India - Gujarat Rural Roads Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 16105 IMPLEMENTATION COMPLETION REPORT INDIA GUJARAT RURAL ROADS PROJECT (Credit 1757-IN) November 8, 1996 Agriculture and Water Operations Division Country Department II South Asia Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Currency Unit = Indian Rupee (Rs) At Appraisal (1986) US$1 = Rsl3 At Completion (1995) US$1 = Rs33 WEIGHTS AND MEASURES 1 meter (m) = 3.281 feet 1 kilometer (km) = 0.621 mile 1 square kilometer (km2)= 0.386 square mile 1 ton-km = 0.621 ton-mile FISCAL YEAR OF BORROWER April 1 - March 31 ACRONYMS AND ABBREVIATIONS CRRI = Central Road Research Institute GOI = Government of India GRRP = Gujarat Rural Roads Project (Credit 1 757-IN) ICR = Implementation Completion Report IDA = International Development Association MPS = Maintenance Planning System NDDB = National Dairy Development Board NDO = New Delhi Office (of the World Bank) NPV = Net Present Value PPMC = Project Planning and Monitoring Cell (R&BD) R&BD = Roads and Buildings Department (of GOG) SAR = Staff Appraisal Report SDR = Special Drawing Rights UK = United Kingdom of Great Britain and Northern Ireland VOC = Vehicle Operating Cost Savings WBM = Waterbound Macadam Vice-President D. Joseph Wood Director : H. Vergin Division Chief/Manager S. Barghouti L_Staff Member F. Galli, Financial Analyst  FOR OFFICIAL USE ONLY IMPLEMENTATION COMPLETION REPORT INDIA GUJARAT RURAL ROADS PROJECT (Credit 1757-IN) Table of Contents PREFACE ...... .... ..... .... ..... .... ..... .... ..... .... ..... .. .. EVALUATION SUMMARY ..................................................... ii PART I: PROJECT IMPLEMENTATION ASSESSMENT ....................................................1 A. STATEMENT/EVALUATION OF OBJECTIVES ......................................................I Objectives ......................................................I B. ACHIEVEMENT OF OBJECTIVES .....................................................1I C. MAJOR FACTORS AFFECTING THE PROJECT ......................................................4 D. PROJECT SUSTAINABILITY ......................................................5 E. PERFORMANCE OF IDA .....................................................5 F. BORROWER PERFORMANCE ......................................................6 G. ASSESSMENT OF OUTCOME ......................................................6 H. FUTURE OPERATIONS ......................................................6 I. KEY LESSONS ......................................................7 PART II: STATISTICAL TABLES .....................................................9 Table 1: Summary of Assessments ......................................................9 Table 2: Related Bank Loans/Credits ...................................................... . 11 Table 3: Project Timetable ...................................................... 12 Table 4: Credit Disbursements: Cumulative Estimated and Actual ...................................................... 12 Table 5: Key Indicators for Project Implementation ...................................................... 13 Table 6: Key Indicators for Project Operation ...................................................... 14 Table 7: Studies Included in Project ...................................................... 14 Table 8A: Project Costs ...................................................... 15 Table 8B: Project Financing ...................................................... 15 Table 9: Economic Costs and Benefits ...................................................... 15 Table 10: Status of Legal Covenants ...................................................... 16 Table 11: Compliance with Operational Manual Statements ...................................................... 16 Table 12: Bank Resources: Staff Inputs ...................................................... 17 Table 13: Bank Resources: Missions ...................................................... 18 APPENDIX A .......................................................19 APPENDIX B .......................................................28 APPENDIX C .......................................................37 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 wiLhout World Bank authorization.  I  i IMPLEMENTATION COMPLETION REPORT INDIA GUJARAT RURAL ROADS PROJECT (CR. 1757-IN) PREFACE This is the Implementation Completion Report (ICR) for the Gujarat Rural Roads Project (Credit 1757-IN) in India, for which Credit 1757-IN in the amount of SDR 101 million (equivalent to US$119.6 million) was approved on February 17, 1987 and made effective on August 27, 1987. The credit was closed on December 31, 1995, one year later than the original closing date. The final disbursement took place on May 21, 1996. The total disbursed under the credit was SDR 71,226,087 and the credit balance of SDR 10,473,913 was cancelled on May 31, 1996. This ICR was prepared by a mission" from the FAO/World Bank Cooperative Programme, which visited India from March 15 to 28, 1996. This ICR is based on information from the project files in Washington D.C. and New Delhi. The implementing agency - Gujarat Roads and Building Department - contributed to the preparation of the ICR by providing information to the ICR mission in the field. The Borrower has submitted its comments on the draft ICR and has prepared its contribution to the ICR (see Appendix B). R.G. Paterson (Civil Engineer, Mission Leader), W. J. Sorrenson (Economist), N. Olsen (Project Analyst) and C.S. Nawathe (World Bank Consultant Engineer).   