India - Kandi Watershed and Area Development Project (Punjab)

Document of The World Bank FOR OFFICIAL USE ONLY M4CRO E[1CE (OPf Report No. 10127 PROJECT PERFORMANCE AUDIT REPORT INDIA KANDI WATERSHED AND AREA DEVELOPMENT PROJECT (LOAN 1897-IN) DECEMBER 16, 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. EXCHANGE RATES Currency Unit Indian Rupee (Re) Appraisal Year Average (1981) Re 9.1 - US$1.00 Completion Year Average (1988) Re 14.9 - US$1.00 WEIGHTS AND MEASURES Metric System ABBREVIATIONS ERR - Economic Rate of Return GOI - Goverment of India GOP - Government of Punjab NDO - World Bank New Delhi Resident Mission DED - Operations Evaluation Department PCR - Project Completion Report PPAR - Project Performance Audit Report FISCAL YEAR Government of India/Government of Punjab April 1 - March 31 FOR OFFICIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A. Office of Director*Ceaerl Opefations Evaluatkn December 16, 1991 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Performance Audit Report on India - Kandi Watershed and Area Development Project (Loan 1897-IN) Attached, for information, is a copy of a report entitled "Project Performance Audit Report on India - Kandi Watershed and Area Development Project (Loan 1897-IN)" prepared by the Operations Evaluation Department. Attachment This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authoritation. FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT INDIA KANDI WATERSHED AND AREA DEVELOPMEUT PROJECT (LOAN 1897-IN) TABLE OF CONTENTS Page No. Preface . . . . . . . . . . . . . . . . . . . . . . .i Basic Data Sheet . . . . . . . . . . . . . . . . . . . . . iii Evaluation Summary . . . . . . . . . . . . . . . . . . . . vii PROJECT PERFORMANCE AUDIT I. BACKGROUND . . . . . . . . . . . . . . . . . . 1 II. IMPLEMENTATION EXPERIENCE . . . . . . . . . . . 5 III. PROJECT OUTCOME . . . . . . . . . . . . . . . . 8 IV. FINDINGS AND ISSUES . . . . . . . . . . . . . . 9 A. Pilot Project Perceptional Differences . 9 B. Dholbaha Dam Problems . . . ...... 11 C. The Positive Role of Foreign Consultants . . . . . . . . . . . . 12 MAP IBRD No. 13695R - Kandi Tract This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. PROJECT PERFORMANCE AUDIT REPORT INDIA KANDI WATERSHED AND AREA DEVELOPMENT PROJECT (LOAN 1897-IN) PREFACE This is a Project Performance Audit Report (PPAR) on the Kandi Watershed and Area Development Project, involving a World Bank Loan in the amount of US$30.0 million (m) to the Government of India (GOI). The project's objective was to lay the basis for sound investment in the future to reverse man-made ecological degradation in the lower foothills of the Himalayas. The loan was approved on September 12, 1980, and became effective on November 18, 1980. The Closing Date of March 31, 1986 was extended to March 31, 1987 and ultimately to March 31, 1988. Final disbursement was made on October 14, 1988 at which time an undisbursed US$3.9 m of the loan was concelled. The PPAR is based on the Project Completion Report (PCR) of the project prepared by the Asia Regional Office and issued in 1990,1 the Staff Appraisal and the President's Reports, the loan documents, the transcript of the Executive Directors' meeting at which the project was considered, on a study of project files, and on discussions with Bank staff. An Operations Evaluation Department (OED) mission visited India in November 1990, and discussed the effectiveness of the Bank's assistance with GOI and Government of Punjab (GOP) staff. Their kind cooperation and valuable assistance in the preparation of this report is gratefully acknowledged. The PCR provides a good account and assessment of the project experience, and discusses the performance of the Bank and the project executing agencins. The PPAR elaborates on particular aspects such as the perceptional differences between Bank and Borrower regarding project objectives, imperfections in the prequalification and selection procedures for civil works contractors, and the constructive role of foreign consultants. Following standard OED procedures, copies of the draft PPAR were sent to the Borrower for comments. No comments were, however, received. 11 Project Completion - sport, India - Kandi Watershed and Area Development Project (Loan 1879- IN), Report #8366, dated February 9, 1990. PROJECT PERFORMANCE AUDIT REPORT INDIA KANDI WATERSHED AND AREA DEVELOPMENT PROJECT (LOAN 1897-IN BASIC DATA SHEET KEY PROJECT DATA Actual as Actual or 2 of Appraisal Current Appraisal Estinate Estimate Estimate Total Project Costs (US$ million) 60.8 48.4 80 Loan Amount (US$ million) 30.0 26.1 87 Economic Rate of Return 16 8 - CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS 1981 1982 1983 1984 1985 1986 1987 1988 Appraisal Estimate (US$ million) 1.8 6.8 14.4 22.4 30.0 30.0 31.0 30.0 Actual (US$ million) 1.0 2.6 6.3 9.3 12.7 16.6 21.1 26.1 Actual as % of Appraisal Estimate 56 38 44 42 42 55 70 87 Date of Final Disbursement: November 14, 1988 PROJECT DATES Original Plan Revisions Actual First Mention in Files 02/77 - - Negotiations - - 10/78 Board Approval - - 07/80 Signing (Loan Agreement Date) - - 09/80 Effectiveness - - 09/80 Closing Date 03/86 - 03/88 - iv - STAFF INPUTS (staff weeks) FY75 FY76 FY77 FY78 FY79 FY80 FY81 FY82 FY83 FY84 FY85 FY86 FY87 FY88 Total Identific/ Preparation 1.3 16.6 7.2 35.9 - - - - - - - - - - 61.1 Appraisal - - - 64.9 27.5 7.6 .9 - - - - - - - 100.9 Negotiation - - - - 6.6 7.3 1.0 - - - - - - - 14.9 Supervision - - - - - - 12.5 11.9 14.8 13.4 14.0 37.2 28.0 27.0 158.8 Other - - - .1 .5 - .2 .2 - - - - - - .9 Total 1.3 16.6 7.2 100.9 34.6 14.9 14.6 12.1 14.8 13.4 14.0 37.2 28.0 27.0 336.6 MISSION DATA Date No. of Specializatiops Performarce Types of Mission _(moLy7r Persons Represented Rating Trend Problems 1/ Identification 06/77 3 F,A,L Preparation Pre-appraisal 02/78 3 F,A,E Appraisal 04/78 6 F,A,E,S,I,L - Supervision I 12/80 2 I,A 1 2 M,F Supervision II 05/81 2 1,A 3 2 T,M,F Supervision III 02/82 3 I,A 2 1 T,M,F Supervision IV 12/82 2 lA 2 1 T,F Supervision V 05/83 1 A 2 1 T,F Supervision VII' 07/84 2 -- - Supervision VII 10/84 3 A,F,I 2 2 T,F Supervision VIII 07/85 5 A,31,W 2 - - Supervision IX 03/86 4 2A,I,W 2 - T,M Supervision X 09/86 5 3A,I,W 2 - T Supervision XI 03/87 5 2A,I,F,S 2 - M Supervision XII 10/87 5 3A,21 2 - T,M Supervision XIII 04/88 5 2A,I,X 2 - T Supervision XIV 06/88 2 21 - - T Supervision XVl 02/89 3 E,I,W 1l A - Agriculturalist/soil conservationist; E = Economist; F - Forester; I = Irrigation Engineer; S = Soil Conservationist; W = Watershed Management Specialist; X = Extension Specialist; L = Livestock Specialist. k1 1 a Problem free or Minor Problems; 2 = Moderate Problems; 3 - Major Problems; 4 = Major Problems, not being addressed adequately. I - Improving; 2 - Stationary; 3 - Deteriorating. 4l F - Financial; M - Managerial; T - Technical. 