India - Second Orissa Irrigation Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 8628 PROJECT COMPLETION REPORT INDIA ORISSA IRRIGATION II PROJECT (CREDIT 1397-IN) MAY 8, 1990 Agriculture Operations Division Country Department IV Asia Regional Office hib doument has a reticted dshistbti and may be used by reipiet only in the pedormne of thir officld duti Its contents may not othewise be discosed wihout Wodd bank atwh ation.  CXBQUbI Nam of Currency: Rupee (Rs) Rate of Exchange: Appraisal (September 1982) US$1.00 = Rs 9.50 Intervening years (average) US$1.00 = Rs 10.40 Compledon year US$1.00 = Rs 12.36 Fiscal Year of BonTower GOI and GOO: April 1 - March 31 Weights and Measures: Metric System GLOSSARY kharif - Wet season; June - October. rab - Dry season; November - ApriL chak - Watercourse CCA. watercourse - Small channel below outlet serving a chak down to S ha blocks (sub-chaks), with capacity to take full outlet discharge from head to tail waabandi - A rotational water supply (RWS) system of alocating water among farmers by fixed turns and time periods; flow rate is constam and time periods are proportional to land holding size. ABBREVIATIONS AC - Appraisal Conmnittee ARDC - Agricultural Refinance and Development Corporation CAD - Command Area Developnent CCA - Culivable Command Anra CWC - Central Water Commission DAFP - Directate of Agriculture and Food Production DRP - Dam (Safety) Review Panel ERR - Economic Rate of Retum GOO - Government of Orissa GOI - Govenmment of India I&PD - lIrigation and Power Department (within GOO) ID - lirigation Department (within I&PD) LC - Land Consolidation M&E - Monitoring and Evaluation MIP - Medium Irrigation Project OFD - On-Farm Development O&M - Operation and Maintenance PCR - Project Completion Report R&D - Research and Development RWS - Rotational Water Supply SCMC - State Coordinating and Monitoring Committee WMP - Water Management Plan  THE WORLD BANK FOR OFFICIA US ONY Washington. D.C. 20433 U.SA. Olfce of OiuctrGa O~ie vluuitiim May 8, 1990 ME4ORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Proj ct Completion Report on India Orissa Irrixation II Proiect (Credit 1397-IN) Attached, for information, is a copy of a report entitled "Project Completion Report on India - Orissa Irrigation II Project (Credit 1397-IN)" prepared by the Asia Regional Office. No audit of this project has been made by the Operations Evaluation Department at this time. Yves Rovani by Ram K. Chopra Attachment This document has a restricted distribution and may be used by recipients only In the perfomtance of their official duties. Its contents may not otherwse be disckwsed without World Blank autttiot*stin.  FOR OFCIAL USE ONLY ORISSA IRRIGATION I PROJECr (Credit 1397-IN) PBOlEC COMPLON REPOR Itae of Coltents Page, No. Preface ........................................................................................... .(i) Evaluation Summary (ii) PART I: PROJECI REVIEW FROM BANKIS PERSPECTIVE Project Identity .1 Background .1 Project Objectives and Desctiption .1 Project Design and Orpanizadon. 2 Project Implementation. 3 Project Results. 6 Project Sustainability. 8 Hank Performance. 8 Borrower Perfornmance .11 Project Relationships .13 Project Documentation and Data .14 Conclusions .14 PART I: PROJECT REVIEW FROM THE BORROWER'S PERSPECTIVE .15 PART I: STATISTICAL INPORMATION Table 1. Related Bank Loans and Credits .............................. 16 Table 2. Project Timetable ........................................ 17 Table 3. Loan Disbursements ........................................ 17 Table 4. Project Implementation ........................................ 18 Table 5. Project Costs and Fmancing. A - Project Costs 19 B - Project Financing 20 C - Loan Disbusement Categories 20 Table 6. Project Results. A - Direct Benefits 21 B - Economic Impart 22 C - Fnancial Impact 22 D - Studies and Institutional Support 23 Table 7. Record of Compliance with Major Covenants .............. 24 Table 8. Use of Staff Resources . A - Staff Inputs 26 B - Bank Missions 27 C - Costs 27 Annex 1 Economic Re-Evaluation of MIPs ........................... 28 map Location of Individual Project Schemes (MIPs) This document has a restricted distribution and may be used by re.ipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  (i) - ~~~~~INDIA ORISSA IRRMAION XI PROJECr CREDIT 1397 - IN PROJECT COQMh QN REPORT This Project Completion Report (PCR) reviews implementaion of the Orissa Irrigadon I Project in India for which Credit 1397-IN in the amount of US$ 105.0 million was approved on July 7,1983. The Credit was closed on March 31,1988 when some US$ 49.7 million (about 38%) of the Credit was cancelled. The Preface, Evaluation Summary, and Parts I and m of this Report were prepared by staff members of the Agricultual Division of Asia Technical Depment (ASTAG), and were reviewed by the staff of the Agriculture Operations Division, Country Department 4, both in headquarters and in New Delhi. The Borwer was requested to prepare a PCR in March 1988, to be available to the Bank in October 1988, but none was received. Arrangements could not be cleared with the Borrower or the Goverment of Orissa (GOO) for a field visit to the project sites. Pzepamtion of this PCR is therefore based on, inter alia, the Staff Appraisal Report (No. 4448-IN, dated May 31, 1983), the President's Report (No. P 3612-IN of the same date), the Development redit and Project Agreements, dated September 16, 1983, Bank supervision reports, correspondence between the Bank and the Borrower, other internal documentation available within the Bank. Ihe staff member responsible for the preartion of the PCR visited the project sites during the Banrcs last full supervision mission in December 1987. The Borrower was requested in January 1990 to provide a review from the Borrower's perspective (Part II), but none was received. This PCR has been read by the Opeatios Evaluation Department (OED).  (ii) INDIA ORSSAS IRRIGATION XI PROJECr CREDIT 1397-N PROlECI COMPLETION REPORT ERvaluation Sumn= Introducton. The project covered by this PCR was the second project in medium scale irngation development financed by the Bank/IDA'/ in the State of Orissa in India. The project followed the Orissa Irrigation I Project (Cr. 740 - IN, US$ 58.OM, 1977) which was on-going at the time of appraisal QbWfe. The four year time slice supported the State's construction program for medium irrigation projects (MIPs), land consolidation and on-farm development. At appraisal, the five main components were: (i) completion of 18 on-going MIPs; (ii) construction of 6 new MIPs; (iii) land consolidation on 160,000 ha of command areas under major irigation projects; (iv) on-farm development on 48,000 ha where LC had been completed; and (v) the support of research and development, training and monitoring and evaluation. The construction of 6 new MIPs was not implemented. Implementadon Experience. Project implementation has been less than successful. The main expenditure item was completion of the 18 MIPs started under the Orissa Irrigation I Project. There has been considerable variation in implementation quality, as would be expected under a time slice operation with many isolated sub-projects spread over the State. Some schemes have deteiorated after construction of main works, because the minor canal systems are incomplete, or failed to function as designed. Most MIPs have not been adequately maintained during the extended construction period. Implementation was hampered by a variety of problems which, inter alia, include: (i) delays in land acquisition and forest clearance; (ii) poor institutional coordinadon; (iii) little priority by GOO to the completion and full development of MIPs; and (iv) the quality of planning and design. Dam safety was a major concern through the project period, due to shortcomings in planning, design and implementation. The GOIVCWC have recognized the potental risks related to dam safety and action is now underway to establish a National Dam Safety Project, which would address problems encountered under this project in Orissa and other Indian States. The land consolidation has been an inportant project achievement, was probably more successful than any simila Program in India and has been implemented in a sensitive and effective way (ref paras 1.08 and 1.17). The on-farm development program was exceeded in terms of physical area, but has made little impact to improving the water use efficiency. Results. The PCR analysis of the ERR is at best indicative, reflecting a similar judgement at project appraisal. The reestinated ERR for the project would average about 6.0 % compared to 14.0% at appWaisal taking into account the PCR aem nt of likely future performance of GOO. The re-evaluated ERR for the individual MIPs are likely to range from 2.4% to 13.5% and only 3 MIPs would have ERRs that exceed 10%. The general shortfall in ERR compared to that evaluated at appaisal (and the PCR of the first Orissa Ifrigation Project) reflects: (i) substantial cost ovenuns; (ii) serious delays in project implementation; (iii) ineffective water management; (iv) incomplete forest and land clearance and agricultural development; and (v) slower 1/ Reference to the Bank in this report includes IDA.  (OiR) than expected response by farmers to inigation, particularly for crops other than paddy. 