Philippines - Rainfed Agricultural Development (ILOILO) Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 7949 PROJECT PERFORMANCE AUDIT REPORT PHILIPPINES RAINFED AGRICULTURAL DEVELOPMENT (ILOILO) PROJECT (LOAN 1815-PH) JUNE 30, 1989 Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the performance of Its contents i i not otherwise be disclosed without World Bank authorization. CUENCI fQUIVALENTS P - Philippine Peso At appraisal 1979 US$1 - P 7.4 P 7.38 (INF) 1980 US$1 - 7.51 1981 - 7.90 1982 - 8.54 1983 - 11.11 1984 - 16.70 1985 - 18.61 1986 - 20.45 1987 = 20.45 WEIGHTS AND MEASURES 1 hectare (ha) - 2.47 acres 1 meter (m) * 39.37 inches 1 kilogram (kg) 2.2 pounds ABBREVIATIONS AFT Agriculture and Food Technician APPU Agricultural Project Preparation Unit ASSP Agricultural Support Services Project (Ln. 2040-PH) BAI Bureau of Animal Industry BPI Bureau of Plant Industry BS Bureau of Soils ICB International Competitive Bidding IRRI International Rice Research Institute, Los Banos, Philippines FAO Food and Agricultural Organization KFC Kabsaka Farming Center MOA Ministry of Agriculture (now Department of Agriculture) NFA National Food Authority PCR Project Completion Report PMU Project Management Unit RADOS Rainfed Agricultural Development Outreach Site SAR Staff Appraisal Report SMS Subject Matter Specialist SN Samahang Nayon (pre-cooperative) SWIM Small Water Impounding Structure T&V Training and Visit (extension system) UFV University of the Philippines Visayas (based in Iloilo Province) USAID United States Agency for International Development WHO World Health Organization FOR FCIAL ON MLY THE WORLD BANK Wash*gton. D.C. 20433 U.S.A. OMC o Discreassl Opeseln avakai.if June 30, 1989 HEHORANDUN TO THE EECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Performance Audit Report on Philippines Rainfed Agricultural Development (Iloilo) Project (Loan 1815-PH) Attached, for information, is a copy of a report entitled "Project Performance Report on Philippines - Rainfed Agricultural Development (Iloilo) Project (Loan 1815-PH)" prepared by the Operations Evaluation Department. Attachment This document has a restricted distribution and may be used by recipients only in the performance Y0M *F3AL USE ONLY PROJECT PERFORIANCE AUDIT REPORT PHILIPPINES RAINFED AGRICULTURAL DEELOPMT (ILOILO) PROJECT (Ln. 1815-PR) TABLE OF CONTENTS Pate No. Preface ..........................................*.***.***.*.** **** * Basic Data Sheet ............................ ........... ... . . . 11 Evaluation Sum ary .................................................. iv PROJECT PERFORMANCE AUDIT REPORT I. BACKGROUND .............................................. 1 A. Objectives .......................................... 1 B. Context ............................................. 1 C. Project Formulation, Processing and Design .......... 2 D. Institutional and Organizational Arrangements ....... 4 II. PROJECT IMPLEMENTATION AND OUTCOME ...................... 5 Multiple Cropping Component ......................... 6 Adaptive Trials ..................................... 8 Other Components ........................................ 8 Economic Rate of Return ............................... 11 III. FINDINGS AND ISSUES ..................................... 11 A. Introduction ........................................ 11 B. Project Design, Management and Coordination Issues................ .......... 12 C. Local Funding Shortages ............................. 14 D. Sustainability ...................................... 14 E. Rainfed vs. Irrigated Cropping Options ....... 16 IV. CONCLUSIONS ................................... 20 COMIENTS FROM THE BORROWER ATTACHMENT: Coments from the Ministry of Agriculture .............. 23 This document has a restricted distribution and may be used by recipients only in the performance PROJECT PERFORMANCE AUDIT REPORT PHILIPPINES RAINFED AGRICULTURAL DEVELOPMENT (ILOILO) PROJECT (Ln. 1815-PH) PREFACE 1. This is a Project Performance Audit Report (PPAR) on the Rainfed Agricultural Development (Iloilo) Project in the Philippines, involving a loan of US$12.0 million to the government, with the main objectives of in- creasing cropping intensity and crop production in the single cropped rain- fed rice areas of Iloilo Province, improving the health and nutrition of the ruz,l population, developing water impoundment schemes and preparing a further rainfed development project. The loan was approved on March 20, 1980 and became effective on June 25, 1980. The loan was closed without extension on June 30, 1986. The undisbursed balance of US$6.74 million was cancelled following final disbursement on February 6, 1987. 2. The PPAR is based on the Project Completion Report (PCR) dated April 1987, prepared by the Ministry of Agriculture in the Philippines, an Overview of the PCR dated June 23, 1988, prepared by the Asia Region, the Appraisal and President's Reports, the Loan Agreement, and correspondence with the Borrower, supervision reports and other internal Bank memoranda on file. Bank staff associated with the project have been interviewed. An audit mission visited the Philippines in October 1988 and had discussions with government staff in Manila and Iloilo and with staff at IRRI, and vi- sited a selection of project sites. Their kind cooperation and valuable assistance in the preparation of this report is gratefully acknowledged. 3. The PCR provides a good account and assessment of the project experience, and discusses the performance of the Bank and the project executing agencies. The PPAR elaborates on particular aspects such as the excellent results achieved by the multiple cropping component, the complex project management design and the relative merits of rainfed and irrigated cropping investment in this high rainfall area of the country. 4. Following standard OED procedures, copies of the draft PPAR were sent to the Borrower's agencies on May 9, 1989. Comments received from the Borrower's Department of Agriculture, dated June 13, 1989 are attached. - It - PROJECT PERFUMANCE AUDIT REPORT PHILIPPINES RAINFED AGRICULTURAL DEVELOPMENT (ILOILO) PROJECT &AN lsf-10 BASIC DATA SHEET KEY PROJECT DATA Appraisal Actual or Actual as 5 of Expectation Current Estimate Appraisal Estimate Project Cost (USS million) 21.5 11.70 55.1 Loan Amount (US million) 12.W6 5.26 48.8 Date Board Approval 03/20/86 Date Effectiven*** 96/26/80 Date Physical Components Completed 6/85 07/865 1M Proportion then Completed (S) 16 67/1 67 /1 Closing Date 66/88 06/86 Economic Rate of Return (3) 37 28 62 Financial rate of return (W) N.A. N.A. Institutional Performance Mediocre Agronomic Performance Good Number of Direct Beneficiaries (Farm Families and Labor) 45,0 38,00 84 CUMULATIVE DISBMSEMENTS FY81 FY82 FY8 FY84 FY85 FY66 FY87 Appraisal Estimate (USS million) 6.8 2.64 6.9 8.29 10.23 12.06 Actual (USS million) - 1.06 3.87 4.67 4.77 6.84 5.26 Actual as S of Estimate 6 49 72 55 47 42 44 Date of Final Disbursement February 1, 1967 Cancellation (USS million) 6.74 STAFF INPUTS (staff weeks) FY77 FY78 FY79 FY80 FY81 FY82 FY83 FY84 FY85 FY8 FY87 FY88 TOTAL Identification/ Preappreisal n.a. n.a. n.a. 52.6 Appraisal 53.4 30.8 64.2 Negotiation .1 6.3 6.4 Supervision .9 32.6 11.3 13.7 11.2 6.6 9.3 3.8 6.7 96.1 Other .3 .3 .0 3.9 .9 5.5 I n.a. n.s. . 8 .3 .8 .4 17.6 ' 1.1 6.8 9.3 3.8 6.7 244.8 - iii - MSSION DATA Date No. of Staff/Days Speeiallzation Performance Types of Iwelar) Paemans to field Ld isls idLt B2shPsM L Identlfletle 1/7 2 Ag (2) Preparation 18/78 7 140 Ec,Cr(2),Ag,H,Nu,E Appreeel f6/79 6 188 Ec(2),Ag,E,Tr,Nu Post Appraisal 1 10/79 1 4 Tr Post Appraieal 2 92/80 1 5 EG Subtotal 16 817 Supervision 1 &4/8 2 4 Ag(2) 2 2 MF Supervision 2 11/8 2 7 Ag(2) 2 1 MF Supervision 3 3/81 8 24 Ag,Cr,Nu 2 1 UF Supervision 4 06/81 8 88 Ag,Tr,Fi 2 2 MF Supervision 5 #2/82 2 21 Ag,FI 2 1 FM Supervision 0 10/82 4 6 Ag,FI,Tr,Nu 2 2 FUT Supervision 7 6/83 1 12 Ag 2 2 FUT Supervision 8 3/64 3 45 Ag,Tr,Nu 2 2 FUT Supervision 9 02/85 1 14 Ag 2 2 FUT Supervision 1 06/88 2 12 Ec,Ag 2 2 - Subtotal 22 228 Total 37 646 OTHER PROJECT DATA Borrower Republic of the Philippines Executing Agency MOA, now DA Fiscal Year of Borrower Calendar Year Name of Currency (abbreviatioG) Peso (P) Currency Exchange Rate: Appraisal Year Average (1979) USS1.W a P 7.37 (IMF) Intervening Yeare Average US81.00 a P 11.73 Completion Year Average (1986) US11.00 a P 26.45 Cofinanciers None Follow-on Project None. Some US$125,000 project expenditures were disbursed under Loan 2049-PH, mostly for consultancies, research and health components. L Original estimated 86,600 he. was revised to target of 52,5600 ha, against which completion was 77X. 6 Ag a agriculturist; Ec = economist; E = engineer; Fi = financial analyst; Cr = credit specialist; Tr a training specialist; No a nutrition specialist; and H a healhtk specialist. A 1 = problem-free or minor problems; 2 = moderate problems; and 8 a major problems. / 1 a Improvi.g; 2 = stationary; 3 a deteriorating. / F z financial; M a managerial; and T = technical. - iv - PROJECT PERFORMANCE AUDIT REPORT PHILIPPINES RAINED AGRICULTURAL DEVELOPMENT (ILOILO) PROJECT (Ln. 1815-PH) EVALUATION SUM MART Introduction 1. Three earlier Bank-assisted rural development projects had included rainfed cropping, but previously country agricultural sector priorities had focussed mainly on rice self-sufficiency and, along with Bank assistance, this meant that agricultural development was concentrated on the irrigated areas. Such irrigated areas were generally better off than rainfed areas. By implication, therefore, this first Bank-assisted project for rainfed areas of the Philippines included equity and poverty alleviation objectives. 