1l No supervision reports on file. V PCR mission. OTHER PROJECT DATA Borrower: Government of India (GOI) and State of Punjab (GOP) Executing Agencies: Department of Forests Soil Conservation and Engineering Department Department of Irrigation Department of Agriculture Department of Animal Husbandry Department of Fisheries Punjab Agricultural University Follow-on Projects: Name Himalayan Watershed Management Loan Number 2295-IN Amount (US$ million) 46.2 Approval Date 06/83 Name Pilot Project for Watershed Development in Rainfed Areas Credit Number 1424-IN Amount (US$ million) 31.0 Approval Date 12/08/83 Name Punjab Irrigation and Drainage Credit/Loan Number 2076-IN/3144-IN Amount (US$ million) 165.0 Approval Date 12/14/89 Name Integrated Watershed Development (Hills) Credit Number 2100-IN Amount (US$ million) 75.0 Approval Date 03/90 Name: Integrated Watershed Development (Plains) Credit Number 2131-IN Amount (US$ million) 55.00 Approval Date 05/15/90 - vii - PROJECT PERFORMANCE AUDIT REPORT INDIA KANDI WATERSHED AND AREA DEVELOPMENT PROJECT (LOAN 1897-IN) EVALUATION SUMMARY 1. Introduction watersheds, to fisheries and horticulture development, gravity and Concern at the degraded condition groundwater irrigation, and flood of the severely eroded hill lands in protection in the lower watersheds. the Himalayan sub-mountainous zone of northern India, known as the Shiwalik 3. ImRlementation EMperience Hills, and at the low income levels and poor quality of life of the Problems in selecting and inhabitants, lay behind the decision appointing consultants, which was a to undertake a pilot rehabilitation condition of Board Presentation, led and development project in the Punjab to a 20-month delay between portion of the zone, known as the negotiations and effectiveness and Kandi Tract. The project, estimated this had a negative impact on project to cost US$60.8 million (m) including momentum in the first year following a US$30.0 m Bank loan became effectiveness, when budgets and effective in November 1980 and was staffing levels were far short of closed in March 1988. appraisal expectationc. Gradually, the upper watershed compo~nents and 2. Oblectives and Setting the horticultural developments in the lower watersheds were implemented and The objectives were to lay the completed satisfactorily, and in some basis for sound investments for cases in excess of the physical reversing man-made ecological targets set at appraisal, after a degradation in the Shiwalik Hills and two-year extension of loan closing. protecting and developing the The irrigation component on the other adjoining agricultural lowlands hand was beset by problems, partly of subject to erosion and flooding. The design but mainly caused by poor project was designed as a pilot performance on the part of an initiative covering five watersheds inexperienced contractor. but was extended to a further six upper watersheds on the basis of the 4. Results project consultant's recommendations, following a mid-term review. The pilot objective of testing and Complementary investments demonstrating sound investment demonstrated or tested ranged from possibilities for reversing man- afforestation, through structural and induced damage in the Kandi Tract vegetative methods of soil were satisfactorily realized in the conservation and water harvesting, upper watersheds through the fodder grass establishment, afforestation, livestock and research livestock, and rainfed farming station sub-components. The more improvement packages in the upper promising dryland rehabilitation or - viii improvement activities have been monitoring and periodic economic shown to be practical and are now analysis of beneficiary acceptance being replicated through other means, rates of the project-promoted including particularly the IDA- investment opportunities is urged as assisted follow-on Integrated a continuing responsibility of the Watershed Development (Hills) Project GOP, in order to confirm their true (Credit Number 2100-IN) which covers validity. four states, including Punjab. Further support for the Ballowal Saunkhari research station provided 6. Findings and Lessons under the project should increase the station's catalytic role through the Although the SAR described the addition of training facilities project as a pilot, aimed at testing serving staff from all four states. and demonstrating options for rainfed development of the upper watersheds, On the other hand, the successful it was in reality a compromise with pilot activities were overshadowed by the GOP's firm commitment to the disappointing outcome of the conventional atorage dam and gravity financially larger and problem- irrigation developments in the lower plagued experiences and outcome of watersheds (paras. 1.13-1.14). The the irrigation-based investments in conceptional differences were a cause the lower watersheds. The for delay in preparation (para. 4.4) unsatisfactory re-estimated ERR of and some confusion and only 2% for this component reflects misunderstanding during the high costs of the storage dams implementation (para. 4.6). The Bank and command area developments apparently misjudged the extent to relative to agricultural benefits which GOP could be persuaded to generated through the irrigation embrace the pilot concept. A facilities. smaller, less costly, more focussed project designed exclusively as a 5. Sustainability pilot could have been more manageable, provided greater GOP and the Bank jointly displayed flexibility in operation, permitted a commendable initiative in tackling more det&iled and accurate monitoring the immense problems of man-induced and evaluation, and better enhanced environmental deterioration in the the prospect of an earlier Shiwalik Hills through direct public- conclusion, leading to more rapid sector interventions of the type replication of the positive findings demonstrated through this pilot in follow-on projects (para. 4.7). project. Sustainability of these investments is uncertain, and will The civil works contractor depend heavily on the maintenance of appointed for the construction of the GOP financial support, the continuing largest project-financed storage dam interest of the departmental proved to be financially and operating staff in the field, and the technically weak, performed extremely prolongation of beneficiary poorly, and was able to complete only enthusiasm after the initial 202 of the agreed works by the expiry subsidized support is withdrawn. date of his contract. The choice of While the outlook at the time of contractor in this instance was audit was encouraging in this unfortunate, and the experience respect, the long-term outcome is points towards imperfections in the difficult to predict.11a Careful GOP prequalification and selection - ix - procedures as they operated in