India - Drought Prone Area Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 3984 PROJECT PERFORMANCE AUDIT REPORT INDIA - DROUGHT PRONE AREAS PROJECT (CREDIT 526-IN) June 16, 1982 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. ABBREVIATIONS AI - Artificial Insemination AICRPDA - All-India Coordinated Research Project for Dryland Agriculture AMUL - Anand Milk Union Ltd. AR - Appraisal Report ARDC - Agricultural Refinance and Development Corporation CAD - Command Area Development DCA - Development Credit Agreement DDA - District Development Authority DEA - Department of Economic Affairs DPAP - Drought Prone Areas Program GOI - Government of India HYV - High-yielding Varieties IAS - India Administrative Service IBRD - International Bank for Reconstruction and Development ICAR - Indian Council of Agricultural Research ICRISAT - International Crops Research Institute for the Semi-arid Tropics IDA - International Development Association IIM - Indian Institute of Management LDB - Land Development Bank M&E - Monitoring and Evaluation MFAL - Marginal Farmers and Agricultural Laborers NDDB - National Dairy Development Board OED - Operations Evaluation Department PCR - Project Completion Report ROR - Rate of Return SFDA - Small Farmers Development Agency SRI - Systems Research Institute T&V - Training and Visit System of Agricultural Extension FISCAL YEAR April 1 - March 31 LOCAL TERMS Bajra - Pearl Millet (Pennisetum typhirdes) Jowar - Sorghum (Sorghum vulgare) Kharif - Monsoon (June - October) (rainy season) Rabi - Low rainfall (October - March) (cool season) Nalla - Small water course Panchayat - A locally-elected legislative body FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT INDIA - DROUGHT PRONE AREAS PROJECT (CREDIT 526-IN) TABLE OF CONTENTS Page No. Preface ................................................................. i Basic Data Sheet ....................................................... ii Highlights ............................................................. iii PROJECT PERFORMANCE AUDIT MEMORANDUM............................ I. SUMMARY ....................................1........ II. ISSUES ....................... . . ................. 7 A. Project Benefits and Monitoring and Evaluation .................................... 7 B. DPAP As A Pilot Project ..................... . 10 C. Other Issues ...11 III. CONCLUSIONS ........................................... 12 ATTACHMENT I - Borrower's Comments ................................... 14 ATTACHMENT II - Borrower's Comments .................................... 18 PROJECT COMPLETION REPORT I. Background ... . ................. ................... 21 Introduction ......................... ........... 21 Agriculture in India .................................. 21 Bank Group Operations in India .................. 22 Drought Prone Areas .......................... ..22 II. Project Formulation ...... . ............... ............. 25 Identification and Preparation ....................... 25 Appraisal, Negotiation and Approval ................... 26 Project Description ............................. 27 General ....................................... 27 Location ............................ 28 Project Components ................................ 28 Organization and Management ......... ..............30 III. Implementation ................................ .....31 Start-up ..................................................... 31 Revision ..................................................... 31 Physical Implementation ..................................... 31 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. TABLE OF CONTENTS (Continued) Page No. IV. Technical Performance ................................. 32 General ...................................................... 32 Project Components ........................................... 32 V. Financial Performance ............................ ... . 37 Project Costs and Financing ........................... 37 Disbursement and Expenditures ........................ 38 Internal and Budgetary Control ........................ 39 Accounts, Audit and Disbursement ..................... . 39 Covenants .......... ...............0.................40 VI. Project Impact ........................... ......40 Introduction . ................. ..............40 Direct Benefits ................................... 41 Indirect Benefits ................................. 44 Rates of Return ............................. ...... 44 Financial ........................... ......... 46 Economic ................................ .... .. 47 VII. Institutional Performance .......................... .48 General ........................ ............... 48 The DPAP Central Unit in GOI .... ........... ........ 48 State Level DPAP Authorities ......................... 49 District Development Authorities ................ 49 Technical Assistance ....... . ........................... 50 Monitoring and Evaluation ........... ........ 51 VIII. Bank Performance ................................ ...52 Appraisal .................... ............. . 52 Supervision ....................................... 52 IX. Conclusions ........................................... 53 Tables 1 - Summary of Project Financing ......................... 56 2 - GOI/State Expenditures by District and Component ...... 57 3 - Project Organization and Objectives ................. 58 4 - Summary of Key Indicators ........................... 59 Map - 1 - PROJECT PERFORMANCE AUDIT REPORT INDIA - DROUGHT PRONE AREAS PROJECT (CREDIT 526-IN) PREFACE This is a performance audit of the Drought Prone Areas Project in India, for which Credit 526-IN in the amount of US$35 million was approved in December 1974. The credit assisted in financing a project with an expected total cost of US$102.7 million. The credit was closed on June 30, 1981 after a delay of one year. The final disbursement was made on May 6, 1981. The audit consists of a memorandum prepared by the Operations Evaluation Department (OED) and a Project Completion Report (PCR) dated January 18, 1981. The PCR was prepared by the South Asia Regional Office and based, inter alia, on a visit to the country in September-October 1981 and on PCRs prepared by the Borrower. The audit is based on a review of the PCR, the Appraisal Report (No. 533a-IN) dated November 14, 1974, the President's Report (P-1522-IN) of November 15, 1974, the Credit and Project Agreements of January 24, 1975. The audit is also based on a review of correspondence with the Borrower and memoranda on project issues as contained in relevant Bank files and on interviews with Bank staff who have been associated with the project. On the basis of this abbreviated procedure, the audit finds that the PCR presents a comprehensive account of the project's achievements and shortcomings. The memorandum elaborates on the issues of project monitoring and evaluation, the design of pilot or experimental projects and problems of accounting, auditing and procurement, which are important for future IBRD/IDA lending. A copy of the draft report was sent to the Borrower on March 31, 1982 for comments. Borrower's comments which were received have been taken into account in the PPAR and are reproduced as Attachments I and II.  PROJECT PERFORMANCE AUDIT BASIC DATA SHEET INDIA - DROUGHT PRONE AREAS PROJECT (CREDIT 526-IN) KEY PROJECT DATA Actual Actual As % of Appraisal or Estimated Appraisal Item Estimate Actual Estimate Total Project Cost (US$ million) 102.7 84.3 82 Credit Amount (US$ million) 35.0 35.0 100 Date Board Approval - 12/05/74 - Date Effectiveness 03/31/75 06/09/75 - Date Physical Components to be Completed 03/31/80 06/30/81 123 Proportion Then Completed (%) 82 95 La- Closing Date 06/30/80 06/30/81 Economic Rate of Return (%) 12-37 neg. - 30 - Incremental Crop Production (tons) 58,400 60,370 103 /b No. of Direct Beneficiaries (number) 262,000 281,000 107 /b Employment Created (Million Man-days) 28.4 35.8 126 71 CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS FY75 FY76 FY77 FY78 FY79 FY80 FY81 Appraisal Estimate (US$ million) 2.35 7.70 17.75 24.50 33.25 35.00 - Actual (US$ million) - 1.49 7.40 13.20 19.90 30.00 35.00 Actual as Z of estimate (US$ million) 0 19 47 54 60 86 100 Date of Final Disbursement: 05/06/81 MISSION DATA Month/ No. of Man-days Specializations Performance Types of Item Year Persons. in Field Represented/c Rating/d Trend/A Problems& Reconnaisance 02/73 6 132 a,b,d - - - Pre-apapraisal 05/73 5 10 a,b - - - Appraisal I 10-11/73 175 a,b,c,d - - - Appraisal II 15/74 4 76 a,b,d - - - Sub-total 393 Supervision I 05/75-L 1 20 d - - - Supervision II 06-07/75 2 38 a,b,d 1 2 - Supervision III 11-12/75 3 54 a,d 1 2 - Supervision IV 05-16/76 1 13 a,d 1 2 - Supervision V 10/76 2 26 b,d 1 2 - Supervision VI 04-05/77 3 57 a,b,d 2 2 T Supervision VII 08-09/77 3 54 a,b,d 2 3 M,T Supervision VIII 10-11/78 3 60 a,d 2 2 0,M,T Supervision IX 08-09/79 3 66 a,b,d 2 2 M,F,T Supervision X 10-11/80 2 44 a,d 1 2 M,T Supervision XI 05/81 1 1 a 1 2 M Sub-total 433 Completion 09-10/81 25 a,b 1 2 M TOTAL 851 OTHER PROJECT DATA Borrower Government of India Executing Agency Ministry of Agriculture/K Fiscal Year of Borrower April 1 to March 31 Name of Currency (Abbreviation) Rupees (Rs.) Currency Exchange Rate: Appraisal Year Average US$1.00 = 8.00 Intervening Years Average US$1.00 = 8.39 Completion Year Average US$1.00 = 8.50 Follow-on Project None /a Some command area development not complete - GOI/States will do with own resources. /b See PCR, para. 6.03. /c a = Agriculturist; b - Agricultural Economist; c = Financial Analyst; d = Livestock Specialist. /d 1 = Problem free or minor problems; 2 = Moderate problems; 3 = Major problems. /e 1 = Improving; 2 = Stationary; 3 = Deteriorating. /f F = Financial; M = Managerial; T = Technical; 0 = Other. / Renamed Ministry of Rural Reconstruction.  - iii - PROJECT PERFORMANCE AUDIT REPORT INDIA - DROUGHT PRONE AREAS PROJECT (CREDIT 526-IN) HIGHLIGHTS The overall objective of the project was to stabilize and increase the production from agriculture and related activities in six drought prone districts of Western and Central India. The project, which was approved in December 1974, was a five-year time slice of an ongoing Drought Prone Areas Program (DPAP) in India, which was initiated as the Rural Works Program in 1970/71 and which covered, at that time, 72 districts in 13 states. It was considered as a pilot project through which project components could be tested and successful ones identified for introduction in a much larger area at a later date. The project was expected to be completed in just over five years at a total cost of US$102.7 million of which IDA would finance US$35.0 million. The investment program was complex and wide-ranging, including water re- sources, dry land farming, livestock, forage processing, forestry, fisheries, sericulture, horticulture and technical assistance. The credit became fully disbursed in May 1981, about one year later than envisaged. The reported total project cost at that time was 18% less than estimated at appraisal, but some command area developments had not been completed. These were expected to be completed with GOI/State funds. The main reason for the underrun was lower disbursement for institutional credit than planned. There is some doubt, however, about the actual level of dis- bursement for institutional credit, and thus, about the true amount of the cost underrun. The impact of the project and its success as a pilot project is uncertain. The reason is that a monitoring and evaluation program (M & E) which was to identify especially productive components for replication and to measure the effectiveness of the project to increase production, raise employ- ment, and to reach poor farmers and improve their incomes was not implemented as planned. The failure of the M & E program likely stems from the fact that M & E concepts were evolving in the Bank Group at the time of appraisal which resulted in no specific program being drawn up and all funding being left to the Borrower. Of the 16 major project components only three have been fully successful, eight partly successful and five unsuccessful or were eliminated (PPAM, para. 19). Some estimates based on incomplete data indicate that the project may have reached about 280,000 farm families or about 8 percent more than estimated at appraisal. Similar estimates indicate that the amount of  - iv - employment created exceeded the target by about one-fourth, and that annual food crop production increases resulting from the project were about on target while milk production exceeded the appraisal target by about one-sixth (PCR, paras. 6.04-6.06). The overall rate of return was not estimated at appraisal or at completion. At appraisal, rates of return were estimated for nine of the major components and ranged from 12 to 37%. Due to the lack of definitive data, re-estimated rates of return could be made only for six of these nine components which range from negative (sheep development) to 30% dry farming. Some lessons learned from the project are: (i) complex on-going programs likely are not suitable to serve as pilot projects, (ii) pilot projects should include a program for mid-term review which would use lessons learned up to that point to redirect funds to the most productive uses, (iii) plans should be made at appraisal for pilot projects which would ensure that both Bank/IDA and the Borrower would focus on the learning process and carry appropriate lessons learned on to other projects, and (iv) for pilot projects it is especially vital that monitoring and evaluation programs be fully designed at preparation/appraisal. Other points of special interest are: - compliance with accounting and auditing covenants was not satisfactory (PPAM, para. 31 and PCR, para. 5.12); - Borrower did not request IDA's further participation in the DPAP program because of differences in opinion on rural develop- ment policy (PPAM, para. 30); - technical assistance generally achieved its objectives (PCR, paras. 4.16 and 7.11); and - a large training program was successfully carried out (PCR, para. 4.18).  - 1 - PROJECT PERFORMANCE AUDIT MEMORANDUM INDIA - DROUGHT PRONE AREAS PROJECT (CREDIT 526-IN) I. SUMMARY-' 1. The project was the first Bank Group operation in India with the purpose of specifically addressing the difficult problems of rural development in the country's arid and semi-arid regions. The main objective of the project was to increase and stabilize production from agriculture and related activities in six drought prone districts2/ of Western and Central India. A corollary objective was to test the usefulness of a complex, multi-component, rural development project for improving the productivity of the drought prone areas and, thus, its usefulness to serve as a model for future projects in other drought prone areas. 2. The project represented a five year "time slice" of an ongoing Drought Prone Areas Program (DPAP) in India, which was initiated as the Rural Works Program in 1970/71 and which covered, at that time, 72 districts in 13 states. Districts selected for inclusion in the program were chosen on the basis of low average annual rainfall, the frequency and severity of droughts, and the relatively low proportions of irrigated land to total cultivated area. The project (Credit 526-IN) was expected to be completed in just over five years at a total cost of US$102.7 million. Of this total, Credit 526-IN would finance US$35.0 million; central and state governments, US$37.1 million; and established credit institutions, US$30.6 million. 3. Activities and investments planned under the Drought Prone Areas Project were ambitious and diverse, reflecting particular district resources and needs of the population. They consisted of both public and on-farm development works and programs covering: minor irrigation; command area devel- opment; watershed management; dry land farming; and dairy, sheep, forestry, and range development. In addition, the project included applied research and training programs to strengthen the technical support provided to farmers and programs to improve the flow of agricultural credit to farmers. The project also included investments through cooperatives in calf raising, conversion of sugar cane residues to fodder, processing and storage of cattle fodder for emergencies, and diversification schemes which included fisheries, sericulture 1/ Based on the PCR and other project documents. 2/ Jodhpur and Naguar districts in the state of Rajasthan, Ahmednagar and Solapur districts in Maharasthra, Bijapur district in Karnataka and Anantapur district in Andhra Pradesh. - 2 - and horticulture. Finally, the project included small development schemes which were unidentified at the time of appraisal, but which were within the objectives and standards of DPAP; and funds for surveys, research, and train- ing (the latter outside of IDA financing). 4. Through the above outlined investments and programs, food production (grains, milk, meat and fish) was expected to be increased by about 150,000 tons per year; this increase in food production would be partially supported by a 70,000 ton increase in fodder production. Through the project, too, it was envisaged that the incomes of 225,000 rural families would be improved by increases in production and employment. During implementation, an additional 85,000 man-years of labor was expected to be required and, following full development, permanent employment would be increased by around 20,000 man- years per annum. 5. The DPAP Central Unit within the Department of Cooperatives and Community Development was charged with overall management of the project. However, an interdepartmental coordination committee would be established under the chairmanship of the Secretary of Cooperative and Community Develop- ment to review performance of the various sub-sectors of the program, to ensure consistency of technical norms with those of other programs, especially irrigation, and to advise DPAP officials of development in respective depart- ments which would have a bearing on project implementation. Because this was considered to be a pilot project and a number of components were experimental, a detailed system of monitoring and evaluation was to be set up under the direction of the DPAP Central Unit to assess the impact of the project. 6. The achievements of the project's objectives appear to have been mixed. Due to a lack of adequate monitoring and evaluation, the magnitude of the project's programs and the impacts of these programs are partially impressionistic or conjectural. The Central DPAP unit worked out detailed guidelines for impact evaluation during the early years of the project. It commissioned 32 studies to be done by its own personnel or by state authori- ties. However, little comprehensive information on the ultimate mpact of the project has been obtained (see PCR, paras. 