2. The project built on experience gained from the earlier successful "Kabsaka Project" in a small rice-growing area of Iloilo Province on Panay Island. Through this project the government, with the assistance of the International Rice Research Institute (IRRI), introduced to the farmers of Iloilo a multiple rainfed cropping technology developed at IRRI. By 1979 some 600 farmers were cultivating about 1,000 ha with average yields of rice per annum of 7.3 tons and a cropping intensity of 180%. This compares with a single cropping production level of about 4.4 tons. Based on field inspection of these activ!ties and discussions with government staff, the Bank took an interest in the proiram from late 1977 and the project was approved with a loan of US$12 million in March 1980. Objectives and Project Components 3. The main objective of the project was to increase cropping intensity and crop production in the traditionally single-cropped rainfed rice arear of Iloilo Province, thereby raising farm incomes. The project was seen as having potential for development of rainfed cropping areas in both the Philippines more generally, as well as in other countries. Other objectives and components for Iloilo included the promotion of related livestock activities, improvement of grain marketing and transport, improved seed supplies, improved human nutrition and health, upgrading of institutional staff resources, construction of 40 scattered small water impounding schemes, a pilot program for village development, and adaptive trials of rainfed cropping. About 45,000 farm families cultivating about 62,000 ha were expected to benefit directly from project investments. National components were included for preparation of future rainfed agricultural development projects and water impoundments, adaptive trials to develop rainfed cropping technologies for other areas, and strengthening of the planning and project preparation capabilities of the Bureau of Soils. The project did not include components for agricultural credit, rural road construction or a significant increase in extension service -v - capability since these aspects were expected to be provided under ongoing or proposed Bank-assisted proJects. Implementation 4. Project implementation took place during an unforeseen period of severe domestic budgetary constraints and political turmoil, declining commodity prices and insufficient official production credit. The achievements of the project in such difficult circumstances point to the good performance of the staff involved, the industry and skills of the Iloilo rural population and, in most cases, the robustness of the approach and technologies, in particular of the central multiple cropping component. 5. Although the project was intended to be implemented over five years, budget constraints resulted in 91% of project expenditures being disbursed in just one year, the second project year (1982). Partly as a result, project expenditures were reduced to 442 of SAR estimates and US$6.7 million (562) of the Bank loan was cancelled. Budgetary constraints were reflected in the field mainly through reduced mobility (only five of the project's vehicles out of 38 proposed were eventually received), low3r staff numbers and less funds for staff operating costs. The loan was closed on schedule in June 1986, although certain high priority activities continued to be financed from surplus funds in the Bank-assisted Agricultural Support Services Project (Loan 2040-PH). Results 6. Despite reduced resources and other implementation problems which reduced coverage, the multiple cropping component was more successful than had been expected at appraisal. This success has attracted much attention, including official visits from several other developing countries and overseas agencies. At the end of the project the new rainfed technology was being used on 40,584 ha, an achievement of 68% of the SAR target of 60,000 ha, or 77% against the agreed revised target of 52,500 ha. Average cropping intensity rose to 200% compared with the SAR projection of 160%. Average yields were 5.6 tons he per year of rice, compared with the appraisal projection of 5.4 tons. Yields increased 47% over pre-project levels. Some farmers follow the two rice crops with a highly profitable upland crop such as mung bean, melons and vegetables. Some 26,315 farmers (66% of the appraisal target or 75% of the revised target) were involved in multiple cropping at the end of the project. Farm incomes at least doubled with adoption of the new technology. Data for the two years following the end of the project in 1985 indicate that multiple rainfed cropping has continued to spread, although at a slightly reduced rate. The human nutrition and health, pilot village development and revised marketing components were also generally very successful. The livestock, adaptive trials, project preparation, and seed production components were less successful, however, although again results were better than might have been expected under the prevailing constraints. The water impounding program was wisely suspended after 16 of the 40 dams had been constructed, given high costs and concentration of benefits on few farmers. Training exceeded appraisal targets handsomely, but monitoring and evaluation of project activities had mixed results. - vi - 7. The assumed levels of institutional agricultural credit, rural road construction and increased extension services to be provided by other Bank-assisted projects did not materialize. 8. The reestimated rate of return of 231 compares with a high rate of 371 estimated at appraisal. This lower rate is attributed in the Overview mainly to world rice prices having about halved during the project period. Sustainability 9. Iloilo farmers will continue to practice multiple cropping as long as crop and livestock production remain profitable. Furthermore, there is already evidence that multiple cropping farmers will shift from rice to higher value upland crops when prices and demand are satisfactory. Extension services are continuing under dynamic provincial management and with improved remunerations (but not increased mobility), and there remains a continuing, albeit reduced, research effort. There is consideraole evidence that the technology, so successfully applied in Iloilo, is also being adopted in other areas where rainfall distribution is satisfactory. 10. Provincial staff arc conrnrned, however, that their reduced ability to follow up on project A 'ities is affecting the sustainability of some activities. In particL . for the rainfed cropping component, reversion to old practices is highfc among the later adopting villages than those earlier ones who received continued support. Also, the dry seeding technology needs to be overhauled for southern Iloilo where soil conditions are less suitable. The human nutrition and health component activities have not continued at project levels, largely because of reduced supplies. Findings and Lessons Learned 11. The excellent results generally achieved, especially for the multiple cropping component and despite severely constrained resources, suggest that the Bank-assisted project may have been over-elaborate. Consequently, attributing the good results achieved to the Bank-assisted project may not be wholly justified. In particular IRRI notes that similar results have been achieved in South Cotabato (in Mindanao) with only inputs and some salary incentives being provided by IRRI. IRRI has suggested (para. 31 of the PPAR) that where strong technologies are available their rapid spread does not require large-scale infrastructure and expenditures. 12. The audit believes that the rainfed cropping component could have been even more successful on a wider scale had various other components not been added to it. The addition of extra components to increase the project size (since the multiple cropping component did not require large-scale assistance) tended to direct attention away from what should have been the preeminent multiple cropping component. 13. The failure of the monitoring and evaluation system to produce timely analysis of field impact, although much data was collected, is a typical failing of such activity. The audit believes that formal M&E arrangements should have been agreed at the outset, possibly with the involvement of IRRI as a partner. - vii - 14. The design of the project management system for such a complex mlti-agency project was typically flawed. Thus the project management unit and committee system proposed did not function as intended, partly because the comittees were too large and unwieldy. 15. Finally the audit raises an issue connected with irrigated cropping investments financed by the Bank in Iloilo Province. Thus the low cost but high benefits of the rainfed cropping technology in this area indicate limited marginal returns to high-cost irrigation investments. Even so the Bank has financed irrigation investments on areas contiguous with the rainfed crop areas. This indicates the need for careful eva- luation of all production options when preparing and appraising irrigation projects. PROJECT PERFORMANCE AUDIT REPORT PHILIPPINES RAINFED AGRICULTURAL DEVELO1MENT (ILOILO) PROJECT (LA. 181W-PHi I. BACKGROUND A. Obiectives 1. The main objective of the project was to iacrease agricultural productivity of lowland areas through the more widespread in;roduction of a proven rainfed multiple cropping technology. Related objectives and components included the promotion of livestock activities integrated with the cropping program, the improvemenz -%f marketing and transport, improving seed supplies, improved human nutrition and health and the upgrading of institutional and staff resources. Other secondLry objectives related to soil and water conservation through scattered water Impounding schemes, a pilot program for village development, and adaptive trials and project preparation for more widespread introduction of multiple rainfed cropping in other suitable areas. 2. The project was seen as the first Bank assistance specifically for rainfed cropping in the Philippines, with potential for development of other rainfed cropping areas in the Philippines as well as in other coun- tries. Three earlier Bank-assisted rural development projects had included rainfed cropping, but previously country agricultural sector priorities had focussed mainly on rice self-sufficiency. This priority, coupled with Bank assistance, had meant that developments were concentrated on the irrigated areas. Such irrigated areas were generally better off than rainfed areas. By implication, therefore, the project and its strategy for the rainfed areas included equity and poverty alleviation objectives, although these were not explicitly stated in the Staff Appraisal Report (SAR). B. Context 3. The project built on research experience gained from the earlier Rainfed Upland Rice Project in small areas of central Luzon. This project was started in 1970 by the then Agricultural Productivity Commission, with support from the International Rice Research Institute (IRRI) and Peacq Corps volunteers from the USA. This project showed that dry seeding of the first rice crop allowed a second rainfed rice crop to be grown. 