that through their detailed planning instance (para. 4.13). activities. Insistence by the Bank on opening the selection process to A foreign consultant team foreign competition (unusual in the performed an important dual role by Indian context) was justified by the supplementing the earlier, inadequate quality of the consultant teams' preparation efforts, and contributing parformance (para. 4.16). significantly to the pilot objectives 1. A recent supervision mission of the follow-on Integrated Watershed Development Project - Hills (Loan 3175/Credit 2100-IN) confirms that GOP has maintained its financial support, and that the interest of local operating staff continues at a high level. PROJECT PERFORMANCE AUDIT REPORT KANDI WATERSHED AND AREA DEVELOPMENT PROJECT (LOAN 1897-IN) I. BACKGROUND Context 1.1 The Kandi Tract is an agriculturally backward and relatively sparsely populated area along the northeastern border of the State of Punjab (SOP). It is part of the strongly dissected Himalayan sub-mountainous zone of northern India and Nepal, known locally as the Shivalik Hills, and is fringed by a narrow band of coarse grained outwash sediments deposited along the southern edge of the Shiwaliks by numerous streams where they emerge from the hills into the agriculturally rich plain of Punjab (see map). 1.2 The natural cover of acacia and pine forest with its undercover of shrubs and grasses was largely destrc:ed by settlers in the latter part of the last century. Unrestricted tree felling and overgrazing led to accelerated erosion of the weakly structured soils of the watershedslV by heavy rains, and the former perennial clear-running streams became seasonal torrents in ever- deepening gullies. The flood waters deposited their coarse-grained loads in f an- like fashion along the edge of the plain, overlaying the rich agricultural soils and ruining thousands of hectares of valuable land. 1.3 Attempts, commencing from the start of the present century, to ameliorate the situation by legislation, which aimed to close certain areas to grazing, control tree felling, and restrict livestock, cultivation and quarrying failed lurgely for lack of enforcement, and further degradation continues under pressure from rising human population and livestock numbers. 1.4 Concern on the part of the Government of India (GOI), GOP and the Bank for advancing the income levels and overall quality of life of the inhabitants of this rainfed area through the introduction of environmental improvements prompted the de',ision to proceed with this project. It was intended to be consistent with a wider effort on the part of GOI, the International Crop Research Institute for the Semi-Arid Tropics (ICRISAT), the Food and Agricultural Organization (FAO) and the Bank to address comprehensively through a pilot operation the problems of rainfed farming in several areas of the sub-continent with rainfalls in the 800 to 1,200 mm range. The term "watershed" as used in this report conforms to local terminology. It refers to the upper catchment or basin of a river or stream rather than the alternative definition of a line of separation between waters flowing into different basins. Objectives 1.5 The primary objective was to lay the basis for sound investments on a larger scale in the future for reversing man-made ecological degradation in che Shiwalik Hills and protecting and developing agricultural land that is presently subject to serious erosion and flooding. Desi&n 1.6 The project was designed as a five-year pilot project with a focus on institution building and emphasis on strengthening GOP's capability to plan, implement and monitor similar projects in the future. It comprised a package of complementary investments, most of which had been tried at one time or another in Punjab, but which had failed to bring about expected results because they had been attempted in isolation. 1.7 The basic units for project planning and implementation were five watersheds, chosen as representative of the 21 major and 120 smaller watersheds within the Punjab Shiwaliks. Only one watershed (Dholbaha) had been planned with full engineering design for a storage dam by the time of appraisal and the remaining four were to be selected, and works planned, during the first two years of project implementation. 1.8 Activities to be supported included (i) rehabilitation in the upper watersheds to reduce runoff and erosion through soil conservation and revegetation, and (ii) flood protection and irrigation of farmland in the lower parts of the watersheds, below the hills. Development packages appropriate to the rainfed and irrigated farmers and livestock owners in the two different agro- ecologic situations were an important part of the project. Organization 1.9 Overall responsibility for project implementation rested with the Financial Commissioner, Development in GOP, assisted by a Project Coordinator heading up a small Program Planning and Coordinating Unit. The GOP agencies involved included the Departments of Agriculture (specifically, Horticulture), Animal Husbandry, Fisheries, Forestry, Irrigation and Soil Conservation and Engineering. The Punjab Agricultural University was given responsibility for establishing and operating an agricultural research station within the project area, and for evaluation. Finance Plan 1.10 The project cost at appraisal was estimated at US$60.8 m of which the Bank was to provide US$30.0 M, or about 49% of total cost. Bank funds were allocated for civilworks, equipment and vehicles, planting materials, livestock, consultants' services and overseas training, and salaries of incremental project staff. About 55% of the base project cost was allocated to the flood control and irrigation component. -3- Pre-implementation Processing 1.11 Early preparation work was carried out by GOP in 1976. The project as initially prepared by GOP staff was described as an integrated land and water development project. Irrigation featured strongly in this proposal, which comprised three components: (i) rehabilitation in five Shiwalik Hills watersheds, including low storage dams and associated command area developments in the middle watershed; reforestation, pasture improvement and construction of check dams in the upper watersheds above the storage areas, and channelization of the river courses (choes) as a protection against floods and installation of irrigation tubewells in the lower watersheds below the command areas; (ii) a reclamation scheme along the Sutlej River, which flows through the southern part of the Kandi Tract, aimed at reclaiming about 2,800 ha for agriculture and preventing flooding on a further 3,200 ha; and (iii) extending an existing irrigated area (the Shar Nehar Scheme) from 30,000 ha to 80,000 ha through the addition of supply canals from an existing dam. 1.12 A Bank project identification mission which visited Punjab in April 1977 recommended dropping the Shar Nehar Scheme on the grounds that unresolved issues of interstate water rights and power allocation would delay its implementation and that it might justify a separate project in its own right at a later date. The mission accepted the other two components of GOP's proposal as a basis for a Bank-assisted project, while emphasizing watershed rehabilitation as the core component and the Sutlej River reclamation component as a subsidiary but complementary component. 