6.01, 6.03 and 9.01).1, 7. According to available information, the planning and development of watersheds have, in general, been piecemeal and cover a wider geographic dispersion of activities than planned. These developments have led to dif- ficulties in assessing the overall benefits, particularly those related to area ecology and to farm production and income. The traditional approach to soil conservation has not been changed as expected, consisting of large bunds which follow field boundaries rather than contours and, thus, are usually ineffective as soil conservation measures.2/ 1/ See Borrower's comments; Attachment I, para. 1. 2/ Additional views on this component are provided by the Borrower in Attachment I, paras. 8 and 9. - 3 - 8. While the total forestation package met the target, progress on reforestation of badly eroded sites of upper catchment areas has been below plans due to unexpected difficulties and costs. These developments have resulted in the selection of re-afforestation sites in areas with less diffi- cult soil conditions.!! Plans to develop village wood lots have not yet been worked out or implemented. New techniques to establish rangelands have been successful but the resulting increases in fodder production to date have been too small to cope with the traditional over-stocking. Thus, other means such as herd and grazing controls are needed in addition to greater fodder produc- tion to restore the ecological balance in rangeland areas. A major accom- plishment in rangeland development has been the testing and introduction on the Deccan Plateau of two new drought-resistant forage legumes that were imported from Australia, which now appear to be capable of greatly increasing feed production on the Deccan Plateau. 9. The dryland farming component has been confined to farmer demonstra- tions combined with farmer extension services based on the new Training and Visit system. Unfortunately, because the technical package to be recommended was not available for the first three years, the overall impact of this component on agricultural production has to date been small. The emphasis has now turned to low cost dryland farming practices that involve small risk for farmers.2! 10. Dairy development has turned out to be the most successful of all components envisaged for the project. Essential milk collection routes and chilling centers have been established and milk produce cooperative societies have been formed in all project districts. However, some technical problems still exist which prevent full attainment of the potential production of the dairy schemes. Among these are low production and low fertility levels, and unreliable artificial insemination services, which have resulted in low and irregular milk production. The proposed calf raising centers were dropped because organizational and technical problems. Sheep growers' societies have been established in all districts, but number about 60 percent of the 270 expected to be established at appraisal. In general, the sheep component has not been successful. Fodder production has remained below expected levels. The 100 ha pasture blocks that were expected to provide the fodder base for the societies could not be developed under the grass establishment technique that was envisaged at appraisal. 11. The minor irrigation development program has met with mixed success. Twenty-three of the planned 29 tanks have been constructed and work on other tanks is at various stages of completion. Investigations have shown, however, that unit costs of tanks are higher than estimated at appraisal, the number of 1/ See Borrower's comments; Attachment I, para. 10. 2/ Additional observations on this component are provided by Borrower in Attachment I, para. 3. beneficiaries per tank has been less than expected, command area development has been slow, water distribution to farmers is uneven, and training of farmers in appropriate irrigation has been disappointing. Even so, IDA agreed to finance additional minor irrigation development costs.-V The number of deep wells drilled in Rajasthan and shallow wells dug in the Deccan plateau were on target. Some problems with the program were encountered, however. On the Deccan plateau wells were dug too close to existing wells. There also was a lack of electricity for pumps, but this has been partially corrected by including an electrification program in the project in Rajasthan. Finally, farmers often lack knowledge of good irrigation management practices which reduce the benefits of the component. 12. The diversification program which included sericulture, inland fisheries, and horticulture achieved some of its goals. Mulberry gardens have been established, but only on irrigated lands. Capacity for the production of fingerlings has been expanded and 41 fishermen societies have been formed. Fruit trees have been distributed to dryland farms. These last two activi- ties, though proceeding well, have benefitted only a limited number of farmers. 13. The research support program envisaged at appraisal has been only partially implemented. Recommendations of the agro-meteorologist to support dryland farming developments have not been implemented by GOI. The work of the pasture research consultant led to the identification and establishment of new drought-resistant forage legumes on the Deccan plateau, which was noted above. The senior agricultural economist in the All India Coordinated Research Project for Dryland Agriculture (AICRPDA) who was hired under the project initiated a number of surveys and studies but a majority of them had not been completed by 19811/. 14. Plans to improve agricultural credit were completed for all dis- tricts, but resulting linkages between district officials and credit institu- tions were much less than envisaged because of established credit institu- tions' skepticism of the financial profitability of farmers who received well development loans, but these institutions did lend for other project activi- ties. The training program envisaged under the project was largely achieved; 118 training courses were designed and conducted for project managers, eco- nomists, foresters animal husbandry and agricultural officers. These courses were attended by over 3,000 persons. 15. At completion, project costs totaled US$84.3 million compared with the US$102.7 million estimated at appraisal. There is, however, some doubt about the actual magnitude of this 18% underrun because it is likely that some 1/ See Borrower's comments, Attachment I, para. 3. 2/ The project economist in the DPAP Central Unit assisted with other studies commission by that Central Unit. - 5 - of the institutional credit that has been provided to project beneficiaries has not been identified and counted in total project cost. The proportion of investments supplied by the Government was on target but institutional credit has been much less than planned at appraisal. In the case of credit institu- tions, the percentage supplied was much less, at just 42% of the level expected (subject, of course, to the comment above). The actual, as a per- centage of the expected, level of lending by the Central and State Governments was about 99%. The large unallocated portion (41%) of the credit, which was to provide for price increases and to finance viable sub-projects not yet identified at appraisal, was used for rural electrification in Rajasthan (see para. 11) and a larger number of irrigation works than planned. The amount of the credit allocated to such works at completion was double the amount planned at appraisal. 16. The total number of families benefitting from the project has been estimated at about 281,000 compared to 262,000 expected at appraisal-/. The amount of short-term employment generated is now estimated to be higher at nearly 36 million man-days compared with about 28 million man-days estimated at appraisal. Information is not available as to the improvements in family incomes that have resulted from the project. Incremental food production is now estimated at about 56,000 tons per annum or just 4 percent short of the appraisal estimate. On the other hand, milk production is estimated to be about 4,000 tons per annum, 5 percent higher than envisaged at appraisal. Finally, incremental annual feed production from pasture development has turned out to be 55 percent of the envisaged level and incremental fish production has fallen far short of appraisal expectations. No overall rate of return to the project was estimated at appraisal or at completion due to the variety of investments and the complexity of measuring their benefits. The estimated rate of return for certain components of the project at appraisal and at completion are as follows: Appraisal Completion_! Sheep Development 17 negative Livestock 37 10 Pasture 12 10 Sericulture 16 ) (not calculated) Horticulture 14 ) (minor component) Tubewell Irrigation 17 20 Tank Irrigation 26 10 Soil Conservation 22 not available Dryland Farming 14 30 1/ This and following estimates are subject to the qualification of para. 6. 2/ Due to the lack of specific data, it was not possible to document the economic rates of return for each project activity. The ones presented are very rough estimates based on the available information. 17. Some of the underachievements of the project might have been. due to deficiencies in project management. There were frequent changes in the administrative staff of DPAP. Some of the administrative staff were not sufficiently instructed about the project procedures and objectives to ensure continuity in the DPAP approach. The DPAP Central Unit in GOI lacked an administrative routine to follow-up on work that was executed with its finan- cial assistance. The Unit also, at times, lacked adequate technical expertise because positions were vacant for long periods. The cooperation of the DPAP Central Unit with the Minister of Agriculture did not develop to the degree that was envisaged. As a result, the major responsibility of developing a technical package fell to the senior technical specialists in the DPAP Central Unit who lacked experience to deal with that task. The performance of the specialists in DPAP Central Unit was further hampered by their low profes- sional grades, which prevented an equal standing in contacts with state directors in line departments. District development personnel lacked experi- ence in managing a multi-disciplinary development task. Because the activi- ties planned under the project often were different from those usually sup- ported by state line departments, line department officers who were seconded to DPAP had a difficult task. The situation was compounded by the fact that the technical staff of the DPAP Central Unit did not have the power to change the approach of state line departments (PCR, para. 7.07). On the other hand, because of its control of GOI's share of development funds, the DPAP Central Unit showed a strong capacity in administering and executing national develop- ment programs. It, however, still lacks an audit procedure for following up on work executed with its financial assistance, apparently due to Indian regulations. In all cases, special DPAP units were established at the state level and they performed their tasks well. Forestry departments performed satisfactorily in all districts due to their strong organization and willing- ness to accept the specialized approach of DPAP activities.l/ 18. For the most part, technical assistance provided by the project achieved its envisaged objectives. Expatriates provided valuable advice on the rangeland improvement, forage seed multiplication, cross-bred calf raising, bagasse treatment, and agrometerology. The monitoring and evaluation program was not designed or implemented as envisaged, (see preceding para. 6). Detailed plans and methodologies for collecting and analyzing project data were not designed or implemented by project authorities. This lack of perfor- mance has resulted in lack of definitive data on the impact of the project, which has been noted in several of the above paragraphs. The absence of definite guidelines and objectives of the monitoring and evaluation component at appraisal is one of the major reason for this deficiency. Its worth noting that no funds for monitoring and evaluation were allocated under the project. 1/ See Borrower's additional observation on project management and staffing; Attachment I, paras. 4 and 5. - 7 - 19. In summary, given the information available, it appears that only a few of the project's many components were successfully implemented, a majority reached part of their implementation goals and others fell short of their targets. The fully successful components are: dairy development, range and pasture development, and technical assistance and training programs. The partially successful set consists of: watershed development, dryland farming, reafforestation, sheep grower societies, minor irrigation (wells), diversifi- cation, agricultural credit, and research. In the unsuccessful or dropped categories are: monitoring and evaluation, calf raising, fodder conservation, agro-meteorology, and command area development. For the estimated quantative levels achieved for these components see PCR table 4 and for their estimated impacts in terms of beneficiaries, employment and production see PCR paras. 6.04 through 6.06. II. ISSUES A. Project Benefits and Monitoring and Evaluation 20. As has been noted in the preceding section, there is a lack of definitive information on project benefits - especially the amounts of in- creased production, income, and employment and this deficiency is due to the fact that the monitoring and evaluation component of the project was not implemented as envisaged (PCR, para. 6.01). The Appraisal Report (AR) states (para. 5.07)"... a detailed system of monitoring and evaluation must be undertaken to assess the impact of the project on the needs of each district as a whole. Furthermore, innovations in the project management must be evaluated to determine the merit of extending the concept to other districts as proposed by GOI." Further, it is stated in the AR (para. 5.10), "An important function of the Central DPAP Unit would be the evaluation of project activities to determine if stated objectives were being realized; for example, whether unemployment was being reduced, if agricultural production was in- creasing and being stabilized, and the numbers and types of farmers benefit- ting under the program. Operating responsbility for this task would be assigned to the project economist of the Unit." While recognizing this need for monitoring and evaluation (M&E), the AR did not present any specifica- tions, terms of reference, or timetable for carrying out this activity. 21. Apparently, little or no attention was paid by IDA to monitoring and evaluation activities until the fifth supervision mission of October 1976 when it was noted that DPAP had engaged the Systems Research Institute (SRI) to design a M&E system for the project. The following supervision mission of April-May 1977 reported that the proposal for M&E which had been submitted by SRI dealt only with monitoring of the project and that the scheme would be meaningless without an adequate evaluation component. This concern was discussed with the borrower and with project officials, but it was not men- tioned in the follow-up letters to the Government. The following mission of August-September 1977 reported that GOI had accepted SRI's proposal for - 8 - project monitoring and that the mission pointed out to GOI the need to measure project impact, and, further, that this work should start immediately. While the supervision report noted that this issue would be raised in a follow-up letter to the Government, it was not. The following mission, over a year later in October-November 1978, noted that progress was being made on monitoring of the physical and financial progress of the project on the basis of the system recommended by SRI, but that there had been little progress on establishing a system for evaluation. This time, however, follow-up letters to the Government stressed that there was virtually no data on project bene- fits. This point was raised mainly in the context of evaluation of innovative project components for continued and future financing rather than project benefit evaluation. The letter went on to suggest that organizations such as the Indian Institute of Management (IIM) might be engaged to assist in this task. 22. About one year later, the next supervision mission of August- September 1979 reported again that the project was still lacking an evaluation system and suggested that responsibility for that system be given to an outside agency such as IIM. The mission follow-up letters also stressed the need for an evaluation system. The next supervision mission over a year later of October-November 1980 reported that it had again stressed to project officials that the experimental value of the project called for a comprehen- sive and reliable evaluation of individual components in order to determine which ones could be suitable for replication. This report went on to outline a detailed program that had been agreed to by project authorities and the mission. The program was designed to provide information needed for evaluat- ing the projects impact. In the mission follow-up letter to high authorities, this problem again was brought to their attention as well as providing them with the agreed program outline. 23. The operational lack of proper evaluation of the project's impact rests with deficiencies in the DPAP Central Unit. This Unit, as noted above, was expected to carry out M&E for the project. But, as noted in PCR, para. 7.05, the DPAP Central Unit lacked technical expertise. A project economist in this unit was to be responsible for project evaluation. It is not clear why he never carried out that task.!l 24. The more fundamental reason for absence of proper project evaluation was the lack of a program being drawn up to carry out this activity at the time of appraisal (see PCR, para. 7.12). At the time this project was approved, the IBRD/IDA program for M&E was still in the development stage and there were no clear views or guidelines on how such activities should be carried out. It was not until March 1977 that guidelines were provided in the Operational Manual, section 3.55. 