 1977 visit in connection with this performance audit. Imple- mentation is further described in PCR, paras. 3.01-3.04 and Table 2. 4. In retrospect, the Maintenance Project seems to have been too ambitious 1/, as discussed in Section III. The expected maintenance improvements were not achieved, consequently roads already in poor condition deteriorated even further. This unsatisfactory outcome resulted primarily because the Government was not able to provide sufficient local funds for recurrent maintenance expenditures in accordance with its agreement with IDA. Of the funds provided, some were diverted to other purposes (PCR, para. 6.03). Project equipment also was shifted to non-project uses (PCR, para. 6.04). To relieve the Government's financial burden, IDa and"the dovernment agreed to scale down the project three times. Only 10% to 20% of the planned mainte- nance program was ultimately completed. In addition, training of local staff in the Department of Public Works (DPW) was below expectations, as discussed in Section III. Nevertheless, the consultants were able to introduce to the DPW a sound organizational structure as well as modern inventory and accounting procedures and to prepare road maintenance manuals. Other consultants satis- factorily completed preinvestment studies of a main road section. The smaller project was finished essentially within the time estimated for the original project. A 6% cost overrun materialized due to increases in equipment costs. Further details on implementation are in PCR, paras. 6.01-6.14 and Tables 3-6. 1/ For example, technical assistance was increased from the consultants' recom- mendation of 36 man-years to 62 man-years, see PGR, Annex A, page 4. -3- Economic Reevaluation 5. An economic rate of return of 14% was estimated at appraisal of the Highway Project based on expected reductions in vehicle operating costs (voc). Using the same methodology as at appraisal, actual construction costs, and revised inputs to traffic, voc, and maintenance cost data, the PCR re-estimation of the rate of return was also 14%. However, the neglect of maintenance in the future may lead to a reduction in the rate of return, as indicated in Section III. 6. At appraisal of the Maintenance Project, reductions in voc due to the maintenance program were expected to yield an economic rate of return of 28%. The PCR re-estimation of the rate of return using the same methodology as at appraisal and allowing for residual values of work done and equipment supplied under the project, concludes that the rate of return is minus 20%. This negative rate of return reflects the fact that the planned maintenance improvements were not implemented. It does not take account of benefits from non-planned works,such as improvement of streets in Bangui. The PCR suggests that some benefits may materialize from the institutional framework established under the project if funds become available for reactivation, and this is discussed in Section III. III. Main Issues Reductions in Benefits from Road Construction Under the Highway Project due to Neglect of Maintenance Operations 7. Vehicle operating cost savings as a result of improvements of the Bangui-M'Baiki Road may be reduced in the future as a result of continued neglect of maintenance operations, such as regravelling of shoulders. Be- cause shoulders have been allowed to wear away, water can collect along the road edge in the rainy season and lead to weakening of the pavement. In addition, pavement surface or subgrade has failed in some sections. To return the road to good condition, maintenance has to be increased substan- tially. Delay for more than two or three years may lead to the need for reconstruction of large sections as a result of further failures and to increased user costs, either of which could reduce the economic rate of return to less than 10%. IDA hopes that the necessary maintenance of this road will be included in the four-year road maintenance program to be ini- tiated in 1978 as a result of arrangements made between the Government and external lenders in April 1977 (PCR, para. 10.03). Lack of Maintenance Improvements Under the Maintenance Project 8. The lack of maintenance improvements raises the question of whether a more cautious approach than that taken at the time of project formulation would have resulted in a better outcome. To determine the answer, examination is necessary of the earlier efforts to improve highway maintenance and the findings of the 1968-69 study by Research and Develop- ment (R&D) (Belgium) on which the maintenance program had been based. Technical assistance in highway maintenance had been provided to the DPW by Fonds d'Aide et de Cooperation (FAC) after independence in 1960. Although IDA's files do not specify the fields for which this assistance had been given, they do indicate that little progress had been made. 9. For its study, R&D developed traffic forecasts, recommended yearly expenditures for highway maintenance, and evaluated DPW maintenance staff. R&D's traffic forecasts for 1969-74 were 9%-11% yearly. They were based on an acceptance of the 1967-70 Development Plan forecast of a yearly increase of 18% in agricultural production for 1966-71, although official figures indicated only 2% for 1959-68. 10. As to finance for highway maintenance, R&D recounted that local funds for recurrent maintenance expenditures had been provided from a Road Fund until 1963 and from general revenue thereafter. This change meant that maintenance had to compete with other demands that might be considered to have higher priority. As a result, R&D considered that the Government spent about 30% less than needed on highway maintenance in 1965-70. R&D estimated that expenditures of CFAF 511 million would be required yearly from 1971-75 for reasonable maintenance operations (including elimination of the backlog that had developed) in contrast to expenditures of CFAF 300 million in 1968 and CFAF 240 million in 1969. 11. On the subject of DPW's maintenance staff, R&D stated that recruitment generally was difficult. In particular, applicants for field positions desired more attractive salaries and living conditions. R&D also described the most important category of staff necessary for efficient maintenance operations, subdivision engineers, as having insufficient capa- bility. Generally they were former qualified workmen who had acquired some experience in subdivision management but not the essential technical back- ground. R&D's findings related to training of DPW technical staff are dis- cussed in para. 19. 12. In retrospect, it seems that problems were created by the accept- ance during project formulation of R&D's optimistic traffic forecasts while disregarding its realistic accounts of the insufficient local funding of maintenance and of the unsatisfactory skills of DPW maintenance staff. Traffic actually grew 5% yearly, or 4%-6% less than R&D had estimated (para. 9). The detrimental effects on the project which resulted are discussed in paras. 13-16 and 20-23. 13. During project implementation (PCR, paras. 6.01-6.08), the Govern- ment was not able to establish promptly the new Road Fund that had been agreed with IDA or to provide sufficient local funds for recurrent mainte- nance expenditures on spare parts and fuel. Insufficient amounts were allo- cated to the Road Fund and some of the limited funds were diverted to other purposes. At this point, it is relevant to remark that the US$6.4 million in local funds required for recurrent costs under this project was consid- erably larger than the US$1.4 million required under the Highway Project. The PCR (para. 9.03) mentions that at appraisal there were doubts as to the availability of local funds. In retrospect, the 20% increase in recurrent - 5 - costs from CFAF 265 million yearly in 1965-69 to CFAF 330 million yearly during the project period seems to have been too heavy a burden for the Government to bear. The consequences for the project of the insufficient local funds were three reductions in scope, rescheduling of work, and transfer of US$652,000 not required as IDA's share of a surplus from the Highway Project to the Maintenance Project for purchases of spare parts and fuel. Even with these measures, only about 10%-20% of the planned maintenance program was achieved. The PCR (para. 9.02) explains the cir- cumstances surrounding IDA's decision not to cancel the credit. 