'tere is insufficient data to quantify the impact of the land consolidation component. However the successful land consolidation program should generate substantial benefits since fragmented land holdings have seriously constrained efficient land use in Orissa. The on-farm development component has been less than successful mainly because of poor planning and design. The component was implemented as a target oriented construction program with little interest in the potential impact. Sustainabilt.. The i cure provided under the project has not been effectively maintained because responsibility for O&M is not well defined and there have been shortage of O&M funding, staffing and resources (including tanspoadon and communications). There are signs that GOO is initiating some coretive actions, but proect sustainability will be difficult because: (i) most inigation schemes are isolated; (ii) funding and staffing are unlikely to improve; and (iii) it will not be easy to change and strengthen the ID. Findings and Lessons Learnt The Bank had reason to assume that a follow-up project in Orissa would be successful given the efforts put into the rirst Ondssa Irigation Project. However project performance depended on GOO taking hard policy decisions on sensitive issues such as institutional strengthening and staffing, budget provision for non-planned expendiures and general improvements in bureaucratuc procedures. The appras also assurmed that under the ine-of-credit' approach, the CWC could effecively appraise MPs and monitor project implementation performance. Both critical appraisal assumptions proved to be opdmistic. The appraisal did not give sufficient attention to the lessons learnt from the on-going first Orissa Irrigation ProjectL The signal given to the Borrower in the rapid processing of the follow-up operation (and confimed in the PCR and PPAR of the Orissa Inigation I Project ) was that the Bank believed that a time slice approach to medium scale irrigation development was woridng. The Bank supervision performance was at best mixed, until May 1987 when the project was rated a problem project and remained such until Credit closure. Lack of reliable information was the dominant reason why many key issues were not addressed earlier in the project cycle. The Bank supervision allowed many problems to develop unchecked, so that the project diverted from the plan envisaged at appraisal. It is understandable that GOVCWCW(;00 found it difficult to read the seriousness of the project situation when it was designated 'serious problem status' after almost a decade of Bank acceptance of implementation progress under Orissa Irrigation Projects. The main lessons to be learnt by the Borrower (GOIAJOO) relate to: (i) local funding and the budgetary process; (ii) institutional capability, resonsibility, accountabiiity and coordination; (iii) manpower and terms of employment; (iv) the role of and need for technical assistance; (v) the quality of planning and design and the appropiateness of technology; (vi) completing development and sustining irrigaton schemes; (vii) monitoring and reporting; and (viii) the need for diagnostic evaluation. There are also important lessons to related to the poor performance of CWC in appraisal of individual MIPs and general project supervision. A fundamental lesson to be leant by the GO/CWC(GOO is that diagnostic analysis of project planning, design and implementation must be introduced, if the substantal investments in the irrigation subsector are to obtain full irrigated agricultural benefits. Clearly the project has not performed as envisaged at appraisal and the major problems, particularly just prior to closure of the Credit, strained the relationship between the Bank and GOVG/OO. However, many project problems could not be tackled under a specific investment project within Orissa and reflect generic/policy issues at national and state leveL particularly those related to institutional capability and performance. Perhaps the most discouraging outcome has been that 01VOO may still not flly recognized the importance of the lessons to be leamnt, not only for the project but for the inrigation subsector. Consequently, there has been litle willingness by G01/GOO to take strong remedial actions necessary to obtain the potential benefits from the project  PART 1: PRQJEC RUEVW EL) BAS PERSPECT IE PepldnfinS Project Name: Orissa Inigation II Project Credit No. : 1397 -IN RVP Unit : Asia Country : India Sector : Agriculture Subsector : Irrigation Backg&nnd 1.01 Seco eeiv The Governmenfs major objectives for the agriculture sector during the project period and successive Five Year Development Plans were to: (i) increase production to meet domestic requirements for food while meeting the needs for industry and expanding agricultural exports; (i) increase farmers' incomes and generate rural employment; (iii) promote a more balanced distibution of economic opportunites through regional development; (iv) contrbute to health and better nutrtonal status of the population; and (v) ensure productive, sustainable use of India's land, water and other naturl resources. The extent of the deficit in food production, the sensitivity of total production to drought, and the rate of increase in demand due to population growth made irrigation, and the introduction of improved agdcultural technology essential elements of India's development strategy. 1.02 Po_ia ne anItese lLink. Agriculture remains a dominant secor of the Indian economy, contributing about 33% of GDP and providing employment to some 64% of the total labor force. Foodgrain production accelerated substandally during the 1960s as a result of the green revolution and the introduction of high-yielding varietieF f wheat and rice. During the fifth (1974-79) and Sixth (1979-84) Plan periods, output has continued oy about 2.7% per year. About 60% of this increase can be attributed to greater yields, and the balance mainly to increased double cropping. Growth in wheat output (5.5% p.a.) reflects the outstanding performance of the Northwest States (Punjab, Haryana, West UP). In contrast, rice production has increased by only 2.0% p.a. owing to the poor performance of the Eastern States, and coarse grain have climbed only (1.5%) because of limited increases under rainfed conditions. The rising output of foodgrains has been accompanied by some diversification into higher-value cash crops (sugarcane, cotton, etc.). At the same time, production of oil seeds and pulses has failed to keep up with the demand. 1. 03 Since 1947, irrigation potential in India has risen from about 19.5 M ha to 61.5 M ha (1983/84), with some 93% of the potential estimated to be utlized. Sixth Plan achievemensts fell short of targets, but potential stll rose at 2.5 M ha per year. Ultimate irrigation potential has been estimated at 113.5 M ha. or 60% of the cultivated area creased irrigation accounts in large part for the improvement in agricultural performance. Inigation has not only increased the area cropped and reduced crop losses in areas of inadequate rainfall, but it has also reduced the risks associated with rainfed farming, and thus encouraged diversified cropping and the use of HYV seeds and chemical inputs. EPrc Objcdv and Deston 1.04 The State of Orissa is one of the poorest States in India. A large segment of the State's population is classified as 'tribal' or 'backward class', which is generally among the less affected by the country's economic progress to date. The State is predominantly rual and agricultural, and progress in general economic development depends on increased crop producton and income in the agriculural sector. Total annual rainfall is high on average, but undependable during monsoon and very limited in the dry season. Expansion and improvement of irrigation is therefore an essential ingredient in the State's economic growth. 1.05 The project objective was to support a second time slice of the Odissa rigation I Project  -2 - (Cr. 740 - IN, US$ 58.OM, 1977) which was on-going at the time of appraisal and closed on October 31, 1983. The four year time slice supported the State's construction program for medium irrigation projects (MWs) and programs for land consolidation and on-farm development. At appraial, the second project had five main components: (i) completion of 18 on-going MIPs; (ii) construction of 6 new MIPs; (iii) land consolidation (LC) on some 160,000 ha of command areas under major irrigation projects; (iv) on-farm development (OFD) on 48,000 ha where LC had been completed; and (v) support of research and development (R&D), training and monitoring and evaluation (M&E). The construction of the 6 new MIPs was conditional on completion of on-going MIPs and not implemented (para 1.26). Prect Design and ranization 1.06 The Orissa Irrigation II Project design followed that of the Orissa Inigation I Project which was conceived as the first 'line-of-credit' project in India for construction of MIPs. The 'line-of-credit' approach was adopted since it was the only pracdcal way that IDA could support a project with a large number of scattered sub-projects. Under this project concept, the appraisal and monitoring of individual MlIPs was to be undertaken by the Central Water Commission (CWC) of GOI in accordance with criteria and procedures agreed with IDA. Ihe project design reflected the lessons learnt by GOI and the Bank under the first time slice and two other similar projects supporting MIP programs in the States of Gujarat and Madhya Pradesh. The five most pressing issues identified during the first time slice were: (i) technical problems related to planning, design and implementation of minor canal systems (watercourses); (ii) cost overruns and delays in completion of MIPs; (iii) institutional capability, (iv) ineffective water management and lack of involvement of farmers; and (v) the need to strengthen the agricultural support services. These lesson learnt from the first project were described in a PCR dated June 18, 1984 (about 10 months after Board presentation of the Orissa U Project). 1.07 Froect rEm atdon. 'The preparation of the Orissa I Project was undertaken by the Irrigation and Power Department (I&PD) of GOO. Relatively little had to be done since most project activities were continuation of works started under the Orissa I project. IDA supported GOO in the preparation, particularly in translating lessons leamt from the first project into addition.0 components. Special attention was given to agricultural support, project commnunications, institutional strengthening, quality control, land acquisiton and forest clearance, and MIP operation and water management based on rotational water supply (RWS). The project preparation covered most areas where improvements were necessary but did not fully address the causes of the problems which affected the first time slice. For example, a major shortcoming was the failure to establish the true status of the on-going MlPs. Most on-going MIPs were originally prepared on an inadequate data base, no real understanding of how the schemes would be operated, and a lack of attention to detailed design. The project costs could not be estimated with any ' certainty even though most MIPs had been under construction for many years. The MIP preparation reports were equivalent to identification or feasibility studies. No MIP appears to have been planned and designed to the standards implicit in the criteria agreed by GO/CWCIIDA. 1.08 Qaization. Implementation anrangernents generally followed the same pattern as in the on- going Orissa I Project. The Irrigation Department (ID) within the I&PD of GOO was responsible for constructon and operation and maintenance of MlPs ard for the training and R&D components; the Diectorate of Land Consolidation in the Revenue Departnent for land consolidation; and the Directrate of Agriculture and Food Production (DAFP) in the Depment of Agriculture and Cooperation for construction of OFD works. An important organizational objective was to inmrove policy level coordination aid overall monitoring of the various GOO agencies involved in the project. A high-level State Coordinating and Monitoring Committee (SCMC) was to be set up chaired by the GOO Development Comnissioner. The role of CWC, through an Appraisal Committee was to take direct responsibility for the appraisal and progress monitoring (in addition to IDA supervision) of the MlPs financed under the project, as agreed under Schedule I in the Development Credit Agreement. 