India - Bihar Rural Roads Project

Document of The World Bank FOR OFFICIL USE ONLY ReportNo. 9910 PROJECT PERFORMANCE AUDIT REPORT INDIA BIHAR RURAL ROADS PROJECT (CREDIT 1072-IN) SEPTEMBER 23, 1991 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. Country Exciange_Rates (Yearly Averages) 1978 US$1.00 - Rupees 8.19 1979 8.13 1980 7.86 1981 8.66 1982 9.45 1983 10.10 1984 11.36 1985 12.37 1986 12.61 1987 12.96 1988 14.60 ABBREVIATIONS ANSIS - A.N. Sinha Institute, Patna ARDC - Agricultural Refinance Development Corporation BOR - Back-to-Office Report DEA - Department of Economic Affairs, 001 ERR - Economic Rate of Return GOB - Government of Bihar GOG - Government of Gujarat GOI Government of India ICB - International Competitive Bidding IDA - International Development Association IRC - Indian Roads Congress OED - Operations Evaluation Department, World Bank p.a. - per annum PCR - Project Completion Report PPAR - Project Performance Audit Report PR - President's Report REO - Rural Engineering Organization, GOB RHD - World Bank Resident Mission, New Delhi SAR - Staff Appraisal Report SR - Supervision Report TD - Transportation Department, World Bank FISCAL YEAR OF THE BORROWER April 1- March 31 THE WORLD BANK FOR OFFICIAL USE ONLY Washington. D.C. 20433 U.S.A. Ooce o# DWettnr.CMMeal Opraone Evaluimin September 23, 1991 E1ORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Performance Audit Report on India Bihar Rural Roads Project (Credit 1072-IN) Attached, for information, is a copy of a report entitled "Project Performance Audit Report on India Bihar Rural Roads Project (Credit 1072-IN)" prepared by the Operations Evaluation Department. Attachment This document has a restricted distribution and may be used by mecipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorisation. I1 FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT BIHAR RURAL ROADS PROJECT (Credit 1072-IN) TABLE OF CONTENTS Page No. Preface ........................................................ i Basic Data ..................................................... iii Evaluation Summary ............................................. v I. FROM INCEPTION TO BOARD APPROVAL .......................... 1 Project Inception ......................................... 1 Identification Mission .................................... 2 Further Contacts .......................................... 3 Appropriate Construction Methods .......................... 3 Project in Suspension ..................................... 4 Successive Preparation Missions .......................... 4 Project Brief ............................................. 5 Further Preparation ....................................... 7 Pre-appraisal Mission ..................................... 8 Project Appraisal ......................................... 10 Project Monitoring ........................................ 11 Issues Paper, Decision Memorandum, Negotiations and Board Approval ...................................... 11 II. FROM BEGINNING OF IMPLEMENTATION TO COMPLETION ............ 12 First Supervision Mission ................................. 12 Feedback .................................................. 12 Second Supervision Mission ................................ 14 Third Supervision Mission ................................. 14 Project Monitoring ........................................ 14 Fourth Supervision Mission ................................ 15 Project Monitoring ........................................ 15 Fifth Supervision Mission ................................. 15 Project Monitoring ........................................ 15 Feedback .................................................. 16 Impact Study .............................................. 16 Sixth Supervision Mission ................................. 18 Reaction at IDA Headquarters .............................. 19 Impact Study .............................................. 20 Supervision and Request for Arbitration ................... 20 Seventh Supervision Mission ............................... 21 Eighth Supervision Mission ................................ 22 Update .................................................... 22 Ninth Supervision Mission ................................. 23 First Extension of the Closing Date ....................... 25 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. TABLE OF CONTENTS (CONT.) PAge No. Tenth Supervision Mission .................................. 26 Eleventh Supervision Mission ............................... 26 Twelfth Supervision Mis!Aon ................................ 27 Thirteenth Supervision Mission ............................. 29 Second Extension of the Closing Date .......................29 Further Supervision and Credit Closing .....................30 Preparation of the Project Completion Report ...............31 III. FINDINGS AND RECOMMENDATIONS .............................. 32 General Findings .......................................... 33 Specific Findings - Inception and Preparation ............ 35 Specific Findings -- Implementation and Monitoring . 37 Specific Findings -- Bank Group Assistance ................ 38 Recommendations ........................................... 39 Annex 1 Labor-based Construction and Maintenance Methods: Notes on the Bank's Research Work ......................... 41 PROJECT PERFORMANCE AUDIT REEORT BIHAR RURAL ROADS-PROJECT (Credit 1072-IN) . PREFACE 1. This is the Project Performance Audit Report (PPAR) on the Bihar Rural Roads Project (Credit 1072-IN of SDR 35.0 million, or US$50.0 million equivalent). praised in March 1980, the project was approved by the Board in November 19,0. The Credit became effective in January 1981 and the project was to be completed in March 1985. The Closing Date was extended twice (to June 30, 1988) but the account was kept open until March 31, 1989, to clear up commitments entered into up to the time of Credit closure. The last disbursement was made on April 18, 1989, and the remaining balance of about SDR 98,000 was cancelled. The Transport and Energy Operations Division of Country Department IV of the Bank's Asia Regional Office prepared a Project Completion Report (PCR) published on December 29, 1989. 2. An OED mission visited India in December 1990, discussed the project with Indian authorities, inspected a large number of roads, and gratefully acknowledges the assistance and hospitality it received. OED prepared the PPAR by studying the project files, the Staff Appraisal Report (SAR), the President's Report (PR), the Credit Agreement, the 1989 PCR, transcripts of the Executive Directors' meeting which considered the project, and published reports on the economy and the country. For the convenience of the reader, the PPAR text is preceded by statistical material extracted from the PCR. 4. According to standard procedure, OED sent copies of the draft PPAR to the Borrower for comments. As of August 5, 1991, no replies had been received. PROJECT PERFORMANCE AUDIT REPORT BIAR RURAL ROADS PROJECT (Credit 1072-IN) BASIC DAT I/ World Bank Regional Unit AOL& Sector Transport Subsector BiWh&ys Borrower Goverement of India Executing Agency Rural Engineering Organization, 00 Prolect Timetable Plan Actual First Mention in Tiles 1/78 Appraisal 3/80 3/80 Negotiarions 9/80 9180 Board Apzroval 11/80 11/80 Credit OArement Date 12/80 12/80 Effectivencs Date 1/80 1/80 Project Completior Date 6/85 6/88 Closing Date 6/86 6/88 Expected and Actual Cost. and ExpenditUes Appraisal Estlmate Actual Expenditures Res million US$ =illion Re million US$ million Construction 382.3 45.5 419.4 36.8 Equipment & Spares 39.9 4.8 47.0 4.2 Supervision 22.2 2.6 23.5 2.1 Total 444.4 52.9 489.9 43.1 Expected and Actual Cumlatie Disersements (USS million) FY81 FY82 TY83 FY84 FY83 TY86 FY87 FY88 FY89 Estimated 0.7 8.3 18.1 27.9 35.0 - * * Actual 0.4 1.3 10.6 13.3 19.1 22.0 22.6 26.9 28.3 Actual/Est. (2) 57 16 59 48 55 63 65 77 81 Expcted and Actual Economic Rate of Return Appraisal Etimaces 22Z Estimate at Completions 0% J./I For much greater detail, please see Part III (Statistical Summary) of Project Comlation Reort - India: Biha .ural .Roads Pro ect (Credit 1072-IN), December 29. 1989. Report Number 8256. iv Staff Input (Staff-Weeks) F78 FY79 FY80 FY81 FY82 FY83 FY84 FY85 F86 FY87 FY88 FY89 TOTAL Preparation 3.5 14.4 53.2 - - - - - - - - - 71.1 Appraisal - - 32.1 11.3 - - - - . - - - 43.4 Negotiations * - - 12.3 - * - - - * * - 12.3 Supervision - - - 5.5 4.7 4.2 6.3 4.4 11.1 9.9 7.9 - 54.0 Completion - - - - 8.7 8.7 GRAND TOTAL 3.5 14.4 85.3 29.1 4.7 .2 6.3 4.4 11.1 9.9 7.9 8.7 189.5 Field Mission Data Month/ No. of No. of Staff Report Year Persons Weeks Weeks Date Identification 1/73 1 1 1.0 3!78 Identification 6/78 1 1 1.0 8/78 Preparation 9/78 3 1 3.0 10/78 Preparation 5/79 1 1 1.0 5/79 Preparation 6/79 2 1 2.0 7/79 Preparation 8/79 1 1 1.0 9/79 Preparation 10/79 1 1 1.0 12/79 Preparation 10/79 1 4 4.0 12/79 Preparation 10/79 4 2 8.0 12/79 SubTotal 22.0 A,#praisal 3/80 2 3 6.0 4/80 Supervision I 11/80 1 1 1.0 1/81 Supervision II 3/81 1 .2 0.2 4/81 Supervision III 6/81 1 1 1.0 7/81 Supervision IV 10/81 2 .6 1.2 11/81 Supervision IV 10/81 1 1 1.0 11/81 Supervision V 5/82 1 1 1.0 6/82 Supervision VI 11182 1 1 1.0 12/82 Supervision VII 3/84 2 .5 1.0 3/84 Supervision VIII 3/85 1 1 1.0 4/85 Supervision IX 1/86 1 1 1.0 2/86 Supervision I 5/86 1 .4 0.4 5/86 Supervision XI 8/86 1 1 1.0 8/86 Supervision XII 2/87 2 1 2.0 5/87 SubTotal 12.8 Completion 10/88 2 2 4.0 12/88 Completion 3/89 1 .4 0.4 5/89 SubTotal 4.4 GRAND TTAL 45.2 v PROJECT PERFORMANCE AUDIT REPORT BIHAR RURAL ROADS PROJECT (Credit 1072-IN) EVALUATION SUMMARY Introduction 1. The project was appraisal), have either failed or identified in January 1978. It was show premature deterioration. On appraised in March 1980, Board many roads there is no motorized approval was given in November traffic. 1980, and the Credit became effective in January 1981. March 4. Actual project cost in 1985 was the planned completion dollars is US$45.7 million (as date. against US$52.9 million estimated at appraisal) because, during Obiectives implementation, the value of the Rupee deteriorated vis-a-vis the US 2. Project objectives were dollar. Actual project cost in provision of all-weather access to rupees is R520.6 million (as farms and villages, and improvement against R444.4 million estimated at of construction and maintenance appraisal). procedures. Project components consisted of construction or 5. Lack of data precluded a strengthening of about 700 km of re-estimation of the economic rate rural roads; improvement of of return (ERR) which, at maintenance of roads administered appraisal, had been estimated at by the Rural Engineering 21%. As constructed, the roads Organization (REO) of the will not last sufficiently long to Government of Bihar (GOB); and generate enough benefits to justify procurement of construction and the investment and the post- maintenance equipment. completion ERR would probably be negativt. Implemettation Exerience and Proiect Results Sustainability 3. Instead of March 1985, 6. The short life of the actual completion took place in K0ads, coupled with tae lack of - June 1988. Institutional effective maintenance, does not objectives were not met but bode well for the sustainability of physical results in terms of km of project benefits. road built were close to appraisal expectations. Because of design, Revie workmanship, and supervision shortcomings, quality of the 7. Review of the project finished works is poor. Mary roads, experience covered two periods: costing about US$50,000 per km (as against US$25,000 estimated at vi (a) From Inception to Board offers specific lessons on: Approval (paras 1-35), Inception and Preparation; and Implementation and Monitoring; and (b) From Beginning of Bank Group Assistance. Implementation to completion (paras 36- 10. Specific Findings -- Inception 104). &nL?reation (paras 123-129) The review revealed some wider (a) Rural road projects are topics and the Audit reathed the to improve rural following conclusions and transport, not just to recommendations. bu4ld roads. (b) Project design must be General Findings (paras 110-122) attuned to realities of time and place. 8. From beginning to end, a (c) Project design must take divergence persisted between what advantage of local GOB wanted and what IDA advocated. expertise. GOB was preoccupied with short-term (d) A distinction must be employment, IDA with long-term drawn between physical development, and the two and institutional preoccupations never meshed. Bihar objectives. did get some employment, IDA did (e) Covenants and conditions disburse most of a US$35 million must relate to credit, but the roads are failing achievements which an and no institutional strengthening Executing Agency is took place. GOB and IDA never came empowered to accomplish. together because of socio-political (f) S t a f f t r a i n i n g conditions in Bihar, because is th institutional project design was inappropriate, objective deserving the and because IDA was unwilling to greatest attention. allocate sufficient resources to (g) Large numbers of civil supervision. Furthermore, during engineers must be given project implementation, GOI the opportunity for on- demonstrated lack of enthusiasm for the-job training in technical assistance. labor-intensive works. Beyond that, proper 9. There has not been a incentives must be follow-on operation to take developed to keep advantage of the experience gained experienced engineers in from the Bihar Project. Bihar in labor-intensive wcrk particular, and India in general, sites. needs to improve rural transport because, for many years to come, 11. SRecifig Eindin&s -- agriculture will provide a Inlementation and Konitorinz livelihood for hundreds of millions (paras 130-133) and better access to rural areas is a precondition for a more (a) Implementation is productive agricultural sector. IDA facilitated when targets can help by facilitating provision and responsibilities are of needed infrastructure, precise and explicit. encouraging staff training, and (b) Flexibility in adjusting strengthening of local both target outputs and institutions. The Bihar Project Closing Dates is vii desirable but must not be Z"omendations (para 139) abused. (c) Experience with exi-ante 12. Thn Audit recommends the and ex-Rost evaluation of following: individual roads suggests that 5-10 km segments are (a) By, say, end 1991, Bank too short for ERR Management discusses with analyses. It might be GOI the latter's preferable to concentrate intentions regarding on least-cost solutions strengthening of rural for construction and transport over the next maintenance and leave it 10 years. at that. (d) Experience with (b) If GOI decides to modify monitoring road impact its policy towards upon the rural, socio- technical assistance, the econom'z environment Bank Group ought to suggests it is present GOI, by June methodologically 1992, with tentative difficult and financially estimates for assistance expensive. needed to improve rural transport in a number of Specific Findings -- Bank States to be identified GrouR AssistAnce (paras 134-138) by GO( (e) The Bihar Project took (c) If GOI agrees that the too long. When it was Bank ought to perform the finished, there was no role of coordinator for follow-up operation to financial and technical make use of the assistance, the Bank experience gained. Group ought to rresent (f) GO! must decide whether GO!, by December 1992,' or not India needs with a detailed plan for foreign techniques and interventions to improve technical assistance for rural transport in the planning and implementing States specified by GOI. programs to strengthen rural transport. Depending on what GO! decides, the Bank Group will have to reconsider its own plans for future support. (g) If the Bank Group is to get seriously involved in strengthening rural transport, appropriate staff time allocations will have to be made. PROJECT PERFORMANCE AUDIT REPORT ID.A& IAR RURAL ROADS PROJECT (Credit 1072-IN) I. FROM INCEPTION TO BOARD APPROVAL January 1978 - November 1980 Project Inception 1. In January 1978, Government of Bihar (GOB) officials discussed with an IDA officer a possible road project. GOB objectives were modest (improve some rural roads to facilitate organization of milk cooperatives), and GOB asked that a road component for Bihar be added to the National Dairy Project, under preparation at the time. The IDA officer tesponded that the project was too advanced, too complex, and too focussed on milk whereas Bihar needed a road network for a variety of goods. GOB was therefore urged to prepare a separate project. 