India - Fourth Agricultural Refinance and Development Corporation Credit Project

Document of The World Bank FOR OFFICIAL USE ONLY py P n Report No. 3629-IN L F C Pt8 STAFF APPRAISAL REPORT INDIA FOURTH AGRICULTURAL REFINANCE AND DEVELOPMENT CORPORATION CREDIT PROJECT February 1, 1982 Agriculture D Division South Asia Projects 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 authorizati6n.  CURRENCY EQUIVALENTS (as of January 20, 1982) US$1.00 = Rs 9.043654 The U.S. Dollar/Rupee exchange rate is subject to change. Conversions in this report were made at US$1.00 to Rs 8.0, which represents the projected average exchange rate over the disbursement period. WEIGHTS AND MEASURES Metric System ABBREVIATIONS ACD - Agricultural Credit Department of the Reserve Bank of India ARDC - Agricultural Refinance and Development Corporation ARDC I - First Agricultural Refinance (and Development) Corporation Credit Project ARDC II - Second Agricultural Refinance and Development Corporation Credit Project ARDC III - Third Agricultural Refinance and Development Corporation Credit Project CALCOB - Committee on Agricultural Loans through Commercial Banks CB - Commercial Bank(s) CRAFICARD - Committee to Review Arrangements for Institutional Credit for Agriculture and Rural Development GOI - Government of India LDB - Land Development Bank(s) (Generic term for all SLDB and PLDB whether operating on a federated or unitary system) NABARD - National Bank for Agriculture and Rural Development PLDB - Primary Land Development Bank(s) RBI - Reserve Bank of India RRB - Regional Rural Bank(s) SCB - State Cooperative Bank(s) SLDB - State Land Development Bank(s) CROPPING SEASONS Kharif - June to September Rabi - October to February Summer - March to May FISCAL YEAR GOI - April 1 to March 31 ARDC and Cooperatives - July 1 to June 30 CB - January 1 to December 31  FOR OFFICIAL USE ONLY INDIA FOURTH AGRICULTURAL REFINANCE AND DEVELOPMENT CORPORATION CREDIT PROJECT Table of Contents Page No. I. THE AGRICULTURAL SECTOR .............................. 1 Introduction ..................1 Land and Water Resources I Production ... .................2 Inputs .... .......... 4 Supporting Services . . .......... 4 II. AGRICULTURAL CREDIT AND AGRICULTURAL CREDIT INSTITUTIONS . 5 A. Background 5..... 5 B. The Cooperative Credit System. 6 C. Land Development Banks ......... .... 7 Review of the Six Selected LDB . 11 Measures to Improve LDB Operations. 12 Special Programs and Participation Conditions for Bihar, Gujarat, Karnataka, Madhya Pradesh, Maharashtra and Tamil Nadu LDB . .13 D. Commercial Banks ............. ... .. . .15 E. Regional Rural Banks .. ..17 III. PERFORMANCE OF BANK GROUP FINANCED AGRICULTURAL CREDIT PROJECTS IN INDIA ...18 Introduction ..18 Results of Completed State and ARDC Credit Projects...... 18 Impact of Indian Credit Projects on Agricul- tural Development . .20 IV. THE PROJECT. . . 21 A. Project Genesis ....21 B. Brief Description and Objectives . 22 C. Detailed Features . . .22 Minor Irrigation ...22 Pumpset Electrification to be Financed from ARDC's Own Resources ...24 Land Development .. ... .......... 25 Soil Conservation and Land Reclamation.. 25 This report is based on the findings of an appraisal mission that visited India during May 1981, consisting of Messrs. G. Stern, B. Argyle, L. Abbie and C. Hachero (Bank); W. Barber and A. Stoneham (Consultants), and P. van Dooren (Netherlands), G. Gwyer (UK) and M. Sorensen (USA), representa- tives of bilateral donors, who participated full time in the appraisal. Messrs. J. Upadhyay, P. Brumby and L. Ljungman (Bank) participated part time in the appraisal, and Mr. T. Turtiainen contributed to the report. 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.  - ii - Table of Contents (cont'd) Page No. Diversified Lending Eligible for IDA/IBRD Refinance ... e...................6..0..0 ........ ....... 25 Diversified Lending to be Financed From ARDC's Own Resources .... . ....... .. .. .......... . 28 Training ............ ............................$ ...... O..... 28 D. Project Cost ..... .... .... ...... 30 E. Financing ......................s..*... ........ 31 F. Procurement . 31 G. Disbursements .*..... .... . .0 ..... . ..0.....#..0.*........ 32 V. PROJECT IMPLEMENTATION ........ . . . . . . . . . ........... .... .*. ..... .. . 32 A. The Agricultural Refinance and Development Corporation ......................................*... 32 B. National Bank for Agriculture and Rural Development . 35 C. Project Lending Procedures ........6... ............... 36 D. Lending Terms and Conditions ..o- ........... ... .. 37 E. Lending to Small Farmers and in Less Developed States ............................. o.....oo*o... *......... - .... 38 F. Accounts and Audit o.. . . . . . . . ... ........ ..... . . .*..... . 39 G. Monitoring, Evaluation and Reporting ................ 39 VI. PRODUCTION, MARKETING, PRICES AND FINANCIAL RETURNS TO BENEFICIARIES . ..... .....-.-... oo ...o 40 Production ....... ...... .. ..... , . . 40 Marketing and Prices o-- ................... 41 Financial Analysis . ......... . ... ..... ........... 42 VII. BENEFITS, RISKS AND JUSTIFICATION ooo ... ....... ... 44 Benefits . ........ 44 Benefits o ..... - -........o.... - ... ............................ 4 Beneficiaries .................................. 44 Environmental Effects ..... .................. oo. 44 Economic Analysis ..... ...................... .45 Other Benefits .......... .............*........ 46 Project Risks .............. . ................... 46 VIII. SUMMARY OF AGREEMENTS ................ o ................ 46 SCHEDULE A Lending Terms and Conditions SCHEDULE B Guidelines for Disbursement Against Public Direct Irrigation and Augmentation Tubewells SCHEDULE C Guidelines for Levels of Evaluation of Groundwater Resources Appropriate to Levels of Resource Development SCHEDULE D Small Farmer Definition SCHEDULE E Criteria for Eligibility of Participating Banks to Receive Refinancing from ARDC  - iii - Table of Contents (cont'd) ANNEXES ANNEX 1 - Table 1 - Loans Issued by State Land Development Banks Table 2 - Loans Outstanding of State Land Development Banks Table 3 - State and Primary Land Developments Banks - Summary of Overdues Table 4 - Eligibility Classification of SLDB Branches and PLDB Table 5 - Ranking of Land Development Banks by Loan Recoveries and Comparative Recoveries of Commercial Banks Table 6 - Commercial Banks: Overdues of Agricultural Advances by States Table 7 - Summary Data on IDA/IBRD Assisted Projects through ARDC Table 8 - ARDC Lending Program by State and Subsector, 1982-1983 Table 9 - Estimated ARDC and IDA/IBRD Disbursements Table 10 - ARDC Growth Since Inception Table 11 - ARDC Disbursements by Subsector 1971-1981 Table 12 - ARDC Refinance by Agencies Table 13 - ARDC Condensed Balance Sheets 197 5/76-1984/85 Table 14 - Condensed Statement of Income and Expenditure 197 5/76-1984/85 Table 15 - ARDC Summarized Cash Flows 197 5/76-1984/85 ANNEX 2 - Operations Review of Six Selected LDB Table 1 - Lending and Recovery Data for Six Selected States Table 2 - Financial and Performance Ratios: Bihar SLDB Table 3 - Financial and Performance Ratios: Gujarat SLDB Table 4 - Financial and Performance Ratios: Karnataka SLDB Table 5 - Financial and Performance Ratios: Madhya Pradesh SLDB Table 6 - Financial and Performance Ratios: Maharashtra SLDB Table 7 - Financial and Performance Ratios: Tamil Nadu SLDB Table 8 - Ownership of Land Development Banks in India (1978-1980) ANNEX 3 - Financial Analysis Table 1 - Financial Rates of Return Table 2 - Farm Cash Flow Model 1. Dugwell with Pumpset Table 3 - Farm Cash Flow Flodel 2. Shallow Tubewell and Pumpset Table 4 - Farm Cash Flow Model 3. Land Development Table 5 - Enterprise Cash Flow Model 4. Coffee Plantation Table 6 - Enterprise Cash Flow Model 5. Dairy (2 Cross-Bred Cows) Table 7 - Enterprise Cash Flow Model 6. Inland Fisheries (Pond 1 ha) Table 8 - Cropping Patterns and Yields Table 9 - Investment Costs, Loan Terms and Budgets Table 10 - Investment Costs, Loan Terms and Budgets ANNEX 4 - Economic Analysis Table 1 - Financial and Economic Prices Table 2 - Economic Rates of Return Table 3 - Economic Costs and Benefits: Minor Irrigation Table 4 - Economic Costs and Benefits: Diversified Lending Table 5 - Summary of Project Impact ANNEX 5 - Riskiness and Drought in Hardrock Areas Table 1 - Financial Impact of Drought in Hardrock Areas  - iv - Table of Contents (cont'd) Annex 6 - Documents Available in the Project File Annex 7 - Organization and Management Study of State Land Development Banks - Terms of Reference LIST OF CHARTS Chart No. WB 23130 - Principal Agricultural Credit Institutions Chart No. WB 23125 - Major Institutional Funding for Agricultural Term Loans Chart No. WB 23412 - Agricultural Refinance and Development Corporation Organization Chart - As on June 30, 1981 MAP IBRD 15972R - India: Generalized Occurrence of Groundwater, Rainfall and Evaporation, and ARDC Regional Offices  INDIA FOURTH AGRICULTURAL REFINANCE AND DEVELOPMENT CORPORATION CREDIT PROJECT 1. THE AGRICULTURAL SECTOR Introduction 1.01 India has a population of about 688 M, growing at an annual rate of about 2%. Since 1960, per capita income has grown at an annual rate of 1.4% and reached US$180 in 1979/80. Much remains to be done to improve the living standards of the vast mass of urban and rural poor, which comprise about 50% of the total population. Accordingly Government of India's (GOI) development plans emphasize maintaining food self-sufficiency, alleviating poverty and creating employment, especially in rural areas. 