1.09 The IDA appraisal documentation provided plausible pioposals to resolve most problems identified under the first time slice. The appraisal and legal documentation confirm that GO recognized most lessons to be learnt from Orissa 1 Project The SAR indicated that the detailed designs of most MIPs were approaching completion and were in the process of being translated into a practical and coordinated work programs. Project implementation proved this not to be the case. The  -3- most important assumption was that the CWC would efficiently fulill its agreed role in respect to MIP appraisal and direct monitoring of the project. The SAR also implies that many basic technical standards and policy, procedural and institutional issues had been fully explored and agreed by GOI/CWC and the partcipating agencies within (OO. Tis proved to be optimistic. Many implementation problems appear to have been caused by only partial evaluation during the preparation and appraisal process of the causes of the problems under the Cr)ssa I Project The MPs started under the Orissa I had been considerably behind schedule even though the main construction activity had been major civil works (i.e. dams and main canal systems). In fact, no on-going MWP was far enough advanced at IDA appraisal to fully assess the practical difficulties and constraints of completing and operating an MIP. Many issues which subsequently became major project problems, should have been given more attention during the IDA appraisal process. Somehow these concerns got lost under the tight time frame (to deliver another lending operadon) as the IDA appraisal moved towards negotiations. Another factor which affected the appraisal was that the IDA appraisal mission followed after very serious monsoon flooding in Orissa when government staff were diverted to relief workl Communications were cut off to most parts of the State, strictly limiting field trips during appraisal Project hnplementation 1.10 Genera. Project implementation has been less than successfua A core problem was the difficulty of managing many isolated MIPs scated over the State by weak government departments. The ID and other departments involved were hampered by a wide range of institutional constraints and problems (para 1.22). These institutional problems were recognized at IDA appraisal and most were covered by appropriate conditionalities, in particular, the monitoring role of CWC. However, the ID appeared to be pnmarily interested in construction of main civil works with limited interest in the overall perfonnance of irigation systems. Consequently, the various work programs continued on separate parallel tracks and reflected un,'redictable budget provisions and general constraints related to staffing, resources and facilities. 1.11 A successful line-of-credit' operation in irrgated a Iculn development depended on a strong coordinating agency able to take action on many issues with statewide implications. A State Coordination and Monitoring Committee (SCMC) of (OO was supposed to take on this responsibility, but was not effective at project implementation level. Many policy issues (ie. land acquisition and forest clearance, water management, and dam safety) were not addressed in a timely manner. Coordinadon problems and delays due to inaction on policy issues were compounded throuh project implementation by poor monitoring and unreliable data collection (para 1.24). The 000/ID and CWC did not '.ilow, or did not seriously question, the project status well enough to idendfy the causes of most problems and plan corrective action. Unfotunately, one indicator of implementation progress, the pace of disbursements, was pardcularly successful under the Orissa I Project and this may have cause both GO/GOO and IDA to underrate the potential problems under the follow-up project 1.12 Construction of MIPs. The main expenditure item was completion of the 18 MIPs started under the Orissa I Project There has been considerable varation in implementation quality, as would be expected under a 'line-of-credit' operation with many isolated sub-prqjects spread over the State. Main civil works were generally of a standard acceptable to the State (and CWC) but not of the quality implied in the Credit and Project Agreements with IDA. Implementation quality varied considerably between the MIPs, reflecting the difficulty of site management and supevision of isolated MIPs. Some schemes have deteriorated after construction of main works, because minor canal systems are incomplete or failed to function as designed. Most MIPs have not been adequately maintained during the extended construction period. A basic variance from project planning has been that main civil works progressed much quicker than minor works and operational principles implicit in the designs of MIPs have not been successfully demonstrated. The work program was also uncoordinated for a variety of reasons including: (i) delays in land acquisition and forest clearance; (ii) poorly institutonal coordination; and (iii) little piority by GOO to the completion and full development of MIPs. A recuing problem through the project period was that schemes were declared and treated as complete in terms of budget provisions when considerable undefined or remedial works still had to be done. This shortcoming occurred mainly because there has been no fomal procedure for testing and verifying that MIPs were complete and that the irrigation service could reach all pars of the MIP command aras.  -4- 1.13 Lack of completion of minor canal works (watercourses) has been widespread, reflecting on the quality of the planning and design of MuPs. It also reflects on the appraisal process by CWC which appraised and approved Mu's with no clcar understanding of how the schemes would be operated. The delays in forest clearance and development have compounded the problems of completing watercourse systems which could not be built and sustained through standing forest or shrub. Experience has shown that if watrcourses are not fully completed, efficient operation of schemes becomes virtually impossible. The incomplete minor irigation systems on most Mu's have also meant that the irrigation senrice has been generally unreliable. As a result many farmers, particularly tail-enders, have not been able to use the imigation service benefiting upstream farnmers. With no mal accountability for project performance, the has been little urgency for the II) to complete the minor works to provide a reliable service to all the command area Incomplete MWP commands have generally obviated the need for efficient O&M because there has usually been sufficient or surplus water for the limited upstream command area which can be irrigated 1.14 Pm Safey. IDA's concern regarding dam safety increased through the project period as it became more evident that there were fundamental technical and quality issues affecting the project These technical concerns related to the quality of planning, design and implementation, poor and incomplete records of completed works, and a general lack of monitoring and reporting. In addition, until October 1986, the GOO had shown nominal interest in dam safety as reflected by the performance of the Dam Safety Review Panel (DRP). The DRP was set up as agreed with IDA but remained dormt The DRP Was only reactivated late in 1986 after IDA prompting, and then without any real support either by GOO, the ID and initially by the CWC The reactivated DRP identified extensive and potentially serious problems with most dams, causing IDA to express grave concern to GO/OO. However, dam safety was a sensitive political and institutional issue. The CWC (responsible for the appraisal of dams and monitoring implementation) reviewed the DSP reports and advised that, with the exception of some outstanding issues related to hydrology, there was no cause for concern as to the safety of the project dams. The CWC advice did not address many of the basic problems identified by the DRP and raised new concms in IDA on the objectivity and competence of CWC staff to perform the required technical audits of dams. 1.15 IDA considered that the situation was inconsistent with obligations under the Bank operational policies and procedures (OMS 3.80) related to the safety of dams. Consequently, IDA sought expert exteral advice on the evidence available from the State's DRP and the CWC reports. As a result, it was decided to mount a high-level independent review by recognized international experts. In December 1987, this review (n contradicton to CWCs position at that time) advised that there were serious unresolved technical issues at each of the six dams visited. The issues meant that at best the safety of the dams was unknown. A core problem which complicated any diagnostic evaluation of the technical issues was extremely poor - sometimes non-existent - documentation of the data related to investigations, designs and construcdon. The independent IDA dam review concluded that CWCfGOO would need substantal technical assistance to resolve the specific technical issues related to dam safety. IDA also concluded that GOO needed technical assistance to address the underlying institutional and other technical problems affecting the project, pardcularly related to MWP operation (para 1.19). 1.16 The differences between IDA and GO1/CWCG000 on the need for extemal technical assistance and the seriousness of the dam safety issues could not be tesolved in the extremely tight time fiame to the Credit closing date of December 31, 1987. As a consequence, and after further efforts to resolve the issues, the Credit was closed on March 31, 1988. The latest situation with respect to dam safety is not known since arrangements for a PCR mission to visit Orissa could not be agreed with GO/00. However, there is encouraging evidence that GO1/CWC/GOO has recognized the importance of the technical and institutional issues related to dam safety and the potential risks involved. Action is now underway to establish a National Dam Safety Project, which wo,Ad address the problems encountered under this project and those in other States of India. 