India - West Bengal Agricultural Development Project

RepRd N&. 12140 India Impact Evaluation Report West Bengal Agricultural Development Project (Credit 541 -IN) (Shallow Tubewells Component) uWW29, 1993 Operations Evaluation EkTartment FOR OFFICIAL USE ONLY e 15- ICROGRAPHICS na :_ M Report No: 12140 IN WR Type: IER X54, 4~ ~ ~ ~ A .~; "Na .5,. At F54 -4 ._ 211. -'xf*; 11' 1 . _,g .4MA. 11 W-Ar - - .'~'-~"~ 4 ~ ~ '~~ ~ - ~4~'.' 4 ~41 .44. 4~4. --4X_. 4~. .. <.'..44N - 4.4~*4g4 *Wgmw MM444 'V I 4~41 TOM4 Km MVg - 4 ~..~~. 1 .44 90 4.-44W- 4 ~ --4-4&~44.~."4'4 ~4.*-xv - die 4 4 .444 ,W-0 Ell. ~~~~~ '.74~ .... ". ~ A-~4. ~~V Q . 4-44 %4' I44 4'Ar4.4-444~1. ~ (~'~N, 2 ABBREVIATIONS AND ACRONYMS ARC - Agriculture Refinance Corporation DTW - Deep Tubewell FRR - Financial Rate of Return GO1 - Government of India GOWB - Government of West Bengal IDA - International Development Association IE * Impact Evaluation MIC - Minor irrigation Corooration PCR - Project Completion Report RLI - River Lift Installation RMC - Regulated Market Committee SAR - Staff Appraisal Report SEB - State Electricity Board SMB - State Marketing Board STW - Shallow Tubewell SWB - State Water Board SWID - State Water Investigation Directorate WBADP - West Bengal Agricultural Development Project CURRENCY EQUIVALENTS The currency is the Indian Rupee (Rs) Average Exchange Rate per US$1 FY75 Rs 8.01 FY76 Rs 8.91 FY77 Rs 8.88 FY78 Rs 8.48 FY79 Rs 8.13 FY80 Rs 8.01 FY81 Rs 8.01 FY82 Rs 9.17 FY83 Re 9.79 FY84 Rs 10.85 FY85 Rs 12.27 FY86 Rs 12.24 FY87 Rs 12.86 FY88 Rs 13.13 FY89 Rs 16.15 FY90 Re 16.97 FY91 Rs 18.74 FOR OIVCIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A. Office of oirector*General Operations Evaluation June 29, 1993 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: impact Evaluation Report an India West Bengal Agricultural Development Project (Credit 541-IN) (Shallow Tbewells Component) Attached is the Impact Evaluation Report on India: West Bengal Agricultural Development Project (Credit 541-IN) (Shallow Tubewells Component) 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 authorization. FOR OFFICIAL USE ONLY IMPACT EVALUATION REPORT INDIA WEST BENGAL AGRICULTURAL DEVELOPMENT PROJECT (CREDIT 541 IN) (SHALLOW TUBEWELLS COMPONENT) TABLE OF CONTENTS Page No. Preface ......................................... i Executive Summary ..................................... iii 1. INTRODUCTION ...................................... 1 PurposefoftheStudy .................................. Background ......................................... 2 Major Conclusions of the PCR/PAR ......................... 3 Focus and Outline of This Report ......................... 3 II. FARMER SURVEYS ................ ................... 4 Ill. SURVEY FINDINGS .................................... 4 STW Use ........................................... 4 Impacton Cropping ................................... 5 Economic and Financial Returns .......................... 5 Social impact ....................................... 7 IV. ISSUES AND LESSONS ................................ 7 A. Groundwater Regulation .............................. 7 B. Diesevs. Electric Power ............................ 9 C. Financial vs. Economic Returns ......................... 10 REFERENCES .......................................... 12 Annexes 1. TheProject ......................................... 13 2. Farm Survey Arrangements ............................. 21 3. Survey Findings ...................................... 23 4. Financial and Economic Analysis .......................... 34 5. Social Impact onWomen ... ............................ 44 Ma IBRD 10905R1 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 authorifation. IMPACT EVALUATION REPORT INDIA WEST BENGAL AGRICULTURAL DEVELOPMENT PROJECT (CREDIT 541-IN) (SHALLOW TUBEWELLS COMPONENT) PREFACE This is an Impact Evaluation Report on the Shallow Tubewells Component of the West Bengal Agricultural Development Project for which Credit 541-IN was approved April 22, 1975, in the amount of US$34.0 million equivalent. The Development Credit Agreement was signed April 28, 1975. The Credit became effective on August 28, 1975 and closed on March 31, 1981, after a 12 month extension. The last disbursement was made on July 15, 1981. The Operations Evaluation Department (OED) audited the project in 1985 following the preparation of a project completion report, dated March 18, 1985. OED selected this project for impact evaluation in 1989, to form part of a series of such evaluations of World Bank experience with irrigation development. After consultation with Bank staff, an OED mission visited India in 1989 to discuss the evaluation with government officials, participating banks, farmers and others; this mission also arranged for a local consultant to carry out surveys of project beneficiaries. Their kind cooperation and valuable assistance in the preparation of this report is gratefully acknowledged. The draft Impact Evaluation was sent to the Borrower on May 28, 1993 for comments. No comments were received. - iii - IMPACT EVALUATION REPORT INDIA - WEST BENGAL AGRICULTURAL DEVELOPMENT PROJECT (Credit 541-IN) (SHALLOW TUBEWELLS COMPONENT) EXECUTIVE SUMMARY INTRODUCTION proximity to the STW farmers in the sample. 1. This study revisits the West Rengal Agricultural Development Project (WBADP) 4. The main findings of this evaluation located in the northeastern region of that confirm that the project's STWs were highly State. Its purpose is to re-evaluate the favorable investments for the economy, shallow tubewells component of the earning economic rates of return of 50 per- project, particularly for its impact on cent or more. They also confirm findings of beneficiaries. A companion study of a previous evaluations of this project, how- shallow tubewell project located close by in ever, that the financial impact of STWs zn Bangladesh has been reported separately. their owners was somewhat less favorable. Financial rates of return were variable, 2. The tubewells consisted of 3 inch ranging from 10 percent to over 50 percent diameter pipe and brass strainer, equipped for two major groups. Interpreting the with small centrifugal pumps and diesel or meaning of such mixed results poses an electric motors. They were expected to interesting challenge for this report. irrigate 7 to 12 acres. Farmers and groups constructed about 33,200 STWs compared 5. The Survey estimate of the net area with the appraisal estimate of 18,000, and irrigated per STW at full development (4.3 fitted an additional 22,500 wells with acres) exceeded the PCR estimates (2.7 pumpsets. The PCR, completed in July acres), but was still much less than 1984, concluded that the project was a appraisal estimates (10.4 acres). STW success and that the shallow tubewell owners accounted for about 75 percent of (STW) component was largely responsible the srea irrigated; sales to water purchasers for this. OED's audit, completed in March accounted for the rest. Each STW serviced 1985, concurred. about two water purchase s on average. The financial analysis reported above 3. An OED mission visited project farms concentrated on measuring returns to in 1989 and made arrangements with a resources under the owner's control; it thus local consultant to organize collection of counts income from water sales, not the farm survey information. This report relies crop income generated by water sales. The mainly on that information. It was obtained economic analysis concentrates on principally from 110 farmers whose STWs measuring all crop income generated by were financed under tlhe project and from STW irrigation, ignoring cash transfers for non-irrigators who were selected for their water trading. - iv - 6. Compared with the PCR findings, the 10. In retrospect, the initial wariness on impact Survey found even greater interest the part of the Bank towards "unregulated" in Boro paddy and even less interest in exploitation of the groundwater resource, wheat. This finding of a strong trend similar in many respects to its wariness also towards Boro paddy agrees with similar in the Shallow Tubewell Project in findings of the impact evaluation of the Bangladesh, was ill-advised. There was a companion project in Bangladesh. tendency on the part of some project planners to propose on-lending criteria and 7. Estimated incremental crop yields are other project regulations, av if the purpose fairly consistent with earlier estimates, of public policy might be to limit the though Boro yields are somewhat higher development of groundwater to that attain- (1.9 to 2.0 metric tons/ac); all estimates of able through STW technologies alone. They boro paddy yields exceeded average yields undervalued the competitive market factors for the districts. Clearly, most other that would discipline such development. methods of irrig,ting the boro crop are (The Regional staff's disagreement with inferior, in their current practices, to STWs these sentiments is flagged in para. 4.8.) (and presumably other tubewells). 11. Fortunately, other planners saw STW MAIN ISSUEs investments as occupying a vital, though likely temporary niche in the evolution of 8. Gromdwater Reguatiron. The SAR groundwater development. Their vision warned that it would be essential in the prevailed, and the original intention of long run for West Bengal to enact appro- promoting rapid installation of STWs was priate legislation to ensure the most met. Subsequent groundwater operations efficient use of the State's groundwater supported by the Bank in West Bengal resources. Confirmation was obtained (Minor Irrigation Development, Credit 1619- during negotiations that GOWB would use IN) made fewer demands for groundwater its best endeavors to introduce this legis- regulation. lation as soon as practicable. Assurances were also obtained at negotiations that 12. Diesl vs. Eectic. Project appraisers detailed well spacing and other technical expected that 75 percent of the wells criteria would be worked out by the State would be electric powered, and 25 percent Water Board and embodied as conditions to diesel powered. Of the pumps energized, be observed by all lending b.nks when 78 percent were in fact diesel powered and approving STW loans. 22 percent electric powered, a complete reversal of what had been forecast. While 9. Some of these prior concerns and this reversal was of no great consequence expectations proved false or misleading. to the wider project performance, it is none- For example, the PCR argues that the use of theless an intriguing feature of its design overly conservative norms for working out and implementation experience. groundwater recharge led to the (false) early conclusion in some quarters that much of 13. Appraisers of expected user prefer- the project area was already over-exploited. ence regarding the two types of power for This regulatory effort needlessly slowed pumping identified a range of factors early project implementation. influencing their choice. The predominant -v- factors, which seemed to dictate in favor of economic returns. In this respect, the choosing electric power, Included the impact evaluation echoes similar observa- government subsidy of electricity tariffs - tions by earlier evaluations of this project. itself a response to some of the other factors identified. Ironically, such subsidies 15. The disparity between the results of had the dual effects of expanding demand the two forms of analysis plus the apparent for electricity while diverting potential public onset of marginal financial returns for some funds away from investment in its supply. farmer groups suggest a need to focus the It was the latter problem - of erratic supply, Bank's analytical efforts in such situations including shortfalls in distribution - that more on the related policy issues. India seems to have had most influence, stands to gain from reducing distortions to eventually, on user choice of power for relative prices. The lower financial returns pumpsets, although the greater number of enjoyed by STW investors illustrate what pumps to be energized was also a factor. Bank policy analysts have said of the Indian This problem is not apparent, however, in agricultural economy in recent years: the follow-on project, in which electric namely, that government should rethink its powered pumpsets predominate. role in output markets, particularly its procurement arrangements for rice, which 14. FAwnesdl versus Economic Asses- depress producer prices. Unless the gap ment. Part of the observed disparity between economic and financial prices for between finticial and economic returns for (project) outputs is narrowed, there is little STWs reflects the deliberately different scope for reducing or eliminating thp perspectives adopted by the financial and electricity subsidies that either undermine economic analyses mentioned in para. 5 the financial strength of state electricity above. But even after accounting for these boards or burden government budgets. The differences in perspective, disparities experience of WBADP suggests that policy remain: economic rates of return still tend reforms are an essential requirement for to be more favorable than financial rates of improving the impact of STW investments return. Differences in yaluations of costs on beneficiaries in India. and benefits in these analyses favor the IMPACT EVALUATION REPORT INDIA - WEST BENGAL AGRICULTURAL DEVELOPMENT PROJECT (Credit 641-IN) (SHALLOW TUBEWELLS COMPONENT) 1. INTRODUCTION PURPOSE OF THE STUDY 1.1 This study revisits the West Bengal Agricultural Development Project l-cated in the northeastern region of that State astride the River Ganga. Its purpose is to re- evaluate the project's shallow tubewells component. It is one of two impact evaluation:) of shallow tubewell irrigation investments financed by the Bank in the lower Gangetic plains. A companion study of a shallow tubewell project located close by In Bangladesh has been reported separately.1 The two studies are part of a series of Impact evaluations whose aim Is to review the sustainability of selected irrigation projects and measure their long-term technical, agricultural, environmental, social, and institutional impact. Other, recent reports in the series have addressed the outcome of (surface) irrigation investments in Mexico, Morocco, Thailand, and the Philippines.2 1.2 Like the shallow tubewell impact evaluation in Bangladesh, this study of shallow tubewells (STWs) in West Bengal deals with groundwater development. STW technology allows the private sector to take the leading role, both in ownership and operation of the wells and pumps and in their manufacture, trade and repair. In contrast, gravity Irrigation schemes, which are the subject of most of the other impact evaluations in this series, typically require the public sector to take the leading role. Even with STWs, however, certain public interests remain evident. The study will expand on the earlier Staff Appraisal Report (SAR), Project Completion Report (PCR), and Performance Audit Report (PAR) accounts of these private and public interests. 