1.02 Agriculture is the dominant sector of the economy. It contributes about 40% of GDP, accounts for about 70% of all employment and provides the base for about 50% of India's exports. During the last decade GOI has concen- trated on raising foodgrain production by increasing use of irrigation, fer- tilizers, plant protection and good seeds, backed by improved extension and research services. Production has responded impressively and over the decade before 1979/80 foodgrain production grew at about 2.75% per annum--sufficient to meet consumer demand, to eliminate imports (which had averaged nearly 5 M tons per year for the 15 years preceding 1976), and to reduce real foodgrain prices to consumers. At the same time India was able to build up buffer stocks which enabled it to overcome the effects of severe drought in 1979/80 with- out food imports. While the performance of the recent past and the probable future trends suggest that on average foodgrain supplies will exceed demand, the balance remains delicate and the need for foodgrain imports to maintain consumer supplies or adequate buffer stocks could arise from time to time. During 1981, for instance, India took advantage of favorable prices to import grain to replenish stocks drawn down by the 1979/80 drought. GOI is therefore continuing to give high priority to grain production and the Sixth Five-Year Plan (1980/81 - 1984/85) aims to maintain grain self-sufficiency and a 15 M tons reserve stock. The Plan is, however, giving greater emphasis now to other crops and other subsectors including animal husbandry and fisheries. Institutional term lending would continue to be a key input to increase pro- duction, through developing minor irrigation in particular. However, in con- formance with GOI plans, increasing amounts of credit would support investments for expanding the livestock, fisheries, plantation/horticulture and forestry subsectors. Land and Water Resources 1.03 India has a total land surface of about 328 M ha of which 175 M ha (53%) are classified as arable. The possibility of reclaiming additional land is limited. In 1977/78 about 143 M ha (82% of arable land) were cultivated but, due to multiple cropping, the gross cropped area was about 171 M ha.  -2- 1.04 The southwest monsoon begins at the end of May or early June and continues until November/December; its duration is at a minimum in the north- west and at a maximum in the south. The wettest months are June-September when most of the country receives 80% or more of the annual rainfall. Mean annual rainfall ranges from 200-300 mm in the Rajasthan desert to more than 5,000 mm in the Western Ghats and in the hills of Assam. The crucial months for rainfed agriculture are July and August. The amount and distribution of rainfall in those months largely decide the fate of kharif 1/ crops. Farmers derive benefits from irrigation in most parts of India, even in high rainfall areas where it allows cropping during the dry season (November/December - May/June). Of the irrigated area 34.6 M ha or 60% are irrigated from surface sources and 22 M ha or 40% from groundwater, mostly pumped or lifted by pri- vately owned irrigation equipment much of which has been financed by agricul- tural credit. An area of 22 M ha is irrigated from groundwater and this is 55% of the about 40 M ha that GOI believes to represent the total area that can be irrigated from groundwater. The present distribution of development relative to resource potential is variable. States such as Gujarat, Haryana, Punjab and Tamil Nadu have reached high levels of development. By contrast, West Bengal, Bihar, Orissa, Madhya Pradesh and the northeastern States have enormous groundwater resources which have barely been tapped. The Sixth Five-Year Plan aims at creating an additional 7 M ha of irrigation from groundwater and the lending program of the Agricultural Refinance and Development Corporation (ARDC), which would be supported by the project, would play an important part in attaining Plan targets. 1.05 In the case of surface water irrigation, GOI estimates that 73.5 M ha can be irrigated compared with the 34.6 M ha or 47% of the potential that has been developed so far. The Sixth Five-Year Plan projects development of 6.7 M ha, to raise the surface irrigated area to 56% of potential. The main works would be financed through State budgets, but ARDC-supported credit to farmers would be important for financing investments such as land levelling and construction of field channels. Production 1.06 Crops. Production of foodgrains dominates the agricultural sec- tor and cereals occupy about 61% of cropped area, followed by pulses (19%), oilseeds (10%) and fibers (.5.5%). The rest of the arable area is devoted to root crops, sugar cane, plantation crops (including important export crops like tea, coffee and cashews), orchards and a wide variety of vege- tables and spices. Foodgrain production has increased steeply from about 60 M tons in the early 1950s to about 132 M tons in 1978/79, mainly due to increase in production of wheat (8 M-35 M tons) and rice (25 M-54 M tons). Due to severe drought, production fell to 108.9 M tons in 1979/80, but recovered in 1980/81, for which year production is estimated at 130 M tons. Increased production resulted mainly from successful varietal im- provements, backed by spread of irrigation (22.6 M ha in 1950/51 to 49.5 M ha in 1977/78) and increase in use of fertilizers (2.23 kg/ha in 1950/51 to 21.11 kg/ha in 1977/78). Production of pulses and oilseeds, by contrast, has been static or increased slowly because of lack of genetic breakthrough. 1/ Rainy season June-September.  - 3 - Cotton production remains constrained by pest problems. Sugar cane experienced rapid expansion up to 1977, when local and export prices collapsed. As a re- sult production has dropped since then, but with recent firmer prices is likely to continue to expand. 1.07 Because of its varied climate, a large range of horticultural and plantation crops is grown in India. There has been a steady expansion of area under orchards, partly assisted by institutional credit. Fruits are mostly for the internal market, but an important export trade to the Middle East is being developed, for instance, for mango. Improvement in the plantation crop sector is also receiving financial assistance from institutional credit for moderniz- ing operations to cater for the expanded local market and to retain the share of world exports. In general the plantation sector is backed by effective research that has identified and proven improved production and processing technology. Improvements of this group of crops would be encouraged by the project. 1.08 Animal Husbandry. India has a large cattle and buffalo population. Of an estimated 253 M in 1972 about 178 M (70%) were cattle used mostly for draft purposes and about 57 M were buffalo used mainly for milk production. Milk production is estimated to have increased from 20 M tons to 30 M tons over the last two decades. This has been the consequence of programs such as crossing with superior breeds, improved fodder production and better animal health services that have been combined in a national dairy development pro- gram, supported by four IDA projects (Credits 482-IN, 521-IN, 522-IN and 824-IN), with successful programs for cooperative milk marketing and pro- cessing. 1.09 The poultry industry has made rapid progress with the emergence of commercial hatchery and poultry production enterprises. Egg production increased from about 3 billion in 1960/61 to over 10 billion in 1977/78. Institutional credit played an important role in developing this industry which aims to increase egg production by another 2 billion eggs per annum. By contrast, there has been little progress in increasing mutton and wool production, both in good demand. 1.10 Fisheries. India's 6,500 km coastline and 1.6 M ha expanse of inland water represent large fisheries resources. The fish catch is reported to have doubled in the period 1960/61-1980/81 from 1.2 M tons to 2.4 M tons, of which about two thirds is the marine catch. The latter is almost completely taken from inshore waters and is dominated by the shrimp catch (10-18% in quantity, over 50%--about US$300 M per annum--in value) which is processed for export. Institutional credit has played an important part in financing the mechanized fishing fleet and to a lesser extent processing facilities. The inshore shrimp potential is fully utilized in most parts of the shoreline and new entrants into the industry are beginning to exploit the non-crustacean catch increas- ingly, for which a local market is emerging. GOI and State governments are supporting development of inland pond fisheries and the IDA supported Inland Fisheries Project (Credit 963-IN) will play an important part by introducing modern large-scale hatchery technology on a commercial scale. 1.11 Forestry. About 74.8 M ha (about 23% of India's land area) is clas- sified as forest. However, between 50%-60% of this has been badly degraded.  - 4 - Value of production has been estimated at about US$340 M in 1974/75 of which about US$240 M was from timber and fuel and the rest from miscellaneous by- products. The demand for forestry products, particularly for fuel is enormous, far greater than available forestry supplies. GOI and State governments are pursuing vigorous tree planting programs many of which benefit tribal groups which are among the weaker section of the population. Increasing use of institutional credit is being made by corporate or private investors for establishing commercial plantations. Inputs 1.12 Provision of inputs to back agricultural capital investment programs is satisfactory. Fertilizer use has increased from 60,000 tons nutrients to over 5 M tons over two decades and continues to grow rapidly. Distribution through more than 100,000 outlets, which also handle seeds and pesticides, is satisfactory in most areas. However, transport and storage has become an enormous operation and plans are under consideration to modernize the systems. Seed supply has not proved a production constraint. Improvements supported by the two National Seed Projects (Ln. 1273-IN and Cr. 816-IN) should ensure adequate future supplies of good quality seed. Use of pesticides has also increased from 100 tons technical grade material to 60,000 tons over the last 20 years. More effective pest control measures are now being taught by improved extension services, which should bring higher benefits to expenditures on pest control. Supporting Services 1.13 Storage and Marketing. Due to increases from 12 M to 24 M tons over the last six years storage capacity of GOI States, cooperatives and the private sector is beginning to catch up with storage needs arising from the greater grain harvests of recent years. Plans to expand capacity by another 7.7 M tons over the next five years, which would be supported by the project, would hetlp to equip India with adequate storage for projected crop production and for GOI grain procurement, price support and food distribution programs. Improvement programs for some of the 4,250 important agricultural produce markets would also be assisted by the project. The programs would help to avoid marketing bottlenecks of increased production that would result from other project invest- ments. 