1/ See Borrower's comments; Attachment I, para. 6. - 9 - 25. The Operations Evaluation Department has undertaken three reviews of "Built-in Project Monitoring and Evaluation"1/. The first report, which was undertaken before OMS 3.55 was issued, notes alternative views regarding the nature of M&E. In the preface of this report, it is stated, "In its Second Annual Review of PPAR Results, OED reported that the performance of several projects had suffered from inadequate attention to project output (as compared with input) by project management, the borrowers and the Bank. This deficiency has been felt to be largely the result of inadequate data on out- put." OED then recommended in the report that the Bank pay special attention to the capacity of the borrowers information gathering and retrieval systems, particularly on project benefits. The OED report goes on to say that the study had difficulties in relating the definition of M&E provided in Bank reports to M&E activities in the six projects reviewed. It further noted that the distinction between monitoring and evaluation is defined in several different ways in Bank reports, i.e., "monitoring and evaluation is variously described in terms of (i) input vs. output measurement, (ii) regular analysis of key criteria vs. occasional in-depth studies, (iii) data collection vs. data analysis, or (iv) assisting project management in its day-to-day opera- tions vs. helping Government with improvements of its long-term development programs." 26. Operations Manual Section 3.55 attempts to clear up this termi- nology and conceptual confusion. It defines monitoring as "...a process consisting of a number of selected summary indicators which, when taken together, provide an adequate index of the project's evolution through execu- tion ("costs") and operation ("benefits") phases." Evaluation, however, is broken down into two components, on-going and ex-post evaluation: "On-going evaluation is a normal part of project management and is primarily designed to analyze and propose solutions to problems of implementation and operations, some of which may have been identified as a result of the monitoring system... the main purpose of ex-post evaluation is to review the experience of a project as a basis for future policy formulation and design." 27. The Second and Third Reviews of M&E by OED, which were completed after OMS 3.55 was issued, recount these definitions, apparently accepting their general usefulness. The Third Review, however, which covered eight irrigation projects, concludes that more clarity of terminology is necessary, particularly in light of different requirements of internal and external M&E and that Bank operations could benefit from better understanding of M&E 1/ OED reports: No. 1758, October 17, 1977; No. 2724, November 5, 1979; and No. 3320, February 1981. A fourth review of monitoring and evalua- tion is in process. The following discussion is meant to relate the M&E problems of this project to the general problem of M&E observed in other Bank/IDA projects. - 10 - concepts by the staff. The audit concurs with this conclusion but believes that perhaps something more is neededl/. 28. What may be needed to ensure that M&E fulfills its objectives with respect to projects is a general terms of reference; one that makes it reason- ably clear (or reasonably so) what is involved and what is required for monitoring and evaluation systems, including examples of working systems. Such general terms of refernce could be augmented by certain details which would meet the needs of each project such as the organization to be responsi- ble, lists of information, and data to be collected and a methodology for its analysis. Any M&E system to be used needs to be in final form (detailed) at the time of negotiations, not let to development after the project is appraised. (See also PCR, para. 7.12). 29. Given the complexity of the DPAP, it was essential that the project have a strong M&E component to manage its progress and measure its benefits; a point that was made in the appraisal report. It is doubtful that without a detailed M&E program, drawn up at appraisal, IDA supervision could have done more than it did to get sound evaluation carried out; other than bringing this deficiency to the attention of the Borrower at an earlier time. Apparently, some kind of a monitoring program was implemented, but information on how it was used in project management is lacking. B. DPAP As A Pilot Project 30. The DPAP Project was presented as a pioneer effort to effect and monitor an investment pattern which contributed to more effective public sec- tor participation in rural economies of backward regions. It was viewed as a pilot project through which successful project components could be identified and then introduced in a much larger area at a later date.2/ 31. As a widely dispersed, multi-faceted project, it seems, at least with hindsight, that DPAP was much too large and complex to usefully serve as a pilot project. The Project, it should be remembered, was a time - and area - slice of an on-going Borrower program augmented with some special 1/ More recently, "A Handbook on Monitoring and Evaluation of Agriculture and Rural Development Projects" was issued by the Bank in November 1981. While this report contains much valuable definitional and procedural information, it is wide-ranging and complex and its guidelines will be difficult to apply. Only time and use will determine how well it meets the present need. OPS wishes to record that the main theme of the Handbook has been to recommend simple pragmatic approaches to monitoring and evaluation, rather than complex surveys of large size; and initial reaction from users in the field has been most supportive. 2/ Presentation to the Board, December 5, 1974. - 11 - technical assistance. As a pilot project, it may have been more practical from IDA's point of view to have supported a smaller project in just one or, at most, two of the six districts where activities and achievements could have been more easily monitored and evaluated. Although supervision missions usually consisted of two to three people, they were only able to visit several of the six districts during each mission. It was a difficult project to supervis2 and the supervision input was high.-' 32. Toward the end of the project, it became evident that the Borrower did not desire, because of a difference of opinion on rural development policy, the continuing involvement of IDA in the Drought Prone Areas Program. Perhaps as a result, IDA came to believe that many of the sub-projects of DPAP could be handled through other projects which focused exclusively on specific components like dairying, extension, soil conservation, etc. IDA did subse- quently help finance such narrowly focused projects (see PCR, para. 9.03). The Drought Prone Areas Project did not provide, however, any useful lessons that guided the design of those projects. With hindsight, it appears that there were no essential linkages or interrelations among the project compo- nents of DPAP which required simultaneous implementation. In other words, the components did not represent a synergistic package, at least from a technical or constraint point of view. Thus, it did not serve as a pilot project to be replicated at a later date as a package. It served mainly to provide funds for an 2-going program that was modified somewhat during the implementation period C. Other Issues (a) Accounting and Auditing 33. According to the Project Agreements (Sec. 3.03) the four states involved in the project were to have their accounts and financial statements for the project (balance sheets, statements of income and expenses, and related statements) for each fiscal year audited, in accordance with sound auditing principles consistently applied, by the accountant general of the states. The states were to furnish to IDA as soon as available, but in any case not later than six months after the end of each fiscal year such certi- fied copies of its financial statements for such year as so audited and the report of such audit by said auditors. This covenant was scarcely complied with by the states involved. The audits that were sent to IDA usually were incomplete and appeared to present the financial statements of the District Development Authorities (DDA) as a whole, not for the transactions related to the project only as was required. The few complete audits that were received (toward the end of the project period) noted a number of irregularities in the 1/ See Borrower's comments on this issue; Attachment I, para. 7. 2/ The Borrower does not fully agree with this view; see Attachment II, para. 2. - 12 - accounting procedures used by the DDA units. Further, these audits were not within the six month grace period allowed for audits, as noted in PCR, para. 5.10. As also noted in PCR, para. 5.16, delays for submitting audits is typical in India.!/ Because of the inadequacy of the formal audits of the district DPAPs, there was no way for IDA to be assured that the credit was being used for designated purposes. Because Credit 526-IN was a complex project, inter alia, IDA should have taken more affirmative action other than reminding the borrower that the auditing covenants were not being complied with. (b) Procurement 34. Contracts to be let or works to be carried under the Credit were to be on the basis of local competitive bidding or force account, except for procurement of well-drilling equipment, DCA schedule 3. However, the audit has not been able to acertain whether well-drilling equipment was contracted for under ICB. Disbursement documents are not useful for making such a determination because only aggregate disbursement requests are received at headquarters. Detailed procurement information is kept by the Borrower and available at any time for IDA staff review. However, IDA staff do not recall the procedures used to obtain well drilling equipment and project files do not mention this item2/. III. CONCLUSIONS AND LESSONS LEARNED 35. The Drought Prone Areas Project represented a time- and area-slice of a huge, on-going Borrower program with supplemented technical assistance Because of its scope and complexity it was not well suited to serve as a pilot project to test the suitability of comprehensive programs to stabilize and increase agricultural production in the drought prone areas of India. 36. The multifaceted program which included some 16 separate, and at most, loosely related components achieved a limited degree of success, according to partial information available. The most successful components were dairy development, range and pasture development, and technical training in management, economics, forestry, and rangeland improvement. These programs 1/ Recently, the Bank has been reviewing this problem and recent appraisal reports have included more specific guidelines and requirements which should significantly reduce such problems in the future. See for example, SA Report 3615, Andra Pradesh Agricultural Extension Project, March 9, 1982. 2/ See also OED Report No. 3557, "An Interim Report on Procurement Issues in Bank-Financed Projects," dated July 15, 1981. - 13 - accounted for about one-fifth of total project cost (PCR table 1). Two other components soil conservation - watershed development, and minor irrigation utilized about forty percent of project funds but made only a limited con- tribution to the project's goals. Minor irrigation received a large share of the unallocated funds, but the soil conservation component also benefitted. Given the lack of success of these two components, the efficiency of these transfers is to be questioned. Alternatively, use of these funds to promote dairying, and range and pasture improvement would have been more productive. Some lessons emerging from the project are: (i) multifaceted, complex on-going programs are not suitable vehicles for pilot projects; (ii) it is vital to design evaluation goals, methodology and organization (monitoring and evaluation) at the preparation/ appraisal stage of projects, but especially vital for pilot projects; (iii) pilot projects should include a program for mid-term review to take advantage of lessons learned up to that point in time in order to cut off unproductive components and redirect available funds; and (iv) special arrangements should be made for pilot projects at appraisal to ensure that IDA and the Borrower focus on the learning experience and carry any lessons learned on to other related projects.  ATTACHMENT I -14 WT: Grams: DRYLAND Borrower's Comments Off.'Phone : 4 9177 Res.Phone:59177 ALL INDIA CO-ORDINATED RESEARCH PROJECT ICA R FOR DRYLAND AGRICULTURE DR. R.P. SINGH PROJECT DIRECTOR Saidabad P.O. =adMPx*miraR HYDERABAD3i@1Wi INDIA* 500659 DO.F.No.IV-3/82.PD May 31, 1982 Dear Shri Kapur, This is in continuation of my letter of even number dated May 6, 1982V/ As desired, I send herewith my comments on "Project Performance Audit Report on India Drought Prone Areas Project (Credit 526-IN)". Kindly acknowledge receipt. With kind regards, Yours sincerely, R.P. Singh) Shri Shiv S. Kapur Director Operations Evaluation Department The World Bank 1818 H Street N.W., Washington, D.C. 20433, U.S.A. Enc: a/a 1/ Refers to letter announcing that comments will be sent. - 15 - qQiENTS ON THE PROJECT PERFORM1,ANCE AUDIT MEMORANDUM, INDIA-DROUGHT PRONE AREAS PROJECT (CREDIT 526-IN) I have gone through the Memorandum (pages 1-39) under reference very carefully. I had a close association with this programme as a member of one of the DDA's, since its very inception. Iy comments are subjoined: 1. On page 3, para 6, it is mentioned that 'the overall objectives of the project appear to have been achieved'. I underline the word appear and leave the rest to the readers to judge for themselves the confidence with which the statement has been made. Lack of adequate monitoring and evaluation of the project has been the main drawback and it is very correct to state that the impacts of these programmes are impressionistic and conjectural. 2. The dryland farming component was not given the importance that it deserved. Whereas the technology was available, the World Bank did not want to take the risk which was involved in the programme. The risk was worth taking, a fact which is substantiated by the data on 'incremental production' set out on page 22. As against the appraisal estimate of 30,000 tons of food crops production, the achievement was 37,000 tons, under dryland programme, proving beyond doubt that there was merit in low cost dryland farming practices that involve small risk for farmers. The soil conservation programmes were over- played at the cost of dryland farming, although both are inseparable. It was just a question of laying emphasis. With traditional approach to soil conservation, consisting of large bunds that follow field boundaries rather than contours, much could have been achieved on la cost dryland farming practices on individual farmers holdings. Timeli- ness and precision (T and P system) are the cardinal principles in dryland farming which promise rich dividends. - 16 - 3. Programmes relating to tank irrigation were not well con7eived and implemented. This is another important programme, where the World Bank has missed the bus. 4. There were too many and frequent changes of Project Officers, which dislocated the programme. Some of the Project Officers were not quite dedicated and they came to make a career and went away. No programme can succeed unless there is a commit- ment and accountability on the part of Project Officers, who are responsible for the implementation of the PrOgrammes. 5. The expertise at the Central DPAP Cell was weak. The cell was not able to give proper guidelines and in time. 6. As rightly mentioned in para 21 on page 13, a 1roject Economist was appointed in each project for project evaluation and it is not clear as to why he never carried out the job assigned to him. The fact remains that the Project Economist had done every thing else except evaluating the benefits of each component of the programme. The Central DPAP Cell should have seen to it that the work assigned to the Project Economists was carried out as per schedule. 7. I agree with the statement made in para 29 on page 16 that DPAP was too large and complex a project to usefully serve as pilot project. The activities were widely dispersed and multifaceted. It would have been appropriate if the IDA would have restricted its support to one or two districts and few activities, so that supervision and monitoring would have been easier and more effective. 8. In the Watershed Development programme, the planning was done all right, but the implementation was faulty, in as much as it was in piecemeal and highly diffused. It is doubtful, therefore, that the benefits, if any, that may have accrued in terms of improvement of ecology, agricultural production and farmers' incomes were properly evaluated. - 17 - 9. It ias been experienced in the past that soil conserva- tion ,rogrammes in this country did not succeed because of poor or no maintenance of soil conservation structures. ,e have, therefore, to rely more on in-field practices such as cross-slope cultivation, sowing on grade with ridging later on or broad beds and furrows, etc. 10. The Forestry programme, by and large, has been quite successful. 11. Range land piogramme, as envisaged in the project apprai- sal, did not deliver the goods because controlled grazing was not ensured. 12. On the Deccan Plateau, Stylosanthes hamata, S. 1cabra have established their creditability on degraded, marginal and sub-marginal lands. 13. I am in agreement with the observations made in para 4.08 on page 14. 