1.3 The primary focus of the study, however, is the impact of STW irrigation on farmer costs and benefits. Because STWs offer quick returns, the PCR and PAR were in a good position to evaluate their short-term effects. The study will reexamine these completion reports' medium- and longer-term projections of farmer costs and benefits, machinery life and repair costs, and the interaction of the STW investments with the durability of the groundwater resource itself. It also examines perceptions of Irrigated agriculture in West Bengal at the times of project approval (1975), and completion reporting (1984). These perceptions show the evolution of government and Bank thinking about shallow tubewells and their role in the subsector. 1. OED Report No. 11031, Impact Evaluation Report: Bangladesh - Shallow Tubewells Project (Credit 724-8D), August 20, 1992. 2. OED Report No. 8494, World Bank Experience with Irrigation Development, SocIo-Economic, Insdtutional and Technical Impact and Lessons, (in six volumes), March 1990. -2- BACKGROUND 1.4 In 1975, the y,ear of project appraisal, nearly two-thirds of the area of West Bengal was cultivable, end cropping intensity was about 127 percent. Rice - the most important crop - accounted for 70 percent of the cropped area; other crops grown were lute, tea, potatoes, chillis and a wide variety of kharif and rabi vegetables. The State had just passed through a period of severe food deficit, and the Government of West Bengal (GOWB) decided to seek improved food security by accelerating development of minor irrigation (Annex 1, paras. 1-4). 1.5 The main thrust of the project proposal that emerged was towards private sector STW development. Other components were to develop some deep tubewells and to complete existing deep tubewells (DTWs) and riverlift installations (RI.s) (Annex 1, pares. 5-10). 1.6 Shallow groundwater levels made most alluvial areas suitable for STW development, particularly in the short term. In the long term (at least 20 years, which is more than the life of a STW), the SAR argued that intensive development would likely lower the water table to a point where typical STWs equipped with centrifugal pumps might fail, and have to be replaced with DTWs. The short- tc medium-term strategy advocated for irrigation investment was to concentrate on rapid development of STWs, and to lay the foundation for future expansion of DTWs and RLis. 1.7 The STWs, consisting of 76 mm (3 in.) diameter pipe and brass strainer, were equipped with small centrifugal pumps and 5 hp diesel or electric motors. These wells had yield capacities, it was thought, to irrigate 3 to 5 ha (7.4 to 12.4 acres). Because of small farm size and fragmentation of holdings, individual farmers who installed wells frequently sold water to neighbors, sometimes to gain access to their own land fragments. 1.8 The project sought to assist in financing 18,000 electric and diesel powered shallow tubewells (75 percent of the wells would be electric-powered, the rest diesel powered), plus the other minor irrigation components. Its total cost was estimated at US$67 million. IDA provided a $34 million credit, about 45 percent of which was earmarked for installation of the STWs. 1.9 The project area covered six districts of Hooghly, Burdwan, Nadia, Murshidabad, Maida, and West Dinajpur. Soils were all alluvial. About 55-60 percent were pervious foams, suitable for wheat, vegetable and olitorius jute, while the rest were impervious clay soils, particularly suitable for Boro paddy. Regarding the size of the area's groundwater resources, the SAR estimated that extraction by a further 18,000 STWs and 300 DTWs would constitute about one-third the calculated recharge, thus projecting that their demands on groundwater were well within available supplies. Other districts were excluded from the project area largely because of uncertainty about their groundwater supplies. The legal agreements contained two provisions thought desirable for sustaining groundwater supplies even in the project area. These involved well spacing criteria and the introduction of appropriate groundwater legislation. .3- MAJoR CONCLUSIONS OF THE PCRIPAR 1.10 The PCR, completed in July 1984, concluded that the project was a success. OED's audit, completed in March 1985, concurred (Annex 1, pares. 11-17). 1.11 According to the PCR, the minor Irrigation components were on the whole implemented as envisaged. *n particular, farmers and groups constructed about 33,200 STWs compared with the appraisal estimate of 18,000, and fitted an additional 22,500 wells with pumosets. Of the 55,700 pumpsets energized, 78 percent were diesel powered and 22 percent etectric powered. 1.12 Following initially slow progress, IDA conducted a mid-term review in January 1979, and this led to reallocations of the credit favoring STWs and pumpsets. It had become apparent, from an all-India groundwater review, that districts previously excluded from STW investments possessed, in fact, adequate groundwater resources. Accordingly, the project area for STWs and pumpsets was expanded to cover the entire state instead of the original six districts. 1.13 Actual project cost was US$59.3 million, or 11 percent below appraisal estimates. The PCR had difficulty in ascertaining the project's production impact because of the dispersed nature of the investments. However, it cited a substantial increase in crop yields, especially for paddy, wheat, potatoes and jute. Farmers proved resistant to reducing cultivation of water-intensive Boro paddy in favor of other crops, mainly wheat. Cropping intensities increased significantly. 1.14 The re-estimated ERRs were 47-50 percent for shallow tubewells, 21 percent to over 50 percent for deep tubewells, and 4-50 percent for river lift installations. The re- estimated average economic rate of return to minor irrigation overall was 42 percent, compared with 55 percent estimated at appraisal. 1.15 For STW development, the main disappointments related to poor extension, a lower net irrigated area per unit, and declining loan recoveries. Focus AND OuuNE oF Tis REPORT 1.16 The study's main thrust is to learn how the passage of five years or more have treated the above PARIPCR conclusions about the project's impact. 1.17 A major reason for locating this evaluation study in India is the lack of previous impact evaluations of any Indian irrigation project. The main reasons for focussing the impact evaluation on the STW component of WBADP are: a prior decision to conduct an Impact evaluation of the Shallow Tubewell Project (Cr. 724-8D) next door in -4- Bangladesh, thus presenting an opportunity to contrast ir ipacts;3 the emphasis given to STWs at the appraisal of WBADP; the even greater emphasis on STWs that emerged during Implementation; and the complexities of attempting a broader, WBADP-wide impact assessment. This does not masn that the DTWs, RLIs, pumpsets, and institutional components of WBADP do not deserve study. The institutional and other problems of the group-oriented technologipe and diffeiing perspectives over what pumpsets were used for (Annex 1, para. 12a), make tnese other components fascinating areas for review. The decision to focus on STWs alone was a pragmatic choice. 1.18 Part II of this report describes the farm survey conducted in the project area in 1989 in terms of its objectives and sampling frame. Part III examines the survey findings, comparing them with expectations at appraisal and completion. Part IV presents issues and lessons arising. 11. FARMER SURVEYS 2.1 Beginning in March 1989, after discussions with GOWS and project agencies, OED contracted with a local consultant to organize collection of farm survey information. This 1989 Survey sampled several categories of STW farmers, from 3-5 development blocks within four districts of West Bengal: Nadia, Malda, Murshidabad, and Burdwan. Non-irrigators and other categories in the sample were selected for their proximity to sampled STWs. As for the latter, this report relies mainly on information obtained from 110 farmers sampled from lists of those who borrowed funds to install STWs under the project (Annex 2). 2.2 Survey investigators advised that survey respondents probably understated the not Incomes they received by at least 10 percent (Annex 2, para. 4). The possible significance of this is discussed in Chapter III. Ill. SURVEY FINDINGS STW USE 3.1 Of the sample of 110 STWs, 80 of their pumpsets were diesel powered, 30 were electric powered. So the sample's revealed preference for diesel as the motive power of choice coincides with the known similar preference on the part of project participants. Part of the reason for the historical dominance of diesel over electric has been problems with electricity supply. 3. A preliminary contrast of the two project outcomes pointed to a wide difference in their areas irrigated per STW, which was intriguing. However, subsequent inspection subdued this early interest in making contrasts. The preliminary review had overlooked design differences between the two projects' STWs: the ones in Bangladesh had 40 diameter wells; the ones In West Bengal, 3* Naturally, there were corresponding differences also in the designed ratings of pumps and motors. .5- 3.2 The number of STWs found to be out of commission suggests that the estimated rate of attrition among STW investments is fairly low, at 3 percent. Other survey indications of the impact of project operations, such as area irrigated, allow for this attrition. Average area irrigated by STW owners was 3.24 acres. 3.3 In addition to the 110 STW owner beneficiaries defined by the sample, their STWs supported 210 water purchasers. The average area irrigated by water purchasers per STW was 1.04 acres, thus bringing total area irrigated per STW to 4.28 acres. This latter estimate of the net area irrigated per STW at full development (4.3 acres) exceeded the PCR estimates (2.7 acres), but was still much less than the appraisal estimates (10.4 acres). The average revenue from water sales was Rs 747 per STW. IMPACT ON CROPPING 3.4 Compared with the PCR findings, the Impact Survey found even greater interest in Boro paddy and even less interest in wheat, yet larger areas irrigated. Considering the Boro paddy crop's heavy demands on water, the larger area irrigated was especially significant. Owners' cropping patterns on clay soils fell somewhere between the SAR and PCR estimates: without STWs, a cropping Intensity of 100 percent - all aman paddy; with STWs, a cropping intensity of 161 percent - mostly Boro (84 percent of net cropped area) and aman paddy, but also some potatoes. On loam soils, the Survey results proclaim a much stronger role than expected for boro, now 61 percent of net cropped area, and a much reduced role for -.heat, now only 3 percent; the increase in cropping Intensity overall, compared to the situation without STWs, is moderate: up 15 points to 162 percent. 3.5 At impact evaluation (E), estimated incremental yields are, again, fairly consistent with earlier estimates. Boro yields are somewhat higher (1.9 to 2.0 tonneslacre), as are potato yields; this time, however, the potatoes are grown on clay soils, not loams. Wheat yields are somewhat lower. 