ii IMPLEMENTATION COMPLETION REPORT INDIA GUJARAT RURAL ROADS PROJECT (Cr. 1757-IN) EVALUATION SUMMARY Introducfton 1. The World Bank's lending to the transport sector in India started in 1949 with the first Loan to Indian Railways (IR). By the time the Gujarat Rural Roads Project (GRRP) was appraised, over 20 loans/credits including one for rural roads (Bihar Rural Roads Project) had been made to the sector. Furthermore, in the agriculture sector several irrigation projects have included the construction/rehabilitation of over 9,000 km of rural roads in various states. The GRRP was in response to the Government of India's (GOI) objective of improving all weather village connective in a key agricultural state like Gujarat. Incidentally, the Gujarat Irrigation II Project which was being implemented at the time had a large rural roads components and suffered initially from poor construction quality due to outdated construction techniques, materials and payment delays. At the beginning, the GOG had also problems in planning and managing the implementation of the project, and this problem was corrected with the formation of a dedicated Project Planning and Monitoring Cell together with the use of a Gujarat-based consultant firm, which developed and introduced computerised project planning techniques. Some of the lessons learned in the implementation of the Gujarat Irrigation II were put to good use in the preparation and implementation of the GRRP. Project Objectives 2. The project development objectives were: (1) provide all weather roads in rural areas; (2) improve rural road construction and maintenance, road planning and road maintenance management; (3) development of a higher level of road safety and; (4) training of local staff in the administration and operation of the road network. At appraisal the project components included: (i) construction, reconstruction and improvement of about 4,040 km of village roads connecting 1,900 villages; (ii) procurement of equipment and vehicles; (iii) construction of houses, offices, warehouses and workshops; (iv) engineering and consulting services for project planning, design, supervision, monitoring and staff training; (v) technical assistance for a road maintenance study; and (vi) preparation of a road safety program.  Implementation Experience and Results 3. Project implementation started off slowly; after two and a half years, the project was almost two years behind schedule2, and was suffering from inadequate funding and poor management (see Part I, para 3). Starting from 1990, implementation progress started to improve due to increased commitment by GOG, appointment of an Officer on Special Duty to take charge of the project, and the lending of funds by the National Dairy Development Board (NDDB). In 1991, GOG requested an increase in the physical scope of the project and IDA after monitoring implementation performance for over a year agreed in 1993 to increase the target length of roads to be improved to about 6,300 km. When ongoing contracts are completed (mid to late 1996) some 6,250 km of all- weather roads will have been completed, 50% more than the original SAR estimate of approximately 4,040 km. 4. As a result of the project, rural road construction techniques have improved through the adoption of more modem contracting methods by the R&BD. At present, even for the construction of rural roads the use of force account has been discontinued. Overall, the construction quality of the project roads has been satisfactory with some having being built to a very high standard. However, on a limited number of contracts, construction quality problems such as inadequate compaction of earthworks, poorly compacted concrete, lack of curing of concrete and cracked or damaged culvert pipes were reported by Bank supervision missions. 5. The road maintenance planning study which includes a computerised maintenance planning system (MPS) was completed substantially behind schedule. Until maintenance funding is increased for all of the road network and the output of the MPS is used in a systematic manner, it is unlikely that the condition of parts of the Gujarat road network will improve. However, it must be stressed that Gujarat even with the current funding constraints is doing a reasonably good job at maintaining its road network. Although it was listed as an objective of the project, there was no specific provision under the project to fund the preparation of a road safety program. Improved traffic segregation, signalling and other road safety improvements introduced by the R&BD during the life of the project were not financed by the project. 