1.17 Iad Consolidaion. The lar;d consolidation has been an important project achievement and probably more successful than any similar program in India. Up to June 1986, the target of 160,000 ha had been successfully consolidated because: (i) the majonty of farmers understood and appreciated their  -5- peonal benefit from the program; and (ii) the program was implemented in a sensitive and effective way. Although the land was successfully consolidated, the programimpact may be less than expected because on-fam development was not undertaken concurrent with the LC program. Thus if at some future date, minor drainage has to be added or the minor irrigation system improved within areas consolidated under the project, further land consolidation may have to take place. In this respect, coordination between the LC and irrigation/drinage progams would have been difflcult to achieve since the LC program was not planned to cover blocks of irigated commands. 1. 18 On--fa= Develogent. The OFD program was exceeded in terms of physical area but the effectiveness of the program has never been really evaluated. There was concern through the project period that the improvement in efficiency of water use was not being achieved. The OFD progmm could and should have been thoroughly evaluated as soon as the works were startd (under the Orissa I Project) since an OFD evaluation could have been done on a small completed area. It is concluded that the main objective of the implementing agency was to complete the physical work targets set in each year. Few OFD works were completed in the MIP command areas although OHD works should have been designed before the watercourse command areas were finalized and watercourses aligned. 1.19 Ogperaion and Maintenance. Risks associated with O&M were foreseen at appraisal and appropriate conditionalities v,ere agreed relating to operating and water management plans. However, the effect on O&M of the long delays in implementation and limited capability of the ID at field level were not anticipated. As yet, no MIP is believed to be fully operational, although GOO has declared 9 MIPs as being complete. Consequently, there is still no conclusive evidence that MPs (some started a decade ago) can be operated as designed. Amnost all MIPs designated as substantially complete appear to be randomly opeaed since there are no explicit operational plans that can be understood and used by typical field staff. Most field staff do not fully understand (and have no practical experience) of rotational water supply (RWS). Consequendy, most MIPs at present are operated either as 'on-demand' systems or 'free flow' with some head regulator control. Operation of MIPs as designed is likely to remain difficult because paddy is the preferred crop in both the kharif and rabi seasons. In fact, there is some doubt whether MIPs can be managed following water allocation principles implicit under RWS operation because O&M staff cannot enforce the required discipline. Thus the O&M role of site staff is mainly limited to maintenance - which is usually remedial or curative. Permanent arrangements for routine O&M are generally not yet finalized orfully funded. 1.20 Ihe reservoir storage and canal system management have not been treated as essential determinants in the planning and design of MlPs. In fact, introduction of the RWS operating principles took place after the MIPs were designed. A basic technical issue, still unresolved, is whether the MIPs as designed and constructed can be operated to provide an inigation service which is timely and reliable thrughout the MIP command areas. For example, there are no specific procedures for dealing with how water shortage should be distributed through a command. Reservoir yields awe not being managed to maximize storage for post-monsoon irrigation. This will inevitably limit the project impact in the dry season which has been assumed to contribute a substantial proportion of the project benefits. 1.21 ITe appraisal placed strong emphasis oa farmer participation. However, most farmers were new to irrigation and simply did not understand the need for cooperation within an MIP command. In this respect, a project conditionality that GOO/ID would produce water management plans (WMPs) for all MIPs with farmer representatves involved in the process proved to be unworkable WMPs were set up as a proforma exercise with no real undemanding of the objectives. The viability of WMPs was questioned through the project period by IDA supervision missions but without any measurable results. It is concluded that WMPs must be demonstated as workable and useful, or farmers will continue to question the credibility of the ID and the project 1.22 Pre Insdtutions. There has been growing evidence that project technical problems and other constraints were simply effects of a more fundamental problem - a basic insdtutional weakness that was not fully recognized either at IDA appraisal or in the PCR of the Orissa 1 Project The ID has limited institutional capability, lacks specialist staff and has an organizational structure that needed to be rationalized. Established management and quality control procedures within the ED either did not work  -6- because they were not practical, or because they could not be (or were not) enforced. Within the ID and other project agencies, policy and technical directives were often misunderstood or were incomplete when they reached the working level. Clearance, certification or verification procedures were regularly treated as proforma exercises rather than essential checks on performance. Extended chains of responsibility and severe restrictions on travel often resulted in misunderstandings and inconsistencies at field level. All these shorcoomings were recognized (and geneally agreed bv IDA and GOO), but could not be resolved at the project level. This was &cuse the causes of the shortcomings were often generic problems and constraints affecting all State institutions. For example, the D:) generally could not place its best staff in key positions because promotion is based on seniority and because staff assignments are often conditioned by patronage outside the department. The terms and conditions of employment are such the govemment staff have virtually permanent tenure. This has stricdy constrained the management options and also has limited accountability for performance. 1.23 The effectiveness of the ID (and other agencies involved) has been restricted by GOO/CWC not recognizing the seriousness of the project situation until the problems became major concerns of IDA. The project was not given sufficient support for non-planned expenditures, particularly if these varied from established statewide nonrs. This has meant dtat the institudonal strengthening components under the project could not be successfully implemented. For example, the ID performance has been severely resticted by poor communications and lack of transportation (essential to reach the isolated MIP sites and offices outside headquarters). There has been insufficient support for non-construction activities such as investigations, planmng and design, R&D, training and M&E. These institutional and staffing risks were not fully identified at IDA appraisal (or in the PCR of the Orissa I Project), but have been critical to shortcomings in project performance. 1.24 Monitoring and Reporfing. A basic reason why project institutions have not be effective has been poor or incomplete monitoring and reporting. This has limited the effectiveness of the State Coordinating and Monitoring Committee. Without information, the SCMC could not make realistic decisions and probably never really knew the actual project status (or the real problems) through Cie project period. The information gap is illustrated by the fact that project status (and thus costs) could not be established with any certainty at IDA appraisal even though Orissa I Project had been underway for five yeas The project status still could not be established with any certainty at Credit closure and was an important reason for not extending the Credit closing date. Almost every progress report revised or contradicted previous progress data. Most information related to physical or financial progress of civil work progams with virtually no reliable data on project performance. In addidon, there has been widespread problems with the quality and quantity of data which has generally been of little value for effective project monitoring or management. The fundamental problem has been insufficient data at the planning and detailed design stage of MIPs (particularly hydrology and structure foundation conditions) and an almost total lack of diagnostic reporting and evaluation. In addition, documentation on completed works has been either incomplete or insufficient - the major reason why the safety of most MIP dams remains unknown. It is concluded that because of poor monitoring and reporting, many project shortomings were not recognized until the situation became critical. The CWC should have focused on this issue since they were responsible for project monitoring under the Credit. 1.25 There are some indications that the importance of information gaps on project implementation and performance has started to be recognized by GOVDO0. Early in 1987, CWC/GOO responded to growing IDA. concerns and made attempts to improve its mitoring of the project. The SCMC was strengthened by putting it under the chairmanship of the Chief Secretary of GOO. However, these actions were too late and could not be fully effective because they depended on GOQ/ID having reliable information available for review. It is concluded that M&E will not be improved in Orissa until staff perceive that GOO and ID management view M&E as useful (through feedback) and essential (through accountability for quality and quantity of data). Projec Results 1.26 Intducimn. The project was conceived as a time slice operation. It was predicted under the Orissa I Project that typical MIPs would take 5 years to fully complete with an overlapping 4 to 8 years to fully develop the land and agriculture. The actual implementation period of a typical MIP is now likely to  -7- average at least ten years. The component to start construction on 6 new MIPs remained inactive because IDA support depended on completion of at least 10 on-going MIPs. The project costs (detailed in Annex 1) for on-going MIPs have steadily risen over the project period from about Rs. 2415 M at appraisal (including contingencies) to about Rs. 3925 M at Credit closure, an increase of some 46 per cent. However, estimated costs of individual MIPs have shown substantial overruns (up to 190%) reflecing on the quality of the original planning and design and the CWC appraisal process. The progress in financial terms has been much slower than antcipated in financial (dollar equivalent) terms as reflected by the cancellaion of about US$ 48.0 M of the Credit. 