FILE COPY. Document of The World Bank FOR OFFICIAL USE ONLY. Public Disclosure Authorized. Report No Public Disclosure Authorized

Transcription

1 Document of The World Bank Public Disclosure Authorized FOR OFFICIAL USE ONLY FILE COPY Report No INDIA: PROJECT PERFORMANCE AUDIT REPORT ANDHRA PRADESH, TAMIL NADU, AND MAHARASHTRA AGRICULTURAL CREDIT PROJECTS (Credits 226, 250 and 293-IN) October 5, 1979 Public Disclosure Authorized Public Disclosure Authorized Public Disclosure Authorized Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.

2 CURRENCY CONVERSIONS Andhra Tamil Pradesh Nadu Maharashtra At appraisal US$1 = Rs7.50 Rs7.50 Rs7.28 At Completion US$1 = Rs8.60 Rs8.60 Rs9.00 ABBREVIATIONS AED - Agricultural Engineering Department (Tamil Nadu) AIC - Andhra Pradesh State Agro-Industries Corporation ARDC - Agricultural Refinance and Development Corporation CBs - Commercial Banks ERR - Economic Rate of Return FAO - Food and Agriculture Organization of the UN FRR - Financial Rate of Return GOAP - Government of Andhra Pradesh GOI - Government of India GOM - Government of Maharashtra GOTN - Government of Tamil Nadu LDA - Land Development Agency (Maharashtra) LDB - The Land Development Bank System, made up of one State (apex) land development bank in each state (SLDB), plus scores or hundreds of SLDB branches ("central" structure) or independent primary land development banks (PLDBs; "federal" structure) spread all throughout each state's territory. They specialize in long term credit. MLDC - Maharashtra Land Development Corporation PCCS - Primary Cooperative Credit Societies (short term credit) PLDBs - Primary Land Development Banks RBI - Reserve Bank of India (India's Central Bank) REC - Rural Electification Corporation SCB - State Cooperative Banks (short and medium-term credit) SEB - State Electricity Board SGD - State Groundwater Department, in Andhra Pradesh; State Groundwater Directorate in Tamil Nadu; and Groundwater Surveys and Development Agency in Maharashtra SLDB - State (Apex) Land Development Bank (Cooperative Central Agricultural Development Bank in Andhra Pradesh)

3 FOR OFFICIAL USE ONLY Project Performance Audit Report INDIA: ANDHRA PRADESH, TAMIL NADU AND MAHARASHTRA AGRICULTURAL CREDIT PROJECTS (Credits 226, 250 and 293-IN) TABLE OF CONTENTS Preface Basic Data Sheets Disbursement Tables Highlights Page No. i iii vi ix I. PROJECTS SUMMARY A. Introduction 1 B. Andhra Pradesh 1 C. Tamil Nadu 3 D. Maharashtra 4 E. General Aspects 7 II. III. MAIN ISSUES IN THE PCRs A. Arrears and Collections 7 1. The Case of Maharashtra 7 2. The Case of Tamil Nadu 9 3. The Case of Andhra Pradesh 9 B. Integration of the Two Cooperative Credit Systems 11 C. Groundwater Development and Control The Need for Control 12 a) Protection of Individual Investments 12 b) Protection against Widespread Overdevelopment The Measures for Control 14 D. Well Failure and Possible Remedies 17 ADDITIONAL ISSUES A. Delays in Project Implementation 17 B. Disbursements against "Irregular" and "Off-limits" Lending 21 C. Groundwater Exploitation Control and Equity Considerations 23 D. A Cautionary Note on the SGDs' Effectiveness 26 E. A Cautionary Note on Statistical and Accounting Data 26 F. Reuse of IDA Funds 27 Annex 1 Comments received from the Government of India 28 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.

4 Project Performance Audit Report INDIA: ANDHRA PRADESH, TAMIL NADU and MAHARASHTRA AGRICULTURAL CREDIT PROJECTS (Credits 226, 250 and 293-IN) PREFACE This is a performance audit of three state agricultural credit projects in India. They are the third, fifth and seventh out of a group of ten such projects, financed by IDA between mid 1970 to late 1973 before switching to the all-india, general agricultural line of credit approach in /. Credits 226, 250 and 293-IN, for a total of US$89.4 million, were signed in January 1971, June 1971 and March 1972, and closed fully disbursed, in June 1977, December 1977 and June 1976, respectively. Two other state agricultural credit projects (Haryana and Punjab) are also closed, as well as ARDC I. They are the subject of two other project performance audit reports: one combining the two state projects and focusing mainly on tractor and tractorization issues, and the other covering ARDC I and treating several financial issues, as well as the convenience of switching from the state-by-state to the all-india lending approach. The first state project (Gujarat), which was closed in March 1975, was the subject of an audit report distributed to the Board on October 4, 1976 (Report No. 1303). The audit report includes a summary of the developments under the projects as reported in the project completion reports (PCRs), the main issues raised in the PCRs and some additional issues raised by the audit mission. The PCRs, dated January 1978, June 1978 and January 1979, were prepared originally by ARDC and put in final form by the Bank's South Asia Regional Office following country visits in August and November The lengths of the PCRs, however, prevent their inclusion in the PPAR2/. A three-member mission from the Operations Evaluation Department visited India in February/March The mission held discussions with officials of the Government of India, ARDC, the governments of the states concerned, the state land development banks (SLDBs), the main commercial banks involved in the projects (CBs) and other relevant state governmental agencies - particularly the state groundwater development agencies and the agencies in charge of tractor procurement. The mission also visited some rural branches of the Maharashtra SLDB and of the CBs, some primary land 1/ ARDC I, Credit 540, April 1975, for US$75 million. An eleventh state project (West Bengal, Credit 690, for US$12 million) and a second ARDC Credit (Credit 715, for US$200 million) followed in June A third ARDC credit was just approved (Credit 947, August 1979, for US$250 million). 2/ They are available from the South Asia Regional Office, Agriculture D Division.

5 - ii - development banks (PLDBs) in Andhra Pradesh and Tamil Nadu, and several farmers who had borrowed under the projects. Comments of the officials interviewed are fully reflected in the report and are gratefully acknowledged. The information obtained during that mission was utilized to test the validity of the analyses and conclusions of the PCRs. The audit report is based on these discussions, on interviews with Bank staff and on a review of the PCRs, the President's Reports No. P-881, 941 and 10271/ as well as the Appraisal Reports (No. PA-59a, 81a, and 116a).V; the Development Credit Agreements, Project Agreements and State Agreementa./ and materials available in the Bank's files. A copy of the draft report was sent to the Borrower. Comments received from the Government of India are in Annex 1. Suggested changes have been introduced or different views footnoted. The PCRs, which are detailed and accurate, focus on and thoroughly cover most aspects of project implementation, analyze the projects' main achievements and shortcomings, and raise some relevant issues. Two issues, however, are only marginally mentioned or not treated at all: delays in project implementation and retroactive financing. These are analyzed in the third chapter of the PPAR, prepared by OED. Also, the equity aspects of groundwater development are further elaborated, and minor comments are included on three other issues. 1/ Dated December 1, 1970; May 17, 1971, and February 16, 1972, respectively. 2/ Dated December 1, 1970; May 14, 1971, and February 14, 1972, respectively. 3/ The set of these for each individual state are dated January 8, 1971; June 11, 1971 and March 29, 1972, respectively.

6 - iii - ANDHRA PRADESR AGRICULTURAL CREDIT PROJECT (Credit 226-IN) Amounts in US$ million as of August 31, 1977 Original Disbursed Cancelled Repaid Outstanding-11 Credit 226-IN Project Data Original Plan Revisions Actual First Reference in Bank files 1968 Board Approval 12/11/70 Credit Agreement 1/08/71 Credit Effectiveness 4/08/71 5/10/71 Physical Completion Z of Original Project Completed by Date Shown - Irrigation Investments U/73 (100) 6/75 (200%) - Land Levelling 11/73 (100) 6/75 ( 50%) - Tractor Purchases 11/73 (100) 7/77 (93%) Loan Closing 6/74 6/75, 6/77 7/77 Total Costs (Rupees_Million) Economic Rate of Return 30% 37% Mission Data Month No. of No. of Man- Date of /Year Days Persons weeks Report Identification 1968 Preparation 6/69 Appraisal 4/ /01/70 Supervision I 3/ /26/71 Supervision II 9/ /21/71 Supervision III 3/ /05/72 Supervision IV 3/ /15/73 Supervision V 11/ /27/74 Supervision VI 7/ /29/75 Supervision VII 9/ /09/76 Supervision VIII (PCR) 11/ /06/78 Sub-total Supervision 16 1/ Before foreign exchange adjustment.

7 - iv - TAMIL NADU AGRICULTURAL CREDIT PROJECT 250-IN Basic Data Sheet Original Disbursed Cancelled Repaid Outstanding Credit 250-IN (US$ M) Project Data Original Actual First Reference in Bank files 1968 Board Approval 06/01/71 Credit Agreement 06/11/71 Credit Effectiveness 09/30/71 11/02/71 Credit Closing Date 12/31/74 12/31/77 Total Costs (US$ M) 62.0 about 62.0 Economic Rate of Return 28% about 35% Mission Data Month/ No. of No. of Man- Date of Year Days Persons Weeks Report Identification 68 Preparation 06/ Appraisal 09/ /71 Supervision I 03/ Supervision II 11/ /06/73 Supervision III 11/ /25/74 Supervision IV 02/ /08/75 Supervision V 10/11/ /19/75 Supervision VI (Completion) 08/ /79 36

8 - V - MAHARASHTRA AGRICULTURAL CREDIT PROJECT 293-IN Basic Data Sheet Original Disbursed Cancelled Repaid Outstanding Credit 293-IN (US$M) Project Data Original Revisions Actual Conception in Bank 1968 Board Approval 3/29/72 Credit Agreement 3/29/72 Credit Effectiveness 9/30/72 1/31/73 1/31/73 Physical Completion 3/76 Credit Closing Date 12/31/75 6/30/76 6/30/76 Total Costs (US$ M) 53.4 /a about 57 Economic Rate of Return 33-over 50% 17-over 50% Mission Data Month/ No. of No. of Man- Date of Year Days Persons Weeks Report Identification 68 Preparation 6/69 4 Appraisal 3-4/71 5 2/14/72 Preliminary Supervision 10/ /30/72 Supervision I 12/ /07/74 Supervision II 12/ /05/75 Supervision III 12/ /22/76 Supervision IV 8/ /78 (PCR) 24 /a Including depreciation of earth-moving machinery.