4. The rainfed technology I/ had then been field tested in Iloilo Province (aLd another site in Luzon). This work started in 1974 on a 1/ The technology used early direct seeding of improved rice varieties to achieve double rainfed cropping, with the passibility of a third upland crop of mung beans, melons etc., in high rainfall areas where formerly only one transplanted rice crop was being grown. For further details see Attachment 1. - 2 - small scale and was then continued more widely from 1976 through a spe- cialized pilot extension service. Experience at Santa Barbara in Iloilo Province over a three-year period indicated that rice yields of 8.7 tons per he per annum were attainable on a farm scale, or four times the average production for such rainfed lowlands. By the third year (1977) of these farm trials, some 89 farmers cultivating 54 ha were involved. The success of the early trials led to the creation of an expanded pilot extension project called "Kabsaka" (an acronym for "bounty in the farm"). Success was such that by 1979/80 there were 600 farmers in the program in Iloilo, cultivating about 1,000 ha. Average yields for the three-year period of the extension project were 7.3 tons per ha per annum with a cropping intensity of 180 percent. Based on field inspection of these activities and 'iscussions with government staff, the Bank took an interest in the program from 1977. C. Project Formulation, Processing and Design 5. The project was identified by a mission in November 1977 and prepared by the Agricultural Project Preparation Unit in the Ministry of Agriculture, assisted by provincial government staff. A Bank preparation mission in October/November 1978 reviewed the broad-based area dev--lopment proposal of the Ministry of Agriculture. Issues addressed included mar- keting problems, poor credit repayments and the interest rates structure, aspects of the water impounding component, the rural road development program, incremental farm labor requirements and the proposals for the health, nutrition and rural water supply components. 6. Throughout project processing, the scope of the project was a major concern, given that existing or proposed projects could provide services related to multiple cropping in particular and could fund other aspects of rural d-elopment activities. As a consequence, once it was agreed that some activities could be funded under other projects, there was relatively little left in the core rainfed cropping component requiring Bank funding. To some extent, therefore, the small size of the project, which was raised as an issue during processing, acted as a stimulus to include other activities which led the SAR to describe the project as a type of area development project. 7. The project was appraised in May 1979 and approved in June 1980. Despite the title of the project 2/ it included significant irrigation de- velopment (a quarter of base costs), other activities which were national in character and not restricted to Iloilo, and some components which were not agricultural production development (nutrition/health and pilot village community development). The project description divided the project into two sets of components, those for Iloilo itself and those classed as na- tional (mainly concerned with the more widespread introduction of multiple cropping) as follows: 2/ Rainfed Agricultural Development (Iloilo) Project. (i) Iloile Project Area Agricultural Production Services 1. Introduction of improved crop and livestock vroduction and animal health through the provision of staff support, vehicles, facilities, supplies, equipment and breeding stock for the extension services. 2. Improvement of seed supplies through production, marketing to improve quality, credit for seed production, processing and storage facilities, and for marketing, and strengthening of the seed quality control system. Adaptive Trials Improve rainfed cropping technology through adaptive trials. Water Impounding Schemes Construction of about 40 small earthen dams with related irrigation and drainage infrastructure. Grain Storage and Shipment Facility Establishment of one grain storage and shipment facility. Human Nutrition Prevention of anemia and control of diarrhea. Project Management Support Establishment of a project coordination comnittee, project secretariat, project management committee and project management unit. Training 1. E-tablishment of a residential trair.Lg facility and provision of fellowships. 2. Introduction of a pilot scheme to test and evaluate an approach to comprehensiie village development by training villages to assess village needs and make use of existing local and external resources and services to meet their needs through community-based action. (ii) National Components Preparation of Future Rainfed Agricultural Development Projects Adaptive Trials Improvement of rainfed cropping through adaptive trials. Project Preparation Capability Strengthening the planning and project preparation capability of the Bureau of Soils in the Ministry of Agriculture.3/ 3/ The intention of the SAR (para. 3.06 (g)) was to fund consultants for the planning of a national water impounding program, but this was not reflected in the Loan Agreement. - 4 - 8. Apart from the central objective of introducing the multiple cropping technology more widely in Iloilo, the project included related off-farm investments, aspects of the social development of rural areas of Iloilo and preparations for the more general introduction of multiple cropping in the Philippines. At appraisal 10.72 of project base costs were allocated to "Agricultural Production Services" (i.e. crops and livestock work in Iloilo). 9. Although the project was referred to in the SAR as a type of "integrated area development" project, certain other important related activities were not supported directly under this project since they were provided for under existing or planned nation-wide projectr which covered Iloilo Province. Thus roads were to be funded under the proposed Rural Roads Project, credit was assumed to be available from existing lines of credit supporting rural banks and development banks, and only a small increment of the extension service was included in the project given that more substantial strengthening of the extension service was provided for under the ongoing National Extension Project.4/ D. Institutional and Organizational Arrangements 10. The project's designers and appraisers were faced with a common organizational problem in such multi-component projects when proposing the project's management structure. Not only was there the dilemma of how to coordinate as many as twelve agencies, 5/ without establishing a new parallel (and enclave) bureaucratic structure, but the Ministry of Agriculture was implementing a decentralization strategy which included appointment of Regional Directors of Agriculture. Selection and appoint- ment of these Regional DLectors was in limbo at the time and a compromise was agreed whereby the Provincial Governor and the Director of the Bureau of Agricultural Extension were to co-chair the Project Management Committee until such time as the Regional Director of Agriculture was appointed. 11. The conventional option adopted to overcome the coordination problem was to set up a project management unit (PMU) in MOA and coor- dinating committees at two levels. The Project Manager was to head this PMU and was to sit in the office of the Regional Director of Agriculture 4/ A commentator from USAID wrote at the time, however, that poor implementation of this extension project cast doubt on the adequacy of the extension services likely to be available to the project, in the absence of more intensive project-specific arrangements. 5/ The twelve agencies requiring coordination were the ministries of Agriculture, Health, Education and Culture, Local Government and Community Development, and also the National Nutrition Council, the Nutrition Center of the Philippines, the Media Production Centers, the National Irrigation authority, the Farm System Development Corporation, the Philippines Training Center for Rural Development, and the Philippines Council for Agricultural Resources and Research. Coordination was also required with the International. Rice Research Institute. - 5 - (when appointed) in Iloilo City. However the Project Manager had no powers, other than those of persuasion and entreaty, to influence the activities of the many other agencies involved in implementation. 12. Coordination in the prolect area was to be achieved by the Project Management Committee, which had a large membership comprising the local di- rectors of other involved agencies, the Governor of Iloilo and other repre- sentatives of provincial authorities, farmers and the marketing sector. This committee was to act as a guiding and coordinating body and could not, of course, be involved in day-to-day management of the project. The inter- agency Projert Coordinating Committee sitting in Manila, chaired by the Minister :: Agriculture, was to coordinate aspects of implementation at the national level. Beneficiary participation was to be achieved by relying on the existing structure of "action committees" at the provincial, municipal and district level to determine participation in the project, location of facilities, etc. Project activities were also expected to support farmers associations and revitalization of the pre-cooperatives (Samahang Nayons). II. PROJECT IMPLEMENTATION AND OUTCOME 13. In addition to the coordination problems inherent in a multi- component project of this type, the project was implemented during a period of severe domestic budgetary constraints and political turmoil, declining international commodity prices and insufficient official production credit. The project had to contend also with drought and with typhoons, which in particular damaged the water impoundments, although these are regular fea- tures of the Philippines' climate. A well guided and committed project management team faced constant shortfalls in budget allocations and delayed releases. Consequently many planned project investments and activities could not take place or were significantly delayed. The achievements of the project in such difficult circumstances point to both the good perfor- mance of the staff involved and, in most cases, the robustness of the approach and technologies, in particular of tl.z central multiple cropping component. 