1.27 MIPs. The command area of the 18 on-going MIPs was estimated at about 108,000 ha at IDA appraisal and revised to about 113,800 ha once the actual schemes were designed. In December 1987, IDA reported that none of the 18 on-going MIPs were fully complete and able to function as designed with some 10 MIPs substantialy complete. However, GOO reported in March 1988 that 9 MIPs were complete and if funds are made available, all on-going MIPs would be comnpleted by March 1992. As of March 1988, GOO has also reported that irrigation works have been completed to serve some 41,950 ha (38%) compared to 25,000 ha recorded as irrigated at IDA appraisal (kharif 1983). About 56,000 ha were reported as irrigated in the khrif/rabi crop years through 1987/88 giving an average cropping intensity of about 50 percent. MIPs which are substantially complete are reported to have average irrigated crop intensities of about 125% compared to about 137% assumed at appraisal. What this means in terms of agricultural production is not known, since there is no reliable performance data on the quality or quantity of the irigation service or irrigated crp yields. However, circumstantial evidence gathered by IDA review missions indicates that average crop yields may be less than predicted at appraisal, particularly for crops other than paddy. Average crop yields are also likely to fall when water becomes more scarce as the full command areas of MIPs are brought under irrigation. Some 24,120 ha (21% of total project area) have still to be cleared and developed for agriculture. About 3,220 families are still to be resetded once new reservoir areas are flooded, but generally the resettlement proram has progressed satisfactorily with 1,480 families resetded to date. On average, about 85% of the main works are reported as complete, together with some 40% of the ninor canal works to watercourse leveL There is insufficient data to fully assess how the various technical problems will effect project impact or why cost estimates have inceased so substantially for some schemes. 1.28 ae ReEvaluadon. There is insufficient data to calculate the financial rate of return, or the impact of the project on poverty alleviation. The PCR analysis of the ERR is at best indicative, reflecting a similr judgement at the I)A appraisal of the project and in the PCR of the Orissa I project (1984). The PCR analysis has been based on full development of the schemes assuming that GOO completes the work program to the latest schedule and at cost esdmates made available to IDA in March 1988. This is probably optimistic. It also assumes that project lessons learnt to date will be effectively applied and that G0T/C WC/GOO will take appropriate and timely actions (particularly related to technical issues such as dam safety, MIP opeation and water management). This is also probably optimistic. Consequendy, a sensitivity analysis has been undertaken to test the project viability ff GOO is not able to resolve the project problems as planned. The re-evaluated ERR for the project MIPs is given in Table 6 (Part III), detailed in Annex 1 of Part III and is summarized below: Evaluation ERR All 18 MEPs Range for individual MIPs SAR 14.0% 7.7% to 23.9% PCR (i) Base Case (no future problems) 9.8% 5.8% to 20.1% (ii) 75% Benefits 7.0% 3.5% to 15.4% (iii) Underfunding of O&M 5.3% 2.0% to 12.6% (iv) Ineffective Management 6.0% 2.4% to 13.5% 1.29 Ras of Retum. The PCR has taken the reestimated ERR of the MlPs under the project as that likely to reflect the future perfomance of 000. It is concluded that the reestimated ERR for the 18 on- going MlPs would reflect the 'ineffective management' case (item (iv) above) and thus average about 6.0 % for all MIPs compared to 14.0% at appraisal. The re-evaluated ERR for the individual MIPs are likely to range from 2.4% to 13.5%. Only 3 MIPs would have ERRs that exceed 10%.  - 8 - 1.30 The PCR re-evaluation has also shown that even if GOO could deliver the project following the assumptions implicit in the SAR, the best likely impact for the 18 MIPS would be an ERR of about 9.8% compared to 14% estimated at appraisal. The shortfall resulting from project performance up to Credit closure were due to unforeseen technical or physical problems (still to be more fully evaluated). lhese have either substantially increased costs or decrased the potential irrigated areas. The general shortfall in ERR conpared to that evaluated at appaisal (and the PCR of the Orissa I Project) reflects the many problems and potential risks identified at appraisal which were not successfully resolved. The main determinants causing these problems include: (i) substantial cost ovemms; (ii) an implementation period to ful development of about double that assumed at IDA (and CWC) appraisal; (iii) ineffective water management; (iv) delays in forest and land clearance and agricultural development; and (v) slower dtan expected response to rabi irrigation, particularly for crops other than paddy. 1.31 There is insufficient data to evaluate the impact of the land consolidation (LC) componen, but the target area to be consolidated was achieved. However, fragmented land holdings have been a serious constraint to efficient land use in Orissa and aU indications are that the LC program has been successfuL Long term, the LC program should generate substantial benefits. Short term, the lack of coordination with OFD will limit the impact. The on-farm development (OFD) component has been less than successful, primaily because technical objectives of OFD were not fully understood and it was ratd as a target oriented construction program. The impact of OFD depends on the successful intoduction of rotational water supply, or at least more efficient management of the major irrigation schemes in which the OFD works were undertaken. This has not happened because of the practical difficulty of treadng one part of a major scheme differently from the rest of a command area (i.e. allocatng water supplies to reflect new cropping patterns). It is therefore concluded that few benefits will be achieved from the ORD component. 1.32 Other ProRect Resus. The institutional strengthening of GOO departments/agencies through training and other facilites were less than successful. The 000/ID) interest in the component is reflected by negligible disbursements under the Credit category (Part UT, Table 5 B&C). The institutional strengthening components have made little impact. There is no record of any achievements under the research and development (R&D) acdvity, primarily because capable ID staff were not interested in R&D which is considered to be a career dead end. Specialized training through technical assistance and the establishmentof a modest R&D and training facility may have a marginal impact There was virually no pogress in the nmodemizaton of the ID central design office mainly because speaized staff were not readily available, lack of budgetary provisions by GOO for equipment and facilities, and a general lack of prionty within the ID. Prject Sustainability 1.33 Te project will not be sustainable unless many institutional and technical issues identified during the last year of the Credit penod are addressed and resolved. GOJ/CWC/GOO must give the simil priorities to delivering project benefits as it does to the investment phase under irrigation development. Schemes need to be venfied and certified as functional before being handed over to O&M authoities. If the original design does not work, history has shown that it is unlikely to be xecdfied at some future date - unless the project is rehabilitated under planned expenditures. The ID must be provided with sufficient funds, staff and resources so that MIPs can be completed, remedial or dam safety works undertaken, and for O&M to be efficient. For MIPs to be sustained in remote and isolated areas, designs may have to be simplified (even sacrificing irrigation efficiency) so that site staff can understand and apply appropriate operadng principles. Fnally, there needs to be much stronger coordination between the agencies responsible for the irrigation service and delivering agriculural benefits. Formunately, there are some signs that GOQ/ID has recognized the importance of completing and sustaining the project and is believed to have initiated some of the necessary actions. Bank Performance 1.34 Project Preparation. The State of Orissa is one of the poorest in India and needed IDA support for the irrigation subsector if irrigation development was to be quickly expanded through the State. The line-  -9 - of credit' approach, although necessary for IDA to support construction of many scattered MIPs, was clearly premature given the institutional capability of the GOO agencies and CWC (with respect to this project). Nonetheless, IDA had reason to assume that a follow-up project would be successful given the efforts put into the Orissa I Project. However, the risk of GOO not being able (or willing) to deliver on the policy issues should have been better recognized at the preparation stage, given the problems already developing under the Orissa I Project. In particular, the preparation unalistically assumed that (OO would take the hard policy decisions related to sensitive issues such as insdtutional strengthening and manpower, budget provision for non-planned expenditures, and general improvements in the bureaucratic procedures. The preparation also underestimated the difficulty in GOO delivering on many non- construction activities. In addition, a critical preparation assumption was that CWC could effectively appraise MIPs and monitor project implementaton performance (para 1.55). Strong CWC supervision was essential to the project success and yet CWC performance under the Orissa I Project should have identified the risks involved. 1.35 Apaisal. The practical difficulties of delivering the project objectives in isolated regions of the State were underrated, particularly with respect to institutional capability at field/scheme level. Most risks were conrecdy identified but treated as project specific. Many risks were really generic subsector issues. The project success depended on the strengthening of weak technical departments without any deep evaluation of why the institutions were weak. It also depended on a basic change in GOO/II attitude towards iigation subsector development, namely to focus on irigated agricultural development rather than implementing civil work programs. IDA should have given more attention to performance problems under the Orissa I Project and the other on-going IDA supported project in Orissa. The real status of on- going MIPs (including other MIPs in Orissa outside the IDA program) should have been established before supporting the new project. For example, there should have been serious questions on why costs of on-going MIPs could not be established with any certainty seven years after the start of Orissa I Project. This should have indicated that basic planning and design of MIPs (and thus CWC appraisals) needed to be critically reexamined. Introduction of RWS operating principles should have been conclusively demonstrated on at least one existing MIP befo'e being put forward for the project. It is concluded that more atiention should have been given to the causes of the problems under the Orissa I Project before appraising a follow up operation. The signal given to the Borrower in the rapid processing the foliow-up operation (and confirmed in the PCR and PPAR of the first stage) was that IDA believed that the 'line-of-credit' approach was working. 