9 - vi - DISBURSEMENTS TABLES ANDHRA PRADESH: Accumulated Disbursements (Credit 226-IN). (US$ million) Appraisal Revised IDA Fiscal Year Estimates Estimates Actual as Percentage and Quarter End (1970) (9/74) Actual of Appraisal Estimate 1971/72 September 30, December 31, March 31, June 30, /73 September 30, December 31, March 31, June 30, /74 September 30, December 31, March 31, June 30, 1974 a/ /75 September 30, December 31, March 31, June 30, 1975 b/ /76 September 30, December 31, March 31, June 30, /77 September 30, December 31, March 31, June 30, /78 August 31, 1977 C/ 24.4 a/ Original Closing Date b/ Revised Closing Date c/ Actual Closing Date: June 30, 1977.

10 - vii - TAMIL NADU: Accumulated Disbursements (Credit 250-IN) IDA Fiscal Year Appraisal Actual as Percentage and Quarter End Estimate Actual of Appraisal Estimate 1972 March 31, June 30, September 30, December March 31, June 30, September 30, December 31, March 31, June 30, September 30, December 31, 1974 a/ March 31, June 30, September 30, December 31, March 31, June 30, September 30, December 31, March 31, June 30, September 30, December 31, 1977 b/ 34.4 March 31, a/ Original Closing Date b/ Actual Closing Date

11 - viii - MAHARASHTRA: Accumulated Disbursements (Credit 293-IN) (US $ M) IDA Fiscal Year Actual as a and Quarter End Appraisal Percentage of Actual Estimate Estimate 1973 March 31, June 30, September 30, December 31, March 31, June 30, September 30, December 31, March 31, June 30, September 30, December 31, 1975 a! March 31, June 30, 1976 b/ September 30, a/ Original Closing Date b/ Actual Closing Date

12 Project Performance Audit Report INDIA: ANDRRA PRADESH, TAMIL NADU AND MAHARASHTRA AGRICULTURAL CREDIT PROJECTS (Credits 226, 250 and 293-IN) HIGHLIGHTS The three state-oriented credit projects financed on-lending activities by the state land development banks and some commercial banks. Funds were channeled through the Agricultural Refinance and Development Corporation (ARDC). They also made provisions for equipment, consultants and supporting services. Sub-loans were primarily for supporting on-farm investments in minor irrigation, with smaller shares assigned to land development, tractors and implements. By and large, the refinancing and on-lending procedures worked as anticipated, although commercial banks' participation was smaller than expected. Serious delays (3 years over the appraisal estimate) occurred in Andhra Pradesh and Tamil Nadu. The project areas were expanded so as to cover the full state territory in both cases. Actual lending veered towards minor irrigation, fueled by strong demand, lower unit costs and difficulties in executing the other components; the number of units financed substantially exceeded appraisal targets. Lending for land development was considerably reduced as a result of lack of progress in major command area development programs, difficulties in consolidation of farm holdings and other problems. About the same number of tractors as anticipated were procured, but tractor deliveries were delayed until 1976/77 as a result of procurement disagreement between the Government of India and IDA affecting a series of similar projects!/. Overall, many more sub-loans were made under the projects (251,000) than anticipated (147,000, a 71% increase). About one-sixth of the projects' lending went to small farmers, defined as cultivators whose potential post development income was Rs2,400 or less. Most of the projects' objectives were broadly achieved. Production increased, though less than anticipated (due to a smaller increase in cropped area in some parts, and to yields lower than appraisal estimates, in others). Farm employment expanded significantly. Investments implemented under the projects were financially and economically viable. Rates of return are similar to or above appraisal estimates; increases in paddy prices contributed to this result. Under the projects, incremental income criteria for sub-loan appraisal were introduced. ARDC expanded and improved its refinancing operations, opening the way to the current all-india line-of-credit approach in IDA lending for agricultural credit in India. The LDB system expanded and developed in Andhra Pradesh but did not improve in Maharashtra. While it developed initially in Tamil Nadu, it later declined. The state groundwater agencies established under or supported by the projects did develop into large and competent groundwater research, development and control organizations. Regulation of groundwater development by administrative/lending procedures was only partly successful. 1/ For a detailed discussion of this issue see PPAR on Punjab and Haryana Agricultural Credit Projects (Credits 203 and 249), Report No. 2684, of October 5, 1979.

13 The following points may be of special interest: - Poor credit discipline and some political interference are the primary reasons for low recoveries (paras. 36 to 44). - Enforcing credit discipline may conflict with investment financing in disadvantaged areas (paras. 45 and 46). - Efforts to prevent well interference and groundwater overdevelopment through lending restrictions were only partly effective and led to unequal water distribution; deferring the establishment of effective controls has made a definitive solution more difficult (paras. 55 to 63, and 80 to 84). - The existence of other sources of funds, which were cheaper and more convenient than the project funds, contributed to the slow pace of disbursements in Andhra Pradesh and Tamil Nadu (paras. 69 to 71). - IDA agreed to disburse credit funds against sub-loans made under the Andhra Pradesh and Tamil Nadu projects which did not conform with the lending conditions stipulated in the credit agreements, or which had been effected beyond the previous, restricted project areas (paras. 74 to 79). - Covenants on reuse of IDA funds recovered from participating lending agencies were not appropriate, for recovered funds are available for a period shorter than the shortest possible loan terms. They have not been complied with (paras. 88 and 89).

14 Project Performance Audit Report INDIA: ANDHRA PRADESH, TAMIL NADU AND MAHARASHTRA AGRICULTURAL CREDIT PROJECTS (Credits 226, 250 and 293-IN) I. PROJECTS SUMMARY 1. A. Introduction 1. The projects were the third, fifth and seventh in a series of ten similarly designed state-oriented credit projects aimed at increasing agricultural production in India. Total costs were estimated at Rs 1195 M (US$160 M) of which Rs 828 M (69%) was for minor irrigation investments, Rs 148 M (12%) for land development, Rs 123 M (10%) for tractors and implements (no tractors were financed in Maharashtra), and Rs 96 M (8%) for equipment, consultants and supporting services. The IDA credits for US$89.4 M were to finance 56% of project costs, the balance being met by the Agricultural Refinance and Development Corporation (ARDC) and the state governments, participating banks and the borrowers themselves. 2. The project's main objectives were to: (a) increase production and employment through financially and economically viable investments in minor irrigation and land development; (b) facilitate expansion of the cropped area by providing support for the state's farm mechanization goals (except in Maharashtra); (c) improve technical services in support of these investments; (d) introduce effective measures to control groundwater development; and (e) improve institutional credit channels. B. Andhra Pradesh 3. The credit became effective in May 1971 and was fully disbursed by August 1977 (July 1975 except for the tractor category). Total costs are estimated at Rs 376 M, of which 67% was for minor irrigation, 25% for tractors, and 8% for land levelling and supporting services. 4. About 80,000 minor irrigation units, twice the appraisal estimate, were financed. This increase was made possible by reallocations from other components and because of lower unit costs than had been estimated at appraisal. The land development program, delayed by lack of progress in major command area development projects, was eventually cut by over 50%. The project area was extended to cover the whole state, thus enabling these components to be fully disbursed by / Adapted from the PCRs.

15 Tractor deliveries were delayed until 1976/77 as a result of a procurement disagreement between GOI and IDA affecting a series of similar projects 1 /. Some 1,400 units were eventually delivered out of the estimated 1,500, at prices far higher than estimated at appraisal. 6. A total of some 95,000 sub-loans (appraisal estimate 52,000) were made under the different project components. The higher than anticipated number of irrigation units easily compensated for the shortfall in land development cases. Both these categories of investment proved to be as viable financially (FRR 16% to over 50%, ERR 30% to over 50%) as predicted at appraisal (FRR 20% to over 50%), although the area benefitted in each farm from either kind of investment was smaller than anticipated. Data on tractors are not conclusive, but point to a satisfactory economic rate of return. The total value of incremental production generated by tractor sub-borrowers in 1976 prices is estimated to be Rs 360 M as compared to appraisal estimates of Rs 460 M. Sixteen per cent of project lending went to small farmers, defined as cultivators whose potential post development income was Rs2,400 or less (no such target was set at appraisal for this project, or the other two projects reviewed here). Further, project investments generated some 8 million man-days of incremental on-farm employment, mainly for landless laborers. 7. The land development banking system (LDB) had at first some difficulty in adapting to the new policies and procedures required under the project, and its lending suffered during the drought and political unrest of 1971/72. Since then it has streamlined its operations in line with project consultants' recommendations, expanded steadily with ARDC support, and consistently maintained one of the best recovery records in the country. Issues it will have to tackle in the near future include: (a) how to address the special problems of drought-prone areas and areas with poor bank coverage; and (b) the long-term viability of a number of small constituent primary banks. 8. Commercial banks' (CBs) participation in the project was smaller than expected, reflecting a lack of expertise in agricultural lending during the early part of the project which has by now mostly been remedied. In addition, there were a number of procedural obstacles placed on their expansion. Now that CBs are becoming more prominent in term lending for agriculture, the need for closer monitoring and guidance by ARDC of this aspect of their services is becoming apparent. 9. The State Groundwater Department (SGD) set up with project assistance also performed well, charting out new areas for possible exploitation and monitoring the use of existing resources. 1/ Disagreement was mainly on procedures and requirements for tractor procurement (interchangeability of individual tractor parts, financing imported vs. domestically manufactured tractors, existence of acceptable pre- and post-sale services, institutions involved, etc.). See PPAR or Punjab and Haryana Agricultural Credit Projects (forthcoming) for a detailed discussion of tractor procurement issues.