14. The observations made in para 4.16 on page 16 regarding the research support are correct. I am in agreement with the same. 15. A mention has been made with regard to training in para 4.18 on page 17. There is need for more training courses on dryland farming, pasture development, agro-forestry etc. I hope my comments will be taken in the right spirit in which they have been made. - 18 - ATTACHMENT II Borrower's Conments No.DPA.1082/322/50/D-41. - GVER1NMET OF MAHARASHrRA, Planning Department, Mantralaya, Bombay-400032. Dated the 449th May,1982. To, Shri Shiv S.Kapur, Director, Operations Evaluation Department, World Bank, 1818 H.Street, N.W., Washington,-D.;C.20433, U.S.A. Subject: Project Performance Audit Report on India Drought Prone Areas Project iCr44it 52-IN)., Dear Sir, Thank you for your letter No.Nil dated the 31st March,1982 enclosing a copy of the draft Project Performance Audit Report, received here in Bombay on 15th April,1982. We have gone through the draft of the Report. Since the,DPA Project covered Six Districts in four States of India and since the draft has been based on an overall review of the Drought Prone Areas Programme in various States our comnents are related to Ahmednagar and Solapur only. These comments are: On page 17, para 30, it has been mentioned that the Project served mainly to provide funds for on-going programme. We feel that this is not entirely true in as much as the prJect was responsible for introducing4new approach to the on-going developmental programmes with their integration on a watershed basis; introduction of dryland technology in the Project areas. New drought-resistant grasses, hamata and scalra imported from Australia, have done a lot to the area and it would have taken many more years for these legumes to reach the drought prone areas perhaps in the absence of Drought Prone Areas Programme assisted b7 I.D.A. This is not a small achievement. Like-wise, there may not be the desired level of integration of all the schemes on a watershed basis, but a beginning has been made through the D.P.A.P. It is seen that the P.C.Report has been finalised at the Bank's level. We do not know for sure whether comments on the P.C.R. are also sought. We however furnish the following comments, which may be considered if the PRC has not yet been finalised. i) Para 4.16: It is not fair to say that the State authorities (Maharashtra) did not go ahead with the bagasse treatment plant because they had little interest in the establi- shment of such a plant. The correct position is that the sugar factories were unable to assure zupply of adequate begasse for treatment plant as it is the cheapest fuel available with them. The Scheme had therefore to be dropped. ii) Para 5.08 : As regards formulation of annual DPAP work programmes, the annual programmes were also examined at the  - 19 - State level keeping in view the 6PA Project. At the District level, the project authorities formulated these programmes and submitted them to the State Government for sanction. (D.T.Jose h ) Copy forwarded to:- 1) Shri D.S.Mehra, Assistant Commissioneid, Drought Prone Areas Programme, Ministry of Rural Development, New Delhi. 2) The Chief Executive Officer, Zilla Parishad, Ahmednagar/Solapur.  - 20 - INDIA DROUGHT PRONE AREA PROJECT (Cr. 526-IN) PROJECT COMPLETION REPORT January 18, 1982 South Asia Project Department Education and Agricultural Institutions Division  - 21 - DROUGHT PRONE AREAS PROJECT (Credit 526-IN) PROJECT COMPLETION REPORT I. BACKGROUND Introduction 1.01 Traditional rainfed agriculture in India is greatly affected by the erratic nature of the monsoons. In some regions, rain failure is frequent and necessitates Government-approved ad-hoc relief operations. The Drought Prone Area Program (DPAP) attempts to mitigate the effects of drought through integrated rural development activities which include: improved crop, land and livestock management, minor irrigation works, additional research and extension activities, and agricultural diver- sification programs. 1.02 The Drought Prone Areas Project, Credit 526-IN, of US$35M, represented the first Bank Group operation in India specifically designed to tackle the difficult problems affecting rural development in drought prone situations. Effective on June 9, 1975, it was administered as part of Government of India's (GOI) own national program. Project implementa- tion was completed on June 30, 1981. 1.03 This report on the Drought Prone Areas Project is based on the findings of Bank supervision missions and discussions held with GOI during a project completion mission consisting of Messrs. J. H. Lindt, H.H. Groenewold (FAO/CP) and 0. Espadas (EDI). Agriculture in India 1.04 India is a large and diverse country with a population of 688M (in mid-1981) and an annual per capita income of US$190. Agriculture continues to dominate India's economy, employing over two-thirds of the labor force. The share of agriculture in GDP at factor cost (1970/71 prices) has declined from 60% to about 40%. The share of industry has increased over the same period from 15% to 24%, but the government and private sector efforts at industrialization have not been able to absorb the growing labor force, nor to bring about the economic transformation needed for higher productivity and to cope with the demands posed by rapid urbanization. 1.05 Economic growth has averaged about 3.5% per annum over the past 30 years. Slow growth of value added in agriculture-- 2.1% per annum-- - 22 - over the last 3 decades has constrained overall growth during this period, not only because of the high share of agriculture in GDP, but also because scarce foreign exchange has been often required to import food. 1.06 In agriculture, the positive results of large investments and appropriate policies in the past years are starting to bear fruit. Investments have been made to increase the rate of expansion in irrigation from 1.3 M ha per year in the early 1970s to about 2.3 m ha in 1978/79. Fertilizer use reached 5.6 14 tons of nutrients in 1980/81. Foodgrain production grew at about 2.75% per annum-- sufficient to meet consumer demand-- to eliminate imports (which had averaged nearly five million tons per year for the 15 years preceeding 1976), and to lower real foodgrain prices for consumers. At the same time, India was able to build up sub- stantial foodgrain buffer stocks which made it possible to weather the effects of the 1979/80 drought with comparative ease, and to export about half-a-million tons of grain in 1980. While the management of the foodgrain economy after the drought was a significant achievement, the effect of the drought on production re-emphasized that India needs to pay continued attention to dryland agriculture. Bank Group Operations in India 1.07 Since 1949, the Bank Group has made 61 loans and 141 development credits to India totaling US$2,833M and US$9,323M (both net of cancella- tion), respectively. In recent years, Bank Group lending has emphasized agriculture. A significant part of the Bank Group support has gone to major and medium irrigation development, and during recent years, increas- ing attention has been given to crop husbandry as a means of improving the efficiency of the use of irrigation water. The Bank Group has also been active in supporting minor irrigation and other on-farm investments through agricultural credit operations. Marketing, seed development, agricultural extension, and dairying are other activities that have been supported by the Bank Group which have been additionally active in financ- ing the expansion of the fertilizer sector. In terms of overall food production, the new technology of fostering the widespread use of high yielding varieties (HYV), more irrigation and increased availability and use of improved inputs has been successful, but shortcomings remain in respect of the equitable and appropriate distribution of the benefits of this technology. Agricultural development and progress in dryland farm- ing, particularly in drought prone areas, has lagged. The DPAP helped to address these shortcomings. Drought Prone Areas 1.08 In 1970, GOI identified fifty-four (later expanded to 74) dis- tricts as being, in whole or in part, drought prone, i.e., areas of highly uncertain rainfall. In 1970-71, GOI introduced the Rural Works Program, later called Drought Prone Areas Program for these districts with the - 23 - objective of creating rural employment through government-executed con- struction of productive assets such as irrigation tanks, conservation works and roads. The urgent need for relief employment during the first three years of the program, especially in the wake of the 1972 monsoon failure, led state governments to place emphasis on employment generation in the design of program components and to play down productive objec- tives. 1.09 India's drought prone areas program has evolved from an employment-oriented program to one which is now intended to reduce the impact of drought on farmer production. To mitigate the effects of future drought on agricultural production and employment, investments are made in all sectors of the rural economies of program districts. However, the program places emphasis on investments to benefit the poorer sections of rural populations and to assist in this, GOI included a range of subsidy schemes. 1.10 Locatign and Population. The drought prone areas of India cover about 600,000 km (IBRD Map 11077PCR) and a population of about 66 mil- lion. The bulk of the population is engaged in a perennial struggle to meet subsistence needs in a harsh environment. 1.11 The 74 districts selected by GOT for DPAP assistance are located throughout the country. Their greatest concentration is in the western and central zones. The total area of drought prone districts represents 19% of the Indian land mass, and their population, about 12% o India's total, resulting in a population density of 112 persons per km compared with an all-India average of 178. The population of a typical drought prone district ranges from 1.2M to 2.OM. Nomadism adds further to the instability of the agricultural economy, and the introduction of measures such as controlled grazing to improve rangeland productivity is exceed- ingly difficult under these circumstances. 1.12 Rainfall and Soils. Most drought prone areas are located in regions with less than 750 mm rainfall, although a few districts are in the hot sub-humid regions with 750-1,125 mm of rain per year. Rainfall is extremely variable and so-called normal rainfall seldom occurs. There is a 50% probability that less than 20 mm rainfall per week will occur (less than the potential evapotranspiration), and only a 20% probability that there will be more than 50 mm of rain per week during the monsoon. Precipitation often occurs at high rates of intensity with little impact on soil moisture availability and with hazards of soil erosion. 1.13 Three major soil types are present in DPAP districts: the sandy soils of northwest India, the red alfisols and the black cotton vertisols. The sandy soils are poor in plant nutrients and have little moisture st--ge ranacitv: on the other hand, Soil erosion is not a major problem as most areas of th soil type are alluv*al and r.-'atively flat. The red - 24 - alfisols are also deficient in soil nutrients and moisture retention is often limited due to their shallowness. During recent decades, more and more alfisols, which are at best only marginally suitable for crop farm- ing, have been brought under cultivation and this has resulted in serious soil erosion. Large tracts of such lands on slopes have now been aban- doned, leaving barren outcrops which provide very limited grazing for village livestock. This forms the typical picture today of many DPAP districts on the Deccan plateau, where about 40% of the surface area is such outcrops with sparse grass and shrub vegetation. The black vertisols of DPAP districts are typically in the lower catchment areas. They are potentially fertile and have a high moisture storage capacity, but are easily erodible. In all areas, traditional cultivation practices prevail. The natural vegetation of non-arable lands has been denuded due to over- grazing and fuelwood cutting, and much of the rainfall that does occur is lost as runoff. The end result is erosion and a continuous degradation of physical resources, resulting in poor groundwater recharge and reduced production potential. 1.14 Agriculture and Livestock. The agricultural systems of drought prone areas are characteristically subsistence-oriented, producing essen- tial food and fodder for local consumption. However, there are some notable exceptions where cash crops such as groundnut and cotton are important to the local economy. Average holding sizes are about 10 ha on the sandy soils of the northwest and about 4 ha on the Deccan plateau. In an attempt to offset low yields, farmers have expanded cultivation into land of poor potential. This has aggravated wind and water erosion problems, with the result that the overall ability of the land to sustain human and animal populations is declining. Pearl millet (bajra) is the staple food in drier areas, while sorghum (jowar) and millets (other than bajra) are grown in areas with more assured rainfall. All soils with low water retention capacity, i.e., the sandy soils and the shallow alfisols can be used only for monsoon (kharif) season crops. Black vertisols are used in many areas for rabi crops planted in residual moisture, i.e., planting starts only when farmers decide that the moisture stored in the soil is sufficient to meet the needs of the crops selected. Cattle is the main source of farm power. They are also a major source of cash income required to the bridge gap when crop production does not meet subsistence needs. To own considerable numbers of livestock is, in fact, an important strategy for survival. To meet stock feed requirements, farmers grow crops with high fodder yields. In spite of this, a large proportion of the feed requirements must come from the non-arable lands. Due to the extremely low levels of income and its uncertainty, an attitude of exploitation of resources prevails and in the absence of effective controls, serious overgrazing and degradation is common. 1.15 Institutional Support to Agriculture. All line departments operate routine field programs in DPAP districts. However, as the produc- tion potential of these districts is limited, departmental presence is - 25 - sometimes less than in areas with higher agricultural productivity. The soil conservation and forestry departments maintain extensive bunding and reafforestation programs in conjunction with drought relief operations. Credit institutions are only weakly represented and the role of private moneylenders is still important. 1.16 The DPAP districts include two types of district administration. The states of Rajasthan and Andhra Pradesh depend traditionally on administration through strong offices of Collectors. Whereas in Maharashtra and Karnataka, administrative and development responsibilities have been increasingly delegated to locally-elected bodies known as Dis- trict Panchayats. 1.17 Research support for dryland farming started some 50 years ago and led to the development of technical packages for dryland farming. However, the estimated benefits of these packages, i.e., 15% to 20% yield increases on a low base of 200 to 400 kg of grain per hectare, have been insufficient to capture farmer interest. Research in dry farming has been intensified since 1970 and the All-India Coordinated Research Project for Dryland Agriculture (AICRPDA) plays a major role in this. Liaison with ICRISAT (International Crops Research Institute for the Semi-Arid Tropics) is close and research programs now give particular attention to crop and land management methods for moisture and soil conservation, and which can be handled by farmers. II. PROJECT FORMULATION Identification and Preparation 2.01 During the visit of a high level World Bank delegation to India in 1972, G01 elicited the Bank's interest in its Drought Prone Areas Program. It was agreed that the Government employment schemes for drought relief, which were often designed on an ad-hoc basis, would be replaced by a package of activities aimed at: (i) reducing the impact of drought on agriculture, or "drought- proofing"; and (ii) improving the agricultural resource base and productivity in the areas concerned. 2.02 Accordingly, a World Bank reconnaissance mission visited India in February 1973. It had at its disposal, the DPAP proposals made by various states for the Fifth Development Plan, and it traveled widely in drought prone areas to delineate the concepts of potential IDA assistance. Fol- low-up discussions with OI took place in May, and a Back-to-Office and Full Report was issued on June 11, 1973. It was agreed that IDA assist- - 26 - ance should be directed to districts representative of the hot arid and semi-arid zones, and would concentrate on economically viable activities for improving the physical resources in a watershed. This included atten- tion to water conservation and development, to dryland farming and the rehabilitation of non-arable lands. It was agreed that details of the extension, dairy and sheep development components would be formulated in liaison with other appraisal teams finalizing projects for World Bank financing at that time. It was also agreed that activities for the development of general infrastructure, i.e., rural electrification, con- struction of roads or village water supplies would not be part of the proposed project, but continue to be carried out under separate programs available for that purpose. 2.03 Further preparatory work was then carried out by GO and four state authorities, with some assistance provided by the Bank's resident mission. This material was made available to IDA in the fall of 1973. Appraisal, Negotiations and Approval 2.04 A pre-appraisal mission was fielded in October 1973. Agreement was reached that the project would be implemented in six districts repre- sentative of the semi-arid and arid zones. The mission established a package of activities to be implemented in catchment areas. The main points of discussion concerned: the approach to soil and water conserva- tion works; the formulation of dryland farming practices and site selec- tion and implementation of minor irrigation works. The mission endeavored to simplify the project by eliminating a number of activities such as assistance to rural artisans. However, even after this, the agreed project components concerned with production support and resource develop- ment still remained wide-ranging. In the absence of reliable data con- cerning improved dryland farming techniques, and considering the acute fertilizer shortage following the 1973/74 oil crisis, the pre-appraisal mission did not propose a dryland farming component. 