1.36 There is evidence that some conditionalities proposed under the IDA appraisal did not survive the negotations stage, partly because of Bank pressure to minimize the number of project covenants in legal agreements. In particular, the appraisal recomwnended a mid-term project review which was not supported. An important conditionality which did survive was that disbursements on the 6 new MIPs would depend on substantial completion of at least 10 on-going MIPs. Consequendy no new MIPs were start In addition, important conditions of disbursement were included in the Schedule 1 of the Development Credit Agreement requiring that Certificates of Completion of irrigation canal systems be issued for each MIP before withdrawals could be made for minor canal systems and land consolidation within MEP conmands. These agreements were not rigidly enforced (para 1.58). 1.37 Sudervision. The supervision perfonnance of WDA was at best mixed. It ranged from almost complacency in the first two years after Credit effectiveness when the project was rated as having 'moderate' problems, yet none of the ten dated covenants were fully met on time. There were also a series of generic subsector problems related to forest clearance, land acquisition and MIP operation which were not being tackled. Lengthy supervision reports were prepared which did not focus on the real problems or issues. For example, the lack of an active Dam Safety Review Panel and the quality of MIP planning and design were treated as irritations rather than basic problems. On the other hand, relatively minor technical issues were reported in exhaustive detail, as was physical and financial progress when the data base was obviously inadequate. A major shortcoming in the first year of project supervision was inadequate attention to performance of the new components and Credit conditionalides which resulted from lessons learnt under the Orissa I Project. An important recommendation in the PPAR of Orissa I Project which was overlooked, was the need for a project impact evaluation. In fact, there is no convincing evidence that the PPAR findings were really followed up under project supervision. It is  - 10- concluded that undl about 1986, the project was supervised to standards and norms which were generally accepted throughout the Indian irigation subsector at that time. However, this intensity of supervision did not met the requirements of the project for a State such as Orissa. 1.38 Once the project was subject to more focused supervision, there was an increasing frustration with the lack of action by GOO/CWC on issues cridcal to project success. In May 1987, the project was justifably rated a problem project and remained such until Credit closure. It is undendable that GOV/GOO found it difficult to read the Bankes position since CWC ap,paed to be satisfied with project progress. In fact, G00/ID and CWC became increasingly bewildered that a whole range of problems had suddenly been designated serious by IDA after almost a decade into the 'line-of-credif operation. 1.39 The supervision of the dam safety issue (para 1.14) deserves special mention because of the seriousness of the problem and the way it was handled. IDA was becoming increasingly concerned about the implementation quality of IDA supported projects throughout the irrigation subsector. Special attention was given to Orissa II Project because of the reported poor quality of civil works, the large number of dams under the project, and nominal compliance with the covenant that GOO would maintain a Dam Safety Review PaneL A high-level and independent Dam Review Mission (DRM) was sent to Orissa by IDA three months before the Credit closing date. The technical findings of the DRM were thorough and professional. However, a DRM recommendation of extensive 'independent' consultancy (723 man months) with international recognized experience resulted in confrontation with GO which could not be resolved in the time frame left under the Credit. The DRM diagnsis was sound but the bed treatment could not be accepted in India (para 1.56). 1.40 IDA had two main agendas with respect to dam safety - dam safety as a subsectorissue and the more narrow problems widtin Orissa. An underlying concem for IDA was a more general issue of quality control and construction standards. The visibility of the dam safety issue at national leveL particularly related to CWC, ensured that project problems got entangled in much broader national issues on dam safety and quality control. These national issues did not have to be resolved under a tight project time frme. It is concluded that the visibility of dam safety issues and sensitivity to the recommended use of extenal technical assistance at the national level, domiinated the last few months before the Credit closing date. Mixing national and project issues hampered GOO in its attempts to take the required actions, not only related to dam safety, but also more general project problems. The outcome has been that GOO and the project may have lost out when there were encouraging signs that GO/IID were at least reconizing the problems and trying to do something about them. 1.41 Lessons Lar. This project appears to be a classic case of IDA committing itself to a follow-up. operation without fully evaluating an on-going operation which had strong indications of basic problems. It is concluded that a subsector type operadon for medium irrigation development in a poor State such as Orissa was premature in 1977 when the first Orissa Irrigation Project was appraised. IDA considered the project as a subsector operation, while GOO and CWC treated it as a specific investment project. The ID tread the project as a civil works construction progrm with little interest in obtaining the potential benefits. With these Perceived differences in project concept between IDA and GO1GOO, project problems inevitably followed as IDA stayed with a 'dme slice' concept which became increasingly less realistic. For example, IDA underesimated a key project risk that successful implementation and sustainability of MIPs would depend on effective and coordinated institutions. The institutional capabiliW for effective irrigated agrcultural development did not exist at appraisal and is still a basic problem (para 1.47). Solutions to many project problems depended on the willingness of GOO to address difficult subsector and statewide issues which could not be resolved at project level. It is also clear that the project was seriously under-supervised during the critical period just after Board approval. The line-of-credit' operation needed strong supervision during the initial implementation stages to confirm that GOO and CWC understood their commitments and obligations and had the resources to function effectively. It is therefore not surprising that many problems developed unchecked and the project diverted from the appraisal plan. Fnally, the way this Credit was closed was an unsatisfactory situation after more than a decade of IDA support to a necessary project in a poor State - and with the basic issues at last identified but unresolved. It may have been worth IDA staying with the project for one more year, if for nothing else to continue the dialogue with GOI/CWC/G0O on the lessons leant.  - 11 - BoQwer,,stnce 1.42 The project perfonrance reflects similar problems which have affected line-of-credit irigation development operations in other States of India. The Borrower's priorty has been construction activities with insufficient attention as to obtaining benefits through increased agncultumal production. There has been no real accountability for project performance since the department spending the money (Irigation Department ) is not responsible for delivering the benefits (Agriculture). Lack of attention to project development impact has meant no urgency to address the c of prHblems which have dominated the Borrower's poor performance through the project period. 1.43 frmowed's Strenths. Clearly the project has not performed as envisaged at appraisal but many of the problems could notbe tackled at a specific project leveL. Theexception to this has been the land consolidation component which was implemented in a sensitive and effective way and was probably more successful than any sinilar program in India. Perhaps the most encouraging outcome has been that GOU1CWC/GOO have stated to recognize the significance of the lessons leamt, not only for the project but through the irrigation subsector. 1.44 Lejjsj The main lessons to be learnt by the Borrower relate to: (i) local funding and the budgetary process; (ii) institutional capability, responsibility and accountability and project coordination; (iii) manpower and terms of employment, (iv) the role of technical assistance; (v) planing and design and appropriate technologyr, (vi) completing development and sustaining irrigation schemes; (vii) monitoring and reporting; and (viii) the need for diagnostic evaluation. ITere are also important lessons to be learnt related to the CWC role under the project (para 1.55). 1.45 LAl Funding and the Budga, Process. Lack of local funding for key activities has been an mportant project constraint, particularly for non-planned expenditures related to institutional sengthening. 011GO agreed to essential components related to strengthening the ID but did approve the local funding, even for items fuly supported by IDA. Budget commitments were generally based on statewide norms and decided at a Government level which was isolated flDm any deep understanding of the problems or requirements of a specific project - particularly for a project with many scattered schemes. It is concluded that the budgetary process has been seen by GOO primarily as an instrument to allocate fands and to control annual spending. The budgetary process has not been viewed or used as an essential tool of management to control and direct project performance. 