16 The design of the project, only slightly modified from a wellestablished pattern to suit local conditions, was basically sound. Weaknesses concern overestimates of command area, unrealistic downpayment requirements, and inconclusive handling of tractor pricing issues. IDA supervision was infrequent but not inadequate in view of the apparent lack of state-level implementation problems. IDA lending to Andhra Pradesh has continued to grow via the three ARDC general lines of credit. C. Tamil Nadu 11. The credit became effective in November 1971 and was fully disbursed by the end of March The delay in project completion was mainly due to the disagreement between Government of India (GOI) and IDA on tractor procurement procedures mentioned above (see para. 5). Total cost was about Rs 524 M of which 70% was for minor irrigation, 2% for land development, 20% for tractors and 8% for equipment and consultants. The IDA credit financed about 53% of project costs; ARDC and participating banks, 29%; and farmers, about 18%. 12. Most of the project objectives were broadly achieved. There was not, however, much progress toward effective control of gr(,andwater development, and the situation of the LDB has deteriorated, being characterized by high overdues and a high proportion of non-viable primary banks. 13. About 52,000 investments in minor irrigation were financed (considerably more than appraisal estimates) of which about 20% were in the sedimentary areas and 80% in the crystalline (hardrock) areas. Of these, about 16% were pumpsets, 14% filter points and tubewells, 41% dugwells and 30% improvements of existing wells. The average area developed under each minor irrigation investment was lower than estimated at appraisal. 14. The land development program (including drainage) was reduced considerably due to difficulties in consolidation of farm holdings and delay in the construction of irrigation dams involving interstate water agreements. Some 4,500 ha were improved as compared with appraisal estimate of 28,000 ha. 15. There was a delay in the supply of tractors due to the disagreement between IDA and GOI on procurement and, thereafter, a slow processing of tender documents by the Government of Tamil Nadu (GOTN). By the time the first tractors arrived (March 1976) many potential buyers had already obtained tractors from non-project sources and the supply shortage which the project was to have relieved had improved significantly. Nevertheless, some 1,627 tractors were provided under the project as compared with an appraisal estimate of 1, The State Groundwater Directorate (SGD) played a central role in the execution of the project. It developed into a large and competent groundwater investigation organization. The requirements and impetus provided by the project both demanded and facilitated not only a substantial increase in staff, but also the budget allocations and equipment needed to support the initiation of a series of important hydrological studies and water resource inventories. The SGD was assisted by three consultants provided under the project.

17 17. Regulation of groundwater development in areas in which overexploitation was a danger was attempted through requiring credit institutions to employ spacing and density criteria. However, since institutional lending accounted for only a small proportion of the investments in minor irrigation (about 13% in the period 1971/72 to 1975/76), the impact of those control measures was very limited. Moreover, employment of the spacing criteria appears to permit inequitable allocation of limited groundwater resources (this issue is discussed at length in paras. 49 to 63, and 80 to 84 below). On the positive side, the project's emphasis on adequate hydrological studies for the appraisal of minor irrigation investments stimulated extensive groundwater investigation and prepared the ground for further progress in this area. 18. During the project investment period there was a deterioration in the performance of the LDB. Overdues increased from about 14% at the beginning of the project to about 40% in 1976/77, primarily because of droughts and their adverse impact on benefits and cash flows. The volume of lending also declined, to a large extent, because in many areas investment in minor irrigation had reached the limit permitted by the spacing criteria. As a result, and because of their high overdues, many of the PLDBs have become non-viable. 19. Despite ARDC's repeated urging, GOTN did not permit the CBs to participate in the minor irrigation component of the project. 20. Investments financed under the project were financially and economically viable. Estimates of average financial rates of return (FRR) of minor irrigation investments vary from 19% to 24% (appraisal estimates were 28% and 29%), those of land development are estimated at over 50% (appraisal estimate was 46%) and that of tractors at 19% (appraisal estimate was 30%). Estimates of economic rates of return (ERR) range from 26% to 41% for minor irrigation (appraisal estimates were 21% and 29%), over 50% for land development and 42% for tractors (appraisal estimates were 37% and 26% respectively). The overall ERR is estimated at 35% as compared with appraisal estimate of 28%. 21. About 56,000 sub-loans (appraisal estimate: 50,000) were made under the different project components. Since the average area benefitted in each farm was much below appraisal targets, overall incremental cropped area is estimated at about 42,000 ha as compared with an appraisal estimate of 60,200 ha. Incremental production at full development, estimated at Rs 250 M at 1976 farmgate prices, is lower than appraisal estimate (Rs 480 M at 1976 farmgate prices) mainly because of the smaller increase in cropped area. About 20% of the lending for minor irrigation was to small farmers. In addition, the poorest segments of the population, namely the landless who provide most of the casual labor in rural areas, benefitted considerably through an increase in the demand for labor generated by the project, which is estimated at about 6 million man-days per annum. D. Maharashtra 22. The credit became effective in January 1973 and was fully disbursed by July 1976, about six months after the original closing date. Total cost was about Rs 509 M of which about 90% was for minor irrigation and 10% for land development (including earth moving machinery).

18 The project's main objectives, i.e. increased production and on farm employment, were achieved. The project also resulted in substantial improvements in the technical services in support of minor irrigation investments. However, the measures taken to control groundwater development were not entirely effective, mainly because the Government of Maharashtra (GOM) has not yet introduced legal measures to control private investments in groundwater development where the groundwater resource is limited. The objective of improving the State Land Development Bank (SLDB), the main institutional credit channel, was not achieved mainly because of severe droughts during the project period and political intervention in SLDB collection activities, which resulted in high overdues. Although the last two objectives were not achieved, lessons were learned during project implementation that have been taken into consideration during later projects. One important, although unplanned, achievement was that, as a result of SLDB not coming up to expectations, the CBs were strengthened and played a much more prominent role than previously in financing agricultural development in Maharashtra. 24. About 73,000 minor irrigation units were financed. Of these about 45% were new wells with pumpsets, 37% new wells, 10% well improvements, and 8% pumpsets only. Because most minor irrigation investments are made in 2 to 3 installments, about 42% of the units financed were incomplete by the end of the project development period and were expected to be completed by mid-1977 with funds provided under the first ARDC general line of credit (ARDC I) funded by Credit 540-IN. In terms of complete units, an equivalent of 60,000 investments were financed, considerably more than the appraisal estimate. Investments envisaged in deep tubewells did not materialize, due mostly to lack of farmer response, while those in lift irrigation were implemented by GOM under drought relief measures. 25. Although generally in accordance with appraisal guidelines, the standard of land development works initially was low but improved as the project progressed. Land development was a new venture in Maharashtra, and there were difficulties with the organization of credit which took a long time to overcome. However, the experience gained from this was later used in designing other IDA financed projects. Land development equipment purchased under the project arrived too late to be utilized in project land development works. However, given GOM plans for future expanded land development, the machinery is expected to be fully utilized. 26. The Groundwater Survey and Development Agency (SGD) was substantially strengthened and played an important role in the project. Some 1,467 groups of watersheds covering the entire state have been delineated and their water resources assessed. Procedures for appraisal of groundwater resources have been modified and substantially improved. However, regulation of groundwater development in areas in which overexploitation is a danger through requiring credit institutions to employ spacing and density criteria was only partially successful, as a great part of minor irrigation investments is financed from private sources. Moreover, the spacing criteria appear to enforce an inequitable allocation of the limited groundwater resources. (see paras and 80-84).

19 Most project lending (95%) was channelled through the SLDB, which was to be rehabilitated under the project. The rehabilitation effort proved unsuccessful despite the fact that GOM introduced several reforms in the LDB system's management and financial structure. Overdues remained at a very high level (63% of demand at June 30, 1977), primarily because of three successive years of drought (1970/71 to 1972/ 73) which reduced farmer repayment capacity and forced a massive, although unsatisfactory, rescheduling of loans. Rescheduling took the form, in most cases, of increasing annual repayments in the remaining years of the original loan period (subject to a maximum increase in annual repayment of 25%), rather than by extending the loan period, thus compounding the problem. The SLDB, heavily engaged in reorganization and rescheduling activities, did not commit sufficient resources to recoveries, and frequent rescheduling and lack of effective action against defaulters generally reduced credit discipline. Default reached epidemic proportions, and, by their sheer numbers, defaulters carry considerable political weight. The overdues situation was aggravated further in that GOM from time to time advised SLDB to avoid taking coercive actions against defaulters. GOM and ARDC are concerned about the poor recoveries and have since agreed on a rehabilitation program. Further, GOM has issued instructions to take coercive action irrespective of the age of overdues and amount involved. 28. Another problem affecting SLDB is the high proportion of subbranches which are non-viable mainly due to insufficient lending volume. This is also a serious problem in the short-term credit system. Integration of the short and long-term systems could improve profitability and, at the same time, enable farmers to obtain their short and long-term credit requirement from one source (see paras. 47 and 48 for further discussions). 29. The CBs gained considerable experience in agricultural lending, strengthened their staff, and expanded their rural branches network. They increased their share in ARDC disbursement in Maharashtra from 5% at the beginning of the project to about 36% in 1976/77 under ARDC I, and in many areas have provided a much needed alternative source of institutional agricultural credit. 30. Investments financed under the project were financially and economically viable. Estimates of average FRR for the main minor irrigation investment (well and pumpset) vary from 36% and 20% in sugarcane growing areas to 11% in areas where the main crops were foodgrains. They are satisfactory, although less than the appraisal estimate of 40% overall for the State. The FRR for land development is estimated at over 50%, the same as the appraisal estimate. Estimates of the ERR range from 18% to over 50% (appraisal estimates 33% to 50%) for minor irrigation and over 50% (same as appraisal) for land development. 31. About 100,000 sub-loans (appraisal estimate 45,000) were made under the different project components. Incremental cropped area resulting from minor irrigation investments is estimated at about 35,000 ha and from land development investments at about 24,000 ha (appraisal estimates were similar). The value of incremental production at full development, estimated at Rs 530 M per annum at 1976 farmgate prices, is lower than appraisal estimate (roughly Rs 900 M in 1976 prices) mainly because yields