1.13 Project preparation by a GOP team continued through the remainder of 1977, with support from the World Bank New Delhi Resident Mission (NDO) staff, a Headquarters irrigation engineer and a grasslands development consultant. The gist of much of the Bank's advice during this phase of preparation was for a relatively small, carefully monitored project with a pilot demonstration or applied research bias, confined to the Shiwalik Hills rehabilitation component (para. 1.11 above) and limited to only one or at the most three watersheds, at least one of which would not include a storage dam. This preference for a small operation on the part of the Bank arose from an awareness that many aspects of the project, particularly the irrigation components, might be difficult to justify on economic grounds, or, in the case of the upper watershed activities, would involve heavy subsidies which could make large-scale replication difficult. These reservations were not shared by GOP which took the view that the storage dams and other irrigation components should be essential elements in the plans for each watershed, and were justifiable in a socio-economic sense, particularly if complemented by the soil conservation and productive land management activities proposed for the upper watersheds. 1.14 The conceptual difference as to whether irrigation storage dams should be essential features and central to each project watershed plan (the strongly held GOP developmental approach), or whether the land rehabilitation activities in the upper watersheds combined with channelization of the choes in the lower watershed, might be justified in isolation but required testing through an actual pilot project experience (the Bank's contention), persisted throughout the preparation and appraisal phases. 1.15 A pre-appraisal mission visited Punjab in February, 1978. The mission reported that preparation of a projeci including rehabilitation in five major watersheds and the Sutlej River component was going quite well. At the same time it identified a large number of items deserving fuller treatment by the GOP preparation team prior to appraisal. Upon consideration of the pre-appraisal mission's report the Bank decided not to finance the Sutlej River component on the grounds that its inclusion would make the project unnecessarily complex, and that it would not contribute towards the basic project objective of watershed rehabilitation and flood control. The Bank at this time also reiterated its preference for a smaller project emphasizing forestry and pasture development in the upper watersheds and flood protection principally through choe canalization in the lower watersheds, while limiting the storage dams and irrigation facilities to only two or, at the most, three watersheds. 1.16 Appraisal followed in April through May 1978. The appraisal mission concluded that the earlier assumption that the irrigation engineering plan for the lower Dholbaha watershed could be readily modified for replication throughout the Shivalik Hills watersheds was incorrect. Since site conditions were unique to each choe and there were few good sites for dams, irrigation developments in the lower watersheds would require considerable investigation and planning for each specific site, which was beyond the capacity of the GOP agencies concerned and would require the assistance of consultants. It was estimated that an internationally recruited consultant team would be needed for a period of two years to prepare adequate plans for the other four watersheds proposed by GOP for development under the project. 1.17 The project as appraised included a component for soil conservation, afforestation and grassland development over about 24,000 ha in the upper parts of the five project watersheds. The appraisal mission expressed confidence that these upper watershed rehabilitation activities, which were to include vegetative and stone check dams, channel grade stabilizers, small drop structures, crate wire dams and debris basins in gullies as well as reforestation and grass planting over the slopes, would be implemented satisfactorily because they would have the full participation of farmers and livestock owners even though the areas would be closed to grazing and unauthorized extraction of fuelwood and timber. Their support would be assured through a variety of development packages including stall-fed cattle, poultry and pig keeping, land leveling of their small cultivated patches, and improved social services including schools, dispensaries, potable water and roads. 1.18 The appraisal recognized that the flood protection and irrigation component in the lower watersheds was more problematic. It was acknowledged that the Dholbaha watershed plan, which included an irrigation storage dam, represented a reasonable investment with an expected economic rate of return (ERR) of 16%, but that the irrigation storage dam-based solution favored by GOP could be sub-optimal in other watersheds. In order to ensure that other options were adequately studied and that appropriate plans, including engineering designs, were prepared in a timely fashion, the Bank therefore decided that appointment of the internationally recruited consultant team (para. 1.16) should be a condition of Board Presentation. - 5 - 1.19 The appraisal mission's decision, upheld by the Bank, to support the GOP proposal for including developments in five watersheds under the project did not go unchallenged within the Bank. One commentator foresaw problems in obtaining commitment from GOI/GOP in the recruitment of foreign consultants, while several others expressed concern that the Bank was backing down from its earlier position of favoring only one, or possibly two or at the most three, watersheds, and doubted that all five watersheds could be planned and developed within the proposed five year time-frame. The Central Project Staff Irrigation Adviser even went so far as to note "...that the status of designs and cost estimates for the four watersheds without detailed plans is that the planning and design must be started anew, with consultants, and that this part of the project clearly does not meet normal Bank appraisal standards." He opinioned that he could "...see no evidence which could give (him) any assurance that the work could be designed and executed in the four watersheds within five years." 