3.6 All three sets of estimates reveal heavy boro paddy yields compared with average yields for the districts. Clearly, most other methods of irrigating the boro crop are inferior, in current practices, to STWs. EcoNomic AND FINANCIAL RETuRNs 3.7 The economic analysis (Annex 4) confirms that STWs were highly attractive investments for the local and Indian economies to pursue. Estimated ERRs exceeded 50 percent for each of the soil type situations and for the survey average. Even allowing for the many approximations that embellish such measures, this report concludes that earlier evaluations were correct in giving strong endorsements to this form of irrigation development. 3.8 The financial analysis for STW owners, on the other hand, suggests that private sector incentives for pursuing STW investments cannot be taken for granted (Annex 4). Financial rates of return for irrigation of clay soils averaged above 50 percent, -6- but the return for loam soils was only 10 percent. These rates compare with PCR estimates of 21 percent and 25 percent respectively (see accompanying table). The overall average financial rate of return (FRR) for the three districts was 23 percent. Being real rates of return, the lowest rate among these was adequate to handle nominal interest charges on loans (10.5 percent per annum), but not some of the owners' own labor and management efforts. FINANCIAL RATES OF RETURN Burdwan District Nadia District 3 IE Survey (clay) (loam) Districts PCR 21% 25% N.A. IE > 50% 10% 23% 3.9 One possible explanation of the low returns may lie with survey investigators' own reservations about the survey data (para. 2.2 and Annex 2). Were it true, for instance, that survey respondents on loams understated yields and overstated operating costs by one-tenth, their FRR would have been 20 percent instead of the 10 percent reported above. Such a more favorable rate is more consistent with the strong interest in STWs taken by farmers. Yet one should not discount other reasons for this interest. A declining recovery rate for credit in West Bengal (and other states), as reported in the Performance Audit Report (para. 34), might offer another credible explanation - borrowers taking advantage of lax credit recoveries. It is also possible that the low FRR estimate is a warning, not about past levels of confidence in STWs, but of future doubts. 3.10 Disparities between economic and financial rates of return also raise an issue. Pai t of the disparity may lie with shortcomings in the analytical conventions used to handle what are obviously delicate financial understandings between STW owners and water purchasers. Had the financial analysis ignored ownership or control of land, the FRRs - for the merged resources of owners and water purchasers - would have risen a few more points.4 Such merged analysis ignores payments for water sales, substituting instead the financial returns and costs of growing water purchasers' crops. These more attractive returns suggest there is room for STW owners and water purchasers to negotiate terms of water sale that make STWs more viable financially. But even allowing for sur-e adjustments, a considerable gap between economic and financial returns still rem-..ts. Tendencies in all three evaluations (SAR, PCR and IE) for FRRs to be less than ERRs establish this as an issue deserving further review. 4. Interestingly, the SAR anticipated the problems facing STW owners who needed to sell water. 'Investors able to use all the water on their own land would have better returns than those selling some water' (SARI para. 6.03). -7- Socal IMPACT 3.11 The Survey did not reveal any information that could describe the impact of STWs on living conditions of women or other discrete segments of rural society. However, other studies as well as the findings of this one suggest that their impact can be mixed. 3.12 Though Irrigation, Including that through STWs, has increased household income, provided new employment opportunities and Improved domestic water supply, it has often rewarded most those fortunate enough to be able to bear the risks of investment. For example, women from differing stratas of rural society have not enjoyed the benefits equally. Since landownership and wealth follow caste lines, studies reveal that such irrigation benefits have even widened the gap between the opportunities available to landless and landed women (Annex 5). That is, although there are thus greater agricultural labor opportunities for landless women - a major plus - many still remain on the margin of survival. 3.13 Offsetting this observation, however, are the quirks of an interdependent rural society, especially those imposed by the spatially fiagmented control and ownership of land. As already remarked In connection with the financial outcomes of STW development (para. 3.9), these social relationships abet the normal market mechanisms that help to spread the private fruits of Investment. IV. ISSUES AND LESSONS A. GROUNDWATER REaULATION 4.1 One of the conditions of project lending to STW buyers to be observed by lending banks was that technical standards, in particular spacing criteria, as laid down by the State Water Board (SWB), be observed (Project Agreement between IDA and the Agricultural Refinance Corporation, Schedule 1). 4.2 The SAR (para. 5.18) observed that state groundwater legislation had been drafted by GOWB based on a model prepared by the Government of India (GOI). While this legislation was not considered important in the short run, because ample groundwater was still available in West Bengal, in the long run it was considered essential to ensure the most efficient use of the State's groundwater resources. Confirmation was obtained during negotiations that GOWB would use its best endeavors to introduce this legislation as soon as practicable. 4.3 Appraisal of groundwater conditions had been considered satisfactory for DTWs but lacking for STWs. At the time of appraisal no criteria were in use for spacing minor irrigation facilities, but with more intensive development the SAR considered that these would be necessary. SAR calculations indicated that on clay soils STWs should normally be more than 350 m apart and not closer than 250 m, and on loam the respective -8- distances were 300 and 200 m. However, these were only average figures for the whole project area, and distances would vary in different localities according to available water resources, localized recharge conditions, and crop rotations. Before assessing the suitability of the area for STWs equipped with centrifugal pumps, it was necessary to apply technical criteria, Including the depth of water below ground level and water level fluctuations. SWB was to be the appropriate authority to devise such criteria for each area, but before adoption by GOWS the criteria were to be reviewed and approved by the Agricultural Refinance Corporation. Assurances were obtained at negotiations that detailed well spacing and technical criteria would be worked out by SWB and embodied as a condition to be observed by all lending banks when approving loans (SAR, para. 5.19). 4.4 The SAR considered that successful installation of STWs required favorable watertable conditions (depths to water less than 3 to 5 m) and favorable geological conditions (productive aquifers at depths less than 30 to 50 m). These conditions were found throughout a large part of the project area. In some areas, however, DTWs were the only feasible Installation because of greater depths to watertable or the presence of thick clay beds at shallower drilling depths (SAR, Annex 2, para. 23). 4.5 Some of these prior concerns and expectations proved false or misleading. For example, the PCR cites, as one of the reasons for initial Implementation delays, the use of overly conservative norms adopted by SWB for working out groundwater recharge from rainfall. It argues that application of these norms led to the (false) early conclusion that much of the project area was already over-exploited (PCR, para. 3.01). Other events, however, gradually overtook these concerns. 4.6 Disbursement delays for other components of the project led GOWS and IDA to conclude that the Credit would not be fully expended without a substantial increase in the STW target. In 1979, therefore, they agreed to enlarge the STW component to cover the entire State and allocated an additional US$8.3 million of the Credit to it (in all, US$21.4 million, or 60 percent of the Credit, was eventually expended on STWs). By this time, the objection to financing STWs throughout the entire state had been removed - the findings of the all-India groundwater review led to a lifting of interdictions on the use of additional groundwater resources in certain areas (PCR, para. 3.04). 4.7 In retrospect, the initial wariness on the part of the Bank towards financing STWs was similar in some respects to the early caution displayed also by the Bank in the Shallow Tubewell Project in Bangladesh. As explained in the impact evaluation of that project, there was a tendency on the part of project planners to propose well spacing criteria for lending and other regulations as if the purpose of public policy might be to limit the development of the groundwater resource to that attainable through the STW technologies alone. 4.8 Fortunately, this wariness on the part of some towards the uncertainties created by STW development in WBADP was moderated by the vision of others, who saw STW development as occupying an important, though likely temporary niche in the evolution of groundwater development. Subsequent thinking, as exhibited for example in the Bank's Minor Irrigation Project (Credit 161 9-IN), confirms the wisdom of those -9- responsible for WBADP who pressed for rapid STW development and reliance on other means, such as competition, for regulating that development.r B. DIESEL vs. ELEcTac PowER 4.9 After careful consideration of the factors involved in determining farmers' likely preferences for different types of equipment, appraisers projected that 75 percent of the STWs in the project would be electric-powered and 25 percent would be diesel- powered (para. 1.8 and SAR, Annex 2, para. 26). In the event, diesel pumpsets were preeminent. Of the 56,000 pumpsets financed, 78 percent were diesel-powered and 22 percent electrically-powered, (para. 1.11 and PCR, para. 3.03). No great consequence hung on this outcome. Nevertheless, It is instructive to review the arguments again just to appreciate the difficulties of forecasting, especially when such forecasts are conditional on investments in, and the performance of, other infrastructure, in this case rural electrification. 4.10 The IDA appraisal team considered that the main factors affecting farmers' choice of power for STWs were the availability of electricity, its reliability relative to that of diesel supplies, comparative capital and recurrent costs of these two power sources, and operating flexibility. On the basis of detailed information provided by the West Bengal State Electricity Board - concerning ongoing schemes to increase power supplies and to expand the transmission and distribution network - the appraisal team concluded that all proposed DTWs, and a substantial proportion of STWs, would be able to connect to the grid. They thought it was unlikely, however, that distribution lines could be provided everywhere for STWs. Unreliability of electricity supplies had been the subject of complaints in parts of the project area, but the situation improved significantly during project preparation. While diesel fuel was available, it was nonetheless scarce, and appraisers feared that this situation could deteriorate in the future because of the dependency of this fuel on foreign exchange resources (SAR, Annex 2, para. 25). 4.11 Electric power connections and charges for electrical use were subsidized, whereas diesel had to that time been taxed. Consequently, the financial cost to STW farmers of water from electric pumps was about two-thirds that of water from diesel pumps. In addition, since West Bengal electricity was generated mainly from coal, it was thought that increasing cost of oil would not affect power generating costs but might cause further increase in diesel costs. Although the relative financial costs of the two power sources would depend on Government taxation and subsidy policies, appraisers assumed that a large margin between electricity and diesel costs would remain - irrespective of government policy - in electricity's favor. 5. The Regional staff ado not agree with the report's statement (pare 10), that competitive market factors alone would discipline the development of groundwater resources. Serious environmental and socioleconomic problems are surfacing fast in India due to indiscriminate tapping of groundwater. There is an urgent need to regulate ground water exploitation through appropriate directives.* Para. 10 in the Executive Summary Is based, of course, on pares. 4.1-4.8. -10- 4.12 Offsetting their disadvantage of higher operating cost, however, diesel pumps were considered more flexible. They could be moved to different tubewells, used to pump surface water, or hired out. The electric pump-motor combiration, by contrast, was static and transformers needed protection from theft. The State Electricity Board had initiated a hiring scheme for mobile electric pump sets, but this had not had much impact. One of the reasons for farmers' eventual preference for diesel power - and another reason for the project's implementation delay - was the long delay of the State Electricity Board (SEB) in providing power for pumps. Another reason for diesel preference was the unexpectedly large number of pumps to be energized. 