20 - 7 - are lower than estimated at appraisal. Only 10% of the lending was to small farmers (as defined above; see para. 6), but the poorest segments of the population, mainly the landless, benefited considerably through an increase in the demand for hired labor estimated at 13 million man-days E. General Aspects 32. An important achievement has been the introduction of the incremental income criteria for evaluation of loans under the projects. It resulted in a change from the security-oriented approach which was predominant before the projects towards a production-oriented one. 33. ARDC played an important role in all aspects of project implementation. It guided participating banks in adjusting to project procedures and assisted them in overcoming bottlenecks and contraints. It carried out its supervision and scheme preparation functions satisfactorily, but needs further strengthening at the regional level to widen its development and evaluation roles. This is being undertaken in the ARDC II General Line of Credit Project. 34. Experience gained under the projects exposed some weaknesses of previous and current approaches to groundwater control and in the agricultural credit system. The lessons learned were used in designing the ARDC III general line of credit. II. MAIN ISSUES IN THE PCRs 35. Several issues are dealt with in the PCRs. The main ones are arrears and collections, integration of the two cooperative credit systems, groundwater development and control, and requirement of insurance for well failure. Since the PCRs are not included as a part of the PPAR, the relevant PCR texts are reproduced in this chapter. A. Arrears and Collections 1. The Case of Maharashtrail/ 36. The management of SLDB attributes the high overdues to the following main reasons: (i) drought conditions; (ii) poor rescheduling method; (iii) unsatisfactory pre-project standards for investments; (iv) incomplete investments and delays in electrical connections; and (v) priority given to short-term credit in recovery from crop proceeds. In a survey covering a sample of SLDB sub-branches, the main reasons for overdues given by sub-branch managers were as follows: crop failure 46%; misuse of loan 23%; unfavorable market conditions 19%; other reasons 12%. Overdues on IDA supported lending (66% of demand as of June 30, 1977) exceeded the state average (63%) and indicate that the improvements in the standard of lending under the project did not result in better recoveries. The tying of recovery to the marketing of cash crops like sugarcane and cotton has also been disappointing mainly due to the low priority given to SLDB in the recovery from sale proceeds. 1/ Taken from PCR on Maharashtra, paras. 3.10, 3.11, 3.14 to 3.17; and 5.03.

21 Drought is clearly one of the main reasons for poor recoveries. Despite the protection provided through irrigation facilities, there have been complaints about reduction in loan repayment capability resulting from drought. Well discharge is affected by variations in rainfall; during drought periods the discharge is reduced and so is the area which can be irrigated and/or the amount of water per unit area 1/. Moreover, most of the investors also had a considerable area of non-irrigated land (65% according to survey results) even after the investment. Rainfed crops remained, therefore, a major source of income and the overall financial position of the household remained sensitive to drought though to a lesser extent than before the investment. However, even in nondrought districts overdues per branch as of June 30, 1977 averaged 57% (range 35-76%), a percentage only slightly lower than the average for branches in drought prone areas which was about 61% (range 45%-81%). 38. Overdues were thus almost as high in non-drought as in drought areas. The recovery rate of loans extended under the project was also low despite the improved lending technique. It appears, therefore, that the primary reason for low recoveries has been the poor credit discipline in Maharashtra resulting from many years (over 15) of poor repayments and the lack of effective action against defaulters. 39. A better understanding of the reasons for overdues can be gained by analyzing the individual farmer's motivations. For most farmers the investment in minor irrigation is a once in a lifetime venture. Therefore, the negative effect of default on potential future borrowing from SLDB is not of great concern to them. Short-term credit is, on the other hand, required every year and the farmer has a built-in interest to repay it in order to ensure a regular supply. Farmers can be expected, therefore, to give priority to repayment of short-term credit-2. Farmers who borrow from private lenders (at a rate of interest estimated to be about 24%) can be expected to give priority to the repayment of such loans because of the high rate of interest. Therefore, borrowers can be expected to give the lowest priority to repayment of SLDB loans. 40. A restoration of credit discipline is a pre-condition to the success of any future rehabilitation of the SLDB and would require tough and unpopular steps which could only be taken effectively with the full support and cooperation of GOM. The number of SLDB defaulters is about 400,000 and the total number of defaulters on loans from cooperative credit institutions (including short-term) exceeds one million. By sheer numbers, the defaulters carry a considerable political weight and actions against them, which are feasible when they represent a small minority, have become politically difficult. There are indications that political interference has affected loan recovery. From time to time GOM signaled to SLDB that it should relax its collection effort (the most recent occurrence was in February 1977) and effectively prevented SLDB from 1/ Presently there are no statistical data to quantify the resulting reduction in net income. 2/ Some staff members have questioned this conclusion, because of the poor repayment record of the short-term credit agencies. See footnote to para. 47.

22 - 9 - introducing a special debt collection drive through provision of financial incentives to its staff. GOM officials have occasionally condemned large and wealthy farmers for willful default and threatened to take strict measures against them. These threats, unfortunately, were not translated into effective actions and have failed to improve loan recoveries. 2. The Case of Tamil Nadul/ 41. The land development bank system (LDB) in Tamil Nadu, which was in the past one of the best in India, has been declining in efficiency and profitability. The main reasons are the poor recoveries and the reduction in the volume of lending for minor irrigation. The primary land development banks (PLDBs) showed satisfactory collection efforts for many years as they enforced timely utilization of loans, kept in close contact with the farmers and took disciplinary action against delinquent staff. However, since 1974, matters took a turn for the worse mainly as a result of the drought; political pressures undermined repayment discipline. Rescheduling has become a routine affair, but while in 1974/75 it was probably justified due to severe drought, there are indications that in the following years rescheduling may have been influenced, in part, by political considerationa-/. The deterioration took place despite the fact that foreclosure actions in case of misutilization and other defaults were prompt and relatively effective. In 1976, there were about 4,000 accounts under foreclosure proceeding and in 200 cases the land was auctioned off. By June 1976, the amount under foreclosure totalled Rs 19 M (down from Rs 31 M in June 1975) compared to a reserve for bad debt of Rs 109 M (another indication of SLDB's financial soundness). Given that Tamil Nadu had a good recovery record prior to 1974/75, it should be possible to reestablish credit discipline and it is possible that the situation could again be brought under control. However, strong commitment on the part of the State government including support of prompt actions against defaulters will be required in order to achieve that. 42. The reduction in the volume of lending and, therefore, profitability of PLDBs is a more serious problem. GOTN plans to reduce the number of PLDBs but this will increase the area of operation per PLDB and make their contact with the individual farmer more difficult, although these difficulties are being partly compensated by increasing the number of field staff. GOTN hopes to expand diversified lending through the PLDBs. However, the possibilities for diversified investments are limited. Moreover, PLDBs would face strong competition from the CBs, which plan to increase agricultural lending, and, as they have been prevented by GOTN from lending for minor irrigation, their increased lending will be mainly for diversified investments. Competition from the State Cooperative Bank can be also expected. 1/ Taken from PCR on Tamil Nadu, para and Annex 3, paras. 18 and 22. 2/ GOTN has stated that relief was given in respect of certain loans both relating to cooperatives as well as governmental dues on account of drought and not otherwise.

23 The Case of Andhra Pradeshl 43. LDB has achieved and maintained one of the highest collection rates in India, both from final borrowers to PLDBs and from these to SLDB. Moreover, this overall performance has held steady between 1975/76 and 1976/77, against an all-india pattern of worsening recovery disciplines. On an individual basis, an impressive 165 out of 205 PLDBs fell into the unrestricted lending category (meaning that they had 25% or less unrecovered) in As many as 82% of these 165 PLDBs were cited as having reached 100% recovery. Even if in some cases year-end window dressing may exaggerate the success of a collection campaign, these are very solid results. Overall, overdues from final borrowers are less than 16% of demand. 44. A combination of several factors underlies this LDB good recovery discipline: (a) Professional tradition. Collection has been stressed by every management since the LDB's foundation and in the Madras State cooperative banking system before that. Loan recovery is the single major target against which PLDB staff and PLDB board performance is measured and rewarded, either tangibly or in terms of status. In the collection season (March to July) this drive is intensified to the point where other work virtually ceases in many PLDBs, even if this means neglecting potential new customers. (b) Cash incentives. Borrowers who repay late, as elsewhere in India, pay penalty interest; borrowers who repay before the well-publicized target date are allowed a discount of 2%, which appears to attract a large number of takers. PLDBs which reach 100% recovery are awarded significant cash bonuses, which compensate for the extra effort to clear the last outstanding accounts. This and a series of other competitive awards help to preserve a good recovery record once it has been achieved. (c) Political climate and LDB management. At the local level, elected PLDB boards reinforce credit discipline by acting as pressure groups to "sponsor" many borrowers and take some responsibility for their repayment performance. This is in net contrast to states where defaulters are a politically significant minority and local PLDB boards tend to act as a shield against the SLDB's collection efforts. At the state government level, there appears to be a strong tradition of noninterference with the internal operation of the banking system, strengthened by solid resistance from the professional management to changes that appear to run counter to perceived sound banking practice. 45. Not all officially declared drought-prone areas in the state have consistently high overdues, and by no means all high overdue PLDBs are in those areas. However, it can be argued that recent discipline of linking new advances by PLDBs to their recovery performance is holding back development in some areas where fresh investment is most needed. These issues were addressed in depth in the preparation of the third ARDC general line of credit. With particular reference to Andhra Pradesh, the following points should also be made: (a) Consistent with its emphasis on credit discipline, the SIDB refuses to authorize rescheduling of 1/ Taken from PCR on Andhra Pradesh Annex 3, paras. 10, 11, 13 and 14.