1.20 Loan negotiations took place in Washington in October 1978. The source and composition of the consultant team proved to be a major discussion point. The Bank felt that a competent team could best be provided by a well experienced foreign consultant firm while the Indian delegation preferred the option of recruiting the five or more team members on an individual basis or through a local firm. Agreement was reached that the consultants could be individually recruited locally by GOP if a mutually acceptable arrangement for back-up services could be provided. However, if the individual recruitment approach was to delay Board Presentation beyond March 1979, GOP would initiate action to recruit through a qualified local or foreign consulting firm. 1.21 Subsequent to negotiation GOP attempted to recruit individual consultants locally, but was unsuccessful. Eventually, in June 1980 and 20 months after loan negotiations, a contract was signed with a foreign consultant firm and the project was presented to the Board the following month. Several executive directors commended the project as a useful initiative for tackling the huge problem of environmental degradation in the Shiwalik Hills. Loan signing took place in Septemter. II. IMPLEMENTATION EXPERIENCE Start-up 2.1 The project was declared effective in November 1980, and the first supervision was carried out in December. The mission judged the start-up to be somewhat problematic, with the majority of new staff positions not yet sanctioned and with the FY80/81 budget allocations for most agencies involved in the implementation substantially below the targets agreed at appraisal. Inter-agency coordination and integ:ation of various agency work programs with the project plans was noted as weak. Delegation of more authority to the head of the Program F.'. .ning and Coordination Unit was recommended as an important step toward accelerating project momentum. 2.2 A second supervision five months later repeated the earlier complaint that budgets remained well below appraisal targets, while many incremental staff positions had still to be sanctioned. The contractor responsible for the - 6 - Dholbaha Dam construction (para. 1.7) was encountering problems, while the consultants were reported to be getting very poor and inadequate support from GOP. The irrigation and flood control component was reported as looking economically marginal at best based on the preliminary findings of the consultant. The project was rated as having major problems of a technical, managerial and financial nature. Sequence and Design Changes 2.3 The third supervision, which was carried out in February 1982, reported considerable improvement, at least in the case of the upper watershed activities. The 1981/82 target for the afforestation component had been achieved and the survival rate of tree platings substantially above that of the previous season, while controlled grazing being effected through the livestock component (reformulated shortly after effectiveness) was demonstrating the value of improved vegetative ground cover in reducing erosion. On the other hand, the irrigation and flood control component in the lower part of the Dholbaha watershed continued to encounter serious problems. The civil works contractor was proving to be financially and technically weak. Serious shortcomings in the original plan for the command area had been revealed upon the realization that a major lined canal associated with an earlier scheme was being constructed through the middle of the area which was capable of providing water to some 40% of the intended irrigation command. A redesign and revised implementation schedule and cost estimation exetcise was therefore initiated. 2.4 The Dholbaha command area had been replanned by the time of the fourth supervision in December 1982. This supervision constituted a mid-term review and afforded an opportunity to assess the extent to which appraisal objectives were being met, and to revise the project scope, scale and timing on the basis of experience during the first two years of implementation. The review benefitted considerably from the work of the consultant team, which at that time had completed field work on ten watersheds and produced final feasibility reports for five. It was concluded that the upper watershed activities, suitably modified on the basis of experience since appraisal, were progressing satisfactorily and could be expanded selectively into additional areas with the expenditure of residual funds. The irrigation and flood control needs or opportunities within the lower watersheds other than Dholbaha, on the other hand, were likely to be more modest in scope and cost than envisaged at appraisal. 2.5 Based on preliminary conclusions from the consultant's studies the mid-term review mission questioned the project concept as presented at appraisal. The seriousness of flood control and degree of sedimentation in the lower watersheds appeared to be much less than anticipated at appraisal, implying that irrigation flood control schemes in such areas would be marginally feasible at best. The supervision mission concluded that it was becoming increasingly clear that the pilot project as designed was in reality two separate projects involving upper and lower watershed activities that had little interrelationship except for their close proximity. The complexity, size and problematic nature of the civil works associated with the lower watersheds tended to overshadow the more modest investment activities in the upper watersheds and obscured their, often quite promisingg achievements. - 7 - 2.6 The final report of the consultants was presented in May 1983.-i It reinforced the utility of an integrated watershed project approach to development while confirming that the irrigation flood control works involving dam construction of the magnitude of the Dholbaha dam were atypical and not applicable to the rest of the project area. Of 10 watersheds studied only two were considered to justify dam construction, and the size of these dams would be only about a third of the size of the Dholbaha dam. 2.7 In January 1984 construction work on the Dholbaha Dam was suspended when the contractor was found to be in default with only 20% of the work completed at the end of the 24-month contract period. A new contractor was engaged to complete the balance of the work in January 1985. A few months later a contract was also awarded for a second dam, in the Maili watershed. 