14. Some benefits may yet be realized from the institutional frame- work established under the Maintenance Project once funds are available to reactivate operations, according to the PCR (para. 7.04). However, although the systems and procedures established by the consultants remain, the knowledge and experience to operate them steadily erodes with inactivity. Retraining of staff and revision of procedures to meet changed circumstances will be required to overcome the present inactivity in highway maintenance that was observed. 15. OED staff visited central workshops and stores and the Bangui- M'Baiki and Bangui-Damara Roads. They also held general discussions about the condition of other roads in the country. In the workshops, a small amount of repairs was being carried out by reasonably competent mechanics, although tools, spares, and consumable items were in short supply. The premises were untidy, with pieces of equipment scattered about awaiting spares or disposal. In the stores, little stock was evident and turnover for available parts seemed to be slow. The stock control system introduced by the consultants was being followed, but reconciliation of stock cards with physical holdings or easy location of those holdings was not always possible. 16. On the Bangui-M'Baiki and Bangui-Damara Roads, maintenance was poor. The situation on the Bangui-M'Baiki Road has already been discussed (para. 7). A similar situation exists on the Bangui-Damara Road, where shoulders were badly eroded and sections of the road had deteriorated as a result of water seepage due to lack of resealing. As to other roads, average speeds on more heavily trafficked unpaved routes, such as Sibut- Bambari, are reported to be significantly lower now than before the project because of the need for maintenance. However, average speeds and comfort on some unpaved roads having low traffic volumes are said to have improved as a result of intensive maintenance carried out during implementation of the project. 17. In retrospect, it seems that a smaller project directed toward fulfillment of institution-building objectives and achievement of urgent physical needs might have led to a better outcome. Lack of Adequate Planning of the Training Program Under the Maintenance Project 18. On the basis of the R&D study for the project and because of the importance of highway maintenance in furthering economic development as well - 6 - as of competent staff in carrying out the work involved, the project included a training component. Through that component, local staff were to acquire specific skills required to maintain highways satisfactorily in the future. 19. R&D's study listed the training establishments and courses avail- able inside and outside CAE as well as their expected output of local staff for 1971-74. As to training inside CAE, R&D identified the most active institution as the Public Works Intermediate Management Training Center, established in Bangui in 1967 and financed by the Fonds Europeen de Development (FED). The Center was providing courses for works and mechanical supervisors. During the project period, the Center was expected to graduate 15 in each category, which would have been sufficient to fill expected vacancies. As to training outside CAE, 15 technical managers were expected to graduate in the project period. This number would not have been sufficient to fill vacancies for experienced managers. 20. R&D's study did not state sufficiently clearly for project defi- nition purposes the numbers, categories, levels, capabilities, or potential of other personnel who required training. Also it did not specify the duration envisaged for the training, the methodology to be used for trans- ferring expertise from consultants to maintenance personnel, or the inclusion of a training expert among the consultants. These information gaps were to carry through the formulation of the training component of the project. 21. This situation resulted because the Terms of Reference for R&D's study had not required an assessment of the overall training needs of the maintenance organization or the design of incentives which would induce staff to remain in Government service after completion of training. No indication that these activities took place appears in IDA files. Infor- mation about many factors crucial to the success of training efforts may be obtained through an assessment of training needs. At the time this project was formulated, IDA had not developed guidelines to assist in examining the status of training needs and in formulating proposals to deal with the findings. Such guidelines have been compiled over the last five years, however, and are attached as an Annex to this memorandum. 22. IDA's Appraisal Report for the project described the training component on a "global" basis within the context and funding of the four- year technical assistance program. Training was to be provided for sub- division and central workshop staff by technical assistance supervisors and technicians, subdivision mechanics on-site by qualified mechanics, and accountants by a qualified accountant provided under technical assis- tance. Although mention also was made that vocational training through the school system should be established to meet long-term requirements for training of mechanics and foremen, no precise program was recommended. Many of the information gaps left by R&D (para. 20) remained, including the duration of training, although 62 man-years were to be devoted to technical assistance. Furthermore, separate costs of training were not mentioned. - 7 - 23. During the project, the technical assistance consultants carried out on-the-job training of counterpart mechanics, a storekeeper, and accountants by daily contacts and by two hour weekly meetings. They also proposed that professional training be given by specialists in Bangui over about 14 months and that visits be made to public works oper- ations and equipment manufacturers abroad. However, a year after the project had begun, no decision had been made by IDA and the Government concerning the proposal and shortly thereafter the scope of the project had to be reduced. Eventually, the proposal had to be abandoned as finance was not available. 24. Lack of any significant amount of maintenance in progress makes evaluation of improvements as a result of training difficult. The DPW feels that the training of works and mechanical supervisors has not been effective in improving their field performance. The on-site training of mechanics appears to have proceeded throughout the project and to have had some effect, although the training of the storekeeper had been less productive. A two-week course given to fourteen accountants in Bangui in 1973 resulted in two graduates ready for promotion and five competent for the job. The remaining seven were rated barely competent or in- competent for accounting work. IV. Future IDA Support for Highway Construction 25. The PCR (para. 10.03) states that IDA is now planning to support a Third Highway Project comprising construction of the main road section for which preinvestment studies were completed under the Maintenance Pro- ject. It also mentions that IDA made financing conditional on the Govern- ment arranging a satisfactory program of minimum highway maintenance sup- ported by external financial and technical assistance. 26. In view of the difficulties encountered in establishing a mainte- nance system under the Maintenance Project, it is important that the eco- nomic analysis for future highway construction projects take account of the definite risk that a completed facility may not be maintained as expected. Without adequate maintenance, the life of the facility would be shortened and voc would be increased. 