24 loans on a block basis in years of drought as is the practice in some neighboring states. It has argued convincingly that such "blanket" measures will lead to expectations of further concessions and ultimately the erosion of discipline state-wide. However, without the buffer of rescheduling, the impact of drought and other calamities is more rapidly reflected in the choking off of new institutional finance. (b) Under existing Debenture Norms Committee rules, the state government is allowed to inject additional capital into those PLDBs which would otherwise not qualify for new lending, this equity being considered equivolent to recoveries up to a certain maximum. This is a temporary measure to give breathing space to PLDBs who can temporarily continue lending under existing programs provided they attempt to bring their collections back into line. To date, the GOAP has not availed itself of this provision, but is now considering doing so. 46. This is an area where developmental objectives and sound banking practices are not easy to reconcile. Given existing regional disparities in Andhra Pradesh as well as the varying profitability of its PLDBs, this issue should be kept under close review. In addition to facilities now allowed by the Debenture Norms Committee (such as concessions to banks lending more than 50% of their total to small farmers), the state government and the banking system should jointly consider what incentives should be provided to spread institutional financing in disadvantaged areas, e.g. by stimulating competition for the restricted PLDBs. B. Integration of the Two Cooperative Credit Systems! 47. Integration of the long and short-term credit systems2was proposed by the Committee on Integration of Cooperative Credit Institutions which studied the problem under the terms of ARDC 1!/. No advance towards that end has been made yet. Allowing PLDBs to extend short-term credit to their long-term borrowers is a further alternative, although the impact of this on the institutions lending short-term would have to be examined carefully. Besides improvements in service to farmers and PLDB viability, either alternative would enable borrowers to obtain short-and long-term credit from one bank and, from the lending banks' point of view, facilitate the recovery of loans. 1/ Taken from PCR on Tamil Nadu, para 4.08 and Annex 3, para. 23; PCR on Andhra Pradesh, para. 3.14, and PCR on Maharashtra, Annex 2, para. 48. For a more detailed discussion of integration, see PPAR on ARDC I, currently under preparation. 2/ Presently the LDB system provides only long-term credit. Short-term credit is provided through a three tier system which includes the primary agricultural cooperative credit societies at the village level, the district cooperative banks at the middle tier and state cooperative bank at the apex level. 3/ Staff members who participated in the appraisal missions mentioned to OED that they had considered the need for a greater integration of both branches of the credit system. But they did not go further on any proposal for integration because at that time the repayment record of the short-term credit institutions was much worse than the LDB's.

25 There is a division of opinion within GOTN and GOM regarding the advantages and disadvantages of integration, and decisions are not expected in the near future. In Maharashtra, the SLDB has a considerable number of non-viable sub-branches. An RBI Study conducted in 1975 estimated that only 54% of SLDB sub-branches were viable, about 30% were potentially viable and 16% were non-viable. As the short-term credit system had, and still has, a large proportion of non-viable primary cooperative credit societies (PCCS), the viability of both credit systems could be improved through a partial or full integration. The GOAP is planning an experiment involving the partial merger of four selected PLDBs with their more numerous PCCS counterparts. This experiment, which amounts to a sub-contracting arrangement, aims at using the PCCS' village level facilities together with PLDBs' specialized areas of expertise. A decision on starting such merger is yet to be adopted. GOI undertook under ARDC III to review the progress made by the cooperative credit system in meeting the farmers' credit needs through one source or closely coordinated sources by December C. Groundwater Development and ControlyI 1. The Need for Control 49. The need to control groundwater development has been advanced on two primary grounds: (a) to protect the cash flow from an individual investment against later interfering development; and (b) to prevent widespread and long-term drawdown trends which might have unacceptable economic effects (particularly increased pumping costs). a) Protection of individual investments. 50. Theoretically, because farms in general are small relative to the physical area of influence of typical wells, the yield of almost any privately owned well in India could be reduced by the development of a new well on an adjacent plot. Prima facie, this is a legitimate area of concern for both the investor (farmer) and his banker. However, the empirical evidence available so far does not support the above concern. 51. Despite the fact that control through institutional credit covered only a small percentage of minor irrigation investments, to date there are not many documented cases of curtailed loan repayment due to sinking of a new well nearby. The main reasons for this may include the following: (a) the geological conditions in the plateau in the three states under analysis (and over most of India) are such that the physical areas influenced by typical wells are not much larger than the fields of those farmers who can afford wells with pumpsets; (b) because of drought conditions in the three states during much of the projects implementation period, it is not possible to distinguish between reductions in discharge due to drought and reductions due to a sinking of a nearby well; (c) the minimum spacing between wells was based on conservative calculations; and (d) it appears reasonable to assume that most farmers will not sink a well too near an existing one since this will put their investment at risk; if correct, this hypothesis would have important policy implications and should therefore be examined through studies and/or surveys. 1/ Taken from the PCRs: (i) Andhra Pradesh, para. 2.06, and Annex 4, para. 16; (ii) Tamil Nadu, paras to 3.15; and (iii) Maharashtra, paras and 2.05, and Annex 3, paras. 43 to 45.

26 Because of the enormous technical and socio-political difficulties, it is presently impractical to impose a comprehensive control on groundwater development although efforts to develop such a system should continue. Consequently for the short-run (3-5 years) the use of pragmatic control measures, such as restrictions on institutional credit, restrictions on electric connections and similar measures (e.g. limiting public drilling operations to wells cleared by SGD) will have to continue. Additional measures of control and protection of institutionally financed investment could be achieved by basing the number of ARDC approved investments on more detailed studies and investigations conducted by SGD and ARDC. Such investigations will have to forecast future levels of groundwater development for each scheme area, including development from noninstitutional sourceql/. b) Protection against widespread overdevelopment 53. Long-term excessive drawdown over large areas could increase pumping costs to economically unacceptable levels. Over most of India, i.e. the hardrock areas as well as the alluvial areas in high rainfall (over 1,000 mm) conditions, this type of drawdown is practically impossible and in the arid states with substantial alluvial deposits (Gujarat, Haryana and Punjab) groundwater is already fully developed. In the latter states overall planning and control of all water resources, surface as well as underground, are required rather than narrowly based legislation limited to groundwater. 54. While some danger of over-exploitation of the groundwater does exist: (a) this is a highly complex question, not only technically, but also economically, socially and legally, and none of these dimensions are yet fully understood; (b) the question was deemed to be of importance to the lending banks primarily because of the possibility that overdevelopment in certain limited areas would lead to reductions in the yield of some wells which, in turn, might reduce the repayment capacity of some well owners; and (c) the imposition of these and the other technical conditions of the credit caused the SGDs to focus their attention on this potential problem area and thus to develop a series of criteria for well site selection and control which are appropriate and have minimized the problem of interference between wells. In the particular case of Maharashtra, it now appears that the size of area in which there is a danger of over-exploitation is far smaller than that of the elemental watersheds which were designated in the Credit Agreement as the basic unit of hydrologic analysis, and that the mechanisms of recharge and discharge are also such as to render inappropriate the formulations imposed under the credit. 1/ Under ARDC III, GOI and ARDC have agreed to introduce a new appraisal system for groundwater development financing in areas with excessive groundwater development. This system would be based on more detailed investigations by SGDs. The choice among three levels of resource evaluation and groundwater investigation will depend on the current groundwater development of the proposed project area. (See also footnote 2/ on page 25).

27 The one serious criticism that might be leveled at the appraisal exercise is that it did not perceive the potentially serious retrogressive effects of the imposed spacing criteria as a mechanism for allocating use of an extremely limited and valuable natural resource. It is clear now that only a very small fraction (10-15%) of the farmers in any hydrologic unit in the Tamil Nadu and Maharashtra plateaus can have a well requiring annual withdrawals of 16,000-24,000 cubic meters per year. As the first borrowers in any area are typically the more well-to-do and aggressive farmers, the conditions imposed by the credit, in the hope of preventing overdevelopment of the groundwater, probably had, as their principal effect, the perpetuation and exacerbation of existing inequities of income distribution. 56. Perceiving this problem is however far easier than solving it, because to share the quantity of say 80,000 m 3 of annual recharge among the 40 farmers occupying a typical square km in Maharashtra would give them only 2,000 m 3 each. Withdrawals at this low level would not support the investment in any kind of well. The logical conclusion is that for development to be equitable as well as economically viable, some form of forced water sharing, from jointly, communally or government owned facilities, is unavoidable. 2. The Measures for Control 57. In the absence of groundwater legislation, a pragmatic effort was made to prevent overdevelopment through stipulation of minimum well spacings and maximum densities. These restrictions, which at appraisal were considered to be only temporary, were also to apply to wells financed through institutional credit not under the project. The spacing and density criteria specified at appraisal and modified during the projects in the light of additional groundwater information obtained by SGDs, were generally observed. They were expected to be effective since a large proportion of minor irrigation investments was assumed to be financed through institutional credit. It appears, however, that non-institutional credit accounted for a greater than envisaged proportion of the investment in minor irrigation 1/. Thus the effectiveness of groundwater 1/ In a survey of villages, conducted by the GOM Bureau of Economics and Statistics on behalf of SLDB, which covered a sample of 90 villages in nine districts, it was found that 25% of the wells constructed in the period 1972/ /76 were financed under IDA projects. The SLDB statistics indicate that about 75% of the wells financed by it in the same period were under IDA supported programs. Assuming that this percentage applies also to the districts surveyed, it would appear that in those districts, SLDB financed only about one third of the wells. The CBs are known to have financed a relatively small number of wells and, assuming that the above percentages hold true for the state, it can be concluded that a large proportion of the wells in Maharashtra were financed from private sources. The data available are insufficient to permit drawing strong conclusions and should be supplemented by additional, carefully planned investigations. There are no reliable data on private money lenders; however, it is clear that, in most cases, farmers preferred institutional credit sources because of their lower interest charges.