1.46 Insitutional Performanc. Many officers in the ID project staff and the CWC demonstrated dedication and produced good results under difficult cdcumstances. However, individual efforts could not resolve th basic institutional problems - neither could the IDA financing for institutional strengthening which was not utilized. Most implementation shortomings stemmed from meffective project organization, burdensome bureaucratic procedures and resource constraints. It is recognized that basic improvements or changes to organizations are extremely difficult to make in any Indian state because any basic changes would have statewide and politcal implications. However, if the development of the rigation subsector in Orissa is to be successful, GOO needs to substandally improve the ID performance. As a first step, GOO should thoroughly evaluate the costs to the State of the insitutional problems, identify the causes of the problems and prepare a realistic plan of action. 1.47 Ihe project developmental impact (and the GOO inigadon subsector) depended on effective coordination and cooperation between the technical departments responsible for irrigadon, agriculture and other related activities. Through project implementain, coordinadon has not been effective, and in paricular, the worlcing relationship between irigation and agriculture has been nominaL The SCMC had litde impact in its task as the authority responsible for project coodination. Generally each organization followed its own budget provision, program and priorites d functioned as if each project acty was self-contained. It is concluded that coordination is weak because the main interest of departments relates to expenditures on target oriented programs with marginal interest in the impact of their progrms. GOO should also better define responsibility with accountability for obtaining acceptable project perfiomance  - 12- 1.48 SWfig. Present terms and conditions of Government employment are not attractive enough to place and keep qualified staffin remote areas or in jobs not perceived as benefiting a career. This would include all MIP locations under the project and almost all activites in the ID which critically neea to be strengthened (i.e. planning and design, M&E, investgatons, hydrology and R&D). Inability to provide staff incentves, limited career prospects and insufficient resources, all contribute to the staffing problems. Staff want to be located where they can live to reasonable standards. Poorly paid national staff are not motvated by project performance, but too often, the need to supplement their Govemment incomes. The preferred actvity of professional staff is construction of large works since most staff see the ID pnrmarily as a constructon agency. The ID staffing problems are compounded by frequent transfers, nearly total job security and no real accountability for perfornance. Staff morale is also affected by limited, unpredictable and often ineffective support for a specific project It is concluded that if the ID is to become an effective organization, the basic issues related to staffing must somehow be addressed by Government. 1.49 Planning and Design. A core problem which has affected both this project (and since 1977 when IDA started to support a 'line-of-credit' operation in Orissa) has been the poor project planning and design. GOO/ID have entered into the implementation stage of MIPs (and probably other irrigation projects) when they are still at the reconnaissance or (at best) the feasibility stage of planning. The one partial exception is the preferred engineering activity of designing dam and other large stuctures. Even in this respect, there is seldom sufficient investigation data (particularly hydrology) to safely move to detailed design of a major structure. All project MIPs entered the construction stage before designs were complete. This has greatly increased the project risks and has inevitably led to disappointing project impact. Many MIPs cost substantially more than predicted at appraisal and it has been virtually impossible for the ID to prepare realistic work programs when the works were not designed in detaiL 1.50 Irrigaion Technology. A disconcerting fact is that the operation of substantially complete MIPs do not follow the agreed design principles, namely rotational water supply (para 1.19). In fact, the RWS principles have not yet been effectively introduced anywhere in Orissa, including the pilot water management component of the Mahanadi Barrages Project (Cr. 1082-IN, 1981) which has been underway since about 1981. One reason for this seems to be lack of understanding and support for RWS within the ID, since RWS requires a level of discipline which apparendy cannot be enforced on most isolated MIPs. The ID needs to thoroughly evaluate what is the appropriate technology and operaing principles for MIPs. In particular, the ID should assess whether the operating principles inherent in the design of isolated schemes are practical. Whatever operating principles are finally adopted for MIPs, they must be demonstrated as workable with the staffimg and resources likely to be available on isolated sites. In addition, the ID should evaluate why few (if any) MIPs have explicit operational plans that can be understood and used by typical site staff and also why water management plans (WMPs) have proved of little value. 1.51 Compledng Developmntj. None of the MIPs have demonstrated that they are complete and fully developed and thus are ab- to deliver predicted benefits. A basic problem has been that too many MIPs were started about the same tme, with implementaton concentrated on main civil works (ie. dams and main canals). The result has been that funding and insttutonal resources have been spread too thinly. This pracdse has persisted because there is not specific accountability for project performance either of the irrigation service or agricultural production. Since there is little urgency to deliver benefits, there has been lack of commitment (and thus funding) to complete nmnor distribution works (particularly watercourses) after the main contracts are completed There are no reliable procedures to verify and certify that MIPs can work (and be managed) as designed. Land and forest clearance have also delayed and complicated completion of MPs. Ineffective O&M, uncoordinated agricultural development and weak extension services have meant that the irrigation services are not efficiently utilized. If the Borrower can resolve these problems, irrigation impact in Orissa could be substantially improved at relatively little cost 1.52 Monitoring and Reporting. A major problem through the project history has been lack of reliable monitoring and reporting. Most project reporting was restricted to unreliable information on physical and financial progress of either schemes or individual contracts with little useful data on perfonnance of investments. Data collected by ID (and CWC) were seldom reviewed in detail or checked for quality or quantity. The agricultual agencies have collected litfle useful data related to the project and thus there is  - 13 - still no reliable data on iTrigated crop yields. Monitoring and reporting was supported under the project yet nominal progress was made. It is concluded that the situation will not improve until Govenment treats monitoting and evaluation as essendal activities. 1.53 Diagnostic Analysis. A fundamental lesson to be learnt by the GOT/CWC/GO0O should be the lack of diagnostic analysis of both project inplementation and impact. The ID (and the CWC) should undertake regulary evaluations of project implementadon perfomance and prject impact and of their institutional capability. For much of the project period, the causes and consequences of most project problems could not established with any certainty. It seems that avoidance of diagnostc analysis is deeply ingrained in the public sector organizations since such analysis (and reliable data) would enable accountability for performance. Nonetheless, diagnostic analysis must be intoduced if the enormous investments in the rrigation subsector are to be effective and efficient. 1.54 Technical Assistance. After more than a decade of IDA support in Orissa, the capability of the ID has shown little or no imprvement. GOQ/ID have agreed that the organization needs to be resucured and modernized if it is to function effectively. It is concluded that this process cannot take place unless there is some input from outside the organization. Most staff have served in the ID since they started their careers and have only experienced the bureaucratic procedures and practices and technical standards of the State. It is concluded that unless the ID seeks technical assistance, there is no way the present shortcomings will be resolved or even fully identified. GOO will continue to get less than satisfactory impact from its investments in irrigation. The perfonmance of CWC also indicates that it would benefit from technical assistance in certain areas (para 1.55). However, for this project, GOI rejected extemal technical assistance on principle without evaluating the substance of the issue. The GOO/ID and CWC have great latent strengtha amongst its dedicated staff. What is lacing is an input from outside to advise and start the necessary esbcntial changes. The basic question for GO/GO is how it can expect an introvert and weak institution such as the ID to improve without assistance from outside the organizadon. Whether technical assistance is external or internal to India is an important but separate issue. 1.55 Role gf CW The CWC has made a major contribution to water resource development in India and remains a model for an apex organization which would gready benefit most other Bomwers supported by the Bank in the irrigation subsector. Yet under this project, the performance of CWC was less than satisfactory. The CWC role to appraise MIPs and monitor project progress was crucial to the success of a line-of credit operation, particularly in Orissa. However, the Appraisal Committee set up in CWC was never given the skilled staff or resources to effectively supervise the project Dedicated staff within CWC clearly tred to do the job but bureaucratic constraints, lack of rsources and support, and the inability of CWC staff to influence policy makers in the State all contributed to a weak performance. A major problem for CWC has been (and stiUl is) the poor standard of preparadon of MIPs sent for appraisal. Most MIP preparsmtions were incomplete and based on a limited and unreliable data base. Some 6 to 10 years after MIPs were appraised by CWC, designs of many sub-projects have still not finalized, there are persistent concerns over dam safety, the operadng principles of MEPs are still to be translated into workable plans, and there is no certainty as to project costs. This has meant that CWC has had to supervise implementation of MIPs still at the design stage, with all of the problems of accommodating a time bound work program It is concluded that if the CWC had enforced the agreed IDA standards for preparation of MIPs, few if any MIPs would have been accepted at appraisal. Unfortunately, it is not usually possible for CWC to reject projects which have been poorly prepared when they are politcally inportant to a