27. To reduce the risk that once highways are constructed they may not be maintained, it is important to establish and enforce workable requirements for Government action before future construction projects are supported. These requirements could include completion of rehabilitation of maintenance equipment purchased under the Maintenance Project, agreement not to divert the rehabilitated equipment to non-maintenance uses, and evidence that funds allocated for highway maintenance have in fact been spent for that purpose. Although the latter two requirements have been provided for in the draft Credit Agreement for the proposed Third Highway Project, IDA recognizes that the most effective safeguard would be a firm Government commitment to mainte- nance of the national highway network. - 8 - V. Conclusions 28. The uncomplicated Highway Project successfully contributed to the development of CAE's road sector. However, the more complex and, in retrospect, too ambitious, Maintenance Project exceeded the capability of the DPW. A smaller project, oriented toward institution-building objectives and physical achievement of urgent needs, might have been more appropriate. The lack of sufficient funds for recurrent maintenance expenditures and reassignment of project equipment prevented completion of at least four-fifths of the planned maintenance operations. Training under the project also resulted in less progress than expected. Neverthe- less, the project did introduce the principle that regular highway mainte- nance is essential. 29. In view of the difficulties experienced by FAC and IDA in their past highway maintenance efforts, it is important that the economic analy- sis of future highway construction projects take account of the definite risk that once facilities are constructed they may not be maintained as expected. ANNEX Page I of 2 PROJECT PERFORMANCE AUDIT MEMORANDUM CENTRAL AFRICAN EMPIRE FIRST HIGHWAY PROJECT (CREDIT 146-CA) AND HIGHWAY MAINTENANCE PROJECT (CREDIT 199-CA) Guidelines for Assessment of Training Needs 1. Preparation of inventory of existing maintenance and mechanical per- sonnel and future personal requirements for the duration of the proposed program, including: (a) established positions broken down by numbers, categories, and levels; (b) positions presently filled and vacant; and (c) additional personnel required by a certain future date broken down by year. 2. Identification of the skills and knowledge required by personnel at all levels for efficient maintenance operations. 3. Evaluation of the capability and potential, including literacy level and language proficiency, of available personnel so as to better assess training needs. 4. Determination of how existing personnel can be made available for retraining and upgrading training without interfering with ongoing operations. 5. Identification of possible sources of recruitment and related levels, and assessment of likely constraints. 6. Evaluation of existing training facilities and programs in the country and determination of their suitability and availability for training of maintenance personnel. 7. Determination of need for training abroad: (a) categories of personnel to be trained; (b) where training is to be carried out and by whom; and (c) duration of programs. 8. Identification of the number and types of courses/programs, short and long-term, for a certain period, including: (a) number of trainees in each class; and (b) duration of classes. ANNEX Page 2 n 2 9. Deter-ination of the number and types of instructors required to implement the program. 10. Determination of the number and types of local technical staff to be trained to become instructors over a certain period. 11. Evaluation of existing personnel/training policies. 12. Definition of the organizational structure, functions, legal frame- work and related operating budget of a permanent Training Section within the highway administration, responsible for the preparation, implementation, and supervision of all training. 13. Estimation of the cost of the training program, broken down between local currency and foreign exchange, as follows: (a) civil works for physical facilities (if applicable and required): (i) classrooms; (ii) workshops; (iii) stores; and (iv) dormitories. (b) training equipment and materials (if applicable and required): (i) heavy road construction/maintenance equipment for the number of operators at the training facilities; (ii) machines and hand tools; (iii) training materials; and (iv) audio-visual equipment. (c) training abroad (if applicable and required): (i) cost of individual programs. (d) operating budget for the training program. 14. Determination of the Government capability to finance recurrent coats of the training program. 15. Determination of the incentives necessary to induce local per- sonnel to remain in Government service after completion of training. ATTACHMENT July 23, 1976 CENTRAL AFRICAN EMPIRE (REPUBLIC) CREDIT 146-CA - FIRST HIGHWAY PROJECT CREDIT 199-CA - HIGHWAY MAINTENANCE PROJECT PROJECT COMPLETION REPORT 1. INTRODUCTION A. Origin of the Projects 1.01 The land-locked country of Central African Republic has approximately 21,000 km of roads and tracks providing reasonable access to most parts of the country. The effectiveness of much of the road network, however, is reduced by a lack of adequate maintenance leaving many areas difficult and costly to reach. The CAR is presently at an early stage of economic development. Its resources in agriculture, forestry and minerals provide the potential for gradually accelerating economic growth, though for the immediate future, growth prospects remain modest. 1.02 The Bangui-M'Baiki (Cr.146-CA) road connects Bangui, the industrial and trade center of CAR, with the Lobaye province, one of the most agricul- turally active and potentially promising regions of the CAR. The province is the main producer of coffee, rubber and timber. Detailed engineering of the reconstruction of the Bangui-M'Baiki road was carried out by the consulting firm of SAUTI (Italy) during 1965 and 1966. Engineering was financed by Fonds Europ6en de Developpement (FED) but it was not sufficiently complete for cost estimates and bidding documents to be prepared. Department of Public Works (DPW) supplemented bridge designs and carried out additional soil topographic surveys in 1968. The appraisal took place in October 1968, and the Credit Agreement was signed on April 3, 1969 for the First Highway Project (Credit 146-CA). 1.03 While the appraisal mission was in the field for the First Highway Project, formulation of what was to become the Highway Maintenance Project (Credit 199-CA) started. The Government requested the UNDP (ith the Bank as executing agency) to finance a study of the road maintenance requirements and to develop an improved highway maintenance program. The study began in October 1968, carried out by the consultants Research and Development (Belgium). Their draft final report was completed in July 1969. 1.04 The UNDP study and subsequent discussions between the Bank and the Government led to a request for IDA assistance to finance a four-year mainte- nance program to provide improved maintenance for the main road network and a feasibility study and detailed engineering for the improvement of the Bangui- Bossemb616 road. The project was appraised in October 1969 and the Credit Agreement signed on June 19, 1970 for the Highway Maintenance Project (Credit 199-CA). - A.2 - (Cr. 146-CA) THE FIRST HIGHWAY PROJECT 2. THE PROJECT A. General Description 2.01 The project consisted of: (i) improvement of the Bangui-M'Baiki road (102 km); (ii) supplementary engineering works and supervision of construction for the above road carried out by the Department of Public Works; (iii) procurement of equipment to maintain the road. The project was amended on July 27, 1973 to include: (iv) procurement of spare parts and fuel for highway maintenance operations carried out under Credit 199-CA. B. Cost Estimates 2.02 The total cost of the project at time of appraisal was estimated at about CFAF 1,390 million (about US$5.6 million at an exchange rate US$1.00 = CFAF 247), taxes and duties included. The reconstruction cost of the Bangui-M'Baiki Road (excluding supervision and contingencies) was estimated at CFAF 1,060 million (US$4.3 million). Association financing was to cover the estimated foreign exchange cost of the project amounting to US$4.2 million (Table 1). - A.3 - 3. PROJECT IMPLEMENTATION A. Construction Works 3.01 The Department of Public Works completed engineering works in 1968 on the Bangui-M'Baiki road. Start of construction was delayed by about five months because the work was retendered after the devaluation of the French and CFA francs in August 1969. On the basis of the new bids, a contract was awarded in February 1970 to the lowest bidder SOFRA TP-Fougerolle (France). 