2.05 A final appraisal mission was fielded in May 1974. On the techni- cal side, discussions centered around such points as design criteria for minor irrigation works, methods of rangeland improvement, possibilities of introducing controlled grazing on improved rangeland, and the role of crossbred cattle for milk production in drought prone areas. GOI exerted considerable pressure for the continuation of traditional approaches to soil conservation and animal husbandry previously adopted by GOI or state governments. These approaches were accepted by the mission, although as discussed in para. 4.04 and 4.12, this proved to be a mistake. Also, on the strong urging of GOI, a pilot program of dryland farming was included and this subsequently proved to be successful. The mission also worked out details concerning project administration and financing, and agreed with GOI on technical assistance. - 27 - 2.06 The Credit was negotiated in October 1974, approved in December 1974 and became effective June 9, 1975. Reconnaissance through Board presentation took 23 months. Project Description General 2.07 The Drought Prone Areas Project was designed as a multi-component pilot project to test and implement technical, economic and organizational means to improve the productivity of drought prone areas in India. Each component, in support of agricultural development, was to be financially and economically viable, but the need of the weaker sections of rural populations for government-supported subsidy schemes was recognized. The project was to provide a "shelf of works" in order to use drought relief operations for the creation of productive assets. 2.08 The project's major thrust was to be on stabilizing and increasing the income of families who were dependent on dryland agriculture and livestock keeping. 2.09 The watershed was taken as the basis for planning land-use systems under rainfed conditions. Soil and water conservation works, together with reforestation and rangeland improvement were to be carried out on public lands in catchment areas unsuitable for farming. These actions were designed to reduce soil erosion, recharge groundwater, increase feed for livestock and fuelwood supply for villagers. Soil and water conserva- tion works were also to be carried out on privately-owned arable lands of the same catchment areas, the expenditure for which was to be recovered from landowners. Rangeland improvement and reforestation works were to be started in ecologically better locations in order to increase the chances of success, leaving the more difficult sites to subsequent years, when experience and skills had been generated by the project. 2.10 The precarious financial position of typical smallholders in drought prone areas, their general inability to make on-farm investments and accept an increased burden of risks were all significant considera- tions in the design of a package to improve dryland farming practices. A Training and Visit (T & V) extension system with demonstrations on farmers' fields was to introduce new methods of land management and crop husbandry. Only low-cost proven technology was to be applied, and research in dryland farming was to be strengthened in order to develop new packages for specific areas. 2.11 Dry farming development was conceived as a pilot scale operation for the reasons that: (i) results from previous research were not economically attractive to farmers; (ii) the extension services needed restructuring and testing; (iii) limited supplies of fertilizer seemed to - 28 - be used to best advantage in irrigated or assured rainfall areas; (iv) credit institutions appeared wary of investing in high-risk dry farming; and (v) institutional credit for farm inputs was not sufficiently avail- able. 2.12 Recognizing the importance of income from farm animals, a number of livestock activities were considered essential. These included a dairy component to improve income through increased sale of milk and the estab- lishment of sheepgrowers' societies to bring about the voluntary par- ticipation of stock owners in the regeneration and controlled use of public rangelands. 2.13 Some agricultural diversification schemes were proposed in order to develop the skills and opportunities of communities in particular areas. These included sericulture and horticulture under dryland condi- tions, and fisheries development to create income opportunities from perennial water impoundments. 2.14 All project works were designed to be executed through existing line departments. Coordination of project activities was to be guaranteed through the strengthening of district level authorities. Project organization and objectives are summarized in Table 3. Location 2.15 Jodhpur and Nagaur Districts of Rajasthan were selected to repre- sent the sandy soils of northwestern India, while Ahmednagar and Solapur Districts in Maharashtra, Bijapur District in Karnataka, and Anantapur District in Andhra Pradesh were chosen to represent the alfisols and vertisols in drought prone areas of the Deccan plateau (Map IBRD 11077 PCR). Project Components 2.16 Watershed Development - Soil Conservation and Forestry The project provided funds for initial land-use capability surveys to deter- mine the conservation measures and land-use practices most suitable to particular watersheds. The appraisal report called for about 15 water- sheds to be surveyed, and subsequently treated. Afforestation was foreseen to be required on about 16,000 ha of public land in the four Deccan Plateau districts. It was further assumed that a total of about 2,000 ha of village woodlots would be planted on public lands and about 1,200 km of windbreaks developed along roadsides and canals. Rangeland improvement was to be carried out on a total of about 80,000 ha of public land, and on 10,000 ha of severely eroded, privately-owned land, unsuitable or marginal for arable farming. Rangeland improvement was based - 29 - on the planting of suitable perennial grass species, closure and sub- sequent controlled grazing. One pasture seed production farm was to be developed in each of the four participating states. 2.17 In conjunction with the above re-vegetation programs, soil and moisture conservation works were to be carried out on all watersheds treated. On arable lands, a system of contour or graded bunds, grassed waterways and farm ponds would be constructed on about 230,000 ha, and bunds already constructed over a further 135,000 ha of already cultivated lands would be repaired. 2.18 Dryland Farming. For the pilot-scale dry farming development, two or three watersheds, embracing about 22,000 smallholdings, would be selected in five project districtsl/ to promote improved dryland farming practices. Extension agents, trained in the T&V System, were to establish demonstration plots on farmers' fields and persuade farmers to use improved practices such as high-yielding varieties of selected crops, application of fertilizers and more care in timing of cultural operations. 2.19 Livestock Development - Sheep. To make a start with the control- led use of rangelands after their improvement, about 270 sheepgrowers' societies would be established, each operating a minimum of 100 ha of improved rangelands and which would be provided by the government. Par- ticipation in societies would be voluntary, with smallholder involvement through provision of shares, generally paid for in kind. Preference was to be given to the poorest segment of village society. The implementation of this component was to be through district task forces, working through sheep service centers. One ram multiplication center would be developed in each state. 2.20 Dairy development under the project was considered as part of the dryland farming program and based on milk production from indigenous animals. District spearhead teams were to establish a total of 480 milk producers' societies in villages with at least 300 mature female cattle or buffaloes. Additional lactating animals would be purchased in line with increased fodder availability expected from progress in the dryland farm- ing component. The project provided finance for milk-cooling facilities and assistance for the transport of milk to existing pasteurization plants. The project provided finance for import of some dairy bulls, for artificial insemination and for construction of raising centers for crossbred calves. Improved veterinary assistance was included in all district proposals, whereas assistance in building up fodder reserves for drought years was foreseen in Rajasthan only, and applied research to 1/ Jodhpur District was not included in the dryland farming program. - 30 - improve the feeding value of sugarcane bagasse through heat treatment, was limited to Maharashtra. 2.21 Water development. The project provided finance for 29 minor irrigation schemes with an expected total command area of about 8,900 ha. Another 7,900 ha was to be brought under irrigation through the estab- lishment of approximately 2,000 wells. Rajasthan, where groundwater is deep, was provided with finance for two air hammer drills for more effi- cient drilling in sandstone. Funds were to be provided for the simul- taneous development of the command areas. 2.22 Diversification Schemes. These were to be implemented in Kar- nataka, and consisted of inland fisheries development through coopera- tives, seasonal silkworm-raising from dryland mulberry gardens, and plant- ing of fruit trees on smallholdings. 2.23 Research, Technical Assistance and Training. The project provided funds for research in the economics of dryland agriculture, pasture development and agro-meteorology. A team of Indian economists, within the All-India Coordinated Research Project for Dryland Agriculture, was to evaluate research results, and on the basis of this, develop packages of dryland farming technology for use by farmers. One expatriate consultant was to assist the same institution with its research into rangeland improvement. A consultant in agro-meteorology was to work on the impact of soil moisture probabilities on cropping patterns. Technical assistance was also provided for teaching district field staff rangeland improvement techniques, and for development of calf-rearing schemes and bagasse treat- ment. 2.24 Various training programs were to be organized by the Central DPAP Unit, and training for dairy development was provided through the National Dairy Development Board. 2.25 Other Activities. A few other catalyctic activities were financed by the project. These included credit surveys and district banking plans, updating of land records, and the establishment of Farmers' Service Societies. Organization and Management. 2.26 The DPAP Central Unit within the Department of Cooperatives and Community Development, Ministry of Rural Reconstruction, New Delhi, was strengthened to provide direction to the project. The important tasks were: (i) ensuring adherence to technical, financial and economic norms; (ii) monitoring of project performance; (iii) execution of training; and (iv) liaison with financial institutions. The Central Unit was strengthened by addition of senior Indian specialists in dry farming and - 31 - soil conservation, minor irrigation, livestock and economics, and an expatriate advisor on rangeland improvement. 2.27 At the state level, small units with responsibility for planning, budgeting and supervision were established. Interdepartmental coordinat- ing committees were also established in order to ensure the participation of line departments. 2.28 District Development Authorities (DDA) were established in each project district to coordinate and supervise development activities. Each Authority consisted of a Managing Board, chaired by a senior government official - in most cases the Collector or Deputy Commissioner. Respons- ibility for planning, organizing, supervising and administering project activities was vested in Project Directors who reported to the Managing Board and who were each assisted by a project economist and an administra- tive staff. All public sector works were to be executed by existing government line departments. Only the drilling of irrigation wells was expected to be undertaken by private contractors. The right of inspection was vested in the Project Director. III. IMPLEMENTATION Start-Up 3.01 Considering its complexity, the project got off to a reasonable start. An intensive dialogue between GOI and state DPAP authorities developed during appraisal and subsequent negotiation of the project, and all states quickly appointed Project Directors for the six IDA-assisted districts. Staff from line departments was promptly seconded to DDA. Revision 3.02 The project did not require major revisions. The only significant change was IDA's agreement to reimburse against expenditures for rural electrification in Jodhpur and Nagaur districts to energize pumps installed for minor irrigation. Physical Implementation 3.03 Since activities foreseen in the Appraisal Report were not location-specific, the translation of models into district development plans and close cooperation with concerned line departments was needed. This took more time than originally expected, thus causing delays in the commencement of work. - 32 - 3.04 Due to the initial lack of technical expertise and understanding of project concepts, project staff tended to adhere rigidly to physical targets as listed in the Appraisal Report and in annual work plans, whereas in practice, these targets were intended to be illustrative. This attitude prevailed at all levels and throughout the project organization, and became generally accepted as an indication of project achievements. 3.05 Furthermore, the need to create employment for rural people, which had always been an important objective of DPAP, led to a strong emphasis on civil works and offered the added advantage of being relatively easily planned and supervised. As a consequence, expenditures for civil works in minor irrigation, soil conservation, reforestation and pasture development tended to exceed target figures. 3.06 The major project targets were achieved or exceeded. Key indicators are tabulated in Table 4. IV. TECHNICAL PERFORMANCE General 4.01 The six IDA-assisted districts were well chosen. They were representative of two main ecological zones of drought prone districts and the development techniques applied were relevant to other districts in the northwestern area, as well as on the Deccan plateau. 4.02 In addition, major project activities for resource and agricul- tural development were well conceived. They embraced all types of work which could reduce the impact of drought, and the project succeeded in testing these under field conditions and at a scale sufficiently large to establish decisions concerning future DPAP policies. Overall, it is considered that the IDA-assisted project in the six districts was success- ful. However, the inclusion of some minor components caused some confu- sion, and sometimes, diffusion of the project's major objectives. Project Components 4.03 Watershed Development. The project envisaged that agricultural and resource development was to be planned and implemented on the basis of watersheds. The planning and development of watersheds was, on the whole, somewhat piecemeal, with a wider than desirable geographic dispersion of activities, and with the result that all components were rarely imple- mented within the same catchment area. This led to difficulties in assessing the level of overall benefits, particularly on the ecology, agricultural production and farmers' incomes. Careful studies over a number of years will be required to assess the full benefit of these investments in 52 watersheds. - 33 - 4.04 Soil Conservation. Practices such as bunding and improvement of waterways were applied to about 335,000 ha in the watersheds above and to other lands. In addition, about 140 farm ponds were constructed. The total investment in soil conservation and soil surveys was US$16.3M. The traditional approach to soil conservation has not changed; it consists of large "contour bunds" which are normally on field boundaries, and not laid out on contours. Farmers have to accept, and pay for these structures of doubtful economic and conservation value. In many instances, farmers have neglected the maintenance of bunds or the repayment of construction costs. It is now much clearer that more effective soil and moisture management is dependent upon in-field practices such as cross-slope cultivation and establishment of broadbeds and furrows supplemented by permanent drainage ways. 4.05 Forestry. State forestry departments were responsible for reforestation and rangeland improvement. Reforestation on badly eroded sites of upper catchment areas has been difficult and costly, and resulted in selection of sites with better soil conditions for reafforestation, whereas rangeland improvement activities were relegated to the most dif- ficult sites which were generally poorly-suited to any form of improve- ment. Forestry Departments have placed increasing emphasis on developing social forestry. However, cooperation with local organizations for the protection and management of such plantations has still to be worked out. About 20,700 ha were planted as against a target of 20,200 ha. 4.06 Rangeland. Initial emphasis was on the establishment of perennial grasses. Establishment techniques promoted by the project were success- ful, but the increase in fodder production was too small to cope with the grave problems created by general overstocking. It is clear, therefore, that establishment of improved grasses alone cannot bring about the chan- ges in the traditional patterns of rangeland exploitation and livestock husbandry. Reductions in livestock numbers are prerequisites for restor- ing the ecological balance of these areas. Reduction in livestock numbers would present a problem in a society which forbids their slaughter. 4.07 On the Deccan Plateau, a major project accomplishment was the introduction and testing of drought-resistant forage legumes. Stylosanthes hamata and S. scabra were introduced from Australia and after an initial period of seed multiplication, the two have been widely planted under a variety of soil, rainfall and land preparation conditions. With an average monsoon, Stylosanthes can be established on better wastelands by broadcasting on unprepared land, and only simple furrows are required for its establishment on badly eroded sites. The immediate impact of Stylosanthes on fodder productivity, its capability to set seed quickly under low rainfall conditions and its relative unpalatability during early growth have created keen interest among farmers as a means of rapid and low-cost rangeland improvement. - 34 - 4.08 In the Rajasthan districts, project activities demonstrated that the reestablishment of perennial grasses alone cannot solve the supply problem created by overstocking. It is now apparent that more emphasis should be placed on the re-introduction of traditional fodder trees and bushes. Techniques for establishing and utilizing "kejari trees" (Prosopis cineraria) and "ber" bushes (Zizyphus spp) are known to farmers, and both grow well in the central and eastern regions of the state. Fod- der production from these plants, additional to that from grasses, is more constant over the years and allows considerably higher stocking rates. In the last two years, extensive efforts were made to introduce Leucaena in all six districts. 