2.8 Based on the consultant's findings that the potential for economically attractive irrigation investments in the four appraisal-designated watersheds other than Dholbaha were less than expected, the GOP in January 1985, proposed adding to the project the six additional watersheds studied by the consultants and extending the project completion date by two years. These changes were introduced by GOP primarily for the purpose of ensuring full utilization of the loan, with top priority to be given to irrigation and flood control through dams in seven of the 11 project watersheds, together with deep tubewells in the remaining four. A few months later GOP proposed constructing dams in three of the four watersheds previously identified by the consultants as suitable only for tubewell irrigation. 2.9 The Bank readily agreed to extending the upper watershed rehabilitation and development activities to all 11 project watersheds on the grounds of satisfactory progress in terms of the project objectives, but requested detailed implementation plans for civil works in the lower watersheds after noting that GOP's proposal included seven more dams than the consultants had recommended. The Bank emphasized that the economic viability and justification for each additional dam would have to be considered on a case by case basis for possible financing from the loan under any extension to the project. 2.10 In the remaining months prior to the original March 1986 closing date and through the period of two annual extensions to March 1988, the Bank declined to accept financing requests for all except a third dam (within the Janauri watershed) on the grounds of inadequate preparation or lack of economic justification, while agreeing to continue reimbursing for the three dams already approved and to support further upper and lower pilot rehabilitation and development activities unrelated to surface irrigation structures. El The consultants recommended a substantial increase in valuable forest products for the upper watersheds through a program of reforestation, combined with fodder grass production and livestock control, supplemented by improvements to the small cultivated areas through limited land levelling, water harvesting techniques and introduction of rainfed horticultural crops. While the provision of irrigation, by gravity or groundwater abstraction was acknowledged as likely to have the most immediate impact in the lower watersheds, its provision was recognized as technically not feasible or excessively costly over large areas, where a more scientific approach to dry land farming, particularly silvicultural and horticultural, was advocated instead. - 8 - 2.11 In keeping with the pilot concept, further improvements were tested or demonstrated within both the upper and lower watersheds in the closing years of the project, with many of the changes derived from earlier project experience. Recommendations arising from applied research conducted at the Ballowal Saunkhri Research Station of Punjab Agricultural University (PAU) were a particularly noteworthy feature of this phase. Supervision missions noted that overall implementation performance improved in the last two years, when line agencies were reported to be cooperating more effectively than previously. This improvement was attributed to better coordination effected by the project's senior management. III. PROJECT OUTCOME 3.1 The audit is reasonably satisfied that, in spite of inefficiencies and much delay caused by shortcomings of a financial, organizational and managerial nature, and notwithstanding the fact that the overall economic rate of return at completion was re-estimated as 8% compared with the appraisal estimate of 16% for Dholbaha and "more than 12%" for the other watersheds, the project can be described as having, in the end, though at high cost, satisfactorily met its pilot objective. That objective was "laying the basis for sound investments on a larger scale in the future for reversing man-made ecological degradation and protecting and developing agricultural land that is presently subject to serious erosion and flooding in the Punjab Shiwalik Hills", as defined at appraisal. Credit for this achievement must be shared by large numbers of staff from the main GOP implementing agencies, aided by a competent team of international consultants in the early years, capable Bank supervision throughout, and helpful technical advice emanating from research staff of the Ballowal Saunkhri research station. Part I Sections C and D, of the PCR provide an excellent 12-page account of the results and lessons learned. 3.2 By concentrating the development efforts of several GOP agencies in a small number of representative watersheds, the project provided the Borrower and the Bank with an opportunity to test and demonstrate sound investment possibilities for reversing man-induced damage on ecologically degraded lands within and immediately below the Shiwalik Hills, including eroding lands of the upper watersheds and flood prone lands of the lower watersheds. 3.3 The very considerable problems to be overcome in meeting project objectives were usefully clarified by the consultants whose detailed watershed feasibility reports and final report afforded invaluable supplements to the Staff Appraisal Report. The consultants' reports were particularly useful for pinpointing the essential differences between watersheds and for the emphasis they gave to developing appropriate patterns of sustainable land use incorporating good soil and water conservation practices in all the upper watersheds, coupled with more selective developments, including tubevell irrigation as well as protective measures against siltation and flood control in the lower watersheds, not all of which could justify a storage/irrigation dam or elaborate channelization works. Furthermore, the consultants' watershed feasibility reports provided the essential detailed plans for guiding the subsequent development of 10 of the 11 project watersheds. - 9 - 3.4 Project activities under rainfed conditions in the upper watersheds have demonstrated the practicality and benefits of reforestation, with a variety of local and exotic, commercially or socially valuable, species, but only when followed by careful maintenance and accompanied by effective control of grazing. Similarly, livestock improvement through animal exchange schemes and artificial insemination services for buffalo and cattle, combined with fodder grass establishment for stall feeding in substitution for grazing, have been successfully demonstrated. Erosion on both fodder grass and field crop lands has been further reduced through extensive use of vegetative methods of soil and moisture conservation with protective grass screens and shrub hedges planted along the contour, combined with minimal use of small check structures in active stream gullies. PCR re-estimated economic rates of return (ERR) of 12%, 9%, and 10% for the forestry, animal husbandry and soil conservation components are encouraging, but realization of such rates over the longer term are heavily dependent on efficient operation and good maintenance standards on the part of beneficiaries in the aftermath of the initial periods of subsidized government assistance. 