2. IDA knew that Bihar roads were in such poor condition that transportation virtually ceased during the rainy season.l/ Since the Bank Group had a number of ongoing ano proposed agricultural schemes,2/ IDA felt that a rural roads project could complement them. GOB did not object to IDA's approach. 3. 'aring January-February 1978, GOB reflected on what it had undertaken to deliver: a proposal for a roads project that would support development financed through a number of other IDA-financed schemes. GOB was unfamiliar with IDA's project preparation procedures and took up the matter with the Government of India (GOI) in Delhi. GOB said it wanted IDA to assist with project preparation, preferably by assigning some of its staff to work with GOB officers in Patna. 4. GOI's policy was to resist technical assistance. However, Bihar insisted and GOI acquiesced. A March 2, 1978, telex from the Bank's Resident Mission in New Delhi (RMD) urged IDA Headquarters to send a reconnaissance mission because both GOI and GOB wanted to identify and prepare a project with IDA staff assistance. 5. It was a significant request but IDA was not geared to respond. A March 8, 1978, internal note concedes that a reconnaissance mission would be desirable but staff was not available. Instead, copies of recent appraisal reports were sent to GOB as models. 1/1 Reference was also made to the May 1977 Eastern States Foodgrain Review - Bihar which stated that "better marketing facilities and feeder roads at village and intermediate levels are needed to improve the efficiency of distributing fbam inputs and marketing farm outputs." 1/ Including the ongoing Bihar Agricultural Credit. Bihar Agricultural Markets, Bihar Agricultural Research and Extension. as well as the proposed Second National Seed and National Dairy projects. 2 6. GOB, supported by GOI, eventually proposed a rural roads project covering the construction, strengthening or improvement of about 12,700 km of village rotads over a three-year period at an estimated cost of US$335 million. Clearly, scope as well as cost were too large and IDA had to react formally. Identification Mission 7. An IDA officer visited India between June 16-21 to discuss the GOB proposal. His August 2, 1978, Project Brief showed that a project in Bihar would be easy neither to plan nor to implement.2/ GOB had prepared an unrealistically large project, counterpart funds would be hard to come by, maintenance would be problematic, qualified staff was in short supply, bureaucratic procedures would delay land acquisition and procurement, the Rural Engineering Organization (REO) of GOB could not supervise properly the works, and beneficiaries were too poor to buy vehicles. 1/ The size of the project was way out of proportion in relation to past expenditures. It was doubtful whether GOB could provide the needed counterpart funds for construction or maintenance, whether GOB had the technical capability to undertake the extra maintenance work, and it was unclear how much work would be done by force account and how much by contract. Engineers needed to be trained. Stone was in short supply. Equipment requirements and ownership had to be reviewed. Development of the contracting industry had to be investigated. Agricultural development, necessary to justify road investments. created problems of coordination. Benefits from increased agricultural production depended on other inputs, principally irrigation. The project ought to complement ongoing IDA and GOB agricultural projects. Road development should therefore concentrate in north Bihar where population pressure was greatest, soils most fertile, and groundwater most plentiful. There was no experienced planning staff in Bihar. An economist was needed to help with route selection. In addition, IDA ought to supply one development economist for three months to train REO staff in simple techniques of rcute appraisal. Contract documents had to be reviewed because they encompassed local practices which IDA might consider unacceptable. Major delays ought to be anticipated with land acquisition and procurement. Project supervision was bound to create difficulties. REO did not have the experience to undertake supervision to the standards that IDA expected, and GOI objected on principle to the employment of consultants on the grounds that sufficient expertise exists in India. Most of the benefits would not reach smaller farmers who are too poor to buy their own vehicles. 3 8. It would have seemed that what was really called for was a project focussed on staff training and inst'tution-building in REO. However, on August 9, 1978, IDA wrote to GOB stressing the need to integrate the proposed project with other IDA-financed initiatives. Further Contacts 9. Meanwhile, GOB began to see the project as more than a means to improve rural roads. An October 12, 1978, memorandum (written after a September visit to Bihar by an IDA officer) said that GOB's interest in the project was related to forthcoming State elections. A project of this size and visibility would demonstrate the incumbent government's commitment to job creation and general improvement of the economic lot of the rural population. 10. The memorandum went on to say that REO would be responsible for project preparation, assisted by the Rural Development Department. However, while both agencies had technically competent staff, political expediency was likely to influence project design, unless the involvement of outside sources of financing served as a moderating influence on political pressures. 11. GOB was itself aware of the need for such a moderating influence and, once again, asked IDA for technical assistance (four experts: in rural planning, highway engineering, economics and agronomy) to assist in project preparation. The GOB request was mentioned by the IDA officer to the Department of Economic Affairs (DEA), Ministry of Finance, GOI. So far, GOI had tried to accommodate GOB. This time, DEA appeared cool and, a few weeks later, IDA was formally informed that GOI would reconsider the project in the context of its overall investment priorities. Even so, GOI stated that continued IDA involvement in project preparation on a technical assistance basis would be welcomed. 12. This was another opportunity for IDA to establish a substantive presence in Bihar but the October 12, 1978, memorandum recommended otherwise: since GOI was changing its investment priorities, IDA should wait for GOI to finalize and clarify its intentions before actively participating in further preparation of the Bihar Project. Nevertheless, since project preparation would be required irrespective of ..he source of finance, IDA should communicate to GOI its endorsement in principle of the approach outlined in the Aide Memoire left behind in Bihar by the September 1978 IDA mission, and emphasize the importance of an early start on project preparation.A/ Ayororiate Construction Methods 13. The October 12, 1978 memorandum provoked an October 18, 1978, note from the Bank's Transportation Department (TD). A/ Project preparation based on this Aide Memoire presupposed the presence of IDA experts in Bihar. How preparation could proceed without them was not elaborated upon in the October 12, 1978, memorandum. 4 14. TD said that the memorandum suggested that IDA finance road construction equipment and training of operators and mechanics. TD conceded that equipment did have a place in a labor-abundant economy but insisted that, in a rural roads project for India, equipment ought to be a minor component. The Bank's research on labor-based construction and maintenance methods had shown that these were technically and economically justified for most aspects of civil construction in India.5/ 15. This finding had been documented in several publications. Furthermore, it had been the topic of extensive discussions at the December 1976 meeting of the Indian Roads Congress.J/ Consequently, TD felt that IDA-financed civil construction pr-jects in India (other than very large and technically complex works) should aim to work with (and hopefully to improve somewhat) the labor-intensive methods currently in use. Rather than equipment procurement, the most important component ought to be work training for foremen and supervisors. Proiect in Suspension 16. TD suggestions were reasonable but came at an inopportune time. On October 19, 1978, IDA wrote to GOI acknowledging that the project had been dropped from the pipeline at GOI request. 17. The file does not explain what prompted the GOI request, nor does it explain what led GOI to revive the project in early 1979. On April 5, 1979, the project was back in the pipeline and the RMD suggested a follow-up mission. Successive Preparation Missions 18. The May 9, 1979, follow-up Mission Report describes the revival of the project. Total estimated cost had been reduced to US$50 million equivalent and the proposed IDA contribution to US$25 million equivalent. Project preparation was to start on June 1, 1979 and be completed by February 25, 1980. Appraisal would be in March 1980. At GOI request, technical assistance for project preparation was dropped. Consequently, increased IDA supervision of project preparation would be required. 19. The next Mission took place in June 7-10, 1979. It left behind an extraordinarily detailed Aide Memoire but did not discuss anticipated traffic. Also, it made no reference to labor-based construction methods I/ At the time, the 3ank was engaged in a 10-year program to explore the feasibility of substituting unskilled labor for mechanical equipment in the civil engineering process. A brief summary is attached as Annex 1: Labor-based Construction and Maintenance Methods - Notes on the Bank's Research Work. g/ Please see: Indian Roads Congress. 37th Annual Session. Bhopal. December 1976. Panel Discussion on: Scope for the substitution of labour and equipment in civil construction and some asiects of the use of labour-intensive methods for road construction (New Delhi. February 1978). 5 and did not take into account TD concerns (expressed in October 1978) that traditional labor-based construction techniques had to be improved. 20. Neglect of this last point is puzzling because the August 3, 1979 Back-to-Office Report (BOR) lists outstanding issues in detailz/ and refers explicitly to GOB concern that labor-based construction methods be utilized to the maximum extent possible. Project Brief 21. The August 7, 1979 Project Brief noted that: (a) REO estimated Rs 10,000 per km per year for rural road maintenance. This was significantly higher than the amount allocated even for national highways. Current expenditures for rural roads were about Rs 1,000 per km. (b) REO estimated an economic rate of return (ERR) of 40%+ for the project as a whole, not counting costs to provide irrigation and other necessary inputs to achieve full agricultural benefits. As there was only partial irrigation in most areas, ERR was unrealistically high. (c) There were major constraints to the achievement of parallel investments: electrical power was a bottleneck; diesel pumps were expensive; credit facilities were faulty; better extension services and improved marketing facilities were needed; transport costs between farm and market were high. (d) The revised project from GOB consisted of 2,816 roads (totalling 19,242 km) in the three parts of the State. A possible US$50 million Credit would be divided at US$25 million for North Bihar, $15 million for South Bihar and $10 million for Chotanagpur/Santhal. This money would be enough for about 1,000 km in all. Preliminary selection out of the REO proposal had yielded about 260 roads with a combined total of about 1400 km, L/ Outstanding issues included: EmloyMent - GOB expressed concern about rural underemployment. REO would have to use casual labor to the maximum, especially in earthworks. Further. GOB felt it might be possible to provide labor under GOI's "Food for Work" Program, thereby reducing direct costs of the project to Bihar. Non-cooperation by other Departments with REO in Bihar. Lack of expertise in the Proiect Preparation Cell - complete lack of any agricultural or economic expertise. Shortage of eauipment - little available for good quality work. Provision of a consultant - to help with project preparation, monitoring, teaching. etc. Project Supervision - IDA would have to decide in what manner this can best be done, given India's dislike of consultants. 6 about 50% more than could be financed. Final route selection would be the task of a second round. There was still no economist to assist with final selection of roads. 22. The Project Brief went on to identify eight main issuesL/ but emphasized three: ProJect Supervision -- REO cannot supervise up to IDA standards. A system ought to be set up, using Bihar as an example, which would eventually allow GOI to prepare and administer similar road projects under IDA program lending. Establishing such a system would require more attention than can be given by normal supervision missions. GOI objects to the employment of consultants on the grounds that sufficient expertise exists in India. Therefore, increased supervision ought to be provided by stationing either in New Delhi or in Patna IDA staff to supervise the project. Coordination of develoRMent -- Accrual of full benefits from increased agricultural development depends on.a number of other inputs. This complicates supervision: progress in activities outside the project would also have to be systematically observed. Farm-to-market transport -- More transport means are needed to encourage competition among transporters. It would be advisable to increase the number of privately owned small motorized or non- motorized vehicles (rickshaws, auto-rickshaws, etc.) but farmers are short of cash. The purchase of small vehicles ought to be financed under the project. 23. The emphasis on supervision underscores the concern of earlier IDA mission that REO could not perform this task adequately. Earlier missions had also emphasized that project benefits would depend on "parallel investments" and that this would complicate project monitoring. What emerges for the first time is the explicit concern that farmers could not afford mechanized vehicles to take advantage of the improved roads. Here was a serious question: if rickshaws were principal means of transport, did Bihar need roads costing $27,000 per km? The record contains no indication that the question was ever asked. These were: Provision of local counterpart funds; Future rural road maintenance; Constraints of manpower and materials; Development of the local construction industry; International conMetitive bidding; Project supervision; Coordination of development; and Farm-to-market transport. 