6. The total project cost to completion is now estimated to be Rs 3,267 million (147% of the SAR Rs estimate) equivalent to US$126.6 million (74% of SAR estimate). Following a mid term review of the project in February 1993, SDR 19.3 million (US$22.85 million) was cancelled from the credit due to project savings caused by the depreciation of the Rs vis-a-vis the US$, and a final undisbursed balance of SDR 10.5 million was cancelled on May 31, 1996. 7. The ERR for the project calculated in the SAR was estimated at 28%, and was based on the economic analysis of 260 individual roads, representing 30% of the total road length (and about 25% of the total estimated project cost). This analysis was based primarily on benefits arising to 2 At the end of FY90, the actual disbursements from the IDA Credit aggregated to just about 30% of the cummulative disbursement for the period until the end of the FY estimated at appraisal (see Part IlI/Table 4).  iv the dairy sector (increased milk procurement, reduced wastage and a reduction in milk animal deaths) and from savings on vehicle operating costs (VOC), but excluded benefits from savings in dairy transport costs and increased agricultural production attributable to the improved project roads. The analysis also assumed that there would be no incremental milk production without the project, and that all incremental milk production was attributable to project roads, although other factors such as investments in more milk chilling centres and collection centres, improved dairy farm management methods and genetic improvement of milk animals would also contribute to increased milk production. 8. The ERR has been re-estimated at project completion correcting for the above deficiencies and has included all project roads. The re-estimated ERR of 16% (see Appendix C for details) though not comparable methodologically to the 28% ERR estimated in the SAR, reconfirms that the project-supported rural roads program is economically viable. 9. Because the economic analysis at appraisal covered only 30% of the total project roads, the SAR expected that the economic returns of the remaining roads would be calculated after credit approval to ensure that only economically viable road proposals would be accepted for financing under the project. In actual fact, the relevant analyses were often undertaken by engineering staff poorly trained in economic analysis and in some cases no economic evaluation was conducted of the rural roads selected for improvement. This has lead to the inclusion in the project of a small percentage of project roads with low rates of return that were included for improvement on the basis of political and connectivity criteria. 10. The performance of IDA during both the project preparation and implementation phases was satisfactory, except that IDA allowed Credit Negotiations to be delayed for over two years after appraisal (see Part II/Table 3) to accommodate the borrower's wish to have the project fully funded by an IDA Credit instead of an IBRD Loan. During project implementation, IDA was responsive to the borrowers needs. IDA agreed to extend the physical scope of the project, increased disbursement percentages and postponed the credit closing date by one year to allow the completion of several hundred km's of rural roads. The R&BD considers that IDA supervision missions gave valuable guidance and support. 11. Delays by GOI in passing on project funds to GOG, slowed down implementation progress during the first five years of the project. After the initial implementation delays and once the funding situation improved, the R&BD performance was good, making this one of the few successful Bank funded road projects in India. However, inadequate planning caused the award of some contracts to be delayed due to land acquisition problems and low quality of engineering designs. In some cases, construction quality fell short of agreed standards. Furthermore, the failure to properly evaluate the economic returns of all roads may have resulted in the inclusion of some roads with low rates of return in the project. Summary of Findings, Future Operation and Key Lessons Learned 12. Whereas the project's dairy benefits are probably less than expected at appraisal, due in part to the introduction of more milk chilling centres, they are still significant. Considerable benefits  v have been generated from incremental agricultural production attributable to the improved roads and there are other social benefits, such as improved access to health and education facilities. Although implementation was initially slow, project outcome can be considered satisfactory since the main physical objectives were substantially exceeded, while the non physical objectives were in large part met, even though there were delays in the implementation of the MPS. 13. Future Oprations. Although all of the project roads are rural roads, which could be maintained at the village Parishad level, it is proposed in the outline operational plan that they be maintained by the R&BD. This would enable a state level maintenance effort, with uniform standards and funding, to meet in a prioritised fashion the maintenance requirements of the whole rural