1.14 Research and Extension. GOI and State governments are giving high priority to ensuring adequate agricultural research and extension. Both are key parameters for improving productivity and for supporting agricultural investment programs. India probably enjoys one of the best agricultural re- search services of any developing country, which is supported by the National Research Project (Cr. 855-IN), designed to strengthen the capability of uni- versities to carry out area-specific research, particularly on foodgrains. 1.15 The extension effort is impressive and benefits from strong IDA support in the form of eleven IDA projects. Under these, improved extension has been introduced into thirteen major States in India and spread to other States is being planned. The strong emphasis on improving extension and research will help farmers derive expected benefits from long-term invest- ments supported by the project.  - 5 - II. AGRICULTURAL CREDIT AND AGRICULTURAL CREDIT INSTITUTIONS A. Background 2.01 Institutional agricultural credit dates back to establishment of credit cooperatives in the early 1900s which, until recently, remained the only source of such credit. Today, however, the rural banking system comprises cooperative institutions--60% of outstanding agricultural loans in 1980-- commercial banks (CB)--about 36%--and more recently established regional rural banks (RRB)--about 4%. Private money lenders are still believed to provide the bulk of rural credit because of easy access and their preparedness to lend for consumption and ceremonial purposes, not covered by institutions. However a Reserve Bank of India survey indicated that their share of outstanding debt is declining, from 90% in the early 1960s to 75% in 1971-72, due to growing availability of cheaper institutional credit and the development of the geo- graphical coverage of the banking system. In 1980 this consisted of over 100,000 outlets and only about 40 blocks 1/ out of 5,000 in India remain with- out institutional credit facilities. Direct agricultural loans by institu- tions increased from Rs 16,750 M in 1976 to Rs 29,400 M 2/ in 1980. Within this, medium-and long-term lending more than doubled from Rs 5,000 M to Rs 11,000 M. GOI's Sixth Plan calls for further doubling of total institutional credit for agriculture by 1984/85. 2.02 Agricultural credit operations are guided and regulated by the Reserve Bank of India (RBI) which ensures that different institutions offer the same lending terms for similar lending purposes. The Agricultural Credit Department (ACD) of RBI is the main source of refinance for short term coop- erative credit institutions, but under its statutes cannot provide medium or long term refinance facilities. Such funding for the agricultural credit system is provided mainly by open market bond operations by the institutions that are supervised by RBI, refinance from the Agricultural Refinance and Development Corporation (ARDC), time deposits, and share capital contributions and loans from State governments and public corporations. 2.03 The main institutions involved in term lending for agricultural development and their principal sources of funds, are shown in Charts WB 23130 and 23125. ARDC, with assistance from IDA, has emerged as a major source of refinance for term lending and the share of such refinance obtained by the different types of institutions is shown below. 1/ Administrative subdivision. 2/ Estimated.  -6- Institution ARDC Refinance Total Outstanding 1979/80 Lending June 30, 1980 Rs M % Rs M % Cooperative Land Development Banks (para 2.05) 9,830 56.6 1,640 39.8 Commercial Banks /a 7,290 41.9 2,390 58.0 Cooperative Banks (para 2.06) 260 1.5 90 2.2 TOTAL 17,380 100 4,120 100 la Includes Regional Rural Banks whose participation in ARDC refinanced schemes is still insignificant. 2.04 In 1979, GOI and RBI appointed a high level "Committee to Review Arrangements for Institutional Credit for Agriculture and Rural Development" (CRAFICARD). Its final report released in May 1981 recommends the establish- ment of a National Bank for Agriculture and Rural Development (NABARD) to combine the agricultural credit activities of RBI and ARDC. The effects of this change are discussed in para 5.08; basically, however, the present insti- tutional structure would be retained. B. The Cooperative Credit System 2.05 Since the mid 1920s, cooperative banks have been organized by indi- vidual States into two independent groups of institutions: (a) land develop- ment banks (LDB), long-term credit system, and (b) cooperative banks, short- and mediumrterm credit system. There are 19 LDB, each consisting of a State Land Development Bank (SLDB), which lends through its own branches or refi- nances affiliated Primary Land Development Banks (PLDB). In 1979, there were 999 LDB branches and 889 PLDB. LDB provide loans for agricultural investments for periods of between 3 and 15 years. They do not make seasonal loans, ac- cept individual deposits, or lend for nonagricultural purposes. They are not classified as scheduled banks and are regulated by RBI's Agricultural Credit Department. 2.06 The cooperative bank structure is much larger and more diversified. In 1979, 26 State Cooperative Banks (SCB) provided funds for 344 district level Central Cooperative Banks which, in turn, supported about 94,000 Primary Agri- cultural Credit Societies. These societies which specialize in seasonal and medium-term (1 to 5 years) credit 1/ are also designed to provide a variety of services to their members, including savings, inputs supply, staple consumer goods supply and crop marketing facilities. Participation of SCB in long-teTm lending has been insignificant, see para 2.03, and confined to fisheries and 1/ A full description of the State Cooperative Bank structure is given in the Staff Appraisal Report on the Second National Cooperative Development Corporation Project, April 1981 (Report No. 3300-IN).  -7- dairy cattle loans. SCB lack staff trained in preparation, appraisal and supervision of long-term lending schemes and the project would pay specific attention to their training needs (para 4.29), to encourage more term lend- ing from this system. 2.07 The main disadvantage of the split cooperative structure is that farmers are unable to obtain all their credit requirements from a single cooperative organization. A study of the advisability of integrating the two systems, conducted by GOI under the first ARDC Credit Project, recommended a gradual merger of the two banking systems, but the recommendation met with considerable opposition from most LDB. Three States planned pilot scale mergers but have found this difficult and have made little progress. More recently, CRAFICARD re-examined the issues and concluded that merger of the two systems would not be practical at this time. Instead, the committee recom- mended that Primary Agricultural Credit Societies should give agency services to LDB for establishing contacts with potential LDB borrowers, for initial scanning of loan applications, loan disbursements, and for receiving repay- ments, and that such agency assistance should be tried by some selected rela- tively strong and well managed primary societies. GOI has accepted these proposals. However, merger of the two credit systems has not been totally abandoned. Measures such as bringing short- medium- and long-term refinanc- ing together at the national level under NABARD (para 5.07); the expected increase in long-term lending by the State Cooperative banking system (para 2.06); and the encouragement being given to LDB to give production credit to their borrowers, are all designed to draw the two cooperative barking systems closer together to prepare the ground for a possible future merger. C. Land Development Banks 2.08 LDB service about six million accounts. Anyone wishing to borrow from an LDB has to become a member and purchase LDB shares. The main LDB shareholders are members and other cooperatives (70%) and State Governments (29%). Both SLDB and PLDB are managed by boards of directors, partly elected and partly State nominated, but State governments may, and do, suspend the boards, sometimes for political reasons. State governments can effectively control LDB, but degree of control actually exercised by them over LDB man- agement varies widely among States. LDB were pioneers of development lending and their work has brought and continues to bring high benefits to rural areas. 2.09 Total LDB lending has been static and their annual share of ARDC refinance declined from nearly 90% in 1974 to about 40% in 1980 with a corres- ponding rise in the share obtained by CB. The main reason for the stagnation is lending restrictions, imposed by ARDC and RBI because of poor loan recovery, particularly of six large LDB--Bihar, Gujarat, Karnataka, Madhya Pradesh, Maharashtra and Tamil Nadu. These six States collectively account for 60% of all LDB loan collection demand 1/ and 80% of all LDB overdues, yet in 1979/80 1/ Principal and interest falling due during the year plus overdues from previous years.  -8- conducted only 29% of all LDB lending compared to nearly 50% in 1975/76. LDB lending, value of outstanding loans, recovery performance, and lending eli- gibility of LDB branches or PLDB are shown in Annex 1, Tables 1-4. The per- formance of the six LDB listed above is compared to that of all-India LDB in the following table: 1975/76 1976/77 1977/78 1978/79 1979/80 A. LDB Lending (Constant 1970/71 terms, Rs M) /a (i) Six States 563 587 344 310 380 (ii) Other States 607 804 952 903 932 (iii) All-India 1,170 1,391 1,296 1,213 1,312 (i)/(iii) % 48 42 27 26 29 B. Loan Repayment Demand (Current Terms, Rs M) (i) Six States 1,711 1,995 2,385 2,694 2,9L2 (ii) All-lndia 2,667 3,305 3,928 4,518 5,003 (i)/(ii) % 64 60 61 60 '58 C. Loan Repayments Overdues (Current Terms, Rs M) (i) Six States 767 1,073 1,399 1,609 2,076 (ii) All-India 919 1,317 1,703 2,003 2,511 (iii) (i)/(ii) % 83 81 82 80 82 D. Overdues as Percent of Demand (i) Six States 45 54 59 60 71 (ii) Other States 16 19 20 22 21 (iii) All-India 34 40 43 42 51 /a Implicit price deflator of gross domestic capital formation -- Economic Situation and Prospects of India -- Report No. 3401-IN, April 1981. The table shows that sluggish lending by the LDB system is due to declining performance of the six LDB listed above. The table also illustrates the impact of their high overdues on the all-India LDB loan collection efforts. A ranking of LDB by collection records into three groups--LDB with collections better than 60% of demand; LDB with collections at about 50%-60% of demand; and other LDB (Annex 1, Table 5) shows that collection records of several other LDB are also unsatisfactory, notably Orissa, Jammu-Kashmir, Assam and Himachal Pradesh. But their operations and impact on national lending stat- istics are small. 