14. Although the project was intended to be implemented over five years, budget constraints resulted in 91% of project expenditures being made in just one year, the second project year (1982). Partly as a result, project expenditures were reduced to 44% of SAR estimates, with crops and livestock activities receiving almost the same one tenth share (of the reduced amount) as allocated at appraisal. This not only dramatizes how difficult implementation was, but adds weight to the possibility that the original project design, at least with respect to the multiple cropping component, may have been over-elaborate given that the impact of this component was still substantial with much lower costs than planned. In effect, for much of the original "project period" resources available to project staff were only a little greater than those that would have been provided in the absence of the project and few of the staff were themselves "incremental". Consequently the achievements of the project must be attri- buted as much to the personal efforts, initiative and drive of project staff as to externally funded incremental resources. - 6 - 15. Details o' implementation problems and resource deficiencies are provided at length in both the PCR and Overview and do not need repeating here. However some illustration of the more dramatic shortfalls serves to indicate how well the project performed in difficult circumstances. First- ly, the key to implementation of the multiple cropping component was mobi- lity of the extension staff, since they needed to move between villages to introduce and assist the change over to the more complex and labor inten- sive new cropping pattern. In the event although some 38 field vehicles were allocated to the project, only five eventually reached the project and staff had to make do with a few existing inadequate vehicles or other means of transport for most of :he project period. (A presidential ban on vehi- cle purchase by government agencies was the main problem.) Secondly, bud- get cuts were extreme and had considerable impact on extension activities since lack of incremental staff operating funds constrained training and other extension activities. Not only were budget allocations cut severely, but often the amounts eventually released were significantly less than the reduced amounts approved. Also, funds became available so late in the budgeting period that frequently unspent funds were forfeited. Implemen- tation performance of the core multiple cropping component and related research is discussed further below, followed by notes on points of inte- rest on other components. Multiple Cropping Component 16. Despite the many implementation problems, this core of the project was more successful than had been expected at appraisal. Success can be attributed mainly to the excellent benefits offered to farmers by the technology, to the enthusiasm and dedication of the staff involved, and to the industry and skill of the farm families. The Bank's funding was seen by project staff as contributing to increasing the expansion rate through the funding of demonstration plots, field days, educational trips, and providing salary incentives. Some of the key staff were contracted and financed by Bank funds. The success of the rainfed technology in Iloilo has attracted much attention, including official visits from several other developing countries and overseas agencies. 17. The following data summarize how the technology spread during the project years. Project Years (c) SAR 1 2 8 4 6 Projection 78/79 79/86 SI/81 81/82 82/88 83/84 84/86 Year 6 (a) (b) Participating Farmers n.a. 600 1,687 4,662 9,769 26,316 81,727 40,000 Cultivated Area (ha) n.a. 1,00 1,497 8,324 13,709 34,627 40,584 s,666 Cropping Intensity X 120 186 194 171 156 188 201 188 Rice Yields tons/ha/yr 4.1 7.3 4.4 n.e. 5.8 n.a. 6.8 6.4 (a) SAR estimate (b) E.V. Sandig, IRRI, Sept., 1988 (c) PCR, Tables 1, 2 and 4 - 7 - 18. In terms of numbers of participants, cropping intensities and yields the component continued, as intended, the good work started in the 1970s, except that the original target of introducing the technology on 60,000 ha was not achieved, the maximum reached in 1985/86 being 40,584 ha, or 682 of the target. (Meanwhile, a revised target of 52,500 was agreed with the Bank. This decision reflected both budget constraints and new land survey data that indica.;ed that the earlier data used in the SAR overstated suitable land areas. The achievement was 77% against this revised target.) Average cropping intensity rose to 2002 and rice yields rose from 3.8 tons/ha/yr to 5.6 tons/ha/year, an increase of 47%. This is well below the ambitious goal of IRRI of 9-10 tons and some 23Z below the average annual yield of 7.3 tons achieved by 600 farmers in the pre-project year of 1979/80. (Yield reduction of about this amount can be considered normal when pilot operations are expanded.) The SAR expectation had been an annual yield of 5.4 tons of rice, but at a cropping intensity of only 160%, and thus the SAR production target per hectare was comfortably exceeded. 19. Implementation of the multiple cropping component was facilitated greatly by the earlier work done under the Kabsaka Project. Thus there had been a steady build-up from a small number of farmer trials by IRR/MOA in the mid 1970s, so that by the late 1970s six hundred farmers in Iloilo had adopted the new technology. By the time the project started, a significant number of farmers had been multiple cropping for several years. The audit would also note that the basic ingredient of the technology, direct seeding rather than transplanting rice, was not new to Iloilo.6/ Direct wet seeding had been the general practice until the 1950s when transplanting of rice was adopted--dry seeding, however, was an innovation.7/ 20. The technology had a number of points strongly in its favor with respect to rate of adoption by farmers, especially that Iloilo's rainfall pattern is probably one of the best in the Philippines for rainfed crop- ping and hence the risks of crop failure are low when appropriate crop varieties and field practices are used. This fact, combined with the good soils and topography, and industrious farm population, made it highly likely that, once appropriate crop varieties and related technology became available, rainfed cropping would intensify rapidly. Income data in the PCR indicate that farmers could typically more than double their net in- comes by adopting the technology and the substantial impact of this in- creased income was evident in the area. Multiple cropping, however, does require much greater labor inputs from the family and the results achieved attest to the industry of the farm families when offered a good opportunity co improve their living standards. 6/ Although direct seeding avoids the heavy labor inputs for transplanting (and also thereby releases funds for purchase of fertilizer and chemicals) the overall labor requirements of double cinpping are higher than for the former less intensive system. 7/ In IRRI's view the main advantages of dry seeding are that (i) it forces early cropping (and hence the possibility of second and third crops); and (ii) a more aggressive root system is established which helps the crop withstand drought. - 8 - 21. Data for the two years following the end of the project, during which extension activities were supported by funds from the Bank-assisted Agricultural Support Services Project (ASSP), are less clear, but multiple rainfed cropping has continued to spread, although at a slightly reduced rate. Adaptive Trials 22. Trials of multiple cropping technology were undertaken both within the Iloilo project area and in eight other provinces to develop cropping technologies for rainfed conditions elsewhere. Despite budget funds being cut and usually arriving late, records show that 155 trials were carried out. Other problems noted were adverse weather and apparently some lack of full commitment to the component's objectives. The component was supported latterly mainly by IRRI funds, with some trimming of staff, but there was insufficient time under the project to develop sound cropping patterns. After the project closed, this was one of the components which continued to function in 1986 and 1987, partly with funding from the ASSP. Other Components 23. The other components of the project will be covered in less detail here as they are discussed adequately in the Overview and PCR. In sumary, the livestock component was generally believed to have had substantial im- pact, with animals produced exceeding appraisal estimates except for goats, but with a low survival rate for improved roosters distributed to villages. The improved breed program's achievements, however, were modest at well below appraisal targets. (Project staff noted that the important marketing aspect of livestock development was clearly demonstrated by the component being successful only where there are cattle auction markets.) The water impounding component was curtailed at the suggestion of Bank supervision staff after only 16 of the 40 dams had been completed and 560 ha irrigated, or 31% of the targetted service area. Although technically in output terms these schemes performed reasonably well, they were judged to be too costly (about US$1,500/ha) and to concentrate their benefits on too few farmers. Several of the schemes were badly damaged by typhoons and reconstruction work was carried out to restore them to full operation (the Bank suggested that repairs should take preference over new schemes). (With respect to impact, project staff noted that, inter alia, flood control and fish farming were more important than irrigation, given high rainfall. This view is consistent with both the SAR (para. 7.01) and the issue raised under E. below on rainfed versus irrigation options in Iloilo.) 