28 control through limitations affecting only institutional credit is questionable. This conclusion is strengthened by statistics from Tamil Nadu which indicate that, as minor irrigation investments approach the limit allowed by the spacing and density criteria, farmers increasingly finance the investment from non-institutional sources, either to avoid conforming to these criteria or because the criteria make them ineligible for institutional loans. 58. Recently, some state governments (as GOTN) have extended controls to electric connections, and only wells which comply with the spacing criteria may be electrified. This measure should discourage uncontrolled investments to some extent, but will not be fully effective since farmers are still free to install diesel engines and many can be expected to do so despite the higher operating costs of such equipment. 59. As in the case of the other agricultural credit projects, provisional controls (pending legislation) were introduced in Andhra Pradesh through spacing and density restrictions. However, unlike the other projects it was left to the SGD to determine the spacing and density norms, subject to ARDC and IDA approval. Control through institutional credit could not have been entirely effective as it did not cover privately financed investments. There are no reliable estimates of the scope of such investments, but it is roughly estimated at 25% of total new wells. According to SGD calculations, groundwater resources in the hard rock area of Andhra Pradesh are sufficient to irrigate about 50% of the cultivated area. In this respect, the situation in Andhra Pradesh appears to be better than in the hard rock areas of Tamil Nadu and Maharashtra, in which groundwater resources are sufficient for roughly 10-15% of the cultivated area. Equitable allocation of water in Andhra Pradesh should, therefore, be easier than in those states. 60. The experience under these projects, and particularly in Tamil Nadu, indicates that by deferring the establishment of effective controls until most of the groundwater resource has been developed makes a definitive solution more difficult. Enforcement of any spacing criteria (if they are effective) means that when a farmer sinks a well, his close neighbors are de facto stripped of the right to dig a well on their land. Despite this consequence of employing the spacing criteria, no guidelines for allocation of groundwater resources among farmers have been prepared and the first to apply had the first entitlement to investment in a well. Understandably farmers who have not yet invested in wells object to such a prior right allocative regulation, while those who have invested can be expected to object to any reallocation of the resource. A more equitable allocation of groundwater development rights, introduced at earlier stages of development, would have reduced conflicting interests and facilitated

29 the establishment of effective controls-1/. Unfortunately, in much of India, that time has passed and the question of the control of groundwater exploitation is becoming an increasingly complex and difficult problem. 61. No state government has yet taken active steps to enact groundwater legislation. Most of them consider it essential to prove and quantify the damage caused in the absence of legislation. Any delay in legislation will, however, make its implementation more difficult. 62. The draft groundwater legislation, under consideration by GOTN for the last 7 years, is based on a model bill published by GOI in December The model bill includes principles which are basic to the orderly development of a groundwater resource. The difficulties lie in the practicability of implementation. The main problem is how to control development in areas in which the groundwater resource is sufficient for irrigating only a small fraction of the cultivated land. Legal controls based on spacing criteria will be especially unpopular since they will be supportive of a basically inequitable allocation of groundwater resources. Unfortunately, however, legislation based on spacing and density criteria with its acceptance of prior right is the only kind of control that GOI or any of the state governments has seriously suggested to date. Withdrawal control, which would involve allocation of irrigation water quotas, would be more appropriate, but the mechanisms for enforcing it remain to be carefully studied. 63. A major constraint on the introduction of groundwater legislation is the difficulty, and danger, of attempting to proceed from the general to the particular in terms of groundwater availability and behavior in any given area. While SGDs have developed considerable overall strength and competence, they have neither the manpower nor the data required to permit precise estimates of the potential for, or constraints on, development in a specific village. In addition, statistical shortcomings make it difficult to be precise about how much water is already being extracted and thus how much remains available. 1/ Staff members who participated in the appraisal missions mentioned to OED that the equity issues were very much in their minds, but they concluded that "a more equitable allocation of groundwater rights" was not an alternative reality at the time for the reasons argued in the subsequent paragraphs. They point out that the projects' design was the correct one to get something going within the realm of the possible. The main objective was to maximize production development (and these three credits along with those to other states have made considerable contributions to India's improved grain position) avoiding over-exploitation of the groundwater resource (which has been achieved to some extent). It was assumed that equity considerations would be taken care of by administrative means, pricing increases, encouragement of water sharing and other direct provisions, but these ones had only limited effect.

30 D. Well Failure and Possible Remediesl 64. Estimates of the proportion of well failure vary from 1% (percentage of claims for well failure subsidy) to about 13% (based on survey). SGDs claim that as a result of groundwater investigations there has been a reduction in the proportion of well failures, but there are no data to verify this. Approved cases of failed wells were entitled to a subsidy which provided for: (i) remission of total interest accrued up to the date of application for subsidy; and (ii) remission of 20% of the outstanding loanl/. As shown above, the subsidy program may have covered only a small proportion of the well failures and only a part of the investment cost. The main risk was still carried by the farmer and may have deterred especially small farmers from investment in minor irrigation. 65. The time appears to be ripe for the introduction of insurance schemes which would cover the loan repayment commitments and thus reduce investors' risk. Insurance to cover cases in which the well dug was found to be dry or nearly dry is feasible since SGDs should be able to provide the data necessary for calculating the insurance premiums. The premium could be paid by the financing bank and added to the loan amount. III. ADDITIONAL ISSUES3/ 66. The PCRs contain a thorough discussion of most project issues. Two issues, however, are only marginally mentioned or ignored: delays in project implementation, and retroactive financing. These are elaborated below. Also, the equity aspects of groundwater development are further elaborated. Minor comments are included on three other issues. A. Delays in Project Implementation 67. Two of the projects: Andhra Pradesh and Tamil Nadu, were delayed from the very beginning; their rates of disbursement fell significantly short of the appraisal estimates throughout the projects' lives. If disbursements on "irregular" and "off-limits" lending (see next section) are not taken into account, the credit funds were disbursed at a rate of 1/ Taken from PCR on Maharashtra, paras and / In Tamil Nadu, the Failed Wells Fund rules provide that up to 90% of the zost of the failed wells (subject to a maximum of Rs. 10,000) can be granted by way of assistance. 3/ A number of other institutional, financial and procedural issues stem from these projects. All of them have been treated together with those arising from other State and all-india projects in the PPAR on ARDC I.

31 US$3.1 million per year in Andhra Pradesh (39% of the appraisal estimate) and of US$3.8 million per year in Tamil Nadu (less than one-third of the estimate). Eventually, the two projects' closing dates had to be postponed by more than three years each. (see Graph 1)1/. 68. Three factors, which are mentioned in the PCRs, contributed to these delays: (i) Disagreement between IDA and GOI on tractor procurement (this issue is analyzed in detail in the PPAR on Punjab and Haryana, currently under preparation). As a result, tractors were delivered starting in March 1976 and up to January 1977 (Andhra Pradesh) and December 1977 (Tamil Nadu). But this factor explains only part of the delay. On the one hand, tractors represented only 12% of total project costs in Andhra Pradesh, and 14% in Tamil Nadu. In terms of credit disbursements, they represented less than US$5 million in Andhra Pradesh, and only US$6.3 million in Tamil Nadu. On the other hand, the disbursement graphs (see Graph 1) show no startling jump in disbursements or in the disbursement rate in late 1976 and through 1977, when disbursements for tractors took place. (ii) Delays in starting the minor irrigation component in Andhra Pradesh; some limitations on groundwater availability in Tamil Nadu and difficulties in obtaining the necessary clearance for new wells from the respective SGDs in either state are also stated in the PCRs or were mentioned to the audit mission. (iii) The restricted area the projects initially covered2/. These factors also contributed to the slow implementation pace, but do not explain such a large discrepancy between actual implementation times and appraisal forecasts either. 69. There is a fourth factor, which is not mentioned in the PCRs, which also contributed to the slow pace of disbursements: LDBs had other sources of funds, which were cheaper and more convenient than the project funds. From the time they started operations until , LDB's main source of funds was the issues of ordinary debentures controlled by RBI. The IDA funds were to be channelled through a different credit instrument: special development debentures, controlled and bought by ARDC. ARDC's funds could be obtained at 6.5%, against 6 to 6.3% for ordinary debentures (earlier costs were even lower). When the IDA credit projects were started, the margins allowed for LDBs were 2.5% for special debentures and 2.7% for ordinary debentures; they increased as rates of interest were raised. Over recent years (including the projects' last years as well as 1978), ordinary debentures would provide LDBs a spread of about 4.25% on minor irrigation loans and 4.5% on diversified lendings, as compared with a uniform spread of 3% under the special debentures. 1/ The Maharashtra project was implemented almost on schedule - in fact, ahead of schedule through the first two years, and behind it thereafter, ending up with a rather minor nine-month delay. A possible explanation for this anomalous behavior is ventured in para / Further, during there was an energy crisis in Tamil Nadu and the energization program could not be expedited.