2.10 Causes of Overdues. The high level of overdues is the most critical problem facing LDB. Over a five year period 1975-1980, all-India LDB overdues increased from about 34% of demand to just over 50%, despite rescheduling 1/ 1/ Spreading repayment for one year over remaining installments.  or rephasing 1/ of part of borrowers repayments by some LDB. This deter- ioration was recognized and the overdues problem received special attention during appraisal of the Third ARDC Credit Project (Credit 947-IN). While collection improved in 1980/81 (para 2.14), overall the results of measures for improving loan recovery, including tighter lending eligibility criteria and plans for rehabilitation of LDB with poor collection records introduced as conditions of that operation, have been disappointing to date (paras 2.15 and 2.16). In fact collection has deteriorated markedly during the life of the Third ARDC Project. 2.11 There are many reasons for high overdues, including willful default; natural factors that result in crop failures; poor LDB organization and man- agement that result in inadequate appraisal, loan supervision and collection efforts; and, in some instances, active discouragement of collection efforts by State governments for political reasons. For example, once overdues have built up, particularly following crop failure, political agitation for debt remission grows and, in some States, governments seek to gain political advantage by remitting collection. Once overdues have built up to high levels, restora- tion of satisfactory collection levels (75% or better) becomes a very difficult task. 2.12 Willful default and unfavorable political conditions for loan col- lection are the major factors causing high overdues. Agro-climatic factors are also important. Since LDB lending has been mainly for minor irrigation from groundwater, it is significant that most States with good loan collection records have reliable aquifers, while five of the six problem States have less reliable groundwater resources, which are adversely affected by periodic droughts (Map IBRD 15972R and Annex 1, Table 5). While dugwell investments in the latter areas show satisfactory rates of return, farmers have more dif- ficulty in meeting loan repayments after bad crop years than those in reliable groundwater areas. Rephasing loans after genuine crop failures is therefore important. In addition, to cater for the two or three droughts that can be expected during the life of a minor irrigation investment, ARDC is evaluating alternatives, including intermediate grace periods, for possible amendment of repayment arrangements for dugwell loans in hardrock areas, to prevent build- up of large repayment arrears. 2.13 The main reasons for deterioration in loan collection performance over the 1979-1980 period are severe drought in 1979 and the several national and State elections that occured during that period. Impact of elections ranged from promises of debt remission by some candidates, to a State government ban on legal action against defaulters, to preoccupation of the administration with elections to the detriment of loan recovery action. In addition, write- off of short-term debt by two States against GOI advice and despite strong opposition by RBI caused even greater reluctance to repay loans in some other States. 2.14 There has been considerable improvement in loan collection climate in 1980/81, due to repeated injunctions from highest GOI levels on the impera- tive need for loan recovery and warnings that poor recovery may mean exclusion 1/ Allowing one or more extra year(s) for repayment.  - 10 - from future Bank Group supported lending. As a result, all State governments have withdrawn bans on legal action against defaulters and have organized vigorous loan recovery drives. Most States have introduced special measures to improve loan recovery and, as a result, collection increased by 50%, from Rs 800 M in 1979/80 to Rs 1,200 M in 1980/81 in the six problem States, al-- though this improvement still only resulted in the collection of about 40% of demand. All-India overdues were reduced to 44% 1/ as of June 30, 1981 from just over 50% in the preceding year, but collection results were uneven and there was a disturbing drop in loan recovery in important States with good collection records to date such as Andhra Pradesh, Uttar Pradesh and West Bengal. In addition, some measures introduced by States to encourage recovery are harmful, for instance, overdue interest remission against repay- ment of overdue principal. That measure may boost recovery this year, but discourage future payments in anticipation of similar concessions. 2.15 Past Measures to Improve Collection. To improve loan recovery, lending eligibility criteria were introduced during the First ARDC Credit Project (Credit 540-IN) which, with some modifications, were retained through- out the two successor projects and are still in force. They regulate the amount of new lending of SLDB branches or PLDB according to loan recovery. The present criteria allow unlimited lending if, in the preceding year, recovery was 75% of demand or better; limited lending, based on a sliding scale percentage of previous lending, if recovery was between 50 and 75%; and no lending for lower recoveries. However, some lending for special GOI small farmer programs and for completing ongoing investments is allowed for outlets below the cut-off point. The criteria appeared to work initially, but in the longer term, have not improved collection, although it is conjectural whether collection would have been worse or better had the criteria not been applied. Whatever the impact on recovery performance, their effect on new lending was highly regressive. In Gujarat, for instance, with the longest history of high overdues, new lending fell from Rs 162 M in 1975 to Rs 40 M in 1980. As a consequence, collections at Rs 843 M, were four times as high as new loans Rs 211 M -- over the same period. Excluding Maharashtra which benefited frcom large-scale rephasing and consequent high lending eligibility from 1979/80 onwards, the five remaining LDB with serious overdues problems collected Rs 2.2 billion over the last three years but lent only Rs 1.4 billion. Lending reduced below recovery levels has caused low staff morale and drop in interest revenue. There is also some evidence that loan collection at outlets with no eligibility has decreased rather than improved. Both GOI and the Bank Group therefore agree that revision of existing criteria is necessary (para 2.20). 2.16 To complement the eligibility criteria, any SLDB with overall over- dues of more than 50% or with more than half of its branches or PLDB without lending eligibility, was required to implement a time-bound rehabilitation program satisfactory to ARDC, to qualify for ARDC refinance. Draft plans were drawn up by senior ARDC and RBI management for five States, 2/ but were not 1/ Without rescheduling and rephasing of loans (para 2.10) and blocking of loans (para. 2.16) all-India overdues would have been 54% in 1979/ 80 and 50% in 1980/81. 2/ Bihar, Gujarat, Karnataka, Maharashtra, and Tamil Nadu.  - 11 - formally agreed between ARDC and State governments. Also, the draft plans lacked implementation timetables. The draft plans provided for rephasing of overdues caused by climatic hazards, blocking of chronic overdues 1/ in some States, injection of capital by State governments into LDB, special measures to recover loans from willful defaulters and for management studies. Most States except Bihar have introduced some of the financial provisions of the plans, but loan recovery has not yet improved. Some management studies have been completed, but have proved disappointing. The rehabilitation programs are important, but measures for ensuring prompt implementation and better supervision of implementation are needed and would be introduced under the project (para 2.28). Review of the Six Selected LDB 2/ 2.17 Because of the high proportion of total overdues accounted for by these LDB, a solution of their problems would considerably improve the all- India LDB loan recovery situation. Consequently their operations have been reviewed in some detail and findings are described in Annex 2. A summary of the findings is given below. 2.18 The poor recovery and management performance in the three LDB with weakest results 3/ have already had repercussions on their operations and financial viability. Lending eligibility has decreased seriously, and aging of overdues has worsened, thus increasing the danger of bad debts. The con- cerned LDB have become difficult and unattractive institutions to manage and work for. Their financial viability has deteriorated and led to curbing of administrative costs, often to the detriment of loan appraisal and collec- tion. Their liquidity has suffered, and to meet their obligations, the LDB have been forced to seek State government assistance or resort to expensive short-term borrowing which has further reduced their profits. Although on the basis of the usual accounting information the six LDB under review appear to have maintained financial soundness (debt equity ratio varying from 10:1 to 6:1 and owned capital to assets ratios from 1:11 to 1:7; Annex 2, Tables 2-7), this has been possible mainly due to help of State government share or loan contributions. The ratio between overdues and assets, however, has weakened considerably during the last five years (from an average 1:12 to 1:5). In addition, their assets are probably overvalued, since they contain lending and receivables which will, with the current poor credit discipline, soon become chronic overdues and gradually bad debts. If these financial trends were to continue, annual losses would erode equity capital, while bad debts in relation to equity would increase to a level where continuation of busi- ness would be impossible without outside assistance. This level is already approaching in Gujarat. Consequently, participation of the six major 1/ Chronic overdues are entered into a blocked account and the State gov- ernment pays LDB any shortfall in collecting the dues, generally over a five-year period. While collection from debtors would continue even after five years, the amount outstanding would no longer appear as such in LDB accounts. 