3.02 The contractor maintained his original work schedule, and con- struction was satisfactorily completed in June 1972. The works were adequately supervised by a DPW team provided by French technical assistance (FAC). DPW soils laboratory was used by the supervising team in checking quality of materials and compaction of earthworks. On expiration of the one-year warranty in July 1973, the Government accepted the work as satisfactory. The equipment provided under the Credit for maintenance of the road was delivered and put into use. B. Special Covenants 3.03 The Government agreed to provide qualified personnel for supervision of construction works, carry out traffic counts and establish a road mainte- nance organization for the Bangui-M'Baiki road and make available local funds as needed. The Government complied with the above covenants. C. Disbursements 3.04 With all project items completed, a total amount of US$3.548 million was disbursed, leaving an undisbursed balance of US$652,000. The Executive Directors approved use of these funds for the Highway Maintenance Project (Cr. 199-CA). (Table 2). 4. ECONOMIC RE-EVALUATION A. General 4.01 The project was re-evaluated using available data, including the only traffic count carried out since 1968. The scarcity of traffic data do not permit detailed comparison with the appraisal evaluation. B. Background 4.02 The project road links Bangui, the capital, with the Lobaye Province, one of the most economically.active provinces of the CAR. Logging is by far the most important economic activity and has been constantly expanding in recent years. The production of logs increased from 170,000 m3 in 1967 to 246,000 m3 in 1974, or by 5.5% annually. Most logs are transported on the project road to Bangui for further processing and export. To facilitate logging operations about 120 km of feeder roads were constructed in the vicinity of the project road under FED financing. A plan to establish a palm oil plantation (mentioned in the Appraisal Report) has not yet been realized, but FED foresees a program to finance a palm oil plantation covering about 2,000 hectares and - A.4 - producing 7,000 tons per year of palm-oil. C. Construction Costs 4.03 Total construction costs (net of taxes) including supervision, totalled US$4,770,000, of which US$3,334,000 in foreign costs and US$1,436,000 in local costs. The total CFAF1,050 million, was 5.3% lower than the appraisal estimate of CFAF 1,071 million, including contingencies (See Table 1). FAC provided an additional CFAF 36 million for supervision. D. Traffic 4.04 In 1968 at the time of the appraisal, the average daily traffic on the road was low - estimated at only 80 vpd of which 70% were trucks, buses and other heavy vehicles. The road was reconstructed to two-lane paved standard in 1972, but until December 1975 there were no traffic counts to provide a reliable trend of traffic development. According to the December 1975 count, there was an average of about 170 vpd. Consequently, it can be assumed that the traffic volume has been growing at an average of about 11% annually, slightly higher than the appraisal estimate of 10%. Results of the 1975 count show a change in the composition of traffic with an increased share of light vehicles, but which probably does not reflect the true composition of the year- round traffic. Since the count is not seasonally adjusted due to a lack of data, the average annual daily traffic most likely includes a larger proportion of heavy vehicles, and therefore the actual traffic composition on the road should not be far from the appraisal estimate. E. Benefits of the Project 4.05 The project produced quantifiable and non-quantifiable benefits. As in the Appraisal Report, the major quantifiable benefits considered in the analysis were savings in vehicle operating cost. In re-evaluation, savings in maintenance cost were also included, although they were excluded in the original analysis since they were considered to be the same before and after the paving of the road. As non-quantifiable benefits, the project has eased traffic flows particularly of big trucks carrying logs to Bangui and has provided better connection between the capital and the populated area in the road vicinity. In addition, numerous car owners have benefitted directly from reduced vehicle operating costs together with passengers who benefit from avoidance of transport fare increases. F. Economic Return 4.06 Based on 10% traffic growth during the first 10 years and 5% there- after, the economic return for paving the road was originally estimated at about 14%. Following the 1975 traffic count, the original traffic growth forecast appears to be somewhat on the low side for the second decade of economic life; therefore, in recomputing the economic return some adjustments were made. On the basis of updated vehicle operating costs, actual construction cost, new economic life 1973-92, and 10% traffic growth in the first six years and 7% growth thereafter, the recomputed economic return is 14%, as in the original estimate (Table 1). The re-evaluation proves therefore, that the reconstruction of the road was justified. Even if the benefits were reduced by 20%, the economic return would still be 11%. - A.5 - (Cr. 199-CA) HIGHWAY MAINTENANCE PROJECT 5. THE PROJECT A. General Description 5.01 The project consisted of: (i) a four-year maintenance program implemented with the assistance of consultants, including the following: (a) purchase of road maintenance equipment, spare parts and materials; (b) establishment of proper accounting and inventory techniques; and (c) training of local staff at all levels in programming, executing and supervising road maintenance works and operating and maintaining equipment; and (ii) feasibility study and detailed engineering of the improvement of the Bangui-Bossemb61 road (156 km). B. Cost Estimates 5.02 The estimated capital cost of the project was CFAF 1,506 million (US$5.42 million) of which US$4.3 million was lent by IDA for the foreign exchange costs. Of the two components of the project, the maintenance program (including technical assistance and contingencies) was estimated to cost CFAF 1,359 million (US$4.89 million); and the feasibility and detailed engineer- ing studies, CFAF 147 million (US$530,000). Recurrent costs for road mainte- nance were estimated at appraisal at about CFAF 330 million per year (plus equipment depreciation on average CFAF 115 million per year). Receipts from fuel taxes to be allocated to the Road Fund were expected to cover this amount. 6. PROJECT IMPLEMENTATION A. Phases of Implementation of Maintenance Program 6.01 Execution of the major component of this project, the four-year maintenance program, was unsuccessful. Difficulties arose from the start with the result that by the end of the second year, it had already become a problem project. One year elapsed between the signing of the Credit Agreement and the date it became effective. There were two conditions of effectiveness: The first was the employment of consultants; the second required establishment of the Road Fund. The former of these two conditions was met in March 1971 when the Government awarded contracts to consultants Lamarre-Valois (Canada) for technical assistance and to Louis Berger (USA) for the Bangui-Bossemb616 road study. However, the second of the two conditions was not met until a year later due to legal and administrative problems. - A.6 - 6.02 A continuous shortage of Government local funds made it impossible to implement the original plan, and it subsequently had to be scaled down on three separate occasions. Consequently, the technical assistance team (Lamarre- Valois) was reduced from 22 to 12, to 8, and finally to 2 team members. Local funds were not available to pay recurrent expenditures for fuel and materials which were needed to carry out maintenance operations outlined in the program and to make full use of the foreign experts and equipment. B. Supervision 6.03 In November 1971 a supervision mission reported that funds from the Road Fund were diverted to finance the operation of a regravelling unit for the improvement of streets in Bangui and other purposes not directly related to the project. The mission recommended Bank participation in semi-annual meetings of a Coordinating Committee composed of Government officials and technical assistance consultants. This recommendation was followed, and supervision missions were frequent. 