7 Further PreRaration 24. An IDA mission visited Bihar in August 1979. It reported that the economist had still not been recruited, that a consultant would be more necessary than ever, and that coordination among GOB agencies was imperfect. The September 13, 1979, BOR recommends financing small freight vehicles for farm-to-market transport ("appropriate vehicles"), a popular notion in the Bank TD 2/ at the time. 25. The September BOR was followed by an October 4, 1979 note on rural vehicles. On November 7, 1979, TOR were written for a British Consultant to look into suitable transport means for the Bihar rural population. The Consultant concluded that the answer was better bullock carts. 12/ This conclusion sprang from well-known facts, and It is puzzling that facts had not so far received the attention they deserved. 26. The Consultant found that the Agricultural Refinance Development Corporation (ARDC) was in a position to refinance bullock carts and tractors and made the following observations: Animal-drawn carts played an important role in Bihar's rural economy because of their farming and transport functions. Therefore, improved animal-drawn carts ought to be a complementary investment to the roads financed under the project The demand for carts in Bihar was not adequately met by currently available institutional financing. 2/1 See, "Appropriate Technology in Rural Development: Vehicles Designed tor On and Off Farm Operations", Rural Development Unit. Transportation Department. World Bank. October 1978. .1./ Please see the December 18. 1979. BOR of the Consultant's November 1979 mission. IJ/ Animal-drawn carts were the most common means of transport in rural Bihar. being used for hauling freight over distances of up to 10-15 km. In 1972 Bihar had 780.000 carts. In the whole of India there were 16 million carts pulled by 70 million bullocks. For a population of about 600 million, this worked out to 1 to 37. whereas the ratio in Bihar was 1 to 72. By 1978. the number had grown to 850,000. most of them bullock carts with a few horse carts and camel carts. The average life of a cart was about 10 years, so annual replacement needs for Bihar were about 85,000 carts. There were two types of carts: the conventional (carrying capacity of 0.7-1.0 ton) and pneumatic tire carts (about 10-15 percent of the fleet) with a carrying capacity of 2.0-2.5 tons which, in addition to better harnesses, drawbars and yokes that allowed greater efficiency in haulage, did less damage to surfaced roads than conventional carts with steel-rimmed wheels because the load, though heavier, was better distributed over the contact surface. 8 Since 1978, local banks had been encouraged by ARDC to finance carts and would need continued support by ARDC. The Bank's association with ARDC through several projects had been satisfactory. ARDC would take responsibility in the preparation and appraisal of the proposed project component. The Consultant concluded by noting that a decision by GOI was expected in January concerning their support for this component.12/ Once again, the question arises: if carts were the most appropriate transport means for the rural areas of Bihar, why did the project want to finance paved roads? The record does not indicate that this question received any attention. Pre-apraisal Mission 27. The December 20, 1979 Preappraisal Mission BOR reported that route selection was proceeding satisfactorily. The idea to assist the local construction industry was abandoned.11/ There was as yet no organiiation for maintenance of REO roads. Funds voted were used mainly in piecemeal repairs by local contractors whose work standards were often poor due to the absence of essential equipment and effective supervision. 12/ On January 31. 1980, the Bank brought to the attention of the GOI the benefits of rubber-tired carts and suggested that the project include an appropriate component. j.1/ For the following reasons: Interviews with contractors (said the BOR). showed that the present system under which contractors hire equipment from Government for use on works operated reasonably satisfactorily. Contractors had no serious complaints about any failure on the part of Government to repair or maintain plant, and if Government did not do repairs, they were done by local firms. Contractors said that although in principle they would like to own their plant and equipment, in practice they preferred the present system where continuity of work and lack of financial liquidity was not a problem. They felt that any technical advice they need on work methods or operations would be provided by the Government engineer supervising the works. They would probably not use the services of an adviser for fear their affairs would become common knowledge. Although the existing system had its disadvantages (the BOR went on), it might be premature to plan greater independence for contractors at this stage. It was therefore proposed to introduce a measure of quality control on works plus limited mechanization where present methods were unsatisfactory. Contract control and procedures were laid down in detail in official codes which, if adhered to, should be satisfactory to IDA. 9 28. Contradictions in the BOR (which at one point said that supervision was adequate and at another that it was not), were noticed ir IDA Headquirters. Indeed, this BOR raised questions in late December 1979 about the goodness of a project whose maintenance and design aspects (as well as the pavement life of its roads), were put in doubt. 29. In a January 3, 1980, memorandum IDA staff responded that maintenance was indeed an issue. However (the staff said), and regardless of any covenants or conditions that might be attached to the Credit Agreement, there was little IDA could do to ensure that maintenance would be either funded or carried out correctly. In view of subsequent problems, two paragraphs (one on design standards and the second on quality control) from the January 3, 1980, note deserve to be quoted in full: "The design standards are such that will, I think, ensure that the great aijority of the road lengths, if well constructed, will survive to serve traffic reasonably well through the analysis period of ten years, even with poor maintenance. W/ Nearly all routes will be surfaced, which is required by poor soils and lack of materials rather than anticipated traffic volusies. To this extent, roads are overdesigned for their traffic load. Roads will be raised to be above normal flood levels but, as was pointed out in the original project brief, it is very difficult to forecast and to design for the very large flows of monsoon flood water moving towards the Ganges. It must, I feel, be accepted that a proportion of the roads we construct will fail or be destroyed due to flooding. The design will attempt to keep this proportion minimal but in designing for normal floods, the abnormal flood is always a possibility next year." "To ensure survival of the average road for ten years, some form of technical audit will be necessary during construction. CO are not enthusiastic about providing this as they have enough trouble trying. to ensure that national routes constructed by States are correctly built. Neither was the use of consultants, even domestic, a popular idea when discussed with CO on my last mission. The GOB regulations define the duties of the engineer very precisely, but given the prevailing standards of contractors in Bihar and the practices in their use, I feel that to ask the engineer to act as his own auditor is to ask too much. The GOB audit regulations require a thorough and independent financial audit of works carried out by departments, which would include the operations of REO, but the Audit Department 14J~ The point is that roads were not well constructed. Para 4 of the March 16, 1980 Aide Memoire on Project Appraisal reads as follows: " In discussions with REO. it was confirmed that the Indian Roads Congress Manual on Route Location. Design. Construction and Maintenance of Rural Roads (1979) was being followed in design, and that their 1980 edition Specifications for Road and Bridge Works would apply for construction. These appear to be satisfactory and appropriate, although on field inspections there appeared to be some cases where work in progress on roads planned for rehabilitation would result in substandard design, or would have to be replaced early in its life." 10 is not concerned with technicai performance. I suggest that we persuade the GOB to include technical personnel in the Audit Department in order to check on the technical, as vell as financial performance cO REO under the -roject. Proiect ADoraisal 30. Appraisal was comprehensive.15/ Two aspects might have caused some apprehension but, at the time, they did not. First, GO insisted that the cart component be excluded from the project. Eecon , the Appraisal Mission indicated that, as part of project monitoring, and in addition to normal engineering progress reports, the effects of rural road improvements on their areas of influence should be closely followed during implementation.j&/ 31. In the Audit's view, the deletion of the cart component was a mistake which -- given the evidence that carts were the predominant mode of rural transport -- IDA should not have agreed to. Insistence upon "impact monitoring" raised a different set of problems. Tried and true impact monitoring techniques had not yet been developed -- as, indeed, they have not been developed down to this day -- and the decision to engage in "monitoring" was reached without sufficient evidence that GOI and GOB were genuinely interested in the exercise. *jg/ Please see the March 16. 1980 Aide Memoire on: (a) Route selection and design. (b) Financing advance purchase of materials for construction. (c) Review of procurement recommendations, (d) Specifications and general conditions of contraet for works. (e) Hire charges for equipment used by contractors, (f) Connection of project roads with the all-weather road network. (g) Technical audit of works. (h) Pavement design for project roads. (i) Food-for-Work Program. (j) Complementary investments and services, (k) Tax element of the project, (1) Non-project program of the REO. (m) Future rural road maintenance program, and (n) Training. .L/ Periodic monitoring of such effects would theoretically have the following benefits: (a) verification of the assumptions made during project preparation and appraisal for the economic analysis of the roads; and (b) collection of information to be used in the economic analysis of future rural road projects, whether financed by IDA as follow-up projects or by other sources. 11 Project Monitoring 32. Iy March 1980, IDA preparation missions had been visiting Bihar for two years and were familiar with the prevailing constraints. Bihar was the second poorest State of the Indian Union. It was not endowed with staff trained in transport economics, civil works quality control, agricultural economics, and cost accounting. Even so, the Appraisal Mission felt that monitoring indicators ought to include at least the following items: (a) progress of roadworks (b) actual costs and unit costs of construction (c) traffic counts on a sample of project roads, including non- motorized vehicles (d) agricultural data within the areas of influence on a sample of project roads before and after the project, including such items as area under cultivation, cropping patterns, utilization of agricultural inputs, yields, external services, marketing, and credit facilities (e) transport charges and crop movements in a sample of districts before and after the project; and (f) impact of the provision of plant and equipment on the local contracting industry and their reaction to stricter standards of quality control. 33. Aware that the plethora of indicators, and the data processing they would require, represented a formidable task, the Appraisal Mission suggested that GOB might wish to consider whether it wished REO to undertake the work in its entirety, or whether it should seek assistance from other branches of government, from universities, or from research institutions. Issues Paer. Decision Memorandum. Negotiations and Board Approval 34. The April 1, 1980, Issues Paper describes the proposed US$25.0 credit in terms identical to that of the March 16 Aide Memoire. Neither the Issues Paper, nor the May 1, 1980, Decision Memorandum, contain any reference either to the project's political dimensions, or to the employment aspects, or to the peculiar geomorphological features of Bihar, or to the fact that most of the traffic consisted of non-motorized vehicles. 35. The September 15, 1980, transmittal note from IDA staff to the Regional Vice President recommended an increase of the credit amount to US$35.0 million. Negotiations were completed in October 1980.1/ The project was approved by the Board on November 11, 1980. The effectiveness date was set at January 15, 1981, project completion was planned for March 1985, and the Closing Date was set at June 30, 1986. 11/ Please see. Summary of Negotiations dated October 20. 1980. 12 II. FROM BEGINNING OF IMPLEMENTATION TO COMPLETION January 1981 - June 1988 36. All preparation missions declared that the nature of the project, GOI unwillingness to accept consultants, and insufficient staffing in REO made it imperative that IDA supervision be unusually frequent and intensive. The Audit agrees and, in this section, will devote particular attention to the frequency, intensity, and thoroughness of supervision. First Supervision Mission 37. This took place from November 28 through December 2, 1980,1/ and the Mission found no major problems. Design work was progressing and a first-year construction program had been proposed by GOB. Prequalification of contractors had been delayed by review of documents by GOB legal authorities, but the hope persisted that construction work could commence by April or May 1981. No mention in the report of any concern for the employment-generation aspects of the project. Project monitoring aspects were expanded: the Supervision Mission handed to GOB draft Terms of Reference for a proposed study of rural transport worldwide, to be carried out and financed by the World Bank. Monitoring of the Bihar Project was to provide inputs for this study. A January 29, 1981, letter from IDA to GOB confirmed this. Feedback 38. A February 10, 1981, letter to IDA from an Indian importer of road construction equipment shows how the project was viewed locally. "We have got honour to state that we are sole distributors of the equipments of ... in India. In endeavour to fulfill your mission, I thr- undersigned, contacted ... (a Senior Officer) of Rural Engineering Organization of Bihar, to sell ... machines, who told me that Motor graders and berm levellers are useless in road construction. Although he needs some road rollers but he is not interested In the cost incurred for its purchase and bank with though (our) roadrollers are much cheaper than Indian counterparts with far better quality. The (Senior Officer's) special remark was that the funds was allocated by IDA for providing jobs for rural unemployed folk and that road construction is secondary matter. However I am of the viev that road roller alone without motor graders and bermlevellers can't help making levelled roads. IDA should keep eyes on competitive prices of the construction machines purchased against their finance. I feel that inspite of best engineers working in Indian Central Road Research Institute for road development our road engineering is not doing well, for short of resources, one of I!/ BOR and Full Report dated January 12, 1981. Performance Rating: 1 (Problem Free). 