3.5 On the other hand, the irrigation component in the lower watersheds, which encountered serious construction delay in the case of the large dam at Dholbaha, and which incurred heavy capital investment costs in relation to the proportion of the command areas actually served, was found to have an ERR of only 2% at completion. The horticulture component, with a very satisfactory re- estimated ERR of 22% at completion, contributed little to the project objectives because it was confined mainly to irrigated lands rather than rainfed areas in the upper watersheds as intended at appraisal. 3.6 The audit views the establishment of the Ballowal Saunkhri Research Station as an especially appropriate pilot activity with great promise for the future. The station staff have quickly assumed a leadership role in adaptive and applied research into the special problems of sustainable renewable resources management in the Shiwalik Hills. Current research on water harvesting in micro watersheds, using small impounding structures combined with tree, fodder grass and dryland crop species and variety selection, and agronomic investigations, are especially relevant to upper watershed rehabilitation activities being supported in Bank/IDA assisted follow-on projects. IV. FINDINGS AND ISSUES A. Pilot Project Perceptional Differences 4.1 The project as conceptualized by GOI, FAO, ICRISAT and Bank staff at the early identification stage of the project cycle was envisioned as a relatively small watershed restoration pilot project intended to demonstrate, for later replication on a larger scale, ways and means for increasing the productivity of rainfed farming in the Himalayan foothill areas through the introduction of improved cropping systems incorporating recently developed soil conservation and water harvesting techniques (para. 1.4). Punjab was identified as an appropriate state to carry out such a project because GOP was at that time (1977) considering a 10-15 year development program for its entire 460,000 ha - 10 - Kandi Tract. It was thought that a project in support of this program would be a suitable vehicle for the pilot initiative, and a Bank mission was accordingly despatched in April 1977 to identify a possible project. 4.2 The mission noted that the Irrigation Department (ID)-inspired GOP program was directed primarily at making water available to the majority of farmers in the tract through gravity, small lift irrigation, or tubewells. At the same time the project -was to include three components, two of which (Sutlej Reclamation and Shar Nehar Irrigation) were conventional gravity or tubewell irrigation schemes in the lower watersheds, and the third a heavily engineering- orientated component involving low storage dams, land reclamation and flood protection in the middle and lower sections of five Kandi tract watersheds and an upper watershed sub-component which gave strong emphasis to check dams intended to provide irrigation water to some 23,000 ha in the upper watersheds. The reforestation and pasture development activities were restricted to more modest investments over approximately 16,000 and 7,000 ha, respectively. 4.3 The identification mission described the project as proposed by GOP as an "integrated land and water development project" rather than the small, pilot, watershed restoration/rainfed farming project as envisaged in earlier discussions (para. 4.1). In its June 1977 back-to-office report however, the mission expressed confidence that the project could be reformulated in the course of preparation and appraisal to conform more closely with the original concept. It anticipated that the GOP project preparation team, with support from Bank staff, could have the project ready for appraisal by January 1978, with a view to the project becoming effective by September, 1978. 4.4 In the event, the project did not become effective until November 1980. The delay was occasioned by strong conceptual differences between GOP and the Bank concerning the project objectives, design and scope. The GOP, led by the ID, was heavily committed to the larger scale, irrigation engineering-based development approach concentrating on the lower watersheds. The Bank adhered to the smaller scale rainfed farming pilot concept for the upper watersheds. The project as appraised in effect therefore represented a compromise between the Bank's strongly held limited pilot objectives (para. 1.5) and the GOP's more ambitious and costly developmental objectives (para. 1.11). 4.5 The essence of the compromise was that the GOP agreed to drop the Sutlej and Shah Nehar components and to accept international consultants to assist with detailed planning of 10 watersheds additional to the only fully prepared watershed plan (for Dholbaha) prior to negotiation. In return the Bank agreed to extend upper and lower watershed investment activities under the pilot concept to a minimum of at least 4 of the 10 additional catchments, to be selected on the basis of the consultants' findings. 4.6 Notwithstanding the compromise, the project is clearly described as a pilot in the text of the Staff Appraisal Report (SAR para. 4.01) and was unequivocally presented to the Board as such. Regrettably, however, the word "pilot" is missing from the project title. The absence of this qualifier is seen by the audit as having reinforced the different perception held by GOP staff as compared to Bank staff, which affected their attitudes towards the project throughout the implementation phase. Only the Forestry and Animal Husbandry - 11 - departments of GOP, and PAU were substantially involved in the pilot rainfed farming and upper watershed rehabilitation aspects which was the element of the project with the greatest interest to the Bank. The other involved GOP agencies (ID, Conservation and Engineering, Agriculture and Fisheries departments) had their attention strongly fixed on the quite separate areas of irrigation potential in the lower watersheds. Supervision missions referred repeatedly in their reports to the tenuous relationship between the upper and lower watershed activities and the generally poor coordination existing between agencies. One commented that the project was in effect two separate projects and another to the fact that "...an expensive irrigation tail was wagging a very thin dog!" These references portray a pilot project which was achieving modest success and a developmental project which was plagued by inadequate prepecation, engineering construction shortcomings, and the prospect of being only marginally justifiable at best, due to high unit area development costs. 