6.04 In May 1972 a supervision mission reported that while the program was proceeding satisfactorily, the CFAF 632 million local component allocated in the 1972 budget had been reduced to CFAF 377 million, and even this amount was not actually available. Only CFAF 275 million was disbursed by the Government that year. Suppliers were reluctant, and subsequently refused to deliver goods until payments were made. As a result, the project almost came to a standstill at the end of 1972 because of the lack of fuel and spare parts. Repairs on existing equipment could not keep up with breakdowns and by late 1972 only slightly over half of the equipment was operational. Some of this equipment was then permanently assigned to the streets of Bangui. Virtually all the equipment procured under the Credit had been delivered by the end of 1972. However, at that time only 3,000 km of rough grading was executed as part of annual routine maintenance, instead of the 13,000 km scheduled for reshaping. The roads scheduled for regravelling were untended and consequently were deteriorating rapidly. 6.05 In view of all this, IDA asked the Government to send a delegation to Washington to review the situation and to agree on remedial measures. The meeting between the delegation, IDA management and staff took place in early June 1973. They agreed on the following: (i) Government committed itself to allocate CFAF 400 million for the road maintenance program in each of the years 1973 and 1974; (ii) IDA agreed to amend Credit 146-CA permitting the use of undisbursed balance (US$652,000) to supplement Credit 199-CA for the acquisition of spare parts, materials and fuel; and (iii) Consultants' technical assistance team would be reduced from 22 to 12 members. - A.7 - 6.06 It was clear by October 1973 that the Government was unable to provide even this reduced contribution to the project. Although Government opened an account for a revolving fund to pay outstanding bills (US$170,000 and US$530,000 equivalent for 1971 and 1972 respectively), no money was actually put into the fund. Of the CFAF 400 million Government had committed to make available for 1973, only CFAF 274 million was disbursed. In February 1974 it was reported that no improvement was expected. The relations between CAR and IDA became strained, leading to temporary suspension of supervision missions until later that year. 6.07 In mid-1974, subsequent to a supervision mission, once again, the Government and IDA agreed to the following measures: (i) Government would make available CFAF 140 million for the maintenance program for the last seven months of 1974, in addition to about CFAF 100 million already spent during the first five months of that year; (ii) consultants' team would be further reduced to 8 members by mid-1974 and to 2 by the end of that year; the assignment would be terminated by the end of January 1975; and (iii) the physical program for 1974 would be further reduced, (see Table 5). 6.08 In August 1974 the consultants reported that a considerable part of the road maintenance equipment had again been diverted to other ministries and organizations. As a result maintenance operations in the Bangui division came to a complete halt and were again very much reduced in the other subdivisions. Later the situation was reported to have improved somewhat. The truncated maintenance program was terminated in January 1975 with the departure of the remaining technical assistance team. C. Pre-Investment Studies Feasibility and Detailed Engineering Studies 6.09 The studies for the Bangui-Bossemb616 road had no major problems. The quality of the work of the consultants Louis Berger (USA) was satisfactory. The feasibility study began later than scheduled because of the delay in Credit effectiveness. An additional five months' delay was necessary when it became clear that forecasted traffic would not justify the original standards and costs. The consultants then made a more thorough study of a low-cost solution. The feasibility study was completed in June 1972 but decision to proceed with the detailed engineering was withheld until October 1974 because of doubtful economic justification of the project. The engineering was finally completed in October 1975. D. Consultant's Performance 6.10 The technical assistance consultants acccmplished certain improvements from a managerial and administrative point of view, although the program was admittedly a difficult one. They established a sound organizational structure, - A.8 - introduced modern accounting and inventory techniques into the Road Maintenance Division and prepared manuals for road maintenance operations and maintenance and repair of equipment. If their recommendations are implemented in the future, their contribution will not have been lost. Workshop improvements were below expectations. Training of lower-level staff was not fully accomplished because of the shortage of operational equipment. Additional training of higher level staff was proposed but not carried out because of shortage of funds. 6.11 The consultants maintained good relations with the Government officials even though their task of technical assistance was frequently frustrating. They had spontaneously suggested withdrawing from the project when it became clear that they could no longer fulfill their objectives efficently as the project was continually hampered. 6.12 The consultants left the CAR at the end of January 1975 and the project was declared completed. Their findings are incorporated in their Final Report, dated August 1975. E. Special Covenants 6.13 In addition to the standard Covenants, the Government also agreed to establish a Road Fund and allocate to it revenues from fuel taxes, to cover road maintenance requirements. The Government, however, was not in a position to fully comply with this commitment. Amounts allocated to the Road Fund were greatly inadequate and considerably less than the revenue collected from fuel taxes during the four-year period of the program. Funds from the Road Fund were also used for purposes other than those directly related to the project. F. Disbursements 6.14 Table 6 reflects the delay in making the Credit effective (para. 6.01). Total disbursements reflect the increase of US$652,000 from the undisbursed balance of funds from the project Cr. 146-CA to the funds available under the project Cr. 199-CA (para. 6.05). 7. ECONOMIC RE-EVALUATION 7.01 The objective of the proposed project was to ensure that economic development would not be hindered by the further deterioration of the primary and secondary road systems, thus leading to rising transport costs. Assuming that the project would have been successful, the savings in vehicle operating cost were calculated to yield economic returns as follow: If Annual Traffic Economic Rate --Growth Rate Is: of Return Will Be: 8% then 28% 5% 15% 4% 9% - A.9 - 7.02 As reported earlier in this PCR, there were various reasons for the unsuccessful execution of the project (mainly lack of Government funds and some misuse of the maintenance equipment). Of the output envisaged in the project, only about 10-20% was actually carried out during implementation of the maintenance program (1971-1974). With the available data on road traffic development it could be estimated that the average annual traffic growth on primary roads was about 5% in 1971-1975. 7.03 By comparing (a) annual traffic growth of 5%, annual benefits of about 25% of the appraisal estimate (assuming vehicle operating costs were correctly estimated), and assuming some residual value of the remaining equipment; with (b) actual capital investments (in line with appraisal estimates) and recurrent maintenance expenditures (lower than ori- ginally estimated), it can be concluded that the estimated rate of return at project completion is actually substantially lower than the 15% estimated at appraisal given an annual traffic growth rate of 5%. However, it should be mentioned that returns in short economic life projects are very sensitive to input changes. Based on above educated guess assumptions and by applying the appraisal report methodology (six-year economic life) a calculation was made showing that an indicative economic rate of return would be negative, about -20% (see attached table). Details are in Table 7. 