24. Almost all the targets of the human nutrition and health component were achieved and substantial impact was reported. Problems encountered relate to reduced funding, including late arrival of drug supplies, and inadequate transport and travel funds, which contributed to poor super- vision of water pump installaticns. The pilot village development component was very successful in identifying needs and mobilizing local resources to make selected improvements. Following early successes, the coverage of the program was expanded to 27 barangays in consultation with the Bank. Expansion of coverage has continued in a small way after the project closed down, so that this activity is now underway in 38 barangays (including starting in the three other provinces of Panay Island). The Overview notes and the audit agrees, that there is cause for concern that - 9 - this component was implemented by MOA personnel whereas Community Deve- lopment Department personnel are more suited to such work. The 10,000-ton rice store was constructed as planned under the storage and marketing component, but the shipping facilities were dropped and by mutual agreement the funds were used to establish 37 farmer center stores (mainly for bulk- ing up purposes) with rice drying floors, and to purchase 50 fuel-fired grain dryers. At the time of the audit's field visit the stores were said to be functioning successfully, except for three where there were security- related problems. These facilities are owned and operated by village coo- peratives. Only 15 of the dryers were sold, and few of these are used, because operating costs proved to be too high. (The Bank here appears to have been ambivalent over the years on the viability of such dryers. An earlier proposed grain dryer component (apparently to fund similar dryers) under the Philippines Rural Infrastructure Project (Credit 790-PH) had been rejected as unviable in 1977, during procesaing of that project.) The seed production and marketing component was concerned with the production of rice seed and seed for upland crops. Whereas adequate rice seed was avail- able to project farmers, the significant shortage of seed 'or upland crop development (mung bean, melons and vegetables) constrained this activity (i.e. the third crop, to take cropping intensities over 200%). Project staff attempted to deal with this situation but with only limited success, partly because of adverse weather. 25. The availability of institutional agricultural credit was consi- dered a constraint throughout the project since, contrary to expectations at appraisal, adequate funds were not available to project participants partly because of both ineligible onlending banks (more than half of them) and ineligible farmer borrowers. Furthermore, the rural banks on which reliance was placed for credit, do not generally lend working capital. In consequence participants had to continue to resort to the informal sources of credit where according to audit mission farmer interviews shor --term operating credit interest rates are the equivalent of 60% per annum or more.8/ Field reports and survey data indicate that over 70% of farmers take such credit. (An alternative view from IRRI, however, is that credit is not a vital part of the package, given the strength of the technology. This view is consistent with experience elsewhere with the adoption of strong new technology, but it would be desirable to investigate the point further through detailed field enquiries.) 26. Tr ining for participants exceeded appraisal targets handsomely. After much delay, construction of the planned training facility was even- tually dropped since provincial authorities considered local facilities adequate. Monitoring and evaluation of project activities was not entirely successful because arrangements were not detailed and agreed at appraisal, the surveys conducted were too complex, and because an attempt at data 8/ Rates reported to the mission were mainly the traditional "five/six" system (take five units and return six in four months time after harvest) and the rice borrowing arrangement (borrow 3 units of rice before harvest and return five after harvest four months later). The first system equates to 60% p.a., whereas the latter is equal to 200% p.a. in physical terms before price changes over the period are taken into account. - 10 - computerization failed. The Loan Agreement contained strong covenants on M&E, but this activity was not identified as a project component.9/ A baseline survey and final evaluat;on of the rainfed cropping component carried out by University of the Philippines staff in 1981 and 1986 res- pectively were quite detailed but were of less usefulness than intended. This because the baseline surve) was not sufficiently comprehensive and there was such a high sample mortality (801), partly because of peace and order problems, that the impact evaluation survey did not readily demons- trate project results. Prolect planning activities for further rainfed cropping projects and a national water impounding program had mixed re- sults. Results from the field trials were insufficient (in such a short time) to finalize recomm-endations for rainfed development projects and in any case project preparation activities under other sponsorships eventually made the work under this project largely obsolescent. Some preparatory work was done on water impounding, but detailed preparation was suspended following review by a FAO/World Bank Cooperative program mission and atten- tion was redirected towards alternative means of providing supplementary water. 27. Project management arrangements did not work out as intended at appraisal, especially as the complex structure planned to overcome coor- dination problems did not function smoothly. Membership of the committees was too large to allow them to function effectively (smaller sub-committees were formed) and there were problems with the allocation of committee functions between Iloilo and Manila (these were eventually rationalized). The Project Management committee met fairly regularly in Iloilo to begin with, but then meetings tailed off. (These problems are fairly typical of experience elsewhere with such coordination arrangements for similar multi- agency projects covering several different activities.) 28. The PCR and comments to the audit mission indicated concern that "the Bank closed the project prematurely". The circumstances are, however, not quite so simple. When the Bank sent a mission to consider, inter alia, the justification for extending the loan, it discovered that the government had already terminated contractual staff, disbanded the project management unit and had not made any budget provisions for the following year. Thus, from the Bank side, the decision was considered to have been already taken, although arrangements were agreed for financing some of the more important continuing activities with surplus funds from the ongoing Agricultural Support Services Project. 9/ The SAR states (para. 4.19) that "Because of the experimental nature of this project, a significant monitoring effort is foreseen. The Project Manager would be responsible for designing and implementing a monitoring and evaluation system for the Project." Furthermore, although there was intended to be a baseline survey under the human nutrition and health component (SAR, para. 3.27), this was not reflected in the SAR project description, nor in the Loan Agreement. The PCR does not refer to this survey, although the results on nutrition aspects of the mid-term Eva- luation Survey by the University of the Philippines in 1984 are summarized in the PCR. - 11 - Economic Rate of Return 29. The PCR Overview reestimated the economic rate of return for the three production components (cropping, livestock and water impounding) at about 232 compared with a high rate of 372 estimated at appraisal. Al- though the achievements of the cropping component exceeded appraisal expec- tations, this lower rate is attributed in the Overview mainly to world rice prices having about halved during the project period. III. FINDINGS AND ISSUES A. Introduction 30. Despite being implemented during a period of economic and poli- tical disruption, and budgetary constraints, the multiple cropping, nu- trition and health, and pilot village development activities with which the project was associated were clearly successful. Livestock activities were also reported to have had significant impact, although clear data on this were not available and breeding results were disappointing. The water impounding component was wisely curtailed once the high costs and concen- trated benefits were fully appreciated. The seed production component did not produce the upland crop seeds in sufficient quantities. Rice handling activities, once it was agreed to change part of the component, were quite successful, but there is still too little post-harvest infrastructure to handle the expanded Iloilo rice crops efficiently. Adaptive trials acti- vities had good initial results, but were not sufficiently advanced during the project period to reach conclusive results. None of the three related supporting projects assisted by the Bank resulted in the provision to Iloilo of the new rural roads, agricultural credit or extension services that the project design assumed would be available. With the possible exception of poor roads in N. Iloilo, these shortfalls are judged not to have greatly affected project success. Overall, therefore, it is tempting to declare the project a great success, especially given the extraordi- narily difficult circumstances under which it was implemented. Mostly this attests to the robustness of the multiple cropping technology and the sound identification of needs with respect to most of the supporting components. 31. Some reservations are held, however, by the audit with respect to attributing these results, without qualification, to the Bank-supported project. These reservations relate most strongly to the multiple cropping component and derive partly from unclear relationship in cause and effect terms between the Bank funds and component activities. The mulliple crop- ping technology had been well tested and tried with hundreds of farmers in Iloilo before the Bank project started, and indeed the PCR comments on the difficulty of establishing clear project impact with respect to cropping intensities during the early 1980s (see PCR Overview, Table 5 and para. 6.3). Similarly 91% of the Bank funds were used to finance activities in only one year, 1982, the second year of the planned five-year project period and in 1983 only 15% of the allocated budget was made available. Thus much of the incremental resources proposed at appraisal were never - 12 - available to the project.10/ The audit judges therefore that the impact of the Bank funding on the multiple cropping technology was probably less dramatic than the PCR or Overview may suggest to the reader. IRRI staff have taken an even firmer position by attributing success mainly to the strength of the technology. Thus in a detailed study published in 1983 (IRRI, Evaluatint Technologies for New Farming Systems), the fol!,wing conclusion is presented with respect to the rapid spread of the rainfed technology in Iloilo and Luzon: "Where technologies appropriate in the sense described here can be introduced to farmers, there should be no need for costly and elaborate extension services and rural development institutions to service their introduction. If such organizations are necessary to support and maintain the widespread adoption of certain new technologies, these are unlikoly to be appropriate." 