13 the reasons, resulting in avoidable excess use of exhausting and soaring priced petroleum products, thus deteriorating our economy. A good levelled surface road saves fuel consumption up to 65%, nullifies wear and tear of vehicles, so reducing laters down time and saving road's own damage due to lateral thrust of vehicle wheels. No doubt assistance granted by IDA is developing rural roads is admirable (but) our economy compels us to let the things going in usual ways as told by ... (the Senior Officer), REO of Bihar, India. Vieving the above facts USSR, Japan, etc. Instead of supplying funds and let authorities decide its use insist to send construction machines and advisers needed for the project financed. I request you to adopt similar attitudes. I therefore request you to review your grant if you really wish to help our roads and our economy as estimated in your news release." 39. What the letter said was that GOB considered the project as a means for creating jobs and that quality of roadworks was irrelevant. This was in complete opposition to the letter and the spirit of the Credit Agreement which assumed that quality roadworks would be the major outcome of the project. Good quality roads were why equipment-based techniques had been espoused in the first place. Nevertheless, the letter was now saying that the wrong equipment was being procured. 40. The March 4, 1981, IDA response said that responsibility for execution of IDA projects and therefore for the award and administration of contracts, rests with the Borrower. Procedures to be followed for procurement of services or equipment had been discussed with GOB and were incorporated in a legal agreement between GOB and the Bank. The next IDA mission to visit Bihar would discuss with GOB the points raised in the letter. 41. This provoked a counter-response dated March 13, 1981, which reads in part: "Thank you for your reply regarding above subject. In fact we really believe in latest technology and more over we are having latest and most sophisticated machines for construction of good roads which is your mission in case of Bihar lural Roads. Unfortunately our relations with (country of manufacture) became sore in pastdays, our engineers are afraid ... for the reasons of embargo of spare parts supply, which would cause machines stop for want of spare parts. Moreover we are here in India making Road Rollers of old technology which has much troubles in operation and maintenance. In the interest of both countries ... we need latest machines ... and want to get rid of troublesome equipments. Local bidding in the case of Bihar would never make fulfill your mission of Roads for rural economy. We therefore hope that you would help us to materialise your mission." 14 42. The record does not contain an IDA response to this second letter and Supervision Mission Reports contain no indication that the issues raised (employment is.more important than roads; local competitive bidding does not work) were ever discussed with GOB authorities. Second Supervision Mission 43. This lasted 24 hours (from Maich 17 to March 18, 1981).1J/ Again, no major problems were identified. Government procedural difficulties had delayed progress with prequalification of contractors, but the mission was assured final GOB approval would be given within a few days. A next mission was planned for mid-June and the hope was expressed that arrangements could be discussed at that time for project monitoring and the possible use of the A.N. Sinha Institute of Social Studies (ANSIS) to assist with this work. Third Supervision Mission 44. This took place from June 22 through June 27, 1981.aQ/ Again, no major problems were identified. The need for Cabinet approval to publication of REO prequalification procedures had caused some delay. Future progress would depend on the abilities of local contractors and on good supervision by REO. The mission noted that REO had more non-project work than had been anticipated, and REO staff numbers allocated to supervision of the IDA project ought to be kept under scrutiny. Bids had been requested for eight roads in the first year's program. Some equipment had been received and was in use. The necessary complementary investment had been or was being provided for twelve roads included in the first year's program, and a program for the construction or improvement of a further 62 roads had been agreed, subject to the required complementary investments being made. Project Monitoring 45. In June 24-26, 1981, the Bank's Regional Transport Advisor visited Patna to investigate possible use of data from Bihar in the Bank's world-wide study of the impact of rural road improvements. He met with senior officials of ANSIS to discuss the Institute's possible participation. The July 9, 1981, Supervision Mission Report (Annex 1, para 10) noted that any such study would be in addition to the requirements for project monitoring as laid out in Annex 2 of the SAR, although some of the data would be common.Z1/ j2/ BOR and Full Report dated April 14, 1981. Performance Rating: 1 (Problem Free). ZO/ BOR and Full Report dated July 9. 1981. Performance Rating: 1 (Problem Free). 21/ An August 28. 1981 letter from IDA to GOB acknowledges GOB assistance to the Regional Transport Adviser's visit to Patna but notes that data complexities would require a further visit by an IDA Rural Transport Specialist. The letter gives the impression that data gathering was now a matter exclusively concerned with the Bihar Project. Similar sentiments were expressed in an August 31. 1981 letter from the Regional Transport Adviser to the Chief Engineer (Roads) of the Ministry of Shipping and Transport in Delhi. 15 Fourth Supervision Mission 46. Tlis took place between October 14-17, 1981.221/ Again, no major problems were identified. Cabinet approval had been given to prequalification procedures and steps had been taken to speed up award of the first batch of contracts. Much of the second year program of work had been approved and bidding was to commence shortly. REO had not given consideration to revisions of the equipment procurement schedule, as was requested by the June 1981 mission, and they were again asked to do so as soon as the results of bidding became reasonably certain. All applicable Credit Covenants had been fulfilled. Proiect Monitorinz 47. A Rural Transport Specialist from the Bank's TD accompanied the October Supervision Mission and in a meeting at ANSIS, where requirements of project monitoring were discussed, explained what his Department would require for their Study. Annex 4 of the Supervision Mission's BOR consists of draft Terms of Reference for Evaluation of the Effects of the Project. 48. This document does not mention any connection between the monitoring mechanism proposed for Bihar and the Bank's worldwide study. However, the proposed extent of data collection was so vast that questions could be raised as to whether GOB needed such a level of detail. The presumption seemed to be that ANSIS would do work for the project but the detailed information would be useful to the Bank in its world-wide study. If so, the matter of who would pay for what was left open. Fifth Supervision Mission 49. This took place between May 31 and June 4, 1982.2_/ Again, no major problems, although para 4 of the BOR notes that both GOB and GOI expressed concern over the apparent lack of progress in civil works. Contracts had been let for a total value of about US$13 million, and further works had been approved for bidding. Some of the equipment had arrived, other was awaited and REO had been asked to submit a list of additional items needed which could be financed out of savings in the original list. All applicable Credit Covenants had been fulfilled. Project Monitorinj 50. In the summer of 1982, a draft proposal, partly for TD purposes, was received in IDA Headquarters from ANSIS. It covered project monitoring and collection of other data (specified in the draft TOR left behind by the October 1981 mission). The ANSIS proposal contained ,a preliminary cost estimate of about US$100,000 over a five-year period which would include a Pilot Survey for six months; "Before the Project Surveys" for twelve L2/ BOR and Full Report dated November 5, 1981. Performance Rating: 1 (Problem Free). L3/ BOR and Full Report dated June 15. 1982. Performance Rating: 1 (Problem Free). 16 months commencing in mid-1983; and "After the Project Surveys" starting in 1985 or 1986, with the Final Report in 1987. 51. IDA Headquarters felt that the ANSIS proposal needed drastic reorientation and communicated its recommendations to GOB in a September 13, 1982, letter. The letter does not make any reference to the split between items of interest to the Project and items of interest to the world-wide study of TD. Feedback 52. While exchanges on impact methodology continued, a letter in Hindi arrived at IDA Headquarters in early September 1982. In translation, it reads as follows: Sir: It is known that the World Bank has approved a road construction from Bushahr to Khijirpura. It is not known how this road was approved. There is no Norm/Criteria to approve it and the public vill not be benefitted from it. The funds spent will be a waste and that is not the purpose of the World Bank. Therefore, it is requested that instead of this road a road from Nayagaon to Khijirpura be approved for construction. The road meets all norms and criterion and public will be benefitted by this road. (Signed by 25 people) 53. IDA responded on September 28, 1982, that roads were selected on the basis of an analysis agreed between IDA and GOB but that, sometimes, circumstances change and routes are re-analyzed before construction begins. Data were not available in Washington on the two roads in question but the matter would be taken up during the next supervision mission. 54. On the same day, IDA wrote to the Chief Engineer of GOB mentioning the contents of the letter and asking that the relative merits of the two roads be discussed during the next supervision mission scheduled for November 1982. The record contains no indication that the matter was ever discussed in the field. Imact Study 55. A Bank Sociologist visited Patna between October 25 and Nivember 5, 1982.2A/ The BOR noted two main objectives of the monitoring exercise: (a) assess the validity of the assumptions made in the original economic analysis; and (b) determine the economic, social and cultural Z_4/ Please see India: Bihar Rural Roads Proect Supervision Mission_- Back to Office and Full Report dated November 11, 1982. Despite the title, this was not a supervision mission but a continuation of discussions between TD. GOB, and ANSIS on the design of a monitoring mechanism that would eventually allow measurement of the impact that the Bihar Project had upon the rural ecrnomy. 17 effects of road improvements in rural areas of India.25/ For the two objectives to be met, the Impact Study would have: (a) to validate *:en key assumptions made during project preparation; (b) to determine thirteen economic effects, and (c) to determine twelve socio-cultural effects. Furthermore, the design of such a study would have to pay particular attention to (a) Sample design, (b) Traffic surveys, (c) Key informant Interviews, and (d) Household Interviews. 56. The November 11, 1982, BOR includes observations about conditions in Bihar: Throughout September and October, there was a strike by Junior Engineers, and little prospect for an early resolution of their grievances. This, combined with religious holidays brought contruction work virtually to a standstill on road and irrigation projects. If the strike was to continue, it would jeopardize completion of the first year's program (para 3). REO did not have anyone who could contribute much to the building of institutional capacity for economic planning and analysis (para 4). Supply of electricity was irregular and uncertain in Patna, which would have a deleterious effect on the ANSIS data processing equipment. There was a severe drought underway and the ineffectiveness of measures designed to combat its effects suggested that the project's production targets would not be met (para 8). Road improvements would likely be followed by a multitude of commercial activities based in small roadside stalls, services such as barber shops and bicycle repair, and "a brisk business in human- powered porterage" (para 9). Annex 1. Attachment A (para 8) of the November 11. 1982 BOR suggests, in terms that may not be thoroughly transparent, how this could be done. The last sentence of the following extract is especially challenging. "This task (i.e., determination of economic, social and cultural effects of road improvements upon rural areas) covers a broad range of topics, some of which have been set forth in the Sinha Institute proposal. Others are suggested here as a result of the finding (sic) of rural road impact studies carried out in other parts of the developing world, including Mexico, Kenya, Thailand. Philippines, Korea, Senegal, Upper Volta and Madagascar. In all cases the research hypotheses take the form of statements that a statistically significant difference will exist between the amount of change that can be measured in a given parameter before and after the project in a project influence area and the change that can be measured in the same parameter over the same time period in a control zone which did not receive a road access improvement." 18 57. The implications of "brisk business in human-powered porterage" do not seem have been adequately grasped. The Impact Study was to gauge the effect of a program building roads at a cost of US$27,000 per km. It seemed that road improvements would lead to an increase of rickshaw traffic. Would rickshaw traffic ever justify such high construction costs? Surprisingly, this inconsistency between high quality infrastructure and low quality transport means continued to escape notice. 58. On the other hand, the cost of the proposed research received ample attention in the BOR. Para 17 says that the amount allocated in the SAR for "impact studies" was about US$10,500. The detailed cost estimate for the proposed research was about ten times that (US$96,000). Discussions in the field between the Sociologist and the IDA Supervising Engineer revealed that savings in equipment procurement could be used to pay for the additional costs of the research. Still, this would not cover everything. The BOR noted that TD had an interest in certain specific research topics which would be identified as a result on work on rural mobility and rural transport services. Data which could be used to address these topics would be collected in the context of the Bihar study. However, given that research was ongoing, data requirements were still unknown. Consequently, they could not be inserted in the terms of reference for the Bihar Impact Study. Therefore, for ANSIS to address the as yet unspecified topics, ANSIS vould require support from one or more consultants. 59. In sum, planning for both the TD i search and for the Bihar Impact Study had hardly begun. Nevertheless, the BOR concluded that "if the Study was executed as planned, it would make a real contribution to the state of the art in rural roads assessment, not only in India but worldwide." Sixth SuRervision Mission 60. The mission took place in November 1982, and the December 22, 1982, Supervision Report gave the project a rating of 1 (No Major Problems). Contracts had been awarded for about 570 I of roads; GOB had so far performed well in selecting routes and awarding contracts for their construction; future progress would depend on the performance of local contractors and on the supervisory ability of REO, both of which showed room for improvement.2,/ 61. Paras 11-15 of Annex 1 of the same Report (Aide Memoire of the November 1982 IDA Highway Mission), offer a less positive account. The mission visited three roads under construction in North Bihar and observed the following: I/ Section 6: Summary of Project Status, Trend and Major Problems. 