4.09 Under the project, state forestry departments developed nurseries for the multiplication of tree material, seed gardens for the propagation of grasses, and legumes for rangeland improvement. During the final project years, much emphasis was put on the multiplication of Stylosanthes seed and a start has been made in Anantapur district to multiply seed on privately-owned marginal lands with a guarantee to buy back seed at fixed prices. About US$4.1M was spent to improve about 47M ha of rangeland. 4.10 Dryland Farming. Overall progress of the dryland farming com- ponent has been satisfactory despite a number of poor monsoons and which has been confined to farmer demonstrations, combined with an extension effort based on the T&V system of extension. Overall impact on agricul- tural production has been moderate, but emphasis is now on simple, low-cost practices which are of little risk to farmers. About 73,000 farmers, for the most part in watershed development areas, were persuaded to apply improved practices such as the selection of better crops and seed, application of fertilizers and improved weed control to about 250,000 ha. The government investment amounted to about US$1.5M. 4.11 Livestock Development. About 160 sheep grower societies were established in all districts and have provided some useful information on the voluntary participation of livestock owners in the controlled grazing of improved rangelands. However, on the whole, this component has not been successful and attracted about one fourth the expected membership. Pasture blocks of 100 ha, which were to become the production base of societies could not be developed using the grass planting technique envisaged as feasible at appraisal; thus, fodder production remained much below expected levels. As a result, the sheep growers' societies had to function below proposed stocking rates and forced the perpetuation of traditional forms of transhumant sheep husbandry. There was also con- siderable resistance to controlled grazing from owners of livestock who were not participants in the project, and thus, not entitled to use lands allocated to societies. Maharashtra, as a result, terminated its support to sheep growers' societies early in 1981, although other states continued to support this component. The continuation of this work appears viable only in combination with rangeland improvement through planting of - 35 - Stylosanthes; this should improve overall forage production on wastelands to the extent that all village livestock could participate. 4.12 Dairy was the most successful project component. About 750 producers cooperative societies were formed in all districts and essential infrastructure in terms of milk collection routes, chilling centers and final product outlets were established. A number of technical problems are still to be solved-- of greatest importance is the need to increase the regularity of milk production at the farm and primary society levels. Better feeding and fertility levels of adult female stock are critical and require the combined attention of agricultural extension staff (to increase availability of fodder as a by-product from dryland farming), and dairy spearhead teams (for advice on management), as well as the provision of quality indigenous dairy bulls. The services of animal husbandry departments also need improvement, as artificial insemination remains unreliable and expensive, requiring improved handling and transport of semen, and better-qualified inseminators. The inclusion under the project of cattle crossbreeding with exotic breeds created considerable problems. First, it allowed the diversion of attention from the wider use of care- fully selected indigenous breeds, and second, it called for the inclusion of AI and calf-raising centers, which attracted too much attention from animal husbandry staff, and were insignificant as far as the overall success of the project was concerned. The concept of calf-raising centers was dropped after discussions with supervision missions. About US$10.7M was invested by Government in dairy development. 4.13 Water Development. The construction of 23 irrigation tanks out of 29 originally planned has been completed and work on others is in various stages of completion. GOI proposed that the number of tanks constructed be increased to 57. Investigations by various supervision missions indicated, however, that: (a) costs of minor tanks were much greater than anticipated; (b) numbers of beneficiaries per tank remained below expectation; (c) command area development (CAD) was slow and efforts to train farmers in irrigation skills limited in their effective- ness; (d) problems of equitable distribution of water to all farmers in the command area persist; and (e) sometimes, political considerations appeared to interfere with technical/economic reasoning in site selection. Repeated requests by GOI for an expansion of the minor irrigation develop- ment program were not accepted by IDA, and supervision missions recom- mended that more emphasis be given to CAD, and to improved water utiliza- tion on the schemes already constructed. 4.14 The target for deep well development in Rajasthan and for dug wells in the Deccan plateau districts was achieved. The problems encountered early in project implementation were the definition of "small farmer" for loan qualification, and lack of electrification; the former being resolved in 1977 and the latter reduced by including a limited electrification program in the project. The major problems that remain - 36 - after electrification are CAD, water management and intensification of extension services. On the Deccan plateau, new wells are often clustered in the vicinity of existing successful wells, and exploitation then exceeds recharge rates. As noted during appraisal, legislation, with suitable enforcement is required to control well placement. While the well program is essential to fully exploit existing groundwater resources, more advice to farmers is required on how to bring about command area development, improved water management, and crop husbandry practices. This requires more training and improved supervision of agricultural extension staff. On many farms, over-irrigation and selection of high water use crops (i.e. paddy, sugarcane) continues, thus reducing the potential command area. Government expenditures of US$16.5M or 23% of of the project investments were made in irrigation which directly benefit an estimated 7,832 farmers. 4.15 Diversification Components. The sericulture program (included in Karnataka and added in Andhra Pradesh) has made good progress. Thus far, mulberry gardens have been established only on irrigated lands, hence, this component has had no impact on dryland farmers other than through employment generation. The fisheries component (added in Andhra Pradesh and Maharashtra) has enlarged the capacity for production of fingerlings for stocking irrigation reservoirs. The horticulture com- ponent in Karnataka continues with the distribution of fruit trees to dryland farmers. These last two components are proceeding well, but are of benefit to about 15,000 families. 4.16 Research Support. The work of the Senior Pasture Research Agronomist has led to the identification of Stylosanthes spp for the Deccan plateau. The DPAP Pasture Advisor completed his assignment and was successful in training foresters and others in pasture development techniques. GOI has not yet developed an agro-meteorological service to support dryland farming development as was recommended in the report of the agro-meteorologist. A final agreement between the Indian Council of Agricultural Research (ICAR), the Indian Meteorological Department and DPAP is still under discussion. The recommendations of the consultant on bagasse treatment were followed by a series of feeding trials. The results have been positive. The authorities in Maharashtra, however, did not go ahead with the installation of a heat-treatment plant at a coopera- tive sugar mill as (i) hard board factories have been established since the inception of the project, tripling the value of untreated bagasse, and (ii) state authorities have little interest in the establishment of fodder banks for years of drought, for which treated bagasse was originally intended.I'The Senior Agriculturist Economist stationed with AICRPDA initiated a number of surveys and studies, the majority of which were completed in 1981. 4.17 Agricultural Credit. Banking plans were completed for all dis- tricts and linkages between district officials and credit institutions 1/ See Borrower's views on why bagasse treatment plant was not installed; Attachment II to PPAM, para. 4. - 37 - improved. On a number of occasions, lending agencies were slow to par- ticipate because of their skepticism regarding the financial viability of wells and farmer ability in water management. Farmer Service Societies were formed in each district and land records were updated in Anantapur, Jodhpur and Nagaur Districts.!/ 4.18 Training. At the initiative of the Central DPAP Unit, 118 train- ing courses designed for various groups such as project managers, economists, foresters and agricultural officers were attended by 3,072 persons. The more successful training courses included pasture develop- ment, project administration and dryland farming. V. FINANCIAL PERFORMANCE Project Costs and Financing 5.01 The allocation of total project costs (US$102.7 million) to dis- tricts and the GOI DPAP Central Unit, and to individual project components is shown in Table 1. Table 2 summarizes the District and GOI Central Unit expenditures. The IDA credit provided US$35 million. 5.02 When appraised, 41% of all funds were left unallocated to cover cost and physical contingencies, as well as to facilitate reallocation of funds during project implementation to successful or additional com- ponents. Schedule A of the Development Credit Agreement (DCA) was sub- sequently amended on February 28, 1980 and funds reallocated within the concept of the original project design (except for US$1.4M which was allocated to rural electrification in Rajasthan to optimize project benefits from minor irrigation investments). 5.03 Total project expenditures are now estimated at US$84.3M, or 82% of appraisal estimate of US$102.7M. Government expenditures were US$71.5M (Table 1), representing 99% of the appraisal estimate, and 85% of actual project costs. Institutional credit, which was expected to account for 34% of total project costs, or US$30.6M amounted to US$12.7M, according to available records. The less than anticipated use of institutional credit was due mainly to the failure to identify funds obtained by individuals as part of DPAP activities. Difficulty was also encountered over the defini- tion of "small farmer" for loan qualification (para. 4.14), which, per- haps, prevented some farmers of this group from participation. 1/ See Borrower's comments, Attachment I to PPAM, para. 15. - 38 - Project Financing US$M GOI/States IDA Credit Inst. Total Appraisal 37.1 35.0 30.6 102.7 Achieved 36.6 35.0 12.7 1/ 84.3 Percent of Appraisal 98.6 100.0 41.5 82.1 1/ Incomplete data. Disbursement and Expenditures 5.04 The rate of disbursement was initially slow (due, for the most part, to laxity in preparing requests at the district level and processing at the GOT level), but picked up during the second year. However, totals remained considerably behind appraisal estimates and the final disburse- ment payment was made on May 9, 1981 about one year after the original project closing date of June 30, 1980. At appraisal, it was estimated that the Credit (US$35M) would finance about 50% of all public sector investments, of which US$9.7M was to be for civil works. GOT, in 1979, proposed disbursement expenditures of US$23.7M in the civil works category, of which US$8.6M were for irrigation and US$5.9M for soil con- servation. 5.05 Eventually the 1979 supervision mission proposed the following disbursement program which was accepted by GOT in February 1980. Disbursement Schedule 1 As per GOT Amendment Actual Category as per DCA Proposal to DCA Disbursement (Sept. 79) (Feb. 25/80) (May 6/81) ------------------- (US$ million)--------------------- 1. Civil Works 9.70 23.75 16.60 18.28 2. Equipment and Vehicles 2.60 6.69 6.25 4.90 3. Operating Expenses 6.90 9.03 9.05 9.54 4. Technical Assistance and Research 1.30 1.70 1.70 0.67 5. Unallocated 14.50 0 0 0 6. Rural Electri- fication 0 1.40 1.40 1.61 Total 35.00 42.57 35.00 35.00 5.06 The aim of the modified schedule was to keep IDA disbursement to about 50% of total project costs and, at the same time, reduce over-emphasis on construction of civil works. Increased spending for civil works resulted from: (a) unit cost increases in minor irrigation - 39 - works; (b) provision of additional facilities for dairy development; and (c) requests for inclusion of soil conservation works on private lands. Internal and Budgetary Control 5.07 The DPAP budgetary procedure gives GOI a controlling position. Participating states submit annual district development plans to the Central DPAP Unit in New Delhi, where the programs proposed by DDA are examined for their eligibility in terms of overall DPAP objectives and conformity with technical norms. The states are then informed by GOI which parts of district proposals will be eligible for GOI's 50% contribu- tion. GOI makes its contribution to the states in quarterly installments upon receipt of the proof of payment made during the preceding three-month period. 5.08 State level authorities played only a minor role in the formula- tion of annual DPAP work programs. The major planning exercise was under- taken at the district level. DDA was informed of the total budget alloca- tion for the year and it was the task of line department officers to prepare detailed annual work plans within the ceilings allocated to their departments by the DPAP Project Director.l/ Accounts, Audit and Disbursements 5.09 The accounting and audit procedures for the project followed regular GOT procedures. DDA maintained records of transactions and trans- fers. The locally responsible officer for each line department maintained accounts and receipts for the schemes or works implemented by his depart- ment and recorded these on forms prescribed by the government. 5.10 The accounts of DDA, as well as those of executing unit officers were audited by a chartered accountant and auditors from the office of the Accountant General of the state concerned. The chartered accountant forwarded the audit reports to GOI and the appropriate state government. As is typical in India, these audits consistently lagged despite repeated requests from supervision missions. At various times, DDA accounts were inspected by missions, and problems of eligibility for disbursements, brought to the attention of GOI's Department of Economic Affairs (DEA), who organized the needed corrective actions. After preparation of IDA reimbursement claims by the project director or line department officer, DDA forwarded IDA reimbursement claims, through the State government, to DEA. DEA further examined claims prior to forwarding to IDA. Respon- sibility for determining the eligibility of items claimed was left to DDA, who acted on guidelines prepared by the DPAP Central Unit and based on the staff Appraisal Report (also see para. 7.04). 1/ See Borrower's views on role played by state authorities in annual DPAP work program development; Attachment I to PPAM, para. 5. - 40 - 5.11 Actual final disbursement (para. 5.05) differed somewhat from the amended DCA because of the lack of financial control of disbursement requests by the DPAP Central Unit. IDA, in turn, exerted little control over the final disbursement in particular categories. Covenants 5.12 Conformity with legal covenants was satisfactory, with the excep- tion of the timely submission of audit reports. Supervision missions repeatedly called attention to this, and all reports and certificates indicating project expenditures were eventually supplied. VI. PROJECT IMPACT Introduction 6.01 The appraisal report and all subsequent supervision reports have stressed that, in addition to the monitoring of physical and financial progress, the experimental nature of the project required a comprehensive evaluation of individual components to determine which of these were sound and suitable for further replication. The DPAP Central Unit worked out detailed guidelines for impact evaluation during the initial project years and at least 32 studies were commissioned either under direct respon- sibility of the DPAP Central Unit or through state authorities. However, little comprehensive impact evaluation data was obtained from these efforts; possibly, more useful information would have resulted had more IDA guidance in evaluation been provided. 