32 AV DM f fm" L5 a 91 ti b t n v4 1t 6.44\ lol-o.s7- -T4.t--AAc o ,, " -.T1.,s -K" i,i

33 Further, the repayment term under IDA supported projects was to approximately match those allowed to the ultimate borrowers. This implied annual repayments by SLDBs to ARDC, while repayments of ordinary debentures to RBI are made in a lump sum after 10 or 15 years. To generate such a lump sum most of the loan collections has to be allocated to a sinking fund of GOI securities, LDB debentures and other securities. At the beginning of project implementation, the sinking fund arrangements were financially marginal for LDBs. Over the years, however, as the returns on investments increased, the income from sinking funds has made a significant contribution to LDB incomes. Currently the amount of investments range from one third to more than half the amounts of loans outstanding for most LDBs, and resulting profits are substantial 1/. Moreover contributions to the sinking funds can be treated with more flexibility than annual payments./. In addition, special debentures introduced new restrictions to regular SLDB and PLDB operations: technical, economic and financial regulations for appraising the sub-projects and subsequent supervision of actual fund utilization. LDBs were therefore hesitant to borrow from ARDC under special debentures because ARDC's innovative rules and control were found to be somewhat constraining. 71. ARDC, with implicit IDA suport, has been trying to reduce the importance of ordinary debentures as a source of funds for LDBs as well as equalize the rules and procedures for issuing both kinds of debentures. Initially, some restrictive measures were recommended - for example, not allowing the refinancing of minor irrigation works through ordinary debentures in those areas where ARDC-financed schemes were operational. Eventually, the debenture eligibility formula was extended to ordinary debentures in 1977, and rates of interest were brought closer to each other in March 19791/. These developments have had different effects. On the positive side, a more technical approach to long-term lending has been established through introducing regulations on subproject analysis and supervision. From this, a more substantial impact resulted on production, employment and income. Also, disbursements under the follow-on 1/ Funds "deposited" in the sinking fund are earning from 6 up to 9 and 9.5% today, to eventually redeem debentures floated at 4% to 6% interest rates. Since almost no cost is incurred in handling investments and risk is nil, that full spread is profit. 2/ The Committee on Integration (see para. 47) has reported that some PLDBs were building their sinking funds at a rate less than proportionate to theit ordinary lending. 3/ Rates of interest on ordinary debentures are currently higher (by 0.25%) than those on special debentures regarding minor irrigation, land development and diversified investments for small farmers, and lower (by 0.75%) regarding diversified investments for other farmers.

34 projects (ARDC I and II) have moved more easily and faster. On the negative side, these developments may have resulted in IDA funds substituting for local capital that ordinary debentures might have raised, thus reducing the overall additional impact IDA funds were supposed to generate in Indian agricultural investments 1/. 72. Why was it that the higher cost of project funds (as compared with the cost of the alternative ordinary debentures) contributed to delaying project implementation in two of the state projects under analysis, and not in the third one, Maharashtra? The mission found no obvious explanation for this anomaly. Some guesses can be ventured, however: First, the Andhra Pradesh and Tamil Nadu LDBs were strongly businessoriented systems (the Andhra Pradesh LDB still is). They were profitminded and readily noticed the difference in operational costs and profit margins and reacted accordingly. The Maharashtra LDB, however, was acting much more like a para-governmental agency; this feature became more apparent when the LDB structure was reorganized and the former PLDBs were made SLDB branches. Second, the Maharashtra project came more than one year after Tamil Nadu and more than a year and a half after Andhra Pradesh. The eventual outcome of the struggle between IDA/ARDC on the one hand and the profit-minded state LDBs on the other was much more evident by then, reducing the temptation to continue resisting the use of IDA funds provided through special debentures. And third, the Maharashtra LDB was in an extremely poor financial situation; the project had to be made conditional on LDB's financial rehabilitation by GOM. LDB was then under pressure to lend, and for this particular purpose, IDA funds were additional financial resources made available to LDB. B. Disbursements against "Irregular" and "Off-limits" Lending2/ 73. To ensure that project funds would be usefully invested, IDA required that new and more sophisticated techniques were to be introduced among the PLDBs for analyzing the feasibility and profitability of the sub-borrowers' proposals. Since effective PLDBs were required for improved analysis of such proposals and for supervising their actual implementation thereafter, project lending was geographically restricted to those districts in which IDA and the state governements agreed that effective PLDBs existed. Also, IDA-financed lending for minor irrigation and for land development was restricted in Andhra Pradesh and in Tamil Nadu to those areas which most obviously lent themselves to such activities. 1/ There are conflicting data on the extent of additionality under the projects. The debenture eligibility formula as well as the additionality (actually, the lack of it) are analyzed in detail in the PPAR, on ARDC I, op.cit. 2/ The terms "irregular" and "off-limits" are not very appropriate to refer to those loans, for they may suggest procedural irregularities. This is not the case; they just did not properly conform to the projects' lending conditions. Since the terms have been extensively utilized in Bank memoranda and in the PCRs, they have been retained in this text.

35 After two years of implementation, GOI and IDA agreed to disburse credit funds against loans made under the Andhra Pradesh and Tamil Nadu projects 1 which did not conform with the lending conditions stipulated in the credit agreements./. These loans ("irregular" loans), which had already been refinanced out of local funds by RBI and ARDC, were retroactively transferred to the project accounts from non-ida ARDC schemes and normal PLDB lending-/. The ordinary debentures which had been previously issued to refinance them were cancelled; special debentures were issued instead under IDA refinancing. Later on, IDA agreed to the State Governments' request to open the full state territory to lending under the projects, and then, to retroactively refinance all lending which had been effected off the previous project limits since the credit agreement had become effective ("off-limits" loans). The same debenture cancellation/debenture issuance routine was carried out again. The switchover of "irregular" and "off-limits" loans to the IDA accounts was suggested as a solution to the slow pace of the disbursement of IDA credits; such solution satisfied both GOI, which was then under extreme need of foreign exchange, and IDA, which wished that sizeable disbursements take place. Had the operation not been carried out, disbursements would have continued to lag behind because of the reluctance of the LDBs to apply the new lending conditions and perhaps not to forego the more attractive terms and conditions of local finance. 75. The discussion between GOI and IDA on the financing of "irregular" loans focussed on the Gujarat case because of the large amount of loans involved (some US$18 million out of an overall amount of about US$27 million loans for the four credits) and of some deliberate indiscipline of the Gujarat LDB in the application of the terms of lending stipulated in the credit agreement A/. IDA recognized that the "irregular" lending in question was in keeping with the project purposes and that the deviation from the new lending norms adopted under the agreements were minor and temporary. It was also appreciated that the Gujarat LDB was the first which had to apply the more stringent lending conditions of the IDA projects and had faced special difficulties in this respect. Consequently, IDA agreed to waive certain terms and conditions of lending 1/ And also under the Gujarat and the Haryana projects. (Credits 191 and 249-IN). 2/ Deviation concerned mainly rates of interest (lower), term of repayment (longer) and down-payment (smaller). 3/ Hence the term "retroactive finance" utilized in the PCR on Tamil Nadu. 4/ In the other states, the amounts involved were small relative to Gujarat, and the lending contrary to the terms of the credit agreements took place only between the dates of the agreements and effectiveness, most likely as a result of some misunderstanding of the date of application of the lending covenants.

36 laid down in the various credit agreements (January 1973) 1/. Information in the project files on the discussions on "off-limits" lending is scanty. Eventually, IDA also issued the letters duly authorizing drawing credit funds against loans made outside the project areas (Februiry 1974). 76. As a result of these agreements, the previously lagging disbursements under both credits suddenly jumped, approaching momentarily appraisal estimates. Then, disbursement rates came down again to their previous level (see Graph 1). 77. The PCR on Gujarat analyzed quite briefly the issue of "irregular" lending. The PPAR, carried out under OED's abbreviated procedures, found the PCR discussion adequate, but did not analyze the issue. At that time, the decision to bypass this and certain other issues was deliberate: the intention being to take advantage of the experience offered by a comparative analysis of this and several other projects in the Indian agricultural credit group, which were expected to be completed shortly (PPAR on Gujarat, op.cit, para in the Memorandum). 78. The audit mission confirmed that Rs 168 million (equivalent to US$21 million; to 35% of total on-lending under the project, and to 56% of total lending for minor irrigation under the project) had been actually disbursed in Tamil Nadu on the basis of retroactively financing: (i) "irregular" loans made within the agricultural area of the project, in an amount of about Rs 31 million, and (ii) loans made outside the project area during the period November 2, 1971 to December 31, 1973, amounting to about Rs 137 million. The PCR on Andhra Pradesh is silent on this issue; Rs 53.6 million in irregular lending were found to have been switched over to the IDA project under the "irregular" concept. No information could be obtained from the files of the Regional Office or Controller's, of IDA, or from ARDC or Andhra Pradesh's SLDB on the amount of "off-limits" lending which was refinanced. Information available in one of SLDB's annual reports seems to indicate a total amount of Rs 20 million. 79. The mission noted a discrepancy in the Tamil Nadu accounts between the amount of special IDA debentures issued under the project (Rs million as of June 30, 1976) and the related amount of loans that should have been financed by the debentures (Rs 207 million as of June 30, 1976). After exhaustive discussions with officials of ARDC and the Tamil Nadu SLDB, the mission found it impossible to have the two figures reconciled by either ARDC or SLDB, because recording of the transfer of the loans to the IDA project loans was never made. In fact, the loans 1/ Amounts and dates are as follows: (a) SLDB Gujarat: Rs M of loans for minor irrigation advanced during the period June 3, July 31, (b) SLDB Haryana: Rs 30.8 M of loans for tubewells, advanced during the period June 11, 1971 to July 31, (c) SLDB Andhra Pradesh: Rs 53.6 M of loans for minor irrigation and land levelling advanced during the period January 6, 1971 to July 31, (d) SLDB Tamil Nadu: Rs 33.2 M of loans for minor irrigation advanced during the period June 11, 1971 to July 31, 1972.