19 (a) there was no setting out of center line and levels; (b) no batter pegs were placed before the start of the earthworks; (c) shoulders and widening were not being done in accordance with the specifications, ane neither was compaction; (d) borrow pits were incorrectly sited and too close to the foot of the embankment; (e) stone stockpiled on site was found to be dirty, poorly graded and/or too flaky; there was no inf-rmation available on test results, apart from grading in one case. 62. The Mission noted that the amount of testing done on materials and for compaction was far less than that recommended by Indian Roads Congress (IRC) standards. This, in spite of the fact that three out of the four mobile laboratories provided under the project were on these three jobs. The Mission made the pointed remark that engineers in charge of works should realize it is a misuse of funds if contractors are paid for work done which is not in accord with the contract and specifications. 63. In addition, the Mission noted that in some cases contracts had been let before any effective action was taken to acquire the necessary land, which led to piecemeal construction around standing crops and encroaching buildings. On one road, the Mission was unable to reconcile the work on site with the drawings. Levels were uncertain and conflicting opinions were expressed on maximum flood levels. In view of these observations, it is curious that the Mission gave the project a rating of 1 (No Major Problems). Reaction at IDA Headauarters 64. The inconsistency was spotted in IDA Headquarters and Management expressed surprise27/ at the high rating. 65. The supervising officer conceded on January 12, 1983, that the rating was too high. He then said that ensuring reasonable work standards had been a concern all along, given India's aversion to consultants, even Indian consultants. To get around the problem, IDA considered having GOI run checks but GOI were against becoming involved. A requirement was therefore included in the Project Agreement that the Technical Examiner's Cell would check at least ten percent of the works. 66. In the light of subsequent problems with quality control, the penultimate paragraph of the January 12, 1983, note deserves to be quoted in full: " The scattered nature of the work, the aversion to consultants, and the radical changes IDA expected in testing, work methods, and supervision, coupled with the standard of local contractors available and the general administration of Bihar, all made close supervision by IDA very important. REO engineers in the project cell appreciate this and expect 17/ The document referred to above, which commented on the December 22. 1982 Supervision Report, is missing from the files. 20 continuing supervision. Hence, the recent apparent trend in reduction of field supervision by the Bank is very disturbing and I fear that this project in particular, and others in general, will suffer as a result. I feel the only effective way to get results on projects such as this one is for Bank staff to arrive as unexpectedly as possible on construction sites and for them to ask lots of detailed questions on operations. This is time consuming, frustrating, and very wearing on Bank staff. However, the marginal cost to the Bank is minimal, and if such supervision does not take place, even the limited improvements we can now expect will certainly not take place. " As events would show, this strong suggestion for very frequent, and very intensive supervision, went unheeded. Impact StudyZ 67. Meanwhile, the Bank was distancing itself from the world-wide Impact Study that had been mentioned to the Bihar authorities during the December 1980 Supervision Mission, and had been elaborated upon during the October 1982 mission. 68. A March 7, 1983, letter from the Bank to the Commissioner for Highway Construction in GOB regretted that the TD review of the state of the art in rural road impact studies was not yet ready for distribution. It also regretted that Bank resources did not permit sending a Bank staff member or consultant to participate in the training program conducted by ANSIS for interviewers. Supervision and Request for Arbitration 69. As already noted, the January 1983 suggestion for more intensive supervision was not followed up. Indeed, the October 28, 1983, Hand-over notes by the supervising officer reveal that no supervision took place between November 1982 and October 1983. Then, on November 23, 1983, an Indian supplier wrote IDA asking for the supervising officer's arbitration in a dispute with REO over the purchase of asphalt mixers financed under the project. 70. The supplier wrote that REO had accepted the IDA officer as an arbitrator, and he attached full documentation to back up his complaints. This was a striking demonstration of faith in the integrity of the IDA officer (who had retired) and, in the Audit's opinion, IDA ought to have immediately retained him in a consultancy capacity and dispatched him to India to arbitrate. 71. Unfortunately, IDA did not choose te- do this. Instead, it sent to REO the documentation received from the supplier. What prompted this is not clear because REO had been given copies before it accepted the former IDA officer as arbitrator. 21 Seventh Sunervision Mission 72. This took place between March 19 and 21, 1984, one and a half years since the last one. Problems with law and order allowed the Mission to inspect only two roads and even this had to be done under police escort. The Mission consisted of two engineers (one from IDA Headquarters and one from RMD), who filed two separate reports. In his March 29, 1984, BOR the RMD engineer noted the following: Cost of roadworks -- All roads are being black-topped with 20 mm premix carpet ... The cost of construction is about Rs.0.55 to 0.6 million per km, quite high compared to the average of Rs. 0.394 million per km on roadworks in progress under irrigation projects (para 8). Supervision, Land acquisition -- Works are scattered and coordination and monitoring are difficult. Difficulties in land acquisition lead to unsatisfactory road alignments are not ideal... Quality of two works inspected was very good 28/ (para 9). Eauipment utilization -- During field inspection a lot of machinery such as scrapers, hot mix plants, trucks, etc. was seen lying idle along the national highway. Even though 3.5 years have passed since the start of the project, important items of machinery have not been procured... It is also surprising that though a lot of earthwork and asphalting work is involved, water tankers, trucks, tractor trailers and stone crushers are not included in the list of the machinery required under the project (para 10). 73. In his April 25, 1984, Supervision Report the IDA Headquarters Engineer noted (para 2 of the covering memorandum) that the project was going well, with no major problems, and he gave it a rating of 1. Section 6 (Summary of Project Status, Trend and Major Problems) notes the following: Output -- The project is progressing well and will achieve a total of about 680 km. 400 km were completed by March 31, 1984, and a further 172 km are under construction... Staf - In the areas visited, site staff appeared keen and competent; test results were produced whenever asked and cost records were also available... L8J This statement is immediately qualified in the sentences that follow: u However in respect of RCC culvert form finish should be aimed at. For hume pipe culverts, costly NP1 pipes are being used whereas cheaper NP2 pipes would have been adequate on most of the rural roads. Presently no good stuff is being used for construction of road berms, however on busy routes provision for hard side shoulders would be necessary." 22 Beneficiaries -- The Mission received many positive unsolicited comments on the impact of the road in the villages recently connected to the highway... Socio-economic impact -- The "baseline" survey and a "before and after" rural impact study by the ANSIS is complete and a report on this phase has been presented ... Eighth Supervision Mission 74. The eighth Supervision Report (April 26, 1985) by the IDA Headquarters Engineer noted that the project was progressing satisfactorily and gave it a 1 rating. With unit costs available from three years of operations involving contracts for 654 km of roads, it seemed likely that the project would achieve its target of about 700 km. In the view of the Supervising Officer, because of strikes and climatic and political disturbances, an extension of the closing date (June 30, 1986) might be required. 75. These positive conclusions are mitigated by para 3 of the Aide Memoire left in Patna at the end of the Mission's March 25-29 visit. The paragraph begins by saying that the general standard of work appeared satisfactory; it was clear that correct amount of supervision was being given to the works, and that quality control test results were always available when requested. But then, it says that the mission did observe many instances of settlement and pavement failure adjacent to cross drainage works resulting from a failure to compact material adjacent to the pipe. 76. Para 5 of the same Aide Memoire touches upon further points. Another general observation (it says) concerns alignment, embankment width and land acquisition. A black-topped road tends to stay where it is first built and serves as an anchor for future development. In the mission's view, it is a short-sighted policy to accept a sub-standard alignment to avoid land acquisition, in the hope that it can be improved later. Furthermore, it is not -echnically feasible to widen earthworks built originally with 1:1 siae slopes in order to restrict the borrow area, and then flatten the slopes later. The wedge of soil added cannot be properly compacted. Embankments must be set out to the full width dictated by the specified side slope and compacted in one operation. 77. Even more disturbingly, para 6 of the same Aide Memoire notes that maintenance leaves much to be desired: some project roads had been extensively damaged by monsoon rain and flood, possibly because of poor design of cross-drainage works; potholes were noted to be already forming on one of the brand new roads; none of the new roads were in regular maintenance; some older REO roads were in very poor condition; the maintenance grant was very low. Ilada.tSg 78. The file contains a Supervision Update (dated December 3, 1985). The IDA Headquarters Engineer stopped in Delhi on his way to Sri Lanka and was visited by the Chief Engineer of the South REO, Patna, who gave him a 23 verbal briefing on the project, plus a set of tables showing the physical and financial status of the project at June 30, 1985. On the basis of this, the IDA Headquarters Engineer completed a Form 590, noted that the project was going well (Section 3: Project Status) aad gave it a rating of 1 (No major problems). A possible reason for the generous treatment is that, at the time, the IDA Headquarters Engineer was deeply into the preparation of a rural roads project for Gujarat and could not afford the time to deal in any great detail with the Bihar Project.2_/ Ninth Supervision Mission 79. There was no field visit between March 1985 and January 1986. The ninth Supervision Mission took place from January 13 to January 22, 1986. It was conducted by the RMD Engineer and his :eport (dated February 12, 1986) notes that the project "is not doing quite well". Specifically, of 657 km under contract, 484 were completed by end December 1985. In the first nine months of FY86 work progress was characterized as "very poor indeed". Out of Rs. 55 million allocated during FY86, only Rs. 11.3 were spent in the first nine months. REO was burdened with additional workload of RS 550 million (against an original workload of Rs.350 million) and this reduced the priority attached to the IDA project which represented only 6% of the REO's new work program. The Aide Memoire left by the Mission in Patna made a number of points: Planning -- Approval of an additional road program (which had been promised more than a year earlier) would be delayed further (para 2). Land acquisition and quality of work -- There was a general tendency to avoid land acquisition to meet the minimum standards of 10 m in built-up areas and 12 m in open areas; non availability of sufficient land has been a major handicap or excuse for not building up the earthwork from the bottom up; when earthwork is undertaken it is rushed through and important aspects like breaking up of clods, removing vegetation, compacting, etc. are all conveniently forgotten; due to limited land-widths, side slopes are often inadequate; (para 4). Monitoring -- Failures were observed in the planning and monitoring of important works; there have been considerable delays in completing many works; even though contractors achieved very little progress during the contract period, no penalties are enforced and extensions are freely granted (para 6). CoMpagtion -- The Mission observed that inadequate attention was given to compaction, resulting in failing of edges and loosely packed strips at the end of the carriageway (para 7). 2_9/ cf. the December 23. 1985, telex from the IDA Headquarters Engineer to the Engineer in the RMD. 24 Ancillary works -- In sandy and silty regions, road embankments have been constructed with the local soils but without any adequate protective-measures; for want of prompt repairs, eroded road embinkments have become quite dangerous (para 8). Field supervision -- Field officers frequently do not prepare detailed working drawings for each site and blindly follow the type drawings, wasting valuable resources (para 10). Major works -- A three-span causeway with raft foundations is under construction on the second kilometer of the Manganj Road. The construction is either improperly done or has been badly damaged in the rainy season (para 12). Easipment -- The Mission tried to collect information about performance of road rollers but was disappointed to see that the log books are not properly maintained. Utilization is also far from satisfactory and outputs are low. Due to the strong opposition of elected representatives for removal of rollers from their zones many machines are idling for lack of work. The current rules and indifferent attitude of the mechanical organization further prevent prompt repairs (para 14). In spite of 115 rollers procured under the project, the Mission observed some of the road works suffering from the lack of rollers (para 15). Maintenance -- The expected increase of the maintenance grant has not come to the REO but gone to the highways department. Maintenance in general continues to be disastrous as the meager grants are not used according to priority... the field engineers are less and less worried about the timely completion of the IDA project or the maintenance responsibilities... Though GOB has agreed to send their maintenance staff for special training courses run by the Central Road Research Institute in Delhi, no progress has been made in this connection (para 16). Selection of new routes -- Though the March 1985 IDA Mission agreed to take up additional works totalling about 92 km, the GOB has not yet finalized the work program (para 18). Progress of work and time extension -- The March 1985 Mission was hopeful about performance during the current year. Actual performance is far from satisfactory. It would be very ambitious to expect full completion of the additional work by June 1987, the end of the one-year extension currently under consideration (para 19). Impact Study -- The ANSIS has completed fieldwork and hopes to prepare a draft report by April 1986. The current delay in preparing 25 the draft report is due to extension of fieldwork and computerization (para 23).I/ 80. In view of the above comments, it is surprising that the Mission gave the project a rating of 2 (Moderate Problems). At any event, the Report caused some consternation in IDA Headquarters which had been under the impression that the project was proceeding well and was now obliged to reconsider this impression. IV First Extension of the Closing Date 81. All the above imperfections notwithstanding, IDA staff decided that the Closing Date ought to be extended to June 30, 1987 and so recommended to IDA Management on April 9, 1986. Justifications included: (a) that the GOB had been asked to immediately restore priority to the project works; (b) that all of the 654 km under the ongoing contracts were now expected to be completed within a one-year extension; (c) that all but US$1.7 million of the credit proceeds would be utilized by the extended closing date; and (d) that enough IDA and GOB funds would be available for project completion. 