2/ Bihar, Gujarat, Karnataka, Madhya Pradesh, Maharashtra and Tamil Nadu. 3/ Bihar, Gujarat, Tamil Nadu.  - 12 - LDB under review in the project needs to be firmly regulated. It should be noted that they accounted for only about 5% of ARDC lending in the last two to three years while not so long ago they were among the strongest LDB in India. The appraisal mission agreed with GOI and State governments that special efforts to rehabilitate these LDB would be justified. 2.19 Participation in Future Lending. Despite the problems that have developed in some cases, LDB remain important rural development institu- tions and both GOI and State governments consider their continued participation essential to their objectives of increasing agricultural production and rural incomes. Most CB, the main alternative lending channel, still operate within a fixed distance of their branches and some of them try to avoid difficult or small farmer lending. However, if future participation of these six LDB in development is to be meaningful, rehabilitation measures more effective than those tried to date are needed. Some of the proposed measures, such as modi- fication of lending eligibility criteria (para. 2.20) and protection of LDB from State actions that endanger efficient operations (para. 2.23), would bene- fit all LDB. Other measures, by way of specific improvement programs would become conditions of participation and would determine the lending level of the six major LDB with serious problems described above. Special improvement programs for the six LDB would be given priority, to allow adequate focus on their effective implementation. Similar programs for smaller LDB with operat- ing problems (para. 2.09) would be designed later, when the priority programs have become well established. Measures to Improve LDB Operations 2.20 Eligibility Criteria - Existing criteria (para 2.15) would be replaced by the following set, regulating lending eligibility of any SLDB branch or PLDB: (a) Loan Collection more than 75% of repayment demand - unlimited lending; (b) Loan Collection 60% - 75% of repayment demand - lending up to highest level in the preceding five years or up to actual repayment collected in the preceding year or average repay- ment collected in the preceding three years, whichever is the higher; and (c) Loan collection less than 60% of repayment demand - lending up to actual repayment collection in the preceding year or aver- age repayment collection in the preceding three years, which- ever is higher. 2.21 State assistance to any outlet by subscription of equity up to 10% of overdues and notional reduction of overdues by that amount would continues. Exemptions to allow lending for Integrated Rural Development Programs or other programs would no longer be available. Exemptions were appropriate when many outlets had lost all lending eligibility, which would no longer be the case in future. In addition, the size of special programs, small in the past, will increase to an extent when exemptions for them would nullify the effect of ithe criteria. The criteria, as well as criteria for SCB and participating Central Cooperative Banks and Primary Agricultural Societies are set out in Schedule E.  - 13 - 2.22 The new LDB criteria were agreed at negotiations. They would be simple and would avoid the regressive effects of present regulations. They would still control lending levels and would provide easily understood incen- tives to improve loan recovery, by directly linking new lending to recovery. To avoid any unforeseen and undesirable effects of applying the criteria, assurances were obtained that a joint GOI/ARDC/ Bank Group review of the operation of the new system would be held by October 31, 1982 and subsequently, as may prove necessary. 2.23 State Intervention. Some measures taken by State governments in the past have caused unfavorable conditions for loan collection (paras 2.11 and 2.13) or have otherwise interfered with LDB operations. Such actions have encouraged willful loan default and are among main reasons for poor collection performance by LDB and other banks. To avoid recurrence of such harmful measures, agreement was reached that all participating banks or their branches would lose eligibility for drawing ARDC refinance, if their State government agrees to the write off of loans or remission or reduction of interest charges; or introduces measures (such as ban on legal action against willful defaulters), if such write offs or other measures would, in the opinion of RBI or ARDC be harmful to effective LDB, SCB, CB or RRB opera- tions. Such disqualification of banks from participation in ARDC refinanced lending would continue until the State government has taken appropriate cor- rective action satisfactory to ARDC (Schedule E). 2.24 Regulation of Rescheduling, Rephasing and Blocking of Accounts. All these measures are useful to alleviate financial distress of borrowers caused by climatic hazards or unavoidable personal circumstances. Timely rescheduling or rephasing should play an important part in preventing build- up of overdues after crop failures. Blocking of accounts, to deal with the hardcore of chronic overdues should be a once-only operation for any LDB. Such measures have been used in the past, but in some cases indiscriminately, to the benefit of willful defaulters. To avoid possible misunderstanding, ARDC would closely monitor the guidelines for rephasing and rescheduling drawn up by RBI, which are satisfactory. Blocking may only be carried out by LDB with express permission from RBI and ARDC. Special Programs and Participation Conditions for Bihar, Gujarat, Karnataka, Madhya Pradesh, Maharashtra and Tamil Nadu LDB 2.25 Operations of the six LDB listed above and particularly those of Bihar, Gujarat and Tamil Nadu have deteriorated considerably. New eligibi- lity criteria (para 2.20) would go some way to ameliorating their operations, but on their own would not be enough. To facilitate meaningful participation of these LDB in the project, would require financial measures and improved management and much closer monitoring of rehabilitation programs by GOI, RBI and ARDC that the Bank Group would also be associated with. The following would form the core of rehabilitation programs, mostly complementary or simi- lar to rehabilitation proposals previously worked out for the six States: 2.26 Financial Measures. - Bihar - Rephasing Rs 82 M of overdues in demand resulting from 1979/80 drought conditions and placing an amount of chronic overdues 1/ agreed by ARDC into a "blocked account" 1/ Generally more than 5 years old.  - 14 - for recovery over five years. Any shortfall in annual collec- tion to this blocked account would be reimbursed by the Government of Bihar to the LDB. (Operation of blocked accounts for chronic overdues in other States would follow similar procedures.) Gujarat - Transfer of part of the Rs 185 M in chronic overdues, acceptable to ARDC and the Bank Group, into a blocked account. Karnataka - Transfer into a blocked account of Rs 54 M, which was the amount of chronic overdues (37% of overdues) as at June 30, 1980. Madhya Pradesh - Acceptance of a financial rehabilitation program satisfactory to the Bank Group. Maharashtra - Transfer of Rs 229 M in overdues, which were rephased in 1979 and which appear difficult to collect, into a blocked account. Tamil Nadu - Transfer of chronic overdues into a blocked account in 1975/76 and rephasing of overdues during 1980/81 are adequate for rehabilitation of the LDB and no additional financial measures are needed. The States concerned have accepted financial rehabilitation proposals for their LDB as described above. However, in some cases, States have proposed to block higher amounts than those outlined above. While the resultant finan- cial injection into LDB resources is welcome, exclusion of the total higher blocked amount from repayment demand would lead to excessive lending eligi- bility for LDB outlets that have been performing badly. ARDC has, therefore, concurred with the Bank Group that no more than the amounts listed above for blocking of chronic overdues would be counted for reducing repayment demand when calculating lending eligibility of outlets of concerned LDB. Execution, of the financial rehabilitation plans by the LDB under review would be a condition of their being able to continue to participate in ARDC refinanced lending (Schedule E). 