6.02 Aware of the problems arising from the lack of effective evalua- tion data, a 1980 supervision mission assisted GO in drafting a set of questionnaires to be used by DDA in order to obtain a rough, but accept- able assessment of project impact. The questionnaires were sent to project districts in May 1981, and the results were made available to the project completion mission. In addition, all districts prepared project completion reports following a standard outline provided by GO. 6.03 These project completion reports prepared by the districts con- centrated on progress towards attaining financial and physical targets. Little useful information was provided on the impact of project activities on agricultural, environmental and social objectives. Data was provided by line departments. Some of the figures did not tally with those provided by DDA for the district as a whole, and some of the impact indicators appear to be expectations rather than field measurements. The project completion mission, therefore, found it difficult to make a com- prehensive assessment of the project's impact. - 41 - Direct Benefits 6.04 Number of Beneficiaries. The total number of families benefiting from the project is estimated at 281,000 (appraisal target was 260,000). The majority of the participants were members of the target group consist- ing of small and marginal farmers, and members of the scheduled castes and tribes. Appraisal Achievement Component Total Target Group 1/ Total Target Group Soil Conservation 84,000 42,000 79,082 54,000 Dryland Farming 37,000 18,500 73,054 37,000 Sheep Raising 30,000 23,000 7,710 6,170 Dairy Improvement 84,000 52,000 74,585 52,200 Minor Irrigation 9,500 7,300 7,832 3,880 Sericulture 400 200 23,335 18,500 Fisheries 1,000 1,000 4,015 3,800 Horticulture 16,000 8,000 11,474 6,000 261,900 152,000 281,087 181,550 As percentage of appraisal 107 119 6.05 Employment Generation. The project provided approximately 35.8M mandays of employment, the main portion of which was for unskilled labor used for the construction of soil conservation works, minor irrigation tanks, and for forestry and pasture improvement. The majority of workers engaged in these activities are from the landless groups or are very small farmers. Additional employment, needed in future years will be required for maintenance of civil works, and for incremental crop and livestock production. Million Mandays of Labor Component Appraisal Estimate Achievement Soil Conservation 15.0 18.8 Irrigation 6.0 6.9 Pasture (Forestry) 2.0 9.3 Other 5.4 .8 Total 28.4 35.8 1/ The target group were small and marginal farmers and members of the scheduled castes and tribes. - 42 - As percentage of appraisal 126 6.06 Incremental Production. The project increased the production of foodgrains, cash crops, milk, fodder, wool, fruit, and fish. Apart from increased output, the project has probably lessened crop yield depression in years of drought. Farmers are beginning to apply improved soil and moisture management practices, as well as crop selection and management practices to reduce the effects of drought. The lack of proper evaluation data and highly disbursed activities made accurate estimates difficult. Incremental Agricultural Production and Income Appraisal Estimate Achievement tons value tons value (RsM) (RsM) Food Crops Soil Conservation 6,400 5.7 3,370 3.4 Dryland 30,000 30.0 37,000 37.0 Irrigation 22,000 20.0 20,000 30.0 58,400 55.7 60,370 70.4 Milk 80,000 97.0 92,400 131.1 Sheep Wool 457 4.2 NA NA Mutton 2,400 6.0 NA NA Fish - - 283 NA Silk (cocoons) - - 74.6 2.3 Pasture 67,500 5.0 37,730 NA 6.07 Watershed Development and Soil Conservation. Because of the time required for surveys, planning and implementation, no watersheds were sufficiently fully developed to permit the evaluation of the benefit of the soil conservation works. Contour and boundary bunding was the prin- cipal soil conservation practice applied to crop lands. This practice, when combined with land shaping and controlled drainage, provides the farmer with increased capability for crop selection, and permits the application of improved cultural practices leading to increase yields. 3 6.08 Forestry. Forest departments estimate that 10-12 m annually of wood are produced on one heStare, but on difficult sites, annual produc- tion is probably only 5-8 m per year. Survival rate to one year after planting has varied between 50% and 80% in below average and normal rain- fall years, respectively. Surviving trees are guarded intensively by forest department staff, but the danger of unauthorized felling increases - 43 - steadily with the growth of trees and is probably a reflection of the huge demands for fuelwood. 6.09 Rangeland. The planting of grasses in contour furrows, as recom- mended during the first few years of rangeland improvement, has had little impact on overall fodder productiion. Furrow plantings occupied only about 20% of the total area and the introduced grasses did not spread into the inter-furrow spaces. Dry matter yields of fodder, therefore, were not significantly different from unimproved areas, i.e., about 200 kg/ha in years of drought, about 500 kg/ha in normal years, and about 1,000 kg/ha in good rainfall years. With the advent of Stylosanthes, in the Deccan Plateau, the situation has changed radically. In the second year after broadcasting, about 800 kg/ha are produced in dry years. With normal rainfall, production ranges between 1,500-2,000 kg/ha and in years of good rainfall, fodder production exceeds 3,000 kg of dry matter per ha. 6.10 Dryland Farming. Demonstrations in farmer fields provide an indication of potential yield increases resulting from low cost activities such as HYV and fertilizer use. For many crops, yields on demonstration plots were 2 to 3 times those obtained from control plots. For the pur- pose of estimating incremental production, crop yields, based upon data from crop demonstrations, were assumed to have increased on average by 150 kg/ha, mostly as a result of planting of HYV, establishment of better stands, application of fertilizer and better weed control. 6.11 Livestock Development - Sheep. The impact of project activities on wool and mutton production has been minimal due to the fact that productive pasture blocks did not develop as was foreseen at appraisal. This was the result of the original technically non-viable approach to rangeland improvement. The small increment of wool and meat production that actually occured was not as a result of increased availability of fodder, but resulted from improvements in management of flocks and veteri- nary services. 6.12 As a consequence of the above, members of sheep producers' societies derived little additional income from their participation. This was due to the fact that they were able only to contribute one or two sheep to the society and this left their remaining flock to exist under traditional husbandry conditions. 6.13 Dairy. Total milk production from producers participating in the project is estimated at 92M liters. No surveys have been made to assess the influence of a regular monetary outlet for fresh liquid milk (milk collection centers) on home consumption. Deliveries of fresh milk to collection centers have built up rapidly and they stood at about 25M liters in 1980/81. Milk production from drought prone areas remain highly seasonal. In Rajasthan, especially, milk deliveries are strongly affected - 44 - by rainfall. In 1980/81 milk deliveries were only about 40% of deliveries two years earlier, when good rains occured. 6.14 The members of milk producers' societies derived an average of Rs760 from the sale of milk during 1980/81. Feeding was based almost exclusively on farm-produced fodder and concentrate feeding remained negligible. Average income per member from milk sales in 1980/81 was lowest in Bijapur District (Rs210) and highest in Ahmednagar District (Rs2,630). 6.15 Water Development. Increased cropping intensity and greater yields, with timely irrigation, account for most of the incremental production to date. As farmers become more skilled in water management and command areas developed, so will production continue to increase. Incremental yields are presently estimated at 500 kg/ha. 6.16 Impact on Environment. Soil conservation, pasture and forest development on 402,000 ha has reduced environmental degradation due to wind and water erosion. The use of better farming practices by par- ticipating dryland farmers has also assisted in reducing erosion losses on productive land, and higher yields lessened the pressure to utilize mar- ginal land for crop production. The success of Stylosanthes plantings in the Deccan plateau for pasture and seed production has interested some farmers in the diversion of marginal crop lands to the growing of this legume. Indirect Benefits 6.17 The package of activities agreed upon for the six IDA-assisted districts has since become the model for GOI-approved DPAP activities in all drought prone districts. Also, the administrative system adopted in project districts has been used in all other DPAP areas. 6.18 The involvement of IDA in DPAP has created an awareness among GOI and state level policy-makers that the technical and economic viability of individual program components must be kept under constant review. A reliable impact evaluation system has still to be established, but govern- ment officials are aware of this need. In this way, IDA assistance to DPAP has prepared the way for a number of policy decisions which will have a bearing on future DPAP approaches. Rates of Return 6.19 Thirteen models were prepared at appraisal to estimate the finan- cial viability of investments for on-farm and public land development. At appraisal, ROR were estimated for the models and ranged from 11 to more than 100 percent. To duplicate the farm models prepared at appraisal, reliable technical and financial information on the impact of project - 45 - investments at the farmer level is essential. Unfortunately, this is not available. Nevertheless, several important project components were selected for examination during the PCR mission visits to project dis- tricts. Thus, it was possible to develop some estimates of the financial impact of the project components at the farmer level. A major difficulty in developing estimates of the future is that crop yields are subject to extreme weather variability in project areas. The range in yields of millet for example, extend from less than 50 kg/ha in an extreme drought year to 1,000 kg/ha in a very good rain year. For sorghum, yields range from 50 to 900 kg/ha. 6.20 The yield estimates used for the farm models were based on an estimation of the expected yield values-- that is, weighting the estimated yields corresponding to specific weather conditions, on the probability that those weather conditions would occur. For example: Pearl Millet (Bijapur-Shallow Soils) Yields Value Weighted Value Performance (kg) Probability (Value times probability) Without Project Very good 400 .20 80 Good 280 .20 56 Not very good 200 .40 80 Poor 50 .20 10 Expected Value: 226* With Project (at full development) Very good 900 .20 180 Good 600 .20 120 Not very good 450 .40 180 Poor 200 .20 40 Expected Value: 520* * The expected values were the figures used in the farm models. The number of years necessary to achieve full development for each techni- cal package, as well as intermediate values, were derived in consult- ation with local project staff. Financial Rates of Return - 46 The financial rates of return for some of the most important project components were: Financial Rates of Return % Net Benefit Increase 1/ (Discounted at 10%) 1. Well (1.5 ha) 39 5052 -- 428% 1181 2. Dry Farming (4-6 ha) 100 603 Red Alfisols --- - 16% 3656 3. Dry Farming (4-6 ha) Black Vertisols 100 895 -- = 173% 517 4. Sheep Societies (100 ha) a. Partioipating Rarmers 1. Assuming Society is liquidated after 5 years (mosly likely). Negative Negative 2. Assuming Society pays dividends of Rs 100, in real terms, for years 5 to 9 (not very likely). 15 Nil b. Society as a whole Negative 2/ 5. Dal~ A landless laborer with 1 cow 100 1109 --- 50% 2250 6. Tank irrigated farms a. Mukarthal Tank 1. 2 ha farm 100 15789 ---- =71% 22255 2. 4 ha farm 17 3318 ----= 5% 62305 b. Jigjivani Tank 1. 2 ha farm 32 4153 ----- = 27% 15602 2. 4 ha farm 56 16302 ---- -= 49% 33517 1/ Net Benefit Increase (NBI) is the ratio, measured in percent of: Ave. Annual Incremental Net Benefit (after financing) ---------------------------------------------------- = NBI Ave. Annual Net Benefit (without project) 2/ The FRR to participating farmers was 15% (in the most optimistic case) and negative (-5%) (in the most likely case). Additional costs of establishing and running the society, not included in the FRR to participating farmers, included rams, fencing and sheds, pasture development, official to supervise (1/3 man year per society) and 2 shepherds. - 47 - 6.21 The NBI ratio represents one measure of the attractiveness of the project to farmers. The FRR tends to be very high (indicating that this type of investment can be attractive to small farmers making little addi- tional investment), but this can be a misleading indicator of the attrac- tiveness of the project. The NBI complements the FRR by measuring the percentage increase in benefits with project, compared to the benefits without project. In other words, NBI indicates how much additional annual income the farmers will have by participating in the project. 6.22 Most selected components show a rather high financial rate of return before financing. Due to subsidized credit, the FRR after financ- ing, would be greater. This indicates that the project was especially beneficial to those who participated and were financed. However, a ques- tion for the future is: by how much can subsidies be reduced in order to spread the benefits among a larger proportion of the target population while at the same time, keeping project activities financially attractive? A serious study of the extent of subsidies-- who gets them and to what extent-- would probably suggest changes in the amount and activities being subsidized, and in the target groups. Economic Rate of Return 6.23 An overall economic rate of return for the project was not calcu- lated at appraisal due to the variety of components and the complexity of measuring some of their benefits. Instead, approximations of economic rates of return were calculated for several project components. The mission had access to only one study on the economic impact of two tank irrigation sub-projects. After making several adjustments to the data and methodology, the estimated economic rates of return were: 1. Mukarthal Tank - 14% 2. Jigjvani Tank - 9% These rates of return are less than the 26 percent EER estimated at appraisal for a 4 ha farm. The mission's impressions, after visits to several tanks, is that the return on this type of investment is marginal on the best assumptions, and perhaps close to zero or even negative, in most cases, as CAD development and water management are often neglected. 6.24 On the positive side, the dryland farming components show extremely high financial, and most likely economic, rates of return. Dairying appears to be financially attractive, but the ERR is much lower due to associated costs of collection, chilling and society management. On the basis of the data collected by the mission and models developed, if the project were to be reformulated now, the dryland farming component, particularly in the black soils areas, would be the central focus of the - 48 - project, and minor irrigation works and sheep societies would not be included. VII. INSTITUTIONAL PERFORMANCE General 7.01 DPAP had good institutional support. There was a genuine interest by Government at all levels to make the complex program work, and a rela- tively intensive dialogue was established between authorities at all levels. This situation did not change during the project implementation period, although the attention given to DPAP at GOI and state levels became less pronounced in later project years. 7.02 The frequent changes of administrative staff were a problem. Only the first Director of the Central DPAP Unit stayed long enough to ensure that comprehensive guidelines were issued in respect of project direction and administration. Administrative staff at the state and district levels were changed frequently and replacements were not fully briefed to ensure continuity in approach. In contrast, there was much greater continuity of professional line department staff seconded to the project. Administra- tive staff, therefore, had difficulty in integrating the work programs of line departments and this explains why: (a) the technical approach of line departments was difficult to change, and (b) physical targets and techniques specified in the appraisal report were given more importance than was envisaged at the time of appraisal. The DPAP Central Unit in GOI 7.03 The strength of the Central Unit was its capacity to execute national development programs. In this, it was obviously aided by its administration of GOI development finance for drought prone areas. This provided it with a strong position vis-a-vis the states in the design of programs eligible for GOI reimbursement. While this role was exercised well in the case of approval of annual district work programs, the Unit was less well able to follow up on the work executed. 7.04 Indian financial regulations, which govern procedures for public sector expenditures produced minor problems in reimbursement procedures. In the case of IDA reimbursement, GOI specified in circulars which com- ponents were eligible. However, claims for IDA reimbursement are sub- mitted directly from the state authority to DEA. The DPAP Central Unit received only a copy of the claims and no action was taken to ascertain the eligibility of individual claims submitted for reimbursement. Such control did not take place at DEA either. The need to improve financial control was brought to the attention of project authorities by supervision missions, and corrective action was taken. - 49 - 7.05 A weakness of the DPAP Central Unit was its lack of technical expertise. Professional expertise in drought prone areas is almost entirely centered in the Ministry of Agriculture. At the beginning of the project, it was planned that essential technical support would be provided in two ways: (a) by appointing a number of senior specialists to the DPAP Central Unit; and (b) by relying on the development of technical packages in close cooperation with the different units of the Ministry of Agriculture. 7.06 Cooperation with the Ministry of Agriculture did not materialize to the extent desirable. However, the DPAP Central Unit was able to obtain the support of ICAR, which with financial support from DPAP, strengthened its activities to identify dryland farming packages and techniques for improving fodder production on non-arable lands. However, the DPAP Central Unit could not rely on routine and timely support from other technical units in the Ministry of Agriculture. 