82. The April 9, 1986, recommendation noted (para 6) that GOB was in compliance with covenants, although maintenance of existing rural roads had only marginally improved and random technical audits were of poor quality. However, the need for improvement in these areas had been stressed in successive supervision missions and would continue to be monitored closely. 83. The Audit has been unable to determine on what evidence the conclusion was reached that maintenance had improved at all. On the 1_/ Please see the February 18, 1986, note from the Bank to REO on the Baseline Survey. The note reveals that the Bank was aware that only 5% of total traffic movements were those of motorized vehicles; 65% were non-motorized vehicles and 30% pedestrians. The following statement is then made: " This underscores the need to study the cost savings accruing to non-motorized vehicles in order to design and evaluate suitable road improvements. Another topic for future study would be the extent to which people shift from non-motorized vehicles to motorized vehicles once a pucca (i.e., paved) road is opened." It seems that the December 1979 Consultant's report recommending better carts had been completely forgotten. 21/ On March 3, 1986. Washington summarized the contents of the Supervision Report in a telex to the IDA Headquarters Engineer, who was on mission in Sri Lanka and India. He replied on March 10. 1986. that the situation had markedly deteriorated since March 1985 and that he agreed with the findings of the RMD Engineer. 26 contrary, supervision missions had repeatedly p,,inted out that maintenance was deteriorating. Tenth Supervision Mission 84. This took place on May 22 and 23, 1986. It was conducted by the RMD Engineer and, in comparison with his January Aide Memoire, his BOR (dated May 28, 1986) is subdued. Under Section 3 (Project Status), the Supervision Summary (dated July 11, 1986) notes the following as major problems: (a) Works are likely to be disrupted when new administrative divisions are established in Bihar, and project roads are transferred from an old division to a new one; (b) As a result of (a), procurement for additional roadworks may not proceed as planned; (c) Encroachment on completed roads may pose a serious threat to traffic; (d) Cement and bitumen shortages may affect timely completion of roadworks; (e) Results of the technical audit, of the Socio-economic Study, and audited certificates for Statements of Expenditure are not forthcoming from GOB. Based on the above, the RMD Engineer gave the project a 2 rating (Moderate Problems). Eleventh Supervision Mission 85. The RMD Engineer visited Bihar from August 4 to 9, 1986, to review progress. His BOR made the following points: Scope of works and ossible new extension -- Credit proceeds contain an estimated unallocated balance of about Rs. 115. Work in hand and commitments amount to about Rs. 72.5 million. "It would therefore appear inevitable to permit GOB to take up some additional road works in the form of extensions or missing links and roads left out from the main and reserved lists. This may involve further extension beyond June 1987." (para 4). Monitoring -- GOB has failed to submit the reports of the technical audit according to the dates promised (2 by February 1986 and 3 by April 1986) (para 7). Costs and finances -- "Cost of some works, according to the administrative sanctions, is not proportionate to their length or scope of work ..." (para 8). "Some cases of expenditures disproportionate to the actual work done were also pointed out ..." (para 9). "The WB Cell reported that a huge amount of expenditure (about Rs. 7.5 million) is locked up in the form of cash security deducted from the contractors' bills (para 10). Quality of work -- in some cases, land acquisition has not been completed and roads continue to be constructed on substandard 27 alignments. Even marginal improvements which could be possible within the available land width are not attempted. Attempts are not made to build the formations to full widths before laying the road surface. Riding quality of the finished surface, including proper camber and super elevation, require improvement (para 12). 86. By December 1986, supervision had been effectively assumed by the RMD Engineer.UZ/ In theory, supervision responsibility still rested with the IDA Headquarters Engineer but, in the Audit's opinion, it would have been difficult for him, even under the best circumstances, to fulfill his duties adequately, given the workload he carried. 2/ Still, in the months that followed, the two engineers did arrange to visit Bihar together, although operational decisions continued to be dealt with largely from the RMD. Twelfth Supervision Mission 87. This was conducted in February 1987 J-/ by the RMD and the IDA Headquarters Engineer together. The Aide Memoire, prepared by the Mission in the field, raises some disturbing issues: Timing -- The mission arrived at an embarrassing time for REO: they had been informed by GOB that the mission had been cancelled and, in addition, all Government officers of Grades III and IV throughout Bihar were on strike. Offices were closed and picketed by strikers. This made access to files, documents, drawings and contracts impossible ... (para 2). Works progress -- The Mission informed GOB that the Credit would formally close on June 30, 1987, but ongoing works could continue after that period if necessary, provided all disbursement certificates were received in Washington before December 24, 1987 (para 3). A2/ cf. letters dated December 23 and December 30, 1986. from RMD to Chief Engineer, REO, Patna. 23/ Consider the January 8. 1987. Terms of Reference issued to the IDA Headquarters Engineer. Between January 25 and March 2. he had to visit Pakistan (to receive the PCR for the Third Highway Project and to discuss the forthcoming negotiations of the proposed Fourth Highway Project); Sri Lanka (to discuss the PCR on the First Roads Project, and to conduct a full review of the implementation status of the Second Roads Project; and India (to discuss the start-up of the Gujarat Rural Roads Project, and to review implementation and outstanding issues related to the Bihar Rural Roads Project). 34/ There is a discrepancy in the file concerning the precise dates. The (undated) Aide Memoire left at Bihar by the two engineers, says February 17-20. The May 14. 1987. BOR on the Supervision Mission. says February 15-23. 28 Construction -- The Mission visited four roads... In one area, the 20 mm road surface was gone ... suggesting insufficient thickness, poor sweeping prior to surfacing, or pocketn of badly mixed material. Stretches of two roads were found to have been completely washed out by the September 1986 floods due to inadequate waterway (in one instance grossly inadequate) and too high an embankment (para 4). The major point of poor construction practice observed is in the manner of bringing up the earthworks. As emphasized in previous aide memoirs, the toe line of the embankment must be set out by the engineer, and rolling must take place across the full width, right up to the shoulder level. Clods must be broken up and the material c<mpactei ... The prevalent practice of placing dry clods loosely on the side slopes after construction of the pavement crust is bad practice and a flagrant waste of money (para 5), Maintenance -- The Mission was very disappointed by the maintenance situation. The assistant engineers questioned stated, quite unequivocally, that they had received no maintenance funds for either the previous or the current fiscal years, and had no permanent organization or planning procedure to carry out any maintenance work. The state of other REO and PWD roads seen on the field visit suggested that maintenance was generally poor (para 6). Audit of Statements of Expenditure (SOEs) -- GOI and GOB have been requested to give audit certificates for the expenditures claimed against SOEs since the beginning of the project. These have not been received even within the grace period of 90 days starting from October 2, 1986. IDA would stop reimbursing against SOEs and would resume after the audit certificates has been received (para 8). Technical audit -- IDA has repeatedly requested to receive technical audits on the quantity and quality and quantity of work done. None have been received so far (para 9). Credit closure and Project Completion Report (PCR) -- The Credit Agreement requires the Borrower to prepare a PCR and the Mission suggested that an immediate start be made. The latest date for receipt of this report in Washington is October 30, 1987 (para 10). Impact StUfy -- The Mission visited ANSIS to discuss progress of their report. ANSIS staff said that the report would be ready by April 1987 and that operators for computer entry of the data had been a cause for delay (para 11). 88. The May 14, 1987 Supervision Report gives the project a 2 rating (Moderate Problems). Three months had passed since the February Mission, but disbursements were still suspended pending receipt of the audit reports. The Supervision Report notes the following: Project Status -- Out of 658 km in the first batch of contracts, 560 km are complete and earthworks are complete on the balance. Contracts for a further 78 km have been let. Maintenance of project roads and of REO roads generally is poor (Section 3). 29 Action Prooosed -- Next mission will concentrate on ascertaining likely final physical and financial status of projects ensuring correct cross-section and compaction of earthworks, reinstatement of flood-damaged roads, and pursuit of technical audits (Section 5). Thirteenth Sugervision Mission 89. This was conducted without a field visit. Instead, during a brief visit to Delhi,15/ en route to Ahmedabad, the IDA Headquarters Engineer, toge1zher with the RMD Engineer, met with REO's Chief Engineer who briefed them on progress. The strike by non-gazetted officers had ended, audit certificates had been received, and disbursements could resume. 90. The REO Chief Engineer argued that in order to speed up completion of the work program of 736 km, REO proposed to do the work departmentally, i.e., by force account through petty contractors. In this way, commitments could be made by June 30, 1987, the date of Credit closure. The two IDA engineers approved this approach. The REO Chief Engineer also said that a program of maintenance had also begun and would be completed by the end of the year with equipment and materials purchased before Credit closure. He also noted that a report entitled Socio-Economic Conditions And Mobility Patterns in some Road Influence Areas* Impact Survey had been received from ANSIS. 91. On the basis of tha above, the June 19, 1987, BOR gave the project a 2 rating (Moderate Problems). Section 5 (Action Proposed) of the Implementation Summary states that the next mission would concentrate on ascertaining likely final physical and financial status of project; ensuring correct cross-section and compaction of earthworks; pursuit of technical audits, remaining SOE audit and PCR. It is possible (the BOR went on) that not all work begun would be completed by December 24, 1987, - but materials and equipment will have been purchased. The Credit would close officially on June 30, 1987. Second Extension of the Closing Date 92. The Credit should have been closed on June 30, 1987, and the file contains no explanation why it was not. About one month after the scheduled Closing Date, there was a July 28, 1987, telex from RMD to Headquarters saying that GOB and DEA would be requesting an additional one- year extension. 93. On September 1, 1987, an internal IDA memorandum recommended that an extension be granted until June 30, 1988. The foilowing reasons were advanced: Back&roun -- At the time of the last closing date extension (April 1986), progress of works had been deteriorating. "This was largely due to the transfer of responsibility for the project from the State's Road construction department to the Rural Development Department, which had a heavy workload of village road construction / No date shown on the Supervision Report. 30 provided for under various central government poverty alleviation programsw" Current status -- Since then, various measures had been adopted to restore progress to previous levels... However, despite the renewed vigor of implementation over the past year, it has not been possible to complete the project as scheduled due to a strike of the GOB non- gazetted officers which, from early January, effectively halted all work for about two months and, more recently, due to floods in Bihar, which has been inundated after prolonged rains. The resulting flooding has put a stop to all works and has damaged some embankments under construction. A total of 736 km of rural roads has been sanctioned, of which 590 km are completed to blacktop. A further 100 km had earthworks completed before the flooding, but some of this length has been damaged. Equipment to facilitate road maintenance and force-account construction is now on order and a crash maintenance program costing Rs 18.6 million to correct previously noted deficiencies is underway. Borrower performance and commitment -- GOB is in compliance with all project covenants. Maintenance of existing rural roads has been poor, but serious efforts are being made to improve the standard of work. However, the frequency and quality of technical audits need improvement, an issue that is being pursued by supervision missions. GOB is clearly concerned that this project, which has a high political visibility, should be completed as planned and without further hindrance. However, to do so, some 6-12 months of additional time is needed to allow for the receipt and payment for equipment on order, the repair of flood damage, and the completion of the sanctioned civil works. GOI has requested a second extension of the Closing Date of one year to June 30, 1987, which this memorandum supports. The Department's Country Operations Division, Loan Administration, and Legal Departments also concur with this request. 94. The Audit notes that strikes and floods are presented as major factors for delay. Also, it notes the delicate handling of numerous aspects (planning, quality of work, maintenance, monitoring), on which successive Supervision Missions had made harsh comments. 