2.27 LDB Organization and Management. The results of past attempts to improve LDB organization and management have been disappointing because of insufficient analysis of LDB problems and inadequate follow-up arrangements to supervise rehabilitation measures. To overcome these shortcomings, a "Standing Committee", satisfactory to the Bank Group, has been established by RBI to superintend rehabilitation of the six LDB. Assurances were obtained that GOI would maintain this committee which., apart from monitoring financial measures, would, by December 31, 1981, appoint a team for each LDB, which should include management consultants, to study LDB organization and manage- ment status and recommend an improvement program, including an implementation timetable. The terms of reference were discussed at negotiations and found satisfactory to the Bank Group (Annex 7). Early completion of the studies anld implementation of rehabilitation programs would be important for improving operations of the six problem LDB. Assurances were therefore obtained that: (i) GOI would ensure completion of the studies by May 31, 1982; and (ii) that ARDC would, by June 30, 1982, prepare programs with implementation time- tables based on the studies, and satisfactory to the Bank Group, for improv- ing operations of the six relevant LDB. After September 1, 1982 acceptance  - 15 - of the program by each of the six State governments and satisfactory imple- mentation of each LDB concerned, would be conditions of continued participa- tion by any of these LDB in ARDC refinanced lending (Schedule E). Assurances were also obtained that similar programs, both for financial rehabilitation and for improving operations, may be drawn up for other LDB, in due course, by mutual agreement and arrangement between GOI, the Bank Group and ARDC. 2.28 Supervision of Rehabilitation Programs and Lending Eligibility. The Standing Committee would continue to supervise the six LDB during the project period. At fairly frequent intervals, at least three times a year, the Com- mittee would carry out a detailed review of LDB progress, in which the Bank Group would participate. On these occasions LDB lending eligibility would be reviewed. Eligibility would initially be based on the criteria described in para 2.20, but, subject to Bank Group agreement, could be changed during the review, but not to allow more lending than under the criteria. Adjustments, mutually acceptable to the Bank Group and ARDC would be based on progress of the LDB in implementing the rehabilitation program and in loan recovery (Sche- dule E). Assurances that the review process and determination of lending eli- gibility of the six LDB would be implemented as outlined, including Bank Group participation, were obtained at negotiations. D. Commercial Banks 2.29 The amount outstanding in agricultural loans by CB grew from Rs 1,900 M lent to 270,000 accounts by June 1969 to Rs 24,600 M for 1.9 million accounts in June 1979. Of the total, about 25% consisted of "indirect" lending for crop storage, input stocks or rural electrification. The remainder was in respect of direct lending to farmers of which Rs 10,650 M (about 60%) was in the form of medium-and long-term loans and the balance for crop production loans. Annual medium/long-term lending levels rose from Rs 1,280 M in 1975 to Rs 4,350 M in 1979. However, even at that time total agricultural lending represented only about 12% of total CB loan business. 2.30 Expansion of CB agricultural lending owes much to GOI requirements for minimum lending levels by CB to priority sectors including agriculture. The Sixth Five Year Plan target calls for 40% of total CB lending for the priority sectors and at least 40% of that or 16% of total lending to agricul- ture. To support this policy, GOI gave preference in licensing new branches to rural or semiurban branches over urban branches, resulting in an increase from about 5,000 rural/semiurban branches in 1969 to over 25,000 in 1980. For development lending, CB are turning increasingly to ARDC for refinance. They have recognized the advantages of using the preparation and appraisal tech- niques required by ARDC, compared to their former scattered lending 1/ and, in early 1978, RBI instructed CB to apply the ARDC prescribed interest rate for all specified agricultural lending even if financed entirely by their own funds. Since these funds are generally more costly than those provided by ARDC, CB have come to rely increasingly on ARDC refinance as the principal means of meeting the priority sector lending targets set by GOI. In calendar 1/ Individual loans for different purposes over a wide area, which were difficult to supervise, compared to single purpose loan schemes in compact areas prepared by CB and appraised by ARDC.  - 16 - year 1979, 1/ ARDC refinance of CB lending was Rs 1,800 M, about 40% of total CB term lending for agriculture. The rate of CB participation in the project will partly depend on differences between interest rates for agricultural and non-agricultural lending. As this difference increases, CB participation in the project could be expected to increase. 2.31 Commercial banks can be divided into two groups - public and private sector banks. The public sector group is the most important and consists of the State Bank of India with seven affiliates and 20 other nationalized banks and accounted for 90% of agricultural advances in 1980. Members of the State Bank of India group are the most active of the CB in agricultural lending. Though operating only about 25% of the some 25,000 rural and semiurban CB branches, they accounted for almost half the total direct agricultural lending and agricultural term lending by CB. 2.32 Most CB have an agricultural lending department and section, appropriately staffed at headquarters, but the deployment of trained staff to evaluate the technical aspects of loan proposals at regional and branch level is variable. The supply of technical staff has been inadequate and CB claim that they have not been able to find all the staff that they wished to employ. Organization of agricultural lending varies. The State Bank group has estab- lished branches that conduct only agricultural business and give assistance to other branches without technical staff. One or two banks are experiment-- ing with a similar system, but most do agricultural lending through ordinary branches that may or may not have technical staff. The CB agricultural port- folio refinanced by ARDC as at June 1980 was composed of minor irrigation, 41%; farm mechanization, 27%; storage and markets, 13%; plantation/horticul-- ture, 4%; fisheries, 4%; dairy schemes, 3%; and 8% for a miscellany of other agricultural investments. The CB portfolio is more diversified from minor irrigation than that of LDB, but loans still tend to be to somewhat larger borrowers than those reached by cooperatives. 2.33 The main advantage of CB in agricultural lending is ability to provide both production and development loans. A second advantage is greater flexibility than the cooperatives with regard to collateral. While LDB are obliged to take a mortgage or charge on land as security for each loan, CB can use their discretion as to the type of loan security. Also, due to size and nature of operations, CB can offer better terms of service and thus attract well-qualified technical staff. Disadvantages are relative newness to rural operations and urban orientation of CB staff. Another important disadvantage, partially overcome in most States, has been the different lending conditions such as stamp and registration fees payable by borrowers from CB, not payable by LDB borrowers. Finally, CB receive relatively little State government assistance for loan recovery compared with that given to cooperatives. 2.34 As with LDB, CB have found difficulties with loan recoveries. The levels of overdues have been static in recent years or improved only marginally and in 1979 stood at just under 50% (Annex 1, Table 6). In general, CB col- lection records in individual States are similar to those of LDB (Annex 1, Table 5). Up to now, accurate comparison of loan collection performance with LDB or between CB has been difficult because of differing reporting and accounting systems. In April 1979, ARDC, with IDA encouragement, establishedi 1! The calendar year is also the financial year for CB.  - 17 - the Committee on Agricultural Loans through Commercial Banks (CALCOB), to advise ARDC on ways of increasing agricultural lending efficiency. A review of the overdues situation was the Committee's priority task. CALCOB studied overdues of 70 branches of 19 major CB in 17 States and found that willful default due to State interference with collection efforts was a major cause and that inadequate appraisal and follow-up due to insufficient or inexperi- enced staff was contributory. The Committee found that in many instances, branch managers had set very tight repayment schedules and had not rephased or rescheduled loans in cases of crop failure. CALCOB considered, but re- jected the introduction of eligibility criteria similar to those employed for LDB as a means of ensuring better collection performance, in favor of standardization of loan collection accounting and reporting; and in cases of applications for new lending, disclosure to ARDC of operational details of any branch with worse than 50% collection record that is included in new schemes; and adoption of measures prescribed by ARDC to improve branch opera- tions as a condition of new ARDC finance. CALCOB also found staffing levels used for agricultural lending to be varied and recommended studies to estab- lish staffing norms. These are in progress. Main reasons for CALCOB advising against eligibility criteria were: the small proportion of agricultural loans to total CB business; existing insurance cover against losses provided by the Deposit Insurance and Guarantee Corporation; comparative newness to agricul- tural lending of many branches; the difference in size, organization, busi- ness and area covered by CB branches compared to LDB outlets and the need to maintain credit flows, particularly in areas where LDB eligibility is low. 2.35 GOI, RBI and ARDC accepted the CALCOB recommendations that eligibi- lity criteria should not be adopted by CB automatically. They also endorsed CALCOB recommendations for operational improvements, which are now being im- plemented. In particular GOI is concerned that CB might welcome an excuse to reduce agricultural lending, and such reduction could slow the pace of rural development. The appraisal mission feels that the time is not opportune for immediate introduction of lending eligibility criteria for CB. However, to ensure uniform credit discipline between institutions, similar criteria to those for LDB should be applied to CB and RRB as soon as possible. Prepara- tions should, therefore, be made for early introduction of such criteria for CB and RRB. Consequently, assurances were obtained that CALCOB would carry out a study to identify any modifications of LDB criteria, acceptable to the Bank Group, that might be considered advisable and draw up measures for intro- duction of the resultant criteria. The study should be completed by December 31, 1982 at the latest, to allow adequate time for introduction of criteria by July 1, 1983 (Schedule E). As an interim measure, before refinancing any scheme by any CB or RRB branch with over 50% overdues, ARDC and the concerned bank would agree on a program to improve that branch's operations (Schedule E). ARDC has already prescribed a standard format for reporting loan repay- ment data, which would be followed by CB and RRB as from June 30, 1982. E. Regional Rural Banks 2.36 Regional Rural Banks (RRB) as yet play a very minor part in ARDC refinancing. Up to June 30, 1980 ARDC had sanctioned 122 lending schemes for RRB and disbursements amounted to Rs 116 M, only about 0.7% of ARDC total disbursements. However, their participation is increasing. For instance, disbursement of ARDC refinanced term lending grew from Rs 30 M in 1978/79 to Rs 90 M in 1979/80, about 2% of ARDC disbursements in that year.  - 18 - 2.37 First established in 1975, there were 73 RRB by June 1980, operat- ing through 2,678 branches. RRB were specifically established to lend to the weakest section of the population, those that do not appear to have access to CB or cooperatives. RRB provide all banking services. They are sponsored by a CB that provides initial management and together with GOI and State gov- ernment contributes to share capital. The objective is to employ and train local staff that know the area well and can truly become "rural bankers". Employment of local staff is intended to keep overheads much lower than those that CB and perhaps even cooperatives can manage. The experience of the last six years has encouraged GOI to expand the RRB system. As it grows, it may become a major channel for development loans to small farmers. As already indicated (para 2.35), lending eligibility criteria, interim branch partici- pation conditions, and loan collection reporting format that would become applicable to CB, would also apply to RRB. III. PERFORMANCE OF BANK GROUP FINANCED AGRICULTURAL CREDIT PROJECTS IN INDIA Introduction 3.01 ARDC's association with the Bank Group dates to the formulation of ten State-based agricultural credit projects between 1969 and 1974. They were implemented through ARDC and are fully disbursed. In the next phase of Bank Group/ARDC association, ARDC was entrusted with three general lines of credit, ARDC I (August 1975 to July 1977); ARDC II (August 1975 to December 1979) and ARDC III which is under implementation. In addition, ARDC has taken an active part in formulation and implementation of another 25 IDA/IBRD pro- jects, by channeling long-term finance to support investments in irrigation including major irrigation, tree crop development, marketing, horticultural processing, inland and marine fisheries, seed production and sericulture (Annex 1, Table 7). The amount committed by IDA/IBRD for those 38 projects totals US$1,991 M. On June 30, 1981 total cumulative IDA/IBRD disbursements to GOI for onlending through ARDC were about US$859 M. Results of Completed State and ARDC Credit Projects. 3.02 Project completion reports and audits by the Operations Evaluation Department have been prepared for the ten State-based credit projects that were financed by IDA, and for the ARDC I and ARDC II credit projects. In addition, the Operations Evaluation Department has reviewed the results of the nine State-based projects and the first ARDC Credit Project, which were all completed between 1975 and 1977. 1/ 3.03 The State-based projects were designed mainly to support lending programs by LDB and CB to farmers for investments in minor irrigation (80% of total program), land leveling, and, in six States, farm mechanization. Most of the projects required the State government concerned to establish or strengthen an agency responsible for monitoring the scale and quality of groundwater investments financed through institutional credit. Each project required participating banks to adopt specific loan appraisal procedures, 1/ Report No. 3415 Agricultural Credit Projects. A Review of Recent Experience in India, April 8, 1981.  - 19 - based on incremental income calculations, to observe uniform lending terms, and for several, to implement improvements in their staffing and management methods. 3.04 Physical targets for minor irrigation components were achieved or surpassed. Farm mechanization targets were also reached in most cases, but land development loans were hampered by organizational or legal difficulties that were solved only after completion of the projects. The main project objectives, to bring about increased agricultural production and improve rural incomes through expansion of minor irrigation and use of modern tech- nology were clearly achieved. The projects stimulated CB to carry out rural lending. Some LDB (Punjab, Haryana, Andhra Pradesh and Uttar Pradesh in particular) performed well. However, the long-term credit system in several other States was adversely affected by the high and persistent overdues dis- cussed earlier. The projects brought about a change from security-oriented lending to appraisals based on incremental income. Groundwater resource evaluation improved in all project States. However, spacing and density criteria used for siting minor irrigation investments were not applicable to privately-financed investments and were therefore only partly effective in regulating groundwater abstraction. 3.05 With the improving capability of ARDC to appraise and supervise lending schemes and with growing experience of participating banks in devel- opment lending, IDA agreed to change the lending pattern to a general line of credit to support long-term lending approved by ARDC throughout India. This was expected to bring better coverage of the whole nation and allow more flexibility in the type of lending, since ARDC can appraise schemes that would be too small for IDA to approve individually. The First ARDC Credit Project for disbursement of US$75 M over two years (US$69 M for minor irrigation, US$5 M for diversified lending and US$1 M for training) was completed three months ahead of schedule. The Second ARDC Credit Project for disbursement of US$200 M (US$175 M for minor irrigation, US$24 M for diversified lending and US$1 M for training) was completed on schedule. Implementation of the Third ARDC Credit Project for US$250 M (US$207.2 M for minor irrigation, US$41.5 M for diversified lending and US$1.3 M for training and equipment) is ahead of schedule. IDA disbursements on June 30, 1981 were US$185 M compared to appraisal forecast of US$155 M. While the projects supported the total ARDC refinancing programs, IDA, for administrative reasons, did not disburse against all components. 3.06 Project results confirm the major contributions towards increased agricultural production and improved rural incomes projected at appraisal. Physical targets were achieved but the trend towards diversified lending for plantation/horticulture, fisheries, forestry, animal husbandry, farm mechanization and a number of other items has been faster than anticipated. CB lending has grown more quickly than expected but LDB participation has stagnated due to mounting overdues, as described in Chapter 2 of this report. A notable feature of the projects was successful direction of lending towards less developed areas and to small farmers. Instead of the 25% lending to less developed areas targeted, over 40% was achieved and from a base of about 25%-30% lending to small farmers prior to ARDC projects, about 50% of the lending of the Second ARDC Credit Project went to small farmers, ARDC continued to expand staff and services particularly its regional and tech- nical services.  - 20 - 3.07 The Operation Evaluations Department review confirmed the important contribution of the credit projects to agricultural development in India. The review found that while, due mainly to methodological shortcomings, project benefits may have been somewhat overestimated, "there is sufficient evidence to conclude that this set of projects was very timely, generally received strong institutional support and achieved an impressive measure of success... The review concluded that the change from State-based to ARDC line of credit projects was timely and justified, despite IDA's loss of influence on State- based institutions. This feature has been taken into account in the design of this project. The review suggested that IDA funding may have replaced some resources that could have been raised locally and that may consequently have been used for less useful purposes. However, GOI and ARDC are convinced that without IDA funding and the other international funding that was attracted as a consequence of the success of the projects, development lending would not have expanded so quickly. 3.08 The review also drew attention to the continuing controversy over benefits of farm mechanization. In general, the review agreed that there are areas in north India where, due to rapid development and increasing labor shortage, farm mechanization, particularly for tillage and transport, has become necessary. In other parts of the country, with the exception of sugar- cane areas, tractor use remains controversial. ARDC is trying to ensure that its farm mechanization lending is channeled to areas where it will be clearly beneficial. 3.09 The review also elaborated on tubewell and dugwell spacing and pos- sible inequities caused by monopoly of resources given to early investors by spacing criteria. However, over the last two years, controversy about appro- priateness of well spacing criteria has practically disappeared with replace- ment of such criteria by more intensive groundwater resource evaluation. Re- sultant studies generally find resources less depleted than previously assumed. In addition, in most cases, resources tapped by an investor, do not extend inea- surably beyond his land holding. At appraisal of this project, well spacing or possible deprivation of some farmers from minor irrigation sources were not considered to be issues. 