7.07 The major responsibility for the development of technical pack- ages, therefore, remained with the senior technical specialists appointed to the Central DPAP Unit. They had a difficult task. This ranged from the establishment and evolution of technical norms, scrutiny of develop- ment proposals and training, to field advice and inspection. The impact of these professionals would have been more significant had they been appointed at civil service grades which permitted meaningful dialogue in their discussions with state directors of line departments. State Level DPAP Authorities 7.08 The state level DPAP authorities were attached to different departments depending on the State. These ranged from the Department of Planning in Maharashtra to the Special Schemes Organization in Rajasthan. In all cases, special state DPAP units were established and the tasks of program planning, financial administration, staff appointment and coor- dination with line departments were conducted well. District Development Authorities (DDA) 7.09 With the exception of Nagaur District and for a period of about two years in Bijapur District, young Indian Administrative Service (IAS) officers headed DDA. Some lacked experience in managing a multi-disciplinary development task force and the majority were not given special training prior to assignment. However, in general, the posting of these IAS officers proved successful. They made strong efforts to inform themselves of DPAP objectives and procedures, and developed considerable skill in coordinating the work of the state line departments. Their close associa- tion with Collectors' (Deputy Commissioners') offices was of considerable help. - 50 - 7.10 Line department officers who were seconded to district DPAP, had a difficult task. Activities foreseen under the project were quite often different from those supported currently by their departments. Par- ticularly strong divergencies existed in respect of soil and water conser- vation techniques, the approach to minor irrigation, the importance of tree planting for wasteland improvement and the role of crossbred cattle in drought prone area milk production. The situation was aggravated by the fact that the technical staff of the DPAP Central Unit were unable to influence the official approach of state line departments except through persuasion. While forestry departments in all districts agreed to include fodder development as part of their activities for wasteland improvement - and actually performed well due to the strong organizational pattern in all forestry departments - the other departments did not really accept the specialized approach needed for DPAP. As a result, line department activities in the project districts often were a mixture of traditional approaches and special DPAP activities. IAS project directors were not in a position to recognize technical shortcomings, or the lukewarm support for DPAP activities shown by some line departments throughout the project implementation period. Technical Assistance 7.11 The impact of technical assistance was positive. Forest depart- ment and animal husbandry staff received detailed and very practical guidance in rangeland improvement and forage seed multiplication, and the introduction of forage legumes under the project has been an event of major significance for rangeland improvement work on the Deccan plateau. Expatriate specialists prepared the groundwork for the implementation of an agro-meteorological service and bagasse treatment. Overall achieve- ments are smmarized below. Speciality Major Accomplishment Sr. Pasture Advisor Pasture establishment and manage- ment techniques and training of district staff. Sr. Pasture Research Agronomist Introduction, testing and develop- ing cultural practices for Stylosanthes (a major accomplish- ment of DPAP). Agro-meteorologist Prepared recommendations for agro-meteorology program which GOI is beginning to implement. - 51 - Calf-rearing Specialist Advised government that large-scale calf-rearing projects were not feasible. Component dropped. Bagasse Consultant Prepared plans for producing bagasse, which government com- pleted. Sr. Agriculture Economist Coordinated five major studies. Monitoring and Evaluation 7.12 At appraisal, it was emphasized that due to the innovative and experimental nature of the project, a special effort would be required to closely monitor project implementation and the impact of investments made under the project. Operating responsibility for this task was assigned to the project economist located at the DPAP Central Unit, but the detailed plans and methodologies for collecting and analyzing data, as well as reporting the findings, were not finalized during appraisal. The Appraisal report (para. 5.10), indicates that "Details of the project evaluation process, including the role of independent institutions, would be finalized in consultation with IDA". This was reflected in the legal documents only in the "boiler plate" clause as "furnish to the Association all such information as the Association shall reasonably request regarding the Project..."(DCA, Section 3.07). Given the everyday pressures of starting up and implementing a very complex project, it is not difficult to understand that the designing of the monitoring system was not given the priority expected at appraisal. A significant lesson from this experience, which reinforces present Bank policy, is that project monitor- ing procedures should be discussed and agreed on at appraisal and just as carefully as the technical, managerial, financial, and economic aspects of the project. 7.13 Another area where project economists in the districts and Central Unit could have made a greater contribution to more efficient utilization of project funds was in the evaluation of specific investment activities. There was no uniform methodology for analyzing or evaluating the financial and economic impact of specific investments. This could have been sig- nificantly improved by a few short training courses for project economists in analyzing investment proposals. IDA could have pushed more in this area by including such expertise in its supervision missions. - 52 - VIII. BANK PERFORMANCE At Appraisal 8.01 The stated objective of the project - to increase and stabilize production from agriculture and related activities in six drought prone districts - was sound. However, at appraisal, how this objective could be achieved was not established with certainty and the appraisal mission was faced with a large number of unrelated development proposals made by Indian authorities. The concept of integrated watershed development was new to India and doubts existed about both the soundness of traditional soil and moisture conservation techniques, and the technical suitability of proposed rangeland improvement measures. There were also different views concerning the possibility of improving dryland farming and develop- ing milk production in drought prone areas. 8.02 Under these circumstances, a major contribution of the project was the use of a time slice of an ongoing national development scheme for testing a number of development concepts under field conditions. 8.03 The appraisal team tried to be realistic in estimating the impact of specific programs on agricultural productivity and farmers' incomes. Reliable baseline data were not available and was from evidence of the likely result of project actions. Furthermore, as all proposed activities were to be explained in terms of location-unspecific models, a con- siderable degree of technical speculation had to be applied. The major mistake was to attempt to apply an approach to rangeland improvement proven in Australia. The re-establishment of grasses did not take place with sufficient speed to meet the social and agricultural demands of highly-populated areas in India. The rangeland improvement technique was successfully changed during the project and led to one of the project's major achievements. However, misjudgment at appraisal not only led to an unsuccessful fodder development program but also caused difficulties with the sheep grower societies. 8.04 Aware of the technical problems inherent in the project, IDA financed a number of expatriate specialists to work on specific problems. It also required appropriate monitoring and evaluation systems, but these were not effectively established. IDA, in retrospect, was probably optimistic in assessing the capacity of Indian authorities to implement such a multi-faceted project. Supervision 8.05 It was intended that this project should be closely supervised so that changes needed in activities could be identified early, and needed technical and policy guidance provided promptly. Eleven supervision missions were fielded, representing 433 mandays. Continuity was provided - 53 - by one man who participated in the reconnaisance, appraisal, five supervi- sions and the PCR mission, another staff member in six supervisions and the PCR, and three others who made two to four visits to the project. IX. CONCLUSIONS 9.01 The project was largely successful in identifying, testing and implementing technical, economic and organizational means to improve the productivity of drought prone areas and at a scale large enough upon which to base future policy decisions. Overall, the project achieved the objec- tive of refocusing attention on "drought-proofing" through economically productive programs rather by labor-intensive construction of public works. In spite of deficient impact evaluation, the project has indicated a number of areas in which the experience gained be used in the design of a future DPAP. The major ones are: (a) Watershed Development. Application of an integrated approach to soil, moisture, crop and livestock management has been one of the major accomplishments of DPAP. Soil surveys have been initiated and taken as a first step towards the planning of proper land use and management coupled with watershed works required to protect crop lands and provide for and alternative uses, i.e., pasture, forage seed production or forestry development on marginal lands. The project estab- lished the need for a central body (DDA) at the district level to coordinate the activities of various line depart- ments and to gain the commitment of policy-makers and farmers. (b) Soil Conservation. Land management, especially changes in soil conservation practices, are dependent upon the greater involvement of farmers. The old concepts of soil conserva- tion, relying for the most part, on contour bunding have been difficult to change. However, ICAR and ICRISAT have demon- strated the value of the graded broadbed-and-furrow system for deep vertisols and which is now becoming more widely considered. (c) Rangeland Improvement. Stylosanthes spp have a definite role in the improvement of wastelands on the Deccan plateau. (d) Dryland Farming. There is now confidence in practices, especially for the medium and deep vertisols, and to some extent, for the shallow soils. Yields have improved as a result of low-cost practices such as selection of crop and variety of greater field potential, establishment of better plant stands, application of small amounts of fertilizers and - 54 - weed control. The states and 001 are ready to proceed with more dryland crop research and extension. (e) Livestock Development. Indian authorities now realize that dairy development in drought prone areas is part of dryland farming and that it requires the keeping of improved indigenous animals rather than exotic half-breeds. (f) Water Development. The development of surface and groundwater potential for irrigation is necessary, but about 23% of project costs were spent to the benefit of only a few farmers. Among the factors to be recognized in future projects are: farmer involvement in planning and operations, development of command area, water management training and provision of extension staff, levying of water charges and on-going and maintenance, and for wells, ground water legis- lation to prevent overexploitation. (g) Diversification. Among diversification schemes, sericulture and fisheries were the most successful. Since mulberries grown for sericulture were an irrigated crop, it appears to have little place in DPAP programs. Fisheries required a large investment for the production of stock fish. Future DPAP activities could include the organization of fishermen societies and marketing cooperatives, but leave the the infrastructure required for the production of stock fish to other resources. Horticulture had a limited benefit to a relatively small number of families. (h) Administration. To assure compliance with technical norms, a routine check indicating compliance should be sent to the DPAP cell by DDA before reimbursement of expenses. The responsibility for providing technical supervision either from the Ministry of Rural Reconstruction, other GOT minis- tries, or from state line departments, is not yet clear. 9.02 The experience gained with IDA Credit 526-IN leads to a number of other major conclusions. (a) DPAP cannot be implemented by individual government agencies alone. It needs the broad participation of the people living in drought prone areas. Too little attention has been given to the formulation of programs which were fully acceptable to and supported by potential beneficiaries; and (c) There has been a tendency to confuse subsidies paid to weaker sections of the community with development justifiable on economic grounds alone. Subsidy support to socially weaker - 55 - sections will, obviously, continue but there is opportunity to use these funds as seed money for the introduction of sound income-generating activities. Some subsidy schemes DPAP are actually counter-productive. This is the case with subsidies given for the purchase of sheep flocks or milking animals to landless laborers. Overstocking is one of the main constraints to in rangeland improvement and the addition of animals to village communities further aggravates the situation. 9.03 GOI has decided to continue implementation of DPAP by allocating Rs. 1750M (US$218M) in the next Five-Year Development Plan. Other Bank group projects which address or continue activities initiated in the six DPAP districts include: Rajasthan Agricultural Extension and Research Project (Cr. 737-IN) Maharashtra Agricultural Extension Project (Cr. 1135-IN) Composite Agricultural Extension Project (Cr. 862-IN) National Agricultural Research Project (Cr. 855-IN) National Dairying Project (Cr. 824-IN) Agricultural Refinance and Development Corporation III (Cr. 947-IN), and IV, negotiated in November 1981 Inland Fisheries (Cr. 963-IN) Karnataka Sericulture (Cr. 1034-IN) National Cooperative Development Corp Project (Cr. 871-IN) improvement of cooperative societies Irrigation (various projects) Upcoming projects include: Rainfed Agriculture Erosion and Flood Control Thus, while IDA is not participating in the continuation of DPAP, many related projects are supportive of GOI activites in these areas. - 56 - Table 1 INDIA DROUGHT PRONE AREA PROJECT Project Completion Report Summary of Project Financing (Rs M) Component Appraisal Actual % of % of Government and IDA Appraisal Total Project Costs -------Rs M------- District Funds Soil Conservation 99.2 134.3 135 19.9 Forestry 18.3 37.0 202 5.5 Pasture Development 37.3 33.1 89 4.9 Dryland Farming 11.4 11.4 100 1.7 Sheep Development 13.1 23.9 183 3.5 Dairy Development 63.6 85.8 135 12.7 Minor Irrigation 47.1 131.9 280 19.6 Diversification 12.7 36.7 289 5.4 Electrification 0 22.1 0 3.3 Updating Land Records 6.2 1.2 18 0.2 Project Management 2.9 12.7 438 1.9 Sub-total - 311.8 530.1 170 78.6 District Core Funds 30.0 40.5 135 6.0 Sub-total - 341.8 570.6 167 84.6 Central Funds 12.8 2.0 16 0.3 Price Contingency 222.4 - - - TOTAL GOV'T/IDA 577.0 572.6 99 84.9 Institutional Credit 244.4 101.6 1/ 42 15.1 TOTAL PROJECT COST 821.4 674.2 82 100.0 US$M 102.7 84.3 - 1/ Incomplete data. INDIA DROUGHT PRONE AREA PROJECT PROJECT COMPLETION REPORT GOL/State Expenditures by District and Component (R9. '000) Districts Anantapur Bijapur Ahmednagar Sholapur Jodhpur Nagaur Project Total Cost Activities A. Distriot Funds Sotl Conservation 27482 23436 31106 34409 9097 8747 134277 Forestry 10707 9075 8805 7592 433 383 36995 Pasture Development 6041 6914 9083 1974 - 9112 33124 Dry Farming 2081 1459 4080 3443 - 293 11356 Sheep Development 5069 2868 758 1183 8378 5638 23894 Dairy Development 25029 6165 21406 14048 10538 8593 85779 (including Cattle Dev.) ui Minor Irrigation 32678 22616 24709 34847 11691 5407 131948 Diversification Schemes 22794 12170 878 892 - - 36734 Electrification - - - 16100 6000 22100 Updating Land Records 575 - - - 368 272 1215 Project Management 2693 1504 1437 1628 2911. 2520 12693 Sub-total 135149 86207 102262 100016 59516 46965 530115 District Core Funds 674 626 1524 587 15641 21419 40471 135823 86833 103786 100603 75157 68384 570586 Total B. Central Funds Research 633 Training 132 Consultants 227 Project Evaluation 259 Central Core Funds 711 Total 1962 TOTAL (Gov't Funds) 572548 C. Institutional Credit 23612 6013 7530 14079 30659 19700 101593 D. GRAND TOTAL (All Funds) 674141 - 58 - Table 3 INDIA DROUGHT PRONE AREAS PROJECT Project Completion Report Project Organization and Objectives Appraisal Actual Project Objectives or Focus: Increase acreage or herd Increased cropping intensity Increased productivity - yield or livestock offtake x x Shift to higher value crops Reduced production cost x x Technological innovation x x Estate/Large scale farming Integrated rural development x x Other - pilot project to test and implement new approaches in drought prone areas x x Project Organization New/Existing Performance District Development Authority new excellent State DPAP Cell new fair to good GOI Central DPAP Cell increased staff good District Coordinating Committee existing under good Collector State Coordinating Committee new fair to good District Banking Plans supplemental satisfactory Annual Plans DPAP supplemented good Annual Reports DPAP new good Monitoring and Evaluation various poor Institution Building and Sector Development The focus was on activities of line departments designed to assist people in rural areas, improve their economic status and reduce the effects of drought. DPAP developed institutional capability and was an overall success. The government will continue DPAP activities without further direct Bank financial assistance. Consultants Use and Performance Five expatriate consultants participated in the project implement- ation and all were technically competent. - 59 - Table 4 INDIA DROUGHT PRONE AREA PROJECT PROJECT COMPLETION REPORT Summary of Key Indicators Appraisal Revised Achieved I of A % of Revisea Soil Conservation Soil Survey Th-Ha 8,983 8,983 9,G39 100 100 Soil Cons. Ha 365,000 280,000 335, 10 92 120 Watershed Dev. 15 NR 52 346 346 Ponds 1,050 180 142 13 78 Forestry Afforestation 16,000 10,500 16,945 106 161 Village Wood 2,000 2,400 2,608 130 109 Wind Breaks - Km 1,200 1,200 1,108 92 92 Stab. Sand Ds-Ha 1,000 0 0 0 NA Pasture Development Past. Improv. - Ha 90,000 42,000 46,803 52 111 Seed Production 0 0 97,732 NA NA Dryland Farming Demonstration - No. 835 NR 13,810 1,653 1,653 Improv. Cu. Th-Ha 200 200 246.3 123 123 Sheep Development Societies - No. 270 270 160 59 59 Dairy Development Societies - No. 480 480 753 157 157 Total Milk Mil. L 0 0 92.4 NA NA Minor Irrigation Wells No. 2,110 2,620 4,384 208 167 Tanks No. 29 29 23 80 80 Land Irrigation - Ha 0 0 16,004 NA NA CAD Ha 12,000 9,000 1,477 12 16 Divers Schemes Sericulture Mulberry Ha 1,000 1,000 15,139 1,514 1,514 Fisheries Societies - No. 0 0 41 NA NA Electrification Wells Conn. - No. 0 0 520 NA NA NA - Not applicable. NR - Not revised.  IBDi77(R E OME M E* U a Ž/*z*--O ---c i i ?i- 5 Js. -l . c. - cx( - - R O -c--..-c¢ t, } o ¾1