32. IRRI staff note furthermore that in South Cotabato (in Mindanao) a similar impact to the project experience in N. Iloilo has been achieved with IRRI providing inputs and salary incentives. Project and IRRI staff recall, however, that Bank funding did facilitate extension activities which otherwise might have been constrained even more than they were, par- ticularly field days, demonstration plots, educational trips and training seminars. 33. A set of issues arising from this project are discussed further below. These relate to the design and management of the project, budgetary constraints and their solutior, sustainability of project investments and the implications of the successful rainfed technology on irrigation deve- lopment priorities. B. Proiect Design, Management and Coordination Issues 34. The audit believes that the rainfed cropping component could have been even more successful on a wider scale had various other components not been added to it (although the additional components were implemented suc- cessfully). The expansion of the project concept occurred, in part at least according to the files and staff comments, to "bump up" project costs. The support needed for the multiple cropping component in Iloilo was relatively small, especially since several related activities were al.ready being funded or would be funded shortly afterwards by other Bank- supported projects. There is a possibility that the Bank's interest in financing large rather than small projects directed attention away from what should have been the preeminent multiple cropping component. In the 10/ It is noteworthy in this connection that the SAR's assumptions for the "without project" situation assumed production growth of 4.8% per annum as the technology continued to spread (with no incremental investments), or in other words that about half of the increased output projected "with project" would occur anyway. In the event that "with project" projection--based on 160% cropping intensity--proved very conservative compared with the 200% intensity achieved with reduced resources. - 13 - process of adding secondary components to the multiple cropping design, the focus of the project was lost somewhat (despite its title remaining speci- fic) and it is probable that achieving the optimum impact of the strong technology was compromised, especially with respect to the adaptive trials and planning work which were intended to pave the way for more widespread introduction of multiple cropping. The spread of the technology to other areas through the adaptive trials and planning activities was not pursued as vigorously as results in Iloilo would suggest was justified. In the audit's view a more straightforward project focussing on spreading the multiple cropping technology (in Iloilo and then elsewhere) could have had even greater impact, given the intrinsic strength of this technology. Consequently a major opportunity to make a substantial contribution to the welfare of large numbers of the rural poor in other suitable areas has not been followed up aggressively. 35. During implementation the Bank was pragmatically flexible with respect to changing the market ng component and agreeing with the govern- ment that the water impounding schemes should not be implemented in the numbers originally proposed. With advantage, however, the Bank could have been even more flexible, despite lending volume concerns, by appreciating at the preparation stage the strong merits of the multiple cropping tech- nology, not only in respect to Iloilo, and pursuing that part of the pro- ject more directly, without requiring various other peripheral components which tended to detract from pursuit of the core multiple cropping work. That this was not done raises questions again about the apparent conflict between the resource transfer function of Bank Group operations and the Bank's interests in supporting sound technically based investment projects. 36. The failure of the monitoring and evaluation system to produce timely analysis of field impact, although much data was collected, is a typical failing of such activities. In this case, given widespread interest in this rainfed cropping technology and the appraisal intent that the project was to be the precursor of other rainfed developments in the Philippines and elsewhere, it is regrettable that firmer M&E arrangements were not agreed at the outset. In particular, it now appears natural and efficient to have co-opted IRRI as a partner in the effort. 37. There were also problems connected with the design of the project management system which are not unique to this particular project. The organization and management system had to deal with difficulties created by more than 12 agencies being involved in project implementation. Planners had to resort to a conventional project management unit and committee approach which has been used on a number of other similarly complex pro- jects with similarly qualified results. Thus the inability of the project unit in MOA to manage and initiate project operations, where other agencies were involved, is a common failing of systems which give responsibility for project implementation to only one of several equally ranked agencies.11/ The two committees proved unwieldy and met infrequently in later years. Organization and management problems were also complicatel by a suggestion on the files that the relevant bureaucracy in the Philippines was not fully 11/ During supervision the possibility was raised of all project funds being passed through MOA to the various agencies to give the PMU more trol ' -- t--' arra - , a , rod, ei. - 14 - supportive of the project objectives and by personality problems between the Manila and Iloilo levels. In the circumstances, when these management problems are added to the other national problems of the time, it is again a tribute to the robustness of the component designs and to the initiative and drive of staff that so much was achieved. 38. There were also problems of integration on both the Bank and country side, given that, apart from the 12 agencies involved in the project in the Philippines, the Brnk also had difficulty dealing with the fact that credit, rural roads and extension services were all either provided under existing projects or expected to receive attention under upcoming projects. This issue took up much time and effort during prepa- ration and appraisal and was settled at a relatively late stage. In the event, for various reasons, the expected credit, rural roads and expanded extension services never materialized. 39. Similarly there were also some coordination problems with other agencies, government and external, with interest in related areas. Thus the files indicate that there was a strong protest from USAID staff with respect to lack of consultation on the extension activity, in which they already had an interest (and in this connection they also questioned whether the project documents fully reflected the problems being expe- rienced on implementing the Bank-supported National Extension Project). On the health and nutrition component the files indicate some unhappiness on the part of WHO and the Ministry of Health with respect to the initiatives for Iloilo under this component overlapping or replacing similar initia- tives provided for un,er a national program in which WHO had an interest. (This suggests that this component may have been substituting for the national program.) C. Local Funding Shortages 40. With worsening budgetary conditions the project was starved of funds throughout and consequently Bank disbursements were correspondingly slow (and eventually 56% of the Bank's funds were cancelled). Thus there was apparently a case for resorting to extraordinary measures to provide local funds to facilitate Bank disbursements. In many other similar cases the Bank has agreed to advancing funds against a "special account", which is then replenished by the Bank against reimbursement applications. No trace of such a possibility being considered appears on file, but the audit understands that such an arrangement was discussed informally but was not requested by the authorities. In the circumstances, the Bank might have been more pressing on this point. Alternatively the Bank may not have felt there was a great need in this case since most project activities, espe- cially the core multiple cropping expansion, seemed to be proceeding quite well anyway. D. Sustainability 41. Iloilo farmers will continue to practice multiple cropping as long an er .1 v pr, 'nt' r n profit hla. 4 .-vi 'are - 15 - continuing under dynamic provincial management and with improved remune- rations (but not increased mobility), and there remains a continuing, al- beit reduced, research effort. There is considerable evidence that the technology, so successfully applied in Iloilo, is also being adopted in other areas where rainfall distribution is satisfactory. 42. Following a period of relatively heightened activity during part of the project period, however, extension activities are again severely constrained by low resources. Provincial staff reported some concern that after the project closed their reduced ability to follow-up with beneficiaries was affecting, and was likely to affect in the future, the sustainability of some of the changes introduced under the project. Data are available for the multiple cropping component indicating that although reversion to old practices is low for earlier adopting villages, it is becoming more significant for later villages. (This problem is mainly noticeable in the south where lighter soil conditions are less ideal for the technology--see below on the need to "overhaul" the technology.) Project staff attribute this, in part at least, to reduced field visits because of reduced travel funds. (This point, however, conflicts somewhat with the view of the rainfed technology as very robust and, because of its obvious merits, largely "self-promoting".) This reversion issue needs to be monitored carefully in the future. 43. Technical staff in Iloilo also believe that for a part of southern Iloilo, the dry seeding technology needs updating because of soils related problems. The need for periodic updating is to be expected for any such advanced agricultural technology. Without specific incremental funding for the Kabsaka program anymore, however, such activity has to be funded through less direct means and this is currently being done through the Bank-assisted Agricultural Support Services Project (22 demonstration plots continue to be operated in Iloilo). 