road network. However, nothwitsanding the recommendations of the operational plan the issue of devolving the maintenance of rural roads at the local level should be further studied. Furthermore, any future road project in Gujarat should make the improvement of maintenance funding and procedures of the whole road network a major development objective of Bank involvement in the sector. This should ensure the long term sustainability of the whole road network in Gujarat, including the 6000+ km of rural roads improved under the GRRP. 14. Key Lessons. Provided strong political support and project ownership exist to overcome fiunding shortages and other constraints, a well targeted and focused rural roads projects can be successfully implemented. However, from a technical point of view, even simple rural roads require proper engineering design and construction techniques, particularly with regards to drainage, compaction and quality control of construction by implementing agency3. The design and construction quality standards required should be clearly and strictly enforced through adequate supervision mechanisms to obtain uniform construction quality. 15. Land required for construction must be fully available and utilities should be relocated before award of contracts for construction or contract terms should be adjusted to take account of the possible non availability of land for some of the works. 16. Post qualification of bidders can be an effective procedure for use when selecting civil works contractors provided the qualification criteria are strict and are clearly defined. However, on larger construction contracts the use of pre qualification is preferable. 17. If non-economist are to assess the economic viability of each rural road before it is sanctioned for construction, clear/specific guidelines on the methodology to use should be provided and its application should be strictly supervised. Such work should be supported by inputs from agriculturists and agricultural economists to fully capture the benefits of increased agricultural production. Selection and design of rural roads would be simplified if up-to-date accurate topographical maps were available together with strip maps. 18. A steady flow of contracts over several years encourages contractors to modernise by adopting more modem construction techniques and through increased mechanisation. These may be Government staff as in this project or consultants appointed for supervision of construction.  IMPLEMENTATION COMPLETION REPORT INDIA GUJARAT RURAL ROADS PROJECT (Cr. 1757-IN) PART I: PROJECT IMPLEMENTATION ASSESSMENT A. STATEMENT/EVALUATION OF OBJECTIVES Objectives 1. The project's development objectives were the: (1) provision of all weather roads in rural areas; (2) improvement of rural road planning, construction and maintenance management; (3) development of a higher level of road safety and; (4) training of local staff in the administration and operation of the road network. At appraisal the project components included: (i) construction, reconstruction and improvement of approximately 4,040 km of village roads connecting 1,900 villages; (ii) procurement of equipment and vehicles; (iii) construction of houses, offices, warehouses and workshops; (iv) engineering and consulting services for project planning, design, supervision, monitoring and staff training; (v) technical assistance for a road maintenance study; and (vi) preparation of a road safety program. In addition the project aimed to improve road construction practices (SAR para 2.18) and improve contract management. This was to be achieved by encouraging larger contract packages including all the works on a stretch of road, the use of bank guarantees and the use of subcontracting. A minimum contract size of Rs 10 million was to be achieved from the second year of the project. In 1993, taking account of increased amounts of rupees available due to the depreciation of the Indian Rupee vis-a-vis the SDR, the scope of the project was increased to include a total of about 6,300 km of rural roads and SDR 19.3 million was cancelled due to project savings. B. ACHIEVEMENT OF OBJECTIVES 2. By 1990, two and a half years after credit effectiveness, implementation progress was almost two years behind schedule and was suffering from inadequate funding ', delays in land acquisition, delays in technical sanctions, poor contract management, weak enforcement of contractual conditions and specifications, inadequate R&BD staff and frequent staff turnover. None of the works started in the previous 30 months had been completed; works were partially completed on about 400 km of road, less than 10% of the work with about 35% of the project duration elapsed. Subsequently, implementation progress improved due to an increased commitment by GOG, appointment of an Officer on Special Duty to take charge of the project and Due in part to financial constraints resulting from three successive drought years.  -2- the lending of funds by the National Dairy Development Board (NDDB)2. In 1991 GOG requested an increase in the physical scope of the project, mainly for new milk-collection roads. IDA afier monitoring implementation performance for over a year, agreed in February 1993, to increase the target length of roads for improvement to about 6,300 km. When the credit closed at the end of December 1995, 5,662 km of road had been completed, 140% of the appraisal target and about 87% of the revised target. It is expected that when all ongoing contracts are completed by mid/end 1996 approximately 6,250 km of rural roads, will have been completed, which is about 155% of the SAR target. 3. As a result of the project, rural road construction techniques have improved through the adoption of modern contracting techniques; contracts are awarded for complete stretches of roads with subcontracting and the use of bank guarantees allowed, which was not previously the case; materials and equipment are now procured by the contractors rather than by the R&BD. Partial mechanisation of construction methods has contributed to higher quality roads with better riding quality than before. Although average contract size has increased because of the project, the minimum size of contract envisaged at appraisal has rarely been achieved, with only about 80 of about 420 contracts (19%) exceeding Rsl 0 million3. Overall construction quality has been good as reported in Bank supervision mission reports. However, there are also references of inconsistent construction quality. Examples are given of inadequate compaction of earthworks, poorly compacted and cured concrete, cracked or damaged culvert pipes. Reasons for this include the wide variety of contractors used, weak post-qualification criteria and, most importantly, inadequate site supervision. Nevertheless, the steady flow of contracts over a period of approximately seven years has enabled a number of contractors to receive a steady stream of work and thereby become established, gain good experience and build up competent staff and good equipment. A system for monitoring physical and financial progress of the project was established by consultants and used effectively throughout project implementation. This was of considerable help to Bank supervision missions (and to the ICR mission) but R&BD staff do not expect to use it for other projects. 4. An objective of the project was to improve road planning and programming through proper economic analysis, including economic analysis of each rural road included in the project. The methods of economic analysis applied in project preparation and throughout the project, while correctly valuing transport and savings in VOCs, overvalued dairy-related benefits and did not adequately quantify all other agricultural benefits generated from the improved rural road network. According to R&BD staff, economic analysis was applied to all roads which were chosen for inclusion in the project. The ICR mission saw evidence of economic analysis for 260 (out of 1,025) roads included in the SAR and of benefit/cost ratio calculations for roads added to the project in 1993. The latter were made by engineering staff in construction divisions, inexperienced in economic analysis and even these analyses were not always used for screening the roads to be 2 Four loans each of Rs 100 million at 17.5% per year interest were made by NDDB to GOG at six monthly intervals starting in 1991/92. If allowance for inflation is made this would reduce to abut 30 contracts (7%) exceeding the equivalent of Rs 10 million in 1989. The Project Agreernent required civil works to be grouped in packages of USS500,000 or more to the extent practicable. Only about 56 contracts (23%) have exceeded this figure.  -3- included in the project. As a result and as reported to the ICR mission by R&BD staff, a few roads with low benefit/cost ratios have been included in the project. 5. Road maintenance management was to have been improved by implementing the recommendations of a road Maintenance Planning System (MPS), which was to have been carried out during the project. The study was completed substantially behind schedule, and the software required to operate the system is not yet fully operational. Furthermore, until an adequate quantum of road condition survey data is inputted into system and until the MPS is used in a systematic manner, it is unlikely to fully generate an improvement in maintenance management practices. However, road maintenance has been improved to some extent by the provision of equipment and training of staff. 6. The quantity of equipment procured is considerably less than estimated at appraisal, due to a change in policy, whereby contractors are now expected to provide all of their equipment. Also some equipment, for example hot-mix asphalt plant, are now readily available in Gujarat. The requirements for buildings were overestimated at appraisal; the number built is about half of appraisal estimates, partly due to the lack of constructable land. 7. Although listed as an objective of the project, there was no specific provision under the project for preparation of a road safety program. Improved traffic segregation, signalling and other road safety improvements introduced by R&BD during the life of the project were not financed by the project. 