4.13 Part of the problem faced by SEBs In India has been the scarcity of funds for investment in electricity generation and distribution. This scarcity is partly a byproduct of subsidies given to electricity consumers. Such scarcities are likely to continue while governments succumb to pressures to subsidize. 4.14 West Bengal's SES has apparently overcome the problems it faced in the second half of the 'seventies. Given farmers' strong preferences for diesel power under WBADP, one might have expected the follow-on Minor Irrigation Project in 1985 to have reflected this. It did not. The SAR for that project projected that 100 percent of STW pumpsets would be electric-powered.6 Judging from recent reports on that project, this projection has proved to be accurate (some river lift schemes have even been converted from diesel to electric). C. F-NANcIAL vERsus EcoNomic RETuRNs 4.15 Disparities between some of the Report's assessments of financial and economic performance of STWs constitute an issue. Such an issue was already apparent in the PCR, though rather more for DTWs and RLI, than for STWs. The PCR reported financial rates of return in the range of 5-10 percent for DTWs on Malda loam soils, and as low as one percent for RLI on loam soils of West Dinajpur. The economic return in each of these cases was two to four times the financial return. In the case of STWs, the PCR's economic rates of return were about twice the financial ones. 4.16 As noted already, part of the observed disparities between financial and economic returns for STWs reflect the deliberately different perspectives adopted by the financial and economic analyses, which measure: on the one hand, the attractiveness of the STW investment to the primary decision maker - its owner; and on the other hand, the joint returns from the STW accruing to wider interests, including the owner and water- purchasing neighbors. But even after accounting for this difference in perspective, certain disparities remain: economic rates of return at impact evaluation tend to be more favorable than financial rates of return; differences in valuations of costs and benefits have favored he economic returns. 4.17 Discussions of this issue with Bank staff suggest that analysts of other groundwater Irrigation situations may disagree with the above reported direction of the disparity between the results of the two forms of analysis. Staff who have worked mainly on electric-powered tubewells report that the subsidies involved in retail electricity prices 6. World Bank Report No.5355-IN, Staff Appraisal Report: INDIA - West Bengal Minor Irrigation Project (Credit 16194IN). February 1. 1986, Pare 3.23. -11 - push financial returns to these wells AhaM the economic returns. The IE's Survey data, however, are dominated by a preponderance of diesel equipment; and project output is dominated by rice, for which domestic prices at impact assessment appear to be only two- thirds the level of border prices. 4.18 Such an issue becomes serious whenever estimated project economic returns, however attractive they might be, are unlikely to be realized because of low financial returns and thus poor incentives for owners to invest in STWs. Fortunately, this did Wt happen in the case of this project. But the evaluation results offer a reminder that such gaps between economic and financial returns cannot be ignored indefinitely. 4.19 What this discussion suggests is the need to focus the Bank's efforts more on the related policy issues. India stands to gain resource use efficiencies from reducing distortions to energy prices and rice prices. The lower, and in some cases unattractive, financial returns that appear to face current replacement of project STW equipment illustrate what Bank policy analysts have said of the Indian agricultural economy in recent years: namely, that government should rethink its role in output markets, particularly its procurement arrangements for rice, which depress producer prices. Unless the gap between economic and financial prices fo !project) outputs is narrowed, there is also less scope for reducing or eliminating the electricity subsidies that either undermine the financial strength of state electricity boards, or burden government budgets, or both.7 7. See World Bank Report No.9412-IN, India 1991 Country Economic Memorandum, Volume 11: Agricultural Challenges and Opportunities, August 23, 1991. Similar remarks apply also, of course, to fertilizer subsidies. * 12- REFERENCES Agricultural Refinance and Development Corporation, Project Completion Report for West Bengal Agricultural Development Project (Cr.541-IN), 1982. Choudhury, S. & Gir, A.K., "Nature and Extent of Female Labour Use in Agriculture", Economic Affairs, Vol.31, Or.2, June 1986. Dixon, R., "Assessing the Impact of Development Projects on Women", AID Program Discussion Paper # 8, Agency for International Development, 1980. GOI, Ministry of Agriculture, Bulletin on Food Statistics (87--89). Kelkar, G., "The Impact of the Green Revolution on Women's Work Participation and Sex Roles, International Labour Organization, 1981. OED Report 1o. 7876, World Bank Experience with Irrigation Development, Socio- Economic, Institutional and Technical Impact and Lessons: Mexico and Morocco, June 1989. OED Report No. 8494, World Bank Experience with Irrigation Development, Socio-Economic, Institutional and Technical Impact and Lessons, (in six volumes), March 1990. OED Report No. 11031, Impact Evaluation Report: Bangladesh - Shallow Tubewells Project (Credit 724-BD), August 20, 1992. Roark, P., "Successful Rural Water Supply Projects and the Concerns of Womenn, Distributed by The Office of Women in Development, Agency for International Development, 1980. Schenk-Sandbergen, L., "Empowerment of Women: Its scope in a bilateral development project", Economic and Political Weekly, April 27, 1991, WS27-WS35. Sharma, U., Women. Work and Property in NorthWest India. Tavistock Publishers, London, 1980. World Bank Report No. 5355-IN, Staff Appraisal Report, India West Bengal Minor Irrigation Project (Credit 1619-IN), February 1, 1985. World Bank Report No. 9412-IN, India 1991 Country Economic Memorandum, Volume II: Agriculture-Challenges and Opportunities, August 23, 1991. -13- Annex I Page 1 of 8 THE PAQ 1. As rgported in the SAR, West Bengal has a gross area of 87,700 km2. of which 55,700 km4 were then classed as cultivable, agricultural land (i.e. 5.6 M ha, 1971 census). The annual cropped area was 7.1 M ha, giving a cropping intensity of about 127 percent. Rice - the most important crop - accounted for 70 percent of the cropped area; other crops grown were jute, tea, potatoes, chillias and a wide variety of kharif and rabi vegetables. The estimated rural population in 1971 was 33 M. out of a State population of 44 M. About 70 percent of the State population, particularly in rural areas, lived below the poverty line (consumption of Rs 20 per month in 1960/61 prices), compared with 40 percent for all India. Between 1.5 to 2.0 million tons of grain were imported annually, mainly for the 8 M Calcutta population. Because of localized floods and rain failures, the State food supply decreased seriously in 1974, and 15 M people reportedly faced severe food shortages. In the face of this situation, the Government of West Bengal (GOWB) was more than ever determined to increase food production rapidly. The quickest way was to accelerate exploitation of the ample minor irrigation potential. 2. Government of India's (GOI's) initial project proposal, in 1972, had emphasized development of rural markets. Subsequent preparation efforts lead to the major emphasis being placed on minor irrigation. The Bank favored this direction, yet resisted efforts by GOWB to expand the number of deep tubewells (DTWs) and riverlift installations (RLI). The issue here, initially, was GOI's reluctance to support procurement through international competitive bidding (ICB) for such an expanded program. Subsequently, the issue became the utilization of these facilities, which the Bank felt were working far below capacity. One of the (several) problematic areas here was water charges. Although GOWB had increased water charges considerably, such charges represented only 2 percent to 5 percent of the annual operating cost of a DTW. This implied a subsidy of at least 95 percent to the Minor Irrigation Corporation (MIC), which would operate the facilities, if current charges were maintained. The compromise reached was to focus these project components on completing existing schemes and, more as a pilot operation, on ways of improving their capacity utilization through extension, etc. The main thrust of the project proposal devolved towards private sector STW development. 3. At appraisal, there were about 61,000 STWs already installed in West Bengal, of which 3,350 were State-owned, the rest privately owned. Most STWs had been installed through concessional credit provided directly by the Agricultural Department, with little attempt to collect repayment or interest charges. Shallow groundwater levels in most alluvial areas made them suitable for STW development, particularly in the short term. In the long term (at least 20 years, which is more than the life of a STW), the SAR argued that intensive development would likely lower the water table to a point where typical STWs equipped with centrifugal pumps might fail, and have to be replaced with DTWs. Such (seasonal) lowering of the groundwater table would reduce runoff and increase recharge, and allow maximum - 14- Annex I Page 2 of 8 groundwater utilization. The irrigation development strategy advocated was to concentrate on rapid development of STWs, and to lay the foundation for future expansion of DTWs and RLIs. 4.Development of STWs had begun only a few years prior to appraisal. They were scattered throughout the alluvial plain, with the greatest concentration in the area south of the Ganges- Padma river. The STWs, consisting of 76 mm (3 in.) diameter pipe and brass or coil strainer, were equipped with small centrifuggl pumps and 5 hp diesel or electric motors. These wells had yield capacities of 20 to 36 m /hr, enough to irrigate 3 to 5 he. Well depths generally ranged from 25 to 45 m, although in some areas in Midnapore and Howrah Districts the depths increased to over 100 m. Unlined ditches to distribute the water were provided by the farmers. Because of small farm size and fragmentation of holdings, individual farmers who installed wells frequently sold water to neighbors, sometimes to gain access to their own land fragments. The Project as Aparoved 5. The project as designed sought to assist in financing: (a) individual farmers or farmers' groups for 18,000 electric and diesel powered shallow tubewells; 75 percent of the wells would be electric-powered, the rest diesel powered; (b) farmers' groups for 200 deep tubewells; (c) the Minor Irrigation Corporation (MIC) for 100 deep tubewells; (d) the Government of West Bengal (GOWB) for completing 600 r**erlift and deep tubewell schemes; (e) private borrowers for equipment for 200 agro-service centers; and (f) Regulated Market Committees (RMC) for the development of 3 markets. The project also intended to support the development of GOWB's maintenance facilities, technical assistance and training for the newly-formed State Water Board (SWB), reorganization and strengthening of the agricultural extension service, and preparation of a second stage project. Its total cost was estimated at US$67 million. IDA provided a $34 million credit, about 45 percent of which was earmarked for installation of the STWs. (Eventual disbursement of 63 percent of the credit for STWs figured in the decision of this report to focus principally on this component). The parties to the Agreement expected the project to be completed by September 30, 1979, i.e. within four and a half years. Annex 1 gives a more detailed description of project components and financing. 1. The expected proportion of wells to be powered by electric motors and diesel engines can be inferred from SAR, para. 4.15 and Annex 2, para. 26. See also PCR, para. 3.03. -15- Anne I Page 3 of 8 6. The project area, covering much of the six districts of Hooghly, Burdwan, Nadia, Murshidabad, Malt a, and West Dinajp, 'out of 15 Districts statewide), has a gross area of about 26,000 km and had in 1971 an estimated population of about 13 M. Soils are all alluvial. About 55-60 percent are pervious looms, suitable for wheat, vegetable and olitorius jute, while the rest are impervious clay soils particularly suitable for boro paddy. On average, 7 percent of the area is annually affected by damaging floods, although during the rains the water table rises to the surface on at least half the land, which may be submerged to about 50 cm for varying periods. This factor determines that over 60 percent of the area is, during the rains, sown to paddy and jute. 7. GO GOWS and the Bank resolved a number of issues during project preparation and appraisal, Including the level of water charges for the recovery of public irrigation investments, the rules for procurement, and the scope for safe levels of ground water extraction. 