7.04 Since the project has not been successful, road conditions have further deteriorated. How much more they would have deteriorated without the project cannot be determined, and thus trying to establish how much benefits can still be allocated to the largely unsuccessful investments and consultants' efforts would be a futile exercise. Part of the benefits may still be obtained later, if and when the framework established under the so far unsuccessful program is revitalized under a subsequent operation, with financing by an international or bilateral organization, such as FED or FAC, as presently envisaged by the Government. 7.05 As to the feasibility and detailed engineering studies for the Bangui-Bossembele road, the internal rate of return on the then estimated investment for improvement of this road to two-lane paved standard was 12%. It appears that the construction of this road is justified at this time since the traffic increased and the economic return is estimated now at 15%. Estimated costs have risen considerably, but so have vehicle operating costs and thus, most likely the savings therein, resulting from this proposed road improvement project. 8. BORROWER'S PERFORMANCE 8.01 The Borrower's Directorate of Planning and Construction (part of DPW) was responsible for project execution under Credit 146-CA and performed the work satisfactorily. - A.10 - 8.02 The Borrower's performance under Credit 199-CA did not, however, meet expectations. The Government was unable to comply with an important commitment under the Credit Agreement, in that it failed to allocate necessary funds for highway maintenance, and both Road Fund money and project equipment were used for other purposes. (See Annex A). 8.03 However, it should be realized that before the project was appraised and the Credit Agreement signed, the Government had spent much less on road maintenance than the amount of revenue collected from fuel taxes. The Govern- ment felt unable to increase the allocations for road maintenance at a time when it faced a huge, overall budget deficit. As to diverting Road Fund revenues and equipment to other works not directly related to the project, this change in work priorities is less understandable. 8.04 It should also be taken into consideration that the program was too ambitious. The Government appeared unable to provide not only the financial means needed for its execution but also to make counterparts and staff available to benefit from the expertise and training provided by the relatively large team of 22 foreign experts. 9. THE ROLE OF IDA 9.01 Project supervision was satisfactory, although for some time its effectiveness was hampered by difficulties arising in relations with the Government. Considering availability of staff, supervision missions were frequent, and reporting adequate until mid-1974. After the four-year mainte- nance program was completed and the consultants left the CAR, IDA's emphasis shifted to the preparation of the next highway project, the construction of the Bangui-Bossemb&16 road which was appraised in March 1976. 9.02 It may also be said that when the Government's financial difficulties became evident, IDA adopted an understanding and flexible position. In order to accommodate the Borrower's problems it agreed to scale down the program three times. It also agreed to the use of undisbursed funds of Credit 146-CA to be disbursed for the project under Credit 199-CA. Although there were sufficient reasons for IDA to cancel the credit, little purpose would have been served by such a measure. Since the new equipment had already been delivered and paid for, premature termination of the project would have only further reduced the benefits from these investments. 9.03 As to the preparation and appraisal of the project, it could be argued that IDA should have recognized at that time that the program was too ambitious. In retrospect it appears that it could possibly have been advisable to first start a pilot project for a few subdivisions closest to Bangui, and to expand the program in a subsequent credit operation. However, at the time of appraisal some doubt already existed that the availability of local funds could create problems. - A.11 - 10. CONCLUSIONS 10.01 A straight-forward road reconstruction project, such as the First Highway Project requiring minimum involvement of the Government, succeeded. However, the Highway Maintenance Project was much more demanding in prepara- tion and implementation. Formulation of the project proved to be overly optimistic and the Government's misconception of its budgetary responsibility towards the project, contributed to the lack of success. In retrospect, the project appears to have been beyond the Government's ability to absorb a program of such size. 10.02 Benefits to the country from road maintenance projects in terms of economic return, institution building and training if properly formulated are clearly high enough to make it worthwhile to take exceptional risks in project implementation. However, as far as any future road maintenance projects in the CAR are concerned, external financing of recurrent costs at an initial stage would substantially reduce risks and eliminate the main cause of past failure. 10.03 Based on the experiences of the First Highway and the Highway Maintenance Projects, a Third Highway Project is now being considered. The main objective of the proposed project is to improve the Bangui-Bossembele road. The Association has made financing of this project contingent upon Government's arranging for a satisfactory road maintenance program, supported by foreign assistance. A more detailed commentary on the Highway Maintenance Project is attached as Annex A.  ANNEX A CENTRAL AFRICAN EMPIRE (REPUBLIC) CREDIT 199-CA HIGHWAY MAINTENANCE PROJECT PROJECT COMPLETION REPORT GENERAL COMMENTS AND OBSERVATIONS Introduction 1. The major component of this project, a four-year (1971- 1974) road maintenance program to be implemented with the tech- nical assistance of consultants, has not been successfully com- pleted. This unfortunate event is largely due to the Government's failure or inability to provide, "promptly as needed," the funds required for recurrent local expenditures, not covered by the IDA Credit. This failure constituted a violation of sections 4.01 and 4.03 of the Credit Agreement. Table 4 shows the amounts estimated in the Appraisal Report, the amounts as budgeted, and promised by the Government after IDA had agreed to scale down the project,as well as the actual expenditures. Because of a back- log of unpaid bills, suppliers of fuel, spare parts and materials to be locally purchased eventually refused to deliver needed supplies. The project would have come to a complete standstill if IDA had not agreed to use US$652,000 of undisbursed funds from Credit 146-CA for payment of local expenditures (largely out- standing bills) needed for the maintenance project. (President's Report to the Board dated July 20, 1973). 2. Another violation of the Credit Agreement occurred when both funds and equipment, intended to be used for the maintenance program for the country's main primary and secondary roads, were diverted to other purposes, not directly related to the project. Some other clauses of the particular covenants were also violated, but these events should be considered as resulting from the pri- mary violation; the shortage of funds. 