44. Sustainability of rainfed multiple cropping of rice will be influenced by the availability of reliable marketing facilities. However, there is already evidence that multiple cropping farmers will shift from rice to higher value upland crops when prices and demand axe satisfactory. During the audit's visit, at the height of heavy rains which coincided with first crop harvesting, there was clearly great need for improved drying and storage facilities (losses through spoilt rice were said to be high in such conditions). (During implementation this need was acknowledged when funds for the cancelled training center were reallocated to grain storage and drying facilities.) A second sustainability factor is that the rainfed technology itself is regarded by project staff as "exacting" with respect to weed control, since there is only a short period of a few days when current control methods are effective and very few herbicides are suitable. Hopeful signs are that farmers have themselves developed a (non-herbicide) water management system for weed control in some parts of the Iloilo area and that international chemical companies are working on weed control problems in Iloilo. 45. The human nutrition and health activities have proved not to be sustainable at project levels, mainly because of shortage of funds and supplies. This is a predictable outcome (based on common experience else- where) and connects with the poirit above on the related national program. - 16 - designed to be self-sustaining using locally available materials, after the technique has been accepted and understood by villagers. E. Rainfed vs. Irrigated Cropping Options 46. The spectacular success of the rainfed multiple cropping techno- logy in Iloilo raises questions about the appropriateness of nearby irri- gated cropping development in such high rainfall areas. Similar concerns have been addressed recently in a PPAR on three Bank assisted irrigation projects in Indonesial2/ and also in a 1984 PPAR on an irrigation project in Mexico.13/ In the case of Iloilo the issues are especially poignant given that the Bank has funded both this rainfed project and surface irrigation systems in contiguous areas and with overlapping implementation periods.14/ 47. The main issue is illustrated most dramatically by the conclusions drawn in the two appraisal reports for the irrigation and rainfed projects from analysis of the same rainfall data series for Iloilo. Thus the SAR (Jan. 5, 1977) for the Jalaur Irrigation Project suggests that "irrigation is essential to ensure a (second rice) crop". In contrast the SAR (Feb. 20, 1980) for the Rainfed Agricultural Development Project deduces from the same data and widespread farm trials by IRRI/MOA that highly productive and profitable multiple cropping without irrigation is possible with only some yield reduction from occasional second crop failures. The difference in approach results largely from the irrigated crops being longer season transplanted rice, whereas the rainfed technology uses early maturing direct-seeded crops. 48. The two project areas concerned are adjacent, indeed the rainfed technology was first successfully field-tested in the mid 1970s with far- mers on the borders of the Aganan and Santa Barbara Irrigation Systems, whi(-h are included in the Jalaur project. Furthermore, rainfed cropping is likely to be even more suited to the irrigated areas since these tend to be the lower-lying heavier soils which are therefore more drought tolerant than some of the non-irrigated land left to rainfed farming. It is also significant that much of the continuing expansion of the rainfed multiple cropping technology (as seen by the audit mission) is taking place in the extremities of the irrigation systems where farmers from experience know that canal supplies are low and unreliable. 12/ PPAR, INDONESIA, Irrigation Projects X, XIV, and XV (Loans 1578, 1811, and Credit 995-IND), dated June 30, 1989, OED. 13/ PPAR, MEXICO, Panuco Irrigation Project (Loan 969-ME) and Rio Sinaloa Irrigation Project (Loan 970-ME), OED Report No. 5191, paras. 26-28. The point refers only to the Panuco Irrigation Project. 14/ The Bank-supported Jalaur Irrigation Project (Loan 1367-PH) was approved in December 1974 and closed in June 1983 (see PPAR, OED Report No. 5969, dated December 12, 1985. The project covered three irrigation systems: Jalaur, Aganan and Santa Barbara, all of which are - Iln0n City. - 17 - 49. Using data extracted from the respective SARs and PCRs (unless indicated), the following table indicates in broad terms the expected and actual performance of these two options for incretasing production in Iloilo. The Aganan component of this project was chosen for the main comparison because it is immediately adjacent to the site of the farmer trials of the rainfed technology in the 1970s and is typical of low river flow irrigation schemes in the Philippines (see footnote 15 below). How*- ver, the rainfed yields used in the comparison below are not the yields achieved in these early farm trials (e.g. 7.3 tons of rice/ha/yr by 600 farmers in 1979/80 - see para. 17), but the 262 lower yields achieved on average for the whole of the rainfed project by over 30,000 farmers. Cropping Intensity Yields (tons/ha/year) Before SAR Before SAR Proiect Projection Actual Proiect Projection Actual Rainfed (Iloilo) Project 120 160 198 (a) 2.1 5.4 5.4 Aganan I.S. component of Jalaur (Iloilo) Irrigation Project 130 149 126 (b) 2.7 6.0 (c) 5.1 (b) (a) Includes some third (upland) crops, which are of higher value than rice. (b) Aganan Irrigation System component--126% cropping intensity (file memo May 20, 1987). This file memo also refers to cropping intensity for the Jalaur-Suague I.S. of 168%, since the Jalaur river has better dry season flows. For the whole project, with three separate irrigation systems, the PCR indicates at completion an average cropping intensity of 167% and a rice yield of 6.6 tons/ha/yr. Slightly higher yields (7.8 tons/ha/yr.) and intensity (170%) are expected at "full agricul- tural development" in 1988. (c) This SAR yield projection for Aganan seems low for irrigated rice, but as things turned out it was probably a realistic reflection of low river flows. 50. The above table shows that where run-of-river surface irrigation schemes have inadequate supplies for high cropping intensities, as for Aganan, the rainfad cropping alternative is superior in production terms alone, without considering ;apital savings on the surface irrigation infrastructure. As a fiit order cost comparison the Jalaur SAR shows rehabilitation base costs for Aganan of US$571/ha (in 1976) compared with SAR base costs for the agricultural production services, all training and project management (pro rata) of US$62/ha (in 1979) for the later rainfed development project. (The rainfed project was contradictory in this res- pect by also supporting irrigation development, through the water impound- ments, at a cost of about US$1,500/ha, although high cost was the main - 18 - reason for suspending the component during implementation.) The rainfed option also has the advantage that double cropping of short-season rice varieties on rainfall permits a third crop of high-value upland crops (such as melons, vegetables and mung bean) which can yield higher income than rice such that some farmers sacrifice the second rice crop (wholly or partly) in favor of the upland crop. The data above also show that the rainfed option was very competitive with the SAR projections for the whole Jalaur scheme, part of which was area expansion through new infrastructure, especially given the high capital, operating and maintenance costs of surface irrigation systems.15/ (In the circumstances it is significant that the Jalaur project was intended as the first phase of a two-phase development, with Phase II providing a storage reservoir. The audit understands, however, that Phase II has not been implemented.) 51. The above comments show that irrigation in Iloilo is not the only route to efficient higher production. A much more detailed cost:benefit analysis of these alternatives than is appropriate here would have to take into account other factors, but, prima facie, the rainfed option appears far superior, in cost:benefit terms, to irrigation in Iloilo conditions. Consequently, it is concluded that the economic rates of return of about 20% reported in the PPAR (both SAR and reestimated) for the Jalaur Project would be substantially reduced if the rainfed cropping option was used as the "without project" situation (the Jalaur SAR assumed a continuation of the current low production rainfed system with only a slow yield increase resulting from introduction of new improved varieties). An alternative method would have been to compare the net present worths of the rainfed and irrigated options against the same "without project" assumption (although the numbers and results would be similar). 52. In commenting on a draft of this memorandum, the region notes that at the cropping intensities achieved by the Jalaur I.S. component of the Jalaur Project (see footnote (b) to the table above) a yield increment to irrigation of a ton of rice indicates that the rainfed comparison is not superior. This ton of rice, however, may be compared with the 3.8 tons increment that was assumed in justifying the Jalaur project (SAR Annex 12, Table 5), the increment of over 3 tons achieved by the rainfed technology and the much higher investment costs of the irrigation schemes. 53. In the Jalaur SAR there is no mention of alternative production strategies for the areas concerned, such that it appears that the appraisal 15/ Although beyond the immediate scope of this audit, it is worth noting that there are a number of irrigation systems elsewhere in the Philippines where low river flows in the drier months constrain cropping intensities such that the national average irrigated cropping intensity (including storage schemes) is only about 130%. Although Iloilo has rainfall amounts and distribution which are particularly favorable to rainfed cropping, such a low national irrigated cropping intensity would suggest that in other areas with good rainfall rainfed cropping could be the most efficient option, rather than high cost, under-utilized surface irrigation systems. - 19 - team took as its main objective the establishment of surface irrigation systems, without examining other options.16/ (The good prospects for rainfed cropping had been demonstrated by IRRI/MOA at the time. Field trials beside the Aganan and Santa Barbara schemes were proceeding and expanding rapidly when the Jalaur Project was being appraised.)17/ Such a broader analysis would have required the examination of at least rainfed cropping potential as well as the groundwater option for irrigation sup plies. In contrast the economic analysis in the Jalaur SAR concerns itself exclusively with demonstrating that the selected surface irrigation option has an adequate rate of return, but does not show that this is the best production option then available. Exploration of options is the true pur- pose of project analysis, but is often overlooked in favor of the less time-consuming "go/no go" approach. One persuasive explanation of the exclusive focus on irrigation, and apparent neglect of alternative pro- duction options, is that the Jalaur project was prepared and appraised by a specialist irrigation division, with the rainfed project being handled by a parallel division handling non-irrigated agriculture in the East Asia and Pacific regional office of that time. At the 1987 reorganization of the Bank such specialist irrigation divisions were merged into agricultural operations divisions and agricultural technical divisions. 