8. Where public irrigation investments were involved, GOWS agreed to increase water charges to recover O&M costs by end of 1975 and a progressively larger portion of their capital costs. 9. Regarding STW procurement, the SAR (para 4.21) points out that execution of this program would involve small investments of about Rs 8,000 (US$1,000) for a tubewell and pumpset, and operations on many individual farms, with purchases spread over four years, so that it would not be practical to bulk these for centralized procurement. An adequate choice of pumpsets was reputed to be available and prices were competitive, so that farmers had a fair choice at reasonable prices. Procurement was to be through normal commercial channels, as in the case of all IDA projects in India involving minor irrigation. None of the equipment purchases for the project was reserved for ICB. 10. Regarding the question of groundwater exploitation, the SAR estimated that extraction by a further 18,000 STWs and 300 DTWs would constitute about one-third the calculated recharge. This was based on supposedly conservative recharge assumptions. The legal agreements provided for the exclusion of DTWs from areas which, according to the technical criteria laid down by the Water Board, were more suitable for STWs. The agreements also required lending banks to follow the technical and siting criteria for STWs to be established by SWO. This was likely to mean, for example, that STWs on clay should be spaced 350 m apart and never less than 250 m and on looms 300 and 200 m respectively. Malor Conclusions of the PCRIPAR 11. The PCR, completed in July 1984, concluded that the project was a success, although overall results were mixed. OED's audit, completed in March 1985, concurred. 12. Imolengtation. According to the PCR, the minor irrigation components were on the whole implemented as envisaged, but other components fell short of expectations: Annex 1 Page 4 of 8 (a) Farmers and groups constructed .about 33,200 STWs compared with the appraisal estimate of 18,000, and fitted an additional 22,500 wells with pumpsets. (This latter characterization by the PCR/PAR is a bit misleading. The main source for its Information reports that farmers who acquired pumpsets used them to irrigate from canals, rIvulets, beels, and tanks, I.e. from surface waters, not from groundwater. ) Of the 55,700 pumpsets energi7ed, 78 percent were diesel powered and 22 percent electric powered (PCR, para. 3.03). (b) The Minor Irrigation Corporation established 300 deep tubewells, as expected. (c) The Agricultural Engineering Directorate improved 546 river lift Irrigation schemes, about 10 percent less than expected, and completed 362 onshore/offshore installations Involving pumping facilities; (d) The State Water Investigation Directorate received equipment, but was unable to make full use of available training funds; its work program was upgraded. (e) Only one of three proposed agricultural markets was completed, but none was operational at project completion; none was financed with funds from the IDA credit. (f) Of the 200 proposed agro-service centers, only 64 were established, also without IDA funds, but many of those were subsequently closed as being non- viable. (g) The institution-building components in the project encountered major difficulties. Improvements were made in MIC and the SWB (later renamed the State Water Investigation Directorate - SWID), but serious deficiencies remained. Implementation of the agricultural extension component was greatly delayed. Following initially slow progress, IDA conducted a mid-term review in January 1979, and this led to reallocations of the credit favoring STWs and pumpsets. It had become obvious that because of the lower expected disbursements for other components, the Credit would not be fully disbursed without a substantial increase in the STW target (PCR, pa-*. 3.04). It also became apparent, from an all-India groundwater review, that districts previously excluded from STW investments possessed adequate groundwater resources. Accordingly, the project 2. ARVC, Project Completon Report tor Credit 5414N, 1982. para. 4.32. - 17 - Page 5 of 8 area for S s and pumpsets was expanded to cover the entire state Instead of the original six districts. 13. Actual project cost was US$59.3 million, or 11 percent below appraisal estimates. The main reasons for the reduced cost were physical under-achievement of several components, and lower-than-expected unit costs for the construction of shallow tubewells. The credit was fully disbursed after a one-year extension of the closing date. 14. Again summarizing the PCR, project implementation experienced some major difficulties: poor coordination among project agencies, insufficient support by some commercial banks, late appointment of agency heads for MIC and SWB, controversy over groundwater exploitation norms, reluctance of farmers to organize in groups, and prolonged litigations related to the markets and agricultural extension components. The changeover to the training and visit (T&V) system of extension under the project was particularly problematic as there was political opposition to this provision for some time, and the existing government staff concerned resisted the required reorganization. Uttle progress was achieved in this regard under the project. Loan recoveries from project beneficiaries deteriorated throughout the project period, although this was a general phenomenon In India at that time and not specific to this project. A major difficulty, and one that resulted in the violation of a Credit Agreement covenant, related to the issue of cost recovery, i.e. the rationalization of water charges. Virtually no progress was made on this point, largely for political reasons. (PAR para 5). 15. Agiultural IM-act. The PCR had difficulty in ascertaining the project's production impact because of the dispersed nature of the investments. However, it cited a substantial increase in crop yields, especially for paddy, wheat, potatoes and jute. The cropping pattern under irrigation had not fully met with expectations, mainly because of farmers' resistance to reducing cultivation of water-intensive boro paddy In favor of other crops, mainly wheat. Cropping intensities increased significantly, e.g. reaching 177 percent for shallow tubewells (from 130 percent without the project). Overall, a shortfall in irrigated area under the project was thought likely, largely due to widespread adherence to the pre-project cropping pattern, but also partly because of design and technical problems with RLis. The reduced irrigation area had an adverse effect on the number of direct beneficiary farmers, which reached only about 120,000 instead of the target of 200,000. 16. Economic Imoact. Both the appraisal report and the PCR give range estimates for the economic rate of return by type of investment. The ranges represented areas with different types of soils. The re-estimated ranges were 47-50 percent for shallow tubewells (compared with > 50 percent at appraisal), 21 percent to over 50 percent for deep tubewells, and 4-50 percent for river lift installations. Taking the mid-points of these ranges and weighting by the cost of the respective components, the re-estimated average economic rate of return to minor 3. ARDC Project Completion Report para. vi, page il. - 18 - Annex 1 Page 6 of 8 irrigation overall was 42 percent, compared with 55 percent estimated at appraisal. No rates of return were calculated for the other project components. 17. For STW development, the main lessons related to - poor extension, for which the project (and a follow-on specialized extension project) were found to be ineffective; - a lower not irrigated area per unit, which reached only about half of the appraisal estimate; and declining loan recoveries, which required solutions at a national level. - 19 - Annex I Page 7 of 8 Anendix 1 IMPACT EVALUATION REPORT INDIA: WEST BENGAL AGRICULTURAL DEVELOPMENT PROJECT (CREDIT 541-IN) (SHALLOW TUBEWELLS COMPONENT) Project Description at Appraisal 1. According to Schedule 2 of the Development Credit Agreement, the project was the first phase of a program to assist West B,-gal with agricultural development in the Project Area and consisted of the following parts: Part A: A four year program for loans through the Participating Banks for: (i) construction and equipping of about 18,000 shallow tubewells including, Inter aia, land development and field channels; (ii) construction and equipping of about 200 deep tubewells for Irrigation Groups and about 100 deep tubewells for MIC including, inter alia, land development and field channels; (iii) construction of facilities (including, intera/ia, roads, fencing, go-downs, auction platforms and shops) and acquisition of equipment for three Markets at Samsi, Karimpur and Katwa; and (iv) acquisition and utilization of machinery and equipment for the establishment of about 200 agro-service centers. Part B: A four-year program for: (I) completion by West Bengal of about 600 partly constructed riverlift and deep tubewell schemes including, Inter a/ia, land development and field channels; (ii) technical assistance, scholarships for staff and acquisition of equipment for the Water Board and the Engineering Directorate of the Department of Agriculture of West Bengal; training of secretaries of three Market Committees located at Sarnsi, Karimpur and Katwa; completion of (a) an economic evaluation of existing agricultural markets and the impact of the Markets on agricultural development in West Bengal, with emphasis on the benefits to farmers, based on the study of the three Markets financed under the Project and selected similar Markets, and (b) feasibility studies for the establishment of further Markets in West Bengal; and preparation of a second phase project; and -20- Annex_1 Page 8 of 8 Apnendix 1 (iii) employment of additional staff, acquisition of equipment and materials, and staff training, for Improvement of extension services of the Department of Agriculture, West Bengal. 2. The Project Area for these investments was designated as the Districts of Hooghly Burdwan, Nadia, Murshidabad, Maida and West Dinajpur in West Bengal. 3. The Credit Agreement allocated its US$34 million (in dollar equivalent) to the following categories: Loans for: Dollar Eauivalent (1) (a) Shallow tubewells 11,400,000 (b) Deep tubewells (i) for MIC 1,000,000 (ii) for Irrigation Groups 1,000,000 (c) Agro-service centers 1,000,000 (d) Markets 600,000 (2) (a) Riverlift and deep tubewell 6,800,000 irrigation schemes (b) Equipment for the Water Board 1,000,000 and Engineering Directorate (c) Improvement of agricultural 2,200,000 extension services (d) Technical assistance and training 700,000 (3) Unallocated 8,300,000 TOTAL 34,000,000 - 21 - Annex2 Page 1 of 2 FARMSUREYLARRANgNWT 1. Beginning In March 1989, after discussions with the Regional Director (Calcutta) of the National Bank for Agriculture and Rural Development and the Chief Engineer, Minor Irrigation Department of Agriculture, OED began collaborating with a local consultant, who agreed to organize collection of farm survey information that would help measure the impact of the STW component of the Project. The basic objectives of the proposed survey were to: (a) collect background and general information regarding the development units in which the project tubewells were located; (b) examine the ownership and use of land of the STW owners/managers and others following operation of the project tubewells; (c) identify technical and socio-economic constraints associated with STW operations; (d) construct Farm Budgets for both irrigators and non-irrigators and Tubewell Budgets for both Owners/Managers and other irrigators on the basis of information for the 1988/89 crop year; and (e) make observations on the spillover effects, technical sustainability, environmental effects and land tenure effects of the areas where project tubewells are located. 