95. The speed with which the one-year extension was processed is remarkable. On the very day when the memorandum recommending approval was written (September 1, 1987), IDA telexed GOI and GOB that the extension had been granted. Further Supervision and Credit Closinjg 96. Despite the statements in the Twelfth and Thirteenth Supervision Reports that "the next mission will look into work procedures, maintenance, etc.", and the flat statement in the September 1, 1987, memorandum that "the frequency and quality of technical audits will be pursued by Supervision Missions", there were no further supervision missions at all. 31 97. The Credit was closed on June 30, 1988, and the GOB authorities were given until December 31, 1988, to clear up commitments entered into up to the time of Credit closure.16/ 98. On November 8, 1988, GOI telexed IDA to state, on behalf of GOB, that in view of the continuing strike of non-gazetted officers, filing of reimbursement claims could not be completed on time. GOB therefore requested that the project account be kept open until March 1, 1989, for processing reimbursements on commitments made up to June 30, 1988. 99. On November 17, 1989, IDA concurred but noted that no further extension would be possible and that any undisbursed balance as of March 31, 1989, would have to be cancelled. The final payment was made on April 18, 1989, and the remaining balance of SDR 97,970.41 was cancelled.IZ/ Preparation of the Project Completion Report 100. In February 1987, the Twelfth Supervision Mission had brought to GOB's attention that the Credit Agreement required the Borrower to prepare a PCR and that if manpower or other constraints made this difficult, the task could be entrusted to a consultant who could be paid out of Credit funds. GOB did not act on the suggestion. 101. On September 26, 1988, IDA issued TOR to one of its economists to proceed to India and carry out the field research necessary for PCR preparation. On October 12, 1988, IDA Headquarters telexed GOB to spell out the nature and extent of data that ought to be collected before the IDA Mission arrived. 102. Accompanied by an IDA Headquarters Engineer and by the RMD Engineer who had supervised the project, the IDA Economist visited Bihar between October 24 and November 4, 1988. By the time of the IDA Mission's arrival, the Borrower PCR ought to have been well advanced.3/ However, on arrival, the IDA Missian was handed copies of a "draft" Borrower PCR which consisted of some Faragraphs of the SAR, with no analysis. The 16/ Memorandum of July 12, 1988. az/ IDA letter dated May 18, 1989 to DEA. 2_/ Detailed guidelines for the preparation of PCRs had been handed over to the REO by the RMD Engineer in March 1987. Explicit instructions were included in Annex 2 of the project's Staff Appraisal Report. In December 1990, the Audit Mission was given a copy of Project ComDletion Report of Bihar Rural Roads Project (Credit 1072-IN). This undated document was prepared by World Bank Project Cell, Rural Engineering Organization. Rural Development Department. GOB. Patna. but was never formally submitted to IDA. The Audit studied the document but found in it no evidence that contradicts, or modifies. the findings of the PCR. 32 Impact Survey which was to have been done of a 10% sample of representative roads had not been done.I/ 103. The Mission encountered what previous Supervision Missions had repeatedly encountered too (a strike by non-gazetted officers) but was not deterred from travelling extensively throughout Bihar and visiting a number of roads built under the project. Given that the impact survey had been contracted, the Mission offered the possibility of postponing the date for submission of the Borrower PCR from December 30, 1988 (as required by the Credit Agreement) to March 31, 1989. No Borrower PCR was submitted by that time. 104. IDA prepared a draft PCR and sent it to the Borrower with the request to prepare Part II (Project Review from Borrower's Perspective). The Borrower did not prepare Part II. The PCR prepared by IDA was published in December 1989. III, FINDINGS AND CONCLUSIONS 105. The project was identified in January 1978. It was appraised in March 1980, Board approval was given in November 1980, the Credit became effective in January 1981, and the project was planned to be completed in March 1985. Project objectives were provision of all-weather access to farms and villages, and improvement of construction and maintenance procedures. Project components consisted of construction or strengthening of about 700 km of rural roads; improvement of maintenance on REO- administered roads; and procurement of road construction and maintenance equipment. 106. Institutional objectives were not met but physical results in terms of kilometers of road constructed were close to appraisal expectations. Because of design, workmanship, and supervision shortcomings, quality of the finished works is not what it should have been. Many roads, costing about US$50,000 per km (as against US$25,000 estimated at appraisal), have either failed or show premature deterioration. On many roads there is no motorized traffic. Actual project cost in dollars is US$45.7 million (as against US$52.9 million estimated at appraisal) because, over the period of implementation, the value of the Rupee deteriorated vis-a-vis the US dollar. Actual project 39/ A contract to carry out the survey had recently been signed with ANSIS but funds had not been appropriated and ANSIS had not been able to commence designing the survey and training the surveyors. The December 1990 Audit Mission was given a copy of the completed survey report: Socio-Economic Conditions and Mobility Pattern in Some Road Influence Areas: Supplementary Impact SurveX - 1989 (A.N. Sinha Institute of Social Studies. Patna. 1990). Study of this document did not lead to conclusions suggesting that the project built roads with a positive economic rate of return; that maintenance procedures have been strengthened; and that sustainability of the project benefits is assured. 33 cost in rupees is R520.6 million (as against R444.4 million estimated at appraisal) 107. Instead of March 1985, actual completion took place in June 1988. Lack of data precluded a re-estimation of the economic rate of return (ERR) which, at appraisal, had been estimated at 21%. As constructed, the roads will not last sufficiently long to generate enough benefits to justify the investment and the post-completion the ERR would probably be negative. The short life of the roads, coupled with the lack of an effective maintenance system, does not bode well for the sustainability of project benefits. 108. Preparatory work for this PPAR included study of the record and interviews with IDA staff in Washington. A December 1990 OED mission to India inspected roads and discussed the project with GOI and GOB officials. Neither the written record, nor field inspections, nor interviews suggested that the project was a success. IDA officers familiar with the Bihar Project and GOI officials in Delhi recommended, for comparison and perspective, a visit to the State of Gujarat where an IDA-supported rural roads project, conceptually similar to that of Bihar, is underway. AO/ The Government of Gujarat (GOG) was kind enough to receive the OED Mission and to offer an extensive briefing (including field visits) on its project. 109. Study of the record, interviews, and field inspections have shown that the PCR account on the Bihar Project is fair and accurate in every particular. In the following paragraphs, the Audit proposes to comment on some wider topics revealed by the Bihar experience. General Findings 110. With about 70 million inhabitants, Bihar is the second most populous State in India. It is also the poorest: about 70% of the people live below the poverty line. Furthermore, it has about 95,000 km of roads (75% unpaved), and most of the goods traffic moves on carts. ill. Being so poor, Bihar is preoccupied with employment creation. Traditionally, what can be built by hand, must. Roadwork is a familiar source of employment but quality of the finished product leaves much to be desired. It therefore seems that the project should have been more responsive to suggestions from IDA Headquarters that efforts be made to improve the efficiency of traditional construction and maintenance techniques. 112. Bihar was not endowed either -ith trained civil servants to plan projects according to IDA guidelines, or with experienced engineers to supervise them, or with small contractors who could do quality work. It therefore seems that the project should have given more attention to staff training, institution-building, and strengthening of small contractors. AO/ Guiarat Rural Roads-Proeat (Credit 1757-IN, US$119.6 million. 1987). 34 113. A REO that knows how to set priorities, engineers who know how to superviser and contractors who can use unskilled labor efficiently, cannot be developed overnight. IDA knew that building up a capacity for planning and implementing rural development projects is a long-term proposition. It therefore seems that the project should have have been designed as the first of a series and that it should have been followed by a second operation to take advantage of the experience gained under the first. 114. From the identification phase onwards, all IDA missions kept repeating this was a project that demanded unusually intensive, unusually frequent supervision. However, IDA officers were burdened with preparation of other operations and could not devote sufficient attention to the Bihar Project. 115. From the beginning, letters from India told IDA that beneficiaries distrusted the local authorities. Also from the beginning, GOB staff wanted an IDA presence during project preparation to minimize political influence on road selection and on allocation of equipment to different districts. All this notwithstanding, IDA refrained from getting involved. 116. From the beginning, IDA knew that GOB would be hard pressed to supervise physical execution of the works and that it was so short of planning staff it could not retain an economist at the project preparation phase. All this notwithstanding, much emphasis was placed on a detailed socio-economic monitoring system, even though IDA itself had not decided what data it wanted, and even though GOB and GOI never showed much interest in the research. 117. From beginning to end, a divergence persisted between what Bihar wanted and what IDA advocated. GOB was preoccupied with short-term employment, IDA with long-term development, and the two preoccupations never meshed. IDA felt that building roads merely to create jobs is wasteful. It is wiser to identify and build roads which would support rural development; to see to it that "parallel investments" (in agriculture, in credit facilities, etc.) are efficiently implemented; and to ensure that, once built, new roads are properly maintained. 118. Bihar did get some employment out of the project, but that was all. IDA did disburse most of a US$35 million credit but the roads are failing and no institutional strengthening took place. The preceding section of this PPAR documented in detail how the project was prepared and implemented. The evidence suggests three explanations why GOB and IDA never really came together. First, because of socio-political conditions in Bihar. Second, because project design was inappropriate. Third, because IDA was unwilling to allocate sufficient resources to supervision. 119. At least once a year, throughout project implementation, GOB came to a stop because of strikes and civil disturbances. This affected the pace of works as well as the spirit and attitudes of the civil service. IDA officers responsible for supervision found it hard to perform their duties. Sometimes they could not get to Patna and REO officers had to fly to Delhi to brief them. When IDA officers did go to Patna, they sometimes 35 could not.get into picketed Government offices. Extensive site visits were often impracticable and, on one occasion, IDA officers went to the field with a police escort. 120. On the other hand, even if socio-political conditions had been ideal, project design would still be inappropriate. Although most of the Bihar traffic is pedestrian, bicycle, or cart,A1/ the project concentrated on construction of paved roads, and although civil works are mostly done by hand, the project relied on equipment-based methods. Furthermore, the resources allocated by IDA for project supervision were inadequate. The first supervising officer recommended unexpected, frequent and intensive visits. Instead, there was a period when IDA did not send a supervision mission for almost eighteen months. Staff discontinuities due to reassignment or retirement, changes in the leading program, the upheaval of the 1987 Bank Reorganization, may explain, but not excuse, the reduced presence of IDA staff during a period when the project was regarded as an important intervention in Bihar and as a possible model for replication in other parts of India. 121. But why did the thought of replication arise in the first instance? Do Indian rural areas really need better road access? The answer is yes. The latest estimatesAZ/ show that India had a population of 814 million in 1988. This is expected to grow to 1,010 million in the year 2000 and, eventually, to 1,766 million of "stationary population".O./ For many years to come, agriculture will provide a livelihood for hundreds of millions and the need for better access to rural areas will remain enormous. IDA can help by facilitating provision of needed infrastructure, encouraging staff training, and strengthening of local institutions. The Bihar Project offers specific lessons which can be grouped under three headings: Inception and Preparation; Implementation and Monitoring; and Bank Group Assistance. Specific Findings -- Inception and Preparation 122. Rural road proiects are to Lmprove rural transort. not Just to build roads. The elements of rural transport are infrastructure, fleet, and users. If the fleet is inadequate, the best infrastructure is of little use. Also, AJ/ Please see. Socio-Economic Conditions and Mobility Pattern in Some Road Influence Areas: Supplementary Impact Survey -- 1989 (A.N. Sinha Institute of Social Studies, Patna. 1990). and, in particular, Chapter V: Mobility Pattern of Household Population. pp.52 ff. A2/ Please see Volume I. Economic Development Data, in India: Trends. Issues and Options (World Bank. Asia Region. India Country Department. Country Operations Division. Washington. D.C., May 1. 1990) AQ/ "Stationary Population" is the projected population level when zero population growth is achieved; i.e., when the birth rate is constant and equal to the death rate and the age structure is stable. 36 if most users live close to subsistence, they cannot afford to pay for a good infrastrOcture and an adequate fleet. Quality of infrastructure, fleet adequacy, and uper ability to pay must develop in tandem. This means that developing rural transport will take far longer than the five or six years of the average construction project. 123. Project design must be attuned to realities of time and place. When most of the traffic moves on carts, paved roads are dispensable. When much of the project area is inundated during the monsoon, bridges and river crossings are more important than roads. When unskilled labor is widely (and inefficiently) used in civil works, the emphasis must be on more efficient use of labor, rather than on equipment-intensive techniques. 