37 transferred had never been individually identified, as in the case of those made directly under IDA credits. Since 1973 SLDB's auditor has repeatedly asked to "reallocate the mortgages under the normal loans and earmark them to the required extent as cover for the special development debentures (IDA) after carrying out necessary adjustments in accounts"; no corrective action has been taken by SLDB, however. This may be due to the fact that the identification of the loans concerned may involve a laborious and time-consumming work. Further, the mission found no indication that the auditor's recommendations had been noted by either ARDC or IDA. The audit mission was assured by ARDC and SLDB, as well as by Bank officials, that the loans refinanced conformed with the project criteria and that deviations regarding lending terms were minor. GOI reiterated that the switchover of loans had been considered appropriate since lending conformed to project criteria. However, as mentioned above, the information required to substantiate these claims is not available. C. Groundwater Exploitation Control and Equity Considerations 80.1/ There is no explicit statement on ownership and use rights of groundwater in India. The right of ownership of groundwater seems to rest with the owner of the overlying land 2/. The right of using the groundwater is acquired through actually capturing it - i.e., drilling and operating a well, and Indian farmers currently have an unrestricted right to use groundwater obtained from a well sunk on their farms. Since groundwater cannot be enclosed, it must be accepted that all landowners above an aquifer share (proportionally to their land holdings) in their rights to own and use the groundwater. However, these private rights of ownership and use are indefinite for any individual user, because other co-owners may take possession of the water at any moment. This situation can lead to depletion of the aquifer over time if eventually groundwater pumping exceeds recharge. Since this is the case in most of the South Indian plateau, the fact that groundwater may be eventually depleted acts as stimulus for farmers to sink their wells as early as possible. The risk of depletion becomes a motivation for depletion, rather than for conservation, of the resource. 1/ This paragraph, and the two following the next one draw some material from Veerman, T.S. - "Water Policy and Water Institutions in Northern India: The Case of Groundwater Rights", Natural Resources Journal, Vol. 18, July 1978, pages 569 to / This is not absolutely clear. Although groundwater has never been declared to be publicly owned, the case can be construed that the Central and State Governments are the ultimate owners of groundwater because ownership rights on groundwater are derived from ownership rights in land, and the governments are the ultimate owners of the land.

38 IDA properly identified this problem and readily realized that public intervention was needed to prevent local well interference as well as long term overexploitation (see paras. 49 to 54). One particular system was selected to perform such control: IDA required that maximum well density and minimum well spacing be observed (paras. 57 to 59)1/. These density and spacing criteria were properly conceived from a strictly technical point of view. If they were properly enforced - currently, they are enforced only for those wells financed by institutional credit sources - they would produce the anticipated long-term effect on groundwater availability and utilization, and on well profitability. 82. But IDA failed to realize at the time that these criteria implied introducing in those states a legal approach to groundwater rights which differed from the present Indian practice, and which leads to increasing wealth and income concentration. The solution adopted was to restrain the right of late-comers to get access to the groundwater resource. But this amounted to acknowledging full private ownership rights to whoever first sank a well over the full quantity of groundwater he could pump, and depriving without compensation all his neighbors (within the area determined by the spacing limit) of their right to use (to own, too?) the groundwater. Thereafter, the "early birds" could draw with impunity the groundwater underlying their and their unfortunate neighbors' land ; their use rights had become definite. Two classes of farmers are thus created: the owners of some land with right of use over all groundwater, and the owners of the rest of the land, with mere rights of ownership over groundwater but no rights of use over it. Even worse, the first to get access to groundwater (i.e., to the credit or other financial resources to finance sinking and motorizing a well) have usually been the better-off farmers.2/ 83. A more equitable solution to the problem - and the problem is indeed serious - would have to recognize and protect the correlative and coequal rights of all overlying landowners to use their proportional share of the total groundwater supply. Such a solution could include one or more of the following elements: (a) water sharing among all farmers, proportionate to either their land holding or their water requirements; (b) water rationing among all those who want to sink a well - or have one already; (c) proper compensation to farmers who lose or must involuntarily surrender their right to use the groundwater underlying their lands; (d) public or social drilling and operation of wells, with water being proportionately distributed or sold to all interested farmers; etc. In either of the first two cases, those receiving water would have to share the investment and operating costs of the owner/operator of the well. In the third case, a decision would be required on whether such compensation should be paid by the neighboring well owner(s) or by the government. A suitable government agency (like the SGDs, which are drilling wells for 1/ Such restrictions were introduced into the states for the first time as a result of the projects. 2/ The comment made by GOI on this paragraph and the following one does not actually refer to these paragraphs but to para. 52. A footnote was added there.

39 drinking water in Tamil Nadu and Andhra Pradesh) or a cooperative/commercial institution should be identified for the latter. Water sharing or rationing could be carried out under either the appropriation system (under which prospective groundwater users must apply for a Government license to draw a certain amount of water), the adjudication system (under which water for quarreling users is apportioned by the judiciary), or traditional systems (under which someone with local authority - local councils, elders, leaders, etc. - iron out discrepancies or impose authoritative decisions, or some traditional consensual mechanisms are put into work). 84. Two additional points - both referring to flexibility - must be borne in mind while devising a more equitable solution. One is to introduce enough flexibility into the system to accommodate future demands, mainly coming from many small farmers (who most certainly will not claim their right to groundwater until they become more aware of their rights and possibilities) and from other potential users (such as nearby villages or far away cities V/). The other one is to consider - where there is such a possibility - conjunctive development and use of surface and groundwater; correlative rights to groundwater, once adjudicated through whatever system, should be flexible enough to be integrated into a rotational or other distribution system for canal water deliveries, or to be shifted to other potential users when the first ones could be more efficiently served with surface water. D. A Cautionary Note on the SGDs' Effectiveness 85. All three PCRs show a well deserved satisfaction with the successful establishment, development or strengthening of the respective SGDs. Indeed, they have become good technical agencies. But some of the SGDs have yet to become fully effective agencies. The PCRs do mention that the human and technical resources of this or that particular SGD have had to be spread too thinly, and that more information on water resources is still required. ARDC III continues to place emphasis on further SGD development and on water resources evaluation. Two examples of limited effectiveness were mentioned to the audit mission: (i) SLDB and PLDB officials expressed the opinion that the Tamil Nadu SGD, for lack of detailed information, was a bit too conservative in its clearance of new well requests. Areas which allegedly could bear additional wells - even areas where the water table is currently rising - were said to be "frozen" for future development for the time being. (ii) ARDC technicians felt that the Andhra Pradesh SGD was some times a bit too lenient in granting clearance for well drilling schemes in some areas; parameters used there by SGD to justify clearance for additional wells were found to be too optimistic when compared with parameters derived for the same areas by GO specialized agencies. 1/ For example, there is an area in Northern Tamil Nadu where groundwater development for farming purposes has been restricted or forbidden to secure sources of water for future demands by Madras, the State capital.

40 E. A Cautionary Note on Statistical and Accounting Data 86. The audit mission was faced with difficulties -i finding consistent and reliable data on loans advanced, and their breakdown according to the resources under which they were financed (ordinary and rural debentures, ARDC non-ida credits, IDA statewide credit or ARDC I). The problem results from the fact that the IDA-supported projects - including the statewide as well as the ARDC projects - are continuing programs where several investments financed under one project were completed under follow-on projects. The sub-loan accounts are only markdd as "IDA projects" - as compared with "normal loans" -, without signaling whether they were financed under the state project or under one of the ARDC projects. Further, the PLDB or SLDB branch staff are not usually aware of the details of the project under which SLDB has floated debentures. The same problems were found regarding the breakdown of loans outstanding and the debentures issued and outstanding. For example, in Tamil,Nadu the annual volume of lending given in the PCR (Annex 3, para. 10) is very differeat from that found in the LDB annual reports and that furnished to the mission (these refer to the cooperative years, the other may refer to the LDB fiscal year but this fact does not explain the differences). Also, "irregular" loans switched over to the state project have not been recorded among the loans made under the project. In Maharashtra no data are available on the number and amounts of loans outstanding under the different IDA-financed projects (the state project, ARDC I and ARDC II); the breakdown of debentures issued and outstanding according to sources of financing is incomplete and some figures furnished to the mission are obviously inaccurate.-v-2/. 87. Regarding the keeping of separate projects accountsr3/ some progress could be made in the standardization of and consistency in accounts; the adoption of uniform recording procedures, more especially regarding interest; the presentation of financial statements and the collection of statistical data. The audit mission was shown no evidence that covenants which stiputate the maintenance of separate accounts for project lending by a participating Bank, and the regular auditing of such accounts, had been strictly adhered to. Commercial banks are annually audited by their statutory auditors. The auditing of PLDB accounts is not fully satisfactory in two respects: quality of auditing and delays 1/ Actually, the Maharashtra and Tamil Nadu SLDBs are not performing properly and are two of the six states which may not be eligible to participate in the ARDC III project. 2/ Prompted by these findings of the OED mission, the last supervision mission of ARDC II (June 1978) checked five different states (including Maharashtra). The mission reported having been able to trace payments through IDA to GOI to ARDC to LDB in every state; the accounts checked in Maharashtra were clearly identifiable. 3/ Discussions on accounting and auditing is absent in the PCRs.

41 (which vary by states, but are usually long)1/. It must be said, however, that the registrars of cooperative societies cannot cope with supervising and auditing the considerable number of cooperative societies, and PLDBs cannot afford to pay for private auditors. F. Reuse of IDA Funds 88. A special covenant concerning the reuse of IDA funds recovered from participating lending agencies was included in the project agreements for Andhra Pradesh and Maharashtra 2/. Under this covenant, repayments to ARDC by participating agencies had to be deposited in a special account to be used only for refinancing ARDC-approved agricultural development schemes, to be carried out in conformity with sound administrative, agricultural and financial practices. In this way, ARDC could count on significant amounts of funds, because the repayment periods under its refinancing arrangements are shorter than GOI's terms of lending. 89. Repayments to ARDC by participating banks have not been deposited in a separate account, and this was never reported by IDA supervision missions. But it was not appropriate to have such covenants included in the project agreements: although the funds involved are substantial, they are available for a period shorter than the shortest term of any loan to be made under any scheme similar to those financed under the projects - even leaving aside the time required to commit and disburse them. Covenants like these do not appear in more recent projects. 1/ The Bank's Regional Office finds that quality of accounting is generally good. 2/ Also in the projects for Punjab and for Karnataka, but not in those for Tamil Nadu, ARDC I and ARDC II.