54. In agreement with the recent PPAM on the Indonesian irrigation projects (OED, op. cit.), the audit suggests that the above issue points to the potential danger, both within the Bank and within countries, of insti- tutionalized irrigation groups which tend to promote and perpetuate public sector irrigation infrastructure without sufficient if any incentive to optimize investment opportunities where there are good prospects for rainfed production. 55. Subsequent to the above analysis and drafting of the preceding paragraphs the following very relevant published comments were brought to the evaluator's attention and merit quotation: "With the commitment of most development assistance agencies to agricultural and rural development, there is great pressure to lend money for agriculture, with irrigation receiving a dispropor- tionate share. Irrigation projects attract development banks and 16/ It is possible that other options were considered at the preappraisal stage, but this is not an audit of the Jalaur project and this point has not been researched. The same comment is also applicable to the appraisal a year later (March 1978), by the same operational division, of the Rural Infrastructure Project (Credit 790-PH) which included a small-scale irrigation systems component for Panay Island (again using Iloilo rainfall data), as well a marginally better rainfall data for northern Panay. 17/ The Bank's regional staff have noted correctly that there were only a few farmers employing the rainfed technology in Iloilo Province (near the irrigated area) at the time of field appraisal in mid-1976 (and by inference that the appraisal team mtght be excused overlooking such an option). The audit's point remains that the organization of the Bank at the time favored supporting irrigation investments, irrespective of - 4rri ti Ar rnat,v '. - 20 - aid agencies because they utilize large amounts of capital, result in highly visible infrastructure and provide a service that nearly everyone agrees is a requirement for development... *Because the two institutions most closely associated with irriga- tion project preparation--the financing development banks and the irrigation authorities--have strong self interests in having favorable project evaluations, it is especially necessary for national planning agencies to be capable of carefully evaluating such projects and comparing them with alternative opportunities."18/ IV. CONCLUSIONS 56. The audit concludes with respect to implementation and Immediate impact on the basis of the above review that: (a) the Kabsaka program assisted by Bank funding continued to be remarkably successful in spreading rainfed multiple cropping in Iloilo Province, with great benefits to participating farmers and labor, despite serious budgetary constraints, political upheaval and weather problems during the project period; (b) the human nutrition and health, pilot village development and revised marketing components were also generally very success- ful, although the nutrition and health activities are not to be sustained at project levels; (c) the livestock, adaptive trials, project preparation, and seed production components were less successful, although again results were better than might have been expected under the prevailing constraints; (d) the water impounding program was wisely suspended given high costs. 57. Sustainability of some components may, however, be threatened by the reduced funding now reported and some attention is warranted by the authorities to follow up along lines discussed above and in the PCR. 53. Finally the audit raises four main concerns with respect to the Bank's role. Firstly, there is a suspicion that the Bank-assisted project may have been over-elaborate in its intended support of the multiple crop- ping expansion. Secondly, the Bank's aversion to small projects may have contributed undesirably to add-on components which were not central to the core objective of expanding rainfed cropping, and thereby diverted the pro- ject away from this very worthwhile activity. Thirdly, the Bank's expec- tations with respect to essential support from three parallel projects 18/ Studies in Water Management Economics at IRRI, Department Paper No. 79-2 (mimeo), Herdt, R.W., March, 1979. (Quoted in Survey of Irrization in Eight Asian Nations, William R. Gasser, USAID, July, 1981.) - 21 - (assisted by the Bank) proved to be over optimistic, indicating perhaps an internal coordination problem in the Bank. Fourthly and lastly, the com- plex PHU and comittee system set up to manage the project proved to have typical drawbacks experienced elsewhere with such multi-component projects. It is surprising that the solution proposed to overcome the project's mana- gemat complexity was exactly that which had been shown to be ineffective elsewhere many times before. 59. The discussion on the rainfed versus irrigated cropping option in Iloilo provides an example of the failure of the Bank to employ project analysis for the original purpose intended--the analysis of options--rather than for making a 'go/no go" decision on a selected option. COMMENTS FROM THE BORROWER 23 - ATTACHMENT REPUBLC OF THE PHLIPPINES DEPARTMENT OF FINANCE MANLA INTERNATIONAL FINANCE SROUP 17 June 1989 Mr. Graham Donaldson Chief, Agriculture, Infrastructure, and Human Resources Division Operations Evaluation Department The World Bank Washington, D.C. 20433 Pe:!ane g Clua Develoome7-t :t 0811-P Draft Project Performance Audit Report Dear Mr. Donaldson: This refers to your letter of 09 May 1989 reauest:ng for comment= on the above document. We are pleased tc transmit the comments of the Deaprtmert o4 Agriculture, impiementing agency of the pro2ect, as submitted to this Department. We trust that these comments shall prove helpful i the preparation of the final evalLiation document. ve . ou vours. CT .INE E CAARoES F71 nial E:conomist - 24 - Comments of the RADIP Component Supervisors on the Final Draft of the Performance Audit Report Loan 1815-PH Page 3 (i) ILOILO PROJECT AREA - Agricultural Production Services 1. Crops Extension 2. Livestock Development 3. Seed Production and Marketing 4. Water Impounding Project 5. Adaptive Trials and Research 6. Marketing 7. Nutrition - include parasitism control 8. Project Management Unit - Training - Pilot Village Development - Monitoring and Evaluation Page 4 (III) PROJECT IMPLEMENTATION - The first six dams were completed by November 1982, and another 10 dams were constructed in the dry -season of 1983. Page 5 AGRICULTURAL SUPPORT SERVICES - 38 proposed vehicles - out of 23 purchased vehicles, 5 reached the project Page 7 PARAGRAPH 19 - The transplanted rice was practiced before, during the late 60's direct seeding (wet) was introduced td minimize labor costs. Direct seeding (dry) however, is an innovation. Page 9 OTHER COMPONENTS - The 37 farmer stores were built from funds originally intended for training facility. - 25 - Page 69 CHAPTER 5 WATER IMPOUNDING COMPONENT - The component contracted the services of the Technosphere Consultant Groups, Inc. (TCGI)in the preparation and actual construction of the dams. Page 71 TABLE 1 PRESENTS THE TARGETS VS. ACCOMPLISHMENTS OF THE WATER IMPOUNING COMPONENTS AS OF DEC, 1986 (See attached sheet) Page 89 AN ADDITIONAL PARAGRAPH UNDER IV ANTHROPOMETRY - Final evaluation report 1986-87 also showed significant increases in the Knowledge and Practice Scores of the mothers related to anemia, diarrhea, and parasitism management control. Likewise, a significantly higher hemoglobin level or better iron nutrition was found among active farmers and a 32% reduction of ascariasis was observed in the intervened areas. Improved dietary patterns as in higher frequency of eating selected foods and increased food intake during pregnancy and lactation was also noted. - 26 - -71- Weble I presents the targets ve accomplishments of the water impounding component as of December 1986 Physical Physical Percent (0 Indicators Activity Accomp. Completion Target lisheent 1. Dame Constructed 40. 16 40 2. Area serviced (ha) 1,800 560 31 3,. Cropping intensity of service area 2.5 2.1 84 4. Watershed development & management a. Preparation of conservation farm plan for WIP & other alternatives 40 16 40 b. Nurseries 1. Maintenance of main nursery 1 1 100 2. No. of seed- lings 75,000 37,350 50 c. Development progran soil conservAtion guides farm concepts 8 - d. Vegetative erosion control measures 1. No. of trees planted 240,000 202,944 85 2. Area planted to trees (ha) 160 101.50 63 3. Bedding of dam embankment (downstream 27 slope) 27 e. Mechanical erosion control (Brush dam) 68 measufe 2 2.94 5. Peasibility studies of proposed WIP a. dam embankment type 42 32 76 b. diversion dams 28 4 14 e. fam pend 28 - - shallow wells 80 - - - 27 - - 72 - Physical Physical Percent () Activity Acromp- Co4letion Indicators Target lishment 6. Institutional work (Phase I) a* hiring of VIP technicians 25 11 44 b. training/seminar of IP techni- cians on insti- tutionalization of 9 & M pro- cedures (no. of training) 20 5 25 7. Institutional Work Phase II (Intervention with farmers) a. training of farmers on insti- tutionalization of 0 & V procedures (no. of farmers trained) 154 498 323 8, Seeding tilepia fingerlings a. No. of fish cages 16 1 6 b. No. of finger- lings distributed to the owners of reservior area 65,000 211,000 325 19. Improved access road (km) 9.04 4.1 45 10. Dams repaired 15 15 100- 11. Training Center constructed - 1 - 12. Rain gauges installed - 16 13. Stop gauges installed - 56 - * Damaged by typhoon "Undang"