124. Project design must take advantage of local expertise. Executing Agencies may be unfamiliar with plann'ng and supervision techniques to make the local administration work better. On the other hand, they are familiar with why the local administration works poorly. This knowledge must be factored into project design, and the time envisaged for implementation expanded accordingly. 125. A distinction must be drawn between physical and institutional objectives. Physical objectives can be expressed in terms of investment: kilometers of road, number of bridges, etc. Institutional objectives (staff training, modification of cumbersome administrative procedures, development of local contracting capacity, etc.) are essentially qualitative and can be achieved only in the long run. It is relatively easy to meet physical objectives. It is not so easy for a project that spans 4-5 years to meet institutional goals that require decades if not generations. 126. Covenants and conditions must relate to achievements which an Executing Agency is empovered to acccmplish. Physical objectives can be spelled out with relative precision, and their accomplishment can be within the power of the Executing Agency. It is not the same with institutional objectives which can be met only when a number of Agencies collaborate. Clear covenants, aimed at specific Agencies, would go a long way to promote achievement of complex, and long-term institutional objectives. 127. Staff training is the institutional obiective deserving the greatest attention. India does not respond well to suggestions that its staff needs exposure to better technical methods. The impression is widespread that the country is amply endowed with highly qualified personnel. At least in the case of rural road construction and maintenance, the impression is not correct. Specifically, technical staff need massive training so that they will be able to plan and supervise labor-based construction operations whose results would be competitive (in terms of both quality and cost) to work done by equipment-intensive methods. 37 128. Massive training must be accomanied by modification of career At present, there is a strong feeling among engineers that becoming involved in labor-based operations will earn them neither professional nor social status. Hence, civil engineers are unwilling to go through the rigors of acquiring on-the-job experience, an indispensable element for the formation of senior staff needed to plan and supervise effectively labor- intensive works. But availability of experienced staff is not, by itself, sufficient. Unless proper incentives are put in place, top-quality staff will avoid working in labor-based rural roads projects. Specific Findings -- Implementation and Monitoring 129. Implementation is facilitated when targets and responsibilities are precise and explicit. Good preparation spells out the volume of physical work, the time it will take to do it, and who is responsible for getting it done. Often, the Executing Agency says it can do a job without considering what has to be done by other Agencies. The Bihar Project has shown that land acquisition may inhibit implementation of an otherwise simple roadbuilding job because the Executing Agency failed to take into account how well other Agencies would cooperate. 130. Flexibility in adjusting targets and Closing Dates is desirable but must not be abused. In the course of implementation, obstacles are bound to emerge: floods, budget cuts, problems with staffing and procurement. Appropriate adjustment of targets ought to be made with a minimum of fuss.AA/ But, by the same token, if implementation suffers because of consistently poor performance of the Executing Agency, extension of the Closing Date must be denied. AJ/ The Gujarat Rural Roads Project offers a good example of appropriate flexibility. In early 1991, GOG felt that its plans to improve the rural network under the ongoing Rural Roads Project in seven Districts could accommodate (thanks to savings realized because some works done by other Agencies, as well as because of favorable foreign exchange variations) the most urgently needed works in five or six additional Districts. The actual volume of extra work would depend on the volume of available Credit funds. To follow up on this, GOG proposed to prepare a new Rural Roads Project intended to cover the balance of works in the five or six Districts mentioned above. and to initiate work in the remaining Districts of the State. 38 131. There are two t=pes of oroect monitoring. The first deals with ex-post evaluation of individual roads. Experience from Bihar (but, also, from Kenya, Kalawi and Benin) suggests that 5-10 km roads are too short for ERR analyses. It might be preferable to concentrate on least-cost solutions for construction and maintenance and leave it at that.Aj/ 132. The second type of .rolect monitoring deals with road impact upon the rural economy. kxperience from Bihar suggests this is methodologically difficult and financially expensive. Only when the Borrower is keenly interested in socio-economic and anthropological research, should future projects devote resources in amassing further information. Seific Findinzs -- Bank Group Assistance 133. The Bihar Prolect took too lome. When it was finished. there was no follow-up o2eration to make use of the exerience gaed. From identification to completion the Bihar Project took ten years to build 700 km of roads. There has been no follow-on operation where lessons of experience might have been applied with respect to project design, choice of technology, and intensity of supervision. India-wide, population growth demands agricultural development which, in turn, demands that rural transport be improved at a faster pace and at a much expanded scale. The problem is that there are not enough domestic resources, and that current administrative practices are not geared for smooth acceleration. 134. GO! is faced with a choice rogarding-technical assistance. Depending on what 90! decides. the Bank Grouo will have to reconsider its own .lans for future suport. Either GOI will continue to insist that Indian administrative and technical services have no need of foreign techniques, or it will acknowledge that experience from abroad can be useful. Whichever the direction, GO's posture ought to determine the Bank's posture. If GOI remains attached to its current position, the Bank Group ought to reconsider whether it should continue with interventions like the unsuccessful Bihar and the apparently more successful Gujarat Projects. LS/ The Audit wishes to point out (because it underscores the importance of training) that agreeing on what constitutes the least-cost solution is not easy. IDA and Indian engineers have different perceptions of good finished quality and a meeting of minds has not yet occurred. 39 135. 001. GOB. and the Bank Grou may have to reconsider their respective stances regarding ways and means to strengthen rural 'transport Assuming that 001 subscribes to the proposition that better rural transport is a prerequisite for accelerated agricultural development, the Audit believes that GOI, GOB, and the Bank Group must reconsider their respective stances. GOI ought to modify its policy regarding technical assistance. GOB specifically (and other States in general) ought to think less in terms of immediate gains and more in terms of long-run development. The Bank Group ought to reconsider its current practices for staff time allocations. 136. If these three conditions are met, the Bank Group will be in a position to play a more substantive role in strengthening rural transport. In the first place, large financial resources will be required to improve rural transport, even in a small number of States. In fact, they will exceed whatever resources GOI, the States and the Bank Group will be able to mobilize. Hence, co-financing will be indispensable and the Bank is the most experienced multilateral development agency in this domain. In the second place, the flow of technical assistance for the modernization of administrative machines (and, especially, the mobilization of resources for staff training) is something complicated, and expensive enough to deserve the most careful planning. Again, the Bank possesses a comparative advantage which could be of value to India. 137. If the Bank Group is to get seriously involved in strengthening rural transport in India. aorooriate staff time allocations will have to be made- Thorough co-financing arrangements and technical assistance planning will be time-consuming. Without a clear understanding of the direction GOI proposes to follow, it is not possible to say exactly how much time, and what kind of staff the Bank Group will have to allocate for preparation. One thing is certain: current allocations are barely sufficient for adequate monitoring even of simple operations such as the Gujarat Rural Roads Project. Recommendations 138. In order to clarify how things stand and where they might go, the Audit recommends the following: * (a) By, say, end 1991, Bank Management discusses with GOI the latter's intentions regarding strengthening of rural transport over the next 10 years. 40 (b) If GOI decides to modify its policy towards technical assistance, the Bank Group ought to present GOI, by June 1992, with tentative estimates for assistance needed to improve rural transport in a number of States to be identified by GOI. (c) If GOI agrees that the Bank ought to perform the role of coordinator for financial and technical assistance, the Bank Group ought to present GOI, by December 1992, with a detailed plan for interventions to improve rural transport in the States specified by GOI. 41 ANNEX 1 LABOR-BASED CONSTRUCTION AND MAINTENANCE METHODS NOTES ON THE BANK'S RESEARCH WORK 1. In the late 1960s, rapid population growth threatened that many developing countries would soon be confronted with rural-urban migrations, decreased standards of living, and political instability. The scope for industrial job creation was limited and, at the same time, agricultural output and social facilities in the rural areas had to cater for an increasing number of people. This would not be easy. Rural areas were not properly equipped, and what infrastructure did exist could not easily support additional numbers. Building roads, irrigation works and other necessary facilities would be expensive and though local currency was not abundant, foreign exchange was truly scarce since most developing countries did not generate enough exports. No matter how large they might be, aid flows could not possibly cover the enormous cost of purchasing and operating the mechanical equipment that would be needed to improve infrastructure. Still, infrastructure had to be built up before rural economies could feed and employ more people.j/ -2. Such were the demographic and economic reasons which prompted the Bank to explore the technical and financial feasibility of "more appropriate" construction methods. The starting point was a supposition: Assume a country with a lot of labor and little capital. Assume that it needs to build and maintain civil works in rural areas. Assume further that construction and maintenance methods must rely more on unskilled labor and less on mechanical equipment. Can the final product be competitive in terms of cost, speed of completion, technical quality, and public acceptability? 3. The supposition did not suggest to abolish but merely to reduce dependence on equipment. Even so, and despite the potential advantages of labor-based methods (employment generation; income distribution; greater reliance on domestic resources; reduction of foreign exchange expenditure for the purchase of machinery, spares and fuel), four considerations discouraged their adoption: j/ For a realistic evaluation of the potential of public works to absorb unemployment, construct needed physical facilities, and improve the living standards of low-income groups. please see Burki, S.J.0 Davies. D.G., Hook. R.H.. and Thomas, J.W., Public Works Programs in Developing Countries: A Comparative Analysis (IBRD Staff Working Paper No 224. Washington. D.C.. 976). 42 (a) as traditionally practiced, labor-based methods were not efficient; they did not make the best use of workers they employed cnd, consequently, were not financially competitive with equipment-based methods. To make them competitive much would have to be done differently in terms of project planning and management, local manufacture of better tools, adoption of incentive payment systems, etc. However, all such changes would adversely affect established sinecures; (b) political and technical leaders in many developing countries felt offended by the suggestion that they ought to sponsor "backward" methods; (c) compared with equipment-based methods, labor-based techniques were more management-intensive, and the one skill that is really scarce in the Third World is management; and (d) it was easier for developing countries to get aid funds for an equipment-based construction program than for a labor-based one. 4. In 1971, the Bank launched The Study of Labor and Capital Substitution in Civil Engineering Construction in order to develop and disseminate nonstruction and maintenance methods that would be more appropriate for tne resource endowment of Third World countries.2/ Over a ten-year period, this Study encompassed research, experimentation, physical construction and monitoring of rural works in Asia, Africa, Latin America and the Caribbean.j/ 5. One finding that emerged fairly quickly was that labor-based methods may be technically and economically feasible and still not be successfully implemented. Efficient large-scale application of such methods demands administrative and organizational arrangements enjoying the full support of a country's technical and political leadership. Actual construction work in a variety of terrain and labor supply conditions permits a real-life feasibility test. Equally important, they help create 2/ That the Study objectives reflected concerns widely shared in the development community is evident from the fact that the Study was financed by the Governments of Canada. Denmark. Finland. the Federal Republic of Germany. Japan. Norway, Sweden. the United Kingdom and the United States, as well as by the Bank. 1./ Technical conclusions from fieldwork were published between January 1975 and June 1976 in 26 Memoranda whose topics range from comparisons of qlternative designs for wheelbarrows to aggregate productfin. Research and implementation work are described in ThU Study of Labor and Capital Substitution: Report on the Bank-Sponsored Seminars in Washington. Cologne. Copenhagen. London and Tokyn (Washington, September 1978). The final outcome of the Bank's work has been summarized in Labor-based Construction Programs: A Planning and Management Handbook (World Bank. Oxford University Press, 1983). 43 the institutional and management stricture for an eventual expansion to full-scale programs. "Demonstration' projects supported by the Bank included road and irrigation works iv Kenya, Honduras, Chad, Lesotho, Benin,./ the Dominican Republic, and Mia lvi.V &/ For a detailed review of IDA-assisted labor-based operations in Benin, see the PPAR on Benin - First and Second Feeder Roads-Proiects (OED Report No. 6105 of March 21, 1986). j/ A full description of the demonstration project is given in Appendix 4 (Malavi: Action plan for a labor-based pilot project) of L4hor- based Construction Programs: A Planning and Management Handbook. op.cit. For a detailed review of IDA-assisted labor based operations in Malawi, see the PPAR on Malawi - Fourth Highway Project (OED Report No. 7568 of December 30. 1988).