3. The unfavorable and unfortunate experience with this pro- ject raises the question whether the four-year road maintenance program prepared by consultants, and adjusted at the time of appraisal, was really sound from a technical, administrative and financial point of view. Since the feasibility and engineer- ing studies for the future construction of the Bangui-Bossemb616 road have not created any major problems, the following observa- tions and comments are directed only to the major component of the project's four-year maintenance program. The Composition of the Program 4. In many developing countries road standards do not meet present day requirements. To the extent that they were built a long time ago, standards were selected to accommodate lower axle loads prevailing at that time, when it also could be assumed that the roads would be reasonably well maintained. Currently, however, sufficient funds for proper maintenance are no longer available. ANNEX A - page 2 As to building new roads, the emphasis is on quantity rather than quality because of the shortage of funds. As a result, Governments of many developing countries, such as the C.A.R., became responsible for the maintenance of an extensive under- designed or non-engineered road network, which was incapable of accommodating growing traffic volumes, carrying trucks with axle loads for which the roads were never built to begin with. 5. This problem was particularly acute in the C.A.R. where a small population (some 1.6 million at the time of appraisal) is spread over an area about twice the size of Italy. Although traffic volumes are low, the road network is large ; some 22,000 km, resulting in a high density of about 14 km per 1,000 inhabitants compared with e.g. 9 km in Chad and 4 km in Cameroon. Because of low traffic volumes, economics call for cheap construction standards, with earth,.laterite or gravel surfacing which requires good, frequent routine and periodic maintenance. Such maintenance is hardly possible, however, if basic engineering requirements, such as proper drainage, properly compacted subgrades and adequate thickness of pavements do not exist. Therefore, before sound maintenance procedures for such below standard roads can be intro- duced, both a certain amount of catching up with deferred main- tenance and improving and upgrading non-engineered roads is needed. 6. For such programs, a part thereof should be considered as capital investment, certainly for the portion identified as "betterment" and for practical reasons, in poor countries such as the C.A.R., also that part of the program which may be considered as "deferred maintenance." This aspect was recongized by the Belgian consultants Research and Development. In their studies, financed by UNDP for which the Bank was the Executing Agency and upon which the IDA project was based, they recommended that both investment for equipment plus an amount for what may be considered as rehabilitation and local improvements should be financed under a 20-year foreign loan. Financing of the Maintenance Program 7. The project as conceived for IDA financing did not make the above distinction between what should be considered as "recurrent expenditures" on the one hand and both foreign and local expenditures which could be considered as "capital expenditures" on the other hand. Even though the IDA Credit finances equipment and consultants' services which have an economic life reaching well beyond the four years of the program, this does not help to solve the problems that still may arise from the point of view of the actual available cash for recurrent expendi- tures during the initial four-year program, in particular the first year(s). In the case of Bank financing of an industrial project, calculation of the cash flow during and after construction of such a project normally forms part of the overall financial analysis. This practice has never been followed in the case of road projects, when it is assumed that Governments "at all times, promptly as needed" will make the local funds available. It should be recognized, however, that in the case of IDA financing for road ANNEX A - page 3 construction projects, the financial burden on the Government budgets can be minimal since IDA will from time to time finance local costs, excluding taxes. In the case of maintenance programs, however, the situation is very much different. 8. The financial burden on the Government's current budget is much larger for maintenance programs since all local labor and locally available materials and consumable supplies have to be paid for from the recurrent expenditures budget. Unless part of such expenditures can be paid from the country's capital budget, (for reasons set forth in para. 6) a cash flow problem may well arise. To take one example, import duties on all new equipment and spare parts will have to be paid in the first year when equip- ment arrives, and also other recurrent local expenditures such as cost of workshops, spares and wages for mechanics for the overhaul of existing equipment, etc. The fact that IDA also finances the value of equipment and of consultants' services still remaining after the completion of a four-year program, does not solve the "cash flow" problem during the first few years. 9. In view of the above, in the case of poor countries, IDA may well give serious consideration to also financing that part of what is called a "maintenance program" which actually can be identified as "betterment" or catching up with "deferred main- tenance." Such a reasonable approach would greatly relieve the financial burden on the Governmntt. It is easy for IDA to include in the Credit Agreement covenants which make the Government re- sponsible to provide at all times all funds needed for the project. What is much more difficult, and possibly sometimes impossible for a Government is to comply with such clauses which instead would give top priority to the IDA project, regardless of avail- able financial resources, commitments and obligations the Govern- ment may have in other fields. One could argue that a Government should not sign a Credit Agreement with such stringent covenants. On the other hand, Governments and IDA know that the practical implications are, and have to be, open for review. This practical approach was shown by IDA also in this case, but too late, when in the face of reality the program was scaled down three times, and surplus funds of Credit 146-CA were used for Credit 199-CA. 10. It is true that at the time of appraisal, estimated revenue from fuel taxes to be put in C A.R.'s re-established road fund would, as an average, have been sufficient to cover the local expenditure for the program, But there would have been a cash flow problem in the first year(s) as mentioned above. It is also true that import duties on new equipment and spares, to be paid by the Equipment Division, were collected by the Government and could be made available to the project. But for such arrangement special budgetary procedures may be needed, since the Equipment Division in principle has its own budget. Finally, for the Government to make an initial budgetary allocation to the road fund of US$1.9 million equivalent (CFAF 530 million or one year fuel tax revenues), ANNEX A - page 4 in accordance with section 4.04 of the Credit Agreement, may have been too heavy a burden on the budget. Most important, if it had been Government's practice in the past to spend less on road maintenance than was collected from fuel taxes, then the consequences of changing such a practice may be serious, particu- larly in bad times caused by drought, recession, inflation of cost of imports (fuel), etc. For instance, for 1972 IMF estimated an overall budget deficit of CFAF 2.8 billion (US$10 million equiva- lent), and in 1973 the situation became even worse. Therefore, in the case of poor countries an analysis of the Government's overall fiscal situation and capability is called for at the time of appraisal. Technical Assistance by Consultants 11. The team of 22 foreign experts was considerably larger that recommended by the consultants Research Development who recommended in their UNDP study a foreign team of 17 in the initial year 1971, gradually to be reduced to 10, 5 and 4 in 1972, 1973 and 1974 respectively, or 36 man-years in total. This compares with 36 man-years for operational experts plus 26 man-years for training, agreed between Government and IDA at the time of apprais- al. The initial recommendation of 36 man-years may well have been insufficient. But the envisaged 62 man-years later decided upon would appear too large and could not be efficiently absorbed Ly the limited and inexperienced staff of C.A.R.'s road department. In the case of road construction projects, the managerial, admini- strative responsibilities for their execution is largely with consultants and contractors, and with IDA for their financing. In the case of maintenance programs which are to be carried out with the road department's own staff and equipment requiring a much larger share of Government finances (para. 8), the situation is completely different. In the case of this C.A.R. project, it became quickly apparent that the expertise and training provided by 22 experts could not be efficiently absorbed and used. Under C.A.R. conditions, it would have been more practical and realistic to start the technical assistance program with a pilot project for only one of the two maintenance divisions, or an even smaller one for only two or three of the seven sub-divisions. After being trained under the pilot project, local staff could then later be transferred to other sub-divisions. 