2. The PCR had drawn on a 1981 field survey of WBADP beneficiaries, with 1979/80 as the reference year. That survey covered 71 STWs, 19 pumpsets, 30 purchasers of water from STWs, and 50 'non-irrigators" or "control" farmers, all located in Burdwan and Nadia Districts only. 3. The 1989 Survey sampled several categories of STW farmers, numbering about 190, taking care to choose the STW loanees at random from 3-5 development blocks within four districts of West Bengal: three bordering Bangladesh, viz Nadia, Malda, and Murshidabad, and one in the south-west, viz Burdwan. It proved difficult to obtain separate lists of loanees under Cr. 541 -IN for all participating banks, which tilted the sample somewhat towards loanees from the district land development banks (numbering 110 loanees) and led to the selection also of some loanees from other schemes. In addition, survey blocks were chosen in consideration of their proximity to district headquarters and of their easier access. Non irrigators and other categories in the sample were selected for their proximity to STW loanees. Table I presents the sample so chosen. In utilizing the STW-specific information -22- annex2 Page 2 of 2 from this sample, this report relies mainly on that pertaining to the 110 loanees under Credit 541-IN. 4. Survey investigators report what may be a decided bias in the results: they argue that Survey farmers habitually understate yields and overstate operating costs by about 10 percent. The investigators offer no motives explaining why farmers would make such evasive responses, but the problem they point to is not unfamiliar. Their warning is a useful one. The Survey results reported herein reflect farmer responses without any adjustment. Survey investigators also judged that STW repair and depreciation costs were too unreliable to report. The implications of these gaps in information and possible biases are discussed in Chapter III. Tabl"1 NUMBERS AND CATEGORIES OF FARMERS SAMPLED Districts SAi Narshidabad Malda Burdwas Total s. so. so. No. No. Project loanee 56 18 24 12 110 Other Scheme loanee 26 Is 16 2 59 Selffinanced 11 2 5 2 20 Purchaser of private water 12 2 - 4 18 Purchaser of institutional water 11 3 2 16 Non-irrigated ('Control") 20 3 8 4 35 Total 136 43 53 26 258 OZZ0= - m 32: 5 5. OED specifically requested that the consultant's Survey report be mainly in the form of tabulations of survey results, with limited supporting text. The report, dated October 1989, is entitled "West Bengal Shallow Tubewell Impact Study (Credit 541-IN)". It is available from OED on request. -23- Annex 3 Page 1 of 11 1. The SAR postulated that two soil types in the project area, loams and clays, would give different responses to irrigation. Because two of the I.E. Survey districts, Nadia and Burdwan, were treated by the PCR as surrogates for loam and clay soils respectively, this Annex includes comparisons of the Survey findings for these two soil types (Districts) with corresponding estimates drawn from the PCR and SAR. Motive Power nd ODerational Status 2. Of the sample of 110 STWs, 80 of their pumpsets were diesel powered, 30 were electric powered (Survey report, Tables 11, 12). This ratio of diesel to electric (73 percent:27 percent) is close to the ratio observed in the PCR for all pumpsets procured under the project (78:22 percent - PCR, para. 3.03). The ratio for 77,402 STWs installed in the four Survey districts as at 1981 was 84:16 percent. So all indications suggest that the project's revealed preference for diesel as the motive power of choice mirrors a similar preference in the wider community. Part of the reason for the historical dominance of diesel over electric has been the delays of the State Electricity Board in providing power connections and, presumably, the prevalence of power disruptions. 3. Of the total number of diesel pumpsets surveyed (144 - which includes non- project STWs), 140 reported the average number of days per year operated to be 109 days. Nineteen STWs, mostly located In Malda District, reported an average of 12 days lost (no discharge) due to low water table. Eleven reported an average of 11 days lost (no discharge) due to machine breakdown. Those operating (140) averaged 6.2 hours per day in operation. Fewer than 13 percent of owners had trouble finding spare parts, but 23 percent had difficulties in effecting repairs. (Survey report, Table 14). 4. Of the total number of electric pumpsets surveyed (48), 46 reported the average number of days per year operated to be 136 days, and the average hours per day operated were 7.7. Six STWs, all in Nadia District, reported an average of 23 days lost (no discharge) due to low water table. Nine reported an average of 13 days lost due to power failures, and two only an average of 15 days lost due to mechanical breakdown. Those operating (46) averaged 7.7 hours per day in operation. Fewer than 6 percent had trouble finding spare parts, but 31 percent had difficulties in effecting repairs. (Survey report, Table 15). 5. The Survey report does not identify directly the proportion of wells out of commission. However, the figures reported in the last two paragraphs suggest that six 4. Source: Survey report, Appendix D. .24- annex 3 Page 2 of 11 wells out of 192 (two of the 189 STW owners accounted for in the first three rows of Annex 2, Table 1 had more than one STW) may have been inoperative. These six reported on the availability of spare parts, but not on days operated. If these six were inoperative, the implied rate of attrition among STW investments (3 percent) is fairly low. Other survey indications of the impact of project operations, such as area irrigated, already allow for this attrition. STW aeneficiaries 6. The average size of STW owners' lands served by sample STWs was small: 3.24 acres for the 110 IDA STWs (3.44 acres and 3.00 acres for the two other categories of STW owners surveyed). Yet such small holdings are about twice the size of both those operated by water purchasers (1.68 acres) and those operated by the control group of non-irrigators (1.81 acres). (Survey report. Table 5). The average area of all holdings in the four sample districts at around the time of appraisal was even smaller than these areas: 1.01 acres (Survey report, Appendix B). 7. In addition to the 110 STW owner beneficiaries defined by the sample, these same STWs supported 210 water purchasers. Only 40 (of the 110) owners sold water. The 210 purchasers of this water irrigated a total of 114.8 acres (net irrigated area), or an average of 2.87 acres per seller. The average charge so realized was Rs 2055 per seller, or Rs 716 per acre irrigated from water sales. When spread over the full sample, the average area Irrigated by water purchasers per STW was 1.04 acres; the average charge per STW was thus Rs 747. 8. Of the total number of plots served by the 110 STWs, 44 percent were classified as being part of low level land (and therefore more deeply flooded), 21 percent medium level land, and 35 percent high level land (Survey report, Table 13). Plots Irrigated were located within 97 to 132 feet from the well, on average. Areas Irriasted ner STW 9. SAR estimates of areas irrigated were derived from forecasts of likely cropping patterns, irrigation efficiencies, and resulting crop water requirements. Using conservative assumptions about peak pumping hours per day, the SAR predicted that a STW would irrigate: 7.4 acres (3 ha) net on clay soils, with annual pumping hours at 1,500 hours and a maximum of 12 hours pumping per day; and 12.4 acres (5 ha) net on loom soils, with annual pumping hours of 1,250 hours and a maximum of 9 hours pumping per day (SAR, Annex 2, part D). 10. The 1989 IE Survey found that the net areas Irrigated per STW at full development exceeded the PCR estimates, but were still much less than appraisal estimates (Tabl9 1). The IE estimate of 4.28 acres is an average for the entire sample of 110 STWs financed under Cr. 541-IN. It comprises 3.24 acres Irrigated on land farmed by the STW proprietor and 1.04 acres irrigated through water sales. There was an average of -25- Annex 3 Page 3 of 11 1.91 buyers of water per STW. IE estimates of average area irrigated per STW proprietor in Nadia and Burdwan were 2.97 acres and 4.25 acres respectively. If we add estimates of the areas irrigated through water sales for these two districts (1.12 acres and 0.86 acres respectively), the areas commanded per STW are the 4.09 acres and 5.11 acres shown In Table 1. The comparable PCR estimates are lower all round: 2.67 acres, comprising 2.26 acres irrigated by the STW proprietor, and 0.41 acres Irrigated through water sales. TabIl: AVERAGE IRRIGATED AREA PER STW BY SOIL TYPE (OR DISTRICT) Loam Clay (Nadia) (Burdwan) All acres SAR 12.35 If 7.41 If 10.37 la PCR (1979-80) 2.71 I 2.61 I 2.67 /a IE (1988-89) 4.09/4 5.11 / 4.28 I ouss: /a The all (soils) figure for the SAR derives from an SAR assumption that 40 percent of STWs would be on clay soils (SAR, Annex 2, page 13). rg The all (soils) figure for the PCR derives either from the SAR assumption of 40 percent of STWs being on clay soils or more directly from the PCR's para. 4.01, provided that one reads sales of water to other farmers as 13,900 agres not hectares as shown there. I The all (district) figure for the IE embraces four districts, not just Nadia and Burdwan. The average of 4.28 acres comprises 3.24 acres by owner, and 1.04 acres by water purchasers. I See text of this Annex, para 10. I PCR, Annex 5. /I SAR. Annex 2, Part D. The origin of the SAR figures for Nadia and Burdwan is discussed in the text. 11. The PCR attributed the large shortfall in area actually irrigated to the failure of farmers to change sufficiently their pre-irrigation cropping pattern. At appraisal, analysts expected that farmers of loam soils, for Instance, would increase their cultivation of irrigated wheat and would not grow Boro paddy, which required much more water. The PCR found that such farmers (on loams) were growing significant amounts of boro paddy by 1979/80 and were expanding their wheat much less than expected. The 1989 IE -26- Annex 3 Page 4 of 11 Survey has found still greater interest in Boro paddy and even less interest in wheat on loam soils, yet larger areas irrigated, which increases the mystery. The difference between PCR survey and IE Survey results here could simply reflect sampling variation. As far as net acres irrigated are concerned, however, it could also reflect improving water use efficiency, which one might expect after a few years of irrigation. As for the puzzling trends in Boro and wheat areas, these simply illustrate the importance of cropping patterns to understanding areas irrigated. .Croppina.Patterns 12. At appraisal, analysts expected that Project STWs would double cropping intensities on clay soils and increase them significantly on loam soils. On clays, STW Irrigation would enable a boro crop to be added to the aman crop, thus increasing cropping intensity from 100 percent to 200 percent. On loams, wheat and vegetables (including potatoes) would be followed by aus paddy, jute, and aman paddy; the increase in cropping intensity, from 150 percent to 190 percent, would come from growing more wheat and vegetables, after eliminating pulses (Annex 3, Table 2). 13. At completion, the PCR confirmed that STWs would increase cropping intensities significantly. On clays, it estimated that irrigation allowed nearly 77 percent of the land to be planted to boro paddy, but that this replaced some area planted to wheat (cropping Intensity rose from 130 percent to 172 percent). On loams, it estimated that increased areas irrigated were shared between aus and boro paddy and wheat (cropping intensity rose from 130 percent to 177 percent) (Annex 3, Table 3). Note that average district-wide cropping intensities in the four Survey districts at about the time of project completion ranged from 138 percent (Burdwan) and 154 percent (Malda), to 171 percent (Nadia) and 172 percent (Murshidabad). (Survey report, Appendix A.) 14. The Survey results for clay soils fall somewhere between the SAR and PCR estimates: without STWs, a cropping intensity of 100 percent - all aman paddy; with STWs, a cropping intensity of 161 percent on owners' land - mostly boro and aman, but also some potatoes. On loam soils, the results proclaim a much stronger role than expected for boro, now 61 percent of net cropped area, and a much reduced role for wheat, now only 3 percent; the increase in cropping intensity overall is moderate: up 15 points to 162 percent. Valid survey results are available for one other district, and Annex 3, Table 4 shows the average cropping pattern for the three districts. Those figures exclude incomplete data on the cropping pattern of survey farmers in Malda district. Cron Yields 15. What counts here are the incma in crop yield (whether expected or achieved) under STW irrigation. At appraisal, these increments were headed by boro and potato yields - 1.6 tonnes of boro paddy per acre (4 tonnesiha) with a STW, nothing - 27- Ann 3 Page 5 of 11 without,?nd 6 tonnes of potatoes per acre (15 tonnes/ha) with a STW. nothing without. Wheat yields were also expected to increase considerably, by 0.7 tonnes/acre (1.7 tonnes/ha). Other crop yields were expected to enjoy lesser improvements (Annex 3, Table 5). 16. At completion, the incremental yields, estimated anew from survey data, confirmed most appraisal estimates, but challenged others (Annex 3, Table 6). Estimates for wheat showed increments of only .08 tonnes/acre (.2 tonnes/ha) moving from local varieties to HYV. Jute showed large increases of between .32 and .40 tonnes/acre (.8 to 1.0 tonnes/ha). Potatoes yielded 6.1 tonnes/acre (15 tonnes/ha) on Burdwan clay Without a STW. 17. At Impact evaluation, the estimated incremental yields are, again, fairly consistent with earlier estimates. Boro yields are somewhat higher 1.9 to 2.0 tonnes/acre (4.6 to 4.9 tonnes/ha), as are potato yields. This time, however, the potatoes are grown with the project and on clay soils not loams. Wheat yields are somewhat low -r .5 tonnes/acre (1.2 tonnes/ha) (Annex 3, Table 7). 