State. It is recommended that before GOT (or IDA) give CWC a similar role under a line- of-credit' operation, they should thoroughly assessed what can reasonably be expected from CWC as it is presently constituted and staffed. Proiec Relationfshp 1.56 Through the project period, the Bank/Borrower relationship was generaly productive. At the working level in CWC and GOO, there was always a readiness to address the issues and attempt to resolve the problems. As would be expected when problems became politically sensitive (i.e. dam safety), the Bank/Borrower relationship became more formal and strained. GO/GOO and the departnents involved in Orissa were also understandably anxious when less than satidsctory perfofmance had to be addressed. However, a considerable amount of effort was spent by GOI defending general  - 14- issues on principle rather than substance of the project problems. This has complicated project relationships because project problems got entangled with national subsector issues. An example has been the treatment of the dam safety and the related issue of external technical assistance. The Bank's agenda on dam safety was simple - IDA wanted to be assured that the dams were safe for the well-being of the population living below the structures. The 0OO/ID shared this concer, but obviously had to consider the political sensitivity of any quesdons related to dam safety. The CWC had the additional problem of having to deal with two agendas in an extremely tight time frame - to tackle project supervision as agreed under the Credit and also consider the implications of the dam safety problems on their national credibility and accountability. Another difficulty which affected pxoject relationships has been that issues such as external technical assistance are debated at GOI level on principle. This approach often strained project relationships because to eliminate any useful dialogue on the substance of critical issues, particularly in the difficult peiod prior to the Credit closing date Proj Docunientation and Data, 1.57 The SAR and legal documents provided a comprehensive and detailed framework for IDA and GOI/CWCJC;OO during implementation. The conditionalities in the Project Agreement covered most of the issues which eventually became serious problems under the project. The Minutes of Negotiations wvere extensive and indicate that each of the core conditionalities had been discussed in detail by those attending negotiations. It is understandable that the documentation prior to Credit effectiveness appeared sufficient to successfully implement the project However, the appraisal documentation underestimated the risks and overestimated G000s ability to strengthen the ID and address the problems and issues already evident under the on-going Orissa I Project. During project implementation, the documentation and data was generally less than satisfactory. Some of the IDA supervision was not sufficiently focused, CWC generally produced superficial monitonng reports and GOO/ID had a basic problem with respect to monitoring and reporting throughout the project period. Lack of reliable infomation was the dominant reason why many project issues became serious problems. 1.58 Cpmpliance with Covenants. The Borrower did not fully comply with most of the project covenants and none of the dated covenants were fully met on time. The main problem has been that GOVGOO appeared to have entered into many project covenants which were extremely difficult for the project authorites to deliver (to IDA's satisfaction) when political determinants had to be taken into accountL Many key project covenants were not project specific and had subsector or statewide implications (ie. local funding; provisions for 0&M; institutional strengthening; provision of qualified staffimg; and cost recovery). Full compliance with these covenants was generally beyond the control of individual departments/agencies since they involved statewide policy issues. It also seems that IDA and Borrower (both 001 and GOO) may have had different understandings as to what specifically was expected under the covenants even though all were carefully drafted and explicit. A disconcering fact is that many covenants were not full complied with by G01/GOO through a decade of IDA support in Orissa. Yet for most of the time, IDA showed little real concern It is concluded that if IDA expects the Borrower to take covenants seriously, IDA must be prepared to enforce them. Conclusions. 1.59 The project has been less than successfuL The lessons to be learnt from this project are important and numerous, but most can only be resolved if GOU O0 is able to akle the underlying cause of most problems, namely institutional weakness. IDA can assist in this process but not through the 'band-aid' components to strengthen institutions. Ihe ir.igation subsector approach in India has been for IDA to support projects with improved 'efficient' technology on the assumption that responsible institutions can be strengthened to match the technology. The results have been disappointing as illustated by the performance of this project It is concluded that a less risky approach would be to support operaions carefully fonrulated to reflect the existing institutional capability.  - 15- PART 1: PROJEC' REVIEW IRBM BORROWER'S PERSPECTIVE (The Borrower was requested to provide a review, but none was received.)  ORISSA IRRTGATION n PROIW (t 1Xn- Bm ii aw LANo, :Lhk Pu Yem 682-N Orissa Agricult Stregth agiukura sporting 1977 Compled Stae-wideoperation. Delopwmn servic particularly tension and w*eseem 740-IN Osarrigan rum aslice ofnmedium irigation schemes. 1977 Completd First ste ofproject undr review. command area developmenL 80B-IN Oujat Mediun Improve peformance of MPs trough 1980 Completed Siilar line of credit opern to Orisa riaion I project Iigaio (1) new plbmhi, design consiuction mnd with CWC responsie for raisal ead supvis O&M cneia.L 843-IN Hiyana rriato (1) bmovam to irrgatin system 1980 Completed Focus was mnwr irgaion systes with rtatoa wae ma _agenent; lidng watcouse channels. suply and rdated imp lmentio and manmet poblm 1082-IN Ma i Ba7Sges Consrctko of two new bawrages on the 1981 Completd Olir Baik supported ao project mn Orissa durin Mahmnadi Rive and wawe mooanagmen project period Watfr management Pilot reslt reported to pilot cmmponet. be used on project 1108-lN MP Madium rime slice of mediun irrigation t981 Compkled Simil in concept and objecives to Oris brigo 11 with Iriguaion deveoepropgam CWC appraisal and supevision 1289-IN Subi ekha Provide iable water supply to inrigate 1983 Completed Tri-state opertion with State of Orissa as one of befitng Irrigation 25QOO O hsha wnd*r ilustiaLsnd munci Staes. users; flood control an Onssa and West Bengal. 1319-lN Haryanlbrigationl mprovement o inor m iuigation sysems 1983 On going Sminar war matamt concept as for Onssa B Pject; wih rottional water supply; intiouonAl a follow up to Hayana Irigaion I (Cr. 843-IN) 1523-IN Natinl Agrcula Mult-state (subsector) opaton to 1985 On going One of benefting sta is Orissa Extension I impove aricultual extesion svices. 1770-IN Natiol Wat Cp an iw ement to existin 1988 on-going o tneu inigae a8iulud prodon thraough rmed Manageme_t Project irigationm schmeu istituional wors man d n t seg . nodn concepas strengtening *i'th focus on operatiot that have couny-wide ppcand rct typicld We manaem t, and Utrng. istutioa capability of stain tmion departents 1/ Only penject relatd to Oriss Irrition II Project ame listed S Iii .~~~~~~~~~~~~~~~~* r  - 17- km~~~~~s RSSA IRIGATION 11 PROECT (iit 13,97-T' Table 2: hisct:Ib lu Date Date Planned Acu - Identification August 1982 August 1982 - Preparation/Preappraisal NA NA - Appraisal Mission 9/12/82 9/12/82 - Lon Negodations May 1983 5/16/83 - Board Approval 6121/83 7/7/83 - Loan Signature 1/30/84 9/16/83 - Loan Effectiveness 5/1/84 12/14/83 - Loan Completion 3/31/87 3/31/88 - Loan Closing 12/31/87 3/3U88 The Project was conceived as a second 'time slice' of the Orissa Irrigation I Project (Ln. 720-IN, 1977, US$ 58.0 M) to continue the on-going medium irigation program. There was no formal identification/preparation missions with these activities covered under expanded supervision of the first stage project. The justification for the second time slice was the need to address constraints/problems identified under the first project and to complete the on-going MIPs stared under Orissa Irrigation L Table 3: Loa Disbusnnt-/ Fiscal Year 1983 1984 1985 1986 1987 1988 --SDR M-----_ ARM"EStimat 6.0 24.9 50.6 77.6 97.1 (97.1) Acual 0.0 9.4 25.3 44.0 57.9 59.2 Acnal asofEiat= - 38% 50% 57% 59% 61% 1/ The amount of the Credit was SDR 97.1 million; the undisbursed balance of SDR 35.9 million was cancelled. Dat- of Final Disbursement was 12/13/88 on recovery of balance in the Special Acomunt.  INI1A ORISSA lRGATION I PROJECr (Credit 1397-IN' Pto iec anetadoe I/ Appraisd Revised Acual 1ndicatUs Ea~uxL E~unsIimat_ PCR Vtc Comments (unit) % Irrigable ona (ha) 108,000 110070 47,410 (43%) Area irrigated is xecorded as geate han de area with completed inigaion works. Dams (No) 18 18 13 72% Noofdamsdesignatedas comnplete by 0R0C03. Construction Main System (ha) 108,000 110,070 79,500 (72%) Consructo Minor Systen (ha) 108,000 110,070 41,950 (38%) LandClmeance(ha) NA 24,123 14679 (61%) Notsteaapraisal. Rsettement (famlies) NA 4705 1480 (31%) NotestitedatappraisaL 2. New MIN (No) 6 - - 0% Excluded fiSm the project4. 3. LandConWUdaf= (ha) 160,000 - 160,000 100% PCRestiinaeofactual performance. 4. On-Farm PaDkXnnent (ha) 48,000 - 53,640 112% No breakdown of works under component. 5. R&D, Training, Technical Assistane, M&E (SDR M) 930 - 4 (negligible) Costs where 100% supported; pefna indaor taken as categoy disbursement. NM 1/ Based on prject progress rep of 3f31/88. Previous supervision reports indicates that performance data may be opimistic. The perfance indicatr does not reflect a consdable amount of remedil works believed to be necessary to make MIPs fully orational. 2/ Perentages in brckets relate 000 reported progress to revised estmate of project imgated area 31 The IDA Dam Safety Review of November 1987 concludes that most dams need additional or remedial work to be designated as safe. A/ The 6 new MIPs were included under the project conditional on completion of 10 on-going MPs. This was not achieved so the component was not started. / No data available of what was achieved under the project, since the component was 100% supported under the Credit, the actual I-' dibursement at Credit closure has been taken as the performance indicator.  ORISSA IRRIGATION II PRO1ECT (edit 1397-INM Proict C9mupletim R=pr Project Costs and Fiani A - P>iec Costs Estimate Total (QWM= ExpendiUves Pknec A In=l azimto IEwjoc rime Slice costsZ/ Lca FC Total LIca FC Total &AtR W