12. In addition, if the suggestion to make a distinction be- tween routine maintenance on the one hand and rehabilitation and/ or upgrading on the other hand is followed, then consideration may be given to having all or part of the latter works carried out by contractors under direct supervision of consultants. Such a procedure would have several advantages: (i) the technical and administratvve burden on the road department would be considerably relieved; (ii) the contractors would be responsible for providing their own equipment, rather than loading up the road department with too ruch equipment it cannot digest, maintain and operate; and (iii) IDA could finance smaller lots of equipment plus the cost (net of taxes) of contractors and supervision, thus relieving the financial burden on the Government's budget. Table 1 CENTRAL AFRICAN EMPIRE (REPUBLIC) CREDIT 146-CA PROJECT COMPLETION REPORT Costs and Basic Data for Calculation of Economic Return, First Highway Project I. Total Construction Cost (net of taxes) Appraisal Estimate Actual Million CFAF 1,071.0 Million CFAF 1,050.0 II. Road Traffic (v-pd) 1968V/ 1975 ii Light vehicles 24 (30%) 110 (64%) Heavy vehicles 56 (70%) 60 (36%) Total 80 170 III. Vehicle Operating CostsL/ (CFAF/km net of taxes) Earth Road Paved Road Appraisal (1968) Actual (1972) Appraisal (1968) Actual (1972) Light vehicles 19.3 26.5 12.5 19.6 Trucks (7t) 70.1 80.2 37.1 55.8 1/ All data in CFAF are in 1972 prices. (US$1.00 = CFAF 247) 2/ Total traffic volume based on counts; traffic composition was estimated by the appraisal mission. 3/ Traffic counts were carried out by PWD in December 1975; the results are not seasonally adjusted. 4/ VOC from the Appraisal Report were updated and adjusted on the basis of Louis Berger's feasibility study of Bangui-Bossemb6l6 Road, 1972. Table 2 CENTRAL AFRICAN EMPIRE (REPUBLIC) CREDIT 146-CA PROJECT COMPLETION REPORT Schedule of Accumulated Disbursement8, First Highway Project (in US$'000) IDA Fiscal Year Appraisal Actual Total and Semester Estimate!/ Disbursements 1969 1st - 2nd 800 152 1970 ist - 220 2nd 2,000 1,615 1971 ist - 2,131 2nd 3,500 2,757 1972 1st - 3,357 2nd 4,200 3,5482/ Closing Date: 3/31/73 12/31/75 I/ Appraisal report did not contain semester figures. 2/ Balance of $652,000 disbursed under Cr. 199-CA as of February 18, 1976. March 1976 CENTRAL AFRICAN EMPIRE (REPUBLIC) CREDIT 199-CA PROJECT COMPLETION REPORT Highway Maintenance Project Statue and Cost Estimates as of May 31, 1976 Total Cost Estimates (US$'000) (including contingencies) Completion Date ... Appraisal Original Contract Present IDA Executed Actual or (US$1 = CFAF 278) (US$1 = CFAF 250) (US$1=CFAF2 0) Cont Project Items by lontract Forecast IDA Govt. Total IDA Govt Total IDA Govt Total % 1. Procure- Various Feb. Feb. 1973 2,130 550 2,680 n.a. n.a. 2,015 2,931 112 3,043 96% ment of equip- suppliers 1973 ment, spare parts,,Tate- rials 2. Consul- Lamarre- Jan. Jan. 31, 1,750 460 2,210 1,694 420 2,114 1,688 488 2,176 78% tants ser- Valois 1975 1975 vices 3. Feasibility Berger Feb. June 1972 240 70 310 184 104 288 241 104 345 70% study and 1972 preliminary engineering 4. Detailed Berger Sept. Nov. 1975 160 60 220 210 Nil 210 184 Nil 184 100 engineering 1975 TOTAL 4,280 1,140 5,420 4,627 5,044 704 5,T48 1/According to the amendment of July 27, 1973 to Credit 146-CA, its balance of US$652,000 was being used for procurement of spare parts and fuel to support this project. 0M 1W. Table 4 CENTRAL AFRICAN EMPIRE (REPUBLIC) CREDIT 199-CA PROJECT COMPLETION REPORT Costs in Local Funds, Highway Maintenance Project (in CFAF million) Required Costs 1/ kppraisal 2/ Initial Budget Revised Budget Actual Estimate before 1971 ---- 200 1971 559 460 n.a. 1972 522 632 377 275 1973 504 590 400./ 216 1974 495 4003/ 245 -/ 245 Total 2,080 2,082 1,022 936 1/ Funds to cover the project's local component such as import duties and taxes on equipment and material, technical assistance, recurrent costs of maintenance, preinvestment studies. 2/ Including amounts of CFAF 60, 120, 140 and 145 resDectivelv, be set aside for equipment renewal. 3/ As revised and agreed with IDA in mid-1973. 4/ As again revised and agreed with IDA in June 1974. CENTRAL AFRICAN EMPIRE (REPUBLIC) CREDIT 199-CA PROJECT COMPLETION REPORT Highway Maintenance Field Operations Revised Initial Pro4ram Annual Program Actual Prograq Approved Pr ram for 1973DJ for 1973 + 1974Vj for 1973 Al for 1974-' (in Km) Backlog regravelling 305 130 90 70 Periodic regravelling 425 150 15 115 Reshaping 11,700 8,800 5,000 5,800 Clearing 26,000 23,000 12,000 17,000 38,430 32,080 17,105 22,985 1/ Based on a Government contribution of CFAF 715 million proposed by consultants, of which only CFAF 590 million was initially budgeted to cover local component of Credit 199-CA. 2/ Based on an annual Government contribution of CFAF 400 million. 3/ Carried out with a Government contribution of CFAF 216 million plus US$350,000 from Credit 146-CA for spare parts and fuel. Based on a Government contribution of CFAF 240 million plus US$300,000 from Credit 146-CA for fuel and US$250,000 from Credit 199-CA for spare parts and materials. Table 6 CENTRAL AFRICAN EMPIRE (REPUBLIC) CREDIT 199-CA PROJECT COMPLETION REPORT Schedule of Accumulated Disbursements, Highway Maintenance Project (in US$'000) Appraisal IDA Fiscal Estimate Actual Year and Original Disbursements Semester Credit from Credit 146-CA from Credit 199-CA 1970 2nd 600 1971 1st 256 2nd 3,500 1,178 1972 1st 2,356 2nd 4,000 2,986 1973 1st 47 3,508 2nd 308 3,707 1974 1st 421 3,993 2nd 4,300 6b.9 4,238 1975 1st 652 4,238 2nd 1976 1st - 4,300 Closing Date 6/30/75 12/31/75 12/31/75 Table 7 CENTRAL AFRICAN EMPIRE (REPUBLIC) CREDIT 199-CA PROJECT COMPLETION REPORT Recomputation of Economic Return, Highway Maintenance Project1/ (in CFAF million) 2/ Incremental 3/ Year Capital Cost- Recurrent Cost Total Costs Total Benefits- 1971 (717) 350 ( 23) - (740) 350 (116) 29 1972 (313) 520 ( 43) 10 (356) 530 (244) 61 1973 (119) 210 ( 63) 20 (158) 230 (387) 97 1974 ( 56) 120 ( 73) 10 (118) 130 (406) 101 1975 0 0 ( 94) 20 ( 94) 20 (427) 107 1976 0 0 (116) 20 (116) 20 (447) 112 + 1304 1/ Figures in parenthesis are original estimates from the Appraisal Report. 2/ Includes cost of equipment, spare parts and of technical assistance. 3/ Traffic growth 5% per annum; about 25% of the estimated benefits in the appraisal were taken into account. 4/ Estimated residual value of road maintenance equipment at the end of 1976 (about 20-25% of its original value). Basic Economic Return: about -20%  IBRD 12228R N I G E R 2JANUAY 1977 Kano - Moldug~ri ~ ~ N'DJAMENA 2K- 5 U D A N BENIN G E R I A Er.H . Minn. N-Gondéré angaf \? Crapel LA -... . . - -. B g.. CENTRAL AFR RCAN EMPIRE CWGHT CIF HE~N..N Sbu:¢ma Pra uc A M E R O N B.ssembele D BANGUl BJGH OF QNNY MALABE Yokado_ PrmararNds.CAE z A I R E EQUATORIAL GUINEA Be Ouess UIBREVILLE B-å C 0 N G 0 CENTRAL. AFRICAN EMPIREo. G A O N THIRD HIGHWAY PROJECT h-,IN ih"m d d e oh, eoa Project road orldankadusafdwe 6 ond'Primary roads, C.A.E. . Nd....... Trans-African Highway Railways Rivers A F R I C A e National capirals --International boundaries BRAZZAVILLE A KINSHASA oi o 10 200 300 400 MILES o0 10 200 300 400 s00 600 KILOMETERS i\ A N G OLA 21  IBRD 12229R JANUARY 17 CENTRAL AFRICAN EMPIRE CNTRIL AFFICAN- THIRD HIGHWAY PROJECT A F R l C A ROAD NETWORK \D BIRAot 10 T. Sari -nd N'DjmeiI To N'DjTmeLa OUADDAROA -CARFPAOUA CRAMPEL - BOSSANGOA tBOUCA IA A INGA BOZOUMDEKOA BACALA Ipp BOLILA1OSSEMTELE •BABOUA SIBUT GRIMARI AMBARI AKOUMIA 0h SE2 ,E RI E LAE, k. tL II P Ö 0 5 5 1 5 0 7 R I 0 5 5 7 0 K RNTDAMAR OUANGO ALINDAO RAFAI E BANGASSOUI t BERB RATl BODA BAN l AYE AN SPI OANGORc ý --PROJECT ROAD † ýPAVED PRIMARY ROADS WA\.OMA O A OTHER PRIMARY ROADS \ NG NIB SECONDARY ROADS '\NOL LOCAL ORI FEFEDE R ROADS *.... CONSTRUCTED UNDER CREDIT 146-CA =NENTONAL AIRPORT -- SCNAY AIRFIELDS ýRIVERS / * .-- .. -INTERNATIONAIL BOUNDARIES ToBrzowe.- / d. ryn m- ac p o,, 15 0 ,_25 50. . 5 10nom]. en 25 50 79 100OKM. PointeNoire~-p. Wobynknrsffhts