8. The training component of the project shown in Table 3.4 (page 68) of the SAR to be carried out at the Engineering Staff College, Ghandinagar has been completed although the total number of staff trained (293) is somewhat less than the 485 shown in the SAR. Overseas training was not approved by GOI. 9. The economic analysis undertaken at appraisal covered 260 individual roads, representing 30% of the total project roads and about 25% of the estimated total project cost. The analysis yielded high returns ranging from 19% to 42% indicating the project's economic viability. Based on the analysis of 260 roads, the weighted average ERR for the project was calculated at 28%. This analysis was based on benefits arising to the dairy sector (increased milk procurement, reduced wastage and a reduction in milk animal deaths) and from savings on VOC, but excluded beiefits from savings in dairy transport costs and increased agricultural production attributable to project roads. Another weakness was that the analysis assumed that there would be no increase in milk procurement without the project and that all incremental milk procurement was attributable to the project roads, although other factors such as investments in more milk chilling centres and collection centres, improved dairy farm management methods and genetic improvement of milk animals would also contribute to increased milk production'. Further, with the benefit of hindsight, 4 Two senior R&BD staff visited UK in 1994/95. 5 Activities initiated under the National Dairy Projects (Credit 824-IN and Credit 1 859-lN/Loan 2893-IN) have had a large impact on dairy production not anticipated in SAR.  -4- the ICR mission found that the assumptions in the SAR regarding the benefits from increased milk procurement, reduced wastage of milk and reduction in milk animal deaths were over estimated. 10. The ERR has been re-estimated at project completion correcting for the above deficiencies and including all completed project roads. The re-estimated ERR of 16% (see Appendix C for details) though not comparable methodologically to the ERR of 28% estimated at appraisal, confirms that the project-supported rural roads program is economically viable. It should be noted, however, that due to the difficulties in precisely segregating increased milk procurement and agricultural production benefits due to the project roads, the re-estimated ERR should be deemed as indicative. However, since the relevant assumptions of the ICR mission were conservative, the actual economic impact of the project would not be less than that reflected in the re-estimated ERR. C. MAJOR FACTORS AFFECTING THE PROJECT 11. Factors not generally subject to government control. Project implementation was severely affected in the early years by drought, which caused resources (financial and manpower) to be diverted to drought relief, and by floods in August 1990, which slowed down construction progress in large part of the state (see para 3). A shortage of bitumen during 1995 further delayed construction progress. Sharp depreciation of the rupee vis-a-vis the SDR increased the amount of rupees available, thereby allowing the physical scope of the project to be substantially increased from about 4,000 km of roads in seven districts to about 6,300 km in 13 districts with substantial savings in SDR terms. 12. Factors generally subject to government control. Particularly in the early stages, the project suffered from inadequate funding, inadequate numbers of staff and frequent staff turnover (para 3). Diversion of staff and vehicles to election duties for two or three months during the working season in 1994/1995 disrupted implementation progress. 13. Factors generally subject to implementing agency control. In some cases, the project suffered from delays in land acquisition and in obtaining technical sanctions, as well as from poor contract execution and weak enforcement of contract conditions and specifications. 14. At appraisal, the total estimated cost of the 7-year project was Rs 2,220.5 million (US$170.8 million). This was to be financed by an IDA credit of SDR1O1 million (US$119.6 million equivalent) and by GOI and GOG contributions (US$51.2 million). In February 1993, SDR19.3 million (US$22.85 million) was cancelled from the credit due to project savings. The Completion cost is now estimated to be Rs 3,267 million (147% of the SAR estimate), equivalent to US$126.6 million (74% of the SAR estimate). While the length of roads has increased by 55% from the SAR estimate, the cost of roads in Rupee terms has increased by about 61%. However, in US$ terms, it is 18% less than what was estimated in the SAR. Even though project implementation took two years longer than estimated, the actual cost in Rs per km of road constructed (Rs 444,000/km) is only slightly (4%) higher than estimated at appraisal (Rs 426,000/km). When expressed in US$ (US$17,000/km) the unit cost was only about 50% of the SAR estimate. This is quite remarkable taking into consideration the relatively high Indian  -5- inflation rate (approximately 10%) during the implementation