3.10 The interest of other countries in supporting ARDC lending is fur-- ther testimony to the success of the projects. The Second ARDC Credit Project attracted US$145 M equivalent from the UK, Canada and EEC, while the Third ARDC Credit Project is supported by contributions totalling US$180 M from the USA, UK, EEC, Canada, Switzerland, Germany and the Netherlands. Most of the contributors show continuing interest in supporting the Fourth ARDC Credit Project, but at appraisal, the amount of contribution by each donor had not been determined. Impact of Indian Credit Projects on Agricultural Development 3.11 The earlier parts of this report elaborated institutional problems particularly of LDB and proposals for correcting them. However, the institu- tional problems should not be allowed to overshadow the fact that Indian credit projects must count among the most successful agricultural development projects financed by the Bank Group. The bulk of funding has been applied to expanding irrigation. High returns on investments in minor irrigation are due to:  - 21 - - quick maturity of benefits--there is no waiting period after investments are completed before benefits accrue; - management of irrigation by the investor--there are no "tailend" farmers or reliance on a water distribution agency; and - availability of appropriate technology and usually satis- factory input supplies to allow high benefits from irri- gation investments. Investments pose few costs to State or central governments. Operation and maintenance are paid entirely by farmers and capital recovery is also high. 3.12 A review of the 12 IDA-financed credit projects completed up to 1980 indicates that about two million investments were made in tubewells, dugwells and pumpsets that resulted in an estimated two million ha incremental irriga- tion. Based on evaluation studies by ARDC, the extra irrigation has brought about production of about 1.8 M tons additional foodgrain per year valued at about Rs 3.8 billion and giving about 2 million beneficiaries total net incre- ental income of about Rs 2 billion. Additional employment resulting from the projects is estimated at about one million man-years. IDA finance also played some part in diversified lending by ARDC that resulted, for instance, in devel- opment of 50,000 ha of plantation or orchards, financing of about 3,500 motor- ized fishing boats, construction of 5 M tons grain storage and a miscellany of animal husbandry investments. 3.13 The emphasis on directing at least half the lending to small farmers and ARDC's success in increasing lending to underdeveloped States have already been indicated (para 3.06). A detailed evaluation of the Uttar Pradesh Credit Project found that farmers too small to install their own equipment derived high benefits from purchase of water from neighbors and that small farmers in general achieve the same or better crop yields than large farmers. Three major institu- tion building accomplishments of the projects were establishment and expansion of groundwater monitoring agencies, initial involvement and rapid expansion of CB in agricultural lending and the expansion of ARDC itself from small beginnings to a major development banking institution with a large professional staff handling annual lending of about Rs 5 billion (US$625 M). The one failure to date relates to mounting overdues and to regression of several formerly strong LDB. As outlined in Chapter 2, this project would pay particular attention to finding and implementing appropriate solutions. IV. THE PROJECT A. Project Genesis 4.01 The first ARDC Credit Project in 1975 was initiated with the under- standing that if successful, repeater projects would be considered by the Bank Group. Since then, two more ARDC credit projects have been approved by IDA. The first two credit projects have been successfully completed while the Third ARDC Credit Project is under implementation. In expectation of its completion on or before the Closing Date (June 30, 1982), GOI has submitted a request for a Fourth Credit Project, to the Bank Group. This project is based on a report "Fourth ARDC Credit Project" (July 1980), prepared by ARDC, and updated in May 1981, and on the findings of a May 1981 appraisal mission.  - 22 - B. Brief Description and Objectives 4.02 The project, in continuation of predecessor projects would, over a two-year period--1982 and 1983--help to finance ARDC lending to borrowers through participating credit institutions (CB, LDB, SCB, and RRB), for minor irrigation and diversified lending. The project would also continue to help finance training programs organized by ARDC mainly for staff of participating banks. 4.03 Main objectives would be to help increase agricultural production, rural incomes and employment; give continuing emphasis to assisting small farmers and less developed areas; continue institution building through further strengthening of ARDC, improving CB agricultural lending and giving particular emphasis to rehabilitation and strengthening of LDB; and continuea programs for improving quality of investments. GOI regards ARDC lending as a key program for its agricultural development plans, and for implementing national policies on Integrated Rural Development, designed to improve the income of the rural poor. 4.04 The project would support the whole of ARDC-s lending operations which cover virtually every district in India. However, for administrative simplicity, the Bank Group would restrict its own disbursements to several major categories including minor irrigation, land development, soil con- servation and land reclamation, plantation and horticulture, livestock, and fisheries. These categories were also financed in earlier projects. Because of increasing emphasis on post harvest activities, the Bank Group would, in this project, also disburse against loans for market and storage construction. Project funds would be channeled through ARDC, which would refinance loans made by participating banks in accordance with specific lending terms and conditions (para 5.12 and Schedule A). Final beneficiaries would be indivi- duals or groups of farmers and fishermen, cooperatives, private and public sector companies, and corporations, with more than 90% of project lending expected to go to individual or small groups of private sector borrowers. As in previous projects, minor irrigation would predominate, representing about 60% of project costs which are estimated at US$2,086 M. ARDC would endeavor to ensure that small farmers receive at least 60% of the Bank Group loan and that about' 45%-50% of project lending would be disbursed to States where agricultural development is lagging. C. Detailed Features Minor Irrigation 4.05 As in the past, the major part of lending would be to individual farmers for construction of dugwells, dug-cum-bore wells, shallow bores and filter points, improvement of existing wells and provision of pumpsets. Costs of various types of wells range from Rs 1,000 for constructing filter points to Rs 16,000 for dugwells in hard rock. Power units would be typically in the range of 3 to 5 hp. Costs of electric powered pumpsets range from Rs 6,000 to Rs 8,500 and of diesel motor powered sets from Rs 6,000 to Rs 10,000. Well construction presents few problems. There are many private contractors with skills, local experience and equipment; in addition, some State government  - 23 - departments or public sector corporations offer custom service to farmers for well construction. 4.06 Suitable equipment, spares and maintenance facilities are plentiful in most areas. However, studies conducted by ARDC, under the Third ARDC Credit project in selected parts of six States 1/, confirmed that equipment purchased by farmers is often substandard. Pumps and power units are frequently mis- matched and pipe work and fittings badly designed. Improvement would result in operating cost savings for farmers and considerable energy savings at State and all-India level. For the study areas, ARDC has published advisory leaflets showing correct type and matching of equipment. Similar investigations, to identify correct specifications for all important minor irrigation areas, are continuing. Several States have specified quality standards to be observed by project borrowers and banks. In view of potentially high cost and energy savings to be achieved, assurances were obtained that by June 30, 1983 ARDC would develop a scheme for minor irrigation equipment improvement in each State in which it supports minor irrigation development. Each scheme plan would indicate the organizations to implement each part of the scheme, studies to be conducted, advisory material to be published and appropriate methods of imple- menting the improvement plan. In drawing up advice on equipment suitability, account would be taken of results of research on pump motor improvement being encouraged by the Rural Electrification Corporation. 4.07 Loans would also be given to groups of farmers for medium-size tube- wells. These serve 10-30 ha and are equipped with diesel or electric powered turbine or electric powered submersible pumps. Costs range from Rs 75,000- Rs 125,000. Private or departmental drilling services and technical standards of installation are satisfactory. 4.08 Loans for larger tubewells would be made Jo public corporations. These would be either deep tubewells with 1I0-250 m /h discharge to irrigate about 100 ha farmland directly or 200-300 m /h augmentation tubewells, pump- ing into surface irrigation channels. Both types would be equipped with either turbine or submersible pumps and piped or lined channel distribution/conveyanc- ing systems. Costs range from Rs 150,000 to Rs 500,000. To ensure as effective operations as possible, appraisal criteria for large tubewells have been drawn up covering power supply, design, and financial criteria, which are listed in Schedule B. Agreement was obtained that these criteria would be applied by ARDC and would form part of the project lending terms and conditions (Schedule A). Drilling and installation services are again considered satisfactory. 4.09 While considerable groundwater resources remain to be developed in most States, high levels of resource utilization are evident in several States, notably Punjab, Haryana and Tamil Nadu. In State-based and First and Second ARDC Credit Projects, well spacing and density criteria were applied as an interim measure to enactment of groundwater legislation. Such legislation has proved socially and politically difficult to introduce and spacing criteria were unsatisfactory because they could not control noncredit investments. Consequently, under the Third ARDC Credit Project, a resource evaluation system was introduced, consisting of three categories of areas based on the level of development, which relates net draft to recoverable recharge. Each corresponds 1/ Andhra Pradesh, Bihar, Maharashtra, Punjab, Rajasthan, and Uttar Prade