18. All three sets of estimates reveal heavy boro paddy yields compared with average yields for the distt.cts. The average district-wide yields for each of the four districts over 1975/76 to 1979/80 ranged from .9 to 1.3 tonnes/acre (2.2 to 3.3 tonnes/ha). The range of average yields for 1980/81 was .9 to 1.1 tonnes/acre (2.3 to 2.7 tonnes/ha) (Survey report, Appendix C). Clearly, most other methods of irrigating the boro crop are inferior to STWs. 5. SAR claims of zero yields with or without a STW do not always mean, of course, zero yields eL.rB More often than not they mean zero plantings. -28 - Anx 3 Page 6 of 11 IaI.l. APPRAISAL ESTIMATES OF CROPPING PATTERNS BY SOIL TYPE (in percentages) Clay 1a Loam it without Wth Without With Project I Project Project I Project (percent) Aus 40 40 Aman 100 100 20 20 Boro .1,g Subtotal 100 200 60 60 Jute 30 30 Wheat 40 90 Pulses 20 Potatoes/Vegetables - ..1 Subtotal 90 130 Total Cropped Area 100 200 150 190 Net Cropped Area 100 100 100 100 Cropping Intensity 100 200 150 190 /a Clay - Burdwan District. rh Loam - Nadia District. fg1rce: SAR, Annex 8, Table 1. -29 - Annex a Page 7 of 11 Taba. PCR ESTIMATES OF CROPPING PATTERNS BY SOIL TYPE (in percentages) Clay /a Loam & Without With Without With Project Project Project Project (percent) Ealid Aus (local) 1 Aus (HYV) 5 25 Aman (HYV) 97 89 44 19 Boro (HYV) ,- Subtotal 98 166 49 80 Jute (local) 32 15 Jute (improved) 9 36 Wheat (HYV) 20 9 36 Wheat (local) 6 4 Pulses & Oilseeds 6 26 10 Potatoes 6 Vegetables Subtotal 32 6 81 9? Total Cropped Area 130 172 130 177 Net Cropped Area 100 100 100 100 Cropping Intensity 130 172 130 177 Ig Clay = Burdwan District. &k Loam = Nadia District. Sgurc PCR, Annexes 4 and 5. -30 - Anonx Page 8 of 11 ]Ahiu2. IE 1988-89 SURVEY ESTIMATES OF CROPPING PATTERNS BY SOIL TYPE * (in percentages) Clay /a Loam & Three Districts I& Without With Without I With Without With Project Project Project Project Project Project Aus 32 10 14 7 Aman 100 53 36 47 62 42 Boro (HYV) .84 ] 59 Subtotal 100 137 68 118 76 108 Other COs Jute 48 23 31 20 Wheat 3 4 11 Pulses & Oilseeds 5 28 11 18 10 Potatoes 20 8 Vegetable 4 5 1 3 Maize & Other -1 - Subtotal 25 80 43 54 52 Total Cropped Area 100 161 148 162 130 160 Net Cropped Area 100 100 100 100 100 100 Cropping Intensity 100 161 147 162 130 180 The IE Survey distinguishes between owner's and water purchaser's cropping patterns. The figures reported here reflect owner's cropping patterns for comparison to the SAR and PCR. /a Clay = Burdwan District. /k Loam = Nadia District. It Nadia, Burdwan and Murshidabad Districts. Sgag Survey report (1989), Tables 28-32 and correspondence. - 31-* Anon 3 Page 9 of I I T[le . APPRAISAL ESTIMATES OF CROP YIELDS BY SOIL TYPE (In tonnes/ha) Clay IA Loam /k Without With Without With Project Project Project Project (tonnes/ ha) Aus 1.5 1.8 Aman 2.0 2.3 2.0 2.3 Boro 4.0 Jute 1.3 1.5 Wheat 0.8 2.5 Pulses 0.7 Potatoes/Vegetables 15.0 Ia Clay - Burdwan District. l Loam = Nadia District. Sgg SAR, Annex 8, Table 1. -32- Anne Page 10 of 11 aM . PCR ESTIMATES OF CROP YIELDS BY SOIL TYPE (in tonnesthal Clay ik Loam ih Without With Without With PProject oject Project Project I(tonnes ha) Aus (local) 1.0 Aus (HYV) 1.5 2.2 Aman (HYV) 1.8 2.0 1.6 2.2 Bor (HYV) 3.9 4.0 Q-2-_- a Jute (local) 0.6 1.7 Jute (Improved) 0.9 1.7 Wheat (HYV) 7 2.1 Wheat (local) 1.4 1.9 Pulses & Oilseeds Potatoes 14.8 Vegetables 1a Clay = Burdwan District. /g Loam = Nadia District ougg: PCR, Annex 6 and para. 4.04. - 33- Annex.3 Page 11 of 11 le 7. IE 1988-89 SURVEY ESTIMATES OF CROP YIELDS BY SOIL TYPE (in tonnes/hal Clay IA Loam /1 Without With Without With project Project Project Project (tonnes/ha) Aus 2.1 2.6 Aman 2.5 3.2 2.4 3.0 Boro (HYV) 4.6 4.9 Jute 1.8 2.2 Wheat 1.2 Pulses & Oilseeds 1.0 0.7 0.6 Potatoes 19.1 Vegetables 6.4 11.1 Maize & Other I- , . Ia Clay = Burdwan District. /k Loam - Nadia District. Aguae Survey report (1989). Tables 28-32 and correspondence. -34- Annex 4 Page 1 of 10 FINANCIAL AND ECONOMIC ANALYSIS 1. This annex undertakes financial and economic analyses at the individual, STW investment level. It explores three models: Model 1 - a STW representing Nadia (loam soils) district; Model 2 - a STW representing Burdwan (clay soils) district; Model 3 - a STW representing Nadia, Burdwan, and Murshidabad districts. The methodology of the analyses follows closely that used in the PARIPCR and the SAR. It draws its data primarily from the IE survey, with alternative sources filling some gaps regarding selected inputs and prices. 2. Due to the Survey report's incomplete information on cropping patterns In the district of Malda, the financial and economic analyses incorporate data for Nadia, Murshidabad, and Burdwan only. The results for Model 3 represent a three district average. This raises the average net area irrigated to 3.39 acres for STW owners and 4.48 acres counting both owners and purchasers relative to the four district averages of 3.24 acres for owners and 4.28 acres for both owners and purchasers reported in Annex 3, Table 1. Insment costs 3. The analysis uses the initial STW investmen cost reported in the PAR/PCR and inflates It to FY89 prices using the GDP deflator. The analysis also gives the pumpset a scrap value of 5 percent of its cost in the final project year, following the lead of the PAR/PCR. However, the PAR/PCR does not account for the replacement cost of the pump and motor half way through the project life of 15 years. Hence, the analysis provides for pump and motor replacement in year 8, at 60 percent of the cost of the pumpset, following the example of the SAR. O&M Cst 4. Because the Survey report could not provide complete information on O&M costs of pumpsets, the analysis relies partly on the PAR/PCR for this information (see Table 1). The SAR envisaged O&M costs of approximately Rs 2 an hour for a diesel pumpset and Rs 1 an hour for electric (in FY74 prices). The PCR reported O&M cost data for diesel pumpsets only (Rs 2.52 an hour in FY80 prices), and argued that damage due to voltage 6. From the Bank Economic and Social Database. - 35- Annex4 Page 2 of 10 fluctuations in electricity supply caused O&M costs for the electric pumpsets to rise closer to that of diesel pumpsets (NABARD PCR pare. 5.5). The IE analysis assumes that 0&M costs for electric pumpsets equal 75 percent of the O&M for diesel pumpsets as opposed to 50 percent in the SAR. Using the GDP deflator, this translates into Rs 3.33 an hour in FY89 prices for diesel and Rs 2.50 an hour for electric. Using annual operating hours from the IE survey, the weighted average annual O&M budget is Rs 2,479 per STW in Nadia, Rs 1,805 per STW in Burdwan and Rs 2,158 for the Survey average. These financial costs of O&M are assumed to represent fully also their economic costs. Receipts from Water Sales 5. The financial and economic analyses differ substantially in their treatment of STW water sales. The financial analysis examines returns to the investment owner, while the economic analysis measures social benefits of the investment regardless of ownership and transfers of water and payments therefor between the owner and water purchasers. Thus, the economic analysis values total agricultural benefits due to STW irrigation by all water users, while the financial analysis values the owner's agricultural benefits plus receipts from water sales collected by the STW owner (in lieu of the crop production due to those sales). This follows the methodology used in the SAR (see SAR notes for Annex 7, Tables 6-7). A. FINANCIAL ANALYSIS Esa 6. All costs and benefits are expressed in FY89 Rs values. The financial prices of both outputs and recurrent inputs, including hired labor, were adjusted to FY81 values using the Agricultural Commodities Index published by the Ministry of Agriculture. Table 2 shows the financial prices used, as derived from the Survey report and partly from the other sources listed with the Table. Farm Models 7. Table 4 brings together the data on cropping pattern, yields, input requirements, and prices into farm incomes per STW for each of the soil (district) models in the survey year (FY89). Farm income per STW represents the gross value of production net of material input, irrigation O&M, and hired farm and animal labor costs. Gross value of production Includes income from water sales, while the contractual crop costs are those reported in Table 3. The net results were as follows: 7. Bulletin on Food Statistics 187891 GOI M. "y of Agriculture, Table 7 (p.27). -36 - Annex 4 Page 3 of 10 INCREMENTAL FARM INCOME, FY89 Rs IE Nadia 1,615 Burdwan 4,658 Survey average /1 2,602 la Nadia, Burdwan, and Murshidabad. Financial Rates of Return 8. Table 5 presents the financial analysis of a STW investment, based on the results obtained for FY89 and revised estimates for earlier years. More favorable not incremental crop income in most earlier years largely reflects higher real rice and potato prices. 9. Financial rates of return for the three STW models are quite varied - 10 percent for loam soils, > 50 percent for clay soils, and 23 percent for the survey average model. The lowest of these (real) returns is marginal. It is large enough to cover the nominal interest rates of 10.5 percent charged on borrowed funds, which are not included in survey costs, but they do not compensate the owner's labor and management inputs or risks taken. The highest returns - on clay soils - can be attributed to two main factors that distinguish this model. First, the cropping intensity In Burdwan district rises by 61 points, from 100 percent without the project to 161 percent with the project; whereas the estimated rise in cropping intensity for Nadia district is only 15 points. Second, the owner's net area irrigated per STW averages 4.25 acres on clay soils (Burdwan), and only 2.97 acres on loam (Nadia). The SAR also expected that STWs on clay soils would perform better than on loam soils, but not to the extent found in the IE. B. ECONOMIC ANALYSIS 10. The economic analysis expresses all costs and benefits in FY89 Rs values. It derives the economic prices for traged commodities from the World Bank's (IECIT) Commodity Prices and Price Forecasts, converting their US dollar values to local currency values at the FY89 prevailing exchange rate and adjusting for transport, quality differences, and other marketing costs. 11. The economic prices for non-traded inputs were derived directly from the financial prices. A conversion factor of 0.95 was applied to the financial prices of pesticides and agrochemicals. The economic prices of seed and manure have been assumed to equal 8. Dstrbution date FebuAy 5, 1992. -37 - Annex4 Page 4 of 10 their financial price. A standard conversion factor of 0.80 was applied to the financial price of draft labor to derive its economic cost. A factor of 0.50 has been applied to convert the average financial wage rate into an economic one. 12. Economic prices for non-traded output (potato, pulses, oilseed, vegetables, and maize were treated as "non-traded") were taken as equal to financial prices. (See Table 2.) Farm Models 13. The economic analysis used similar farm models to the financial analysis, except for the treatment of water sales referred to earlier. The analysis includes agricultural benefits from areas irrigated by water sales, resulting in a greater net area irrigated per STW for measuring cropping benefits. Economic Rates of Return 14. Using the IE net benefits per STW as indicative of full development, Table 6 shows the economic net benefits over the project life of the investment. The economic rates of return remain excellent - over 50 percent for all three STW models. -38- Anne= Page 5 of 10 INDIA: WEST BENGAL AGRICULTURAL DEVELOPMENT PROJECT (Cr.541-IN) Table 1: Tubewell Operating and Maintenance Financial Cost 11 (FY89 Rs.) Input Diesel Cost Electric CoS LOAM Operating Cost (Rs/hour) 3.33 2.50 Hours pumping per year 728 1,022 Total annual O&M charge 2,423 2,549 Weighted average /2 2,479 CLAY Operating Cost (Rs/hour) 3.33 2.50 Hours pumping per year 539 780 Total annual O&M charge 1,793 1,946 Weighted average /2 1,805 SURVEY Operating Cost (Rs/hour) 3.33 2.50 AVG. Hours pumping per year 615 984 Total annual O&M charge 2,047 2,455 Weighted average /2 2,158 /1 Per hour operating charges include fuel, lubricants, and annual repairs and are taken from the PCR, to FY89 prices using the GDP deflator. The pumping hours are from the IE Survey report, Tables 14 and 15. Higher total O&M charges for electrically powered STWs may strike some as being a curious anomaly of the method of estimating these costs. (One reviewer suggests, however, that such higher charges are a consequence of pricing incentives that encourage STW owners to operate electric pumps for longer periods than diesel pumps.) /2 Averages per district reflect the survey distribution of diesel and electric pumpsets per district. -39- Ann 4 Page 6 of 10 INDIA: WEST BENGAL AGRICULTURAL DEVELOPMENT PROJECT (Cr.541-IN) Table 2: OUTPUT PRICES, FY89 /1 Financial /2 Economic PADDY 2,060 4,199 JUTE 3,159 4,188 WHEAT 2,313 3,268 POTATO 978 978 PULSES 3,502 3,502 OILSEEDS 5,326 5,326 VEG. 712 712 MAIZE 1,683 1,683 /1 Representative of FY89 and expressed in FY89 Rs per metric ton. Financial paddy prices represent averages, the IE Survey reported separate figures for owners and non-irrigators. /2 Financial price sources: paddy - IE Survey; wheat, potato, pulses, oilseeds - West Bengal Minor Irrigation Project, Credit 1619, Staff Appraisal Report (1985); vegetable, maize - Bihar Public Tubewells Project, Credit 1737, Staff Appraisal Report (1986). -40- Annex4 Page 7 of 10 INDIA: WEST BENGAL AGRICULTURAL DEVELOPMENT P~'OJECT (Cr.541-IN) Table 3: Crop Input Financial Cost per STW, FY89 (FY89 fs.) Loam Clay Survey Average Without With Without With Without With Project Project Project Project Project Project Costs Irrigation /1 10 2,479 10 1,805 10 2,158 Other Inputs Seed 398 564 476 2,291 406 1,163 Fertilizer 499 1,416 599 2,910 508 1,804 Pest/Agrochem 212 330 255 575 228 410 Manure 149 239 332 400 206 259 Draft 1,361 1,645 1,245 2,567 1,289 1,914 Hired /2 642 -677 774 1,020 684 789 Labor 3,626 5,644 3,870 9,030 3,611 6,410 Hired /2 3,306 5,340 3,484 8,564 3,282 6,066 Gross Cost /3 6,256 12,317 6,787 19,578 6,258 14,118 Contractual Cost /3 5,216 11,046 5,929 17,565 5,324 12,649 /1 There were minor instances of irrigation among "non-Irrigators" surveyed. 