period and the long implementation period (8 years) for the project. D. PROJECT SUSTAINABILITY 15. Sustainability of the main project component - about 6,000 km of rural roads (out of a total road network in Gujarat of over 70,000 kIn) - will depend on; (a) the R&BD adequately monitoring the road conditions to optimise the priority maintenance needs for the network; (b) the continued availability of adequate funds for the maintenance and repair of the roads and; (c) the efficient use of the funds available for maintenance (see para 20). For the project roads, sufficient funds for maintenance have generally been made available6. R&BD has given assurances that the maintenance requirements of the project roads would be met in full. While the capability to maintain the roads exists, it is likely that the R&BD maintenance budgets will continue to-be insufficient and that the maintenance of some of the less trafficked project roads could possibly be neglected. However public interest and pressure can be expected to ensure that the more heavily used roads are maintained in reasonable condition and overall it is likely that the project roads will be sustained. Roads which are at present the responsibility of the project are being transferred to the State R&BD in good condition. An outline operational plan is shown in Attachment 2 to Appendix A. E. PERFORMANCE OF IDA 16. Performance of IDA during both the project processing and implementation phases has been satisfactory, except that IDA allowed Credit Negotiations to be delayed for over two years after appraisal (see Part II/Table 3) to accommodate the borrower's wish to have the project fully financed by an IDA Credit rather than by an IBRD Loan. An objective of the project was to increase the size of civil works contracts but this was discouraged by the SAR requirement that contracts which exceeded US$ 700,000 require prior review, which entailed additional administrative clearances (and hence delays) that R&BD were reluctant to undertake. IDA was very responsive to the borrowers needs. IDA agreed to extend the scope of the project, increase disbursement percentages to take account of drought and the effects of the Gulf war and to postpone the credit closing date by one year to enable the completion of several hundred km's of rural roads. The implementing agency considers that IDA supervision missions gave valuable guidance and support. However, some supervision missions did not adequately monitor whether the correct economic evaluation criteria was being used to select roads to be funded under the project. 6 Funds made available are less than the Eighth Finance Commission's norms (see part IVTable 6) but this is to be expected in the early years of the project. One supervision mission expressed satisfaction with maintenance but others have only drawn attention to budgetary allocations being less than the norm.  -6- F. BORROWER PERFORMANCE 17. GOI's delays in passing on project funds to GOG, slowed down implementation progress during the first five years of the project, and the funding situation did not really improve until 1993. After the initial delays (para 3) and once funding problems had been largely overcome, the borrower and the implementing agency (R&BD) generally performed satisfactorily, making this one of the few successful Bank funded road projects in India. The SAR target of physical works improvements was achieved by mid-1993, six months before the completion date in the SAR, and was exceeded by more than 40% by the time the project closed. The cost of construction per km of road was extremely close to appraisal estimates (see para 15) and GOG is to be commended for obtaining bridging financing from NDDB to speed up project implementation. However, in some cases, inadequate planning caused some contracts to be delayed due to lack of timely acquisition of land; the quality of design and construction sometimes fell short of acceptable standards; and the failure to correctly evaluate the economic benefits of all roads to be funded under the project may have resulted in the inclusion of some roads (see para 5) with low rates of return. G. ASSESSMENT OF OUTCOME 18. Even though project implementation was initially slow, the 'physical' targets of the project were substantially exceeded. Thanks to a steady supply of contracts over a seven year period, several small contractors were able to modernise and to improve their road construction capabilities. However, had the contracts for road construction been larger, as envisaged at appraisal, larger contractors would have been established with higher levels of mechanisation and more technical expertise, which would have probably resulted in better quality construction. The equipment and buildings provided under the project, although less than envisaged at appraisal, are sufficient to meet the present day requirements of the R&BD. Whereas the benefits from increased milk production, lower milk-animal deaths, reduced souring of milk and savings on vehicle operating costs are probably less than expected at appraisal, due to the introduction of more milk chilling centres, they are