4.7 During th preparation phase Bank staff apparently misjudged the extent to which the GOP could be persuaded to embrace the pilot concept (para. 4.3). As a result the Bank erred at appraisal in accepting too large a project than was justified by the pilot concept alone. The audit concludes that the pilot objectives could have been equally well, or better, attained through support of one, or at most two, pilot watersheds. A smaller, less costly, more focussed and manageable project could have provided greater flexibility in operation, permitted more detailed and accurate monitoring and evaluation, and better enhanced the prospect of an earlier conclusion, leading to more rapid replication of the positive findings in follow-on projects--all highly desirable features of a pilot project. B. Dholbaha Dam Problems 4.8 Only one watershed (Dholbaha) had been planned, with full engineering design for dam and spillway completed, before project effectiveness in September 1980 (para. 1.19). Its construction was extremely problematic (paras. 2.2-2.3, and 2.7). During the 23-month period between negotiations (October 1978) and ]-an signing (September 1980) GOP independently proceeded to prequalify, and to finalize tender documents and invite tenders through local competitive bidding procedures, without the benefit of an ex-ante review by the Bank as agreed at loan negotiations and specified in the draft loan agreement document. The Bank was only advised of these actions on August 12, 1980, after tenders had been invited and only three weeks before the advertized bid opening. Four bids were received, ranging from Rupees 37.9 to 48.2 m. The absence of an ex-ante review by the Bank before the issuance of tender document was a matter of concern to the Bank, which noted that no price escalation clause had been included in the tender document, and, more seriously, that certain aspects of the prequalification procedure appeared irregular. 4.9 The Bank's concern over the prequalification arose from the fact that the contract award, to the lowest bidder, was made to a local company which was originally deemed not qualified by GOP. However, the firm's bid was later accepted by GOP after its management had announced a collaborative arrangement with another, more experienced, New Delhi-based firm which was to provide capable personnel and equipment. - 12 - 4.10 But the contract agreement between GOP and the successful bidder, signed in early December, 1980, again without giving the Bank the benefit of prior review as specified in the Loan Agreement, made no mention of the collaborative agreement, which prompted the Bank to query GOP's decision in accepting the bid, and to seek an assurance that the two companies were in fact legally collaborating in the construction. 4.11 No such assurance was received. Instead, in January, 1981 GOP advised the Bank that the local company had, since signing the agreement, considerably improved its manpower and fleet of construction equipment and submitted a construction schedule which would permit timely execution of works according to contract specifications and without the need for collaboration with the New Delhi firm. At this stage, the Bank, which viewed the development as a probable case of misprocurement, expressed concern about the manner in which the local company was able to qualify itself for the contract, and surprise that it had been able to strengthen its staff and equipment inventory so quickly. Nevertheless the Bank, conscious of the fact that the contractor had already commenced work and that a rebidding exercise would incur a delay of one construction year, did not formally object to the contract. 4.12 In fact, the choice of contractor was most unfortunate. The firm proved to be financially and technically weak and it performed extremely poorly, to the extent that by the end of the contract period in December, 1983 only 20% of the expected construction had been completed (para 2.7). A new agreement then had to be made with another contractor to finish the works. in the event, civil works construction took eight years rather than the two to three years estimated at appraisal, and at project completion most of the tertiary irrigation system was still not operational. 4.13 In retrospect, the skepticism shown by the Bank regarding the capability and capacity of the originally selected contractor to perform satisfactorily the Dholbaha civil works can be seen to have been fully warranted. The Bank would have been well justified in declaring the contract award a case of misprocurement and declining to reimburse the borrower from loan funds. The GOP, on its part, would do well, if it has not already done so, to str ngthen its tender procedures in respect to prequalification of bidders to prevent any repetition of this unfortunate experience. C. The Positive Role of Foreign Consultants 4.14 Persistence on the part of the Bank was needed to persuade GOI and GOP into accepting a foreign consultant firm to augment the weak GOP capacity for preparation of plans and feasibility studies for the ten watersheds other than Dholbaha (paras. 1.16, 1.18, and 1.20-1.21). 4.15 The appraisal mission recognized that the Dholbaha watershed, with its favorable dam site and unusually large irrigable command area, would have limited value as a model for the other watersheds. Each watershed presented a unique set of problems, the solution to which called for a high level of individual planning and engineering design, which in the mission's view could best be provided by an experienced design team, and that such experience might not be available locally. 13 - 4.16 Insistence by the Bank on opening the selection process to foreign competition (unusual in the Indian context) was justified by the quality of the selected consultant's performance. The consultant's reports were instrumental in providing the case for realigning the project from its initial emphasis on surface irrigation and flood control in the lower watersheds towards groundwater development in the lower and water harvesting in the upper watersheds, and the extension of the afforestation and other soil conservation activities to all eleven watersheds (paras. 2.6, and 2.8-2.9). This realignment of the project more closely towards its appraisal objectives, introduced at the suggestion of the Bank during a mid-term review following the completion of the consultants' assignment, constituted an important turning point in terms of successful project implementation, which, in the view of the audit, is insufficiently highlighted in the PCR. In this instance the consultants performed an important dual role of supplementing the earlier preparation efforts, which were recognized as inadequate at appraisal (para. 1.19), and of contributing significantly to the pilot objectives through their detailed planning activities (para. 2.6). 不