12 Survey results allowed the IE to distinguish both the hired and household components of farm labor for the paddy crops. The average ratio observed for paddy was applied to the other crops. /3 Gross cost includes imputed value of all household labor (human + animal) while contractual cost includes only hired labor. -41- Annex Page 8 of 10 INDIA: WEST BENGAL AGRICULTURAL DEVELOPMENT PROJECT (Cr.541-N) Table 4: Farm Income Per STW. FY89 (FY89 Rs.) Without With With Over Project Project Without LOAM Gross Value of Production I1 8,206 15,651 7,448 Cost of Farm inputs 12 6,256 12,317 6,061 of which farm labor 3,626 5,644 2.017 of which animal labor 1,361 1,645 283 Net Value of Production 1,949 3,334 1,385 hired farm labor 13 3,306 5,340 2,035 hired animal labor /3 642 677 36 Farm Income /4 2,989 4,605 1,615 CLAY Gross Value of Production /1 8,837 25,131 16,294 Cost of Farm Inputs 12 6,787 19,578 12,791 of which farm labor 3.870 9,030 5,180 of which animal labor 1,245 2.567 1,322 Net Value of Production 2,050 5,553 3,503 hired farm labor 13 3,484 8,564 5,080 hired animal labor /3 774 1,020 247 Farm Income /4 2,908 7,566 4,658 SURVEY AVERAGE Gross Value of Production 11 9,081 19,008 9,927 Cost of Farm Inputs 12 6,258 14,118 7,860 of which farm labor 3,611 6,410 2,800 of which animal labor 1,289 1,914 625 Net Value of Production 2,823 4,890 2,067 hired farm labor 13 3,282 6,066 2,785 hired animal labor /3 684 789 104 Farm Income 14 3,757 6,360 2,602 11 Includes income from water sales. /2 Annex 4, Table 3. t3 Survey results allowed the IE to distinguish both the hired and household components of farm labor. /4 Farm Income - Gross Value of Production net of contractual costs (i.e. net of material input, irrigation O&M, and hired farm and animal labor costs). евРас�t fи�гдлоы яегОкг � ав�1г+: w8st еаlвм. двюав.nачи ос�rеtАм<в�гг г�ст ог�рт заз.вq � �g. wlewaa. !ое►теа aF ябпиw �гвs яе лег атv�ь ' . <.о�М IЧвоаU - Q32 ffi! � tiffi f�. Е�3 flL4� FY�4 t7� ЕУffi !�2 Е`� � !� Е�! � � о ! 2 s а s е т е s зо !з и з8 14 эб � Ип 6иnгдвМ�! Grog ttwaaв • 2,188 2.488 8,т03 4,твв 4,889 4,490 4.841 4,l82 4.01 t 8,в8з 8.в50 8,28в 8.28в 8.288 3.289 • lдаам Ftam веtв о1 УУдRег • 408 в04 вОв 806 ввв 806 806 8вб в06 806 80б в0В в0В в0В в06 Amw{ овМ • l1.2ао1 11.24о1 l1.�в1 tй4теl l2.47ek i2.4?e! 12.47в1 t2.47e1 G't.4те1 i?.4'7е1 12.4та1 12.4791 12.4тв1 I2.47б1 12.47е1 8ог�р V11ив • . ввз а�.имп соа nа2зв) р.а8з1 Чпь�г.амц�саhikw ре,2зв1 1.ва2 з.ее8 2.е� а.1зв s.ses 2,е1в �.2еа! 2.взе �ют 2.:вт 1.sтв з.азв 1.взв 1.вis г.2тв Аи�Мрц МТЕ aF f@iWO� зв9б С1АУ tBuNOwлN1 EY�Z �ffi fl� дСS� fl�t �& Еffi2 f�4 FY� � t� �ffi £� �4 i�t II'�3 � о 1 2 а а в е т е s 1о 1з з2 1в t4 зв гик гпог«пмгr� сгоо ь,�п,. • e.2s4 в.2ва s.ааз зз.4вз за81s а.зте е.оее е.воз е.вю в.атs е.2ве в.вов бsае в.вав в.а�е моамг�оп,в.иогw.аг • 8ов 4ее взэ взт sзт в1т в1т в1т вtт sn мт мт азт взт езэ , м�аа1 в�1r • 1вое1 1saa1 1t.esal 11.eos! 11saN ll.еовl 1з.еовl Мsов1 11.еов! р.еовl 1з.еовf p.eosl р.ва� 11so� 11.еов1 • 8onp Уtlи • в1з Ч мм�мwп сме ав.2зю ns8u � wlар.пмпwс.а�гюw 11а,21ю s.тоо s.ев4 а.твв lа2в8 s,з2в тsее 1t.oeol в.т:, в.во2 а.2sз s.oeo a.+ase 4,еsв а,взе в,з1в : Rииси� wиt oF пеп,п� > so9i 8ilRVlгУ AV�iAOH f�Z ffi� flC� �ffi ЕУ�1. ЕУ�2 EYS� !� !� fl�ffi Е�l2 f� � � � � � 0 1 2 8 4 в 8 7 в в 10 11 12 13 14 1В Nвtlдawмattl fkарЬюам • 8.А8в S.4M в.з8в 0.4в8 О.Ов8 5.в21 б,звв 4.767 4.вОв 4,4в8 4.811 4.018 4.018 4.013 4,018 моааи Fгап 8�М af WiNr • 874 SBO 747 747 7i7 т4т т47 М7 747 тq тдт 7Ч т4т тq т47 Мiт10бМ • (1.Отв1 И.07в) (1.в19{ 1й1вв1 12,1Qв, (2,11i8) (2,1ЕВ! (йlвв} 1й16в) (2.16� (2,18в1 C�158I 12.158з l2.18в1 (�18в) 8аир 11trв - ла1 lмaibnrк соп (1J 2fe0 t� вS11 • 1Мlbхвм�вf Св�АlЭж (18.2зв! 2.тВ8 ?.в40 4,818 в.От4 4.842 4.210 (41в81 8.84в 8.1в7 8,01в 2.ввв $Q02 2.вD2 2.Q02 2.002 � �111UCf11l RI►T� df I1E1111�k 2� 1р S � О 州O 司9 鳥物黝繼陣吧戶馴網•馴州• C以隊’O亂制膩，．館亂，o界．亂I,o亂乙離，．必奮9,e帕C鸞亂飼冷〕亂‘亂奮伯露閱酒亂乙以，oI幼亂．皂亂鉤啄勿馴細勿滷乙口，細矚勺磚頗綢』磚”訪開崗網頃•嗎網闖 勾時域個魷自嗡0萬”O勻••神鹹間 勸·馴，汎自” 審鷓鳥勵唱驢亂觔奮胞亂勵奮鵝r幼唱鷓亂勵唱閱亂嚇細9取勵個的亂劬攤閱亂淤審閱寡劬細9亂淤電閱忽嗡細亂『亂寧輾卹•畸他切，U·調奮O戶細卹 寥細祖亂露開．亂亂飼認鸞！口纏露絮亂驢C.t亂個口《．！奉藝卹〕亂蓄開勿亂6亂O》亂O心0鸞奉•亂e鸞ld驢勿亂l辟響基，O口化煙助伯·劉•馴開戶閑的．•網由q調開 嗚鳥令亂霧亂拙亂亂亂0亂6•之，9令t蓄10瀾岷 啊屆訪右編右編U編右1刃口開勾馴（細q細q讓田q倆口投編視勾翁咸抑初口 仲綢開奮戶啊州開馴牌，州 切雊曰騰A鄙團騙 倡儲綢闖•：山憚馴戶•網闢神綢綢叫 馴論伶亂U方露亂系颼勿馳以電亂馴州電亂闢卹忿亂鸛膠針以留戲以r•！以勿亂例州O亂胡9化亂亂切乍‘，坤勿劇詩勿胎噱9嗡0頗鳥口磚•O開即網U．•喝神闢 細吋嗡細頃，o縴網•。電調開，“叫 喲·劍，為自” 一臘‘，團I參韌t，團‘，幼’&，願”亂，．州”勾州亂，附u&&”除’&，爍’”婉•一開，麒州 n＿一，一＿,- 方麟‘硼‘&&，映‘臢〝‘喲‘幼‘喲‘粤”‘麒‘一‘一‘戲，&‘•，o‘州，o&”一一－ 協！令取寥亂寥亂亂籐。亂二必9唱令當寡亂0細哎 啊U詣啊祠口編眉細眉1勵q頸調發口祠勾細q取旬口〕眉編勾視啊龍眉左勾 辭啊單（•闈甲中口•奮馮義箏。 偶闖•員•捆憚馴戶•神闖調戶．啊細 朧劊勿馴細勿飽I俗口遞O唱亂卹勿馴抽勿坤『亂亂純必嗆t細磚勺卹露，．俗電6導劇坤伯以留讓口緘伯劇細勿幼喊皂劬幼州。“開．，寫神劉目•悶州目 鑰崤瀾輪嘆•勻畸寫”。•調叨”“閑 網·“，為自” 開膩f粤編偉馴開膩令t訪編鈴觔細膩申淤自以奮衫細鱸令划緬除嗡楓除划馴誡申酒騙鱸令划個膩俗划翰幼化寧和坤藝．幼初闐．幼·網奮O戶劉•甲 潤f亂亂婉州亂亂驢口．亂【細榆亂亂個州亂亂劇叮亂亂劍膩弋亂電價嗡亂憫叮寫亂胡C亂鳥t州鳥亂寥亂卹鳥亂騰f6馴州•幼亂鴛。啊•以屹戶卹鬨••網面勸奮州 。籐會亂零。g亂t么’0亂••必9•令t藝亂0萬岷 實司ql蔔q開必’祠口卻跑斤紹編觔審勾祠U開U細ql旬口騙必視屆取勾1刀口 ．你叩”他•戶關審，馮調審州 駟“觔劇－ 綠州名劉滷O田幼爛馴網馴戲幼〕．妒碩叮 網•亂帕』口劉訓幼討煩細幼曰陪勵唱呢（T州戲幼州滅勵彎馴，喊馮二織劉曉二娥關曦 』州細劉開中瞌滷n切編么O唱綢間 -44- Angn 5 Page 1 of 3 SOCIALMEACT ON WOMEN 1. Unfortunately, the terms-of-reference given to the Survey consultant overlooked the objective of measuring separately the project's Impact on women. Accounts of this aspect of STW development rely, therefore, on other sources. 2. The implementation of irrigation facilities impacts significantly on both men and women; both have an interest in more water to increase agricultural production. In addition, the role of women as water-carriers, water-managers and family health educators makes them an integral part of any successful rural water supply project. 3. However, in West Bengal, as In the other states of northern and eastern India, cultural norms severely limit women's roles. These norms are IT anifest within the home in women's primary, and often sole, responsibility for 'bari-based activities, and outside the home in women being confined to certain agricultural tasks and barred from others. For example, in areas of extensive irrigation systems, women perform the heavier work of weeding and fodder collection, and the transplanting of paddy, while men are occupied with the digging of irrigation ditches in the fields, the lifting of water from wells and canal's and the repair of terraces. 4. In order to understand the effects on women of irrigation, and STWs in particular, one must first realize that there are considerable variations in participation rates for paid arid/or unpaid agricultural labor depending on class, caste or community. In general, the following three broad classes of rural households, roughly categorized in terms of women's labor and income, rest also on those differences in labor participation rates: I) Women of landless households: households which cannot subsist without female and children's paid village labor. ii) Marginal farm women: the households can subsist because of the female unpaid family labor and some paid Village labor during harvest periods. These women work outside the house in the vegetable garden and many of them also work in their family paddy fields. Their husbands are potential beneficiaries of STWs if their land is In the project area. It I) Small farm women: households which, for reasons of social esteem, withdraw their women from paid village labor or even from work in their own crop lands. These women are indirectly affected by STW irrigation if their husband's land Is in the project area. They are mainly involved in `barl -besed activities, which include tending vegetable gardens, taking care of livestock, and postharvest operations. Postharvest work includes threshing, separating, 9. -,UWaW meaning ftwe-ba=f. -45- Ane f Page 2 of 3 parboiling, and storing paddy and winnowing and storing rice. Sometimet they will hire women from the first category to help in postharvest work. 5. Therefore, as the three categories clearly show, female labor force participation is inversely related to farm size and land ownership. The poorest group - women of landless households - have the largest percentage of females working. This should come as no surprise since in many cases, these women are the primary or sole income earners in their families, and their access to employment and income is crucial for their own and their family's survival. 6. Earlier studies of labor use patterns suggest that in rice-growing areas, the introduction of new HYV varieties of rice leads to an increased demand for female agricultural labor.11 Typically, this derives from a higher cropping intensity, in turn made possible by a greater proportion of land being under an effective irrigation facility. Increased productivity in turn leads to an increase in household incoipj, causing some landed women from category two to withdraw from the labor force.* Thus, landed women have felt the direct effect of increased productivity in their family fields. So have previously underemployed women in category one. Whereas, the changes for landless women already employed have been more in degree than in kind, making them only marginally better off. Aside from the direct income generating aspects, however, irrigation also has other positive effects. 7. Before the advent of hand pumps and STWs, the water required for domestic uses often came from surface water, which was frequently contaminated and sometimes located at a great distance. Installation of pumps has relieved women's domestic labor and improved the health condition of household members, especially children, who are more susceptible to water-borne diseases. In addition, other field surveys reveal that vegetable gardens, which are tended by women, have been expanding. These typically contribute to better family nutrition and augment family income in instances where women sell their produce on the market. 8. The location and usage of new water sources is a controversial subject among sociologists many of whom claim that, given several choices of traditional and improved water sources, users almost always choose the closest source. In many areas, this means 10. "Empowerment of Women: its Scope in a Bilateral Development Project', Loes Schenk-Sandbergen, April 1991. 11. See Ima of the GMe Revolution on Women's Wrk Participation and Sex Roles by Govind Kelkar, April 1981, and 'Nature and Extent of Female Labour Use in Agriculture', by S. Chowdhury & Ai.Gid. 12. For example, Sharma (1980) observed that in Punjab, increased prosperity as a result of the Green Revolution has been accompanied by a withdrawal of women from the agricultural labour force. Dixon (1980) explained this phenomenon by observing that in traditional rural 'purdah' societies, an increase in household income is usually accompanied by the withdrawal of women from the labour force to the more socially esteemed practice of female seclusion. -46- Annex Page 3 of 3 that the improved source is used by all during the dry season, but those further from the Improved soyr e will go back to seasonal, traditional sources that are closer ... during the wet season. Convenience, they argue, should be a major factor in decisions pertaining to the location of an improved source of water. While this is a valid argument, geotechnological factors may not always make this possible, in which case, women need to be made aware of the advantages of clean water through health education. 9. Finally, though irrigation, including that through STWs, has increased household income, provided new employment opportunities and improved domestic water supply, It is clear that women from differing stratas of rural society have not enjoyed the benefits equally. Since landownership and wealth follow caste lines, studies reveal that these irrigation benefits have widened the gap between the opportunities available to landless and landed women. That is, although there are thus greater agricultural labor opportunities for landless women - a major plus - this does not substantially alter the fact that they remain on the margin of survival. 13. Se Wome* in Development: Successful fural Water Supply Projects and the Concems of Women', by Paula Roawk Septemb 1980. "*" " ,-h-? ,a N E P AI B H U T A N NEPALDRJEEL N H U A ond J. b.opnd~d hw ...4æm~ ~ A L P A I G U R i - COOCH BEHAR 26' INDIA WEST BENGAL AGRICULTURAL DEVELOPMENT PROJECT P~oec Area Nect markets BoAk Lotiom ------.. - ,~ G-tic Alk~ (to the East) ond Vndhya Alhjmium (ro %h ~e) Delimiting Bondary of Physicøl Fea%res 0 District He rtr --- Dstrict Bo~ndories - - S.te aoundrie Intemational eundaries 2s4 s ir 8 A N K U R øf D N A P G R A 2 st 00 T m~ O__R_/_S __S_A S 0 una