still significant. However, considerable benefits have been generated from incremental agricultural production attributable to the improved roads. Producers' incomes have risen dramatically through increased mobility and access to primary markets. There have been other social benefits difficult to quantify, such as improved access to health and education facilities. Overall the project outcome can be considered satisfactory since the main physical objectives were substantially exceeded, while the non physical objectives were largely meet even though there were delays in the implementation of the MPS. H. FUTURE OPERATIONS 19. Although most of the project roads are rural roads, which could be maintained at the village Parishad level, it is proposed in the outline operational plan (Attachment 2 to Appendix A) that they be the responsibility of the state R&BD. This is to ensure uniform state level maintenance standards and funding levels to meet in prioritised manner the maintenance requirements of the whole rural road network. Since funds available to the R&BD may be used for any road contract, a  -7- method of prioritising the use of the funds is needed. Even though the MPS completed in 1994 is supposed to provide a rational system to prioritise maintenance needs, it still requires the input in a systematic and timely manner of the results of periodic road condition surveys to be fully effective. To do this for the whole state road network is likely to require a substantial investment by the GOG in terms of time and money. An initial 'operationalization' of the MPS covering about 2,700 km of the state road network is currently underway on a pilot basis. 20. Furthermore, any future road project in Gujarat should make the improvement of maintenance funding and procedures of the whole road network a major development objective of Bank involvement in the sector. This should ensure the long term sustainability of the whole road network in Gujarat including the 6000+ km of rural roads improved under the GRRP and the roads that would be improved under any future road project in the State. I. KEY LESSONS 21. Provided strong political support and project ownership exist to overcome funding shortages and other constraints7, a well targeted rural roads project a la GRR can be successful in a progressive state like Gujarat. Without political support and project ownership, even simple projects can suffer from delays and weak implementation, as was the case during the first two and a half years of implementation of this project. 22. From a technical point of view, even the construction of simple roads like rural roads require proper engineering design, particularly with regard to drainage, compaction and quality control of construction by implementing agency. These design and construction quality standards required should be clearly and strictly enforced through adequate supervision mechanisms to obtain uniform construction quality. 23. Land required for construction must be available and utilities must be relocated before award of contracts for construction or the contract terms must be adjusted to take account of the possible non-availability of land for some of the works. 24. Post qualification of bidders can be an effective procedure for use when selecting civil works contractors provided the qualification criteria are strict and are clearly defined. However, on larger construction contracts the use of pre qualification is preferable. 25. If non-economist are to assess the economic viability of each rural road before it is sanctioned for construction, clear/specific guidelines on the methodology to use must be provided and its application should be strictly supervised. Such work should be supported by inputs from agriculturists and agricultural economists. Selection and design of rural roads would be simplified if up-to-date accurate topographical maps were available. 7 Such as flood and drought.  -8- 26. A steady flow of contracts over a long period, such as seven years, enables contractors to receive regular work and thereby become established, gain good experience. Also it encourages them to mechanise and adopt more modem construction techniques.  -9- PART II: STATISTICAL TABLES Table 1: Summary of Assessments A. Achievement of objective Substantial Partial Negligrible hlc Applicabl Macro policies FI D EF l Sector policies El LII El Financial objectives El LIm [In Institutional development El [3 E El Physical objectives [l El El Poverty reduction El E El Gender issues El El El E Other social objectives F E E l Environmental objectives El El E [ Public sector management E [ El El Private sector development m []E El El Other (specify) El El El El B. Projectsustainabilv Lily Unlikel Uncertin Fx1 F- ED  -10- EighLy C. Bank performne satisfactoQv Satisfactory Deficient Identification z El : Preparation assistance E= E E Appraisal IZ [I [ Supervision f El Highly D. Borrower performance satisfactory Satisfy Deficin Preparation If El Implementation Covenant compliance EL FE Operation (if applicable) Ei E Highly Highly E. Assessment of outcome satisfactory Satisfactorv Insatisfactow unsaIif